diff --git a/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/added_tokens.json b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..c9d3d3a1b74d87e381e471f7b33784015d2dc0ea --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_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_adv_competesmoe/config.json b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/config.json new file mode 100644 index 0000000000000000000000000000000000000000..73ab92e38f84782475b94b505806c28379a3720e --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_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": 0.01, + "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_adv_competesmoe/generation_config.json b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..dad5c4578f0dc5969b38755d095fc30c368bb54a --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_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_adv_competesmoe/model-00001-of-00003.safetensors b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8c1af57948147ff18257357fa98acb202372edc8 --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:76be3f26b6d71fd43af65c062018e6eed055c1a9f54deef6fd3f17c1bb453a87 +size 4972489328 diff --git a/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/model-00002-of-00003.safetensors b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e25f101408d9c220431328cb41703b82c4dc59af --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1b4525210f81e36d6ab0be9661d6f27d9ab6fe71e00334d611ddbd50303f0842 +size 4985754844 diff --git a/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/model-00003-of-00003.safetensors b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..230587cf548beb0a0fb13448a8706fdd881d338b --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b7835661b178d74e6ad75b74e9b65661e8a196b4d889d2f0cca04adc5b262782 +size 248943552 diff --git a/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/model.safetensors.index.json b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..a900cb68b39c8fe6eedc011196340060a750c9c8 --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_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_adv_competesmoe/special_tokens_map.json b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..3e4d5a5bc1cb51753cc9ae0305ece0da60052b10 --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_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_adv_competesmoe/tokenizer.model b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/tokenizer.model new file mode 100644 index 0000000000000000000000000000000000000000..6c00c742ce03c627d6cd5b795984876fa49fa899 --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_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_adv_competesmoe/tokenizer_config.json b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..d579bb0b91b24b214ea3c2e487e27a65017cdc4a --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_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_adv_competesmoe/trainer_state.json b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..66c8c5c1dae737a65b6a914a909d4237d420869d --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_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": 0.06611694, + "auxiliary_loss_mlp": 0.02537191, + "balance_loss_clip": 2.63839054, + "balance_loss_mlp": 2.03670216, + "epoch": 0.00012024289063909097, + "flos": 24932483919360.0, + "grad_norm": 39.88656985738595, + "language_loss": 2.58406591, + "learning_rate": 0.0, + "loss": 1.91742194, + "num_input_tokens_seen": 20375, + "step": 1, + "time_per_iteration": 15.3322594165802 + }, + { + "auxiliary_loss_clip": 0.0440361, + "auxiliary_loss_mlp": 0.0167635, + "balance_loss_clip": 1.75535691, + "balance_loss_mlp": 1.34523463, + "epoch": 0.00024048578127818193, + "flos": 30664624377600.0, + "grad_norm": 55.415347047541715, + "language_loss": 1.88945651, + "learning_rate": 5.021476677069823e-07, + "loss": 1.95025611, + "num_input_tokens_seen": 39035, + "step": 2, + "time_per_iteration": 2.712953567504883 + }, + { + "auxiliary_loss_clip": 0.0439425, + "auxiliary_loss_mlp": 0.01668898, + "balance_loss_clip": 1.76197433, + "balance_loss_mlp": 1.33701921, + "epoch": 0.0003607286719172729, + "flos": 19026227969280.0, + "grad_norm": 40.15924109223687, + "language_loss": 1.6169138, + "learning_rate": 7.958852231401551e-07, + "loss": 1.67754531, + "num_input_tokens_seen": 57600, + "step": 3, + "time_per_iteration": 2.6141245365142822 + }, + { + "auxiliary_loss_clip": 0.04443379, + "auxiliary_loss_mlp": 0.01669364, + "balance_loss_clip": 1.76140428, + "balance_loss_mlp": 1.34530532, + "epoch": 0.00048097156255636386, + "flos": 19316314206720.0, + "grad_norm": 36.84608737932661, + "language_loss": 1.64669359, + "learning_rate": 1.0042953354139647e-06, + "loss": 1.70782089, + "num_input_tokens_seen": 76465, + "step": 4, + "time_per_iteration": 2.6353821754455566 + }, + { + "auxiliary_loss_clip": 0.04411855, + "auxiliary_loss_mlp": 0.01654159, + "balance_loss_clip": 1.76032114, + "balance_loss_mlp": 1.33200765, + "epoch": 0.0006012144531954548, + "flos": 13991264893440.0, + "grad_norm": 54.77765439565791, + "language_loss": 1.93802142, + "learning_rate": 1.1659507774310057e-06, + "loss": 1.99868155, + "num_input_tokens_seen": 94350, + "step": 5, + "time_per_iteration": 2.8461031913757324 + }, + { + "auxiliary_loss_clip": 0.04458215, + "auxiliary_loss_mlp": 0.01693767, + "balance_loss_clip": 1.76005554, + "balance_loss_mlp": 1.3582648, + "epoch": 0.0007214573438345458, + "flos": 23148988225920.0, + "grad_norm": 47.254691852508614, + "language_loss": 1.61150467, + "learning_rate": 1.2980328908471373e-06, + "loss": 1.67302454, + "num_input_tokens_seen": 114595, + "step": 6, + "time_per_iteration": 2.9388697147369385 + }, + { + "auxiliary_loss_clip": 0.04189422, + "auxiliary_loss_mlp": 0.01787674, + "balance_loss_clip": 1.86930847, + "balance_loss_mlp": 1.4626615, + "epoch": 0.0008417002344736367, + "flos": 67663246170240.0, + "grad_norm": 4.623217876081487, + "language_loss": 0.81474757, + "learning_rate": 1.4097067265369432e-06, + "loss": 0.87451851, + "num_input_tokens_seen": 179590, + "step": 7, + "time_per_iteration": 3.361647367477417 + }, + { + "auxiliary_loss_clip": 0.04370199, + "auxiliary_loss_mlp": 0.01700638, + "balance_loss_clip": 1.75338912, + "balance_loss_mlp": 1.36589801, + "epoch": 0.0009619431251127277, + "flos": 21281381504640.0, + "grad_norm": 41.48469888293885, + "language_loss": 1.59074008, + "learning_rate": 1.506443003120947e-06, + "loss": 1.65144849, + "num_input_tokens_seen": 195090, + "step": 8, + "time_per_iteration": 2.9003889560699463 + }, + { + "auxiliary_loss_clip": 0.04392865, + "auxiliary_loss_mlp": 0.01702704, + "balance_loss_clip": 1.75577652, + "balance_loss_mlp": 1.36414957, + "epoch": 0.0010821860157518186, + "flos": 23331342597120.0, + "grad_norm": 17.64537028200169, + "language_loss": 1.47883904, + "learning_rate": 1.5917704462803102e-06, + "loss": 1.53979468, + "num_input_tokens_seen": 211635, + "step": 9, + "time_per_iteration": 2.8687214851379395 + }, + { + "auxiliary_loss_clip": 0.04422557, + "auxiliary_loss_mlp": 0.01682041, + "balance_loss_clip": 1.76195204, + "balance_loss_mlp": 1.35512125, + "epoch": 0.0012024289063909096, + "flos": 17010166337280.0, + "grad_norm": 13.195804262213434, + "language_loss": 1.53035188, + "learning_rate": 1.6680984451379884e-06, + "loss": 1.591398, + "num_input_tokens_seen": 224705, + "step": 10, + "time_per_iteration": 2.930316925048828 + }, + { + "auxiliary_loss_clip": 0.04360861, + "auxiliary_loss_mlp": 0.0167489, + "balance_loss_clip": 1.75462794, + "balance_loss_mlp": 1.34739852, + "epoch": 0.0013226717970300007, + "flos": 21288133261440.0, + "grad_norm": 14.111005766461565, + "language_loss": 1.32554746, + "learning_rate": 1.7371455188905097e-06, + "loss": 1.38590491, + "num_input_tokens_seen": 244635, + "step": 11, + "time_per_iteration": 2.831874132156372 + }, + { + "auxiliary_loss_clip": 0.0438629, + "auxiliary_loss_mlp": 0.01670457, + "balance_loss_clip": 1.75637448, + "balance_loss_mlp": 1.33953142, + "epoch": 0.0014429146876690916, + "flos": 27237884935680.0, + "grad_norm": 12.90176252338624, + "language_loss": 1.253088, + "learning_rate": 1.8001805585541196e-06, + "loss": 1.31365561, + "num_input_tokens_seen": 265765, + "step": 12, + "time_per_iteration": 2.886225938796997 + }, + { + "auxiliary_loss_clip": 0.04354782, + "auxiliary_loss_mlp": 0.01638038, + "balance_loss_clip": 1.75425172, + "balance_loss_mlp": 1.3269496, + "epoch": 0.0015631575783081825, + "flos": 19062174504960.0, + "grad_norm": 6.78991217392871, + "language_loss": 1.2924726, + "learning_rate": 1.8581671739548328e-06, + "loss": 1.35240078, + "num_input_tokens_seen": 283500, + "step": 13, + "time_per_iteration": 2.774747848510742 + }, + { + "auxiliary_loss_clip": 0.04327024, + "auxiliary_loss_mlp": 0.01646839, + "balance_loss_clip": 1.7487967, + "balance_loss_mlp": 1.32621336, + "epoch": 0.0016834004689472734, + "flos": 48139473985920.0, + "grad_norm": 6.191995036714096, + "language_loss": 1.1345371, + "learning_rate": 1.9118543942439254e-06, + "loss": 1.19427574, + "num_input_tokens_seen": 305685, + "step": 14, + "time_per_iteration": 4.153839111328125 + }, + { + "auxiliary_loss_clip": 0.04305815, + "auxiliary_loss_mlp": 0.01667771, + "balance_loss_clip": 1.75140166, + "balance_loss_mlp": 1.34542894, + "epoch": 0.0018036433595863645, + "flos": 34970026314240.0, + "grad_norm": 5.815752761050017, + "language_loss": 1.12745321, + "learning_rate": 1.961836000571161e-06, + "loss": 1.1871891, + "num_input_tokens_seen": 327340, + "step": 15, + "time_per_iteration": 3.908928155899048 + }, + { + "auxiliary_loss_clip": 0.04012989, + "auxiliary_loss_mlp": 0.01672899, + "balance_loss_clip": 1.85007048, + "balance_loss_mlp": 1.36009359, + "epoch": 0.0019238862502254555, + "flos": 59768284440960.0, + "grad_norm": 3.8176595006077316, + "language_loss": 0.64706701, + "learning_rate": 2.0085906708279293e-06, + "loss": 0.70392591, + "num_input_tokens_seen": 382710, + "step": 16, + "time_per_iteration": 3.275775909423828 + }, + { + "auxiliary_loss_clip": 0.04318209, + "auxiliary_loss_mlp": 0.01614332, + "balance_loss_clip": 1.7492615, + "balance_loss_mlp": 1.3032434, + "epoch": 0.0020441291408645466, + "flos": 20814543417600.0, + "grad_norm": 4.241054689496294, + "language_loss": 1.15947795, + "learning_rate": 2.0525099325728135e-06, + "loss": 1.21880341, + "num_input_tokens_seen": 400890, + "step": 17, + "time_per_iteration": 2.8636486530303955 + }, + { + "auxiliary_loss_clip": 0.03951266, + "auxiliary_loss_mlp": 0.01634243, + "balance_loss_clip": 1.84240294, + "balance_loss_mlp": 1.32601547, + "epoch": 0.0021643720315036373, + "flos": 63857001582720.0, + "grad_norm": 3.53942936408498, + "language_loss": 0.72176963, + "learning_rate": 2.0939181139872922e-06, + "loss": 0.77762473, + "num_input_tokens_seen": 462605, + "step": 18, + "time_per_iteration": 3.308528184890747 + }, + { + "auxiliary_loss_clip": 0.04256603, + "auxiliary_loss_mlp": 0.01628536, + "balance_loss_clip": 1.74103308, + "balance_loss_mlp": 1.3050499, + "epoch": 0.0022846149221427284, + "flos": 31284981192960.0, + "grad_norm": 5.569925843079255, + "language_loss": 1.01597548, + "learning_rate": 2.1330868934640175e-06, + "loss": 1.07482672, + "num_input_tokens_seen": 483280, + "step": 19, + "time_per_iteration": 3.010676145553589 + }, + { + "auxiliary_loss_clip": 0.03864964, + "auxiliary_loss_mlp": 0.01600496, + "balance_loss_clip": 1.83259273, + "balance_loss_mlp": 1.29989743, + "epoch": 0.002404857812781819, + "flos": 51083648161920.0, + "grad_norm": 3.5875357352662576, + "language_loss": 0.76448566, + "learning_rate": 2.170246112844971e-06, + "loss": 0.81914026, + "num_input_tokens_seen": 537620, + "step": 20, + "time_per_iteration": 3.076911687850952 + }, + { + "auxiliary_loss_clip": 0.04201567, + "auxiliary_loss_mlp": 0.01632013, + "balance_loss_clip": 1.73901224, + "balance_loss_mlp": 1.315202, + "epoch": 0.0025251007034209102, + "flos": 15815347309440.0, + "grad_norm": 4.27416864062976, + "language_loss": 1.01715279, + "learning_rate": 2.2055919496770983e-06, + "loss": 1.07548857, + "num_input_tokens_seen": 555760, + "step": 21, + "time_per_iteration": 2.9327187538146973 + }, + { + "auxiliary_loss_clip": 0.04134433, + "auxiliary_loss_mlp": 0.01629394, + "balance_loss_clip": 1.72642171, + "balance_loss_mlp": 1.32097614, + "epoch": 0.0026453435940600014, + "flos": 37851857458560.0, + "grad_norm": 5.5440140646503515, + "language_loss": 0.89554203, + "learning_rate": 2.2392931865974923e-06, + "loss": 0.95318031, + "num_input_tokens_seen": 578450, + "step": 22, + "time_per_iteration": 3.0610547065734863 + }, + { + "auxiliary_loss_clip": 0.04083547, + "auxiliary_loss_mlp": 0.01597519, + "balance_loss_clip": 1.72180581, + "balance_loss_mlp": 1.29425061, + "epoch": 0.002765586484699092, + "flos": 21141976821120.0, + "grad_norm": 4.533276043663761, + "language_loss": 1.02000523, + "learning_rate": 2.271496085962064e-06, + "loss": 1.07681584, + "num_input_tokens_seen": 596145, + "step": 23, + "time_per_iteration": 2.975431442260742 + }, + { + "auxiliary_loss_clip": 0.04101684, + "auxiliary_loss_mlp": 0.0161366, + "balance_loss_clip": 1.7161243, + "balance_loss_mlp": 1.31039155, + "epoch": 0.002885829375338183, + "flos": 20667381396480.0, + "grad_norm": 3.0585544069513917, + "language_loss": 1.0264374, + "learning_rate": 2.3023282262611022e-06, + "loss": 1.08359087, + "num_input_tokens_seen": 614920, + "step": 24, + "time_per_iteration": 2.9276862144470215 + }, + { + "auxiliary_loss_clip": 0.04015572, + "auxiliary_loss_mlp": 0.01558253, + "balance_loss_clip": 1.714432, + "balance_loss_mlp": 1.26223254, + "epoch": 0.003006072265977274, + "flos": 34823869873920.0, + "grad_norm": 5.897246610295519, + "language_loss": 0.92500865, + "learning_rate": 2.3319015548620114e-06, + "loss": 0.98074698, + "num_input_tokens_seen": 636060, + "step": 25, + "time_per_iteration": 3.008657932281494 + }, + { + "auxiliary_loss_clip": 0.04007208, + "auxiliary_loss_mlp": 0.0156095, + "balance_loss_clip": 1.71641111, + "balance_loss_mlp": 1.27313066, + "epoch": 0.003126315156616365, + "flos": 24422021118720.0, + "grad_norm": 2.6014475973266586, + "language_loss": 0.92865956, + "learning_rate": 2.3603148416618152e-06, + "loss": 0.98434108, + "num_input_tokens_seen": 655575, + "step": 26, + "time_per_iteration": 2.913837194442749 + }, + { + "auxiliary_loss_clip": 0.03937785, + "auxiliary_loss_mlp": 0.01558402, + "balance_loss_clip": 1.70668387, + "balance_loss_mlp": 1.27554262, + "epoch": 0.003246558047255456, + "flos": 23622326674560.0, + "grad_norm": 2.4992817601738935, + "language_loss": 1.0105226, + "learning_rate": 2.3876556694204647e-06, + "loss": 1.06548452, + "num_input_tokens_seen": 675730, + "step": 27, + "time_per_iteration": 2.7904489040374756 + }, + { + "auxiliary_loss_clip": 0.03910412, + "auxiliary_loss_mlp": 0.01573428, + "balance_loss_clip": 1.70628357, + "balance_loss_mlp": 1.28236675, + "epoch": 0.003366800937894547, + "flos": 17820275725440.0, + "grad_norm": 3.695145776780836, + "language_loss": 0.9071995, + "learning_rate": 2.414002061950908e-06, + "loss": 0.96203786, + "num_input_tokens_seen": 694605, + "step": 28, + "time_per_iteration": 2.9004316329956055 + }, + { + "auxiliary_loss_clip": 0.03869791, + "auxiliary_loss_mlp": 0.01541681, + "balance_loss_clip": 1.69560826, + "balance_loss_mlp": 1.2693119, + "epoch": 0.003487043828533638, + "flos": 24426115269120.0, + "grad_norm": 2.394380086802325, + "language_loss": 1.00005507, + "learning_rate": 2.4394238264681557e-06, + "loss": 1.05416989, + "num_input_tokens_seen": 714340, + "step": 29, + "time_per_iteration": 2.923031806945801 + }, + { + "auxiliary_loss_clip": 0.03868448, + "auxiliary_loss_mlp": 0.01551557, + "balance_loss_clip": 1.69112277, + "balance_loss_mlp": 1.26717091, + "epoch": 0.003607286719172729, + "flos": 26140311002880.0, + "grad_norm": 2.0982191340105434, + "language_loss": 0.99608958, + "learning_rate": 2.4639836682781433e-06, + "loss": 1.05028963, + "num_input_tokens_seen": 734470, + "step": 30, + "time_per_iteration": 2.928041696548462 + }, + { + "auxiliary_loss_clip": 0.03781854, + "auxiliary_loss_mlp": 0.01523815, + "balance_loss_clip": 1.69147444, + "balance_loss_mlp": 1.25030136, + "epoch": 0.00372752960981182, + "flos": 20593082113920.0, + "grad_norm": 2.649361784092686, + "language_loss": 1.00433648, + "learning_rate": 2.487738122623307e-06, + "loss": 1.05739307, + "num_input_tokens_seen": 753380, + "step": 31, + "time_per_iteration": 2.8465628623962402 + }, + { + "auxiliary_loss_clip": 0.03736402, + "auxiliary_loss_mlp": 0.01519329, + "balance_loss_clip": 1.67627287, + "balance_loss_mlp": 1.24409842, + "epoch": 0.003847772500450911, + "flos": 22674608282880.0, + "grad_norm": 2.4179201712214202, + "language_loss": 0.9893167, + "learning_rate": 2.510738338534912e-06, + "loss": 1.04187393, + "num_input_tokens_seen": 772105, + "step": 32, + "time_per_iteration": 2.869523525238037 + }, + { + "auxiliary_loss_clip": 0.03590878, + "auxiliary_loss_mlp": 0.01509911, + "balance_loss_clip": 1.65113974, + "balance_loss_mlp": 1.23925805, + "epoch": 0.003968015391090002, + "flos": 17967796882560.0, + "grad_norm": 2.85518944121255, + "language_loss": 1.02639556, + "learning_rate": 2.5330307420306648e-06, + "loss": 1.07740331, + "num_input_tokens_seen": 788955, + "step": 33, + "time_per_iteration": 2.8940393924713135 + }, + { + "auxiliary_loss_clip": 0.03510087, + "auxiliary_loss_mlp": 0.01498441, + "balance_loss_clip": 1.64156914, + "balance_loss_mlp": 1.24266529, + "epoch": 0.004088258281729093, + "flos": 27304103658240.0, + "grad_norm": 2.248911870783041, + "language_loss": 0.88131452, + "learning_rate": 2.554657600279796e-06, + "loss": 0.93139982, + "num_input_tokens_seen": 810230, + "step": 34, + "time_per_iteration": 2.926544427871704 + }, + { + "auxiliary_loss_clip": 0.03450083, + "auxiliary_loss_mlp": 0.01504498, + "balance_loss_clip": 1.62058091, + "balance_loss_mlp": 1.24967599, + "epoch": 0.004208501172368184, + "flos": 23258587599360.0, + "grad_norm": 2.2703391730504046, + "language_loss": 1.03546894, + "learning_rate": 2.5756575039679493e-06, + "loss": 1.0850147, + "num_input_tokens_seen": 829780, + "step": 35, + "time_per_iteration": 2.866823434829712 + }, + { + "auxiliary_loss_clip": 0.03387856, + "auxiliary_loss_mlp": 0.01489291, + "balance_loss_clip": 1.6135807, + "balance_loss_mlp": 1.23160803, + "epoch": 0.0043287440630072746, + "flos": 17312104062720.0, + "grad_norm": 1.9936792732809108, + "language_loss": 0.95085973, + "learning_rate": 2.5960657816942747e-06, + "loss": 0.99963123, + "num_input_tokens_seen": 848695, + "step": 36, + "time_per_iteration": 2.879009962081909 + }, + { + "auxiliary_loss_clip": 0.02664295, + "auxiliary_loss_mlp": 0.01518805, + "balance_loss_clip": 1.57439375, + "balance_loss_mlp": 1.28687119, + "epoch": 0.004448986953646365, + "flos": 53092491160320.0, + "grad_norm": 1.4061246103939655, + "language_loss": 0.60961747, + "learning_rate": 2.6159148575788668e-06, + "loss": 0.65144849, + "num_input_tokens_seen": 906730, + "step": 37, + "time_per_iteration": 3.3135838508605957 + }, + { + "auxiliary_loss_clip": 0.0326435, + "auxiliary_loss_mlp": 0.01443924, + "balance_loss_clip": 1.59548712, + "balance_loss_mlp": 1.20111895, + "epoch": 0.004569229844285457, + "flos": 13444165866240.0, + "grad_norm": 2.288405264507453, + "language_loss": 0.98747444, + "learning_rate": 2.635234561171e-06, + "loss": 1.0345571, + "num_input_tokens_seen": 925125, + "step": 38, + "time_per_iteration": 2.860964298248291 + }, + { + "auxiliary_loss_clip": 0.03208183, + "auxiliary_loss_mlp": 0.01425635, + "balance_loss_clip": 1.58114576, + "balance_loss_mlp": 1.20342886, + "epoch": 0.0046894727349245475, + "flos": 16209609966720.0, + "grad_norm": 2.3803086652878633, + "language_loss": 0.94175255, + "learning_rate": 2.6540523970949877e-06, + "loss": 0.98809081, + "num_input_tokens_seen": 939970, + "step": 39, + "time_per_iteration": 2.808168649673462 + }, + { + "auxiliary_loss_clip": 0.03136876, + "auxiliary_loss_mlp": 0.01409062, + "balance_loss_clip": 1.57150829, + "balance_loss_mlp": 1.18609285, + "epoch": 0.004809715625563638, + "flos": 23914244505600.0, + "grad_norm": 2.583965033669434, + "language_loss": 0.92303169, + "learning_rate": 2.6723937805519533e-06, + "loss": 0.96849114, + "num_input_tokens_seen": 957470, + "step": 40, + "time_per_iteration": 3.814314603805542 + }, + { + "auxiliary_loss_clip": 0.03118283, + "auxiliary_loss_mlp": 0.01405024, + "balance_loss_clip": 1.56453264, + "balance_loss_mlp": 1.18110156, + "epoch": 0.00492995851620273, + "flos": 20773030273920.0, + "grad_norm": 2.033239281135742, + "language_loss": 0.93118036, + "learning_rate": 2.690282243737839e-06, + "loss": 0.97641337, + "num_input_tokens_seen": 976405, + "step": 41, + "time_per_iteration": 3.8798575401306152 + }, + { + "auxiliary_loss_clip": 0.03068277, + "auxiliary_loss_mlp": 0.01369013, + "balance_loss_clip": 1.55184674, + "balance_loss_mlp": 1.15252876, + "epoch": 0.0050502014068418205, + "flos": 20338655103360.0, + "grad_norm": 2.984929726441715, + "language_loss": 0.99363613, + "learning_rate": 2.7077396173840807e-06, + "loss": 1.03800905, + "num_input_tokens_seen": 994690, + "step": 42, + "time_per_iteration": 2.969465494155884 + }, + { + "auxiliary_loss_clip": 0.03049999, + "auxiliary_loss_mlp": 0.01397825, + "balance_loss_clip": 1.54701853, + "balance_loss_mlp": 1.18286693, + "epoch": 0.005170444297480911, + "flos": 25994872834560.0, + "grad_norm": 3.3710060483694337, + "language_loss": 0.92714715, + "learning_rate": 2.7247861909342594e-06, + "loss": 0.97162539, + "num_input_tokens_seen": 1015615, + "step": 43, + "time_per_iteration": 2.9625837802886963 + }, + { + "auxiliary_loss_clip": 0.029301, + "auxiliary_loss_mlp": 0.01350821, + "balance_loss_clip": 1.52174091, + "balance_loss_mlp": 1.15073991, + "epoch": 0.005290687188120003, + "flos": 20954055841920.0, + "grad_norm": 2.554429216195915, + "language_loss": 0.83197427, + "learning_rate": 2.7414408543044743e-06, + "loss": 0.87478346, + "num_input_tokens_seen": 1031255, + "step": 44, + "time_per_iteration": 2.9961631298065186 + }, + { + "auxiliary_loss_clip": 0.02897669, + "auxiliary_loss_mlp": 0.01331936, + "balance_loss_clip": 1.50919688, + "balance_loss_mlp": 1.13433456, + "epoch": 0.005410930078759093, + "flos": 15851401585920.0, + "grad_norm": 5.006323059511652, + "language_loss": 0.79348284, + "learning_rate": 2.7577212237113157e-06, + "loss": 0.83577889, + "num_input_tokens_seen": 1048295, + "step": 45, + "time_per_iteration": 2.7773003578186035 + }, + { + "auxiliary_loss_clip": 0.02872042, + "auxiliary_loss_mlp": 0.01326209, + "balance_loss_clip": 1.51389003, + "balance_loss_mlp": 1.12708187, + "epoch": 0.005531172969398184, + "flos": 21104988791040.0, + "grad_norm": 5.809991266010175, + "language_loss": 1.04217339, + "learning_rate": 2.7736437536690466e-06, + "loss": 1.0841558, + "num_input_tokens_seen": 1067925, + "step": 46, + "time_per_iteration": 2.853935480117798 + }, + { + "auxiliary_loss_clip": 0.02854679, + "auxiliary_loss_mlp": 0.01312159, + "balance_loss_clip": 1.50221133, + "balance_loss_mlp": 1.12991166, + "epoch": 0.005651415860037276, + "flos": 20844887431680.0, + "grad_norm": 2.180347849804833, + "language_loss": 1.07868528, + "learning_rate": 2.789223836941131e-06, + "loss": 1.12035358, + "num_input_tokens_seen": 1088060, + "step": 47, + "time_per_iteration": 2.851883888244629 + }, + { + "auxiliary_loss_clip": 0.02779858, + "auxiliary_loss_mlp": 0.01326256, + "balance_loss_clip": 1.48417282, + "balance_loss_mlp": 1.14048004, + "epoch": 0.005771658750676366, + "flos": 13260195383040.0, + "grad_norm": 3.9591069067748315, + "language_loss": 1.08753204, + "learning_rate": 2.8044758939680847e-06, + "loss": 1.12859321, + "num_input_tokens_seen": 1104130, + "step": 48, + "time_per_iteration": 2.7923743724823 + }, + { + "auxiliary_loss_clip": 0.02768096, + "auxiliary_loss_mlp": 0.01312523, + "balance_loss_clip": 1.4825964, + "balance_loss_mlp": 1.12522173, + "epoch": 0.005891901641315457, + "flos": 24425396997120.0, + "grad_norm": 3.382659031465294, + "language_loss": 1.02062011, + "learning_rate": 2.8194134530738863e-06, + "loss": 1.06142628, + "num_input_tokens_seen": 1122900, + "step": 49, + "time_per_iteration": 2.9603431224823 + }, + { + "auxiliary_loss_clip": 0.02707533, + "auxiliary_loss_mlp": 0.0130456, + "balance_loss_clip": 1.47490382, + "balance_loss_mlp": 1.12937021, + "epoch": 0.006012144531954548, + "flos": 23076197314560.0, + "grad_norm": 2.9081347451537867, + "language_loss": 0.90397507, + "learning_rate": 2.834049222568994e-06, + "loss": 0.94409609, + "num_input_tokens_seen": 1140250, + "step": 50, + "time_per_iteration": 2.858058214187622 + }, + { + "auxiliary_loss_clip": 0.02719164, + "auxiliary_loss_mlp": 0.01266274, + "balance_loss_clip": 1.4745729, + "balance_loss_mlp": 1.09251451, + "epoch": 0.006132387422593639, + "flos": 22528775064960.0, + "grad_norm": 1.8186639237326832, + "language_loss": 0.92643154, + "learning_rate": 2.848395155712969e-06, + "loss": 0.96628594, + "num_input_tokens_seen": 1160470, + "step": 51, + "time_per_iteration": 2.8594319820404053 + }, + { + "auxiliary_loss_clip": 0.02666009, + "auxiliary_loss_mlp": 0.01288152, + "balance_loss_clip": 1.46203375, + "balance_loss_mlp": 1.11038709, + "epoch": 0.00625263031323273, + "flos": 27628340751360.0, + "grad_norm": 2.6064261359299326, + "language_loss": 0.97687614, + "learning_rate": 2.8624625093687977e-06, + "loss": 1.01641774, + "num_input_tokens_seen": 1177605, + "step": 52, + "time_per_iteration": 2.889244318008423 + }, + { + "auxiliary_loss_clip": 0.02646851, + "auxiliary_loss_mlp": 0.01261601, + "balance_loss_clip": 1.45617282, + "balance_loss_mlp": 1.0959475, + "epoch": 0.006372873203871821, + "flos": 23110671392640.0, + "grad_norm": 2.23278468165917, + "language_loss": 0.89031088, + "learning_rate": 2.876261897070029e-06, + "loss": 0.92939532, + "num_input_tokens_seen": 1197735, + "step": 53, + "time_per_iteration": 2.9232065677642822 + }, + { + "auxiliary_loss_clip": 0.02627762, + "auxiliary_loss_mlp": 0.01252549, + "balance_loss_clip": 1.45061541, + "balance_loss_mlp": 1.09004283, + "epoch": 0.006493116094510912, + "flos": 22856028900480.0, + "grad_norm": 3.151317409547167, + "language_loss": 0.92673397, + "learning_rate": 2.889803337127447e-06, + "loss": 0.96553707, + "num_input_tokens_seen": 1216335, + "step": 54, + "time_per_iteration": 2.966370105743408 + }, + { + "auxiliary_loss_clip": 0.02565656, + "auxiliary_loss_mlp": 0.01265377, + "balance_loss_clip": 1.43674588, + "balance_loss_mlp": 1.09714866, + "epoch": 0.006613358985150003, + "flos": 23071708114560.0, + "grad_norm": 3.6049201010748897, + "language_loss": 0.84811682, + "learning_rate": 2.903096296321516e-06, + "loss": 0.88642716, + "num_input_tokens_seen": 1234480, + "step": 55, + "time_per_iteration": 2.8994176387786865 + }, + { + "auxiliary_loss_clip": 0.02553699, + "auxiliary_loss_mlp": 0.01227015, + "balance_loss_clip": 1.43419647, + "balance_loss_mlp": 1.07252002, + "epoch": 0.006733601875789094, + "flos": 26537662229760.0, + "grad_norm": 2.3029726664102377, + "language_loss": 0.91650665, + "learning_rate": 2.9161497296578907e-06, + "loss": 0.95431387, + "num_input_tokens_seen": 1253870, + "step": 56, + "time_per_iteration": 2.9433486461639404 + }, + { + "auxiliary_loss_clip": 0.02552635, + "auxiliary_loss_mlp": 0.01236835, + "balance_loss_clip": 1.43897462, + "balance_loss_mlp": 1.07852554, + "epoch": 0.006853844766428185, + "flos": 15523178083200.0, + "grad_norm": 2.390580754614046, + "language_loss": 0.85888171, + "learning_rate": 2.928972116604173e-06, + "loss": 0.89677638, + "num_input_tokens_seen": 1270145, + "step": 57, + "time_per_iteration": 2.784221887588501 + }, + { + "auxiliary_loss_clip": 0.02505199, + "auxiliary_loss_mlp": 0.01214585, + "balance_loss_clip": 1.42187095, + "balance_loss_mlp": 1.0684824, + "epoch": 0.006974087657067276, + "flos": 24243760897920.0, + "grad_norm": 10.261608857386644, + "language_loss": 1.02151275, + "learning_rate": 2.9415714941751377e-06, + "loss": 1.05871058, + "num_input_tokens_seen": 1291365, + "step": 58, + "time_per_iteration": 2.930610179901123 + }, + { + "auxiliary_loss_clip": 0.02518014, + "auxiliary_loss_mlp": 0.01219343, + "balance_loss_clip": 1.42605317, + "balance_loss_mlp": 1.07028413, + "epoch": 0.007094330547706367, + "flos": 25772513690880.0, + "grad_norm": 2.559329446548028, + "language_loss": 0.93550479, + "learning_rate": 2.9539554871897396e-06, + "loss": 0.97287834, + "num_input_tokens_seen": 1311535, + "step": 59, + "time_per_iteration": 2.8629820346832275 + }, + { + "auxiliary_loss_clip": 0.02446574, + "auxiliary_loss_mlp": 0.0124349, + "balance_loss_clip": 1.40756702, + "balance_loss_mlp": 1.09299994, + "epoch": 0.007214573438345458, + "flos": 21319015979520.0, + "grad_norm": 2.271827301581079, + "language_loss": 0.97431481, + "learning_rate": 2.9661313359851253e-06, + "loss": 1.01121545, + "num_input_tokens_seen": 1329420, + "step": 60, + "time_per_iteration": 2.9655628204345703 + }, + { + "auxiliary_loss_clip": 0.02413297, + "auxiliary_loss_mlp": 0.01206203, + "balance_loss_clip": 1.39979613, + "balance_loss_mlp": 1.06019497, + "epoch": 0.007334816328984549, + "flos": 24937088192640.0, + "grad_norm": 1.9155951185775038, + "language_loss": 0.94132715, + "learning_rate": 2.978105921839922e-06, + "loss": 0.97752213, + "num_input_tokens_seen": 1349965, + "step": 61, + "time_per_iteration": 2.89477276802063 + }, + { + "auxiliary_loss_clip": 0.02392328, + "auxiliary_loss_mlp": 0.01190254, + "balance_loss_clip": 1.39389777, + "balance_loss_mlp": 1.05759764, + "epoch": 0.00745505921962364, + "flos": 18510586277760.0, + "grad_norm": 2.3424259985133724, + "language_loss": 0.72171021, + "learning_rate": 2.9898857903302893e-06, + "loss": 0.75753605, + "num_input_tokens_seen": 1368915, + "step": 62, + "time_per_iteration": 2.801839590072632 + }, + { + "auxiliary_loss_clip": 0.02339784, + "auxiliary_loss_mlp": 0.01191185, + "balance_loss_clip": 1.38660741, + "balance_loss_mlp": 1.05528665, + "epoch": 0.007575302110262731, + "flos": 18477656484480.0, + "grad_norm": 3.4834549710338787, + "language_loss": 0.87974012, + "learning_rate": 3.001477172817253e-06, + "loss": 0.91504979, + "num_input_tokens_seen": 1386805, + "step": 63, + "time_per_iteration": 2.833784341812134 + }, + { + "auxiliary_loss_clip": 0.0235811, + "auxiliary_loss_mlp": 0.01195623, + "balance_loss_clip": 1.38883984, + "balance_loss_mlp": 1.05667281, + "epoch": 0.007695545000901822, + "flos": 24973178382720.0, + "grad_norm": 2.7759693136368773, + "language_loss": 0.96443444, + "learning_rate": 3.012886006241894e-06, + "loss": 0.99997175, + "num_input_tokens_seen": 1406190, + "step": 64, + "time_per_iteration": 2.813711166381836 + }, + { + "auxiliary_loss_clip": 0.02349609, + "auxiliary_loss_mlp": 0.01202858, + "balance_loss_clip": 1.38447952, + "balance_loss_mlp": 1.06629205, + "epoch": 0.007815787891540913, + "flos": 21324223451520.0, + "grad_norm": 2.825098346476438, + "language_loss": 0.88241518, + "learning_rate": 3.0241179513858383e-06, + "loss": 0.9179399, + "num_input_tokens_seen": 1425500, + "step": 65, + "time_per_iteration": 2.864471673965454 + }, + { + "auxiliary_loss_clip": 0.02312913, + "auxiliary_loss_mlp": 0.01190589, + "balance_loss_clip": 1.37100244, + "balance_loss_mlp": 1.05354595, + "epoch": 0.007936030782180003, + "flos": 21575777374080.0, + "grad_norm": 2.548461562202195, + "language_loss": 0.87785137, + "learning_rate": 3.035178409737647e-06, + "loss": 0.91288632, + "num_input_tokens_seen": 1442950, + "step": 66, + "time_per_iteration": 3.800933599472046 + }, + { + "auxiliary_loss_clip": 0.02268805, + "auxiliary_loss_mlp": 0.0118392, + "balance_loss_clip": 1.36356378, + "balance_loss_mlp": 1.0617547, + "epoch": 0.008056273672819095, + "flos": 20120785159680.0, + "grad_norm": 2.734116917580113, + "language_loss": 0.88674796, + "learning_rate": 3.046072539090907e-06, + "loss": 0.9212752, + "num_input_tokens_seen": 1460915, + "step": 67, + "time_per_iteration": 5.739641904830933 + }, + { + "auxiliary_loss_clip": 0.02250473, + "auxiliary_loss_mlp": 0.01200032, + "balance_loss_clip": 1.35887909, + "balance_loss_mlp": 1.0694741, + "epoch": 0.008176516563458186, + "flos": 18333116156160.0, + "grad_norm": 2.9064945576069667, + "language_loss": 1.04755163, + "learning_rate": 3.056805267986779e-06, + "loss": 1.08205676, + "num_input_tokens_seen": 1478385, + "step": 68, + "time_per_iteration": 2.8098576068878174 + }, + { + "auxiliary_loss_clip": 0.02215037, + "auxiliary_loss_mlp": 0.01178779, + "balance_loss_clip": 1.34829926, + "balance_loss_mlp": 1.0630033, + "epoch": 0.008296759454097276, + "flos": 21872076664320.0, + "grad_norm": 2.973782197327817, + "language_loss": 0.95220184, + "learning_rate": 3.0673813091022194e-06, + "loss": 0.98613995, + "num_input_tokens_seen": 1497605, + "step": 69, + "time_per_iteration": 2.857753038406372 + }, + { + "auxiliary_loss_clip": 0.01843555, + "auxiliary_loss_mlp": 0.01168497, + "balance_loss_clip": 1.31418836, + "balance_loss_mlp": 1.05787075, + "epoch": 0.008417002344736368, + "flos": 63408228036480.0, + "grad_norm": 1.2684428895189213, + "language_loss": 0.62050712, + "learning_rate": 3.0778051716749317e-06, + "loss": 0.65062761, + "num_input_tokens_seen": 1561150, + "step": 70, + "time_per_iteration": 3.393846273422241 + }, + { + "auxiliary_loss_clip": 0.02196016, + "auxiliary_loss_mlp": 0.0118775, + "balance_loss_clip": 1.3364079, + "balance_loss_mlp": 1.06720591, + "epoch": 0.008537245235375458, + "flos": 22966454286720.0, + "grad_norm": 2.274283904986527, + "language_loss": 0.90547884, + "learning_rate": 3.0880811730470094e-06, + "loss": 0.93931651, + "num_input_tokens_seen": 1580605, + "step": 71, + "time_per_iteration": 2.8481554985046387 + }, + { + "auxiliary_loss_clip": 0.0180564, + "auxiliary_loss_mlp": 0.01143519, + "balance_loss_clip": 1.299119, + "balance_loss_mlp": 1.0405221, + "epoch": 0.008657488126014549, + "flos": 61984046712960.0, + "grad_norm": 1.1873882515600478, + "language_loss": 0.58612949, + "learning_rate": 3.098213449401257e-06, + "loss": 0.61562109, + "num_input_tokens_seen": 1647535, + "step": 72, + "time_per_iteration": 3.311267137527466 + }, + { + "auxiliary_loss_clip": 0.02166303, + "auxiliary_loss_mlp": 0.01172776, + "balance_loss_clip": 1.33302104, + "balance_loss_mlp": 1.06100559, + "epoch": 0.00877773101665364, + "flos": 30296791152000.0, + "grad_norm": 2.97565243953495, + "language_loss": 0.98960364, + "learning_rate": 3.1082059657570015e-06, + "loss": 1.0229944, + "num_input_tokens_seen": 1666770, + "step": 73, + "time_per_iteration": 2.8828492164611816 + }, + { + "auxiliary_loss_clip": 0.02144077, + "auxiliary_loss_mlp": 0.01170435, + "balance_loss_clip": 1.32168639, + "balance_loss_mlp": 1.05914092, + "epoch": 0.00889797390729273, + "flos": 23514056104320.0, + "grad_norm": 2.1698822182567583, + "language_loss": 0.96720028, + "learning_rate": 3.1180625252858496e-06, + "loss": 1.00034547, + "num_input_tokens_seen": 1685200, + "step": 74, + "time_per_iteration": 2.832388162612915 + }, + { + "auxiliary_loss_clip": 0.02085628, + "auxiliary_loss_mlp": 0.01163386, + "balance_loss_clip": 1.3081677, + "balance_loss_mlp": 1.05871987, + "epoch": 0.009018216797931822, + "flos": 23075838178560.0, + "grad_norm": 2.9476700089150194, + "language_loss": 0.80119717, + "learning_rate": 3.1277867780021663e-06, + "loss": 0.83368725, + "num_input_tokens_seen": 1701835, + "step": 75, + "time_per_iteration": 2.819680690765381 + }, + { + "auxiliary_loss_clip": 0.02055518, + "auxiliary_loss_mlp": 0.01147223, + "balance_loss_clip": 1.29908836, + "balance_loss_mlp": 1.05118775, + "epoch": 0.009138459688570914, + "flos": 15918877284480.0, + "grad_norm": 1.9952258219490144, + "language_loss": 0.95734775, + "learning_rate": 3.1373822288779824e-06, + "loss": 0.98937511, + "num_input_tokens_seen": 1718415, + "step": 76, + "time_per_iteration": 2.890353202819824 + }, + { + "auxiliary_loss_clip": 0.02082179, + "auxiliary_loss_mlp": 0.01178075, + "balance_loss_clip": 1.30896616, + "balance_loss_mlp": 1.06907046, + "epoch": 0.009258702579210003, + "flos": 27016531372800.0, + "grad_norm": 2.5369785093940527, + "language_loss": 0.79671478, + "learning_rate": 3.1468522454274533e-06, + "loss": 0.82931733, + "num_input_tokens_seen": 1738770, + "step": 77, + "time_per_iteration": 2.9673922061920166 + }, + { + "auxiliary_loss_clip": 0.0205111, + "auxiliary_loss_mlp": 0.01152871, + "balance_loss_clip": 1.29363847, + "balance_loss_mlp": 1.05564368, + "epoch": 0.009378945469849095, + "flos": 26903196984960.0, + "grad_norm": 2.05966247727897, + "language_loss": 0.91806144, + "learning_rate": 3.15620006480197e-06, + "loss": 0.95010126, + "num_input_tokens_seen": 1758040, + "step": 78, + "time_per_iteration": 2.864126205444336 + }, + { + "auxiliary_loss_clip": 0.0204541, + "auxiliary_loss_mlp": 0.01154434, + "balance_loss_clip": 1.30035186, + "balance_loss_mlp": 1.05515671, + "epoch": 0.009499188360488187, + "flos": 35694236327040.0, + "grad_norm": 2.908428738448974, + "language_loss": 0.75055116, + "learning_rate": 3.1654288004333087e-06, + "loss": 0.78254962, + "num_input_tokens_seen": 1776705, + "step": 79, + "time_per_iteration": 3.079855442047119 + }, + { + "auxiliary_loss_clip": 0.02007487, + "auxiliary_loss_mlp": 0.01153843, + "balance_loss_clip": 1.28878975, + "balance_loss_mlp": 1.06147981, + "epoch": 0.009619431251127276, + "flos": 21503201944320.0, + "grad_norm": 177.08524412737856, + "language_loss": 0.76235282, + "learning_rate": 3.1745414482589353e-06, + "loss": 0.79396617, + "num_input_tokens_seen": 1795915, + "step": 80, + "time_per_iteration": 2.839477777481079 + }, + { + "auxiliary_loss_clip": 0.0200029, + "auxiliary_loss_mlp": 0.01127513, + "balance_loss_clip": 1.28645921, + "balance_loss_mlp": 1.0409193, + "epoch": 0.009739674141766368, + "flos": 17421056991360.0, + "grad_norm": 2.323297459755085, + "language_loss": 0.87131763, + "learning_rate": 3.1835408925606204e-06, + "loss": 0.90259564, + "num_input_tokens_seen": 1814055, + "step": 81, + "time_per_iteration": 2.904207944869995 + }, + { + "auxiliary_loss_clip": 0.01973942, + "auxiliary_loss_mlp": 0.01144906, + "balance_loss_clip": 1.28141224, + "balance_loss_mlp": 1.05578518, + "epoch": 0.00985991703240546, + "flos": 27527109246720.0, + "grad_norm": 2.4212653796518673, + "language_loss": 0.89502436, + "learning_rate": 3.1924299114448214e-06, + "loss": 0.92621279, + "num_input_tokens_seen": 1834535, + "step": 82, + "time_per_iteration": 2.972604513168335 + }, + { + "auxiliary_loss_clip": 0.01982454, + "auxiliary_loss_mlp": 0.01167349, + "balance_loss_clip": 1.28358531, + "balance_loss_mlp": 1.07617807, + "epoch": 0.00998015992304455, + "flos": 13808084509440.0, + "grad_norm": 2.050421141840248, + "language_loss": 0.83608508, + "learning_rate": 3.2012111819909055e-06, + "loss": 0.8675831, + "num_input_tokens_seen": 1851865, + "step": 83, + "time_per_iteration": 2.877596139907837 + }, + { + "auxiliary_loss_clip": 0.01953863, + "auxiliary_loss_mlp": 0.0114406, + "balance_loss_clip": 1.27436876, + "balance_loss_mlp": 1.05794287, + "epoch": 0.010100402813683641, + "flos": 20191385341440.0, + "grad_norm": 2.311314887519752, + "language_loss": 0.95158905, + "learning_rate": 3.2098872850910627e-06, + "loss": 0.98256826, + "num_input_tokens_seen": 1868540, + "step": 84, + "time_per_iteration": 2.8618783950805664 + }, + { + "auxiliary_loss_clip": 0.01950255, + "auxiliary_loss_mlp": 0.01148633, + "balance_loss_clip": 1.27695727, + "balance_loss_mlp": 1.06642652, + "epoch": 0.010220645704322733, + "flos": 17201642762880.0, + "grad_norm": 2.1007786358137, + "language_loss": 0.89126474, + "learning_rate": 3.2184607100038194e-06, + "loss": 0.92225361, + "num_input_tokens_seen": 1887180, + "step": 85, + "time_per_iteration": 2.8909735679626465 + }, + { + "auxiliary_loss_clip": 0.01931204, + "auxiliary_loss_mlp": 0.01132112, + "balance_loss_clip": 1.26729667, + "balance_loss_mlp": 1.05186081, + "epoch": 0.010340888594961822, + "flos": 21470415805440.0, + "grad_norm": 2.2679684674957445, + "language_loss": 0.93237174, + "learning_rate": 3.2269338586412414e-06, + "loss": 0.96300489, + "num_input_tokens_seen": 1904765, + "step": 86, + "time_per_iteration": 2.896695137023926 + }, + { + "auxiliary_loss_clip": 0.01905038, + "auxiliary_loss_mlp": 0.01131935, + "balance_loss_clip": 1.26301765, + "balance_loss_mlp": 1.05149245, + "epoch": 0.010461131485600914, + "flos": 23002831785600.0, + "grad_norm": 2.179760397555901, + "language_loss": 0.96714067, + "learning_rate": 3.2353090496083106e-06, + "loss": 0.99751043, + "num_input_tokens_seen": 1922600, + "step": 87, + "time_per_iteration": 3.0500197410583496 + }, + { + "auxiliary_loss_clip": 0.01882702, + "auxiliary_loss_mlp": 0.01133125, + "balance_loss_clip": 1.25637484, + "balance_loss_mlp": 1.06212437, + "epoch": 0.010581374376240005, + "flos": 33546850571520.0, + "grad_norm": 1.8988791662177795, + "language_loss": 0.81396639, + "learning_rate": 3.2435885220114572e-06, + "loss": 0.84412467, + "num_input_tokens_seen": 1943950, + "step": 88, + "time_per_iteration": 3.007077932357788 + }, + { + "auxiliary_loss_clip": 0.01896446, + "auxiliary_loss_mlp": 0.01127016, + "balance_loss_clip": 1.26183939, + "balance_loss_mlp": 1.05210519, + "epoch": 0.010701617266879095, + "flos": 21763087822080.0, + "grad_norm": 3.8659531063837265, + "language_loss": 0.94111735, + "learning_rate": 3.2517744390519113e-06, + "loss": 0.97135198, + "num_input_tokens_seen": 1962815, + "step": 89, + "time_per_iteration": 2.916466236114502 + }, + { + "auxiliary_loss_clip": 0.01866079, + "auxiliary_loss_mlp": 0.01114854, + "balance_loss_clip": 1.24812484, + "balance_loss_mlp": 1.04523587, + "epoch": 0.010821860157518187, + "flos": 19060199256960.0, + "grad_norm": 2.346972414604285, + "language_loss": 0.75183761, + "learning_rate": 3.259868891418298e-06, + "loss": 0.78164697, + "num_input_tokens_seen": 1980580, + "step": 90, + "time_per_iteration": 2.8287370204925537 + }, + { + "auxiliary_loss_clip": 0.01855754, + "auxiliary_loss_mlp": 0.01119872, + "balance_loss_clip": 1.25183642, + "balance_loss_mlp": 1.04758394, + "epoch": 0.010942103048157278, + "flos": 25447378757760.0, + "grad_norm": 2.3877116113791983, + "language_loss": 0.85122633, + "learning_rate": 3.2678739004917757e-06, + "loss": 0.88098258, + "num_input_tokens_seen": 2000315, + "step": 91, + "time_per_iteration": 2.949753761291504 + }, + { + "auxiliary_loss_clip": 0.01849353, + "auxiliary_loss_mlp": 0.01124357, + "balance_loss_clip": 1.25200522, + "balance_loss_mlp": 1.05392861, + "epoch": 0.011062345938796368, + "flos": 27493928058240.0, + "grad_norm": 1.939202895292928, + "language_loss": 0.92197025, + "learning_rate": 3.275791421376029e-06, + "loss": 0.9517073, + "num_input_tokens_seen": 2023760, + "step": 92, + "time_per_iteration": 3.910886764526367 + }, + { + "auxiliary_loss_clip": 0.01833921, + "auxiliary_loss_mlp": 0.01118425, + "balance_loss_clip": 1.24209213, + "balance_loss_mlp": 1.05061913, + "epoch": 0.01118258882943546, + "flos": 16071210864000.0, + "grad_norm": 2.415185460834783, + "language_loss": 0.96309072, + "learning_rate": 3.2836233457634622e-06, + "loss": 0.99261415, + "num_input_tokens_seen": 2041895, + "step": 93, + "time_per_iteration": 5.610487937927246 + }, + { + "auxiliary_loss_clip": 0.01844651, + "auxiliary_loss_mlp": 0.01126797, + "balance_loss_clip": 1.24468756, + "balance_loss_mlp": 1.05617785, + "epoch": 0.011302831720074551, + "flos": 20668602458880.0, + "grad_norm": 2.0983156667121636, + "language_loss": 0.85708869, + "learning_rate": 3.2913715046481135e-06, + "loss": 0.88680315, + "num_input_tokens_seen": 2061640, + "step": 94, + "time_per_iteration": 2.833284378051758 + }, + { + "auxiliary_loss_clip": 0.01817868, + "auxiliary_loss_mlp": 0.01112619, + "balance_loss_clip": 1.23828125, + "balance_loss_mlp": 1.04843664, + "epoch": 0.011423074610713641, + "flos": 13072238490240.0, + "grad_norm": 2.3191067791545477, + "language_loss": 0.88915437, + "learning_rate": 3.299037670895023e-06, + "loss": 0.9184593, + "num_input_tokens_seen": 2078255, + "step": 95, + "time_per_iteration": 2.9152004718780518 + }, + { + "auxiliary_loss_clip": 0.01817126, + "auxiliary_loss_mlp": 0.01127189, + "balance_loss_clip": 1.24020958, + "balance_loss_mlp": 1.06119514, + "epoch": 0.011543317501352733, + "flos": 30335646689280.0, + "grad_norm": 1.8459995520226995, + "language_loss": 0.80584377, + "learning_rate": 3.3066235616750667e-06, + "loss": 0.83528697, + "num_input_tokens_seen": 2099490, + "step": 96, + "time_per_iteration": 2.9001660346984863 + }, + { + "auxiliary_loss_clip": 0.01796534, + "auxiliary_loss_mlp": 0.01119641, + "balance_loss_clip": 1.23297417, + "balance_loss_mlp": 1.05932128, + "epoch": 0.011663560391991824, + "flos": 15522962601600.0, + "grad_norm": 2.4854148391473574, + "language_loss": 0.92327529, + "learning_rate": 3.3141308407736276e-06, + "loss": 0.95243704, + "num_input_tokens_seen": 2116125, + "step": 97, + "time_per_iteration": 2.9227845668792725 + }, + { + "auxiliary_loss_clip": 0.01796754, + "auxiliary_loss_mlp": 0.01120087, + "balance_loss_clip": 1.2274847, + "balance_loss_mlp": 1.05576193, + "epoch": 0.011783803282630914, + "flos": 19902125116800.0, + "grad_norm": 4.77580729770178, + "language_loss": 0.86508834, + "learning_rate": 3.321561120780869e-06, + "loss": 0.89425671, + "num_input_tokens_seen": 2134835, + "step": 98, + "time_per_iteration": 2.9349730014801025 + }, + { + "auxiliary_loss_clip": 0.01771852, + "auxiliary_loss_mlp": 0.01101202, + "balance_loss_clip": 1.22798181, + "balance_loss_mlp": 1.04391003, + "epoch": 0.011904046173270006, + "flos": 22340674517760.0, + "grad_norm": 2.9334982993365775, + "language_loss": 1.01704717, + "learning_rate": 3.3289159651708192e-06, + "loss": 1.04577768, + "num_input_tokens_seen": 2152410, + "step": 99, + "time_per_iteration": 2.9245223999023438 + }, + { + "auxiliary_loss_clip": 0.0176877, + "auxiliary_loss_mlp": 0.01127535, + "balance_loss_clip": 1.22425938, + "balance_loss_mlp": 1.06511652, + "epoch": 0.012024289063909096, + "flos": 19100060375040.0, + "grad_norm": 4.523290765070524, + "language_loss": 0.9770605, + "learning_rate": 3.3361968902759768e-06, + "loss": 1.00602353, + "num_input_tokens_seen": 2172090, + "step": 100, + "time_per_iteration": 3.086963653564453 + }, + { + "auxiliary_loss_clip": 0.01753364, + "auxiliary_loss_mlp": 0.01107052, + "balance_loss_clip": 1.22050524, + "balance_loss_mlp": 1.05238318, + "epoch": 0.012144531954548187, + "flos": 15012205159680.0, + "grad_norm": 2.966774445776538, + "language_loss": 0.94167423, + "learning_rate": 3.343405367163663e-06, + "loss": 0.97027838, + "num_input_tokens_seen": 2189020, + "step": 101, + "time_per_iteration": 2.8214471340179443 + }, + { + "auxiliary_loss_clip": 0.01773157, + "auxiliary_loss_mlp": 0.01105641, + "balance_loss_clip": 1.23076344, + "balance_loss_mlp": 1.04813457, + "epoch": 0.012264774845187279, + "flos": 15122020014720.0, + "grad_norm": 8.327537710960701, + "language_loss": 0.81526434, + "learning_rate": 3.350542823419951e-06, + "loss": 0.84405231, + "num_input_tokens_seen": 2205620, + "step": 102, + "time_per_iteration": 2.9061503410339355 + }, + { + "auxiliary_loss_clip": 0.01767721, + "auxiliary_loss_mlp": 0.01134199, + "balance_loss_clip": 1.22671592, + "balance_loss_mlp": 1.07421279, + "epoch": 0.012385017735826368, + "flos": 13949248959360.0, + "grad_norm": 3.713133068838989, + "language_loss": 0.87367296, + "learning_rate": 3.3576106448465615e-06, + "loss": 0.9026922, + "num_input_tokens_seen": 2219000, + "step": 103, + "time_per_iteration": 2.895533561706543 + }, + { + "auxiliary_loss_clip": 0.01744501, + "auxiliary_loss_mlp": 0.01105582, + "balance_loss_clip": 1.21836376, + "balance_loss_mlp": 1.05196166, + "epoch": 0.01250526062646546, + "flos": 23623260428160.0, + "grad_norm": 2.5076650286069277, + "language_loss": 0.88291979, + "learning_rate": 3.3646101770757797e-06, + "loss": 0.9114207, + "num_input_tokens_seen": 2237790, + "step": 104, + "time_per_iteration": 2.8173410892486572 + }, + { + "auxiliary_loss_clip": 0.01736289, + "auxiliary_loss_mlp": 0.01107087, + "balance_loss_clip": 1.21826708, + "balance_loss_mlp": 1.05384851, + "epoch": 0.012625503517104552, + "flos": 34640078958720.0, + "grad_norm": 1.7292029941321962, + "language_loss": 0.85586727, + "learning_rate": 3.371542727108104e-06, + "loss": 0.88430107, + "num_input_tokens_seen": 2259965, + "step": 105, + "time_per_iteration": 2.971675157546997 + }, + { + "auxiliary_loss_clip": 0.01744184, + "auxiliary_loss_mlp": 0.01120497, + "balance_loss_clip": 1.21718216, + "balance_loss_mlp": 1.066257, + "epoch": 0.012745746407743641, + "flos": 17821891837440.0, + "grad_norm": 2.847317598399419, + "language_loss": 0.89882004, + "learning_rate": 3.3784095647770114e-06, + "loss": 0.92746687, + "num_input_tokens_seen": 2278610, + "step": 106, + "time_per_iteration": 2.744131326675415 + }, + { + "auxiliary_loss_clip": 0.01725952, + "auxiliary_loss_mlp": 0.01110401, + "balance_loss_clip": 1.21048927, + "balance_loss_mlp": 1.05828226, + "epoch": 0.012865989298382733, + "flos": 20595057361920.0, + "grad_norm": 2.1630126858892784, + "language_loss": 0.8884232, + "learning_rate": 3.3852119241449547e-06, + "loss": 0.91678667, + "num_input_tokens_seen": 2297730, + "step": 107, + "time_per_iteration": 2.9160947799682617 + }, + { + "auxiliary_loss_clip": 0.01725144, + "auxiliary_loss_mlp": 0.01114877, + "balance_loss_clip": 1.21300721, + "balance_loss_mlp": 1.06278229, + "epoch": 0.012986232189021825, + "flos": 23948969978880.0, + "grad_norm": 2.750155618421439, + "language_loss": 0.96518624, + "learning_rate": 3.3919510048344295e-06, + "loss": 0.99358648, + "num_input_tokens_seen": 2315740, + "step": 108, + "time_per_iteration": 2.836552143096924 + }, + { + "auxiliary_loss_clip": 0.01698516, + "auxiliary_loss_mlp": 0.01102355, + "balance_loss_clip": 1.20795512, + "balance_loss_mlp": 1.05192995, + "epoch": 0.013106475079660914, + "flos": 23725425686400.0, + "grad_norm": 2.0148591503011097, + "language_loss": 0.86830062, + "learning_rate": 3.3986279732976907e-06, + "loss": 0.89630938, + "num_input_tokens_seen": 2334215, + "step": 109, + "time_per_iteration": 2.9095427989959717 + }, + { + "auxiliary_loss_clip": 0.01697269, + "auxiliary_loss_mlp": 0.01117036, + "balance_loss_clip": 1.20575905, + "balance_loss_mlp": 1.06870866, + "epoch": 0.013226717970300006, + "flos": 21102438925440.0, + "grad_norm": 2.1064167816851747, + "language_loss": 0.95488256, + "learning_rate": 3.4052439640284983e-06, + "loss": 0.98302561, + "num_input_tokens_seen": 2353130, + "step": 110, + "time_per_iteration": 2.8440945148468018 + }, + { + "auxiliary_loss_clip": 0.01706017, + "auxiliary_loss_mlp": 0.01107014, + "balance_loss_clip": 1.21043634, + "balance_loss_mlp": 1.05568254, + "epoch": 0.013346960860939098, + "flos": 24863902231680.0, + "grad_norm": 2.7771946554819555, + "language_loss": 0.81191248, + "learning_rate": 3.4118000807190217e-06, + "loss": 0.84004277, + "num_input_tokens_seen": 2374010, + "step": 111, + "time_per_iteration": 3.0114586353302 + }, + { + "auxiliary_loss_clip": 0.01701037, + "auxiliary_loss_mlp": 0.01101982, + "balance_loss_clip": 1.20599902, + "balance_loss_mlp": 1.05065072, + "epoch": 0.013467203751578187, + "flos": 28181940140160.0, + "grad_norm": 2.157468390387504, + "language_loss": 0.76089084, + "learning_rate": 3.4182973973648723e-06, + "loss": 0.78892112, + "num_input_tokens_seen": 2395220, + "step": 112, + "time_per_iteration": 2.8955211639404297 + }, + { + "auxiliary_loss_clip": 0.01685203, + "auxiliary_loss_mlp": 0.01094167, + "balance_loss_clip": 1.20265818, + "balance_loss_mlp": 1.04822385, + "epoch": 0.013587446642217279, + "flos": 18916233546240.0, + "grad_norm": 3.190213960814635, + "language_loss": 0.95087552, + "learning_rate": 3.424736959321014e-06, + "loss": 0.97866923, + "num_input_tokens_seen": 2413025, + "step": 113, + "time_per_iteration": 2.8034987449645996 + }, + { + "auxiliary_loss_clip": 0.01683879, + "auxiliary_loss_mlp": 0.01117024, + "balance_loss_clip": 1.20208323, + "balance_loss_mlp": 1.06676507, + "epoch": 0.01370768953285637, + "flos": 23988615615360.0, + "grad_norm": 2.0281561986934, + "language_loss": 0.88702184, + "learning_rate": 3.431119784311155e-06, + "loss": 0.9150309, + "num_input_tokens_seen": 2432700, + "step": 114, + "time_per_iteration": 2.8672354221343994 + }, + { + "auxiliary_loss_clip": 0.01675255, + "auxiliary_loss_mlp": 0.01092916, + "balance_loss_clip": 1.20215988, + "balance_loss_mlp": 1.05047727, + "epoch": 0.01382793242349546, + "flos": 39202565512320.0, + "grad_norm": 1.8918843302235515, + "language_loss": 0.77644074, + "learning_rate": 3.43744686339307e-06, + "loss": 0.80412251, + "num_input_tokens_seen": 2455020, + "step": 115, + "time_per_iteration": 3.0261874198913574 + }, + { + "auxiliary_loss_clip": 0.01670201, + "auxiliary_loss_mlp": 0.01090531, + "balance_loss_clip": 1.19472861, + "balance_loss_mlp": 1.04783022, + "epoch": 0.013948175314134552, + "flos": 41353506714240.0, + "grad_norm": 2.1498606397136433, + "language_loss": 0.90924799, + "learning_rate": 3.44371916188212e-06, + "loss": 0.93685532, + "num_input_tokens_seen": 2475775, + "step": 116, + "time_per_iteration": 2.972459316253662 + }, + { + "auxiliary_loss_clip": 0.0165046, + "auxiliary_loss_mlp": 0.01102428, + "balance_loss_clip": 1.18840849, + "balance_loss_mlp": 1.05867815, + "epoch": 0.014068418204773643, + "flos": 22453542028800.0, + "grad_norm": 2.3321626841283227, + "language_loss": 0.86364788, + "learning_rate": 3.449937620235143e-06, + "loss": 0.89117676, + "num_input_tokens_seen": 2496370, + "step": 117, + "time_per_iteration": 2.734614849090576 + }, + { + "auxiliary_loss_clip": 0.01631738, + "auxiliary_loss_mlp": 0.01096286, + "balance_loss_clip": 1.18279803, + "balance_loss_mlp": 1.05382371, + "epoch": 0.014188661095412733, + "flos": 23805147922560.0, + "grad_norm": 2.8445538416828877, + "language_loss": 0.89372754, + "learning_rate": 3.456103154896722e-06, + "loss": 0.92100775, + "num_input_tokens_seen": 2517645, + "step": 118, + "time_per_iteration": 3.6214444637298584 + }, + { + "auxiliary_loss_clip": 0.0162602, + "auxiliary_loss_mlp": 0.01085218, + "balance_loss_clip": 1.17934871, + "balance_loss_mlp": 1.04609323, + "epoch": 0.014308903986051825, + "flos": 23660248458240.0, + "grad_norm": 2.8065118377454215, + "language_loss": 0.92490685, + "learning_rate": 3.462216659109757e-06, + "loss": 0.95201921, + "num_input_tokens_seen": 2537825, + "step": 119, + "time_per_iteration": 4.38528299331665 + }, + { + "auxiliary_loss_clip": 0.01648112, + "auxiliary_loss_mlp": 0.01109184, + "balance_loss_clip": 1.18763208, + "balance_loss_mlp": 1.06672144, + "epoch": 0.014429146876690916, + "flos": 20667991927680.0, + "grad_norm": 2.2247996552603295, + "language_loss": 0.85487992, + "learning_rate": 3.4682790036921077e-06, + "loss": 0.88245291, + "num_input_tokens_seen": 2556485, + "step": 120, + "time_per_iteration": 3.4908976554870605 + }, + { + "auxiliary_loss_clip": 0.01620957, + "auxiliary_loss_mlp": 0.01089048, + "balance_loss_clip": 1.18085909, + "balance_loss_mlp": 1.0522604, + "epoch": 0.014549389767330006, + "flos": 20229199384320.0, + "grad_norm": 1.9318521045703645, + "language_loss": 0.83289957, + "learning_rate": 3.4742910377810193e-06, + "loss": 0.8599996, + "num_input_tokens_seen": 2573945, + "step": 121, + "time_per_iteration": 2.6236488819122314 + }, + { + "auxiliary_loss_clip": 0.01619245, + "auxiliary_loss_mlp": 0.01095237, + "balance_loss_clip": 1.18184626, + "balance_loss_mlp": 1.05797267, + "epoch": 0.014669632657969098, + "flos": 18004174381440.0, + "grad_norm": 2.2003289628894107, + "language_loss": 0.88876235, + "learning_rate": 3.4802535895469042e-06, + "loss": 0.9159072, + "num_input_tokens_seen": 2592695, + "step": 122, + "time_per_iteration": 2.8908443450927734 + }, + { + "auxiliary_loss_clip": 0.01619814, + "auxiliary_loss_mlp": 0.01089176, + "balance_loss_clip": 1.18047225, + "balance_loss_mlp": 1.04983735, + "epoch": 0.01478987554860819, + "flos": 22741796672640.0, + "grad_norm": 2.219056659886043, + "language_loss": 0.89719105, + "learning_rate": 3.4861674668779934e-06, + "loss": 0.924281, + "num_input_tokens_seen": 2610925, + "step": 123, + "time_per_iteration": 2.9571430683135986 + }, + { + "auxiliary_loss_clip": 0.01614107, + "auxiliary_loss_mlp": 0.01079705, + "balance_loss_clip": 1.17966223, + "balance_loss_mlp": 1.04289365, + "epoch": 0.01491011843924728, + "flos": 17198590106880.0, + "grad_norm": 1.9689790624855354, + "language_loss": 0.84253132, + "learning_rate": 3.492033458037272e-06, + "loss": 0.8694694, + "num_input_tokens_seen": 2629495, + "step": 124, + "time_per_iteration": 2.8682000637054443 + }, + { + "auxiliary_loss_clip": 0.01607913, + "auxiliary_loss_mlp": 0.01082563, + "balance_loss_clip": 1.17588103, + "balance_loss_mlp": 1.04525042, + "epoch": 0.01503036132988637, + "flos": 17673867889920.0, + "grad_norm": 2.709267598283612, + "language_loss": 0.86959577, + "learning_rate": 3.497852332293018e-06, + "loss": 0.89650047, + "num_input_tokens_seen": 2645070, + "step": 125, + "time_per_iteration": 3.1706278324127197 + }, + { + "auxiliary_loss_clip": 0.01606685, + "auxiliary_loss_mlp": 0.01093558, + "balance_loss_clip": 1.17655134, + "balance_loss_mlp": 1.05586433, + "epoch": 0.015150604220525462, + "flos": 18878239935360.0, + "grad_norm": 2.0249592637053535, + "language_loss": 0.96622074, + "learning_rate": 3.5036248405242356e-06, + "loss": 0.99322313, + "num_input_tokens_seen": 2663825, + "step": 126, + "time_per_iteration": 2.961616277694702 + }, + { + "auxiliary_loss_clip": 0.01587828, + "auxiliary_loss_mlp": 0.01082715, + "balance_loss_clip": 1.16523218, + "balance_loss_mlp": 1.04521179, + "epoch": 0.015270847111164552, + "flos": 39420184060800.0, + "grad_norm": 2.154276332999132, + "language_loss": 0.82782698, + "learning_rate": 3.509351715802146e-06, + "loss": 0.85453236, + "num_input_tokens_seen": 2684710, + "step": 127, + "time_per_iteration": 2.975443124771118 + }, + { + "auxiliary_loss_clip": 0.01595379, + "auxiliary_loss_mlp": 0.0108219, + "balance_loss_clip": 1.16924405, + "balance_loss_mlp": 1.04533005, + "epoch": 0.015391090001803644, + "flos": 43762466286720.0, + "grad_norm": 1.9937851508716855, + "language_loss": 0.78679806, + "learning_rate": 3.5150336739488763e-06, + "loss": 0.81357372, + "num_input_tokens_seen": 2706995, + "step": 128, + "time_per_iteration": 2.942589282989502 + }, + { + "auxiliary_loss_clip": 0.01578685, + "auxiliary_loss_mlp": 0.01082054, + "balance_loss_clip": 1.16533589, + "balance_loss_mlp": 1.04877067, + "epoch": 0.015511332892442733, + "flos": 18916341287040.0, + "grad_norm": 2.510206725430682, + "language_loss": 0.83916473, + "learning_rate": 3.5206714140744143e-06, + "loss": 0.86577213, + "num_input_tokens_seen": 2727050, + "step": 129, + "time_per_iteration": 2.7786996364593506 + }, + { + "auxiliary_loss_clip": 0.01588199, + "auxiliary_loss_mlp": 0.01096032, + "balance_loss_clip": 1.17536926, + "balance_loss_mlp": 1.06261802, + "epoch": 0.015631575783081827, + "flos": 24535283679360.0, + "grad_norm": 3.1384735745914374, + "language_loss": 0.87431794, + "learning_rate": 3.5262656190928208e-06, + "loss": 0.90116024, + "num_input_tokens_seen": 2745350, + "step": 130, + "time_per_iteration": 2.821514368057251 + }, + { + "auxiliary_loss_clip": 0.01533008, + "auxiliary_loss_mlp": 0.01105246, + "balance_loss_clip": 1.25950885, + "balance_loss_mlp": 1.08159482, + "epoch": 0.015751818673720917, + "flos": 62328536098560.0, + "grad_norm": 1.0851531219570272, + "language_loss": 0.71472144, + "learning_rate": 3.5318169562186737e-06, + "loss": 0.74110395, + "num_input_tokens_seen": 2814195, + "step": 131, + "time_per_iteration": 3.3543856143951416 + }, + { + "auxiliary_loss_clip": 0.01572846, + "auxiliary_loss_mlp": 0.01079667, + "balance_loss_clip": 1.16747212, + "balance_loss_mlp": 1.04464328, + "epoch": 0.015872061564360006, + "flos": 23878549365120.0, + "grad_norm": 2.135411786007225, + "language_loss": 0.82086813, + "learning_rate": 3.5373260774446292e-06, + "loss": 0.84739327, + "num_input_tokens_seen": 2834645, + "step": 132, + "time_per_iteration": 2.9627573490142822 + }, + { + "auxiliary_loss_clip": 0.01562694, + "auxiliary_loss_mlp": 0.01077121, + "balance_loss_clip": 1.1600585, + "balance_loss_mlp": 1.04531622, + "epoch": 0.0159923044549991, + "flos": 23367899664000.0, + "grad_norm": 2.9564805091875406, + "language_loss": 0.90336812, + "learning_rate": 3.542793620000961e-06, + "loss": 0.92976624, + "num_input_tokens_seen": 2854120, + "step": 133, + "time_per_iteration": 2.956651449203491 + }, + { + "auxiliary_loss_clip": 0.01564991, + "auxiliary_loss_mlp": 0.01092815, + "balance_loss_clip": 1.16585314, + "balance_loss_mlp": 1.06005657, + "epoch": 0.01611254734563819, + "flos": 17858305249920.0, + "grad_norm": 2.392201358136586, + "language_loss": 0.87067306, + "learning_rate": 3.5482202067978894e-06, + "loss": 0.89725113, + "num_input_tokens_seen": 2871330, + "step": 134, + "time_per_iteration": 2.8045053482055664 + }, + { + "auxiliary_loss_clip": 0.01560146, + "auxiliary_loss_mlp": 0.01085923, + "balance_loss_clip": 1.16101742, + "balance_loss_mlp": 1.05163884, + "epoch": 0.01623279023627728, + "flos": 20954774113920.0, + "grad_norm": 2.0587044149136218, + "language_loss": 0.76160657, + "learning_rate": 3.553606446851471e-06, + "loss": 0.78806728, + "num_input_tokens_seen": 2888070, + "step": 135, + "time_per_iteration": 2.8532583713531494 + }, + { + "auxiliary_loss_clip": 0.01551569, + "auxiliary_loss_mlp": 0.01083786, + "balance_loss_clip": 1.15797663, + "balance_loss_mlp": 1.0519805, + "epoch": 0.016353033126916373, + "flos": 15742412743680.0, + "grad_norm": 2.545937568758635, + "language_loss": 0.83439672, + "learning_rate": 3.5589529356937613e-06, + "loss": 0.86075032, + "num_input_tokens_seen": 2906465, + "step": 136, + "time_per_iteration": 2.786659002304077 + }, + { + "auxiliary_loss_clip": 0.01546201, + "auxiliary_loss_mlp": 0.0108236, + "balance_loss_clip": 1.15524244, + "balance_loss_mlp": 1.04828978, + "epoch": 0.016473276017555463, + "flos": 18807280617600.0, + "grad_norm": 3.3563629930188967, + "language_loss": 0.76955062, + "learning_rate": 3.5642602557679627e-06, + "loss": 0.79583621, + "num_input_tokens_seen": 2924915, + "step": 137, + "time_per_iteration": 2.8525450229644775 + }, + { + "auxiliary_loss_clip": 0.01541736, + "auxiliary_loss_mlp": 0.01076428, + "balance_loss_clip": 1.16239309, + "balance_loss_mlp": 1.04661322, + "epoch": 0.016593518908194552, + "flos": 24352641999360.0, + "grad_norm": 2.238760919368466, + "language_loss": 0.84395123, + "learning_rate": 3.569528976809202e-06, + "loss": 0.87013292, + "num_input_tokens_seen": 2942130, + "step": 138, + "time_per_iteration": 2.834343910217285 + }, + { + "auxiliary_loss_clip": 0.01545138, + "auxiliary_loss_mlp": 0.0108683, + "balance_loss_clip": 1.15521908, + "balance_loss_mlp": 1.0541904, + "epoch": 0.016713761798833646, + "flos": 22346133384960.0, + "grad_norm": 1.8345121550229704, + "language_loss": 0.89927751, + "learning_rate": 3.5747596562115522e-06, + "loss": 0.92559719, + "num_input_tokens_seen": 2962745, + "step": 139, + "time_per_iteration": 2.865997791290283 + }, + { + "auxiliary_loss_clip": 0.01553756, + "auxiliary_loss_mlp": 0.01081285, + "balance_loss_clip": 1.15975261, + "balance_loss_mlp": 1.04843044, + "epoch": 0.016834004689472735, + "flos": 17821820010240.0, + "grad_norm": 4.436930754828113, + "language_loss": 0.90716445, + "learning_rate": 3.5799528393819138e-06, + "loss": 0.93351483, + "num_input_tokens_seen": 2981825, + "step": 140, + "time_per_iteration": 2.8358347415924072 + }, + { + "auxiliary_loss_clip": 0.01525362, + "auxiliary_loss_mlp": 0.01079123, + "balance_loss_clip": 1.14582968, + "balance_loss_mlp": 1.04755652, + "epoch": 0.016954247580111825, + "flos": 20519501103360.0, + "grad_norm": 1.891440527476482, + "language_loss": 0.87925696, + "learning_rate": 3.585109060081286e-06, + "loss": 0.90530181, + "num_input_tokens_seen": 3001625, + "step": 141, + "time_per_iteration": 2.8310019969940186 + }, + { + "auxiliary_loss_clip": 0.01536883, + "auxiliary_loss_mlp": 0.01075352, + "balance_loss_clip": 1.15146601, + "balance_loss_mlp": 1.0440948, + "epoch": 0.017074490470750915, + "flos": 22088869200000.0, + "grad_norm": 1.7090677887130854, + "language_loss": 0.78599691, + "learning_rate": 3.590228840753992e-06, + "loss": 0.81211925, + "num_input_tokens_seen": 3022055, + "step": 142, + "time_per_iteration": 2.887864351272583 + }, + { + "auxiliary_loss_clip": 0.01529863, + "auxiliary_loss_mlp": 0.01076894, + "balance_loss_clip": 1.152282, + "balance_loss_mlp": 1.04671001, + "epoch": 0.01719473336139001, + "flos": 15997270717440.0, + "grad_norm": 10.119339631153917, + "language_loss": 0.87442905, + "learning_rate": 3.5953126928453423e-06, + "loss": 0.9004966, + "num_input_tokens_seen": 3039605, + "step": 143, + "time_per_iteration": 2.7431161403656006 + }, + { + "auxiliary_loss_clip": 0.01516771, + "auxiliary_loss_mlp": 0.01083038, + "balance_loss_clip": 1.143538, + "balance_loss_mlp": 1.05137587, + "epoch": 0.017314976252029098, + "flos": 22492038430080.0, + "grad_norm": 2.810337996587641, + "language_loss": 0.80578268, + "learning_rate": 3.600361117108239e-06, + "loss": 0.83178073, + "num_input_tokens_seen": 3059405, + "step": 144, + "time_per_iteration": 2.866790771484375 + }, + { + "auxiliary_loss_clip": 0.01510352, + "auxiliary_loss_mlp": 0.01066983, + "balance_loss_clip": 1.14029503, + "balance_loss_mlp": 1.03826571, + "epoch": 0.017435219142668188, + "flos": 22018053536640.0, + "grad_norm": 5.968963857509597, + "language_loss": 0.97408831, + "learning_rate": 3.6053746038991616e-06, + "loss": 0.9998616, + "num_input_tokens_seen": 3078490, + "step": 145, + "time_per_iteration": 3.8916687965393066 + }, + { + "auxiliary_loss_clip": 0.01440211, + "auxiliary_loss_mlp": 0.01044385, + "balance_loss_clip": 1.19816113, + "balance_loss_mlp": 1.0244534, + "epoch": 0.01755546203330728, + "flos": 72240526149120.0, + "grad_norm": 1.0644661048560689, + "language_loss": 0.58436435, + "learning_rate": 3.6103536334639843e-06, + "loss": 0.60921025, + "num_input_tokens_seen": 3131755, + "step": 146, + "time_per_iteration": 4.1846349239349365 + }, + { + "auxiliary_loss_clip": 0.01505719, + "auxiliary_loss_mlp": 0.01079707, + "balance_loss_clip": 1.14128637, + "balance_loss_mlp": 1.04988039, + "epoch": 0.01767570492394637, + "flos": 25337061112320.0, + "grad_norm": 2.3587881886339, + "language_loss": 0.85594177, + "learning_rate": 3.615298676214041e-06, + "loss": 0.881796, + "num_input_tokens_seen": 3152035, + "step": 147, + "time_per_iteration": 3.7961108684539795 + }, + { + "auxiliary_loss_clip": 0.01506015, + "auxiliary_loss_mlp": 0.01082292, + "balance_loss_clip": 1.14193356, + "balance_loss_mlp": 1.05346739, + "epoch": 0.01779594781458546, + "flos": 20449188230400.0, + "grad_norm": 9.563308534732206, + "language_loss": 0.89159584, + "learning_rate": 3.6202101929928317e-06, + "loss": 0.91747892, + "num_input_tokens_seen": 3170625, + "step": 148, + "time_per_iteration": 2.8399922847747803 + }, + { + "auxiliary_loss_clip": 0.01501753, + "auxiliary_loss_mlp": 0.01077598, + "balance_loss_clip": 1.14067602, + "balance_loss_mlp": 1.04935741, + "epoch": 0.017916190705224554, + "flos": 16253601148800.0, + "grad_norm": 2.0245537848067903, + "language_loss": 0.8828181, + "learning_rate": 3.6250886353337413e-06, + "loss": 0.90861166, + "num_input_tokens_seen": 3188155, + "step": 149, + "time_per_iteration": 2.768321990966797 + }, + { + "auxiliary_loss_clip": 0.01509424, + "auxiliary_loss_mlp": 0.010787, + "balance_loss_clip": 1.14230216, + "balance_loss_mlp": 1.04942238, + "epoch": 0.018036433595863644, + "flos": 23330588411520.0, + "grad_norm": 1.8488202216691074, + "language_loss": 0.8651706, + "learning_rate": 3.6299344457091488e-06, + "loss": 0.89105189, + "num_input_tokens_seen": 3209015, + "step": 150, + "time_per_iteration": 2.9016122817993164 + }, + { + "auxiliary_loss_clip": 0.01497494, + "auxiliary_loss_mlp": 0.01070302, + "balance_loss_clip": 1.13829648, + "balance_loss_mlp": 1.04216838, + "epoch": 0.018156676486502734, + "flos": 18588010043520.0, + "grad_norm": 2.2944817092522602, + "language_loss": 0.9397285, + "learning_rate": 3.634748057771256e-06, + "loss": 0.96540654, + "num_input_tokens_seen": 3224955, + "step": 151, + "time_per_iteration": 2.937983751296997 + }, + { + "auxiliary_loss_clip": 0.01495689, + "auxiliary_loss_mlp": 0.01082522, + "balance_loss_clip": 1.13911057, + "balance_loss_mlp": 1.05496073, + "epoch": 0.018276919377141827, + "flos": 25448707560960.0, + "grad_norm": 1.8523801635592174, + "language_loss": 0.85851711, + "learning_rate": 3.639529896584965e-06, + "loss": 0.88429922, + "num_input_tokens_seen": 3246330, + "step": 152, + "time_per_iteration": 2.9097015857696533 + }, + { + "auxiliary_loss_clip": 0.0149799, + "auxiliary_loss_mlp": 0.01072724, + "balance_loss_clip": 1.14018059, + "balance_loss_mlp": 1.04327893, + "epoch": 0.018397162267780917, + "flos": 20047311889920.0, + "grad_norm": 2.6921135680098565, + "language_loss": 0.8915081, + "learning_rate": 3.6442803788531233e-06, + "loss": 0.91721523, + "num_input_tokens_seen": 3264290, + "step": 153, + "time_per_iteration": 2.7813470363616943 + }, + { + "auxiliary_loss_clip": 0.01500703, + "auxiliary_loss_mlp": 0.01083233, + "balance_loss_clip": 1.1355828, + "balance_loss_mlp": 1.0537405, + "epoch": 0.018517405158420007, + "flos": 27565282425600.0, + "grad_norm": 2.5589742060666265, + "language_loss": 0.96080375, + "learning_rate": 3.6489999131344357e-06, + "loss": 0.98664308, + "num_input_tokens_seen": 3287065, + "step": 154, + "time_per_iteration": 2.8730313777923584 + }, + { + "auxiliary_loss_clip": 0.01476658, + "auxiliary_loss_mlp": 0.01071814, + "balance_loss_clip": 1.13200712, + "balance_loss_mlp": 1.04632711, + "epoch": 0.0186376480490591, + "flos": 19354056422400.0, + "grad_norm": 2.295835809850064, + "language_loss": 0.90600318, + "learning_rate": 3.653688900054313e-06, + "loss": 0.93148792, + "num_input_tokens_seen": 3305595, + "step": 155, + "time_per_iteration": 2.7736480236053467 + }, + { + "auxiliary_loss_clip": 0.0148726, + "auxiliary_loss_mlp": 0.01073286, + "balance_loss_clip": 1.13012815, + "balance_loss_mlp": 1.04443717, + "epoch": 0.01875789093969819, + "flos": 26687840993280.0, + "grad_norm": 2.2839238182703636, + "language_loss": 0.76135886, + "learning_rate": 3.6583477325089526e-06, + "loss": 0.7869643, + "num_input_tokens_seen": 3326135, + "step": 156, + "time_per_iteration": 2.803480386734009 + }, + { + "auxiliary_loss_clip": 0.01479848, + "auxiliary_loss_mlp": 0.01075177, + "balance_loss_clip": 1.13285005, + "balance_loss_mlp": 1.04824781, + "epoch": 0.01887813383033728, + "flos": 24353001135360.0, + "grad_norm": 3.3290723322358917, + "language_loss": 1.04281259, + "learning_rate": 3.6629767958628916e-06, + "loss": 1.06836295, + "num_input_tokens_seen": 3343510, + "step": 157, + "time_per_iteration": 2.8329901695251465 + }, + { + "auxiliary_loss_clip": 0.01477487, + "auxiliary_loss_mlp": 0.01083112, + "balance_loss_clip": 1.12957788, + "balance_loss_mlp": 1.05447829, + "epoch": 0.018998376720976373, + "flos": 14647532330880.0, + "grad_norm": 2.244985276432113, + "language_loss": 0.85813504, + "learning_rate": 3.667576468140291e-06, + "loss": 0.88374108, + "num_input_tokens_seen": 3361325, + "step": 158, + "time_per_iteration": 2.7759499549865723 + }, + { + "auxiliary_loss_clip": 0.0146423, + "auxiliary_loss_mlp": 0.01072832, + "balance_loss_clip": 1.12372541, + "balance_loss_mlp": 1.04541409, + "epoch": 0.019118619611615463, + "flos": 29305261146240.0, + "grad_norm": 2.224228547293952, + "language_loss": 0.8928073, + "learning_rate": 3.672147120210184e-06, + "loss": 0.91817796, + "num_input_tokens_seen": 3377925, + "step": 159, + "time_per_iteration": 2.8901162147521973 + }, + { + "auxiliary_loss_clip": 0.01475224, + "auxiliary_loss_mlp": 0.01076481, + "balance_loss_clip": 1.13593853, + "balance_loss_mlp": 1.04982555, + "epoch": 0.019238862502254553, + "flos": 20886723797760.0, + "grad_norm": 2.6857182733204095, + "language_loss": 0.86284649, + "learning_rate": 3.6766891159659177e-06, + "loss": 0.88836348, + "num_input_tokens_seen": 3396335, + "step": 160, + "time_per_iteration": 2.819340467453003 + }, + { + "auxiliary_loss_clip": 0.01466904, + "auxiliary_loss_mlp": 0.01080116, + "balance_loss_clip": 1.1292218, + "balance_loss_mlp": 1.05336595, + "epoch": 0.019359105392893646, + "flos": 21360672777600.0, + "grad_norm": 2.610324925709763, + "language_loss": 0.87936765, + "learning_rate": 3.6812028124990075e-06, + "loss": 0.90483785, + "num_input_tokens_seen": 3413605, + "step": 161, + "time_per_iteration": 2.857987642288208 + }, + { + "auxiliary_loss_clip": 0.01465676, + "auxiliary_loss_mlp": 0.01069535, + "balance_loss_clip": 1.1268245, + "balance_loss_mlp": 1.0436312, + "epoch": 0.019479348283532736, + "flos": 16283729681280.0, + "grad_norm": 3.2488013416296, + "language_loss": 0.81591547, + "learning_rate": 3.6856885602676016e-06, + "loss": 0.84126759, + "num_input_tokens_seen": 3429640, + "step": 162, + "time_per_iteration": 2.8105757236480713 + }, + { + "auxiliary_loss_clip": 0.01464672, + "auxiliary_loss_mlp": 0.01076645, + "balance_loss_clip": 1.12694061, + "balance_loss_mlp": 1.04919076, + "epoch": 0.019599591174171826, + "flos": 22091239497600.0, + "grad_norm": 3.8165813726602105, + "language_loss": 0.94068241, + "learning_rate": 3.6901467032597733e-06, + "loss": 0.96609563, + "num_input_tokens_seen": 3448125, + "step": 163, + "time_per_iteration": 2.7521276473999023 + }, + { + "auxiliary_loss_clip": 0.01461975, + "auxiliary_loss_mlp": 0.01071747, + "balance_loss_clip": 1.12466931, + "balance_loss_mlp": 1.04459095, + "epoch": 0.01971983406481092, + "flos": 19609668581760.0, + "grad_norm": 2.3214544030235795, + "language_loss": 0.87523496, + "learning_rate": 3.694577579151804e-06, + "loss": 0.90057212, + "num_input_tokens_seen": 3466535, + "step": 164, + "time_per_iteration": 2.8044791221618652 + }, + { + "auxiliary_loss_clip": 0.01460371, + "auxiliary_loss_mlp": 0.01084312, + "balance_loss_clip": 1.12772441, + "balance_loss_mlp": 1.05763328, + "epoch": 0.01984007695545001, + "flos": 19099342103040.0, + "grad_norm": 2.4879388092944166, + "language_loss": 0.73774558, + "learning_rate": 3.6989815194616703e-06, + "loss": 0.76319242, + "num_input_tokens_seen": 3483730, + "step": 165, + "time_per_iteration": 2.7785608768463135 + }, + { + "auxiliary_loss_clip": 0.01464487, + "auxiliary_loss_mlp": 0.01088296, + "balance_loss_clip": 1.12410414, + "balance_loss_mlp": 1.06112814, + "epoch": 0.0199603198460891, + "flos": 20848406964480.0, + "grad_norm": 2.4907599725626834, + "language_loss": 0.80048156, + "learning_rate": 3.703358849697888e-06, + "loss": 0.82600939, + "num_input_tokens_seen": 3503640, + "step": 166, + "time_per_iteration": 2.8378443717956543 + }, + { + "auxiliary_loss_clip": 0.01465005, + "auxiliary_loss_mlp": 0.01074891, + "balance_loss_clip": 1.13008165, + "balance_loss_mlp": 1.04949975, + "epoch": 0.020080562736728192, + "flos": 21870747861120.0, + "grad_norm": 1.923651069482366, + "language_loss": 0.82742894, + "learning_rate": 3.7077098895038803e-06, + "loss": 0.85282791, + "num_input_tokens_seen": 3523010, + "step": 167, + "time_per_iteration": 2.8078861236572266 + }, + { + "auxiliary_loss_clip": 0.01456452, + "auxiliary_loss_mlp": 0.01063829, + "balance_loss_clip": 1.12347579, + "balance_loss_mlp": 1.03772271, + "epoch": 0.020200805627367282, + "flos": 21688788539520.0, + "grad_norm": 4.233522238481811, + "language_loss": 0.96930397, + "learning_rate": 3.712034952798045e-06, + "loss": 0.99450678, + "num_input_tokens_seen": 3541125, + "step": 168, + "time_per_iteration": 2.8224587440490723 + }, + { + "auxiliary_loss_clip": 0.01451672, + "auxiliary_loss_mlp": 0.01063925, + "balance_loss_clip": 1.11919081, + "balance_loss_mlp": 1.03656626, + "epoch": 0.02032104851800637, + "flos": 33543043729920.0, + "grad_norm": 2.5218863826864943, + "language_loss": 0.8450613, + "learning_rate": 3.7163343479096656e-06, + "loss": 0.87021732, + "num_input_tokens_seen": 3562700, + "step": 169, + "time_per_iteration": 2.9674715995788574 + }, + { + "auxiliary_loss_clip": 0.01449771, + "auxiliary_loss_mlp": 0.01074738, + "balance_loss_clip": 1.12306547, + "balance_loss_mlp": 1.05047917, + "epoch": 0.020441291408645465, + "flos": 31686965274240.0, + "grad_norm": 2.4643108426282994, + "language_loss": 0.82998967, + "learning_rate": 3.720608377710802e-06, + "loss": 0.85523486, + "num_input_tokens_seen": 3582790, + "step": 170, + "time_per_iteration": 2.982860803604126 + }, + { + "auxiliary_loss_clip": 0.01444218, + "auxiliary_loss_mlp": 0.01056939, + "balance_loss_clip": 1.11787224, + "balance_loss_mlp": 1.03006864, + "epoch": 0.020561534299284555, + "flos": 20886687884160.0, + "grad_norm": 2.11889465879208, + "language_loss": 0.8654629, + "learning_rate": 3.7248573397443277e-06, + "loss": 0.89047456, + "num_input_tokens_seen": 3601715, + "step": 171, + "time_per_iteration": 2.8592700958251953 + }, + { + "auxiliary_loss_clip": 0.01448551, + "auxiliary_loss_mlp": 0.01078996, + "balance_loss_clip": 1.1247313, + "balance_loss_mlp": 1.05360484, + "epoch": 0.020681777189923645, + "flos": 20996610480000.0, + "grad_norm": 1.9227935477421132, + "language_loss": 0.97552627, + "learning_rate": 3.729081526348224e-06, + "loss": 1.0008018, + "num_input_tokens_seen": 3620245, + "step": 172, + "time_per_iteration": 3.769394874572754 + }, + { + "auxiliary_loss_clip": 0.01450173, + "auxiliary_loss_mlp": 0.01073165, + "balance_loss_clip": 1.12180424, + "balance_loss_mlp": 1.04674828, + "epoch": 0.020802020080562738, + "flos": 28257532312320.0, + "grad_norm": 1.7840643838812456, + "language_loss": 0.85011172, + "learning_rate": 3.7332812247762777e-06, + "loss": 0.87534511, + "num_input_tokens_seen": 3641545, + "step": 173, + "time_per_iteration": 3.9681553840637207 + }, + { + "auxiliary_loss_clip": 0.01441701, + "auxiliary_loss_mlp": 0.01065327, + "balance_loss_clip": 1.12153959, + "balance_loss_mlp": 1.03951859, + "epoch": 0.020922262971201828, + "flos": 19681274344320.0, + "grad_norm": 2.3478241396996355, + "language_loss": 0.95355904, + "learning_rate": 3.737456717315293e-06, + "loss": 0.97862929, + "num_input_tokens_seen": 3660510, + "step": 174, + "time_per_iteration": 2.821349620819092 + }, + { + "auxiliary_loss_clip": 0.01440628, + "auxiliary_loss_mlp": 0.01073059, + "balance_loss_clip": 1.12255979, + "balance_loss_mlp": 1.04868126, + "epoch": 0.021042505861840918, + "flos": 15666353694720.0, + "grad_norm": 1.816902416598986, + "language_loss": 0.90777099, + "learning_rate": 3.7416082813989552e-06, + "loss": 0.93290794, + "num_input_tokens_seen": 3677505, + "step": 175, + "time_per_iteration": 2.7769110202789307 + }, + { + "auxiliary_loss_clip": 0.01444829, + "auxiliary_loss_mlp": 0.0107903, + "balance_loss_clip": 1.11920178, + "balance_loss_mlp": 1.05279219, + "epoch": 0.02116274875248001, + "flos": 21142012734720.0, + "grad_norm": 2.1669573542840297, + "language_loss": 0.8931728, + "learning_rate": 3.745736189718439e-06, + "loss": 0.91841143, + "num_input_tokens_seen": 3696760, + "step": 176, + "time_per_iteration": 2.7534215450286865 + }, + { + "auxiliary_loss_clip": 0.01430209, + "auxiliary_loss_mlp": 0.01069424, + "balance_loss_clip": 1.1164068, + "balance_loss_mlp": 1.04599941, + "epoch": 0.0212829916431191, + "flos": 24715770543360.0, + "grad_norm": 2.5593524883772814, + "language_loss": 0.72433007, + "learning_rate": 3.749840710329894e-06, + "loss": 0.74932647, + "num_input_tokens_seen": 3717465, + "step": 177, + "time_per_iteration": 2.8180084228515625 + }, + { + "auxiliary_loss_clip": 0.01446296, + "auxiliary_loss_mlp": 0.01072251, + "balance_loss_clip": 1.12075973, + "balance_loss_mlp": 1.04641867, + "epoch": 0.02140323453375819, + "flos": 16645493508480.0, + "grad_norm": 2.775931115228702, + "language_loss": 0.98373759, + "learning_rate": 3.7539221067588938e-06, + "loss": 1.00892305, + "num_input_tokens_seen": 3731440, + "step": 178, + "time_per_iteration": 2.715641498565674 + }, + { + "auxiliary_loss_clip": 0.01444643, + "auxiliary_loss_mlp": 0.01074733, + "balance_loss_clip": 1.12173939, + "balance_loss_mlp": 1.04856682, + "epoch": 0.021523477424397284, + "flos": 20299332689280.0, + "grad_norm": 4.447951599993028, + "language_loss": 0.93837237, + "learning_rate": 3.757980638101964e-06, + "loss": 0.96356612, + "num_input_tokens_seen": 3744935, + "step": 179, + "time_per_iteration": 2.7750511169433594 + }, + { + "auxiliary_loss_clip": 0.01443034, + "auxiliary_loss_mlp": 0.01070731, + "balance_loss_clip": 1.12120163, + "balance_loss_mlp": 1.04399264, + "epoch": 0.021643720315036374, + "flos": 26104005331200.0, + "grad_norm": 2.1688820580905457, + "language_loss": 0.89568222, + "learning_rate": 3.7620165591252806e-06, + "loss": 0.92081988, + "num_input_tokens_seen": 3763035, + "step": 180, + "time_per_iteration": 2.8418807983398438 + }, + { + "auxiliary_loss_clip": 0.01431313, + "auxiliary_loss_mlp": 0.01071599, + "balance_loss_clip": 1.11804461, + "balance_loss_mlp": 1.04878283, + "epoch": 0.021763963205675464, + "flos": 24787663614720.0, + "grad_norm": 2.8028590721825255, + "language_loss": 0.94687277, + "learning_rate": 3.766030120360636e-06, + "loss": 0.97190189, + "num_input_tokens_seen": 3782665, + "step": 181, + "time_per_iteration": 2.950482130050659 + }, + { + "auxiliary_loss_clip": 0.01428953, + "auxiliary_loss_mlp": 0.01073957, + "balance_loss_clip": 1.1172086, + "balance_loss_mlp": 1.05040169, + "epoch": 0.021884206096314557, + "flos": 25813559957760.0, + "grad_norm": 2.021086913133383, + "language_loss": 0.90591586, + "learning_rate": 3.7700215681987578e-06, + "loss": 0.93094498, + "num_input_tokens_seen": 3802435, + "step": 182, + "time_per_iteration": 2.7555129528045654 + }, + { + "auxiliary_loss_clip": 0.01426275, + "auxiliary_loss_mlp": 0.01070879, + "balance_loss_clip": 1.11383569, + "balance_loss_mlp": 1.04604745, + "epoch": 0.022004448986953647, + "flos": 20082719721600.0, + "grad_norm": 1.9428963221708409, + "language_loss": 0.82348675, + "learning_rate": 3.7739911449800767e-06, + "loss": 0.84845829, + "num_input_tokens_seen": 3822490, + "step": 183, + "time_per_iteration": 2.720649242401123 + }, + { + "auxiliary_loss_clip": 0.01428912, + "auxiliary_loss_mlp": 0.01066552, + "balance_loss_clip": 1.11483932, + "balance_loss_mlp": 1.04074383, + "epoch": 0.022124691877592736, + "flos": 20480609652480.0, + "grad_norm": 1.8428330840338056, + "language_loss": 0.80856758, + "learning_rate": 3.7779390890830114e-06, + "loss": 0.8335222, + "num_input_tokens_seen": 3841140, + "step": 184, + "time_per_iteration": 2.900942325592041 + }, + { + "auxiliary_loss_clip": 0.01429181, + "auxiliary_loss_mlp": 0.01061094, + "balance_loss_clip": 1.11426282, + "balance_loss_mlp": 1.03585708, + "epoch": 0.02224493476823183, + "flos": 23586847015680.0, + "grad_norm": 1.8458759420148025, + "language_loss": 0.86044616, + "learning_rate": 3.7818656350098723e-06, + "loss": 0.88534892, + "num_input_tokens_seen": 3862090, + "step": 185, + "time_per_iteration": 2.85626482963562 + }, + { + "auxiliary_loss_clip": 0.01426469, + "auxiliary_loss_mlp": 0.0106519, + "balance_loss_clip": 1.11334145, + "balance_loss_mlp": 1.03914309, + "epoch": 0.02236517765887092, + "flos": 16909940413440.0, + "grad_norm": 2.6391705556988048, + "language_loss": 0.77130353, + "learning_rate": 3.7857710134704447e-06, + "loss": 0.79622012, + "num_input_tokens_seen": 3881025, + "step": 186, + "time_per_iteration": 2.802680253982544 + }, + { + "auxiliary_loss_clip": 0.01416192, + "auxiliary_loss_mlp": 0.01074182, + "balance_loss_clip": 1.11478353, + "balance_loss_mlp": 1.0499351, + "epoch": 0.02248542054951001, + "flos": 43508182930560.0, + "grad_norm": 3.444727523851284, + "language_loss": 0.79723728, + "learning_rate": 3.7896554514633234e-06, + "loss": 0.82214105, + "num_input_tokens_seen": 3905310, + "step": 187, + "time_per_iteration": 2.9336109161376953 + }, + { + "auxiliary_loss_clip": 0.01416025, + "auxiliary_loss_mlp": 0.0106388, + "balance_loss_clip": 1.11126375, + "balance_loss_mlp": 1.03912008, + "epoch": 0.022605663440149103, + "flos": 23367648268800.0, + "grad_norm": 3.6784803643912305, + "language_loss": 0.84182835, + "learning_rate": 3.7935191723550955e-06, + "loss": 0.86662734, + "num_input_tokens_seen": 3924265, + "step": 188, + "time_per_iteration": 2.759539842605591 + }, + { + "auxiliary_loss_clip": 0.01415388, + "auxiliary_loss_mlp": 0.01072397, + "balance_loss_clip": 1.11199737, + "balance_loss_mlp": 1.05022407, + "epoch": 0.022725906330788193, + "flos": 29019915504000.0, + "grad_norm": 2.7139122037464287, + "language_loss": 0.88436687, + "learning_rate": 3.797362395957408e-06, + "loss": 0.90924478, + "num_input_tokens_seen": 3944830, + "step": 189, + "time_per_iteration": 2.79717755317688 + }, + { + "auxiliary_loss_clip": 0.01426843, + "auxiliary_loss_mlp": 0.01073777, + "balance_loss_clip": 1.11816049, + "balance_loss_mlp": 1.04709864, + "epoch": 0.022846149221427282, + "flos": 24496176746880.0, + "grad_norm": 2.3526372618400218, + "language_loss": 0.78197742, + "learning_rate": 3.8011853386020055e-06, + "loss": 0.80698359, + "num_input_tokens_seen": 3965735, + "step": 190, + "time_per_iteration": 2.819347620010376 + }, + { + "auxiliary_loss_clip": 0.01418794, + "auxiliary_loss_mlp": 0.01069363, + "balance_loss_clip": 1.11492252, + "balance_loss_mlp": 1.04504478, + "epoch": 0.022966392112066376, + "flos": 15523537219200.0, + "grad_norm": 4.539970575028367, + "language_loss": 0.89796519, + "learning_rate": 3.804988213213804e-06, + "loss": 0.92284679, + "num_input_tokens_seen": 3983975, + "step": 191, + "time_per_iteration": 2.7052812576293945 + }, + { + "auxiliary_loss_clip": 0.01367594, + "auxiliary_loss_mlp": 0.0101883, + "balance_loss_clip": 1.16988015, + "balance_loss_mlp": 1.00376189, + "epoch": 0.023086635002705466, + "flos": 55650408433920.0, + "grad_norm": 1.0185117977316498, + "language_loss": 0.63124174, + "learning_rate": 3.808771229382049e-06, + "loss": 0.65510595, + "num_input_tokens_seen": 4043440, + "step": 192, + "time_per_iteration": 3.2353522777557373 + }, + { + "auxiliary_loss_clip": 0.01416001, + "auxiliary_loss_mlp": 0.01058974, + "balance_loss_clip": 1.11241341, + "balance_loss_mlp": 1.03595519, + "epoch": 0.023206877893344555, + "flos": 19313441118720.0, + "grad_norm": 2.0030920812112, + "language_loss": 0.84609503, + "learning_rate": 3.8125345934296324e-06, + "loss": 0.87084472, + "num_input_tokens_seen": 4061750, + "step": 193, + "time_per_iteration": 2.851071357727051 + }, + { + "auxiliary_loss_clip": 0.01416362, + "auxiliary_loss_mlp": 0.01069418, + "balance_loss_clip": 1.11108804, + "balance_loss_mlp": 1.04424119, + "epoch": 0.02332712078398365, + "flos": 23072965090560.0, + "grad_norm": 2.091037988516509, + "language_loss": 0.88219011, + "learning_rate": 3.81627850848061e-06, + "loss": 0.90704787, + "num_input_tokens_seen": 4082345, + "step": 194, + "time_per_iteration": 2.8150198459625244 + }, + { + "auxiliary_loss_clip": 0.01411348, + "auxiliary_loss_mlp": 0.01063459, + "balance_loss_clip": 1.11170173, + "balance_loss_mlp": 1.03813934, + "epoch": 0.02344736367462274, + "flos": 24425971614720.0, + "grad_norm": 2.413886517772201, + "language_loss": 0.8651197, + "learning_rate": 3.820003174525994e-06, + "loss": 0.88986772, + "num_input_tokens_seen": 4101770, + "step": 195, + "time_per_iteration": 2.9112706184387207 + }, + { + "auxiliary_loss_clip": 0.01407461, + "auxiliary_loss_mlp": 0.01065027, + "balance_loss_clip": 1.11164284, + "balance_loss_mlp": 1.04220986, + "epoch": 0.02356760656526183, + "flos": 21579799697280.0, + "grad_norm": 2.4806922628634243, + "language_loss": 0.82955158, + "learning_rate": 3.823708788487851e-06, + "loss": 0.85427648, + "num_input_tokens_seen": 4118770, + "step": 196, + "time_per_iteration": 2.842906951904297 + }, + { + "auxiliary_loss_clip": 0.01409377, + "auxiliary_loss_mlp": 0.01058573, + "balance_loss_clip": 1.11452162, + "balance_loss_mlp": 1.03543425, + "epoch": 0.02368784945590092, + "flos": 25193598192000.0, + "grad_norm": 1.7919600226248324, + "language_loss": 0.84531999, + "learning_rate": 3.827395544281781e-06, + "loss": 0.86999953, + "num_input_tokens_seen": 4141110, + "step": 197, + "time_per_iteration": 2.851987838745117 + }, + { + "auxiliary_loss_clip": 0.01412574, + "auxiliary_loss_mlp": 0.01066331, + "balance_loss_clip": 1.11251998, + "balance_loss_mlp": 1.04222715, + "epoch": 0.02380809234654001, + "flos": 27562481164800.0, + "grad_norm": 1.7435772665244405, + "language_loss": 0.78743082, + "learning_rate": 3.831063632877802e-06, + "loss": 0.81221986, + "num_input_tokens_seen": 4161430, + "step": 198, + "time_per_iteration": 4.638232707977295 + }, + { + "auxiliary_loss_clip": 0.01410703, + "auxiliary_loss_mlp": 0.01066868, + "balance_loss_clip": 1.11985958, + "balance_loss_mlp": 1.04493392, + "epoch": 0.0239283352371791, + "flos": 18259786540800.0, + "grad_norm": 2.3367671010918505, + "language_loss": 0.76033068, + "learning_rate": 3.834713242359712e-06, + "loss": 0.78510636, + "num_input_tokens_seen": 4179260, + "step": 199, + "time_per_iteration": 2.804870128631592 + }, + { + "auxiliary_loss_clip": 0.0141328, + "auxiliary_loss_mlp": 0.01076869, + "balance_loss_clip": 1.11441278, + "balance_loss_mlp": 1.05175161, + "epoch": 0.02404857812781819, + "flos": 21395110942080.0, + "grad_norm": 3.2612131670876825, + "language_loss": 0.87400782, + "learning_rate": 3.838344557982959e-06, + "loss": 0.89890933, + "num_input_tokens_seen": 4200640, + "step": 200, + "time_per_iteration": 3.589111804962158 + }, + { + "auxiliary_loss_clip": 0.01396557, + "auxiliary_loss_mlp": 0.01070216, + "balance_loss_clip": 1.10543084, + "balance_loss_mlp": 1.04635024, + "epoch": 0.024168821018457284, + "flos": 16654256426880.0, + "grad_norm": 2.8297874601275814, + "language_loss": 0.84564036, + "learning_rate": 3.841957762231063e-06, + "loss": 0.87030804, + "num_input_tokens_seen": 4218170, + "step": 201, + "time_per_iteration": 2.7873778343200684 + }, + { + "auxiliary_loss_clip": 0.01396987, + "auxiliary_loss_mlp": 0.01067626, + "balance_loss_clip": 1.10419536, + "balance_loss_mlp": 1.04491663, + "epoch": 0.024289063909096374, + "flos": 22820872464000.0, + "grad_norm": 2.7031006905523842, + "language_loss": 0.87854159, + "learning_rate": 3.8455530348706454e-06, + "loss": 0.90318775, + "num_input_tokens_seen": 4237770, + "step": 202, + "time_per_iteration": 2.7676217555999756 + }, + { + "auxiliary_loss_clip": 0.01398497, + "auxiliary_loss_mlp": 0.01069535, + "balance_loss_clip": 1.10974455, + "balance_loss_mlp": 1.0465157, + "epoch": 0.024409306799735464, + "flos": 17748598135680.0, + "grad_norm": 1.9103593416342377, + "language_loss": 0.77479357, + "learning_rate": 3.849130553005099e-06, + "loss": 0.79947388, + "num_input_tokens_seen": 4255985, + "step": 203, + "time_per_iteration": 2.8371636867523193 + }, + { + "auxiliary_loss_clip": 0.01402591, + "auxiliary_loss_mlp": 0.01063222, + "balance_loss_clip": 1.11003852, + "balance_loss_mlp": 1.03941572, + "epoch": 0.024529549690374557, + "flos": 21616213109760.0, + "grad_norm": 1.7302166061437363, + "language_loss": 0.83820605, + "learning_rate": 3.852690491126933e-06, + "loss": 0.86286426, + "num_input_tokens_seen": 4276035, + "step": 204, + "time_per_iteration": 2.7771992683410645 + }, + { + "auxiliary_loss_clip": 0.01399984, + "auxiliary_loss_mlp": 0.01061137, + "balance_loss_clip": 1.10729802, + "balance_loss_mlp": 1.03730714, + "epoch": 0.024649792581013647, + "flos": 25551662918400.0, + "grad_norm": 2.4387818671507904, + "language_loss": 0.91712296, + "learning_rate": 3.856233021168845e-06, + "loss": 0.94173419, + "num_input_tokens_seen": 4295730, + "step": 205, + "time_per_iteration": 2.876668930053711 + }, + { + "auxiliary_loss_clip": 0.01387834, + "auxiliary_loss_mlp": 0.01064386, + "balance_loss_clip": 1.10368359, + "balance_loss_mlp": 1.04327369, + "epoch": 0.024770035471652737, + "flos": 34495574544000.0, + "grad_norm": 2.2445670642641966, + "language_loss": 0.91495413, + "learning_rate": 3.859758312553544e-06, + "loss": 0.93947637, + "num_input_tokens_seen": 4317950, + "step": 206, + "time_per_iteration": 2.934697389602661 + }, + { + "auxiliary_loss_clip": 0.01400654, + "auxiliary_loss_mlp": 0.01070092, + "balance_loss_clip": 1.11088049, + "balance_loss_mlp": 1.04663205, + "epoch": 0.02489027836229183, + "flos": 21505428587520.0, + "grad_norm": 3.521562265401886, + "language_loss": 0.91978037, + "learning_rate": 3.8632665322423735e-06, + "loss": 0.94448775, + "num_input_tokens_seen": 4337605, + "step": 207, + "time_per_iteration": 2.7552506923675537 + }, + { + "auxiliary_loss_clip": 0.0139731, + "auxiliary_loss_mlp": 0.01065218, + "balance_loss_clip": 1.10997951, + "balance_loss_mlp": 1.04286647, + "epoch": 0.02501052125293092, + "flos": 23219013790080.0, + "grad_norm": 2.4167898685961835, + "language_loss": 0.86080587, + "learning_rate": 3.866757844782762e-06, + "loss": 0.88543111, + "num_input_tokens_seen": 4358110, + "step": 208, + "time_per_iteration": 2.886643171310425 + }, + { + "auxiliary_loss_clip": 0.01393581, + "auxiliary_loss_mlp": 0.01060819, + "balance_loss_clip": 1.10894608, + "balance_loss_mlp": 1.0379312, + "epoch": 0.02513076414357001, + "flos": 26388920010240.0, + "grad_norm": 5.987807771429256, + "language_loss": 0.91489172, + "learning_rate": 3.870232412354527e-06, + "loss": 0.93943578, + "num_input_tokens_seen": 4374955, + "step": 209, + "time_per_iteration": 2.7972395420074463 + }, + { + "auxiliary_loss_clip": 0.01392429, + "auxiliary_loss_mlp": 0.01065525, + "balance_loss_clip": 1.10448718, + "balance_loss_mlp": 1.04257715, + "epoch": 0.025251007034209103, + "flos": 13590430047360.0, + "grad_norm": 2.4912555912435574, + "language_loss": 0.92857635, + "learning_rate": 3.873690394815086e-06, + "loss": 0.95315588, + "num_input_tokens_seen": 4391535, + "step": 210, + "time_per_iteration": 2.8232717514038086 + }, + { + "auxiliary_loss_clip": 0.01396291, + "auxiliary_loss_mlp": 0.01064001, + "balance_loss_clip": 1.10597181, + "balance_loss_mlp": 1.04098213, + "epoch": 0.025371249924848193, + "flos": 15049229103360.0, + "grad_norm": 2.805248130830032, + "language_loss": 0.91161638, + "learning_rate": 3.877131949743587e-06, + "loss": 0.93621933, + "num_input_tokens_seen": 4408400, + "step": 211, + "time_per_iteration": 2.691856622695923 + }, + { + "auxiliary_loss_clip": 0.01388256, + "auxiliary_loss_mlp": 0.0106702, + "balance_loss_clip": 1.10438275, + "balance_loss_mlp": 1.04404795, + "epoch": 0.025491492815487283, + "flos": 25553853648000.0, + "grad_norm": 2.3850908306845815, + "language_loss": 0.78146487, + "learning_rate": 3.880557232483993e-06, + "loss": 0.80601764, + "num_input_tokens_seen": 4427840, + "step": 212, + "time_per_iteration": 2.808873176574707 + }, + { + "auxiliary_loss_clip": 0.01382408, + "auxiliary_loss_mlp": 0.01076157, + "balance_loss_clip": 1.09944427, + "balance_loss_mlp": 1.05311394, + "epoch": 0.025611735706126376, + "flos": 20630752502400.0, + "grad_norm": 1.9123716385012282, + "language_loss": 0.87127995, + "learning_rate": 3.883966396187164e-06, + "loss": 0.89586562, + "num_input_tokens_seen": 4447110, + "step": 213, + "time_per_iteration": 2.778974771499634 + }, + { + "auxiliary_loss_clip": 0.01385285, + "auxiliary_loss_mlp": 0.01068115, + "balance_loss_clip": 1.10461497, + "balance_loss_mlp": 1.04494107, + "epoch": 0.025731978596765466, + "flos": 19062282245760.0, + "grad_norm": 2.0326926147979005, + "language_loss": 0.90061307, + "learning_rate": 3.887359591851937e-06, + "loss": 0.92514712, + "num_input_tokens_seen": 4464715, + "step": 214, + "time_per_iteration": 2.7777915000915527 + }, + { + "auxiliary_loss_clip": 0.01380067, + "auxiliary_loss_mlp": 0.010492, + "balance_loss_clip": 1.10350609, + "balance_loss_mlp": 1.02655017, + "epoch": 0.025852221487404556, + "flos": 22163814927360.0, + "grad_norm": 1.94842715032364, + "language_loss": 0.92331302, + "learning_rate": 3.890736968365265e-06, + "loss": 0.94760573, + "num_input_tokens_seen": 4485030, + "step": 215, + "time_per_iteration": 2.741098403930664 + }, + { + "auxiliary_loss_clip": 0.01383076, + "auxiliary_loss_mlp": 0.01047939, + "balance_loss_clip": 1.10177922, + "balance_loss_mlp": 1.02501488, + "epoch": 0.02597246437804365, + "flos": 26541971861760.0, + "grad_norm": 1.9013826912666458, + "language_loss": 0.85352027, + "learning_rate": 3.894098672541412e-06, + "loss": 0.87783039, + "num_input_tokens_seen": 4505935, + "step": 216, + "time_per_iteration": 2.7544333934783936 + }, + { + "auxiliary_loss_clip": 0.01383158, + "auxiliary_loss_mlp": 0.01060955, + "balance_loss_clip": 1.1036396, + "balance_loss_mlp": 1.0374229, + "epoch": 0.02609270726868274, + "flos": 32671671696000.0, + "grad_norm": 1.781607611974906, + "language_loss": 0.7542991, + "learning_rate": 3.89744484916025e-06, + "loss": 0.77874029, + "num_input_tokens_seen": 4527045, + "step": 217, + "time_per_iteration": 2.9025824069976807 + }, + { + "auxiliary_loss_clip": 0.01385207, + "auxiliary_loss_mlp": 0.0106619, + "balance_loss_clip": 1.10285783, + "balance_loss_mlp": 1.04069149, + "epoch": 0.02621295015932183, + "flos": 26243553669120.0, + "grad_norm": 2.2966915722857038, + "language_loss": 0.87487435, + "learning_rate": 3.900775641004673e-06, + "loss": 0.89938837, + "num_input_tokens_seen": 4546360, + "step": 218, + "time_per_iteration": 2.8278231620788574 + }, + { + "auxiliary_loss_clip": 0.01393559, + "auxiliary_loss_mlp": 0.01071088, + "balance_loss_clip": 1.10780752, + "balance_loss_mlp": 1.04626846, + "epoch": 0.026333193049960922, + "flos": 42921402353280.0, + "grad_norm": 2.6020782973650527, + "language_loss": 0.74166918, + "learning_rate": 3.904091188897156e-06, + "loss": 0.76631558, + "num_input_tokens_seen": 4565495, + "step": 219, + "time_per_iteration": 2.894645929336548 + }, + { + "auxiliary_loss_clip": 0.01387813, + "auxiliary_loss_mlp": 0.01074733, + "balance_loss_clip": 1.10522532, + "balance_loss_mlp": 1.05086732, + "epoch": 0.026453435940600012, + "flos": 17963846386560.0, + "grad_norm": 2.385698537480927, + "language_loss": 0.81663162, + "learning_rate": 3.90739163173548e-06, + "loss": 0.8412571, + "num_input_tokens_seen": 4583330, + "step": 220, + "time_per_iteration": 2.7084672451019287 + }, + { + "auxiliary_loss_clip": 0.01381334, + "auxiliary_loss_mlp": 0.0107027, + "balance_loss_clip": 1.10223019, + "balance_loss_mlp": 1.0474894, + "epoch": 0.026573678831239102, + "flos": 18984319776000.0, + "grad_norm": 3.360047666381712, + "language_loss": 0.88736272, + "learning_rate": 3.910677106527646e-06, + "loss": 0.91187871, + "num_input_tokens_seen": 4600520, + "step": 221, + "time_per_iteration": 2.778254985809326 + }, + { + "auxiliary_loss_clip": 0.01378987, + "auxiliary_loss_mlp": 0.01072486, + "balance_loss_clip": 1.10307848, + "balance_loss_mlp": 1.05008638, + "epoch": 0.026693921721878195, + "flos": 29241448634880.0, + "grad_norm": 2.2771468819604945, + "language_loss": 0.8449918, + "learning_rate": 3.913947748426004e-06, + "loss": 0.86950648, + "num_input_tokens_seen": 4617340, + "step": 222, + "time_per_iteration": 2.830904722213745 + }, + { + "auxiliary_loss_clip": 0.01387284, + "auxiliary_loss_mlp": 0.01063837, + "balance_loss_clip": 1.10936165, + "balance_loss_mlp": 1.04110408, + "epoch": 0.026814164612517285, + "flos": 14128083797760.0, + "grad_norm": 2.571426754751136, + "language_loss": 0.76276571, + "learning_rate": 3.9172036907606136e-06, + "loss": 0.78727692, + "num_input_tokens_seen": 4630820, + "step": 223, + "time_per_iteration": 2.6647069454193115 + }, + { + "auxiliary_loss_clip": 0.0138154, + "auxiliary_loss_mlp": 0.01073959, + "balance_loss_clip": 1.10210228, + "balance_loss_mlp": 1.0499264, + "epoch": 0.026934407503156375, + "flos": 23511973115520.0, + "grad_norm": 1.8635296192335196, + "language_loss": 0.95203149, + "learning_rate": 3.920445065071855e-06, + "loss": 0.97658646, + "num_input_tokens_seen": 4651985, + "step": 224, + "time_per_iteration": 3.5859642028808594 + }, + { + "auxiliary_loss_clip": 0.01385625, + "auxiliary_loss_mlp": 0.01068021, + "balance_loss_clip": 1.10679841, + "balance_loss_mlp": 1.04488277, + "epoch": 0.027054650393795468, + "flos": 28950356816640.0, + "grad_norm": 2.589861678631667, + "language_loss": 0.79747212, + "learning_rate": 3.923672001142322e-06, + "loss": 0.82200861, + "num_input_tokens_seen": 4672295, + "step": 225, + "time_per_iteration": 3.7104554176330566 + }, + { + "auxiliary_loss_clip": 0.01377951, + "auxiliary_loss_mlp": 0.0107943, + "balance_loss_clip": 1.10262549, + "balance_loss_mlp": 1.05630326, + "epoch": 0.027174893284434558, + "flos": 31431568596480.0, + "grad_norm": 3.1585664432928136, + "language_loss": 0.84516412, + "learning_rate": 3.926884627027996e-06, + "loss": 0.86973786, + "num_input_tokens_seen": 4696065, + "step": 226, + "time_per_iteration": 3.768834114074707 + }, + { + "auxiliary_loss_clip": 0.01376381, + "auxiliary_loss_mlp": 0.01061343, + "balance_loss_clip": 1.10027754, + "balance_loss_mlp": 1.03721523, + "epoch": 0.027295136175073648, + "flos": 22054466949120.0, + "grad_norm": 1.9929314424001636, + "language_loss": 0.77547914, + "learning_rate": 3.930083069088744e-06, + "loss": 0.79985636, + "num_input_tokens_seen": 4716065, + "step": 227, + "time_per_iteration": 2.7433767318725586 + }, + { + "auxiliary_loss_clip": 0.01321461, + "auxiliary_loss_mlp": 0.01017065, + "balance_loss_clip": 1.14554703, + "balance_loss_mlp": 1.00276017, + "epoch": 0.02741537906571274, + "flos": 60800752972800.0, + "grad_norm": 0.9829547924581161, + "language_loss": 0.59328783, + "learning_rate": 3.933267452018137e-06, + "loss": 0.61667299, + "num_input_tokens_seen": 4775860, + "step": 228, + "time_per_iteration": 3.335780382156372 + }, + { + "auxiliary_loss_clip": 0.01374716, + "auxiliary_loss_mlp": 0.01056114, + "balance_loss_clip": 1.1041019, + "balance_loss_mlp": 1.03491879, + "epoch": 0.02753562195635183, + "flos": 24606278910720.0, + "grad_norm": 2.6120362862235647, + "language_loss": 0.84556186, + "learning_rate": 3.936437898872622e-06, + "loss": 0.86987019, + "num_input_tokens_seen": 4795835, + "step": 229, + "time_per_iteration": 2.9109737873077393 + }, + { + "auxiliary_loss_clip": 0.01375492, + "auxiliary_loss_mlp": 0.01070131, + "balance_loss_clip": 1.10190892, + "balance_loss_mlp": 1.04771996, + "epoch": 0.02765586484699092, + "flos": 34094236907520.0, + "grad_norm": 2.325895107759685, + "language_loss": 0.79619896, + "learning_rate": 3.9395945311000525e-06, + "loss": 0.82065523, + "num_input_tokens_seen": 4817460, + "step": 230, + "time_per_iteration": 2.90592098236084 + }, + { + "auxiliary_loss_clip": 0.01375324, + "auxiliary_loss_mlp": 0.01063827, + "balance_loss_clip": 1.1023078, + "balance_loss_mlp": 1.04140353, + "epoch": 0.027776107737630014, + "flos": 14829922615680.0, + "grad_norm": 5.992723633155718, + "language_loss": 0.9099735, + "learning_rate": 3.942737468567608e-06, + "loss": 0.93436503, + "num_input_tokens_seen": 4835475, + "step": 231, + "time_per_iteration": 2.8586225509643555 + }, + { + "auxiliary_loss_clip": 0.01372735, + "auxiliary_loss_mlp": 0.01057308, + "balance_loss_clip": 1.10276723, + "balance_loss_mlp": 1.03608882, + "epoch": 0.027896350628269104, + "flos": 47920347066240.0, + "grad_norm": 2.276395046735701, + "language_loss": 0.86267757, + "learning_rate": 3.9458668295891026e-06, + "loss": 0.88697809, + "num_input_tokens_seen": 4857760, + "step": 232, + "time_per_iteration": 3.1828384399414062 + }, + { + "auxiliary_loss_clip": 0.01368695, + "auxiliary_loss_mlp": 0.01068913, + "balance_loss_clip": 1.09924376, + "balance_loss_mlp": 1.0452615, + "epoch": 0.028016593518908194, + "flos": 21684550734720.0, + "grad_norm": 4.183160178377811, + "language_loss": 0.86676693, + "learning_rate": 3.948982730951712e-06, + "loss": 0.89114296, + "num_input_tokens_seen": 4875855, + "step": 233, + "time_per_iteration": 2.7776193618774414 + }, + { + "auxiliary_loss_clip": 0.01373818, + "auxiliary_loss_mlp": 0.01056482, + "balance_loss_clip": 1.10051751, + "balance_loss_mlp": 1.03366482, + "epoch": 0.028136836409547287, + "flos": 18439483305600.0, + "grad_norm": 4.800805606546107, + "language_loss": 0.82361042, + "learning_rate": 3.9520852879421254e-06, + "loss": 0.84791338, + "num_input_tokens_seen": 4893200, + "step": 234, + "time_per_iteration": 2.7692408561706543 + }, + { + "auxiliary_loss_clip": 0.01372171, + "auxiliary_loss_mlp": 0.01057887, + "balance_loss_clip": 1.10324597, + "balance_loss_mlp": 1.03518915, + "epoch": 0.028257079300186377, + "flos": 31576934937600.0, + "grad_norm": 2.167536019618749, + "language_loss": 0.81481659, + "learning_rate": 3.955174614372137e-06, + "loss": 0.83911717, + "num_input_tokens_seen": 4912965, + "step": 235, + "time_per_iteration": 2.9328269958496094 + }, + { + "auxiliary_loss_clip": 0.01369897, + "auxiliary_loss_mlp": 0.01062615, + "balance_loss_clip": 1.10127807, + "balance_loss_mlp": 1.04004884, + "epoch": 0.028377322190825467, + "flos": 23513337832320.0, + "grad_norm": 3.877259141532309, + "language_loss": 0.84468615, + "learning_rate": 3.9582508226037045e-06, + "loss": 0.86901122, + "num_input_tokens_seen": 4933105, + "step": 236, + "time_per_iteration": 2.822815179824829 + }, + { + "auxiliary_loss_clip": 0.01379314, + "auxiliary_loss_mlp": 0.01057962, + "balance_loss_clip": 1.10394382, + "balance_loss_mlp": 1.03251123, + "epoch": 0.02849756508146456, + "flos": 20479604071680.0, + "grad_norm": 2.69343619099514, + "language_loss": 0.94124687, + "learning_rate": 3.9613140235734636e-06, + "loss": 0.96561968, + "num_input_tokens_seen": 4950085, + "step": 237, + "time_per_iteration": 2.766449451446533 + }, + { + "auxiliary_loss_clip": 0.01370084, + "auxiliary_loss_mlp": 0.01065948, + "balance_loss_clip": 1.10182214, + "balance_loss_mlp": 1.04357243, + "epoch": 0.02861780797210365, + "flos": 14283362292480.0, + "grad_norm": 2.244400109658795, + "language_loss": 0.80989301, + "learning_rate": 3.96436432681674e-06, + "loss": 0.83425331, + "num_input_tokens_seen": 4968075, + "step": 238, + "time_per_iteration": 2.7031772136688232 + }, + { + "auxiliary_loss_clip": 0.01368362, + "auxiliary_loss_mlp": 0.01074656, + "balance_loss_clip": 1.09967875, + "balance_loss_mlp": 1.05194712, + "epoch": 0.02873805086274274, + "flos": 25808532053760.0, + "grad_norm": 2.1006884647723525, + "language_loss": 0.89331377, + "learning_rate": 3.967401840491044e-06, + "loss": 0.91774392, + "num_input_tokens_seen": 4987355, + "step": 239, + "time_per_iteration": 2.830690622329712 + }, + { + "auxiliary_loss_clip": 0.01360761, + "auxiliary_loss_mlp": 0.0106603, + "balance_loss_clip": 1.09869599, + "balance_loss_mlp": 1.04444122, + "epoch": 0.028858293753381833, + "flos": 17304238984320.0, + "grad_norm": 2.351905022961887, + "language_loss": 0.88007069, + "learning_rate": 3.97042667139909e-06, + "loss": 0.9043386, + "num_input_tokens_seen": 5004680, + "step": 240, + "time_per_iteration": 2.7648637294769287 + }, + { + "auxiliary_loss_clip": 0.01363973, + "auxiliary_loss_mlp": 0.01062241, + "balance_loss_clip": 1.10007, + "balance_loss_mlp": 1.0394243, + "epoch": 0.028978536644020923, + "flos": 23038347358080.0, + "grad_norm": 2.331329495469027, + "language_loss": 0.87718546, + "learning_rate": 3.973438925011327e-06, + "loss": 0.90144765, + "num_input_tokens_seen": 5022965, + "step": 241, + "time_per_iteration": 2.832559823989868 + }, + { + "auxiliary_loss_clip": 0.01369, + "auxiliary_loss_mlp": 0.0106034, + "balance_loss_clip": 1.10069335, + "balance_loss_mlp": 1.03803611, + "epoch": 0.029098779534660012, + "flos": 28329712692480.0, + "grad_norm": 2.6497571846542827, + "language_loss": 0.91326672, + "learning_rate": 3.976438705488002e-06, + "loss": 0.93756008, + "num_input_tokens_seen": 5042625, + "step": 242, + "time_per_iteration": 2.915846824645996 + }, + { + "auxiliary_loss_clip": 0.01365986, + "auxiliary_loss_mlp": 0.01062611, + "balance_loss_clip": 1.10343397, + "balance_loss_mlp": 1.04081988, + "epoch": 0.029219022425299106, + "flos": 13881665520000.0, + "grad_norm": 3.000934664938178, + "language_loss": 0.92882472, + "learning_rate": 3.9794261157007744e-06, + "loss": 0.95311069, + "num_input_tokens_seen": 5060380, + "step": 243, + "time_per_iteration": 2.7642717361450195 + }, + { + "auxiliary_loss_clip": 0.01373433, + "auxiliary_loss_mlp": 0.01056888, + "balance_loss_clip": 1.10505044, + "balance_loss_mlp": 1.03286755, + "epoch": 0.029339265315938196, + "flos": 19422501788160.0, + "grad_norm": 2.3804583426529646, + "language_loss": 0.8480171, + "learning_rate": 3.982401257253887e-06, + "loss": 0.87232029, + "num_input_tokens_seen": 5078720, + "step": 244, + "time_per_iteration": 2.8568665981292725 + }, + { + "auxiliary_loss_clip": 0.01364019, + "auxiliary_loss_mlp": 0.01066563, + "balance_loss_clip": 1.09974432, + "balance_loss_mlp": 1.04449749, + "epoch": 0.029459508206577285, + "flos": 15669550005120.0, + "grad_norm": 3.0176210538416184, + "language_loss": 0.89847082, + "learning_rate": 3.985364230504893e-06, + "loss": 0.92277664, + "num_input_tokens_seen": 5096605, + "step": 245, + "time_per_iteration": 2.884274482727051 + }, + { + "auxiliary_loss_clip": 0.013648, + "auxiliary_loss_mlp": 0.01070739, + "balance_loss_clip": 1.10228813, + "balance_loss_mlp": 1.04781532, + "epoch": 0.02957975109721638, + "flos": 28220975245440.0, + "grad_norm": 2.008284884919104, + "language_loss": 0.84388685, + "learning_rate": 3.988315134584976e-06, + "loss": 0.86824226, + "num_input_tokens_seen": 5116285, + "step": 246, + "time_per_iteration": 2.8922431468963623 + }, + { + "auxiliary_loss_clip": 0.01365459, + "auxiliary_loss_mlp": 0.01058922, + "balance_loss_clip": 1.10063434, + "balance_loss_mlp": 1.03680837, + "epoch": 0.02969999398785547, + "flos": 24315869450880.0, + "grad_norm": 3.764505881567228, + "language_loss": 0.8070035, + "learning_rate": 3.991254067418851e-06, + "loss": 0.83124727, + "num_input_tokens_seen": 5136825, + "step": 247, + "time_per_iteration": 2.926658868789673 + }, + { + "auxiliary_loss_clip": 0.01360572, + "auxiliary_loss_mlp": 0.01061711, + "balance_loss_clip": 1.10065973, + "balance_loss_mlp": 1.04057479, + "epoch": 0.02982023687849456, + "flos": 35078584193280.0, + "grad_norm": 2.1449905574265893, + "language_loss": 0.82816005, + "learning_rate": 3.994181125744254e-06, + "loss": 0.8523829, + "num_input_tokens_seen": 5158630, + "step": 248, + "time_per_iteration": 2.9924397468566895 + }, + { + "auxiliary_loss_clip": 0.01356418, + "auxiliary_loss_mlp": 0.01059552, + "balance_loss_clip": 1.0988462, + "balance_loss_mlp": 1.03817749, + "epoch": 0.02994047976913365, + "flos": 26177155378560.0, + "grad_norm": 2.5876566433898187, + "language_loss": 0.74184215, + "learning_rate": 3.99709640513106e-06, + "loss": 0.76600182, + "num_input_tokens_seen": 5179510, + "step": 249, + "time_per_iteration": 2.8038177490234375 + }, + { + "auxiliary_loss_clip": 0.01365149, + "auxiliary_loss_mlp": 0.0106757, + "balance_loss_clip": 1.10063839, + "balance_loss_mlp": 1.0438478, + "epoch": 0.03006072265977274, + "flos": 25625028447360.0, + "grad_norm": 2.4702832227830402, + "language_loss": 0.85389149, + "learning_rate": 4e-06, + "loss": 0.87821877, + "num_input_tokens_seen": 5199345, + "step": 250, + "time_per_iteration": 2.7469286918640137 + }, + { + "auxiliary_loss_clip": 0.0136025, + "auxiliary_loss_mlp": 0.01069784, + "balance_loss_clip": 1.10237575, + "balance_loss_mlp": 1.04793274, + "epoch": 0.03018096555041183, + "flos": 22127078292480.0, + "grad_norm": 2.9047663867791655, + "language_loss": 0.88867974, + "learning_rate": 3.999999848300794e-06, + "loss": 0.91298008, + "num_input_tokens_seen": 5218330, + "step": 251, + "time_per_iteration": 4.904852867126465 + }, + { + "auxiliary_loss_clip": 0.01352939, + "auxiliary_loss_mlp": 0.01064738, + "balance_loss_clip": 1.09663248, + "balance_loss_mlp": 1.04204106, + "epoch": 0.030301208441050925, + "flos": 30188197359360.0, + "grad_norm": 1.6398248663817958, + "language_loss": 0.89215541, + "learning_rate": 3.999999393203203e-06, + "loss": 0.91633219, + "num_input_tokens_seen": 5240740, + "step": 252, + "time_per_iteration": 3.764141321182251 + }, + { + "auxiliary_loss_clip": 0.01355086, + "auxiliary_loss_mlp": 0.01055498, + "balance_loss_clip": 1.09846532, + "balance_loss_mlp": 1.03426671, + "epoch": 0.030421451331690014, + "flos": 23621392920960.0, + "grad_norm": 1.8905158483654614, + "language_loss": 0.85092127, + "learning_rate": 3.999998634707293e-06, + "loss": 0.87502712, + "num_input_tokens_seen": 5260290, + "step": 253, + "time_per_iteration": 3.858558416366577 + }, + { + "auxiliary_loss_clip": 0.01365135, + "auxiliary_loss_mlp": 0.01068499, + "balance_loss_clip": 1.10519028, + "balance_loss_mlp": 1.0460515, + "epoch": 0.030541694222329104, + "flos": 27928446883200.0, + "grad_norm": 2.6479935745129763, + "language_loss": 0.96831506, + "learning_rate": 3.999997572813182e-06, + "loss": 0.99265146, + "num_input_tokens_seen": 5278100, + "step": 254, + "time_per_iteration": 2.7879209518432617 + }, + { + "auxiliary_loss_clip": 0.01357463, + "auxiliary_loss_mlp": 0.01067376, + "balance_loss_clip": 1.09865141, + "balance_loss_mlp": 1.04510748, + "epoch": 0.030661937112968194, + "flos": 18588441006720.0, + "grad_norm": 2.4433155772157997, + "language_loss": 0.87562621, + "learning_rate": 3.999996207521028e-06, + "loss": 0.89987463, + "num_input_tokens_seen": 5296810, + "step": 255, + "time_per_iteration": 2.733639717102051 + }, + { + "auxiliary_loss_clip": 0.0135599, + "auxiliary_loss_mlp": 0.01055703, + "balance_loss_clip": 1.09837687, + "balance_loss_mlp": 1.03289866, + "epoch": 0.030782180003607287, + "flos": 12969139478400.0, + "grad_norm": 2.7049534453486936, + "language_loss": 0.82061249, + "learning_rate": 3.999994538831039e-06, + "loss": 0.84472942, + "num_input_tokens_seen": 5313395, + "step": 256, + "time_per_iteration": 2.7434544563293457 + }, + { + "auxiliary_loss_clip": 0.01355348, + "auxiliary_loss_mlp": 0.01058927, + "balance_loss_clip": 1.09900177, + "balance_loss_mlp": 1.0361104, + "epoch": 0.030902422894246377, + "flos": 23335364920320.0, + "grad_norm": 2.6233183884577316, + "language_loss": 0.85982406, + "learning_rate": 3.99999256674347e-06, + "loss": 0.8839668, + "num_input_tokens_seen": 5333545, + "step": 257, + "time_per_iteration": 2.808690071105957 + }, + { + "auxiliary_loss_clip": 0.0129443, + "auxiliary_loss_mlp": 0.01021776, + "balance_loss_clip": 1.12921786, + "balance_loss_mlp": 1.00871086, + "epoch": 0.031022665784885467, + "flos": 55094151438720.0, + "grad_norm": 1.028745840194357, + "language_loss": 0.53510737, + "learning_rate": 3.999990291258618e-06, + "loss": 0.5582695, + "num_input_tokens_seen": 5392235, + "step": 258, + "time_per_iteration": 3.269774913787842 + }, + { + "auxiliary_loss_clip": 0.01359651, + "auxiliary_loss_mlp": 0.01061369, + "balance_loss_clip": 1.10387909, + "balance_loss_mlp": 1.0400064, + "epoch": 0.03114290867552456, + "flos": 19317786664320.0, + "grad_norm": 3.7367197256819242, + "language_loss": 0.86748183, + "learning_rate": 3.999987712376829e-06, + "loss": 0.89169204, + "num_input_tokens_seen": 5410555, + "step": 259, + "time_per_iteration": 2.7814457416534424 + }, + { + "auxiliary_loss_clip": 0.01353967, + "auxiliary_loss_mlp": 0.01062255, + "balance_loss_clip": 1.10061407, + "balance_loss_mlp": 1.04058242, + "epoch": 0.031263151566163654, + "flos": 20959442881920.0, + "grad_norm": 3.187839031290697, + "language_loss": 0.82096189, + "learning_rate": 3.999984830098494e-06, + "loss": 0.84512413, + "num_input_tokens_seen": 5430135, + "step": 260, + "time_per_iteration": 2.858309268951416 + }, + { + "auxiliary_loss_clip": 0.01352818, + "auxiliary_loss_mlp": 0.010612, + "balance_loss_clip": 1.0987227, + "balance_loss_mlp": 1.03742933, + "epoch": 0.03138339445680274, + "flos": 14793006412800.0, + "grad_norm": 5.528773930121376, + "language_loss": 0.97874993, + "learning_rate": 3.999981644424051e-06, + "loss": 1.00289011, + "num_input_tokens_seen": 5444935, + "step": 261, + "time_per_iteration": 2.816800832748413 + }, + { + "auxiliary_loss_clip": 0.0135563, + "auxiliary_loss_mlp": 0.01060053, + "balance_loss_clip": 1.10225332, + "balance_loss_mlp": 1.03722489, + "epoch": 0.03150363734744183, + "flos": 11655599022720.0, + "grad_norm": 2.333673802479587, + "language_loss": 0.86408353, + "learning_rate": 3.999978155353982e-06, + "loss": 0.88824046, + "num_input_tokens_seen": 5462080, + "step": 262, + "time_per_iteration": 2.776834011077881 + }, + { + "auxiliary_loss_clip": 0.01353551, + "auxiliary_loss_mlp": 0.01064296, + "balance_loss_clip": 1.09797359, + "balance_loss_mlp": 1.04195571, + "epoch": 0.03162388023808092, + "flos": 33727732485120.0, + "grad_norm": 32.793083107928, + "language_loss": 0.80159098, + "learning_rate": 3.9999743628888186e-06, + "loss": 0.82576948, + "num_input_tokens_seen": 5483870, + "step": 263, + "time_per_iteration": 2.7940826416015625 + }, + { + "auxiliary_loss_clip": 0.01356477, + "auxiliary_loss_mlp": 0.01060008, + "balance_loss_clip": 1.10097468, + "balance_loss_mlp": 1.03735793, + "epoch": 0.03174412312872001, + "flos": 20810952057600.0, + "grad_norm": 2.3687334243515514, + "language_loss": 0.89619911, + "learning_rate": 3.999970267029133e-06, + "loss": 0.92036402, + "num_input_tokens_seen": 5502830, + "step": 264, + "time_per_iteration": 2.8570754528045654 + }, + { + "auxiliary_loss_clip": 0.0134924, + "auxiliary_loss_mlp": 0.01059741, + "balance_loss_clip": 1.098037, + "balance_loss_mlp": 1.03798509, + "epoch": 0.0318643660193591, + "flos": 23727939638400.0, + "grad_norm": 1.9205355710700047, + "language_loss": 0.80085242, + "learning_rate": 3.999965867775548e-06, + "loss": 0.82494229, + "num_input_tokens_seen": 5523225, + "step": 265, + "time_per_iteration": 2.8022778034210205 + }, + { + "auxiliary_loss_clip": 0.01350171, + "auxiliary_loss_mlp": 0.01056887, + "balance_loss_clip": 1.09808254, + "balance_loss_mlp": 1.03477407, + "epoch": 0.0319846089099982, + "flos": 13917863450880.0, + "grad_norm": 3.5215047338027747, + "language_loss": 0.86944151, + "learning_rate": 3.9999611651287315e-06, + "loss": 0.89351213, + "num_input_tokens_seen": 5541380, + "step": 266, + "time_per_iteration": 2.716967821121216 + }, + { + "auxiliary_loss_clip": 0.01350763, + "auxiliary_loss_mlp": 0.01055629, + "balance_loss_clip": 1.0979048, + "balance_loss_mlp": 1.03361154, + "epoch": 0.03210485180063729, + "flos": 14753253035520.0, + "grad_norm": 3.0068766790577213, + "language_loss": 0.78745675, + "learning_rate": 3.999956159089396e-06, + "loss": 0.81152064, + "num_input_tokens_seen": 5558830, + "step": 267, + "time_per_iteration": 2.6796367168426514 + }, + { + "auxiliary_loss_clip": 0.01351422, + "auxiliary_loss_mlp": 0.01057932, + "balance_loss_clip": 1.10044408, + "balance_loss_mlp": 1.03614068, + "epoch": 0.03222509469127638, + "flos": 28913153304960.0, + "grad_norm": 4.415113641262616, + "language_loss": 0.80089998, + "learning_rate": 3.999950849658302e-06, + "loss": 0.82499349, + "num_input_tokens_seen": 5577750, + "step": 268, + "time_per_iteration": 2.8208749294281006 + }, + { + "auxiliary_loss_clip": 0.01355862, + "auxiliary_loss_mlp": 0.01063404, + "balance_loss_clip": 1.10212672, + "balance_loss_mlp": 1.04123068, + "epoch": 0.03234533758191547, + "flos": 16946389739520.0, + "grad_norm": 2.2494287492346072, + "language_loss": 0.84200835, + "learning_rate": 3.999945236836254e-06, + "loss": 0.86620104, + "num_input_tokens_seen": 5596715, + "step": 269, + "time_per_iteration": 2.8871030807495117 + }, + { + "auxiliary_loss_clip": 0.0135664, + "auxiliary_loss_mlp": 0.0107212, + "balance_loss_clip": 1.10140347, + "balance_loss_mlp": 1.04843283, + "epoch": 0.03246558047255456, + "flos": 18989096284800.0, + "grad_norm": 2.611389367072968, + "language_loss": 0.95168245, + "learning_rate": 3.999939320624103e-06, + "loss": 0.97597003, + "num_input_tokens_seen": 5611865, + "step": 270, + "time_per_iteration": 2.6613051891326904 + }, + { + "auxiliary_loss_clip": 0.01356519, + "auxiliary_loss_mlp": 0.0105748, + "balance_loss_clip": 1.10235476, + "balance_loss_mlp": 1.03584361, + "epoch": 0.03258582336319365, + "flos": 23728334688000.0, + "grad_norm": 1.9823608796956476, + "language_loss": 0.90031874, + "learning_rate": 3.999933101022749e-06, + "loss": 0.92445868, + "num_input_tokens_seen": 5632270, + "step": 271, + "time_per_iteration": 2.7515692710876465 + }, + { + "auxiliary_loss_clip": 0.01352322, + "auxiliary_loss_mlp": 0.01068769, + "balance_loss_clip": 1.10049558, + "balance_loss_mlp": 1.04713237, + "epoch": 0.032706066253832745, + "flos": 27670823562240.0, + "grad_norm": 2.1001526245950624, + "language_loss": 0.86987191, + "learning_rate": 3.999926578033132e-06, + "loss": 0.89408278, + "num_input_tokens_seen": 5652085, + "step": 272, + "time_per_iteration": 2.7899973392486572 + }, + { + "auxiliary_loss_clip": 0.01354741, + "auxiliary_loss_mlp": 0.01062734, + "balance_loss_clip": 1.09949708, + "balance_loss_mlp": 1.03960741, + "epoch": 0.032826309144471835, + "flos": 45624685968000.0, + "grad_norm": 2.7945282553269255, + "language_loss": 0.63184726, + "learning_rate": 3.999919751656244e-06, + "loss": 0.65602195, + "num_input_tokens_seen": 5678985, + "step": 273, + "time_per_iteration": 3.0391132831573486 + }, + { + "auxiliary_loss_clip": 0.01351586, + "auxiliary_loss_mlp": 0.0107422, + "balance_loss_clip": 1.0999788, + "balance_loss_mlp": 1.05136764, + "epoch": 0.032946552035110925, + "flos": 25812374808960.0, + "grad_norm": 3.0773827169814174, + "language_loss": 0.75583768, + "learning_rate": 3.9999126218931195e-06, + "loss": 0.7800957, + "num_input_tokens_seen": 5697020, + "step": 274, + "time_per_iteration": 2.737847328186035 + }, + { + "auxiliary_loss_clip": 0.01350969, + "auxiliary_loss_mlp": 0.01062388, + "balance_loss_clip": 1.10146308, + "balance_loss_mlp": 1.04075122, + "epoch": 0.033066794925750015, + "flos": 15121984101120.0, + "grad_norm": 2.1230825275419534, + "language_loss": 0.8979792, + "learning_rate": 3.99990518874484e-06, + "loss": 0.92211282, + "num_input_tokens_seen": 5713460, + "step": 275, + "time_per_iteration": 2.7965989112854004 + }, + { + "auxiliary_loss_clip": 0.01352515, + "auxiliary_loss_mlp": 0.01071809, + "balance_loss_clip": 1.10116649, + "balance_loss_mlp": 1.05037498, + "epoch": 0.033187037816389105, + "flos": 22776593973120.0, + "grad_norm": 2.3046502287572608, + "language_loss": 0.92382687, + "learning_rate": 3.999897452212534e-06, + "loss": 0.94807011, + "num_input_tokens_seen": 5730790, + "step": 276, + "time_per_iteration": 2.920086145401001 + }, + { + "auxiliary_loss_clip": 0.01353269, + "auxiliary_loss_mlp": 0.01063471, + "balance_loss_clip": 1.10138369, + "balance_loss_mlp": 1.0419184, + "epoch": 0.033307280707028195, + "flos": 23331414424320.0, + "grad_norm": 2.50999356332136, + "language_loss": 1.00022149, + "learning_rate": 3.999889412297374e-06, + "loss": 1.02438879, + "num_input_tokens_seen": 5750215, + "step": 277, + "time_per_iteration": 3.7583773136138916 + }, + { + "auxiliary_loss_clip": 0.01352401, + "auxiliary_loss_mlp": 0.01068337, + "balance_loss_clip": 1.09894395, + "balance_loss_mlp": 1.047225, + "epoch": 0.03342752359766729, + "flos": 28840290566400.0, + "grad_norm": 44.15409377154289, + "language_loss": 0.78812158, + "learning_rate": 3.999881069000581e-06, + "loss": 0.81232899, + "num_input_tokens_seen": 5769945, + "step": 278, + "time_per_iteration": 4.643393278121948 + }, + { + "auxiliary_loss_clip": 0.01346107, + "auxiliary_loss_mlp": 0.01058878, + "balance_loss_clip": 1.09750366, + "balance_loss_mlp": 1.03666937, + "epoch": 0.03354776648830638, + "flos": 19384544090880.0, + "grad_norm": 2.676088739838586, + "language_loss": 0.8719945, + "learning_rate": 3.99987242232342e-06, + "loss": 0.89604437, + "num_input_tokens_seen": 5784950, + "step": 279, + "time_per_iteration": 2.8238868713378906 + }, + { + "auxiliary_loss_clip": 0.01350578, + "auxiliary_loss_mlp": 0.01057786, + "balance_loss_clip": 1.10150433, + "balance_loss_mlp": 1.03629303, + "epoch": 0.03366800937894547, + "flos": 17858628472320.0, + "grad_norm": 2.9187366471637506, + "language_loss": 0.79674542, + "learning_rate": 3.9998634722672026e-06, + "loss": 0.82082903, + "num_input_tokens_seen": 5805005, + "step": 280, + "time_per_iteration": 3.790874719619751 + }, + { + "auxiliary_loss_clip": 0.01355515, + "auxiliary_loss_mlp": 0.01070109, + "balance_loss_clip": 1.103338, + "balance_loss_mlp": 1.04867482, + "epoch": 0.03378825226958456, + "flos": 35951033635200.0, + "grad_norm": 2.2655683569774245, + "language_loss": 0.78687632, + "learning_rate": 3.999854218833286e-06, + "loss": 0.81113255, + "num_input_tokens_seen": 5825825, + "step": 281, + "time_per_iteration": 2.8354454040527344 + }, + { + "auxiliary_loss_clip": 0.01346721, + "auxiliary_loss_mlp": 0.01058668, + "balance_loss_clip": 1.10018206, + "balance_loss_mlp": 1.03773475, + "epoch": 0.03390849516022365, + "flos": 25702488126720.0, + "grad_norm": 3.466014287107408, + "language_loss": 0.82139361, + "learning_rate": 3.999844662023075e-06, + "loss": 0.84544754, + "num_input_tokens_seen": 5845700, + "step": 282, + "time_per_iteration": 2.7849478721618652 + }, + { + "auxiliary_loss_clip": 0.0134231, + "auxiliary_loss_mlp": 0.01062742, + "balance_loss_clip": 1.09727812, + "balance_loss_mlp": 1.0422262, + "epoch": 0.03402873805086274, + "flos": 21284505987840.0, + "grad_norm": 1.9166849315144951, + "language_loss": 0.92316222, + "learning_rate": 3.999834801838018e-06, + "loss": 0.94721276, + "num_input_tokens_seen": 5864680, + "step": 283, + "time_per_iteration": 2.8189237117767334 + }, + { + "auxiliary_loss_clip": 0.01344743, + "auxiliary_loss_mlp": 0.01055826, + "balance_loss_clip": 1.10003591, + "balance_loss_mlp": 1.03552485, + "epoch": 0.03414898094150183, + "flos": 22710913954560.0, + "grad_norm": 4.750469188227962, + "language_loss": 0.74421084, + "learning_rate": 3.9998246382796115e-06, + "loss": 0.76821649, + "num_input_tokens_seen": 5884260, + "step": 284, + "time_per_iteration": 2.8487789630889893 + }, + { + "auxiliary_loss_clip": 0.01349618, + "auxiliary_loss_mlp": 0.0106768, + "balance_loss_clip": 1.09809089, + "balance_loss_mlp": 1.04542315, + "epoch": 0.03426922383214093, + "flos": 18879927874560.0, + "grad_norm": 2.4363476021332535, + "language_loss": 0.90830904, + "learning_rate": 3.999814171349399e-06, + "loss": 0.932482, + "num_input_tokens_seen": 5902120, + "step": 285, + "time_per_iteration": 2.8151450157165527 + }, + { + "auxiliary_loss_clip": 0.01342044, + "auxiliary_loss_mlp": 0.01059827, + "balance_loss_clip": 1.09790385, + "balance_loss_mlp": 1.03783274, + "epoch": 0.03438946672278002, + "flos": 34752012716160.0, + "grad_norm": 1.7950634696076457, + "language_loss": 0.73620963, + "learning_rate": 3.9998034010489655e-06, + "loss": 0.76022834, + "num_input_tokens_seen": 5925810, + "step": 286, + "time_per_iteration": 2.83011531829834 + }, + { + "auxiliary_loss_clip": 0.01345515, + "auxiliary_loss_mlp": 0.01066668, + "balance_loss_clip": 1.09934604, + "balance_loss_mlp": 1.04610443, + "epoch": 0.03450970961341911, + "flos": 22164102236160.0, + "grad_norm": 2.2881893725637554, + "language_loss": 0.75822443, + "learning_rate": 3.999792327379946e-06, + "loss": 0.78234625, + "num_input_tokens_seen": 5945185, + "step": 287, + "time_per_iteration": 2.675440549850464 + }, + { + "auxiliary_loss_clip": 0.0134693, + "auxiliary_loss_mlp": 0.01062334, + "balance_loss_clip": 1.10310769, + "balance_loss_mlp": 1.04192495, + "epoch": 0.034629952504058197, + "flos": 21725740656000.0, + "grad_norm": 2.3067376373421578, + "language_loss": 0.96573311, + "learning_rate": 3.999780950344021e-06, + "loss": 0.98982584, + "num_input_tokens_seen": 5963375, + "step": 288, + "time_per_iteration": 2.9048638343811035 + }, + { + "auxiliary_loss_clip": 0.01351635, + "auxiliary_loss_mlp": 0.01059672, + "balance_loss_clip": 1.10145819, + "balance_loss_mlp": 1.03743923, + "epoch": 0.034750195394697286, + "flos": 20047994248320.0, + "grad_norm": 1.8800881008425148, + "language_loss": 0.82360733, + "learning_rate": 3.999769269942916e-06, + "loss": 0.84772038, + "num_input_tokens_seen": 5983415, + "step": 289, + "time_per_iteration": 2.7303645610809326 + }, + { + "auxiliary_loss_clip": 0.01339837, + "auxiliary_loss_mlp": 0.01061767, + "balance_loss_clip": 1.09628916, + "balance_loss_mlp": 1.04172802, + "epoch": 0.034870438285336376, + "flos": 27965865876480.0, + "grad_norm": 2.169201784807765, + "language_loss": 0.81038392, + "learning_rate": 3.999757286178402e-06, + "loss": 0.83439994, + "num_input_tokens_seen": 6005850, + "step": 290, + "time_per_iteration": 2.780451536178589 + }, + { + "auxiliary_loss_clip": 0.01345846, + "auxiliary_loss_mlp": 0.01064081, + "balance_loss_clip": 1.098037, + "balance_loss_mlp": 1.04201579, + "epoch": 0.03499068117597547, + "flos": 22017514832640.0, + "grad_norm": 3.0898367383748444, + "language_loss": 0.90567905, + "learning_rate": 3.999744999052299e-06, + "loss": 0.92977834, + "num_input_tokens_seen": 6027240, + "step": 291, + "time_per_iteration": 2.802639961242676 + }, + { + "auxiliary_loss_clip": 0.01268995, + "auxiliary_loss_mlp": 0.010125, + "balance_loss_clip": 1.11223173, + "balance_loss_mlp": 0.99981576, + "epoch": 0.03511092406661456, + "flos": 57242147725440.0, + "grad_norm": 0.9538287437743445, + "language_loss": 0.61180317, + "learning_rate": 3.9997324085664675e-06, + "loss": 0.63461816, + "num_input_tokens_seen": 6087470, + "step": 292, + "time_per_iteration": 3.2386155128479004 + }, + { + "auxiliary_loss_clip": 0.01337894, + "auxiliary_loss_mlp": 0.01052311, + "balance_loss_clip": 1.0929755, + "balance_loss_mlp": 1.0294826, + "epoch": 0.03523116695725365, + "flos": 22928065626240.0, + "grad_norm": 2.5602264612892562, + "language_loss": 0.92081583, + "learning_rate": 3.999719514722821e-06, + "loss": 0.94471788, + "num_input_tokens_seen": 6107600, + "step": 293, + "time_per_iteration": 2.7446491718292236 + }, + { + "auxiliary_loss_clip": 0.01339637, + "auxiliary_loss_mlp": 0.01061349, + "balance_loss_clip": 1.09533906, + "balance_loss_mlp": 1.04090452, + "epoch": 0.03535140984789274, + "flos": 36903241226880.0, + "grad_norm": 2.51081170408304, + "language_loss": 0.74524993, + "learning_rate": 3.999706317523314e-06, + "loss": 0.76925981, + "num_input_tokens_seen": 6126160, + "step": 294, + "time_per_iteration": 2.8639438152313232 + }, + { + "auxiliary_loss_clip": 0.01339007, + "auxiliary_loss_mlp": 0.01063167, + "balance_loss_clip": 1.09906912, + "balance_loss_mlp": 1.04291284, + "epoch": 0.03547165273853183, + "flos": 20449152316800.0, + "grad_norm": 1.8747296553356019, + "language_loss": 0.86110842, + "learning_rate": 3.999692816969948e-06, + "loss": 0.88513017, + "num_input_tokens_seen": 6145695, + "step": 295, + "time_per_iteration": 2.79636812210083 + }, + { + "auxiliary_loss_clip": 0.01261845, + "auxiliary_loss_mlp": 0.01012227, + "balance_loss_clip": 1.10610318, + "balance_loss_mlp": 0.99973345, + "epoch": 0.03559189562917092, + "flos": 69850564871040.0, + "grad_norm": 1.0060920907897017, + "language_loss": 0.69361067, + "learning_rate": 3.999679013064772e-06, + "loss": 0.71635127, + "num_input_tokens_seen": 6212440, + "step": 296, + "time_per_iteration": 3.302973508834839 + }, + { + "auxiliary_loss_clip": 0.01336203, + "auxiliary_loss_mlp": 0.01063967, + "balance_loss_clip": 1.09318137, + "balance_loss_mlp": 1.0430336, + "epoch": 0.03571213851981002, + "flos": 21651944163840.0, + "grad_norm": 3.387804507837493, + "language_loss": 0.85609639, + "learning_rate": 3.99966490580988e-06, + "loss": 0.8800981, + "num_input_tokens_seen": 6229800, + "step": 297, + "time_per_iteration": 2.7328696250915527 + }, + { + "auxiliary_loss_clip": 0.0134436, + "auxiliary_loss_mlp": 0.01060315, + "balance_loss_clip": 1.09919453, + "balance_loss_mlp": 1.03882194, + "epoch": 0.03583238141044911, + "flos": 43945610757120.0, + "grad_norm": 2.593126620307902, + "language_loss": 0.66243714, + "learning_rate": 3.999650495207411e-06, + "loss": 0.68648386, + "num_input_tokens_seen": 6255825, + "step": 298, + "time_per_iteration": 2.9789772033691406 + }, + { + "auxiliary_loss_clip": 0.01340717, + "auxiliary_loss_mlp": 0.01054613, + "balance_loss_clip": 1.09691286, + "balance_loss_mlp": 1.03431165, + "epoch": 0.0359526243010882, + "flos": 18910810592640.0, + "grad_norm": 2.8375207048023943, + "language_loss": 0.90654325, + "learning_rate": 3.999635781259553e-06, + "loss": 0.93049657, + "num_input_tokens_seen": 6271090, + "step": 299, + "time_per_iteration": 2.6831860542297363 + }, + { + "auxiliary_loss_clip": 0.0125058, + "auxiliary_loss_mlp": 0.01019568, + "balance_loss_clip": 1.09735608, + "balance_loss_mlp": 1.00717032, + "epoch": 0.03607286719172729, + "flos": 61668892782720.0, + "grad_norm": 0.9207771400289333, + "language_loss": 0.52288783, + "learning_rate": 3.999620763968535e-06, + "loss": 0.54558933, + "num_input_tokens_seen": 6329965, + "step": 300, + "time_per_iteration": 3.1241395473480225 + }, + { + "auxiliary_loss_clip": 0.01338648, + "auxiliary_loss_mlp": 0.01065927, + "balance_loss_clip": 1.09709454, + "balance_loss_mlp": 1.04615021, + "epoch": 0.03619311008236638, + "flos": 27819062991360.0, + "grad_norm": 1.6369096569096417, + "language_loss": 0.86697447, + "learning_rate": 3.999605443336638e-06, + "loss": 0.89102024, + "num_input_tokens_seen": 6352095, + "step": 301, + "time_per_iteration": 2.8078360557556152 + }, + { + "auxiliary_loss_clip": 0.01341646, + "auxiliary_loss_mlp": 0.01062842, + "balance_loss_clip": 1.09611869, + "balance_loss_mlp": 1.04131317, + "epoch": 0.03631335297300547, + "flos": 13621133197440.0, + "grad_norm": 2.3321798910634373, + "language_loss": 0.8923406, + "learning_rate": 3.999589819366185e-06, + "loss": 0.91638553, + "num_input_tokens_seen": 6365885, + "step": 302, + "time_per_iteration": 2.697671413421631 + }, + { + "auxiliary_loss_clip": 0.01339411, + "auxiliary_loss_mlp": 0.01056107, + "balance_loss_clip": 1.09641647, + "balance_loss_mlp": 1.0346489, + "epoch": 0.036433595863644565, + "flos": 27631788456960.0, + "grad_norm": 2.3903967917118782, + "language_loss": 0.84731823, + "learning_rate": 3.999573892059547e-06, + "loss": 0.8712734, + "num_input_tokens_seen": 6385015, + "step": 303, + "time_per_iteration": 2.8571083545684814 + }, + { + "auxiliary_loss_clip": 0.01344895, + "auxiliary_loss_mlp": 0.01064787, + "balance_loss_clip": 1.09551954, + "balance_loss_mlp": 1.0441041, + "epoch": 0.036553838754283655, + "flos": 24572020314240.0, + "grad_norm": 2.8856185133522825, + "language_loss": 0.81130373, + "learning_rate": 3.999557661419138e-06, + "loss": 0.83540052, + "num_input_tokens_seen": 6405165, + "step": 304, + "time_per_iteration": 3.6647679805755615 + }, + { + "auxiliary_loss_clip": 0.01341946, + "auxiliary_loss_mlp": 0.01059095, + "balance_loss_clip": 1.09841311, + "balance_loss_mlp": 1.03764939, + "epoch": 0.036674081644922744, + "flos": 23404313076480.0, + "grad_norm": 2.0510085212225224, + "language_loss": 0.81544805, + "learning_rate": 3.9995411274474225e-06, + "loss": 0.83945847, + "num_input_tokens_seen": 6424445, + "step": 305, + "time_per_iteration": 3.7488181591033936 + }, + { + "auxiliary_loss_clip": 0.01341791, + "auxiliary_loss_mlp": 0.01061679, + "balance_loss_clip": 1.09809446, + "balance_loss_mlp": 1.03907704, + "epoch": 0.036794324535561834, + "flos": 27489690253440.0, + "grad_norm": 2.587423276590776, + "language_loss": 0.81447887, + "learning_rate": 3.999524290146908e-06, + "loss": 0.83851361, + "num_input_tokens_seen": 6444650, + "step": 306, + "time_per_iteration": 3.7994515895843506 + }, + { + "auxiliary_loss_clip": 0.01342793, + "auxiliary_loss_mlp": 0.01056829, + "balance_loss_clip": 1.10041142, + "balance_loss_mlp": 1.03721952, + "epoch": 0.036914567426200924, + "flos": 19463476227840.0, + "grad_norm": 2.682723661773303, + "language_loss": 0.92377293, + "learning_rate": 3.9995071495201485e-06, + "loss": 0.94776917, + "num_input_tokens_seen": 6461755, + "step": 307, + "time_per_iteration": 2.6925790309906006 + }, + { + "auxiliary_loss_clip": 0.01340237, + "auxiliary_loss_mlp": 0.01062457, + "balance_loss_clip": 1.09855449, + "balance_loss_mlp": 1.04175019, + "epoch": 0.037034810316840014, + "flos": 22309324922880.0, + "grad_norm": 8.74816199502439, + "language_loss": 0.98135746, + "learning_rate": 3.999489705569744e-06, + "loss": 1.00538445, + "num_input_tokens_seen": 6479455, + "step": 308, + "time_per_iteration": 2.7272562980651855 + }, + { + "auxiliary_loss_clip": 0.01337719, + "auxiliary_loss_mlp": 0.01058766, + "balance_loss_clip": 1.09583879, + "balance_loss_mlp": 1.03798807, + "epoch": 0.03715505320747911, + "flos": 18588333265920.0, + "grad_norm": 3.994759884077687, + "language_loss": 0.86647117, + "learning_rate": 3.999471958298341e-06, + "loss": 0.89043605, + "num_input_tokens_seen": 6498365, + "step": 309, + "time_per_iteration": 2.7114155292510986 + }, + { + "auxiliary_loss_clip": 0.01343318, + "auxiliary_loss_mlp": 0.01057618, + "balance_loss_clip": 1.10025835, + "balance_loss_mlp": 1.03676796, + "epoch": 0.0372752960981182, + "flos": 35955343267200.0, + "grad_norm": 1.94013931885305, + "language_loss": 0.76264942, + "learning_rate": 3.999453907708631e-06, + "loss": 0.78665882, + "num_input_tokens_seen": 6520770, + "step": 310, + "time_per_iteration": 2.8743767738342285 + }, + { + "auxiliary_loss_clip": 0.01336513, + "auxiliary_loss_mlp": 0.01063271, + "balance_loss_clip": 1.09637427, + "balance_loss_mlp": 1.04351807, + "epoch": 0.03739553898875729, + "flos": 20814040627200.0, + "grad_norm": 1.7772673528643932, + "language_loss": 0.81453449, + "learning_rate": 3.999435553803353e-06, + "loss": 0.83853239, + "num_input_tokens_seen": 6540170, + "step": 311, + "time_per_iteration": 2.7649006843566895 + }, + { + "auxiliary_loss_clip": 0.01332229, + "auxiliary_loss_mlp": 0.0106504, + "balance_loss_clip": 1.09341693, + "balance_loss_mlp": 1.04416668, + "epoch": 0.03751578187939638, + "flos": 20264140339200.0, + "grad_norm": 2.2740017649130566, + "language_loss": 0.84145308, + "learning_rate": 3.999416896585292e-06, + "loss": 0.86542583, + "num_input_tokens_seen": 6557200, + "step": 312, + "time_per_iteration": 2.7643895149230957 + }, + { + "auxiliary_loss_clip": 0.01331097, + "auxiliary_loss_mlp": 0.0105449, + "balance_loss_clip": 1.093274, + "balance_loss_mlp": 1.03368759, + "epoch": 0.03763602477003547, + "flos": 20668063754880.0, + "grad_norm": 2.7108083437267387, + "language_loss": 0.85451448, + "learning_rate": 3.9993979360572775e-06, + "loss": 0.8783704, + "num_input_tokens_seen": 6577340, + "step": 313, + "time_per_iteration": 2.815175771713257 + }, + { + "auxiliary_loss_clip": 0.01342418, + "auxiliary_loss_mlp": 0.01063243, + "balance_loss_clip": 1.09863281, + "balance_loss_mlp": 1.04245269, + "epoch": 0.03775626766067456, + "flos": 16691352197760.0, + "grad_norm": 2.788831865973569, + "language_loss": 0.82955551, + "learning_rate": 3.999378672222185e-06, + "loss": 0.85361212, + "num_input_tokens_seen": 6595125, + "step": 314, + "time_per_iteration": 2.7011399269104004 + }, + { + "auxiliary_loss_clip": 0.01331199, + "auxiliary_loss_mlp": 0.01070285, + "balance_loss_clip": 1.0949223, + "balance_loss_mlp": 1.04945922, + "epoch": 0.03787651055131366, + "flos": 21141797253120.0, + "grad_norm": 1.9608973613652931, + "language_loss": 0.8283484, + "learning_rate": 3.9993591050829385e-06, + "loss": 0.85236323, + "num_input_tokens_seen": 6612990, + "step": 315, + "time_per_iteration": 2.7512335777282715 + }, + { + "auxiliary_loss_clip": 0.01336505, + "auxiliary_loss_mlp": 0.01065175, + "balance_loss_clip": 1.09562016, + "balance_loss_mlp": 1.04417026, + "epoch": 0.037996753441952746, + "flos": 22018089450240.0, + "grad_norm": 1.9155564047750737, + "language_loss": 0.79447627, + "learning_rate": 3.999339234642506e-06, + "loss": 0.81849301, + "num_input_tokens_seen": 6632740, + "step": 316, + "time_per_iteration": 2.698234796524048 + }, + { + "auxiliary_loss_clip": 0.013365, + "auxiliary_loss_mlp": 0.01068079, + "balance_loss_clip": 1.09527993, + "balance_loss_mlp": 1.04730058, + "epoch": 0.038116996332591836, + "flos": 27709391790720.0, + "grad_norm": 1.9584106063370974, + "language_loss": 0.83969688, + "learning_rate": 3.9993190609038994e-06, + "loss": 0.86374271, + "num_input_tokens_seen": 6651505, + "step": 317, + "time_per_iteration": 2.7509067058563232 + }, + { + "auxiliary_loss_clip": 0.0133239, + "auxiliary_loss_mlp": 0.0105554, + "balance_loss_clip": 1.09318781, + "balance_loss_mlp": 1.03337944, + "epoch": 0.038237239223230926, + "flos": 21178067011200.0, + "grad_norm": 1.9533444471709895, + "language_loss": 0.83258933, + "learning_rate": 3.999298583870182e-06, + "loss": 0.85646868, + "num_input_tokens_seen": 6671090, + "step": 318, + "time_per_iteration": 2.775290012359619 + }, + { + "auxiliary_loss_clip": 0.01331441, + "auxiliary_loss_mlp": 0.01067162, + "balance_loss_clip": 1.09299111, + "balance_loss_mlp": 1.04620516, + "epoch": 0.038357482113870016, + "flos": 25556618995200.0, + "grad_norm": 2.0921848694948646, + "language_loss": 0.77528846, + "learning_rate": 3.999277803544458e-06, + "loss": 0.79927444, + "num_input_tokens_seen": 6691245, + "step": 319, + "time_per_iteration": 2.7082948684692383 + }, + { + "auxiliary_loss_clip": 0.01240904, + "auxiliary_loss_mlp": 0.01012352, + "balance_loss_clip": 1.09043193, + "balance_loss_mlp": 1.00090837, + "epoch": 0.038477725004509106, + "flos": 59227578034560.0, + "grad_norm": 0.9608088321285161, + "language_loss": 0.62389374, + "learning_rate": 3.999256719929882e-06, + "loss": 0.64642632, + "num_input_tokens_seen": 6752520, + "step": 320, + "time_per_iteration": 3.2306923866271973 + }, + { + "auxiliary_loss_clip": 0.01240846, + "auxiliary_loss_mlp": 0.01010941, + "balance_loss_clip": 1.09066677, + "balance_loss_mlp": 0.99954492, + "epoch": 0.0385979678951482, + "flos": 67317676398720.0, + "grad_norm": 1.2119261135356116, + "language_loss": 0.67056632, + "learning_rate": 3.999235333029651e-06, + "loss": 0.69308412, + "num_input_tokens_seen": 6806460, + "step": 321, + "time_per_iteration": 3.1336050033569336 + }, + { + "auxiliary_loss_clip": 0.01327656, + "auxiliary_loss_mlp": 0.01066119, + "balance_loss_clip": 1.09436846, + "balance_loss_mlp": 1.04642558, + "epoch": 0.03871821078578729, + "flos": 22746752749440.0, + "grad_norm": 2.2372292406085865, + "language_loss": 0.82257974, + "learning_rate": 3.999213642847009e-06, + "loss": 0.84651744, + "num_input_tokens_seen": 6827045, + "step": 322, + "time_per_iteration": 2.810230016708374 + }, + { + "auxiliary_loss_clip": 0.01325154, + "auxiliary_loss_mlp": 0.01066042, + "balance_loss_clip": 1.09141994, + "balance_loss_mlp": 1.04625368, + "epoch": 0.03883845367642638, + "flos": 26280613526400.0, + "grad_norm": 2.0282585769189403, + "language_loss": 0.91337591, + "learning_rate": 3.999191649385247e-06, + "loss": 0.93728787, + "num_input_tokens_seen": 6848220, + "step": 323, + "time_per_iteration": 2.794156312942505 + }, + { + "auxiliary_loss_clip": 0.01229181, + "auxiliary_loss_mlp": 0.01015347, + "balance_loss_clip": 1.08069754, + "balance_loss_mlp": 1.00428426, + "epoch": 0.03895869656706547, + "flos": 56962835568000.0, + "grad_norm": 0.9019572665667719, + "language_loss": 0.59769344, + "learning_rate": 3.999169352647702e-06, + "loss": 0.6201387, + "num_input_tokens_seen": 6909400, + "step": 324, + "time_per_iteration": 3.2429563999176025 + }, + { + "auxiliary_loss_clip": 0.01334624, + "auxiliary_loss_mlp": 0.01059986, + "balance_loss_clip": 1.09367776, + "balance_loss_mlp": 1.03898156, + "epoch": 0.03907893945770456, + "flos": 24863363527680.0, + "grad_norm": 2.026178464747914, + "language_loss": 0.82945055, + "learning_rate": 3.999146752637755e-06, + "loss": 0.85339665, + "num_input_tokens_seen": 6930445, + "step": 325, + "time_per_iteration": 2.823441505432129 + }, + { + "auxiliary_loss_clip": 0.0133106, + "auxiliary_loss_mlp": 0.01056297, + "balance_loss_clip": 1.0933162, + "balance_loss_mlp": 1.03507769, + "epoch": 0.03919918234834365, + "flos": 18368595815040.0, + "grad_norm": 4.456714474131061, + "language_loss": 0.89771736, + "learning_rate": 3.999123849358836e-06, + "loss": 0.92159098, + "num_input_tokens_seen": 6948110, + "step": 326, + "time_per_iteration": 2.7871432304382324 + }, + { + "auxiliary_loss_clip": 0.01326971, + "auxiliary_loss_mlp": 0.01065773, + "balance_loss_clip": 1.09166551, + "balance_loss_mlp": 1.04545939, + "epoch": 0.03931942523898275, + "flos": 25225414663680.0, + "grad_norm": 1.9167841828160812, + "language_loss": 0.75089043, + "learning_rate": 3.999100642814418e-06, + "loss": 0.77481782, + "num_input_tokens_seen": 6968550, + "step": 327, + "time_per_iteration": 2.867877721786499 + }, + { + "auxiliary_loss_clip": 0.01326779, + "auxiliary_loss_mlp": 0.01064169, + "balance_loss_clip": 1.09379983, + "balance_loss_mlp": 1.0447619, + "epoch": 0.03943966812962184, + "flos": 23257905240960.0, + "grad_norm": 3.8668931410166683, + "language_loss": 0.88479519, + "learning_rate": 3.999077133008022e-06, + "loss": 0.9087047, + "num_input_tokens_seen": 6987135, + "step": 328, + "time_per_iteration": 2.720881223678589 + }, + { + "auxiliary_loss_clip": 0.01330713, + "auxiliary_loss_mlp": 0.01061352, + "balance_loss_clip": 1.09319615, + "balance_loss_mlp": 1.04146743, + "epoch": 0.03955991102026093, + "flos": 29168837291520.0, + "grad_norm": 1.829534677546803, + "language_loss": 0.90728623, + "learning_rate": 3.9990533199432145e-06, + "loss": 0.93120694, + "num_input_tokens_seen": 7008630, + "step": 329, + "time_per_iteration": 2.7622830867767334 + }, + { + "auxiliary_loss_clip": 0.01336001, + "auxiliary_loss_mlp": 0.01059937, + "balance_loss_clip": 1.09704888, + "balance_loss_mlp": 1.03875303, + "epoch": 0.03968015391090002, + "flos": 17602441695360.0, + "grad_norm": 2.4083316307353404, + "language_loss": 0.7600404, + "learning_rate": 3.999029203623608e-06, + "loss": 0.7839998, + "num_input_tokens_seen": 7026350, + "step": 330, + "time_per_iteration": 2.8470025062561035 + }, + { + "auxiliary_loss_clip": 0.01320943, + "auxiliary_loss_mlp": 0.01050093, + "balance_loss_clip": 1.08932614, + "balance_loss_mlp": 1.029971, + "epoch": 0.03980039680153911, + "flos": 21799285752960.0, + "grad_norm": 2.476393974769263, + "language_loss": 0.87024945, + "learning_rate": 3.99900478405286e-06, + "loss": 0.89395976, + "num_input_tokens_seen": 7045660, + "step": 331, + "time_per_iteration": 3.795971155166626 + }, + { + "auxiliary_loss_clip": 0.01327577, + "auxiliary_loss_mlp": 0.01067845, + "balance_loss_clip": 1.09393394, + "balance_loss_mlp": 1.04884267, + "epoch": 0.0399206396921782, + "flos": 15195134148480.0, + "grad_norm": 2.233339809433114, + "language_loss": 0.82229209, + "learning_rate": 3.998980061234676e-06, + "loss": 0.8462463, + "num_input_tokens_seen": 7063575, + "step": 332, + "time_per_iteration": 4.679368019104004 + }, + { + "auxiliary_loss_clip": 0.0132495, + "auxiliary_loss_mlp": 0.01060349, + "balance_loss_clip": 1.09025371, + "balance_loss_mlp": 1.03918898, + "epoch": 0.040040882582817294, + "flos": 14422910630400.0, + "grad_norm": 2.82107379732585, + "language_loss": 0.76011217, + "learning_rate": 3.9989550351728055e-06, + "loss": 0.78396511, + "num_input_tokens_seen": 7080505, + "step": 333, + "time_per_iteration": 2.8642921447753906 + }, + { + "auxiliary_loss_clip": 0.01323691, + "auxiliary_loss_mlp": 0.01055506, + "balance_loss_clip": 1.09155393, + "balance_loss_mlp": 1.03512096, + "epoch": 0.040161125473456384, + "flos": 19280906375040.0, + "grad_norm": 3.4080576291943423, + "language_loss": 0.84294111, + "learning_rate": 3.998929705871046e-06, + "loss": 0.86673307, + "num_input_tokens_seen": 7097860, + "step": 334, + "time_per_iteration": 2.767526149749756 + }, + { + "auxiliary_loss_clip": 0.0132746, + "auxiliary_loss_mlp": 0.01058172, + "balance_loss_clip": 1.09351945, + "balance_loss_mlp": 1.03740621, + "epoch": 0.040281368364095474, + "flos": 17821101738240.0, + "grad_norm": 2.7504702783398867, + "language_loss": 0.88986802, + "learning_rate": 3.99890407333324e-06, + "loss": 0.9137243, + "num_input_tokens_seen": 7116390, + "step": 335, + "time_per_iteration": 2.7646539211273193 + }, + { + "auxiliary_loss_clip": 0.01322298, + "auxiliary_loss_mlp": 0.01057466, + "balance_loss_clip": 1.08955348, + "balance_loss_mlp": 1.03746247, + "epoch": 0.040401611254734564, + "flos": 19573757959680.0, + "grad_norm": 1.7182947652484732, + "language_loss": 0.87321675, + "learning_rate": 3.998878137563275e-06, + "loss": 0.89701438, + "num_input_tokens_seen": 7135940, + "step": 336, + "time_per_iteration": 2.7144289016723633 + }, + { + "auxiliary_loss_clip": 0.01325181, + "auxiliary_loss_mlp": 0.01068272, + "balance_loss_clip": 1.09292603, + "balance_loss_mlp": 1.04863858, + "epoch": 0.040521854145373654, + "flos": 22054466949120.0, + "grad_norm": 2.8120060952967134, + "language_loss": 0.8535682, + "learning_rate": 3.998851898565085e-06, + "loss": 0.87750274, + "num_input_tokens_seen": 7155745, + "step": 337, + "time_per_iteration": 2.9164724349975586 + }, + { + "auxiliary_loss_clip": 0.01323484, + "auxiliary_loss_mlp": 0.01056753, + "balance_loss_clip": 1.08962083, + "balance_loss_mlp": 1.03633296, + "epoch": 0.04064209703601274, + "flos": 22674644196480.0, + "grad_norm": 2.078411522177721, + "language_loss": 0.83167839, + "learning_rate": 3.998825356342653e-06, + "loss": 0.85548085, + "num_input_tokens_seen": 7175920, + "step": 338, + "time_per_iteration": 2.9187772274017334 + }, + { + "auxiliary_loss_clip": 0.01323806, + "auxiliary_loss_mlp": 0.01050897, + "balance_loss_clip": 1.0905807, + "balance_loss_mlp": 1.03034532, + "epoch": 0.04076233992665183, + "flos": 38582172783360.0, + "grad_norm": 2.5483911720180545, + "language_loss": 0.73124033, + "learning_rate": 3.998798510900003e-06, + "loss": 0.75498736, + "num_input_tokens_seen": 7198720, + "step": 339, + "time_per_iteration": 2.9891746044158936 + }, + { + "auxiliary_loss_clip": 0.01319985, + "auxiliary_loss_mlp": 0.0104956, + "balance_loss_clip": 1.08706605, + "balance_loss_mlp": 1.02912784, + "epoch": 0.04088258281729093, + "flos": 25885309374720.0, + "grad_norm": 3.3258911496282666, + "language_loss": 0.83896065, + "learning_rate": 3.998771362241207e-06, + "loss": 0.86265612, + "num_input_tokens_seen": 7219125, + "step": 340, + "time_per_iteration": 2.8364968299865723 + }, + { + "auxiliary_loss_clip": 0.01324159, + "auxiliary_loss_mlp": 0.01058316, + "balance_loss_clip": 1.09254301, + "balance_loss_mlp": 1.03751445, + "epoch": 0.04100282570793002, + "flos": 19789832223360.0, + "grad_norm": 2.9096028615816554, + "language_loss": 0.8800571, + "learning_rate": 3.998743910370385e-06, + "loss": 0.90388191, + "num_input_tokens_seen": 7237985, + "step": 341, + "time_per_iteration": 2.7964837551116943 + }, + { + "auxiliary_loss_clip": 0.01328074, + "auxiliary_loss_mlp": 0.01059768, + "balance_loss_clip": 1.09731317, + "balance_loss_mlp": 1.03963375, + "epoch": 0.04112306859856911, + "flos": 22565152563840.0, + "grad_norm": 2.261374425059689, + "language_loss": 0.73533535, + "learning_rate": 3.998716155291702e-06, + "loss": 0.75921381, + "num_input_tokens_seen": 7255825, + "step": 342, + "time_per_iteration": 2.8233630657196045 + }, + { + "auxiliary_loss_clip": 0.01325174, + "auxiliary_loss_mlp": 0.01058144, + "balance_loss_clip": 1.09426701, + "balance_loss_mlp": 1.0385108, + "epoch": 0.0412433114892082, + "flos": 25040654081280.0, + "grad_norm": 1.6347382297869062, + "language_loss": 0.90437925, + "learning_rate": 3.998688097009366e-06, + "loss": 0.9282124, + "num_input_tokens_seen": 7276590, + "step": 343, + "time_per_iteration": 2.7669060230255127 + }, + { + "auxiliary_loss_clip": 0.01323767, + "auxiliary_loss_mlp": 0.01063056, + "balance_loss_clip": 1.09220481, + "balance_loss_mlp": 1.04238558, + "epoch": 0.04136355437984729, + "flos": 25191371548800.0, + "grad_norm": 2.496908019152404, + "language_loss": 0.79716194, + "learning_rate": 3.998659735527636e-06, + "loss": 0.82103014, + "num_input_tokens_seen": 7295680, + "step": 344, + "time_per_iteration": 2.7389817237854004 + }, + { + "auxiliary_loss_clip": 0.01321285, + "auxiliary_loss_mlp": 0.01062317, + "balance_loss_clip": 1.09170294, + "balance_loss_mlp": 1.04256439, + "epoch": 0.04148379727048638, + "flos": 22966777509120.0, + "grad_norm": 2.636333289279536, + "language_loss": 0.77996337, + "learning_rate": 3.998631070850813e-06, + "loss": 0.80379939, + "num_input_tokens_seen": 7316300, + "step": 345, + "time_per_iteration": 2.6926681995391846 + }, + { + "auxiliary_loss_clip": 0.01324455, + "auxiliary_loss_mlp": 0.01060854, + "balance_loss_clip": 1.09509516, + "balance_loss_mlp": 1.04191172, + "epoch": 0.041604040161125476, + "flos": 14063481187200.0, + "grad_norm": 2.159340594773399, + "language_loss": 0.83481467, + "learning_rate": 3.9986021029832455e-06, + "loss": 0.85866773, + "num_input_tokens_seen": 7333615, + "step": 346, + "time_per_iteration": 2.700479030609131 + }, + { + "auxiliary_loss_clip": 0.01321421, + "auxiliary_loss_mlp": 0.01061089, + "balance_loss_clip": 1.08937407, + "balance_loss_mlp": 1.04034626, + "epoch": 0.041724283051764566, + "flos": 12091877614080.0, + "grad_norm": 3.218930203015286, + "language_loss": 0.91607869, + "learning_rate": 3.9985728319293285e-06, + "loss": 0.9399038, + "num_input_tokens_seen": 7347590, + "step": 347, + "time_per_iteration": 2.655008554458618 + }, + { + "auxiliary_loss_clip": 0.01326347, + "auxiliary_loss_mlp": 0.01058984, + "balance_loss_clip": 1.09086752, + "balance_loss_mlp": 1.03821802, + "epoch": 0.041844525942403656, + "flos": 12385303816320.0, + "grad_norm": 2.87764849628161, + "language_loss": 0.85309178, + "learning_rate": 3.998543257693501e-06, + "loss": 0.87694502, + "num_input_tokens_seen": 7364345, + "step": 348, + "time_per_iteration": 2.654771566390991 + }, + { + "auxiliary_loss_clip": 0.01323548, + "auxiliary_loss_mlp": 0.0106689, + "balance_loss_clip": 1.09291303, + "balance_loss_mlp": 1.04723239, + "epoch": 0.041964768833042745, + "flos": 23769345041280.0, + "grad_norm": 1.79615047846503, + "language_loss": 0.87836206, + "learning_rate": 3.998513380280251e-06, + "loss": 0.90226638, + "num_input_tokens_seen": 7384625, + "step": 349, + "time_per_iteration": 2.849649429321289 + }, + { + "auxiliary_loss_clip": 0.01325145, + "auxiliary_loss_mlp": 0.01060356, + "balance_loss_clip": 1.09150147, + "balance_loss_mlp": 1.03973269, + "epoch": 0.042085011723681835, + "flos": 11875336473600.0, + "grad_norm": 2.4186546522097507, + "language_loss": 0.94984376, + "learning_rate": 3.99848319969411e-06, + "loss": 0.97369874, + "num_input_tokens_seen": 7402225, + "step": 350, + "time_per_iteration": 2.797288656234741 + }, + { + "auxiliary_loss_clip": 0.01328033, + "auxiliary_loss_mlp": 0.01053667, + "balance_loss_clip": 1.09545207, + "balance_loss_mlp": 1.03234053, + "epoch": 0.042205254614320925, + "flos": 16873957964160.0, + "grad_norm": 8.97585293603189, + "language_loss": 0.79114461, + "learning_rate": 3.9984527159396564e-06, + "loss": 0.81496161, + "num_input_tokens_seen": 7420865, + "step": 351, + "time_per_iteration": 2.953317642211914 + }, + { + "auxiliary_loss_clip": 0.01319539, + "auxiliary_loss_mlp": 0.01050859, + "balance_loss_clip": 1.08828664, + "balance_loss_mlp": 1.03037882, + "epoch": 0.04232549750496002, + "flos": 25118508810240.0, + "grad_norm": 2.1177011179409093, + "language_loss": 0.84701467, + "learning_rate": 3.9984219290215154e-06, + "loss": 0.8707186, + "num_input_tokens_seen": 7441040, + "step": 352, + "time_per_iteration": 2.7623138427734375 + }, + { + "auxiliary_loss_clip": 0.01320102, + "auxiliary_loss_mlp": 0.01052066, + "balance_loss_clip": 1.09357381, + "balance_loss_mlp": 1.03362489, + "epoch": 0.04244574039559911, + "flos": 26724541714560.0, + "grad_norm": 1.6346365151131195, + "language_loss": 0.89364076, + "learning_rate": 3.998390838944356e-06, + "loss": 0.91736245, + "num_input_tokens_seen": 7462545, + "step": 353, + "time_per_iteration": 2.8600618839263916 + }, + { + "auxiliary_loss_clip": 0.01321263, + "auxiliary_loss_mlp": 0.01064861, + "balance_loss_clip": 1.09015012, + "balance_loss_mlp": 1.04464364, + "epoch": 0.0425659832862382, + "flos": 20923244951040.0, + "grad_norm": 2.3100613768294576, + "language_loss": 0.90560657, + "learning_rate": 3.998359445712895e-06, + "loss": 0.9294678, + "num_input_tokens_seen": 7481650, + "step": 354, + "time_per_iteration": 2.692152500152588 + }, + { + "auxiliary_loss_clip": 0.0132012, + "auxiliary_loss_mlp": 0.01058332, + "balance_loss_clip": 1.08803225, + "balance_loss_mlp": 1.03782773, + "epoch": 0.04268622617687729, + "flos": 23331127115520.0, + "grad_norm": 2.2598593894045056, + "language_loss": 0.81416154, + "learning_rate": 3.9983277493318955e-06, + "loss": 0.83794606, + "num_input_tokens_seen": 7500945, + "step": 355, + "time_per_iteration": 2.7208034992218018 + }, + { + "auxiliary_loss_clip": 0.01320452, + "auxiliary_loss_mlp": 0.01054201, + "balance_loss_clip": 1.08954346, + "balance_loss_mlp": 1.03312492, + "epoch": 0.04280646906751638, + "flos": 25994010908160.0, + "grad_norm": 1.9611226456458128, + "language_loss": 0.81467915, + "learning_rate": 3.998295749806165e-06, + "loss": 0.83842576, + "num_input_tokens_seen": 7522170, + "step": 356, + "time_per_iteration": 2.8957018852233887 + }, + { + "auxiliary_loss_clip": 0.01318761, + "auxiliary_loss_mlp": 0.01054154, + "balance_loss_clip": 1.08995223, + "balance_loss_mlp": 1.03504467, + "epoch": 0.04292671195815547, + "flos": 26906824258560.0, + "grad_norm": 1.9825109957358247, + "language_loss": 0.83236092, + "learning_rate": 3.998263447140558e-06, + "loss": 0.85609013, + "num_input_tokens_seen": 7542370, + "step": 357, + "time_per_iteration": 3.6091461181640625 + }, + { + "auxiliary_loss_clip": 0.01314208, + "auxiliary_loss_mlp": 0.01059238, + "balance_loss_clip": 1.08674622, + "balance_loss_mlp": 1.0394609, + "epoch": 0.04304695484879457, + "flos": 39457315745280.0, + "grad_norm": 2.0158660819813505, + "language_loss": 0.8168242, + "learning_rate": 3.998230841339976e-06, + "loss": 0.84055865, + "num_input_tokens_seen": 7564380, + "step": 358, + "time_per_iteration": 5.579439163208008 + }, + { + "auxiliary_loss_clip": 0.01315995, + "auxiliary_loss_mlp": 0.01059063, + "balance_loss_clip": 1.09049869, + "balance_loss_mlp": 1.0395366, + "epoch": 0.04316719773943366, + "flos": 19646297475840.0, + "grad_norm": 2.1495753786687968, + "language_loss": 0.85191691, + "learning_rate": 3.998197932409363e-06, + "loss": 0.87566757, + "num_input_tokens_seen": 7582390, + "step": 359, + "time_per_iteration": 2.7912285327911377 + }, + { + "auxiliary_loss_clip": 0.01312485, + "auxiliary_loss_mlp": 0.0105666, + "balance_loss_clip": 1.08718061, + "balance_loss_mlp": 1.03716934, + "epoch": 0.04328744063007275, + "flos": 22452320966400.0, + "grad_norm": 4.443122611811507, + "language_loss": 0.86320841, + "learning_rate": 3.9981647203537125e-06, + "loss": 0.88689983, + "num_input_tokens_seen": 7599890, + "step": 360, + "time_per_iteration": 2.761695623397827 + }, + { + "auxiliary_loss_clip": 0.01321158, + "auxiliary_loss_mlp": 0.01064543, + "balance_loss_clip": 1.09025133, + "balance_loss_mlp": 1.04517198, + "epoch": 0.04340768352071184, + "flos": 21283033530240.0, + "grad_norm": 2.864162759870375, + "language_loss": 0.95602667, + "learning_rate": 3.998131205178063e-06, + "loss": 0.97988361, + "num_input_tokens_seen": 7618360, + "step": 361, + "time_per_iteration": 2.77510404586792 + }, + { + "auxiliary_loss_clip": 0.01321308, + "auxiliary_loss_mlp": 0.01056956, + "balance_loss_clip": 1.08990073, + "balance_loss_mlp": 1.0370481, + "epoch": 0.04352792641135093, + "flos": 11583705951360.0, + "grad_norm": 3.422290636468516, + "language_loss": 0.76767254, + "learning_rate": 3.998097386887498e-06, + "loss": 0.79145515, + "num_input_tokens_seen": 7635435, + "step": 362, + "time_per_iteration": 2.7113537788391113 + }, + { + "auxiliary_loss_clip": 0.01313211, + "auxiliary_loss_mlp": 0.01062473, + "balance_loss_clip": 1.0892415, + "balance_loss_mlp": 1.04337561, + "epoch": 0.04364816930199002, + "flos": 23623547736960.0, + "grad_norm": 1.676167580788389, + "language_loss": 0.84906125, + "learning_rate": 3.998063265487148e-06, + "loss": 0.87281805, + "num_input_tokens_seen": 7656485, + "step": 363, + "time_per_iteration": 2.7467949390411377 + }, + { + "auxiliary_loss_clip": 0.0131522, + "auxiliary_loss_mlp": 0.01054801, + "balance_loss_clip": 1.08857727, + "balance_loss_mlp": 1.03523898, + "epoch": 0.043768412192629114, + "flos": 14429734214400.0, + "grad_norm": 3.8532358317294624, + "language_loss": 0.80789065, + "learning_rate": 3.99802884098219e-06, + "loss": 0.83159089, + "num_input_tokens_seen": 7674595, + "step": 364, + "time_per_iteration": 2.856938362121582 + }, + { + "auxiliary_loss_clip": 0.0131523, + "auxiliary_loss_mlp": 0.01063051, + "balance_loss_clip": 1.08571863, + "balance_loss_mlp": 1.04301238, + "epoch": 0.043888655083268203, + "flos": 26468893641600.0, + "grad_norm": 2.3046767346549637, + "language_loss": 0.82646489, + "learning_rate": 3.997994113377845e-06, + "loss": 0.85024774, + "num_input_tokens_seen": 7693495, + "step": 365, + "time_per_iteration": 2.7638134956359863 + }, + { + "auxiliary_loss_clip": 0.01319195, + "auxiliary_loss_mlp": 0.01053776, + "balance_loss_clip": 1.08878291, + "balance_loss_mlp": 1.03241384, + "epoch": 0.04400889797390729, + "flos": 27235263242880.0, + "grad_norm": 2.2257418603982475, + "language_loss": 0.83464801, + "learning_rate": 3.9979590826793815e-06, + "loss": 0.8583777, + "num_input_tokens_seen": 7714685, + "step": 366, + "time_per_iteration": 2.7865946292877197 + }, + { + "auxiliary_loss_clip": 0.0132675, + "auxiliary_loss_mlp": 0.01048667, + "balance_loss_clip": 1.0935607, + "balance_loss_mlp": 1.02825809, + "epoch": 0.04412914086454638, + "flos": 20119528183680.0, + "grad_norm": 1.996303361525515, + "language_loss": 0.80957246, + "learning_rate": 3.997923748892113e-06, + "loss": 0.83332664, + "num_input_tokens_seen": 7734005, + "step": 367, + "time_per_iteration": 2.702887535095215 + }, + { + "auxiliary_loss_clip": 0.01314489, + "auxiliary_loss_mlp": 0.01061314, + "balance_loss_clip": 1.08910882, + "balance_loss_mlp": 1.04199016, + "epoch": 0.04424938375518547, + "flos": 22604618632320.0, + "grad_norm": 1.9012151566440196, + "language_loss": 0.88749415, + "learning_rate": 3.9978881120214015e-06, + "loss": 0.9112522, + "num_input_tokens_seen": 7755525, + "step": 368, + "time_per_iteration": 2.7361152172088623 + }, + { + "auxiliary_loss_clip": 0.01317264, + "auxiliary_loss_mlp": 0.01057722, + "balance_loss_clip": 1.08598363, + "balance_loss_mlp": 1.03737307, + "epoch": 0.04436962664582456, + "flos": 24132365844480.0, + "grad_norm": 1.9299696752147513, + "language_loss": 0.7940824, + "learning_rate": 3.997852172072652e-06, + "loss": 0.81783223, + "num_input_tokens_seen": 7776740, + "step": 369, + "time_per_iteration": 2.718546152114868 + }, + { + "auxiliary_loss_clip": 0.01319873, + "auxiliary_loss_mlp": 0.01062846, + "balance_loss_clip": 1.08858871, + "balance_loss_mlp": 1.04244959, + "epoch": 0.04448986953646366, + "flos": 18222906251520.0, + "grad_norm": 2.618780099859536, + "language_loss": 0.89540976, + "learning_rate": 3.9978159290513155e-06, + "loss": 0.91923696, + "num_input_tokens_seen": 7794820, + "step": 370, + "time_per_iteration": 2.7575066089630127 + }, + { + "auxiliary_loss_clip": 0.0132362, + "auxiliary_loss_mlp": 0.01066652, + "balance_loss_clip": 1.09065056, + "balance_loss_mlp": 1.04478884, + "epoch": 0.04461011242710275, + "flos": 30117920400000.0, + "grad_norm": 1.7363881987312513, + "language_loss": 0.80073917, + "learning_rate": 3.997779382962892e-06, + "loss": 0.82464182, + "num_input_tokens_seen": 7817705, + "step": 371, + "time_per_iteration": 2.8287193775177 + }, + { + "auxiliary_loss_clip": 0.01315211, + "auxiliary_loss_mlp": 0.01062096, + "balance_loss_clip": 1.08746028, + "balance_loss_mlp": 1.04103184, + "epoch": 0.04473035531774184, + "flos": 29752529299200.0, + "grad_norm": 2.0886023301260055, + "language_loss": 0.73845303, + "learning_rate": 3.997742533812924e-06, + "loss": 0.7622261, + "num_input_tokens_seen": 7840970, + "step": 372, + "time_per_iteration": 2.7348997592926025 + }, + { + "auxiliary_loss_clip": 0.01315131, + "auxiliary_loss_mlp": 0.01061665, + "balance_loss_clip": 1.08785021, + "balance_loss_mlp": 1.04113722, + "epoch": 0.04485059820838093, + "flos": 13151565676800.0, + "grad_norm": 2.626085289907967, + "language_loss": 0.92400616, + "learning_rate": 3.997705381607001e-06, + "loss": 0.94777411, + "num_input_tokens_seen": 7857785, + "step": 373, + "time_per_iteration": 2.6989965438842773 + }, + { + "auxiliary_loss_clip": 0.01233461, + "auxiliary_loss_mlp": 0.01013261, + "balance_loss_clip": 1.09422612, + "balance_loss_mlp": 1.00238919, + "epoch": 0.04497084109902002, + "flos": 68094209548800.0, + "grad_norm": 1.016230466556008, + "language_loss": 0.6025176, + "learning_rate": 3.997667926350761e-06, + "loss": 0.62498486, + "num_input_tokens_seen": 7916115, + "step": 374, + "time_per_iteration": 3.171409845352173 + }, + { + "auxiliary_loss_clip": 0.01228874, + "auxiliary_loss_mlp": 0.01011625, + "balance_loss_clip": 1.08977175, + "balance_loss_mlp": 1.00056291, + "epoch": 0.04509108398965911, + "flos": 64342263346560.0, + "grad_norm": 0.9140006293316637, + "language_loss": 0.57743645, + "learning_rate": 3.997630168049886e-06, + "loss": 0.59984148, + "num_input_tokens_seen": 7974480, + "step": 375, + "time_per_iteration": 3.263284921646118 + }, + { + "auxiliary_loss_clip": 0.01324066, + "auxiliary_loss_mlp": 0.01065303, + "balance_loss_clip": 1.09095335, + "balance_loss_mlp": 1.04396462, + "epoch": 0.045211326880298205, + "flos": 22271115830400.0, + "grad_norm": 2.0272831592115486, + "language_loss": 0.7746917, + "learning_rate": 3.997592106710101e-06, + "loss": 0.79858541, + "num_input_tokens_seen": 7993940, + "step": 376, + "time_per_iteration": 2.738095283508301 + }, + { + "auxiliary_loss_clip": 0.01311176, + "auxiliary_loss_mlp": 0.01056494, + "balance_loss_clip": 1.08410227, + "balance_loss_mlp": 1.03693223, + "epoch": 0.045331569770937295, + "flos": 32159441796480.0, + "grad_norm": 2.777038586916122, + "language_loss": 0.66052932, + "learning_rate": 3.997553742337182e-06, + "loss": 0.68420601, + "num_input_tokens_seen": 8013365, + "step": 377, + "time_per_iteration": 2.794957399368286 + }, + { + "auxiliary_loss_clip": 0.01318632, + "auxiliary_loss_mlp": 0.01056811, + "balance_loss_clip": 1.08894467, + "balance_loss_mlp": 1.03768992, + "epoch": 0.045451812661576385, + "flos": 22163455791360.0, + "grad_norm": 1.8935432582584735, + "language_loss": 0.91496515, + "learning_rate": 3.997515074936949e-06, + "loss": 0.93871963, + "num_input_tokens_seen": 8034240, + "step": 378, + "time_per_iteration": 2.81182861328125 + }, + { + "auxiliary_loss_clip": 0.01315687, + "auxiliary_loss_mlp": 0.01061668, + "balance_loss_clip": 1.08858597, + "balance_loss_mlp": 1.04204631, + "epoch": 0.045572055552215475, + "flos": 16581968305920.0, + "grad_norm": 2.860401415508225, + "language_loss": 0.87165308, + "learning_rate": 3.997476104515268e-06, + "loss": 0.89542663, + "num_input_tokens_seen": 8052430, + "step": 379, + "time_per_iteration": 2.7003026008605957 + }, + { + "auxiliary_loss_clip": 0.01310318, + "auxiliary_loss_mlp": 0.01056105, + "balance_loss_clip": 1.0881772, + "balance_loss_mlp": 1.03692365, + "epoch": 0.045692298442854565, + "flos": 17603375448960.0, + "grad_norm": 3.067950414479415, + "language_loss": 0.77863157, + "learning_rate": 3.9974368310780485e-06, + "loss": 0.8022958, + "num_input_tokens_seen": 8069605, + "step": 380, + "time_per_iteration": 2.7309327125549316 + }, + { + "auxiliary_loss_clip": 0.01325674, + "auxiliary_loss_mlp": 0.0107291, + "balance_loss_clip": 1.09115982, + "balance_loss_mlp": 1.05065393, + "epoch": 0.045812541333493655, + "flos": 26761098781440.0, + "grad_norm": 2.447714933528025, + "language_loss": 0.7453723, + "learning_rate": 3.997397254631251e-06, + "loss": 0.76935822, + "num_input_tokens_seen": 8090225, + "step": 381, + "time_per_iteration": 2.757856607437134 + }, + { + "auxiliary_loss_clip": 0.01212271, + "auxiliary_loss_mlp": 0.01026329, + "balance_loss_clip": 1.07574368, + "balance_loss_mlp": 1.01574314, + "epoch": 0.04593278422413275, + "flos": 60250349894400.0, + "grad_norm": 0.8220420262431702, + "language_loss": 0.60056806, + "learning_rate": 3.997357375180878e-06, + "loss": 0.62295401, + "num_input_tokens_seen": 8154505, + "step": 382, + "time_per_iteration": 3.4319684505462646 + }, + { + "auxiliary_loss_clip": 0.01311723, + "auxiliary_loss_mlp": 0.01051335, + "balance_loss_clip": 1.08457041, + "balance_loss_mlp": 1.03044963, + "epoch": 0.04605302711477184, + "flos": 21799249839360.0, + "grad_norm": 1.7768934866529567, + "language_loss": 0.7529608, + "learning_rate": 3.997317192732979e-06, + "loss": 0.77659142, + "num_input_tokens_seen": 8173285, + "step": 383, + "time_per_iteration": 2.7476491928100586 + }, + { + "auxiliary_loss_clip": 0.01317949, + "auxiliary_loss_mlp": 0.01065901, + "balance_loss_clip": 1.08783901, + "balance_loss_mlp": 1.04475284, + "epoch": 0.04617327000541093, + "flos": 19459705299840.0, + "grad_norm": 2.111716110064841, + "language_loss": 0.8273797, + "learning_rate": 3.99727670729365e-06, + "loss": 0.85121822, + "num_input_tokens_seen": 8191845, + "step": 384, + "time_per_iteration": 3.6467041969299316 + }, + { + "auxiliary_loss_clip": 0.01313455, + "auxiliary_loss_mlp": 0.01067653, + "balance_loss_clip": 1.08956075, + "balance_loss_mlp": 1.04729176, + "epoch": 0.04629351289605002, + "flos": 25411468135680.0, + "grad_norm": 1.7901443200165736, + "language_loss": 0.78130591, + "learning_rate": 3.997235918869033e-06, + "loss": 0.80511701, + "num_input_tokens_seen": 8212880, + "step": 385, + "time_per_iteration": 4.768811464309692 + }, + { + "auxiliary_loss_clip": 0.0131325, + "auxiliary_loss_mlp": 0.01058386, + "balance_loss_clip": 1.08869934, + "balance_loss_mlp": 1.03946722, + "epoch": 0.04641375578668911, + "flos": 20558284813440.0, + "grad_norm": 1.9113891567335224, + "language_loss": 0.82571852, + "learning_rate": 3.997194827465315e-06, + "loss": 0.84943485, + "num_input_tokens_seen": 8231475, + "step": 386, + "time_per_iteration": 2.8070335388183594 + }, + { + "auxiliary_loss_clip": 0.01312301, + "auxiliary_loss_mlp": 0.01053075, + "balance_loss_clip": 1.08576536, + "balance_loss_mlp": 1.03284574, + "epoch": 0.0465339986773282, + "flos": 13188661447680.0, + "grad_norm": 2.5181106261308295, + "language_loss": 0.91504562, + "learning_rate": 3.997153433088728e-06, + "loss": 0.93869948, + "num_input_tokens_seen": 8248600, + "step": 387, + "time_per_iteration": 2.6797595024108887 + }, + { + "auxiliary_loss_clip": 0.01313323, + "auxiliary_loss_mlp": 0.01065219, + "balance_loss_clip": 1.08641613, + "balance_loss_mlp": 1.04405951, + "epoch": 0.0466542415679673, + "flos": 25556547168000.0, + "grad_norm": 2.0163557051080625, + "language_loss": 0.8105154, + "learning_rate": 3.997111735745554e-06, + "loss": 0.83430082, + "num_input_tokens_seen": 8271570, + "step": 388, + "time_per_iteration": 2.886796712875366 + }, + { + "auxiliary_loss_clip": 0.01309066, + "auxiliary_loss_mlp": 0.01060327, + "balance_loss_clip": 1.08650589, + "balance_loss_mlp": 1.04058611, + "epoch": 0.04677448445860639, + "flos": 22236749493120.0, + "grad_norm": 1.826436349286185, + "language_loss": 0.82873106, + "learning_rate": 3.997069735442118e-06, + "loss": 0.85242498, + "num_input_tokens_seen": 8291265, + "step": 389, + "time_per_iteration": 2.7509095668792725 + }, + { + "auxiliary_loss_clip": 0.01307702, + "auxiliary_loss_mlp": 0.01051659, + "balance_loss_clip": 1.08493602, + "balance_loss_mlp": 1.03407574, + "epoch": 0.04689472734924548, + "flos": 28147825198080.0, + "grad_norm": 1.5244477896889779, + "language_loss": 0.80404478, + "learning_rate": 3.997027432184792e-06, + "loss": 0.82763839, + "num_input_tokens_seen": 8315925, + "step": 390, + "time_per_iteration": 2.8380801677703857 + }, + { + "auxiliary_loss_clip": 0.01312661, + "auxiliary_loss_mlp": 0.01058837, + "balance_loss_clip": 1.08612609, + "balance_loss_mlp": 1.0388217, + "epoch": 0.04701497023988457, + "flos": 23148952312320.0, + "grad_norm": 1.8839822174416205, + "language_loss": 0.89587486, + "learning_rate": 3.99698482597999e-06, + "loss": 0.91958988, + "num_input_tokens_seen": 8333605, + "step": 391, + "time_per_iteration": 2.70863938331604 + }, + { + "auxiliary_loss_clip": 0.01204588, + "auxiliary_loss_mlp": 0.01014302, + "balance_loss_clip": 1.06885207, + "balance_loss_mlp": 1.00390744, + "epoch": 0.04713521313052366, + "flos": 64827668764800.0, + "grad_norm": 0.8711849814867207, + "language_loss": 0.63924253, + "learning_rate": 3.99694191683418e-06, + "loss": 0.66143143, + "num_input_tokens_seen": 8394405, + "step": 392, + "time_per_iteration": 3.2751317024230957 + }, + { + "auxiliary_loss_clip": 0.01317791, + "auxiliary_loss_mlp": 0.01060724, + "balance_loss_clip": 1.09199882, + "balance_loss_mlp": 1.04106629, + "epoch": 0.047255456021162746, + "flos": 18771585477120.0, + "grad_norm": 1.9344489083161778, + "language_loss": 0.81597817, + "learning_rate": 3.996898704753867e-06, + "loss": 0.8397634, + "num_input_tokens_seen": 8412355, + "step": 393, + "time_per_iteration": 2.756197690963745 + }, + { + "auxiliary_loss_clip": 0.01310188, + "auxiliary_loss_mlp": 0.01052804, + "balance_loss_clip": 1.08530807, + "balance_loss_mlp": 1.03346801, + "epoch": 0.04737569891180184, + "flos": 22053820504320.0, + "grad_norm": 2.0822741701678646, + "language_loss": 0.87498134, + "learning_rate": 3.996855189745609e-06, + "loss": 0.89861131, + "num_input_tokens_seen": 8431620, + "step": 394, + "time_per_iteration": 2.7031054496765137 + }, + { + "auxiliary_loss_clip": 0.01305361, + "auxiliary_loss_mlp": 0.01052565, + "balance_loss_clip": 1.08318925, + "balance_loss_mlp": 1.03423095, + "epoch": 0.04749594180244093, + "flos": 29057370410880.0, + "grad_norm": 2.269781578583948, + "language_loss": 0.9285121, + "learning_rate": 3.996811371816007e-06, + "loss": 0.9520914, + "num_input_tokens_seen": 8454045, + "step": 395, + "time_per_iteration": 2.7799808979034424 + }, + { + "auxiliary_loss_clip": 0.01304342, + "auxiliary_loss_mlp": 0.01053734, + "balance_loss_clip": 1.08477974, + "balance_loss_mlp": 1.03454113, + "epoch": 0.04761618469308002, + "flos": 35112268172160.0, + "grad_norm": 2.3204270795751363, + "language_loss": 0.77812028, + "learning_rate": 3.996767250971707e-06, + "loss": 0.80170101, + "num_input_tokens_seen": 8476785, + "step": 396, + "time_per_iteration": 2.7999236583709717 + }, + { + "auxiliary_loss_clip": 0.01310221, + "auxiliary_loss_mlp": 0.01050846, + "balance_loss_clip": 1.08632314, + "balance_loss_mlp": 1.03278613, + "epoch": 0.04773642758371911, + "flos": 25630702796160.0, + "grad_norm": 3.778506085264851, + "language_loss": 0.8736375, + "learning_rate": 3.996722827219403e-06, + "loss": 0.89724815, + "num_input_tokens_seen": 8498400, + "step": 397, + "time_per_iteration": 2.82930064201355 + }, + { + "auxiliary_loss_clip": 0.01313111, + "auxiliary_loss_mlp": 0.01060095, + "balance_loss_clip": 1.08868504, + "balance_loss_mlp": 1.03905487, + "epoch": 0.0478566704743582, + "flos": 20631506688000.0, + "grad_norm": 2.531263477454368, + "language_loss": 0.82687449, + "learning_rate": 3.996678100565833e-06, + "loss": 0.85060662, + "num_input_tokens_seen": 8517455, + "step": 398, + "time_per_iteration": 2.754387617111206 + }, + { + "auxiliary_loss_clip": 0.01304014, + "auxiliary_loss_mlp": 0.01053113, + "balance_loss_clip": 1.0838182, + "balance_loss_mlp": 1.03383684, + "epoch": 0.04797691336499729, + "flos": 18835721210880.0, + "grad_norm": 2.457597648937002, + "language_loss": 0.8908484, + "learning_rate": 3.996633071017783e-06, + "loss": 0.91441965, + "num_input_tokens_seen": 8534085, + "step": 399, + "time_per_iteration": 2.6910440921783447 + }, + { + "auxiliary_loss_clip": 0.01304904, + "auxiliary_loss_mlp": 0.01055161, + "balance_loss_clip": 1.08566952, + "balance_loss_mlp": 1.03452635, + "epoch": 0.04809715625563638, + "flos": 21099673578240.0, + "grad_norm": 2.0099698241525457, + "language_loss": 0.82116508, + "learning_rate": 3.996587738582084e-06, + "loss": 0.84476572, + "num_input_tokens_seen": 8550885, + "step": 400, + "time_per_iteration": 2.739224672317505 + }, + { + "auxiliary_loss_clip": 0.01299129, + "auxiliary_loss_mlp": 0.01051041, + "balance_loss_clip": 1.08222747, + "balance_loss_mlp": 1.03248024, + "epoch": 0.04821739914627548, + "flos": 23805650712960.0, + "grad_norm": 3.2687538678068773, + "language_loss": 0.86189866, + "learning_rate": 3.9965421032656115e-06, + "loss": 0.88540035, + "num_input_tokens_seen": 8570815, + "step": 401, + "time_per_iteration": 2.7221744060516357 + }, + { + "auxiliary_loss_clip": 0.01305365, + "auxiliary_loss_mlp": 0.01057102, + "balance_loss_clip": 1.08517444, + "balance_loss_mlp": 1.03819561, + "epoch": 0.04833764203691457, + "flos": 22200587475840.0, + "grad_norm": 2.6957319081941375, + "language_loss": 0.94550067, + "learning_rate": 3.99649616507529e-06, + "loss": 0.96912533, + "num_input_tokens_seen": 8589910, + "step": 402, + "time_per_iteration": 2.735646963119507 + }, + { + "auxiliary_loss_clip": 0.01193576, + "auxiliary_loss_mlp": 0.01009957, + "balance_loss_clip": 1.05989003, + "balance_loss_mlp": 0.99946612, + "epoch": 0.04845788492755366, + "flos": 65904376896000.0, + "grad_norm": 0.8931728364427178, + "language_loss": 0.63167411, + "learning_rate": 3.996449924018088e-06, + "loss": 0.65370947, + "num_input_tokens_seen": 8650370, + "step": 403, + "time_per_iteration": 3.298811912536621 + }, + { + "auxiliary_loss_clip": 0.0130091, + "auxiliary_loss_mlp": 0.01048096, + "balance_loss_clip": 1.08510196, + "balance_loss_mlp": 1.03050077, + "epoch": 0.04857812781819275, + "flos": 19281301424640.0, + "grad_norm": 2.0112887941136046, + "language_loss": 0.79637831, + "learning_rate": 3.99640338010102e-06, + "loss": 0.81986833, + "num_input_tokens_seen": 8669475, + "step": 404, + "time_per_iteration": 2.8528900146484375 + }, + { + "auxiliary_loss_clip": 0.01302696, + "auxiliary_loss_mlp": 0.01052323, + "balance_loss_clip": 1.08281374, + "balance_loss_mlp": 1.03367829, + "epoch": 0.04869837070883184, + "flos": 24062376193920.0, + "grad_norm": 1.9425408257587642, + "language_loss": 0.78808403, + "learning_rate": 3.996356533331146e-06, + "loss": 0.8116343, + "num_input_tokens_seen": 8691345, + "step": 405, + "time_per_iteration": 2.888857126235962 + }, + { + "auxiliary_loss_clip": 0.01312541, + "auxiliary_loss_mlp": 0.01054462, + "balance_loss_clip": 1.0836525, + "balance_loss_mlp": 1.03499544, + "epoch": 0.04881861359947093, + "flos": 25187169657600.0, + "grad_norm": 2.2043185945431665, + "language_loss": 0.61805987, + "learning_rate": 3.996309383715573e-06, + "loss": 0.64172989, + "num_input_tokens_seen": 8710125, + "step": 406, + "time_per_iteration": 2.7922611236572266 + }, + { + "auxiliary_loss_clip": 0.01305792, + "auxiliary_loss_mlp": 0.01059475, + "balance_loss_clip": 1.08377492, + "balance_loss_mlp": 1.03999615, + "epoch": 0.048938856490110025, + "flos": 16362913213440.0, + "grad_norm": 4.599237448024428, + "language_loss": 0.73821115, + "learning_rate": 3.996261931261454e-06, + "loss": 0.76186383, + "num_input_tokens_seen": 8728705, + "step": 407, + "time_per_iteration": 2.8200042247772217 + }, + { + "auxiliary_loss_clip": 0.01306579, + "auxiliary_loss_mlp": 0.01057744, + "balance_loss_clip": 1.0853157, + "balance_loss_mlp": 1.03837299, + "epoch": 0.049059099380749115, + "flos": 29895094379520.0, + "grad_norm": 4.320705048634516, + "language_loss": 0.86638862, + "learning_rate": 3.996214175975987e-06, + "loss": 0.89003181, + "num_input_tokens_seen": 8749225, + "step": 408, + "time_per_iteration": 2.7878687381744385 + }, + { + "auxiliary_loss_clip": 0.01308567, + "auxiliary_loss_mlp": 0.01057915, + "balance_loss_clip": 1.08787215, + "balance_loss_mlp": 1.03861547, + "epoch": 0.049179342271388204, + "flos": 35918858027520.0, + "grad_norm": 2.121735919563555, + "language_loss": 0.79305094, + "learning_rate": 3.996166117866417e-06, + "loss": 0.81671578, + "num_input_tokens_seen": 8771160, + "step": 409, + "time_per_iteration": 2.8999216556549072 + }, + { + "auxiliary_loss_clip": 0.01296636, + "auxiliary_loss_mlp": 0.01057468, + "balance_loss_clip": 1.07980847, + "balance_loss_mlp": 1.03902602, + "epoch": 0.049299585162027294, + "flos": 14611226659200.0, + "grad_norm": 2.1045434579457014, + "language_loss": 0.86804891, + "learning_rate": 3.996117756940035e-06, + "loss": 0.89158994, + "num_input_tokens_seen": 8787845, + "step": 410, + "time_per_iteration": 2.7731008529663086 + }, + { + "auxiliary_loss_clip": 0.0130147, + "auxiliary_loss_mlp": 0.01056812, + "balance_loss_clip": 1.08370399, + "balance_loss_mlp": 1.03899026, + "epoch": 0.049419828052666384, + "flos": 19567939956480.0, + "grad_norm": 1.9830652218414808, + "language_loss": 0.97556639, + "learning_rate": 3.996069093204175e-06, + "loss": 0.99914926, + "num_input_tokens_seen": 8803805, + "step": 411, + "time_per_iteration": 5.435424327850342 + }, + { + "auxiliary_loss_clip": 0.0130528, + "auxiliary_loss_mlp": 0.01052217, + "balance_loss_clip": 1.08560252, + "balance_loss_mlp": 1.03352499, + "epoch": 0.049540070943305474, + "flos": 13659916907520.0, + "grad_norm": 2.9350841701144774, + "language_loss": 0.88073713, + "learning_rate": 3.996020126666221e-06, + "loss": 0.90431207, + "num_input_tokens_seen": 8820785, + "step": 412, + "time_per_iteration": 3.632485866546631 + }, + { + "auxiliary_loss_clip": 0.01304153, + "auxiliary_loss_mlp": 0.01055816, + "balance_loss_clip": 1.08363473, + "balance_loss_mlp": 1.03715944, + "epoch": 0.04966031383394457, + "flos": 21832035978240.0, + "grad_norm": 2.1728715921229425, + "language_loss": 0.82509875, + "learning_rate": 3.995970857333601e-06, + "loss": 0.84869844, + "num_input_tokens_seen": 8841195, + "step": 413, + "time_per_iteration": 2.82668399810791 + }, + { + "auxiliary_loss_clip": 0.01301334, + "auxiliary_loss_mlp": 0.01062675, + "balance_loss_clip": 1.08193338, + "balance_loss_mlp": 1.04401839, + "epoch": 0.04978055672458366, + "flos": 28618793349120.0, + "grad_norm": 1.8097929551260032, + "language_loss": 0.79802549, + "learning_rate": 3.995921285213789e-06, + "loss": 0.82166559, + "num_input_tokens_seen": 8861455, + "step": 414, + "time_per_iteration": 2.7667176723480225 + }, + { + "auxiliary_loss_clip": 0.01295552, + "auxiliary_loss_mlp": 0.01055177, + "balance_loss_clip": 1.08032835, + "balance_loss_mlp": 1.0367955, + "epoch": 0.04990079961522275, + "flos": 19828220883840.0, + "grad_norm": 2.475792110945682, + "language_loss": 0.80970728, + "learning_rate": 3.995871410314305e-06, + "loss": 0.83321464, + "num_input_tokens_seen": 8880015, + "step": 415, + "time_per_iteration": 2.833432197570801 + }, + { + "auxiliary_loss_clip": 0.01183083, + "auxiliary_loss_mlp": 0.01012813, + "balance_loss_clip": 1.05939269, + "balance_loss_mlp": 1.00241792, + "epoch": 0.05002104250586184, + "flos": 62735045293440.0, + "grad_norm": 0.9065494288866045, + "language_loss": 0.59606773, + "learning_rate": 3.995821232642714e-06, + "loss": 0.61802661, + "num_input_tokens_seen": 8938420, + "step": 416, + "time_per_iteration": 3.4458656311035156 + }, + { + "auxiliary_loss_clip": 0.0128589, + "auxiliary_loss_mlp": 0.0105136, + "balance_loss_clip": 1.08169866, + "balance_loss_mlp": 1.03259671, + "epoch": 0.05014128539650093, + "flos": 27928518710400.0, + "grad_norm": 2.1076200429555487, + "language_loss": 0.82927072, + "learning_rate": 3.995770752206629e-06, + "loss": 0.85264325, + "num_input_tokens_seen": 8959495, + "step": 417, + "time_per_iteration": 2.81653094291687 + }, + { + "auxiliary_loss_clip": 0.01300902, + "auxiliary_loss_mlp": 0.01054918, + "balance_loss_clip": 1.084108, + "balance_loss_mlp": 1.03687012, + "epoch": 0.05026152828714002, + "flos": 17705576620800.0, + "grad_norm": 2.9603229448408808, + "language_loss": 0.97453207, + "learning_rate": 3.995719969013709e-06, + "loss": 0.99809033, + "num_input_tokens_seen": 8976675, + "step": 418, + "time_per_iteration": 2.790189027786255 + }, + { + "auxiliary_loss_clip": 0.01273463, + "auxiliary_loss_mlp": 0.01058723, + "balance_loss_clip": 1.078089, + "balance_loss_mlp": 1.03943515, + "epoch": 0.05038177117777912, + "flos": 19133277477120.0, + "grad_norm": 2.6232513864259905, + "language_loss": 0.85967869, + "learning_rate": 3.995668883071655e-06, + "loss": 0.88300055, + "num_input_tokens_seen": 8992900, + "step": 419, + "time_per_iteration": 2.8986825942993164 + }, + { + "auxiliary_loss_clip": 0.01302603, + "auxiliary_loss_mlp": 0.01060864, + "balance_loss_clip": 1.08389735, + "balance_loss_mlp": 1.04143238, + "epoch": 0.050502014068418206, + "flos": 20667704618880.0, + "grad_norm": 2.541217296523194, + "language_loss": 0.9115473, + "learning_rate": 3.995617494388219e-06, + "loss": 0.93518198, + "num_input_tokens_seen": 9011020, + "step": 420, + "time_per_iteration": 2.785429000854492 + }, + { + "auxiliary_loss_clip": 0.01270472, + "auxiliary_loss_mlp": 0.01052764, + "balance_loss_clip": 1.07480001, + "balance_loss_mlp": 1.03391719, + "epoch": 0.050622256959057296, + "flos": 21361103740800.0, + "grad_norm": 1.993202415821963, + "language_loss": 0.80617005, + "learning_rate": 3.995565802971196e-06, + "loss": 0.82940239, + "num_input_tokens_seen": 9030995, + "step": 421, + "time_per_iteration": 2.894571542739868 + }, + { + "auxiliary_loss_clip": 0.01271778, + "auxiliary_loss_mlp": 0.01064385, + "balance_loss_clip": 1.07627547, + "balance_loss_mlp": 1.04583573, + "epoch": 0.050742499849696386, + "flos": 27673588909440.0, + "grad_norm": 1.9251990448269989, + "language_loss": 0.67728567, + "learning_rate": 3.995513808828427e-06, + "loss": 0.70064723, + "num_input_tokens_seen": 9053790, + "step": 422, + "time_per_iteration": 2.9749534130096436 + }, + { + "auxiliary_loss_clip": 0.01276468, + "auxiliary_loss_mlp": 0.01056993, + "balance_loss_clip": 1.07746482, + "balance_loss_mlp": 1.03923082, + "epoch": 0.050862742740335476, + "flos": 19865999013120.0, + "grad_norm": 1.8321620346084926, + "language_loss": 0.76633549, + "learning_rate": 3.9954615119678e-06, + "loss": 0.78967011, + "num_input_tokens_seen": 9072345, + "step": 423, + "time_per_iteration": 2.878063440322876 + }, + { + "auxiliary_loss_clip": 0.01268809, + "auxiliary_loss_mlp": 0.0105251, + "balance_loss_clip": 1.07411218, + "balance_loss_mlp": 1.03427052, + "epoch": 0.050982985630974566, + "flos": 22085098272000.0, + "grad_norm": 2.0069294309927512, + "language_loss": 0.80694592, + "learning_rate": 3.995408912397248e-06, + "loss": 0.83015913, + "num_input_tokens_seen": 9090240, + "step": 424, + "time_per_iteration": 2.794396162033081 + }, + { + "auxiliary_loss_clip": 0.01277377, + "auxiliary_loss_mlp": 0.01057933, + "balance_loss_clip": 1.07905507, + "balance_loss_mlp": 1.04003942, + "epoch": 0.05110322852161366, + "flos": 20740962407040.0, + "grad_norm": 2.1414928685826937, + "language_loss": 0.93463027, + "learning_rate": 3.99535601012475e-06, + "loss": 0.95798331, + "num_input_tokens_seen": 9105570, + "step": 425, + "time_per_iteration": 2.834705352783203 + }, + { + "auxiliary_loss_clip": 0.01255444, + "auxiliary_loss_mlp": 0.00891482, + "balance_loss_clip": 1.07508945, + "balance_loss_mlp": 1.00065184, + "epoch": 0.05122347141225275, + "flos": 28547295327360.0, + "grad_norm": 2.5645541881853204, + "language_loss": 0.75761014, + "learning_rate": 3.995302805158333e-06, + "loss": 0.77907944, + "num_input_tokens_seen": 9128225, + "step": 426, + "time_per_iteration": 2.8655710220336914 + }, + { + "auxiliary_loss_clip": 0.01259211, + "auxiliary_loss_mlp": 0.01057934, + "balance_loss_clip": 1.07287192, + "balance_loss_mlp": 1.03732228, + "epoch": 0.05134371430289184, + "flos": 19722679747200.0, + "grad_norm": 3.1893656760556652, + "language_loss": 0.83365607, + "learning_rate": 3.9952492975060665e-06, + "loss": 0.85682744, + "num_input_tokens_seen": 9148295, + "step": 427, + "time_per_iteration": 2.9070041179656982 + }, + { + "auxiliary_loss_clip": 0.01281651, + "auxiliary_loss_mlp": 0.01046916, + "balance_loss_clip": 1.07643986, + "balance_loss_mlp": 1.02898717, + "epoch": 0.05146395719353093, + "flos": 34458945649920.0, + "grad_norm": 2.4254835678456432, + "language_loss": 0.8523733, + "learning_rate": 3.995195487176067e-06, + "loss": 0.87565899, + "num_input_tokens_seen": 9168525, + "step": 428, + "time_per_iteration": 2.8355207443237305 + }, + { + "auxiliary_loss_clip": 0.01297401, + "auxiliary_loss_mlp": 0.01056739, + "balance_loss_clip": 1.08171356, + "balance_loss_mlp": 1.03704548, + "epoch": 0.05158420008417002, + "flos": 21760286561280.0, + "grad_norm": 1.9903641270355548, + "language_loss": 0.86072361, + "learning_rate": 3.995141374176499e-06, + "loss": 0.88426501, + "num_input_tokens_seen": 9186920, + "step": 429, + "time_per_iteration": 2.7631044387817383 + }, + { + "auxiliary_loss_clip": 0.01162317, + "auxiliary_loss_mlp": 0.00882074, + "balance_loss_clip": 1.05790091, + "balance_loss_mlp": 1.00105429, + "epoch": 0.05170444297480911, + "flos": 72553956226560.0, + "grad_norm": 0.8716986575773545, + "language_loss": 0.63102359, + "learning_rate": 3.995086958515572e-06, + "loss": 0.65146744, + "num_input_tokens_seen": 9244940, + "step": 430, + "time_per_iteration": 3.399465560913086 + }, + { + "auxiliary_loss_clip": 0.01188825, + "auxiliary_loss_mlp": 0.00881948, + "balance_loss_clip": 1.06146049, + "balance_loss_mlp": 1.00117421, + "epoch": 0.05182468586544821, + "flos": 62416159326720.0, + "grad_norm": 0.8672507175794422, + "language_loss": 0.59927493, + "learning_rate": 3.995032240201538e-06, + "loss": 0.61998266, + "num_input_tokens_seen": 9307335, + "step": 431, + "time_per_iteration": 3.1989779472351074 + }, + { + "auxiliary_loss_clip": 0.01173442, + "auxiliary_loss_mlp": 0.01010982, + "balance_loss_clip": 1.05618227, + "balance_loss_mlp": 1.0004437, + "epoch": 0.0519449287560873, + "flos": 41225989432320.0, + "grad_norm": 0.9321827874631332, + "language_loss": 0.63102198, + "learning_rate": 3.9949772192427e-06, + "loss": 0.65286624, + "num_input_tokens_seen": 9353960, + "step": 432, + "time_per_iteration": 3.0701475143432617 + }, + { + "auxiliary_loss_clip": 0.01266744, + "auxiliary_loss_mlp": 0.01060276, + "balance_loss_clip": 1.07384956, + "balance_loss_mlp": 1.04142904, + "epoch": 0.05206517164672639, + "flos": 17494530261120.0, + "grad_norm": 1.8777490322032935, + "language_loss": 0.79671741, + "learning_rate": 3.994921895647405e-06, + "loss": 0.81998754, + "num_input_tokens_seen": 9372130, + "step": 433, + "time_per_iteration": 2.840967893600464 + }, + { + "auxiliary_loss_clip": 0.01183186, + "auxiliary_loss_mlp": 0.0101056, + "balance_loss_clip": 1.0568614, + "balance_loss_mlp": 1.00002217, + "epoch": 0.05218541453736548, + "flos": 64002762973440.0, + "grad_norm": 0.8401368102644036, + "language_loss": 0.55318326, + "learning_rate": 3.994866269424043e-06, + "loss": 0.57512075, + "num_input_tokens_seen": 9428500, + "step": 434, + "time_per_iteration": 3.2485063076019287 + }, + { + "auxiliary_loss_clip": 0.01231882, + "auxiliary_loss_mlp": 0.01046127, + "balance_loss_clip": 1.06072795, + "balance_loss_mlp": 1.02745867, + "epoch": 0.05230565742800457, + "flos": 19317319787520.0, + "grad_norm": 2.499999339362704, + "language_loss": 0.7876029, + "learning_rate": 3.9948103405810545e-06, + "loss": 0.81038308, + "num_input_tokens_seen": 9447450, + "step": 435, + "time_per_iteration": 3.066112995147705 + }, + { + "auxiliary_loss_clip": 0.01240477, + "auxiliary_loss_mlp": 0.01052345, + "balance_loss_clip": 1.06845379, + "balance_loss_mlp": 1.03526187, + "epoch": 0.05242590031864366, + "flos": 25298636538240.0, + "grad_norm": 1.8634222121900679, + "language_loss": 0.86012882, + "learning_rate": 3.994754109126923e-06, + "loss": 0.88305706, + "num_input_tokens_seen": 9468945, + "step": 436, + "time_per_iteration": 3.747920513153076 + }, + { + "auxiliary_loss_clip": 0.01225241, + "auxiliary_loss_mlp": 0.01051818, + "balance_loss_clip": 1.06774926, + "balance_loss_mlp": 1.03353143, + "epoch": 0.052546143209282754, + "flos": 26211629456640.0, + "grad_norm": 3.242795873595335, + "language_loss": 0.93356633, + "learning_rate": 3.994697575070181e-06, + "loss": 0.95633698, + "num_input_tokens_seen": 9488405, + "step": 437, + "time_per_iteration": 3.8668625354766846 + }, + { + "auxiliary_loss_clip": 0.01273561, + "auxiliary_loss_mlp": 0.0105632, + "balance_loss_clip": 1.07982945, + "balance_loss_mlp": 1.03721082, + "epoch": 0.052666386099921844, + "flos": 22158140578560.0, + "grad_norm": 2.0575729965857517, + "language_loss": 0.91555762, + "learning_rate": 3.994640738419402e-06, + "loss": 0.93885636, + "num_input_tokens_seen": 9507780, + "step": 438, + "time_per_iteration": 2.8024299144744873 + }, + { + "auxiliary_loss_clip": 0.01283118, + "auxiliary_loss_mlp": 0.01053741, + "balance_loss_clip": 1.07908297, + "balance_loss_mlp": 1.03457177, + "epoch": 0.052786628990560934, + "flos": 23881817502720.0, + "grad_norm": 2.2790456216233945, + "language_loss": 0.81224239, + "learning_rate": 3.9945835991832075e-06, + "loss": 0.83561099, + "num_input_tokens_seen": 9529665, + "step": 439, + "time_per_iteration": 2.7858524322509766 + }, + { + "auxiliary_loss_clip": 0.01299896, + "auxiliary_loss_mlp": 0.0105941, + "balance_loss_clip": 1.08653033, + "balance_loss_mlp": 1.0409205, + "epoch": 0.052906871881200024, + "flos": 24605021934720.0, + "grad_norm": 1.9556942870497018, + "language_loss": 0.92954946, + "learning_rate": 3.994526157370268e-06, + "loss": 0.95314252, + "num_input_tokens_seen": 9548280, + "step": 440, + "time_per_iteration": 2.775719404220581 + }, + { + "auxiliary_loss_clip": 0.01164465, + "auxiliary_loss_mlp": 0.01014736, + "balance_loss_clip": 1.04863143, + "balance_loss_mlp": 1.00415063, + "epoch": 0.053027114771839114, + "flos": 56461631143680.0, + "grad_norm": 0.8944197323996272, + "language_loss": 0.59254909, + "learning_rate": 3.994468412989296e-06, + "loss": 0.61434114, + "num_input_tokens_seen": 9609690, + "step": 441, + "time_per_iteration": 3.4604201316833496 + }, + { + "auxiliary_loss_clip": 0.01248959, + "auxiliary_loss_mlp": 0.01050185, + "balance_loss_clip": 1.07174265, + "balance_loss_mlp": 1.03123057, + "epoch": 0.053147357662478203, + "flos": 17311098481920.0, + "grad_norm": 2.3943228629488806, + "language_loss": 0.9262175, + "learning_rate": 3.994410366049052e-06, + "loss": 0.94920897, + "num_input_tokens_seen": 9627550, + "step": 442, + "time_per_iteration": 2.8084332942962646 + }, + { + "auxiliary_loss_clip": 0.01279747, + "auxiliary_loss_mlp": 0.01053769, + "balance_loss_clip": 1.07836795, + "balance_loss_mlp": 1.03575635, + "epoch": 0.0532676005531173, + "flos": 17164977955200.0, + "grad_norm": 2.372556072326585, + "language_loss": 0.83233154, + "learning_rate": 3.994352016558341e-06, + "loss": 0.8556667, + "num_input_tokens_seen": 9644855, + "step": 443, + "time_per_iteration": 2.791114330291748 + }, + { + "auxiliary_loss_clip": 0.0128237, + "auxiliary_loss_mlp": 0.01055588, + "balance_loss_clip": 1.08090878, + "balance_loss_mlp": 1.03714609, + "epoch": 0.05338784344375639, + "flos": 27819960831360.0, + "grad_norm": 2.080411539558911, + "language_loss": 0.74031889, + "learning_rate": 3.994293364526014e-06, + "loss": 0.76369846, + "num_input_tokens_seen": 9665740, + "step": 444, + "time_per_iteration": 2.8290562629699707 + }, + { + "auxiliary_loss_clip": 0.01260037, + "auxiliary_loss_mlp": 0.01051875, + "balance_loss_clip": 1.07620502, + "balance_loss_mlp": 1.03283763, + "epoch": 0.05350808633439548, + "flos": 21507691144320.0, + "grad_norm": 2.4149650205706363, + "language_loss": 0.84965861, + "learning_rate": 3.99423440996097e-06, + "loss": 0.8727777, + "num_input_tokens_seen": 9685280, + "step": 445, + "time_per_iteration": 2.9099981784820557 + }, + { + "auxiliary_loss_clip": 0.01272755, + "auxiliary_loss_mlp": 0.01053518, + "balance_loss_clip": 1.08008814, + "balance_loss_mlp": 1.03531444, + "epoch": 0.05362832922503457, + "flos": 20084299920000.0, + "grad_norm": 3.2875777668113617, + "language_loss": 0.81589663, + "learning_rate": 3.994175152872152e-06, + "loss": 0.83915937, + "num_input_tokens_seen": 9704365, + "step": 446, + "time_per_iteration": 2.7253592014312744 + }, + { + "auxiliary_loss_clip": 0.01288547, + "auxiliary_loss_mlp": 0.01053765, + "balance_loss_clip": 1.07997966, + "balance_loss_mlp": 1.03574049, + "epoch": 0.05374857211567366, + "flos": 26137222433280.0, + "grad_norm": 2.011107914048173, + "language_loss": 0.78846788, + "learning_rate": 3.994115593268548e-06, + "loss": 0.81189108, + "num_input_tokens_seen": 9724145, + "step": 447, + "time_per_iteration": 2.832960605621338 + }, + { + "auxiliary_loss_clip": 0.01295729, + "auxiliary_loss_mlp": 0.01049205, + "balance_loss_clip": 1.08223248, + "balance_loss_mlp": 1.03166914, + "epoch": 0.05386881500631275, + "flos": 27486817165440.0, + "grad_norm": 1.8856797729997392, + "language_loss": 0.82282901, + "learning_rate": 3.994055731159195e-06, + "loss": 0.84627831, + "num_input_tokens_seen": 9741615, + "step": 448, + "time_per_iteration": 2.86344575881958 + }, + { + "auxiliary_loss_clip": 0.01286176, + "auxiliary_loss_mlp": 0.01068775, + "balance_loss_clip": 1.08261275, + "balance_loss_mlp": 1.05212188, + "epoch": 0.053989057896951846, + "flos": 23585087249280.0, + "grad_norm": 2.7841382897434617, + "language_loss": 0.86961693, + "learning_rate": 3.993995566553172e-06, + "loss": 0.89316648, + "num_input_tokens_seen": 9760580, + "step": 449, + "time_per_iteration": 2.9555959701538086 + }, + { + "auxiliary_loss_clip": 0.0125906, + "auxiliary_loss_mlp": 0.01056998, + "balance_loss_clip": 1.07171988, + "balance_loss_mlp": 1.03866339, + "epoch": 0.054109300787590936, + "flos": 25228862369280.0, + "grad_norm": 1.6493392780635916, + "language_loss": 0.77353495, + "learning_rate": 3.993935099459607e-06, + "loss": 0.79669553, + "num_input_tokens_seen": 9782195, + "step": 450, + "time_per_iteration": 2.9308276176452637 + }, + { + "auxiliary_loss_clip": 0.01286729, + "auxiliary_loss_mlp": 0.01048883, + "balance_loss_clip": 1.0785147, + "balance_loss_mlp": 1.03125179, + "epoch": 0.054229543678230026, + "flos": 23841525421440.0, + "grad_norm": 2.075918270701772, + "language_loss": 0.74216038, + "learning_rate": 3.993874329887673e-06, + "loss": 0.76551652, + "num_input_tokens_seen": 9800850, + "step": 451, + "time_per_iteration": 2.731997013092041 + }, + { + "auxiliary_loss_clip": 0.01280423, + "auxiliary_loss_mlp": 0.01048531, + "balance_loss_clip": 1.07945657, + "balance_loss_mlp": 1.03004169, + "epoch": 0.054349786568869116, + "flos": 16320933192960.0, + "grad_norm": 2.1792880677304796, + "language_loss": 0.86706686, + "learning_rate": 3.993813257846589e-06, + "loss": 0.89035642, + "num_input_tokens_seen": 9817605, + "step": 452, + "time_per_iteration": 2.7715468406677246 + }, + { + "auxiliary_loss_clip": 0.01285481, + "auxiliary_loss_mlp": 0.01059597, + "balance_loss_clip": 1.08043671, + "balance_loss_mlp": 1.04231119, + "epoch": 0.054470029459508205, + "flos": 18660729127680.0, + "grad_norm": 2.0627981911298336, + "language_loss": 0.92844695, + "learning_rate": 3.993751883345619e-06, + "loss": 0.95189774, + "num_input_tokens_seen": 9835965, + "step": 453, + "time_per_iteration": 2.783092737197876 + }, + { + "auxiliary_loss_clip": 0.01266931, + "auxiliary_loss_mlp": 0.01053397, + "balance_loss_clip": 1.0767324, + "balance_loss_mlp": 1.03632593, + "epoch": 0.054590272350147295, + "flos": 17785298856960.0, + "grad_norm": 3.2834796158031176, + "language_loss": 0.87697518, + "learning_rate": 3.993690206394073e-06, + "loss": 0.90017843, + "num_input_tokens_seen": 9852265, + "step": 454, + "time_per_iteration": 2.8520941734313965 + }, + { + "auxiliary_loss_clip": 0.01282305, + "auxiliary_loss_mlp": 0.01051053, + "balance_loss_clip": 1.08032179, + "balance_loss_mlp": 1.03286171, + "epoch": 0.054710515240786385, + "flos": 17785945301760.0, + "grad_norm": 2.1691710634738817, + "language_loss": 0.87715507, + "learning_rate": 3.993628227001307e-06, + "loss": 0.90048867, + "num_input_tokens_seen": 9870465, + "step": 455, + "time_per_iteration": 2.8361713886260986 + }, + { + "auxiliary_loss_clip": 0.01264912, + "auxiliary_loss_mlp": 0.01050337, + "balance_loss_clip": 1.07569981, + "balance_loss_mlp": 1.03201449, + "epoch": 0.05483075813142548, + "flos": 48210900180480.0, + "grad_norm": 1.953807308476536, + "language_loss": 0.71603131, + "learning_rate": 3.993565945176726e-06, + "loss": 0.73918378, + "num_input_tokens_seen": 9891490, + "step": 456, + "time_per_iteration": 3.1444969177246094 + }, + { + "auxiliary_loss_clip": 0.01265191, + "auxiliary_loss_mlp": 0.01053856, + "balance_loss_clip": 1.07904315, + "balance_loss_mlp": 1.03671408, + "epoch": 0.05495100102206457, + "flos": 19682244011520.0, + "grad_norm": 2.9311927430644804, + "language_loss": 0.84289652, + "learning_rate": 3.993503360929776e-06, + "loss": 0.86608702, + "num_input_tokens_seen": 9910375, + "step": 457, + "time_per_iteration": 2.8097400665283203 + }, + { + "auxiliary_loss_clip": 0.01225408, + "auxiliary_loss_mlp": 0.0105724, + "balance_loss_clip": 1.06845927, + "balance_loss_mlp": 1.03881061, + "epoch": 0.05507124391270366, + "flos": 26360048453760.0, + "grad_norm": 1.642913628414985, + "language_loss": 0.81263548, + "learning_rate": 3.99344047426995e-06, + "loss": 0.83546197, + "num_input_tokens_seen": 9931635, + "step": 458, + "time_per_iteration": 3.0960588455200195 + }, + { + "auxiliary_loss_clip": 0.01246656, + "auxiliary_loss_mlp": 0.01053295, + "balance_loss_clip": 1.07071257, + "balance_loss_mlp": 1.0346148, + "epoch": 0.05519148680334275, + "flos": 22601314581120.0, + "grad_norm": 2.387089608819481, + "language_loss": 0.9347645, + "learning_rate": 3.993377285206789e-06, + "loss": 0.95776403, + "num_input_tokens_seen": 9951420, + "step": 459, + "time_per_iteration": 3.268312931060791 + }, + { + "auxiliary_loss_clip": 0.01234036, + "auxiliary_loss_mlp": 0.01059049, + "balance_loss_clip": 1.06808496, + "balance_loss_mlp": 1.04166818, + "epoch": 0.05531172969398184, + "flos": 40552519380480.0, + "grad_norm": 1.855937754039239, + "language_loss": 0.86776578, + "learning_rate": 3.99331379374988e-06, + "loss": 0.89069664, + "num_input_tokens_seen": 9975025, + "step": 460, + "time_per_iteration": 2.9879391193389893 + }, + { + "auxiliary_loss_clip": 0.01261327, + "auxiliary_loss_mlp": 0.0105052, + "balance_loss_clip": 1.06798172, + "balance_loss_mlp": 1.0319953, + "epoch": 0.05543197258462093, + "flos": 23477894087040.0, + "grad_norm": 1.986771632644713, + "language_loss": 0.80085182, + "learning_rate": 3.993249999908852e-06, + "loss": 0.82397032, + "num_input_tokens_seen": 9995175, + "step": 461, + "time_per_iteration": 2.8012495040893555 + }, + { + "auxiliary_loss_clip": 0.01295209, + "auxiliary_loss_mlp": 0.01053147, + "balance_loss_clip": 1.08127666, + "balance_loss_mlp": 1.03478897, + "epoch": 0.05555221547526003, + "flos": 18624603024000.0, + "grad_norm": 1.893306292578156, + "language_loss": 0.87272042, + "learning_rate": 3.993185903693384e-06, + "loss": 0.89620399, + "num_input_tokens_seen": 10011975, + "step": 462, + "time_per_iteration": 2.73835825920105 + }, + { + "auxiliary_loss_clip": 0.01269646, + "auxiliary_loss_mlp": 0.01051822, + "balance_loss_clip": 1.07604909, + "balance_loss_mlp": 1.03466821, + "epoch": 0.05567245836589912, + "flos": 23587098410880.0, + "grad_norm": 1.9919112723571157, + "language_loss": 0.8232094, + "learning_rate": 3.9931215051131995e-06, + "loss": 0.84642404, + "num_input_tokens_seen": 10032620, + "step": 463, + "time_per_iteration": 4.684478044509888 + }, + { + "auxiliary_loss_clip": 0.01273212, + "auxiliary_loss_mlp": 0.01061554, + "balance_loss_clip": 1.07411122, + "balance_loss_mlp": 1.04481697, + "epoch": 0.05579270125653821, + "flos": 27746667129600.0, + "grad_norm": 1.6462848129133911, + "language_loss": 0.80224693, + "learning_rate": 3.993056804178068e-06, + "loss": 0.82559454, + "num_input_tokens_seen": 10054165, + "step": 464, + "time_per_iteration": 4.684109926223755 + }, + { + "auxiliary_loss_clip": 0.01242786, + "auxiliary_loss_mlp": 0.01045198, + "balance_loss_clip": 1.0714345, + "balance_loss_mlp": 1.02703023, + "epoch": 0.0559129441471773, + "flos": 27014161075200.0, + "grad_norm": 2.125961585119213, + "language_loss": 0.8450768, + "learning_rate": 3.992991800897803e-06, + "loss": 0.8679567, + "num_input_tokens_seen": 10073970, + "step": 465, + "time_per_iteration": 2.9637796878814697 + }, + { + "auxiliary_loss_clip": 0.01294201, + "auxiliary_loss_mlp": 0.01048473, + "balance_loss_clip": 1.08188462, + "balance_loss_mlp": 1.03050864, + "epoch": 0.05603318703781639, + "flos": 15229787794560.0, + "grad_norm": 2.2381214716624873, + "language_loss": 0.90165424, + "learning_rate": 3.9929264952822665e-06, + "loss": 0.92508101, + "num_input_tokens_seen": 10091505, + "step": 466, + "time_per_iteration": 2.716233968734741 + }, + { + "auxiliary_loss_clip": 0.01286184, + "auxiliary_loss_mlp": 0.01043482, + "balance_loss_clip": 1.07673478, + "balance_loss_mlp": 1.02524257, + "epoch": 0.05615342992845548, + "flos": 22266482976000.0, + "grad_norm": 2.185959521241165, + "language_loss": 0.88445622, + "learning_rate": 3.992860887341366e-06, + "loss": 0.90775287, + "num_input_tokens_seen": 10109675, + "step": 467, + "time_per_iteration": 2.713494300842285 + }, + { + "auxiliary_loss_clip": 0.01250179, + "auxiliary_loss_mlp": 0.01055193, + "balance_loss_clip": 1.07196939, + "balance_loss_mlp": 1.03728795, + "epoch": 0.056273672819094574, + "flos": 23584979508480.0, + "grad_norm": 3.219256755410986, + "language_loss": 0.80970126, + "learning_rate": 3.992794977085052e-06, + "loss": 0.83275497, + "num_input_tokens_seen": 10127675, + "step": 468, + "time_per_iteration": 2.850949764251709 + }, + { + "auxiliary_loss_clip": 0.01260407, + "auxiliary_loss_mlp": 0.01055505, + "balance_loss_clip": 1.07588339, + "balance_loss_mlp": 1.03719497, + "epoch": 0.056393915709733664, + "flos": 19858708552320.0, + "grad_norm": 1.9948277068651883, + "language_loss": 0.84939659, + "learning_rate": 3.992728764523326e-06, + "loss": 0.87255567, + "num_input_tokens_seen": 10146620, + "step": 469, + "time_per_iteration": 2.8037917613983154 + }, + { + "auxiliary_loss_clip": 0.01276443, + "auxiliary_loss_mlp": 0.01053541, + "balance_loss_clip": 1.08087468, + "balance_loss_mlp": 1.03540957, + "epoch": 0.05651415860037275, + "flos": 22163779013760.0, + "grad_norm": 1.9092408841372344, + "language_loss": 0.80663896, + "learning_rate": 3.99266224966623e-06, + "loss": 0.82993877, + "num_input_tokens_seen": 10167535, + "step": 470, + "time_per_iteration": 2.806256055831909 + }, + { + "auxiliary_loss_clip": 0.01264742, + "auxiliary_loss_mlp": 0.01055209, + "balance_loss_clip": 1.07908821, + "balance_loss_mlp": 1.03716123, + "epoch": 0.05663440149101184, + "flos": 19463548055040.0, + "grad_norm": 1.9053631404787, + "language_loss": 0.88084108, + "learning_rate": 3.992595432523855e-06, + "loss": 0.90404058, + "num_input_tokens_seen": 10184825, + "step": 471, + "time_per_iteration": 2.7543437480926514 + }, + { + "auxiliary_loss_clip": 0.01245722, + "auxiliary_loss_mlp": 0.01053034, + "balance_loss_clip": 1.07299256, + "balance_loss_mlp": 1.0357964, + "epoch": 0.05675464438165093, + "flos": 22670226823680.0, + "grad_norm": 2.1643810010184046, + "language_loss": 0.85927927, + "learning_rate": 3.992528313106338e-06, + "loss": 0.88226682, + "num_input_tokens_seen": 10203025, + "step": 472, + "time_per_iteration": 2.8420464992523193 + }, + { + "auxiliary_loss_clip": 0.01296174, + "auxiliary_loss_mlp": 0.00890657, + "balance_loss_clip": 1.08350396, + "balance_loss_mlp": 1.00043225, + "epoch": 0.05687488727229002, + "flos": 16901177495040.0, + "grad_norm": 2.8926947275988426, + "language_loss": 0.82296503, + "learning_rate": 3.9924608914238595e-06, + "loss": 0.84483337, + "num_input_tokens_seen": 10218020, + "step": 473, + "time_per_iteration": 2.7435660362243652 + }, + { + "auxiliary_loss_clip": 0.01284214, + "auxiliary_loss_mlp": 0.01049431, + "balance_loss_clip": 1.0810163, + "balance_loss_mlp": 1.03023863, + "epoch": 0.05699513016292912, + "flos": 29168980945920.0, + "grad_norm": 2.469637209700021, + "language_loss": 0.84011793, + "learning_rate": 3.992393167486648e-06, + "loss": 0.86345428, + "num_input_tokens_seen": 10237170, + "step": 474, + "time_per_iteration": 2.85355281829834 + }, + { + "auxiliary_loss_clip": 0.01294256, + "auxiliary_loss_mlp": 0.0104998, + "balance_loss_clip": 1.08127069, + "balance_loss_mlp": 1.03190827, + "epoch": 0.05711537305356821, + "flos": 18916197632640.0, + "grad_norm": 2.5183468816616226, + "language_loss": 0.80369401, + "learning_rate": 3.992325141304977e-06, + "loss": 0.8271364, + "num_input_tokens_seen": 10255125, + "step": 475, + "time_per_iteration": 2.7190775871276855 + }, + { + "auxiliary_loss_clip": 0.01242869, + "auxiliary_loss_mlp": 0.01049445, + "balance_loss_clip": 1.07147229, + "balance_loss_mlp": 1.03131342, + "epoch": 0.0572356159442073, + "flos": 26758979879040.0, + "grad_norm": 2.744632694027087, + "language_loss": 0.86626816, + "learning_rate": 3.992256812889166e-06, + "loss": 0.88919133, + "num_input_tokens_seen": 10271230, + "step": 476, + "time_per_iteration": 2.9577910900115967 + }, + { + "auxiliary_loss_clip": 0.01298779, + "auxiliary_loss_mlp": 0.01049444, + "balance_loss_clip": 1.08494616, + "balance_loss_mlp": 1.03162205, + "epoch": 0.05735585883484639, + "flos": 35116146840960.0, + "grad_norm": 2.0768164904418, + "language_loss": 0.77068853, + "learning_rate": 3.992188182249582e-06, + "loss": 0.7941708, + "num_input_tokens_seen": 10293125, + "step": 477, + "time_per_iteration": 2.8433127403259277 + }, + { + "auxiliary_loss_clip": 0.0127179, + "auxiliary_loss_mlp": 0.01055314, + "balance_loss_clip": 1.0808084, + "balance_loss_mlp": 1.03538191, + "epoch": 0.05747610172548548, + "flos": 18734381965440.0, + "grad_norm": 2.5895462084414707, + "language_loss": 0.90737438, + "learning_rate": 3.992119249396633e-06, + "loss": 0.93064547, + "num_input_tokens_seen": 10311810, + "step": 478, + "time_per_iteration": 2.7954154014587402 + }, + { + "auxiliary_loss_clip": 0.01259246, + "auxiliary_loss_mlp": 0.00890455, + "balance_loss_clip": 1.07141161, + "balance_loss_mlp": 1.00048852, + "epoch": 0.05759634461612457, + "flos": 27964752554880.0, + "grad_norm": 2.809340756593063, + "language_loss": 0.81819785, + "learning_rate": 3.992050014340778e-06, + "loss": 0.83969486, + "num_input_tokens_seen": 10332165, + "step": 479, + "time_per_iteration": 2.924403429031372 + }, + { + "auxiliary_loss_clip": 0.01200941, + "auxiliary_loss_mlp": 0.01028334, + "balance_loss_clip": 1.06873274, + "balance_loss_mlp": 1.01574564, + "epoch": 0.057716587506763666, + "flos": 69292009405440.0, + "grad_norm": 0.8481748768811833, + "language_loss": 0.5503574, + "learning_rate": 3.99198047709252e-06, + "loss": 0.57265013, + "num_input_tokens_seen": 10393685, + "step": 480, + "time_per_iteration": 3.397468328475952 + }, + { + "auxiliary_loss_clip": 0.01254951, + "auxiliary_loss_mlp": 0.0105096, + "balance_loss_clip": 1.07028031, + "balance_loss_mlp": 1.03179145, + "epoch": 0.057836830397402755, + "flos": 25009196745600.0, + "grad_norm": 1.8670484183732456, + "language_loss": 0.78963649, + "learning_rate": 3.991910637662408e-06, + "loss": 0.81269562, + "num_input_tokens_seen": 10413975, + "step": 481, + "time_per_iteration": 2.797656536102295 + }, + { + "auxiliary_loss_clip": 0.01290497, + "auxiliary_loss_mlp": 0.01048207, + "balance_loss_clip": 1.07987094, + "balance_loss_mlp": 1.0306828, + "epoch": 0.057957073288041845, + "flos": 25593894334080.0, + "grad_norm": 2.5788800929792552, + "language_loss": 0.80702841, + "learning_rate": 3.9918404960610355e-06, + "loss": 0.83041549, + "num_input_tokens_seen": 10433005, + "step": 482, + "time_per_iteration": 2.840378761291504 + }, + { + "auxiliary_loss_clip": 0.01288594, + "auxiliary_loss_mlp": 0.01056073, + "balance_loss_clip": 1.08071232, + "balance_loss_mlp": 1.038311, + "epoch": 0.058077316178680935, + "flos": 20777411733120.0, + "grad_norm": 2.2845890360851002, + "language_loss": 0.77640003, + "learning_rate": 3.991770052299043e-06, + "loss": 0.79984671, + "num_input_tokens_seen": 10451235, + "step": 483, + "time_per_iteration": 2.7974371910095215 + }, + { + "auxiliary_loss_clip": 0.01269929, + "auxiliary_loss_mlp": 0.0104345, + "balance_loss_clip": 1.07463062, + "balance_loss_mlp": 1.02565205, + "epoch": 0.058197559069320025, + "flos": 18916484941440.0, + "grad_norm": 2.2175254713688135, + "language_loss": 0.87668091, + "learning_rate": 3.991699306387118e-06, + "loss": 0.89981467, + "num_input_tokens_seen": 10469705, + "step": 484, + "time_per_iteration": 2.9872634410858154 + }, + { + "auxiliary_loss_clip": 0.01284858, + "auxiliary_loss_mlp": 0.01062422, + "balance_loss_clip": 1.07950282, + "balance_loss_mlp": 1.04471946, + "epoch": 0.058317801959959115, + "flos": 24863327614080.0, + "grad_norm": 2.014939067016797, + "language_loss": 0.78282583, + "learning_rate": 3.991628258335991e-06, + "loss": 0.80629861, + "num_input_tokens_seen": 10491910, + "step": 485, + "time_per_iteration": 2.8621816635131836 + }, + { + "auxiliary_loss_clip": 0.01242617, + "auxiliary_loss_mlp": 0.01050592, + "balance_loss_clip": 1.06606483, + "balance_loss_mlp": 1.03175688, + "epoch": 0.05843804485059821, + "flos": 23257977068160.0, + "grad_norm": 3.5736330854093694, + "language_loss": 0.87617999, + "learning_rate": 3.991556908156442e-06, + "loss": 0.89911211, + "num_input_tokens_seen": 10508435, + "step": 486, + "time_per_iteration": 2.983043670654297 + }, + { + "auxiliary_loss_clip": 0.01268626, + "auxiliary_loss_mlp": 0.0105517, + "balance_loss_clip": 1.07656837, + "balance_loss_mlp": 1.03709793, + "epoch": 0.0585582877412373, + "flos": 23150532510720.0, + "grad_norm": 5.266564876198789, + "language_loss": 0.88012028, + "learning_rate": 3.9914852558592914e-06, + "loss": 0.90335822, + "num_input_tokens_seen": 10529485, + "step": 487, + "time_per_iteration": 2.966735363006592 + }, + { + "auxiliary_loss_clip": 0.01279734, + "auxiliary_loss_mlp": 0.01050157, + "balance_loss_clip": 1.07947874, + "balance_loss_mlp": 1.03216851, + "epoch": 0.05867853063187639, + "flos": 23506406507520.0, + "grad_norm": 3.7769942542967856, + "language_loss": 0.81461543, + "learning_rate": 3.991413301455413e-06, + "loss": 0.83791435, + "num_input_tokens_seen": 10545935, + "step": 488, + "time_per_iteration": 2.895078420639038 + }, + { + "auxiliary_loss_clip": 0.01252163, + "auxiliary_loss_mlp": 0.0104749, + "balance_loss_clip": 1.07002902, + "balance_loss_mlp": 1.03056204, + "epoch": 0.05879877352251548, + "flos": 29495803818240.0, + "grad_norm": 3.8706663847159324, + "language_loss": 0.77979171, + "learning_rate": 3.991341044955719e-06, + "loss": 0.80278826, + "num_input_tokens_seen": 10565690, + "step": 489, + "time_per_iteration": 5.902601003646851 + }, + { + "auxiliary_loss_clip": 0.01279025, + "auxiliary_loss_mlp": 0.0089137, + "balance_loss_clip": 1.07653117, + "balance_loss_mlp": 1.00057626, + "epoch": 0.05891901641315457, + "flos": 20157485880960.0, + "grad_norm": 2.101190101829867, + "language_loss": 0.81723285, + "learning_rate": 3.991268486371172e-06, + "loss": 0.83893681, + "num_input_tokens_seen": 10584245, + "step": 490, + "time_per_iteration": 3.7425692081451416 + }, + { + "auxiliary_loss_clip": 0.01268204, + "auxiliary_loss_mlp": 0.01048525, + "balance_loss_clip": 1.07309282, + "balance_loss_mlp": 1.02964199, + "epoch": 0.05903925930379366, + "flos": 24644200694400.0, + "grad_norm": 2.5065234796896028, + "language_loss": 0.87830818, + "learning_rate": 3.991195625712779e-06, + "loss": 0.90147543, + "num_input_tokens_seen": 10601210, + "step": 491, + "time_per_iteration": 2.855098009109497 + }, + { + "auxiliary_loss_clip": 0.01290188, + "auxiliary_loss_mlp": 0.01061194, + "balance_loss_clip": 1.08102429, + "balance_loss_mlp": 1.042907, + "epoch": 0.05915950219443276, + "flos": 21250391045760.0, + "grad_norm": 2.2335328954356948, + "language_loss": 0.81747961, + "learning_rate": 3.991122462991592e-06, + "loss": 0.8409934, + "num_input_tokens_seen": 10620730, + "step": 492, + "time_per_iteration": 2.7005257606506348 + }, + { + "auxiliary_loss_clip": 0.01294268, + "auxiliary_loss_mlp": 0.0105058, + "balance_loss_clip": 1.08047414, + "balance_loss_mlp": 1.03195953, + "epoch": 0.05927974508507185, + "flos": 9902727319680.0, + "grad_norm": 3.1754426963257454, + "language_loss": 0.80932027, + "learning_rate": 3.991048998218712e-06, + "loss": 0.83276874, + "num_input_tokens_seen": 10634035, + "step": 493, + "time_per_iteration": 2.754230499267578 + }, + { + "auxiliary_loss_clip": 0.01276709, + "auxiliary_loss_mlp": 0.01059606, + "balance_loss_clip": 1.07618582, + "balance_loss_mlp": 1.04098535, + "epoch": 0.05939998797571094, + "flos": 18259499232000.0, + "grad_norm": 2.624998075745842, + "language_loss": 0.76288533, + "learning_rate": 3.990975231405281e-06, + "loss": 0.78624851, + "num_input_tokens_seen": 10652485, + "step": 494, + "time_per_iteration": 2.807300567626953 + }, + { + "auxiliary_loss_clip": 0.01276207, + "auxiliary_loss_mlp": 0.0106352, + "balance_loss_clip": 1.07788742, + "balance_loss_mlp": 1.04441106, + "epoch": 0.05952023086635003, + "flos": 28256598558720.0, + "grad_norm": 2.2136399331874297, + "language_loss": 0.78705621, + "learning_rate": 3.990901162562491e-06, + "loss": 0.81045353, + "num_input_tokens_seen": 10673175, + "step": 495, + "time_per_iteration": 2.811816930770874 + }, + { + "auxiliary_loss_clip": 0.01249232, + "auxiliary_loss_mlp": 0.0089116, + "balance_loss_clip": 1.06807709, + "balance_loss_mlp": 1.00057387, + "epoch": 0.05964047375698912, + "flos": 14902498045440.0, + "grad_norm": 3.2028394190661804, + "language_loss": 0.90594816, + "learning_rate": 3.9908267917015765e-06, + "loss": 0.92735207, + "num_input_tokens_seen": 10691235, + "step": 496, + "time_per_iteration": 2.7993199825286865 + }, + { + "auxiliary_loss_clip": 0.01269509, + "auxiliary_loss_mlp": 0.01054405, + "balance_loss_clip": 1.07368767, + "balance_loss_mlp": 1.03660727, + "epoch": 0.059760716647628206, + "flos": 23185581206400.0, + "grad_norm": 2.236076891764289, + "language_loss": 0.93302572, + "learning_rate": 3.990752118833821e-06, + "loss": 0.95626485, + "num_input_tokens_seen": 10708675, + "step": 497, + "time_per_iteration": 2.8143718242645264 + }, + { + "auxiliary_loss_clip": 0.01290941, + "auxiliary_loss_mlp": 0.01047447, + "balance_loss_clip": 1.08192754, + "balance_loss_mlp": 1.02975595, + "epoch": 0.0598809595382673, + "flos": 22746968231040.0, + "grad_norm": 1.796234546568289, + "language_loss": 0.78133214, + "learning_rate": 3.990677143970553e-06, + "loss": 0.80471599, + "num_input_tokens_seen": 10729485, + "step": 498, + "time_per_iteration": 2.8120274543762207 + }, + { + "auxiliary_loss_clip": 0.01253753, + "auxiliary_loss_mlp": 0.01049213, + "balance_loss_clip": 1.07664049, + "balance_loss_mlp": 1.0311172, + "epoch": 0.06000120242890639, + "flos": 22127221946880.0, + "grad_norm": 2.19262440555516, + "language_loss": 0.81302524, + "learning_rate": 3.990601867123144e-06, + "loss": 0.83605492, + "num_input_tokens_seen": 10749210, + "step": 499, + "time_per_iteration": 2.8123607635498047 + }, + { + "auxiliary_loss_clip": 0.01245587, + "auxiliary_loss_mlp": 0.01060277, + "balance_loss_clip": 1.07564461, + "balance_loss_mlp": 1.04265761, + "epoch": 0.06012144531954548, + "flos": 19171773878400.0, + "grad_norm": 2.1756782243377697, + "language_loss": 0.84645486, + "learning_rate": 3.990526288303014e-06, + "loss": 0.86951351, + "num_input_tokens_seen": 10768000, + "step": 500, + "time_per_iteration": 2.881009101867676 + }, + { + "auxiliary_loss_clip": 0.01262349, + "auxiliary_loss_mlp": 0.00890163, + "balance_loss_clip": 1.07403064, + "balance_loss_mlp": 1.00063741, + "epoch": 0.06024168821018457, + "flos": 22783345729920.0, + "grad_norm": 5.980105493584853, + "language_loss": 0.90900725, + "learning_rate": 3.9904504075216295e-06, + "loss": 0.93053234, + "num_input_tokens_seen": 10788760, + "step": 501, + "time_per_iteration": 2.814502239227295 + }, + { + "auxiliary_loss_clip": 0.01251546, + "auxiliary_loss_mlp": 0.01062245, + "balance_loss_clip": 1.06875134, + "balance_loss_mlp": 1.04333794, + "epoch": 0.06036193110082366, + "flos": 18770687637120.0, + "grad_norm": 3.15224478081202, + "language_loss": 0.94251251, + "learning_rate": 3.990374224790501e-06, + "loss": 0.96565032, + "num_input_tokens_seen": 10806965, + "step": 502, + "time_per_iteration": 2.910358190536499 + }, + { + "auxiliary_loss_clip": 0.01268089, + "auxiliary_loss_mlp": 0.01056749, + "balance_loss_clip": 1.07790112, + "balance_loss_mlp": 1.03905821, + "epoch": 0.06048217399146275, + "flos": 17201570935680.0, + "grad_norm": 1.9928420533249198, + "language_loss": 0.71304476, + "learning_rate": 3.990297740121185e-06, + "loss": 0.7362932, + "num_input_tokens_seen": 10824900, + "step": 503, + "time_per_iteration": 2.7835512161254883 + }, + { + "auxiliary_loss_clip": 0.0127081, + "auxiliary_loss_mlp": 0.00890374, + "balance_loss_clip": 1.07488883, + "balance_loss_mlp": 1.00058913, + "epoch": 0.06060241688210185, + "flos": 24024131187840.0, + "grad_norm": 2.079969600374517, + "language_loss": 0.78028995, + "learning_rate": 3.990220953525284e-06, + "loss": 0.80190182, + "num_input_tokens_seen": 10842010, + "step": 504, + "time_per_iteration": 2.781562566757202 + }, + { + "auxiliary_loss_clip": 0.01250015, + "auxiliary_loss_mlp": 0.0105655, + "balance_loss_clip": 1.06955719, + "balance_loss_mlp": 1.0369159, + "epoch": 0.06072265977274094, + "flos": 14611190745600.0, + "grad_norm": 2.67475906414214, + "language_loss": 0.74717784, + "learning_rate": 3.9901438650144465e-06, + "loss": 0.77024341, + "num_input_tokens_seen": 10858260, + "step": 505, + "time_per_iteration": 2.7634875774383545 + }, + { + "auxiliary_loss_clip": 0.01265072, + "auxiliary_loss_mlp": 0.01049102, + "balance_loss_clip": 1.07233775, + "balance_loss_mlp": 1.03145885, + "epoch": 0.06084290266338003, + "flos": 20558284813440.0, + "grad_norm": 9.273620684262074, + "language_loss": 0.91756523, + "learning_rate": 3.990066474600367e-06, + "loss": 0.94070691, + "num_input_tokens_seen": 10876230, + "step": 506, + "time_per_iteration": 2.7068076133728027 + }, + { + "auxiliary_loss_clip": 0.01252385, + "auxiliary_loss_mlp": 0.0105309, + "balance_loss_clip": 1.06671441, + "balance_loss_mlp": 1.03535223, + "epoch": 0.06096314555401912, + "flos": 22309217182080.0, + "grad_norm": 1.8383669187992837, + "language_loss": 0.68106198, + "learning_rate": 3.989988782294786e-06, + "loss": 0.70411682, + "num_input_tokens_seen": 10896320, + "step": 507, + "time_per_iteration": 2.809418201446533 + }, + { + "auxiliary_loss_clip": 0.01224111, + "auxiliary_loss_mlp": 0.01054401, + "balance_loss_clip": 1.06388247, + "balance_loss_mlp": 1.03777099, + "epoch": 0.06108338844465821, + "flos": 19131374056320.0, + "grad_norm": 1.750910769148142, + "language_loss": 0.94912148, + "learning_rate": 3.989910788109489e-06, + "loss": 0.97190666, + "num_input_tokens_seen": 10912970, + "step": 508, + "time_per_iteration": 2.8578243255615234 + }, + { + "auxiliary_loss_clip": 0.0125273, + "auxiliary_loss_mlp": 0.01056866, + "balance_loss_clip": 1.07312822, + "balance_loss_mlp": 1.03860307, + "epoch": 0.0612036313352973, + "flos": 33584018169600.0, + "grad_norm": 3.157953749713791, + "language_loss": 0.75462145, + "learning_rate": 3.989832492056307e-06, + "loss": 0.77771747, + "num_input_tokens_seen": 10933995, + "step": 509, + "time_per_iteration": 3.0538299083709717 + }, + { + "auxiliary_loss_clip": 0.01276919, + "auxiliary_loss_mlp": 0.01044891, + "balance_loss_clip": 1.07920933, + "balance_loss_mlp": 1.02664042, + "epoch": 0.06132387422593639, + "flos": 27490552179840.0, + "grad_norm": 2.410799717893846, + "language_loss": 0.81028569, + "learning_rate": 3.989753894147119e-06, + "loss": 0.83350372, + "num_input_tokens_seen": 10954120, + "step": 510, + "time_per_iteration": 2.8311586380004883 + }, + { + "auxiliary_loss_clip": 0.01270409, + "auxiliary_loss_mlp": 0.01053433, + "balance_loss_clip": 1.07810771, + "balance_loss_mlp": 1.03637481, + "epoch": 0.061444117116575485, + "flos": 25885057979520.0, + "grad_norm": 1.945716081728192, + "language_loss": 0.79903638, + "learning_rate": 3.989674994393846e-06, + "loss": 0.8222748, + "num_input_tokens_seen": 10973595, + "step": 511, + "time_per_iteration": 2.8415889739990234 + }, + { + "auxiliary_loss_clip": 0.01273563, + "auxiliary_loss_mlp": 0.01045055, + "balance_loss_clip": 1.07899487, + "balance_loss_mlp": 1.02802014, + "epoch": 0.061564360007214575, + "flos": 28512031150080.0, + "grad_norm": 2.1797089585357265, + "language_loss": 0.93873954, + "learning_rate": 3.98959579280846e-06, + "loss": 0.96192575, + "num_input_tokens_seen": 10991995, + "step": 512, + "time_per_iteration": 2.8526206016540527 + }, + { + "auxiliary_loss_clip": 0.01227265, + "auxiliary_loss_mlp": 0.0106396, + "balance_loss_clip": 1.07263184, + "balance_loss_mlp": 1.04661536, + "epoch": 0.061684602897853665, + "flos": 12094355652480.0, + "grad_norm": 2.0100558309114955, + "language_loss": 0.83186877, + "learning_rate": 3.989516289402973e-06, + "loss": 0.85478103, + "num_input_tokens_seen": 11007625, + "step": 513, + "time_per_iteration": 2.7766246795654297 + }, + { + "auxiliary_loss_clip": 0.01203728, + "auxiliary_loss_mlp": 0.01041078, + "balance_loss_clip": 1.05818462, + "balance_loss_mlp": 1.02364945, + "epoch": 0.061804845788492754, + "flos": 19532639865600.0, + "grad_norm": 2.43181620843611, + "language_loss": 0.80878383, + "learning_rate": 3.989436484189447e-06, + "loss": 0.83123195, + "num_input_tokens_seen": 11025570, + "step": 514, + "time_per_iteration": 2.9253275394439697 + }, + { + "auxiliary_loss_clip": 0.01269572, + "auxiliary_loss_mlp": 0.01049804, + "balance_loss_clip": 1.07237279, + "balance_loss_mlp": 1.0322082, + "epoch": 0.061925088679131844, + "flos": 15341111020800.0, + "grad_norm": 2.942566191340735, + "language_loss": 0.80800718, + "learning_rate": 3.9893563771799885e-06, + "loss": 0.83120096, + "num_input_tokens_seen": 11042045, + "step": 515, + "time_per_iteration": 4.732289791107178 + }, + { + "auxiliary_loss_clip": 0.01286129, + "auxiliary_loss_mlp": 0.01054749, + "balance_loss_clip": 1.0781225, + "balance_loss_mlp": 1.03697515, + "epoch": 0.062045331569770934, + "flos": 25919927107200.0, + "grad_norm": 2.080989205699523, + "language_loss": 0.86334455, + "learning_rate": 3.989275968386749e-06, + "loss": 0.88675332, + "num_input_tokens_seen": 11059955, + "step": 516, + "time_per_iteration": 3.599297285079956 + }, + { + "auxiliary_loss_clip": 0.01241881, + "auxiliary_loss_mlp": 0.01049718, + "balance_loss_clip": 1.0659616, + "balance_loss_mlp": 1.03243315, + "epoch": 0.06216557446041003, + "flos": 28110621686400.0, + "grad_norm": 2.086526332319496, + "language_loss": 0.76867074, + "learning_rate": 3.989195257821926e-06, + "loss": 0.79158676, + "num_input_tokens_seen": 11078440, + "step": 517, + "time_per_iteration": 2.807288885116577 + }, + { + "auxiliary_loss_clip": 0.01264065, + "auxiliary_loss_mlp": 0.01054878, + "balance_loss_clip": 1.07939911, + "balance_loss_mlp": 1.03694844, + "epoch": 0.06228581735104912, + "flos": 23478181395840.0, + "grad_norm": 2.0212532694571297, + "language_loss": 0.84547454, + "learning_rate": 3.989114245497765e-06, + "loss": 0.86866397, + "num_input_tokens_seen": 11098240, + "step": 518, + "time_per_iteration": 2.795191526412964 + }, + { + "auxiliary_loss_clip": 0.01277322, + "auxiliary_loss_mlp": 0.01044398, + "balance_loss_clip": 1.0750463, + "balance_loss_mlp": 1.02667105, + "epoch": 0.06240606024168821, + "flos": 15195205975680.0, + "grad_norm": 2.096987537787674, + "language_loss": 0.95196342, + "learning_rate": 3.989032931426554e-06, + "loss": 0.97518063, + "num_input_tokens_seen": 11115395, + "step": 519, + "time_per_iteration": 2.718248128890991 + }, + { + "auxiliary_loss_clip": 0.01259143, + "auxiliary_loss_mlp": 0.01052258, + "balance_loss_clip": 1.07425117, + "balance_loss_mlp": 1.03546143, + "epoch": 0.06252630313232731, + "flos": 20631829910400.0, + "grad_norm": 2.153373943415168, + "language_loss": 0.86392742, + "learning_rate": 3.9889513156206295e-06, + "loss": 0.88704145, + "num_input_tokens_seen": 11134835, + "step": 520, + "time_per_iteration": 2.8002161979675293 + }, + { + "auxiliary_loss_clip": 0.01249713, + "auxiliary_loss_mlp": 0.01048328, + "balance_loss_clip": 1.06967378, + "balance_loss_mlp": 1.02946949, + "epoch": 0.06264654602296639, + "flos": 20778058177920.0, + "grad_norm": 2.871438402932833, + "language_loss": 0.73947215, + "learning_rate": 3.988869398092371e-06, + "loss": 0.76245248, + "num_input_tokens_seen": 11154745, + "step": 521, + "time_per_iteration": 2.7786343097686768 + }, + { + "auxiliary_loss_clip": 0.01262632, + "auxiliary_loss_mlp": 0.01057299, + "balance_loss_clip": 1.07482696, + "balance_loss_mlp": 1.03903615, + "epoch": 0.06276678891360549, + "flos": 29605798241280.0, + "grad_norm": 2.669158721223016, + "language_loss": 0.78988087, + "learning_rate": 3.988787178854206e-06, + "loss": 0.81308025, + "num_input_tokens_seen": 11174280, + "step": 522, + "time_per_iteration": 2.8524160385131836 + }, + { + "auxiliary_loss_clip": 0.01288619, + "auxiliary_loss_mlp": 0.0106184, + "balance_loss_clip": 1.08206129, + "balance_loss_mlp": 1.0436604, + "epoch": 0.06288703180424457, + "flos": 22126288193280.0, + "grad_norm": 2.266511139328637, + "language_loss": 0.87846327, + "learning_rate": 3.988704657918608e-06, + "loss": 0.90196788, + "num_input_tokens_seen": 11193340, + "step": 523, + "time_per_iteration": 2.6736457347869873 + }, + { + "auxiliary_loss_clip": 0.01273025, + "auxiliary_loss_mlp": 0.0106431, + "balance_loss_clip": 1.07908785, + "balance_loss_mlp": 1.04686952, + "epoch": 0.06300727469488367, + "flos": 14976689587200.0, + "grad_norm": 2.6045794598867924, + "language_loss": 0.79885876, + "learning_rate": 3.988621835298094e-06, + "loss": 0.82223207, + "num_input_tokens_seen": 11210555, + "step": 524, + "time_per_iteration": 2.7782108783721924 + }, + { + "auxiliary_loss_clip": 0.01281585, + "auxiliary_loss_mlp": 0.01051646, + "balance_loss_clip": 1.07768416, + "balance_loss_mlp": 1.0342052, + "epoch": 0.06312751758552275, + "flos": 24535391420160.0, + "grad_norm": 2.4004906197393727, + "language_loss": 0.91791099, + "learning_rate": 3.988538711005229e-06, + "loss": 0.94124329, + "num_input_tokens_seen": 11230010, + "step": 525, + "time_per_iteration": 2.7472128868103027 + }, + { + "auxiliary_loss_clip": 0.01265277, + "auxiliary_loss_mlp": 0.01050638, + "balance_loss_clip": 1.07716203, + "balance_loss_mlp": 1.03359139, + "epoch": 0.06324776047616185, + "flos": 21507008785920.0, + "grad_norm": 2.5356924367253364, + "language_loss": 0.88201439, + "learning_rate": 3.988455285052622e-06, + "loss": 0.90517354, + "num_input_tokens_seen": 11246190, + "step": 526, + "time_per_iteration": 2.8135526180267334 + }, + { + "auxiliary_loss_clip": 0.01268448, + "auxiliary_loss_mlp": 0.01049997, + "balance_loss_clip": 1.07533967, + "balance_loss_mlp": 1.03135324, + "epoch": 0.06336800336680094, + "flos": 21688034353920.0, + "grad_norm": 2.088030135535211, + "language_loss": 0.84088314, + "learning_rate": 3.98837155745293e-06, + "loss": 0.86406755, + "num_input_tokens_seen": 11264230, + "step": 527, + "time_per_iteration": 2.847708225250244 + }, + { + "auxiliary_loss_clip": 0.01273952, + "auxiliary_loss_mlp": 0.0105435, + "balance_loss_clip": 1.08002853, + "balance_loss_mlp": 1.03693366, + "epoch": 0.06348824625744003, + "flos": 19500895221120.0, + "grad_norm": 2.2206682538188898, + "language_loss": 0.75766289, + "learning_rate": 3.988287528218854e-06, + "loss": 0.7809459, + "num_input_tokens_seen": 11283015, + "step": 528, + "time_per_iteration": 2.8481240272521973 + }, + { + "auxiliary_loss_clip": 0.01273283, + "auxiliary_loss_mlp": 0.01053865, + "balance_loss_clip": 1.0806098, + "balance_loss_mlp": 1.03719985, + "epoch": 0.06360848914807912, + "flos": 15481233976320.0, + "grad_norm": 2.2029722470995576, + "language_loss": 0.90553463, + "learning_rate": 3.98820319736314e-06, + "loss": 0.92880607, + "num_input_tokens_seen": 11299630, + "step": 529, + "time_per_iteration": 2.6919891834259033 + }, + { + "auxiliary_loss_clip": 0.0124774, + "auxiliary_loss_mlp": 0.01054009, + "balance_loss_clip": 1.07130206, + "balance_loss_mlp": 1.03774846, + "epoch": 0.0637287320387182, + "flos": 20593369422720.0, + "grad_norm": 1.8555425750038026, + "language_loss": 0.85353804, + "learning_rate": 3.988118564898582e-06, + "loss": 0.8765555, + "num_input_tokens_seen": 11319170, + "step": 530, + "time_per_iteration": 2.8543429374694824 + }, + { + "auxiliary_loss_clip": 0.01234244, + "auxiliary_loss_mlp": 0.00891017, + "balance_loss_clip": 1.07044029, + "balance_loss_mlp": 1.00078082, + "epoch": 0.0638489749293573, + "flos": 17412222245760.0, + "grad_norm": 2.4693533517769075, + "language_loss": 0.89459825, + "learning_rate": 3.988033630838019e-06, + "loss": 0.91585088, + "num_input_tokens_seen": 11333210, + "step": 531, + "time_per_iteration": 2.9825119972229004 + }, + { + "auxiliary_loss_clip": 0.0127512, + "auxiliary_loss_mlp": 0.01046097, + "balance_loss_clip": 1.07894969, + "balance_loss_mlp": 1.02838254, + "epoch": 0.0639692178199964, + "flos": 23807661874560.0, + "grad_norm": 2.0737442347562824, + "language_loss": 0.8843829, + "learning_rate": 3.987948395194334e-06, + "loss": 0.90759516, + "num_input_tokens_seen": 11355590, + "step": 532, + "time_per_iteration": 2.7986972332000732 + }, + { + "auxiliary_loss_clip": 0.01260113, + "auxiliary_loss_mlp": 0.01048616, + "balance_loss_clip": 1.07064736, + "balance_loss_mlp": 1.03171229, + "epoch": 0.06408946071063548, + "flos": 18477225521280.0, + "grad_norm": 2.2345135583263183, + "language_loss": 0.7691853, + "learning_rate": 3.987862857980458e-06, + "loss": 0.79227263, + "num_input_tokens_seen": 11371535, + "step": 533, + "time_per_iteration": 2.739337921142578 + }, + { + "auxiliary_loss_clip": 0.01251497, + "auxiliary_loss_mlp": 0.01052667, + "balance_loss_clip": 1.07430196, + "balance_loss_mlp": 1.03515482, + "epoch": 0.06420970360127458, + "flos": 27162220936320.0, + "grad_norm": 2.2092410792364534, + "language_loss": 0.76906455, + "learning_rate": 3.987777019209368e-06, + "loss": 0.79210615, + "num_input_tokens_seen": 11392050, + "step": 534, + "time_per_iteration": 2.8665904998779297 + }, + { + "auxiliary_loss_clip": 0.01279886, + "auxiliary_loss_mlp": 0.01053168, + "balance_loss_clip": 1.07728457, + "balance_loss_mlp": 1.0352869, + "epoch": 0.06432994649191366, + "flos": 23659673840640.0, + "grad_norm": 1.6857840417413976, + "language_loss": 0.81114882, + "learning_rate": 3.987690878894084e-06, + "loss": 0.83447939, + "num_input_tokens_seen": 11411765, + "step": 535, + "time_per_iteration": 2.7629079818725586 + }, + { + "auxiliary_loss_clip": 0.01262208, + "auxiliary_loss_mlp": 0.01052699, + "balance_loss_clip": 1.07564116, + "balance_loss_mlp": 1.03568828, + "epoch": 0.06445018938255276, + "flos": 23403953940480.0, + "grad_norm": 2.3365177800760755, + "language_loss": 0.84983933, + "learning_rate": 3.987604437047673e-06, + "loss": 0.8729884, + "num_input_tokens_seen": 11431565, + "step": 536, + "time_per_iteration": 2.858464241027832 + }, + { + "auxiliary_loss_clip": 0.01268624, + "auxiliary_loss_mlp": 0.01050557, + "balance_loss_clip": 1.0752461, + "balance_loss_mlp": 1.03280711, + "epoch": 0.06457043227319184, + "flos": 19646692525440.0, + "grad_norm": 3.051106420531132, + "language_loss": 0.77666438, + "learning_rate": 3.987517693683251e-06, + "loss": 0.79985619, + "num_input_tokens_seen": 11450140, + "step": 537, + "time_per_iteration": 2.718548536300659 + }, + { + "auxiliary_loss_clip": 0.01256717, + "auxiliary_loss_mlp": 0.01045047, + "balance_loss_clip": 1.07564306, + "balance_loss_mlp": 1.02797663, + "epoch": 0.06469067516383094, + "flos": 16978744915200.0, + "grad_norm": 2.8003241978625093, + "language_loss": 0.95931643, + "learning_rate": 3.9874306488139745e-06, + "loss": 0.98233408, + "num_input_tokens_seen": 11465400, + "step": 538, + "time_per_iteration": 2.7951138019561768 + }, + { + "auxiliary_loss_clip": 0.01247415, + "auxiliary_loss_mlp": 0.01051498, + "balance_loss_clip": 1.07287395, + "balance_loss_mlp": 1.03302073, + "epoch": 0.06481091805447003, + "flos": 23296401642240.0, + "grad_norm": 1.9671453659364377, + "language_loss": 0.88053405, + "learning_rate": 3.987343302453049e-06, + "loss": 0.90352321, + "num_input_tokens_seen": 11486675, + "step": 539, + "time_per_iteration": 2.9844858646392822 + }, + { + "auxiliary_loss_clip": 0.01257633, + "auxiliary_loss_mlp": 0.01050554, + "balance_loss_clip": 1.07461643, + "balance_loss_mlp": 1.03359056, + "epoch": 0.06493116094510912, + "flos": 29172356824320.0, + "grad_norm": 1.8211612197284262, + "language_loss": 0.82378507, + "learning_rate": 3.987255654613724e-06, + "loss": 0.84686697, + "num_input_tokens_seen": 11510440, + "step": 540, + "time_per_iteration": 2.8908231258392334 + }, + { + "auxiliary_loss_clip": 0.01244913, + "auxiliary_loss_mlp": 0.01046897, + "balance_loss_clip": 1.06734085, + "balance_loss_mlp": 1.02890897, + "epoch": 0.06505140383574821, + "flos": 19865065259520.0, + "grad_norm": 2.2320603565526023, + "language_loss": 0.70274568, + "learning_rate": 3.987167705309296e-06, + "loss": 0.72566372, + "num_input_tokens_seen": 11529715, + "step": 541, + "time_per_iteration": 4.764929294586182 + }, + { + "auxiliary_loss_clip": 0.01275774, + "auxiliary_loss_mlp": 0.00890868, + "balance_loss_clip": 1.07734168, + "balance_loss_mlp": 1.00067961, + "epoch": 0.0651716467263873, + "flos": 17924703540480.0, + "grad_norm": 2.048418405628991, + "language_loss": 0.95207107, + "learning_rate": 3.987079454553108e-06, + "loss": 0.97373748, + "num_input_tokens_seen": 11547665, + "step": 542, + "time_per_iteration": 2.715369701385498 + }, + { + "auxiliary_loss_clip": 0.01247302, + "auxiliary_loss_mlp": 0.01047156, + "balance_loss_clip": 1.07519746, + "balance_loss_mlp": 1.02911937, + "epoch": 0.0652918896170264, + "flos": 20842840356480.0, + "grad_norm": 1.9776380247418603, + "language_loss": 0.91182393, + "learning_rate": 3.986990902358546e-06, + "loss": 0.9347685, + "num_input_tokens_seen": 11564605, + "step": 543, + "time_per_iteration": 2.8419432640075684 + }, + { + "auxiliary_loss_clip": 0.01271924, + "auxiliary_loss_mlp": 0.01055669, + "balance_loss_clip": 1.07357168, + "balance_loss_mlp": 1.03748918, + "epoch": 0.06541213250766549, + "flos": 21872507627520.0, + "grad_norm": 2.024123617007609, + "language_loss": 0.93250924, + "learning_rate": 3.986902048739045e-06, + "loss": 0.95578516, + "num_input_tokens_seen": 11584550, + "step": 544, + "time_per_iteration": 2.800537109375 + }, + { + "auxiliary_loss_clip": 0.01259477, + "auxiliary_loss_mlp": 0.01057369, + "balance_loss_clip": 1.07355928, + "balance_loss_mlp": 1.0381999, + "epoch": 0.06553237539830457, + "flos": 23110743219840.0, + "grad_norm": 2.652946294091431, + "language_loss": 0.80185437, + "learning_rate": 3.986812893708082e-06, + "loss": 0.82502282, + "num_input_tokens_seen": 11600740, + "step": 545, + "time_per_iteration": 2.7268996238708496 + }, + { + "auxiliary_loss_clip": 0.01260727, + "auxiliary_loss_mlp": 0.01050808, + "balance_loss_clip": 1.06856382, + "balance_loss_mlp": 1.03157949, + "epoch": 0.06565261828894367, + "flos": 17923769786880.0, + "grad_norm": 2.3267346654412155, + "language_loss": 0.81671131, + "learning_rate": 3.9867234372791826e-06, + "loss": 0.83982658, + "num_input_tokens_seen": 11618695, + "step": 546, + "time_per_iteration": 2.762632131576538 + }, + { + "auxiliary_loss_clip": 0.01272501, + "auxiliary_loss_mlp": 0.01071018, + "balance_loss_clip": 1.07701707, + "balance_loss_mlp": 1.05333972, + "epoch": 0.06577286117958275, + "flos": 22783058421120.0, + "grad_norm": 1.6205615206641493, + "language_loss": 0.87087989, + "learning_rate": 3.986633679465918e-06, + "loss": 0.89431512, + "num_input_tokens_seen": 11638850, + "step": 547, + "time_per_iteration": 2.741506814956665 + }, + { + "auxiliary_loss_clip": 0.01235073, + "auxiliary_loss_mlp": 0.01050038, + "balance_loss_clip": 1.06972408, + "balance_loss_mlp": 1.03246653, + "epoch": 0.06589310407022185, + "flos": 23696194993920.0, + "grad_norm": 2.3036465838738827, + "language_loss": 0.80516356, + "learning_rate": 3.986543620281904e-06, + "loss": 0.82801461, + "num_input_tokens_seen": 11658500, + "step": 548, + "time_per_iteration": 3.0112433433532715 + }, + { + "auxiliary_loss_clip": 0.01242894, + "auxiliary_loss_mlp": 0.0105041, + "balance_loss_clip": 1.06851292, + "balance_loss_mlp": 1.03335094, + "epoch": 0.06601334696086093, + "flos": 26864772410880.0, + "grad_norm": 2.0196755203453933, + "language_loss": 0.91182411, + "learning_rate": 3.986453259740802e-06, + "loss": 0.93475711, + "num_input_tokens_seen": 11676670, + "step": 549, + "time_per_iteration": 2.840754270553589 + }, + { + "auxiliary_loss_clip": 0.0126364, + "auxiliary_loss_mlp": 0.01048352, + "balance_loss_clip": 1.07836771, + "balance_loss_mlp": 1.03024447, + "epoch": 0.06613358985150003, + "flos": 12567694101120.0, + "grad_norm": 2.546596364416662, + "language_loss": 0.79332721, + "learning_rate": 3.986362597856319e-06, + "loss": 0.81644714, + "num_input_tokens_seen": 11693170, + "step": 550, + "time_per_iteration": 2.7690701484680176 + }, + { + "auxiliary_loss_clip": 0.01251541, + "auxiliary_loss_mlp": 0.00892248, + "balance_loss_clip": 1.06897175, + "balance_loss_mlp": 1.00070596, + "epoch": 0.06625383274213913, + "flos": 18332505624960.0, + "grad_norm": 2.569249395894229, + "language_loss": 0.82108867, + "learning_rate": 3.986271634642211e-06, + "loss": 0.84252656, + "num_input_tokens_seen": 11710150, + "step": 551, + "time_per_iteration": 2.7630040645599365 + }, + { + "auxiliary_loss_clip": 0.01282145, + "auxiliary_loss_mlp": 0.01059457, + "balance_loss_clip": 1.07910407, + "balance_loss_mlp": 1.0419457, + "epoch": 0.06637407563277821, + "flos": 15375585098880.0, + "grad_norm": 4.132041454197683, + "language_loss": 0.81471908, + "learning_rate": 3.986180370112274e-06, + "loss": 0.83813512, + "num_input_tokens_seen": 11726670, + "step": 552, + "time_per_iteration": 2.7859740257263184 + }, + { + "auxiliary_loss_clip": 0.01274435, + "auxiliary_loss_mlp": 0.00891311, + "balance_loss_clip": 1.07677293, + "balance_loss_mlp": 1.00070238, + "epoch": 0.0664943185234173, + "flos": 24025244509440.0, + "grad_norm": 1.8656935164888218, + "language_loss": 0.747913, + "learning_rate": 3.986088804280354e-06, + "loss": 0.76957053, + "num_input_tokens_seen": 11746400, + "step": 553, + "time_per_iteration": 2.8351306915283203 + }, + { + "auxiliary_loss_clip": 0.01266463, + "auxiliary_loss_mlp": 0.01046881, + "balance_loss_clip": 1.07745862, + "balance_loss_mlp": 1.02882063, + "epoch": 0.06661456141405639, + "flos": 20957503547520.0, + "grad_norm": 2.8160186321401026, + "language_loss": 0.94058299, + "learning_rate": 3.985996937160342e-06, + "loss": 0.96371645, + "num_input_tokens_seen": 11765590, + "step": 554, + "time_per_iteration": 2.8789448738098145 + }, + { + "auxiliary_loss_clip": 0.01268304, + "auxiliary_loss_mlp": 0.01049374, + "balance_loss_clip": 1.07847714, + "balance_loss_mlp": 1.03130209, + "epoch": 0.06673480430469549, + "flos": 52223953322880.0, + "grad_norm": 2.6806299496234747, + "language_loss": 0.69059861, + "learning_rate": 3.985904768766173e-06, + "loss": 0.7137754, + "num_input_tokens_seen": 11788365, + "step": 555, + "time_per_iteration": 3.0819480419158936 + }, + { + "auxiliary_loss_clip": 0.01245069, + "auxiliary_loss_mlp": 0.01054538, + "balance_loss_clip": 1.07108605, + "balance_loss_mlp": 1.03561974, + "epoch": 0.06685504719533458, + "flos": 16217079995520.0, + "grad_norm": 2.2165255515588766, + "language_loss": 0.76402873, + "learning_rate": 3.98581229911183e-06, + "loss": 0.7870248, + "num_input_tokens_seen": 11807285, + "step": 556, + "time_per_iteration": 2.837153434753418 + }, + { + "auxiliary_loss_clip": 0.01266142, + "auxiliary_loss_mlp": 0.01052302, + "balance_loss_clip": 1.07168365, + "balance_loss_mlp": 1.03513551, + "epoch": 0.06697529008597367, + "flos": 22491535639680.0, + "grad_norm": 2.1533057772659374, + "language_loss": 0.92044878, + "learning_rate": 3.985719528211341e-06, + "loss": 0.9436332, + "num_input_tokens_seen": 11826655, + "step": 557, + "time_per_iteration": 2.7875537872314453 + }, + { + "auxiliary_loss_clip": 0.01190358, + "auxiliary_loss_mlp": 0.01014054, + "balance_loss_clip": 1.07028246, + "balance_loss_mlp": 1.00442171, + "epoch": 0.06709553297661276, + "flos": 62688216936960.0, + "grad_norm": 0.8453221543255592, + "language_loss": 0.62965727, + "learning_rate": 3.985626456078777e-06, + "loss": 0.65170145, + "num_input_tokens_seen": 11891310, + "step": 558, + "time_per_iteration": 3.439613103866577 + }, + { + "auxiliary_loss_clip": 0.0125244, + "auxiliary_loss_mlp": 0.01070294, + "balance_loss_clip": 1.07422733, + "balance_loss_mlp": 1.05110097, + "epoch": 0.06721577586725185, + "flos": 11216590997760.0, + "grad_norm": 2.289875290779408, + "language_loss": 0.86615127, + "learning_rate": 3.985533082728259e-06, + "loss": 0.88937867, + "num_input_tokens_seen": 11906965, + "step": 559, + "time_per_iteration": 2.7988409996032715 + }, + { + "auxiliary_loss_clip": 0.01285596, + "auxiliary_loss_mlp": 0.0105318, + "balance_loss_clip": 1.07760143, + "balance_loss_mlp": 1.03410625, + "epoch": 0.06733601875789094, + "flos": 25922189664000.0, + "grad_norm": 2.103677743239658, + "language_loss": 0.74642575, + "learning_rate": 3.985439408173951e-06, + "loss": 0.76981348, + "num_input_tokens_seen": 11927190, + "step": 560, + "time_per_iteration": 2.770798683166504 + }, + { + "auxiliary_loss_clip": 0.01285886, + "auxiliary_loss_mlp": 0.01051821, + "balance_loss_clip": 1.08016741, + "balance_loss_mlp": 1.03401136, + "epoch": 0.06745626164853002, + "flos": 20813645577600.0, + "grad_norm": 3.6623286688806993, + "language_loss": 0.70688766, + "learning_rate": 3.9853454324300634e-06, + "loss": 0.73026472, + "num_input_tokens_seen": 11946400, + "step": 561, + "time_per_iteration": 2.6946911811828613 + }, + { + "auxiliary_loss_clip": 0.01229482, + "auxiliary_loss_mlp": 0.01052479, + "balance_loss_clip": 1.06691098, + "balance_loss_mlp": 1.03463316, + "epoch": 0.06757650453916912, + "flos": 19829262378240.0, + "grad_norm": 1.901540776732772, + "language_loss": 0.7794677, + "learning_rate": 3.985251155510852e-06, + "loss": 0.80228734, + "num_input_tokens_seen": 11965430, + "step": 562, + "time_per_iteration": 2.8862416744232178 + }, + { + "auxiliary_loss_clip": 0.01238483, + "auxiliary_loss_mlp": 0.01045412, + "balance_loss_clip": 1.07380331, + "balance_loss_mlp": 1.02794802, + "epoch": 0.06769674742980822, + "flos": 25739224761600.0, + "grad_norm": 2.393138290560479, + "language_loss": 0.80180746, + "learning_rate": 3.98515657743062e-06, + "loss": 0.82464641, + "num_input_tokens_seen": 11984895, + "step": 563, + "time_per_iteration": 2.8678483963012695 + }, + { + "auxiliary_loss_clip": 0.01261952, + "auxiliary_loss_mlp": 0.01049788, + "balance_loss_clip": 1.07296288, + "balance_loss_mlp": 1.03046441, + "epoch": 0.0678169903204473, + "flos": 13074788355840.0, + "grad_norm": 2.0241533151202633, + "language_loss": 0.77770627, + "learning_rate": 3.985061698203711e-06, + "loss": 0.80082369, + "num_input_tokens_seen": 12002010, + "step": 564, + "time_per_iteration": 2.7784037590026855 + }, + { + "auxiliary_loss_clip": 0.01194196, + "auxiliary_loss_mlp": 0.01017473, + "balance_loss_clip": 1.06345057, + "balance_loss_mlp": 1.00764978, + "epoch": 0.0679372332110864, + "flos": 70865830788480.0, + "grad_norm": 0.8876037098864563, + "language_loss": 0.63837034, + "learning_rate": 3.984966517844523e-06, + "loss": 0.660487, + "num_input_tokens_seen": 12057255, + "step": 565, + "time_per_iteration": 3.2121357917785645 + }, + { + "auxiliary_loss_clip": 0.01287671, + "auxiliary_loss_mlp": 0.01048544, + "balance_loss_clip": 1.08017254, + "balance_loss_mlp": 1.03029323, + "epoch": 0.06805747610172548, + "flos": 28256418990720.0, + "grad_norm": 2.46288387335201, + "language_loss": 0.8057512, + "learning_rate": 3.984871036367492e-06, + "loss": 0.82911336, + "num_input_tokens_seen": 12077280, + "step": 566, + "time_per_iteration": 2.7114107608795166 + }, + { + "auxiliary_loss_clip": 0.01269994, + "auxiliary_loss_mlp": 0.00891061, + "balance_loss_clip": 1.07734215, + "balance_loss_mlp": 1.00060964, + "epoch": 0.06817771899236458, + "flos": 20120533764480.0, + "grad_norm": 1.787228612145513, + "language_loss": 0.83406305, + "learning_rate": 3.984775253787102e-06, + "loss": 0.85567361, + "num_input_tokens_seen": 12095570, + "step": 567, + "time_per_iteration": 4.657917261123657 + }, + { + "auxiliary_loss_clip": 0.01271255, + "auxiliary_loss_mlp": 0.01057158, + "balance_loss_clip": 1.07198668, + "balance_loss_mlp": 1.03913379, + "epoch": 0.06829796188300366, + "flos": 17930629284480.0, + "grad_norm": 3.7915762588243163, + "language_loss": 0.87996244, + "learning_rate": 3.984679170117885e-06, + "loss": 0.90324664, + "num_input_tokens_seen": 12111775, + "step": 568, + "time_per_iteration": 3.724475383758545 + }, + { + "auxiliary_loss_clip": 0.01265346, + "auxiliary_loss_mlp": 0.0104754, + "balance_loss_clip": 1.07277989, + "balance_loss_mlp": 1.03006339, + "epoch": 0.06841820477364276, + "flos": 14501627285760.0, + "grad_norm": 2.7629385102072486, + "language_loss": 0.7860834, + "learning_rate": 3.984582785374415e-06, + "loss": 0.80921233, + "num_input_tokens_seen": 12129215, + "step": 569, + "time_per_iteration": 2.69675350189209 + }, + { + "auxiliary_loss_clip": 0.01257135, + "auxiliary_loss_mlp": 0.0089157, + "balance_loss_clip": 1.07557678, + "balance_loss_mlp": 1.00065994, + "epoch": 0.06853844766428185, + "flos": 21938474954880.0, + "grad_norm": 2.1101767059503724, + "language_loss": 0.80479884, + "learning_rate": 3.9844860995713155e-06, + "loss": 0.8262859, + "num_input_tokens_seen": 12148755, + "step": 570, + "time_per_iteration": 2.8581748008728027 + }, + { + "auxiliary_loss_clip": 0.01274745, + "auxiliary_loss_mlp": 0.0105297, + "balance_loss_clip": 1.08110452, + "balance_loss_mlp": 1.03527892, + "epoch": 0.06865869055492094, + "flos": 16800628348800.0, + "grad_norm": 2.293866338365156, + "language_loss": 0.83181626, + "learning_rate": 3.9843891127232524e-06, + "loss": 0.85509342, + "num_input_tokens_seen": 12166290, + "step": 571, + "time_per_iteration": 2.697740077972412 + }, + { + "auxiliary_loss_clip": 0.01218932, + "auxiliary_loss_mlp": 0.0104702, + "balance_loss_clip": 1.06683826, + "balance_loss_mlp": 1.02942443, + "epoch": 0.06877893344556003, + "flos": 19937281553280.0, + "grad_norm": 2.527990697175696, + "language_loss": 0.67398083, + "learning_rate": 3.984291824844938e-06, + "loss": 0.69664037, + "num_input_tokens_seen": 12181385, + "step": 572, + "time_per_iteration": 2.8607587814331055 + }, + { + "auxiliary_loss_clip": 0.01281991, + "auxiliary_loss_mlp": 0.01046065, + "balance_loss_clip": 1.07715297, + "balance_loss_mlp": 1.02686071, + "epoch": 0.06889917633619912, + "flos": 23039388852480.0, + "grad_norm": 2.3482236558578875, + "language_loss": 0.85301453, + "learning_rate": 3.984194235951132e-06, + "loss": 0.87629509, + "num_input_tokens_seen": 12197530, + "step": 573, + "time_per_iteration": 2.6802978515625 + }, + { + "auxiliary_loss_clip": 0.01285069, + "auxiliary_loss_mlp": 0.01048805, + "balance_loss_clip": 1.08179092, + "balance_loss_mlp": 1.03155518, + "epoch": 0.06901941922683821, + "flos": 20960556203520.0, + "grad_norm": 3.5849391813401303, + "language_loss": 0.85229862, + "learning_rate": 3.9840963460566375e-06, + "loss": 0.87563735, + "num_input_tokens_seen": 12216310, + "step": 574, + "time_per_iteration": 2.7228102684020996 + }, + { + "auxiliary_loss_clip": 0.0119956, + "auxiliary_loss_mlp": 0.01053356, + "balance_loss_clip": 1.06265593, + "balance_loss_mlp": 1.03574896, + "epoch": 0.06913966211747731, + "flos": 24821850384000.0, + "grad_norm": 3.236160214950937, + "language_loss": 0.89566374, + "learning_rate": 3.983998155176305e-06, + "loss": 0.91819298, + "num_input_tokens_seen": 12236670, + "step": 575, + "time_per_iteration": 2.9474291801452637 + }, + { + "auxiliary_loss_clip": 0.01185024, + "auxiliary_loss_mlp": 0.01021081, + "balance_loss_clip": 1.05635738, + "balance_loss_mlp": 1.01159239, + "epoch": 0.06925990500811639, + "flos": 58367446957440.0, + "grad_norm": 0.8216918847258273, + "language_loss": 0.57018614, + "learning_rate": 3.9838996633250305e-06, + "loss": 0.59224719, + "num_input_tokens_seen": 12297185, + "step": 576, + "time_per_iteration": 3.2726686000823975 + }, + { + "auxiliary_loss_clip": 0.01271333, + "auxiliary_loss_mlp": 0.01062346, + "balance_loss_clip": 1.07480729, + "balance_loss_mlp": 1.04552507, + "epoch": 0.06938014789875549, + "flos": 12749940731520.0, + "grad_norm": 2.0492643820401963, + "language_loss": 0.88136649, + "learning_rate": 3.983800870517753e-06, + "loss": 0.90470326, + "num_input_tokens_seen": 12313975, + "step": 577, + "time_per_iteration": 2.807347059249878 + }, + { + "auxiliary_loss_clip": 0.01264398, + "auxiliary_loss_mlp": 0.01054148, + "balance_loss_clip": 1.07672822, + "balance_loss_mlp": 1.03763759, + "epoch": 0.06950039078939457, + "flos": 22820226019200.0, + "grad_norm": 2.8190028997163923, + "language_loss": 0.77907914, + "learning_rate": 3.983701776769463e-06, + "loss": 0.80226457, + "num_input_tokens_seen": 12331385, + "step": 578, + "time_per_iteration": 2.7798588275909424 + }, + { + "auxiliary_loss_clip": 0.01258607, + "auxiliary_loss_mlp": 0.01048575, + "balance_loss_clip": 1.07375908, + "balance_loss_mlp": 1.03162313, + "epoch": 0.06962063368003367, + "flos": 21941348042880.0, + "grad_norm": 4.338440386862095, + "language_loss": 0.85469747, + "learning_rate": 3.9836023820951885e-06, + "loss": 0.87776923, + "num_input_tokens_seen": 12350600, + "step": 579, + "time_per_iteration": 2.754405975341797 + }, + { + "auxiliary_loss_clip": 0.01241127, + "auxiliary_loss_mlp": 0.01054537, + "balance_loss_clip": 1.07088602, + "balance_loss_mlp": 1.03727531, + "epoch": 0.06974087657067275, + "flos": 20706021452160.0, + "grad_norm": 1.985962973158645, + "language_loss": 0.68826586, + "learning_rate": 3.983502686510011e-06, + "loss": 0.71122253, + "num_input_tokens_seen": 12371430, + "step": 580, + "time_per_iteration": 2.8797905445098877 + }, + { + "auxiliary_loss_clip": 0.01273973, + "auxiliary_loss_mlp": 0.00891461, + "balance_loss_clip": 1.07527709, + "balance_loss_mlp": 1.00055504, + "epoch": 0.06986111946131185, + "flos": 22638230784000.0, + "grad_norm": 1.9873134650288455, + "language_loss": 0.73893046, + "learning_rate": 3.9834026900290525e-06, + "loss": 0.76058483, + "num_input_tokens_seen": 12390825, + "step": 581, + "time_per_iteration": 2.8005244731903076 + }, + { + "auxiliary_loss_clip": 0.01282074, + "auxiliary_loss_mlp": 0.01052151, + "balance_loss_clip": 1.07799149, + "balance_loss_mlp": 1.03525877, + "epoch": 0.06998136235195095, + "flos": 26943453152640.0, + "grad_norm": 1.9298496794461688, + "language_loss": 1.00178337, + "learning_rate": 3.983302392667482e-06, + "loss": 1.0251255, + "num_input_tokens_seen": 12411670, + "step": 582, + "time_per_iteration": 2.7526073455810547 + }, + { + "auxiliary_loss_clip": 0.01270832, + "auxiliary_loss_mlp": 0.01066056, + "balance_loss_clip": 1.08116198, + "balance_loss_mlp": 1.04892623, + "epoch": 0.07010160524259003, + "flos": 22492505306880.0, + "grad_norm": 1.7345113759644033, + "language_loss": 0.93620467, + "learning_rate": 3.983201794440517e-06, + "loss": 0.95957357, + "num_input_tokens_seen": 12431245, + "step": 583, + "time_per_iteration": 2.745638608932495 + }, + { + "auxiliary_loss_clip": 0.01247923, + "auxiliary_loss_mlp": 0.01051532, + "balance_loss_clip": 1.07085085, + "balance_loss_mlp": 1.03578401, + "epoch": 0.07022184813322913, + "flos": 18332541538560.0, + "grad_norm": 2.2449667060369536, + "language_loss": 0.67234677, + "learning_rate": 3.9831008953634165e-06, + "loss": 0.69534135, + "num_input_tokens_seen": 12450535, + "step": 584, + "time_per_iteration": 2.7894234657287598 + }, + { + "auxiliary_loss_clip": 0.01213683, + "auxiliary_loss_mlp": 0.01059372, + "balance_loss_clip": 1.06285417, + "balance_loss_mlp": 1.04205072, + "epoch": 0.07034209102386821, + "flos": 24675550289280.0, + "grad_norm": 2.165742519180429, + "language_loss": 0.8108837, + "learning_rate": 3.9829996954514864e-06, + "loss": 0.83361423, + "num_input_tokens_seen": 12469675, + "step": 585, + "time_per_iteration": 2.9212868213653564 + }, + { + "auxiliary_loss_clip": 0.01252934, + "auxiliary_loss_mlp": 0.01063809, + "balance_loss_clip": 1.06994843, + "balance_loss_mlp": 1.0455699, + "epoch": 0.0704623339145073, + "flos": 25995878415360.0, + "grad_norm": 1.8575684438582576, + "language_loss": 0.8436538, + "learning_rate": 3.982898194720079e-06, + "loss": 0.86682129, + "num_input_tokens_seen": 12490405, + "step": 586, + "time_per_iteration": 2.829254388809204 + }, + { + "auxiliary_loss_clip": 0.01242785, + "auxiliary_loss_mlp": 0.00891769, + "balance_loss_clip": 1.06905544, + "balance_loss_mlp": 1.00062513, + "epoch": 0.0705825768051464, + "flos": 25338318088320.0, + "grad_norm": 1.9649238988738484, + "language_loss": 0.82225049, + "learning_rate": 3.982796393184592e-06, + "loss": 0.84359598, + "num_input_tokens_seen": 12509485, + "step": 587, + "time_per_iteration": 2.95477557182312 + }, + { + "auxiliary_loss_clip": 0.01158608, + "auxiliary_loss_mlp": 0.01012738, + "balance_loss_clip": 1.04088855, + "balance_loss_mlp": 1.00305808, + "epoch": 0.07070281969578548, + "flos": 66047552507520.0, + "grad_norm": 0.7873363947144904, + "language_loss": 0.62540227, + "learning_rate": 3.98269429086047e-06, + "loss": 0.64711571, + "num_input_tokens_seen": 12567325, + "step": 588, + "time_per_iteration": 3.2662582397460938 + }, + { + "auxiliary_loss_clip": 0.01245183, + "auxiliary_loss_mlp": 0.01058151, + "balance_loss_clip": 1.07022309, + "balance_loss_mlp": 1.03969753, + "epoch": 0.07082306258642458, + "flos": 23653568528640.0, + "grad_norm": 3.016722864170082, + "language_loss": 0.86274111, + "learning_rate": 3.982591887763199e-06, + "loss": 0.88577449, + "num_input_tokens_seen": 12584785, + "step": 589, + "time_per_iteration": 2.8926100730895996 + }, + { + "auxiliary_loss_clip": 0.01215405, + "auxiliary_loss_mlp": 0.01048261, + "balance_loss_clip": 1.0597806, + "balance_loss_mlp": 1.03111887, + "epoch": 0.07094330547706366, + "flos": 13880049408000.0, + "grad_norm": 2.081591657626029, + "language_loss": 0.81903183, + "learning_rate": 3.982489183908316e-06, + "loss": 0.84166855, + "num_input_tokens_seen": 12601205, + "step": 590, + "time_per_iteration": 2.8414082527160645 + }, + { + "auxiliary_loss_clip": 0.01187398, + "auxiliary_loss_mlp": 0.01052422, + "balance_loss_clip": 1.05151868, + "balance_loss_mlp": 1.03490973, + "epoch": 0.07106354836770276, + "flos": 24645098534400.0, + "grad_norm": 1.9191995561627546, + "language_loss": 0.84847575, + "learning_rate": 3.982386179311399e-06, + "loss": 0.87087393, + "num_input_tokens_seen": 12621725, + "step": 591, + "time_per_iteration": 2.936094284057617 + }, + { + "auxiliary_loss_clip": 0.01269298, + "auxiliary_loss_mlp": 0.01057832, + "balance_loss_clip": 1.07225871, + "balance_loss_mlp": 1.03884244, + "epoch": 0.07118379125834184, + "flos": 16217223649920.0, + "grad_norm": 8.211049875622509, + "language_loss": 0.8740505, + "learning_rate": 3.982282873988075e-06, + "loss": 0.89732182, + "num_input_tokens_seen": 12639600, + "step": 592, + "time_per_iteration": 2.736774444580078 + }, + { + "auxiliary_loss_clip": 0.01251491, + "auxiliary_loss_mlp": 0.01056245, + "balance_loss_clip": 1.07351637, + "balance_loss_mlp": 1.03885269, + "epoch": 0.07130403414898094, + "flos": 19719986227200.0, + "grad_norm": 2.188013997416938, + "language_loss": 0.87174094, + "learning_rate": 3.982179267954016e-06, + "loss": 0.89481831, + "num_input_tokens_seen": 12660030, + "step": 593, + "time_per_iteration": 4.5909507274627686 + }, + { + "auxiliary_loss_clip": 0.01277478, + "auxiliary_loss_mlp": 0.01054325, + "balance_loss_clip": 1.07568753, + "balance_loss_mlp": 1.03626513, + "epoch": 0.07142427703962004, + "flos": 21871933009920.0, + "grad_norm": 2.090623331460828, + "language_loss": 0.96162409, + "learning_rate": 3.982075361224937e-06, + "loss": 0.98494208, + "num_input_tokens_seen": 12678395, + "step": 594, + "time_per_iteration": 3.675307512283325 + }, + { + "auxiliary_loss_clip": 0.0126155, + "auxiliary_loss_mlp": 0.00891164, + "balance_loss_clip": 1.07690012, + "balance_loss_mlp": 1.00066614, + "epoch": 0.07154451993025912, + "flos": 18296595002880.0, + "grad_norm": 7.433711230428423, + "language_loss": 0.88167328, + "learning_rate": 3.981971153816602e-06, + "loss": 0.90320039, + "num_input_tokens_seen": 12696000, + "step": 595, + "time_per_iteration": 2.724254608154297 + }, + { + "auxiliary_loss_clip": 0.01273956, + "auxiliary_loss_mlp": 0.01061448, + "balance_loss_clip": 1.07889438, + "balance_loss_mlp": 1.04444909, + "epoch": 0.07166476282089822, + "flos": 22160690444160.0, + "grad_norm": 1.5557107747848138, + "language_loss": 0.96367419, + "learning_rate": 3.981866645744819e-06, + "loss": 0.98702824, + "num_input_tokens_seen": 12716715, + "step": 596, + "time_per_iteration": 2.826185464859009 + }, + { + "auxiliary_loss_clip": 0.01276967, + "auxiliary_loss_mlp": 0.00891415, + "balance_loss_clip": 1.07564902, + "balance_loss_mlp": 1.00059915, + "epoch": 0.0717850057115373, + "flos": 14136343925760.0, + "grad_norm": 2.4553168525345574, + "language_loss": 0.81396067, + "learning_rate": 3.9817618370254416e-06, + "loss": 0.83564448, + "num_input_tokens_seen": 12733370, + "step": 597, + "time_per_iteration": 2.696153163909912 + }, + { + "auxiliary_loss_clip": 0.01278562, + "auxiliary_loss_mlp": 0.01045989, + "balance_loss_clip": 1.07751262, + "balance_loss_mlp": 1.02966964, + "epoch": 0.0719052486021764, + "flos": 30917794412160.0, + "grad_norm": 3.1843904198721322, + "language_loss": 0.87429047, + "learning_rate": 3.9816567276743684e-06, + "loss": 0.89753598, + "num_input_tokens_seen": 12753235, + "step": 598, + "time_per_iteration": 2.831683874130249 + }, + { + "auxiliary_loss_clip": 0.01249147, + "auxiliary_loss_mlp": 0.01053148, + "balance_loss_clip": 1.07151508, + "balance_loss_mlp": 1.03642249, + "epoch": 0.0720254914928155, + "flos": 21287019939840.0, + "grad_norm": 1.921110104989164, + "language_loss": 0.77359045, + "learning_rate": 3.9815513177075466e-06, + "loss": 0.79661334, + "num_input_tokens_seen": 12772020, + "step": 599, + "time_per_iteration": 2.8664560317993164 + }, + { + "auxiliary_loss_clip": 0.01257441, + "auxiliary_loss_mlp": 0.01053978, + "balance_loss_clip": 1.07355511, + "balance_loss_mlp": 1.03740788, + "epoch": 0.07214573438345458, + "flos": 27819170732160.0, + "grad_norm": 1.6325868659967246, + "language_loss": 0.69964898, + "learning_rate": 3.9814456071409646e-06, + "loss": 0.72276312, + "num_input_tokens_seen": 12792555, + "step": 600, + "time_per_iteration": 2.9225964546203613 + }, + { + "auxiliary_loss_clip": 0.01237549, + "auxiliary_loss_mlp": 0.01054821, + "balance_loss_clip": 1.06897271, + "balance_loss_mlp": 1.03790486, + "epoch": 0.07226597727409367, + "flos": 25483576688640.0, + "grad_norm": 3.131524992731022, + "language_loss": 0.85476947, + "learning_rate": 3.981339595990659e-06, + "loss": 0.87769312, + "num_input_tokens_seen": 12811085, + "step": 601, + "time_per_iteration": 2.879068374633789 + }, + { + "auxiliary_loss_clip": 0.01260816, + "auxiliary_loss_mlp": 0.01049429, + "balance_loss_clip": 1.07268357, + "balance_loss_mlp": 1.03214359, + "epoch": 0.07238622016473276, + "flos": 23513840622720.0, + "grad_norm": 2.061518039378868, + "language_loss": 0.81138748, + "learning_rate": 3.981233284272713e-06, + "loss": 0.83448994, + "num_input_tokens_seen": 12830830, + "step": 602, + "time_per_iteration": 2.850726842880249 + }, + { + "auxiliary_loss_clip": 0.01241952, + "auxiliary_loss_mlp": 0.01056685, + "balance_loss_clip": 1.07235849, + "balance_loss_mlp": 1.04075837, + "epoch": 0.07250646305537185, + "flos": 25453519983360.0, + "grad_norm": 1.6615350819092047, + "language_loss": 0.90172827, + "learning_rate": 3.981126672003253e-06, + "loss": 0.92471462, + "num_input_tokens_seen": 12853505, + "step": 603, + "time_per_iteration": 2.851315975189209 + }, + { + "auxiliary_loss_clip": 0.01254972, + "auxiliary_loss_mlp": 0.01048268, + "balance_loss_clip": 1.06820643, + "balance_loss_mlp": 1.03089941, + "epoch": 0.07262670594601094, + "flos": 27155038216320.0, + "grad_norm": 5.585861649723421, + "language_loss": 0.78100288, + "learning_rate": 3.981019759198451e-06, + "loss": 0.80403537, + "num_input_tokens_seen": 12872455, + "step": 604, + "time_per_iteration": 2.859997272491455 + }, + { + "auxiliary_loss_clip": 0.01252425, + "auxiliary_loss_mlp": 0.01045332, + "balance_loss_clip": 1.07118738, + "balance_loss_mlp": 1.02851152, + "epoch": 0.07274694883665003, + "flos": 26651607148800.0, + "grad_norm": 2.0466946881440027, + "language_loss": 0.83755076, + "learning_rate": 3.980912545874528e-06, + "loss": 0.86052835, + "num_input_tokens_seen": 12892620, + "step": 605, + "time_per_iteration": 2.8583219051361084 + }, + { + "auxiliary_loss_clip": 0.01253471, + "auxiliary_loss_mlp": 0.00890779, + "balance_loss_clip": 1.07105708, + "balance_loss_mlp": 1.00059414, + "epoch": 0.07286719172728913, + "flos": 29862344154240.0, + "grad_norm": 2.1199180267593376, + "language_loss": 0.85880166, + "learning_rate": 3.980805032047746e-06, + "loss": 0.88024414, + "num_input_tokens_seen": 12914090, + "step": 606, + "time_per_iteration": 2.8415541648864746 + }, + { + "auxiliary_loss_clip": 0.01249736, + "auxiliary_loss_mlp": 0.01056727, + "balance_loss_clip": 1.07258558, + "balance_loss_mlp": 1.040097, + "epoch": 0.07298743461792821, + "flos": 17382057799680.0, + "grad_norm": 1.7931595382744778, + "language_loss": 0.81148517, + "learning_rate": 3.980697217734415e-06, + "loss": 0.83454978, + "num_input_tokens_seen": 12931830, + "step": 607, + "time_per_iteration": 2.8627750873565674 + }, + { + "auxiliary_loss_clip": 0.01235929, + "auxiliary_loss_mlp": 0.00891087, + "balance_loss_clip": 1.06958258, + "balance_loss_mlp": 1.00068331, + "epoch": 0.07310767750856731, + "flos": 19498201701120.0, + "grad_norm": 1.8960724632019819, + "language_loss": 0.91630709, + "learning_rate": 3.980589102950891e-06, + "loss": 0.93757725, + "num_input_tokens_seen": 12949995, + "step": 608, + "time_per_iteration": 2.9015390872955322 + }, + { + "auxiliary_loss_clip": 0.01251816, + "auxiliary_loss_mlp": 0.01046195, + "balance_loss_clip": 1.07432258, + "balance_loss_mlp": 1.02881467, + "epoch": 0.07322792039920639, + "flos": 29168693637120.0, + "grad_norm": 2.6219121461162227, + "language_loss": 0.76093006, + "learning_rate": 3.9804806877135755e-06, + "loss": 0.7839101, + "num_input_tokens_seen": 12968040, + "step": 609, + "time_per_iteration": 2.8654420375823975 + }, + { + "auxiliary_loss_clip": 0.01260407, + "auxiliary_loss_mlp": 0.00891272, + "balance_loss_clip": 1.06870079, + "balance_loss_mlp": 1.00059569, + "epoch": 0.07334816328984549, + "flos": 23477822259840.0, + "grad_norm": 2.0659429280118875, + "language_loss": 0.86027992, + "learning_rate": 3.980371972038915e-06, + "loss": 0.88179672, + "num_input_tokens_seen": 12988530, + "step": 610, + "time_per_iteration": 2.8639416694641113 + }, + { + "auxiliary_loss_clip": 0.01276748, + "auxiliary_loss_mlp": 0.01049066, + "balance_loss_clip": 1.07797587, + "balance_loss_mlp": 1.03172135, + "epoch": 0.07346840618048459, + "flos": 22962467877120.0, + "grad_norm": 2.2301191330457413, + "language_loss": 0.84365082, + "learning_rate": 3.980262955943399e-06, + "loss": 0.86690891, + "num_input_tokens_seen": 13008195, + "step": 611, + "time_per_iteration": 2.82161021232605 + }, + { + "auxiliary_loss_clip": 0.01246643, + "auxiliary_loss_mlp": 0.01052829, + "balance_loss_clip": 1.0762943, + "balance_loss_mlp": 1.03647387, + "epoch": 0.07358864907112367, + "flos": 17673903803520.0, + "grad_norm": 2.525884355632581, + "language_loss": 0.87047338, + "learning_rate": 3.980153639443569e-06, + "loss": 0.89346814, + "num_input_tokens_seen": 13024180, + "step": 612, + "time_per_iteration": 2.7931082248687744 + }, + { + "auxiliary_loss_clip": 0.01259471, + "auxiliary_loss_mlp": 0.01046092, + "balance_loss_clip": 1.07370198, + "balance_loss_mlp": 1.02991486, + "epoch": 0.07370889196176277, + "flos": 24097029840000.0, + "grad_norm": 2.326269116491681, + "language_loss": 0.79946315, + "learning_rate": 3.980044022556005e-06, + "loss": 0.82251871, + "num_input_tokens_seen": 13043865, + "step": 613, + "time_per_iteration": 2.8427622318267822 + }, + { + "auxiliary_loss_clip": 0.01266412, + "auxiliary_loss_mlp": 0.01053913, + "balance_loss_clip": 1.07590246, + "balance_loss_mlp": 1.03688979, + "epoch": 0.07382913485240185, + "flos": 25885919905920.0, + "grad_norm": 2.317468094866122, + "language_loss": 0.72717166, + "learning_rate": 3.9799341052973375e-06, + "loss": 0.75037491, + "num_input_tokens_seen": 13063700, + "step": 614, + "time_per_iteration": 2.831352472305298 + }, + { + "auxiliary_loss_clip": 0.0125792, + "auxiliary_loss_mlp": 0.01043792, + "balance_loss_clip": 1.07651055, + "balance_loss_mlp": 1.02661359, + "epoch": 0.07394937774304094, + "flos": 16873850223360.0, + "grad_norm": 2.8729905656928896, + "language_loss": 0.74902928, + "learning_rate": 3.979823887684241e-06, + "loss": 0.77204645, + "num_input_tokens_seen": 13082640, + "step": 615, + "time_per_iteration": 2.794083595275879 + }, + { + "auxiliary_loss_clip": 0.01276986, + "auxiliary_loss_mlp": 0.01050174, + "balance_loss_clip": 1.07736397, + "balance_loss_mlp": 1.03478384, + "epoch": 0.07406962063368003, + "flos": 20703471586560.0, + "grad_norm": 2.5472821004270165, + "language_loss": 0.84609371, + "learning_rate": 3.979713369733434e-06, + "loss": 0.86936527, + "num_input_tokens_seen": 13100505, + "step": 616, + "time_per_iteration": 2.759019136428833 + }, + { + "auxiliary_loss_clip": 0.01257013, + "auxiliary_loss_mlp": 0.01058117, + "balance_loss_clip": 1.07158303, + "balance_loss_mlp": 1.03980684, + "epoch": 0.07418986352431912, + "flos": 21430985650560.0, + "grad_norm": 2.2912880404804628, + "language_loss": 0.84742731, + "learning_rate": 3.979602551461683e-06, + "loss": 0.87057859, + "num_input_tokens_seen": 13121285, + "step": 617, + "time_per_iteration": 2.7976958751678467 + }, + { + "auxiliary_loss_clip": 0.01250607, + "auxiliary_loss_mlp": 0.01054489, + "balance_loss_clip": 1.0742625, + "balance_loss_mlp": 1.03789508, + "epoch": 0.07431010641495822, + "flos": 12021133777920.0, + "grad_norm": 2.410191415087528, + "language_loss": 0.91679531, + "learning_rate": 3.979491432885799e-06, + "loss": 0.93984628, + "num_input_tokens_seen": 13137550, + "step": 618, + "time_per_iteration": 2.9190680980682373 + }, + { + "auxiliary_loss_clip": 0.01223023, + "auxiliary_loss_mlp": 0.00890231, + "balance_loss_clip": 1.06712317, + "balance_loss_mlp": 1.00056839, + "epoch": 0.0744303493055973, + "flos": 20957575374720.0, + "grad_norm": 1.9843085244980456, + "language_loss": 0.82958448, + "learning_rate": 3.97938001402264e-06, + "loss": 0.85071701, + "num_input_tokens_seen": 13156675, + "step": 619, + "time_per_iteration": 4.768043041229248 + }, + { + "auxiliary_loss_clip": 0.01232878, + "auxiliary_loss_mlp": 0.01053476, + "balance_loss_clip": 1.07042289, + "balance_loss_mlp": 1.03749037, + "epoch": 0.0745505921962364, + "flos": 16253134272000.0, + "grad_norm": 5.043001820606104, + "language_loss": 0.7972784, + "learning_rate": 3.979268294889105e-06, + "loss": 0.82014197, + "num_input_tokens_seen": 13172225, + "step": 620, + "time_per_iteration": 2.8965721130371094 + }, + { + "auxiliary_loss_clip": 0.01275734, + "auxiliary_loss_mlp": 0.01048759, + "balance_loss_clip": 1.07712865, + "balance_loss_mlp": 1.032547, + "epoch": 0.07467083508687548, + "flos": 50944635550080.0, + "grad_norm": 1.8665483277756458, + "language_loss": 0.73875558, + "learning_rate": 3.979156275502143e-06, + "loss": 0.7620005, + "num_input_tokens_seen": 13195885, + "step": 621, + "time_per_iteration": 2.9603326320648193 + }, + { + "auxiliary_loss_clip": 0.0123854, + "auxiliary_loss_mlp": 0.01043746, + "balance_loss_clip": 1.06991851, + "balance_loss_mlp": 1.02791524, + "epoch": 0.07479107797751458, + "flos": 17529686697600.0, + "grad_norm": 2.3293595834386425, + "language_loss": 0.91992515, + "learning_rate": 3.979043955878749e-06, + "loss": 0.94274795, + "num_input_tokens_seen": 13213730, + "step": 622, + "time_per_iteration": 2.846784830093384 + }, + { + "auxiliary_loss_clip": 0.01245476, + "auxiliary_loss_mlp": 0.01048105, + "balance_loss_clip": 1.07192993, + "balance_loss_mlp": 1.03244066, + "epoch": 0.07491132086815366, + "flos": 23473943591040.0, + "grad_norm": 9.972080468130128, + "language_loss": 0.83149964, + "learning_rate": 3.978931336035959e-06, + "loss": 0.85443544, + "num_input_tokens_seen": 13232540, + "step": 623, + "time_per_iteration": 2.830287218093872 + }, + { + "auxiliary_loss_clip": 0.01265416, + "auxiliary_loss_mlp": 0.0105634, + "balance_loss_clip": 1.07847846, + "balance_loss_mlp": 1.03922176, + "epoch": 0.07503156375879276, + "flos": 20157557708160.0, + "grad_norm": 2.988222208536173, + "language_loss": 0.82365364, + "learning_rate": 3.9788184159908595e-06, + "loss": 0.8468712, + "num_input_tokens_seen": 13249670, + "step": 624, + "time_per_iteration": 2.766327381134033 + }, + { + "auxiliary_loss_clip": 0.01238042, + "auxiliary_loss_mlp": 0.01043076, + "balance_loss_clip": 1.06575286, + "balance_loss_mlp": 1.02769816, + "epoch": 0.07515180664943186, + "flos": 15115519653120.0, + "grad_norm": 5.599131407769983, + "language_loss": 0.83034664, + "learning_rate": 3.97870519576058e-06, + "loss": 0.85315782, + "num_input_tokens_seen": 13266095, + "step": 625, + "time_per_iteration": 2.7220284938812256 + }, + { + "auxiliary_loss_clip": 0.01236321, + "auxiliary_loss_mlp": 0.00891032, + "balance_loss_clip": 1.06888556, + "balance_loss_mlp": 1.00048602, + "epoch": 0.07527204954007094, + "flos": 21287702298240.0, + "grad_norm": 4.3127979137304715, + "language_loss": 0.81353378, + "learning_rate": 3.978591675362295e-06, + "loss": 0.83480734, + "num_input_tokens_seen": 13284810, + "step": 626, + "time_per_iteration": 2.8658998012542725 + }, + { + "auxiliary_loss_clip": 0.0123102, + "auxiliary_loss_mlp": 0.01051985, + "balance_loss_clip": 1.07442677, + "balance_loss_mlp": 1.03572512, + "epoch": 0.07539229243071004, + "flos": 21324187537920.0, + "grad_norm": 1.6849055515590827, + "language_loss": 0.87692046, + "learning_rate": 3.978477854813226e-06, + "loss": 0.89975047, + "num_input_tokens_seen": 13304150, + "step": 627, + "time_per_iteration": 2.872030735015869 + }, + { + "auxiliary_loss_clip": 0.01259516, + "auxiliary_loss_mlp": 0.01042805, + "balance_loss_clip": 1.07229865, + "balance_loss_mlp": 1.02692604, + "epoch": 0.07551253532134912, + "flos": 13042540920960.0, + "grad_norm": 2.145513211538433, + "language_loss": 0.82500553, + "learning_rate": 3.97836373413064e-06, + "loss": 0.84802872, + "num_input_tokens_seen": 13322205, + "step": 628, + "time_per_iteration": 2.8364956378936768 + }, + { + "auxiliary_loss_clip": 0.01270283, + "auxiliary_loss_mlp": 0.01047055, + "balance_loss_clip": 1.07352805, + "balance_loss_mlp": 1.03025889, + "epoch": 0.07563277821198822, + "flos": 19208761908480.0, + "grad_norm": 2.041679584739414, + "language_loss": 0.74382544, + "learning_rate": 3.978249313331848e-06, + "loss": 0.76699883, + "num_input_tokens_seen": 13340435, + "step": 629, + "time_per_iteration": 2.732545852661133 + }, + { + "auxiliary_loss_clip": 0.01262002, + "auxiliary_loss_mlp": 0.0089042, + "balance_loss_clip": 1.07037449, + "balance_loss_mlp": 1.00042713, + "epoch": 0.07575302110262731, + "flos": 19537200892800.0, + "grad_norm": 2.974480153701802, + "language_loss": 0.62319589, + "learning_rate": 3.978134592434208e-06, + "loss": 0.64472002, + "num_input_tokens_seen": 13358185, + "step": 630, + "time_per_iteration": 2.901632070541382 + }, + { + "auxiliary_loss_clip": 0.01138637, + "auxiliary_loss_mlp": 0.01008565, + "balance_loss_clip": 1.05514038, + "balance_loss_mlp": 0.99936181, + "epoch": 0.0758732639932664, + "flos": 67961808017280.0, + "grad_norm": 1.0094006138890992, + "language_loss": 0.59431154, + "learning_rate": 3.978019571455123e-06, + "loss": 0.61578357, + "num_input_tokens_seen": 13410130, + "step": 631, + "time_per_iteration": 3.5843355655670166 + }, + { + "auxiliary_loss_clip": 0.01270448, + "auxiliary_loss_mlp": 0.01044932, + "balance_loss_clip": 1.07740664, + "balance_loss_mlp": 1.03003097, + "epoch": 0.07599350688390549, + "flos": 18989204025600.0, + "grad_norm": 2.2856242308535153, + "language_loss": 0.84233379, + "learning_rate": 3.977904250412042e-06, + "loss": 0.86548758, + "num_input_tokens_seen": 13429085, + "step": 632, + "time_per_iteration": 2.9053356647491455 + }, + { + "auxiliary_loss_clip": 0.01255381, + "auxiliary_loss_mlp": 0.01053299, + "balance_loss_clip": 1.0739727, + "balance_loss_mlp": 1.03743207, + "epoch": 0.07611374977454458, + "flos": 21069006341760.0, + "grad_norm": 2.208517611995791, + "language_loss": 0.85588729, + "learning_rate": 3.97778862932246e-06, + "loss": 0.87897414, + "num_input_tokens_seen": 13446250, + "step": 633, + "time_per_iteration": 2.8897860050201416 + }, + { + "auxiliary_loss_clip": 0.01166054, + "auxiliary_loss_mlp": 0.01049402, + "balance_loss_clip": 1.04756427, + "balance_loss_mlp": 1.03346419, + "epoch": 0.07623399266518367, + "flos": 18514536773760.0, + "grad_norm": 2.330400331128123, + "language_loss": 0.94134778, + "learning_rate": 3.9776727082039144e-06, + "loss": 0.96350241, + "num_input_tokens_seen": 13463220, + "step": 634, + "time_per_iteration": 3.2319693565368652 + }, + { + "auxiliary_loss_clip": 0.01179632, + "auxiliary_loss_mlp": 0.01010089, + "balance_loss_clip": 1.05590129, + "balance_loss_mlp": 1.00131559, + "epoch": 0.07635423555582276, + "flos": 44663036077440.0, + "grad_norm": 0.8153170134298788, + "language_loss": 0.55425251, + "learning_rate": 3.977556487073991e-06, + "loss": 0.5761497, + "num_input_tokens_seen": 13517775, + "step": 635, + "time_per_iteration": 3.674396514892578 + }, + { + "auxiliary_loss_clip": 0.01237655, + "auxiliary_loss_mlp": 0.01045714, + "balance_loss_clip": 1.06441617, + "balance_loss_mlp": 1.0302527, + "epoch": 0.07647447844646185, + "flos": 21761148487680.0, + "grad_norm": 1.7060374129954352, + "language_loss": 0.81431198, + "learning_rate": 3.97743996595032e-06, + "loss": 0.83714569, + "num_input_tokens_seen": 13537815, + "step": 636, + "time_per_iteration": 2.8682901859283447 + }, + { + "auxiliary_loss_clip": 0.01269362, + "auxiliary_loss_mlp": 0.01041158, + "balance_loss_clip": 1.07473397, + "balance_loss_mlp": 1.02438474, + "epoch": 0.07659472133710095, + "flos": 23806799948160.0, + "grad_norm": 2.927183352042614, + "language_loss": 0.81633335, + "learning_rate": 3.9773231448505804e-06, + "loss": 0.83943856, + "num_input_tokens_seen": 13559605, + "step": 637, + "time_per_iteration": 2.7363996505737305 + }, + { + "auxiliary_loss_clip": 0.01245738, + "auxiliary_loss_mlp": 0.00889727, + "balance_loss_clip": 1.07144451, + "balance_loss_mlp": 1.00032115, + "epoch": 0.07671496422774003, + "flos": 21469984842240.0, + "grad_norm": 2.3180360437929215, + "language_loss": 0.78195918, + "learning_rate": 3.977206023792491e-06, + "loss": 0.80331385, + "num_input_tokens_seen": 13579495, + "step": 638, + "time_per_iteration": 2.9981977939605713 + }, + { + "auxiliary_loss_clip": 0.01256531, + "auxiliary_loss_mlp": 0.01045735, + "balance_loss_clip": 1.07260847, + "balance_loss_mlp": 1.02983296, + "epoch": 0.07683520711837913, + "flos": 16980971558400.0, + "grad_norm": 2.080743036315984, + "language_loss": 0.81321627, + "learning_rate": 3.97708860279382e-06, + "loss": 0.83623886, + "num_input_tokens_seen": 13597605, + "step": 639, + "time_per_iteration": 2.9168102741241455 + }, + { + "auxiliary_loss_clip": 0.01223733, + "auxiliary_loss_mlp": 0.01046473, + "balance_loss_clip": 1.06289673, + "balance_loss_mlp": 1.030141, + "epoch": 0.07695545000901821, + "flos": 23476744851840.0, + "grad_norm": 1.9231932579521278, + "language_loss": 0.78009927, + "learning_rate": 3.97697088187238e-06, + "loss": 0.80280137, + "num_input_tokens_seen": 13618120, + "step": 640, + "time_per_iteration": 2.962900400161743 + }, + { + "auxiliary_loss_clip": 0.01244226, + "auxiliary_loss_mlp": 0.01045921, + "balance_loss_clip": 1.07266855, + "balance_loss_mlp": 1.03015018, + "epoch": 0.07707569289965731, + "flos": 17634258167040.0, + "grad_norm": 2.3143348850321606, + "language_loss": 0.92216259, + "learning_rate": 3.976852861046029e-06, + "loss": 0.94506407, + "num_input_tokens_seen": 13634735, + "step": 641, + "time_per_iteration": 2.8286402225494385 + }, + { + "auxiliary_loss_clip": 0.01217135, + "auxiliary_loss_mlp": 0.01044917, + "balance_loss_clip": 1.06618142, + "balance_loss_mlp": 1.02838314, + "epoch": 0.0771959357902964, + "flos": 25775674087680.0, + "grad_norm": 1.724619487847824, + "language_loss": 0.80307788, + "learning_rate": 3.97673454033267e-06, + "loss": 0.82569844, + "num_input_tokens_seen": 13656835, + "step": 642, + "time_per_iteration": 2.981877565383911 + }, + { + "auxiliary_loss_clip": 0.01250529, + "auxiliary_loss_mlp": 0.01046882, + "balance_loss_clip": 1.07092643, + "balance_loss_mlp": 1.03113472, + "epoch": 0.07731617868093549, + "flos": 19828651847040.0, + "grad_norm": 2.0180213906737885, + "language_loss": 0.83022434, + "learning_rate": 3.976615919750254e-06, + "loss": 0.85319841, + "num_input_tokens_seen": 13674535, + "step": 643, + "time_per_iteration": 2.8600680828094482 + }, + { + "auxiliary_loss_clip": 0.01256157, + "auxiliary_loss_mlp": 0.01053406, + "balance_loss_clip": 1.07414532, + "balance_loss_mlp": 1.03707409, + "epoch": 0.07743642157157458, + "flos": 21324654414720.0, + "grad_norm": 2.057270223749695, + "language_loss": 0.86544442, + "learning_rate": 3.976496999316775e-06, + "loss": 0.88854009, + "num_input_tokens_seen": 13693290, + "step": 644, + "time_per_iteration": 2.742877244949341 + }, + { + "auxiliary_loss_clip": 0.0125144, + "auxiliary_loss_mlp": 0.01054799, + "balance_loss_clip": 1.07697546, + "balance_loss_mlp": 1.03846741, + "epoch": 0.07755666446221367, + "flos": 19969133938560.0, + "grad_norm": 9.712948586478053, + "language_loss": 0.84123266, + "learning_rate": 3.976377779050271e-06, + "loss": 0.86429507, + "num_input_tokens_seen": 13711420, + "step": 645, + "time_per_iteration": 4.614662408828735 + }, + { + "auxiliary_loss_clip": 0.01248527, + "auxiliary_loss_mlp": 0.01054205, + "balance_loss_clip": 1.0682807, + "balance_loss_mlp": 1.03813601, + "epoch": 0.07767690735285276, + "flos": 23623224514560.0, + "grad_norm": 2.3100825340383793, + "language_loss": 0.8443985, + "learning_rate": 3.976258258968831e-06, + "loss": 0.8674258, + "num_input_tokens_seen": 13729965, + "step": 646, + "time_per_iteration": 3.642000913619995 + }, + { + "auxiliary_loss_clip": 0.01233152, + "auxiliary_loss_mlp": 0.01057052, + "balance_loss_clip": 1.07168126, + "balance_loss_mlp": 1.04176939, + "epoch": 0.07779715024349185, + "flos": 22236246702720.0, + "grad_norm": 2.1325809546040313, + "language_loss": 0.74354255, + "learning_rate": 3.976138439090583e-06, + "loss": 0.76644456, + "num_input_tokens_seen": 13748045, + "step": 647, + "time_per_iteration": 2.865926742553711 + }, + { + "auxiliary_loss_clip": 0.01236658, + "auxiliary_loss_mlp": 0.01047388, + "balance_loss_clip": 1.06946075, + "balance_loss_mlp": 1.03130674, + "epoch": 0.07791739313413094, + "flos": 20955097336320.0, + "grad_norm": 4.780755946311576, + "language_loss": 0.8513692, + "learning_rate": 3.976018319433706e-06, + "loss": 0.87420964, + "num_input_tokens_seen": 13765590, + "step": 648, + "time_per_iteration": 2.7770071029663086 + }, + { + "auxiliary_loss_clip": 0.01258986, + "auxiliary_loss_mlp": 0.01050298, + "balance_loss_clip": 1.07463968, + "balance_loss_mlp": 1.03422904, + "epoch": 0.07803763602477004, + "flos": 19312327797120.0, + "grad_norm": 2.3330223092869526, + "language_loss": 0.9132092, + "learning_rate": 3.9758979000164205e-06, + "loss": 0.93630207, + "num_input_tokens_seen": 13782410, + "step": 649, + "time_per_iteration": 2.7139663696289062 + }, + { + "auxiliary_loss_clip": 0.01241805, + "auxiliary_loss_mlp": 0.01051411, + "balance_loss_clip": 1.07017207, + "balance_loss_mlp": 1.03432798, + "epoch": 0.07815787891540912, + "flos": 22710806213760.0, + "grad_norm": 1.8746003224239838, + "language_loss": 0.717246, + "learning_rate": 3.975777180856995e-06, + "loss": 0.74017817, + "num_input_tokens_seen": 13801530, + "step": 650, + "time_per_iteration": 2.871065855026245 + }, + { + "auxiliary_loss_clip": 0.01274441, + "auxiliary_loss_mlp": 0.01055045, + "balance_loss_clip": 1.07753479, + "balance_loss_mlp": 1.03862977, + "epoch": 0.07827812180604822, + "flos": 22711129436160.0, + "grad_norm": 1.9871854213433209, + "language_loss": 0.86163425, + "learning_rate": 3.975656161973742e-06, + "loss": 0.88492918, + "num_input_tokens_seen": 13820615, + "step": 651, + "time_per_iteration": 2.7059223651885986 + }, + { + "auxiliary_loss_clip": 0.01269318, + "auxiliary_loss_mlp": 0.01050191, + "balance_loss_clip": 1.07193923, + "balance_loss_mlp": 1.03364456, + "epoch": 0.0783983646966873, + "flos": 21725597001600.0, + "grad_norm": 2.7430875607038367, + "language_loss": 0.88895464, + "learning_rate": 3.9755348433850194e-06, + "loss": 0.91214985, + "num_input_tokens_seen": 13835955, + "step": 652, + "time_per_iteration": 2.641709089279175 + }, + { + "auxiliary_loss_clip": 0.01147692, + "auxiliary_loss_mlp": 0.01011489, + "balance_loss_clip": 1.04388356, + "balance_loss_mlp": 1.00271535, + "epoch": 0.0785186075873264, + "flos": 60640877537280.0, + "grad_norm": 0.9563741426222923, + "language_loss": 0.63603157, + "learning_rate": 3.975413225109232e-06, + "loss": 0.65762341, + "num_input_tokens_seen": 13896505, + "step": 653, + "time_per_iteration": 3.3515307903289795 + }, + { + "auxiliary_loss_clip": 0.01262782, + "auxiliary_loss_mlp": 0.01049278, + "balance_loss_clip": 1.07600427, + "balance_loss_mlp": 1.03254092, + "epoch": 0.0786388504779655, + "flos": 23877902920320.0, + "grad_norm": 3.3322950520125767, + "language_loss": 0.93212008, + "learning_rate": 3.975291307164829e-06, + "loss": 0.95524073, + "num_input_tokens_seen": 13915150, + "step": 654, + "time_per_iteration": 2.7897372245788574 + }, + { + "auxiliary_loss_clip": 0.01224091, + "auxiliary_loss_mlp": 0.01050239, + "balance_loss_clip": 1.06378174, + "balance_loss_mlp": 1.03422952, + "epoch": 0.07875909336860458, + "flos": 15158684822400.0, + "grad_norm": 2.2376381753893604, + "language_loss": 0.85277522, + "learning_rate": 3.975169089570306e-06, + "loss": 0.8755185, + "num_input_tokens_seen": 13933525, + "step": 655, + "time_per_iteration": 2.7133078575134277 + }, + { + "auxiliary_loss_clip": 0.01247656, + "auxiliary_loss_mlp": 0.010449, + "balance_loss_clip": 1.07135808, + "balance_loss_mlp": 1.02931952, + "epoch": 0.07887933625924368, + "flos": 22236857233920.0, + "grad_norm": 2.276579477382778, + "language_loss": 0.91487241, + "learning_rate": 3.975046572344202e-06, + "loss": 0.93779802, + "num_input_tokens_seen": 13949985, + "step": 656, + "time_per_iteration": 2.7508912086486816 + }, + { + "auxiliary_loss_clip": 0.0123261, + "auxiliary_loss_mlp": 0.01048257, + "balance_loss_clip": 1.06703448, + "balance_loss_mlp": 1.03210449, + "epoch": 0.07899957914988276, + "flos": 20777734955520.0, + "grad_norm": 1.9038615932951228, + "language_loss": 0.71236181, + "learning_rate": 3.974923755505103e-06, + "loss": 0.73517054, + "num_input_tokens_seen": 13969215, + "step": 657, + "time_per_iteration": 2.7622082233428955 + }, + { + "auxiliary_loss_clip": 0.01225991, + "auxiliary_loss_mlp": 0.01048506, + "balance_loss_clip": 1.06703091, + "balance_loss_mlp": 1.03275847, + "epoch": 0.07911982204052186, + "flos": 23003047267200.0, + "grad_norm": 1.7007850007579501, + "language_loss": 0.91278988, + "learning_rate": 3.974800639071641e-06, + "loss": 0.93553483, + "num_input_tokens_seen": 13989935, + "step": 658, + "time_per_iteration": 2.852032423019409 + }, + { + "auxiliary_loss_clip": 0.01195283, + "auxiliary_loss_mlp": 0.00890612, + "balance_loss_clip": 1.05878961, + "balance_loss_mlp": 1.00041747, + "epoch": 0.07924006493116094, + "flos": 23111389664640.0, + "grad_norm": 2.3279092630702154, + "language_loss": 1.00986922, + "learning_rate": 3.974677223062492e-06, + "loss": 1.0307281, + "num_input_tokens_seen": 14007150, + "step": 659, + "time_per_iteration": 2.8263494968414307 + }, + { + "auxiliary_loss_clip": 0.01245551, + "auxiliary_loss_mlp": 0.01049269, + "balance_loss_clip": 1.07278609, + "balance_loss_mlp": 1.03293741, + "epoch": 0.07936030782180004, + "flos": 16472153450880.0, + "grad_norm": 2.405692046038198, + "language_loss": 0.74271142, + "learning_rate": 3.974553507496378e-06, + "loss": 0.76565957, + "num_input_tokens_seen": 14025725, + "step": 660, + "time_per_iteration": 2.6538310050964355 + }, + { + "auxiliary_loss_clip": 0.01239666, + "auxiliary_loss_mlp": 0.01048321, + "balance_loss_clip": 1.06998062, + "balance_loss_mlp": 1.03154874, + "epoch": 0.07948055071243913, + "flos": 23733290764800.0, + "grad_norm": 2.948498602372952, + "language_loss": 0.89297742, + "learning_rate": 3.974429492392068e-06, + "loss": 0.91585732, + "num_input_tokens_seen": 14045750, + "step": 661, + "time_per_iteration": 2.677290439605713 + }, + { + "auxiliary_loss_clip": 0.01270352, + "auxiliary_loss_mlp": 0.00890549, + "balance_loss_clip": 1.07665706, + "balance_loss_mlp": 1.00043535, + "epoch": 0.07960079360307822, + "flos": 19573326996480.0, + "grad_norm": 2.2869022199912403, + "language_loss": 0.9105553, + "learning_rate": 3.974305177768373e-06, + "loss": 0.93216431, + "num_input_tokens_seen": 14063960, + "step": 662, + "time_per_iteration": 2.5877699851989746 + }, + { + "auxiliary_loss_clip": 0.01222644, + "auxiliary_loss_mlp": 0.01053589, + "balance_loss_clip": 1.06852257, + "balance_loss_mlp": 1.03797245, + "epoch": 0.07972103649371731, + "flos": 23513409659520.0, + "grad_norm": 2.236565716191757, + "language_loss": 0.86765862, + "learning_rate": 3.974180563644152e-06, + "loss": 0.89042103, + "num_input_tokens_seen": 14082525, + "step": 663, + "time_per_iteration": 2.669466733932495 + }, + { + "auxiliary_loss_clip": 0.01247957, + "auxiliary_loss_mlp": 0.01047091, + "balance_loss_clip": 1.07156992, + "balance_loss_mlp": 1.03092623, + "epoch": 0.0798412793843564, + "flos": 16726867770240.0, + "grad_norm": 2.3063322837145686, + "language_loss": 0.8929683, + "learning_rate": 3.97405565003831e-06, + "loss": 0.91591883, + "num_input_tokens_seen": 14098610, + "step": 664, + "time_per_iteration": 2.5699660778045654 + }, + { + "auxiliary_loss_clip": 0.01238714, + "auxiliary_loss_mlp": 0.01048755, + "balance_loss_clip": 1.06982446, + "balance_loss_mlp": 1.03219676, + "epoch": 0.07996152227499549, + "flos": 18223337214720.0, + "grad_norm": 2.2583798679866094, + "language_loss": 0.78445494, + "learning_rate": 3.973930436969794e-06, + "loss": 0.80732965, + "num_input_tokens_seen": 14117065, + "step": 665, + "time_per_iteration": 2.6451408863067627 + }, + { + "auxiliary_loss_clip": 0.01238779, + "auxiliary_loss_mlp": 0.01053261, + "balance_loss_clip": 1.06865621, + "balance_loss_mlp": 1.0366199, + "epoch": 0.08008176516563459, + "flos": 20594877793920.0, + "grad_norm": 2.0771695406100474, + "language_loss": 0.85963649, + "learning_rate": 3.973804924457602e-06, + "loss": 0.88255692, + "num_input_tokens_seen": 14135145, + "step": 666, + "time_per_iteration": 2.6167564392089844 + }, + { + "auxiliary_loss_clip": 0.01242917, + "auxiliary_loss_mlp": 0.01054063, + "balance_loss_clip": 1.06916094, + "balance_loss_mlp": 1.03825545, + "epoch": 0.08020200805627367, + "flos": 31834306863360.0, + "grad_norm": 1.6925785840415875, + "language_loss": 0.85722637, + "learning_rate": 3.973679112520771e-06, + "loss": 0.88019621, + "num_input_tokens_seen": 14156860, + "step": 667, + "time_per_iteration": 3.0140202045440674 + }, + { + "auxiliary_loss_clip": 0.01220611, + "auxiliary_loss_mlp": 0.01048069, + "balance_loss_clip": 1.06360698, + "balance_loss_mlp": 1.03205919, + "epoch": 0.08032225094691277, + "flos": 17783503176960.0, + "grad_norm": 1.8288361406155464, + "language_loss": 0.98766512, + "learning_rate": 3.973553001178389e-06, + "loss": 1.0103519, + "num_input_tokens_seen": 14174365, + "step": 668, + "time_per_iteration": 2.9016637802124023 + }, + { + "auxiliary_loss_clip": 0.0123142, + "auxiliary_loss_mlp": 0.01050142, + "balance_loss_clip": 1.06914842, + "balance_loss_mlp": 1.03401327, + "epoch": 0.08044249383755185, + "flos": 24061693835520.0, + "grad_norm": 1.961839555624524, + "language_loss": 0.75931716, + "learning_rate": 3.973426590449585e-06, + "loss": 0.78213274, + "num_input_tokens_seen": 14192320, + "step": 669, + "time_per_iteration": 2.941957473754883 + }, + { + "auxiliary_loss_clip": 0.0122235, + "auxiliary_loss_mlp": 0.01043372, + "balance_loss_clip": 1.06696224, + "balance_loss_mlp": 1.027565, + "epoch": 0.08056273672819095, + "flos": 18223624523520.0, + "grad_norm": 2.189541335515063, + "language_loss": 0.75197738, + "learning_rate": 3.9732998803535364e-06, + "loss": 0.7746346, + "num_input_tokens_seen": 14210380, + "step": 670, + "time_per_iteration": 3.9074485301971436 + }, + { + "auxiliary_loss_clip": 0.01268656, + "auxiliary_loss_mlp": 0.01047268, + "balance_loss_clip": 1.07385325, + "balance_loss_mlp": 1.03140092, + "epoch": 0.08068297961883003, + "flos": 19676856971520.0, + "grad_norm": 2.6756195282344937, + "language_loss": 0.85313845, + "learning_rate": 3.973172870909465e-06, + "loss": 0.87629771, + "num_input_tokens_seen": 14225145, + "step": 671, + "time_per_iteration": 3.6866090297698975 + }, + { + "auxiliary_loss_clip": 0.01249882, + "auxiliary_loss_mlp": 0.01050032, + "balance_loss_clip": 1.07001019, + "balance_loss_mlp": 1.03292584, + "epoch": 0.08080322250946913, + "flos": 23148736830720.0, + "grad_norm": 2.3070250322912935, + "language_loss": 0.8033309, + "learning_rate": 3.973045562136638e-06, + "loss": 0.82633007, + "num_input_tokens_seen": 14241960, + "step": 672, + "time_per_iteration": 3.655219078063965 + }, + { + "auxiliary_loss_clip": 0.01257437, + "auxiliary_loss_mlp": 0.01059559, + "balance_loss_clip": 1.0712893, + "balance_loss_mlp": 1.04375172, + "epoch": 0.08092346540010822, + "flos": 21763626526080.0, + "grad_norm": 2.127388111615969, + "language_loss": 0.91222745, + "learning_rate": 3.972917954054368e-06, + "loss": 0.93539739, + "num_input_tokens_seen": 14260515, + "step": 673, + "time_per_iteration": 2.670186996459961 + }, + { + "auxiliary_loss_clip": 0.01247556, + "auxiliary_loss_mlp": 0.01058866, + "balance_loss_clip": 1.07364035, + "balance_loss_mlp": 1.04108047, + "epoch": 0.08104370829074731, + "flos": 21032485188480.0, + "grad_norm": 2.6556302436208967, + "language_loss": 0.81837189, + "learning_rate": 3.972790046682013e-06, + "loss": 0.84143615, + "num_input_tokens_seen": 14279190, + "step": 674, + "time_per_iteration": 2.6907427310943604 + }, + { + "auxiliary_loss_clip": 0.01232517, + "auxiliary_loss_mlp": 0.01052663, + "balance_loss_clip": 1.06604707, + "balance_loss_mlp": 1.03662992, + "epoch": 0.0811639511813864, + "flos": 20083186598400.0, + "grad_norm": 2.2183703543728783, + "language_loss": 0.79266238, + "learning_rate": 3.972661840038977e-06, + "loss": 0.81551421, + "num_input_tokens_seen": 14299480, + "step": 675, + "time_per_iteration": 2.848518133163452 + }, + { + "auxiliary_loss_clip": 0.01257302, + "auxiliary_loss_mlp": 0.01054795, + "balance_loss_clip": 1.07333028, + "balance_loss_mlp": 1.03771234, + "epoch": 0.08128419407202549, + "flos": 16836718538880.0, + "grad_norm": 2.592267494967769, + "language_loss": 0.83432257, + "learning_rate": 3.972533334144707e-06, + "loss": 0.85744351, + "num_input_tokens_seen": 14316405, + "step": 676, + "time_per_iteration": 2.6431796550750732 + }, + { + "auxiliary_loss_clip": 0.01257213, + "auxiliary_loss_mlp": 0.01050752, + "balance_loss_clip": 1.06979465, + "balance_loss_mlp": 1.03406227, + "epoch": 0.08140443696266458, + "flos": 23769273214080.0, + "grad_norm": 2.17258337081746, + "language_loss": 0.78241134, + "learning_rate": 3.972404529018699e-06, + "loss": 0.80549103, + "num_input_tokens_seen": 14336265, + "step": 677, + "time_per_iteration": 2.723311185836792 + }, + { + "auxiliary_loss_clip": 0.01234736, + "auxiliary_loss_mlp": 0.010485, + "balance_loss_clip": 1.06226683, + "balance_loss_mlp": 1.03215671, + "epoch": 0.08152467985330367, + "flos": 24390132819840.0, + "grad_norm": 1.809151363087497, + "language_loss": 0.85450971, + "learning_rate": 3.972275424680493e-06, + "loss": 0.87734205, + "num_input_tokens_seen": 14356375, + "step": 678, + "time_per_iteration": 2.7069103717803955 + }, + { + "auxiliary_loss_clip": 0.01269561, + "auxiliary_loss_mlp": 0.01048178, + "balance_loss_clip": 1.07477641, + "balance_loss_mlp": 1.0326333, + "epoch": 0.08164492274394276, + "flos": 19317750750720.0, + "grad_norm": 2.0426017281619404, + "language_loss": 0.92024338, + "learning_rate": 3.972146021149673e-06, + "loss": 0.94342077, + "num_input_tokens_seen": 14374650, + "step": 679, + "time_per_iteration": 2.6461639404296875 + }, + { + "auxiliary_loss_clip": 0.01230921, + "auxiliary_loss_mlp": 0.0104859, + "balance_loss_clip": 1.06842721, + "balance_loss_mlp": 1.03266358, + "epoch": 0.08176516563458186, + "flos": 14830461319680.0, + "grad_norm": 2.245323694326945, + "language_loss": 0.79213965, + "learning_rate": 3.972016318445868e-06, + "loss": 0.81493473, + "num_input_tokens_seen": 14392650, + "step": 680, + "time_per_iteration": 2.872514009475708 + }, + { + "auxiliary_loss_clip": 0.01259519, + "auxiliary_loss_mlp": 0.01050015, + "balance_loss_clip": 1.07213497, + "balance_loss_mlp": 1.03332543, + "epoch": 0.08188540852522094, + "flos": 22602320161920.0, + "grad_norm": 1.985952639105405, + "language_loss": 0.92277586, + "learning_rate": 3.971886316588757e-06, + "loss": 0.94587111, + "num_input_tokens_seen": 14413155, + "step": 681, + "time_per_iteration": 2.800013303756714 + }, + { + "auxiliary_loss_clip": 0.01225607, + "auxiliary_loss_mlp": 0.01045914, + "balance_loss_clip": 1.06919217, + "balance_loss_mlp": 1.0292604, + "epoch": 0.08200565141586004, + "flos": 19463727623040.0, + "grad_norm": 2.6490517578879813, + "language_loss": 0.73298723, + "learning_rate": 3.9717560155980595e-06, + "loss": 0.75570244, + "num_input_tokens_seen": 14428805, + "step": 682, + "time_per_iteration": 2.8394906520843506 + }, + { + "auxiliary_loss_clip": 0.01256036, + "auxiliary_loss_mlp": 0.01049895, + "balance_loss_clip": 1.07157373, + "balance_loss_mlp": 1.0349226, + "epoch": 0.08212589430649912, + "flos": 20594662312320.0, + "grad_norm": 2.0607554920061713, + "language_loss": 0.9222213, + "learning_rate": 3.971625415493542e-06, + "loss": 0.94528061, + "num_input_tokens_seen": 14447125, + "step": 683, + "time_per_iteration": 2.7241170406341553 + }, + { + "auxiliary_loss_clip": 0.0122795, + "auxiliary_loss_mlp": 0.01039507, + "balance_loss_clip": 1.06488705, + "balance_loss_mlp": 1.02430749, + "epoch": 0.08224613719713822, + "flos": 25953611086080.0, + "grad_norm": 2.240957900946278, + "language_loss": 0.87537241, + "learning_rate": 3.971494516295017e-06, + "loss": 0.89804697, + "num_input_tokens_seen": 14466575, + "step": 684, + "time_per_iteration": 2.927699327468872 + }, + { + "auxiliary_loss_clip": 0.01235142, + "auxiliary_loss_mlp": 0.01055715, + "balance_loss_clip": 1.06529653, + "balance_loss_mlp": 1.03996754, + "epoch": 0.08236638008777732, + "flos": 23768734510080.0, + "grad_norm": 2.0065333743087166, + "language_loss": 0.85633564, + "learning_rate": 3.971363318022341e-06, + "loss": 0.87924421, + "num_input_tokens_seen": 14487915, + "step": 685, + "time_per_iteration": 2.8738858699798584 + }, + { + "auxiliary_loss_clip": 0.01242801, + "auxiliary_loss_mlp": 0.01046143, + "balance_loss_clip": 1.06526709, + "balance_loss_mlp": 1.02990663, + "epoch": 0.0824866229784164, + "flos": 38799144887040.0, + "grad_norm": 2.5233440243764, + "language_loss": 0.68570203, + "learning_rate": 3.971231820695417e-06, + "loss": 0.70859146, + "num_input_tokens_seen": 14511530, + "step": 686, + "time_per_iteration": 2.9635632038116455 + }, + { + "auxiliary_loss_clip": 0.01251166, + "auxiliary_loss_mlp": 0.01050411, + "balance_loss_clip": 1.07052755, + "balance_loss_mlp": 1.03483057, + "epoch": 0.0826068658690555, + "flos": 23107762391040.0, + "grad_norm": 2.3309149941221423, + "language_loss": 0.81305599, + "learning_rate": 3.971100024334193e-06, + "loss": 0.83607179, + "num_input_tokens_seen": 14529050, + "step": 687, + "time_per_iteration": 2.799417734146118 + }, + { + "auxiliary_loss_clip": 0.01214349, + "auxiliary_loss_mlp": 0.01056192, + "balance_loss_clip": 1.06029248, + "balance_loss_mlp": 1.04052782, + "epoch": 0.08272710875969458, + "flos": 21136374299520.0, + "grad_norm": 2.472930104986218, + "language_loss": 0.86406636, + "learning_rate": 3.970967928958663e-06, + "loss": 0.8867718, + "num_input_tokens_seen": 14546165, + "step": 688, + "time_per_iteration": 2.90533185005188 + }, + { + "auxiliary_loss_clip": 0.0122289, + "auxiliary_loss_mlp": 0.01044668, + "balance_loss_clip": 1.06317854, + "balance_loss_mlp": 1.02918315, + "epoch": 0.08284735165033368, + "flos": 19063000517760.0, + "grad_norm": 1.8198373599188413, + "language_loss": 0.8354829, + "learning_rate": 3.970835534588865e-06, + "loss": 0.85815847, + "num_input_tokens_seen": 14563660, + "step": 689, + "time_per_iteration": 2.87064790725708 + }, + { + "auxiliary_loss_clip": 0.01248367, + "auxiliary_loss_mlp": 0.01056211, + "balance_loss_clip": 1.07347751, + "balance_loss_mlp": 1.04018903, + "epoch": 0.08296759454097276, + "flos": 16727442387840.0, + "grad_norm": 1.9216971442817035, + "language_loss": 0.86025774, + "learning_rate": 3.970702841244883e-06, + "loss": 0.88330346, + "num_input_tokens_seen": 14581980, + "step": 690, + "time_per_iteration": 2.8099722862243652 + }, + { + "auxiliary_loss_clip": 0.01259552, + "auxiliary_loss_mlp": 0.01049353, + "balance_loss_clip": 1.07293797, + "balance_loss_mlp": 1.03393888, + "epoch": 0.08308783743161186, + "flos": 18004928567040.0, + "grad_norm": 1.8548995546674607, + "language_loss": 0.82366198, + "learning_rate": 3.970569848946847e-06, + "loss": 0.84675097, + "num_input_tokens_seen": 14601795, + "step": 691, + "time_per_iteration": 2.7832839488983154 + }, + { + "auxiliary_loss_clip": 0.01243723, + "auxiliary_loss_mlp": 0.01054617, + "balance_loss_clip": 1.06911135, + "balance_loss_mlp": 1.03926313, + "epoch": 0.08320808032225095, + "flos": 15079788599040.0, + "grad_norm": 2.0569081809624894, + "language_loss": 0.82591188, + "learning_rate": 3.970436557714932e-06, + "loss": 0.84889531, + "num_input_tokens_seen": 14618315, + "step": 692, + "time_per_iteration": 2.7441654205322266 + }, + { + "auxiliary_loss_clip": 0.01243948, + "auxiliary_loss_mlp": 0.01054067, + "balance_loss_clip": 1.06988859, + "balance_loss_mlp": 1.03848648, + "epoch": 0.08332832321289003, + "flos": 22383085501440.0, + "grad_norm": 2.4230430553197406, + "language_loss": 0.86379242, + "learning_rate": 3.970302967569358e-06, + "loss": 0.88677257, + "num_input_tokens_seen": 14636905, + "step": 693, + "time_per_iteration": 2.826948404312134 + }, + { + "auxiliary_loss_clip": 0.0125313, + "auxiliary_loss_mlp": 0.0104151, + "balance_loss_clip": 1.07292664, + "balance_loss_mlp": 1.02527416, + "epoch": 0.08344856610352913, + "flos": 24717386655360.0, + "grad_norm": 2.249193218914879, + "language_loss": 0.68142027, + "learning_rate": 3.9701690785303896e-06, + "loss": 0.70436668, + "num_input_tokens_seen": 14656100, + "step": 694, + "time_per_iteration": 2.7581539154052734 + }, + { + "auxiliary_loss_clip": 0.01253003, + "auxiliary_loss_mlp": 0.01052174, + "balance_loss_clip": 1.06833839, + "balance_loss_mlp": 1.03659344, + "epoch": 0.08356880899416821, + "flos": 25370206387200.0, + "grad_norm": 2.896748973979637, + "language_loss": 0.8812387, + "learning_rate": 3.970034890618339e-06, + "loss": 0.9042905, + "num_input_tokens_seen": 14675790, + "step": 695, + "time_per_iteration": 2.7655749320983887 + }, + { + "auxiliary_loss_clip": 0.01237717, + "auxiliary_loss_mlp": 0.01052135, + "balance_loss_clip": 1.06521118, + "balance_loss_mlp": 1.03633976, + "epoch": 0.08368905188480731, + "flos": 24353072962560.0, + "grad_norm": 2.137389962608884, + "language_loss": 0.8823204, + "learning_rate": 3.969900403853562e-06, + "loss": 0.90521896, + "num_input_tokens_seen": 14694830, + "step": 696, + "time_per_iteration": 4.6304662227630615 + }, + { + "auxiliary_loss_clip": 0.01265878, + "auxiliary_loss_mlp": 0.01052829, + "balance_loss_clip": 1.07371807, + "balance_loss_mlp": 1.037058, + "epoch": 0.08380929477544641, + "flos": 18037319656320.0, + "grad_norm": 1.6405637932685189, + "language_loss": 0.7814756, + "learning_rate": 3.96976561825646e-06, + "loss": 0.80466264, + "num_input_tokens_seen": 14711920, + "step": 697, + "time_per_iteration": 3.766326665878296 + }, + { + "auxiliary_loss_clip": 0.01227456, + "auxiliary_loss_mlp": 0.01053713, + "balance_loss_clip": 1.0668757, + "balance_loss_mlp": 1.03797698, + "epoch": 0.08392953766608549, + "flos": 26286287875200.0, + "grad_norm": 1.987821627010404, + "language_loss": 0.8703289, + "learning_rate": 3.969630533847479e-06, + "loss": 0.89314055, + "num_input_tokens_seen": 14730880, + "step": 698, + "time_per_iteration": 3.782953977584839 + }, + { + "auxiliary_loss_clip": 0.01251841, + "auxiliary_loss_mlp": 0.01047438, + "balance_loss_clip": 1.06903744, + "balance_loss_mlp": 1.03233457, + "epoch": 0.08404978055672459, + "flos": 22492146170880.0, + "grad_norm": 1.909738908103795, + "language_loss": 0.83971685, + "learning_rate": 3.969495150647113e-06, + "loss": 0.86270964, + "num_input_tokens_seen": 14749050, + "step": 699, + "time_per_iteration": 2.756000518798828 + }, + { + "auxiliary_loss_clip": 0.01228545, + "auxiliary_loss_mlp": 0.01048114, + "balance_loss_clip": 1.07011545, + "balance_loss_mlp": 1.03295016, + "epoch": 0.08417002344736367, + "flos": 24826878288000.0, + "grad_norm": 1.8656011993309534, + "language_loss": 0.76380259, + "learning_rate": 3.969359468675899e-06, + "loss": 0.78656918, + "num_input_tokens_seen": 14769180, + "step": 700, + "time_per_iteration": 2.845797538757324 + }, + { + "auxiliary_loss_clip": 0.01247168, + "auxiliary_loss_mlp": 0.01045859, + "balance_loss_clip": 1.06944299, + "balance_loss_mlp": 1.03083825, + "epoch": 0.08429026633800277, + "flos": 16945922862720.0, + "grad_norm": 1.9610867256141506, + "language_loss": 0.89157963, + "learning_rate": 3.969223487954418e-06, + "loss": 0.91450989, + "num_input_tokens_seen": 14786640, + "step": 701, + "time_per_iteration": 2.810748815536499 + }, + { + "auxiliary_loss_clip": 0.01217231, + "auxiliary_loss_mlp": 0.01048039, + "balance_loss_clip": 1.06574523, + "balance_loss_mlp": 1.03276789, + "epoch": 0.08441050922864185, + "flos": 23841920471040.0, + "grad_norm": 2.7355030573542933, + "language_loss": 0.82542014, + "learning_rate": 3.969087208503301e-06, + "loss": 0.84807289, + "num_input_tokens_seen": 14806720, + "step": 702, + "time_per_iteration": 2.951634645462036 + }, + { + "auxiliary_loss_clip": 0.01220659, + "auxiliary_loss_mlp": 0.01039164, + "balance_loss_clip": 1.06707644, + "balance_loss_mlp": 1.02335691, + "epoch": 0.08453075211928095, + "flos": 25520205582720.0, + "grad_norm": 2.5621947161100116, + "language_loss": 0.84443474, + "learning_rate": 3.968950630343219e-06, + "loss": 0.86703295, + "num_input_tokens_seen": 14823705, + "step": 703, + "time_per_iteration": 2.886800765991211 + }, + { + "auxiliary_loss_clip": 0.0123302, + "auxiliary_loss_mlp": 0.01043295, + "balance_loss_clip": 1.06374633, + "balance_loss_mlp": 1.02826214, + "epoch": 0.08465099500992004, + "flos": 19532496211200.0, + "grad_norm": 2.1004163290577527, + "language_loss": 0.93989098, + "learning_rate": 3.968813753494892e-06, + "loss": 0.96265417, + "num_input_tokens_seen": 14841865, + "step": 704, + "time_per_iteration": 2.841679334640503 + }, + { + "auxiliary_loss_clip": 0.01212866, + "auxiliary_loss_mlp": 0.00890057, + "balance_loss_clip": 1.06013358, + "balance_loss_mlp": 1.00026679, + "epoch": 0.08477123790055913, + "flos": 29351299403520.0, + "grad_norm": 2.301642356911419, + "language_loss": 0.76050764, + "learning_rate": 3.968676577979084e-06, + "loss": 0.78153688, + "num_input_tokens_seen": 14861415, + "step": 705, + "time_per_iteration": 2.9518544673919678 + }, + { + "auxiliary_loss_clip": 0.01213729, + "auxiliary_loss_mlp": 0.01049014, + "balance_loss_clip": 1.06160772, + "balance_loss_mlp": 1.03348064, + "epoch": 0.08489148079119822, + "flos": 18624495283200.0, + "grad_norm": 2.167558242517389, + "language_loss": 0.78368938, + "learning_rate": 3.968539103816605e-06, + "loss": 0.80631673, + "num_input_tokens_seen": 14879215, + "step": 706, + "time_per_iteration": 2.878042459487915 + }, + { + "auxiliary_loss_clip": 0.01234034, + "auxiliary_loss_mlp": 0.00889628, + "balance_loss_clip": 1.06678534, + "balance_loss_mlp": 1.00031805, + "epoch": 0.0850117236818373, + "flos": 23471393725440.0, + "grad_norm": 2.060428452485985, + "language_loss": 0.89346349, + "learning_rate": 3.9684013310283085e-06, + "loss": 0.91470015, + "num_input_tokens_seen": 14897900, + "step": 707, + "time_per_iteration": 2.7693824768066406 + }, + { + "auxiliary_loss_clip": 0.01239386, + "auxiliary_loss_mlp": 0.01051723, + "balance_loss_clip": 1.07077432, + "balance_loss_mlp": 1.03722751, + "epoch": 0.0851319665724764, + "flos": 40625058896640.0, + "grad_norm": 1.8121329050978008, + "language_loss": 0.6402061, + "learning_rate": 3.9682632596350956e-06, + "loss": 0.66311723, + "num_input_tokens_seen": 14919065, + "step": 708, + "time_per_iteration": 2.896587610244751 + }, + { + "auxiliary_loss_clip": 0.01248374, + "auxiliary_loss_mlp": 0.01044395, + "balance_loss_clip": 1.07103157, + "balance_loss_mlp": 1.02849221, + "epoch": 0.0852522094631155, + "flos": 15879554870400.0, + "grad_norm": 2.0904203745410572, + "language_loss": 0.7871626, + "learning_rate": 3.968124889657911e-06, + "loss": 0.81009024, + "num_input_tokens_seen": 14934165, + "step": 709, + "time_per_iteration": 2.7143921852111816 + }, + { + "auxiliary_loss_clip": 0.01214578, + "auxiliary_loss_mlp": 0.01044188, + "balance_loss_clip": 1.06254065, + "balance_loss_mlp": 1.02880979, + "epoch": 0.08537245235375458, + "flos": 14567091822720.0, + "grad_norm": 2.6144728555198755, + "language_loss": 0.90902781, + "learning_rate": 3.967986221117746e-06, + "loss": 0.93161547, + "num_input_tokens_seen": 14950105, + "step": 710, + "time_per_iteration": 2.753303289413452 + }, + { + "auxiliary_loss_clip": 0.01201822, + "auxiliary_loss_mlp": 0.01054803, + "balance_loss_clip": 1.06305432, + "balance_loss_mlp": 1.04011691, + "epoch": 0.08549269524439368, + "flos": 26468929555200.0, + "grad_norm": 1.8965699508437053, + "language_loss": 0.86574948, + "learning_rate": 3.967847254035635e-06, + "loss": 0.8883158, + "num_input_tokens_seen": 14969490, + "step": 711, + "time_per_iteration": 2.9861793518066406 + }, + { + "auxiliary_loss_clip": 0.01228059, + "auxiliary_loss_mlp": 0.010485, + "balance_loss_clip": 1.0673511, + "balance_loss_mlp": 1.03224003, + "epoch": 0.08561293813503276, + "flos": 13590214565760.0, + "grad_norm": 2.5980975709559635, + "language_loss": 0.86460239, + "learning_rate": 3.967707988432661e-06, + "loss": 0.88736796, + "num_input_tokens_seen": 14987195, + "step": 712, + "time_per_iteration": 2.6962106227874756 + }, + { + "auxiliary_loss_clip": 0.01262545, + "auxiliary_loss_mlp": 0.0104441, + "balance_loss_clip": 1.07106948, + "balance_loss_mlp": 1.02874589, + "epoch": 0.08573318102567186, + "flos": 26943524979840.0, + "grad_norm": 10.222617934423562, + "language_loss": 0.87754905, + "learning_rate": 3.967568424329949e-06, + "loss": 0.90061867, + "num_input_tokens_seen": 15007620, + "step": 713, + "time_per_iteration": 2.7704014778137207 + }, + { + "auxiliary_loss_clip": 0.0115106, + "auxiliary_loss_mlp": 0.0101551, + "balance_loss_clip": 1.04182696, + "balance_loss_mlp": 1.00726056, + "epoch": 0.08585342391631094, + "flos": 67302739319040.0, + "grad_norm": 0.8198391960484412, + "language_loss": 0.55531979, + "learning_rate": 3.967428561748671e-06, + "loss": 0.57698554, + "num_input_tokens_seen": 15075590, + "step": 714, + "time_per_iteration": 3.488640308380127 + }, + { + "auxiliary_loss_clip": 0.01214832, + "auxiliary_loss_mlp": 0.01054243, + "balance_loss_clip": 1.06147146, + "balance_loss_mlp": 1.0372436, + "epoch": 0.08597366680695004, + "flos": 22456594684800.0, + "grad_norm": 2.284644475267123, + "language_loss": 0.87410671, + "learning_rate": 3.967288400710045e-06, + "loss": 0.89679742, + "num_input_tokens_seen": 15095055, + "step": 715, + "time_per_iteration": 2.9031553268432617 + }, + { + "auxiliary_loss_clip": 0.01225658, + "auxiliary_loss_mlp": 0.01045542, + "balance_loss_clip": 1.06864786, + "balance_loss_mlp": 1.0313319, + "epoch": 0.08609390969758914, + "flos": 23550505430400.0, + "grad_norm": 1.9987467984304697, + "language_loss": 0.88621873, + "learning_rate": 3.9671479412353335e-06, + "loss": 0.90893066, + "num_input_tokens_seen": 15113520, + "step": 716, + "time_per_iteration": 2.904381036758423 + }, + { + "auxiliary_loss_clip": 0.01253372, + "auxiliary_loss_mlp": 0.01050226, + "balance_loss_clip": 1.07122564, + "balance_loss_mlp": 1.03536057, + "epoch": 0.08621415258822822, + "flos": 25885848078720.0, + "grad_norm": 1.958035223987232, + "language_loss": 0.74401385, + "learning_rate": 3.967007183345843e-06, + "loss": 0.76704979, + "num_input_tokens_seen": 15133375, + "step": 717, + "time_per_iteration": 2.9307799339294434 + }, + { + "auxiliary_loss_clip": 0.01250395, + "auxiliary_loss_mlp": 0.01050028, + "balance_loss_clip": 1.07259965, + "balance_loss_mlp": 1.03451848, + "epoch": 0.08633439547886732, + "flos": 13589568120960.0, + "grad_norm": 2.1084163770438606, + "language_loss": 0.89270991, + "learning_rate": 3.966866127062927e-06, + "loss": 0.91571414, + "num_input_tokens_seen": 15150500, + "step": 718, + "time_per_iteration": 2.7656142711639404 + }, + { + "auxiliary_loss_clip": 0.01150602, + "auxiliary_loss_mlp": 0.01010539, + "balance_loss_clip": 1.04214311, + "balance_loss_mlp": 1.00224173, + "epoch": 0.0864546383695064, + "flos": 57767342434560.0, + "grad_norm": 0.8651055534644017, + "language_loss": 0.62647694, + "learning_rate": 3.966724772407982e-06, + "loss": 0.64808834, + "num_input_tokens_seen": 15208015, + "step": 719, + "time_per_iteration": 3.2151246070861816 + }, + { + "auxiliary_loss_clip": 0.01226241, + "auxiliary_loss_mlp": 0.01048802, + "balance_loss_clip": 1.0680337, + "balance_loss_mlp": 1.03325737, + "epoch": 0.0865748812601455, + "flos": 20046952753920.0, + "grad_norm": 2.0739073298814144, + "language_loss": 0.88798052, + "learning_rate": 3.966583119402454e-06, + "loss": 0.91073096, + "num_input_tokens_seen": 15224780, + "step": 720, + "time_per_iteration": 2.804593324661255 + }, + { + "auxiliary_loss_clip": 0.01250927, + "auxiliary_loss_mlp": 0.00889824, + "balance_loss_clip": 1.07203472, + "balance_loss_mlp": 1.00015903, + "epoch": 0.08669512415078459, + "flos": 35262446935680.0, + "grad_norm": 1.7720718826875812, + "language_loss": 0.82007074, + "learning_rate": 3.9664411680678305e-06, + "loss": 0.84147823, + "num_input_tokens_seen": 15246535, + "step": 721, + "time_per_iteration": 2.8489151000976562 + }, + { + "auxiliary_loss_clip": 0.01131747, + "auxiliary_loss_mlp": 0.01009009, + "balance_loss_clip": 1.03736937, + "balance_loss_mlp": 1.00061631, + "epoch": 0.08681536704142367, + "flos": 65654870048640.0, + "grad_norm": 4.743420288984984, + "language_loss": 0.61486244, + "learning_rate": 3.966298918425644e-06, + "loss": 0.63627005, + "num_input_tokens_seen": 15304025, + "step": 722, + "time_per_iteration": 5.048830032348633 + }, + { + "auxiliary_loss_clip": 0.01256257, + "auxiliary_loss_mlp": 0.01045674, + "balance_loss_clip": 1.07150578, + "balance_loss_mlp": 1.03066611, + "epoch": 0.08693560993206277, + "flos": 34529940881280.0, + "grad_norm": 1.819715782779334, + "language_loss": 0.82762706, + "learning_rate": 3.966156370497476e-06, + "loss": 0.85064638, + "num_input_tokens_seen": 15327635, + "step": 723, + "time_per_iteration": 4.6660685539245605 + }, + { + "auxiliary_loss_clip": 0.01255914, + "auxiliary_loss_mlp": 0.01048733, + "balance_loss_clip": 1.07083941, + "balance_loss_mlp": 1.03325939, + "epoch": 0.08705585282270185, + "flos": 23149419189120.0, + "grad_norm": 1.7196847817414032, + "language_loss": 0.88790882, + "learning_rate": 3.96601352430495e-06, + "loss": 0.91095531, + "num_input_tokens_seen": 15347405, + "step": 724, + "time_per_iteration": 2.797360420227051 + }, + { + "auxiliary_loss_clip": 0.01240559, + "auxiliary_loss_mlp": 0.01052149, + "balance_loss_clip": 1.07158256, + "balance_loss_mlp": 1.037045, + "epoch": 0.08717609571334095, + "flos": 29497599498240.0, + "grad_norm": 1.759492870524377, + "language_loss": 0.83303159, + "learning_rate": 3.965870379869735e-06, + "loss": 0.8559587, + "num_input_tokens_seen": 15369450, + "step": 725, + "time_per_iteration": 2.841618776321411 + }, + { + "auxiliary_loss_clip": 0.01249328, + "auxiliary_loss_mlp": 0.01044802, + "balance_loss_clip": 1.0676477, + "balance_loss_mlp": 1.03022289, + "epoch": 0.08729633860398003, + "flos": 20667489137280.0, + "grad_norm": 2.2226019272790545, + "language_loss": 0.86870241, + "learning_rate": 3.965726937213547e-06, + "loss": 0.89164376, + "num_input_tokens_seen": 15388085, + "step": 726, + "time_per_iteration": 2.7104084491729736 + }, + { + "auxiliary_loss_clip": 0.01247056, + "auxiliary_loss_mlp": 0.01050974, + "balance_loss_clip": 1.06569207, + "balance_loss_mlp": 1.03408194, + "epoch": 0.08741658149461913, + "flos": 18369493655040.0, + "grad_norm": 2.513804224681894, + "language_loss": 0.81005114, + "learning_rate": 3.965583196358144e-06, + "loss": 0.83303148, + "num_input_tokens_seen": 15407120, + "step": 727, + "time_per_iteration": 2.733971118927002 + }, + { + "auxiliary_loss_clip": 0.01263988, + "auxiliary_loss_mlp": 0.0105259, + "balance_loss_clip": 1.07202077, + "balance_loss_mlp": 1.03584087, + "epoch": 0.08753682438525823, + "flos": 18729677283840.0, + "grad_norm": 2.360456507690023, + "language_loss": 0.74797571, + "learning_rate": 3.965439157325335e-06, + "loss": 0.77114153, + "num_input_tokens_seen": 15424485, + "step": 728, + "time_per_iteration": 2.6735479831695557 + }, + { + "auxiliary_loss_clip": 0.01229157, + "auxiliary_loss_mlp": 0.01048833, + "balance_loss_clip": 1.06232357, + "balance_loss_mlp": 1.0335381, + "epoch": 0.08765706727589731, + "flos": 27776113303680.0, + "grad_norm": 2.5479432139177964, + "language_loss": 0.76106459, + "learning_rate": 3.965294820136968e-06, + "loss": 0.78384447, + "num_input_tokens_seen": 15446285, + "step": 729, + "time_per_iteration": 2.9186699390411377 + }, + { + "auxiliary_loss_clip": 0.01239294, + "auxiliary_loss_mlp": 0.01051399, + "balance_loss_clip": 1.06761718, + "balance_loss_mlp": 1.03550863, + "epoch": 0.08777731016653641, + "flos": 24389127239040.0, + "grad_norm": 2.0452850187666662, + "language_loss": 0.86719781, + "learning_rate": 3.965150184814938e-06, + "loss": 0.89010471, + "num_input_tokens_seen": 15465770, + "step": 730, + "time_per_iteration": 2.7695956230163574 + }, + { + "auxiliary_loss_clip": 0.01219946, + "auxiliary_loss_mlp": 0.01045501, + "balance_loss_clip": 1.06213856, + "balance_loss_mlp": 1.0299921, + "epoch": 0.08789755305717549, + "flos": 21981855605760.0, + "grad_norm": 2.107178943412937, + "language_loss": 0.7629174, + "learning_rate": 3.965005251381189e-06, + "loss": 0.78557187, + "num_input_tokens_seen": 15483705, + "step": 731, + "time_per_iteration": 2.9589526653289795 + }, + { + "auxiliary_loss_clip": 0.01153789, + "auxiliary_loss_mlp": 0.01010635, + "balance_loss_clip": 1.0418489, + "balance_loss_mlp": 1.00252903, + "epoch": 0.08801779594781459, + "flos": 58360120583040.0, + "grad_norm": 0.8792805446981076, + "language_loss": 0.64526153, + "learning_rate": 3.964860019857705e-06, + "loss": 0.66690576, + "num_input_tokens_seen": 15548620, + "step": 732, + "time_per_iteration": 3.345219373703003 + }, + { + "auxiliary_loss_clip": 0.01262951, + "auxiliary_loss_mlp": 0.01051046, + "balance_loss_clip": 1.07554138, + "balance_loss_mlp": 1.03585887, + "epoch": 0.08813803883845367, + "flos": 23294785530240.0, + "grad_norm": 1.903556882070955, + "language_loss": 0.83789992, + "learning_rate": 3.964714490266518e-06, + "loss": 0.86103988, + "num_input_tokens_seen": 15569265, + "step": 733, + "time_per_iteration": 2.7033324241638184 + }, + { + "auxiliary_loss_clip": 0.01141576, + "auxiliary_loss_mlp": 0.01008234, + "balance_loss_clip": 1.03690815, + "balance_loss_mlp": 1.00017536, + "epoch": 0.08825828172909277, + "flos": 63424924882560.0, + "grad_norm": 0.8794022904369317, + "language_loss": 0.64646107, + "learning_rate": 3.964568662629706e-06, + "loss": 0.66795921, + "num_input_tokens_seen": 15630570, + "step": 734, + "time_per_iteration": 3.2159390449523926 + }, + { + "auxiliary_loss_clip": 0.01245112, + "auxiliary_loss_mlp": 0.01044852, + "balance_loss_clip": 1.06596863, + "balance_loss_mlp": 1.02933061, + "epoch": 0.08837852461973186, + "flos": 26720986268160.0, + "grad_norm": 2.505355447290584, + "language_loss": 0.84323835, + "learning_rate": 3.9644225369693895e-06, + "loss": 0.86613798, + "num_input_tokens_seen": 15650870, + "step": 735, + "time_per_iteration": 2.7939021587371826 + }, + { + "auxiliary_loss_clip": 0.01264552, + "auxiliary_loss_mlp": 0.01056751, + "balance_loss_clip": 1.07432532, + "balance_loss_mlp": 1.04175425, + "epoch": 0.08849876751037095, + "flos": 27265427688960.0, + "grad_norm": 2.0910630397221146, + "language_loss": 0.86740768, + "learning_rate": 3.964276113307735e-06, + "loss": 0.89062071, + "num_input_tokens_seen": 15670835, + "step": 736, + "time_per_iteration": 2.725900411605835 + }, + { + "auxiliary_loss_clip": 0.01225693, + "auxiliary_loss_mlp": 0.01049418, + "balance_loss_clip": 1.06655812, + "balance_loss_mlp": 1.0334444, + "epoch": 0.08861901040101004, + "flos": 19828759587840.0, + "grad_norm": 2.3390654828339037, + "language_loss": 0.80935633, + "learning_rate": 3.9641293916669574e-06, + "loss": 0.83210742, + "num_input_tokens_seen": 15689795, + "step": 737, + "time_per_iteration": 2.799311399459839 + }, + { + "auxiliary_loss_clip": 0.01224926, + "auxiliary_loss_mlp": 0.01050131, + "balance_loss_clip": 1.0656147, + "balance_loss_mlp": 1.03453803, + "epoch": 0.08873925329164913, + "flos": 23658704173440.0, + "grad_norm": 2.071060695469302, + "language_loss": 0.82528782, + "learning_rate": 3.9639823720693115e-06, + "loss": 0.84803838, + "num_input_tokens_seen": 15711650, + "step": 738, + "time_per_iteration": 2.848102331161499 + }, + { + "auxiliary_loss_clip": 0.01120148, + "auxiliary_loss_mlp": 0.01015793, + "balance_loss_clip": 1.03265738, + "balance_loss_mlp": 1.00806773, + "epoch": 0.08885949618228822, + "flos": 71831541893760.0, + "grad_norm": 0.916713494592456, + "language_loss": 0.6001687, + "learning_rate": 3.963835054537102e-06, + "loss": 0.62152815, + "num_input_tokens_seen": 15780615, + "step": 739, + "time_per_iteration": 3.469749927520752 + }, + { + "auxiliary_loss_clip": 0.0123237, + "auxiliary_loss_mlp": 0.01048948, + "balance_loss_clip": 1.06235158, + "balance_loss_mlp": 1.03376102, + "epoch": 0.08897973907292732, + "flos": 22346169298560.0, + "grad_norm": 3.2411327679199755, + "language_loss": 0.60701305, + "learning_rate": 3.963687439092676e-06, + "loss": 0.62982619, + "num_input_tokens_seen": 15801300, + "step": 740, + "time_per_iteration": 2.8464765548706055 + }, + { + "auxiliary_loss_clip": 0.01249142, + "auxiliary_loss_mlp": 0.01046719, + "balance_loss_clip": 1.07058835, + "balance_loss_mlp": 1.03011346, + "epoch": 0.0890999819635664, + "flos": 21251827589760.0, + "grad_norm": 2.1462698083995164, + "language_loss": 0.80633199, + "learning_rate": 3.963539525758427e-06, + "loss": 0.82929063, + "num_input_tokens_seen": 15820860, + "step": 741, + "time_per_iteration": 2.810689687728882 + }, + { + "auxiliary_loss_clip": 0.01241394, + "auxiliary_loss_mlp": 0.01046633, + "balance_loss_clip": 1.06906974, + "balance_loss_mlp": 1.03114724, + "epoch": 0.0892202248542055, + "flos": 25370888745600.0, + "grad_norm": 2.3412926802623413, + "language_loss": 0.67978942, + "learning_rate": 3.9633913145567925e-06, + "loss": 0.70266968, + "num_input_tokens_seen": 15841350, + "step": 742, + "time_per_iteration": 2.8154871463775635 + }, + { + "auxiliary_loss_clip": 0.01242019, + "auxiliary_loss_mlp": 0.01051808, + "balance_loss_clip": 1.07020617, + "balance_loss_mlp": 1.03666878, + "epoch": 0.08934046774484458, + "flos": 24457895827200.0, + "grad_norm": 2.1157596529425597, + "language_loss": 0.81730306, + "learning_rate": 3.9632428055102575e-06, + "loss": 0.84024131, + "num_input_tokens_seen": 15861360, + "step": 743, + "time_per_iteration": 2.834132194519043 + }, + { + "auxiliary_loss_clip": 0.01256936, + "auxiliary_loss_mlp": 0.01049514, + "balance_loss_clip": 1.07483578, + "balance_loss_mlp": 1.0333848, + "epoch": 0.08946071063548368, + "flos": 35772773414400.0, + "grad_norm": 2.9475059985531873, + "language_loss": 0.664895, + "learning_rate": 3.9630939986413495e-06, + "loss": 0.68795943, + "num_input_tokens_seen": 15883160, + "step": 744, + "time_per_iteration": 2.8705251216888428 + }, + { + "auxiliary_loss_clip": 0.01205453, + "auxiliary_loss_mlp": 0.01050318, + "balance_loss_clip": 1.05827153, + "balance_loss_mlp": 1.03538084, + "epoch": 0.08958095352612276, + "flos": 14356584167040.0, + "grad_norm": 1.6295743414995072, + "language_loss": 0.7820884, + "learning_rate": 3.962944893972643e-06, + "loss": 0.80464613, + "num_input_tokens_seen": 15901610, + "step": 745, + "time_per_iteration": 2.7668344974517822 + }, + { + "auxiliary_loss_clip": 0.01242797, + "auxiliary_loss_mlp": 0.010407, + "balance_loss_clip": 1.07164598, + "balance_loss_mlp": 1.02529764, + "epoch": 0.08970119641676186, + "flos": 17853277345920.0, + "grad_norm": 2.4882475485583373, + "language_loss": 0.90246773, + "learning_rate": 3.962795491526756e-06, + "loss": 0.92530274, + "num_input_tokens_seen": 15918770, + "step": 746, + "time_per_iteration": 2.840421199798584 + }, + { + "auxiliary_loss_clip": 0.0126863, + "auxiliary_loss_mlp": 0.01046844, + "balance_loss_clip": 1.07525289, + "balance_loss_mlp": 1.03035736, + "epoch": 0.08982143930740095, + "flos": 20811670329600.0, + "grad_norm": 2.5726390497993297, + "language_loss": 0.8911835, + "learning_rate": 3.962645791326354e-06, + "loss": 0.91433829, + "num_input_tokens_seen": 15938025, + "step": 747, + "time_per_iteration": 2.6955862045288086 + }, + { + "auxiliary_loss_clip": 0.01248218, + "auxiliary_loss_mlp": 0.01048808, + "balance_loss_clip": 1.07182777, + "balance_loss_mlp": 1.03347814, + "epoch": 0.08994168219804004, + "flos": 24097712198400.0, + "grad_norm": 1.8218119234415362, + "language_loss": 0.8307004, + "learning_rate": 3.962495793394146e-06, + "loss": 0.85367072, + "num_input_tokens_seen": 15957215, + "step": 748, + "time_per_iteration": 3.8246419429779053 + }, + { + "auxiliary_loss_clip": 0.01150981, + "auxiliary_loss_mlp": 0.01016462, + "balance_loss_clip": 1.03551662, + "balance_loss_mlp": 1.00883234, + "epoch": 0.09006192508867913, + "flos": 57188893812480.0, + "grad_norm": 0.7532323873800524, + "language_loss": 0.61186814, + "learning_rate": 3.9623454977528864e-06, + "loss": 0.63354254, + "num_input_tokens_seen": 16015870, + "step": 749, + "time_per_iteration": 3.9912233352661133 + }, + { + "auxiliary_loss_clip": 0.01236097, + "auxiliary_loss_mlp": 0.01051616, + "balance_loss_clip": 1.06763244, + "balance_loss_mlp": 1.03471184, + "epoch": 0.09018216797931822, + "flos": 20487505063680.0, + "grad_norm": 1.7732121213290595, + "language_loss": 0.85015464, + "learning_rate": 3.962194904425375e-06, + "loss": 0.87303174, + "num_input_tokens_seen": 16036500, + "step": 750, + "time_per_iteration": 3.8077704906463623 + }, + { + "auxiliary_loss_clip": 0.01249869, + "auxiliary_loss_mlp": 0.01050171, + "balance_loss_clip": 1.07092607, + "balance_loss_mlp": 1.03530526, + "epoch": 0.09030241086995731, + "flos": 22638123043200.0, + "grad_norm": 1.964034508636556, + "language_loss": 0.68122107, + "learning_rate": 3.9620440134344566e-06, + "loss": 0.70422149, + "num_input_tokens_seen": 16054655, + "step": 751, + "time_per_iteration": 2.769324779510498 + }, + { + "auxiliary_loss_clip": 0.01231926, + "auxiliary_loss_mlp": 0.01047103, + "balance_loss_clip": 1.06893325, + "balance_loss_mlp": 1.03024626, + "epoch": 0.09042265376059641, + "flos": 21871502046720.0, + "grad_norm": 3.0269852307706, + "language_loss": 0.82688522, + "learning_rate": 3.9618928248030215e-06, + "loss": 0.84967554, + "num_input_tokens_seen": 16074165, + "step": 752, + "time_per_iteration": 2.9069581031799316 + }, + { + "auxiliary_loss_clip": 0.01251636, + "auxiliary_loss_mlp": 0.01056555, + "balance_loss_clip": 1.07307613, + "balance_loss_mlp": 1.04102242, + "epoch": 0.0905428966512355, + "flos": 24316192673280.0, + "grad_norm": 2.312554853034376, + "language_loss": 0.82879514, + "learning_rate": 3.961741338554005e-06, + "loss": 0.85187703, + "num_input_tokens_seen": 16092505, + "step": 753, + "time_per_iteration": 2.7411866188049316 + }, + { + "auxiliary_loss_clip": 0.01251567, + "auxiliary_loss_mlp": 0.01046701, + "balance_loss_clip": 1.07234824, + "balance_loss_mlp": 1.03082275, + "epoch": 0.09066313954187459, + "flos": 35845061535360.0, + "grad_norm": 3.2658051822889917, + "language_loss": 0.7553432, + "learning_rate": 3.9615895547103865e-06, + "loss": 0.7783258, + "num_input_tokens_seen": 16116150, + "step": 754, + "time_per_iteration": 2.9112093448638916 + }, + { + "auxiliary_loss_clip": 0.01236959, + "auxiliary_loss_mlp": 0.01043806, + "balance_loss_clip": 1.06869316, + "balance_loss_mlp": 1.0282371, + "epoch": 0.09078338243251367, + "flos": 29168729550720.0, + "grad_norm": 1.9317246333045732, + "language_loss": 0.77869391, + "learning_rate": 3.961437473295193e-06, + "loss": 0.80150151, + "num_input_tokens_seen": 16136295, + "step": 755, + "time_per_iteration": 2.848872184753418 + }, + { + "auxiliary_loss_clip": 0.01194076, + "auxiliary_loss_mlp": 0.01041365, + "balance_loss_clip": 1.05433965, + "balance_loss_mlp": 1.02602315, + "epoch": 0.09090362532315277, + "flos": 21907699977600.0, + "grad_norm": 2.40421992190154, + "language_loss": 0.72240269, + "learning_rate": 3.961285094331495e-06, + "loss": 0.74475706, + "num_input_tokens_seen": 16154210, + "step": 756, + "time_per_iteration": 2.9203197956085205 + }, + { + "auxiliary_loss_clip": 0.01257578, + "auxiliary_loss_mlp": 0.01045263, + "balance_loss_clip": 1.07050979, + "balance_loss_mlp": 1.03052843, + "epoch": 0.09102386821379185, + "flos": 27344503480320.0, + "grad_norm": 2.6265770652357143, + "language_loss": 0.86017573, + "learning_rate": 3.961132417842406e-06, + "loss": 0.88320422, + "num_input_tokens_seen": 16173995, + "step": 757, + "time_per_iteration": 2.802367925643921 + }, + { + "auxiliary_loss_clip": 0.01243355, + "auxiliary_loss_mlp": 0.01046975, + "balance_loss_clip": 1.06957626, + "balance_loss_mlp": 1.03081012, + "epoch": 0.09114411110443095, + "flos": 20813501923200.0, + "grad_norm": 3.7061543056371336, + "language_loss": 0.74870145, + "learning_rate": 3.960979443851089e-06, + "loss": 0.77160484, + "num_input_tokens_seen": 16191020, + "step": 758, + "time_per_iteration": 2.848604202270508 + }, + { + "auxiliary_loss_clip": 0.0123262, + "auxiliary_loss_mlp": 0.01047199, + "balance_loss_clip": 1.06653953, + "balance_loss_mlp": 1.03112972, + "epoch": 0.09126435399507005, + "flos": 26145949438080.0, + "grad_norm": 1.6636441808892477, + "language_loss": 0.79137176, + "learning_rate": 3.96082617238075e-06, + "loss": 0.81416994, + "num_input_tokens_seen": 16213645, + "step": 759, + "time_per_iteration": 2.8647572994232178 + }, + { + "auxiliary_loss_clip": 0.01240491, + "auxiliary_loss_mlp": 0.01051561, + "balance_loss_clip": 1.06949139, + "balance_loss_mlp": 1.03699374, + "epoch": 0.09138459688570913, + "flos": 24388911757440.0, + "grad_norm": 2.8658605668411243, + "language_loss": 0.80065811, + "learning_rate": 3.960672603454639e-06, + "loss": 0.8235786, + "num_input_tokens_seen": 16233625, + "step": 760, + "time_per_iteration": 2.8164784908294678 + }, + { + "auxiliary_loss_clip": 0.01246488, + "auxiliary_loss_mlp": 0.01046978, + "balance_loss_clip": 1.06895781, + "balance_loss_mlp": 1.03157663, + "epoch": 0.09150483977634823, + "flos": 21032664756480.0, + "grad_norm": 2.7131133264874756, + "language_loss": 0.76879913, + "learning_rate": 3.960518737096054e-06, + "loss": 0.79173374, + "num_input_tokens_seen": 16253255, + "step": 761, + "time_per_iteration": 2.8196334838867188 + }, + { + "auxiliary_loss_clip": 0.0125102, + "auxiliary_loss_mlp": 0.01046409, + "balance_loss_clip": 1.07306004, + "balance_loss_mlp": 1.03191352, + "epoch": 0.09162508266698731, + "flos": 22856998567680.0, + "grad_norm": 2.1168462221267013, + "language_loss": 0.73192084, + "learning_rate": 3.960364573328334e-06, + "loss": 0.75489521, + "num_input_tokens_seen": 16272580, + "step": 762, + "time_per_iteration": 2.812077283859253 + }, + { + "auxiliary_loss_clip": 0.01228014, + "auxiliary_loss_mlp": 0.01044835, + "balance_loss_clip": 1.06482899, + "balance_loss_mlp": 1.02893281, + "epoch": 0.0917453255576264, + "flos": 21724411852800.0, + "grad_norm": 2.68078478489033, + "language_loss": 0.88483316, + "learning_rate": 3.9602101121748675e-06, + "loss": 0.90756166, + "num_input_tokens_seen": 16293075, + "step": 763, + "time_per_iteration": 2.8058369159698486 + }, + { + "auxiliary_loss_clip": 0.01237198, + "auxiliary_loss_mlp": 0.01048427, + "balance_loss_clip": 1.07136106, + "balance_loss_mlp": 1.033705, + "epoch": 0.0918655684482655, + "flos": 14609215497600.0, + "grad_norm": 2.637602254295478, + "language_loss": 0.72877115, + "learning_rate": 3.960055353659085e-06, + "loss": 0.75162739, + "num_input_tokens_seen": 16310185, + "step": 764, + "time_per_iteration": 2.8046751022338867 + }, + { + "auxiliary_loss_clip": 0.01233412, + "auxiliary_loss_mlp": 0.01046472, + "balance_loss_clip": 1.07034039, + "balance_loss_mlp": 1.031106, + "epoch": 0.09198581133890459, + "flos": 23435016226560.0, + "grad_norm": 1.7999875509177825, + "language_loss": 0.8370679, + "learning_rate": 3.959900297804465e-06, + "loss": 0.85986674, + "num_input_tokens_seen": 16330355, + "step": 765, + "time_per_iteration": 2.771111011505127 + }, + { + "auxiliary_loss_clip": 0.0121757, + "auxiliary_loss_mlp": 0.0104758, + "balance_loss_clip": 1.06288505, + "balance_loss_mlp": 1.03234529, + "epoch": 0.09210605422954368, + "flos": 16795887753600.0, + "grad_norm": 2.36920350193241, + "language_loss": 0.77409554, + "learning_rate": 3.9597449446345276e-06, + "loss": 0.79674697, + "num_input_tokens_seen": 16347600, + "step": 766, + "time_per_iteration": 2.8039233684539795 + }, + { + "auxiliary_loss_clip": 0.01221847, + "auxiliary_loss_mlp": 0.01045552, + "balance_loss_clip": 1.06162834, + "balance_loss_mlp": 1.03010297, + "epoch": 0.09222629712018277, + "flos": 22674249146880.0, + "grad_norm": 2.6723183457505963, + "language_loss": 0.832609, + "learning_rate": 3.95958929417284e-06, + "loss": 0.85528302, + "num_input_tokens_seen": 16365755, + "step": 767, + "time_per_iteration": 2.78657865524292 + }, + { + "auxiliary_loss_clip": 0.01141232, + "auxiliary_loss_mlp": 0.01007743, + "balance_loss_clip": 1.03274679, + "balance_loss_mlp": 1.0002569, + "epoch": 0.09234654001082186, + "flos": 69976756327680.0, + "grad_norm": 0.7305392538811882, + "language_loss": 0.5878461, + "learning_rate": 3.9594333464430145e-06, + "loss": 0.60933584, + "num_input_tokens_seen": 16435245, + "step": 768, + "time_per_iteration": 3.4665029048919678 + }, + { + "auxiliary_loss_clip": 0.01179705, + "auxiliary_loss_mlp": 0.01044671, + "balance_loss_clip": 1.05318856, + "balance_loss_mlp": 1.02988935, + "epoch": 0.09246678290146094, + "flos": 20011437181440.0, + "grad_norm": 2.079652049689508, + "language_loss": 0.88062155, + "learning_rate": 3.959277101468709e-06, + "loss": 0.90286541, + "num_input_tokens_seen": 16454795, + "step": 769, + "time_per_iteration": 2.9897308349609375 + }, + { + "auxiliary_loss_clip": 0.01221796, + "auxiliary_loss_mlp": 0.01045217, + "balance_loss_clip": 1.06415749, + "balance_loss_mlp": 1.03041768, + "epoch": 0.09258702579210004, + "flos": 17747448900480.0, + "grad_norm": 2.8185995898033416, + "language_loss": 0.78713745, + "learning_rate": 3.959120559273624e-06, + "loss": 0.8098076, + "num_input_tokens_seen": 16472580, + "step": 770, + "time_per_iteration": 2.8231728076934814 + }, + { + "auxiliary_loss_clip": 0.01220573, + "auxiliary_loss_mlp": 0.01042041, + "balance_loss_clip": 1.06641209, + "balance_loss_mlp": 1.02641273, + "epoch": 0.09270726868273914, + "flos": 20886544229760.0, + "grad_norm": 1.9418668154022665, + "language_loss": 0.83655787, + "learning_rate": 3.958963719881509e-06, + "loss": 0.85918403, + "num_input_tokens_seen": 16490670, + "step": 771, + "time_per_iteration": 2.817777633666992 + }, + { + "auxiliary_loss_clip": 0.01250739, + "auxiliary_loss_mlp": 0.01038379, + "balance_loss_clip": 1.07293594, + "balance_loss_mlp": 1.02264321, + "epoch": 0.09282751157337822, + "flos": 17015697031680.0, + "grad_norm": 1.9719374991632677, + "language_loss": 0.93810499, + "learning_rate": 3.958806583316154e-06, + "loss": 0.96099621, + "num_input_tokens_seen": 16508640, + "step": 772, + "time_per_iteration": 2.7540085315704346 + }, + { + "auxiliary_loss_clip": 0.01261706, + "auxiliary_loss_mlp": 0.01047362, + "balance_loss_clip": 1.07381451, + "balance_loss_mlp": 1.03211474, + "epoch": 0.09294775446401732, + "flos": 32523647748480.0, + "grad_norm": 1.807607978760691, + "language_loss": 0.7900939, + "learning_rate": 3.9586491496013985e-06, + "loss": 0.81318462, + "num_input_tokens_seen": 16531035, + "step": 773, + "time_per_iteration": 2.8872933387756348 + }, + { + "auxiliary_loss_clip": 0.01252642, + "auxiliary_loss_mlp": 0.01059399, + "balance_loss_clip": 1.07149374, + "balance_loss_mlp": 1.04400885, + "epoch": 0.0930679973546564, + "flos": 18259750627200.0, + "grad_norm": 1.9745247574264322, + "language_loss": 0.83099091, + "learning_rate": 3.958491418761124e-06, + "loss": 0.85411125, + "num_input_tokens_seen": 16548605, + "step": 774, + "time_per_iteration": 3.714738607406616 + }, + { + "auxiliary_loss_clip": 0.0123443, + "auxiliary_loss_mlp": 0.01047529, + "balance_loss_clip": 1.06103134, + "balance_loss_mlp": 1.03222275, + "epoch": 0.0931882402452955, + "flos": 21099745405440.0, + "grad_norm": 3.082745285544825, + "language_loss": 0.7273016, + "learning_rate": 3.958333390819258e-06, + "loss": 0.75012124, + "num_input_tokens_seen": 16565535, + "step": 775, + "time_per_iteration": 3.831162929534912 + }, + { + "auxiliary_loss_clip": 0.01257778, + "auxiliary_loss_mlp": 0.01047669, + "balance_loss_clip": 1.07239914, + "balance_loss_mlp": 1.03307807, + "epoch": 0.0933084831359346, + "flos": 24207275658240.0, + "grad_norm": 2.1299005545179672, + "language_loss": 0.80275846, + "learning_rate": 3.9581750657997754e-06, + "loss": 0.82581294, + "num_input_tokens_seen": 16584900, + "step": 776, + "time_per_iteration": 3.735755205154419 + }, + { + "auxiliary_loss_clip": 0.01231699, + "auxiliary_loss_mlp": 0.01038537, + "balance_loss_clip": 1.06424844, + "balance_loss_mlp": 1.02343297, + "epoch": 0.09342872602657368, + "flos": 25480272637440.0, + "grad_norm": 1.7912268136039042, + "language_loss": 0.89583385, + "learning_rate": 3.95801644372669e-06, + "loss": 0.91853619, + "num_input_tokens_seen": 16604805, + "step": 777, + "time_per_iteration": 2.8621530532836914 + }, + { + "auxiliary_loss_clip": 0.0124353, + "auxiliary_loss_mlp": 0.01045278, + "balance_loss_clip": 1.06521606, + "balance_loss_mlp": 1.02935123, + "epoch": 0.09354896891721277, + "flos": 23149060053120.0, + "grad_norm": 2.0170708549298704, + "language_loss": 0.84274185, + "learning_rate": 3.957857524624068e-06, + "loss": 0.86562991, + "num_input_tokens_seen": 16623685, + "step": 778, + "time_per_iteration": 2.8496057987213135 + }, + { + "auxiliary_loss_clip": 0.01238986, + "auxiliary_loss_mlp": 0.01046842, + "balance_loss_clip": 1.06675839, + "balance_loss_mlp": 1.03176177, + "epoch": 0.09366921180785186, + "flos": 24279563779200.0, + "grad_norm": 1.5515359342474702, + "language_loss": 0.89423633, + "learning_rate": 3.957698308516016e-06, + "loss": 0.91709465, + "num_input_tokens_seen": 16644985, + "step": 779, + "time_per_iteration": 2.940058469772339 + }, + { + "auxiliary_loss_clip": 0.0124934, + "auxiliary_loss_mlp": 0.00889557, + "balance_loss_clip": 1.07437539, + "balance_loss_mlp": 1.00031388, + "epoch": 0.09378945469849095, + "flos": 18730036419840.0, + "grad_norm": 2.2511042943616184, + "language_loss": 0.8271023, + "learning_rate": 3.957538795426688e-06, + "loss": 0.84849131, + "num_input_tokens_seen": 16662410, + "step": 780, + "time_per_iteration": 2.897397994995117 + }, + { + "auxiliary_loss_clip": 0.01238176, + "auxiliary_loss_mlp": 0.010419, + "balance_loss_clip": 1.06760001, + "balance_loss_mlp": 1.02699864, + "epoch": 0.09390969758913004, + "flos": 23218834222080.0, + "grad_norm": 2.222510948508622, + "language_loss": 0.77499831, + "learning_rate": 3.9573789853802804e-06, + "loss": 0.79779911, + "num_input_tokens_seen": 16680885, + "step": 781, + "time_per_iteration": 2.8792364597320557 + }, + { + "auxiliary_loss_clip": 0.01236453, + "auxiliary_loss_mlp": 0.00889581, + "balance_loss_clip": 1.06963277, + "balance_loss_mlp": 1.00022531, + "epoch": 0.09402994047976913, + "flos": 19646728439040.0, + "grad_norm": 2.0787893165370503, + "language_loss": 0.74289638, + "learning_rate": 3.957218878401037e-06, + "loss": 0.7641567, + "num_input_tokens_seen": 16699375, + "step": 782, + "time_per_iteration": 2.860769271850586 + }, + { + "auxiliary_loss_clip": 0.01260009, + "auxiliary_loss_mlp": 0.01048636, + "balance_loss_clip": 1.0732367, + "balance_loss_mlp": 1.03248322, + "epoch": 0.09415018337040823, + "flos": 29420463041280.0, + "grad_norm": 2.4775623853451707, + "language_loss": 0.89477146, + "learning_rate": 3.957058474513246e-06, + "loss": 0.91785789, + "num_input_tokens_seen": 16719230, + "step": 783, + "time_per_iteration": 2.968515634536743 + }, + { + "auxiliary_loss_clip": 0.01244692, + "auxiliary_loss_mlp": 0.01042638, + "balance_loss_clip": 1.07120168, + "balance_loss_mlp": 1.02745056, + "epoch": 0.09427042626104731, + "flos": 24572092141440.0, + "grad_norm": 1.7975023714710852, + "language_loss": 0.78612304, + "learning_rate": 3.956897773741241e-06, + "loss": 0.80899632, + "num_input_tokens_seen": 16738220, + "step": 784, + "time_per_iteration": 2.8701815605163574 + }, + { + "auxiliary_loss_clip": 0.01222962, + "auxiliary_loss_mlp": 0.01052133, + "balance_loss_clip": 1.06409752, + "balance_loss_mlp": 1.0366478, + "epoch": 0.09439066915168641, + "flos": 26359581576960.0, + "grad_norm": 2.042500853698433, + "language_loss": 0.71490282, + "learning_rate": 3.956736776109398e-06, + "loss": 0.73765379, + "num_input_tokens_seen": 16759395, + "step": 785, + "time_per_iteration": 2.80995774269104 + }, + { + "auxiliary_loss_clip": 0.01241031, + "auxiliary_loss_mlp": 0.00890116, + "balance_loss_clip": 1.06608343, + "balance_loss_mlp": 1.00030184, + "epoch": 0.09451091204232549, + "flos": 19427278296960.0, + "grad_norm": 1.9684720360924761, + "language_loss": 0.83609927, + "learning_rate": 3.956575481642143e-06, + "loss": 0.85741079, + "num_input_tokens_seen": 16778285, + "step": 786, + "time_per_iteration": 2.776045560836792 + }, + { + "auxiliary_loss_clip": 0.01208927, + "auxiliary_loss_mlp": 0.01049447, + "balance_loss_clip": 1.058303, + "balance_loss_mlp": 1.03443861, + "epoch": 0.09463115493296459, + "flos": 25368051571200.0, + "grad_norm": 2.570156202393541, + "language_loss": 0.75095099, + "learning_rate": 3.956413890363943e-06, + "loss": 0.77353477, + "num_input_tokens_seen": 16795265, + "step": 787, + "time_per_iteration": 2.896606683731079 + }, + { + "auxiliary_loss_clip": 0.01252454, + "auxiliary_loss_mlp": 0.01054508, + "balance_loss_clip": 1.07323766, + "balance_loss_mlp": 1.03817606, + "epoch": 0.09475139782360369, + "flos": 10123254869760.0, + "grad_norm": 2.0310434762665057, + "language_loss": 0.82168025, + "learning_rate": 3.956252002299312e-06, + "loss": 0.84474981, + "num_input_tokens_seen": 16811165, + "step": 788, + "time_per_iteration": 2.7576093673706055 + }, + { + "auxiliary_loss_clip": 0.01257096, + "auxiliary_loss_mlp": 0.0104459, + "balance_loss_clip": 1.07122242, + "balance_loss_mlp": 1.03005886, + "epoch": 0.09487164071424277, + "flos": 17231088936960.0, + "grad_norm": 2.4277975216926593, + "language_loss": 0.91121858, + "learning_rate": 3.956089817472807e-06, + "loss": 0.93423551, + "num_input_tokens_seen": 16828470, + "step": 789, + "time_per_iteration": 2.6801035404205322 + }, + { + "auxiliary_loss_clip": 0.0123437, + "auxiliary_loss_mlp": 0.0104459, + "balance_loss_clip": 1.07023358, + "balance_loss_mlp": 1.02838898, + "epoch": 0.09499188360488187, + "flos": 30849564528000.0, + "grad_norm": 2.1090012827183724, + "language_loss": 0.85440964, + "learning_rate": 3.955927335909032e-06, + "loss": 0.87719923, + "num_input_tokens_seen": 16851680, + "step": 790, + "time_per_iteration": 2.974940776824951 + }, + { + "auxiliary_loss_clip": 0.01210738, + "auxiliary_loss_mlp": 0.01048271, + "balance_loss_clip": 1.06606388, + "balance_loss_mlp": 1.03353667, + "epoch": 0.09511212649552095, + "flos": 29351694453120.0, + "grad_norm": 2.605986451475929, + "language_loss": 0.75889587, + "learning_rate": 3.955764557632634e-06, + "loss": 0.78148592, + "num_input_tokens_seen": 16871490, + "step": 791, + "time_per_iteration": 3.046452283859253 + }, + { + "auxiliary_loss_clip": 0.01227845, + "auxiliary_loss_mlp": 0.01042236, + "balance_loss_clip": 1.06389916, + "balance_loss_mlp": 1.02746606, + "epoch": 0.09523236938616005, + "flos": 10378687461120.0, + "grad_norm": 2.1620632068404135, + "language_loss": 0.94897735, + "learning_rate": 3.955601482668309e-06, + "loss": 0.97167814, + "num_input_tokens_seen": 16889350, + "step": 792, + "time_per_iteration": 2.7871525287628174 + }, + { + "auxiliary_loss_clip": 0.01207175, + "auxiliary_loss_mlp": 0.01046509, + "balance_loss_clip": 1.05850101, + "balance_loss_mlp": 1.03103614, + "epoch": 0.09535261227679913, + "flos": 19061815368960.0, + "grad_norm": 2.04714525452623, + "language_loss": 0.8897016, + "learning_rate": 3.955438111040794e-06, + "loss": 0.91223848, + "num_input_tokens_seen": 16907625, + "step": 793, + "time_per_iteration": 2.8845102787017822 + }, + { + "auxiliary_loss_clip": 0.01210162, + "auxiliary_loss_mlp": 0.01037792, + "balance_loss_clip": 1.06132817, + "balance_loss_mlp": 1.02250969, + "epoch": 0.09547285516743823, + "flos": 20922993555840.0, + "grad_norm": 4.358235720683282, + "language_loss": 0.8035208, + "learning_rate": 3.955274442774873e-06, + "loss": 0.82600033, + "num_input_tokens_seen": 16926205, + "step": 794, + "time_per_iteration": 2.8401622772216797 + }, + { + "auxiliary_loss_clip": 0.01245379, + "auxiliary_loss_mlp": 0.01044452, + "balance_loss_clip": 1.06737638, + "balance_loss_mlp": 1.0286808, + "epoch": 0.09559309805807732, + "flos": 30154405639680.0, + "grad_norm": 2.215780250133231, + "language_loss": 0.70274144, + "learning_rate": 3.9551104778953725e-06, + "loss": 0.72563976, + "num_input_tokens_seen": 16946500, + "step": 795, + "time_per_iteration": 2.888043165206909 + }, + { + "auxiliary_loss_clip": 0.01222019, + "auxiliary_loss_mlp": 0.01039537, + "balance_loss_clip": 1.06148982, + "balance_loss_mlp": 1.0244925, + "epoch": 0.0957133409487164, + "flos": 21066743784960.0, + "grad_norm": 1.7711420251775745, + "language_loss": 0.85132337, + "learning_rate": 3.954946216427167e-06, + "loss": 0.87393892, + "num_input_tokens_seen": 16966960, + "step": 796, + "time_per_iteration": 2.8322088718414307 + }, + { + "auxiliary_loss_clip": 0.01123234, + "auxiliary_loss_mlp": 0.0101662, + "balance_loss_clip": 1.03811359, + "balance_loss_mlp": 1.00908554, + "epoch": 0.0958335838393555, + "flos": 71297979315840.0, + "grad_norm": 0.8991804794506677, + "language_loss": 0.61630696, + "learning_rate": 3.954781658395176e-06, + "loss": 0.6377055, + "num_input_tokens_seen": 17023215, + "step": 797, + "time_per_iteration": 3.288205146789551 + }, + { + "auxiliary_loss_clip": 0.01247894, + "auxiliary_loss_mlp": 0.01046407, + "balance_loss_clip": 1.07082677, + "balance_loss_mlp": 1.0306952, + "epoch": 0.09595382672999458, + "flos": 21872974504320.0, + "grad_norm": 1.8114835266790945, + "language_loss": 0.92172134, + "learning_rate": 3.95461680382436e-06, + "loss": 0.94466436, + "num_input_tokens_seen": 17042140, + "step": 798, + "time_per_iteration": 2.850768566131592 + }, + { + "auxiliary_loss_clip": 0.01250722, + "auxiliary_loss_mlp": 0.01051919, + "balance_loss_clip": 1.07169366, + "balance_loss_mlp": 1.03558767, + "epoch": 0.09607406962063368, + "flos": 18695562341760.0, + "grad_norm": 6.628687621089067, + "language_loss": 0.86371124, + "learning_rate": 3.9544516527397295e-06, + "loss": 0.88673764, + "num_input_tokens_seen": 17058490, + "step": 799, + "time_per_iteration": 2.8225040435791016 + }, + { + "auxiliary_loss_clip": 0.01221801, + "auxiliary_loss_mlp": 0.01042289, + "balance_loss_clip": 1.0643357, + "balance_loss_mlp": 1.02733994, + "epoch": 0.09619431251127276, + "flos": 22568456615040.0, + "grad_norm": 1.6778880471713336, + "language_loss": 0.80611992, + "learning_rate": 3.954286205166338e-06, + "loss": 0.82876086, + "num_input_tokens_seen": 17079655, + "step": 800, + "time_per_iteration": 4.748415231704712 + }, + { + "auxiliary_loss_clip": 0.01252223, + "auxiliary_loss_mlp": 0.01052083, + "balance_loss_clip": 1.07427955, + "balance_loss_mlp": 1.03670454, + "epoch": 0.09631455540191186, + "flos": 14246230608000.0, + "grad_norm": 2.145457457104979, + "language_loss": 0.83872926, + "learning_rate": 3.954120461129282e-06, + "loss": 0.8617723, + "num_input_tokens_seen": 17097065, + "step": 801, + "time_per_iteration": 4.634357213973999 + }, + { + "auxiliary_loss_clip": 0.01263024, + "auxiliary_loss_mlp": 0.01056156, + "balance_loss_clip": 1.07635736, + "balance_loss_mlp": 1.04083765, + "epoch": 0.09643479829255096, + "flos": 20740387789440.0, + "grad_norm": 2.039040493000454, + "language_loss": 0.83778834, + "learning_rate": 3.953954420653706e-06, + "loss": 0.86098015, + "num_input_tokens_seen": 17114090, + "step": 802, + "time_per_iteration": 2.709735155105591 + }, + { + "auxiliary_loss_clip": 0.0124845, + "auxiliary_loss_mlp": 0.01051206, + "balance_loss_clip": 1.07263398, + "balance_loss_mlp": 1.03644776, + "epoch": 0.09655504118319004, + "flos": 24420476833920.0, + "grad_norm": 2.521466392761962, + "language_loss": 0.88155448, + "learning_rate": 3.953788083764798e-06, + "loss": 0.90455109, + "num_input_tokens_seen": 17133325, + "step": 803, + "time_per_iteration": 2.7358953952789307 + }, + { + "auxiliary_loss_clip": 0.01209738, + "auxiliary_loss_mlp": 0.01051238, + "balance_loss_clip": 1.06416106, + "balance_loss_mlp": 1.03601551, + "epoch": 0.09667528407382914, + "flos": 18441961344000.0, + "grad_norm": 2.0853252124281347, + "language_loss": 0.92729789, + "learning_rate": 3.953621450487792e-06, + "loss": 0.94990766, + "num_input_tokens_seen": 17151945, + "step": 804, + "time_per_iteration": 2.8673834800720215 + }, + { + "auxiliary_loss_clip": 0.01145838, + "auxiliary_loss_mlp": 0.0101062, + "balance_loss_clip": 1.03533411, + "balance_loss_mlp": 1.00332451, + "epoch": 0.09679552696446822, + "flos": 70816455544320.0, + "grad_norm": 0.8472782342597798, + "language_loss": 0.61256462, + "learning_rate": 3.953454520847964e-06, + "loss": 0.63412923, + "num_input_tokens_seen": 17216790, + "step": 805, + "time_per_iteration": 3.509460210800171 + }, + { + "auxiliary_loss_clip": 0.01223265, + "auxiliary_loss_mlp": 0.01042098, + "balance_loss_clip": 1.06174088, + "balance_loss_mlp": 1.02601635, + "epoch": 0.09691576985510732, + "flos": 21945514020480.0, + "grad_norm": 3.751373825792708, + "language_loss": 0.7384904, + "learning_rate": 3.9532872948706395e-06, + "loss": 0.76114404, + "num_input_tokens_seen": 17236285, + "step": 806, + "time_per_iteration": 2.8529891967773438 + }, + { + "auxiliary_loss_clip": 0.01239432, + "auxiliary_loss_mlp": 0.01049133, + "balance_loss_clip": 1.06696773, + "balance_loss_mlp": 1.03376663, + "epoch": 0.09703601274574641, + "flos": 17965211103360.0, + "grad_norm": 2.3996661307136873, + "language_loss": 0.82553506, + "learning_rate": 3.9531197725811845e-06, + "loss": 0.84842074, + "num_input_tokens_seen": 17251670, + "step": 807, + "time_per_iteration": 2.778944253921509 + }, + { + "auxiliary_loss_clip": 0.01259881, + "auxiliary_loss_mlp": 0.01050698, + "balance_loss_clip": 1.07666039, + "balance_loss_mlp": 1.03597546, + "epoch": 0.0971562556363855, + "flos": 22162162901760.0, + "grad_norm": 1.8152400019818205, + "language_loss": 0.87961447, + "learning_rate": 3.952951954005013e-06, + "loss": 0.90272027, + "num_input_tokens_seen": 17271355, + "step": 808, + "time_per_iteration": 2.7360634803771973 + }, + { + "auxiliary_loss_clip": 0.01227208, + "auxiliary_loss_mlp": 0.01040145, + "balance_loss_clip": 1.06255627, + "balance_loss_mlp": 1.02524436, + "epoch": 0.0972764985270246, + "flos": 25848716394240.0, + "grad_norm": 1.8652244821980872, + "language_loss": 0.84751797, + "learning_rate": 3.952783839167584e-06, + "loss": 0.87019145, + "num_input_tokens_seen": 17291400, + "step": 809, + "time_per_iteration": 2.8967065811157227 + }, + { + "auxiliary_loss_clip": 0.01243652, + "auxiliary_loss_mlp": 0.01045685, + "balance_loss_clip": 1.0687288, + "balance_loss_mlp": 1.03078425, + "epoch": 0.09739674141766368, + "flos": 20339373375360.0, + "grad_norm": 2.8477547270175196, + "language_loss": 0.74699992, + "learning_rate": 3.952615428094398e-06, + "loss": 0.76989329, + "num_input_tokens_seen": 17310920, + "step": 810, + "time_per_iteration": 2.749201536178589 + }, + { + "auxiliary_loss_clip": 0.01202974, + "auxiliary_loss_mlp": 0.01047743, + "balance_loss_clip": 1.05928445, + "balance_loss_mlp": 1.03211451, + "epoch": 0.09751698430830277, + "flos": 15743059188480.0, + "grad_norm": 1.6313008844980998, + "language_loss": 0.7335012, + "learning_rate": 3.952446720811004e-06, + "loss": 0.75600839, + "num_input_tokens_seen": 17329245, + "step": 811, + "time_per_iteration": 2.920262575149536 + }, + { + "auxiliary_loss_clip": 0.0111604, + "auxiliary_loss_mlp": 0.01010705, + "balance_loss_clip": 1.03220081, + "balance_loss_mlp": 1.00321865, + "epoch": 0.09763722719894186, + "flos": 63716806800000.0, + "grad_norm": 0.8471209934675574, + "language_loss": 0.63702524, + "learning_rate": 3.952277717342995e-06, + "loss": 0.65829265, + "num_input_tokens_seen": 17395680, + "step": 812, + "time_per_iteration": 3.5048067569732666 + }, + { + "auxiliary_loss_clip": 0.01238189, + "auxiliary_loss_mlp": 0.01044212, + "balance_loss_clip": 1.06782532, + "balance_loss_mlp": 1.02925158, + "epoch": 0.09775747008958095, + "flos": 22090916275200.0, + "grad_norm": 2.9512638787634615, + "language_loss": 0.85374576, + "learning_rate": 3.952108417716009e-06, + "loss": 0.87656975, + "num_input_tokens_seen": 17415135, + "step": 813, + "time_per_iteration": 2.820711612701416 + }, + { + "auxiliary_loss_clip": 0.01248923, + "auxiliary_loss_mlp": 0.01063886, + "balance_loss_clip": 1.07281578, + "balance_loss_mlp": 1.04850829, + "epoch": 0.09787771298022005, + "flos": 21286050272640.0, + "grad_norm": 1.8120881096697992, + "language_loss": 0.85039806, + "learning_rate": 3.951938821955727e-06, + "loss": 0.8735261, + "num_input_tokens_seen": 17434535, + "step": 814, + "time_per_iteration": 2.832993268966675 + }, + { + "auxiliary_loss_clip": 0.01234656, + "auxiliary_loss_mlp": 0.01049329, + "balance_loss_clip": 1.06967139, + "balance_loss_mlp": 1.03386748, + "epoch": 0.09799795587085913, + "flos": 22054574689920.0, + "grad_norm": 1.6633736814610727, + "language_loss": 0.76633453, + "learning_rate": 3.9517689300878786e-06, + "loss": 0.78917438, + "num_input_tokens_seen": 17454270, + "step": 815, + "time_per_iteration": 2.832449197769165 + }, + { + "auxiliary_loss_clip": 0.01254491, + "auxiliary_loss_mlp": 0.01047486, + "balance_loss_clip": 1.0701189, + "balance_loss_mlp": 1.03232265, + "epoch": 0.09811819876149823, + "flos": 22163743100160.0, + "grad_norm": 1.736512451134163, + "language_loss": 0.78517234, + "learning_rate": 3.951598742138236e-06, + "loss": 0.80819213, + "num_input_tokens_seen": 17472995, + "step": 816, + "time_per_iteration": 2.7905242443084717 + }, + { + "auxiliary_loss_clip": 0.01237016, + "auxiliary_loss_mlp": 0.01052984, + "balance_loss_clip": 1.0652895, + "balance_loss_mlp": 1.03835738, + "epoch": 0.09823844165213731, + "flos": 22231111057920.0, + "grad_norm": 3.28878013417125, + "language_loss": 0.79912257, + "learning_rate": 3.951428258132615e-06, + "loss": 0.82202262, + "num_input_tokens_seen": 17491115, + "step": 817, + "time_per_iteration": 2.8396666049957275 + }, + { + "auxiliary_loss_clip": 0.01237964, + "auxiliary_loss_mlp": 0.01044194, + "balance_loss_clip": 1.07019472, + "balance_loss_mlp": 1.02924538, + "epoch": 0.09835868454277641, + "flos": 22487728798080.0, + "grad_norm": 1.9707631530267444, + "language_loss": 0.84389734, + "learning_rate": 3.951257478096879e-06, + "loss": 0.86671883, + "num_input_tokens_seen": 17509480, + "step": 818, + "time_per_iteration": 2.7494661808013916 + }, + { + "auxiliary_loss_clip": 0.01239458, + "auxiliary_loss_mlp": 0.0089031, + "balance_loss_clip": 1.06981933, + "balance_loss_mlp": 1.00036287, + "epoch": 0.0984789274334155, + "flos": 16362554077440.0, + "grad_norm": 2.6919693742092536, + "language_loss": 0.68899441, + "learning_rate": 3.951086402056936e-06, + "loss": 0.7102921, + "num_input_tokens_seen": 17524080, + "step": 819, + "time_per_iteration": 2.8136165142059326 + }, + { + "auxiliary_loss_clip": 0.01189281, + "auxiliary_loss_mlp": 0.00889719, + "balance_loss_clip": 1.06629372, + "balance_loss_mlp": 1.00042677, + "epoch": 0.09859917032405459, + "flos": 24243545416320.0, + "grad_norm": 1.896475510375433, + "language_loss": 0.83580226, + "learning_rate": 3.950915030038735e-06, + "loss": 0.8565923, + "num_input_tokens_seen": 17543875, + "step": 820, + "time_per_iteration": 3.057159900665283 + }, + { + "auxiliary_loss_clip": 0.01240461, + "auxiliary_loss_mlp": 0.01049095, + "balance_loss_clip": 1.07121122, + "balance_loss_mlp": 1.03518379, + "epoch": 0.09871941321469369, + "flos": 17420195064960.0, + "grad_norm": 2.1767666713623157, + "language_loss": 0.83491695, + "learning_rate": 3.9507433620682765e-06, + "loss": 0.85781246, + "num_input_tokens_seen": 17560810, + "step": 821, + "time_per_iteration": 3.090733051300049 + }, + { + "auxiliary_loss_clip": 0.01222449, + "auxiliary_loss_mlp": 0.0104541, + "balance_loss_clip": 1.06627405, + "balance_loss_mlp": 1.03055668, + "epoch": 0.09883965610533277, + "flos": 28477341590400.0, + "grad_norm": 2.0337192660479078, + "language_loss": 0.88391685, + "learning_rate": 3.9505713981716e-06, + "loss": 0.90659547, + "num_input_tokens_seen": 17583640, + "step": 822, + "time_per_iteration": 2.8560776710510254 + }, + { + "auxiliary_loss_clip": 0.01228598, + "auxiliary_loss_mlp": 0.01035921, + "balance_loss_clip": 1.06817126, + "balance_loss_mlp": 1.02218843, + "epoch": 0.09895989899597187, + "flos": 23693932437120.0, + "grad_norm": 2.3160468558743044, + "language_loss": 0.81173491, + "learning_rate": 3.950399138374795e-06, + "loss": 0.83438009, + "num_input_tokens_seen": 17602720, + "step": 823, + "time_per_iteration": 2.7554945945739746 + }, + { + "auxiliary_loss_clip": 0.01241468, + "auxiliary_loss_mlp": 0.01043483, + "balance_loss_clip": 1.06726205, + "balance_loss_mlp": 1.02890372, + "epoch": 0.09908014188661095, + "flos": 24679608526080.0, + "grad_norm": 2.080166071123522, + "language_loss": 0.74481332, + "learning_rate": 3.95022658270399e-06, + "loss": 0.76766288, + "num_input_tokens_seen": 17623085, + "step": 824, + "time_per_iteration": 2.764794111251831 + }, + { + "auxiliary_loss_clip": 0.01226159, + "auxiliary_loss_mlp": 0.01046394, + "balance_loss_clip": 1.06616211, + "balance_loss_mlp": 1.03180325, + "epoch": 0.09920038477725004, + "flos": 14064307200000.0, + "grad_norm": 1.8449416924871145, + "language_loss": 0.78312922, + "learning_rate": 3.9500537311853635e-06, + "loss": 0.8058548, + "num_input_tokens_seen": 17641040, + "step": 825, + "time_per_iteration": 2.7435660362243652 + }, + { + "auxiliary_loss_clip": 0.01239773, + "auxiliary_loss_mlp": 0.01048344, + "balance_loss_clip": 1.06491089, + "balance_loss_mlp": 1.03300142, + "epoch": 0.09932062766788914, + "flos": 13407070095360.0, + "grad_norm": 2.2826553421439564, + "language_loss": 0.83486998, + "learning_rate": 3.949880583845136e-06, + "loss": 0.85775113, + "num_input_tokens_seen": 17659115, + "step": 826, + "time_per_iteration": 4.606433153152466 + }, + { + "auxiliary_loss_clip": 0.01229995, + "auxiliary_loss_mlp": 0.01048207, + "balance_loss_clip": 1.06741273, + "balance_loss_mlp": 1.03236389, + "epoch": 0.09944087055852822, + "flos": 19500751566720.0, + "grad_norm": 1.9723481835718009, + "language_loss": 0.81469488, + "learning_rate": 3.949707140709575e-06, + "loss": 0.83747697, + "num_input_tokens_seen": 17678845, + "step": 827, + "time_per_iteration": 3.7536494731903076 + }, + { + "auxiliary_loss_clip": 0.01242558, + "auxiliary_loss_mlp": 0.01038787, + "balance_loss_clip": 1.06700003, + "balance_loss_mlp": 1.02400482, + "epoch": 0.09956111344916732, + "flos": 17749100926080.0, + "grad_norm": 2.5597071255220962, + "language_loss": 0.83516639, + "learning_rate": 3.949533401804991e-06, + "loss": 0.85797983, + "num_input_tokens_seen": 17695750, + "step": 828, + "time_per_iteration": 2.711315870285034 + }, + { + "auxiliary_loss_clip": 0.012434, + "auxiliary_loss_mlp": 0.00889795, + "balance_loss_clip": 1.07091904, + "balance_loss_mlp": 1.00033736, + "epoch": 0.0996813563398064, + "flos": 17967581400960.0, + "grad_norm": 2.3999986516875533, + "language_loss": 0.90592778, + "learning_rate": 3.949359367157739e-06, + "loss": 0.92725968, + "num_input_tokens_seen": 17714445, + "step": 829, + "time_per_iteration": 2.722165107727051 + }, + { + "auxiliary_loss_clip": 0.01249867, + "auxiliary_loss_mlp": 0.01042999, + "balance_loss_clip": 1.07090592, + "balance_loss_mlp": 1.02744186, + "epoch": 0.0998015992304455, + "flos": 17457039440640.0, + "grad_norm": 3.3082018789030254, + "language_loss": 0.75496578, + "learning_rate": 3.949185036794222e-06, + "loss": 0.7778945, + "num_input_tokens_seen": 17732455, + "step": 830, + "time_per_iteration": 2.719695568084717 + }, + { + "auxiliary_loss_clip": 0.0125398, + "auxiliary_loss_mlp": 0.01041273, + "balance_loss_clip": 1.07267129, + "balance_loss_mlp": 1.02698004, + "epoch": 0.0999218421210846, + "flos": 25888757080320.0, + "grad_norm": 1.8694632051131, + "language_loss": 0.78654927, + "learning_rate": 3.949010410740884e-06, + "loss": 0.80950183, + "num_input_tokens_seen": 17755280, + "step": 831, + "time_per_iteration": 2.801335334777832 + }, + { + "auxiliary_loss_clip": 0.01222205, + "auxiliary_loss_mlp": 0.00890088, + "balance_loss_clip": 1.0657351, + "balance_loss_mlp": 1.0003376, + "epoch": 0.10004208501172368, + "flos": 21215916967680.0, + "grad_norm": 2.130098212704189, + "language_loss": 0.86447388, + "learning_rate": 3.948835489024216e-06, + "loss": 0.88559675, + "num_input_tokens_seen": 17775015, + "step": 832, + "time_per_iteration": 2.861726999282837 + }, + { + "auxiliary_loss_clip": 0.01242865, + "auxiliary_loss_mlp": 0.01038539, + "balance_loss_clip": 1.06706977, + "balance_loss_mlp": 1.02475226, + "epoch": 0.10016232790236278, + "flos": 17348409734400.0, + "grad_norm": 2.4857605486715317, + "language_loss": 0.90471542, + "learning_rate": 3.948660271670755e-06, + "loss": 0.92752945, + "num_input_tokens_seen": 17792165, + "step": 833, + "time_per_iteration": 2.739617347717285 + }, + { + "auxiliary_loss_clip": 0.01227603, + "auxiliary_loss_mlp": 0.01044254, + "balance_loss_clip": 1.06648719, + "balance_loss_mlp": 1.02980018, + "epoch": 0.10028257079300186, + "flos": 25666541591040.0, + "grad_norm": 2.2677576785498417, + "language_loss": 0.84484076, + "learning_rate": 3.948484758707079e-06, + "loss": 0.86755931, + "num_input_tokens_seen": 17811765, + "step": 834, + "time_per_iteration": 2.847352981567383 + }, + { + "auxiliary_loss_clip": 0.0120465, + "auxiliary_loss_mlp": 0.0104099, + "balance_loss_clip": 1.05886197, + "balance_loss_mlp": 1.02543294, + "epoch": 0.10040281368364096, + "flos": 25156035544320.0, + "grad_norm": 2.2243236034400233, + "language_loss": 0.83591628, + "learning_rate": 3.948308950159815e-06, + "loss": 0.85837275, + "num_input_tokens_seen": 17830445, + "step": 835, + "time_per_iteration": 2.8378920555114746 + }, + { + "auxiliary_loss_clip": 0.01209068, + "auxiliary_loss_mlp": 0.01040874, + "balance_loss_clip": 1.05886209, + "balance_loss_mlp": 1.02536511, + "epoch": 0.10052305657428004, + "flos": 17603303621760.0, + "grad_norm": 2.5184912671292317, + "language_loss": 0.76077014, + "learning_rate": 3.9481328460556326e-06, + "loss": 0.78326958, + "num_input_tokens_seen": 17847665, + "step": 836, + "time_per_iteration": 2.9392306804656982 + }, + { + "auxiliary_loss_clip": 0.01224131, + "auxiliary_loss_mlp": 0.01043068, + "balance_loss_clip": 1.0650351, + "balance_loss_mlp": 1.02821469, + "epoch": 0.10064329946491914, + "flos": 18660154510080.0, + "grad_norm": 2.702765070781847, + "language_loss": 0.89537919, + "learning_rate": 3.9479564464212455e-06, + "loss": 0.91805112, + "num_input_tokens_seen": 17866825, + "step": 837, + "time_per_iteration": 2.7974631786346436 + }, + { + "auxiliary_loss_clip": 0.01257962, + "auxiliary_loss_mlp": 0.01035209, + "balance_loss_clip": 1.07078803, + "balance_loss_mlp": 1.02024877, + "epoch": 0.10076354235555823, + "flos": 17199056983680.0, + "grad_norm": 2.3663137150069673, + "language_loss": 0.76073968, + "learning_rate": 3.947779751283414e-06, + "loss": 0.78367138, + "num_input_tokens_seen": 17883995, + "step": 838, + "time_per_iteration": 2.734694242477417 + }, + { + "auxiliary_loss_clip": 0.01245277, + "auxiliary_loss_mlp": 0.00889694, + "balance_loss_clip": 1.07574272, + "balance_loss_mlp": 1.00026631, + "epoch": 0.10088378524619732, + "flos": 22962252395520.0, + "grad_norm": 1.9101743866962997, + "language_loss": 0.75863898, + "learning_rate": 3.947602760668944e-06, + "loss": 0.77998865, + "num_input_tokens_seen": 17903785, + "step": 839, + "time_per_iteration": 2.7627360820770264 + }, + { + "auxiliary_loss_clip": 0.01242696, + "auxiliary_loss_mlp": 0.01054249, + "balance_loss_clip": 1.07296145, + "balance_loss_mlp": 1.0395329, + "epoch": 0.10100402813683641, + "flos": 37885828746240.0, + "grad_norm": 1.845136649074915, + "language_loss": 0.71561563, + "learning_rate": 3.947425474604684e-06, + "loss": 0.73858511, + "num_input_tokens_seen": 17927720, + "step": 840, + "time_per_iteration": 2.9025042057037354 + }, + { + "auxiliary_loss_clip": 0.01227459, + "auxiliary_loss_mlp": 0.01036908, + "balance_loss_clip": 1.06602311, + "balance_loss_mlp": 1.02241194, + "epoch": 0.1011242710274755, + "flos": 21543458112000.0, + "grad_norm": 3.7995009235004407, + "language_loss": 0.92312402, + "learning_rate": 3.947247893117528e-06, + "loss": 0.9457677, + "num_input_tokens_seen": 17946225, + "step": 841, + "time_per_iteration": 2.7601523399353027 + }, + { + "auxiliary_loss_clip": 0.01237129, + "auxiliary_loss_mlp": 0.01045265, + "balance_loss_clip": 1.06521928, + "balance_loss_mlp": 1.03048301, + "epoch": 0.10124451391811459, + "flos": 13621456419840.0, + "grad_norm": 3.0910479542246203, + "language_loss": 0.69569111, + "learning_rate": 3.947070016234413e-06, + "loss": 0.71851504, + "num_input_tokens_seen": 17962015, + "step": 842, + "time_per_iteration": 2.706696033477783 + }, + { + "auxiliary_loss_clip": 0.01242043, + "auxiliary_loss_mlp": 0.01058275, + "balance_loss_clip": 1.07010269, + "balance_loss_mlp": 1.04289675, + "epoch": 0.10136475680875369, + "flos": 16649228522880.0, + "grad_norm": 2.294628215510907, + "language_loss": 0.75350046, + "learning_rate": 3.946891843982326e-06, + "loss": 0.77650368, + "num_input_tokens_seen": 17979680, + "step": 843, + "time_per_iteration": 2.8483712673187256 + }, + { + "auxiliary_loss_clip": 0.01244783, + "auxiliary_loss_mlp": 0.01042609, + "balance_loss_clip": 1.07247138, + "balance_loss_mlp": 1.02782702, + "epoch": 0.10148499969939277, + "flos": 19461034103040.0, + "grad_norm": 2.1600389982769874, + "language_loss": 0.74590707, + "learning_rate": 3.9467133763882935e-06, + "loss": 0.76878095, + "num_input_tokens_seen": 17998145, + "step": 844, + "time_per_iteration": 2.749509811401367 + }, + { + "auxiliary_loss_clip": 0.01234261, + "auxiliary_loss_mlp": 0.01040946, + "balance_loss_clip": 1.06752205, + "balance_loss_mlp": 1.02581215, + "epoch": 0.10160524259003187, + "flos": 21104988791040.0, + "grad_norm": 2.153139098419383, + "language_loss": 0.86315286, + "learning_rate": 3.9465346134793905e-06, + "loss": 0.88590491, + "num_input_tokens_seen": 18017955, + "step": 845, + "time_per_iteration": 2.8258488178253174 + }, + { + "auxiliary_loss_clip": 0.01217566, + "auxiliary_loss_mlp": 0.0104553, + "balance_loss_clip": 1.06661868, + "balance_loss_mlp": 1.03102183, + "epoch": 0.10172548548067095, + "flos": 17712687513600.0, + "grad_norm": 2.051943623230818, + "language_loss": 0.79429209, + "learning_rate": 3.9463555552827335e-06, + "loss": 0.81692302, + "num_input_tokens_seen": 18035125, + "step": 846, + "time_per_iteration": 2.803755044937134 + }, + { + "auxiliary_loss_clip": 0.01227804, + "auxiliary_loss_mlp": 0.01048543, + "balance_loss_clip": 1.06290317, + "balance_loss_mlp": 1.03433347, + "epoch": 0.10184572837131005, + "flos": 21104845136640.0, + "grad_norm": 2.5299682380449275, + "language_loss": 0.86517525, + "learning_rate": 3.946176201825487e-06, + "loss": 0.88793874, + "num_input_tokens_seen": 18053160, + "step": 847, + "time_per_iteration": 2.787379264831543 + }, + { + "auxiliary_loss_clip": 0.01233468, + "auxiliary_loss_mlp": 0.01051141, + "balance_loss_clip": 1.06997418, + "balance_loss_mlp": 1.03651369, + "epoch": 0.10196597126194913, + "flos": 26067591918720.0, + "grad_norm": 1.9389095999525081, + "language_loss": 0.83313775, + "learning_rate": 3.9459965531348575e-06, + "loss": 0.85598385, + "num_input_tokens_seen": 18072815, + "step": 848, + "time_per_iteration": 2.8097667694091797 + }, + { + "auxiliary_loss_clip": 0.01230098, + "auxiliary_loss_mlp": 0.00890544, + "balance_loss_clip": 1.06781316, + "balance_loss_mlp": 1.00040495, + "epoch": 0.10208621415258823, + "flos": 29314634595840.0, + "grad_norm": 2.094313393553291, + "language_loss": 0.85428691, + "learning_rate": 3.945816609238098e-06, + "loss": 0.87549329, + "num_input_tokens_seen": 18092225, + "step": 849, + "time_per_iteration": 2.86724853515625 + }, + { + "auxiliary_loss_clip": 0.01204708, + "auxiliary_loss_mlp": 0.01055036, + "balance_loss_clip": 1.06377125, + "balance_loss_mlp": 1.0400517, + "epoch": 0.10220645704322733, + "flos": 23805794367360.0, + "grad_norm": 2.009878589842106, + "language_loss": 0.85391223, + "learning_rate": 3.945636370162507e-06, + "loss": 0.87650967, + "num_input_tokens_seen": 18112335, + "step": 850, + "time_per_iteration": 2.8720133304595947 + }, + { + "auxiliary_loss_clip": 0.0124071, + "auxiliary_loss_mlp": 0.01042953, + "balance_loss_clip": 1.0690366, + "balance_loss_mlp": 1.02901793, + "epoch": 0.10232669993386641, + "flos": 23218546913280.0, + "grad_norm": 1.8907182125145756, + "language_loss": 0.79480428, + "learning_rate": 3.945455835935425e-06, + "loss": 0.8176409, + "num_input_tokens_seen": 18131520, + "step": 851, + "time_per_iteration": 3.598325729370117 + }, + { + "auxiliary_loss_clip": 0.01226607, + "auxiliary_loss_mlp": 0.01043937, + "balance_loss_clip": 1.06329751, + "balance_loss_mlp": 1.02903605, + "epoch": 0.1024469428245055, + "flos": 22922929981440.0, + "grad_norm": 2.3988926629409626, + "language_loss": 0.75072831, + "learning_rate": 3.94527500658424e-06, + "loss": 0.77343374, + "num_input_tokens_seen": 18149185, + "step": 852, + "time_per_iteration": 3.7042813301086426 + }, + { + "auxiliary_loss_clip": 0.01211353, + "auxiliary_loss_mlp": 0.01042038, + "balance_loss_clip": 1.06491506, + "balance_loss_mlp": 1.02708948, + "epoch": 0.10256718571514459, + "flos": 31359495957120.0, + "grad_norm": 1.9577810455185416, + "language_loss": 0.81281447, + "learning_rate": 3.945093882136382e-06, + "loss": 0.83534837, + "num_input_tokens_seen": 18172960, + "step": 853, + "time_per_iteration": 4.903201580047607 + }, + { + "auxiliary_loss_clip": 0.01231646, + "auxiliary_loss_mlp": 0.00889508, + "balance_loss_clip": 1.070081, + "balance_loss_mlp": 1.0004046, + "epoch": 0.10268742860578368, + "flos": 23474877344640.0, + "grad_norm": 1.8056740752817593, + "language_loss": 0.84671056, + "learning_rate": 3.944912462619329e-06, + "loss": 0.86792207, + "num_input_tokens_seen": 18191925, + "step": 854, + "time_per_iteration": 2.792247772216797 + }, + { + "auxiliary_loss_clip": 0.01234821, + "auxiliary_loss_mlp": 0.01045593, + "balance_loss_clip": 1.06743324, + "balance_loss_mlp": 1.02985704, + "epoch": 0.10280767149642277, + "flos": 25520313323520.0, + "grad_norm": 3.684509116906937, + "language_loss": 0.8063547, + "learning_rate": 3.9447307480606025e-06, + "loss": 0.8291589, + "num_input_tokens_seen": 18212010, + "step": 855, + "time_per_iteration": 2.797795057296753 + }, + { + "auxiliary_loss_clip": 0.01221183, + "auxiliary_loss_mlp": 0.01048316, + "balance_loss_clip": 1.06618118, + "balance_loss_mlp": 1.03244925, + "epoch": 0.10292791438706186, + "flos": 17347691462400.0, + "grad_norm": 1.9521000918135079, + "language_loss": 0.89833891, + "learning_rate": 3.944548738487767e-06, + "loss": 0.92103386, + "num_input_tokens_seen": 18229525, + "step": 856, + "time_per_iteration": 2.8400068283081055 + }, + { + "auxiliary_loss_clip": 0.01261821, + "auxiliary_loss_mlp": 0.01040886, + "balance_loss_clip": 1.07709324, + "balance_loss_mlp": 1.02641368, + "epoch": 0.10304815727770096, + "flos": 27052693390080.0, + "grad_norm": 2.859931465552654, + "language_loss": 0.90655863, + "learning_rate": 3.944366433928434e-06, + "loss": 0.9295857, + "num_input_tokens_seen": 18249505, + "step": 857, + "time_per_iteration": 2.69419527053833 + }, + { + "auxiliary_loss_clip": 0.01220045, + "auxiliary_loss_mlp": 0.01041857, + "balance_loss_clip": 1.0627172, + "balance_loss_mlp": 1.02783775, + "epoch": 0.10316840016834004, + "flos": 22782591544320.0, + "grad_norm": 1.5503329902235137, + "language_loss": 0.83457488, + "learning_rate": 3.9441838344102594e-06, + "loss": 0.85719395, + "num_input_tokens_seen": 18269230, + "step": 858, + "time_per_iteration": 2.8447206020355225 + }, + { + "auxiliary_loss_clip": 0.01233301, + "auxiliary_loss_mlp": 0.01048555, + "balance_loss_clip": 1.06829476, + "balance_loss_mlp": 1.0327841, + "epoch": 0.10328864305897914, + "flos": 20704584908160.0, + "grad_norm": 2.435862700590273, + "language_loss": 0.67079437, + "learning_rate": 3.944000939960943e-06, + "loss": 0.69361293, + "num_input_tokens_seen": 18287955, + "step": 859, + "time_per_iteration": 2.771266222000122 + }, + { + "auxiliary_loss_clip": 0.0124108, + "auxiliary_loss_mlp": 0.01048636, + "balance_loss_clip": 1.06692636, + "balance_loss_mlp": 1.0342828, + "epoch": 0.10340888594961822, + "flos": 28478814048000.0, + "grad_norm": 1.6293416277322166, + "language_loss": 0.80017066, + "learning_rate": 3.943817750608229e-06, + "loss": 0.82306772, + "num_input_tokens_seen": 18310505, + "step": 860, + "time_per_iteration": 2.8335938453674316 + }, + { + "auxiliary_loss_clip": 0.01246085, + "auxiliary_loss_mlp": 0.01041332, + "balance_loss_clip": 1.07231367, + "balance_loss_mlp": 1.02687216, + "epoch": 0.10352912884025732, + "flos": 13370333460480.0, + "grad_norm": 2.2209323957949074, + "language_loss": 0.81791121, + "learning_rate": 3.943634266379908e-06, + "loss": 0.84078538, + "num_input_tokens_seen": 18327400, + "step": 861, + "time_per_iteration": 2.7960963249206543 + }, + { + "auxiliary_loss_clip": 0.01245882, + "auxiliary_loss_mlp": 0.01049327, + "balance_loss_clip": 1.06842017, + "balance_loss_mlp": 1.03409219, + "epoch": 0.10364937173089642, + "flos": 25558558329600.0, + "grad_norm": 2.519613094311446, + "language_loss": 0.84802091, + "learning_rate": 3.943450487303815e-06, + "loss": 0.87097299, + "num_input_tokens_seen": 18347895, + "step": 862, + "time_per_iteration": 2.7794623374938965 + }, + { + "auxiliary_loss_clip": 0.01239675, + "auxiliary_loss_mlp": 0.01052535, + "balance_loss_clip": 1.06917143, + "balance_loss_mlp": 1.0371809, + "epoch": 0.1037696146215355, + "flos": 21215486004480.0, + "grad_norm": 1.8156060612099831, + "language_loss": 0.85178572, + "learning_rate": 3.943266413407827e-06, + "loss": 0.87470782, + "num_input_tokens_seen": 18367170, + "step": 863, + "time_per_iteration": 2.7799062728881836 + }, + { + "auxiliary_loss_clip": 0.012451, + "auxiliary_loss_mlp": 0.01053902, + "balance_loss_clip": 1.07053399, + "balance_loss_mlp": 1.03962648, + "epoch": 0.1038898575121746, + "flos": 25807382818560.0, + "grad_norm": 5.652374398669534, + "language_loss": 0.84955001, + "learning_rate": 3.94308204471987e-06, + "loss": 0.87254, + "num_input_tokens_seen": 18386185, + "step": 864, + "time_per_iteration": 2.8068840503692627 + }, + { + "auxiliary_loss_clip": 0.01215324, + "auxiliary_loss_mlp": 0.01036774, + "balance_loss_clip": 1.06264424, + "balance_loss_mlp": 1.02223015, + "epoch": 0.10401010040281368, + "flos": 19062425900160.0, + "grad_norm": 2.3214765651753417, + "language_loss": 0.74581552, + "learning_rate": 3.942897381267912e-06, + "loss": 0.76833642, + "num_input_tokens_seen": 18402550, + "step": 865, + "time_per_iteration": 2.8602845668792725 + }, + { + "auxiliary_loss_clip": 0.0124305, + "auxiliary_loss_mlp": 0.01056426, + "balance_loss_clip": 1.06757855, + "balance_loss_mlp": 1.04079723, + "epoch": 0.10413034329345278, + "flos": 16355119962240.0, + "grad_norm": 3.089181712521007, + "language_loss": 0.66126013, + "learning_rate": 3.942712423079965e-06, + "loss": 0.68425488, + "num_input_tokens_seen": 18418940, + "step": 866, + "time_per_iteration": 2.880354642868042 + }, + { + "auxiliary_loss_clip": 0.01200912, + "auxiliary_loss_mlp": 0.01045912, + "balance_loss_clip": 1.05558562, + "balance_loss_mlp": 1.0317142, + "epoch": 0.10425058618409186, + "flos": 17236511890560.0, + "grad_norm": 2.5964401231166394, + "language_loss": 0.89786196, + "learning_rate": 3.942527170184088e-06, + "loss": 0.92033017, + "num_input_tokens_seen": 18435560, + "step": 867, + "time_per_iteration": 2.8693485260009766 + }, + { + "auxiliary_loss_clip": 0.0125473, + "auxiliary_loss_mlp": 0.01047595, + "balance_loss_clip": 1.07216728, + "balance_loss_mlp": 1.03206205, + "epoch": 0.10437082907473096, + "flos": 17967365919360.0, + "grad_norm": 2.3806742313894222, + "language_loss": 0.77495122, + "learning_rate": 3.942341622608385e-06, + "loss": 0.79797453, + "num_input_tokens_seen": 18452590, + "step": 868, + "time_per_iteration": 2.614001512527466 + }, + { + "auxiliary_loss_clip": 0.01238409, + "auxiliary_loss_mlp": 0.01041801, + "balance_loss_clip": 1.07196403, + "balance_loss_mlp": 1.02740026, + "epoch": 0.10449107196537005, + "flos": 36283315374720.0, + "grad_norm": 1.5949516287715788, + "language_loss": 0.77716565, + "learning_rate": 3.942155780381001e-06, + "loss": 0.79996777, + "num_input_tokens_seen": 18476325, + "step": 869, + "time_per_iteration": 2.8985512256622314 + }, + { + "auxiliary_loss_clip": 0.01230503, + "auxiliary_loss_mlp": 0.01052799, + "balance_loss_clip": 1.06509364, + "balance_loss_mlp": 1.03763592, + "epoch": 0.10461131485600914, + "flos": 23802095266560.0, + "grad_norm": 2.4949942562943135, + "language_loss": 0.75805241, + "learning_rate": 3.94196964353013e-06, + "loss": 0.7808854, + "num_input_tokens_seen": 18495775, + "step": 870, + "time_per_iteration": 2.8881452083587646 + }, + { + "auxiliary_loss_clip": 0.01224881, + "auxiliary_loss_mlp": 0.00888952, + "balance_loss_clip": 1.06285405, + "balance_loss_mlp": 1.00046134, + "epoch": 0.10473155774664823, + "flos": 18405476104320.0, + "grad_norm": 2.6436396582564066, + "language_loss": 0.80979073, + "learning_rate": 3.941783212084008e-06, + "loss": 0.83092904, + "num_input_tokens_seen": 18513530, + "step": 871, + "time_per_iteration": 2.7108139991760254 + }, + { + "auxiliary_loss_clip": 0.01212367, + "auxiliary_loss_mlp": 0.01038543, + "balance_loss_clip": 1.06140459, + "balance_loss_mlp": 1.02462482, + "epoch": 0.10485180063728732, + "flos": 25592637358080.0, + "grad_norm": 2.548544537381103, + "language_loss": 0.78921723, + "learning_rate": 3.941596486070916e-06, + "loss": 0.81172633, + "num_input_tokens_seen": 18531575, + "step": 872, + "time_per_iteration": 2.803384780883789 + }, + { + "auxiliary_loss_clip": 0.01206351, + "auxiliary_loss_mlp": 0.01045681, + "balance_loss_clip": 1.06489837, + "balance_loss_mlp": 1.03017139, + "epoch": 0.10497204352792641, + "flos": 27088747666560.0, + "grad_norm": 2.89598140937977, + "language_loss": 0.58888596, + "learning_rate": 3.941409465519182e-06, + "loss": 0.61140633, + "num_input_tokens_seen": 18552100, + "step": 873, + "time_per_iteration": 2.8364999294281006 + }, + { + "auxiliary_loss_clip": 0.01227574, + "auxiliary_loss_mlp": 0.01044785, + "balance_loss_clip": 1.06291747, + "balance_loss_mlp": 1.0293119, + "epoch": 0.10509228641856551, + "flos": 32858479353600.0, + "grad_norm": 1.8189856557705955, + "language_loss": 0.85319304, + "learning_rate": 3.941222150457176e-06, + "loss": 0.87591666, + "num_input_tokens_seen": 18575355, + "step": 874, + "time_per_iteration": 2.8619189262390137 + }, + { + "auxiliary_loss_clip": 0.01246265, + "auxiliary_loss_mlp": 0.01043378, + "balance_loss_clip": 1.06610835, + "balance_loss_mlp": 1.02804732, + "epoch": 0.10521252930920459, + "flos": 14319165173760.0, + "grad_norm": 3.012874641755252, + "language_loss": 0.71256763, + "learning_rate": 3.941034540913311e-06, + "loss": 0.73546404, + "num_input_tokens_seen": 18592885, + "step": 875, + "time_per_iteration": 2.7782888412475586 + }, + { + "auxiliary_loss_clip": 0.01242576, + "auxiliary_loss_mlp": 0.00890384, + "balance_loss_clip": 1.06878114, + "balance_loss_mlp": 1.00035214, + "epoch": 0.10533277219984369, + "flos": 21687028773120.0, + "grad_norm": 1.7514833772093572, + "language_loss": 0.82326353, + "learning_rate": 3.940846636916051e-06, + "loss": 0.84459305, + "num_input_tokens_seen": 18612920, + "step": 876, + "time_per_iteration": 2.8932676315307617 + }, + { + "auxiliary_loss_clip": 0.01228504, + "auxiliary_loss_mlp": 0.0104553, + "balance_loss_clip": 1.06953442, + "balance_loss_mlp": 1.03108156, + "epoch": 0.10545301509048277, + "flos": 22269787027200.0, + "grad_norm": 2.0061845796504016, + "language_loss": 0.86530465, + "learning_rate": 3.940658438493899e-06, + "loss": 0.88804501, + "num_input_tokens_seen": 18630765, + "step": 877, + "time_per_iteration": 3.7336232662200928 + }, + { + "auxiliary_loss_clip": 0.01253301, + "auxiliary_loss_mlp": 0.01050609, + "balance_loss_clip": 1.06668365, + "balance_loss_mlp": 1.03467071, + "epoch": 0.10557325798112187, + "flos": 22199725549440.0, + "grad_norm": 2.0670911283788396, + "language_loss": 0.75787091, + "learning_rate": 3.940469945675405e-06, + "loss": 0.78091002, + "num_input_tokens_seen": 18649150, + "step": 878, + "time_per_iteration": 2.760542154312134 + }, + { + "auxiliary_loss_clip": 0.01186208, + "auxiliary_loss_mlp": 0.01041614, + "balance_loss_clip": 1.05814362, + "balance_loss_mlp": 1.02825058, + "epoch": 0.10569350087176095, + "flos": 25775889569280.0, + "grad_norm": 2.289514550831609, + "language_loss": 0.91490829, + "learning_rate": 3.940281158489163e-06, + "loss": 0.93718654, + "num_input_tokens_seen": 18668380, + "step": 879, + "time_per_iteration": 4.851658582687378 + }, + { + "auxiliary_loss_clip": 0.01191396, + "auxiliary_loss_mlp": 0.01041096, + "balance_loss_clip": 1.05630016, + "balance_loss_mlp": 1.02656436, + "epoch": 0.10581374376240005, + "flos": 17311385790720.0, + "grad_norm": 1.8993084735639074, + "language_loss": 0.82779813, + "learning_rate": 3.940092076963812e-06, + "loss": 0.85012299, + "num_input_tokens_seen": 18685875, + "step": 880, + "time_per_iteration": 2.9002585411071777 + }, + { + "auxiliary_loss_clip": 0.01228611, + "auxiliary_loss_mlp": 0.01042015, + "balance_loss_clip": 1.06311691, + "balance_loss_mlp": 1.02709591, + "epoch": 0.10593398665303914, + "flos": 34349454017280.0, + "grad_norm": 2.040963792990023, + "language_loss": 0.78808796, + "learning_rate": 3.9399027011280355e-06, + "loss": 0.81079423, + "num_input_tokens_seen": 18707970, + "step": 881, + "time_per_iteration": 2.8634519577026367 + }, + { + "auxiliary_loss_clip": 0.0123044, + "auxiliary_loss_mlp": 0.01044562, + "balance_loss_clip": 1.07059574, + "balance_loss_mlp": 1.02910006, + "epoch": 0.10605422954367823, + "flos": 23257977068160.0, + "grad_norm": 2.134606065267226, + "language_loss": 0.77334517, + "learning_rate": 3.939713031010561e-06, + "loss": 0.79609519, + "num_input_tokens_seen": 18726335, + "step": 882, + "time_per_iteration": 2.819000005722046 + }, + { + "auxiliary_loss_clip": 0.01217057, + "auxiliary_loss_mlp": 0.01046016, + "balance_loss_clip": 1.06567001, + "balance_loss_mlp": 1.0305779, + "epoch": 0.10617447243431732, + "flos": 22820118278400.0, + "grad_norm": 2.119774385647235, + "language_loss": 0.7772789, + "learning_rate": 3.939523066640163e-06, + "loss": 0.79990965, + "num_input_tokens_seen": 18745230, + "step": 883, + "time_per_iteration": 2.7513275146484375 + }, + { + "auxiliary_loss_clip": 0.01238856, + "auxiliary_loss_mlp": 0.01044774, + "balance_loss_clip": 1.06559253, + "balance_loss_mlp": 1.03058779, + "epoch": 0.10629471532495641, + "flos": 24386577373440.0, + "grad_norm": 5.67746765656349, + "language_loss": 0.81254375, + "learning_rate": 3.939332808045657e-06, + "loss": 0.83538008, + "num_input_tokens_seen": 18764880, + "step": 884, + "time_per_iteration": 2.7531650066375732 + }, + { + "auxiliary_loss_clip": 0.01221935, + "auxiliary_loss_mlp": 0.01046728, + "balance_loss_clip": 1.06605148, + "balance_loss_mlp": 1.03201795, + "epoch": 0.1064149582155955, + "flos": 21105491581440.0, + "grad_norm": 1.8045473196862032, + "language_loss": 0.84765875, + "learning_rate": 3.939142255255906e-06, + "loss": 0.87034535, + "num_input_tokens_seen": 18785765, + "step": 885, + "time_per_iteration": 2.83636474609375 + }, + { + "auxiliary_loss_clip": 0.01234701, + "auxiliary_loss_mlp": 0.01046894, + "balance_loss_clip": 1.06380224, + "balance_loss_mlp": 1.03249335, + "epoch": 0.1065352011062346, + "flos": 20702035042560.0, + "grad_norm": 2.2236481212769363, + "language_loss": 0.86885679, + "learning_rate": 3.938951408299817e-06, + "loss": 0.89167273, + "num_input_tokens_seen": 18804605, + "step": 886, + "time_per_iteration": 2.7779948711395264 + }, + { + "auxiliary_loss_clip": 0.01114165, + "auxiliary_loss_mlp": 0.01008034, + "balance_loss_clip": 1.04516792, + "balance_loss_mlp": 1.00121522, + "epoch": 0.10665544399687368, + "flos": 62659632689280.0, + "grad_norm": 0.8049116304787608, + "language_loss": 0.54500079, + "learning_rate": 3.938760267206342e-06, + "loss": 0.56622279, + "num_input_tokens_seen": 18866425, + "step": 887, + "time_per_iteration": 3.3890275955200195 + }, + { + "auxiliary_loss_clip": 0.01249825, + "auxiliary_loss_mlp": 0.01038524, + "balance_loss_clip": 1.06944585, + "balance_loss_mlp": 1.02412391, + "epoch": 0.10677568688751278, + "flos": 26140382830080.0, + "grad_norm": 4.965102462824858, + "language_loss": 0.78875065, + "learning_rate": 3.938568832004475e-06, + "loss": 0.81163418, + "num_input_tokens_seen": 18885130, + "step": 888, + "time_per_iteration": 2.682617664337158 + }, + { + "auxiliary_loss_clip": 0.01216412, + "auxiliary_loss_mlp": 0.0104598, + "balance_loss_clip": 1.06037438, + "balance_loss_mlp": 1.03199697, + "epoch": 0.10689592977815186, + "flos": 12786533712000.0, + "grad_norm": 1.9876799530684623, + "language_loss": 0.75208139, + "learning_rate": 3.938377102723257e-06, + "loss": 0.77470529, + "num_input_tokens_seen": 18902265, + "step": 889, + "time_per_iteration": 2.9927263259887695 + }, + { + "auxiliary_loss_clip": 0.01190962, + "auxiliary_loss_mlp": 0.01051941, + "balance_loss_clip": 1.05531561, + "balance_loss_mlp": 1.03688526, + "epoch": 0.10701617266879096, + "flos": 22126683242880.0, + "grad_norm": 2.2300964798089167, + "language_loss": 0.83323342, + "learning_rate": 3.938185079391774e-06, + "loss": 0.85566247, + "num_input_tokens_seen": 18919310, + "step": 890, + "time_per_iteration": 2.918800115585327 + }, + { + "auxiliary_loss_clip": 0.01252998, + "auxiliary_loss_mlp": 0.01039057, + "balance_loss_clip": 1.06938934, + "balance_loss_mlp": 1.02339292, + "epoch": 0.10713641555943004, + "flos": 19745625559680.0, + "grad_norm": 2.9925492206499067, + "language_loss": 1.06370831, + "learning_rate": 3.937992762039157e-06, + "loss": 1.08662879, + "num_input_tokens_seen": 18932635, + "step": 891, + "time_per_iteration": 2.6432950496673584 + }, + { + "auxiliary_loss_clip": 0.01236209, + "auxiliary_loss_mlp": 0.01049944, + "balance_loss_clip": 1.0668087, + "balance_loss_mlp": 1.03642011, + "epoch": 0.10725665845006914, + "flos": 23952992302080.0, + "grad_norm": 2.9379872768212145, + "language_loss": 0.80626082, + "learning_rate": 3.937800150694577e-06, + "loss": 0.82912236, + "num_input_tokens_seen": 18953810, + "step": 892, + "time_per_iteration": 2.8639628887176514 + }, + { + "auxiliary_loss_clip": 0.01209051, + "auxiliary_loss_mlp": 0.01047043, + "balance_loss_clip": 1.06191468, + "balance_loss_mlp": 1.03203487, + "epoch": 0.10737690134070824, + "flos": 18551704371840.0, + "grad_norm": 2.2215829626183683, + "language_loss": 0.76156968, + "learning_rate": 3.937607245387255e-06, + "loss": 0.78413057, + "num_input_tokens_seen": 18973175, + "step": 893, + "time_per_iteration": 2.934842348098755 + }, + { + "auxiliary_loss_clip": 0.01235149, + "auxiliary_loss_mlp": 0.010373, + "balance_loss_clip": 1.06386852, + "balance_loss_mlp": 1.02277994, + "epoch": 0.10749714423134732, + "flos": 22707609903360.0, + "grad_norm": 2.8196698329876395, + "language_loss": 0.7226609, + "learning_rate": 3.937414046146455e-06, + "loss": 0.74538541, + "num_input_tokens_seen": 18991130, + "step": 894, + "time_per_iteration": 2.905452013015747 + }, + { + "auxiliary_loss_clip": 0.01253658, + "auxiliary_loss_mlp": 0.01040463, + "balance_loss_clip": 1.07128835, + "balance_loss_mlp": 1.02638412, + "epoch": 0.10761738712198642, + "flos": 21106066199040.0, + "grad_norm": 2.8793639646582623, + "language_loss": 0.75474948, + "learning_rate": 3.9372205530014845e-06, + "loss": 0.77769071, + "num_input_tokens_seen": 19009610, + "step": 895, + "time_per_iteration": 2.7655951976776123 + }, + { + "auxiliary_loss_clip": 0.01250091, + "auxiliary_loss_mlp": 0.01044934, + "balance_loss_clip": 1.066522, + "balance_loss_mlp": 1.03061116, + "epoch": 0.1077376300126255, + "flos": 23766723348480.0, + "grad_norm": 2.212761776874843, + "language_loss": 0.71369958, + "learning_rate": 3.937026765981696e-06, + "loss": 0.73664981, + "num_input_tokens_seen": 19029680, + "step": 896, + "time_per_iteration": 2.7680349349975586 + }, + { + "auxiliary_loss_clip": 0.01219641, + "auxiliary_loss_mlp": 0.01041784, + "balance_loss_clip": 1.06501365, + "balance_loss_mlp": 1.02850401, + "epoch": 0.1078578729032646, + "flos": 20919581763840.0, + "grad_norm": 1.8264779113737872, + "language_loss": 0.79546332, + "learning_rate": 3.936832685116488e-06, + "loss": 0.81807756, + "num_input_tokens_seen": 19047775, + "step": 897, + "time_per_iteration": 2.9595842361450195 + }, + { + "auxiliary_loss_clip": 0.01247263, + "auxiliary_loss_mlp": 0.01048717, + "balance_loss_clip": 1.06663108, + "balance_loss_mlp": 1.03406, + "epoch": 0.10797811579390369, + "flos": 14829886702080.0, + "grad_norm": 2.3273130707943137, + "language_loss": 0.899454, + "learning_rate": 3.936638310435301e-06, + "loss": 0.92241377, + "num_input_tokens_seen": 19065640, + "step": 898, + "time_per_iteration": 2.763903856277466 + }, + { + "auxiliary_loss_clip": 0.01244143, + "auxiliary_loss_mlp": 0.01042213, + "balance_loss_clip": 1.06878495, + "balance_loss_mlp": 1.02778828, + "epoch": 0.10809835868454278, + "flos": 19536985411200.0, + "grad_norm": 1.8999522779469222, + "language_loss": 0.81393856, + "learning_rate": 3.936443641967623e-06, + "loss": 0.83680212, + "num_input_tokens_seen": 19084470, + "step": 899, + "time_per_iteration": 2.759519100189209 + }, + { + "auxiliary_loss_clip": 0.01229876, + "auxiliary_loss_mlp": 0.01042987, + "balance_loss_clip": 1.06402326, + "balance_loss_mlp": 1.02880144, + "epoch": 0.10821860157518187, + "flos": 18442320480000.0, + "grad_norm": 2.4026740686221446, + "language_loss": 0.83085084, + "learning_rate": 3.936248679742983e-06, + "loss": 0.85357952, + "num_input_tokens_seen": 19102965, + "step": 900, + "time_per_iteration": 2.738675594329834 + }, + { + "auxiliary_loss_clip": 0.01124946, + "auxiliary_loss_mlp": 0.01013514, + "balance_loss_clip": 1.04449892, + "balance_loss_mlp": 1.00707674, + "epoch": 0.10833884446582095, + "flos": 49359468447360.0, + "grad_norm": 1.0844304654802612, + "language_loss": 0.70212615, + "learning_rate": 3.936053423790959e-06, + "loss": 0.72351074, + "num_input_tokens_seen": 19151285, + "step": 901, + "time_per_iteration": 3.236013650894165 + }, + { + "auxiliary_loss_clip": 0.01248923, + "auxiliary_loss_mlp": 0.01042859, + "balance_loss_clip": 1.06851721, + "balance_loss_mlp": 1.02886415, + "epoch": 0.10845908735646005, + "flos": 20411912891520.0, + "grad_norm": 1.8632201216741993, + "language_loss": 0.77604938, + "learning_rate": 3.935857874141168e-06, + "loss": 0.79896712, + "num_input_tokens_seen": 19170120, + "step": 902, + "time_per_iteration": 2.776557683944702 + }, + { + "auxiliary_loss_clip": 0.01225234, + "auxiliary_loss_mlp": 0.01042561, + "balance_loss_clip": 1.06673384, + "balance_loss_mlp": 1.02818418, + "epoch": 0.10857933024709913, + "flos": 14027750133120.0, + "grad_norm": 2.763461038344769, + "language_loss": 0.84016854, + "learning_rate": 3.935662030823279e-06, + "loss": 0.86284649, + "num_input_tokens_seen": 19186305, + "step": 903, + "time_per_iteration": 3.7617173194885254 + }, + { + "auxiliary_loss_clip": 0.01239904, + "auxiliary_loss_mlp": 0.01043622, + "balance_loss_clip": 1.06674266, + "balance_loss_mlp": 1.02848268, + "epoch": 0.10869957313773823, + "flos": 13369004657280.0, + "grad_norm": 2.5787818906644104, + "language_loss": 0.72848678, + "learning_rate": 3.935465893866998e-06, + "loss": 0.75132203, + "num_input_tokens_seen": 19204530, + "step": 904, + "time_per_iteration": 3.682278871536255 + }, + { + "auxiliary_loss_clip": 0.0122701, + "auxiliary_loss_mlp": 0.01048584, + "balance_loss_clip": 1.06520295, + "balance_loss_mlp": 1.03353953, + "epoch": 0.10881981602837733, + "flos": 25807095509760.0, + "grad_norm": 1.9592381104106793, + "language_loss": 0.80029905, + "learning_rate": 3.935269463302079e-06, + "loss": 0.82305497, + "num_input_tokens_seen": 19222735, + "step": 905, + "time_per_iteration": 3.840993642807007 + }, + { + "auxiliary_loss_clip": 0.01244842, + "auxiliary_loss_mlp": 0.01044517, + "balance_loss_clip": 1.06877601, + "balance_loss_mlp": 1.03023624, + "epoch": 0.10894005891901641, + "flos": 20777555387520.0, + "grad_norm": 1.826439129292967, + "language_loss": 0.7656936, + "learning_rate": 3.935072739158322e-06, + "loss": 0.78858721, + "num_input_tokens_seen": 19242445, + "step": 906, + "time_per_iteration": 3.5979039669036865 + }, + { + "auxiliary_loss_clip": 0.01226595, + "auxiliary_loss_mlp": 0.01041305, + "balance_loss_clip": 1.06336021, + "balance_loss_mlp": 1.02729809, + "epoch": 0.10906030180965551, + "flos": 26649883296000.0, + "grad_norm": 1.5554095509110824, + "language_loss": 0.79565275, + "learning_rate": 3.934875721465569e-06, + "loss": 0.81833172, + "num_input_tokens_seen": 19262865, + "step": 907, + "time_per_iteration": 3.009371042251587 + }, + { + "auxiliary_loss_clip": 0.01225268, + "auxiliary_loss_mlp": 0.01047596, + "balance_loss_clip": 1.06160164, + "balance_loss_mlp": 1.03249216, + "epoch": 0.10918054470029459, + "flos": 36534402420480.0, + "grad_norm": 3.0404354954751254, + "language_loss": 0.7180419, + "learning_rate": 3.9346784102537076e-06, + "loss": 0.74077052, + "num_input_tokens_seen": 19285000, + "step": 908, + "time_per_iteration": 2.933147668838501 + }, + { + "auxiliary_loss_clip": 0.01247241, + "auxiliary_loss_mlp": 0.01046982, + "balance_loss_clip": 1.06638265, + "balance_loss_mlp": 1.03261709, + "epoch": 0.10930078759093369, + "flos": 21762549118080.0, + "grad_norm": 2.4494055594471296, + "language_loss": 0.78328919, + "learning_rate": 3.934480805552669e-06, + "loss": 0.80623138, + "num_input_tokens_seen": 19306010, + "step": 909, + "time_per_iteration": 2.738286018371582 + }, + { + "auxiliary_loss_clip": 0.01245737, + "auxiliary_loss_mlp": 0.00889523, + "balance_loss_clip": 1.06631374, + "balance_loss_mlp": 1.00032961, + "epoch": 0.10942103048157277, + "flos": 22601781457920.0, + "grad_norm": 3.5677335570230717, + "language_loss": 0.88172424, + "learning_rate": 3.93428290739243e-06, + "loss": 0.90307689, + "num_input_tokens_seen": 19325380, + "step": 910, + "time_per_iteration": 2.7627463340759277 + }, + { + "auxiliary_loss_clip": 0.01225845, + "auxiliary_loss_mlp": 0.0104217, + "balance_loss_clip": 1.06347287, + "balance_loss_mlp": 1.02882433, + "epoch": 0.10954127337221187, + "flos": 15045781397760.0, + "grad_norm": 2.3038183322391848, + "language_loss": 0.80339295, + "learning_rate": 3.9340847158030125e-06, + "loss": 0.82607305, + "num_input_tokens_seen": 19338960, + "step": 911, + "time_per_iteration": 2.756404399871826 + }, + { + "auxiliary_loss_clip": 0.01243235, + "auxiliary_loss_mlp": 0.01052094, + "balance_loss_clip": 1.06624818, + "balance_loss_mlp": 1.03808677, + "epoch": 0.10966151626285096, + "flos": 21650974496640.0, + "grad_norm": 1.7626798591095094, + "language_loss": 0.75641876, + "learning_rate": 3.9338862308144814e-06, + "loss": 0.7793721, + "num_input_tokens_seen": 19357780, + "step": 912, + "time_per_iteration": 2.780594825744629 + }, + { + "auxiliary_loss_clip": 0.01248119, + "auxiliary_loss_mlp": 0.01045335, + "balance_loss_clip": 1.06853628, + "balance_loss_mlp": 1.03171492, + "epoch": 0.10978175915349005, + "flos": 20121359777280.0, + "grad_norm": 1.654402690682711, + "language_loss": 0.8460691, + "learning_rate": 3.933687452456946e-06, + "loss": 0.86900365, + "num_input_tokens_seen": 19377680, + "step": 913, + "time_per_iteration": 2.7177634239196777 + }, + { + "auxiliary_loss_clip": 0.01218528, + "auxiliary_loss_mlp": 0.01038679, + "balance_loss_clip": 1.06204295, + "balance_loss_mlp": 1.02331269, + "epoch": 0.10990200204412914, + "flos": 20412667077120.0, + "grad_norm": 2.033200216788641, + "language_loss": 0.85938966, + "learning_rate": 3.933488380760562e-06, + "loss": 0.8819617, + "num_input_tokens_seen": 19397040, + "step": 914, + "time_per_iteration": 2.7883498668670654 + }, + { + "auxiliary_loss_clip": 0.01248119, + "auxiliary_loss_mlp": 0.00889992, + "balance_loss_clip": 1.06611526, + "balance_loss_mlp": 1.00033164, + "epoch": 0.11002224493476823, + "flos": 17530117660800.0, + "grad_norm": 1.9473962169767005, + "language_loss": 0.87358356, + "learning_rate": 3.9332890157555286e-06, + "loss": 0.89496464, + "num_input_tokens_seen": 19413975, + "step": 915, + "time_per_iteration": 2.6486217975616455 + }, + { + "auxiliary_loss_clip": 0.0123226, + "auxiliary_loss_mlp": 0.01038601, + "balance_loss_clip": 1.0657618, + "balance_loss_mlp": 1.02384341, + "epoch": 0.11014248782540732, + "flos": 12203093099520.0, + "grad_norm": 1.8531292508122785, + "language_loss": 0.76372737, + "learning_rate": 3.933089357472088e-06, + "loss": 0.78643596, + "num_input_tokens_seen": 19432005, + "step": 916, + "time_per_iteration": 2.8235690593719482 + }, + { + "auxiliary_loss_clip": 0.01249184, + "auxiliary_loss_mlp": 0.01040235, + "balance_loss_clip": 1.06978798, + "balance_loss_mlp": 1.02681768, + "epoch": 0.11026273071604642, + "flos": 22382977760640.0, + "grad_norm": 1.8544603249691805, + "language_loss": 0.85741591, + "learning_rate": 3.932889405940529e-06, + "loss": 0.88031006, + "num_input_tokens_seen": 19450100, + "step": 917, + "time_per_iteration": 2.7084615230560303 + }, + { + "auxiliary_loss_clip": 0.01235683, + "auxiliary_loss_mlp": 0.01042257, + "balance_loss_clip": 1.07277393, + "balance_loss_mlp": 1.02794552, + "epoch": 0.1103829736066855, + "flos": 19829046896640.0, + "grad_norm": 2.3061125197471855, + "language_loss": 0.80249369, + "learning_rate": 3.932689161191184e-06, + "loss": 0.8252731, + "num_input_tokens_seen": 19467805, + "step": 918, + "time_per_iteration": 2.8891851902008057 + }, + { + "auxiliary_loss_clip": 0.01235189, + "auxiliary_loss_mlp": 0.01040001, + "balance_loss_clip": 1.06300175, + "balance_loss_mlp": 1.02596974, + "epoch": 0.1105032164973246, + "flos": 22669616292480.0, + "grad_norm": 6.694246791923, + "language_loss": 0.87684536, + "learning_rate": 3.93248862325443e-06, + "loss": 0.89959729, + "num_input_tokens_seen": 19486710, + "step": 919, + "time_per_iteration": 2.7533938884735107 + }, + { + "auxiliary_loss_clip": 0.01133899, + "auxiliary_loss_mlp": 0.01012746, + "balance_loss_clip": 1.03332138, + "balance_loss_mlp": 1.00506866, + "epoch": 0.11062345938796368, + "flos": 66483507876480.0, + "grad_norm": 0.942752578915378, + "language_loss": 0.64453471, + "learning_rate": 3.932287792160688e-06, + "loss": 0.6660012, + "num_input_tokens_seen": 19545170, + "step": 920, + "time_per_iteration": 3.210653781890869 + }, + { + "auxiliary_loss_clip": 0.01238731, + "auxiliary_loss_mlp": 0.01056735, + "balance_loss_clip": 1.06454611, + "balance_loss_mlp": 1.04200625, + "epoch": 0.11074370227860278, + "flos": 21907771804800.0, + "grad_norm": 3.082781222597812, + "language_loss": 0.809304, + "learning_rate": 3.932086667940424e-06, + "loss": 0.8322587, + "num_input_tokens_seen": 19561875, + "step": 921, + "time_per_iteration": 2.9033942222595215 + }, + { + "auxiliary_loss_clip": 0.01232887, + "auxiliary_loss_mlp": 0.00889187, + "balance_loss_clip": 1.06607127, + "balance_loss_mlp": 1.0003922, + "epoch": 0.11086394516924186, + "flos": 28658115763200.0, + "grad_norm": 1.9979762835393424, + "language_loss": 0.8212285, + "learning_rate": 3.93188525062415e-06, + "loss": 0.84244919, + "num_input_tokens_seen": 19582340, + "step": 922, + "time_per_iteration": 2.873680830001831 + }, + { + "auxiliary_loss_clip": 0.0124437, + "auxiliary_loss_mlp": 0.01043359, + "balance_loss_clip": 1.07019973, + "balance_loss_mlp": 1.02820706, + "epoch": 0.11098418805988096, + "flos": 24535247765760.0, + "grad_norm": 2.38528795456072, + "language_loss": 0.86230797, + "learning_rate": 3.931683540242418e-06, + "loss": 0.88518524, + "num_input_tokens_seen": 19603405, + "step": 923, + "time_per_iteration": 2.7846105098724365 + }, + { + "auxiliary_loss_clip": 0.01233171, + "auxiliary_loss_mlp": 0.01040569, + "balance_loss_clip": 1.06383538, + "balance_loss_mlp": 1.02669311, + "epoch": 0.11110443095052006, + "flos": 22960384888320.0, + "grad_norm": 2.9159699358366367, + "language_loss": 0.9128058, + "learning_rate": 3.9314815368258295e-06, + "loss": 0.93554318, + "num_input_tokens_seen": 19619885, + "step": 924, + "time_per_iteration": 2.858534097671509 + }, + { + "auxiliary_loss_clip": 0.01245289, + "auxiliary_loss_mlp": 0.01052207, + "balance_loss_clip": 1.0725553, + "balance_loss_mlp": 1.03750277, + "epoch": 0.11122467384115914, + "flos": 18950025265920.0, + "grad_norm": 2.088494035918748, + "language_loss": 0.7881754, + "learning_rate": 3.9312792404050275e-06, + "loss": 0.81115037, + "num_input_tokens_seen": 19637940, + "step": 925, + "time_per_iteration": 2.7647759914398193 + }, + { + "auxiliary_loss_clip": 0.01245585, + "auxiliary_loss_mlp": 0.01041197, + "balance_loss_clip": 1.0685339, + "balance_loss_mlp": 1.02696347, + "epoch": 0.11134491673179824, + "flos": 25082957324160.0, + "grad_norm": 1.9033771998387932, + "language_loss": 0.77125359, + "learning_rate": 3.9310766510107e-06, + "loss": 0.79412138, + "num_input_tokens_seen": 19657115, + "step": 926, + "time_per_iteration": 2.694040298461914 + }, + { + "auxiliary_loss_clip": 0.01222229, + "auxiliary_loss_mlp": 0.01045956, + "balance_loss_clip": 1.06253541, + "balance_loss_mlp": 1.030864, + "epoch": 0.11146515962243732, + "flos": 24499121662080.0, + "grad_norm": 1.9021701242760138, + "language_loss": 0.91873848, + "learning_rate": 3.9308737686735806e-06, + "loss": 0.94142038, + "num_input_tokens_seen": 19677075, + "step": 927, + "time_per_iteration": 2.7873852252960205 + }, + { + "auxiliary_loss_clip": 0.0125418, + "auxiliary_loss_mlp": 0.01038909, + "balance_loss_clip": 1.07044697, + "balance_loss_mlp": 1.02539647, + "epoch": 0.11158540251307641, + "flos": 22343763087360.0, + "grad_norm": 2.124269302140775, + "language_loss": 0.82999218, + "learning_rate": 3.9306705934244455e-06, + "loss": 0.8529231, + "num_input_tokens_seen": 19697155, + "step": 928, + "time_per_iteration": 2.7339437007904053 + }, + { + "auxiliary_loss_clip": 0.0121499, + "auxiliary_loss_mlp": 0.0105643, + "balance_loss_clip": 1.0628314, + "balance_loss_mlp": 1.04244685, + "epoch": 0.11170564540371551, + "flos": 19902304684800.0, + "grad_norm": 1.825098415668274, + "language_loss": 0.88267601, + "learning_rate": 3.930467125294116e-06, + "loss": 0.90539026, + "num_input_tokens_seen": 19716705, + "step": 929, + "time_per_iteration": 2.8132712841033936 + }, + { + "auxiliary_loss_clip": 0.01092196, + "auxiliary_loss_mlp": 0.0101267, + "balance_loss_clip": 1.03026354, + "balance_loss_mlp": 1.0062331, + "epoch": 0.1118258882943546, + "flos": 64586239499520.0, + "grad_norm": 0.9199954998542199, + "language_loss": 0.60493958, + "learning_rate": 3.930263364313458e-06, + "loss": 0.62598825, + "num_input_tokens_seen": 19767275, + "step": 930, + "time_per_iteration": 5.289549350738525 + }, + { + "auxiliary_loss_clip": 0.0121998, + "auxiliary_loss_mlp": 0.01048142, + "balance_loss_clip": 1.06481385, + "balance_loss_mlp": 1.03309178, + "epoch": 0.11194613118499369, + "flos": 17201965985280.0, + "grad_norm": 2.0241633222937803, + "language_loss": 0.83093387, + "learning_rate": 3.930059310513384e-06, + "loss": 0.85361505, + "num_input_tokens_seen": 19786315, + "step": 931, + "time_per_iteration": 4.822428464889526 + }, + { + "auxiliary_loss_clip": 0.01202604, + "auxiliary_loss_mlp": 0.0088976, + "balance_loss_clip": 1.06001425, + "balance_loss_mlp": 1.00038052, + "epoch": 0.11206637407563277, + "flos": 31863465728640.0, + "grad_norm": 1.8795251147513816, + "language_loss": 0.8383249, + "learning_rate": 3.929854963924846e-06, + "loss": 0.85924852, + "num_input_tokens_seen": 19806580, + "step": 932, + "time_per_iteration": 2.8924989700317383 + }, + { + "auxiliary_loss_clip": 0.01211711, + "auxiliary_loss_mlp": 0.01048344, + "balance_loss_clip": 1.05803335, + "balance_loss_mlp": 1.03409851, + "epoch": 0.11218661696627187, + "flos": 21945621761280.0, + "grad_norm": 2.0631909768936425, + "language_loss": 0.7748673, + "learning_rate": 3.929650324578845e-06, + "loss": 0.79746783, + "num_input_tokens_seen": 19826045, + "step": 933, + "time_per_iteration": 2.774959087371826 + }, + { + "auxiliary_loss_clip": 0.01239513, + "auxiliary_loss_mlp": 0.01047182, + "balance_loss_clip": 1.06935072, + "balance_loss_mlp": 1.03110015, + "epoch": 0.11230685985691095, + "flos": 25878198481920.0, + "grad_norm": 2.262177427022899, + "language_loss": 0.81909782, + "learning_rate": 3.929445392506423e-06, + "loss": 0.84196478, + "num_input_tokens_seen": 19843985, + "step": 934, + "time_per_iteration": 2.7619643211364746 + }, + { + "auxiliary_loss_clip": 0.01240043, + "auxiliary_loss_mlp": 0.01039014, + "balance_loss_clip": 1.07099497, + "balance_loss_mlp": 1.02528083, + "epoch": 0.11242710274755005, + "flos": 22231506107520.0, + "grad_norm": 2.3528393590923296, + "language_loss": 0.75631791, + "learning_rate": 3.92924016773867e-06, + "loss": 0.77910852, + "num_input_tokens_seen": 19860480, + "step": 935, + "time_per_iteration": 2.799530029296875 + }, + { + "auxiliary_loss_clip": 0.01226288, + "auxiliary_loss_mlp": 0.00889317, + "balance_loss_clip": 1.06244349, + "balance_loss_mlp": 1.00033641, + "epoch": 0.11254734563818915, + "flos": 17712184723200.0, + "grad_norm": 2.3633241540059196, + "language_loss": 0.73498476, + "learning_rate": 3.9290346503067175e-06, + "loss": 0.75614083, + "num_input_tokens_seen": 19877145, + "step": 936, + "time_per_iteration": 2.7430365085601807 + }, + { + "auxiliary_loss_clip": 0.01242596, + "auxiliary_loss_mlp": 0.01044915, + "balance_loss_clip": 1.06665671, + "balance_loss_mlp": 1.03040707, + "epoch": 0.11266758852882823, + "flos": 54930397334400.0, + "grad_norm": 1.9555498262247533, + "language_loss": 0.78799176, + "learning_rate": 3.9288288402417415e-06, + "loss": 0.81086683, + "num_input_tokens_seen": 19903405, + "step": 937, + "time_per_iteration": 3.0689749717712402 + }, + { + "auxiliary_loss_clip": 0.01247182, + "auxiliary_loss_mlp": 0.01044857, + "balance_loss_clip": 1.07038343, + "balance_loss_mlp": 1.02961004, + "epoch": 0.11278783141946733, + "flos": 18878132194560.0, + "grad_norm": 2.988802479591511, + "language_loss": 0.70430499, + "learning_rate": 3.928622737574964e-06, + "loss": 0.72722542, + "num_input_tokens_seen": 19918740, + "step": 938, + "time_per_iteration": 2.745634078979492 + }, + { + "auxiliary_loss_clip": 0.01225811, + "auxiliary_loss_mlp": 0.01048655, + "balance_loss_clip": 1.06287813, + "balance_loss_mlp": 1.03444564, + "epoch": 0.11290807431010641, + "flos": 26469252777600.0, + "grad_norm": 1.96336584815214, + "language_loss": 0.90749812, + "learning_rate": 3.928416342337652e-06, + "loss": 0.93024272, + "num_input_tokens_seen": 19938475, + "step": 939, + "time_per_iteration": 2.8939316272735596 + }, + { + "auxiliary_loss_clip": 0.01226286, + "auxiliary_loss_mlp": 0.01044113, + "balance_loss_clip": 1.06342435, + "balance_loss_mlp": 1.03024936, + "epoch": 0.1130283172007455, + "flos": 22710590732160.0, + "grad_norm": 1.87571717401337, + "language_loss": 0.82828379, + "learning_rate": 3.928209654561113e-06, + "loss": 0.85098773, + "num_input_tokens_seen": 19959310, + "step": 940, + "time_per_iteration": 2.793246030807495 + }, + { + "auxiliary_loss_clip": 0.01220639, + "auxiliary_loss_mlp": 0.01045556, + "balance_loss_clip": 1.06487989, + "balance_loss_mlp": 1.03088164, + "epoch": 0.1131485600913846, + "flos": 23219911630080.0, + "grad_norm": 2.3759027875177776, + "language_loss": 0.8170262, + "learning_rate": 3.928002674276703e-06, + "loss": 0.83968818, + "num_input_tokens_seen": 19978700, + "step": 941, + "time_per_iteration": 2.8859753608703613 + }, + { + "auxiliary_loss_clip": 0.01187416, + "auxiliary_loss_mlp": 0.01042755, + "balance_loss_clip": 1.05326629, + "balance_loss_mlp": 1.02767491, + "epoch": 0.11326880298202369, + "flos": 14064271286400.0, + "grad_norm": 2.3771201126731096, + "language_loss": 0.75953937, + "learning_rate": 3.92779540151582e-06, + "loss": 0.7818411, + "num_input_tokens_seen": 19995785, + "step": 942, + "time_per_iteration": 2.84700608253479 + }, + { + "auxiliary_loss_clip": 0.01233562, + "auxiliary_loss_mlp": 0.0105041, + "balance_loss_clip": 1.0662539, + "balance_loss_mlp": 1.03624845, + "epoch": 0.11338904587266278, + "flos": 16325386479360.0, + "grad_norm": 10.058665328818968, + "language_loss": 0.85598195, + "learning_rate": 3.927587836309907e-06, + "loss": 0.87882173, + "num_input_tokens_seen": 20013615, + "step": 943, + "time_per_iteration": 2.7525620460510254 + }, + { + "auxiliary_loss_clip": 0.01218878, + "auxiliary_loss_mlp": 0.01048734, + "balance_loss_clip": 1.05875897, + "balance_loss_mlp": 1.03408289, + "epoch": 0.11350928876330187, + "flos": 24426258923520.0, + "grad_norm": 1.9073790712927758, + "language_loss": 0.78040004, + "learning_rate": 3.927379978690452e-06, + "loss": 0.80307615, + "num_input_tokens_seen": 20032880, + "step": 944, + "time_per_iteration": 2.811901807785034 + }, + { + "auxiliary_loss_clip": 0.01202738, + "auxiliary_loss_mlp": 0.01044772, + "balance_loss_clip": 1.05469346, + "balance_loss_mlp": 1.03015721, + "epoch": 0.11362953165394096, + "flos": 24497074586880.0, + "grad_norm": 2.776347098134001, + "language_loss": 0.87367541, + "learning_rate": 3.927171828688987e-06, + "loss": 0.89615059, + "num_input_tokens_seen": 20052405, + "step": 945, + "time_per_iteration": 2.8247642517089844 + }, + { + "auxiliary_loss_clip": 0.01249424, + "auxiliary_loss_mlp": 0.01041407, + "balance_loss_clip": 1.0683645, + "balance_loss_mlp": 1.02655339, + "epoch": 0.11374977454458005, + "flos": 24060831909120.0, + "grad_norm": 2.209001471638725, + "language_loss": 0.82344204, + "learning_rate": 3.926963386337088e-06, + "loss": 0.84635037, + "num_input_tokens_seen": 20070635, + "step": 946, + "time_per_iteration": 2.7435011863708496 + }, + { + "auxiliary_loss_clip": 0.0125062, + "auxiliary_loss_mlp": 0.01047207, + "balance_loss_clip": 1.06650591, + "balance_loss_mlp": 1.03186512, + "epoch": 0.11387001743521914, + "flos": 39457638967680.0, + "grad_norm": 3.8455872717806465, + "language_loss": 0.70172805, + "learning_rate": 3.926754651666375e-06, + "loss": 0.72470629, + "num_input_tokens_seen": 20091195, + "step": 947, + "time_per_iteration": 2.774960994720459 + }, + { + "auxiliary_loss_clip": 0.01217643, + "auxiliary_loss_mlp": 0.01045994, + "balance_loss_clip": 1.06473899, + "balance_loss_mlp": 1.03126025, + "epoch": 0.11399026032585824, + "flos": 25082454533760.0, + "grad_norm": 2.8237633881333073, + "language_loss": 0.78134048, + "learning_rate": 3.926545624708513e-06, + "loss": 0.80397689, + "num_input_tokens_seen": 20110435, + "step": 948, + "time_per_iteration": 2.8308706283569336 + }, + { + "auxiliary_loss_clip": 0.01211797, + "auxiliary_loss_mlp": 0.01044719, + "balance_loss_clip": 1.05998075, + "balance_loss_mlp": 1.03074217, + "epoch": 0.11411050321649732, + "flos": 17961835224960.0, + "grad_norm": 1.8205919966005975, + "language_loss": 0.85813224, + "learning_rate": 3.926336305495213e-06, + "loss": 0.88069737, + "num_input_tokens_seen": 20128995, + "step": 949, + "time_per_iteration": 2.752286434173584 + }, + { + "auxiliary_loss_clip": 0.01197709, + "auxiliary_loss_mlp": 0.01045756, + "balance_loss_clip": 1.05636764, + "balance_loss_mlp": 1.03168321, + "epoch": 0.11423074610713642, + "flos": 22455409536000.0, + "grad_norm": 2.786416654773633, + "language_loss": 0.88960993, + "learning_rate": 3.926126694058226e-06, + "loss": 0.91204458, + "num_input_tokens_seen": 20148145, + "step": 950, + "time_per_iteration": 2.9090845584869385 + }, + { + "auxiliary_loss_clip": 0.01202054, + "auxiliary_loss_mlp": 0.01043261, + "balance_loss_clip": 1.06257069, + "balance_loss_mlp": 1.0292064, + "epoch": 0.1143509889977755, + "flos": 19717687756800.0, + "grad_norm": 1.5596750290509258, + "language_loss": 0.81968141, + "learning_rate": 3.92591679042935e-06, + "loss": 0.84213459, + "num_input_tokens_seen": 20168035, + "step": 951, + "time_per_iteration": 2.929340124130249 + }, + { + "auxiliary_loss_clip": 0.01236665, + "auxiliary_loss_mlp": 0.01038236, + "balance_loss_clip": 1.06784439, + "balance_loss_mlp": 1.02353752, + "epoch": 0.1144712318884146, + "flos": 19822869757440.0, + "grad_norm": 1.9635554039140353, + "language_loss": 0.82335246, + "learning_rate": 3.92570659464043e-06, + "loss": 0.84610152, + "num_input_tokens_seen": 20186095, + "step": 952, + "time_per_iteration": 2.8394935131073 + }, + { + "auxiliary_loss_clip": 0.01232303, + "auxiliary_loss_mlp": 0.00889235, + "balance_loss_clip": 1.06662822, + "balance_loss_mlp": 1.00028157, + "epoch": 0.1145914747790537, + "flos": 14939198766720.0, + "grad_norm": 2.018200691510421, + "language_loss": 0.79772019, + "learning_rate": 3.925496106723349e-06, + "loss": 0.81893551, + "num_input_tokens_seen": 20203535, + "step": 953, + "time_per_iteration": 2.7850184440612793 + }, + { + "auxiliary_loss_clip": 0.01236214, + "auxiliary_loss_mlp": 0.01043386, + "balance_loss_clip": 1.06504285, + "balance_loss_mlp": 1.02957606, + "epoch": 0.11471171766969278, + "flos": 19865029345920.0, + "grad_norm": 1.9290315002093756, + "language_loss": 0.84254491, + "learning_rate": 3.9252853267100405e-06, + "loss": 0.86534095, + "num_input_tokens_seen": 20222780, + "step": 954, + "time_per_iteration": 2.7979133129119873 + }, + { + "auxiliary_loss_clip": 0.01210622, + "auxiliary_loss_mlp": 0.01055662, + "balance_loss_clip": 1.0613904, + "balance_loss_mlp": 1.040344, + "epoch": 0.11483196056033187, + "flos": 22526476594560.0, + "grad_norm": 1.899943776221632, + "language_loss": 0.83773017, + "learning_rate": 3.9250742546324786e-06, + "loss": 0.86039305, + "num_input_tokens_seen": 20243015, + "step": 955, + "time_per_iteration": 3.7774267196655273 + }, + { + "auxiliary_loss_clip": 0.01224055, + "auxiliary_loss_mlp": 0.01041639, + "balance_loss_clip": 1.06132221, + "balance_loss_mlp": 1.0283891, + "epoch": 0.11495220345097096, + "flos": 28220292887040.0, + "grad_norm": 1.7765408077098708, + "language_loss": 0.8657707, + "learning_rate": 3.924862890522683e-06, + "loss": 0.88842762, + "num_input_tokens_seen": 20263025, + "step": 956, + "time_per_iteration": 3.777254104614258 + }, + { + "auxiliary_loss_clip": 0.01234487, + "auxiliary_loss_mlp": 0.0104748, + "balance_loss_clip": 1.06301498, + "balance_loss_mlp": 1.03318715, + "epoch": 0.11507244634161005, + "flos": 17492267704320.0, + "grad_norm": 2.4736867907702194, + "language_loss": 0.86323357, + "learning_rate": 3.9246512344127174e-06, + "loss": 0.88605326, + "num_input_tokens_seen": 20280685, + "step": 957, + "time_per_iteration": 3.7014052867889404 + }, + { + "auxiliary_loss_clip": 0.01172168, + "auxiliary_loss_mlp": 0.01043873, + "balance_loss_clip": 1.05410385, + "balance_loss_mlp": 1.03041387, + "epoch": 0.11519268923224914, + "flos": 22564937082240.0, + "grad_norm": 2.2501951731165533, + "language_loss": 0.8193543, + "learning_rate": 3.9244392863346895e-06, + "loss": 0.84151471, + "num_input_tokens_seen": 20300090, + "step": 958, + "time_per_iteration": 3.7743093967437744 + }, + { + "auxiliary_loss_clip": 0.01231109, + "auxiliary_loss_mlp": 0.01050689, + "balance_loss_clip": 1.06914568, + "balance_loss_mlp": 1.0349052, + "epoch": 0.11531293212288823, + "flos": 16982839065600.0, + "grad_norm": 2.312030540382754, + "language_loss": 0.92195791, + "learning_rate": 3.9242270463207524e-06, + "loss": 0.94477588, + "num_input_tokens_seen": 20318480, + "step": 959, + "time_per_iteration": 2.8276751041412354 + }, + { + "auxiliary_loss_clip": 0.01193726, + "auxiliary_loss_mlp": 0.01043765, + "balance_loss_clip": 1.05885184, + "balance_loss_mlp": 1.02949536, + "epoch": 0.11543317501352733, + "flos": 12422004537600.0, + "grad_norm": 3.2297614521580207, + "language_loss": 0.8474195, + "learning_rate": 3.924014514403102e-06, + "loss": 0.86979437, + "num_input_tokens_seen": 20334635, + "step": 960, + "time_per_iteration": 2.9068758487701416 + }, + { + "auxiliary_loss_clip": 0.01197601, + "auxiliary_loss_mlp": 0.01041367, + "balance_loss_clip": 1.05789816, + "balance_loss_mlp": 1.02625096, + "epoch": 0.11555341790416641, + "flos": 19821648695040.0, + "grad_norm": 2.446062865503036, + "language_loss": 0.9122088, + "learning_rate": 3.92380169061398e-06, + "loss": 0.93459845, + "num_input_tokens_seen": 20352415, + "step": 961, + "time_per_iteration": 2.852588415145874 + }, + { + "auxiliary_loss_clip": 0.01201167, + "auxiliary_loss_mlp": 0.00889045, + "balance_loss_clip": 1.05300093, + "balance_loss_mlp": 1.00024247, + "epoch": 0.11567366079480551, + "flos": 25738865625600.0, + "grad_norm": 2.277853029319161, + "language_loss": 0.84313607, + "learning_rate": 3.9235885749856705e-06, + "loss": 0.86403823, + "num_input_tokens_seen": 20371095, + "step": 962, + "time_per_iteration": 2.8492825031280518 + }, + { + "auxiliary_loss_clip": 0.01226312, + "auxiliary_loss_mlp": 0.0104557, + "balance_loss_clip": 1.06606746, + "balance_loss_mlp": 1.03075194, + "epoch": 0.1157939036854446, + "flos": 18223301301120.0, + "grad_norm": 1.816741444766388, + "language_loss": 0.82470274, + "learning_rate": 3.9233751675505035e-06, + "loss": 0.84742165, + "num_input_tokens_seen": 20389805, + "step": 963, + "time_per_iteration": 2.77134108543396 + }, + { + "auxiliary_loss_clip": 0.01218468, + "auxiliary_loss_mlp": 0.01045435, + "balance_loss_clip": 1.06447315, + "balance_loss_mlp": 1.0314697, + "epoch": 0.11591414657608369, + "flos": 23073755189760.0, + "grad_norm": 2.1061353596358683, + "language_loss": 0.85035944, + "learning_rate": 3.923161468340853e-06, + "loss": 0.87299848, + "num_input_tokens_seen": 20409640, + "step": 964, + "time_per_iteration": 2.8355531692504883 + }, + { + "auxiliary_loss_clip": 0.01189243, + "auxiliary_loss_mlp": 0.01043371, + "balance_loss_clip": 1.05227876, + "balance_loss_mlp": 1.02806461, + "epoch": 0.11603438946672277, + "flos": 19461716461440.0, + "grad_norm": 1.8160846596311806, + "language_loss": 0.81575882, + "learning_rate": 3.9229474773891374e-06, + "loss": 0.83808494, + "num_input_tokens_seen": 20428180, + "step": 965, + "time_per_iteration": 2.9270360469818115 + }, + { + "auxiliary_loss_clip": 0.01220802, + "auxiliary_loss_mlp": 0.01042473, + "balance_loss_clip": 1.0593816, + "balance_loss_mlp": 1.02714252, + "epoch": 0.11615463235736187, + "flos": 26831986272000.0, + "grad_norm": 2.3672566846319683, + "language_loss": 0.83571392, + "learning_rate": 3.922733194727818e-06, + "loss": 0.8583467, + "num_input_tokens_seen": 20447975, + "step": 966, + "time_per_iteration": 2.9411463737487793 + }, + { + "auxiliary_loss_clip": 0.01241823, + "auxiliary_loss_mlp": 0.01046353, + "balance_loss_clip": 1.06800926, + "balance_loss_mlp": 1.03184533, + "epoch": 0.11627487524800097, + "flos": 18580324533120.0, + "grad_norm": 1.9133893519936174, + "language_loss": 0.87317282, + "learning_rate": 3.922518620389402e-06, + "loss": 0.89605451, + "num_input_tokens_seen": 20464840, + "step": 967, + "time_per_iteration": 2.7733147144317627 + }, + { + "auxiliary_loss_clip": 0.01147866, + "auxiliary_loss_mlp": 0.01050511, + "balance_loss_clip": 1.04902661, + "balance_loss_mlp": 1.0369451, + "epoch": 0.11639511813864005, + "flos": 18150474476160.0, + "grad_norm": 1.939690243068435, + "language_loss": 0.89436054, + "learning_rate": 3.922303754406439e-06, + "loss": 0.91634429, + "num_input_tokens_seen": 20482680, + "step": 968, + "time_per_iteration": 2.9637413024902344 + }, + { + "auxiliary_loss_clip": 0.01189287, + "auxiliary_loss_mlp": 0.01050692, + "balance_loss_clip": 1.05198896, + "balance_loss_mlp": 1.036816, + "epoch": 0.11651536102927915, + "flos": 20922023888640.0, + "grad_norm": 5.75089702902508, + "language_loss": 0.79190588, + "learning_rate": 3.922088596811526e-06, + "loss": 0.81430566, + "num_input_tokens_seen": 20501810, + "step": 969, + "time_per_iteration": 2.8348491191864014 + }, + { + "auxiliary_loss_clip": 0.01225342, + "auxiliary_loss_mlp": 0.01049833, + "balance_loss_clip": 1.06213355, + "balance_loss_mlp": 1.03570664, + "epoch": 0.11663560391991823, + "flos": 16508602776960.0, + "grad_norm": 2.2211587738530376, + "language_loss": 0.86928546, + "learning_rate": 3.9218731476373e-06, + "loss": 0.89203715, + "num_input_tokens_seen": 20517995, + "step": 970, + "time_per_iteration": 2.7944960594177246 + }, + { + "auxiliary_loss_clip": 0.01239824, + "auxiliary_loss_mlp": 0.0105175, + "balance_loss_clip": 1.06782413, + "balance_loss_mlp": 1.03662801, + "epoch": 0.11675584681055733, + "flos": 19865029345920.0, + "grad_norm": 5.093025831804058, + "language_loss": 0.84939861, + "learning_rate": 3.9216574069164455e-06, + "loss": 0.87231439, + "num_input_tokens_seen": 20536970, + "step": 971, + "time_per_iteration": 2.782179355621338 + }, + { + "auxiliary_loss_clip": 0.01242038, + "auxiliary_loss_mlp": 0.0104201, + "balance_loss_clip": 1.06676853, + "balance_loss_mlp": 1.02845633, + "epoch": 0.11687608970119642, + "flos": 21944364785280.0, + "grad_norm": 1.5997712660719763, + "language_loss": 0.80227923, + "learning_rate": 3.921441374681691e-06, + "loss": 0.82511973, + "num_input_tokens_seen": 20557030, + "step": 972, + "time_per_iteration": 2.6662020683288574 + }, + { + "auxiliary_loss_clip": 0.01215947, + "auxiliary_loss_mlp": 0.01050097, + "balance_loss_clip": 1.0617255, + "balance_loss_mlp": 1.03603077, + "epoch": 0.1169963325918355, + "flos": 24061155131520.0, + "grad_norm": 1.8875191561268698, + "language_loss": 0.65062505, + "learning_rate": 3.921225050965808e-06, + "loss": 0.67328548, + "num_input_tokens_seen": 20576915, + "step": 973, + "time_per_iteration": 2.774291515350342 + }, + { + "auxiliary_loss_clip": 0.01215612, + "auxiliary_loss_mlp": 0.01048638, + "balance_loss_clip": 1.06254447, + "balance_loss_mlp": 1.03453517, + "epoch": 0.1171165754824746, + "flos": 23368151059200.0, + "grad_norm": 2.7857173463088953, + "language_loss": 0.7537117, + "learning_rate": 3.921008435801612e-06, + "loss": 0.77635419, + "num_input_tokens_seen": 20596000, + "step": 974, + "time_per_iteration": 2.7693264484405518 + }, + { + "auxiliary_loss_clip": 0.01226715, + "auxiliary_loss_mlp": 0.01042991, + "balance_loss_clip": 1.06462181, + "balance_loss_mlp": 1.02828074, + "epoch": 0.11723681837311369, + "flos": 18552243075840.0, + "grad_norm": 2.2555672170736862, + "language_loss": 0.75877583, + "learning_rate": 3.920791529221963e-06, + "loss": 0.78147292, + "num_input_tokens_seen": 20614675, + "step": 975, + "time_per_iteration": 2.742429256439209 + }, + { + "auxiliary_loss_clip": 0.01223943, + "auxiliary_loss_mlp": 0.0088884, + "balance_loss_clip": 1.06119502, + "balance_loss_mlp": 1.00034201, + "epoch": 0.11735706126375278, + "flos": 23550541344000.0, + "grad_norm": 1.937239143380487, + "language_loss": 0.7629717, + "learning_rate": 3.920574331259768e-06, + "loss": 0.78409946, + "num_input_tokens_seen": 20635875, + "step": 976, + "time_per_iteration": 2.850330352783203 + }, + { + "auxiliary_loss_clip": 0.01213381, + "auxiliary_loss_mlp": 0.01046587, + "balance_loss_clip": 1.05961204, + "balance_loss_mlp": 1.03241873, + "epoch": 0.11747730415439187, + "flos": 22381541216640.0, + "grad_norm": 2.4295035671688754, + "language_loss": 0.79355645, + "learning_rate": 3.9203568419479716e-06, + "loss": 0.81615615, + "num_input_tokens_seen": 20656430, + "step": 977, + "time_per_iteration": 2.910594940185547 + }, + { + "auxiliary_loss_clip": 0.01224631, + "auxiliary_loss_mlp": 0.01045651, + "balance_loss_clip": 1.06413257, + "balance_loss_mlp": 1.0330565, + "epoch": 0.11759754704503096, + "flos": 22200731130240.0, + "grad_norm": 1.7636676154689426, + "language_loss": 0.75368166, + "learning_rate": 3.92013906131957e-06, + "loss": 0.77638453, + "num_input_tokens_seen": 20675360, + "step": 978, + "time_per_iteration": 2.8347928524017334 + }, + { + "auxiliary_loss_clip": 0.01206024, + "auxiliary_loss_mlp": 0.0104076, + "balance_loss_clip": 1.05982828, + "balance_loss_mlp": 1.02680087, + "epoch": 0.11771778993567006, + "flos": 22309755886080.0, + "grad_norm": 1.6516214096886017, + "language_loss": 0.82579249, + "learning_rate": 3.9199209894076e-06, + "loss": 0.84826028, + "num_input_tokens_seen": 20695675, + "step": 979, + "time_per_iteration": 2.7767794132232666 + }, + { + "auxiliary_loss_clip": 0.01248366, + "auxiliary_loss_mlp": 0.01049574, + "balance_loss_clip": 1.06597185, + "balance_loss_mlp": 1.03429127, + "epoch": 0.11783803282630914, + "flos": 21288169175040.0, + "grad_norm": 2.438176424638343, + "language_loss": 0.90207112, + "learning_rate": 3.919702626245142e-06, + "loss": 0.92505056, + "num_input_tokens_seen": 20715330, + "step": 980, + "time_per_iteration": 2.721912145614624 + }, + { + "auxiliary_loss_clip": 0.01210938, + "auxiliary_loss_mlp": 0.01044301, + "balance_loss_clip": 1.05955601, + "balance_loss_mlp": 1.03072906, + "epoch": 0.11795827571694824, + "flos": 25371535190400.0, + "grad_norm": 2.2568032679414483, + "language_loss": 0.6601783, + "learning_rate": 3.919483971865322e-06, + "loss": 0.68273073, + "num_input_tokens_seen": 20735325, + "step": 981, + "time_per_iteration": 2.757075071334839 + }, + { + "auxiliary_loss_clip": 0.01217662, + "auxiliary_loss_mlp": 0.01045211, + "balance_loss_clip": 1.06300533, + "balance_loss_mlp": 1.03118038, + "epoch": 0.11807851860758732, + "flos": 23622218933760.0, + "grad_norm": 2.032833927099876, + "language_loss": 0.87919688, + "learning_rate": 3.91926502630131e-06, + "loss": 0.90182561, + "num_input_tokens_seen": 20755940, + "step": 982, + "time_per_iteration": 4.638034820556641 + }, + { + "auxiliary_loss_clip": 0.01238251, + "auxiliary_loss_mlp": 0.01048628, + "balance_loss_clip": 1.06791568, + "balance_loss_mlp": 1.03533006, + "epoch": 0.11819876149822642, + "flos": 24972496024320.0, + "grad_norm": 1.752556986512156, + "language_loss": 0.72093314, + "learning_rate": 3.91904578958632e-06, + "loss": 0.74380195, + "num_input_tokens_seen": 20775355, + "step": 983, + "time_per_iteration": 4.602415323257446 + }, + { + "auxiliary_loss_clip": 0.01247237, + "auxiliary_loss_mlp": 0.01044204, + "balance_loss_clip": 1.06795466, + "balance_loss_mlp": 1.03072166, + "epoch": 0.11831900438886551, + "flos": 23003226835200.0, + "grad_norm": 2.5518412751300383, + "language_loss": 0.84331167, + "learning_rate": 3.918826261753608e-06, + "loss": 0.86622608, + "num_input_tokens_seen": 20794935, + "step": 984, + "time_per_iteration": 2.719599962234497 + }, + { + "auxiliary_loss_clip": 0.01222253, + "auxiliary_loss_mlp": 0.01038283, + "balance_loss_clip": 1.06324911, + "balance_loss_mlp": 1.02495563, + "epoch": 0.1184392472795046, + "flos": 27965147604480.0, + "grad_norm": 2.81098087339723, + "language_loss": 0.71153557, + "learning_rate": 3.918606442836478e-06, + "loss": 0.73414093, + "num_input_tokens_seen": 20817155, + "step": 985, + "time_per_iteration": 2.799292802810669 + }, + { + "auxiliary_loss_clip": 0.01231457, + "auxiliary_loss_mlp": 0.0104544, + "balance_loss_clip": 1.06721818, + "balance_loss_mlp": 1.03298879, + "epoch": 0.1185594901701437, + "flos": 19898497843200.0, + "grad_norm": 1.862748173567425, + "language_loss": 0.77294147, + "learning_rate": 3.918386332868277e-06, + "loss": 0.79571044, + "num_input_tokens_seen": 20835125, + "step": 986, + "time_per_iteration": 2.784477472305298 + }, + { + "auxiliary_loss_clip": 0.01228526, + "auxiliary_loss_mlp": 0.01041414, + "balance_loss_clip": 1.06655848, + "balance_loss_mlp": 1.02764583, + "epoch": 0.11867973306078278, + "flos": 18912354877440.0, + "grad_norm": 1.8161586885675383, + "language_loss": 0.94593012, + "learning_rate": 3.918165931882394e-06, + "loss": 0.9686296, + "num_input_tokens_seen": 20853525, + "step": 987, + "time_per_iteration": 2.7383522987365723 + }, + { + "auxiliary_loss_clip": 0.01173573, + "auxiliary_loss_mlp": 0.01042833, + "balance_loss_clip": 1.05055594, + "balance_loss_mlp": 1.02970779, + "epoch": 0.11879997595142187, + "flos": 16982803152000.0, + "grad_norm": 2.306271777887355, + "language_loss": 0.75812006, + "learning_rate": 3.917945239912264e-06, + "loss": 0.78028411, + "num_input_tokens_seen": 20871000, + "step": 988, + "time_per_iteration": 2.9186031818389893 + }, + { + "auxiliary_loss_clip": 0.01196963, + "auxiliary_loss_mlp": 0.01037809, + "balance_loss_clip": 1.05853009, + "balance_loss_mlp": 1.02373004, + "epoch": 0.11892021884206096, + "flos": 17530369056000.0, + "grad_norm": 5.250065839107923, + "language_loss": 0.75537753, + "learning_rate": 3.917724256991367e-06, + "loss": 0.77772528, + "num_input_tokens_seen": 20889745, + "step": 989, + "time_per_iteration": 2.8712096214294434 + }, + { + "auxiliary_loss_clip": 0.01215126, + "auxiliary_loss_mlp": 0.01039226, + "balance_loss_clip": 1.06225193, + "balance_loss_mlp": 1.02538586, + "epoch": 0.11904046173270005, + "flos": 30955895763840.0, + "grad_norm": 2.555917348169514, + "language_loss": 0.81451309, + "learning_rate": 3.9175029831532245e-06, + "loss": 0.83705664, + "num_input_tokens_seen": 20909260, + "step": 990, + "time_per_iteration": 2.922546863555908 + }, + { + "auxiliary_loss_clip": 0.01216176, + "auxiliary_loss_mlp": 0.01040458, + "balance_loss_clip": 1.06356907, + "balance_loss_mlp": 1.02750635, + "epoch": 0.11916070462333915, + "flos": 20157234485760.0, + "grad_norm": 2.1144259595676536, + "language_loss": 0.88537025, + "learning_rate": 3.917281418431404e-06, + "loss": 0.90793657, + "num_input_tokens_seen": 20928305, + "step": 991, + "time_per_iteration": 2.8997726440429688 + }, + { + "auxiliary_loss_clip": 0.0122216, + "auxiliary_loss_mlp": 0.01038838, + "balance_loss_clip": 1.06501293, + "balance_loss_mlp": 1.02524781, + "epoch": 0.11928094751397823, + "flos": 23551115961600.0, + "grad_norm": 1.8623340226467562, + "language_loss": 0.76896703, + "learning_rate": 3.917059562859516e-06, + "loss": 0.79157698, + "num_input_tokens_seen": 20947630, + "step": 992, + "time_per_iteration": 2.8184866905212402 + }, + { + "auxiliary_loss_clip": 0.01210661, + "auxiliary_loss_mlp": 0.0104725, + "balance_loss_clip": 1.06111264, + "balance_loss_mlp": 1.03405988, + "epoch": 0.11940119040461733, + "flos": 23908426502400.0, + "grad_norm": 2.0629100740774424, + "language_loss": 0.88802761, + "learning_rate": 3.916837416471218e-06, + "loss": 0.91060674, + "num_input_tokens_seen": 20964250, + "step": 993, + "time_per_iteration": 2.863542318344116 + }, + { + "auxiliary_loss_clip": 0.01225736, + "auxiliary_loss_mlp": 0.01037007, + "balance_loss_clip": 1.06119883, + "balance_loss_mlp": 1.02378106, + "epoch": 0.11952143329525641, + "flos": 13844533835520.0, + "grad_norm": 2.4570193286550612, + "language_loss": 0.72157401, + "learning_rate": 3.916614979300207e-06, + "loss": 0.74420142, + "num_input_tokens_seen": 20979095, + "step": 994, + "time_per_iteration": 2.6969094276428223 + }, + { + "auxiliary_loss_clip": 0.01195711, + "auxiliary_loss_mlp": 0.01046817, + "balance_loss_clip": 1.06092489, + "balance_loss_mlp": 1.03310239, + "epoch": 0.11964167618589551, + "flos": 27015525792000.0, + "grad_norm": 1.6386969762291412, + "language_loss": 0.78577256, + "learning_rate": 3.9163922513802274e-06, + "loss": 0.8081978, + "num_input_tokens_seen": 21001430, + "step": 995, + "time_per_iteration": 2.907346725463867 + }, + { + "auxiliary_loss_clip": 0.01246887, + "auxiliary_loss_mlp": 0.01041874, + "balance_loss_clip": 1.06620312, + "balance_loss_mlp": 1.02745581, + "epoch": 0.1197619190765346, + "flos": 12567622273920.0, + "grad_norm": 2.745843219787548, + "language_loss": 0.82301271, + "learning_rate": 3.916169232745067e-06, + "loss": 0.8459003, + "num_input_tokens_seen": 21019105, + "step": 996, + "time_per_iteration": 2.7033956050872803 + }, + { + "auxiliary_loss_clip": 0.01212031, + "auxiliary_loss_mlp": 0.01035001, + "balance_loss_clip": 1.06092465, + "balance_loss_mlp": 1.02101803, + "epoch": 0.11988216196717369, + "flos": 16909437623040.0, + "grad_norm": 2.8152445935137744, + "language_loss": 0.91736573, + "learning_rate": 3.915945923428559e-06, + "loss": 0.93983603, + "num_input_tokens_seen": 21035630, + "step": 997, + "time_per_iteration": 2.77656888961792 + }, + { + "auxiliary_loss_clip": 0.01228464, + "auxiliary_loss_mlp": 0.01035379, + "balance_loss_clip": 1.06238878, + "balance_loss_mlp": 1.02133656, + "epoch": 0.12000240485781279, + "flos": 16216577205120.0, + "grad_norm": 2.180675360058231, + "language_loss": 0.83234155, + "learning_rate": 3.915722323464577e-06, + "loss": 0.85498005, + "num_input_tokens_seen": 21054235, + "step": 998, + "time_per_iteration": 2.7702701091766357 + }, + { + "auxiliary_loss_clip": 0.01236477, + "auxiliary_loss_mlp": 0.01042197, + "balance_loss_clip": 1.06649709, + "balance_loss_mlp": 1.02779698, + "epoch": 0.12012264774845187, + "flos": 49344887525760.0, + "grad_norm": 2.376653193828638, + "language_loss": 0.70558137, + "learning_rate": 3.91549843288704e-06, + "loss": 0.72836804, + "num_input_tokens_seen": 21077915, + "step": 999, + "time_per_iteration": 2.97861385345459 + }, + { + "auxiliary_loss_clip": 0.01205713, + "auxiliary_loss_mlp": 0.00889115, + "balance_loss_clip": 1.05555773, + "balance_loss_mlp": 1.00026083, + "epoch": 0.12024289063909097, + "flos": 26979435601920.0, + "grad_norm": 1.8310116635851064, + "language_loss": 0.78937608, + "learning_rate": 3.915274251729916e-06, + "loss": 0.81032443, + "num_input_tokens_seen": 21099205, + "step": 1000, + "time_per_iteration": 2.9305481910705566 + }, + { + "auxiliary_loss_clip": 0.01214805, + "auxiliary_loss_mlp": 0.01035096, + "balance_loss_clip": 1.06197929, + "balance_loss_mlp": 1.02160132, + "epoch": 0.12036313352973005, + "flos": 19537308633600.0, + "grad_norm": 2.1676658237156574, + "language_loss": 0.9006995, + "learning_rate": 3.91504978002721e-06, + "loss": 0.92319858, + "num_input_tokens_seen": 21118260, + "step": 1001, + "time_per_iteration": 2.8194987773895264 + }, + { + "auxiliary_loss_clip": 0.01225275, + "auxiliary_loss_mlp": 0.00889207, + "balance_loss_clip": 1.06161213, + "balance_loss_mlp": 1.00022519, + "epoch": 0.12048337642036915, + "flos": 17268256535040.0, + "grad_norm": 5.8686856964811405, + "language_loss": 0.76285303, + "learning_rate": 3.914825017812974e-06, + "loss": 0.78399783, + "num_input_tokens_seen": 21134910, + "step": 1002, + "time_per_iteration": 2.90708589553833 + }, + { + "auxiliary_loss_clip": 0.01225431, + "auxiliary_loss_mlp": 0.01047112, + "balance_loss_clip": 1.06658912, + "balance_loss_mlp": 1.03475642, + "epoch": 0.12060361931100824, + "flos": 22856962654080.0, + "grad_norm": 2.3197525490084105, + "language_loss": 0.72353661, + "learning_rate": 3.9145999651213065e-06, + "loss": 0.74626207, + "num_input_tokens_seen": 21154150, + "step": 1003, + "time_per_iteration": 2.8329756259918213 + }, + { + "auxiliary_loss_clip": 0.01236483, + "auxiliary_loss_mlp": 0.01039652, + "balance_loss_clip": 1.06782269, + "balance_loss_mlp": 1.02549016, + "epoch": 0.12072386220164733, + "flos": 16726795943040.0, + "grad_norm": 3.1284301488753417, + "language_loss": 0.88375628, + "learning_rate": 3.9143746219863465e-06, + "loss": 0.90651762, + "num_input_tokens_seen": 21171255, + "step": 1004, + "time_per_iteration": 2.797086715698242 + }, + { + "auxiliary_loss_clip": 0.01135889, + "auxiliary_loss_mlp": 0.01014818, + "balance_loss_clip": 1.04138303, + "balance_loss_mlp": 1.00790429, + "epoch": 0.12084410509228642, + "flos": 55144176105600.0, + "grad_norm": 0.9305286048725123, + "language_loss": 0.64763099, + "learning_rate": 3.914148988442278e-06, + "loss": 0.66913807, + "num_input_tokens_seen": 21227045, + "step": 1005, + "time_per_iteration": 3.314077615737915 + }, + { + "auxiliary_loss_clip": 0.01216961, + "auxiliary_loss_mlp": 0.01039894, + "balance_loss_clip": 1.06564403, + "balance_loss_mlp": 1.02459967, + "epoch": 0.1209643479829255, + "flos": 26760236855040.0, + "grad_norm": 2.394954534884688, + "language_loss": 0.94905496, + "learning_rate": 3.91392306452333e-06, + "loss": 0.97162354, + "num_input_tokens_seen": 21244120, + "step": 1006, + "time_per_iteration": 2.799278736114502 + }, + { + "auxiliary_loss_clip": 0.0125238, + "auxiliary_loss_mlp": 0.01044122, + "balance_loss_clip": 1.0696342, + "balance_loss_mlp": 1.03003168, + "epoch": 0.1210845908735646, + "flos": 11035026725760.0, + "grad_norm": 2.783406606999136, + "language_loss": 0.66523433, + "learning_rate": 3.913696850263774e-06, + "loss": 0.68819934, + "num_input_tokens_seen": 21258485, + "step": 1007, + "time_per_iteration": 3.605086088180542 + }, + { + "auxiliary_loss_clip": 0.01234785, + "auxiliary_loss_mlp": 0.01041011, + "balance_loss_clip": 1.06479716, + "balance_loss_mlp": 1.02720642, + "epoch": 0.1212048337642037, + "flos": 20484631975680.0, + "grad_norm": 5.408376943293974, + "language_loss": 0.78922874, + "learning_rate": 3.913470345697929e-06, + "loss": 0.81198668, + "num_input_tokens_seen": 21277115, + "step": 1008, + "time_per_iteration": 3.630481719970703 + }, + { + "auxiliary_loss_clip": 0.01202481, + "auxiliary_loss_mlp": 0.01039748, + "balance_loss_clip": 1.06070769, + "balance_loss_mlp": 1.02622962, + "epoch": 0.12132507665484278, + "flos": 22346061557760.0, + "grad_norm": 2.630489994730872, + "language_loss": 0.85990053, + "learning_rate": 3.913243550860153e-06, + "loss": 0.88232285, + "num_input_tokens_seen": 21294880, + "step": 1009, + "time_per_iteration": 4.757833957672119 + }, + { + "auxiliary_loss_clip": 0.01243539, + "auxiliary_loss_mlp": 0.01051287, + "balance_loss_clip": 1.0726006, + "balance_loss_mlp": 1.03731573, + "epoch": 0.12144531954548188, + "flos": 29314957818240.0, + "grad_norm": 4.099942003722154, + "language_loss": 0.76277673, + "learning_rate": 3.913016465784852e-06, + "loss": 0.785725, + "num_input_tokens_seen": 21315555, + "step": 1010, + "time_per_iteration": 2.753202438354492 + }, + { + "auxiliary_loss_clip": 0.01205251, + "auxiliary_loss_mlp": 0.01050431, + "balance_loss_clip": 1.06066275, + "balance_loss_mlp": 1.03666222, + "epoch": 0.12156556243612096, + "flos": 20485242506880.0, + "grad_norm": 2.3076701813462583, + "language_loss": 0.71858966, + "learning_rate": 3.912789090506474e-06, + "loss": 0.74114645, + "num_input_tokens_seen": 21334815, + "step": 1011, + "time_per_iteration": 2.856842041015625 + }, + { + "auxiliary_loss_clip": 0.01218912, + "auxiliary_loss_mlp": 0.01042055, + "balance_loss_clip": 1.05929303, + "balance_loss_mlp": 1.02849531, + "epoch": 0.12168580532676006, + "flos": 16472009796480.0, + "grad_norm": 2.688027895644034, + "language_loss": 0.72467983, + "learning_rate": 3.9125614250595114e-06, + "loss": 0.74728948, + "num_input_tokens_seen": 21351025, + "step": 1012, + "time_per_iteration": 2.777045965194702 + }, + { + "auxiliary_loss_clip": 0.01238505, + "auxiliary_loss_mlp": 0.01042974, + "balance_loss_clip": 1.0670501, + "balance_loss_mlp": 1.02857387, + "epoch": 0.12180604821739914, + "flos": 15341290588800.0, + "grad_norm": 2.708445753640447, + "language_loss": 0.88940942, + "learning_rate": 3.912333469478502e-06, + "loss": 0.91222423, + "num_input_tokens_seen": 21368990, + "step": 1013, + "time_per_iteration": 2.8117923736572266 + }, + { + "auxiliary_loss_clip": 0.01223837, + "auxiliary_loss_mlp": 0.01045547, + "balance_loss_clip": 1.06281865, + "balance_loss_mlp": 1.03181386, + "epoch": 0.12192629110803824, + "flos": 19318038059520.0, + "grad_norm": 2.3098827713647645, + "language_loss": 0.7829994, + "learning_rate": 3.912105223798025e-06, + "loss": 0.80569327, + "num_input_tokens_seen": 21388410, + "step": 1014, + "time_per_iteration": 2.7886099815368652 + }, + { + "auxiliary_loss_clip": 0.01124007, + "auxiliary_loss_mlp": 0.01010662, + "balance_loss_clip": 1.03599799, + "balance_loss_mlp": 1.00379515, + "epoch": 0.12204653399867733, + "flos": 47725354085760.0, + "grad_norm": 1.0017781170193867, + "language_loss": 0.67725909, + "learning_rate": 3.9118766880527065e-06, + "loss": 0.69860578, + "num_input_tokens_seen": 21442845, + "step": 1015, + "time_per_iteration": 3.227062702178955 + }, + { + "auxiliary_loss_clip": 0.01196394, + "auxiliary_loss_mlp": 0.01042932, + "balance_loss_clip": 1.06005383, + "balance_loss_mlp": 1.02958083, + "epoch": 0.12216677688931642, + "flos": 18221936584320.0, + "grad_norm": 1.6605767598144698, + "language_loss": 0.74038863, + "learning_rate": 3.9116478622772145e-06, + "loss": 0.76278186, + "num_input_tokens_seen": 21461420, + "step": 1016, + "time_per_iteration": 2.856274127960205 + }, + { + "auxiliary_loss_clip": 0.01232015, + "auxiliary_loss_mlp": 0.0104113, + "balance_loss_clip": 1.06804204, + "balance_loss_mlp": 1.026932, + "epoch": 0.12228701977995551, + "flos": 27525636789120.0, + "grad_norm": 2.346340559775487, + "language_loss": 0.88277483, + "learning_rate": 3.911418746506261e-06, + "loss": 0.90550625, + "num_input_tokens_seen": 21481550, + "step": 1017, + "time_per_iteration": 2.7863478660583496 + }, + { + "auxiliary_loss_clip": 0.01241566, + "auxiliary_loss_mlp": 0.01038905, + "balance_loss_clip": 1.07245636, + "balance_loss_mlp": 1.02588773, + "epoch": 0.1224072626705946, + "flos": 21798136517760.0, + "grad_norm": 1.707312350102687, + "language_loss": 0.78183532, + "learning_rate": 3.911189340774604e-06, + "loss": 0.80464, + "num_input_tokens_seen": 21501680, + "step": 1018, + "time_per_iteration": 2.7876148223876953 + }, + { + "auxiliary_loss_clip": 0.01232424, + "auxiliary_loss_mlp": 0.01044123, + "balance_loss_clip": 1.06448317, + "balance_loss_mlp": 1.0300684, + "epoch": 0.1225275055612337, + "flos": 20703758895360.0, + "grad_norm": 2.6942189814197612, + "language_loss": 0.79334474, + "learning_rate": 3.910959645117043e-06, + "loss": 0.81611025, + "num_input_tokens_seen": 21521015, + "step": 1019, + "time_per_iteration": 2.7620651721954346 + }, + { + "auxiliary_loss_clip": 0.01121719, + "auxiliary_loss_mlp": 0.0087978, + "balance_loss_clip": 1.03586245, + "balance_loss_mlp": 1.00021684, + "epoch": 0.12264774845187278, + "flos": 57745294462080.0, + "grad_norm": 0.813599301702528, + "language_loss": 0.56716597, + "learning_rate": 3.910729659568423e-06, + "loss": 0.58718097, + "num_input_tokens_seen": 21578200, + "step": 1020, + "time_per_iteration": 3.3386449813842773 + }, + { + "auxiliary_loss_clip": 0.01226579, + "auxiliary_loss_mlp": 0.01038998, + "balance_loss_clip": 1.06741452, + "balance_loss_mlp": 1.02614141, + "epoch": 0.12276799134251187, + "flos": 26396282298240.0, + "grad_norm": 2.871965424961315, + "language_loss": 0.82415968, + "learning_rate": 3.9104993841636344e-06, + "loss": 0.84681547, + "num_input_tokens_seen": 21598770, + "step": 1021, + "time_per_iteration": 2.841958999633789 + }, + { + "auxiliary_loss_clip": 0.01218446, + "auxiliary_loss_mlp": 0.00888176, + "balance_loss_clip": 1.0669117, + "balance_loss_mlp": 1.00027859, + "epoch": 0.12288823423315097, + "flos": 21064193919360.0, + "grad_norm": 1.9888633176090038, + "language_loss": 0.80745959, + "learning_rate": 3.910268818937608e-06, + "loss": 0.82852578, + "num_input_tokens_seen": 21616925, + "step": 1022, + "time_per_iteration": 2.89363431930542 + }, + { + "auxiliary_loss_clip": 0.01205026, + "auxiliary_loss_mlp": 0.01037313, + "balance_loss_clip": 1.06340361, + "balance_loss_mlp": 1.02454591, + "epoch": 0.12300847712379005, + "flos": 12312441077760.0, + "grad_norm": 2.6669662354110537, + "language_loss": 0.87417424, + "learning_rate": 3.9100379639253196e-06, + "loss": 0.89659768, + "num_input_tokens_seen": 21633645, + "step": 1023, + "time_per_iteration": 2.814981698989868 + }, + { + "auxiliary_loss_clip": 0.0121919, + "auxiliary_loss_mlp": 0.01045696, + "balance_loss_clip": 1.05953646, + "balance_loss_mlp": 1.03192103, + "epoch": 0.12312872001442915, + "flos": 16762239688320.0, + "grad_norm": 3.1717692554148633, + "language_loss": 0.86712223, + "learning_rate": 3.909806819161791e-06, + "loss": 0.8897711, + "num_input_tokens_seen": 21649120, + "step": 1024, + "time_per_iteration": 2.7646074295043945 + }, + { + "auxiliary_loss_clip": 0.0121823, + "auxiliary_loss_mlp": 0.01041807, + "balance_loss_clip": 1.06075025, + "balance_loss_mlp": 1.02885485, + "epoch": 0.12324896290506823, + "flos": 18404937400320.0, + "grad_norm": 2.362564025553954, + "language_loss": 0.86071253, + "learning_rate": 3.909575384682086e-06, + "loss": 0.88331282, + "num_input_tokens_seen": 21668000, + "step": 1025, + "time_per_iteration": 2.9124794006347656 + }, + { + "auxiliary_loss_clip": 0.01239009, + "auxiliary_loss_mlp": 0.01041199, + "balance_loss_clip": 1.06491923, + "balance_loss_mlp": 1.02815175, + "epoch": 0.12336920579570733, + "flos": 18915407533440.0, + "grad_norm": 1.8194543534905294, + "language_loss": 0.69341898, + "learning_rate": 3.9093436605213144e-06, + "loss": 0.71622109, + "num_input_tokens_seen": 21688500, + "step": 1026, + "time_per_iteration": 2.783292531967163 + }, + { + "auxiliary_loss_clip": 0.01221639, + "auxiliary_loss_mlp": 0.01040221, + "balance_loss_clip": 1.06324029, + "balance_loss_mlp": 1.02651191, + "epoch": 0.12348944868634643, + "flos": 23878369797120.0, + "grad_norm": 1.7596958361332826, + "language_loss": 0.79278302, + "learning_rate": 3.909111646714627e-06, + "loss": 0.81540161, + "num_input_tokens_seen": 21709345, + "step": 1027, + "time_per_iteration": 2.828688859939575 + }, + { + "auxiliary_loss_clip": 0.01238769, + "auxiliary_loss_mlp": 0.01036653, + "balance_loss_clip": 1.06440103, + "balance_loss_mlp": 1.02368307, + "epoch": 0.12360969157698551, + "flos": 19026084314880.0, + "grad_norm": 2.1047619723650626, + "language_loss": 0.72553653, + "learning_rate": 3.9088793432972206e-06, + "loss": 0.74829078, + "num_input_tokens_seen": 21728165, + "step": 1028, + "time_per_iteration": 2.7414424419403076 + }, + { + "auxiliary_loss_clip": 0.01202331, + "auxiliary_loss_mlp": 0.01041044, + "balance_loss_clip": 1.06071961, + "balance_loss_mlp": 1.02809775, + "epoch": 0.1237299344676246, + "flos": 13224607983360.0, + "grad_norm": 2.107462791985471, + "language_loss": 0.82181728, + "learning_rate": 3.908646750304336e-06, + "loss": 0.84425104, + "num_input_tokens_seen": 21745850, + "step": 1029, + "time_per_iteration": 2.850050449371338 + }, + { + "auxiliary_loss_clip": 0.01226694, + "auxiliary_loss_mlp": 0.01042615, + "balance_loss_clip": 1.06777835, + "balance_loss_mlp": 1.02960944, + "epoch": 0.12385017735826369, + "flos": 20485673470080.0, + "grad_norm": 1.7872014049759475, + "language_loss": 0.87323892, + "learning_rate": 3.908413867771257e-06, + "loss": 0.89593202, + "num_input_tokens_seen": 21764760, + "step": 1030, + "time_per_iteration": 2.7404696941375732 + }, + { + "auxiliary_loss_clip": 0.01230705, + "auxiliary_loss_mlp": 0.01041159, + "balance_loss_clip": 1.06675839, + "balance_loss_mlp": 1.02775383, + "epoch": 0.12397042024890279, + "flos": 17347835116800.0, + "grad_norm": 2.272002819292274, + "language_loss": 0.80795842, + "learning_rate": 3.908180695733311e-06, + "loss": 0.83067703, + "num_input_tokens_seen": 21784250, + "step": 1031, + "time_per_iteration": 2.7321054935455322 + }, + { + "auxiliary_loss_clip": 0.01182185, + "auxiliary_loss_mlp": 0.01047562, + "balance_loss_clip": 1.05620456, + "balance_loss_mlp": 1.03311419, + "epoch": 0.12409066313954187, + "flos": 20412343854720.0, + "grad_norm": 1.7775639017813405, + "language_loss": 0.82824302, + "learning_rate": 3.907947234225871e-06, + "loss": 0.85054052, + "num_input_tokens_seen": 21803260, + "step": 1032, + "time_per_iteration": 2.9211466312408447 + }, + { + "auxiliary_loss_clip": 0.01183867, + "auxiliary_loss_mlp": 0.01035018, + "balance_loss_clip": 1.05523527, + "balance_loss_mlp": 1.02150583, + "epoch": 0.12421090603018096, + "flos": 20736688688640.0, + "grad_norm": 2.0513927288501876, + "language_loss": 0.87507093, + "learning_rate": 3.907713483284352e-06, + "loss": 0.89725977, + "num_input_tokens_seen": 21822735, + "step": 1033, + "time_per_iteration": 2.8888750076293945 + }, + { + "auxiliary_loss_clip": 0.01160609, + "auxiliary_loss_mlp": 0.01043779, + "balance_loss_clip": 1.05111814, + "balance_loss_mlp": 1.02996278, + "epoch": 0.12433114892082006, + "flos": 24498834353280.0, + "grad_norm": 2.3049561814512556, + "language_loss": 0.97526979, + "learning_rate": 3.907479442944216e-06, + "loss": 0.99731368, + "num_input_tokens_seen": 21841140, + "step": 1034, + "time_per_iteration": 4.725017070770264 + }, + { + "auxiliary_loss_clip": 0.0123193, + "auxiliary_loss_mlp": 0.01043686, + "balance_loss_clip": 1.06645083, + "balance_loss_mlp": 1.02993524, + "epoch": 0.12445139181145914, + "flos": 19682315838720.0, + "grad_norm": 2.6006760735526098, + "language_loss": 0.92751694, + "learning_rate": 3.907245113240963e-06, + "loss": 0.95027304, + "num_input_tokens_seen": 21859260, + "step": 1035, + "time_per_iteration": 4.593508720397949 + }, + { + "auxiliary_loss_clip": 0.01208851, + "auxiliary_loss_mlp": 0.01038631, + "balance_loss_clip": 1.05624568, + "balance_loss_mlp": 1.02550614, + "epoch": 0.12457163470209824, + "flos": 46423087522560.0, + "grad_norm": 2.1451362846526396, + "language_loss": 0.73460484, + "learning_rate": 3.907010494210144e-06, + "loss": 0.75707972, + "num_input_tokens_seen": 21881920, + "step": 1036, + "time_per_iteration": 3.1018824577331543 + }, + { + "auxiliary_loss_clip": 0.01235929, + "auxiliary_loss_mlp": 0.01041824, + "balance_loss_clip": 1.06582856, + "balance_loss_mlp": 1.0279839, + "epoch": 0.12469187759273732, + "flos": 20376289578240.0, + "grad_norm": 2.2226320330055387, + "language_loss": 0.91793108, + "learning_rate": 3.9067755858873495e-06, + "loss": 0.94070864, + "num_input_tokens_seen": 21898720, + "step": 1037, + "time_per_iteration": 2.9940943717956543 + }, + { + "auxiliary_loss_clip": 0.01117925, + "auxiliary_loss_mlp": 0.01009672, + "balance_loss_clip": 1.03434062, + "balance_loss_mlp": 1.0031395, + "epoch": 0.12481212048337642, + "flos": 69224641447680.0, + "grad_norm": 0.8592278330271347, + "language_loss": 0.62741423, + "learning_rate": 3.906540388308214e-06, + "loss": 0.64869022, + "num_input_tokens_seen": 21958305, + "step": 1038, + "time_per_iteration": 3.384443759918213 + }, + { + "auxiliary_loss_clip": 0.01187874, + "auxiliary_loss_mlp": 0.01042607, + "balance_loss_clip": 1.05886567, + "balance_loss_mlp": 1.02949417, + "epoch": 0.12493236337401552, + "flos": 18223696350720.0, + "grad_norm": 1.7294255194326609, + "language_loss": 0.81479812, + "learning_rate": 3.906304901508417e-06, + "loss": 0.83710289, + "num_input_tokens_seen": 21977205, + "step": 1039, + "time_per_iteration": 2.979522466659546 + }, + { + "auxiliary_loss_clip": 0.01235364, + "auxiliary_loss_mlp": 0.01040514, + "balance_loss_clip": 1.06839037, + "balance_loss_mlp": 1.02818167, + "epoch": 0.12505260626465461, + "flos": 30044375303040.0, + "grad_norm": 3.0666773542836685, + "language_loss": 0.75453907, + "learning_rate": 3.9060691255236835e-06, + "loss": 0.77729785, + "num_input_tokens_seen": 21997770, + "step": 1040, + "time_per_iteration": 2.74432110786438 + }, + { + "auxiliary_loss_clip": 0.01228468, + "auxiliary_loss_mlp": 0.01044987, + "balance_loss_clip": 1.06315923, + "balance_loss_mlp": 1.03230333, + "epoch": 0.1251728491552937, + "flos": 24433980347520.0, + "grad_norm": 1.9242798865424617, + "language_loss": 0.80502898, + "learning_rate": 3.905833060389778e-06, + "loss": 0.8277635, + "num_input_tokens_seen": 22021890, + "step": 1041, + "time_per_iteration": 3.0039353370666504 + }, + { + "auxiliary_loss_clip": 0.01246787, + "auxiliary_loss_mlp": 0.00889464, + "balance_loss_clip": 1.06783104, + "balance_loss_mlp": 1.00031996, + "epoch": 0.12529309204593278, + "flos": 27119809952640.0, + "grad_norm": 2.978152581429204, + "language_loss": 0.78224695, + "learning_rate": 3.905596706142513e-06, + "loss": 0.80360949, + "num_input_tokens_seen": 22043300, + "step": 1042, + "time_per_iteration": 2.8157544136047363 + }, + { + "auxiliary_loss_clip": 0.01207508, + "auxiliary_loss_mlp": 0.01040747, + "balance_loss_clip": 1.05797088, + "balance_loss_mlp": 1.02710938, + "epoch": 0.12541333493657186, + "flos": 30774151923840.0, + "grad_norm": 2.206537112941419, + "language_loss": 0.86025017, + "learning_rate": 3.9053600628177435e-06, + "loss": 0.88273275, + "num_input_tokens_seen": 22062910, + "step": 1043, + "time_per_iteration": 2.945664882659912 + }, + { + "auxiliary_loss_clip": 0.01240865, + "auxiliary_loss_mlp": 0.01038252, + "balance_loss_clip": 1.06419539, + "balance_loss_mlp": 1.02541304, + "epoch": 0.12553357782721097, + "flos": 23659566099840.0, + "grad_norm": 2.0315415620280923, + "language_loss": 0.84736693, + "learning_rate": 3.905123130451367e-06, + "loss": 0.87015814, + "num_input_tokens_seen": 22084010, + "step": 1044, + "time_per_iteration": 2.739647626876831 + }, + { + "auxiliary_loss_clip": 0.01243961, + "auxiliary_loss_mlp": 0.01042874, + "balance_loss_clip": 1.06745243, + "balance_loss_mlp": 1.02986264, + "epoch": 0.12565382071785006, + "flos": 24863758577280.0, + "grad_norm": 2.3830631788243957, + "language_loss": 0.79421502, + "learning_rate": 3.904885909079326e-06, + "loss": 0.81708336, + "num_input_tokens_seen": 22102795, + "step": 1045, + "time_per_iteration": 2.677277088165283 + }, + { + "auxiliary_loss_clip": 0.01229269, + "auxiliary_loss_mlp": 0.0104653, + "balance_loss_clip": 1.06060946, + "balance_loss_mlp": 1.03273153, + "epoch": 0.12577406360848914, + "flos": 21360780518400.0, + "grad_norm": 2.786542879286287, + "language_loss": 0.78098166, + "learning_rate": 3.904648398737607e-06, + "loss": 0.80373961, + "num_input_tokens_seen": 22121360, + "step": 1046, + "time_per_iteration": 2.7782435417175293 + }, + { + "auxiliary_loss_clip": 0.01239253, + "auxiliary_loss_mlp": 0.01039563, + "balance_loss_clip": 1.06400275, + "balance_loss_mlp": 1.02650404, + "epoch": 0.12589430649912825, + "flos": 36138056774400.0, + "grad_norm": 2.504897331348878, + "language_loss": 0.78076828, + "learning_rate": 3.9044105994622406e-06, + "loss": 0.8035565, + "num_input_tokens_seen": 22142505, + "step": 1047, + "time_per_iteration": 2.854555130004883 + }, + { + "auxiliary_loss_clip": 0.01227678, + "auxiliary_loss_mlp": 0.00889429, + "balance_loss_clip": 1.06405866, + "balance_loss_mlp": 1.00031877, + "epoch": 0.12601454938976733, + "flos": 25337671643520.0, + "grad_norm": 2.11925393797114, + "language_loss": 0.81658947, + "learning_rate": 3.9041725112893005e-06, + "loss": 0.83776063, + "num_input_tokens_seen": 22163730, + "step": 1048, + "time_per_iteration": 2.8341710567474365 + }, + { + "auxiliary_loss_clip": 0.01206305, + "auxiliary_loss_mlp": 0.010517, + "balance_loss_clip": 1.06163502, + "balance_loss_mlp": 1.03906941, + "epoch": 0.12613479228040642, + "flos": 15560094286080.0, + "grad_norm": 1.9124947849368346, + "language_loss": 0.74830335, + "learning_rate": 3.903934134254904e-06, + "loss": 0.77088344, + "num_input_tokens_seen": 22181520, + "step": 1049, + "time_per_iteration": 2.800267457962036 + }, + { + "auxiliary_loss_clip": 0.01239492, + "auxiliary_loss_mlp": 0.01048466, + "balance_loss_clip": 1.06471217, + "balance_loss_mlp": 1.03501296, + "epoch": 0.1262550351710455, + "flos": 21470595373440.0, + "grad_norm": 2.1111085484110657, + "language_loss": 0.84985846, + "learning_rate": 3.903695468395213e-06, + "loss": 0.872738, + "num_input_tokens_seen": 22199390, + "step": 1050, + "time_per_iteration": 2.7313926219940186 + }, + { + "auxiliary_loss_clip": 0.01226293, + "auxiliary_loss_mlp": 0.01052096, + "balance_loss_clip": 1.05969143, + "balance_loss_mlp": 1.03873825, + "epoch": 0.1263752780616846, + "flos": 31576719456000.0, + "grad_norm": 3.030882712385015, + "language_loss": 0.55942476, + "learning_rate": 3.903456513746434e-06, + "loss": 0.58220863, + "num_input_tokens_seen": 22220365, + "step": 1051, + "time_per_iteration": 2.803844690322876 + }, + { + "auxiliary_loss_clip": 0.01244622, + "auxiliary_loss_mlp": 0.01041703, + "balance_loss_clip": 1.067173, + "balance_loss_mlp": 1.02866197, + "epoch": 0.1264955209523237, + "flos": 28768217927040.0, + "grad_norm": 1.819173332629291, + "language_loss": 0.87487906, + "learning_rate": 3.903217270344815e-06, + "loss": 0.89774233, + "num_input_tokens_seen": 22240615, + "step": 1052, + "time_per_iteration": 2.7385432720184326 + }, + { + "auxiliary_loss_clip": 0.01206501, + "auxiliary_loss_mlp": 0.01040032, + "balance_loss_clip": 1.06010365, + "balance_loss_mlp": 1.02743208, + "epoch": 0.12661576384296278, + "flos": 29241125412480.0, + "grad_norm": 2.1407250385442, + "language_loss": 0.82166094, + "learning_rate": 3.902977738226648e-06, + "loss": 0.84412628, + "num_input_tokens_seen": 22261350, + "step": 1053, + "time_per_iteration": 2.868030548095703 + }, + { + "auxiliary_loss_clip": 0.01237507, + "auxiliary_loss_mlp": 0.01044132, + "balance_loss_clip": 1.0685643, + "balance_loss_mlp": 1.02976775, + "epoch": 0.12673600673360189, + "flos": 20850346298880.0, + "grad_norm": 1.8955648906880753, + "language_loss": 0.91451353, + "learning_rate": 3.902737917428273e-06, + "loss": 0.93732989, + "num_input_tokens_seen": 22279515, + "step": 1054, + "time_per_iteration": 2.698744773864746 + }, + { + "auxiliary_loss_clip": 0.01246019, + "auxiliary_loss_mlp": 0.01043843, + "balance_loss_clip": 1.06747115, + "balance_loss_mlp": 1.03030074, + "epoch": 0.12685624962424097, + "flos": 25263695583360.0, + "grad_norm": 1.7463275949871746, + "language_loss": 0.84364921, + "learning_rate": 3.902497807986068e-06, + "loss": 0.86654782, + "num_input_tokens_seen": 22299535, + "step": 1055, + "time_per_iteration": 2.7444581985473633 + }, + { + "auxiliary_loss_clip": 0.01215749, + "auxiliary_loss_mlp": 0.01046442, + "balance_loss_clip": 1.05952287, + "balance_loss_mlp": 1.03265572, + "epoch": 0.12697649251488005, + "flos": 27527109246720.0, + "grad_norm": 1.6156615535497643, + "language_loss": 0.84136713, + "learning_rate": 3.902257409936458e-06, + "loss": 0.86398906, + "num_input_tokens_seen": 22320300, + "step": 1056, + "time_per_iteration": 2.840363025665283 + }, + { + "auxiliary_loss_clip": 0.0122249, + "auxiliary_loss_mlp": 0.0104175, + "balance_loss_clip": 1.06604624, + "balance_loss_mlp": 1.02811241, + "epoch": 0.12709673540551916, + "flos": 21251863503360.0, + "grad_norm": 1.9082040741674913, + "language_loss": 0.8417055, + "learning_rate": 3.902016723315912e-06, + "loss": 0.86434793, + "num_input_tokens_seen": 22338240, + "step": 1057, + "time_per_iteration": 2.8012595176696777 + }, + { + "auxiliary_loss_clip": 0.01228805, + "auxiliary_loss_mlp": 0.01037245, + "balance_loss_clip": 1.0612967, + "balance_loss_mlp": 1.02424538, + "epoch": 0.12721697829615825, + "flos": 25337707557120.0, + "grad_norm": 5.2872173589117555, + "language_loss": 0.69169015, + "learning_rate": 3.901775748160941e-06, + "loss": 0.71435064, + "num_input_tokens_seen": 22357420, + "step": 1058, + "time_per_iteration": 2.731339693069458 + }, + { + "auxiliary_loss_clip": 0.01119922, + "auxiliary_loss_mlp": 0.01010527, + "balance_loss_clip": 1.0339725, + "balance_loss_mlp": 1.00487602, + "epoch": 0.12733722118679733, + "flos": 61943287754880.0, + "grad_norm": 0.8274023043719753, + "language_loss": 0.60879487, + "learning_rate": 3.901534484508101e-06, + "loss": 0.63009936, + "num_input_tokens_seen": 22420095, + "step": 1059, + "time_per_iteration": 3.3348865509033203 + }, + { + "auxiliary_loss_clip": 0.0121068, + "auxiliary_loss_mlp": 0.01044887, + "balance_loss_clip": 1.05862689, + "balance_loss_mlp": 1.03120148, + "epoch": 0.1274574640774364, + "flos": 26976742081920.0, + "grad_norm": 2.129192779807206, + "language_loss": 0.74731487, + "learning_rate": 3.901292932393991e-06, + "loss": 0.76987052, + "num_input_tokens_seen": 22438975, + "step": 1060, + "time_per_iteration": 3.768359899520874 + }, + { + "auxiliary_loss_clip": 0.01245845, + "auxiliary_loss_mlp": 0.01047381, + "balance_loss_clip": 1.06696117, + "balance_loss_mlp": 1.03439355, + "epoch": 0.12757770696807552, + "flos": 22236318529920.0, + "grad_norm": 2.1752171413029058, + "language_loss": 0.85488766, + "learning_rate": 3.9010510918552555e-06, + "loss": 0.8778199, + "num_input_tokens_seen": 22458050, + "step": 1061, + "time_per_iteration": 3.859382152557373 + }, + { + "auxiliary_loss_clip": 0.01216188, + "auxiliary_loss_mlp": 0.01054881, + "balance_loss_clip": 1.05801415, + "balance_loss_mlp": 1.04074264, + "epoch": 0.1276979498587146, + "flos": 28547905858560.0, + "grad_norm": 2.396572276463483, + "language_loss": 0.74813056, + "learning_rate": 3.900808962928581e-06, + "loss": 0.77084124, + "num_input_tokens_seen": 22475665, + "step": 1062, + "time_per_iteration": 3.7535271644592285 + }, + { + "auxiliary_loss_clip": 0.01241335, + "auxiliary_loss_mlp": 0.01043294, + "balance_loss_clip": 1.06641853, + "balance_loss_mlp": 1.03097987, + "epoch": 0.1278181927493537, + "flos": 17420338719360.0, + "grad_norm": 2.4790412977866763, + "language_loss": 0.89660037, + "learning_rate": 3.900566545650698e-06, + "loss": 0.91944665, + "num_input_tokens_seen": 22493335, + "step": 1063, + "time_per_iteration": 2.7397620677948 + }, + { + "auxiliary_loss_clip": 0.01234194, + "auxiliary_loss_mlp": 0.01040735, + "balance_loss_clip": 1.06559229, + "balance_loss_mlp": 1.02709723, + "epoch": 0.1279384356399928, + "flos": 21138636856320.0, + "grad_norm": 2.331824930037059, + "language_loss": 0.81814361, + "learning_rate": 3.900323840058381e-06, + "loss": 0.84089291, + "num_input_tokens_seen": 22511045, + "step": 1064, + "time_per_iteration": 2.6846773624420166 + }, + { + "auxiliary_loss_clip": 0.0123033, + "auxiliary_loss_mlp": 0.0104082, + "balance_loss_clip": 1.06170964, + "balance_loss_mlp": 1.02766526, + "epoch": 0.12805867853063188, + "flos": 26576733248640.0, + "grad_norm": 1.8398973449664924, + "language_loss": 0.8188566, + "learning_rate": 3.900080846188449e-06, + "loss": 0.84156799, + "num_input_tokens_seen": 22529635, + "step": 1065, + "time_per_iteration": 2.7630205154418945 + }, + { + "auxiliary_loss_clip": 0.01242538, + "auxiliary_loss_mlp": 0.01043847, + "balance_loss_clip": 1.06483912, + "balance_loss_mlp": 1.03138328, + "epoch": 0.12817892142127096, + "flos": 16436206915200.0, + "grad_norm": 1.8171405009066683, + "language_loss": 0.81547296, + "learning_rate": 3.8998375640777625e-06, + "loss": 0.83833683, + "num_input_tokens_seen": 22547505, + "step": 1066, + "time_per_iteration": 2.6927173137664795 + }, + { + "auxiliary_loss_clip": 0.01136357, + "auxiliary_loss_mlp": 0.01010083, + "balance_loss_clip": 1.05192804, + "balance_loss_mlp": 1.00426531, + "epoch": 0.12829916431191005, + "flos": 60757049099520.0, + "grad_norm": 0.705855824696854, + "language_loss": 0.52622652, + "learning_rate": 3.899593993763229e-06, + "loss": 0.54769093, + "num_input_tokens_seen": 22608465, + "step": 1067, + "time_per_iteration": 3.3554110527038574 + }, + { + "auxiliary_loss_clip": 0.01200032, + "auxiliary_loss_mlp": 0.01037537, + "balance_loss_clip": 1.05803633, + "balance_loss_mlp": 1.02381039, + "epoch": 0.12841940720254916, + "flos": 29786895636480.0, + "grad_norm": 2.0138658013034103, + "language_loss": 0.81390566, + "learning_rate": 3.899350135281796e-06, + "loss": 0.83628136, + "num_input_tokens_seen": 22629465, + "step": 1068, + "time_per_iteration": 2.9167160987854004 + }, + { + "auxiliary_loss_clip": 0.01209329, + "auxiliary_loss_mlp": 0.01045593, + "balance_loss_clip": 1.05958056, + "balance_loss_mlp": 1.03265309, + "epoch": 0.12853965009318824, + "flos": 25951851319680.0, + "grad_norm": 1.938970958309172, + "language_loss": 0.79804194, + "learning_rate": 3.8991059886704585e-06, + "loss": 0.82059121, + "num_input_tokens_seen": 22648970, + "step": 1069, + "time_per_iteration": 2.7943108081817627 + }, + { + "auxiliary_loss_clip": 0.01203512, + "auxiliary_loss_mlp": 0.01055599, + "balance_loss_clip": 1.06051147, + "balance_loss_mlp": 1.0417769, + "epoch": 0.12865989298382732, + "flos": 30846871008000.0, + "grad_norm": 2.299319311202515, + "language_loss": 0.83310968, + "learning_rate": 3.898861553966252e-06, + "loss": 0.85570073, + "num_input_tokens_seen": 22668620, + "step": 1070, + "time_per_iteration": 2.8836538791656494 + }, + { + "auxiliary_loss_clip": 0.01176479, + "auxiliary_loss_mlp": 0.01051145, + "balance_loss_clip": 1.05306911, + "balance_loss_mlp": 1.03762102, + "epoch": 0.12878013587446643, + "flos": 25885776251520.0, + "grad_norm": 1.7358094552867576, + "language_loss": 0.88171428, + "learning_rate": 3.898616831206257e-06, + "loss": 0.90399051, + "num_input_tokens_seen": 22689045, + "step": 1071, + "time_per_iteration": 2.9075469970703125 + }, + { + "auxiliary_loss_clip": 0.0119895, + "auxiliary_loss_mlp": 0.01044372, + "balance_loss_clip": 1.05320263, + "balance_loss_mlp": 1.03045988, + "epoch": 0.12890037876510552, + "flos": 23333138277120.0, + "grad_norm": 3.3699031024662527, + "language_loss": 0.76643878, + "learning_rate": 3.8983718204276e-06, + "loss": 0.788872, + "num_input_tokens_seen": 22711265, + "step": 1072, + "time_per_iteration": 2.8095333576202393 + }, + { + "auxiliary_loss_clip": 0.01219551, + "auxiliary_loss_mlp": 0.0104882, + "balance_loss_clip": 1.06232834, + "balance_loss_mlp": 1.03514624, + "epoch": 0.1290206216557446, + "flos": 23587242065280.0, + "grad_norm": 2.1152142635185816, + "language_loss": 0.82525086, + "learning_rate": 3.898126521667446e-06, + "loss": 0.8479346, + "num_input_tokens_seen": 22731420, + "step": 1073, + "time_per_iteration": 2.776231527328491 + }, + { + "auxiliary_loss_clip": 0.01229318, + "auxiliary_loss_mlp": 0.01042098, + "balance_loss_clip": 1.06173897, + "balance_loss_mlp": 1.02863932, + "epoch": 0.12914086454638368, + "flos": 24170610850560.0, + "grad_norm": 1.6949800312538401, + "language_loss": 0.8362298, + "learning_rate": 3.897880934963007e-06, + "loss": 0.85894394, + "num_input_tokens_seen": 22750970, + "step": 1074, + "time_per_iteration": 2.7481374740600586 + }, + { + "auxiliary_loss_clip": 0.01213562, + "auxiliary_loss_mlp": 0.01045803, + "balance_loss_clip": 1.05855548, + "balance_loss_mlp": 1.03283334, + "epoch": 0.1292611074370228, + "flos": 20267157081600.0, + "grad_norm": 2.0589039640891347, + "language_loss": 0.78648162, + "learning_rate": 3.89763506035154e-06, + "loss": 0.80907524, + "num_input_tokens_seen": 22768820, + "step": 1075, + "time_per_iteration": 2.7491354942321777 + }, + { + "auxiliary_loss_clip": 0.01220501, + "auxiliary_loss_mlp": 0.01041703, + "balance_loss_clip": 1.06078053, + "balance_loss_mlp": 1.02918589, + "epoch": 0.12938135032766188, + "flos": 27377684668800.0, + "grad_norm": 2.0590518823446833, + "language_loss": 0.81174386, + "learning_rate": 3.897388897870343e-06, + "loss": 0.83436584, + "num_input_tokens_seen": 22789460, + "step": 1076, + "time_per_iteration": 2.8112354278564453 + }, + { + "auxiliary_loss_clip": 0.01231972, + "auxiliary_loss_mlp": 0.01041903, + "balance_loss_clip": 1.06168866, + "balance_loss_mlp": 1.02766955, + "epoch": 0.12950159321830096, + "flos": 29277107861760.0, + "grad_norm": 1.8888278141438877, + "language_loss": 0.74993491, + "learning_rate": 3.89714244755676e-06, + "loss": 0.77267361, + "num_input_tokens_seen": 22810820, + "step": 1077, + "time_per_iteration": 2.851081609725952 + }, + { + "auxiliary_loss_clip": 0.01181305, + "auxiliary_loss_mlp": 0.01041239, + "balance_loss_clip": 1.04937387, + "balance_loss_mlp": 1.02782774, + "epoch": 0.12962183610894007, + "flos": 24534888629760.0, + "grad_norm": 2.1560342379196133, + "language_loss": 0.86188209, + "learning_rate": 3.896895709448175e-06, + "loss": 0.88410747, + "num_input_tokens_seen": 22830570, + "step": 1078, + "time_per_iteration": 2.927457094192505 + }, + { + "auxiliary_loss_clip": 0.0117699, + "auxiliary_loss_mlp": 0.01043482, + "balance_loss_clip": 1.05179572, + "balance_loss_mlp": 1.03058338, + "epoch": 0.12974207899957915, + "flos": 11215944552960.0, + "grad_norm": 2.5807118716604736, + "language_loss": 0.76800144, + "learning_rate": 3.896648683582019e-06, + "loss": 0.79020607, + "num_input_tokens_seen": 22845905, + "step": 1079, + "time_per_iteration": 2.9823193550109863 + }, + { + "auxiliary_loss_clip": 0.01197707, + "auxiliary_loss_mlp": 0.01048141, + "balance_loss_clip": 1.0590359, + "balance_loss_mlp": 1.03425264, + "epoch": 0.12986232189021824, + "flos": 24717889445760.0, + "grad_norm": 1.9445806690642102, + "language_loss": 0.80799073, + "learning_rate": 3.896401369995766e-06, + "loss": 0.83044916, + "num_input_tokens_seen": 22865710, + "step": 1080, + "time_per_iteration": 2.8980188369750977 + }, + { + "auxiliary_loss_clip": 0.01242032, + "auxiliary_loss_mlp": 0.01039249, + "balance_loss_clip": 1.06546867, + "balance_loss_mlp": 1.02596271, + "epoch": 0.12998256478085732, + "flos": 23915357827200.0, + "grad_norm": 1.8370415627538725, + "language_loss": 0.79782784, + "learning_rate": 3.896153768726932e-06, + "loss": 0.82064068, + "num_input_tokens_seen": 22886020, + "step": 1081, + "time_per_iteration": 2.703900098800659 + }, + { + "auxiliary_loss_clip": 0.01229204, + "auxiliary_loss_mlp": 0.01041654, + "balance_loss_clip": 1.06506228, + "balance_loss_mlp": 1.02858293, + "epoch": 0.13010280767149643, + "flos": 18624207974400.0, + "grad_norm": 2.7889800601874746, + "language_loss": 0.8805809, + "learning_rate": 3.8959058798130806e-06, + "loss": 0.9032895, + "num_input_tokens_seen": 22903995, + "step": 1082, + "time_per_iteration": 2.815101146697998 + }, + { + "auxiliary_loss_clip": 0.01220892, + "auxiliary_loss_mlp": 0.00889933, + "balance_loss_clip": 1.06323266, + "balance_loss_mlp": 1.00034153, + "epoch": 0.1302230505621355, + "flos": 22783992174720.0, + "grad_norm": 1.5777651004667896, + "language_loss": 0.74877352, + "learning_rate": 3.895657703291814e-06, + "loss": 0.76988173, + "num_input_tokens_seen": 22924100, + "step": 1083, + "time_per_iteration": 2.771735191345215 + }, + { + "auxiliary_loss_clip": 0.01226746, + "auxiliary_loss_mlp": 0.01046817, + "balance_loss_clip": 1.06001389, + "balance_loss_mlp": 1.03317928, + "epoch": 0.1303432934527746, + "flos": 21323612920320.0, + "grad_norm": 3.15824164544387, + "language_loss": 0.79583669, + "learning_rate": 3.895409239200781e-06, + "loss": 0.81857228, + "num_input_tokens_seen": 22939985, + "step": 1084, + "time_per_iteration": 2.8390798568725586 + }, + { + "auxiliary_loss_clip": 0.01220773, + "auxiliary_loss_mlp": 0.0103784, + "balance_loss_clip": 1.06038964, + "balance_loss_mlp": 1.02335036, + "epoch": 0.1304635363434137, + "flos": 20922490765440.0, + "grad_norm": 2.181485456179922, + "language_loss": 0.91595763, + "learning_rate": 3.895160487577673e-06, + "loss": 0.9385438, + "num_input_tokens_seen": 22957555, + "step": 1085, + "time_per_iteration": 2.655203104019165 + }, + { + "auxiliary_loss_clip": 0.01128018, + "auxiliary_loss_mlp": 0.01012163, + "balance_loss_clip": 1.03433609, + "balance_loss_mlp": 1.00591648, + "epoch": 0.1305837792340528, + "flos": 63245659080960.0, + "grad_norm": 0.7808157172259936, + "language_loss": 0.60892963, + "learning_rate": 3.894911448460226e-06, + "loss": 0.63033146, + "num_input_tokens_seen": 23016870, + "step": 1086, + "time_per_iteration": 5.893057107925415 + }, + { + "auxiliary_loss_clip": 0.01159122, + "auxiliary_loss_mlp": 0.01047256, + "balance_loss_clip": 1.05170226, + "balance_loss_mlp": 1.03361273, + "epoch": 0.13070402212469187, + "flos": 26428852955520.0, + "grad_norm": 2.131392986053182, + "language_loss": 0.72640342, + "learning_rate": 3.8946621218862195e-06, + "loss": 0.74846721, + "num_input_tokens_seen": 23037870, + "step": 1087, + "time_per_iteration": 4.024510383605957 + }, + { + "auxiliary_loss_clip": 0.01204108, + "auxiliary_loss_mlp": 0.01038951, + "balance_loss_clip": 1.05973983, + "balance_loss_mlp": 1.02517653, + "epoch": 0.13082426501533098, + "flos": 27673409341440.0, + "grad_norm": 2.279846814730726, + "language_loss": 0.88951689, + "learning_rate": 3.894412507893475e-06, + "loss": 0.91194749, + "num_input_tokens_seen": 23058150, + "step": 1088, + "time_per_iteration": 3.0301826000213623 + }, + { + "auxiliary_loss_clip": 0.01205753, + "auxiliary_loss_mlp": 0.01042688, + "balance_loss_clip": 1.05721033, + "balance_loss_mlp": 1.02906787, + "epoch": 0.13094450790597006, + "flos": 24826770547200.0, + "grad_norm": 3.26465468881156, + "language_loss": 0.72143656, + "learning_rate": 3.894162606519859e-06, + "loss": 0.74392098, + "num_input_tokens_seen": 23077100, + "step": 1089, + "time_per_iteration": 2.874648332595825 + }, + { + "auxiliary_loss_clip": 0.011982, + "auxiliary_loss_mlp": 0.01041164, + "balance_loss_clip": 1.05876088, + "balance_loss_mlp": 1.02875423, + "epoch": 0.13106475079660915, + "flos": 19062605468160.0, + "grad_norm": 4.038948505635244, + "language_loss": 0.77264035, + "learning_rate": 3.893912417803282e-06, + "loss": 0.79503393, + "num_input_tokens_seen": 23096815, + "step": 1090, + "time_per_iteration": 2.7959444522857666 + }, + { + "auxiliary_loss_clip": 0.01192452, + "auxiliary_loss_mlp": 0.01049436, + "balance_loss_clip": 1.0522964, + "balance_loss_mlp": 1.03477883, + "epoch": 0.13118499368724823, + "flos": 28913189218560.0, + "grad_norm": 1.8231907460714256, + "language_loss": 0.76957178, + "learning_rate": 3.8936619417816975e-06, + "loss": 0.79199064, + "num_input_tokens_seen": 23117145, + "step": 1091, + "time_per_iteration": 2.8849809169769287 + }, + { + "auxiliary_loss_clip": 0.01214691, + "auxiliary_loss_mlp": 0.01043392, + "balance_loss_clip": 1.06437874, + "balance_loss_mlp": 1.02965927, + "epoch": 0.13130523657788734, + "flos": 14283398206080.0, + "grad_norm": 1.8068257194104722, + "language_loss": 0.71606433, + "learning_rate": 3.8934111784931015e-06, + "loss": 0.7386452, + "num_input_tokens_seen": 23134595, + "step": 1092, + "time_per_iteration": 2.771524429321289 + }, + { + "auxiliary_loss_clip": 0.01121199, + "auxiliary_loss_mlp": 0.01006684, + "balance_loss_clip": 1.03277659, + "balance_loss_mlp": 1.00038946, + "epoch": 0.13142547946852642, + "flos": 70174155519360.0, + "grad_norm": 0.9098525601693481, + "language_loss": 0.59049428, + "learning_rate": 3.893160127975535e-06, + "loss": 0.61177307, + "num_input_tokens_seen": 23195285, + "step": 1093, + "time_per_iteration": 3.523820400238037 + }, + { + "auxiliary_loss_clip": 0.01202811, + "auxiliary_loss_mlp": 0.01047288, + "balance_loss_clip": 1.05802453, + "balance_loss_mlp": 1.03433037, + "epoch": 0.1315457223591655, + "flos": 45805998844800.0, + "grad_norm": 2.6109188876632565, + "language_loss": 0.81095946, + "learning_rate": 3.8929087902670826e-06, + "loss": 0.83346045, + "num_input_tokens_seen": 23216915, + "step": 1094, + "time_per_iteration": 3.0785489082336426 + }, + { + "auxiliary_loss_clip": 0.0113195, + "auxiliary_loss_mlp": 0.01008394, + "balance_loss_clip": 1.03506207, + "balance_loss_mlp": 1.00190914, + "epoch": 0.13166596524980462, + "flos": 62881165820160.0, + "grad_norm": 0.9402893131319495, + "language_loss": 0.60694122, + "learning_rate": 3.8926571654058715e-06, + "loss": 0.62834466, + "num_input_tokens_seen": 23273560, + "step": 1095, + "time_per_iteration": 3.2739102840423584 + }, + { + "auxiliary_loss_clip": 0.01205833, + "auxiliary_loss_mlp": 0.01053379, + "balance_loss_clip": 1.059587, + "balance_loss_mlp": 1.0401473, + "epoch": 0.1317862081404437, + "flos": 23586523793280.0, + "grad_norm": 2.596130024596334, + "language_loss": 0.76155454, + "learning_rate": 3.892405253430074e-06, + "loss": 0.78414667, + "num_input_tokens_seen": 23291080, + "step": 1096, + "time_per_iteration": 2.8936941623687744 + }, + { + "auxiliary_loss_clip": 0.01229076, + "auxiliary_loss_mlp": 0.00890229, + "balance_loss_clip": 1.06638241, + "balance_loss_mlp": 1.00030637, + "epoch": 0.13190645103108278, + "flos": 20260764460800.0, + "grad_norm": 2.8110272043065105, + "language_loss": 0.82310832, + "learning_rate": 3.892153054377904e-06, + "loss": 0.84430134, + "num_input_tokens_seen": 23308485, + "step": 1097, + "time_per_iteration": 2.8216824531555176 + }, + { + "auxiliary_loss_clip": 0.01115735, + "auxiliary_loss_mlp": 0.01007043, + "balance_loss_clip": 1.05685842, + "balance_loss_mlp": 1.0005101, + "epoch": 0.13202669392172187, + "flos": 53455440136320.0, + "grad_norm": 0.9359456928574615, + "language_loss": 0.59422225, + "learning_rate": 3.891900568287619e-06, + "loss": 0.61545002, + "num_input_tokens_seen": 23360870, + "step": 1098, + "time_per_iteration": 3.303612232208252 + }, + { + "auxiliary_loss_clip": 0.01218767, + "auxiliary_loss_mlp": 0.01049303, + "balance_loss_clip": 1.06160593, + "balance_loss_mlp": 1.03463995, + "epoch": 0.13214693681236098, + "flos": 15851293845120.0, + "grad_norm": 2.6547935765133146, + "language_loss": 0.72143775, + "learning_rate": 3.891647795197523e-06, + "loss": 0.74411845, + "num_input_tokens_seen": 23376910, + "step": 1099, + "time_per_iteration": 2.776374340057373 + }, + { + "auxiliary_loss_clip": 0.01217317, + "auxiliary_loss_mlp": 0.01048279, + "balance_loss_clip": 1.05898476, + "balance_loss_mlp": 1.03427243, + "epoch": 0.13226717970300006, + "flos": 19353840940800.0, + "grad_norm": 2.2503409542567594, + "language_loss": 0.68898034, + "learning_rate": 3.8913947351459605e-06, + "loss": 0.7116363, + "num_input_tokens_seen": 23394450, + "step": 1100, + "time_per_iteration": 2.8291022777557373 + }, + { + "auxiliary_loss_clip": 0.01245138, + "auxiliary_loss_mlp": 0.01046672, + "balance_loss_clip": 1.06929374, + "balance_loss_mlp": 1.03306437, + "epoch": 0.13238742259363914, + "flos": 20698084546560.0, + "grad_norm": 2.1027846266810997, + "language_loss": 0.67750895, + "learning_rate": 3.89114138817132e-06, + "loss": 0.70042706, + "num_input_tokens_seen": 23411115, + "step": 1101, + "time_per_iteration": 2.7112019062042236 + }, + { + "auxiliary_loss_clip": 0.01233024, + "auxiliary_loss_mlp": 0.01044369, + "balance_loss_clip": 1.067716, + "balance_loss_mlp": 1.03114915, + "epoch": 0.13250766548427825, + "flos": 21032449274880.0, + "grad_norm": 1.9901928523061823, + "language_loss": 0.84435105, + "learning_rate": 3.890887754312035e-06, + "loss": 0.86712497, + "num_input_tokens_seen": 23429360, + "step": 1102, + "time_per_iteration": 2.738863945007324 + }, + { + "auxiliary_loss_clip": 0.01217629, + "auxiliary_loss_mlp": 0.01047667, + "balance_loss_clip": 1.061239, + "balance_loss_mlp": 1.03436971, + "epoch": 0.13262790837491734, + "flos": 22637871648000.0, + "grad_norm": 1.8904144600014525, + "language_loss": 0.882698, + "learning_rate": 3.890633833606581e-06, + "loss": 0.90535092, + "num_input_tokens_seen": 23449050, + "step": 1103, + "time_per_iteration": 2.838380813598633 + }, + { + "auxiliary_loss_clip": 0.01229809, + "auxiliary_loss_mlp": 0.01048897, + "balance_loss_clip": 1.06733131, + "balance_loss_mlp": 1.03550386, + "epoch": 0.13274815126555642, + "flos": 19683141851520.0, + "grad_norm": 1.9809760509119574, + "language_loss": 0.69734359, + "learning_rate": 3.890379626093477e-06, + "loss": 0.72013062, + "num_input_tokens_seen": 23468800, + "step": 1104, + "time_per_iteration": 2.689885377883911 + }, + { + "auxiliary_loss_clip": 0.01187738, + "auxiliary_loss_mlp": 0.01046138, + "balance_loss_clip": 1.05722356, + "balance_loss_mlp": 1.03244042, + "epoch": 0.1328683941561955, + "flos": 21317687176320.0, + "grad_norm": 2.173366969456979, + "language_loss": 0.92439717, + "learning_rate": 3.890125131811287e-06, + "loss": 0.94673598, + "num_input_tokens_seen": 23486850, + "step": 1105, + "time_per_iteration": 2.854620933532715 + }, + { + "auxiliary_loss_clip": 0.01205271, + "auxiliary_loss_mlp": 0.01046292, + "balance_loss_clip": 1.05628204, + "balance_loss_mlp": 1.03375125, + "epoch": 0.1329886370468346, + "flos": 13699131580800.0, + "grad_norm": 2.0061914249455524, + "language_loss": 0.75842434, + "learning_rate": 3.889870350798618e-06, + "loss": 0.78093994, + "num_input_tokens_seen": 23504195, + "step": 1106, + "time_per_iteration": 2.695797920227051 + }, + { + "auxiliary_loss_clip": 0.01238932, + "auxiliary_loss_mlp": 0.0104468, + "balance_loss_clip": 1.06346333, + "balance_loss_mlp": 1.03160298, + "epoch": 0.1331088799374737, + "flos": 21032413361280.0, + "grad_norm": 1.5597158686371786, + "language_loss": 0.78585768, + "learning_rate": 3.889615283094119e-06, + "loss": 0.80869377, + "num_input_tokens_seen": 23523385, + "step": 1107, + "time_per_iteration": 2.6960458755493164 + }, + { + "auxiliary_loss_clip": 0.01248848, + "auxiliary_loss_mlp": 0.01048161, + "balance_loss_clip": 1.06716895, + "balance_loss_mlp": 1.03339648, + "epoch": 0.13322912282811278, + "flos": 18260432985600.0, + "grad_norm": 2.397255337097181, + "language_loss": 0.84151113, + "learning_rate": 3.889359928736485e-06, + "loss": 0.86448121, + "num_input_tokens_seen": 23541330, + "step": 1108, + "time_per_iteration": 2.634706974029541 + }, + { + "auxiliary_loss_clip": 0.0121884, + "auxiliary_loss_mlp": 0.00888955, + "balance_loss_clip": 1.06528521, + "balance_loss_mlp": 1.00025809, + "epoch": 0.1333493657187519, + "flos": 24460876656000.0, + "grad_norm": 2.2350864594249225, + "language_loss": 0.90909481, + "learning_rate": 3.889104287764451e-06, + "loss": 0.9301728, + "num_input_tokens_seen": 23561705, + "step": 1109, + "time_per_iteration": 2.8669369220733643 + }, + { + "auxiliary_loss_clip": 0.01218644, + "auxiliary_loss_mlp": 0.01050111, + "balance_loss_clip": 1.06432569, + "balance_loss_mlp": 1.03638959, + "epoch": 0.13346960860939097, + "flos": 22158930677760.0, + "grad_norm": 2.1973078033810793, + "language_loss": 0.9073298, + "learning_rate": 3.888848360216798e-06, + "loss": 0.93001735, + "num_input_tokens_seen": 23579350, + "step": 1110, + "time_per_iteration": 2.791289806365967 + }, + { + "auxiliary_loss_clip": 0.01124347, + "auxiliary_loss_mlp": 0.01009546, + "balance_loss_clip": 1.03325534, + "balance_loss_mlp": 1.0033474, + "epoch": 0.13358985150003005, + "flos": 67931212608000.0, + "grad_norm": 0.8008585786975797, + "language_loss": 0.56566936, + "learning_rate": 3.888592146132351e-06, + "loss": 0.5870083, + "num_input_tokens_seen": 23640620, + "step": 1111, + "time_per_iteration": 3.4370524883270264 + }, + { + "auxiliary_loss_clip": 0.01230801, + "auxiliary_loss_mlp": 0.0104317, + "balance_loss_clip": 1.06681013, + "balance_loss_mlp": 1.03015268, + "epoch": 0.13371009439066917, + "flos": 26834284742400.0, + "grad_norm": 1.8309265372515005, + "language_loss": 0.78032553, + "learning_rate": 3.888335645549978e-06, + "loss": 0.8030653, + "num_input_tokens_seen": 23661040, + "step": 1112, + "time_per_iteration": 5.5661301612854 + }, + { + "auxiliary_loss_clip": 0.01240012, + "auxiliary_loss_mlp": 0.01043492, + "balance_loss_clip": 1.06606531, + "balance_loss_mlp": 1.03003287, + "epoch": 0.13383033728130825, + "flos": 26322844942080.0, + "grad_norm": 2.2042317761279175, + "language_loss": 0.81395608, + "learning_rate": 3.888078858508588e-06, + "loss": 0.83679116, + "num_input_tokens_seen": 23680900, + "step": 1113, + "time_per_iteration": 3.6799111366271973 + }, + { + "auxiliary_loss_clip": 0.01216512, + "auxiliary_loss_mlp": 0.01040941, + "balance_loss_clip": 1.06599569, + "balance_loss_mlp": 1.02739918, + "epoch": 0.13395058017194733, + "flos": 22563931501440.0, + "grad_norm": 1.737258285497413, + "language_loss": 0.84634566, + "learning_rate": 3.8878217850471365e-06, + "loss": 0.86892021, + "num_input_tokens_seen": 23700815, + "step": 1114, + "time_per_iteration": 2.767498731613159 + }, + { + "auxiliary_loss_clip": 0.01245422, + "auxiliary_loss_mlp": 0.01037741, + "balance_loss_clip": 1.06978893, + "balance_loss_mlp": 1.02413917, + "epoch": 0.13407082306258641, + "flos": 25810938264960.0, + "grad_norm": 2.210132831375369, + "language_loss": 0.74146038, + "learning_rate": 3.887564425204621e-06, + "loss": 0.764292, + "num_input_tokens_seen": 23722500, + "step": 1115, + "time_per_iteration": 2.710089921951294 + }, + { + "auxiliary_loss_clip": 0.01106433, + "auxiliary_loss_mlp": 0.01006958, + "balance_loss_clip": 1.03214896, + "balance_loss_mlp": 1.00087821, + "epoch": 0.13419106595322552, + "flos": 68338365269760.0, + "grad_norm": 0.8450984328704004, + "language_loss": 0.54709101, + "learning_rate": 3.887306779020083e-06, + "loss": 0.56822491, + "num_input_tokens_seen": 23777155, + "step": 1116, + "time_per_iteration": 3.3002896308898926 + }, + { + "auxiliary_loss_clip": 0.01234185, + "auxiliary_loss_mlp": 0.01038549, + "balance_loss_clip": 1.06644499, + "balance_loss_mlp": 1.02550173, + "epoch": 0.1343113088438646, + "flos": 20449080489600.0, + "grad_norm": 2.359816117710648, + "language_loss": 0.69837344, + "learning_rate": 3.887048846532608e-06, + "loss": 0.72110075, + "num_input_tokens_seen": 23794130, + "step": 1117, + "time_per_iteration": 2.6751489639282227 + }, + { + "auxiliary_loss_clip": 0.01109757, + "auxiliary_loss_mlp": 0.01005992, + "balance_loss_clip": 1.03144455, + "balance_loss_mlp": 0.99998397, + "epoch": 0.1344315517345037, + "flos": 67389784951680.0, + "grad_norm": 0.7664164377116053, + "language_loss": 0.58168983, + "learning_rate": 3.8867906277813224e-06, + "loss": 0.60284722, + "num_input_tokens_seen": 23852285, + "step": 1118, + "time_per_iteration": 3.244703769683838 + }, + { + "auxiliary_loss_clip": 0.01235335, + "auxiliary_loss_mlp": 0.00888869, + "balance_loss_clip": 1.06605327, + "balance_loss_mlp": 1.00038052, + "epoch": 0.1345517946251428, + "flos": 40734442788480.0, + "grad_norm": 2.0782959812559776, + "language_loss": 0.73941517, + "learning_rate": 3.886532122805399e-06, + "loss": 0.76065719, + "num_input_tokens_seen": 23874765, + "step": 1119, + "time_per_iteration": 2.96406626701355 + }, + { + "auxiliary_loss_clip": 0.01171251, + "auxiliary_loss_mlp": 0.01047524, + "balance_loss_clip": 1.05144954, + "balance_loss_mlp": 1.0337553, + "epoch": 0.13467203751578188, + "flos": 22816850140800.0, + "grad_norm": 1.8555192066603063, + "language_loss": 0.89762866, + "learning_rate": 3.886273331644053e-06, + "loss": 0.91981637, + "num_input_tokens_seen": 23893635, + "step": 1120, + "time_per_iteration": 2.9161105155944824 + }, + { + "auxiliary_loss_clip": 0.01195499, + "auxiliary_loss_mlp": 0.01043001, + "balance_loss_clip": 1.06024361, + "balance_loss_mlp": 1.02935147, + "epoch": 0.13479228040642097, + "flos": 17091576512640.0, + "grad_norm": 2.2514640705997873, + "language_loss": 0.82351542, + "learning_rate": 3.886014254336542e-06, + "loss": 0.84590042, + "num_input_tokens_seen": 23910110, + "step": 1121, + "time_per_iteration": 2.8530642986297607 + }, + { + "auxiliary_loss_clip": 0.01226404, + "auxiliary_loss_mlp": 0.01043142, + "balance_loss_clip": 1.06149089, + "balance_loss_mlp": 1.03112531, + "epoch": 0.13491252329706005, + "flos": 23730525417600.0, + "grad_norm": 1.7081425063903677, + "language_loss": 0.92256796, + "learning_rate": 3.885754890922168e-06, + "loss": 0.94526339, + "num_input_tokens_seen": 23930440, + "step": 1122, + "time_per_iteration": 2.7543699741363525 + }, + { + "auxiliary_loss_clip": 0.01162975, + "auxiliary_loss_mlp": 0.01048986, + "balance_loss_clip": 1.05473304, + "balance_loss_mlp": 1.03534818, + "epoch": 0.13503276618769916, + "flos": 34127058960000.0, + "grad_norm": 1.7990773442896568, + "language_loss": 0.78493428, + "learning_rate": 3.885495241440277e-06, + "loss": 0.80705386, + "num_input_tokens_seen": 23954535, + "step": 1123, + "time_per_iteration": 3.219191074371338 + }, + { + "auxiliary_loss_clip": 0.01240744, + "auxiliary_loss_mlp": 0.010464, + "balance_loss_clip": 1.06506002, + "balance_loss_mlp": 1.03233337, + "epoch": 0.13515300907833824, + "flos": 17712328377600.0, + "grad_norm": 1.9492172790443745, + "language_loss": 0.74347579, + "learning_rate": 3.885235305930257e-06, + "loss": 0.76634723, + "num_input_tokens_seen": 23972735, + "step": 1124, + "time_per_iteration": 2.9964053630828857 + }, + { + "auxiliary_loss_clip": 0.01201197, + "auxiliary_loss_mlp": 0.0104078, + "balance_loss_clip": 1.0611546, + "balance_loss_mlp": 1.02802467, + "epoch": 0.13527325196897733, + "flos": 20260872201600.0, + "grad_norm": 1.939952579225078, + "language_loss": 0.8574838, + "learning_rate": 3.884975084431539e-06, + "loss": 0.87990355, + "num_input_tokens_seen": 23987685, + "step": 1125, + "time_per_iteration": 2.795454263687134 + }, + { + "auxiliary_loss_clip": 0.01213827, + "auxiliary_loss_mlp": 0.00889042, + "balance_loss_clip": 1.06066656, + "balance_loss_mlp": 1.00035787, + "epoch": 0.13539349485961644, + "flos": 18186492839040.0, + "grad_norm": 2.5571430093937555, + "language_loss": 0.91593409, + "learning_rate": 3.8847145769836e-06, + "loss": 0.93696272, + "num_input_tokens_seen": 24004105, + "step": 1126, + "time_per_iteration": 2.799912929534912 + }, + { + "auxiliary_loss_clip": 0.01244047, + "auxiliary_loss_mlp": 0.01048605, + "balance_loss_clip": 1.06671059, + "balance_loss_mlp": 1.03471708, + "epoch": 0.13551373775025552, + "flos": 19317463441920.0, + "grad_norm": 3.5527853765310966, + "language_loss": 0.66695559, + "learning_rate": 3.884453783625959e-06, + "loss": 0.6898821, + "num_input_tokens_seen": 24021715, + "step": 1127, + "time_per_iteration": 2.783360481262207 + }, + { + "auxiliary_loss_clip": 0.0121678, + "auxiliary_loss_mlp": 0.01049843, + "balance_loss_clip": 1.063766, + "balance_loss_mlp": 1.03624165, + "epoch": 0.1356339806408946, + "flos": 20850813175680.0, + "grad_norm": 2.6336929548190726, + "language_loss": 0.84814829, + "learning_rate": 3.884192704398176e-06, + "loss": 0.8708145, + "num_input_tokens_seen": 24038915, + "step": 1128, + "time_per_iteration": 2.833833932876587 + }, + { + "auxiliary_loss_clip": 0.01229803, + "auxiliary_loss_mlp": 0.0104748, + "balance_loss_clip": 1.06346536, + "balance_loss_mlp": 1.0348978, + "epoch": 0.13575422353153369, + "flos": 50476037696640.0, + "grad_norm": 3.327668737745151, + "language_loss": 0.74584591, + "learning_rate": 3.883931339339858e-06, + "loss": 0.76861876, + "num_input_tokens_seen": 24063300, + "step": 1129, + "time_per_iteration": 2.966177225112915 + }, + { + "auxiliary_loss_clip": 0.01236118, + "auxiliary_loss_mlp": 0.01042766, + "balance_loss_clip": 1.06459916, + "balance_loss_mlp": 1.02819836, + "epoch": 0.1358744664221728, + "flos": 18150797698560.0, + "grad_norm": 2.0242498380846214, + "language_loss": 0.7894572, + "learning_rate": 3.883669688490654e-06, + "loss": 0.81224608, + "num_input_tokens_seen": 24081070, + "step": 1130, + "time_per_iteration": 2.775104284286499 + }, + { + "auxiliary_loss_clip": 0.0120739, + "auxiliary_loss_mlp": 0.00889505, + "balance_loss_clip": 1.06030977, + "balance_loss_mlp": 1.00043368, + "epoch": 0.13599470931281188, + "flos": 18442966924800.0, + "grad_norm": 2.1268095838506893, + "language_loss": 0.85449493, + "learning_rate": 3.883407751890256e-06, + "loss": 0.87546384, + "num_input_tokens_seen": 24099675, + "step": 1131, + "time_per_iteration": 2.7761175632476807 + }, + { + "auxiliary_loss_clip": 0.01200405, + "auxiliary_loss_mlp": 0.01047207, + "balance_loss_clip": 1.05637789, + "balance_loss_mlp": 1.03324795, + "epoch": 0.13611495220345096, + "flos": 26680766014080.0, + "grad_norm": 1.7460285449979096, + "language_loss": 0.85729504, + "learning_rate": 3.8831455295783994e-06, + "loss": 0.87977111, + "num_input_tokens_seen": 24118925, + "step": 1132, + "time_per_iteration": 2.875253200531006 + }, + { + "auxiliary_loss_clip": 0.01213185, + "auxiliary_loss_mlp": 0.01043895, + "balance_loss_clip": 1.06041014, + "balance_loss_mlp": 1.03028107, + "epoch": 0.13623519509409007, + "flos": 21686238673920.0, + "grad_norm": 1.936512555237891, + "language_loss": 0.74048012, + "learning_rate": 3.882883021594864e-06, + "loss": 0.76305091, + "num_input_tokens_seen": 24137065, + "step": 1133, + "time_per_iteration": 2.7946131229400635 + }, + { + "auxiliary_loss_clip": 0.01198869, + "auxiliary_loss_mlp": 0.01044398, + "balance_loss_clip": 1.06141043, + "balance_loss_mlp": 1.03118324, + "epoch": 0.13635543798472916, + "flos": 14830389492480.0, + "grad_norm": 2.4403494624208184, + "language_loss": 0.86800504, + "learning_rate": 3.8826202279794705e-06, + "loss": 0.89043772, + "num_input_tokens_seen": 24154125, + "step": 1134, + "time_per_iteration": 2.751809597015381 + }, + { + "auxiliary_loss_clip": 0.01242179, + "auxiliary_loss_mlp": 0.01047297, + "balance_loss_clip": 1.06778264, + "balance_loss_mlp": 1.03398108, + "epoch": 0.13647568087536824, + "flos": 22890323410560.0, + "grad_norm": 1.9968548701608426, + "language_loss": 0.70028365, + "learning_rate": 3.882357148772085e-06, + "loss": 0.72317839, + "num_input_tokens_seen": 24171550, + "step": 1135, + "time_per_iteration": 2.69492506980896 + }, + { + "auxiliary_loss_clip": 0.01193424, + "auxiliary_loss_mlp": 0.01036606, + "balance_loss_clip": 1.05844784, + "balance_loss_mlp": 1.02376735, + "epoch": 0.13659592376600732, + "flos": 19937927998080.0, + "grad_norm": 2.301014355134209, + "language_loss": 0.84404129, + "learning_rate": 3.882093784012617e-06, + "loss": 0.86634159, + "num_input_tokens_seen": 24190190, + "step": 1136, + "time_per_iteration": 2.8125953674316406 + }, + { + "auxiliary_loss_clip": 0.01216568, + "auxiliary_loss_mlp": 0.01044388, + "balance_loss_clip": 1.06273913, + "balance_loss_mlp": 1.03075659, + "epoch": 0.13671616665664643, + "flos": 21428579439360.0, + "grad_norm": 1.8136553626939862, + "language_loss": 0.84309173, + "learning_rate": 3.881830133741019e-06, + "loss": 0.86570126, + "num_input_tokens_seen": 24209055, + "step": 1137, + "time_per_iteration": 3.6798455715179443 + }, + { + "auxiliary_loss_clip": 0.01208129, + "auxiliary_loss_mlp": 0.01047221, + "balance_loss_clip": 1.06486487, + "balance_loss_mlp": 1.03383994, + "epoch": 0.13683640954728551, + "flos": 22778138257920.0, + "grad_norm": 2.2950225170196137, + "language_loss": 0.75805432, + "learning_rate": 3.881566197997285e-06, + "loss": 0.78060782, + "num_input_tokens_seen": 24225490, + "step": 1138, + "time_per_iteration": 4.7101781368255615 + }, + { + "auxiliary_loss_clip": 0.01208117, + "auxiliary_loss_mlp": 0.01041724, + "balance_loss_clip": 1.06175804, + "balance_loss_mlp": 1.03007674, + "epoch": 0.1369566524379246, + "flos": 21725884310400.0, + "grad_norm": 1.648205497928934, + "language_loss": 0.748918, + "learning_rate": 3.881301976821456e-06, + "loss": 0.77141643, + "num_input_tokens_seen": 24245520, + "step": 1139, + "time_per_iteration": 3.764559030532837 + }, + { + "auxiliary_loss_clip": 0.01221875, + "auxiliary_loss_mlp": 0.0103647, + "balance_loss_clip": 1.06318712, + "balance_loss_mlp": 1.02401876, + "epoch": 0.1370768953285637, + "flos": 18624459369600.0, + "grad_norm": 1.8016824899245025, + "language_loss": 0.90691376, + "learning_rate": 3.881037470253612e-06, + "loss": 0.92949718, + "num_input_tokens_seen": 24265035, + "step": 1140, + "time_per_iteration": 2.792802572250366 + }, + { + "auxiliary_loss_clip": 0.01195691, + "auxiliary_loss_mlp": 0.01043805, + "balance_loss_clip": 1.06020451, + "balance_loss_mlp": 1.03127003, + "epoch": 0.1371971382192028, + "flos": 14939521989120.0, + "grad_norm": 3.3535089710464465, + "language_loss": 0.79625326, + "learning_rate": 3.88077267833388e-06, + "loss": 0.81864822, + "num_input_tokens_seen": 24281550, + "step": 1141, + "time_per_iteration": 3.0121700763702393 + }, + { + "auxiliary_loss_clip": 0.01186669, + "auxiliary_loss_mlp": 0.01039456, + "balance_loss_clip": 1.0567944, + "balance_loss_mlp": 1.02659333, + "epoch": 0.13731738110984187, + "flos": 19023785844480.0, + "grad_norm": 2.1477791890934124, + "language_loss": 0.83888263, + "learning_rate": 3.880507601102427e-06, + "loss": 0.86114383, + "num_input_tokens_seen": 24299485, + "step": 1142, + "time_per_iteration": 2.917687177658081 + }, + { + "auxiliary_loss_clip": 0.01236422, + "auxiliary_loss_mlp": 0.01035863, + "balance_loss_clip": 1.06643426, + "balance_loss_mlp": 1.0238111, + "epoch": 0.13743762400048098, + "flos": 18187462506240.0, + "grad_norm": 1.8271371214560157, + "language_loss": 0.82337356, + "learning_rate": 3.880242238599467e-06, + "loss": 0.8460964, + "num_input_tokens_seen": 24316010, + "step": 1143, + "time_per_iteration": 2.7177116870880127 + }, + { + "auxiliary_loss_clip": 0.01235656, + "auxiliary_loss_mlp": 0.01037222, + "balance_loss_clip": 1.0654304, + "balance_loss_mlp": 1.02462113, + "epoch": 0.13755786689112007, + "flos": 21031982398080.0, + "grad_norm": 1.8686087597558978, + "language_loss": 0.83149123, + "learning_rate": 3.879976590865254e-06, + "loss": 0.85422003, + "num_input_tokens_seen": 24335465, + "step": 1144, + "time_per_iteration": 2.6698837280273438 + }, + { + "auxiliary_loss_clip": 0.01213767, + "auxiliary_loss_mlp": 0.01037424, + "balance_loss_clip": 1.06281626, + "balance_loss_mlp": 1.02531838, + "epoch": 0.13767810978175915, + "flos": 21360636864000.0, + "grad_norm": 1.9408869914948985, + "language_loss": 0.87943107, + "learning_rate": 3.879710657940087e-06, + "loss": 0.90194297, + "num_input_tokens_seen": 24354415, + "step": 1145, + "time_per_iteration": 2.843904972076416 + }, + { + "auxiliary_loss_clip": 0.01227477, + "auxiliary_loss_mlp": 0.01043159, + "balance_loss_clip": 1.06359458, + "balance_loss_mlp": 1.0293839, + "epoch": 0.13779835267239823, + "flos": 30592084861440.0, + "grad_norm": 2.1911625614975545, + "language_loss": 0.70291954, + "learning_rate": 3.879444439864308e-06, + "loss": 0.72562593, + "num_input_tokens_seen": 24373990, + "step": 1146, + "time_per_iteration": 2.7956771850585938 + }, + { + "auxiliary_loss_clip": 0.01223394, + "auxiliary_loss_mlp": 0.00888684, + "balance_loss_clip": 1.06194139, + "balance_loss_mlp": 1.00040448, + "epoch": 0.13791859556303734, + "flos": 22669867687680.0, + "grad_norm": 2.6289802138848475, + "language_loss": 0.85906899, + "learning_rate": 3.879177936678301e-06, + "loss": 0.88018978, + "num_input_tokens_seen": 24392995, + "step": 1147, + "time_per_iteration": 2.6802985668182373 + }, + { + "auxiliary_loss_clip": 0.01218794, + "auxiliary_loss_mlp": 0.01044387, + "balance_loss_clip": 1.06046557, + "balance_loss_mlp": 1.03199482, + "epoch": 0.13803883845367643, + "flos": 35224166016000.0, + "grad_norm": 1.9236711261448218, + "language_loss": 0.77268785, + "learning_rate": 3.878911148422496e-06, + "loss": 0.79531968, + "num_input_tokens_seen": 24414470, + "step": 1148, + "time_per_iteration": 2.884675979614258 + }, + { + "auxiliary_loss_clip": 0.01227535, + "auxiliary_loss_mlp": 0.01039214, + "balance_loss_clip": 1.06321812, + "balance_loss_mlp": 1.02524269, + "epoch": 0.1381590813443155, + "flos": 32014542332160.0, + "grad_norm": 2.045409112305915, + "language_loss": 0.7003057, + "learning_rate": 3.878644075137364e-06, + "loss": 0.72297317, + "num_input_tokens_seen": 24435120, + "step": 1149, + "time_per_iteration": 2.867983341217041 + }, + { + "auxiliary_loss_clip": 0.01184778, + "auxiliary_loss_mlp": 0.01046267, + "balance_loss_clip": 1.05462563, + "balance_loss_mlp": 1.03335655, + "epoch": 0.13827932423495462, + "flos": 17821855923840.0, + "grad_norm": 1.8652087658373464, + "language_loss": 0.79299641, + "learning_rate": 3.878376716863418e-06, + "loss": 0.8153069, + "num_input_tokens_seen": 24451420, + "step": 1150, + "time_per_iteration": 2.8346941471099854 + }, + { + "auxiliary_loss_clip": 0.0121182, + "auxiliary_loss_mlp": 0.01044993, + "balance_loss_clip": 1.06117809, + "balance_loss_mlp": 1.03151083, + "epoch": 0.1383995671255937, + "flos": 19427098728960.0, + "grad_norm": 2.3122620281561725, + "language_loss": 0.71868676, + "learning_rate": 3.878109073641219e-06, + "loss": 0.74125493, + "num_input_tokens_seen": 24470450, + "step": 1151, + "time_per_iteration": 2.7793002128601074 + }, + { + "auxiliary_loss_clip": 0.0119092, + "auxiliary_loss_mlp": 0.01040756, + "balance_loss_clip": 1.06104553, + "balance_loss_mlp": 1.02733946, + "epoch": 0.13851981001623279, + "flos": 28296603331200.0, + "grad_norm": 2.8464840802252738, + "language_loss": 0.81338775, + "learning_rate": 3.877841145511366e-06, + "loss": 0.83570457, + "num_input_tokens_seen": 24493190, + "step": 1152, + "time_per_iteration": 2.911499500274658 + }, + { + "auxiliary_loss_clip": 0.01228669, + "auxiliary_loss_mlp": 0.01036907, + "balance_loss_clip": 1.06342578, + "balance_loss_mlp": 1.02441406, + "epoch": 0.13864005290687187, + "flos": 21213079793280.0, + "grad_norm": 1.8811112580282696, + "language_loss": 0.82971066, + "learning_rate": 3.8775729325145035e-06, + "loss": 0.85236645, + "num_input_tokens_seen": 24512425, + "step": 1153, + "time_per_iteration": 2.860275983810425 + }, + { + "auxiliary_loss_clip": 0.01103134, + "auxiliary_loss_mlp": 0.0100517, + "balance_loss_clip": 1.03600287, + "balance_loss_mlp": 1.00002003, + "epoch": 0.13876029579751098, + "flos": 71653389413760.0, + "grad_norm": 0.789669974727848, + "language_loss": 0.6471113, + "learning_rate": 3.877304434691321e-06, + "loss": 0.66819441, + "num_input_tokens_seen": 24579275, + "step": 1154, + "time_per_iteration": 3.51410174369812 + }, + { + "auxiliary_loss_clip": 0.0120494, + "auxiliary_loss_mlp": 0.01036887, + "balance_loss_clip": 1.06368959, + "balance_loss_mlp": 1.0244298, + "epoch": 0.13888053868815006, + "flos": 21941348042880.0, + "grad_norm": 1.8156144089436248, + "language_loss": 0.7949425, + "learning_rate": 3.877035652082548e-06, + "loss": 0.81736088, + "num_input_tokens_seen": 24598720, + "step": 1155, + "time_per_iteration": 2.8843252658843994 + }, + { + "auxiliary_loss_clip": 0.01205911, + "auxiliary_loss_mlp": 0.01047926, + "balance_loss_clip": 1.06213641, + "balance_loss_mlp": 1.03430009, + "epoch": 0.13900078157878915, + "flos": 19608627087360.0, + "grad_norm": 1.7443935065245029, + "language_loss": 0.85700631, + "learning_rate": 3.87676658472896e-06, + "loss": 0.87954468, + "num_input_tokens_seen": 24617530, + "step": 1156, + "time_per_iteration": 2.7452139854431152 + }, + { + "auxiliary_loss_clip": 0.01227167, + "auxiliary_loss_mlp": 0.01048032, + "balance_loss_clip": 1.06298006, + "balance_loss_mlp": 1.03497815, + "epoch": 0.13912102446942826, + "flos": 22638051216000.0, + "grad_norm": 1.7584240048447641, + "language_loss": 0.85337341, + "learning_rate": 3.876497232671372e-06, + "loss": 0.8761254, + "num_input_tokens_seen": 24637485, + "step": 1157, + "time_per_iteration": 2.869722366333008 + }, + { + "auxiliary_loss_clip": 0.01198386, + "auxiliary_loss_mlp": 0.01043636, + "balance_loss_clip": 1.05839562, + "balance_loss_mlp": 1.03082669, + "epoch": 0.13924126736006734, + "flos": 29643324975360.0, + "grad_norm": 3.1854877026269692, + "language_loss": 0.83514559, + "learning_rate": 3.876227595950647e-06, + "loss": 0.85756582, + "num_input_tokens_seen": 24656915, + "step": 1158, + "time_per_iteration": 3.0047292709350586 + }, + { + "auxiliary_loss_clip": 0.01236318, + "auxiliary_loss_mlp": 0.01044102, + "balance_loss_clip": 1.06592095, + "balance_loss_mlp": 1.03177559, + "epoch": 0.13936151025070642, + "flos": 27417653527680.0, + "grad_norm": 4.282637959287355, + "language_loss": 0.79012907, + "learning_rate": 3.875957674607686e-06, + "loss": 0.81293321, + "num_input_tokens_seen": 24679190, + "step": 1159, + "time_per_iteration": 2.7490432262420654 + }, + { + "auxiliary_loss_clip": 0.01210286, + "auxiliary_loss_mlp": 0.0088905, + "balance_loss_clip": 1.05639911, + "balance_loss_mlp": 1.00028276, + "epoch": 0.1394817531413455, + "flos": 16399326625920.0, + "grad_norm": 1.7631526819200227, + "language_loss": 0.87972295, + "learning_rate": 3.8756874686834386e-06, + "loss": 0.9007163, + "num_input_tokens_seen": 24697405, + "step": 1160, + "time_per_iteration": 2.7528326511383057 + }, + { + "auxiliary_loss_clip": 0.01230463, + "auxiliary_loss_mlp": 0.00888984, + "balance_loss_clip": 1.06287503, + "balance_loss_mlp": 1.00035632, + "epoch": 0.13960199603198462, + "flos": 30922319525760.0, + "grad_norm": 1.7387455354827, + "language_loss": 0.80513203, + "learning_rate": 3.875416978218893e-06, + "loss": 0.82632649, + "num_input_tokens_seen": 24720600, + "step": 1161, + "time_per_iteration": 2.7713937759399414 + }, + { + "auxiliary_loss_clip": 0.01210703, + "auxiliary_loss_mlp": 0.01037962, + "balance_loss_clip": 1.05646372, + "balance_loss_mlp": 1.02520037, + "epoch": 0.1397222389226237, + "flos": 18113773754880.0, + "grad_norm": 2.5492696833535744, + "language_loss": 0.83477342, + "learning_rate": 3.8751462032550835e-06, + "loss": 0.85725999, + "num_input_tokens_seen": 24737605, + "step": 1162, + "time_per_iteration": 2.7835938930511475 + }, + { + "auxiliary_loss_clip": 0.01214073, + "auxiliary_loss_mlp": 0.0103919, + "balance_loss_clip": 1.06477332, + "balance_loss_mlp": 1.02652359, + "epoch": 0.13984248181326278, + "flos": 16872772815360.0, + "grad_norm": 2.421106282934928, + "language_loss": 0.82996017, + "learning_rate": 3.874875143833085e-06, + "loss": 0.85249281, + "num_input_tokens_seen": 24755845, + "step": 1163, + "time_per_iteration": 4.631893634796143 + }, + { + "auxiliary_loss_clip": 0.01227087, + "auxiliary_loss_mlp": 0.01035317, + "balance_loss_clip": 1.06382835, + "balance_loss_mlp": 1.02162623, + "epoch": 0.1399627247039019, + "flos": 54121401267840.0, + "grad_norm": 1.8390774053429106, + "language_loss": 0.69128644, + "learning_rate": 3.874603799994019e-06, + "loss": 0.71391046, + "num_input_tokens_seen": 24779380, + "step": 1164, + "time_per_iteration": 3.929532527923584 + }, + { + "auxiliary_loss_clip": 0.01195764, + "auxiliary_loss_mlp": 0.01037988, + "balance_loss_clip": 1.05778813, + "balance_loss_mlp": 1.02545953, + "epoch": 0.14008296759454097, + "flos": 11765521618560.0, + "grad_norm": 2.0967213411528167, + "language_loss": 0.86563885, + "learning_rate": 3.874332171779046e-06, + "loss": 0.88797641, + "num_input_tokens_seen": 24794260, + "step": 1165, + "time_per_iteration": 3.7376859188079834 + }, + { + "auxiliary_loss_clip": 0.01198656, + "auxiliary_loss_mlp": 0.01036094, + "balance_loss_clip": 1.05955505, + "balance_loss_mlp": 1.02277279, + "epoch": 0.14020321048518006, + "flos": 22017514832640.0, + "grad_norm": 1.7712797513482355, + "language_loss": 0.75491834, + "learning_rate": 3.874060259229373e-06, + "loss": 0.77726579, + "num_input_tokens_seen": 24815835, + "step": 1166, + "time_per_iteration": 2.8695971965789795 + }, + { + "auxiliary_loss_clip": 0.01230608, + "auxiliary_loss_mlp": 0.0104321, + "balance_loss_clip": 1.06641984, + "balance_loss_mlp": 1.03025162, + "epoch": 0.14032345337581917, + "flos": 23404313076480.0, + "grad_norm": 2.3280680689321573, + "language_loss": 0.93414575, + "learning_rate": 3.873788062386249e-06, + "loss": 0.95688391, + "num_input_tokens_seen": 24834095, + "step": 1167, + "time_per_iteration": 2.6830525398254395 + }, + { + "auxiliary_loss_clip": 0.01211071, + "auxiliary_loss_mlp": 0.01045693, + "balance_loss_clip": 1.06308675, + "balance_loss_mlp": 1.03198433, + "epoch": 0.14044369626645825, + "flos": 29645767100160.0, + "grad_norm": 1.784706251423246, + "language_loss": 0.82171565, + "learning_rate": 3.873515581290965e-06, + "loss": 0.84428334, + "num_input_tokens_seen": 24858900, + "step": 1168, + "time_per_iteration": 2.9061808586120605 + }, + { + "auxiliary_loss_clip": 0.01205627, + "auxiliary_loss_mlp": 0.01046141, + "balance_loss_clip": 1.06259346, + "balance_loss_mlp": 1.03361797, + "epoch": 0.14056393915709733, + "flos": 18332972501760.0, + "grad_norm": 3.5913232654180045, + "language_loss": 0.75799745, + "learning_rate": 3.8732428159848575e-06, + "loss": 0.78051513, + "num_input_tokens_seen": 24877875, + "step": 1169, + "time_per_iteration": 2.785369634628296 + }, + { + "auxiliary_loss_clip": 0.01226659, + "auxiliary_loss_mlp": 0.01040889, + "balance_loss_clip": 1.0671165, + "balance_loss_mlp": 1.02784169, + "epoch": 0.14068418204773642, + "flos": 26687517770880.0, + "grad_norm": 1.9652462990136104, + "language_loss": 0.78370225, + "learning_rate": 3.872969766509304e-06, + "loss": 0.80637777, + "num_input_tokens_seen": 24898430, + "step": 1170, + "time_per_iteration": 2.8933329582214355 + }, + { + "auxiliary_loss_clip": 0.01097804, + "auxiliary_loss_mlp": 0.01006351, + "balance_loss_clip": 1.03173542, + "balance_loss_mlp": 1.00086772, + "epoch": 0.14080442493837553, + "flos": 65259314501760.0, + "grad_norm": 0.7638151890779793, + "language_loss": 0.55648422, + "learning_rate": 3.872696432905726e-06, + "loss": 0.57752579, + "num_input_tokens_seen": 24959250, + "step": 1171, + "time_per_iteration": 3.423584222793579 + }, + { + "auxiliary_loss_clip": 0.01233623, + "auxiliary_loss_mlp": 0.01047341, + "balance_loss_clip": 1.06313419, + "balance_loss_mlp": 1.03311324, + "epoch": 0.1409246678290146, + "flos": 25776715582080.0, + "grad_norm": 2.4742794059447695, + "language_loss": 0.71497393, + "learning_rate": 3.872422815215589e-06, + "loss": 0.73778355, + "num_input_tokens_seen": 24978330, + "step": 1172, + "time_per_iteration": 2.768357992172241 + }, + { + "auxiliary_loss_clip": 0.01221077, + "auxiliary_loss_mlp": 0.01039429, + "balance_loss_clip": 1.05771494, + "balance_loss_mlp": 1.025172, + "epoch": 0.1410449107196537, + "flos": 21868521217920.0, + "grad_norm": 1.8392440176369884, + "language_loss": 0.74278569, + "learning_rate": 3.8721489134803994e-06, + "loss": 0.76539075, + "num_input_tokens_seen": 24997120, + "step": 1173, + "time_per_iteration": 2.82464599609375 + }, + { + "auxiliary_loss_clip": 0.01227319, + "auxiliary_loss_mlp": 0.01042934, + "balance_loss_clip": 1.06727755, + "balance_loss_mlp": 1.02957654, + "epoch": 0.1411651536102928, + "flos": 16684133564160.0, + "grad_norm": 2.6433246750583987, + "language_loss": 0.72704625, + "learning_rate": 3.871874727741707e-06, + "loss": 0.74974877, + "num_input_tokens_seen": 25014350, + "step": 1174, + "time_per_iteration": 2.8105294704437256 + }, + { + "auxiliary_loss_clip": 0.01223958, + "auxiliary_loss_mlp": 0.01042786, + "balance_loss_clip": 1.06662881, + "balance_loss_mlp": 1.0306623, + "epoch": 0.1412853965009319, + "flos": 20992264934400.0, + "grad_norm": 1.7453013837518863, + "language_loss": 0.96653211, + "learning_rate": 3.871600258041108e-06, + "loss": 0.98919958, + "num_input_tokens_seen": 25033875, + "step": 1175, + "time_per_iteration": 2.893118381500244 + }, + { + "auxiliary_loss_clip": 0.01212866, + "auxiliary_loss_mlp": 0.01042012, + "balance_loss_clip": 1.06040454, + "balance_loss_mlp": 1.02737319, + "epoch": 0.14140563939157097, + "flos": 20335279224960.0, + "grad_norm": 2.105704153753896, + "language_loss": 0.85606688, + "learning_rate": 3.871325504420238e-06, + "loss": 0.87861568, + "num_input_tokens_seen": 25052865, + "step": 1176, + "time_per_iteration": 2.832235097885132 + }, + { + "auxiliary_loss_clip": 0.01237741, + "auxiliary_loss_mlp": 0.01052422, + "balance_loss_clip": 1.06659615, + "balance_loss_mlp": 1.03850973, + "epoch": 0.14152588228221005, + "flos": 21068826773760.0, + "grad_norm": 1.8652755865572088, + "language_loss": 0.81752348, + "learning_rate": 3.871050466920776e-06, + "loss": 0.84042507, + "num_input_tokens_seen": 25072770, + "step": 1177, + "time_per_iteration": 2.7448229789733887 + }, + { + "auxiliary_loss_clip": 0.01194299, + "auxiliary_loss_mlp": 0.01046909, + "balance_loss_clip": 1.05551934, + "balance_loss_mlp": 1.03435659, + "epoch": 0.14164612517284916, + "flos": 18223157646720.0, + "grad_norm": 2.0594930173297956, + "language_loss": 0.79715002, + "learning_rate": 3.870775145584447e-06, + "loss": 0.81956208, + "num_input_tokens_seen": 25090550, + "step": 1178, + "time_per_iteration": 2.8192923069000244 + }, + { + "auxiliary_loss_clip": 0.01223577, + "auxiliary_loss_mlp": 0.01045984, + "balance_loss_clip": 1.06159699, + "balance_loss_mlp": 1.03183949, + "epoch": 0.14176636806348825, + "flos": 22744454279040.0, + "grad_norm": 4.028209988424469, + "language_loss": 0.65242362, + "learning_rate": 3.8704995404530145e-06, + "loss": 0.67511928, + "num_input_tokens_seen": 25106175, + "step": 1179, + "time_per_iteration": 2.750140905380249 + }, + { + "auxiliary_loss_clip": 0.01234139, + "auxiliary_loss_mlp": 0.01048793, + "balance_loss_clip": 1.06576347, + "balance_loss_mlp": 1.03650212, + "epoch": 0.14188661095412733, + "flos": 22091095843200.0, + "grad_norm": 1.7884855280826664, + "language_loss": 0.85269761, + "learning_rate": 3.87022365156829e-06, + "loss": 0.87552691, + "num_input_tokens_seen": 25126890, + "step": 1180, + "time_per_iteration": 2.717054843902588 + }, + { + "auxiliary_loss_clip": 0.01170748, + "auxiliary_loss_mlp": 0.01038581, + "balance_loss_clip": 1.05206037, + "balance_loss_mlp": 1.02514052, + "epoch": 0.14200685384476644, + "flos": 24352390604160.0, + "grad_norm": 1.951754713029812, + "language_loss": 0.80966759, + "learning_rate": 3.869947478972123e-06, + "loss": 0.83176088, + "num_input_tokens_seen": 25147915, + "step": 1181, + "time_per_iteration": 2.912196636199951 + }, + { + "auxiliary_loss_clip": 0.01213305, + "auxiliary_loss_mlp": 0.0103729, + "balance_loss_clip": 1.05794919, + "balance_loss_mlp": 1.023664, + "epoch": 0.14212709673540552, + "flos": 24022048199040.0, + "grad_norm": 3.516066074110615, + "language_loss": 0.82652593, + "learning_rate": 3.869671022706412e-06, + "loss": 0.84903181, + "num_input_tokens_seen": 25166645, + "step": 1182, + "time_per_iteration": 2.772092342376709 + }, + { + "auxiliary_loss_clip": 0.01180681, + "auxiliary_loss_mlp": 0.01045741, + "balance_loss_clip": 1.05269134, + "balance_loss_mlp": 1.03171039, + "epoch": 0.1422473396260446, + "flos": 26431797870720.0, + "grad_norm": 1.90859010205142, + "language_loss": 0.65121937, + "learning_rate": 3.869394282813092e-06, + "loss": 0.67348361, + "num_input_tokens_seen": 25185845, + "step": 1183, + "time_per_iteration": 2.927764892578125 + }, + { + "auxiliary_loss_clip": 0.01211966, + "auxiliary_loss_mlp": 0.01046343, + "balance_loss_clip": 1.05664861, + "balance_loss_mlp": 1.03282452, + "epoch": 0.1423675825166837, + "flos": 17055306754560.0, + "grad_norm": 5.045811023055941, + "language_loss": 0.89635915, + "learning_rate": 3.869117259334147e-06, + "loss": 0.91894221, + "num_input_tokens_seen": 25203770, + "step": 1184, + "time_per_iteration": 2.870500326156616 + }, + { + "auxiliary_loss_clip": 0.01217024, + "auxiliary_loss_mlp": 0.01047165, + "balance_loss_clip": 1.05840242, + "balance_loss_mlp": 1.03418326, + "epoch": 0.1424878254073228, + "flos": 17929480049280.0, + "grad_norm": 1.7246154749512377, + "language_loss": 0.82054383, + "learning_rate": 3.868839952311599e-06, + "loss": 0.84318572, + "num_input_tokens_seen": 25221725, + "step": 1185, + "time_per_iteration": 2.7469582557678223 + }, + { + "auxiliary_loss_clip": 0.01214892, + "auxiliary_loss_mlp": 0.01038136, + "balance_loss_clip": 1.06362009, + "balance_loss_mlp": 1.02515996, + "epoch": 0.14260806829796188, + "flos": 20303606407680.0, + "grad_norm": 2.222659332430551, + "language_loss": 0.80420208, + "learning_rate": 3.868562361787516e-06, + "loss": 0.8267324, + "num_input_tokens_seen": 25240855, + "step": 1186, + "time_per_iteration": 2.7540526390075684 + }, + { + "auxiliary_loss_clip": 0.01166331, + "auxiliary_loss_mlp": 0.01034619, + "balance_loss_clip": 1.05030882, + "balance_loss_mlp": 1.02113032, + "epoch": 0.14272831118860096, + "flos": 23185724860800.0, + "grad_norm": 4.053452240904763, + "language_loss": 0.690341, + "learning_rate": 3.868284487804009e-06, + "loss": 0.71235049, + "num_input_tokens_seen": 25260085, + "step": 1187, + "time_per_iteration": 2.798476219177246 + }, + { + "auxiliary_loss_clip": 0.0122202, + "auxiliary_loss_mlp": 0.01046832, + "balance_loss_clip": 1.06142485, + "balance_loss_mlp": 1.03398693, + "epoch": 0.14284855407924008, + "flos": 27232210586880.0, + "grad_norm": 1.678851157106376, + "language_loss": 0.78105068, + "learning_rate": 3.86800633040323e-06, + "loss": 0.80373919, + "num_input_tokens_seen": 25280675, + "step": 1188, + "time_per_iteration": 2.651773691177368 + }, + { + "auxiliary_loss_clip": 0.01214869, + "auxiliary_loss_mlp": 0.0088915, + "balance_loss_clip": 1.06566954, + "balance_loss_mlp": 1.00038123, + "epoch": 0.14296879696987916, + "flos": 28184202696960.0, + "grad_norm": 2.757626892320752, + "language_loss": 0.78307611, + "learning_rate": 3.867727889627376e-06, + "loss": 0.80411631, + "num_input_tokens_seen": 25300290, + "step": 1189, + "time_per_iteration": 3.5485329627990723 + }, + { + "auxiliary_loss_clip": 0.01194778, + "auxiliary_loss_mlp": 0.0103956, + "balance_loss_clip": 1.05933726, + "balance_loss_mlp": 1.02683413, + "epoch": 0.14308903986051824, + "flos": 19390290266880.0, + "grad_norm": 2.055330017033283, + "language_loss": 0.78776079, + "learning_rate": 3.867449165518687e-06, + "loss": 0.81010419, + "num_input_tokens_seen": 25316760, + "step": 1190, + "time_per_iteration": 3.401397466659546 + }, + { + "auxiliary_loss_clip": 0.01241343, + "auxiliary_loss_mlp": 0.00889251, + "balance_loss_clip": 1.06553996, + "balance_loss_mlp": 1.00028133, + "epoch": 0.14320928275115732, + "flos": 17457506317440.0, + "grad_norm": 1.8038299927047383, + "language_loss": 0.71392238, + "learning_rate": 3.867170158119444e-06, + "loss": 0.73522836, + "num_input_tokens_seen": 25335760, + "step": 1191, + "time_per_iteration": 3.4888064861297607 + }, + { + "auxiliary_loss_clip": 0.0123893, + "auxiliary_loss_mlp": 0.01043418, + "balance_loss_clip": 1.06589437, + "balance_loss_mlp": 1.02981663, + "epoch": 0.14332952564179643, + "flos": 21466070259840.0, + "grad_norm": 2.045456374794452, + "language_loss": 0.75417238, + "learning_rate": 3.866890867471972e-06, + "loss": 0.7769959, + "num_input_tokens_seen": 25354230, + "step": 1192, + "time_per_iteration": 2.7593677043914795 + }, + { + "auxiliary_loss_clip": 0.01216024, + "auxiliary_loss_mlp": 0.010504, + "balance_loss_clip": 1.06045556, + "balance_loss_mlp": 1.03528404, + "epoch": 0.14344976853243552, + "flos": 16396992241920.0, + "grad_norm": 2.997038802295845, + "language_loss": 0.89694619, + "learning_rate": 3.86661129361864e-06, + "loss": 0.91961044, + "num_input_tokens_seen": 25368720, + "step": 1193, + "time_per_iteration": 2.854881525039673 + }, + { + "auxiliary_loss_clip": 0.01216813, + "auxiliary_loss_mlp": 0.01048022, + "balance_loss_clip": 1.06448889, + "balance_loss_mlp": 1.03475356, + "epoch": 0.1435700114230746, + "flos": 18916736336640.0, + "grad_norm": 2.101505891885439, + "language_loss": 0.86629713, + "learning_rate": 3.866331436601859e-06, + "loss": 0.88894546, + "num_input_tokens_seen": 25386715, + "step": 1194, + "time_per_iteration": 2.903803586959839 + }, + { + "auxiliary_loss_clip": 0.01239203, + "auxiliary_loss_mlp": 0.01049873, + "balance_loss_clip": 1.06763744, + "balance_loss_mlp": 1.0365572, + "epoch": 0.1436902543137137, + "flos": 19755394058880.0, + "grad_norm": 2.3770366371235685, + "language_loss": 0.73682904, + "learning_rate": 3.866051296464083e-06, + "loss": 0.75971985, + "num_input_tokens_seen": 25405550, + "step": 1195, + "time_per_iteration": 2.86773419380188 + }, + { + "auxiliary_loss_clip": 0.01242312, + "auxiliary_loss_mlp": 0.00889363, + "balance_loss_clip": 1.06615579, + "balance_loss_mlp": 1.00035667, + "epoch": 0.1438104972043528, + "flos": 14684807669760.0, + "grad_norm": 2.77028843774137, + "language_loss": 0.85120696, + "learning_rate": 3.86577087324781e-06, + "loss": 0.87252367, + "num_input_tokens_seen": 25422040, + "step": 1196, + "time_per_iteration": 2.671142578125 + }, + { + "auxiliary_loss_clip": 0.01226427, + "auxiliary_loss_mlp": 0.01041552, + "balance_loss_clip": 1.06744933, + "balance_loss_mlp": 1.02917206, + "epoch": 0.14393074009499188, + "flos": 17092330698240.0, + "grad_norm": 3.066413549194419, + "language_loss": 0.77273035, + "learning_rate": 3.865490166995578e-06, + "loss": 0.79541016, + "num_input_tokens_seen": 25440270, + "step": 1197, + "time_per_iteration": 2.7909798622131348 + }, + { + "auxiliary_loss_clip": 0.0122909, + "auxiliary_loss_mlp": 0.01047278, + "balance_loss_clip": 1.06581342, + "balance_loss_mlp": 1.0335393, + "epoch": 0.144050982985631, + "flos": 30476200608000.0, + "grad_norm": 2.8443116440934735, + "language_loss": 0.84009796, + "learning_rate": 3.86520917774997e-06, + "loss": 0.86286163, + "num_input_tokens_seen": 25459705, + "step": 1198, + "time_per_iteration": 2.651167869567871 + }, + { + "auxiliary_loss_clip": 0.01225406, + "auxiliary_loss_mlp": 0.01036592, + "balance_loss_clip": 1.06592464, + "balance_loss_mlp": 1.02323449, + "epoch": 0.14417122587627007, + "flos": 17858484817920.0, + "grad_norm": 2.3451551884053132, + "language_loss": 0.74987257, + "learning_rate": 3.864927905553614e-06, + "loss": 0.77249253, + "num_input_tokens_seen": 25477615, + "step": 1199, + "time_per_iteration": 2.5940706729888916 + }, + { + "auxiliary_loss_clip": 0.01203974, + "auxiliary_loss_mlp": 0.01038959, + "balance_loss_clip": 1.06044388, + "balance_loss_mlp": 1.02561903, + "epoch": 0.14429146876690915, + "flos": 21613914639360.0, + "grad_norm": 1.8170660579662465, + "language_loss": 0.88976711, + "learning_rate": 3.8646463504491765e-06, + "loss": 0.91219652, + "num_input_tokens_seen": 25497750, + "step": 1200, + "time_per_iteration": 2.738807201385498 + }, + { + "auxiliary_loss_clip": 0.01234698, + "auxiliary_loss_mlp": 0.01041612, + "balance_loss_clip": 1.07034159, + "balance_loss_mlp": 1.02825451, + "epoch": 0.14441171165754824, + "flos": 23258120722560.0, + "grad_norm": 1.766763918145261, + "language_loss": 0.83230621, + "learning_rate": 3.8643645124793705e-06, + "loss": 0.85506928, + "num_input_tokens_seen": 25516650, + "step": 1201, + "time_per_iteration": 2.737372636795044 + }, + { + "auxiliary_loss_clip": 0.01222996, + "auxiliary_loss_mlp": 0.01041313, + "balance_loss_clip": 1.06265402, + "balance_loss_mlp": 1.02876639, + "epoch": 0.14453195454818735, + "flos": 42854213963520.0, + "grad_norm": 2.3535385575342884, + "language_loss": 0.75311732, + "learning_rate": 3.8640823916869515e-06, + "loss": 0.77576041, + "num_input_tokens_seen": 25540960, + "step": 1202, + "time_per_iteration": 2.99652099609375 + }, + { + "auxiliary_loss_clip": 0.01241994, + "auxiliary_loss_mlp": 0.01041079, + "balance_loss_clip": 1.06796193, + "balance_loss_mlp": 1.02846682, + "epoch": 0.14465219743882643, + "flos": 27235873774080.0, + "grad_norm": 1.6140798998572474, + "language_loss": 0.78443944, + "learning_rate": 3.863799988114714e-06, + "loss": 0.80727017, + "num_input_tokens_seen": 25562990, + "step": 1203, + "time_per_iteration": 2.825899839401245 + }, + { + "auxiliary_loss_clip": 0.01242281, + "auxiliary_loss_mlp": 0.01042706, + "balance_loss_clip": 1.06857467, + "balance_loss_mlp": 1.02930665, + "epoch": 0.1447724403294655, + "flos": 16690705752960.0, + "grad_norm": 2.6655117748861548, + "language_loss": 0.70953405, + "learning_rate": 3.863517301805502e-06, + "loss": 0.73238391, + "num_input_tokens_seen": 25581380, + "step": 1204, + "time_per_iteration": 2.88631534576416 + }, + { + "auxiliary_loss_clip": 0.01213653, + "auxiliary_loss_mlp": 0.01044606, + "balance_loss_clip": 1.06873715, + "balance_loss_mlp": 1.03130269, + "epoch": 0.14489268322010462, + "flos": 20073741321600.0, + "grad_norm": 3.054375037036088, + "language_loss": 0.97260845, + "learning_rate": 3.863234332802196e-06, + "loss": 0.99519104, + "num_input_tokens_seen": 25593585, + "step": 1205, + "time_per_iteration": 2.7720282077789307 + }, + { + "auxiliary_loss_clip": 0.0121686, + "auxiliary_loss_mlp": 0.0104169, + "balance_loss_clip": 1.06315601, + "balance_loss_mlp": 1.02845812, + "epoch": 0.1450129261107437, + "flos": 27125627955840.0, + "grad_norm": 2.5008676222343795, + "language_loss": 0.74036252, + "learning_rate": 3.862951081147723e-06, + "loss": 0.76294804, + "num_input_tokens_seen": 25613750, + "step": 1206, + "time_per_iteration": 2.8305468559265137 + }, + { + "auxiliary_loss_clip": 0.01232496, + "auxiliary_loss_mlp": 0.01042477, + "balance_loss_clip": 1.06956398, + "balance_loss_mlp": 1.02982855, + "epoch": 0.1451331690013828, + "flos": 25702344472320.0, + "grad_norm": 2.095604980019857, + "language_loss": 0.78568554, + "learning_rate": 3.862667546885053e-06, + "loss": 0.80843532, + "num_input_tokens_seen": 25632300, + "step": 1207, + "time_per_iteration": 2.7900137901306152 + }, + { + "auxiliary_loss_clip": 0.01225012, + "auxiliary_loss_mlp": 0.01048882, + "balance_loss_clip": 1.0658052, + "balance_loss_mlp": 1.0351553, + "epoch": 0.14525341189202187, + "flos": 25737393168000.0, + "grad_norm": 1.9062653338154847, + "language_loss": 0.73784494, + "learning_rate": 3.8623837300571965e-06, + "loss": 0.76058388, + "num_input_tokens_seen": 25651285, + "step": 1208, + "time_per_iteration": 2.970167398452759 + }, + { + "auxiliary_loss_clip": 0.01240106, + "auxiliary_loss_mlp": 0.01040833, + "balance_loss_clip": 1.06754827, + "balance_loss_mlp": 1.02738059, + "epoch": 0.14537365478266098, + "flos": 23073898844160.0, + "grad_norm": 1.8507796164984398, + "language_loss": 0.84222126, + "learning_rate": 3.8620996307072085e-06, + "loss": 0.86503065, + "num_input_tokens_seen": 25671990, + "step": 1209, + "time_per_iteration": 2.6689884662628174 + }, + { + "auxiliary_loss_clip": 0.01210023, + "auxiliary_loss_mlp": 0.01048132, + "balance_loss_clip": 1.06014156, + "balance_loss_mlp": 1.03444076, + "epoch": 0.14549389767330007, + "flos": 20595021448320.0, + "grad_norm": 1.7779467473114432, + "language_loss": 0.64279687, + "learning_rate": 3.861815248878188e-06, + "loss": 0.66537845, + "num_input_tokens_seen": 25689475, + "step": 1210, + "time_per_iteration": 2.829664945602417 + }, + { + "auxiliary_loss_clip": 0.01214734, + "auxiliary_loss_mlp": 0.01036996, + "balance_loss_clip": 1.06686544, + "balance_loss_mlp": 1.0248425, + "epoch": 0.14561414056393915, + "flos": 15121804533120.0, + "grad_norm": 2.7218545418413407, + "language_loss": 0.79636598, + "learning_rate": 3.861530584613274e-06, + "loss": 0.8188833, + "num_input_tokens_seen": 25707475, + "step": 1211, + "time_per_iteration": 2.7304372787475586 + }, + { + "auxiliary_loss_clip": 0.01231382, + "auxiliary_loss_mlp": 0.00889956, + "balance_loss_clip": 1.06590199, + "balance_loss_mlp": 1.00028944, + "epoch": 0.14573438345457826, + "flos": 19427493778560.0, + "grad_norm": 3.8697360111284222, + "language_loss": 0.82687575, + "learning_rate": 3.86124563795565e-06, + "loss": 0.8480891, + "num_input_tokens_seen": 25726290, + "step": 1212, + "time_per_iteration": 2.7909672260284424 + }, + { + "auxiliary_loss_clip": 0.01237763, + "auxiliary_loss_mlp": 0.01047141, + "balance_loss_clip": 1.06739688, + "balance_loss_mlp": 1.03460681, + "epoch": 0.14585462634521734, + "flos": 24828422572800.0, + "grad_norm": 2.0040976679311924, + "language_loss": 0.70220655, + "learning_rate": 3.860960408948543e-06, + "loss": 0.72505558, + "num_input_tokens_seen": 25748040, + "step": 1213, + "time_per_iteration": 2.8009438514709473 + }, + { + "auxiliary_loss_clip": 0.0122272, + "auxiliary_loss_mlp": 0.0103999, + "balance_loss_clip": 1.06942964, + "balance_loss_mlp": 1.02810538, + "epoch": 0.14597486923585642, + "flos": 15448627405440.0, + "grad_norm": 2.490225788515729, + "language_loss": 0.89838243, + "learning_rate": 3.860674897635222e-06, + "loss": 0.92100954, + "num_input_tokens_seen": 25764525, + "step": 1214, + "time_per_iteration": 2.6597964763641357 + }, + { + "auxiliary_loss_clip": 0.01229374, + "auxiliary_loss_mlp": 0.01046231, + "balance_loss_clip": 1.06675482, + "balance_loss_mlp": 1.03342187, + "epoch": 0.1460951121264955, + "flos": 16655154266880.0, + "grad_norm": 1.96030532849945, + "language_loss": 0.83459431, + "learning_rate": 3.860389104058998e-06, + "loss": 0.85735035, + "num_input_tokens_seen": 25782755, + "step": 1215, + "time_per_iteration": 3.669438123703003 + }, + { + "auxiliary_loss_clip": 0.0121979, + "auxiliary_loss_mlp": 0.01036442, + "balance_loss_clip": 1.06402898, + "balance_loss_mlp": 1.0243603, + "epoch": 0.14621535501713462, + "flos": 24863291700480.0, + "grad_norm": 3.789740179659038, + "language_loss": 0.72647285, + "learning_rate": 3.860103028263227e-06, + "loss": 0.74903524, + "num_input_tokens_seen": 25805860, + "step": 1216, + "time_per_iteration": 4.7079081535339355 + }, + { + "auxiliary_loss_clip": 0.01188462, + "auxiliary_loss_mlp": 0.01042008, + "balance_loss_clip": 1.05556035, + "balance_loss_mlp": 1.02692771, + "epoch": 0.1463355979077737, + "flos": 25228000442880.0, + "grad_norm": 2.0820154565584277, + "language_loss": 0.70635068, + "learning_rate": 3.859816670291304e-06, + "loss": 0.72865546, + "num_input_tokens_seen": 25824955, + "step": 1217, + "time_per_iteration": 3.816027879714966 + }, + { + "auxiliary_loss_clip": 0.01174331, + "auxiliary_loss_mlp": 0.01041627, + "balance_loss_clip": 1.05838943, + "balance_loss_mlp": 1.02857327, + "epoch": 0.14645584079841278, + "flos": 22054143726720.0, + "grad_norm": 2.054820233777038, + "language_loss": 0.90001112, + "learning_rate": 3.859530030186672e-06, + "loss": 0.9221707, + "num_input_tokens_seen": 25841965, + "step": 1218, + "time_per_iteration": 2.915442943572998 + }, + { + "auxiliary_loss_clip": 0.01221385, + "auxiliary_loss_mlp": 0.01048411, + "balance_loss_clip": 1.06592333, + "balance_loss_mlp": 1.03340292, + "epoch": 0.1465760836890519, + "flos": 23623870959360.0, + "grad_norm": 2.8436060245523693, + "language_loss": 0.82828099, + "learning_rate": 3.859243107992813e-06, + "loss": 0.85097891, + "num_input_tokens_seen": 25860770, + "step": 1219, + "time_per_iteration": 2.828697443008423 + }, + { + "auxiliary_loss_clip": 0.01206538, + "auxiliary_loss_mlp": 0.01048007, + "balance_loss_clip": 1.05606151, + "balance_loss_mlp": 1.03417909, + "epoch": 0.14669632657969098, + "flos": 37407893356800.0, + "grad_norm": 2.853290779084078, + "language_loss": 0.78047019, + "learning_rate": 3.858955903753252e-06, + "loss": 0.80301559, + "num_input_tokens_seen": 25879410, + "step": 1220, + "time_per_iteration": 2.9379377365112305 + }, + { + "auxiliary_loss_clip": 0.01228998, + "auxiliary_loss_mlp": 0.01039603, + "balance_loss_clip": 1.06487513, + "balance_loss_mlp": 1.02706814, + "epoch": 0.14681656947033006, + "flos": 28365910623360.0, + "grad_norm": 1.4840512069824106, + "language_loss": 0.83576441, + "learning_rate": 3.858668417511559e-06, + "loss": 0.85845041, + "num_input_tokens_seen": 25902160, + "step": 1221, + "time_per_iteration": 2.8612074851989746 + }, + { + "auxiliary_loss_clip": 0.01221155, + "auxiliary_loss_mlp": 0.01047986, + "balance_loss_clip": 1.06584334, + "balance_loss_mlp": 1.03506947, + "epoch": 0.14693681236096917, + "flos": 18479488078080.0, + "grad_norm": 2.4328691452028415, + "language_loss": 0.76528764, + "learning_rate": 3.8583806493113445e-06, + "loss": 0.78797907, + "num_input_tokens_seen": 25920505, + "step": 1222, + "time_per_iteration": 2.867345094680786 + }, + { + "auxiliary_loss_clip": 0.01227797, + "auxiliary_loss_mlp": 0.0104369, + "balance_loss_clip": 1.06821966, + "balance_loss_mlp": 1.03028512, + "epoch": 0.14705705525160825, + "flos": 20777806782720.0, + "grad_norm": 2.38903835728879, + "language_loss": 0.82552272, + "learning_rate": 3.858092599196263e-06, + "loss": 0.84823757, + "num_input_tokens_seen": 25938460, + "step": 1223, + "time_per_iteration": 2.7119739055633545 + }, + { + "auxiliary_loss_clip": 0.0122648, + "auxiliary_loss_mlp": 0.01046163, + "balance_loss_clip": 1.06630003, + "balance_loss_mlp": 1.03397977, + "epoch": 0.14717729814224734, + "flos": 29932944336000.0, + "grad_norm": 3.4937673691483355, + "language_loss": 0.82687557, + "learning_rate": 3.857804267210012e-06, + "loss": 0.84960198, + "num_input_tokens_seen": 25957760, + "step": 1224, + "time_per_iteration": 2.85760498046875 + }, + { + "auxiliary_loss_clip": 0.01193089, + "auxiliary_loss_mlp": 0.01045252, + "balance_loss_clip": 1.05697024, + "balance_loss_mlp": 1.03256822, + "epoch": 0.14729754103288642, + "flos": 20047491457920.0, + "grad_norm": 2.013327797957913, + "language_loss": 0.88250375, + "learning_rate": 3.857515653396331e-06, + "loss": 0.90488714, + "num_input_tokens_seen": 25974970, + "step": 1225, + "time_per_iteration": 2.7592594623565674 + }, + { + "auxiliary_loss_clip": 0.01195508, + "auxiliary_loss_mlp": 0.0103653, + "balance_loss_clip": 1.05961061, + "balance_loss_mlp": 1.02429366, + "epoch": 0.14741778392352553, + "flos": 19281516906240.0, + "grad_norm": 4.2182167816071034, + "language_loss": 0.86774683, + "learning_rate": 3.857226757799002e-06, + "loss": 0.89006722, + "num_input_tokens_seen": 25992525, + "step": 1226, + "time_per_iteration": 2.8025641441345215 + }, + { + "auxiliary_loss_clip": 0.01213677, + "auxiliary_loss_mlp": 0.01044387, + "balance_loss_clip": 1.06018806, + "balance_loss_mlp": 1.03095222, + "epoch": 0.1475380268141646, + "flos": 25411108999680.0, + "grad_norm": 2.322110692832406, + "language_loss": 0.73832768, + "learning_rate": 3.85693758046185e-06, + "loss": 0.76090831, + "num_input_tokens_seen": 26010815, + "step": 1227, + "time_per_iteration": 2.7866299152374268 + }, + { + "auxiliary_loss_clip": 0.01238128, + "auxiliary_loss_mlp": 0.01046048, + "balance_loss_clip": 1.06982422, + "balance_loss_mlp": 1.03363276, + "epoch": 0.1476582697048037, + "flos": 20847652778880.0, + "grad_norm": 1.7802186836893323, + "language_loss": 0.82454342, + "learning_rate": 3.8566481214287435e-06, + "loss": 0.84738517, + "num_input_tokens_seen": 26028935, + "step": 1228, + "time_per_iteration": 2.6996877193450928 + }, + { + "auxiliary_loss_clip": 0.01195985, + "auxiliary_loss_mlp": 0.01044148, + "balance_loss_clip": 1.05576158, + "balance_loss_mlp": 1.03175008, + "epoch": 0.1477785125954428, + "flos": 14028109269120.0, + "grad_norm": 4.180478539773092, + "language_loss": 0.91075462, + "learning_rate": 3.8563583807435935e-06, + "loss": 0.93315595, + "num_input_tokens_seen": 26045080, + "step": 1229, + "time_per_iteration": 2.7940688133239746 + }, + { + "auxiliary_loss_clip": 0.01229823, + "auxiliary_loss_mlp": 0.00889574, + "balance_loss_clip": 1.06630433, + "balance_loss_mlp": 1.00026178, + "epoch": 0.1478987554860819, + "flos": 20516699842560.0, + "grad_norm": 2.1246863239724183, + "language_loss": 0.77684247, + "learning_rate": 3.856068358450353e-06, + "loss": 0.79803646, + "num_input_tokens_seen": 26065030, + "step": 1230, + "time_per_iteration": 2.7987935543060303 + }, + { + "auxiliary_loss_clip": 0.01216129, + "auxiliary_loss_mlp": 0.0103781, + "balance_loss_clip": 1.06918049, + "balance_loss_mlp": 1.02522194, + "epoch": 0.14801899837672097, + "flos": 17857012360320.0, + "grad_norm": 1.6883524257042362, + "language_loss": 0.85749274, + "learning_rate": 3.8557780545930186e-06, + "loss": 0.88003212, + "num_input_tokens_seen": 26083445, + "step": 1231, + "time_per_iteration": 2.788973569869995 + }, + { + "auxiliary_loss_clip": 0.01214296, + "auxiliary_loss_mlp": 0.01041713, + "balance_loss_clip": 1.06545937, + "balance_loss_mlp": 1.02893949, + "epoch": 0.14813924126736006, + "flos": 20881408584960.0, + "grad_norm": 2.0925555162734817, + "language_loss": 0.79522085, + "learning_rate": 3.855487469215628e-06, + "loss": 0.81778097, + "num_input_tokens_seen": 26102375, + "step": 1232, + "time_per_iteration": 2.7334561347961426 + }, + { + "auxiliary_loss_clip": 0.01205056, + "auxiliary_loss_mlp": 0.01046694, + "balance_loss_clip": 1.06342137, + "balance_loss_mlp": 1.03392625, + "epoch": 0.14825948415799917, + "flos": 37414070496000.0, + "grad_norm": 2.810666459867504, + "language_loss": 0.72053099, + "learning_rate": 3.855196602362264e-06, + "loss": 0.74304843, + "num_input_tokens_seen": 26125295, + "step": 1233, + "time_per_iteration": 3.014928102493286 + }, + { + "auxiliary_loss_clip": 0.01226678, + "auxiliary_loss_mlp": 0.01040132, + "balance_loss_clip": 1.06587613, + "balance_loss_mlp": 1.02713776, + "epoch": 0.14837972704863825, + "flos": 22014641744640.0, + "grad_norm": 1.9047020008285047, + "language_loss": 0.94031477, + "learning_rate": 3.854905454077051e-06, + "loss": 0.96298277, + "num_input_tokens_seen": 26142905, + "step": 1234, + "time_per_iteration": 2.910959243774414 + }, + { + "auxiliary_loss_clip": 0.01170698, + "auxiliary_loss_mlp": 0.01044751, + "balance_loss_clip": 1.05552983, + "balance_loss_mlp": 1.03130984, + "epoch": 0.14849996993927733, + "flos": 20996323171200.0, + "grad_norm": 2.133460773329784, + "language_loss": 0.87996125, + "learning_rate": 3.854614024404155e-06, + "loss": 0.90211576, + "num_input_tokens_seen": 26161215, + "step": 1235, + "time_per_iteration": 3.090954303741455 + }, + { + "auxiliary_loss_clip": 0.01205129, + "auxiliary_loss_mlp": 0.01039185, + "balance_loss_clip": 1.05967689, + "balance_loss_mlp": 1.02728808, + "epoch": 0.14862021282991644, + "flos": 20047994248320.0, + "grad_norm": 1.8088004165179479, + "language_loss": 0.89243054, + "learning_rate": 3.8543223133877865e-06, + "loss": 0.91487372, + "num_input_tokens_seen": 26179810, + "step": 1236, + "time_per_iteration": 2.8556976318359375 + }, + { + "auxiliary_loss_clip": 0.01193386, + "auxiliary_loss_mlp": 0.01040094, + "balance_loss_clip": 1.05576348, + "balance_loss_mlp": 1.02682006, + "epoch": 0.14874045572055553, + "flos": 22712027276160.0, + "grad_norm": 1.7229465448270234, + "language_loss": 0.88120633, + "learning_rate": 3.854030321072198e-06, + "loss": 0.90354109, + "num_input_tokens_seen": 26199715, + "step": 1237, + "time_per_iteration": 2.8579773902893066 + }, + { + "auxiliary_loss_clip": 0.01208867, + "auxiliary_loss_mlp": 0.01034985, + "balance_loss_clip": 1.06102943, + "balance_loss_mlp": 1.02212262, + "epoch": 0.1488606986111946, + "flos": 25411288567680.0, + "grad_norm": 2.172066813683627, + "language_loss": 0.73595679, + "learning_rate": 3.853738047501682e-06, + "loss": 0.75839531, + "num_input_tokens_seen": 26220275, + "step": 1238, + "time_per_iteration": 2.9585206508636475 + }, + { + "auxiliary_loss_clip": 0.01227024, + "auxiliary_loss_mlp": 0.01037323, + "balance_loss_clip": 1.0662868, + "balance_loss_mlp": 1.02481806, + "epoch": 0.1489809415018337, + "flos": 17018749687680.0, + "grad_norm": 2.4918015929508064, + "language_loss": 0.77687657, + "learning_rate": 3.85344549272058e-06, + "loss": 0.79952002, + "num_input_tokens_seen": 26238255, + "step": 1239, + "time_per_iteration": 2.7152421474456787 + }, + { + "auxiliary_loss_clip": 0.01217814, + "auxiliary_loss_mlp": 0.01042409, + "balance_loss_clip": 1.0601716, + "balance_loss_mlp": 1.03008318, + "epoch": 0.1491011843924728, + "flos": 33659394860160.0, + "grad_norm": 1.7527560178726902, + "language_loss": 0.82243633, + "learning_rate": 3.853152656773269e-06, + "loss": 0.84503859, + "num_input_tokens_seen": 26259690, + "step": 1240, + "time_per_iteration": 2.8596110343933105 + }, + { + "auxiliary_loss_clip": 0.01208939, + "auxiliary_loss_mlp": 0.01038255, + "balance_loss_clip": 1.05984402, + "balance_loss_mlp": 1.02589941, + "epoch": 0.14922142728311188, + "flos": 21179000764800.0, + "grad_norm": 1.7208776007472915, + "language_loss": 0.85194397, + "learning_rate": 3.852859539704174e-06, + "loss": 0.87441587, + "num_input_tokens_seen": 26278990, + "step": 1241, + "time_per_iteration": 3.783085346221924 + }, + { + "auxiliary_loss_clip": 0.01190379, + "auxiliary_loss_mlp": 0.01043392, + "balance_loss_clip": 1.05681181, + "balance_loss_mlp": 1.03122091, + "epoch": 0.14934167017375097, + "flos": 29860548474240.0, + "grad_norm": 2.1581209486112303, + "language_loss": 0.76173151, + "learning_rate": 3.85256614155776e-06, + "loss": 0.78406918, + "num_input_tokens_seen": 26299120, + "step": 1242, + "time_per_iteration": 3.8268983364105225 + }, + { + "auxiliary_loss_clip": 0.01224288, + "auxiliary_loss_mlp": 0.01047752, + "balance_loss_clip": 1.06271613, + "balance_loss_mlp": 1.03427553, + "epoch": 0.14946191306439008, + "flos": 17019216564480.0, + "grad_norm": 1.9975124197688905, + "language_loss": 0.74407542, + "learning_rate": 3.852272462378535e-06, + "loss": 0.76679575, + "num_input_tokens_seen": 26316995, + "step": 1243, + "time_per_iteration": 3.6070988178253174 + }, + { + "auxiliary_loss_clip": 0.01214511, + "auxiliary_loss_mlp": 0.01044859, + "balance_loss_clip": 1.06456375, + "balance_loss_mlp": 1.03245521, + "epoch": 0.14958215595502916, + "flos": 15669047214720.0, + "grad_norm": 2.317132047191413, + "language_loss": 0.78005648, + "learning_rate": 3.85197850221105e-06, + "loss": 0.80265015, + "num_input_tokens_seen": 26333295, + "step": 1244, + "time_per_iteration": 2.8092753887176514 + }, + { + "auxiliary_loss_clip": 0.01220346, + "auxiliary_loss_mlp": 0.01035954, + "balance_loss_clip": 1.06582594, + "balance_loss_mlp": 1.02424765, + "epoch": 0.14970239884566824, + "flos": 33108560818560.0, + "grad_norm": 1.8862165481216442, + "language_loss": 0.76048076, + "learning_rate": 3.851684261099899e-06, + "loss": 0.78304374, + "num_input_tokens_seen": 26355035, + "step": 1245, + "time_per_iteration": 2.789294719696045 + }, + { + "auxiliary_loss_clip": 0.01203417, + "auxiliary_loss_mlp": 0.01042215, + "balance_loss_clip": 1.05709672, + "balance_loss_mlp": 1.02858925, + "epoch": 0.14982264173630733, + "flos": 17821245392640.0, + "grad_norm": 1.9092056371017878, + "language_loss": 0.86615837, + "learning_rate": 3.851389739089718e-06, + "loss": 0.88861471, + "num_input_tokens_seen": 26371655, + "step": 1246, + "time_per_iteration": 2.7648658752441406 + }, + { + "auxiliary_loss_clip": 0.01226063, + "auxiliary_loss_mlp": 0.01041462, + "balance_loss_clip": 1.06870675, + "balance_loss_mlp": 1.02790165, + "epoch": 0.14994288462694644, + "flos": 32409559175040.0, + "grad_norm": 2.1996515623677806, + "language_loss": 0.80501771, + "learning_rate": 3.851094936225186e-06, + "loss": 0.82769299, + "num_input_tokens_seen": 26392540, + "step": 1247, + "time_per_iteration": 2.8154139518737793 + }, + { + "auxiliary_loss_clip": 0.01207447, + "auxiliary_loss_mlp": 0.010387, + "balance_loss_clip": 1.06427026, + "balance_loss_mlp": 1.02674317, + "epoch": 0.15006312751758552, + "flos": 31794661226880.0, + "grad_norm": 1.6633538317329144, + "language_loss": 0.76672721, + "learning_rate": 3.850799852551024e-06, + "loss": 0.78918868, + "num_input_tokens_seen": 26414960, + "step": 1248, + "time_per_iteration": 2.87284255027771 + }, + { + "auxiliary_loss_clip": 0.01210608, + "auxiliary_loss_mlp": 0.01041803, + "balance_loss_clip": 1.05916595, + "balance_loss_mlp": 1.02849948, + "epoch": 0.1501833704082246, + "flos": 16618022582400.0, + "grad_norm": 2.3969345385342864, + "language_loss": 0.85903955, + "learning_rate": 3.850504488111995e-06, + "loss": 0.88156366, + "num_input_tokens_seen": 26431635, + "step": 1249, + "time_per_iteration": 2.630230188369751 + }, + { + "auxiliary_loss_clip": 0.01208028, + "auxiliary_loss_mlp": 0.01040259, + "balance_loss_clip": 1.06211972, + "balance_loss_mlp": 1.02857637, + "epoch": 0.15030361329886371, + "flos": 23471178243840.0, + "grad_norm": 2.0266384454428708, + "language_loss": 0.82809067, + "learning_rate": 3.850208842952907e-06, + "loss": 0.85057354, + "num_input_tokens_seen": 26450440, + "step": 1250, + "time_per_iteration": 2.9300971031188965 + }, + { + "auxiliary_loss_clip": 0.01203448, + "auxiliary_loss_mlp": 0.01040713, + "balance_loss_clip": 1.06231058, + "balance_loss_mlp": 1.0279398, + "epoch": 0.1504238561895028, + "flos": 25629409906560.0, + "grad_norm": 1.9239537652408403, + "language_loss": 0.79281449, + "learning_rate": 3.849912917118608e-06, + "loss": 0.81525612, + "num_input_tokens_seen": 26471480, + "step": 1251, + "time_per_iteration": 2.860166311264038 + }, + { + "auxiliary_loss_clip": 0.01134443, + "auxiliary_loss_mlp": 0.01013473, + "balance_loss_clip": 1.04464746, + "balance_loss_mlp": 1.00825202, + "epoch": 0.15054409908014188, + "flos": 52095146129280.0, + "grad_norm": 0.8839112068364529, + "language_loss": 0.59289634, + "learning_rate": 3.849616710653992e-06, + "loss": 0.61437553, + "num_input_tokens_seen": 26532950, + "step": 1252, + "time_per_iteration": 3.273015022277832 + }, + { + "auxiliary_loss_clip": 0.01222784, + "auxiliary_loss_mlp": 0.0103745, + "balance_loss_clip": 1.06400633, + "balance_loss_mlp": 1.02526665, + "epoch": 0.150664341970781, + "flos": 18880251096960.0, + "grad_norm": 1.9859779113339366, + "language_loss": 0.74885184, + "learning_rate": 3.84932022360399e-06, + "loss": 0.77145416, + "num_input_tokens_seen": 26551615, + "step": 1253, + "time_per_iteration": 2.7019824981689453 + }, + { + "auxiliary_loss_clip": 0.01211047, + "auxiliary_loss_mlp": 0.01042825, + "balance_loss_clip": 1.06683421, + "balance_loss_mlp": 1.0305469, + "epoch": 0.15078458486142007, + "flos": 22163240309760.0, + "grad_norm": 2.8833544889109803, + "language_loss": 0.84674472, + "learning_rate": 3.849023456013581e-06, + "loss": 0.86928344, + "num_input_tokens_seen": 26569175, + "step": 1254, + "time_per_iteration": 2.738767623901367 + }, + { + "auxiliary_loss_clip": 0.0122784, + "auxiliary_loss_mlp": 0.01044712, + "balance_loss_clip": 1.06397426, + "balance_loss_mlp": 1.03224897, + "epoch": 0.15090482775205916, + "flos": 26651894457600.0, + "grad_norm": 2.3283031753687227, + "language_loss": 0.62286073, + "learning_rate": 3.848726407927784e-06, + "loss": 0.64558625, + "num_input_tokens_seen": 26589560, + "step": 1255, + "time_per_iteration": 2.838686943054199 + }, + { + "auxiliary_loss_clip": 0.01217198, + "auxiliary_loss_mlp": 0.01035235, + "balance_loss_clip": 1.06754136, + "balance_loss_mlp": 1.02280104, + "epoch": 0.15102507064269824, + "flos": 21798998444160.0, + "grad_norm": 2.401730236564763, + "language_loss": 0.86462903, + "learning_rate": 3.84842907939166e-06, + "loss": 0.88715339, + "num_input_tokens_seen": 26608785, + "step": 1256, + "time_per_iteration": 2.7422361373901367 + }, + { + "auxiliary_loss_clip": 0.0119411, + "auxiliary_loss_mlp": 0.01036451, + "balance_loss_clip": 1.06022811, + "balance_loss_mlp": 1.0248456, + "epoch": 0.15114531353333735, + "flos": 22820908377600.0, + "grad_norm": 2.3145408054071686, + "language_loss": 0.71432173, + "learning_rate": 3.8481314704503146e-06, + "loss": 0.7366274, + "num_input_tokens_seen": 26628615, + "step": 1257, + "time_per_iteration": 2.9352777004241943 + }, + { + "auxiliary_loss_clip": 0.01221796, + "auxiliary_loss_mlp": 0.01044228, + "balance_loss_clip": 1.06758571, + "balance_loss_mlp": 1.0322175, + "epoch": 0.15126555642397643, + "flos": 19682674974720.0, + "grad_norm": 2.2946906768761, + "language_loss": 0.88662279, + "learning_rate": 3.847833581148895e-06, + "loss": 0.90928304, + "num_input_tokens_seen": 26647525, + "step": 1258, + "time_per_iteration": 2.798933506011963 + }, + { + "auxiliary_loss_clip": 0.01231722, + "auxiliary_loss_mlp": 0.01040126, + "balance_loss_clip": 1.06243539, + "balance_loss_mlp": 1.02692986, + "epoch": 0.15138579931461552, + "flos": 28726022424960.0, + "grad_norm": 2.422637428193649, + "language_loss": 0.80856657, + "learning_rate": 3.84753541153259e-06, + "loss": 0.83128512, + "num_input_tokens_seen": 26667095, + "step": 1259, + "time_per_iteration": 2.7131803035736084 + }, + { + "auxiliary_loss_clip": 0.01223329, + "auxiliary_loss_mlp": 0.01035878, + "balance_loss_clip": 1.06532788, + "balance_loss_mlp": 1.02420187, + "epoch": 0.15150604220525463, + "flos": 22127006465280.0, + "grad_norm": 1.652297959787507, + "language_loss": 0.83065766, + "learning_rate": 3.847236961646633e-06, + "loss": 0.85324973, + "num_input_tokens_seen": 26686075, + "step": 1260, + "time_per_iteration": 2.758850574493408 + }, + { + "auxiliary_loss_clip": 0.01208547, + "auxiliary_loss_mlp": 0.01051928, + "balance_loss_clip": 1.06091619, + "balance_loss_mlp": 1.03897548, + "epoch": 0.1516262850958937, + "flos": 12968708515200.0, + "grad_norm": 2.882006783214971, + "language_loss": 0.78149319, + "learning_rate": 3.846938231536296e-06, + "loss": 0.80409795, + "num_input_tokens_seen": 26701695, + "step": 1261, + "time_per_iteration": 2.686199426651001 + }, + { + "auxiliary_loss_clip": 0.01227845, + "auxiliary_loss_mlp": 0.01043831, + "balance_loss_clip": 1.06693256, + "balance_loss_mlp": 1.03078938, + "epoch": 0.1517465279865328, + "flos": 21797130936960.0, + "grad_norm": 1.787301123635627, + "language_loss": 0.80668479, + "learning_rate": 3.8466392212468995e-06, + "loss": 0.82940155, + "num_input_tokens_seen": 26721885, + "step": 1262, + "time_per_iteration": 2.7560019493103027 + }, + { + "auxiliary_loss_clip": 0.01116427, + "auxiliary_loss_mlp": 0.01007322, + "balance_loss_clip": 1.03534508, + "balance_loss_mlp": 1.00195718, + "epoch": 0.15186677087717187, + "flos": 58174569901440.0, + "grad_norm": 0.814500062292813, + "language_loss": 0.61895305, + "learning_rate": 3.8463399308238e-06, + "loss": 0.64019054, + "num_input_tokens_seen": 26780990, + "step": 1263, + "time_per_iteration": 3.2919154167175293 + }, + { + "auxiliary_loss_clip": 0.01221258, + "auxiliary_loss_mlp": 0.01038257, + "balance_loss_clip": 1.0649879, + "balance_loss_mlp": 1.02577031, + "epoch": 0.15198701376781099, + "flos": 32669696448000.0, + "grad_norm": 1.92765022856974, + "language_loss": 0.64019012, + "learning_rate": 3.846040360312402e-06, + "loss": 0.66278529, + "num_input_tokens_seen": 26804250, + "step": 1264, + "time_per_iteration": 2.8729567527770996 + }, + { + "auxiliary_loss_clip": 0.01233966, + "auxiliary_loss_mlp": 0.01031013, + "balance_loss_clip": 1.06588519, + "balance_loss_mlp": 1.01875865, + "epoch": 0.15210725665845007, + "flos": 28402575431040.0, + "grad_norm": 2.049725669225994, + "language_loss": 0.81270826, + "learning_rate": 3.8457405097581485e-06, + "loss": 0.83535814, + "num_input_tokens_seen": 26823240, + "step": 1265, + "time_per_iteration": 2.671912670135498 + }, + { + "auxiliary_loss_clip": 0.01193073, + "auxiliary_loss_mlp": 0.01045295, + "balance_loss_clip": 1.05536318, + "balance_loss_mlp": 1.03262877, + "epoch": 0.15222749954908915, + "flos": 19938179393280.0, + "grad_norm": 2.1199498358608713, + "language_loss": 0.78371388, + "learning_rate": 3.8454403792065275e-06, + "loss": 0.80609751, + "num_input_tokens_seen": 26842060, + "step": 1266, + "time_per_iteration": 2.8217251300811768 + }, + { + "auxiliary_loss_clip": 0.01195776, + "auxiliary_loss_mlp": 0.01040734, + "balance_loss_clip": 1.06128168, + "balance_loss_mlp": 1.02878904, + "epoch": 0.15234774243972826, + "flos": 21324223451520.0, + "grad_norm": 1.9927618044405222, + "language_loss": 0.85563695, + "learning_rate": 3.845139968703068e-06, + "loss": 0.87800205, + "num_input_tokens_seen": 26859580, + "step": 1267, + "time_per_iteration": 3.7706539630889893 + }, + { + "auxiliary_loss_clip": 0.01189521, + "auxiliary_loss_mlp": 0.01041084, + "balance_loss_clip": 1.05593884, + "balance_loss_mlp": 1.02808428, + "epoch": 0.15246798533036734, + "flos": 25957812977280.0, + "grad_norm": 3.339667397157656, + "language_loss": 0.83168131, + "learning_rate": 3.844839278293342e-06, + "loss": 0.8539874, + "num_input_tokens_seen": 26880430, + "step": 1268, + "time_per_iteration": 3.839219808578491 + }, + { + "auxiliary_loss_clip": 0.01235308, + "auxiliary_loss_mlp": 0.01038488, + "balance_loss_clip": 1.06643176, + "balance_loss_mlp": 1.02565503, + "epoch": 0.15258822822100643, + "flos": 25811907932160.0, + "grad_norm": 2.0419736740570213, + "language_loss": 0.77433658, + "learning_rate": 3.8445383080229654e-06, + "loss": 0.79707456, + "num_input_tokens_seen": 26896445, + "step": 1269, + "time_per_iteration": 3.6415998935699463 + }, + { + "auxiliary_loss_clip": 0.01202066, + "auxiliary_loss_mlp": 0.01044774, + "balance_loss_clip": 1.05630708, + "balance_loss_mlp": 1.03151774, + "epoch": 0.1527084711116455, + "flos": 25265455349760.0, + "grad_norm": 2.768300698528739, + "language_loss": 0.73688817, + "learning_rate": 3.844237057937593e-06, + "loss": 0.75935662, + "num_input_tokens_seen": 26915450, + "step": 1270, + "time_per_iteration": 2.8373730182647705 + }, + { + "auxiliary_loss_clip": 0.01224229, + "auxiliary_loss_mlp": 0.01046175, + "balance_loss_clip": 1.06200004, + "balance_loss_mlp": 1.03312182, + "epoch": 0.15282871400228462, + "flos": 29240227572480.0, + "grad_norm": 2.14507963071148, + "language_loss": 0.77641881, + "learning_rate": 3.843935528082926e-06, + "loss": 0.79912287, + "num_input_tokens_seen": 26936475, + "step": 1271, + "time_per_iteration": 2.763481855392456 + }, + { + "auxiliary_loss_clip": 0.0122256, + "auxiliary_loss_mlp": 0.01043909, + "balance_loss_clip": 1.06225634, + "balance_loss_mlp": 1.0311836, + "epoch": 0.1529489568929237, + "flos": 20882952869760.0, + "grad_norm": 2.1470588486142663, + "language_loss": 0.8472572, + "learning_rate": 3.843633718504704e-06, + "loss": 0.86992192, + "num_input_tokens_seen": 26954920, + "step": 1272, + "time_per_iteration": 2.6907503604888916 + }, + { + "auxiliary_loss_clip": 0.01202557, + "auxiliary_loss_mlp": 0.01041658, + "balance_loss_clip": 1.06089067, + "balance_loss_mlp": 1.02875996, + "epoch": 0.1530691997835628, + "flos": 20083833043200.0, + "grad_norm": 2.406877343250315, + "language_loss": 0.90233469, + "learning_rate": 3.843331629248715e-06, + "loss": 0.92477685, + "num_input_tokens_seen": 26972520, + "step": 1273, + "time_per_iteration": 2.7784526348114014 + }, + { + "auxiliary_loss_clip": 0.01231766, + "auxiliary_loss_mlp": 0.01047366, + "balance_loss_clip": 1.06677532, + "balance_loss_mlp": 1.03569579, + "epoch": 0.1531894426742019, + "flos": 28759814144640.0, + "grad_norm": 2.4620954767481322, + "language_loss": 0.7641027, + "learning_rate": 3.843029260360782e-06, + "loss": 0.78689396, + "num_input_tokens_seen": 26990890, + "step": 1274, + "time_per_iteration": 2.7499442100524902 + }, + { + "auxiliary_loss_clip": 0.01216636, + "auxiliary_loss_mlp": 0.01037877, + "balance_loss_clip": 1.06181002, + "balance_loss_mlp": 1.02582514, + "epoch": 0.15330968556484098, + "flos": 22236282616320.0, + "grad_norm": 1.9934575891814952, + "language_loss": 0.79048038, + "learning_rate": 3.8427266118867755e-06, + "loss": 0.81302553, + "num_input_tokens_seen": 27010640, + "step": 1275, + "time_per_iteration": 2.732208728790283 + }, + { + "auxiliary_loss_clip": 0.01210685, + "auxiliary_loss_mlp": 0.0104184, + "balance_loss_clip": 1.06173897, + "balance_loss_mlp": 1.02914989, + "epoch": 0.15342992845548006, + "flos": 27527504296320.0, + "grad_norm": 2.0092727320122776, + "language_loss": 0.82892203, + "learning_rate": 3.842423683872608e-06, + "loss": 0.85144722, + "num_input_tokens_seen": 27031215, + "step": 1276, + "time_per_iteration": 2.7573461532592773 + }, + { + "auxiliary_loss_clip": 0.01219699, + "auxiliary_loss_mlp": 0.01038205, + "balance_loss_clip": 1.06135225, + "balance_loss_mlp": 1.02488911, + "epoch": 0.15355017134611917, + "flos": 19609596754560.0, + "grad_norm": 2.417283046459591, + "language_loss": 0.78036761, + "learning_rate": 3.842120476364232e-06, + "loss": 0.80294669, + "num_input_tokens_seen": 27049665, + "step": 1277, + "time_per_iteration": 2.74166202545166 + }, + { + "auxiliary_loss_clip": 0.01224692, + "auxiliary_loss_mlp": 0.01041829, + "balance_loss_clip": 1.06036794, + "balance_loss_mlp": 1.02877581, + "epoch": 0.15367041423675826, + "flos": 18478590238080.0, + "grad_norm": 2.1130286520028645, + "language_loss": 0.83655536, + "learning_rate": 3.841816989407644e-06, + "loss": 0.85922062, + "num_input_tokens_seen": 27065155, + "step": 1278, + "time_per_iteration": 2.6600558757781982 + }, + { + "auxiliary_loss_clip": 0.01193285, + "auxiliary_loss_mlp": 0.01039229, + "balance_loss_clip": 1.05595493, + "balance_loss_mlp": 1.02758217, + "epoch": 0.15379065712739734, + "flos": 41427662342400.0, + "grad_norm": 2.737347719316161, + "language_loss": 0.76939905, + "learning_rate": 3.841513223048884e-06, + "loss": 0.79172415, + "num_input_tokens_seen": 27085840, + "step": 1279, + "time_per_iteration": 3.0081002712249756 + }, + { + "auxiliary_loss_clip": 0.01198586, + "auxiliary_loss_mlp": 0.01036198, + "balance_loss_clip": 1.0574944, + "balance_loss_mlp": 1.02386606, + "epoch": 0.15391090001803642, + "flos": 22054215553920.0, + "grad_norm": 2.397666383632697, + "language_loss": 0.78610897, + "learning_rate": 3.841209177334031e-06, + "loss": 0.80845678, + "num_input_tokens_seen": 27104200, + "step": 1280, + "time_per_iteration": 2.755502700805664 + }, + { + "auxiliary_loss_clip": 0.01217896, + "auxiliary_loss_mlp": 0.01044042, + "balance_loss_clip": 1.06205654, + "balance_loss_mlp": 1.03199029, + "epoch": 0.15403114290867553, + "flos": 15450351258240.0, + "grad_norm": 1.9450495283520006, + "language_loss": 0.74845302, + "learning_rate": 3.84090485230921e-06, + "loss": 0.77107239, + "num_input_tokens_seen": 27122440, + "step": 1281, + "time_per_iteration": 2.68007493019104 + }, + { + "auxiliary_loss_clip": 0.01233667, + "auxiliary_loss_mlp": 0.01039894, + "balance_loss_clip": 1.06561041, + "balance_loss_mlp": 1.02749014, + "epoch": 0.15415138579931462, + "flos": 17929156826880.0, + "grad_norm": 4.030233945331351, + "language_loss": 0.76386249, + "learning_rate": 3.840600248020588e-06, + "loss": 0.78659815, + "num_input_tokens_seen": 27139380, + "step": 1282, + "time_per_iteration": 2.583329677581787 + }, + { + "auxiliary_loss_clip": 0.01212628, + "auxiliary_loss_mlp": 0.01040879, + "balance_loss_clip": 1.05610037, + "balance_loss_mlp": 1.02823675, + "epoch": 0.1542716286899537, + "flos": 11429325296640.0, + "grad_norm": 31.663046160599503, + "language_loss": 0.79883236, + "learning_rate": 3.840295364514371e-06, + "loss": 0.82136744, + "num_input_tokens_seen": 27156760, + "step": 1283, + "time_per_iteration": 2.766342878341675 + }, + { + "auxiliary_loss_clip": 0.01210357, + "auxiliary_loss_mlp": 0.01040156, + "balance_loss_clip": 1.06041145, + "balance_loss_mlp": 1.02784145, + "epoch": 0.1543918715805928, + "flos": 17420338719360.0, + "grad_norm": 5.221622931478999, + "language_loss": 0.78753209, + "learning_rate": 3.83999020183681e-06, + "loss": 0.8100372, + "num_input_tokens_seen": 27175455, + "step": 1284, + "time_per_iteration": 2.7400529384613037 + }, + { + "auxiliary_loss_clip": 0.01175531, + "auxiliary_loss_mlp": 0.01042608, + "balance_loss_clip": 1.05457544, + "balance_loss_mlp": 1.02965605, + "epoch": 0.1545121144712319, + "flos": 17786376264960.0, + "grad_norm": 2.0671186022634798, + "language_loss": 0.78729856, + "learning_rate": 3.839684760034199e-06, + "loss": 0.80947995, + "num_input_tokens_seen": 27193660, + "step": 1285, + "time_per_iteration": 2.9139420986175537 + }, + { + "auxiliary_loss_clip": 0.011917, + "auxiliary_loss_mlp": 0.01042705, + "balance_loss_clip": 1.05724883, + "balance_loss_mlp": 1.03005075, + "epoch": 0.15463235736187098, + "flos": 28220185146240.0, + "grad_norm": 2.3309326184827577, + "language_loss": 0.65321046, + "learning_rate": 3.8393790391528716e-06, + "loss": 0.67555451, + "num_input_tokens_seen": 27214355, + "step": 1286, + "time_per_iteration": 2.90378737449646 + }, + { + "auxiliary_loss_clip": 0.01204842, + "auxiliary_loss_mlp": 0.0103586, + "balance_loss_clip": 1.05735183, + "balance_loss_mlp": 1.02339053, + "epoch": 0.15475260025251006, + "flos": 22856890826880.0, + "grad_norm": 2.88101464704374, + "language_loss": 0.89249825, + "learning_rate": 3.8390730392392075e-06, + "loss": 0.91490525, + "num_input_tokens_seen": 27234335, + "step": 1287, + "time_per_iteration": 2.8224728107452393 + }, + { + "auxiliary_loss_clip": 0.01232149, + "auxiliary_loss_mlp": 0.01037365, + "balance_loss_clip": 1.06425619, + "balance_loss_mlp": 1.02496767, + "epoch": 0.15487284314314917, + "flos": 17602872658560.0, + "grad_norm": 2.136848827990743, + "language_loss": 0.7930693, + "learning_rate": 3.838766760339626e-06, + "loss": 0.81576443, + "num_input_tokens_seen": 27252860, + "step": 1288, + "time_per_iteration": 2.7097795009613037 + }, + { + "auxiliary_loss_clip": 0.01176088, + "auxiliary_loss_mlp": 0.01038591, + "balance_loss_clip": 1.05061531, + "balance_loss_mlp": 1.02572811, + "epoch": 0.15499308603378825, + "flos": 20082037363200.0, + "grad_norm": 2.944857325864811, + "language_loss": 0.79246497, + "learning_rate": 3.838460202500587e-06, + "loss": 0.81461173, + "num_input_tokens_seen": 27268650, + "step": 1289, + "time_per_iteration": 2.7670650482177734 + }, + { + "auxiliary_loss_clip": 0.01198256, + "auxiliary_loss_mlp": 0.01043264, + "balance_loss_clip": 1.06549978, + "balance_loss_mlp": 1.03081822, + "epoch": 0.15511332892442733, + "flos": 15918051271680.0, + "grad_norm": 3.884152811490073, + "language_loss": 0.73938555, + "learning_rate": 3.838153365768599e-06, + "loss": 0.76180077, + "num_input_tokens_seen": 27285160, + "step": 1290, + "time_per_iteration": 2.8146097660064697 + }, + { + "auxiliary_loss_clip": 0.01202673, + "auxiliary_loss_mlp": 0.01041059, + "balance_loss_clip": 1.06523657, + "balance_loss_mlp": 1.02752852, + "epoch": 0.15523357181506645, + "flos": 41282475569280.0, + "grad_norm": 2.3365252346636467, + "language_loss": 0.75674582, + "learning_rate": 3.837846250190206e-06, + "loss": 0.77918315, + "num_input_tokens_seen": 27308025, + "step": 1291, + "time_per_iteration": 2.9919378757476807 + }, + { + "auxiliary_loss_clip": 0.01181845, + "auxiliary_loss_mlp": 0.00888716, + "balance_loss_clip": 1.05483127, + "balance_loss_mlp": 1.00053346, + "epoch": 0.15535381470570553, + "flos": 18478769806080.0, + "grad_norm": 1.8865471264051143, + "language_loss": 0.77008766, + "learning_rate": 3.837538855811998e-06, + "loss": 0.7907933, + "num_input_tokens_seen": 27326200, + "step": 1292, + "time_per_iteration": 2.9083287715911865 + }, + { + "auxiliary_loss_clip": 0.0121273, + "auxiliary_loss_mlp": 0.01045322, + "balance_loss_clip": 1.06093144, + "balance_loss_mlp": 1.03283477, + "epoch": 0.1554740575963446, + "flos": 13918150759680.0, + "grad_norm": 2.405684068301502, + "language_loss": 0.71155125, + "learning_rate": 3.837231182680606e-06, + "loss": 0.73413169, + "num_input_tokens_seen": 27344165, + "step": 1293, + "time_per_iteration": 4.6443634033203125 + }, + { + "auxiliary_loss_clip": 0.01224348, + "auxiliary_loss_mlp": 0.01041542, + "balance_loss_clip": 1.0626514, + "balance_loss_mlp": 1.03010368, + "epoch": 0.1555943004869837, + "flos": 20847078161280.0, + "grad_norm": 1.6376455983590346, + "language_loss": 0.75841689, + "learning_rate": 3.836923230842706e-06, + "loss": 0.78107578, + "num_input_tokens_seen": 27363280, + "step": 1294, + "time_per_iteration": 2.790461540222168 + }, + { + "auxiliary_loss_clip": 0.01184827, + "auxiliary_loss_mlp": 0.01041761, + "balance_loss_clip": 1.05045605, + "balance_loss_mlp": 1.02892828, + "epoch": 0.1557145433776228, + "flos": 22085888371200.0, + "grad_norm": 2.734724631141997, + "language_loss": 0.81133795, + "learning_rate": 3.836615000345011e-06, + "loss": 0.83360386, + "num_input_tokens_seen": 27381460, + "step": 1295, + "time_per_iteration": 4.81325626373291 + }, + { + "auxiliary_loss_clip": 0.01229072, + "auxiliary_loss_mlp": 0.01037751, + "balance_loss_clip": 1.06336021, + "balance_loss_mlp": 1.0262593, + "epoch": 0.1558347862682619, + "flos": 19791987039360.0, + "grad_norm": 2.8374506640901416, + "language_loss": 0.78078288, + "learning_rate": 3.836306491234282e-06, + "loss": 0.80345112, + "num_input_tokens_seen": 27399310, + "step": 1296, + "time_per_iteration": 2.642982244491577 + }, + { + "auxiliary_loss_clip": 0.01207278, + "auxiliary_loss_mlp": 0.01042791, + "balance_loss_clip": 1.06671464, + "balance_loss_mlp": 1.0307864, + "epoch": 0.15595502915890097, + "flos": 17237086508160.0, + "grad_norm": 2.425853838311709, + "language_loss": 0.75632507, + "learning_rate": 3.835997703557317e-06, + "loss": 0.77882576, + "num_input_tokens_seen": 27416050, + "step": 1297, + "time_per_iteration": 2.774153709411621 + }, + { + "auxiliary_loss_clip": 0.01187976, + "auxiliary_loss_mlp": 0.01039294, + "balance_loss_clip": 1.05243123, + "balance_loss_mlp": 1.02684832, + "epoch": 0.15607527204954008, + "flos": 19719519350400.0, + "grad_norm": 1.7784563356401621, + "language_loss": 0.80002689, + "learning_rate": 3.83568863736096e-06, + "loss": 0.8222996, + "num_input_tokens_seen": 27434920, + "step": 1298, + "time_per_iteration": 2.9188718795776367 + }, + { + "auxiliary_loss_clip": 0.01198819, + "auxiliary_loss_mlp": 0.01040501, + "balance_loss_clip": 1.05500495, + "balance_loss_mlp": 1.02846658, + "epoch": 0.15619551494017916, + "flos": 18515650095360.0, + "grad_norm": 2.277051305842478, + "language_loss": 0.89065397, + "learning_rate": 3.8353792926920975e-06, + "loss": 0.91304708, + "num_input_tokens_seen": 27453570, + "step": 1299, + "time_per_iteration": 2.855633020401001 + }, + { + "auxiliary_loss_clip": 0.0122664, + "auxiliary_loss_mlp": 0.01044629, + "balance_loss_clip": 1.06381392, + "balance_loss_mlp": 1.0321238, + "epoch": 0.15631575783081825, + "flos": 19902125116800.0, + "grad_norm": 2.1783560635555097, + "language_loss": 0.81628716, + "learning_rate": 3.835069669597655e-06, + "loss": 0.83899987, + "num_input_tokens_seen": 27471960, + "step": 1300, + "time_per_iteration": 2.7486696243286133 + }, + { + "auxiliary_loss_clip": 0.01226778, + "auxiliary_loss_mlp": 0.00888465, + "balance_loss_clip": 1.06309175, + "balance_loss_mlp": 1.00036478, + "epoch": 0.15643600072145733, + "flos": 20777663128320.0, + "grad_norm": 1.9515869721844779, + "language_loss": 0.80210418, + "learning_rate": 3.834759768124603e-06, + "loss": 0.82325661, + "num_input_tokens_seen": 27490835, + "step": 1301, + "time_per_iteration": 2.7616031169891357 + }, + { + "auxiliary_loss_clip": 0.01202208, + "auxiliary_loss_mlp": 0.01045057, + "balance_loss_clip": 1.0607686, + "balance_loss_mlp": 1.03254008, + "epoch": 0.15655624361209644, + "flos": 18546389159040.0, + "grad_norm": 4.679590255035991, + "language_loss": 0.76421821, + "learning_rate": 3.834449588319953e-06, + "loss": 0.78669083, + "num_input_tokens_seen": 27508870, + "step": 1302, + "time_per_iteration": 2.763270378112793 + }, + { + "auxiliary_loss_clip": 0.01218507, + "auxiliary_loss_mlp": 0.01035115, + "balance_loss_clip": 1.06703639, + "balance_loss_mlp": 1.02426147, + "epoch": 0.15667648650273552, + "flos": 25229544727680.0, + "grad_norm": 2.4544353808494566, + "language_loss": 0.8527022, + "learning_rate": 3.834139130230758e-06, + "loss": 0.87523842, + "num_input_tokens_seen": 27528175, + "step": 1303, + "time_per_iteration": 2.802480936050415 + }, + { + "auxiliary_loss_clip": 0.01212887, + "auxiliary_loss_mlp": 0.01036949, + "balance_loss_clip": 1.05961013, + "balance_loss_mlp": 1.02555871, + "epoch": 0.1567967293933746, + "flos": 24827093769600.0, + "grad_norm": 1.6546517731702697, + "language_loss": 0.81383991, + "learning_rate": 3.833828393904117e-06, + "loss": 0.83633828, + "num_input_tokens_seen": 27548455, + "step": 1304, + "time_per_iteration": 2.8806729316711426 + }, + { + "auxiliary_loss_clip": 0.01178095, + "auxiliary_loss_mlp": 0.01044718, + "balance_loss_clip": 1.05117118, + "balance_loss_mlp": 1.031569, + "epoch": 0.15691697228401372, + "flos": 19164555244800.0, + "grad_norm": 2.2434951272355956, + "language_loss": 0.77541316, + "learning_rate": 3.833517379387165e-06, + "loss": 0.79764128, + "num_input_tokens_seen": 27564910, + "step": 1305, + "time_per_iteration": 2.8369154930114746 + }, + { + "auxiliary_loss_clip": 0.01223638, + "auxiliary_loss_mlp": 0.01050245, + "balance_loss_clip": 1.06358874, + "balance_loss_mlp": 1.03871155, + "epoch": 0.1570372151746528, + "flos": 24790931752320.0, + "grad_norm": 2.0429357185654937, + "language_loss": 0.89268988, + "learning_rate": 3.833206086727085e-06, + "loss": 0.91542876, + "num_input_tokens_seen": 27584260, + "step": 1306, + "time_per_iteration": 2.779527425765991 + }, + { + "auxiliary_loss_clip": 0.01202786, + "auxiliary_loss_mlp": 0.01036404, + "balance_loss_clip": 1.05667973, + "balance_loss_mlp": 1.02416074, + "epoch": 0.15715745806529188, + "flos": 24863650836480.0, + "grad_norm": 2.03784394679431, + "language_loss": 0.70384276, + "learning_rate": 3.8328945159710994e-06, + "loss": 0.72623467, + "num_input_tokens_seen": 27604440, + "step": 1307, + "time_per_iteration": 2.8500845432281494 + }, + { + "auxiliary_loss_clip": 0.01228007, + "auxiliary_loss_mlp": 0.00888409, + "balance_loss_clip": 1.06618273, + "balance_loss_mlp": 1.00022292, + "epoch": 0.157277700955931, + "flos": 21872148491520.0, + "grad_norm": 1.903296500765648, + "language_loss": 0.88716435, + "learning_rate": 3.832582667166473e-06, + "loss": 0.90832847, + "num_input_tokens_seen": 27624250, + "step": 1308, + "time_per_iteration": 2.8181474208831787 + }, + { + "auxiliary_loss_clip": 0.01203189, + "auxiliary_loss_mlp": 0.01041337, + "balance_loss_clip": 1.05651522, + "balance_loss_mlp": 1.02879667, + "epoch": 0.15739794384657008, + "flos": 24533344344960.0, + "grad_norm": 1.7648371337460202, + "language_loss": 0.81310594, + "learning_rate": 3.8322705403605125e-06, + "loss": 0.83555126, + "num_input_tokens_seen": 27644595, + "step": 1309, + "time_per_iteration": 2.7864291667938232 + }, + { + "auxiliary_loss_clip": 0.01201152, + "auxiliary_loss_mlp": 0.01033181, + "balance_loss_clip": 1.05820608, + "balance_loss_mlp": 1.02222586, + "epoch": 0.15751818673720916, + "flos": 17745329998080.0, + "grad_norm": 2.0512573821938975, + "language_loss": 0.81406105, + "learning_rate": 3.831958135600568e-06, + "loss": 0.83640432, + "num_input_tokens_seen": 27662145, + "step": 1310, + "time_per_iteration": 2.7758681774139404 + }, + { + "auxiliary_loss_clip": 0.01222467, + "auxiliary_loss_mlp": 0.01041059, + "balance_loss_clip": 1.06383014, + "balance_loss_mlp": 1.03003812, + "epoch": 0.15763842962784824, + "flos": 17858520731520.0, + "grad_norm": 1.7965357075751, + "language_loss": 0.79616416, + "learning_rate": 3.831645452934032e-06, + "loss": 0.81879938, + "num_input_tokens_seen": 27680575, + "step": 1311, + "time_per_iteration": 2.7285730838775635 + }, + { + "auxiliary_loss_clip": 0.01230905, + "auxiliary_loss_mlp": 0.01042533, + "balance_loss_clip": 1.06547737, + "balance_loss_mlp": 1.029742, + "epoch": 0.15775867251848735, + "flos": 26980908059520.0, + "grad_norm": 2.0163995101313534, + "language_loss": 0.79765427, + "learning_rate": 3.831332492408336e-06, + "loss": 0.82038867, + "num_input_tokens_seen": 27701985, + "step": 1312, + "time_per_iteration": 2.7296392917633057 + }, + { + "auxiliary_loss_clip": 0.01202094, + "auxiliary_loss_mlp": 0.01033045, + "balance_loss_clip": 1.05792797, + "balance_loss_mlp": 1.02136874, + "epoch": 0.15787891540912644, + "flos": 19240398812160.0, + "grad_norm": 1.9794701521591787, + "language_loss": 0.69265682, + "learning_rate": 3.831019254070957e-06, + "loss": 0.7150082, + "num_input_tokens_seen": 27719770, + "step": 1313, + "time_per_iteration": 2.7547314167022705 + }, + { + "auxiliary_loss_clip": 0.011916, + "auxiliary_loss_mlp": 0.01040788, + "balance_loss_clip": 1.05672407, + "balance_loss_mlp": 1.0285337, + "epoch": 0.15799915829976552, + "flos": 27271102037760.0, + "grad_norm": 2.6140131854437345, + "language_loss": 0.95279998, + "learning_rate": 3.8307057379694135e-06, + "loss": 0.97512382, + "num_input_tokens_seen": 27739105, + "step": 1314, + "time_per_iteration": 2.868438720703125 + }, + { + "auxiliary_loss_clip": 0.01229832, + "auxiliary_loss_mlp": 0.01045412, + "balance_loss_clip": 1.06174242, + "balance_loss_mlp": 1.03269875, + "epoch": 0.15811940119040463, + "flos": 20405520270720.0, + "grad_norm": 2.452860317895828, + "language_loss": 0.82399565, + "learning_rate": 3.830391944151264e-06, + "loss": 0.84674811, + "num_input_tokens_seen": 27754985, + "step": 1315, + "time_per_iteration": 2.6519601345062256 + }, + { + "auxiliary_loss_clip": 0.01212599, + "auxiliary_loss_mlp": 0.01035959, + "balance_loss_clip": 1.06091118, + "balance_loss_mlp": 1.02456236, + "epoch": 0.1582396440810437, + "flos": 32599347661440.0, + "grad_norm": 2.1289260976306656, + "language_loss": 0.6737197, + "learning_rate": 3.830077872664114e-06, + "loss": 0.69620526, + "num_input_tokens_seen": 27776110, + "step": 1316, + "time_per_iteration": 2.814115524291992 + }, + { + "auxiliary_loss_clip": 0.01182168, + "auxiliary_loss_mlp": 0.01042294, + "balance_loss_clip": 1.05672252, + "balance_loss_mlp": 1.03027773, + "epoch": 0.1583598869716828, + "flos": 33800559310080.0, + "grad_norm": 1.9395596965293023, + "language_loss": 0.73065513, + "learning_rate": 3.829763523555604e-06, + "loss": 0.75289965, + "num_input_tokens_seen": 27796510, + "step": 1317, + "time_per_iteration": 2.9858083724975586 + }, + { + "auxiliary_loss_clip": 0.01216603, + "auxiliary_loss_mlp": 0.0104161, + "balance_loss_clip": 1.0682621, + "balance_loss_mlp": 1.03090537, + "epoch": 0.15848012986232188, + "flos": 24681332378880.0, + "grad_norm": 2.261148857864305, + "language_loss": 0.77878362, + "learning_rate": 3.829448896873423e-06, + "loss": 0.80136573, + "num_input_tokens_seen": 27815610, + "step": 1318, + "time_per_iteration": 2.809680223464966 + }, + { + "auxiliary_loss_clip": 0.01180279, + "auxiliary_loss_mlp": 0.00889006, + "balance_loss_clip": 1.06172299, + "balance_loss_mlp": 1.00027585, + "epoch": 0.158600372752961, + "flos": 22602068766720.0, + "grad_norm": 3.9127115628281386, + "language_loss": 0.79336607, + "learning_rate": 3.829133992665299e-06, + "loss": 0.81405896, + "num_input_tokens_seen": 27834735, + "step": 1319, + "time_per_iteration": 4.79924201965332 + }, + { + "auxiliary_loss_clip": 0.01207951, + "auxiliary_loss_mlp": 0.01046479, + "balance_loss_clip": 1.06007302, + "balance_loss_mlp": 1.03453457, + "epoch": 0.15872061564360007, + "flos": 27927944092800.0, + "grad_norm": 2.2479716615571745, + "language_loss": 0.89022118, + "learning_rate": 3.828818810979002e-06, + "loss": 0.9127655, + "num_input_tokens_seen": 27853065, + "step": 1320, + "time_per_iteration": 3.6807851791381836 + }, + { + "auxiliary_loss_clip": 0.01231001, + "auxiliary_loss_mlp": 0.01042757, + "balance_loss_clip": 1.06591475, + "balance_loss_mlp": 1.02976918, + "epoch": 0.15884085853423915, + "flos": 23696805525120.0, + "grad_norm": 1.9053940763413268, + "language_loss": 0.80726582, + "learning_rate": 3.8285033518623454e-06, + "loss": 0.83000338, + "num_input_tokens_seen": 27873315, + "step": 1321, + "time_per_iteration": 3.5720434188842773 + }, + { + "auxiliary_loss_clip": 0.01229146, + "auxiliary_loss_mlp": 0.01040692, + "balance_loss_clip": 1.06608105, + "balance_loss_mlp": 1.02779341, + "epoch": 0.15896110142487826, + "flos": 23112359331840.0, + "grad_norm": 2.8636517410797415, + "language_loss": 0.81115413, + "learning_rate": 3.8281876153631845e-06, + "loss": 0.83385253, + "num_input_tokens_seen": 27890070, + "step": 1322, + "time_per_iteration": 2.705967903137207 + }, + { + "auxiliary_loss_clip": 0.01182739, + "auxiliary_loss_mlp": 0.01041611, + "balance_loss_clip": 1.05492496, + "balance_loss_mlp": 1.0285809, + "epoch": 0.15908134431551735, + "flos": 14685238632960.0, + "grad_norm": 1.8228600236773715, + "language_loss": 0.64776242, + "learning_rate": 3.827871601529416e-06, + "loss": 0.67000592, + "num_input_tokens_seen": 27908590, + "step": 1323, + "time_per_iteration": 2.987637519836426 + }, + { + "auxiliary_loss_clip": 0.01196413, + "auxiliary_loss_mlp": 0.01042671, + "balance_loss_clip": 1.05916965, + "balance_loss_mlp": 1.03060699, + "epoch": 0.15920158720615643, + "flos": 20193611984640.0, + "grad_norm": 1.7348083383594948, + "language_loss": 0.80530334, + "learning_rate": 3.827555310408979e-06, + "loss": 0.82769418, + "num_input_tokens_seen": 27927985, + "step": 1324, + "time_per_iteration": 2.8072011470794678 + }, + { + "auxiliary_loss_clip": 0.01199302, + "auxiliary_loss_mlp": 0.01039207, + "balance_loss_clip": 1.06541383, + "balance_loss_mlp": 1.02630806, + "epoch": 0.1593218300967955, + "flos": 24826626892800.0, + "grad_norm": 2.030201836735062, + "language_loss": 0.8278392, + "learning_rate": 3.827238742049854e-06, + "loss": 0.85022426, + "num_input_tokens_seen": 27948280, + "step": 1325, + "time_per_iteration": 2.893493413925171 + }, + { + "auxiliary_loss_clip": 0.01229729, + "auxiliary_loss_mlp": 0.010402, + "balance_loss_clip": 1.06295776, + "balance_loss_mlp": 1.02749801, + "epoch": 0.15944207298743462, + "flos": 28328707111680.0, + "grad_norm": 2.0402379590141666, + "language_loss": 0.51805496, + "learning_rate": 3.826921896500066e-06, + "loss": 0.5407542, + "num_input_tokens_seen": 27969565, + "step": 1326, + "time_per_iteration": 2.8085286617279053 + }, + { + "auxiliary_loss_clip": 0.01203986, + "auxiliary_loss_mlp": 0.01042389, + "balance_loss_clip": 1.06155443, + "balance_loss_mlp": 1.02915716, + "epoch": 0.1595623158780737, + "flos": 22964838174720.0, + "grad_norm": 2.490378126622421, + "language_loss": 0.78214687, + "learning_rate": 3.826604773807678e-06, + "loss": 0.80461061, + "num_input_tokens_seen": 27987540, + "step": 1327, + "time_per_iteration": 2.7864623069763184 + }, + { + "auxiliary_loss_clip": 0.0120926, + "auxiliary_loss_mlp": 0.01046745, + "balance_loss_clip": 1.05704868, + "balance_loss_mlp": 1.03297019, + "epoch": 0.1596825587687128, + "flos": 19710540950400.0, + "grad_norm": 2.4985338143294946, + "language_loss": 0.73469889, + "learning_rate": 3.826287374020798e-06, + "loss": 0.75725895, + "num_input_tokens_seen": 28002345, + "step": 1328, + "time_per_iteration": 2.749290943145752 + }, + { + "auxiliary_loss_clip": 0.01233333, + "auxiliary_loss_mlp": 0.01039612, + "balance_loss_clip": 1.06730604, + "balance_loss_mlp": 1.0277983, + "epoch": 0.1598028016593519, + "flos": 22637727993600.0, + "grad_norm": 2.441967094585239, + "language_loss": 0.81920862, + "learning_rate": 3.825969697187575e-06, + "loss": 0.84193802, + "num_input_tokens_seen": 28021675, + "step": 1329, + "time_per_iteration": 2.7363879680633545 + }, + { + "auxiliary_loss_clip": 0.01196134, + "auxiliary_loss_mlp": 0.01039488, + "balance_loss_clip": 1.0568856, + "balance_loss_mlp": 1.02684021, + "epoch": 0.15992304454999098, + "flos": 20482908122880.0, + "grad_norm": 2.111359730458386, + "language_loss": 0.69878936, + "learning_rate": 3.8256517433562015e-06, + "loss": 0.72114557, + "num_input_tokens_seen": 28039615, + "step": 1330, + "time_per_iteration": 2.842862129211426 + }, + { + "auxiliary_loss_clip": 0.01232385, + "auxiliary_loss_mlp": 0.01036694, + "balance_loss_clip": 1.06735981, + "balance_loss_mlp": 1.02432585, + "epoch": 0.16004328744063007, + "flos": 17676094533120.0, + "grad_norm": 2.3636922532639897, + "language_loss": 0.91668117, + "learning_rate": 3.82533351257491e-06, + "loss": 0.93937194, + "num_input_tokens_seen": 28057565, + "step": 1331, + "time_per_iteration": 2.6495089530944824 + }, + { + "auxiliary_loss_clip": 0.01219721, + "auxiliary_loss_mlp": 0.01044105, + "balance_loss_clip": 1.06483889, + "balance_loss_mlp": 1.03189194, + "epoch": 0.16016353033126918, + "flos": 24098717779200.0, + "grad_norm": 1.767013854510471, + "language_loss": 0.88810158, + "learning_rate": 3.825015004891975e-06, + "loss": 0.91073978, + "num_input_tokens_seen": 28076305, + "step": 1332, + "time_per_iteration": 2.7723891735076904 + }, + { + "auxiliary_loss_clip": 0.01217196, + "auxiliary_loss_mlp": 0.01034312, + "balance_loss_clip": 1.06177783, + "balance_loss_mlp": 1.02205741, + "epoch": 0.16028377322190826, + "flos": 27634841112960.0, + "grad_norm": 1.8444949567824362, + "language_loss": 0.75885075, + "learning_rate": 3.824696220355716e-06, + "loss": 0.78136587, + "num_input_tokens_seen": 28097895, + "step": 1333, + "time_per_iteration": 2.7987782955169678 + }, + { + "auxiliary_loss_clip": 0.01207471, + "auxiliary_loss_mlp": 0.01044247, + "balance_loss_clip": 1.06250811, + "balance_loss_mlp": 1.03247559, + "epoch": 0.16040401611254734, + "flos": 20961202648320.0, + "grad_norm": 1.6399444085743156, + "language_loss": 0.78850329, + "learning_rate": 3.824377159014491e-06, + "loss": 0.81102043, + "num_input_tokens_seen": 28118790, + "step": 1334, + "time_per_iteration": 2.742734432220459 + }, + { + "auxiliary_loss_clip": 0.01222117, + "auxiliary_loss_mlp": 0.01041592, + "balance_loss_clip": 1.0662353, + "balance_loss_mlp": 1.02981997, + "epoch": 0.16052425900318643, + "flos": 21247051080960.0, + "grad_norm": 1.8491818859637401, + "language_loss": 0.84645307, + "learning_rate": 3.824057820916702e-06, + "loss": 0.86909014, + "num_input_tokens_seen": 28135995, + "step": 1335, + "time_per_iteration": 2.8294053077697754 + }, + { + "auxiliary_loss_clip": 0.01213305, + "auxiliary_loss_mlp": 0.0103485, + "balance_loss_clip": 1.0640614, + "balance_loss_mlp": 1.02140355, + "epoch": 0.16064450189382554, + "flos": 15524004096000.0, + "grad_norm": 2.9753734697244343, + "language_loss": 0.7208662, + "learning_rate": 3.8237382061107904e-06, + "loss": 0.74334776, + "num_input_tokens_seen": 28152715, + "step": 1336, + "time_per_iteration": 2.736938238143921 + }, + { + "auxiliary_loss_clip": 0.01156884, + "auxiliary_loss_mlp": 0.01036021, + "balance_loss_clip": 1.04808021, + "balance_loss_mlp": 1.02429664, + "epoch": 0.16076474478446462, + "flos": 21178497974400.0, + "grad_norm": 2.3461907608467327, + "language_loss": 0.7865774, + "learning_rate": 3.823418314645243e-06, + "loss": 0.80850649, + "num_input_tokens_seen": 28171590, + "step": 1337, + "time_per_iteration": 2.9745988845825195 + }, + { + "auxiliary_loss_clip": 0.01180081, + "auxiliary_loss_mlp": 0.0103812, + "balance_loss_clip": 1.06217408, + "balance_loss_mlp": 1.02624035, + "epoch": 0.1608849876751037, + "flos": 18366476912640.0, + "grad_norm": 2.5535386859318705, + "language_loss": 0.75602245, + "learning_rate": 3.823098146568588e-06, + "loss": 0.77820444, + "num_input_tokens_seen": 28191295, + "step": 1338, + "time_per_iteration": 2.805983304977417 + }, + { + "auxiliary_loss_clip": 0.01223624, + "auxiliary_loss_mlp": 0.01035644, + "balance_loss_clip": 1.06726027, + "balance_loss_mlp": 1.02346039, + "epoch": 0.1610052305657428, + "flos": 29497024880640.0, + "grad_norm": 3.7882708701684193, + "language_loss": 0.71638942, + "learning_rate": 3.822777701929394e-06, + "loss": 0.73898208, + "num_input_tokens_seen": 28213120, + "step": 1339, + "time_per_iteration": 2.8053793907165527 + }, + { + "auxiliary_loss_clip": 0.01209704, + "auxiliary_loss_mlp": 0.01038067, + "balance_loss_clip": 1.05941248, + "balance_loss_mlp": 1.02546644, + "epoch": 0.1611254734563819, + "flos": 26797871329920.0, + "grad_norm": 1.9790312176414944, + "language_loss": 0.73545659, + "learning_rate": 3.8224569807762714e-06, + "loss": 0.75793433, + "num_input_tokens_seen": 28232440, + "step": 1340, + "time_per_iteration": 2.7255008220672607 + }, + { + "auxiliary_loss_clip": 0.01166339, + "auxiliary_loss_mlp": 0.01039844, + "balance_loss_clip": 1.04957306, + "balance_loss_mlp": 1.027565, + "epoch": 0.16124571634702098, + "flos": 22419570741120.0, + "grad_norm": 2.6422383199785378, + "language_loss": 0.7662462, + "learning_rate": 3.822135983157873e-06, + "loss": 0.78830802, + "num_input_tokens_seen": 28251715, + "step": 1341, + "time_per_iteration": 2.9649245738983154 + }, + { + "auxiliary_loss_clip": 0.01228809, + "auxiliary_loss_mlp": 0.0088768, + "balance_loss_clip": 1.06481576, + "balance_loss_mlp": 1.00013661, + "epoch": 0.16136595923766006, + "flos": 10999116103680.0, + "grad_norm": 2.3828043965049557, + "language_loss": 0.84281319, + "learning_rate": 3.821814709122896e-06, + "loss": 0.86397809, + "num_input_tokens_seen": 28269765, + "step": 1342, + "time_per_iteration": 2.65364146232605 + }, + { + "auxiliary_loss_clip": 0.01207558, + "auxiliary_loss_mlp": 0.01043133, + "balance_loss_clip": 1.06239653, + "balance_loss_mlp": 1.03134274, + "epoch": 0.16148620212829917, + "flos": 21214983214080.0, + "grad_norm": 2.120965796574138, + "language_loss": 0.84824193, + "learning_rate": 3.821493158720076e-06, + "loss": 0.87074882, + "num_input_tokens_seen": 28288870, + "step": 1343, + "time_per_iteration": 2.8452799320220947 + }, + { + "auxiliary_loss_clip": 0.01191694, + "auxiliary_loss_mlp": 0.01045748, + "balance_loss_clip": 1.05198836, + "balance_loss_mlp": 1.03282559, + "epoch": 0.16160644501893826, + "flos": 16758468760320.0, + "grad_norm": 3.4659900068456806, + "language_loss": 0.73370808, + "learning_rate": 3.821171331998191e-06, + "loss": 0.75608253, + "num_input_tokens_seen": 28305400, + "step": 1344, + "time_per_iteration": 2.8235034942626953 + }, + { + "auxiliary_loss_clip": 0.01117772, + "auxiliary_loss_mlp": 0.01008741, + "balance_loss_clip": 1.04295611, + "balance_loss_mlp": 1.00371063, + "epoch": 0.16172668790957734, + "flos": 64444967308800.0, + "grad_norm": 0.7128759205558246, + "language_loss": 0.54444909, + "learning_rate": 3.820849229006064e-06, + "loss": 0.56571424, + "num_input_tokens_seen": 28373150, + "step": 1345, + "time_per_iteration": 4.387451887130737 + }, + { + "auxiliary_loss_clip": 0.01229993, + "auxiliary_loss_mlp": 0.0103696, + "balance_loss_clip": 1.06449127, + "balance_loss_mlp": 1.02455688, + "epoch": 0.16184693080021645, + "flos": 23257689759360.0, + "grad_norm": 2.1780380348501667, + "language_loss": 0.71193111, + "learning_rate": 3.8205268497925564e-06, + "loss": 0.73460066, + "num_input_tokens_seen": 28393620, + "step": 1346, + "time_per_iteration": 3.6269772052764893 + }, + { + "auxiliary_loss_clip": 0.01230078, + "auxiliary_loss_mlp": 0.01042818, + "balance_loss_clip": 1.06593907, + "balance_loss_mlp": 1.03119504, + "epoch": 0.16196717369085553, + "flos": 17451113696640.0, + "grad_norm": 2.1506232590250787, + "language_loss": 0.78577626, + "learning_rate": 3.8202041944065725e-06, + "loss": 0.80850518, + "num_input_tokens_seen": 28409440, + "step": 1347, + "time_per_iteration": 3.619598150253296 + }, + { + "auxiliary_loss_clip": 0.01229927, + "auxiliary_loss_mlp": 0.01037674, + "balance_loss_clip": 1.06572175, + "balance_loss_mlp": 1.02599716, + "epoch": 0.16208741658149461, + "flos": 23873377806720.0, + "grad_norm": 2.481702607790926, + "language_loss": 0.73989552, + "learning_rate": 3.819881262897061e-06, + "loss": 0.76257145, + "num_input_tokens_seen": 28427575, + "step": 1348, + "time_per_iteration": 2.7366645336151123 + }, + { + "auxiliary_loss_clip": 0.01207537, + "auxiliary_loss_mlp": 0.01046949, + "balance_loss_clip": 1.06742191, + "balance_loss_mlp": 1.03448009, + "epoch": 0.1622076594721337, + "flos": 25884806584320.0, + "grad_norm": 1.837011701874461, + "language_loss": 0.73477125, + "learning_rate": 3.819558055313008e-06, + "loss": 0.75731611, + "num_input_tokens_seen": 28448260, + "step": 1349, + "time_per_iteration": 2.991185426712036 + }, + { + "auxiliary_loss_clip": 0.01226589, + "auxiliary_loss_mlp": 0.01038204, + "balance_loss_clip": 1.06589031, + "balance_loss_mlp": 1.02596152, + "epoch": 0.1623279023627728, + "flos": 21539759011200.0, + "grad_norm": 3.231495559670143, + "language_loss": 0.77261025, + "learning_rate": 3.819234571703444e-06, + "loss": 0.79525816, + "num_input_tokens_seen": 28467085, + "step": 1350, + "time_per_iteration": 2.76924729347229 + }, + { + "auxiliary_loss_clip": 0.01215421, + "auxiliary_loss_mlp": 0.01038466, + "balance_loss_clip": 1.06188416, + "balance_loss_mlp": 1.02638435, + "epoch": 0.1624481452534119, + "flos": 22085421494400.0, + "grad_norm": 2.1672006077482346, + "language_loss": 0.8572666, + "learning_rate": 3.8189108121174435e-06, + "loss": 0.87980545, + "num_input_tokens_seen": 28486850, + "step": 1351, + "time_per_iteration": 2.8650245666503906 + }, + { + "auxiliary_loss_clip": 0.01196504, + "auxiliary_loss_mlp": 0.01037671, + "balance_loss_clip": 1.06576395, + "balance_loss_mlp": 1.0259645, + "epoch": 0.16256838814405097, + "flos": 27087490690560.0, + "grad_norm": 1.6801716063152154, + "language_loss": 0.83503342, + "learning_rate": 3.818586776604118e-06, + "loss": 0.85737514, + "num_input_tokens_seen": 28507490, + "step": 1352, + "time_per_iteration": 2.9227051734924316 + }, + { + "auxiliary_loss_clip": 0.01202461, + "auxiliary_loss_mlp": 0.0104124, + "balance_loss_clip": 1.0589335, + "balance_loss_mlp": 1.02981973, + "epoch": 0.16268863103469008, + "flos": 20120354196480.0, + "grad_norm": 2.1143164150096707, + "language_loss": 0.61516839, + "learning_rate": 3.818262465212625e-06, + "loss": 0.63760543, + "num_input_tokens_seen": 28527615, + "step": 1353, + "time_per_iteration": 2.7350075244903564 + }, + { + "auxiliary_loss_clip": 0.0121454, + "auxiliary_loss_mlp": 0.01046455, + "balance_loss_clip": 1.06439626, + "balance_loss_mlp": 1.03264475, + "epoch": 0.16280887392532917, + "flos": 18332792933760.0, + "grad_norm": 2.2135022071076382, + "language_loss": 0.77168727, + "learning_rate": 3.817937877992161e-06, + "loss": 0.79429722, + "num_input_tokens_seen": 28544910, + "step": 1354, + "time_per_iteration": 2.7191452980041504 + }, + { + "auxiliary_loss_clip": 0.01199812, + "auxiliary_loss_mlp": 0.00888281, + "balance_loss_clip": 1.05798602, + "balance_loss_mlp": 1.00030255, + "epoch": 0.16292911681596825, + "flos": 11874330892800.0, + "grad_norm": 3.564390015282601, + "language_loss": 0.85737109, + "learning_rate": 3.817613014991967e-06, + "loss": 0.87825203, + "num_input_tokens_seen": 28561050, + "step": 1355, + "time_per_iteration": 2.854111909866333 + }, + { + "auxiliary_loss_clip": 0.01186342, + "auxiliary_loss_mlp": 0.01039445, + "balance_loss_clip": 1.05467176, + "balance_loss_mlp": 1.0274229, + "epoch": 0.16304935970660733, + "flos": 26103466627200.0, + "grad_norm": 2.6683281360562923, + "language_loss": 0.76771605, + "learning_rate": 3.817287876261323e-06, + "loss": 0.78997386, + "num_input_tokens_seen": 28581385, + "step": 1356, + "time_per_iteration": 2.8559346199035645 + }, + { + "auxiliary_loss_clip": 0.01208804, + "auxiliary_loss_mlp": 0.01042504, + "balance_loss_clip": 1.06417418, + "balance_loss_mlp": 1.02996945, + "epoch": 0.16316960259724644, + "flos": 29351945848320.0, + "grad_norm": 1.7962774104355617, + "language_loss": 0.80094278, + "learning_rate": 3.816962461849553e-06, + "loss": 0.82345587, + "num_input_tokens_seen": 28603255, + "step": 1357, + "time_per_iteration": 2.8598055839538574 + }, + { + "auxiliary_loss_clip": 0.01200331, + "auxiliary_loss_mlp": 0.01043319, + "balance_loss_clip": 1.06305552, + "balance_loss_mlp": 1.03160071, + "epoch": 0.16328984548788553, + "flos": 20886759711360.0, + "grad_norm": 2.438587941467794, + "language_loss": 0.84377384, + "learning_rate": 3.8166367718060235e-06, + "loss": 0.86621034, + "num_input_tokens_seen": 28623145, + "step": 1358, + "time_per_iteration": 2.7263901233673096 + }, + { + "auxiliary_loss_clip": 0.01213271, + "auxiliary_loss_mlp": 0.01041718, + "balance_loss_clip": 1.0586338, + "balance_loss_mlp": 1.03003526, + "epoch": 0.1634100883785246, + "flos": 18041090584320.0, + "grad_norm": 2.825785204639359, + "language_loss": 0.76231194, + "learning_rate": 3.816310806180139e-06, + "loss": 0.7848618, + "num_input_tokens_seen": 28641555, + "step": 1359, + "time_per_iteration": 2.7685439586639404 + }, + { + "auxiliary_loss_clip": 0.01207292, + "auxiliary_loss_mlp": 0.01042582, + "balance_loss_clip": 1.06302953, + "balance_loss_mlp": 1.03020799, + "epoch": 0.16353033126916372, + "flos": 24572128055040.0, + "grad_norm": 1.7933763844310602, + "language_loss": 0.81137621, + "learning_rate": 3.81598456502135e-06, + "loss": 0.83387494, + "num_input_tokens_seen": 28661575, + "step": 1360, + "time_per_iteration": 2.787200689315796 + }, + { + "auxiliary_loss_clip": 0.012037, + "auxiliary_loss_mlp": 0.01044843, + "balance_loss_clip": 1.06257653, + "balance_loss_mlp": 1.03335738, + "epoch": 0.1636505741598028, + "flos": 19892895321600.0, + "grad_norm": 2.2564155208390066, + "language_loss": 0.86972982, + "learning_rate": 3.8156580483791455e-06, + "loss": 0.89221525, + "num_input_tokens_seen": 28676765, + "step": 1361, + "time_per_iteration": 2.7455883026123047 + }, + { + "auxiliary_loss_clip": 0.01229281, + "auxiliary_loss_mlp": 0.01044416, + "balance_loss_clip": 1.065081, + "balance_loss_mlp": 1.03294837, + "epoch": 0.16377081705044189, + "flos": 28402611344640.0, + "grad_norm": 2.140279564278239, + "language_loss": 0.76942241, + "learning_rate": 3.815331256303059e-06, + "loss": 0.79215938, + "num_input_tokens_seen": 28696795, + "step": 1362, + "time_per_iteration": 2.8541505336761475 + }, + { + "auxiliary_loss_clip": 0.01197671, + "auxiliary_loss_mlp": 0.01042072, + "balance_loss_clip": 1.06462348, + "balance_loss_mlp": 1.03023422, + "epoch": 0.163891059941081, + "flos": 21908059113600.0, + "grad_norm": 2.0608895040463695, + "language_loss": 0.77077484, + "learning_rate": 3.815004188842665e-06, + "loss": 0.79317224, + "num_input_tokens_seen": 28714835, + "step": 1363, + "time_per_iteration": 2.8190720081329346 + }, + { + "auxiliary_loss_clip": 0.01197756, + "auxiliary_loss_mlp": 0.01036237, + "balance_loss_clip": 1.0559032, + "balance_loss_mlp": 1.02434039, + "epoch": 0.16401130283172008, + "flos": 26797619934720.0, + "grad_norm": 1.9099203178096564, + "language_loss": 0.79775882, + "learning_rate": 3.814676846047578e-06, + "loss": 0.82009876, + "num_input_tokens_seen": 28735710, + "step": 1364, + "time_per_iteration": 2.8084239959716797 + }, + { + "auxiliary_loss_clip": 0.01216387, + "auxiliary_loss_mlp": 0.01038795, + "balance_loss_clip": 1.06243873, + "balance_loss_mlp": 1.0268743, + "epoch": 0.16413154572235916, + "flos": 32997417160320.0, + "grad_norm": 2.0155268454505486, + "language_loss": 0.6976949, + "learning_rate": 3.8143492279674565e-06, + "loss": 0.72024667, + "num_input_tokens_seen": 28758405, + "step": 1365, + "time_per_iteration": 2.827901601791382 + }, + { + "auxiliary_loss_clip": 0.01105561, + "auxiliary_loss_mlp": 0.01013724, + "balance_loss_clip": 1.03503299, + "balance_loss_mlp": 1.00886011, + "epoch": 0.16425178861299825, + "flos": 40113622074240.0, + "grad_norm": 0.8444122404740748, + "language_loss": 0.58419514, + "learning_rate": 3.8140213346519997e-06, + "loss": 0.60538793, + "num_input_tokens_seen": 28809000, + "step": 1366, + "time_per_iteration": 3.185053825378418 + }, + { + "auxiliary_loss_clip": 0.01192107, + "auxiliary_loss_mlp": 0.01040825, + "balance_loss_clip": 1.0601635, + "balance_loss_mlp": 1.02955985, + "epoch": 0.16437203150363736, + "flos": 25447486498560.0, + "grad_norm": 1.7819225552737816, + "language_loss": 0.77093583, + "learning_rate": 3.813693166150948e-06, + "loss": 0.79326516, + "num_input_tokens_seen": 28829210, + "step": 1367, + "time_per_iteration": 2.8490047454833984 + }, + { + "auxiliary_loss_clip": 0.01186002, + "auxiliary_loss_mlp": 0.01047644, + "balance_loss_clip": 1.05627108, + "balance_loss_mlp": 1.03576517, + "epoch": 0.16449227439427644, + "flos": 23476888506240.0, + "grad_norm": 2.498469542041247, + "language_loss": 0.85594088, + "learning_rate": 3.813364722514086e-06, + "loss": 0.8782773, + "num_input_tokens_seen": 28847545, + "step": 1368, + "time_per_iteration": 2.945223808288574 + }, + { + "auxiliary_loss_clip": 0.01214773, + "auxiliary_loss_mlp": 0.01044674, + "balance_loss_clip": 1.06024265, + "balance_loss_mlp": 1.03253245, + "epoch": 0.16461251728491552, + "flos": 13545217802880.0, + "grad_norm": 2.012892888388323, + "language_loss": 0.80644763, + "learning_rate": 3.8130360037912368e-06, + "loss": 0.82904208, + "num_input_tokens_seen": 28863990, + "step": 1369, + "time_per_iteration": 2.686236619949341 + }, + { + "auxiliary_loss_clip": 0.01215712, + "auxiliary_loss_mlp": 0.01037008, + "balance_loss_clip": 1.06163514, + "balance_loss_mlp": 1.02480662, + "epoch": 0.16473276017555463, + "flos": 23003298662400.0, + "grad_norm": 2.163116564039326, + "language_loss": 0.82024568, + "learning_rate": 3.812707010032268e-06, + "loss": 0.84277284, + "num_input_tokens_seen": 28883045, + "step": 1370, + "time_per_iteration": 2.685913562774658 + }, + { + "auxiliary_loss_clip": 0.01222949, + "auxiliary_loss_mlp": 0.0104332, + "balance_loss_clip": 1.06558752, + "balance_loss_mlp": 1.03059459, + "epoch": 0.16485300306619372, + "flos": 24790680357120.0, + "grad_norm": 1.9049620349272518, + "language_loss": 0.78998268, + "learning_rate": 3.8123777412870863e-06, + "loss": 0.81264538, + "num_input_tokens_seen": 28902545, + "step": 1371, + "time_per_iteration": 3.6230359077453613 + }, + { + "auxiliary_loss_clip": 0.0121211, + "auxiliary_loss_mlp": 0.01038197, + "balance_loss_clip": 1.06008816, + "balance_loss_mlp": 1.02656841, + "epoch": 0.1649732459568328, + "flos": 21106497162240.0, + "grad_norm": 2.342522587514715, + "language_loss": 0.78611809, + "learning_rate": 3.812048197605643e-06, + "loss": 0.80862117, + "num_input_tokens_seen": 28921440, + "step": 1372, + "time_per_iteration": 3.6406304836273193 + }, + { + "auxiliary_loss_clip": 0.01216258, + "auxiliary_loss_mlp": 0.01040213, + "balance_loss_clip": 1.05983281, + "balance_loss_mlp": 1.02852428, + "epoch": 0.16509348884747188, + "flos": 20266726118400.0, + "grad_norm": 1.9672564485365436, + "language_loss": 0.81385285, + "learning_rate": 3.8117183790379277e-06, + "loss": 0.83641762, + "num_input_tokens_seen": 28939890, + "step": 1373, + "time_per_iteration": 3.6647560596466064 + }, + { + "auxiliary_loss_clip": 0.01229602, + "auxiliary_loss_mlp": 0.01037944, + "balance_loss_clip": 1.06418836, + "balance_loss_mlp": 1.02611816, + "epoch": 0.165213731738111, + "flos": 11035493602560.0, + "grad_norm": 3.099040118511013, + "language_loss": 0.94169515, + "learning_rate": 3.811388285633976e-06, + "loss": 0.96437061, + "num_input_tokens_seen": 28955875, + "step": 1374, + "time_per_iteration": 2.7026405334472656 + }, + { + "auxiliary_loss_clip": 0.01191309, + "auxiliary_loss_mlp": 0.01041396, + "balance_loss_clip": 1.05886757, + "balance_loss_mlp": 1.02846205, + "epoch": 0.16533397462875007, + "flos": 29972051268480.0, + "grad_norm": 1.9853958470218316, + "language_loss": 0.62449527, + "learning_rate": 3.811057917443861e-06, + "loss": 0.64682233, + "num_input_tokens_seen": 28975140, + "step": 1375, + "time_per_iteration": 2.868889093399048 + }, + { + "auxiliary_loss_clip": 0.01119421, + "auxiliary_loss_mlp": 0.01009062, + "balance_loss_clip": 1.03748369, + "balance_loss_mlp": 1.00424635, + "epoch": 0.16545421751938916, + "flos": 65556763027200.0, + "grad_norm": 0.8553194553782684, + "language_loss": 0.68307328, + "learning_rate": 3.8107272745177e-06, + "loss": 0.7043581, + "num_input_tokens_seen": 29047470, + "step": 1376, + "time_per_iteration": 3.436652421951294 + }, + { + "auxiliary_loss_clip": 0.01195615, + "auxiliary_loss_mlp": 0.0103907, + "balance_loss_clip": 1.05788338, + "balance_loss_mlp": 1.02651715, + "epoch": 0.16557446041002827, + "flos": 22492361652480.0, + "grad_norm": 1.725287330464412, + "language_loss": 0.78595757, + "learning_rate": 3.8103963569056513e-06, + "loss": 0.80830443, + "num_input_tokens_seen": 29066605, + "step": 1377, + "time_per_iteration": 2.854123115539551 + }, + { + "auxiliary_loss_clip": 0.0119503, + "auxiliary_loss_mlp": 0.01039693, + "balance_loss_clip": 1.05383253, + "balance_loss_mlp": 1.02824295, + "epoch": 0.16569470330066735, + "flos": 24602723464320.0, + "grad_norm": 1.7052799287912934, + "language_loss": 0.87977815, + "learning_rate": 3.8100651646579146e-06, + "loss": 0.90212536, + "num_input_tokens_seen": 29085815, + "step": 1378, + "time_per_iteration": 2.7965712547302246 + }, + { + "auxiliary_loss_clip": 0.01195327, + "auxiliary_loss_mlp": 0.01038695, + "balance_loss_clip": 1.05290306, + "balance_loss_mlp": 1.02681541, + "epoch": 0.16581494619130643, + "flos": 15006207588480.0, + "grad_norm": 2.2972077964371658, + "language_loss": 0.92709422, + "learning_rate": 3.8097336978247317e-06, + "loss": 0.94943446, + "num_input_tokens_seen": 29102520, + "step": 1379, + "time_per_iteration": 2.7325642108917236 + }, + { + "auxiliary_loss_clip": 0.0119483, + "auxiliary_loss_mlp": 0.01036764, + "balance_loss_clip": 1.05760777, + "balance_loss_mlp": 1.02413368, + "epoch": 0.16593518908194552, + "flos": 17420338719360.0, + "grad_norm": 2.0487976302379196, + "language_loss": 0.88925564, + "learning_rate": 3.8094019564563854e-06, + "loss": 0.9115715, + "num_input_tokens_seen": 29119450, + "step": 1380, + "time_per_iteration": 2.716857433319092 + }, + { + "auxiliary_loss_clip": 0.01227373, + "auxiliary_loss_mlp": 0.0088845, + "balance_loss_clip": 1.06337607, + "balance_loss_mlp": 1.00022078, + "epoch": 0.16605543197258463, + "flos": 20412631163520.0, + "grad_norm": 2.2749139479177813, + "language_loss": 0.75442302, + "learning_rate": 3.809069940603201e-06, + "loss": 0.77558124, + "num_input_tokens_seen": 29137405, + "step": 1381, + "time_per_iteration": 2.7692532539367676 + }, + { + "auxiliary_loss_clip": 0.01198198, + "auxiliary_loss_mlp": 0.01038332, + "balance_loss_clip": 1.06114399, + "balance_loss_mlp": 1.02661395, + "epoch": 0.1661756748632237, + "flos": 14209745368320.0, + "grad_norm": 2.1423909562914667, + "language_loss": 0.78475845, + "learning_rate": 3.8087376503155452e-06, + "loss": 0.80712378, + "num_input_tokens_seen": 29154890, + "step": 1382, + "time_per_iteration": 2.749048948287964 + }, + { + "auxiliary_loss_clip": 0.01113549, + "auxiliary_loss_mlp": 0.01007058, + "balance_loss_clip": 1.0348146, + "balance_loss_mlp": 1.00240886, + "epoch": 0.1662959177538628, + "flos": 66080877350400.0, + "grad_norm": 0.8974320572449438, + "language_loss": 0.56263232, + "learning_rate": 3.808405085643826e-06, + "loss": 0.58383846, + "num_input_tokens_seen": 29219770, + "step": 1383, + "time_per_iteration": 3.433664321899414 + }, + { + "auxiliary_loss_clip": 0.01226955, + "auxiliary_loss_mlp": 0.00887821, + "balance_loss_clip": 1.06406748, + "balance_loss_mlp": 1.00024772, + "epoch": 0.1664161606445019, + "flos": 20740567357440.0, + "grad_norm": 2.1604502376240595, + "language_loss": 0.8865658, + "learning_rate": 3.8080722466384925e-06, + "loss": 0.90771365, + "num_input_tokens_seen": 29237620, + "step": 1384, + "time_per_iteration": 2.716423273086548 + }, + { + "auxiliary_loss_clip": 0.01228378, + "auxiliary_loss_mlp": 0.01041554, + "balance_loss_clip": 1.06228209, + "balance_loss_mlp": 1.02985334, + "epoch": 0.166536403535141, + "flos": 25260930236160.0, + "grad_norm": 2.5706942440836493, + "language_loss": 0.70828152, + "learning_rate": 3.8077391333500376e-06, + "loss": 0.73098081, + "num_input_tokens_seen": 29256760, + "step": 1385, + "time_per_iteration": 2.693199634552002 + }, + { + "auxiliary_loss_clip": 0.0120542, + "auxiliary_loss_mlp": 0.01038305, + "balance_loss_clip": 1.06082702, + "balance_loss_mlp": 1.02757049, + "epoch": 0.16665664642578007, + "flos": 25447450584960.0, + "grad_norm": 1.6480671888937664, + "language_loss": 0.76843441, + "learning_rate": 3.8074057458289934e-06, + "loss": 0.79087168, + "num_input_tokens_seen": 29277450, + "step": 1386, + "time_per_iteration": 2.820972204208374 + }, + { + "auxiliary_loss_clip": 0.01200534, + "auxiliary_loss_mlp": 0.01040744, + "balance_loss_clip": 1.05627251, + "balance_loss_mlp": 1.02862692, + "epoch": 0.16677688931641918, + "flos": 22200767043840.0, + "grad_norm": 2.0442642503118087, + "language_loss": 0.82501823, + "learning_rate": 3.807072084125934e-06, + "loss": 0.847431, + "num_input_tokens_seen": 29299300, + "step": 1387, + "time_per_iteration": 2.8100745677948 + }, + { + "auxiliary_loss_clip": 0.01200949, + "auxiliary_loss_mlp": 0.01045276, + "balance_loss_clip": 1.0610168, + "balance_loss_mlp": 1.03311026, + "epoch": 0.16689713220705826, + "flos": 16945958776320.0, + "grad_norm": 5.308552536653996, + "language_loss": 0.80614543, + "learning_rate": 3.806738148291477e-06, + "loss": 0.82860768, + "num_input_tokens_seen": 29316125, + "step": 1388, + "time_per_iteration": 2.875710964202881 + }, + { + "auxiliary_loss_clip": 0.01174683, + "auxiliary_loss_mlp": 0.01045893, + "balance_loss_clip": 1.05172324, + "balance_loss_mlp": 1.03363276, + "epoch": 0.16701737509769735, + "flos": 36244423923840.0, + "grad_norm": 1.9143503470946643, + "language_loss": 0.71382964, + "learning_rate": 3.8064039383762793e-06, + "loss": 0.73603541, + "num_input_tokens_seen": 29338490, + "step": 1389, + "time_per_iteration": 3.005352735519409 + }, + { + "auxiliary_loss_clip": 0.01219373, + "auxiliary_loss_mlp": 0.01036277, + "balance_loss_clip": 1.06818032, + "balance_loss_mlp": 1.02487481, + "epoch": 0.16713761798833643, + "flos": 23258659426560.0, + "grad_norm": 2.676552142556664, + "language_loss": 0.77121741, + "learning_rate": 3.8060694544310396e-06, + "loss": 0.79377389, + "num_input_tokens_seen": 29357000, + "step": 1390, + "time_per_iteration": 2.7701756954193115 + }, + { + "auxiliary_loss_clip": 0.01228533, + "auxiliary_loss_mlp": 0.01039739, + "balance_loss_clip": 1.06330776, + "balance_loss_mlp": 1.02800274, + "epoch": 0.16725786087897554, + "flos": 25302515207040.0, + "grad_norm": 1.7010822080383547, + "language_loss": 0.78256238, + "learning_rate": 3.8057346965065006e-06, + "loss": 0.80524504, + "num_input_tokens_seen": 29378230, + "step": 1391, + "time_per_iteration": 2.806638240814209 + }, + { + "auxiliary_loss_clip": 0.01208904, + "auxiliary_loss_mlp": 0.01043971, + "balance_loss_clip": 1.06383955, + "balance_loss_mlp": 1.03186536, + "epoch": 0.16737810376961462, + "flos": 31831541516160.0, + "grad_norm": 1.678050258658315, + "language_loss": 0.84876138, + "learning_rate": 3.805399664653443e-06, + "loss": 0.87129021, + "num_input_tokens_seen": 29400370, + "step": 1392, + "time_per_iteration": 2.8266708850860596 + }, + { + "auxiliary_loss_clip": 0.01228286, + "auxiliary_loss_mlp": 0.01042956, + "balance_loss_clip": 1.06253946, + "balance_loss_mlp": 1.03148198, + "epoch": 0.1674983466602537, + "flos": 27961843553280.0, + "grad_norm": 2.8812688078208653, + "language_loss": 0.74315202, + "learning_rate": 3.805064358922692e-06, + "loss": 0.76586449, + "num_input_tokens_seen": 29418660, + "step": 1393, + "time_per_iteration": 2.734316110610962 + }, + { + "auxiliary_loss_clip": 0.01220112, + "auxiliary_loss_mlp": 0.01039395, + "balance_loss_clip": 1.06142688, + "balance_loss_mlp": 1.02701521, + "epoch": 0.16761858955089282, + "flos": 21762656858880.0, + "grad_norm": 2.787242877560518, + "language_loss": 0.81058538, + "learning_rate": 3.8047287793651136e-06, + "loss": 0.83318043, + "num_input_tokens_seen": 29440105, + "step": 1394, + "time_per_iteration": 2.813469886779785 + }, + { + "auxiliary_loss_clip": 0.01194829, + "auxiliary_loss_mlp": 0.01041396, + "balance_loss_clip": 1.05782557, + "balance_loss_mlp": 1.02974892, + "epoch": 0.1677388324415319, + "flos": 23805507058560.0, + "grad_norm": 2.2057702778988086, + "language_loss": 0.88605213, + "learning_rate": 3.8043929260316137e-06, + "loss": 0.90841436, + "num_input_tokens_seen": 29458260, + "step": 1395, + "time_per_iteration": 2.8673365116119385 + }, + { + "auxiliary_loss_clip": 0.01207971, + "auxiliary_loss_mlp": 0.01045073, + "balance_loss_clip": 1.06336498, + "balance_loss_mlp": 1.03290129, + "epoch": 0.16785907533217098, + "flos": 20558859431040.0, + "grad_norm": 1.981009561773525, + "language_loss": 0.83578807, + "learning_rate": 3.8040567989731417e-06, + "loss": 0.85831851, + "num_input_tokens_seen": 29476205, + "step": 1396, + "time_per_iteration": 2.8375234603881836 + }, + { + "auxiliary_loss_clip": 0.0121192, + "auxiliary_loss_mlp": 0.01036945, + "balance_loss_clip": 1.0624603, + "balance_loss_mlp": 1.02598345, + "epoch": 0.16797931822281006, + "flos": 15669657745920.0, + "grad_norm": 4.227967925122095, + "language_loss": 0.79574454, + "learning_rate": 3.8037203982406876e-06, + "loss": 0.81823325, + "num_input_tokens_seen": 29494370, + "step": 1397, + "time_per_iteration": 4.501558303833008 + }, + { + "auxiliary_loss_clip": 0.01228663, + "auxiliary_loss_mlp": 0.01040456, + "balance_loss_clip": 1.06601548, + "balance_loss_mlp": 1.02920866, + "epoch": 0.16809956111344918, + "flos": 16541101607040.0, + "grad_norm": 2.0435512136880027, + "language_loss": 0.73237908, + "learning_rate": 3.8033837238852835e-06, + "loss": 0.75507033, + "num_input_tokens_seen": 29511070, + "step": 1398, + "time_per_iteration": 3.567761182785034 + }, + { + "auxiliary_loss_clip": 0.01196531, + "auxiliary_loss_mlp": 0.01039282, + "balance_loss_clip": 1.05696225, + "balance_loss_mlp": 1.02739692, + "epoch": 0.16821980400408826, + "flos": 23258084808960.0, + "grad_norm": 2.025340327093348, + "language_loss": 0.69888532, + "learning_rate": 3.8030467759580017e-06, + "loss": 0.72124344, + "num_input_tokens_seen": 29531990, + "step": 1399, + "time_per_iteration": 3.707157611846924 + }, + { + "auxiliary_loss_clip": 0.01217665, + "auxiliary_loss_mlp": 0.01047317, + "balance_loss_clip": 1.06037903, + "balance_loss_mlp": 1.03531814, + "epoch": 0.16834004689472734, + "flos": 20774754126720.0, + "grad_norm": 1.933728233203081, + "language_loss": 0.87210739, + "learning_rate": 3.802709554509958e-06, + "loss": 0.89475721, + "num_input_tokens_seen": 29549790, + "step": 1400, + "time_per_iteration": 2.7956202030181885 + }, + { + "auxiliary_loss_clip": 0.01203325, + "auxiliary_loss_mlp": 0.01037077, + "balance_loss_clip": 1.05645514, + "balance_loss_mlp": 1.02592528, + "epoch": 0.16846028978536645, + "flos": 26687302289280.0, + "grad_norm": 1.7754809711133186, + "language_loss": 0.79449677, + "learning_rate": 3.8023720595923083e-06, + "loss": 0.81690085, + "num_input_tokens_seen": 29569045, + "step": 1401, + "time_per_iteration": 2.996199369430542 + }, + { + "auxiliary_loss_clip": 0.01183484, + "auxiliary_loss_mlp": 0.01037888, + "balance_loss_clip": 1.05493104, + "balance_loss_mlp": 1.02597284, + "epoch": 0.16858053267600553, + "flos": 18843298980480.0, + "grad_norm": 2.1172827701623493, + "language_loss": 0.87525058, + "learning_rate": 3.80203429125625e-06, + "loss": 0.89746428, + "num_input_tokens_seen": 29587220, + "step": 1402, + "time_per_iteration": 2.7954437732696533 + }, + { + "auxiliary_loss_clip": 0.01166317, + "auxiliary_loss_mlp": 0.01040716, + "balance_loss_clip": 1.05581784, + "balance_loss_mlp": 1.0295459, + "epoch": 0.16870077556664462, + "flos": 27744548227200.0, + "grad_norm": 1.8867384093690638, + "language_loss": 0.70462382, + "learning_rate": 3.8016962495530225e-06, + "loss": 0.72669411, + "num_input_tokens_seen": 29606410, + "step": 1403, + "time_per_iteration": 2.9604439735412598 + }, + { + "auxiliary_loss_clip": 0.01226879, + "auxiliary_loss_mlp": 0.01039607, + "balance_loss_clip": 1.0618124, + "balance_loss_mlp": 1.02810335, + "epoch": 0.1688210184572837, + "flos": 13730768484480.0, + "grad_norm": 2.635885951596904, + "language_loss": 0.76635242, + "learning_rate": 3.8013579345339063e-06, + "loss": 0.7890172, + "num_input_tokens_seen": 29621275, + "step": 1404, + "time_per_iteration": 2.714897394180298 + }, + { + "auxiliary_loss_clip": 0.01199887, + "auxiliary_loss_mlp": 0.01036918, + "balance_loss_clip": 1.05859959, + "balance_loss_mlp": 1.02581954, + "epoch": 0.1689412613479228, + "flos": 26468785900800.0, + "grad_norm": 2.021878462871995, + "language_loss": 0.68994427, + "learning_rate": 3.801019346250224e-06, + "loss": 0.71231234, + "num_input_tokens_seen": 29641420, + "step": 1405, + "time_per_iteration": 2.822312355041504 + }, + { + "auxiliary_loss_clip": 0.01215739, + "auxiliary_loss_mlp": 0.0103835, + "balance_loss_clip": 1.06470561, + "balance_loss_mlp": 1.02623248, + "epoch": 0.1690615042385619, + "flos": 21138852337920.0, + "grad_norm": 6.84675828155939, + "language_loss": 0.83636129, + "learning_rate": 3.8006804847533395e-06, + "loss": 0.85890222, + "num_input_tokens_seen": 29660935, + "step": 1406, + "time_per_iteration": 2.7655117511749268 + }, + { + "auxiliary_loss_clip": 0.01224687, + "auxiliary_loss_mlp": 0.0103662, + "balance_loss_clip": 1.06252742, + "balance_loss_mlp": 1.02434707, + "epoch": 0.16918174712920098, + "flos": 20849340718080.0, + "grad_norm": 2.158712910232086, + "language_loss": 0.85905451, + "learning_rate": 3.8003413500946556e-06, + "loss": 0.88166755, + "num_input_tokens_seen": 29681045, + "step": 1407, + "time_per_iteration": 2.6778764724731445 + }, + { + "auxiliary_loss_clip": 0.01208917, + "auxiliary_loss_mlp": 0.01046159, + "balance_loss_clip": 1.06201649, + "balance_loss_mlp": 1.03320694, + "epoch": 0.1693019900198401, + "flos": 16983270028800.0, + "grad_norm": 2.7043984557362166, + "language_loss": 0.83189929, + "learning_rate": 3.8000019423256216e-06, + "loss": 0.85445005, + "num_input_tokens_seen": 29698810, + "step": 1408, + "time_per_iteration": 2.8078453540802 + }, + { + "auxiliary_loss_clip": 0.01195262, + "auxiliary_loss_mlp": 0.01040804, + "balance_loss_clip": 1.05824351, + "balance_loss_mlp": 1.02896714, + "epoch": 0.16942223291047917, + "flos": 26796901662720.0, + "grad_norm": 5.403282036626413, + "language_loss": 0.88023126, + "learning_rate": 3.7996622614977234e-06, + "loss": 0.90259194, + "num_input_tokens_seen": 29720000, + "step": 1409, + "time_per_iteration": 2.801546096801758 + }, + { + "auxiliary_loss_clip": 0.01208644, + "auxiliary_loss_mlp": 0.0104632, + "balance_loss_clip": 1.06547856, + "balance_loss_mlp": 1.03474474, + "epoch": 0.16954247580111825, + "flos": 18583700411520.0, + "grad_norm": 2.399850325344972, + "language_loss": 0.79183686, + "learning_rate": 3.799322307662492e-06, + "loss": 0.81438655, + "num_input_tokens_seen": 29737820, + "step": 1410, + "time_per_iteration": 2.822844982147217 + }, + { + "auxiliary_loss_clip": 0.01190946, + "auxiliary_loss_mlp": 0.0103656, + "balance_loss_clip": 1.05890322, + "balance_loss_mlp": 1.024436, + "epoch": 0.16966271869175734, + "flos": 13983651210240.0, + "grad_norm": 2.1520614609185227, + "language_loss": 0.83579463, + "learning_rate": 3.798982080871496e-06, + "loss": 0.85806972, + "num_input_tokens_seen": 29752960, + "step": 1411, + "time_per_iteration": 2.853403329849243 + }, + { + "auxiliary_loss_clip": 0.012304, + "auxiliary_loss_mlp": 0.01046502, + "balance_loss_clip": 1.06534839, + "balance_loss_mlp": 1.0339613, + "epoch": 0.16978296158239645, + "flos": 37487328284160.0, + "grad_norm": 2.5524323083091263, + "language_loss": 0.67834604, + "learning_rate": 3.798641581176349e-06, + "loss": 0.70111507, + "num_input_tokens_seen": 29775240, + "step": 1412, + "time_per_iteration": 2.896503210067749 + }, + { + "auxiliary_loss_clip": 0.01208692, + "auxiliary_loss_mlp": 0.01036517, + "balance_loss_clip": 1.05994689, + "balance_loss_mlp": 1.02382123, + "epoch": 0.16990320447303553, + "flos": 28328958506880.0, + "grad_norm": 2.670255795762667, + "language_loss": 0.74702871, + "learning_rate": 3.7983008086287044e-06, + "loss": 0.76948082, + "num_input_tokens_seen": 29796560, + "step": 1413, + "time_per_iteration": 2.8614368438720703 + }, + { + "auxiliary_loss_clip": 0.01207866, + "auxiliary_loss_mlp": 0.01042901, + "balance_loss_clip": 1.06020296, + "balance_loss_mlp": 1.02997911, + "epoch": 0.1700234473636746, + "flos": 20188189031040.0, + "grad_norm": 2.2606163878706202, + "language_loss": 0.79312199, + "learning_rate": 3.797959763280257e-06, + "loss": 0.81562966, + "num_input_tokens_seen": 29815245, + "step": 1414, + "time_per_iteration": 2.770942449569702 + }, + { + "auxiliary_loss_clip": 0.01218193, + "auxiliary_loss_mlp": 0.01039174, + "balance_loss_clip": 1.06173122, + "balance_loss_mlp": 1.02697909, + "epoch": 0.17014369025431372, + "flos": 24858658846080.0, + "grad_norm": 2.1273302639822758, + "language_loss": 0.7937153, + "learning_rate": 3.797618445182743e-06, + "loss": 0.81628895, + "num_input_tokens_seen": 29836640, + "step": 1415, + "time_per_iteration": 2.8081672191619873 + }, + { + "auxiliary_loss_clip": 0.01181046, + "auxiliary_loss_mlp": 0.01042951, + "balance_loss_clip": 1.0536567, + "balance_loss_mlp": 1.03175092, + "epoch": 0.1702639331449528, + "flos": 16467233287680.0, + "grad_norm": 2.1948835294512294, + "language_loss": 0.85090458, + "learning_rate": 3.79727685438794e-06, + "loss": 0.87314451, + "num_input_tokens_seen": 29850830, + "step": 1416, + "time_per_iteration": 2.880753755569458 + }, + { + "auxiliary_loss_clip": 0.01122536, + "auxiliary_loss_mlp": 0.01005047, + "balance_loss_clip": 1.03431153, + "balance_loss_mlp": 1.00018299, + "epoch": 0.1703841760355919, + "flos": 52508870979840.0, + "grad_norm": 0.8400853339344341, + "language_loss": 0.61661816, + "learning_rate": 3.796934990947667e-06, + "loss": 0.63789397, + "num_input_tokens_seen": 29912515, + "step": 1417, + "time_per_iteration": 3.3184807300567627 + }, + { + "auxiliary_loss_clip": 0.01121228, + "auxiliary_loss_mlp": 0.01005387, + "balance_loss_clip": 1.03519011, + "balance_loss_mlp": 1.0004276, + "epoch": 0.170504418926231, + "flos": 49370637576960.0, + "grad_norm": 0.882460789711294, + "language_loss": 0.62517869, + "learning_rate": 3.7965928549137854e-06, + "loss": 0.64644486, + "num_input_tokens_seen": 29969330, + "step": 1418, + "time_per_iteration": 3.2884511947631836 + }, + { + "auxiliary_loss_clip": 0.01203487, + "auxiliary_loss_mlp": 0.01038747, + "balance_loss_clip": 1.05568457, + "balance_loss_mlp": 1.02595007, + "epoch": 0.17062466181687008, + "flos": 25849219184640.0, + "grad_norm": 2.3588272688681253, + "language_loss": 0.77545571, + "learning_rate": 3.7962504463381953e-06, + "loss": 0.79787809, + "num_input_tokens_seen": 29990820, + "step": 1419, + "time_per_iteration": 2.9340245723724365 + }, + { + "auxiliary_loss_clip": 0.01198621, + "auxiliary_loss_mlp": 0.00888817, + "balance_loss_clip": 1.06184769, + "balance_loss_mlp": 1.00035191, + "epoch": 0.17074490470750917, + "flos": 20960412549120.0, + "grad_norm": 2.1643961241022027, + "language_loss": 0.78847611, + "learning_rate": 3.7959077652728412e-06, + "loss": 0.80935061, + "num_input_tokens_seen": 30009275, + "step": 1420, + "time_per_iteration": 2.8627355098724365 + }, + { + "auxiliary_loss_clip": 0.01209446, + "auxiliary_loss_mlp": 0.01041554, + "balance_loss_clip": 1.05931854, + "balance_loss_mlp": 1.02957332, + "epoch": 0.17086514759814825, + "flos": 20959766104320.0, + "grad_norm": 2.2504120109609165, + "language_loss": 0.78108871, + "learning_rate": 3.795564811769707e-06, + "loss": 0.80359864, + "num_input_tokens_seen": 30027630, + "step": 1421, + "time_per_iteration": 2.810136556625366 + }, + { + "auxiliary_loss_clip": 0.0121296, + "auxiliary_loss_mlp": 0.01038435, + "balance_loss_clip": 1.06627011, + "balance_loss_mlp": 1.02662778, + "epoch": 0.17098539048878736, + "flos": 28474073452800.0, + "grad_norm": 2.087767829422252, + "language_loss": 0.78161919, + "learning_rate": 3.795221585880818e-06, + "loss": 0.80413318, + "num_input_tokens_seen": 30048310, + "step": 1422, + "time_per_iteration": 2.8882505893707275 + }, + { + "auxiliary_loss_clip": 0.01199454, + "auxiliary_loss_mlp": 0.01040681, + "balance_loss_clip": 1.06512368, + "balance_loss_mlp": 1.02824712, + "epoch": 0.17110563337942644, + "flos": 16290014561280.0, + "grad_norm": 1.863831861995175, + "language_loss": 0.91165429, + "learning_rate": 3.794878087658242e-06, + "loss": 0.93405557, + "num_input_tokens_seen": 30066080, + "step": 1423, + "time_per_iteration": 4.84348201751709 + }, + { + "auxiliary_loss_clip": 0.01220007, + "auxiliary_loss_mlp": 0.01045262, + "balance_loss_clip": 1.06135201, + "balance_loss_mlp": 1.0331738, + "epoch": 0.17122587627006552, + "flos": 29674207693440.0, + "grad_norm": 1.926399101833357, + "language_loss": 0.78574342, + "learning_rate": 3.7945343171540873e-06, + "loss": 0.8083961, + "num_input_tokens_seen": 30086955, + "step": 1424, + "time_per_iteration": 3.738327980041504 + }, + { + "auxiliary_loss_clip": 0.01230811, + "auxiliary_loss_mlp": 0.01041811, + "balance_loss_clip": 1.06363118, + "balance_loss_mlp": 1.02940679, + "epoch": 0.17134611916070464, + "flos": 25338389915520.0, + "grad_norm": 2.018743944020469, + "language_loss": 0.78859049, + "learning_rate": 3.7941902744205033e-06, + "loss": 0.81131673, + "num_input_tokens_seen": 30107990, + "step": 1425, + "time_per_iteration": 2.749796152114868 + }, + { + "auxiliary_loss_clip": 0.01219378, + "auxiliary_loss_mlp": 0.01048716, + "balance_loss_clip": 1.06497884, + "balance_loss_mlp": 1.03532839, + "epoch": 0.17146636205134372, + "flos": 13953845900160.0, + "grad_norm": 2.824990695672281, + "language_loss": 0.83643281, + "learning_rate": 3.7938459595096817e-06, + "loss": 0.85911375, + "num_input_tokens_seen": 30126535, + "step": 1426, + "time_per_iteration": 2.813532590866089 + }, + { + "auxiliary_loss_clip": 0.01227366, + "auxiliary_loss_mlp": 0.01048639, + "balance_loss_clip": 1.06338382, + "balance_loss_mlp": 1.03538918, + "epoch": 0.1715866049419828, + "flos": 23915214172800.0, + "grad_norm": 1.7729726010912952, + "language_loss": 0.86195076, + "learning_rate": 3.7935013724738545e-06, + "loss": 0.88471079, + "num_input_tokens_seen": 30147035, + "step": 1427, + "time_per_iteration": 2.7987849712371826 + }, + { + "auxiliary_loss_clip": 0.01215028, + "auxiliary_loss_mlp": 0.01046098, + "balance_loss_clip": 1.06192493, + "balance_loss_mlp": 1.03415298, + "epoch": 0.17170684783262188, + "flos": 22709369669760.0, + "grad_norm": 1.8579607550617563, + "language_loss": 0.78383803, + "learning_rate": 3.7931565133652945e-06, + "loss": 0.80644929, + "num_input_tokens_seen": 30167110, + "step": 1428, + "time_per_iteration": 2.8740928173065186 + }, + { + "auxiliary_loss_clip": 0.01228147, + "auxiliary_loss_mlp": 0.01034702, + "balance_loss_clip": 1.06209993, + "balance_loss_mlp": 1.0226258, + "epoch": 0.171827090723261, + "flos": 26613290315520.0, + "grad_norm": 2.2325539291166234, + "language_loss": 0.67827088, + "learning_rate": 3.792811382236317e-06, + "loss": 0.70089936, + "num_input_tokens_seen": 30185620, + "step": 1429, + "time_per_iteration": 2.7062766551971436 + }, + { + "auxiliary_loss_clip": 0.01220905, + "auxiliary_loss_mlp": 0.0104464, + "balance_loss_clip": 1.06171465, + "balance_loss_mlp": 1.03121066, + "epoch": 0.17194733361390008, + "flos": 28148507556480.0, + "grad_norm": 1.9095627198862961, + "language_loss": 0.78697646, + "learning_rate": 3.792465979139279e-06, + "loss": 0.80963194, + "num_input_tokens_seen": 30208225, + "step": 1430, + "time_per_iteration": 2.8459856510162354 + }, + { + "auxiliary_loss_clip": 0.01093671, + "auxiliary_loss_mlp": 0.01010605, + "balance_loss_clip": 1.02597594, + "balance_loss_mlp": 1.006194, + "epoch": 0.17206757650453916, + "flos": 65530689753600.0, + "grad_norm": 0.9241116780858771, + "language_loss": 0.65674967, + "learning_rate": 3.792120304126576e-06, + "loss": 0.67779243, + "num_input_tokens_seen": 30271600, + "step": 1431, + "time_per_iteration": 3.364241361618042 + }, + { + "auxiliary_loss_clip": 0.01170383, + "auxiliary_loss_mlp": 0.01042763, + "balance_loss_clip": 1.05529499, + "balance_loss_mlp": 1.03011537, + "epoch": 0.17218781939517827, + "flos": 22273486128000.0, + "grad_norm": 1.9004510631759677, + "language_loss": 0.83421373, + "learning_rate": 3.791774357250649e-06, + "loss": 0.85634524, + "num_input_tokens_seen": 30290430, + "step": 1432, + "time_per_iteration": 2.892085313796997 + }, + { + "auxiliary_loss_clip": 0.01206515, + "auxiliary_loss_mlp": 0.0104822, + "balance_loss_clip": 1.06275868, + "balance_loss_mlp": 1.03591132, + "epoch": 0.17230806228581735, + "flos": 14137313592960.0, + "grad_norm": 2.3455139084492256, + "language_loss": 0.79551435, + "learning_rate": 3.7914281385639757e-06, + "loss": 0.81806171, + "num_input_tokens_seen": 30308305, + "step": 1433, + "time_per_iteration": 2.8412482738494873 + }, + { + "auxiliary_loss_clip": 0.01216098, + "auxiliary_loss_mlp": 0.0103405, + "balance_loss_clip": 1.05831361, + "balance_loss_mlp": 1.02295208, + "epoch": 0.17242830517645644, + "flos": 20704836303360.0, + "grad_norm": 1.886784186705972, + "language_loss": 0.79483932, + "learning_rate": 3.7910816481190784e-06, + "loss": 0.81734085, + "num_input_tokens_seen": 30328120, + "step": 1434, + "time_per_iteration": 2.7031056880950928 + }, + { + "auxiliary_loss_clip": 0.01198085, + "auxiliary_loss_mlp": 0.01044798, + "balance_loss_clip": 1.05807018, + "balance_loss_mlp": 1.03256691, + "epoch": 0.17254854806709552, + "flos": 30774582887040.0, + "grad_norm": 2.09608501852623, + "language_loss": 0.75050008, + "learning_rate": 3.7907348859685193e-06, + "loss": 0.77292883, + "num_input_tokens_seen": 30349825, + "step": 1435, + "time_per_iteration": 2.7887299060821533 + }, + { + "auxiliary_loss_clip": 0.01216708, + "auxiliary_loss_mlp": 0.01042629, + "balance_loss_clip": 1.06574976, + "balance_loss_mlp": 1.02994514, + "epoch": 0.17266879095773463, + "flos": 26614726859520.0, + "grad_norm": 1.9325783040585647, + "language_loss": 0.80338985, + "learning_rate": 3.790387852164902e-06, + "loss": 0.82598323, + "num_input_tokens_seen": 30370555, + "step": 1436, + "time_per_iteration": 2.754953145980835 + }, + { + "auxiliary_loss_clip": 0.0121738, + "auxiliary_loss_mlp": 0.01043401, + "balance_loss_clip": 1.06104302, + "balance_loss_mlp": 1.03079498, + "epoch": 0.1727890338483737, + "flos": 20266295155200.0, + "grad_norm": 2.1309789956797887, + "language_loss": 0.77186507, + "learning_rate": 3.7900405467608707e-06, + "loss": 0.79447293, + "num_input_tokens_seen": 30390100, + "step": 1437, + "time_per_iteration": 2.668574333190918 + }, + { + "auxiliary_loss_clip": 0.0117017, + "auxiliary_loss_mlp": 0.01032269, + "balance_loss_clip": 1.0484246, + "balance_loss_mlp": 1.02022922, + "epoch": 0.1729092767390128, + "flos": 18179812909440.0, + "grad_norm": 3.071552526415478, + "language_loss": 0.79761791, + "learning_rate": 3.7896929698091114e-06, + "loss": 0.81964231, + "num_input_tokens_seen": 30402915, + "step": 1438, + "time_per_iteration": 2.908543825149536 + }, + { + "auxiliary_loss_clip": 0.01234835, + "auxiliary_loss_mlp": 0.01043143, + "balance_loss_clip": 1.06860328, + "balance_loss_mlp": 1.03094172, + "epoch": 0.1730295196296519, + "flos": 26759518583040.0, + "grad_norm": 2.970720122595041, + "language_loss": 0.67908531, + "learning_rate": 3.7893451213623518e-06, + "loss": 0.70186508, + "num_input_tokens_seen": 30420145, + "step": 1439, + "time_per_iteration": 2.686133861541748 + }, + { + "auxiliary_loss_clip": 0.01220878, + "auxiliary_loss_mlp": 0.00889351, + "balance_loss_clip": 1.06683481, + "balance_loss_mlp": 1.0004791, + "epoch": 0.173149762520291, + "flos": 23842531002240.0, + "grad_norm": 2.3166507826263145, + "language_loss": 0.81965673, + "learning_rate": 3.7889970014733606e-06, + "loss": 0.84075904, + "num_input_tokens_seen": 30439250, + "step": 1440, + "time_per_iteration": 2.7074742317199707 + }, + { + "auxiliary_loss_clip": 0.01168051, + "auxiliary_loss_mlp": 0.01042553, + "balance_loss_clip": 1.05110073, + "balance_loss_mlp": 1.03018522, + "epoch": 0.17327000541093007, + "flos": 23368186972800.0, + "grad_norm": 1.579997961354916, + "language_loss": 0.77668542, + "learning_rate": 3.7886486101949463e-06, + "loss": 0.79879141, + "num_input_tokens_seen": 30460430, + "step": 1441, + "time_per_iteration": 2.8627004623413086 + }, + { + "auxiliary_loss_clip": 0.01168168, + "auxiliary_loss_mlp": 0.01039425, + "balance_loss_clip": 1.04798651, + "balance_loss_mlp": 1.02666378, + "epoch": 0.17339024830156918, + "flos": 18221290139520.0, + "grad_norm": 2.194049848987385, + "language_loss": 0.88351357, + "learning_rate": 3.7882999475799594e-06, + "loss": 0.90558952, + "num_input_tokens_seen": 30478465, + "step": 1442, + "time_per_iteration": 2.8858096599578857 + }, + { + "auxiliary_loss_clip": 0.01170491, + "auxiliary_loss_mlp": 0.01039586, + "balance_loss_clip": 1.05544388, + "balance_loss_mlp": 1.02699161, + "epoch": 0.17351049119220827, + "flos": 23332024955520.0, + "grad_norm": 1.6939682030402103, + "language_loss": 0.8128351, + "learning_rate": 3.787951013681293e-06, + "loss": 0.8349359, + "num_input_tokens_seen": 30496510, + "step": 1443, + "time_per_iteration": 2.8636634349823 + }, + { + "auxiliary_loss_clip": 0.01216826, + "auxiliary_loss_mlp": 0.01037088, + "balance_loss_clip": 1.05975747, + "balance_loss_mlp": 1.02480936, + "epoch": 0.17363073408284735, + "flos": 23803495896960.0, + "grad_norm": 2.0999649633474298, + "language_loss": 0.7738589, + "learning_rate": 3.787601808551879e-06, + "loss": 0.79639804, + "num_input_tokens_seen": 30516325, + "step": 1444, + "time_per_iteration": 2.7401716709136963 + }, + { + "auxiliary_loss_clip": 0.01197014, + "auxiliary_loss_mlp": 0.01037333, + "balance_loss_clip": 1.05701029, + "balance_loss_mlp": 1.02516782, + "epoch": 0.17375097697348643, + "flos": 18515290959360.0, + "grad_norm": 2.7321449447614516, + "language_loss": 0.84215462, + "learning_rate": 3.7872523322446926e-06, + "loss": 0.86449802, + "num_input_tokens_seen": 30535210, + "step": 1445, + "time_per_iteration": 2.8459811210632324 + }, + { + "auxiliary_loss_clip": 0.01190143, + "auxiliary_loss_mlp": 0.01042477, + "balance_loss_clip": 1.05209255, + "balance_loss_mlp": 1.02993631, + "epoch": 0.17387121986412554, + "flos": 38877897456000.0, + "grad_norm": 1.7189857858502129, + "language_loss": 0.60094851, + "learning_rate": 3.7869025848127478e-06, + "loss": 0.62327474, + "num_input_tokens_seen": 30559405, + "step": 1446, + "time_per_iteration": 3.0350492000579834 + }, + { + "auxiliary_loss_clip": 0.0121961, + "auxiliary_loss_mlp": 0.01039134, + "balance_loss_clip": 1.06098938, + "balance_loss_mlp": 1.02709389, + "epoch": 0.17399146275476463, + "flos": 20375714960640.0, + "grad_norm": 3.1164383724160936, + "language_loss": 0.80712819, + "learning_rate": 3.786552566309102e-06, + "loss": 0.82971561, + "num_input_tokens_seen": 30577615, + "step": 1447, + "time_per_iteration": 2.7774600982666016 + }, + { + "auxiliary_loss_clip": 0.01204144, + "auxiliary_loss_mlp": 0.00888835, + "balance_loss_clip": 1.06259024, + "balance_loss_mlp": 1.00046146, + "epoch": 0.1741117056454037, + "flos": 19164339763200.0, + "grad_norm": 2.2515270633343465, + "language_loss": 0.86172402, + "learning_rate": 3.7862022767868517e-06, + "loss": 0.88265377, + "num_input_tokens_seen": 30595205, + "step": 1448, + "time_per_iteration": 2.8345460891723633 + }, + { + "auxiliary_loss_clip": 0.01195037, + "auxiliary_loss_mlp": 0.01040987, + "balance_loss_clip": 1.06525099, + "balance_loss_mlp": 1.02876806, + "epoch": 0.17423194853604282, + "flos": 25374300537600.0, + "grad_norm": 2.1215801568814663, + "language_loss": 0.84248418, + "learning_rate": 3.7858517162991367e-06, + "loss": 0.86484438, + "num_input_tokens_seen": 30615280, + "step": 1449, + "time_per_iteration": 3.8151345252990723 + }, + { + "auxiliary_loss_clip": 0.01193713, + "auxiliary_loss_mlp": 0.01046764, + "balance_loss_clip": 1.05390608, + "balance_loss_mlp": 1.0350219, + "epoch": 0.1743521914266819, + "flos": 25191874339200.0, + "grad_norm": 2.364868652589402, + "language_loss": 0.61219251, + "learning_rate": 3.7855008848991363e-06, + "loss": 0.6345973, + "num_input_tokens_seen": 30633485, + "step": 1450, + "time_per_iteration": 4.6712446212768555 + }, + { + "auxiliary_loss_clip": 0.01204922, + "auxiliary_loss_mlp": 0.010391, + "balance_loss_clip": 1.06249285, + "balance_loss_mlp": 1.02725649, + "epoch": 0.17447243431732098, + "flos": 25666577504640.0, + "grad_norm": 1.8457113956423055, + "language_loss": 0.78005463, + "learning_rate": 3.7851497826400714e-06, + "loss": 0.80249488, + "num_input_tokens_seen": 30653625, + "step": 1451, + "time_per_iteration": 2.7590856552124023 + }, + { + "auxiliary_loss_clip": 0.01227562, + "auxiliary_loss_mlp": 0.01038955, + "balance_loss_clip": 1.06451106, + "balance_loss_mlp": 1.02668786, + "epoch": 0.17459267720796007, + "flos": 36281950657920.0, + "grad_norm": 16.337235678642912, + "language_loss": 0.75905061, + "learning_rate": 3.7847984095752034e-06, + "loss": 0.78171575, + "num_input_tokens_seen": 30677080, + "step": 1452, + "time_per_iteration": 2.7959203720092773 + }, + { + "auxiliary_loss_clip": 0.01226047, + "auxiliary_loss_mlp": 0.0103997, + "balance_loss_clip": 1.06163216, + "balance_loss_mlp": 1.02751875, + "epoch": 0.17471292009859918, + "flos": 20011113959040.0, + "grad_norm": 2.0106082238065577, + "language_loss": 0.80549753, + "learning_rate": 3.784446765757836e-06, + "loss": 0.82815766, + "num_input_tokens_seen": 30695725, + "step": 1453, + "time_per_iteration": 2.7236979007720947 + }, + { + "auxiliary_loss_clip": 0.011827, + "auxiliary_loss_mlp": 0.01046184, + "balance_loss_clip": 1.05579925, + "balance_loss_mlp": 1.03419757, + "epoch": 0.17483316298923826, + "flos": 27819242559360.0, + "grad_norm": 2.570417362349676, + "language_loss": 0.77793491, + "learning_rate": 3.7840948512413133e-06, + "loss": 0.80022371, + "num_input_tokens_seen": 30713310, + "step": 1454, + "time_per_iteration": 2.8270363807678223 + }, + { + "auxiliary_loss_clip": 0.01192828, + "auxiliary_loss_mlp": 0.0104303, + "balance_loss_clip": 1.06235123, + "balance_loss_mlp": 1.03038168, + "epoch": 0.17495340587987734, + "flos": 44017934791680.0, + "grad_norm": 1.9039877451613845, + "language_loss": 0.78746969, + "learning_rate": 3.7837426660790196e-06, + "loss": 0.80982828, + "num_input_tokens_seen": 30734725, + "step": 1455, + "time_per_iteration": 2.9752604961395264 + }, + { + "auxiliary_loss_clip": 0.0122437, + "auxiliary_loss_mlp": 0.01044232, + "balance_loss_clip": 1.06163096, + "balance_loss_mlp": 1.03242469, + "epoch": 0.17507364877051645, + "flos": 20885825957760.0, + "grad_norm": 2.7622975000682146, + "language_loss": 0.81527174, + "learning_rate": 3.783390210324382e-06, + "loss": 0.83795774, + "num_input_tokens_seen": 30754450, + "step": 1456, + "time_per_iteration": 2.662802219390869 + }, + { + "auxiliary_loss_clip": 0.01197417, + "auxiliary_loss_mlp": 0.01034775, + "balance_loss_clip": 1.06100142, + "balance_loss_mlp": 1.02294326, + "epoch": 0.17519389166115554, + "flos": 24717602136960.0, + "grad_norm": 2.0233725861739162, + "language_loss": 0.72755814, + "learning_rate": 3.7830374840308676e-06, + "loss": 0.74988008, + "num_input_tokens_seen": 30774605, + "step": 1457, + "time_per_iteration": 2.838620901107788 + }, + { + "auxiliary_loss_clip": 0.01218769, + "auxiliary_loss_mlp": 0.01038076, + "balance_loss_clip": 1.06238067, + "balance_loss_mlp": 1.02549934, + "epoch": 0.17531413455179462, + "flos": 23798144770560.0, + "grad_norm": 2.4939324658628035, + "language_loss": 0.8262676, + "learning_rate": 3.7826844872519842e-06, + "loss": 0.84883606, + "num_input_tokens_seen": 30792460, + "step": 1458, + "time_per_iteration": 2.7543747425079346 + }, + { + "auxiliary_loss_clip": 0.01205466, + "auxiliary_loss_mlp": 0.01039555, + "balance_loss_clip": 1.0626595, + "balance_loss_mlp": 1.02840257, + "epoch": 0.1754343774424337, + "flos": 24572379450240.0, + "grad_norm": 2.1502747417953207, + "language_loss": 0.72778141, + "learning_rate": 3.782331220041282e-06, + "loss": 0.75023162, + "num_input_tokens_seen": 30812525, + "step": 1459, + "time_per_iteration": 2.8784027099609375 + }, + { + "auxiliary_loss_clip": 0.01203171, + "auxiliary_loss_mlp": 0.01040545, + "balance_loss_clip": 1.05630207, + "balance_loss_mlp": 1.02772951, + "epoch": 0.17555462033307281, + "flos": 18114599767680.0, + "grad_norm": 3.0628424906423324, + "language_loss": 0.82683706, + "learning_rate": 3.7819776824523504e-06, + "loss": 0.84927416, + "num_input_tokens_seen": 30830390, + "step": 1460, + "time_per_iteration": 2.868727684020996 + }, + { + "auxiliary_loss_clip": 0.01213458, + "auxiliary_loss_mlp": 0.01043704, + "balance_loss_clip": 1.05994582, + "balance_loss_mlp": 1.03149652, + "epoch": 0.1756748632237119, + "flos": 28366018364160.0, + "grad_norm": 4.089937532880594, + "language_loss": 0.83994114, + "learning_rate": 3.7816238745388213e-06, + "loss": 0.86251277, + "num_input_tokens_seen": 30849935, + "step": 1461, + "time_per_iteration": 2.831122398376465 + }, + { + "auxiliary_loss_clip": 0.01208971, + "auxiliary_loss_mlp": 0.01044406, + "balance_loss_clip": 1.05618536, + "balance_loss_mlp": 1.03216863, + "epoch": 0.17579510611435098, + "flos": 25732939881600.0, + "grad_norm": 1.8651600994764577, + "language_loss": 0.87137526, + "learning_rate": 3.781269796354367e-06, + "loss": 0.89390904, + "num_input_tokens_seen": 30869555, + "step": 1462, + "time_per_iteration": 2.7806246280670166 + }, + { + "auxiliary_loss_clip": 0.0120713, + "auxiliary_loss_mlp": 0.01041887, + "balance_loss_clip": 1.06052327, + "balance_loss_mlp": 1.03032386, + "epoch": 0.1759153490049901, + "flos": 18588081870720.0, + "grad_norm": 1.9375953316293728, + "language_loss": 0.8611719, + "learning_rate": 3.7809154479527006e-06, + "loss": 0.8836621, + "num_input_tokens_seen": 30888760, + "step": 1463, + "time_per_iteration": 2.7720818519592285 + }, + { + "auxiliary_loss_clip": 0.01192264, + "auxiliary_loss_mlp": 0.01042292, + "balance_loss_clip": 1.05961716, + "balance_loss_mlp": 1.03065705, + "epoch": 0.17603559189562917, + "flos": 18619323724800.0, + "grad_norm": 2.594950162380946, + "language_loss": 0.84153622, + "learning_rate": 3.780560829387577e-06, + "loss": 0.86388183, + "num_input_tokens_seen": 30907260, + "step": 1464, + "time_per_iteration": 2.8086302280426025 + }, + { + "auxiliary_loss_clip": 0.01119385, + "auxiliary_loss_mlp": 0.01009872, + "balance_loss_clip": 1.03612137, + "balance_loss_mlp": 1.00577128, + "epoch": 0.17615583478626826, + "flos": 60530775373440.0, + "grad_norm": 0.8754094300831374, + "language_loss": 0.57977492, + "learning_rate": 3.7802059407127915e-06, + "loss": 0.60106748, + "num_input_tokens_seen": 30965810, + "step": 1465, + "time_per_iteration": 3.280996084213257 + }, + { + "auxiliary_loss_clip": 0.01193778, + "auxiliary_loss_mlp": 0.01036639, + "balance_loss_clip": 1.05330777, + "balance_loss_mlp": 1.02360392, + "epoch": 0.17627607767690734, + "flos": 23616221362560.0, + "grad_norm": 2.277324348317436, + "language_loss": 0.86316681, + "learning_rate": 3.7798507819821797e-06, + "loss": 0.88547099, + "num_input_tokens_seen": 30982935, + "step": 1466, + "time_per_iteration": 2.9042470455169678 + }, + { + "auxiliary_loss_clip": 0.01182921, + "auxiliary_loss_mlp": 0.01052424, + "balance_loss_clip": 1.05540657, + "balance_loss_mlp": 1.04019928, + "epoch": 0.17639632056754645, + "flos": 17639070589440.0, + "grad_norm": 2.179785481477165, + "language_loss": 0.78856993, + "learning_rate": 3.7794953532496197e-06, + "loss": 0.81092334, + "num_input_tokens_seen": 30998840, + "step": 1467, + "time_per_iteration": 2.814340829849243 + }, + { + "auxiliary_loss_clip": 0.01101253, + "auxiliary_loss_mlp": 0.00879575, + "balance_loss_clip": 1.04702258, + "balance_loss_mlp": 1.00081909, + "epoch": 0.17651656345818553, + "flos": 57932604910080.0, + "grad_norm": 0.8654031078585924, + "language_loss": 0.57916772, + "learning_rate": 3.7791396545690295e-06, + "loss": 0.59897602, + "num_input_tokens_seen": 31060075, + "step": 1468, + "time_per_iteration": 3.3814163208007812 + }, + { + "auxiliary_loss_clip": 0.0121758, + "auxiliary_loss_mlp": 0.01042076, + "balance_loss_clip": 1.06393337, + "balance_loss_mlp": 1.02979708, + "epoch": 0.17663680634882462, + "flos": 22929502170240.0, + "grad_norm": 2.4912475688841704, + "language_loss": 0.81247663, + "learning_rate": 3.7787836859943685e-06, + "loss": 0.83507323, + "num_input_tokens_seen": 31078800, + "step": 1469, + "time_per_iteration": 2.8219566345214844 + }, + { + "auxiliary_loss_clip": 0.0121523, + "auxiliary_loss_mlp": 0.01036035, + "balance_loss_clip": 1.05990994, + "balance_loss_mlp": 1.02358925, + "epoch": 0.17675704923946373, + "flos": 22637979388800.0, + "grad_norm": 2.670463265538787, + "language_loss": 0.79600191, + "learning_rate": 3.7784274475796363e-06, + "loss": 0.81851459, + "num_input_tokens_seen": 31097430, + "step": 1470, + "time_per_iteration": 2.700810670852661 + }, + { + "auxiliary_loss_clip": 0.01197567, + "auxiliary_loss_mlp": 0.01040918, + "balance_loss_clip": 1.05575287, + "balance_loss_mlp": 1.02785897, + "epoch": 0.1768772921301028, + "flos": 27126525795840.0, + "grad_norm": 2.2730812598973453, + "language_loss": 0.76222754, + "learning_rate": 3.7780709393788745e-06, + "loss": 0.78461242, + "num_input_tokens_seen": 31117905, + "step": 1471, + "time_per_iteration": 2.8221044540405273 + }, + { + "auxiliary_loss_clip": 0.0122564, + "auxiliary_loss_mlp": 0.01042584, + "balance_loss_clip": 1.06254101, + "balance_loss_mlp": 1.03023362, + "epoch": 0.1769975350207419, + "flos": 19172133014400.0, + "grad_norm": 2.104866717391452, + "language_loss": 0.75392592, + "learning_rate": 3.777714161446165e-06, + "loss": 0.77660823, + "num_input_tokens_seen": 31137610, + "step": 1472, + "time_per_iteration": 2.6303935050964355 + }, + { + "auxiliary_loss_clip": 0.01216264, + "auxiliary_loss_mlp": 0.0103595, + "balance_loss_clip": 1.06074691, + "balance_loss_mlp": 1.02438641, + "epoch": 0.177117777911381, + "flos": 36134932291200.0, + "grad_norm": 2.4082472187457404, + "language_loss": 0.69591796, + "learning_rate": 3.7773571138356304e-06, + "loss": 0.71844006, + "num_input_tokens_seen": 31157780, + "step": 1473, + "time_per_iteration": 2.8613717555999756 + }, + { + "auxiliary_loss_clip": 0.01168683, + "auxiliary_loss_mlp": 0.01034328, + "balance_loss_clip": 1.0558908, + "balance_loss_mlp": 1.02297354, + "epoch": 0.17723802080202009, + "flos": 22090593052800.0, + "grad_norm": 2.685061525404297, + "language_loss": 0.89106297, + "learning_rate": 3.776999796601435e-06, + "loss": 0.91309321, + "num_input_tokens_seen": 31176540, + "step": 1474, + "time_per_iteration": 2.9017598628997803 + }, + { + "auxiliary_loss_clip": 0.01218606, + "auxiliary_loss_mlp": 0.01044143, + "balance_loss_clip": 1.06092727, + "balance_loss_mlp": 1.03222799, + "epoch": 0.17735826369265917, + "flos": 30222671437440.0, + "grad_norm": 1.990095953116785, + "language_loss": 0.72814244, + "learning_rate": 3.776642209797783e-06, + "loss": 0.75076991, + "num_input_tokens_seen": 31198370, + "step": 1475, + "time_per_iteration": 5.559632062911987 + }, + { + "auxiliary_loss_clip": 0.01211487, + "auxiliary_loss_mlp": 0.01050942, + "balance_loss_clip": 1.0590837, + "balance_loss_mlp": 1.03727448, + "epoch": 0.17747850658329825, + "flos": 21397588980480.0, + "grad_norm": 2.2934143636746045, + "language_loss": 0.78010654, + "learning_rate": 3.7762843534789205e-06, + "loss": 0.80273092, + "num_input_tokens_seen": 31217120, + "step": 1476, + "time_per_iteration": 3.618450164794922 + }, + { + "auxiliary_loss_clip": 0.01210264, + "auxiliary_loss_mlp": 0.01038451, + "balance_loss_clip": 1.05812931, + "balance_loss_mlp": 1.02632165, + "epoch": 0.17759874947393736, + "flos": 16983341856000.0, + "grad_norm": 3.20542513204412, + "language_loss": 0.88570058, + "learning_rate": 3.7759262276991343e-06, + "loss": 0.90818769, + "num_input_tokens_seen": 31234730, + "step": 1477, + "time_per_iteration": 2.786414384841919 + }, + { + "auxiliary_loss_clip": 0.01209281, + "auxiliary_loss_mlp": 0.01038478, + "balance_loss_clip": 1.05979455, + "balance_loss_mlp": 1.02570498, + "epoch": 0.17771899236457644, + "flos": 11546107390080.0, + "grad_norm": 2.35775929230562, + "language_loss": 0.80732381, + "learning_rate": 3.7755678325127506e-06, + "loss": 0.82980144, + "num_input_tokens_seen": 31252410, + "step": 1478, + "time_per_iteration": 2.7496137619018555 + }, + { + "auxiliary_loss_clip": 0.01180918, + "auxiliary_loss_mlp": 0.01038413, + "balance_loss_clip": 1.05912888, + "balance_loss_mlp": 1.02729034, + "epoch": 0.17783923525521553, + "flos": 18807747494400.0, + "grad_norm": 1.920754154892285, + "language_loss": 0.75720763, + "learning_rate": 3.7752091679741393e-06, + "loss": 0.77940094, + "num_input_tokens_seen": 31270200, + "step": 1479, + "time_per_iteration": 2.885176658630371 + }, + { + "auxiliary_loss_clip": 0.01213959, + "auxiliary_loss_mlp": 0.01038699, + "balance_loss_clip": 1.06254125, + "balance_loss_mlp": 1.02715957, + "epoch": 0.17795947814585464, + "flos": 30408365773440.0, + "grad_norm": 5.171938424731877, + "language_loss": 0.76944971, + "learning_rate": 3.774850234137708e-06, + "loss": 0.79197633, + "num_input_tokens_seen": 31287495, + "step": 1480, + "time_per_iteration": 2.806206703186035 + }, + { + "auxiliary_loss_clip": 0.01209028, + "auxiliary_loss_mlp": 0.01041522, + "balance_loss_clip": 1.0575968, + "balance_loss_mlp": 1.02914166, + "epoch": 0.17807972103649372, + "flos": 24389055411840.0, + "grad_norm": 2.4825708105142694, + "language_loss": 0.82551265, + "learning_rate": 3.7744910310579076e-06, + "loss": 0.84801817, + "num_input_tokens_seen": 31306420, + "step": 1481, + "time_per_iteration": 2.7384753227233887 + }, + { + "auxiliary_loss_clip": 0.01223197, + "auxiliary_loss_mlp": 0.01045993, + "balance_loss_clip": 1.06183696, + "balance_loss_mlp": 1.0340302, + "epoch": 0.1781999639271328, + "flos": 20301559332480.0, + "grad_norm": 1.949750201683896, + "language_loss": 0.85339344, + "learning_rate": 3.774131558789229e-06, + "loss": 0.87608534, + "num_input_tokens_seen": 31325750, + "step": 1482, + "time_per_iteration": 2.754702091217041 + }, + { + "auxiliary_loss_clip": 0.01224698, + "auxiliary_loss_mlp": 0.00888437, + "balance_loss_clip": 1.06270301, + "balance_loss_mlp": 1.00064588, + "epoch": 0.1783202068177719, + "flos": 15924479806080.0, + "grad_norm": 3.345200743342026, + "language_loss": 0.70277977, + "learning_rate": 3.773771817386203e-06, + "loss": 0.72391105, + "num_input_tokens_seen": 31343080, + "step": 1483, + "time_per_iteration": 2.6813154220581055 + }, + { + "auxiliary_loss_clip": 0.01202259, + "auxiliary_loss_mlp": 0.0104103, + "balance_loss_clip": 1.0588994, + "balance_loss_mlp": 1.02970552, + "epoch": 0.178440449708411, + "flos": 20631758083200.0, + "grad_norm": 1.8344748984392638, + "language_loss": 0.79036462, + "learning_rate": 3.773411806903403e-06, + "loss": 0.81279749, + "num_input_tokens_seen": 31362160, + "step": 1484, + "time_per_iteration": 2.8152191638946533 + }, + { + "auxiliary_loss_clip": 0.01175119, + "auxiliary_loss_mlp": 0.01040942, + "balance_loss_clip": 1.05334139, + "balance_loss_mlp": 1.02883649, + "epoch": 0.17856069259905008, + "flos": 21686059105920.0, + "grad_norm": 1.9239391254818707, + "language_loss": 0.94840467, + "learning_rate": 3.7730515273954415e-06, + "loss": 0.97056532, + "num_input_tokens_seen": 31380770, + "step": 1485, + "time_per_iteration": 2.8860437870025635 + }, + { + "auxiliary_loss_clip": 0.0122347, + "auxiliary_loss_mlp": 0.01039267, + "balance_loss_clip": 1.06148779, + "balance_loss_mlp": 1.02745914, + "epoch": 0.17868093548968916, + "flos": 26572962320640.0, + "grad_norm": 2.1166539546117944, + "language_loss": 0.85360277, + "learning_rate": 3.772690978916973e-06, + "loss": 0.87623012, + "num_input_tokens_seen": 31400525, + "step": 1486, + "time_per_iteration": 2.734508752822876 + }, + { + "auxiliary_loss_clip": 0.01210661, + "auxiliary_loss_mlp": 0.0104039, + "balance_loss_clip": 1.06022763, + "balance_loss_mlp": 1.02866578, + "epoch": 0.17880117838032827, + "flos": 18581006891520.0, + "grad_norm": 6.228205135926535, + "language_loss": 0.86412245, + "learning_rate": 3.772330161522693e-06, + "loss": 0.88663292, + "num_input_tokens_seen": 31418435, + "step": 1487, + "time_per_iteration": 2.7585062980651855 + }, + { + "auxiliary_loss_clip": 0.01201462, + "auxiliary_loss_mlp": 0.01041459, + "balance_loss_clip": 1.06024051, + "balance_loss_mlp": 1.0291146, + "epoch": 0.17892142127096736, + "flos": 26541217676160.0, + "grad_norm": 2.199953373546828, + "language_loss": 0.79638362, + "learning_rate": 3.7719690752673365e-06, + "loss": 0.81881285, + "num_input_tokens_seen": 31439230, + "step": 1488, + "time_per_iteration": 2.8205416202545166 + }, + { + "auxiliary_loss_clip": 0.01192994, + "auxiliary_loss_mlp": 0.01048409, + "balance_loss_clip": 1.05735672, + "balance_loss_mlp": 1.03663707, + "epoch": 0.17904166416160644, + "flos": 23872623621120.0, + "grad_norm": 2.371518618548893, + "language_loss": 0.78332996, + "learning_rate": 3.7716077202056796e-06, + "loss": 0.80574399, + "num_input_tokens_seen": 31457705, + "step": 1489, + "time_per_iteration": 2.854801654815674 + }, + { + "auxiliary_loss_clip": 0.01189793, + "auxiliary_loss_mlp": 0.01041293, + "balance_loss_clip": 1.05710864, + "balance_loss_mlp": 1.02908051, + "epoch": 0.17916190705224552, + "flos": 19134426712320.0, + "grad_norm": 2.631632264888276, + "language_loss": 0.93632919, + "learning_rate": 3.7712460963925404e-06, + "loss": 0.9586401, + "num_input_tokens_seen": 31473645, + "step": 1490, + "time_per_iteration": 2.833702802658081 + }, + { + "auxiliary_loss_clip": 0.01193168, + "auxiliary_loss_mlp": 0.01047262, + "balance_loss_clip": 1.05345595, + "balance_loss_mlp": 1.03633595, + "epoch": 0.17928214994288463, + "flos": 25152120961920.0, + "grad_norm": 2.3862060516080494, + "language_loss": 0.75426948, + "learning_rate": 3.7708842038827775e-06, + "loss": 0.77667379, + "num_input_tokens_seen": 31492605, + "step": 1491, + "time_per_iteration": 2.8500983715057373 + }, + { + "auxiliary_loss_clip": 0.01212345, + "auxiliary_loss_mlp": 0.01037588, + "balance_loss_clip": 1.05940914, + "balance_loss_mlp": 1.0259825, + "epoch": 0.17940239283352372, + "flos": 22384629786240.0, + "grad_norm": 2.379526910580724, + "language_loss": 0.85741436, + "learning_rate": 3.770522042731288e-06, + "loss": 0.87991369, + "num_input_tokens_seen": 31514500, + "step": 1492, + "time_per_iteration": 2.8045036792755127 + }, + { + "auxiliary_loss_clip": 0.01175092, + "auxiliary_loss_mlp": 0.0103936, + "balance_loss_clip": 1.05392802, + "balance_loss_mlp": 1.02683091, + "epoch": 0.1795226357241628, + "flos": 23178685795200.0, + "grad_norm": 6.184715918820152, + "language_loss": 0.87673676, + "learning_rate": 3.7701596129930122e-06, + "loss": 0.89888132, + "num_input_tokens_seen": 31533225, + "step": 1493, + "time_per_iteration": 2.899286985397339 + }, + { + "auxiliary_loss_clip": 0.01198618, + "auxiliary_loss_mlp": 0.01033434, + "balance_loss_clip": 1.05674839, + "balance_loss_mlp": 1.02184153, + "epoch": 0.1796428786148019, + "flos": 22090413484800.0, + "grad_norm": 2.0989578192198213, + "language_loss": 0.73790836, + "learning_rate": 3.7697969147229315e-06, + "loss": 0.76022887, + "num_input_tokens_seen": 31551385, + "step": 1494, + "time_per_iteration": 2.839174747467041 + }, + { + "auxiliary_loss_clip": 0.01208267, + "auxiliary_loss_mlp": 0.01031179, + "balance_loss_clip": 1.05779922, + "balance_loss_mlp": 1.02037835, + "epoch": 0.179763121505441, + "flos": 21324618501120.0, + "grad_norm": 2.453502132188676, + "language_loss": 0.85246265, + "learning_rate": 3.7694339479760647e-06, + "loss": 0.87485707, + "num_input_tokens_seen": 31570415, + "step": 1495, + "time_per_iteration": 2.718094825744629 + }, + { + "auxiliary_loss_clip": 0.01103448, + "auxiliary_loss_mlp": 0.01017374, + "balance_loss_clip": 1.02998662, + "balance_loss_mlp": 1.0127728, + "epoch": 0.17988336439608008, + "flos": 68161864815360.0, + "grad_norm": 0.974133471508356, + "language_loss": 0.57330519, + "learning_rate": 3.769070712807476e-06, + "loss": 0.59451336, + "num_input_tokens_seen": 31632445, + "step": 1496, + "time_per_iteration": 3.392759323120117 + }, + { + "auxiliary_loss_clip": 0.01158035, + "auxiliary_loss_mlp": 0.01044974, + "balance_loss_clip": 1.05126441, + "balance_loss_mlp": 1.03291583, + "epoch": 0.18000360728671919, + "flos": 21945047143680.0, + "grad_norm": 1.9179530382556964, + "language_loss": 0.79331505, + "learning_rate": 3.768707209272266e-06, + "loss": 0.81534517, + "num_input_tokens_seen": 31652575, + "step": 1497, + "time_per_iteration": 2.87998104095459 + }, + { + "auxiliary_loss_clip": 0.01193998, + "auxiliary_loss_mlp": 0.0104301, + "balance_loss_clip": 1.05432868, + "balance_loss_mlp": 1.03128529, + "epoch": 0.18012385017735827, + "flos": 18986330937600.0, + "grad_norm": 2.6919700946984766, + "language_loss": 0.76810825, + "learning_rate": 3.768343437425579e-06, + "loss": 0.79047835, + "num_input_tokens_seen": 31671145, + "step": 1498, + "time_per_iteration": 2.852261781692505 + }, + { + "auxiliary_loss_clip": 0.01149665, + "auxiliary_loss_mlp": 0.01041212, + "balance_loss_clip": 1.05058622, + "balance_loss_mlp": 1.02849793, + "epoch": 0.18024409306799735, + "flos": 19748103598080.0, + "grad_norm": 2.471218019080731, + "language_loss": 0.85865819, + "learning_rate": 3.7679793973225987e-06, + "loss": 0.8805669, + "num_input_tokens_seen": 31686955, + "step": 1499, + "time_per_iteration": 2.9374773502349854 + }, + { + "auxiliary_loss_clip": 0.0108637, + "auxiliary_loss_mlp": 0.01005434, + "balance_loss_clip": 1.02695417, + "balance_loss_mlp": 1.0011425, + "epoch": 0.18036433595863643, + "flos": 67227183060480.0, + "grad_norm": 0.8484712543911963, + "language_loss": 0.61596072, + "learning_rate": 3.767615089018549e-06, + "loss": 0.63687879, + "num_input_tokens_seen": 31749300, + "step": 1500, + "time_per_iteration": 3.4560751914978027 + }, + { + "auxiliary_loss_clip": 0.01197143, + "auxiliary_loss_mlp": 0.01038239, + "balance_loss_clip": 1.05715239, + "balance_loss_mlp": 1.02606201, + "epoch": 0.18048457884927555, + "flos": 18181464935040.0, + "grad_norm": 2.6477752309822375, + "language_loss": 0.86417681, + "learning_rate": 3.7672505125686966e-06, + "loss": 0.88653064, + "num_input_tokens_seen": 31765665, + "step": 1501, + "time_per_iteration": 3.9581139087677 + }, + { + "auxiliary_loss_clip": 0.01179392, + "auxiliary_loss_mlp": 0.0104396, + "balance_loss_clip": 1.05244803, + "balance_loss_mlp": 1.03206861, + "epoch": 0.18060482173991463, + "flos": 15813767111040.0, + "grad_norm": 4.321653159871716, + "language_loss": 0.84727591, + "learning_rate": 3.7668856680283455e-06, + "loss": 0.8695094, + "num_input_tokens_seen": 31782690, + "step": 1502, + "time_per_iteration": 3.8216118812561035 + }, + { + "auxiliary_loss_clip": 0.0120794, + "auxiliary_loss_mlp": 0.01037206, + "balance_loss_clip": 1.05969453, + "balance_loss_mlp": 1.02524376, + "epoch": 0.1807250646305537, + "flos": 18587399512320.0, + "grad_norm": 1.8443622084921778, + "language_loss": 0.82572305, + "learning_rate": 3.7665205554528437e-06, + "loss": 0.84817451, + "num_input_tokens_seen": 31802045, + "step": 1503, + "time_per_iteration": 2.7345359325408936 + }, + { + "auxiliary_loss_clip": 0.01200784, + "auxiliary_loss_mlp": 0.01041927, + "balance_loss_clip": 1.05811226, + "balance_loss_mlp": 1.03075099, + "epoch": 0.18084530752119282, + "flos": 23149131880320.0, + "grad_norm": 1.8159100276196445, + "language_loss": 0.73908103, + "learning_rate": 3.7661551748975782e-06, + "loss": 0.76150805, + "num_input_tokens_seen": 31820220, + "step": 1504, + "time_per_iteration": 2.810887098312378 + }, + { + "auxiliary_loss_clip": 0.01102285, + "auxiliary_loss_mlp": 0.01004865, + "balance_loss_clip": 1.02753127, + "balance_loss_mlp": 1.00043094, + "epoch": 0.1809655504118319, + "flos": 59803153568640.0, + "grad_norm": 0.8193095168263518, + "language_loss": 0.60401046, + "learning_rate": 3.7657895264179772e-06, + "loss": 0.62508196, + "num_input_tokens_seen": 31876195, + "step": 1505, + "time_per_iteration": 3.307446002960205 + }, + { + "auxiliary_loss_clip": 0.01194358, + "auxiliary_loss_mlp": 0.01035024, + "balance_loss_clip": 1.05511069, + "balance_loss_mlp": 1.02318668, + "epoch": 0.181085793302471, + "flos": 44201941188480.0, + "grad_norm": 2.0980763921771186, + "language_loss": 0.74628675, + "learning_rate": 3.765423610069509e-06, + "loss": 0.76858062, + "num_input_tokens_seen": 31901585, + "step": 1506, + "time_per_iteration": 2.9353675842285156 + }, + { + "auxiliary_loss_clip": 0.01203734, + "auxiliary_loss_mlp": 0.01038286, + "balance_loss_clip": 1.06000721, + "balance_loss_mlp": 1.02712774, + "epoch": 0.18120603619311007, + "flos": 34898384638080.0, + "grad_norm": 2.2915570379156986, + "language_loss": 0.72612095, + "learning_rate": 3.765057425907683e-06, + "loss": 0.74854124, + "num_input_tokens_seen": 31923045, + "step": 1507, + "time_per_iteration": 2.892399787902832 + }, + { + "auxiliary_loss_clip": 0.01215492, + "auxiliary_loss_mlp": 0.01040041, + "balance_loss_clip": 1.05959463, + "balance_loss_mlp": 1.0289371, + "epoch": 0.18132627908374918, + "flos": 21506757390720.0, + "grad_norm": 1.858905210757723, + "language_loss": 0.78532791, + "learning_rate": 3.764690973988048e-06, + "loss": 0.80788326, + "num_input_tokens_seen": 31943385, + "step": 1508, + "time_per_iteration": 2.776662826538086 + }, + { + "auxiliary_loss_clip": 0.01194232, + "auxiliary_loss_mlp": 0.0104272, + "balance_loss_clip": 1.05594921, + "balance_loss_mlp": 1.03157449, + "epoch": 0.18144652197438826, + "flos": 29057693633280.0, + "grad_norm": 1.8401307568892817, + "language_loss": 0.73714358, + "learning_rate": 3.7643242543661967e-06, + "loss": 0.75951308, + "num_input_tokens_seen": 31966045, + "step": 1509, + "time_per_iteration": 2.895097255706787 + }, + { + "auxiliary_loss_clip": 0.01092025, + "auxiliary_loss_mlp": 0.01015154, + "balance_loss_clip": 1.02414405, + "balance_loss_mlp": 1.01079142, + "epoch": 0.18156676486502735, + "flos": 68675064382080.0, + "grad_norm": 0.8169634733804053, + "language_loss": 0.60570645, + "learning_rate": 3.7639572670977573e-06, + "loss": 0.62677824, + "num_input_tokens_seen": 32021540, + "step": 1510, + "time_per_iteration": 3.202047348022461 + }, + { + "auxiliary_loss_clip": 0.01193609, + "auxiliary_loss_mlp": 0.01036469, + "balance_loss_clip": 1.05543303, + "balance_loss_mlp": 1.02407753, + "epoch": 0.18168700775566646, + "flos": 26471515334400.0, + "grad_norm": 1.711641947567614, + "language_loss": 0.76883322, + "learning_rate": 3.7635900122384042e-06, + "loss": 0.791134, + "num_input_tokens_seen": 32044535, + "step": 1511, + "time_per_iteration": 2.9288878440856934 + }, + { + "auxiliary_loss_clip": 0.01205615, + "auxiliary_loss_mlp": 0.01037201, + "balance_loss_clip": 1.05738091, + "balance_loss_mlp": 1.02486908, + "epoch": 0.18180725064630554, + "flos": 15005668884480.0, + "grad_norm": 2.5361348954756875, + "language_loss": 0.87059295, + "learning_rate": 3.7632224898438477e-06, + "loss": 0.89302117, + "num_input_tokens_seen": 32061010, + "step": 1512, + "time_per_iteration": 2.7151551246643066 + }, + { + "auxiliary_loss_clip": 0.01193616, + "auxiliary_loss_mlp": 0.01040272, + "balance_loss_clip": 1.05438161, + "balance_loss_mlp": 1.02888131, + "epoch": 0.18192749353694462, + "flos": 19682387665920.0, + "grad_norm": 1.6443451454967934, + "language_loss": 0.7939198, + "learning_rate": 3.762854699969842e-06, + "loss": 0.81625867, + "num_input_tokens_seen": 32081520, + "step": 1513, + "time_per_iteration": 2.8389909267425537 + }, + { + "auxiliary_loss_clip": 0.01208908, + "auxiliary_loss_mlp": 0.01039117, + "balance_loss_clip": 1.05980456, + "balance_loss_mlp": 1.02738118, + "epoch": 0.1820477364275837, + "flos": 20702717400960.0, + "grad_norm": 2.1295748006166577, + "language_loss": 0.7312386, + "learning_rate": 3.762486642672179e-06, + "loss": 0.75371885, + "num_input_tokens_seen": 32098460, + "step": 1514, + "time_per_iteration": 2.838003158569336 + }, + { + "auxiliary_loss_clip": 0.01207802, + "auxiliary_loss_mlp": 0.01041382, + "balance_loss_clip": 1.06197691, + "balance_loss_mlp": 1.02906823, + "epoch": 0.18216797931822282, + "flos": 17128708197120.0, + "grad_norm": 2.3232208821467712, + "language_loss": 0.86947489, + "learning_rate": 3.7621183180066946e-06, + "loss": 0.8919667, + "num_input_tokens_seen": 32116420, + "step": 1515, + "time_per_iteration": 2.7461957931518555 + }, + { + "auxiliary_loss_clip": 0.01195892, + "auxiliary_loss_mlp": 0.01038069, + "balance_loss_clip": 1.05315161, + "balance_loss_mlp": 1.02658904, + "epoch": 0.1822882222088619, + "flos": 29242561956480.0, + "grad_norm": 1.6040165114318774, + "language_loss": 0.73904896, + "learning_rate": 3.7617497260292625e-06, + "loss": 0.76138854, + "num_input_tokens_seen": 32138475, + "step": 1516, + "time_per_iteration": 2.8602147102355957 + }, + { + "auxiliary_loss_clip": 0.01198461, + "auxiliary_loss_mlp": 0.01036268, + "balance_loss_clip": 1.06014276, + "balance_loss_mlp": 1.02445412, + "epoch": 0.18240846509950098, + "flos": 17702739446400.0, + "grad_norm": 3.016480916676586, + "language_loss": 0.78775418, + "learning_rate": 3.7613808667957967e-06, + "loss": 0.81010139, + "num_input_tokens_seen": 32151165, + "step": 1517, + "time_per_iteration": 2.859668254852295 + }, + { + "auxiliary_loss_clip": 0.0120152, + "auxiliary_loss_mlp": 0.01038861, + "balance_loss_clip": 1.0551523, + "balance_loss_mlp": 1.02618921, + "epoch": 0.1825287079901401, + "flos": 14790025584000.0, + "grad_norm": 2.903981378259532, + "language_loss": 0.91519618, + "learning_rate": 3.7610117403622547e-06, + "loss": 0.93760002, + "num_input_tokens_seen": 32167725, + "step": 1518, + "time_per_iteration": 2.7619423866271973 + }, + { + "auxiliary_loss_clip": 0.01179464, + "auxiliary_loss_mlp": 0.01037058, + "balance_loss_clip": 1.05205107, + "balance_loss_mlp": 1.02488697, + "epoch": 0.18264895088077918, + "flos": 21946232292480.0, + "grad_norm": 1.7849005240839668, + "language_loss": 0.89996731, + "learning_rate": 3.7606423467846313e-06, + "loss": 0.92213249, + "num_input_tokens_seen": 32187330, + "step": 1519, + "time_per_iteration": 2.8196635246276855 + }, + { + "auxiliary_loss_clip": 0.01197806, + "auxiliary_loss_mlp": 0.01043951, + "balance_loss_clip": 1.05969071, + "balance_loss_mlp": 1.03238201, + "epoch": 0.18276919377141826, + "flos": 20886759711360.0, + "grad_norm": 1.7576227076754265, + "language_loss": 0.7952953, + "learning_rate": 3.760272686118964e-06, + "loss": 0.8177129, + "num_input_tokens_seen": 32205550, + "step": 1520, + "time_per_iteration": 2.986266613006592 + }, + { + "auxiliary_loss_clip": 0.01202947, + "auxiliary_loss_mlp": 0.01038016, + "balance_loss_clip": 1.05781102, + "balance_loss_mlp": 1.0266614, + "epoch": 0.18288943666205737, + "flos": 21469877101440.0, + "grad_norm": 2.828133670712384, + "language_loss": 0.92319858, + "learning_rate": 3.7599027584213297e-06, + "loss": 0.9456082, + "num_input_tokens_seen": 32224430, + "step": 1521, + "time_per_iteration": 2.756992816925049 + }, + { + "auxiliary_loss_clip": 0.0121541, + "auxiliary_loss_mlp": 0.0103452, + "balance_loss_clip": 1.05962288, + "balance_loss_mlp": 1.0225935, + "epoch": 0.18300967955269645, + "flos": 21539363961600.0, + "grad_norm": 2.5834466654355355, + "language_loss": 0.78301466, + "learning_rate": 3.7595325637478465e-06, + "loss": 0.80551398, + "num_input_tokens_seen": 32242455, + "step": 1522, + "time_per_iteration": 2.7625184059143066 + }, + { + "auxiliary_loss_clip": 0.01191892, + "auxiliary_loss_mlp": 0.01041753, + "balance_loss_clip": 1.05724835, + "balance_loss_mlp": 1.02958775, + "epoch": 0.18312992244333554, + "flos": 28876237102080.0, + "grad_norm": 2.2115338949676846, + "language_loss": 0.81729805, + "learning_rate": 3.7591621021546723e-06, + "loss": 0.83963454, + "num_input_tokens_seen": 32264450, + "step": 1523, + "time_per_iteration": 2.831477642059326 + }, + { + "auxiliary_loss_clip": 0.01201363, + "auxiliary_loss_mlp": 0.01047324, + "balance_loss_clip": 1.05521679, + "balance_loss_mlp": 1.03513527, + "epoch": 0.18325016533397462, + "flos": 20120102801280.0, + "grad_norm": 1.749782094253012, + "language_loss": 0.81733382, + "learning_rate": 3.7587913736980062e-06, + "loss": 0.83982068, + "num_input_tokens_seen": 32284090, + "step": 1524, + "time_per_iteration": 2.7103097438812256 + }, + { + "auxiliary_loss_clip": 0.0115412, + "auxiliary_loss_mlp": 0.01039511, + "balance_loss_clip": 1.04719937, + "balance_loss_mlp": 1.02800131, + "epoch": 0.18337040822461373, + "flos": 23329187781120.0, + "grad_norm": 1.8440565404832427, + "language_loss": 0.84531856, + "learning_rate": 3.7584203784340865e-06, + "loss": 0.86725485, + "num_input_tokens_seen": 32303260, + "step": 1525, + "time_per_iteration": 2.9300191402435303 + }, + { + "auxiliary_loss_clip": 0.01194004, + "auxiliary_loss_mlp": 0.01033193, + "balance_loss_clip": 1.0543381, + "balance_loss_mlp": 1.02167714, + "epoch": 0.1834906511152528, + "flos": 25009555881600.0, + "grad_norm": 2.8204529149139606, + "language_loss": 0.86005324, + "learning_rate": 3.7580491164191938e-06, + "loss": 0.88232523, + "num_input_tokens_seen": 32321570, + "step": 1526, + "time_per_iteration": 2.794649124145508 + }, + { + "auxiliary_loss_clip": 0.01108143, + "auxiliary_loss_mlp": 0.01007084, + "balance_loss_clip": 1.02726245, + "balance_loss_mlp": 1.00322211, + "epoch": 0.1836108940058919, + "flos": 67251493589760.0, + "grad_norm": 0.7476205648465274, + "language_loss": 0.6123659, + "learning_rate": 3.757677587709648e-06, + "loss": 0.63351822, + "num_input_tokens_seen": 32384835, + "step": 1527, + "time_per_iteration": 5.317498207092285 + }, + { + "auxiliary_loss_clip": 0.01183962, + "auxiliary_loss_mlp": 0.01035965, + "balance_loss_clip": 1.05566883, + "balance_loss_mlp": 1.02410352, + "epoch": 0.183731136896531, + "flos": 25738721971200.0, + "grad_norm": 2.274389040701668, + "language_loss": 0.75574303, + "learning_rate": 3.7573057923618095e-06, + "loss": 0.7779423, + "num_input_tokens_seen": 32404930, + "step": 1528, + "time_per_iteration": 3.7460875511169434 + }, + { + "auxiliary_loss_clip": 0.01173798, + "auxiliary_loss_mlp": 0.01039519, + "balance_loss_clip": 1.05020869, + "balance_loss_mlp": 1.02688873, + "epoch": 0.1838513797871701, + "flos": 20449403712000.0, + "grad_norm": 2.2299844887232774, + "language_loss": 0.74242854, + "learning_rate": 3.7569337304320793e-06, + "loss": 0.76456165, + "num_input_tokens_seen": 32424515, + "step": 1529, + "time_per_iteration": 2.955972194671631 + }, + { + "auxiliary_loss_clip": 0.0109787, + "auxiliary_loss_mlp": 0.01006522, + "balance_loss_clip": 1.02498448, + "balance_loss_mlp": 1.00256431, + "epoch": 0.18397162267780917, + "flos": 68565141786240.0, + "grad_norm": 0.837433517230852, + "language_loss": 0.6446448, + "learning_rate": 3.756561401976899e-06, + "loss": 0.66568869, + "num_input_tokens_seen": 32484220, + "step": 1530, + "time_per_iteration": 3.2361674308776855 + }, + { + "auxiliary_loss_clip": 0.01222168, + "auxiliary_loss_mlp": 0.01038161, + "balance_loss_clip": 1.06150186, + "balance_loss_mlp": 1.02668762, + "epoch": 0.18409186556844825, + "flos": 31941104976000.0, + "grad_norm": 1.876302699476226, + "language_loss": 0.8255161, + "learning_rate": 3.7561888070527514e-06, + "loss": 0.84811938, + "num_input_tokens_seen": 32506260, + "step": 1531, + "time_per_iteration": 2.900223970413208 + }, + { + "auxiliary_loss_clip": 0.01175222, + "auxiliary_loss_mlp": 0.00888203, + "balance_loss_clip": 1.05281901, + "balance_loss_mlp": 1.00080955, + "epoch": 0.18421210845908736, + "flos": 20120533764480.0, + "grad_norm": 4.48746780997002, + "language_loss": 0.801283, + "learning_rate": 3.7558159457161577e-06, + "loss": 0.8219173, + "num_input_tokens_seen": 32524225, + "step": 1532, + "time_per_iteration": 2.84513783454895 + }, + { + "auxiliary_loss_clip": 0.0119966, + "auxiliary_loss_mlp": 0.00888105, + "balance_loss_clip": 1.05613947, + "balance_loss_mlp": 1.00059795, + "epoch": 0.18433235134972645, + "flos": 23110491824640.0, + "grad_norm": 3.2373634490246723, + "language_loss": 0.77525437, + "learning_rate": 3.755442818023681e-06, + "loss": 0.79613197, + "num_input_tokens_seen": 32543850, + "step": 1533, + "time_per_iteration": 2.747309684753418 + }, + { + "auxiliary_loss_clip": 0.01187379, + "auxiliary_loss_mlp": 0.01040952, + "balance_loss_clip": 1.05471742, + "balance_loss_mlp": 1.02935851, + "epoch": 0.18445259424036553, + "flos": 18291351617280.0, + "grad_norm": 2.1605172743893855, + "language_loss": 0.76196808, + "learning_rate": 3.7550694240319246e-06, + "loss": 0.78425133, + "num_input_tokens_seen": 32561725, + "step": 1534, + "time_per_iteration": 2.778454542160034 + }, + { + "auxiliary_loss_clip": 0.01214698, + "auxiliary_loss_mlp": 0.01042436, + "balance_loss_clip": 1.06018436, + "balance_loss_mlp": 1.03098571, + "epoch": 0.18457283713100464, + "flos": 21324079797120.0, + "grad_norm": 2.2574798562508804, + "language_loss": 0.76619804, + "learning_rate": 3.7546957637975326e-06, + "loss": 0.78876948, + "num_input_tokens_seen": 32579135, + "step": 1535, + "time_per_iteration": 2.7676947116851807 + }, + { + "auxiliary_loss_clip": 0.0115508, + "auxiliary_loss_mlp": 0.01040958, + "balance_loss_clip": 1.04372907, + "balance_loss_mlp": 1.02998519, + "epoch": 0.18469308002164372, + "flos": 20375679047040.0, + "grad_norm": 1.5723407326001146, + "language_loss": 0.74166363, + "learning_rate": 3.7543218373771873e-06, + "loss": 0.76362407, + "num_input_tokens_seen": 32598460, + "step": 1536, + "time_per_iteration": 2.9386613368988037 + }, + { + "auxiliary_loss_clip": 0.0115725, + "auxiliary_loss_mlp": 0.00887686, + "balance_loss_clip": 1.04747152, + "balance_loss_mlp": 1.00075519, + "epoch": 0.1848133229122828, + "flos": 26435892021120.0, + "grad_norm": 1.4555815932050873, + "language_loss": 0.78263241, + "learning_rate": 3.753947644827615e-06, + "loss": 0.80308175, + "num_input_tokens_seen": 32621920, + "step": 1537, + "time_per_iteration": 3.0396924018859863 + }, + { + "auxiliary_loss_clip": 0.01098003, + "auxiliary_loss_mlp": 0.0100499, + "balance_loss_clip": 1.02315605, + "balance_loss_mlp": 1.00096059, + "epoch": 0.1849335658029219, + "flos": 70547447612160.0, + "grad_norm": 0.922449186617107, + "language_loss": 0.57165146, + "learning_rate": 3.753573186205579e-06, + "loss": 0.59268129, + "num_input_tokens_seen": 32690040, + "step": 1538, + "time_per_iteration": 3.5640013217926025 + }, + { + "auxiliary_loss_clip": 0.01187681, + "auxiliary_loss_mlp": 0.00888102, + "balance_loss_clip": 1.05276608, + "balance_loss_mlp": 1.00057542, + "epoch": 0.185053808693561, + "flos": 17384140788480.0, + "grad_norm": 2.4823598662170987, + "language_loss": 0.77741289, + "learning_rate": 3.753198461567885e-06, + "loss": 0.79817069, + "num_input_tokens_seen": 32707285, + "step": 1539, + "time_per_iteration": 2.8929247856140137 + }, + { + "auxiliary_loss_clip": 0.01179275, + "auxiliary_loss_mlp": 0.01034021, + "balance_loss_clip": 1.05637002, + "balance_loss_mlp": 1.02398992, + "epoch": 0.18517405158420008, + "flos": 28986159697920.0, + "grad_norm": 1.8737908885315961, + "language_loss": 0.91968799, + "learning_rate": 3.7528234709713783e-06, + "loss": 0.94182092, + "num_input_tokens_seen": 32730030, + "step": 1540, + "time_per_iteration": 2.831134796142578 + }, + { + "auxiliary_loss_clip": 0.01214092, + "auxiliary_loss_mlp": 0.01039787, + "balance_loss_clip": 1.06125617, + "balance_loss_mlp": 1.0290345, + "epoch": 0.18529429447483917, + "flos": 26794962328320.0, + "grad_norm": 2.3572518443248205, + "language_loss": 0.84541345, + "learning_rate": 3.7524482144729447e-06, + "loss": 0.86795229, + "num_input_tokens_seen": 32749485, + "step": 1541, + "time_per_iteration": 2.8129401206970215 + }, + { + "auxiliary_loss_clip": 0.01178051, + "auxiliary_loss_mlp": 0.0104232, + "balance_loss_clip": 1.05003393, + "balance_loss_mlp": 1.03104234, + "epoch": 0.18541453736547828, + "flos": 13581595301760.0, + "grad_norm": 3.01789868629759, + "language_loss": 0.83382285, + "learning_rate": 3.7520726921295106e-06, + "loss": 0.85602659, + "num_input_tokens_seen": 32766205, + "step": 1542, + "time_per_iteration": 2.8382785320281982 + }, + { + "auxiliary_loss_clip": 0.01201107, + "auxiliary_loss_mlp": 0.0103856, + "balance_loss_clip": 1.05278873, + "balance_loss_mlp": 1.0271039, + "epoch": 0.18553478025611736, + "flos": 24025424077440.0, + "grad_norm": 2.43308293112884, + "language_loss": 0.7239911, + "learning_rate": 3.751696903998042e-06, + "loss": 0.74638778, + "num_input_tokens_seen": 32784840, + "step": 1543, + "time_per_iteration": 2.8117613792419434 + }, + { + "auxiliary_loss_clip": 0.01207164, + "auxiliary_loss_mlp": 0.01034861, + "balance_loss_clip": 1.05940664, + "balance_loss_mlp": 1.02408409, + "epoch": 0.18565502314675644, + "flos": 25885165720320.0, + "grad_norm": 1.7626103281015744, + "language_loss": 0.70061255, + "learning_rate": 3.7513208501355456e-06, + "loss": 0.72303277, + "num_input_tokens_seen": 32805945, + "step": 1544, + "time_per_iteration": 2.7745087146759033 + }, + { + "auxiliary_loss_clip": 0.01193837, + "auxiliary_loss_mlp": 0.01037061, + "balance_loss_clip": 1.05612528, + "balance_loss_mlp": 1.02702928, + "epoch": 0.18577526603739553, + "flos": 19610063631360.0, + "grad_norm": 1.942575668166768, + "language_loss": 0.84103769, + "learning_rate": 3.750944530599069e-06, + "loss": 0.8633467, + "num_input_tokens_seen": 32825515, + "step": 1545, + "time_per_iteration": 2.833986520767212 + }, + { + "auxiliary_loss_clip": 0.01217668, + "auxiliary_loss_mlp": 0.01041107, + "balance_loss_clip": 1.06229258, + "balance_loss_mlp": 1.02953136, + "epoch": 0.18589550892803464, + "flos": 18474891137280.0, + "grad_norm": 2.149847045945045, + "language_loss": 0.80576378, + "learning_rate": 3.7505679454456992e-06, + "loss": 0.8283515, + "num_input_tokens_seen": 32842125, + "step": 1546, + "time_per_iteration": 2.728579521179199 + }, + { + "auxiliary_loss_clip": 0.01152861, + "auxiliary_loss_mlp": 0.01037402, + "balance_loss_clip": 1.04751778, + "balance_loss_mlp": 1.02583885, + "epoch": 0.18601575181867372, + "flos": 23549966726400.0, + "grad_norm": 2.169749943496553, + "language_loss": 0.70240486, + "learning_rate": 3.750191094732564e-06, + "loss": 0.72430742, + "num_input_tokens_seen": 32862990, + "step": 1547, + "time_per_iteration": 2.9606196880340576 + }, + { + "auxiliary_loss_clip": 0.01148653, + "auxiliary_loss_mlp": 0.00888076, + "balance_loss_clip": 1.04637098, + "balance_loss_mlp": 1.00069594, + "epoch": 0.1861359947093128, + "flos": 26360192108160.0, + "grad_norm": 2.4006453552988187, + "language_loss": 0.75272775, + "learning_rate": 3.7498139785168313e-06, + "loss": 0.77309501, + "num_input_tokens_seen": 32883595, + "step": 1548, + "time_per_iteration": 2.996678590774536 + }, + { + "auxiliary_loss_clip": 0.01205702, + "auxiliary_loss_mlp": 0.01032617, + "balance_loss_clip": 1.05857587, + "balance_loss_mlp": 1.02203703, + "epoch": 0.1862562375999519, + "flos": 23331198942720.0, + "grad_norm": 2.0409709110082095, + "language_loss": 0.77652436, + "learning_rate": 3.749436596855709e-06, + "loss": 0.79890758, + "num_input_tokens_seen": 32902895, + "step": 1549, + "time_per_iteration": 2.73966383934021 + }, + { + "auxiliary_loss_clip": 0.01199217, + "auxiliary_loss_mlp": 0.01048295, + "balance_loss_clip": 1.05479801, + "balance_loss_mlp": 1.03692269, + "epoch": 0.186376480490591, + "flos": 16648222942080.0, + "grad_norm": 2.2937654655458806, + "language_loss": 0.90728807, + "learning_rate": 3.749058949806446e-06, + "loss": 0.9297632, + "num_input_tokens_seen": 32919620, + "step": 1550, + "time_per_iteration": 2.785390853881836 + }, + { + "auxiliary_loss_clip": 0.01208424, + "auxiliary_loss_mlp": 0.01031212, + "balance_loss_clip": 1.05761588, + "balance_loss_mlp": 1.02003002, + "epoch": 0.18649672338123008, + "flos": 21468656039040.0, + "grad_norm": 1.729406241416172, + "language_loss": 0.84263897, + "learning_rate": 3.748681037426331e-06, + "loss": 0.8650353, + "num_input_tokens_seen": 32938830, + "step": 1551, + "time_per_iteration": 2.821143865585327 + }, + { + "auxiliary_loss_clip": 0.01219884, + "auxiliary_loss_mlp": 0.01039016, + "balance_loss_clip": 1.06102824, + "balance_loss_mlp": 1.02805448, + "epoch": 0.1866169662718692, + "flos": 12312728386560.0, + "grad_norm": 2.3227230230171996, + "language_loss": 0.91503668, + "learning_rate": 3.7483028597726936e-06, + "loss": 0.93762571, + "num_input_tokens_seen": 32955600, + "step": 1552, + "time_per_iteration": 2.694141387939453 + }, + { + "auxiliary_loss_clip": 0.01190614, + "auxiliary_loss_mlp": 0.01035021, + "balance_loss_clip": 1.0587616, + "balance_loss_mlp": 1.02387524, + "epoch": 0.18673720916250827, + "flos": 23581280407680.0, + "grad_norm": 1.7612221932904508, + "language_loss": 0.62404937, + "learning_rate": 3.7479244169029017e-06, + "loss": 0.64630568, + "num_input_tokens_seen": 32975390, + "step": 1553, + "time_per_iteration": 5.572770833969116 + }, + { + "auxiliary_loss_clip": 0.01207174, + "auxiliary_loss_mlp": 0.01031852, + "balance_loss_clip": 1.05453587, + "balance_loss_mlp": 1.02096272, + "epoch": 0.18685745205314735, + "flos": 19718370115200.0, + "grad_norm": 3.6744733396040576, + "language_loss": 0.73109549, + "learning_rate": 3.7475457088743658e-06, + "loss": 0.7534858, + "num_input_tokens_seen": 32992640, + "step": 1554, + "time_per_iteration": 2.7391865253448486 + }, + { + "auxiliary_loss_clip": 0.01187743, + "auxiliary_loss_mlp": 0.01032117, + "balance_loss_clip": 1.05629098, + "balance_loss_mlp": 1.02110243, + "epoch": 0.18697769494378644, + "flos": 34204123589760.0, + "grad_norm": 1.8806173481373398, + "language_loss": 0.74444592, + "learning_rate": 3.7471667357445348e-06, + "loss": 0.76664454, + "num_input_tokens_seen": 33012470, + "step": 1555, + "time_per_iteration": 2.871788263320923 + }, + { + "auxiliary_loss_clip": 0.01167762, + "auxiliary_loss_mlp": 0.01034294, + "balance_loss_clip": 1.05343533, + "balance_loss_mlp": 1.02313018, + "epoch": 0.18709793783442555, + "flos": 34241327101440.0, + "grad_norm": 2.0849643231424237, + "language_loss": 0.72350907, + "learning_rate": 3.7467874975709e-06, + "loss": 0.74552965, + "num_input_tokens_seen": 33033275, + "step": 1556, + "time_per_iteration": 3.0158474445343018 + }, + { + "auxiliary_loss_clip": 0.01213199, + "auxiliary_loss_mlp": 0.01033744, + "balance_loss_clip": 1.06065524, + "balance_loss_mlp": 1.02308106, + "epoch": 0.18721818072506463, + "flos": 40734550529280.0, + "grad_norm": 2.8698088648023745, + "language_loss": 0.78218842, + "learning_rate": 3.7464079944109904e-06, + "loss": 0.80465782, + "num_input_tokens_seen": 33055135, + "step": 1557, + "time_per_iteration": 2.9282655715942383 + }, + { + "auxiliary_loss_clip": 0.01190303, + "auxiliary_loss_mlp": 0.01029291, + "balance_loss_clip": 1.05508542, + "balance_loss_mlp": 1.01851439, + "epoch": 0.18733842361570371, + "flos": 22157386392960.0, + "grad_norm": 2.085436862119647, + "language_loss": 0.77434683, + "learning_rate": 3.746028226322376e-06, + "loss": 0.79654276, + "num_input_tokens_seen": 33071015, + "step": 1558, + "time_per_iteration": 2.8433730602264404 + }, + { + "auxiliary_loss_clip": 0.01192681, + "auxiliary_loss_mlp": 0.01030379, + "balance_loss_clip": 1.05461121, + "balance_loss_mlp": 1.0202049, + "epoch": 0.18745866650634282, + "flos": 18914940656640.0, + "grad_norm": 2.035106695605152, + "language_loss": 0.75398171, + "learning_rate": 3.745648193362669e-06, + "loss": 0.77621233, + "num_input_tokens_seen": 33090370, + "step": 1559, + "time_per_iteration": 2.8354744911193848 + }, + { + "auxiliary_loss_clip": 0.01195691, + "auxiliary_loss_mlp": 0.0104179, + "balance_loss_clip": 1.05700016, + "balance_loss_mlp": 1.03082895, + "epoch": 0.1875789093969819, + "flos": 19314626267520.0, + "grad_norm": 2.130268576748478, + "language_loss": 0.72021705, + "learning_rate": 3.745267895589518e-06, + "loss": 0.74259186, + "num_input_tokens_seen": 33108910, + "step": 1560, + "time_per_iteration": 2.786878824234009 + }, + { + "auxiliary_loss_clip": 0.01198173, + "auxiliary_loss_mlp": 0.01034725, + "balance_loss_clip": 1.05832005, + "balance_loss_mlp": 1.02379346, + "epoch": 0.187699152287621, + "flos": 17018965169280.0, + "grad_norm": 2.0165828778932, + "language_loss": 0.82242113, + "learning_rate": 3.7448873330606154e-06, + "loss": 0.84475011, + "num_input_tokens_seen": 33126680, + "step": 1561, + "time_per_iteration": 2.806884527206421 + }, + { + "auxiliary_loss_clip": 0.0117651, + "auxiliary_loss_mlp": 0.01035659, + "balance_loss_clip": 1.05429506, + "balance_loss_mlp": 1.02482879, + "epoch": 0.18781939517826007, + "flos": 22346384780160.0, + "grad_norm": 2.5470756707220295, + "language_loss": 0.87547386, + "learning_rate": 3.7445065058336914e-06, + "loss": 0.89759558, + "num_input_tokens_seen": 33145550, + "step": 1562, + "time_per_iteration": 2.833564519882202 + }, + { + "auxiliary_loss_clip": 0.01163754, + "auxiliary_loss_mlp": 0.01035512, + "balance_loss_clip": 1.04963517, + "balance_loss_mlp": 1.02416968, + "epoch": 0.18793963806889918, + "flos": 14611478054400.0, + "grad_norm": 1.924896261849397, + "language_loss": 0.86725307, + "learning_rate": 3.7441254139665176e-06, + "loss": 0.88924575, + "num_input_tokens_seen": 33161735, + "step": 1563, + "time_per_iteration": 2.806732177734375 + }, + { + "auxiliary_loss_clip": 0.01220503, + "auxiliary_loss_mlp": 0.01036414, + "balance_loss_clip": 1.06409216, + "balance_loss_mlp": 1.02637029, + "epoch": 0.18805988095953827, + "flos": 17457075354240.0, + "grad_norm": 1.9489424450519104, + "language_loss": 0.8267414, + "learning_rate": 3.743744057516905e-06, + "loss": 0.84931052, + "num_input_tokens_seen": 33179795, + "step": 1564, + "time_per_iteration": 2.6774561405181885 + }, + { + "auxiliary_loss_clip": 0.01172474, + "auxiliary_loss_mlp": 0.01033699, + "balance_loss_clip": 1.05075598, + "balance_loss_mlp": 1.02323234, + "epoch": 0.18818012385017735, + "flos": 15043877976960.0, + "grad_norm": 4.288825187802543, + "language_loss": 0.8742851, + "learning_rate": 3.743362436542706e-06, + "loss": 0.89634693, + "num_input_tokens_seen": 33194485, + "step": 1565, + "time_per_iteration": 2.825833797454834 + }, + { + "auxiliary_loss_clip": 0.01215252, + "auxiliary_loss_mlp": 0.01044205, + "balance_loss_clip": 1.05772853, + "balance_loss_mlp": 1.03310084, + "epoch": 0.18830036674081646, + "flos": 47551975136640.0, + "grad_norm": 2.347815931914824, + "language_loss": 0.76921475, + "learning_rate": 3.7429805511018115e-06, + "loss": 0.79180932, + "num_input_tokens_seen": 33216145, + "step": 1566, + "time_per_iteration": 2.9322359561920166 + }, + { + "auxiliary_loss_clip": 0.0118325, + "auxiliary_loss_mlp": 0.0088843, + "balance_loss_clip": 1.05732894, + "balance_loss_mlp": 1.00068641, + "epoch": 0.18842060963145554, + "flos": 30044626698240.0, + "grad_norm": 2.1378023364618883, + "language_loss": 0.78215367, + "learning_rate": 3.7425984012521524e-06, + "loss": 0.80287051, + "num_input_tokens_seen": 33236345, + "step": 1567, + "time_per_iteration": 2.9051971435546875 + }, + { + "auxiliary_loss_clip": 0.010862, + "auxiliary_loss_mlp": 0.00878045, + "balance_loss_clip": 1.02350438, + "balance_loss_mlp": 1.00055885, + "epoch": 0.18854085252209463, + "flos": 70318372625280.0, + "grad_norm": 0.7364084094102432, + "language_loss": 0.60393059, + "learning_rate": 3.7422159870517025e-06, + "loss": 0.62357295, + "num_input_tokens_seen": 33301600, + "step": 1568, + "time_per_iteration": 3.3985414505004883 + }, + { + "auxiliary_loss_clip": 0.0119255, + "auxiliary_loss_mlp": 0.01035298, + "balance_loss_clip": 1.0545342, + "balance_loss_mlp": 1.02499282, + "epoch": 0.1886610954127337, + "flos": 21289318410240.0, + "grad_norm": 1.689158311796734, + "language_loss": 0.7870366, + "learning_rate": 3.7418333085584717e-06, + "loss": 0.80931509, + "num_input_tokens_seen": 33322785, + "step": 1569, + "time_per_iteration": 2.9364538192749023 + }, + { + "auxiliary_loss_clip": 0.01188122, + "auxiliary_loss_mlp": 0.01035549, + "balance_loss_clip": 1.0568639, + "balance_loss_mlp": 1.02471948, + "epoch": 0.18878133830337282, + "flos": 17266819991040.0, + "grad_norm": 2.1543989194475355, + "language_loss": 0.90384054, + "learning_rate": 3.7414503658305128e-06, + "loss": 0.92607731, + "num_input_tokens_seen": 33340020, + "step": 1570, + "time_per_iteration": 3.0545363426208496 + }, + { + "auxiliary_loss_clip": 0.011833, + "auxiliary_loss_mlp": 0.01034363, + "balance_loss_clip": 1.05210686, + "balance_loss_mlp": 1.02277613, + "epoch": 0.1889015811940119, + "flos": 25775207210880.0, + "grad_norm": 2.192660468297324, + "language_loss": 0.77996361, + "learning_rate": 3.7410671589259185e-06, + "loss": 0.80214024, + "num_input_tokens_seen": 33358620, + "step": 1571, + "time_per_iteration": 2.9937283992767334 + }, + { + "auxiliary_loss_clip": 0.01220969, + "auxiliary_loss_mlp": 0.01041788, + "balance_loss_clip": 1.06331956, + "balance_loss_mlp": 1.03066003, + "epoch": 0.18902182408465099, + "flos": 21032197879680.0, + "grad_norm": 2.129374667389495, + "language_loss": 0.79689759, + "learning_rate": 3.7406836879028205e-06, + "loss": 0.81952518, + "num_input_tokens_seen": 33378845, + "step": 1572, + "time_per_iteration": 2.708677053451538 + }, + { + "auxiliary_loss_clip": 0.01201752, + "auxiliary_loss_mlp": 0.01031595, + "balance_loss_clip": 1.05695939, + "balance_loss_mlp": 1.02123535, + "epoch": 0.1891420669752901, + "flos": 22272121411200.0, + "grad_norm": 2.158116227588561, + "language_loss": 0.76775062, + "learning_rate": 3.7402999528193907e-06, + "loss": 0.79008412, + "num_input_tokens_seen": 33398345, + "step": 1573, + "time_per_iteration": 2.7355117797851562 + }, + { + "auxiliary_loss_clip": 0.01174309, + "auxiliary_loss_mlp": 0.00887985, + "balance_loss_clip": 1.05370665, + "balance_loss_mlp": 1.00061226, + "epoch": 0.18926230986592918, + "flos": 22017802141440.0, + "grad_norm": 2.3896191693840128, + "language_loss": 0.85516739, + "learning_rate": 3.739915953733842e-06, + "loss": 0.87579036, + "num_input_tokens_seen": 33416390, + "step": 1574, + "time_per_iteration": 2.8221592903137207 + }, + { + "auxiliary_loss_clip": 0.0121188, + "auxiliary_loss_mlp": 0.01042233, + "balance_loss_clip": 1.05588317, + "balance_loss_mlp": 1.03179622, + "epoch": 0.18938255275656826, + "flos": 24462672336000.0, + "grad_norm": 1.8138758911518729, + "language_loss": 0.81759584, + "learning_rate": 3.7395316907044264e-06, + "loss": 0.840137, + "num_input_tokens_seen": 33437175, + "step": 1575, + "time_per_iteration": 2.6559395790100098 + }, + { + "auxiliary_loss_clip": 0.01207128, + "auxiliary_loss_mlp": 0.01034741, + "balance_loss_clip": 1.05880117, + "balance_loss_mlp": 1.02339864, + "epoch": 0.18950279564720737, + "flos": 24427049022720.0, + "grad_norm": 1.5544251329093202, + "language_loss": 0.7962817, + "learning_rate": 3.7391471637894364e-06, + "loss": 0.81870031, + "num_input_tokens_seen": 33459440, + "step": 1576, + "time_per_iteration": 2.70817494392395 + }, + { + "auxiliary_loss_clip": 0.01185158, + "auxiliary_loss_mlp": 0.01033302, + "balance_loss_clip": 1.05190194, + "balance_loss_mlp": 1.02176285, + "epoch": 0.18962303853784646, + "flos": 19756291898880.0, + "grad_norm": 1.8772310624367163, + "language_loss": 0.84501326, + "learning_rate": 3.738762373047205e-06, + "loss": 0.86719781, + "num_input_tokens_seen": 33479360, + "step": 1577, + "time_per_iteration": 2.7826600074768066 + }, + { + "auxiliary_loss_clip": 0.01190952, + "auxiliary_loss_mlp": 0.01037881, + "balance_loss_clip": 1.05747962, + "balance_loss_mlp": 1.02707493, + "epoch": 0.18974328142848554, + "flos": 21032054225280.0, + "grad_norm": 1.6563135402867544, + "language_loss": 0.83548987, + "learning_rate": 3.738377318536103e-06, + "loss": 0.85777819, + "num_input_tokens_seen": 33499245, + "step": 1578, + "time_per_iteration": 3.7802815437316895 + }, + { + "auxiliary_loss_clip": 0.0121397, + "auxiliary_loss_mlp": 0.01036207, + "balance_loss_clip": 1.06026602, + "balance_loss_mlp": 1.02577019, + "epoch": 0.18986352431912462, + "flos": 12966122736000.0, + "grad_norm": 2.675010348504252, + "language_loss": 0.71570534, + "learning_rate": 3.7379920003145447e-06, + "loss": 0.7382071, + "num_input_tokens_seen": 33513520, + "step": 1579, + "time_per_iteration": 5.291112899780273 + }, + { + "auxiliary_loss_clip": 0.01192016, + "auxiliary_loss_mlp": 0.0103814, + "balance_loss_clip": 1.05824399, + "balance_loss_mlp": 1.02652347, + "epoch": 0.18998376720976373, + "flos": 23767908497280.0, + "grad_norm": 1.7960366598396154, + "language_loss": 0.8341074, + "learning_rate": 3.7376064184409817e-06, + "loss": 0.85640895, + "num_input_tokens_seen": 33533100, + "step": 1580, + "time_per_iteration": 2.714984178543091 + }, + { + "auxiliary_loss_clip": 0.01192097, + "auxiliary_loss_mlp": 0.01032682, + "balance_loss_clip": 1.05412984, + "balance_loss_mlp": 1.0215776, + "epoch": 0.19010401010040281, + "flos": 22966023323520.0, + "grad_norm": 1.4561128963898526, + "language_loss": 0.87092435, + "learning_rate": 3.7372205729739063e-06, + "loss": 0.89317214, + "num_input_tokens_seen": 33554915, + "step": 1581, + "time_per_iteration": 2.741854667663574 + }, + { + "auxiliary_loss_clip": 0.01210709, + "auxiliary_loss_mlp": 0.01033943, + "balance_loss_clip": 1.05854607, + "balance_loss_mlp": 1.02262998, + "epoch": 0.1902242529910419, + "flos": 19135647774720.0, + "grad_norm": 2.4042071471694144, + "language_loss": 0.71971357, + "learning_rate": 3.7368344639718514e-06, + "loss": 0.74216008, + "num_input_tokens_seen": 33572850, + "step": 1582, + "time_per_iteration": 2.688023805618286 + }, + { + "auxiliary_loss_clip": 0.01205747, + "auxiliary_loss_mlp": 0.01042349, + "balance_loss_clip": 1.05694914, + "balance_loss_mlp": 1.03240728, + "epoch": 0.190344495881681, + "flos": 25483935824640.0, + "grad_norm": 1.612523269327134, + "language_loss": 0.8030895, + "learning_rate": 3.7364480914933895e-06, + "loss": 0.82557046, + "num_input_tokens_seen": 33593090, + "step": 1583, + "time_per_iteration": 2.725003480911255 + }, + { + "auxiliary_loss_clip": 0.01173884, + "auxiliary_loss_mlp": 0.00887752, + "balance_loss_clip": 1.05508351, + "balance_loss_mlp": 1.00062466, + "epoch": 0.1904647387723201, + "flos": 26792843425920.0, + "grad_norm": 2.141409223441894, + "language_loss": 0.80744845, + "learning_rate": 3.7360614555971325e-06, + "loss": 0.8280648, + "num_input_tokens_seen": 33612745, + "step": 1584, + "time_per_iteration": 2.903458595275879 + }, + { + "auxiliary_loss_clip": 0.01204807, + "auxiliary_loss_mlp": 0.00887471, + "balance_loss_clip": 1.05797565, + "balance_loss_mlp": 1.00059843, + "epoch": 0.19058498166295917, + "flos": 23987753688960.0, + "grad_norm": 1.9637135090539677, + "language_loss": 0.85007071, + "learning_rate": 3.735674556341733e-06, + "loss": 0.87099349, + "num_input_tokens_seen": 33632360, + "step": 1585, + "time_per_iteration": 2.8158411979675293 + }, + { + "auxiliary_loss_clip": 0.01192068, + "auxiliary_loss_mlp": 0.01039996, + "balance_loss_clip": 1.05810678, + "balance_loss_mlp": 1.02951121, + "epoch": 0.19070522455359826, + "flos": 28293299280000.0, + "grad_norm": 2.168435299762481, + "language_loss": 0.82985485, + "learning_rate": 3.7352873937858835e-06, + "loss": 0.85217547, + "num_input_tokens_seen": 33653895, + "step": 1586, + "time_per_iteration": 2.931562662124634 + }, + { + "auxiliary_loss_clip": 0.01181123, + "auxiliary_loss_mlp": 0.00887613, + "balance_loss_clip": 1.05532622, + "balance_loss_mlp": 1.00050771, + "epoch": 0.19082546744423737, + "flos": 25660220797440.0, + "grad_norm": 1.9216967154237583, + "language_loss": 0.72085309, + "learning_rate": 3.734899967988316e-06, + "loss": 0.74154049, + "num_input_tokens_seen": 33672075, + "step": 1587, + "time_per_iteration": 2.8138914108276367 + }, + { + "auxiliary_loss_clip": 0.01178651, + "auxiliary_loss_mlp": 0.01031646, + "balance_loss_clip": 1.05179048, + "balance_loss_mlp": 1.02085757, + "epoch": 0.19094571033487645, + "flos": 19719483436800.0, + "grad_norm": 1.991539465736078, + "language_loss": 0.83936423, + "learning_rate": 3.7345122790078026e-06, + "loss": 0.86146718, + "num_input_tokens_seen": 33689640, + "step": 1588, + "time_per_iteration": 2.799954414367676 + }, + { + "auxiliary_loss_clip": 0.01204427, + "auxiliary_loss_mlp": 0.01043355, + "balance_loss_clip": 1.05734015, + "balance_loss_mlp": 1.03238738, + "epoch": 0.19106595322551553, + "flos": 21616320850560.0, + "grad_norm": 2.802479387625443, + "language_loss": 0.92736578, + "learning_rate": 3.7341243269031556e-06, + "loss": 0.94984359, + "num_input_tokens_seen": 33708630, + "step": 1589, + "time_per_iteration": 2.8324358463287354 + }, + { + "auxiliary_loss_clip": 0.01187578, + "auxiliary_loss_mlp": 0.0103745, + "balance_loss_clip": 1.05392992, + "balance_loss_mlp": 1.02741838, + "epoch": 0.19118619611615464, + "flos": 29896890059520.0, + "grad_norm": 1.8497129819748122, + "language_loss": 0.7744031, + "learning_rate": 3.7337361117332275e-06, + "loss": 0.79665339, + "num_input_tokens_seen": 33730370, + "step": 1590, + "time_per_iteration": 2.8437390327453613 + }, + { + "auxiliary_loss_clip": 0.01187221, + "auxiliary_loss_mlp": 0.01040474, + "balance_loss_clip": 1.0535295, + "balance_loss_mlp": 1.02983499, + "epoch": 0.19130643900679373, + "flos": 17273428093440.0, + "grad_norm": 2.1180804072102264, + "language_loss": 0.77224779, + "learning_rate": 3.7333476335569087e-06, + "loss": 0.79452479, + "num_input_tokens_seen": 33748370, + "step": 1591, + "time_per_iteration": 2.7574996948242188 + }, + { + "auxiliary_loss_clip": 0.01192711, + "auxiliary_loss_mlp": 0.01038027, + "balance_loss_clip": 1.05399966, + "balance_loss_mlp": 1.02658844, + "epoch": 0.1914266818974328, + "flos": 24826339584000.0, + "grad_norm": 3.113696589308348, + "language_loss": 0.66877568, + "learning_rate": 3.7329588924331325e-06, + "loss": 0.69108307, + "num_input_tokens_seen": 33769575, + "step": 1592, + "time_per_iteration": 2.786750078201294 + }, + { + "auxiliary_loss_clip": 0.0117538, + "auxiliary_loss_mlp": 0.01033649, + "balance_loss_clip": 1.04967165, + "balance_loss_mlp": 1.02261066, + "epoch": 0.1915469247880719, + "flos": 18952467390720.0, + "grad_norm": 2.1240823112255978, + "language_loss": 0.82760608, + "learning_rate": 3.732569888420871e-06, + "loss": 0.84969634, + "num_input_tokens_seen": 33789110, + "step": 1593, + "time_per_iteration": 2.8208401203155518 + }, + { + "auxiliary_loss_clip": 0.01214987, + "auxiliary_loss_mlp": 0.01033109, + "balance_loss_clip": 1.05615747, + "balance_loss_mlp": 1.02176642, + "epoch": 0.191667167678711, + "flos": 21032952065280.0, + "grad_norm": 5.451315041088976, + "language_loss": 0.82331485, + "learning_rate": 3.732180621579134e-06, + "loss": 0.84579575, + "num_input_tokens_seen": 33808325, + "step": 1594, + "time_per_iteration": 2.664910316467285 + }, + { + "auxiliary_loss_clip": 0.0119426, + "auxiliary_loss_mlp": 0.01045103, + "balance_loss_clip": 1.0553174, + "balance_loss_mlp": 1.03375971, + "epoch": 0.1917874105693501, + "flos": 34237663914240.0, + "grad_norm": 2.0120750915405274, + "language_loss": 0.81351191, + "learning_rate": 3.7317910919669745e-06, + "loss": 0.83590543, + "num_input_tokens_seen": 33829520, + "step": 1595, + "time_per_iteration": 2.905365467071533 + }, + { + "auxiliary_loss_clip": 0.0120453, + "auxiliary_loss_mlp": 0.01035136, + "balance_loss_clip": 1.05815315, + "balance_loss_mlp": 1.02367365, + "epoch": 0.19190765345998917, + "flos": 23550613171200.0, + "grad_norm": 2.876678973910871, + "language_loss": 0.76791984, + "learning_rate": 3.7314012996434826e-06, + "loss": 0.79031646, + "num_input_tokens_seen": 33848250, + "step": 1596, + "time_per_iteration": 2.833301067352295 + }, + { + "auxiliary_loss_clip": 0.01191657, + "auxiliary_loss_mlp": 0.01032329, + "balance_loss_clip": 1.0547502, + "balance_loss_mlp": 1.02146912, + "epoch": 0.19202789635062828, + "flos": 19861330245120.0, + "grad_norm": 2.1910338062488677, + "language_loss": 0.80996346, + "learning_rate": 3.7310112446677907e-06, + "loss": 0.83220327, + "num_input_tokens_seen": 33866160, + "step": 1597, + "time_per_iteration": 2.790447235107422 + }, + { + "auxiliary_loss_clip": 0.01218488, + "auxiliary_loss_mlp": 0.01039853, + "balance_loss_clip": 1.06000519, + "balance_loss_mlp": 1.0283556, + "epoch": 0.19214813924126736, + "flos": 20922957642240.0, + "grad_norm": 2.5340917087374257, + "language_loss": 0.69015968, + "learning_rate": 3.7306209270990695e-06, + "loss": 0.71274316, + "num_input_tokens_seen": 33884165, + "step": 1598, + "time_per_iteration": 2.667245864868164 + }, + { + "auxiliary_loss_clip": 0.01195387, + "auxiliary_loss_mlp": 0.01036211, + "balance_loss_clip": 1.05505109, + "balance_loss_mlp": 1.02541649, + "epoch": 0.19226838213190645, + "flos": 26359725231360.0, + "grad_norm": 1.9377565428246772, + "language_loss": 0.86580324, + "learning_rate": 3.7302303469965292e-06, + "loss": 0.88811922, + "num_input_tokens_seen": 33903705, + "step": 1599, + "time_per_iteration": 2.879495859146118 + }, + { + "auxiliary_loss_clip": 0.01206031, + "auxiliary_loss_mlp": 0.01035588, + "balance_loss_clip": 1.05884099, + "balance_loss_mlp": 1.02465105, + "epoch": 0.19238862502254553, + "flos": 20850525866880.0, + "grad_norm": 2.7817562545323082, + "language_loss": 0.70821738, + "learning_rate": 3.7298395044194206e-06, + "loss": 0.73063362, + "num_input_tokens_seen": 33922515, + "step": 1600, + "time_per_iteration": 2.7339956760406494 + }, + { + "auxiliary_loss_clip": 0.01218746, + "auxiliary_loss_mlp": 0.01035328, + "balance_loss_clip": 1.06018686, + "balance_loss_mlp": 1.02427149, + "epoch": 0.19250886791318464, + "flos": 21726063878400.0, + "grad_norm": 2.0215668631044346, + "language_loss": 0.94248301, + "learning_rate": 3.7294483994270356e-06, + "loss": 0.96502376, + "num_input_tokens_seen": 33940840, + "step": 1601, + "time_per_iteration": 2.6200954914093018 + }, + { + "auxiliary_loss_clip": 0.01164723, + "auxiliary_loss_mlp": 0.01033807, + "balance_loss_clip": 1.04937792, + "balance_loss_mlp": 1.02279782, + "epoch": 0.19262911080382372, + "flos": 23367827836800.0, + "grad_norm": 2.0154463943478316, + "language_loss": 0.78154325, + "learning_rate": 3.7290570320787033e-06, + "loss": 0.80352855, + "num_input_tokens_seen": 33960420, + "step": 1602, + "time_per_iteration": 2.909834861755371 + }, + { + "auxiliary_loss_clip": 0.01203216, + "auxiliary_loss_mlp": 0.01036485, + "balance_loss_clip": 1.05801582, + "balance_loss_mlp": 1.02479017, + "epoch": 0.1927493536944628, + "flos": 21943502858880.0, + "grad_norm": 2.0675899558679496, + "language_loss": 0.71359456, + "learning_rate": 3.728665402433793e-06, + "loss": 0.73599154, + "num_input_tokens_seen": 33978990, + "step": 1603, + "time_per_iteration": 2.721071720123291 + }, + { + "auxiliary_loss_clip": 0.01194573, + "auxiliary_loss_mlp": 0.01033844, + "balance_loss_clip": 1.05760145, + "balance_loss_mlp": 1.02248979, + "epoch": 0.19286959658510192, + "flos": 16545590807040.0, + "grad_norm": 2.3608062837838473, + "language_loss": 0.85928559, + "learning_rate": 3.7282735105517164e-06, + "loss": 0.88156974, + "num_input_tokens_seen": 33997115, + "step": 1604, + "time_per_iteration": 2.785475254058838 + }, + { + "auxiliary_loss_clip": 0.01181495, + "auxiliary_loss_mlp": 0.01037214, + "balance_loss_clip": 1.05098653, + "balance_loss_mlp": 1.02615762, + "epoch": 0.192989839475741, + "flos": 21616967295360.0, + "grad_norm": 2.6242957229506563, + "language_loss": 0.67027849, + "learning_rate": 3.727881356491922e-06, + "loss": 0.6924656, + "num_input_tokens_seen": 34015525, + "step": 1605, + "time_per_iteration": 4.831786870956421 + }, + { + "auxiliary_loss_clip": 0.01213766, + "auxiliary_loss_mlp": 0.01031964, + "balance_loss_clip": 1.05899107, + "balance_loss_mlp": 1.02147377, + "epoch": 0.19311008236638008, + "flos": 19281516906240.0, + "grad_norm": 1.8628383511653073, + "language_loss": 0.75932986, + "learning_rate": 3.7274889403139002e-06, + "loss": 0.78178716, + "num_input_tokens_seen": 34033150, + "step": 1606, + "time_per_iteration": 2.693182945251465 + }, + { + "auxiliary_loss_clip": 0.01170109, + "auxiliary_loss_mlp": 0.01036154, + "balance_loss_clip": 1.05136156, + "balance_loss_mlp": 1.02528191, + "epoch": 0.1932303252570192, + "flos": 28652369587200.0, + "grad_norm": 2.8618455179164397, + "language_loss": 0.78910863, + "learning_rate": 3.727096262077179e-06, + "loss": 0.81117123, + "num_input_tokens_seen": 34052145, + "step": 1607, + "time_per_iteration": 3.035170078277588 + }, + { + "auxiliary_loss_clip": 0.01206372, + "auxiliary_loss_mlp": 0.01036731, + "balance_loss_clip": 1.05851078, + "balance_loss_mlp": 1.0259012, + "epoch": 0.19335056814765827, + "flos": 18368990864640.0, + "grad_norm": 1.9360193717988563, + "language_loss": 0.85101247, + "learning_rate": 3.7267033218413285e-06, + "loss": 0.87344348, + "num_input_tokens_seen": 34069940, + "step": 1608, + "time_per_iteration": 2.6394999027252197 + }, + { + "auxiliary_loss_clip": 0.01161494, + "auxiliary_loss_mlp": 0.01038679, + "balance_loss_clip": 1.0456624, + "balance_loss_mlp": 1.02771759, + "epoch": 0.19347081103829736, + "flos": 13260877741440.0, + "grad_norm": 2.091487317289526, + "language_loss": 0.81360382, + "learning_rate": 3.726310119665957e-06, + "loss": 0.8356055, + "num_input_tokens_seen": 34086275, + "step": 1609, + "time_per_iteration": 2.8457343578338623 + }, + { + "auxiliary_loss_clip": 0.01206151, + "auxiliary_loss_mlp": 0.01032575, + "balance_loss_clip": 1.05704236, + "balance_loss_mlp": 1.02175069, + "epoch": 0.19359105392893644, + "flos": 20300122788480.0, + "grad_norm": 1.9423621715652872, + "language_loss": 0.85189342, + "learning_rate": 3.725916655610713e-06, + "loss": 0.87428069, + "num_input_tokens_seen": 34105605, + "step": 1610, + "time_per_iteration": 2.707721710205078 + }, + { + "auxiliary_loss_clip": 0.01187475, + "auxiliary_loss_mlp": 0.01037557, + "balance_loss_clip": 1.05351126, + "balance_loss_mlp": 1.02579653, + "epoch": 0.19371129681957555, + "flos": 20484596062080.0, + "grad_norm": 4.859936791470899, + "language_loss": 0.75630212, + "learning_rate": 3.725522929735284e-06, + "loss": 0.77855241, + "num_input_tokens_seen": 34122540, + "step": 1611, + "time_per_iteration": 2.792729139328003 + }, + { + "auxiliary_loss_clip": 0.01201758, + "auxiliary_loss_mlp": 0.01033622, + "balance_loss_clip": 1.05423176, + "balance_loss_mlp": 1.02299416, + "epoch": 0.19383153971021463, + "flos": 30445497457920.0, + "grad_norm": 2.499899666808962, + "language_loss": 0.74460077, + "learning_rate": 3.725128942099399e-06, + "loss": 0.7669546, + "num_input_tokens_seen": 34142940, + "step": 1612, + "time_per_iteration": 2.853876829147339 + }, + { + "auxiliary_loss_clip": 0.0118661, + "auxiliary_loss_mlp": 0.010324, + "balance_loss_clip": 1.05215025, + "balance_loss_mlp": 1.02112877, + "epoch": 0.19395178260085372, + "flos": 24569937325440.0, + "grad_norm": 1.6603315299931782, + "language_loss": 0.80209374, + "learning_rate": 3.7247346927628245e-06, + "loss": 0.82428384, + "num_input_tokens_seen": 34162875, + "step": 1613, + "time_per_iteration": 2.7402985095977783 + }, + { + "auxiliary_loss_clip": 0.01196419, + "auxiliary_loss_mlp": 0.00888707, + "balance_loss_clip": 1.05791473, + "balance_loss_mlp": 1.00041461, + "epoch": 0.19407202549149283, + "flos": 28950608211840.0, + "grad_norm": 1.7777493119978685, + "language_loss": 0.7951861, + "learning_rate": 3.7243401817853694e-06, + "loss": 0.8160373, + "num_input_tokens_seen": 34183565, + "step": 1614, + "time_per_iteration": 2.8522908687591553 + }, + { + "auxiliary_loss_clip": 0.01198052, + "auxiliary_loss_mlp": 0.01032533, + "balance_loss_clip": 1.05465841, + "balance_loss_mlp": 1.02133369, + "epoch": 0.1941922683821319, + "flos": 18004497603840.0, + "grad_norm": 1.8857473236432023, + "language_loss": 0.71582639, + "learning_rate": 3.723945409226879e-06, + "loss": 0.73813224, + "num_input_tokens_seen": 34202055, + "step": 1615, + "time_per_iteration": 2.780548095703125 + }, + { + "auxiliary_loss_clip": 0.01205854, + "auxiliary_loss_mlp": 0.01028141, + "balance_loss_clip": 1.05695486, + "balance_loss_mlp": 1.0173583, + "epoch": 0.194312511272771, + "flos": 9720337034880.0, + "grad_norm": 3.500823546662552, + "language_loss": 0.80260038, + "learning_rate": 3.723550375147241e-06, + "loss": 0.82494038, + "num_input_tokens_seen": 34216830, + "step": 1616, + "time_per_iteration": 2.719881296157837 + }, + { + "auxiliary_loss_clip": 0.01175316, + "auxiliary_loss_mlp": 0.01032987, + "balance_loss_clip": 1.05111361, + "balance_loss_mlp": 1.02153111, + "epoch": 0.19443275416341008, + "flos": 27016208150400.0, + "grad_norm": 1.7610609663303085, + "language_loss": 0.80044639, + "learning_rate": 3.7231550796063816e-06, + "loss": 0.82252938, + "num_input_tokens_seen": 34236840, + "step": 1617, + "time_per_iteration": 2.788393497467041 + }, + { + "auxiliary_loss_clip": 0.01202876, + "auxiliary_loss_mlp": 0.01029007, + "balance_loss_clip": 1.05932915, + "balance_loss_mlp": 1.01800966, + "epoch": 0.1945529970540492, + "flos": 15846625077120.0, + "grad_norm": 2.5737558621019065, + "language_loss": 0.65274894, + "learning_rate": 3.722759522664266e-06, + "loss": 0.67506778, + "num_input_tokens_seen": 34254140, + "step": 1618, + "time_per_iteration": 2.7672173976898193 + }, + { + "auxiliary_loss_clip": 0.01177048, + "auxiliary_loss_mlp": 0.01040825, + "balance_loss_clip": 1.05261374, + "balance_loss_mlp": 1.03016806, + "epoch": 0.19467323994468827, + "flos": 19314985403520.0, + "grad_norm": 3.3348858210241303, + "language_loss": 0.81998122, + "learning_rate": 3.7223637043809016e-06, + "loss": 0.84215999, + "num_input_tokens_seen": 34273120, + "step": 1619, + "time_per_iteration": 2.908998966217041 + }, + { + "auxiliary_loss_clip": 0.01192179, + "auxiliary_loss_mlp": 0.01039152, + "balance_loss_clip": 1.05955052, + "balance_loss_mlp": 1.02780914, + "epoch": 0.19479348283532735, + "flos": 24133227770880.0, + "grad_norm": 2.177016649587707, + "language_loss": 0.86403662, + "learning_rate": 3.7219676248163322e-06, + "loss": 0.88634992, + "num_input_tokens_seen": 34290285, + "step": 1620, + "time_per_iteration": 2.792642593383789 + }, + { + "auxiliary_loss_clip": 0.0120905, + "auxiliary_loss_mlp": 0.01035665, + "balance_loss_clip": 1.05905449, + "balance_loss_mlp": 1.02451301, + "epoch": 0.19491372572596646, + "flos": 25775638174080.0, + "grad_norm": 2.155900951252186, + "language_loss": 0.93463039, + "learning_rate": 3.721571284030643e-06, + "loss": 0.9570775, + "num_input_tokens_seen": 34310095, + "step": 1621, + "time_per_iteration": 2.7508621215820312 + }, + { + "auxiliary_loss_clip": 0.01209692, + "auxiliary_loss_mlp": 0.01030892, + "balance_loss_clip": 1.05850506, + "balance_loss_mlp": 1.01968026, + "epoch": 0.19503396861660555, + "flos": 19645220067840.0, + "grad_norm": 2.3806691320378106, + "language_loss": 0.79629993, + "learning_rate": 3.7211746820839587e-06, + "loss": 0.8187058, + "num_input_tokens_seen": 34327190, + "step": 1622, + "time_per_iteration": 2.6938157081604004 + }, + { + "auxiliary_loss_clip": 0.01140732, + "auxiliary_loss_mlp": 0.01036517, + "balance_loss_clip": 1.04576468, + "balance_loss_mlp": 1.02630043, + "epoch": 0.19515421150724463, + "flos": 21033023892480.0, + "grad_norm": 1.7454476793219729, + "language_loss": 0.81178021, + "learning_rate": 3.7207778190364437e-06, + "loss": 0.83355272, + "num_input_tokens_seen": 34345615, + "step": 1623, + "time_per_iteration": 2.932553768157959 + }, + { + "auxiliary_loss_clip": 0.0114838, + "auxiliary_loss_mlp": 0.01033157, + "balance_loss_clip": 1.04713166, + "balance_loss_mlp": 1.02241063, + "epoch": 0.1952744543978837, + "flos": 32961255143040.0, + "grad_norm": 1.6387459781176452, + "language_loss": 0.73901224, + "learning_rate": 3.720380694948302e-06, + "loss": 0.7608276, + "num_input_tokens_seen": 34368500, + "step": 1624, + "time_per_iteration": 2.9295616149902344 + }, + { + "auxiliary_loss_clip": 0.01096263, + "auxiliary_loss_mlp": 0.01010312, + "balance_loss_clip": 1.03137279, + "balance_loss_mlp": 1.00640225, + "epoch": 0.19539469728852282, + "flos": 64044312030720.0, + "grad_norm": 1.0390186187883457, + "language_loss": 0.71236169, + "learning_rate": 3.719983309879777e-06, + "loss": 0.73342741, + "num_input_tokens_seen": 34428280, + "step": 1625, + "time_per_iteration": 3.3820741176605225 + }, + { + "auxiliary_loss_clip": 0.01177842, + "auxiliary_loss_mlp": 0.01031827, + "balance_loss_clip": 1.05380082, + "balance_loss_mlp": 1.02047276, + "epoch": 0.1955149401791619, + "flos": 13370908078080.0, + "grad_norm": 1.664806230025361, + "language_loss": 0.775913, + "learning_rate": 3.719585663891151e-06, + "loss": 0.79800969, + "num_input_tokens_seen": 34445815, + "step": 1626, + "time_per_iteration": 2.760028600692749 + }, + { + "auxiliary_loss_clip": 0.01168049, + "auxiliary_loss_mlp": 0.01040055, + "balance_loss_clip": 1.05317283, + "balance_loss_mlp": 1.02834284, + "epoch": 0.195635183069801, + "flos": 18728887184640.0, + "grad_norm": 1.9578764536407407, + "language_loss": 0.79230332, + "learning_rate": 3.719187757042747e-06, + "loss": 0.81438434, + "num_input_tokens_seen": 34463635, + "step": 1627, + "time_per_iteration": 2.7762022018432617 + }, + { + "auxiliary_loss_clip": 0.01102857, + "auxiliary_loss_mlp": 0.0100495, + "balance_loss_clip": 1.02927566, + "balance_loss_mlp": 1.00113571, + "epoch": 0.1957554259604401, + "flos": 69313952615040.0, + "grad_norm": 0.7257733372771753, + "language_loss": 0.54937172, + "learning_rate": 3.7187895893949275e-06, + "loss": 0.57044983, + "num_input_tokens_seen": 34530105, + "step": 1628, + "time_per_iteration": 3.431568145751953 + }, + { + "auxiliary_loss_clip": 0.01160655, + "auxiliary_loss_mlp": 0.01035527, + "balance_loss_clip": 1.04708946, + "balance_loss_mlp": 1.02349889, + "epoch": 0.19587566885107918, + "flos": 21069257736960.0, + "grad_norm": 2.468354601181935, + "language_loss": 0.76168275, + "learning_rate": 3.7183911610080937e-06, + "loss": 0.78364462, + "num_input_tokens_seen": 34546970, + "step": 1629, + "time_per_iteration": 2.8043015003204346 + }, + { + "auxiliary_loss_clip": 0.01184836, + "auxiliary_loss_mlp": 0.01035037, + "balance_loss_clip": 1.05196023, + "balance_loss_mlp": 1.02344418, + "epoch": 0.19599591174171827, + "flos": 22194661731840.0, + "grad_norm": 2.9029220857517024, + "language_loss": 0.74641526, + "learning_rate": 3.7179924719426872e-06, + "loss": 0.76861399, + "num_input_tokens_seen": 34564865, + "step": 1630, + "time_per_iteration": 3.734065294265747 + }, + { + "auxiliary_loss_clip": 0.01209465, + "auxiliary_loss_mlp": 0.01036987, + "balance_loss_clip": 1.05935669, + "balance_loss_mlp": 1.02496457, + "epoch": 0.19611615463235738, + "flos": 23768375374080.0, + "grad_norm": 5.150480913143544, + "language_loss": 0.75938946, + "learning_rate": 3.7175935222591885e-06, + "loss": 0.78185397, + "num_input_tokens_seen": 34584165, + "step": 1631, + "time_per_iteration": 5.473925352096558 + }, + { + "auxiliary_loss_clip": 0.01202614, + "auxiliary_loss_mlp": 0.01038303, + "balance_loss_clip": 1.06355095, + "balance_loss_mlp": 1.02721035, + "epoch": 0.19623639752299646, + "flos": 28618218731520.0, + "grad_norm": 1.9043043379363682, + "language_loss": 0.74048489, + "learning_rate": 3.717194312018118e-06, + "loss": 0.76289403, + "num_input_tokens_seen": 34603150, + "step": 1632, + "time_per_iteration": 2.827998161315918 + }, + { + "auxiliary_loss_clip": 0.01206077, + "auxiliary_loss_mlp": 0.01034508, + "balance_loss_clip": 1.0563122, + "balance_loss_mlp": 1.02250957, + "epoch": 0.19635664041363554, + "flos": 21032700670080.0, + "grad_norm": 1.9121719178929766, + "language_loss": 0.76172638, + "learning_rate": 3.716794841280036e-06, + "loss": 0.78413224, + "num_input_tokens_seen": 34621855, + "step": 1633, + "time_per_iteration": 2.773416519165039 + }, + { + "auxiliary_loss_clip": 0.01210263, + "auxiliary_loss_mlp": 0.01035589, + "balance_loss_clip": 1.05657864, + "balance_loss_mlp": 1.02439523, + "epoch": 0.19647688330427462, + "flos": 18879748306560.0, + "grad_norm": 2.0058493652014815, + "language_loss": 0.77354491, + "learning_rate": 3.7163951101055407e-06, + "loss": 0.79600346, + "num_input_tokens_seen": 34639915, + "step": 1634, + "time_per_iteration": 2.8631832599639893 + }, + { + "auxiliary_loss_clip": 0.01190269, + "auxiliary_loss_mlp": 0.01029189, + "balance_loss_clip": 1.05660021, + "balance_loss_mlp": 1.0177567, + "epoch": 0.19659712619491373, + "flos": 24242503921920.0, + "grad_norm": 1.9014926655244335, + "language_loss": 0.7877959, + "learning_rate": 3.715995118555273e-06, + "loss": 0.80999047, + "num_input_tokens_seen": 34659890, + "step": 1635, + "time_per_iteration": 2.789545774459839 + }, + { + "auxiliary_loss_clip": 0.01174991, + "auxiliary_loss_mlp": 0.01035903, + "balance_loss_clip": 1.05184293, + "balance_loss_mlp": 1.02361846, + "epoch": 0.19671736908555282, + "flos": 24717422568960.0, + "grad_norm": 2.1709871960420384, + "language_loss": 0.85910612, + "learning_rate": 3.71559486668991e-06, + "loss": 0.8812151, + "num_input_tokens_seen": 34678750, + "step": 1636, + "time_per_iteration": 3.0393099784851074 + }, + { + "auxiliary_loss_clip": 0.01213752, + "auxiliary_loss_mlp": 0.0088778, + "balance_loss_clip": 1.06098986, + "balance_loss_mlp": 1.00047743, + "epoch": 0.1968376119761919, + "flos": 23842279607040.0, + "grad_norm": 1.655937605622227, + "language_loss": 0.77365863, + "learning_rate": 3.715194354570169e-06, + "loss": 0.79467398, + "num_input_tokens_seen": 34698755, + "step": 1637, + "time_per_iteration": 2.781557559967041 + }, + { + "auxiliary_loss_clip": 0.01209484, + "auxiliary_loss_mlp": 0.01035997, + "balance_loss_clip": 1.0622493, + "balance_loss_mlp": 1.02543461, + "epoch": 0.196957854866831, + "flos": 18113917409280.0, + "grad_norm": 2.988885528667205, + "language_loss": 0.83411682, + "learning_rate": 3.714793582256809e-06, + "loss": 0.85657161, + "num_input_tokens_seen": 34715820, + "step": 1638, + "time_per_iteration": 2.7220029830932617 + }, + { + "auxiliary_loss_clip": 0.01215285, + "auxiliary_loss_mlp": 0.01036495, + "balance_loss_clip": 1.05887783, + "balance_loss_mlp": 1.02560532, + "epoch": 0.1970780977574701, + "flos": 21653129312640.0, + "grad_norm": 2.808985554472987, + "language_loss": 0.85114074, + "learning_rate": 3.7143925498106253e-06, + "loss": 0.87365854, + "num_input_tokens_seen": 34734360, + "step": 1639, + "time_per_iteration": 2.744521379470825 + }, + { + "auxiliary_loss_clip": 0.0119303, + "auxiliary_loss_mlp": 0.01051817, + "balance_loss_clip": 1.05030596, + "balance_loss_mlp": 1.04051602, + "epoch": 0.19719834064810918, + "flos": 20811813984000.0, + "grad_norm": 1.7472801502820476, + "language_loss": 0.79012001, + "learning_rate": 3.7139912572924558e-06, + "loss": 0.81256855, + "num_input_tokens_seen": 34753390, + "step": 1640, + "time_per_iteration": 2.75117826461792 + }, + { + "auxiliary_loss_clip": 0.01202726, + "auxiliary_loss_mlp": 0.01034473, + "balance_loss_clip": 1.05338764, + "balance_loss_mlp": 1.02258217, + "epoch": 0.19731858353874826, + "flos": 23434800744960.0, + "grad_norm": 2.9530132595263914, + "language_loss": 0.808577, + "learning_rate": 3.7135897047631744e-06, + "loss": 0.83094901, + "num_input_tokens_seen": 34771275, + "step": 1641, + "time_per_iteration": 2.875089645385742 + }, + { + "auxiliary_loss_clip": 0.0119892, + "auxiliary_loss_mlp": 0.01031801, + "balance_loss_clip": 1.05592537, + "balance_loss_mlp": 1.02023816, + "epoch": 0.19743882642938737, + "flos": 23988184652160.0, + "grad_norm": 2.0228303930318945, + "language_loss": 0.75777715, + "learning_rate": 3.713187892283698e-06, + "loss": 0.78008437, + "num_input_tokens_seen": 34790885, + "step": 1642, + "time_per_iteration": 2.753239154815674 + }, + { + "auxiliary_loss_clip": 0.01173886, + "auxiliary_loss_mlp": 0.01039444, + "balance_loss_clip": 1.04943442, + "balance_loss_mlp": 1.02753484, + "epoch": 0.19755906932002645, + "flos": 15004340081280.0, + "grad_norm": 2.4741281845798984, + "language_loss": 0.87172973, + "learning_rate": 3.71278581991498e-06, + "loss": 0.89386308, + "num_input_tokens_seen": 34806745, + "step": 1643, + "time_per_iteration": 2.915102005004883 + }, + { + "auxiliary_loss_clip": 0.01191355, + "auxiliary_loss_mlp": 0.00888603, + "balance_loss_clip": 1.06123543, + "balance_loss_mlp": 1.00049233, + "epoch": 0.19767931221066554, + "flos": 19494466686720.0, + "grad_norm": 1.846126891318427, + "language_loss": 0.79251528, + "learning_rate": 3.712383487718015e-06, + "loss": 0.8133148, + "num_input_tokens_seen": 34824985, + "step": 1644, + "time_per_iteration": 2.806057929992676 + }, + { + "auxiliary_loss_clip": 0.0115972, + "auxiliary_loss_mlp": 0.01042052, + "balance_loss_clip": 1.05178034, + "balance_loss_mlp": 1.03069162, + "epoch": 0.19779955510130465, + "flos": 25737895958400.0, + "grad_norm": 1.912962437860544, + "language_loss": 0.86708713, + "learning_rate": 3.7119808957538365e-06, + "loss": 0.88910484, + "num_input_tokens_seen": 34843980, + "step": 1645, + "time_per_iteration": 2.9059815406799316 + }, + { + "auxiliary_loss_clip": 0.0119006, + "auxiliary_loss_mlp": 0.01043629, + "balance_loss_clip": 1.05266523, + "balance_loss_mlp": 1.03141594, + "epoch": 0.19791979799194373, + "flos": 20777699041920.0, + "grad_norm": 2.2400128558267927, + "language_loss": 0.79912484, + "learning_rate": 3.711578044083517e-06, + "loss": 0.82146168, + "num_input_tokens_seen": 34860780, + "step": 1646, + "time_per_iteration": 2.749063491821289 + }, + { + "auxiliary_loss_clip": 0.01197403, + "auxiliary_loss_mlp": 0.01045777, + "balance_loss_clip": 1.05511558, + "balance_loss_mlp": 1.03485775, + "epoch": 0.1980400408825828, + "flos": 25589010084480.0, + "grad_norm": 1.8823185308236678, + "language_loss": 0.74772429, + "learning_rate": 3.7111749327681698e-06, + "loss": 0.77015603, + "num_input_tokens_seen": 34880815, + "step": 1647, + "time_per_iteration": 2.784701108932495 + }, + { + "auxiliary_loss_clip": 0.01213443, + "auxiliary_loss_mlp": 0.01036755, + "balance_loss_clip": 1.06063056, + "balance_loss_mlp": 1.02590108, + "epoch": 0.1981602837732219, + "flos": 23513840622720.0, + "grad_norm": 1.9599672258339444, + "language_loss": 0.86081862, + "learning_rate": 3.7107715618689455e-06, + "loss": 0.88332057, + "num_input_tokens_seen": 34899790, + "step": 1648, + "time_per_iteration": 2.805577039718628 + }, + { + "auxiliary_loss_clip": 0.01206924, + "auxiliary_loss_mlp": 0.01037714, + "balance_loss_clip": 1.05983818, + "balance_loss_mlp": 1.02613914, + "epoch": 0.198280526663861, + "flos": 23185365724800.0, + "grad_norm": 1.438070819082153, + "language_loss": 0.83299726, + "learning_rate": 3.710367931447035e-06, + "loss": 0.8554436, + "num_input_tokens_seen": 34921570, + "step": 1649, + "time_per_iteration": 2.713690996170044 + }, + { + "auxiliary_loss_clip": 0.01216141, + "auxiliary_loss_mlp": 0.01047954, + "balance_loss_clip": 1.0598942, + "balance_loss_mlp": 1.03617013, + "epoch": 0.1984007695545001, + "flos": 21689470897920.0, + "grad_norm": 4.31283172481, + "language_loss": 0.86334896, + "learning_rate": 3.70996404156367e-06, + "loss": 0.8859899, + "num_input_tokens_seen": 34941205, + "step": 1650, + "time_per_iteration": 2.740767002105713 + }, + { + "auxiliary_loss_clip": 0.01165333, + "auxiliary_loss_mlp": 0.01042255, + "balance_loss_clip": 1.04945135, + "balance_loss_mlp": 1.03044105, + "epoch": 0.19852101244513917, + "flos": 36064008887040.0, + "grad_norm": 1.6601427922504153, + "language_loss": 0.72871196, + "learning_rate": 3.7095598922801187e-06, + "loss": 0.75078773, + "num_input_tokens_seen": 34963280, + "step": 1651, + "time_per_iteration": 2.910996913909912 + }, + { + "auxiliary_loss_clip": 0.01218222, + "auxiliary_loss_mlp": 0.01038351, + "balance_loss_clip": 1.06150925, + "balance_loss_mlp": 1.02718759, + "epoch": 0.19864125533577828, + "flos": 23105894883840.0, + "grad_norm": 2.752369261259059, + "language_loss": 0.76344824, + "learning_rate": 3.7091554836576914e-06, + "loss": 0.78601396, + "num_input_tokens_seen": 34979955, + "step": 1652, + "time_per_iteration": 2.7895450592041016 + }, + { + "auxiliary_loss_clip": 0.01207029, + "auxiliary_loss_mlp": 0.00888181, + "balance_loss_clip": 1.06148756, + "balance_loss_mlp": 1.00047827, + "epoch": 0.19876149822641737, + "flos": 24608505553920.0, + "grad_norm": 4.3819923118599124, + "language_loss": 0.82701302, + "learning_rate": 3.708750815757736e-06, + "loss": 0.84796512, + "num_input_tokens_seen": 35000725, + "step": 1653, + "time_per_iteration": 2.8136539459228516 + }, + { + "auxiliary_loss_clip": 0.01210065, + "auxiliary_loss_mlp": 0.01039521, + "balance_loss_clip": 1.05784285, + "balance_loss_mlp": 1.02750492, + "epoch": 0.19888174111705645, + "flos": 32196645308160.0, + "grad_norm": 2.3009565115528567, + "language_loss": 0.73244494, + "learning_rate": 3.7083458886416407e-06, + "loss": 0.75494081, + "num_input_tokens_seen": 35019920, + "step": 1654, + "time_per_iteration": 2.8196964263916016 + }, + { + "auxiliary_loss_clip": 0.01167846, + "auxiliary_loss_mlp": 0.01041068, + "balance_loss_clip": 1.05359101, + "balance_loss_mlp": 1.02906919, + "epoch": 0.19900198400769553, + "flos": 24608469640320.0, + "grad_norm": 2.185074021429238, + "language_loss": 0.8807466, + "learning_rate": 3.707940702370832e-06, + "loss": 0.90283573, + "num_input_tokens_seen": 35040765, + "step": 1655, + "time_per_iteration": 2.9664812088012695 + }, + { + "auxiliary_loss_clip": 0.01103424, + "auxiliary_loss_mlp": 0.0101553, + "balance_loss_clip": 1.02660036, + "balance_loss_mlp": 1.01126218, + "epoch": 0.19912222689833464, + "flos": 67915805673600.0, + "grad_norm": 0.7659876520369404, + "language_loss": 0.58284706, + "learning_rate": 3.707535257006777e-06, + "loss": 0.60403663, + "num_input_tokens_seen": 35106390, + "step": 1656, + "time_per_iteration": 4.224287271499634 + }, + { + "auxiliary_loss_clip": 0.01200494, + "auxiliary_loss_mlp": 0.01039871, + "balance_loss_clip": 1.05839324, + "balance_loss_mlp": 1.02835512, + "epoch": 0.19924246978897373, + "flos": 15742340916480.0, + "grad_norm": 3.2159096933484523, + "language_loss": 0.88571858, + "learning_rate": 3.707129552610981e-06, + "loss": 0.90812224, + "num_input_tokens_seen": 35125040, + "step": 1657, + "time_per_iteration": 4.684798002243042 + }, + { + "auxiliary_loss_clip": 0.01192226, + "auxiliary_loss_mlp": 0.01039991, + "balance_loss_clip": 1.05896306, + "balance_loss_mlp": 1.02846324, + "epoch": 0.1993627126796128, + "flos": 17566566986880.0, + "grad_norm": 2.105204750607636, + "language_loss": 0.73700261, + "learning_rate": 3.70672358924499e-06, + "loss": 0.75932479, + "num_input_tokens_seen": 35144280, + "step": 1658, + "time_per_iteration": 2.7391412258148193 + }, + { + "auxiliary_loss_clip": 0.01187784, + "auxiliary_loss_mlp": 0.01034415, + "balance_loss_clip": 1.05790687, + "balance_loss_mlp": 1.02347171, + "epoch": 0.19948295557025192, + "flos": 40843826680320.0, + "grad_norm": 2.1209373675138266, + "language_loss": 0.78897655, + "learning_rate": 3.706317366970386e-06, + "loss": 0.81119847, + "num_input_tokens_seen": 35165280, + "step": 1659, + "time_per_iteration": 3.023848056793213 + }, + { + "auxiliary_loss_clip": 0.01220276, + "auxiliary_loss_mlp": 0.00888434, + "balance_loss_clip": 1.0591104, + "balance_loss_mlp": 1.00041962, + "epoch": 0.199603198460891, + "flos": 25082418620160.0, + "grad_norm": 1.8670170330632223, + "language_loss": 0.83387816, + "learning_rate": 3.705910885848795e-06, + "loss": 0.85496533, + "num_input_tokens_seen": 35183655, + "step": 1660, + "time_per_iteration": 2.689789056777954 + }, + { + "auxiliary_loss_clip": 0.01205465, + "auxiliary_loss_mlp": 0.01036461, + "balance_loss_clip": 1.05809391, + "balance_loss_mlp": 1.02538657, + "epoch": 0.19972344135153008, + "flos": 20084120352000.0, + "grad_norm": 2.1944693016628967, + "language_loss": 0.84759295, + "learning_rate": 3.705504145941879e-06, + "loss": 0.87001222, + "num_input_tokens_seen": 35201825, + "step": 1661, + "time_per_iteration": 2.7201805114746094 + }, + { + "auxiliary_loss_clip": 0.01215918, + "auxiliary_loss_mlp": 0.01035048, + "balance_loss_clip": 1.05802345, + "balance_loss_mlp": 1.02395546, + "epoch": 0.1998436842421692, + "flos": 23727472761600.0, + "grad_norm": 3.5954226579391455, + "language_loss": 0.78524303, + "learning_rate": 3.7050971473113403e-06, + "loss": 0.80775273, + "num_input_tokens_seen": 35221600, + "step": 1662, + "time_per_iteration": 2.6657330989837646 + }, + { + "auxiliary_loss_clip": 0.01200214, + "auxiliary_loss_mlp": 0.00887369, + "balance_loss_clip": 1.0532186, + "balance_loss_mlp": 1.00042772, + "epoch": 0.19996392713280828, + "flos": 36102361633920.0, + "grad_norm": 1.6931760709134578, + "language_loss": 0.80077785, + "learning_rate": 3.7046898900189196e-06, + "loss": 0.82165366, + "num_input_tokens_seen": 35245935, + "step": 1663, + "time_per_iteration": 2.873002290725708 + }, + { + "auxiliary_loss_clip": 0.01194824, + "auxiliary_loss_mlp": 0.01040879, + "balance_loss_clip": 1.05931854, + "balance_loss_mlp": 1.02879095, + "epoch": 0.20008417002344736, + "flos": 23657662679040.0, + "grad_norm": 1.8637041988616492, + "language_loss": 0.82872427, + "learning_rate": 3.704282374126398e-06, + "loss": 0.85108137, + "num_input_tokens_seen": 35265615, + "step": 1664, + "time_per_iteration": 2.8430304527282715 + }, + { + "auxiliary_loss_clip": 0.01187551, + "auxiliary_loss_mlp": 0.01039314, + "balance_loss_clip": 1.05437732, + "balance_loss_mlp": 1.02705312, + "epoch": 0.20020441291408644, + "flos": 21872076664320.0, + "grad_norm": 1.939042243096584, + "language_loss": 0.87262422, + "learning_rate": 3.7038745996955954e-06, + "loss": 0.89489287, + "num_input_tokens_seen": 35284960, + "step": 1665, + "time_per_iteration": 323.99078392982483 + }, + { + "auxiliary_loss_clip": 0.01189421, + "auxiliary_loss_mlp": 0.01035284, + "balance_loss_clip": 1.05190873, + "balance_loss_mlp": 1.02361894, + "epoch": 0.20032465580472555, + "flos": 23179691376000.0, + "grad_norm": 2.530574435371799, + "language_loss": 0.71813059, + "learning_rate": 3.703466566788371e-06, + "loss": 0.74037766, + "num_input_tokens_seen": 35304090, + "step": 1666, + "time_per_iteration": 2.992252826690674 + }, + { + "auxiliary_loss_clip": 0.01188576, + "auxiliary_loss_mlp": 0.01033016, + "balance_loss_clip": 1.05676723, + "balance_loss_mlp": 1.02108359, + "epoch": 0.20044489869536464, + "flos": 23873521461120.0, + "grad_norm": 1.977896365548778, + "language_loss": 0.74485749, + "learning_rate": 3.703058275466622e-06, + "loss": 0.76707339, + "num_input_tokens_seen": 35323325, + "step": 1667, + "time_per_iteration": 2.971245288848877 + }, + { + "auxiliary_loss_clip": 0.01195143, + "auxiliary_loss_mlp": 0.01039494, + "balance_loss_clip": 1.05596256, + "balance_loss_mlp": 1.02791286, + "epoch": 0.20056514158600372, + "flos": 21945226711680.0, + "grad_norm": 1.8558494463540451, + "language_loss": 0.77683878, + "learning_rate": 3.7026497257922877e-06, + "loss": 0.79918516, + "num_input_tokens_seen": 35343635, + "step": 1668, + "time_per_iteration": 2.875978708267212 + }, + { + "auxiliary_loss_clip": 0.01170836, + "auxiliary_loss_mlp": 0.01037519, + "balance_loss_clip": 1.04925728, + "balance_loss_mlp": 1.02592587, + "epoch": 0.20068538447664283, + "flos": 23879159896320.0, + "grad_norm": 1.7889880616946787, + "language_loss": 0.8512522, + "learning_rate": 3.7022409178273436e-06, + "loss": 0.87333572, + "num_input_tokens_seen": 35364615, + "step": 1669, + "time_per_iteration": 2.915839195251465 + }, + { + "auxiliary_loss_clip": 0.01200174, + "auxiliary_loss_mlp": 0.01032172, + "balance_loss_clip": 1.05425048, + "balance_loss_mlp": 1.02053761, + "epoch": 0.2008056273672819, + "flos": 18442823270400.0, + "grad_norm": 2.2885034435247946, + "language_loss": 0.787395, + "learning_rate": 3.7018318516338054e-06, + "loss": 0.80971843, + "num_input_tokens_seen": 35383775, + "step": 1670, + "time_per_iteration": 2.6959869861602783 + }, + { + "auxiliary_loss_clip": 0.0121285, + "auxiliary_loss_mlp": 0.01035113, + "balance_loss_clip": 1.05851483, + "balance_loss_mlp": 1.02347183, + "epoch": 0.200925870257921, + "flos": 23659530186240.0, + "grad_norm": 2.558336405002953, + "language_loss": 0.81563759, + "learning_rate": 3.7014225272737284e-06, + "loss": 0.83811724, + "num_input_tokens_seen": 35403000, + "step": 1671, + "time_per_iteration": 2.763674020767212 + }, + { + "auxiliary_loss_clip": 0.01199635, + "auxiliary_loss_mlp": 0.01034017, + "balance_loss_clip": 1.05472755, + "balance_loss_mlp": 1.02250099, + "epoch": 0.20104611314856008, + "flos": 16217115909120.0, + "grad_norm": 2.2949597605222904, + "language_loss": 0.739416, + "learning_rate": 3.701012944809207e-06, + "loss": 0.76175255, + "num_input_tokens_seen": 35420115, + "step": 1672, + "time_per_iteration": 2.7497506141662598 + }, + { + "auxiliary_loss_clip": 0.01196365, + "auxiliary_loss_mlp": 0.00888347, + "balance_loss_clip": 1.05744052, + "balance_loss_mlp": 1.00041556, + "epoch": 0.2011663560391992, + "flos": 21397373498880.0, + "grad_norm": 2.0001468908245754, + "language_loss": 0.78783357, + "learning_rate": 3.700603104302374e-06, + "loss": 0.80868065, + "num_input_tokens_seen": 35439925, + "step": 1673, + "time_per_iteration": 2.8791415691375732 + }, + { + "auxiliary_loss_clip": 0.01071123, + "auxiliary_loss_mlp": 0.01004404, + "balance_loss_clip": 1.02220201, + "balance_loss_mlp": 1.00044656, + "epoch": 0.20128659892983827, + "flos": 62229459409920.0, + "grad_norm": 0.9013679160181302, + "language_loss": 0.55936301, + "learning_rate": 3.7001930058154027e-06, + "loss": 0.5801183, + "num_input_tokens_seen": 35504885, + "step": 1674, + "time_per_iteration": 3.3867311477661133 + }, + { + "auxiliary_loss_clip": 0.01186684, + "auxiliary_loss_mlp": 0.01037999, + "balance_loss_clip": 1.05511546, + "balance_loss_mlp": 1.02635789, + "epoch": 0.20140684182047736, + "flos": 28438737448320.0, + "grad_norm": 2.912669585629389, + "language_loss": 0.79395592, + "learning_rate": 3.6997826494105037e-06, + "loss": 0.8162027, + "num_input_tokens_seen": 35525330, + "step": 1675, + "time_per_iteration": 3.047508955001831 + }, + { + "auxiliary_loss_clip": 0.01195601, + "auxiliary_loss_mlp": 0.0103331, + "balance_loss_clip": 1.0558238, + "balance_loss_mlp": 1.02166915, + "epoch": 0.20152708471111647, + "flos": 28074064619520.0, + "grad_norm": 2.0998532003468617, + "language_loss": 0.69778615, + "learning_rate": 3.6993720351499286e-06, + "loss": 0.72007525, + "num_input_tokens_seen": 35546455, + "step": 1676, + "time_per_iteration": 2.8647217750549316 + }, + { + "auxiliary_loss_clip": 0.0119402, + "auxiliary_loss_mlp": 0.0103414, + "balance_loss_clip": 1.06070423, + "balance_loss_mlp": 1.02320242, + "epoch": 0.20164732760175555, + "flos": 23549751244800.0, + "grad_norm": 1.9817477364667502, + "language_loss": 0.77444804, + "learning_rate": 3.6989611630959666e-06, + "loss": 0.79672968, + "num_input_tokens_seen": 35565010, + "step": 1677, + "time_per_iteration": 2.754204750061035 + }, + { + "auxiliary_loss_clip": 0.01099846, + "auxiliary_loss_mlp": 0.0100461, + "balance_loss_clip": 1.02386713, + "balance_loss_mlp": 1.00019979, + "epoch": 0.20176757049239463, + "flos": 71100616037760.0, + "grad_norm": 0.6886441767300082, + "language_loss": 0.58271754, + "learning_rate": 3.6985500333109474e-06, + "loss": 0.60376215, + "num_input_tokens_seen": 35633340, + "step": 1678, + "time_per_iteration": 3.4038796424865723 + }, + { + "auxiliary_loss_clip": 0.01175934, + "auxiliary_loss_mlp": 0.01030118, + "balance_loss_clip": 1.05058742, + "balance_loss_mlp": 1.01984215, + "epoch": 0.20188781338303372, + "flos": 21430159637760.0, + "grad_norm": 2.1531643188331184, + "language_loss": 0.76411396, + "learning_rate": 3.6981386458572385e-06, + "loss": 0.78617454, + "num_input_tokens_seen": 35651315, + "step": 1679, + "time_per_iteration": 2.8188536167144775 + }, + { + "auxiliary_loss_clip": 0.01183096, + "auxiliary_loss_mlp": 0.0104204, + "balance_loss_clip": 1.05317891, + "balance_loss_mlp": 1.03034568, + "epoch": 0.20200805627367283, + "flos": 11546215130880.0, + "grad_norm": 2.9396975440704596, + "language_loss": 0.75937676, + "learning_rate": 3.6977270007972468e-06, + "loss": 0.78162813, + "num_input_tokens_seen": 35668850, + "step": 1680, + "time_per_iteration": 2.8471763134002686 + }, + { + "auxiliary_loss_clip": 0.01199618, + "auxiliary_loss_mlp": 0.01035436, + "balance_loss_clip": 1.05665672, + "balance_loss_mlp": 1.02426052, + "epoch": 0.2021282991643119, + "flos": 28545391906560.0, + "grad_norm": 2.3807012350207315, + "language_loss": 0.72936517, + "learning_rate": 3.6973150981934196e-06, + "loss": 0.75171572, + "num_input_tokens_seen": 35690080, + "step": 1681, + "time_per_iteration": 3.7950634956359863 + }, + { + "auxiliary_loss_clip": 0.01222134, + "auxiliary_loss_mlp": 0.01041104, + "balance_loss_clip": 1.06048393, + "balance_loss_mlp": 1.02909386, + "epoch": 0.202248542054951, + "flos": 17923446564480.0, + "grad_norm": 2.524870166407546, + "language_loss": 0.83990735, + "learning_rate": 3.6969029381082415e-06, + "loss": 0.86253977, + "num_input_tokens_seen": 35706075, + "step": 1682, + "time_per_iteration": 4.611663818359375 + }, + { + "auxiliary_loss_clip": 0.01193352, + "auxiliary_loss_mlp": 0.01042041, + "balance_loss_clip": 1.05494404, + "balance_loss_mlp": 1.03127658, + "epoch": 0.2023687849455901, + "flos": 19864634296320.0, + "grad_norm": 3.5596566533069707, + "language_loss": 0.79532075, + "learning_rate": 3.696490520604237e-06, + "loss": 0.81767464, + "num_input_tokens_seen": 35724765, + "step": 1683, + "time_per_iteration": 2.9522624015808105 + }, + { + "auxiliary_loss_clip": 0.01203161, + "auxiliary_loss_mlp": 0.0104037, + "balance_loss_clip": 1.05690718, + "balance_loss_mlp": 1.02932501, + "epoch": 0.20248902783622919, + "flos": 22564721600640.0, + "grad_norm": 1.8116261333523855, + "language_loss": 0.80803365, + "learning_rate": 3.696077845743968e-06, + "loss": 0.83046895, + "num_input_tokens_seen": 35744355, + "step": 1684, + "time_per_iteration": 2.792714834213257 + }, + { + "auxiliary_loss_clip": 0.01220271, + "auxiliary_loss_mlp": 0.01038158, + "balance_loss_clip": 1.05997467, + "balance_loss_mlp": 1.02683377, + "epoch": 0.20260927072686827, + "flos": 22709728805760.0, + "grad_norm": 2.639898050082237, + "language_loss": 0.72972685, + "learning_rate": 3.69566491359004e-06, + "loss": 0.75231111, + "num_input_tokens_seen": 35761000, + "step": 1685, + "time_per_iteration": 2.7257862091064453 + }, + { + "auxiliary_loss_clip": 0.01192479, + "auxiliary_loss_mlp": 0.01040795, + "balance_loss_clip": 1.05371141, + "balance_loss_mlp": 1.02856433, + "epoch": 0.20272951361750738, + "flos": 51023998650240.0, + "grad_norm": 1.7735884435917468, + "language_loss": 0.69319308, + "learning_rate": 3.695251724205092e-06, + "loss": 0.71552575, + "num_input_tokens_seen": 35785360, + "step": 1686, + "time_per_iteration": 3.1011500358581543 + }, + { + "auxiliary_loss_clip": 0.01216887, + "auxiliary_loss_mlp": 0.01032871, + "balance_loss_clip": 1.05868149, + "balance_loss_mlp": 1.02171254, + "epoch": 0.20284975650814646, + "flos": 26578133879040.0, + "grad_norm": 1.8738800814540046, + "language_loss": 0.86552715, + "learning_rate": 3.6948382776518054e-06, + "loss": 0.88802475, + "num_input_tokens_seen": 35806065, + "step": 1687, + "time_per_iteration": 2.8289589881896973 + }, + { + "auxiliary_loss_clip": 0.01195901, + "auxiliary_loss_mlp": 0.01038755, + "balance_loss_clip": 1.05474889, + "balance_loss_mlp": 1.02651215, + "epoch": 0.20296999939878554, + "flos": 16034222833920.0, + "grad_norm": 2.482761761742558, + "language_loss": 0.79482949, + "learning_rate": 3.6944245739929e-06, + "loss": 0.81717604, + "num_input_tokens_seen": 35822225, + "step": 1688, + "time_per_iteration": 2.7835066318511963 + }, + { + "auxiliary_loss_clip": 0.01208386, + "auxiliary_loss_mlp": 0.01042723, + "balance_loss_clip": 1.05934048, + "balance_loss_mlp": 1.03024769, + "epoch": 0.20309024228942463, + "flos": 19203374868480.0, + "grad_norm": 2.250305939421092, + "language_loss": 0.72144634, + "learning_rate": 3.6940106132911332e-06, + "loss": 0.7439574, + "num_input_tokens_seen": 35839410, + "step": 1689, + "time_per_iteration": 2.8648245334625244 + }, + { + "auxiliary_loss_clip": 0.01207807, + "auxiliary_loss_mlp": 0.01036915, + "balance_loss_clip": 1.05745053, + "balance_loss_mlp": 1.02541137, + "epoch": 0.20321048518006374, + "flos": 22821087945600.0, + "grad_norm": 3.9579331638984496, + "language_loss": 0.88982278, + "learning_rate": 3.6935963956093037e-06, + "loss": 0.91226995, + "num_input_tokens_seen": 35859495, + "step": 1690, + "time_per_iteration": 2.6973981857299805 + }, + { + "auxiliary_loss_clip": 0.01199685, + "auxiliary_loss_mlp": 0.01037532, + "balance_loss_clip": 1.05705845, + "balance_loss_mlp": 1.02613544, + "epoch": 0.20333072807070282, + "flos": 19096397187840.0, + "grad_norm": 2.070397153908224, + "language_loss": 0.6887213, + "learning_rate": 3.6931819210102474e-06, + "loss": 0.71109343, + "num_input_tokens_seen": 35878890, + "step": 1691, + "time_per_iteration": 2.7361011505126953 + }, + { + "auxiliary_loss_clip": 0.01221378, + "auxiliary_loss_mlp": 0.01049596, + "balance_loss_clip": 1.06046295, + "balance_loss_mlp": 1.03743672, + "epoch": 0.2034509709613419, + "flos": 18180962144640.0, + "grad_norm": 1.9285653335822988, + "language_loss": 0.84449977, + "learning_rate": 3.6927671895568402e-06, + "loss": 0.86720955, + "num_input_tokens_seen": 35897950, + "step": 1692, + "time_per_iteration": 2.764232873916626 + }, + { + "auxiliary_loss_clip": 0.01221747, + "auxiliary_loss_mlp": 0.01040149, + "balance_loss_clip": 1.06254101, + "balance_loss_mlp": 1.02866316, + "epoch": 0.20357121385198101, + "flos": 22923899648640.0, + "grad_norm": 2.486677076026541, + "language_loss": 0.86835098, + "learning_rate": 3.692352201311996e-06, + "loss": 0.89096987, + "num_input_tokens_seen": 35916800, + "step": 1693, + "time_per_iteration": 2.6378989219665527 + }, + { + "auxiliary_loss_clip": 0.01181295, + "auxiliary_loss_mlp": 0.01039195, + "balance_loss_clip": 1.0540241, + "balance_loss_mlp": 1.02708983, + "epoch": 0.2036914567426201, + "flos": 20922131629440.0, + "grad_norm": 1.9825920873308467, + "language_loss": 0.7633996, + "learning_rate": 3.6919369563386687e-06, + "loss": 0.78560448, + "num_input_tokens_seen": 35936600, + "step": 1694, + "time_per_iteration": 2.786708354949951 + }, + { + "auxiliary_loss_clip": 0.01192924, + "auxiliary_loss_mlp": 0.01036681, + "balance_loss_clip": 1.0562396, + "balance_loss_mlp": 1.02471805, + "epoch": 0.20381169963325918, + "flos": 15519155760000.0, + "grad_norm": 4.029897016760565, + "language_loss": 0.79117703, + "learning_rate": 3.69152145469985e-06, + "loss": 0.81347311, + "num_input_tokens_seen": 35953645, + "step": 1695, + "time_per_iteration": 2.745020627975464 + }, + { + "auxiliary_loss_clip": 0.01182133, + "auxiliary_loss_mlp": 0.01038712, + "balance_loss_clip": 1.05283141, + "balance_loss_mlp": 1.02631962, + "epoch": 0.20393194252389826, + "flos": 28833143760000.0, + "grad_norm": 2.2954758315858697, + "language_loss": 0.8221035, + "learning_rate": 3.691105696458572e-06, + "loss": 0.84431195, + "num_input_tokens_seen": 35970940, + "step": 1696, + "time_per_iteration": 2.9441375732421875 + }, + { + "auxiliary_loss_clip": 0.01220597, + "auxiliary_loss_mlp": 0.01039884, + "balance_loss_clip": 1.06368446, + "balance_loss_mlp": 1.02798128, + "epoch": 0.20405218541453737, + "flos": 22488554810880.0, + "grad_norm": 2.5512049498180884, + "language_loss": 0.67917025, + "learning_rate": 3.690689681677904e-06, + "loss": 0.70177507, + "num_input_tokens_seen": 35989410, + "step": 1697, + "time_per_iteration": 2.651315689086914 + }, + { + "auxiliary_loss_clip": 0.01199532, + "auxiliary_loss_mlp": 0.01033145, + "balance_loss_clip": 1.05583632, + "balance_loss_mlp": 1.02177835, + "epoch": 0.20417242830517646, + "flos": 25374408278400.0, + "grad_norm": 1.7477347479528027, + "language_loss": 0.88795328, + "learning_rate": 3.690273410420956e-06, + "loss": 0.91028011, + "num_input_tokens_seen": 36009175, + "step": 1698, + "time_per_iteration": 2.857733964920044 + }, + { + "auxiliary_loss_clip": 0.01203895, + "auxiliary_loss_mlp": 0.0103516, + "balance_loss_clip": 1.05545974, + "balance_loss_mlp": 1.02291679, + "epoch": 0.20429267119581554, + "flos": 14793078240000.0, + "grad_norm": 2.400004721468116, + "language_loss": 0.76451707, + "learning_rate": 3.689856882750875e-06, + "loss": 0.78690767, + "num_input_tokens_seen": 36024375, + "step": 1699, + "time_per_iteration": 2.7172436714172363 + }, + { + "auxiliary_loss_clip": 0.01204058, + "auxiliary_loss_mlp": 0.01029064, + "balance_loss_clip": 1.05831909, + "balance_loss_mlp": 1.01777506, + "epoch": 0.20441291408645465, + "flos": 17781851151360.0, + "grad_norm": 1.8184869789681288, + "language_loss": 0.78429008, + "learning_rate": 3.6894400987308486e-06, + "loss": 0.80662131, + "num_input_tokens_seen": 36041895, + "step": 1700, + "time_per_iteration": 2.835535764694214 + }, + { + "auxiliary_loss_clip": 0.01212338, + "auxiliary_loss_mlp": 0.01041247, + "balance_loss_clip": 1.05839181, + "balance_loss_mlp": 1.02917743, + "epoch": 0.20453315697709373, + "flos": 16435668211200.0, + "grad_norm": 2.102628018954177, + "language_loss": 0.84807521, + "learning_rate": 3.6890230584241024e-06, + "loss": 0.87061107, + "num_input_tokens_seen": 36058825, + "step": 1701, + "time_per_iteration": 2.6623761653900146 + }, + { + "auxiliary_loss_clip": 0.01111877, + "auxiliary_loss_mlp": 0.01007844, + "balance_loss_clip": 1.02611899, + "balance_loss_mlp": 1.00336194, + "epoch": 0.20465339986773282, + "flos": 66713085653760.0, + "grad_norm": 1.0957383172555284, + "language_loss": 0.66370416, + "learning_rate": 3.6886057618939016e-06, + "loss": 0.68490136, + "num_input_tokens_seen": 36121645, + "step": 1702, + "time_per_iteration": 3.29964017868042 + }, + { + "auxiliary_loss_clip": 0.01177846, + "auxiliary_loss_mlp": 0.01035087, + "balance_loss_clip": 1.05102682, + "balance_loss_mlp": 1.0222367, + "epoch": 0.2047736427583719, + "flos": 41974114924800.0, + "grad_norm": 2.299441215441631, + "language_loss": 0.69170439, + "learning_rate": 3.6881882092035492e-06, + "loss": 0.71383381, + "num_input_tokens_seen": 36143030, + "step": 1703, + "time_per_iteration": 3.0201728343963623 + }, + { + "auxiliary_loss_clip": 0.01095156, + "auxiliary_loss_mlp": 0.00878296, + "balance_loss_clip": 1.02514625, + "balance_loss_mlp": 1.00012767, + "epoch": 0.204893885649011, + "flos": 69940878641280.0, + "grad_norm": 0.9271458397257468, + "language_loss": 0.61232209, + "learning_rate": 3.6877704004163873e-06, + "loss": 0.63205665, + "num_input_tokens_seen": 36203435, + "step": 1704, + "time_per_iteration": 3.4790055751800537 + }, + { + "auxiliary_loss_clip": 0.01220926, + "auxiliary_loss_mlp": 0.01044191, + "balance_loss_clip": 1.0613569, + "balance_loss_mlp": 1.03172135, + "epoch": 0.2050141285396501, + "flos": 22200012858240.0, + "grad_norm": 2.6544098710774433, + "language_loss": 0.77632195, + "learning_rate": 3.6873523355957984e-06, + "loss": 0.79897308, + "num_input_tokens_seen": 36222435, + "step": 1705, + "time_per_iteration": 2.706390619277954 + }, + { + "auxiliary_loss_clip": 0.01109813, + "auxiliary_loss_mlp": 0.01004699, + "balance_loss_clip": 1.02357268, + "balance_loss_mlp": 1.00035942, + "epoch": 0.20513437143028918, + "flos": 46283721730560.0, + "grad_norm": 0.9869309800328843, + "language_loss": 0.64072692, + "learning_rate": 3.686934014805201e-06, + "loss": 0.66187203, + "num_input_tokens_seen": 36273065, + "step": 1706, + "time_per_iteration": 3.060594320297241 + }, + { + "auxiliary_loss_clip": 0.01207004, + "auxiliary_loss_mlp": 0.01042128, + "balance_loss_clip": 1.05808687, + "balance_loss_mlp": 1.02986693, + "epoch": 0.20525461432092829, + "flos": 21904324099200.0, + "grad_norm": 1.9771803172325995, + "language_loss": 0.80917752, + "learning_rate": 3.6865154381080552e-06, + "loss": 0.83166879, + "num_input_tokens_seen": 36293750, + "step": 1707, + "time_per_iteration": 3.6347832679748535 + }, + { + "auxiliary_loss_clip": 0.01157628, + "auxiliary_loss_mlp": 0.01033538, + "balance_loss_clip": 1.04894733, + "balance_loss_mlp": 1.02159286, + "epoch": 0.20537485721156737, + "flos": 21214264942080.0, + "grad_norm": 2.199145954090321, + "language_loss": 0.8253122, + "learning_rate": 3.6860966055678585e-06, + "loss": 0.84722388, + "num_input_tokens_seen": 36310105, + "step": 1708, + "time_per_iteration": 5.766002416610718 + }, + { + "auxiliary_loss_clip": 0.01208163, + "auxiliary_loss_mlp": 0.01040484, + "balance_loss_clip": 1.05848789, + "balance_loss_mlp": 1.02837813, + "epoch": 0.20549510010220645, + "flos": 20191205773440.0, + "grad_norm": 1.8547135898009155, + "language_loss": 0.86310041, + "learning_rate": 3.685677517248147e-06, + "loss": 0.88558686, + "num_input_tokens_seen": 36328995, + "step": 1709, + "time_per_iteration": 2.8093502521514893 + }, + { + "auxiliary_loss_clip": 0.01200459, + "auxiliary_loss_mlp": 0.00888157, + "balance_loss_clip": 1.06252646, + "balance_loss_mlp": 1.0005939, + "epoch": 0.20561534299284553, + "flos": 17016702612480.0, + "grad_norm": 3.9074483557966824, + "language_loss": 0.80467981, + "learning_rate": 3.6852581732124967e-06, + "loss": 0.82556593, + "num_input_tokens_seen": 36346340, + "step": 1710, + "time_per_iteration": 2.8872227668762207 + }, + { + "auxiliary_loss_clip": 0.01209015, + "auxiliary_loss_mlp": 0.0104701, + "balance_loss_clip": 1.05779219, + "balance_loss_mlp": 1.03480327, + "epoch": 0.20573558588348465, + "flos": 22890467064960.0, + "grad_norm": 1.8371126087250498, + "language_loss": 0.76177818, + "learning_rate": 3.6848385735245213e-06, + "loss": 0.78433847, + "num_input_tokens_seen": 36365430, + "step": 1711, + "time_per_iteration": 2.7710983753204346 + }, + { + "auxiliary_loss_clip": 0.01193389, + "auxiliary_loss_mlp": 0.01034997, + "balance_loss_clip": 1.05234885, + "balance_loss_mlp": 1.02346361, + "epoch": 0.20585582877412373, + "flos": 24643123286400.0, + "grad_norm": 1.9375033195023983, + "language_loss": 0.8592751, + "learning_rate": 3.6844187182478734e-06, + "loss": 0.88155895, + "num_input_tokens_seen": 36386285, + "step": 1712, + "time_per_iteration": 2.852233409881592 + }, + { + "auxiliary_loss_clip": 0.01184644, + "auxiliary_loss_mlp": 0.01041519, + "balance_loss_clip": 1.05218601, + "balance_loss_mlp": 1.02885306, + "epoch": 0.2059760716647628, + "flos": 24206952435840.0, + "grad_norm": 1.8746630253281515, + "language_loss": 0.75460792, + "learning_rate": 3.683998607446246e-06, + "loss": 0.7768696, + "num_input_tokens_seen": 36404935, + "step": 1713, + "time_per_iteration": 2.819978713989258 + }, + { + "auxiliary_loss_clip": 0.01210807, + "auxiliary_loss_mlp": 0.01038492, + "balance_loss_clip": 1.0610311, + "balance_loss_mlp": 1.02717304, + "epoch": 0.20609631455540192, + "flos": 20229522606720.0, + "grad_norm": 2.2166217959294556, + "language_loss": 0.75388312, + "learning_rate": 3.6835782411833686e-06, + "loss": 0.77637613, + "num_input_tokens_seen": 36424455, + "step": 1714, + "time_per_iteration": 2.75138521194458 + }, + { + "auxiliary_loss_clip": 0.01175504, + "auxiliary_loss_mlp": 0.0104743, + "balance_loss_clip": 1.05056834, + "balance_loss_mlp": 1.03629017, + "epoch": 0.206216557446041, + "flos": 19864957518720.0, + "grad_norm": 1.950129922955988, + "language_loss": 0.74515641, + "learning_rate": 3.68315761952301e-06, + "loss": 0.76738572, + "num_input_tokens_seen": 36441685, + "step": 1715, + "time_per_iteration": 2.7300291061401367 + }, + { + "auxiliary_loss_clip": 0.01216761, + "auxiliary_loss_mlp": 0.010333, + "balance_loss_clip": 1.05806255, + "balance_loss_mlp": 1.02176678, + "epoch": 0.2063368003366801, + "flos": 24096311568000.0, + "grad_norm": 2.4824017938947, + "language_loss": 0.82879698, + "learning_rate": 3.6827367425289797e-06, + "loss": 0.85129762, + "num_input_tokens_seen": 36461460, + "step": 1716, + "time_per_iteration": 2.8547556400299072 + }, + { + "auxiliary_loss_clip": 0.01197618, + "auxiliary_loss_mlp": 0.01046223, + "balance_loss_clip": 1.05722082, + "balance_loss_mlp": 1.03451657, + "epoch": 0.2064570432273192, + "flos": 20340163474560.0, + "grad_norm": 2.408257032372282, + "language_loss": 0.72732097, + "learning_rate": 3.6823156102651225e-06, + "loss": 0.74975932, + "num_input_tokens_seen": 36479615, + "step": 1717, + "time_per_iteration": 2.7261483669281006 + }, + { + "auxiliary_loss_clip": 0.01154177, + "auxiliary_loss_mlp": 0.01036027, + "balance_loss_clip": 1.05105877, + "balance_loss_mlp": 1.02479172, + "epoch": 0.20657728611795828, + "flos": 20520363029760.0, + "grad_norm": 1.949988559806639, + "language_loss": 0.70591629, + "learning_rate": 3.6818942227953257e-06, + "loss": 0.72781825, + "num_input_tokens_seen": 36500160, + "step": 1718, + "time_per_iteration": 2.9393720626831055 + }, + { + "auxiliary_loss_clip": 0.01189362, + "auxiliary_loss_mlp": 0.01036036, + "balance_loss_clip": 1.05605054, + "balance_loss_mlp": 1.02399564, + "epoch": 0.20669752900859736, + "flos": 21799285752960.0, + "grad_norm": 1.859230392192973, + "language_loss": 0.69010311, + "learning_rate": 3.681472580183512e-06, + "loss": 0.71235704, + "num_input_tokens_seen": 36518810, + "step": 1719, + "time_per_iteration": 2.8928232192993164 + }, + { + "auxiliary_loss_clip": 0.01204045, + "auxiliary_loss_mlp": 0.01035048, + "balance_loss_clip": 1.05860841, + "balance_loss_mlp": 1.02400875, + "epoch": 0.20681777189923645, + "flos": 15122020014720.0, + "grad_norm": 1.793127533177127, + "language_loss": 0.86251944, + "learning_rate": 3.6810506824936455e-06, + "loss": 0.88491035, + "num_input_tokens_seen": 36536890, + "step": 1720, + "time_per_iteration": 2.705306053161621 + }, + { + "auxiliary_loss_clip": 0.01092548, + "auxiliary_loss_mlp": 0.01010802, + "balance_loss_clip": 1.02023113, + "balance_loss_mlp": 1.00660551, + "epoch": 0.20693801478987556, + "flos": 56481021509760.0, + "grad_norm": 1.0565000141825793, + "language_loss": 0.62564421, + "learning_rate": 3.680628529789726e-06, + "loss": 0.64667767, + "num_input_tokens_seen": 36589300, + "step": 1721, + "time_per_iteration": 3.1544976234436035 + }, + { + "auxiliary_loss_clip": 0.01225071, + "auxiliary_loss_mlp": 0.01045013, + "balance_loss_clip": 1.0623467, + "balance_loss_mlp": 1.03254402, + "epoch": 0.20705825768051464, + "flos": 21614201948160.0, + "grad_norm": 1.7749920495785556, + "language_loss": 0.86386627, + "learning_rate": 3.680206122135796e-06, + "loss": 0.88656706, + "num_input_tokens_seen": 36609905, + "step": 1722, + "time_per_iteration": 2.721586227416992 + }, + { + "auxiliary_loss_clip": 0.01185618, + "auxiliary_loss_mlp": 0.01049994, + "balance_loss_clip": 1.05706346, + "balance_loss_mlp": 1.03790045, + "epoch": 0.20717850057115372, + "flos": 25848895962240.0, + "grad_norm": 1.8552222590045462, + "language_loss": 0.78073478, + "learning_rate": 3.6797834595959323e-06, + "loss": 0.80309093, + "num_input_tokens_seen": 36629805, + "step": 1723, + "time_per_iteration": 2.9317715167999268 + }, + { + "auxiliary_loss_clip": 0.01166424, + "auxiliary_loss_mlp": 0.01040952, + "balance_loss_clip": 1.0492065, + "balance_loss_mlp": 1.02802348, + "epoch": 0.20729874346179283, + "flos": 29130807767040.0, + "grad_norm": 2.7333266705588253, + "language_loss": 0.77995133, + "learning_rate": 3.679360542234254e-06, + "loss": 0.80202508, + "num_input_tokens_seen": 36649150, + "step": 1724, + "time_per_iteration": 2.9578003883361816 + }, + { + "auxiliary_loss_clip": 0.01189254, + "auxiliary_loss_mlp": 0.00888791, + "balance_loss_clip": 1.05173194, + "balance_loss_mlp": 1.00037158, + "epoch": 0.20741898635243192, + "flos": 29023363209600.0, + "grad_norm": 1.8994738528104973, + "language_loss": 0.72315443, + "learning_rate": 3.678937370114916e-06, + "loss": 0.74393487, + "num_input_tokens_seen": 36668955, + "step": 1725, + "time_per_iteration": 2.828587770462036 + }, + { + "auxiliary_loss_clip": 0.01191187, + "auxiliary_loss_mlp": 0.01041817, + "balance_loss_clip": 1.05553865, + "balance_loss_mlp": 1.03099895, + "epoch": 0.207539229243071, + "flos": 15559447841280.0, + "grad_norm": 2.1564852274012094, + "language_loss": 0.78629792, + "learning_rate": 3.678513943302114e-06, + "loss": 0.80862802, + "num_input_tokens_seen": 36685730, + "step": 1726, + "time_per_iteration": 2.7595579624176025 + }, + { + "auxiliary_loss_clip": 0.01215542, + "auxiliary_loss_mlp": 0.01034546, + "balance_loss_clip": 1.05928636, + "balance_loss_mlp": 1.02342415, + "epoch": 0.20765947213371008, + "flos": 20521081301760.0, + "grad_norm": 1.6666975790249245, + "language_loss": 0.85456562, + "learning_rate": 3.678090261860082e-06, + "loss": 0.87706649, + "num_input_tokens_seen": 36705460, + "step": 1727, + "time_per_iteration": 2.7851343154907227 + }, + { + "auxiliary_loss_clip": 0.01186182, + "auxiliary_loss_mlp": 0.01038543, + "balance_loss_clip": 1.05098534, + "balance_loss_mlp": 1.0269084, + "epoch": 0.2077797150243492, + "flos": 19354415558400.0, + "grad_norm": 2.060815337342906, + "language_loss": 0.77892303, + "learning_rate": 3.6776663258530906e-06, + "loss": 0.80117029, + "num_input_tokens_seen": 36724110, + "step": 1728, + "time_per_iteration": 2.887725591659546 + }, + { + "auxiliary_loss_clip": 0.01209831, + "auxiliary_loss_mlp": 0.01042378, + "balance_loss_clip": 1.05829453, + "balance_loss_mlp": 1.0309813, + "epoch": 0.20789995791498828, + "flos": 21829952989440.0, + "grad_norm": 1.9796428694984305, + "language_loss": 0.71167588, + "learning_rate": 3.6772421353454516e-06, + "loss": 0.73419797, + "num_input_tokens_seen": 36742705, + "step": 1729, + "time_per_iteration": 2.858142614364624 + }, + { + "auxiliary_loss_clip": 0.01206514, + "auxiliary_loss_mlp": 0.01038771, + "balance_loss_clip": 1.06103969, + "balance_loss_mlp": 1.02643251, + "epoch": 0.20802020080562736, + "flos": 23148844571520.0, + "grad_norm": 2.3782318648316862, + "language_loss": 0.88903403, + "learning_rate": 3.6768176904015153e-06, + "loss": 0.91148686, + "num_input_tokens_seen": 36762510, + "step": 1730, + "time_per_iteration": 2.8341400623321533 + }, + { + "auxiliary_loss_clip": 0.01207359, + "auxiliary_loss_mlp": 0.01033673, + "balance_loss_clip": 1.05594158, + "balance_loss_mlp": 1.02262247, + "epoch": 0.20814044369626647, + "flos": 23072677781760.0, + "grad_norm": 1.946007950498287, + "language_loss": 0.60585785, + "learning_rate": 3.6763929910856674e-06, + "loss": 0.62826812, + "num_input_tokens_seen": 36780960, + "step": 1731, + "time_per_iteration": 2.682110548019409 + }, + { + "auxiliary_loss_clip": 0.01203698, + "auxiliary_loss_mlp": 0.0103958, + "balance_loss_clip": 1.05873859, + "balance_loss_mlp": 1.02838671, + "epoch": 0.20826068658690555, + "flos": 19608016556160.0, + "grad_norm": 2.0901079000007035, + "language_loss": 0.77445829, + "learning_rate": 3.6759680374623365e-06, + "loss": 0.79689103, + "num_input_tokens_seen": 36798875, + "step": 1732, + "time_per_iteration": 2.6898961067199707 + }, + { + "auxiliary_loss_clip": 0.01216614, + "auxiliary_loss_mlp": 0.01034563, + "balance_loss_clip": 1.06090462, + "balance_loss_mlp": 1.02346492, + "epoch": 0.20838092947754464, + "flos": 25374049142400.0, + "grad_norm": 2.4367314913956153, + "language_loss": 0.75702989, + "learning_rate": 3.675542829595986e-06, + "loss": 0.77954173, + "num_input_tokens_seen": 36818540, + "step": 1733, + "time_per_iteration": 3.569962739944458 + }, + { + "auxiliary_loss_clip": 0.01197221, + "auxiliary_loss_mlp": 0.01033028, + "balance_loss_clip": 1.05602467, + "balance_loss_mlp": 1.02187562, + "epoch": 0.20850117236818372, + "flos": 24061729749120.0, + "grad_norm": 1.4710685468806277, + "language_loss": 0.79311895, + "learning_rate": 3.6751173675511213e-06, + "loss": 0.8154214, + "num_input_tokens_seen": 36840585, + "step": 1734, + "time_per_iteration": 5.69762659072876 + }, + { + "auxiliary_loss_clip": 0.01193014, + "auxiliary_loss_mlp": 0.01038551, + "balance_loss_clip": 1.05053091, + "balance_loss_mlp": 1.02739334, + "epoch": 0.20862141525882283, + "flos": 20077799558400.0, + "grad_norm": 2.559295406172323, + "language_loss": 0.87541139, + "learning_rate": 3.674691651392283e-06, + "loss": 0.89772701, + "num_input_tokens_seen": 36858255, + "step": 1735, + "time_per_iteration": 2.8700764179229736 + }, + { + "auxiliary_loss_clip": 0.01201646, + "auxiliary_loss_mlp": 0.01040989, + "balance_loss_clip": 1.05722976, + "balance_loss_mlp": 1.02919865, + "epoch": 0.2087416581494619, + "flos": 39015183237120.0, + "grad_norm": 2.0605026353445366, + "language_loss": 0.75803351, + "learning_rate": 3.674265681184053e-06, + "loss": 0.78045988, + "num_input_tokens_seen": 36881515, + "step": 1736, + "time_per_iteration": 2.925290822982788 + }, + { + "auxiliary_loss_clip": 0.01196735, + "auxiliary_loss_mlp": 0.0103878, + "balance_loss_clip": 1.05552077, + "balance_loss_mlp": 1.0275743, + "epoch": 0.208861901040101, + "flos": 26101994169600.0, + "grad_norm": 2.9415767711877994, + "language_loss": 0.86025369, + "learning_rate": 3.6738394569910504e-06, + "loss": 0.88260877, + "num_input_tokens_seen": 36902055, + "step": 1737, + "time_per_iteration": 2.8004214763641357 + }, + { + "auxiliary_loss_clip": 0.01207228, + "auxiliary_loss_mlp": 0.01034937, + "balance_loss_clip": 1.06000948, + "balance_loss_mlp": 1.02299237, + "epoch": 0.2089821439307401, + "flos": 28398732675840.0, + "grad_norm": 2.126242049283218, + "language_loss": 0.82642877, + "learning_rate": 3.6734129788779333e-06, + "loss": 0.84885043, + "num_input_tokens_seen": 36921230, + "step": 1738, + "time_per_iteration": 2.8433427810668945 + }, + { + "auxiliary_loss_clip": 0.01185592, + "auxiliary_loss_mlp": 0.01036583, + "balance_loss_clip": 1.05670333, + "balance_loss_mlp": 1.02583623, + "epoch": 0.2091023868213792, + "flos": 21069616872960.0, + "grad_norm": 1.6632637403731731, + "language_loss": 0.90578115, + "learning_rate": 3.6729862469093976e-06, + "loss": 0.92800289, + "num_input_tokens_seen": 36940325, + "step": 1739, + "time_per_iteration": 2.808664321899414 + }, + { + "auxiliary_loss_clip": 0.01183179, + "auxiliary_loss_mlp": 0.0103306, + "balance_loss_clip": 1.05173326, + "balance_loss_mlp": 1.02196121, + "epoch": 0.20922262971201827, + "flos": 22455481363200.0, + "grad_norm": 3.202220145777171, + "language_loss": 0.82775974, + "learning_rate": 3.6725592611501782e-06, + "loss": 0.84992212, + "num_input_tokens_seen": 36959000, + "step": 1740, + "time_per_iteration": 2.8704729080200195 + }, + { + "auxiliary_loss_clip": 0.01203811, + "auxiliary_loss_mlp": 0.01031596, + "balance_loss_clip": 1.05549586, + "balance_loss_mlp": 1.02102196, + "epoch": 0.20934287260265738, + "flos": 27852244179840.0, + "grad_norm": 2.333390969806885, + "language_loss": 0.77158695, + "learning_rate": 3.6721320216650496e-06, + "loss": 0.79394102, + "num_input_tokens_seen": 36979615, + "step": 1741, + "time_per_iteration": 2.833756446838379 + }, + { + "auxiliary_loss_clip": 0.01193066, + "auxiliary_loss_mlp": 0.01036, + "balance_loss_clip": 1.05576575, + "balance_loss_mlp": 1.02455592, + "epoch": 0.20946311549329646, + "flos": 16435309075200.0, + "grad_norm": 1.9463734490167783, + "language_loss": 0.83462524, + "learning_rate": 3.6717045285188215e-06, + "loss": 0.85691589, + "num_input_tokens_seen": 36997310, + "step": 1742, + "time_per_iteration": 2.781101942062378 + }, + { + "auxiliary_loss_clip": 0.01159864, + "auxiliary_loss_mlp": 0.01031118, + "balance_loss_clip": 1.04731989, + "balance_loss_mlp": 1.01943588, + "epoch": 0.20958335838393555, + "flos": 22492720788480.0, + "grad_norm": 12.302075467425382, + "language_loss": 0.86913961, + "learning_rate": 3.671276781776346e-06, + "loss": 0.89104939, + "num_input_tokens_seen": 37015965, + "step": 1743, + "time_per_iteration": 2.9262585639953613 + }, + { + "auxiliary_loss_clip": 0.01190779, + "auxiliary_loss_mlp": 0.01034314, + "balance_loss_clip": 1.05190921, + "balance_loss_mlp": 1.02357888, + "epoch": 0.20970360127457463, + "flos": 25224768218880.0, + "grad_norm": 2.1985715198171634, + "language_loss": 0.6732626, + "learning_rate": 3.6708487815025128e-06, + "loss": 0.69551361, + "num_input_tokens_seen": 37036545, + "step": 1744, + "time_per_iteration": 2.919337272644043 + }, + { + "auxiliary_loss_clip": 0.01182808, + "auxiliary_loss_mlp": 0.01036206, + "balance_loss_clip": 1.05103135, + "balance_loss_mlp": 1.02454138, + "epoch": 0.20982384416521374, + "flos": 18479164855680.0, + "grad_norm": 2.3573622153288083, + "language_loss": 0.74823171, + "learning_rate": 3.6704205277622463e-06, + "loss": 0.77042186, + "num_input_tokens_seen": 37054985, + "step": 1745, + "time_per_iteration": 2.806293487548828 + }, + { + "auxiliary_loss_clip": 0.01196069, + "auxiliary_loss_mlp": 0.01040603, + "balance_loss_clip": 1.05332506, + "balance_loss_mlp": 1.02819324, + "epoch": 0.20994408705585282, + "flos": 25373546352000.0, + "grad_norm": 1.8136147134506628, + "language_loss": 0.80557644, + "learning_rate": 3.6699920206205146e-06, + "loss": 0.82794315, + "num_input_tokens_seen": 37075725, + "step": 1746, + "time_per_iteration": 2.916473150253296 + }, + { + "auxiliary_loss_clip": 0.01206504, + "auxiliary_loss_mlp": 0.01032831, + "balance_loss_clip": 1.05584002, + "balance_loss_mlp": 1.02164924, + "epoch": 0.2100643299464919, + "flos": 21320955313920.0, + "grad_norm": 1.7412998043277297, + "language_loss": 0.81741613, + "learning_rate": 3.669563260142321e-06, + "loss": 0.83980948, + "num_input_tokens_seen": 37094615, + "step": 1747, + "time_per_iteration": 2.713815212249756 + }, + { + "auxiliary_loss_clip": 0.01192119, + "auxiliary_loss_mlp": 0.01042836, + "balance_loss_clip": 1.05850863, + "balance_loss_mlp": 1.03158259, + "epoch": 0.21018457283713102, + "flos": 19354379644800.0, + "grad_norm": 3.2815756871112565, + "language_loss": 0.84082651, + "learning_rate": 3.6691342463927083e-06, + "loss": 0.86317599, + "num_input_tokens_seen": 37113610, + "step": 1748, + "time_per_iteration": 2.9102115631103516 + }, + { + "auxiliary_loss_clip": 0.01189592, + "auxiliary_loss_mlp": 0.01042579, + "balance_loss_clip": 1.05307114, + "balance_loss_mlp": 1.0313549, + "epoch": 0.2103048157277701, + "flos": 28330035914880.0, + "grad_norm": 1.662459703931402, + "language_loss": 0.81622493, + "learning_rate": 3.668704979436758e-06, + "loss": 0.83854663, + "num_input_tokens_seen": 37133705, + "step": 1749, + "time_per_iteration": 2.9721415042877197 + }, + { + "auxiliary_loss_clip": 0.01187678, + "auxiliary_loss_mlp": 0.01039207, + "balance_loss_clip": 1.05216193, + "balance_loss_mlp": 1.02803063, + "epoch": 0.21042505861840918, + "flos": 17457290835840.0, + "grad_norm": 1.9567973883676462, + "language_loss": 0.78631902, + "learning_rate": 3.668275459339588e-06, + "loss": 0.80858791, + "num_input_tokens_seen": 37152185, + "step": 1750, + "time_per_iteration": 2.862520456314087 + }, + { + "auxiliary_loss_clip": 0.01216509, + "auxiliary_loss_mlp": 0.01037448, + "balance_loss_clip": 1.06106853, + "balance_loss_mlp": 1.02639759, + "epoch": 0.21054530150904827, + "flos": 14209817195520.0, + "grad_norm": 2.015421869493543, + "language_loss": 0.80460608, + "learning_rate": 3.667845686166358e-06, + "loss": 0.8271457, + "num_input_tokens_seen": 37169110, + "step": 1751, + "time_per_iteration": 2.650535821914673 + }, + { + "auxiliary_loss_clip": 0.01168554, + "auxiliary_loss_mlp": 0.01038997, + "balance_loss_clip": 1.05007815, + "balance_loss_mlp": 1.0264802, + "epoch": 0.21066554439968738, + "flos": 18618210403200.0, + "grad_norm": 1.9729128508679692, + "language_loss": 0.86021423, + "learning_rate": 3.6674156599822634e-06, + "loss": 0.88228977, + "num_input_tokens_seen": 37184905, + "step": 1752, + "time_per_iteration": 2.7723352909088135 + }, + { + "auxiliary_loss_clip": 0.01177114, + "auxiliary_loss_mlp": 0.01038086, + "balance_loss_clip": 1.05043399, + "balance_loss_mlp": 1.0266304, + "epoch": 0.21078578729032646, + "flos": 23658883741440.0, + "grad_norm": 1.9412157334823688, + "language_loss": 0.81783676, + "learning_rate": 3.666985380852539e-06, + "loss": 0.83998871, + "num_input_tokens_seen": 37203910, + "step": 1753, + "time_per_iteration": 2.9207143783569336 + }, + { + "auxiliary_loss_clip": 0.01195852, + "auxiliary_loss_mlp": 0.01037952, + "balance_loss_clip": 1.05628979, + "balance_loss_mlp": 1.02585244, + "epoch": 0.21090603018096554, + "flos": 29346379240320.0, + "grad_norm": 2.5108664113327928, + "language_loss": 0.74594378, + "learning_rate": 3.6665548488424576e-06, + "loss": 0.76828182, + "num_input_tokens_seen": 37222670, + "step": 1754, + "time_per_iteration": 2.8455939292907715 + }, + { + "auxiliary_loss_clip": 0.01217791, + "auxiliary_loss_mlp": 0.01039129, + "balance_loss_clip": 1.06033683, + "balance_loss_mlp": 1.02665424, + "epoch": 0.21102627307160465, + "flos": 23261245205760.0, + "grad_norm": 1.6233700902845118, + "language_loss": 0.87803602, + "learning_rate": 3.6661240640173307e-06, + "loss": 0.9006052, + "num_input_tokens_seen": 37244140, + "step": 1755, + "time_per_iteration": 2.692734479904175 + }, + { + "auxiliary_loss_clip": 0.01090562, + "auxiliary_loss_mlp": 0.01008066, + "balance_loss_clip": 1.02634239, + "balance_loss_mlp": 1.00384617, + "epoch": 0.21114651596224374, + "flos": 54633454577280.0, + "grad_norm": 0.8684823080609767, + "language_loss": 0.57974803, + "learning_rate": 3.6656930264425085e-06, + "loss": 0.60073435, + "num_input_tokens_seen": 37308185, + "step": 1756, + "time_per_iteration": 3.357358932495117 + }, + { + "auxiliary_loss_clip": 0.01213992, + "auxiliary_loss_mlp": 0.01040138, + "balance_loss_clip": 1.05872893, + "balance_loss_mlp": 1.02860439, + "epoch": 0.21126675885288282, + "flos": 21543314457600.0, + "grad_norm": 6.8752956212905385, + "language_loss": 0.75728214, + "learning_rate": 3.665261736183378e-06, + "loss": 0.77982348, + "num_input_tokens_seen": 37328220, + "step": 1757, + "time_per_iteration": 2.7481439113616943 + }, + { + "auxiliary_loss_clip": 0.01187607, + "auxiliary_loss_mlp": 0.01036471, + "balance_loss_clip": 1.05758893, + "balance_loss_mlp": 1.02548003, + "epoch": 0.2113870017435219, + "flos": 10961876678400.0, + "grad_norm": 2.5703199488280917, + "language_loss": 0.8860808, + "learning_rate": 3.664830193305366e-06, + "loss": 0.90832156, + "num_input_tokens_seen": 37345995, + "step": 1758, + "time_per_iteration": 2.8825745582580566 + }, + { + "auxiliary_loss_clip": 0.01179587, + "auxiliary_loss_mlp": 0.01035837, + "balance_loss_clip": 1.05095959, + "balance_loss_mlp": 1.02369547, + "epoch": 0.211507244634161, + "flos": 16653825463680.0, + "grad_norm": 7.875634767023557, + "language_loss": 0.77222502, + "learning_rate": 3.6643983978739373e-06, + "loss": 0.79437923, + "num_input_tokens_seen": 37362610, + "step": 1759, + "time_per_iteration": 3.71994948387146 + }, + { + "auxiliary_loss_clip": 0.01187329, + "auxiliary_loss_mlp": 0.01039261, + "balance_loss_clip": 1.05649447, + "balance_loss_mlp": 1.02658868, + "epoch": 0.2116274875248001, + "flos": 20954091755520.0, + "grad_norm": 1.8410349587253698, + "language_loss": 0.82332635, + "learning_rate": 3.663966349954596e-06, + "loss": 0.8455922, + "num_input_tokens_seen": 37382790, + "step": 1760, + "time_per_iteration": 5.587176084518433 + }, + { + "auxiliary_loss_clip": 0.01103622, + "auxiliary_loss_mlp": 0.01004128, + "balance_loss_clip": 1.02575946, + "balance_loss_mlp": 0.99988407, + "epoch": 0.21174773041543918, + "flos": 68196949424640.0, + "grad_norm": 0.7828710375257059, + "language_loss": 0.59763348, + "learning_rate": 3.6635340496128816e-06, + "loss": 0.61871099, + "num_input_tokens_seen": 37439720, + "step": 1761, + "time_per_iteration": 3.212275743484497 + }, + { + "auxiliary_loss_clip": 0.01168945, + "auxiliary_loss_mlp": 0.01041495, + "balance_loss_clip": 1.05139351, + "balance_loss_mlp": 1.03047991, + "epoch": 0.2118679733060783, + "flos": 20668315150080.0, + "grad_norm": 1.774408496064292, + "language_loss": 0.92907786, + "learning_rate": 3.6631014969143747e-06, + "loss": 0.95118225, + "num_input_tokens_seen": 37459410, + "step": 1762, + "time_per_iteration": 2.920410633087158 + }, + { + "auxiliary_loss_clip": 0.01206596, + "auxiliary_loss_mlp": 0.01041286, + "balance_loss_clip": 1.05956841, + "balance_loss_mlp": 1.02980614, + "epoch": 0.21198821619671737, + "flos": 23223431162880.0, + "grad_norm": 1.9459086304108315, + "language_loss": 0.88916039, + "learning_rate": 3.662668691924693e-06, + "loss": 0.91163921, + "num_input_tokens_seen": 37480460, + "step": 1763, + "time_per_iteration": 2.7437524795532227 + }, + { + "auxiliary_loss_clip": 0.01180161, + "auxiliary_loss_mlp": 0.01041906, + "balance_loss_clip": 1.05224335, + "balance_loss_mlp": 1.03047979, + "epoch": 0.21210845908735645, + "flos": 24498547044480.0, + "grad_norm": 3.1720168687789396, + "language_loss": 0.71387535, + "learning_rate": 3.6622356347094927e-06, + "loss": 0.73609602, + "num_input_tokens_seen": 37502025, + "step": 1764, + "time_per_iteration": 2.870091438293457 + }, + { + "auxiliary_loss_clip": 0.01182339, + "auxiliary_loss_mlp": 0.01039687, + "balance_loss_clip": 1.05132771, + "balance_loss_mlp": 1.02729511, + "epoch": 0.21222870197799554, + "flos": 27089789160960.0, + "grad_norm": 2.02753824765573, + "language_loss": 0.78572655, + "learning_rate": 3.6618023253344684e-06, + "loss": 0.8079468, + "num_input_tokens_seen": 37520885, + "step": 1765, + "time_per_iteration": 2.811317205429077 + }, + { + "auxiliary_loss_clip": 0.01201254, + "auxiliary_loss_mlp": 0.01048731, + "balance_loss_clip": 1.0538522, + "balance_loss_mlp": 1.03654218, + "epoch": 0.21234894486863465, + "flos": 16873850223360.0, + "grad_norm": 1.64684431343361, + "language_loss": 0.83256245, + "learning_rate": 3.6613687638653527e-06, + "loss": 0.85506225, + "num_input_tokens_seen": 37539055, + "step": 1766, + "time_per_iteration": 2.7117021083831787 + }, + { + "auxiliary_loss_clip": 0.01191218, + "auxiliary_loss_mlp": 0.0104623, + "balance_loss_clip": 1.0563947, + "balance_loss_mlp": 1.03465462, + "epoch": 0.21246918775927373, + "flos": 23474949171840.0, + "grad_norm": 1.7564742252800762, + "language_loss": 0.77655458, + "learning_rate": 3.660934950367916e-06, + "loss": 0.7989291, + "num_input_tokens_seen": 37558300, + "step": 1767, + "time_per_iteration": 2.8122899532318115 + }, + { + "auxiliary_loss_clip": 0.01206419, + "auxiliary_loss_mlp": 0.01036445, + "balance_loss_clip": 1.05787563, + "balance_loss_mlp": 1.02487612, + "epoch": 0.21258943064991281, + "flos": 22382295402240.0, + "grad_norm": 2.896946700137205, + "language_loss": 0.83451474, + "learning_rate": 3.660500884907968e-06, + "loss": 0.85694337, + "num_input_tokens_seen": 37579040, + "step": 1768, + "time_per_iteration": 2.7299320697784424 + }, + { + "auxiliary_loss_clip": 0.01070718, + "auxiliary_loss_mlp": 0.01006151, + "balance_loss_clip": 1.02049947, + "balance_loss_mlp": 1.0020262, + "epoch": 0.21270967354055192, + "flos": 59440168679040.0, + "grad_norm": 0.82846731063255, + "language_loss": 0.6001091, + "learning_rate": 3.660066567551356e-06, + "loss": 0.62087774, + "num_input_tokens_seen": 37639185, + "step": 1769, + "time_per_iteration": 3.289867877960205 + }, + { + "auxiliary_loss_clip": 0.0120554, + "auxiliary_loss_mlp": 0.00888062, + "balance_loss_clip": 1.05946684, + "balance_loss_mlp": 1.00038946, + "epoch": 0.212829916431191, + "flos": 21544032729600.0, + "grad_norm": 2.4681787124212287, + "language_loss": 0.84276325, + "learning_rate": 3.6596319983639657e-06, + "loss": 0.86369926, + "num_input_tokens_seen": 37657765, + "step": 1770, + "time_per_iteration": 2.6962485313415527 + }, + { + "auxiliary_loss_clip": 0.01182959, + "auxiliary_loss_mlp": 0.00888387, + "balance_loss_clip": 1.05415618, + "balance_loss_mlp": 1.00044572, + "epoch": 0.2129501593218301, + "flos": 28987739896320.0, + "grad_norm": 1.771194731167034, + "language_loss": 0.86395347, + "learning_rate": 3.6591971774117214e-06, + "loss": 0.88466692, + "num_input_tokens_seen": 37680740, + "step": 1771, + "time_per_iteration": 2.9596879482269287 + }, + { + "auxiliary_loss_clip": 0.01208053, + "auxiliary_loss_mlp": 0.01040572, + "balance_loss_clip": 1.05835128, + "balance_loss_mlp": 1.03056443, + "epoch": 0.2130704022124692, + "flos": 18806993308800.0, + "grad_norm": 2.0308057012380853, + "language_loss": 0.80317801, + "learning_rate": 3.6587621047605833e-06, + "loss": 0.82566428, + "num_input_tokens_seen": 37697910, + "step": 1772, + "time_per_iteration": 2.721759080886841 + }, + { + "auxiliary_loss_clip": 0.01203082, + "auxiliary_loss_mlp": 0.01043165, + "balance_loss_clip": 1.057621, + "balance_loss_mlp": 1.03231657, + "epoch": 0.21319064510310828, + "flos": 13918150759680.0, + "grad_norm": 1.9301096584346793, + "language_loss": 0.8657161, + "learning_rate": 3.6583267804765542e-06, + "loss": 0.88817853, + "num_input_tokens_seen": 37712245, + "step": 1773, + "time_per_iteration": 2.8058183193206787 + }, + { + "auxiliary_loss_clip": 0.01203816, + "auxiliary_loss_mlp": 0.01028211, + "balance_loss_clip": 1.05675888, + "balance_loss_mlp": 1.01665378, + "epoch": 0.21331088799374737, + "flos": 20959694277120.0, + "grad_norm": 1.8759118594941269, + "language_loss": 0.85855687, + "learning_rate": 3.6578912046256702e-06, + "loss": 0.88087708, + "num_input_tokens_seen": 37730765, + "step": 1774, + "time_per_iteration": 2.7892093658447266 + }, + { + "auxiliary_loss_clip": 0.01172318, + "auxiliary_loss_mlp": 0.01038623, + "balance_loss_clip": 1.04733717, + "balance_loss_mlp": 1.0266664, + "epoch": 0.21343113088438645, + "flos": 18624638937600.0, + "grad_norm": 2.3017381653995415, + "language_loss": 0.76468617, + "learning_rate": 3.6574553772740083e-06, + "loss": 0.78679562, + "num_input_tokens_seen": 37748695, + "step": 1775, + "time_per_iteration": 2.891944169998169 + }, + { + "auxiliary_loss_clip": 0.0111319, + "auxiliary_loss_mlp": 0.01009028, + "balance_loss_clip": 1.04135942, + "balance_loss_mlp": 1.00511789, + "epoch": 0.21355137377502556, + "flos": 67413128791680.0, + "grad_norm": 0.8476689691623603, + "language_loss": 0.61869138, + "learning_rate": 3.657019298487684e-06, + "loss": 0.63991356, + "num_input_tokens_seen": 37813705, + "step": 1776, + "time_per_iteration": 3.344867706298828 + }, + { + "auxiliary_loss_clip": 0.01209393, + "auxiliary_loss_mlp": 0.00888552, + "balance_loss_clip": 1.05680394, + "balance_loss_mlp": 1.00039339, + "epoch": 0.21367161666566464, + "flos": 34532095697280.0, + "grad_norm": 1.729129030417503, + "language_loss": 0.83305931, + "learning_rate": 3.6565829683328495e-06, + "loss": 0.85403877, + "num_input_tokens_seen": 37836330, + "step": 1777, + "time_per_iteration": 2.965576648712158 + }, + { + "auxiliary_loss_clip": 0.0119746, + "auxiliary_loss_mlp": 0.01032911, + "balance_loss_clip": 1.05718374, + "balance_loss_mlp": 1.02172291, + "epoch": 0.21379185955630373, + "flos": 18989347680000.0, + "grad_norm": 2.266729341360853, + "language_loss": 0.85572183, + "learning_rate": 3.6561463868756965e-06, + "loss": 0.87802553, + "num_input_tokens_seen": 37855030, + "step": 1778, + "time_per_iteration": 2.7483153343200684 + }, + { + "auxiliary_loss_clip": 0.01202047, + "auxiliary_loss_mlp": 0.01040367, + "balance_loss_clip": 1.05680752, + "balance_loss_mlp": 1.0290122, + "epoch": 0.21391210244694284, + "flos": 28218497207040.0, + "grad_norm": 1.5150062714107624, + "language_loss": 0.781322, + "learning_rate": 3.655709554182452e-06, + "loss": 0.8037461, + "num_input_tokens_seen": 37875370, + "step": 1779, + "time_per_iteration": 2.7653145790100098 + }, + { + "auxiliary_loss_clip": 0.01203402, + "auxiliary_loss_mlp": 0.01038248, + "balance_loss_clip": 1.05442882, + "balance_loss_mlp": 1.02628505, + "epoch": 0.21403234533758192, + "flos": 17455064192640.0, + "grad_norm": 1.9220939443142901, + "language_loss": 0.84495407, + "learning_rate": 3.6552724703193855e-06, + "loss": 0.86737055, + "num_input_tokens_seen": 37892560, + "step": 1780, + "time_per_iteration": 2.6984591484069824 + }, + { + "auxiliary_loss_clip": 0.01078141, + "auxiliary_loss_mlp": 0.01004326, + "balance_loss_clip": 1.02155733, + "balance_loss_mlp": 0.99993926, + "epoch": 0.214152588228221, + "flos": 51637606686720.0, + "grad_norm": 0.783437124201437, + "language_loss": 0.55889857, + "learning_rate": 3.654835135352801e-06, + "loss": 0.57972324, + "num_input_tokens_seen": 37947370, + "step": 1781, + "time_per_iteration": 3.248147487640381 + }, + { + "auxiliary_loss_clip": 0.01170016, + "auxiliary_loss_mlp": 0.01036139, + "balance_loss_clip": 1.04650652, + "balance_loss_mlp": 1.02411723, + "epoch": 0.21427283111886009, + "flos": 19496154625920.0, + "grad_norm": 1.9144160397338799, + "language_loss": 0.8718276, + "learning_rate": 3.654397549349043e-06, + "loss": 0.89388913, + "num_input_tokens_seen": 37964745, + "step": 1782, + "time_per_iteration": 2.9088029861450195 + }, + { + "auxiliary_loss_clip": 0.01188801, + "auxiliary_loss_mlp": 0.01034969, + "balance_loss_clip": 1.05561018, + "balance_loss_mlp": 1.0238589, + "epoch": 0.2143930740094992, + "flos": 20084802710400.0, + "grad_norm": 2.1547949084864273, + "language_loss": 0.75241733, + "learning_rate": 3.653959712374491e-06, + "loss": 0.77465504, + "num_input_tokens_seen": 37982850, + "step": 1783, + "time_per_iteration": 2.803536891937256 + }, + { + "auxiliary_loss_clip": 0.01178036, + "auxiliary_loss_mlp": 0.01037266, + "balance_loss_clip": 1.05821681, + "balance_loss_mlp": 1.02596486, + "epoch": 0.21451331690013828, + "flos": 21798603394560.0, + "grad_norm": 1.7330015207769278, + "language_loss": 0.82649493, + "learning_rate": 3.6535216244955663e-06, + "loss": 0.84864789, + "num_input_tokens_seen": 38002745, + "step": 1784, + "time_per_iteration": 2.8289852142333984 + }, + { + "auxiliary_loss_clip": 0.01189409, + "auxiliary_loss_mlp": 0.01033909, + "balance_loss_clip": 1.05381966, + "balance_loss_mlp": 1.02300096, + "epoch": 0.21463355979077736, + "flos": 32853882412800.0, + "grad_norm": 1.8181584007565355, + "language_loss": 0.71047813, + "learning_rate": 3.653083285778726e-06, + "loss": 0.73271132, + "num_input_tokens_seen": 38024115, + "step": 1785, + "time_per_iteration": 4.79051661491394 + }, + { + "auxiliary_loss_clip": 0.01205822, + "auxiliary_loss_mlp": 0.01034957, + "balance_loss_clip": 1.05684924, + "balance_loss_mlp": 1.02298856, + "epoch": 0.21475380268141647, + "flos": 21543817248000.0, + "grad_norm": 2.1922401852275115, + "language_loss": 0.81298691, + "learning_rate": 3.6526446962904653e-06, + "loss": 0.83539474, + "num_input_tokens_seen": 38042830, + "step": 1786, + "time_per_iteration": 3.641723155975342 + }, + { + "auxiliary_loss_clip": 0.01202375, + "auxiliary_loss_mlp": 0.01037318, + "balance_loss_clip": 1.05944109, + "balance_loss_mlp": 1.02645791, + "epoch": 0.21487404557205556, + "flos": 32159082660480.0, + "grad_norm": 1.4590753624942148, + "language_loss": 0.74228942, + "learning_rate": 3.652205856097318e-06, + "loss": 0.76468629, + "num_input_tokens_seen": 38066015, + "step": 1787, + "time_per_iteration": 2.7950174808502197 + }, + { + "auxiliary_loss_clip": 0.01193311, + "auxiliary_loss_mlp": 0.00888203, + "balance_loss_clip": 1.0544076, + "balance_loss_mlp": 1.00051427, + "epoch": 0.21499428846269464, + "flos": 12673091583360.0, + "grad_norm": 1.944901448037603, + "language_loss": 0.79284245, + "learning_rate": 3.651766765265856e-06, + "loss": 0.81365758, + "num_input_tokens_seen": 38083025, + "step": 1788, + "time_per_iteration": 2.9336729049682617 + }, + { + "auxiliary_loss_clip": 0.011863, + "auxiliary_loss_mlp": 0.01042103, + "balance_loss_clip": 1.05300748, + "balance_loss_mlp": 1.03102291, + "epoch": 0.21511453135333372, + "flos": 23471573293440.0, + "grad_norm": 3.0006779942841737, + "language_loss": 0.81311232, + "learning_rate": 3.65132742386269e-06, + "loss": 0.83539635, + "num_input_tokens_seen": 38098245, + "step": 1789, + "time_per_iteration": 2.848037004470825 + }, + { + "auxiliary_loss_clip": 0.01212198, + "auxiliary_loss_mlp": 0.01033201, + "balance_loss_clip": 1.05750048, + "balance_loss_mlp": 1.02244258, + "epoch": 0.21523477424397283, + "flos": 26943560893440.0, + "grad_norm": 1.8358233469494996, + "language_loss": 0.84599888, + "learning_rate": 3.6508878319544656e-06, + "loss": 0.86845291, + "num_input_tokens_seen": 38118460, + "step": 1790, + "time_per_iteration": 2.716984272003174 + }, + { + "auxiliary_loss_clip": 0.0118576, + "auxiliary_loss_mlp": 0.01038313, + "balance_loss_clip": 1.05564904, + "balance_loss_mlp": 1.02691686, + "epoch": 0.21535501713461191, + "flos": 18916161719040.0, + "grad_norm": 2.4322495229339394, + "language_loss": 0.81462765, + "learning_rate": 3.65044798960787e-06, + "loss": 0.83686835, + "num_input_tokens_seen": 38136800, + "step": 1791, + "time_per_iteration": 2.7711195945739746 + }, + { + "auxiliary_loss_clip": 0.01175839, + "auxiliary_loss_mlp": 0.01039427, + "balance_loss_clip": 1.05228806, + "balance_loss_mlp": 1.02808464, + "epoch": 0.215475260025251, + "flos": 17895113712000.0, + "grad_norm": 1.7771012951849758, + "language_loss": 0.78194618, + "learning_rate": 3.650007896889627e-06, + "loss": 0.80409884, + "num_input_tokens_seen": 38155380, + "step": 1792, + "time_per_iteration": 2.804373264312744 + }, + { + "auxiliary_loss_clip": 0.01212148, + "auxiliary_loss_mlp": 0.01036484, + "balance_loss_clip": 1.05824518, + "balance_loss_mlp": 1.02455091, + "epoch": 0.2155955029158901, + "flos": 16654292340480.0, + "grad_norm": 1.7625238522158073, + "language_loss": 0.80685687, + "learning_rate": 3.6495675538664974e-06, + "loss": 0.82934308, + "num_input_tokens_seen": 38174395, + "step": 1793, + "time_per_iteration": 2.7294328212738037 + }, + { + "auxiliary_loss_clip": 0.0119384, + "auxiliary_loss_mlp": 0.01036463, + "balance_loss_clip": 1.05371821, + "balance_loss_mlp": 1.02498877, + "epoch": 0.2157157458065292, + "flos": 23621213352960.0, + "grad_norm": 1.6822166736362527, + "language_loss": 0.82267809, + "learning_rate": 3.649126960605282e-06, + "loss": 0.84498119, + "num_input_tokens_seen": 38195380, + "step": 1794, + "time_per_iteration": 2.7412750720977783 + }, + { + "auxiliary_loss_clip": 0.01187601, + "auxiliary_loss_mlp": 0.01037039, + "balance_loss_clip": 1.05599904, + "balance_loss_mlp": 1.02567852, + "epoch": 0.21583598869716827, + "flos": 22127078292480.0, + "grad_norm": 2.32204857810271, + "language_loss": 0.83508193, + "learning_rate": 3.6486861171728174e-06, + "loss": 0.8573283, + "num_input_tokens_seen": 38213775, + "step": 1795, + "time_per_iteration": 2.8439226150512695 + }, + { + "auxiliary_loss_clip": 0.01184111, + "auxiliary_loss_mlp": 0.01039507, + "balance_loss_clip": 1.04974329, + "balance_loss_mlp": 1.02774692, + "epoch": 0.21595623158780738, + "flos": 23441229279360.0, + "grad_norm": 1.5931123927526836, + "language_loss": 0.78436279, + "learning_rate": 3.6482450236359803e-06, + "loss": 0.80659902, + "num_input_tokens_seen": 38235630, + "step": 1796, + "time_per_iteration": 2.8113317489624023 + }, + { + "auxiliary_loss_clip": 0.01199569, + "auxiliary_loss_mlp": 0.01039071, + "balance_loss_clip": 1.05616808, + "balance_loss_mlp": 1.02861595, + "epoch": 0.21607647447844647, + "flos": 26906501036160.0, + "grad_norm": 2.478167059122672, + "language_loss": 0.78038168, + "learning_rate": 3.647803680061683e-06, + "loss": 0.80276811, + "num_input_tokens_seen": 38256045, + "step": 1797, + "time_per_iteration": 2.72345232963562 + }, + { + "auxiliary_loss_clip": 0.01194992, + "auxiliary_loss_mlp": 0.01035896, + "balance_loss_clip": 1.05522954, + "balance_loss_mlp": 1.02395165, + "epoch": 0.21619671736908555, + "flos": 14495378319360.0, + "grad_norm": 2.4717611179304186, + "language_loss": 0.74687111, + "learning_rate": 3.6473620865168776e-06, + "loss": 0.76917994, + "num_input_tokens_seen": 38272915, + "step": 1798, + "time_per_iteration": 2.7343218326568604 + }, + { + "auxiliary_loss_clip": 0.01193223, + "auxiliary_loss_mlp": 0.01039055, + "balance_loss_clip": 1.05734658, + "balance_loss_mlp": 1.02812386, + "epoch": 0.21631696025972463, + "flos": 17931096161280.0, + "grad_norm": 1.837909051186264, + "language_loss": 0.81821078, + "learning_rate": 3.646920243068554e-06, + "loss": 0.84053361, + "num_input_tokens_seen": 38290810, + "step": 1799, + "time_per_iteration": 2.7352569103240967 + }, + { + "auxiliary_loss_clip": 0.01183739, + "auxiliary_loss_mlp": 0.01041269, + "balance_loss_clip": 1.05387831, + "balance_loss_mlp": 1.02912712, + "epoch": 0.21643720315036374, + "flos": 24462385027200.0, + "grad_norm": 1.5613051427549074, + "language_loss": 0.74428934, + "learning_rate": 3.6464781497837384e-06, + "loss": 0.76653945, + "num_input_tokens_seen": 38312785, + "step": 1800, + "time_per_iteration": 2.865537643432617 + }, + { + "auxiliary_loss_clip": 0.01196368, + "auxiliary_loss_mlp": 0.01039493, + "balance_loss_clip": 1.05434012, + "balance_loss_mlp": 1.02773356, + "epoch": 0.21655744604100283, + "flos": 28474432588800.0, + "grad_norm": 1.7588073321630349, + "language_loss": 0.72904533, + "learning_rate": 3.6460358067294965e-06, + "loss": 0.75140393, + "num_input_tokens_seen": 38334015, + "step": 1801, + "time_per_iteration": 2.810230255126953 + }, + { + "auxiliary_loss_clip": 0.01214149, + "auxiliary_loss_mlp": 0.01033464, + "balance_loss_clip": 1.0565058, + "balance_loss_mlp": 1.02187729, + "epoch": 0.2166776889316419, + "flos": 20152960767360.0, + "grad_norm": 2.3767080048950016, + "language_loss": 0.77841222, + "learning_rate": 3.645593213972932e-06, + "loss": 0.8008883, + "num_input_tokens_seen": 38352920, + "step": 1802, + "time_per_iteration": 2.7406444549560547 + }, + { + "auxiliary_loss_clip": 0.01196323, + "auxiliary_loss_mlp": 0.01039313, + "balance_loss_clip": 1.05534387, + "balance_loss_mlp": 1.02822018, + "epoch": 0.21679793182228102, + "flos": 15193482122880.0, + "grad_norm": 2.851956014487186, + "language_loss": 0.79814816, + "learning_rate": 3.6451503715811852e-06, + "loss": 0.82050449, + "num_input_tokens_seen": 38371230, + "step": 1803, + "time_per_iteration": 2.682096004486084 + }, + { + "auxiliary_loss_clip": 0.0119146, + "auxiliary_loss_mlp": 0.01033572, + "balance_loss_clip": 1.0571295, + "balance_loss_mlp": 1.02296841, + "epoch": 0.2169181747129201, + "flos": 17384464010880.0, + "grad_norm": 2.476196633604181, + "language_loss": 0.80243355, + "learning_rate": 3.6447072796214345e-06, + "loss": 0.8246839, + "num_input_tokens_seen": 38389795, + "step": 1804, + "time_per_iteration": 2.807445764541626 + }, + { + "auxiliary_loss_clip": 0.01075344, + "auxiliary_loss_mlp": 0.0100487, + "balance_loss_clip": 1.02025151, + "balance_loss_mlp": 1.0007453, + "epoch": 0.21703841760355919, + "flos": 58760955429120.0, + "grad_norm": 0.9398555539375524, + "language_loss": 0.63291728, + "learning_rate": 3.644263938160898e-06, + "loss": 0.65371943, + "num_input_tokens_seen": 38445760, + "step": 1805, + "time_per_iteration": 3.3327934741973877 + }, + { + "auxiliary_loss_clip": 0.01180315, + "auxiliary_loss_mlp": 0.01037884, + "balance_loss_clip": 1.05343056, + "balance_loss_mlp": 1.02645195, + "epoch": 0.21715866049419827, + "flos": 22418457419520.0, + "grad_norm": 1.9448443239015052, + "language_loss": 0.72071314, + "learning_rate": 3.6438203472668293e-06, + "loss": 0.74289513, + "num_input_tokens_seen": 38465405, + "step": 1806, + "time_per_iteration": 2.8218913078308105 + }, + { + "auxiliary_loss_clip": 0.01194744, + "auxiliary_loss_mlp": 0.01037539, + "balance_loss_clip": 1.05562294, + "balance_loss_mlp": 1.02739453, + "epoch": 0.21727890338483738, + "flos": 17237732952960.0, + "grad_norm": 2.1003221826066603, + "language_loss": 0.81885755, + "learning_rate": 3.6433765070065206e-06, + "loss": 0.84118032, + "num_input_tokens_seen": 38483195, + "step": 1807, + "time_per_iteration": 2.7597601413726807 + }, + { + "auxiliary_loss_clip": 0.01214466, + "auxiliary_loss_mlp": 0.01041681, + "balance_loss_clip": 1.05732942, + "balance_loss_mlp": 1.02958155, + "epoch": 0.21739914627547646, + "flos": 13434792416640.0, + "grad_norm": 2.9920264086421127, + "language_loss": 0.87485051, + "learning_rate": 3.6429324174473025e-06, + "loss": 0.89741194, + "num_input_tokens_seen": 38496735, + "step": 1808, + "time_per_iteration": 2.652397632598877 + }, + { + "auxiliary_loss_clip": 0.01204501, + "auxiliary_loss_mlp": 0.01029918, + "balance_loss_clip": 1.05632639, + "balance_loss_mlp": 1.01889157, + "epoch": 0.21751938916611555, + "flos": 20959514709120.0, + "grad_norm": 2.194330191743805, + "language_loss": 0.85075259, + "learning_rate": 3.6424880786565425e-06, + "loss": 0.87309682, + "num_input_tokens_seen": 38512880, + "step": 1809, + "time_per_iteration": 2.689438581466675 + }, + { + "auxiliary_loss_clip": 0.0117076, + "auxiliary_loss_mlp": 0.01036073, + "balance_loss_clip": 1.05422211, + "balance_loss_mlp": 1.0243187, + "epoch": 0.21763963205675466, + "flos": 27599936071680.0, + "grad_norm": 2.1313861387737396, + "language_loss": 0.79881418, + "learning_rate": 3.6420434907016482e-06, + "loss": 0.8208825, + "num_input_tokens_seen": 38532570, + "step": 1810, + "time_per_iteration": 2.8347432613372803 + }, + { + "auxiliary_loss_clip": 0.01204017, + "auxiliary_loss_mlp": 0.01036231, + "balance_loss_clip": 1.05853736, + "balance_loss_mlp": 1.02523446, + "epoch": 0.21775987494739374, + "flos": 21430411032960.0, + "grad_norm": 1.6092146380177164, + "language_loss": 0.81158054, + "learning_rate": 3.6415986536500606e-06, + "loss": 0.83398306, + "num_input_tokens_seen": 38550900, + "step": 1811, + "time_per_iteration": 4.589010953903198 + }, + { + "auxiliary_loss_clip": 0.0116387, + "auxiliary_loss_mlp": 0.01032716, + "balance_loss_clip": 1.05465937, + "balance_loss_mlp": 1.02162361, + "epoch": 0.21788011783803282, + "flos": 18332972501760.0, + "grad_norm": 1.75211778868207, + "language_loss": 0.8036797, + "learning_rate": 3.641153567569263e-06, + "loss": 0.82564557, + "num_input_tokens_seen": 38569215, + "step": 1812, + "time_per_iteration": 4.752429485321045 + }, + { + "auxiliary_loss_clip": 0.01196831, + "auxiliary_loss_mlp": 0.01030028, + "balance_loss_clip": 1.05557883, + "balance_loss_mlp": 1.01885772, + "epoch": 0.2180003607286719, + "flos": 30262748037120.0, + "grad_norm": 2.4455872571051716, + "language_loss": 0.95288956, + "learning_rate": 3.640708232526774e-06, + "loss": 0.97515821, + "num_input_tokens_seen": 38587870, + "step": 1813, + "time_per_iteration": 2.7589006423950195 + }, + { + "auxiliary_loss_clip": 0.01156522, + "auxiliary_loss_mlp": 0.01043614, + "balance_loss_clip": 1.044559, + "balance_loss_mlp": 1.03227711, + "epoch": 0.21812060361931102, + "flos": 25480272637440.0, + "grad_norm": 1.9012956899831193, + "language_loss": 0.78723192, + "learning_rate": 3.6402626485901504e-06, + "loss": 0.80923325, + "num_input_tokens_seen": 38606965, + "step": 1814, + "time_per_iteration": 2.899986982345581 + }, + { + "auxiliary_loss_clip": 0.01197967, + "auxiliary_loss_mlp": 0.01033567, + "balance_loss_clip": 1.05817962, + "balance_loss_mlp": 1.02294517, + "epoch": 0.2182408465099501, + "flos": 21908166854400.0, + "grad_norm": 2.2678814454288654, + "language_loss": 0.78082132, + "learning_rate": 3.639816815826988e-06, + "loss": 0.80313671, + "num_input_tokens_seen": 38626290, + "step": 1815, + "time_per_iteration": 2.84028697013855 + }, + { + "auxiliary_loss_clip": 0.01191556, + "auxiliary_loss_mlp": 0.01033713, + "balance_loss_clip": 1.05526173, + "balance_loss_mlp": 1.0236938, + "epoch": 0.21836108940058918, + "flos": 23657339456640.0, + "grad_norm": 1.8616541426962134, + "language_loss": 0.78139842, + "learning_rate": 3.6393707343049176e-06, + "loss": 0.80365109, + "num_input_tokens_seen": 38646620, + "step": 1816, + "time_per_iteration": 2.736565351486206 + }, + { + "auxiliary_loss_clip": 0.01205652, + "auxiliary_loss_mlp": 0.01031611, + "balance_loss_clip": 1.05615938, + "balance_loss_mlp": 1.02133536, + "epoch": 0.2184813322912283, + "flos": 24681009156480.0, + "grad_norm": 2.3512362795665105, + "language_loss": 0.73300314, + "learning_rate": 3.6389244040916104e-06, + "loss": 0.7553758, + "num_input_tokens_seen": 38665695, + "step": 1817, + "time_per_iteration": 2.7392048835754395 + }, + { + "auxiliary_loss_clip": 0.01179432, + "auxiliary_loss_mlp": 0.00888, + "balance_loss_clip": 1.05314255, + "balance_loss_mlp": 1.00041366, + "epoch": 0.21860157518186737, + "flos": 26574650259840.0, + "grad_norm": 1.966476792805467, + "language_loss": 0.79113704, + "learning_rate": 3.6384778252547747e-06, + "loss": 0.81181133, + "num_input_tokens_seen": 38681575, + "step": 1818, + "time_per_iteration": 2.764925003051758 + }, + { + "auxiliary_loss_clip": 0.01192445, + "auxiliary_loss_mlp": 0.00888255, + "balance_loss_clip": 1.06004906, + "balance_loss_mlp": 1.00052428, + "epoch": 0.21872181807250646, + "flos": 20886292834560.0, + "grad_norm": 2.31155856576812, + "language_loss": 0.78430516, + "learning_rate": 3.638030997862155e-06, + "loss": 0.80511212, + "num_input_tokens_seen": 38700510, + "step": 1819, + "time_per_iteration": 2.714677095413208 + }, + { + "auxiliary_loss_clip": 0.01082179, + "auxiliary_loss_mlp": 0.01003421, + "balance_loss_clip": 1.01919055, + "balance_loss_mlp": 0.99951047, + "epoch": 0.21884206096314554, + "flos": 61209452897280.0, + "grad_norm": 0.759674480453687, + "language_loss": 0.59429103, + "learning_rate": 3.6375839219815356e-06, + "loss": 0.61514699, + "num_input_tokens_seen": 38758310, + "step": 1820, + "time_per_iteration": 3.2191097736358643 + }, + { + "auxiliary_loss_clip": 0.01211767, + "auxiliary_loss_mlp": 0.01036114, + "balance_loss_clip": 1.05790913, + "balance_loss_mlp": 1.02531385, + "epoch": 0.21896230385378465, + "flos": 23473835850240.0, + "grad_norm": 2.033559964165276, + "language_loss": 0.82950675, + "learning_rate": 3.6371365976807375e-06, + "loss": 0.85198557, + "num_input_tokens_seen": 38778705, + "step": 1821, + "time_per_iteration": 2.719204902648926 + }, + { + "auxiliary_loss_clip": 0.01164763, + "auxiliary_loss_mlp": 0.01034525, + "balance_loss_clip": 1.05129623, + "balance_loss_mlp": 1.02297997, + "epoch": 0.21908254674442373, + "flos": 25081915829760.0, + "grad_norm": 1.7317276756823297, + "language_loss": 0.83411241, + "learning_rate": 3.6366890250276185e-06, + "loss": 0.85610527, + "num_input_tokens_seen": 38799660, + "step": 1822, + "time_per_iteration": 2.9469704627990723 + }, + { + "auxiliary_loss_clip": 0.0121123, + "auxiliary_loss_mlp": 0.01033368, + "balance_loss_clip": 1.05824172, + "balance_loss_mlp": 1.02251434, + "epoch": 0.21920278963506282, + "flos": 23513768795520.0, + "grad_norm": 1.8440680938763043, + "language_loss": 0.90112448, + "learning_rate": 3.6362412040900764e-06, + "loss": 0.92357051, + "num_input_tokens_seen": 38819450, + "step": 1823, + "time_per_iteration": 2.789201021194458 + }, + { + "auxiliary_loss_clip": 0.01202597, + "auxiliary_loss_mlp": 0.01040413, + "balance_loss_clip": 1.05392754, + "balance_loss_mlp": 1.02904058, + "epoch": 0.21932303252570193, + "flos": 29242238734080.0, + "grad_norm": 1.9412888627577436, + "language_loss": 0.80572343, + "learning_rate": 3.635793134936044e-06, + "loss": 0.82815349, + "num_input_tokens_seen": 38840460, + "step": 1824, + "time_per_iteration": 2.8008391857147217 + }, + { + "auxiliary_loss_clip": 0.01199024, + "auxiliary_loss_mlp": 0.0103257, + "balance_loss_clip": 1.05547404, + "balance_loss_mlp": 1.02084601, + "epoch": 0.219443275416341, + "flos": 20806857907200.0, + "grad_norm": 1.950477522073272, + "language_loss": 0.7328282, + "learning_rate": 3.635344817633494e-06, + "loss": 0.75514412, + "num_input_tokens_seen": 38859775, + "step": 1825, + "time_per_iteration": 2.692525863647461 + }, + { + "auxiliary_loss_clip": 0.01195386, + "auxiliary_loss_mlp": 0.01032689, + "balance_loss_clip": 1.05405641, + "balance_loss_mlp": 1.02195966, + "epoch": 0.2195635183069801, + "flos": 14501555458560.0, + "grad_norm": 2.1222550233948283, + "language_loss": 0.75431871, + "learning_rate": 3.634896252250436e-06, + "loss": 0.77659941, + "num_input_tokens_seen": 38876540, + "step": 1826, + "time_per_iteration": 2.7133142948150635 + }, + { + "auxiliary_loss_clip": 0.01216218, + "auxiliary_loss_mlp": 0.01041115, + "balance_loss_clip": 1.06113982, + "balance_loss_mlp": 1.03046393, + "epoch": 0.2196837611976192, + "flos": 24243473589120.0, + "grad_norm": 1.77091886631815, + "language_loss": 0.82529163, + "learning_rate": 3.6344474388549157e-06, + "loss": 0.84786499, + "num_input_tokens_seen": 38896195, + "step": 1827, + "time_per_iteration": 2.74601674079895 + }, + { + "auxiliary_loss_clip": 0.01202422, + "auxiliary_loss_mlp": 0.01044192, + "balance_loss_clip": 1.05727625, + "balance_loss_mlp": 1.03297448, + "epoch": 0.2198040040882583, + "flos": 18074523168000.0, + "grad_norm": 2.0051775056797982, + "language_loss": 0.80756497, + "learning_rate": 3.6339983775150183e-06, + "loss": 0.8300311, + "num_input_tokens_seen": 38912755, + "step": 1828, + "time_per_iteration": 2.6798245906829834 + }, + { + "auxiliary_loss_clip": 0.01201321, + "auxiliary_loss_mlp": 0.01037806, + "balance_loss_clip": 1.05735588, + "balance_loss_mlp": 1.02642703, + "epoch": 0.21992424697889737, + "flos": 17784185535360.0, + "grad_norm": 2.5531761965367976, + "language_loss": 0.84359932, + "learning_rate": 3.6335490682988664e-06, + "loss": 0.86599058, + "num_input_tokens_seen": 38928365, + "step": 1829, + "time_per_iteration": 2.779924154281616 + }, + { + "auxiliary_loss_clip": 0.01155123, + "auxiliary_loss_mlp": 0.01039664, + "balance_loss_clip": 1.04838991, + "balance_loss_mlp": 1.02755249, + "epoch": 0.22004448986953645, + "flos": 17638495971840.0, + "grad_norm": 2.401673100557863, + "language_loss": 0.82985187, + "learning_rate": 3.63309951127462e-06, + "loss": 0.85179979, + "num_input_tokens_seen": 38945275, + "step": 1830, + "time_per_iteration": 2.8491146564483643 + }, + { + "auxiliary_loss_clip": 0.0118458, + "auxiliary_loss_mlp": 0.01040293, + "balance_loss_clip": 1.05561817, + "balance_loss_mlp": 1.02893817, + "epoch": 0.22016473276017556, + "flos": 22275533203200.0, + "grad_norm": 1.9481265342389071, + "language_loss": 0.74938905, + "learning_rate": 3.6326497065104757e-06, + "loss": 0.77163774, + "num_input_tokens_seen": 38965740, + "step": 1831, + "time_per_iteration": 2.8748090267181396 + }, + { + "auxiliary_loss_clip": 0.01208123, + "auxiliary_loss_mlp": 0.01039954, + "balance_loss_clip": 1.05714417, + "balance_loss_mlp": 1.02896309, + "epoch": 0.22028497565081465, + "flos": 25556259859200.0, + "grad_norm": 2.0060585280978134, + "language_loss": 0.78415942, + "learning_rate": 3.6321996540746697e-06, + "loss": 0.80664021, + "num_input_tokens_seen": 38984815, + "step": 1832, + "time_per_iteration": 2.721464157104492 + }, + { + "auxiliary_loss_clip": 0.01181808, + "auxiliary_loss_mlp": 0.01038, + "balance_loss_clip": 1.05482543, + "balance_loss_mlp": 1.02668667, + "epoch": 0.22040521854145373, + "flos": 36247332925440.0, + "grad_norm": 1.6476325835622045, + "language_loss": 0.80713373, + "learning_rate": 3.6317493540354733e-06, + "loss": 0.82933187, + "num_input_tokens_seen": 39008230, + "step": 1833, + "time_per_iteration": 2.878600597381592 + }, + { + "auxiliary_loss_clip": 0.01196278, + "auxiliary_loss_mlp": 0.01044851, + "balance_loss_clip": 1.05327654, + "balance_loss_mlp": 1.03266227, + "epoch": 0.22052546143209284, + "flos": 11838420270720.0, + "grad_norm": 2.0443016531785103, + "language_loss": 0.76484191, + "learning_rate": 3.6312988064611976e-06, + "loss": 0.78725326, + "num_input_tokens_seen": 39026540, + "step": 1834, + "time_per_iteration": 2.6552038192749023 + }, + { + "auxiliary_loss_clip": 0.01185483, + "auxiliary_loss_mlp": 0.01035853, + "balance_loss_clip": 1.05136204, + "balance_loss_mlp": 1.02473104, + "epoch": 0.22064570432273192, + "flos": 24209250906240.0, + "grad_norm": 1.7552862457190368, + "language_loss": 0.81244856, + "learning_rate": 3.6308480114201896e-06, + "loss": 0.83466196, + "num_input_tokens_seen": 39048460, + "step": 1835, + "time_per_iteration": 2.930180311203003 + }, + { + "auxiliary_loss_clip": 0.0121658, + "auxiliary_loss_mlp": 0.01042891, + "balance_loss_clip": 1.06137943, + "balance_loss_mlp": 1.03152978, + "epoch": 0.220765947213371, + "flos": 17931347556480.0, + "grad_norm": 2.0941220923283637, + "language_loss": 0.76456547, + "learning_rate": 3.630396968980835e-06, + "loss": 0.78716016, + "num_input_tokens_seen": 39066335, + "step": 1836, + "time_per_iteration": 2.778707265853882 + }, + { + "auxiliary_loss_clip": 0.0119544, + "auxiliary_loss_mlp": 0.01037305, + "balance_loss_clip": 1.05495596, + "balance_loss_mlp": 1.02602232, + "epoch": 0.2208861901040101, + "flos": 26757040544640.0, + "grad_norm": 3.7328676160478103, + "language_loss": 0.83904552, + "learning_rate": 3.6299456792115575e-06, + "loss": 0.86137301, + "num_input_tokens_seen": 39087590, + "step": 1837, + "time_per_iteration": 5.59220552444458 + }, + { + "auxiliary_loss_clip": 0.01133546, + "auxiliary_loss_mlp": 0.01038268, + "balance_loss_clip": 1.04159069, + "balance_loss_mlp": 1.027843, + "epoch": 0.2210064329946492, + "flos": 17817977255040.0, + "grad_norm": 2.0067827209426436, + "language_loss": 0.80897468, + "learning_rate": 3.629494142180815e-06, + "loss": 0.83069283, + "num_input_tokens_seen": 39106335, + "step": 1838, + "time_per_iteration": 3.9038686752319336 + }, + { + "auxiliary_loss_clip": 0.01212776, + "auxiliary_loss_mlp": 0.01044962, + "balance_loss_clip": 1.0595063, + "balance_loss_mlp": 1.0342685, + "epoch": 0.22112667588528828, + "flos": 17967401832960.0, + "grad_norm": 2.211671122634858, + "language_loss": 0.8486467, + "learning_rate": 3.6290423579571075e-06, + "loss": 0.87122405, + "num_input_tokens_seen": 39122875, + "step": 1839, + "time_per_iteration": 2.6980297565460205 + }, + { + "auxiliary_loss_clip": 0.01202514, + "auxiliary_loss_mlp": 0.01039551, + "balance_loss_clip": 1.06059051, + "balance_loss_mlp": 1.02761197, + "epoch": 0.22124691877592736, + "flos": 18369206346240.0, + "grad_norm": 1.6358313727257972, + "language_loss": 0.80278778, + "learning_rate": 3.6285903266089694e-06, + "loss": 0.82520849, + "num_input_tokens_seen": 39142150, + "step": 1840, + "time_per_iteration": 2.863635778427124 + }, + { + "auxiliary_loss_clip": 0.01193918, + "auxiliary_loss_mlp": 0.01046451, + "balance_loss_clip": 1.05747378, + "balance_loss_mlp": 1.03520989, + "epoch": 0.22136716166656648, + "flos": 20813286441600.0, + "grad_norm": 1.7916771120895651, + "language_loss": 0.77339929, + "learning_rate": 3.628138048204974e-06, + "loss": 0.79580307, + "num_input_tokens_seen": 39162835, + "step": 1841, + "time_per_iteration": 2.772670269012451 + }, + { + "auxiliary_loss_clip": 0.01166167, + "auxiliary_loss_mlp": 0.01041885, + "balance_loss_clip": 1.05287552, + "balance_loss_mlp": 1.02965391, + "epoch": 0.22148740455720556, + "flos": 17675699483520.0, + "grad_norm": 2.042980310830267, + "language_loss": 0.76317251, + "learning_rate": 3.6276855228137304e-06, + "loss": 0.78525305, + "num_input_tokens_seen": 39181040, + "step": 1842, + "time_per_iteration": 2.894709348678589 + }, + { + "auxiliary_loss_clip": 0.01214702, + "auxiliary_loss_mlp": 0.00888367, + "balance_loss_clip": 1.05961668, + "balance_loss_mlp": 1.00037277, + "epoch": 0.22160764744784464, + "flos": 21726710323200.0, + "grad_norm": 2.0336761810084427, + "language_loss": 0.82588291, + "learning_rate": 3.6272327505038874e-06, + "loss": 0.84691358, + "num_input_tokens_seen": 39197505, + "step": 1843, + "time_per_iteration": 2.7057082653045654 + }, + { + "auxiliary_loss_clip": 0.01177784, + "auxiliary_loss_mlp": 0.01030056, + "balance_loss_clip": 1.05277824, + "balance_loss_mlp": 1.01936257, + "epoch": 0.22172789033848372, + "flos": 23764712186880.0, + "grad_norm": 1.9044309161927735, + "language_loss": 0.78529024, + "learning_rate": 3.626779731344131e-06, + "loss": 0.8073687, + "num_input_tokens_seen": 39217295, + "step": 1844, + "time_per_iteration": 2.9607956409454346 + }, + { + "auxiliary_loss_clip": 0.01211839, + "auxiliary_loss_mlp": 0.01034188, + "balance_loss_clip": 1.05992854, + "balance_loss_mlp": 1.02276158, + "epoch": 0.22184813322912283, + "flos": 16982300361600.0, + "grad_norm": 2.537089452971504, + "language_loss": 0.85330588, + "learning_rate": 3.6263264654031814e-06, + "loss": 0.87576616, + "num_input_tokens_seen": 39234195, + "step": 1845, + "time_per_iteration": 2.689768075942993 + }, + { + "auxiliary_loss_clip": 0.01083605, + "auxiliary_loss_mlp": 0.01005292, + "balance_loss_clip": 1.02416158, + "balance_loss_mlp": 1.00133455, + "epoch": 0.22196837611976192, + "flos": 61823740314240.0, + "grad_norm": 0.7015809973163472, + "language_loss": 0.5917204, + "learning_rate": 3.6258729527498008e-06, + "loss": 0.61260939, + "num_input_tokens_seen": 39295040, + "step": 1846, + "time_per_iteration": 3.357877492904663 + }, + { + "auxiliary_loss_clip": 0.01198089, + "auxiliary_loss_mlp": 0.01033986, + "balance_loss_clip": 1.05602813, + "balance_loss_mlp": 1.02257228, + "epoch": 0.222088619010401, + "flos": 25558019625600.0, + "grad_norm": 2.9686277306828157, + "language_loss": 0.65165573, + "learning_rate": 3.6254191934527854e-06, + "loss": 0.67397648, + "num_input_tokens_seen": 39314395, + "step": 1847, + "time_per_iteration": 2.7939908504486084 + }, + { + "auxiliary_loss_clip": 0.01185866, + "auxiliary_loss_mlp": 0.01044351, + "balance_loss_clip": 1.05999303, + "balance_loss_mlp": 1.03318691, + "epoch": 0.2222088619010401, + "flos": 19318612677120.0, + "grad_norm": 2.2751841039081, + "language_loss": 0.64645517, + "learning_rate": 3.6249651875809715e-06, + "loss": 0.66875732, + "num_input_tokens_seen": 39334275, + "step": 1848, + "time_per_iteration": 2.7356314659118652 + }, + { + "auxiliary_loss_clip": 0.01189697, + "auxiliary_loss_mlp": 0.01034765, + "balance_loss_clip": 1.05936658, + "balance_loss_mlp": 1.02402973, + "epoch": 0.2223291047916792, + "flos": 19099342103040.0, + "grad_norm": 1.8172489759184893, + "language_loss": 0.89100623, + "learning_rate": 3.62451093520323e-06, + "loss": 0.91325086, + "num_input_tokens_seen": 39352180, + "step": 1849, + "time_per_iteration": 2.7402660846710205 + }, + { + "auxiliary_loss_clip": 0.01169396, + "auxiliary_loss_mlp": 0.01035936, + "balance_loss_clip": 1.05269754, + "balance_loss_mlp": 1.0241996, + "epoch": 0.22244934768231828, + "flos": 20850418126080.0, + "grad_norm": 2.671036051644632, + "language_loss": 0.90506804, + "learning_rate": 3.6240564363884714e-06, + "loss": 0.92712128, + "num_input_tokens_seen": 39372125, + "step": 1850, + "time_per_iteration": 2.850820541381836 + }, + { + "auxiliary_loss_clip": 0.01208348, + "auxiliary_loss_mlp": 0.01041342, + "balance_loss_clip": 1.05641139, + "balance_loss_mlp": 1.02995718, + "epoch": 0.2225695905729574, + "flos": 15632921111040.0, + "grad_norm": 2.2766493948632105, + "language_loss": 0.70387602, + "learning_rate": 3.623601691205643e-06, + "loss": 0.72637296, + "num_input_tokens_seen": 39391200, + "step": 1851, + "time_per_iteration": 2.6804778575897217 + }, + { + "auxiliary_loss_clip": 0.01200686, + "auxiliary_loss_mlp": 0.01036988, + "balance_loss_clip": 1.05574155, + "balance_loss_mlp": 1.02577674, + "epoch": 0.22268983346359647, + "flos": 25373582265600.0, + "grad_norm": 2.663309364282831, + "language_loss": 0.81515437, + "learning_rate": 3.623146699723729e-06, + "loss": 0.83753109, + "num_input_tokens_seen": 39410660, + "step": 1852, + "time_per_iteration": 2.703944206237793 + }, + { + "auxiliary_loss_clip": 0.01196638, + "auxiliary_loss_mlp": 0.01032339, + "balance_loss_clip": 1.06195736, + "balance_loss_mlp": 1.02156842, + "epoch": 0.22281007635423555, + "flos": 13261452359040.0, + "grad_norm": 1.9604895788812422, + "language_loss": 0.77648342, + "learning_rate": 3.6226914620117507e-06, + "loss": 0.79877323, + "num_input_tokens_seen": 39429280, + "step": 1853, + "time_per_iteration": 2.7598843574523926 + }, + { + "auxiliary_loss_clip": 0.01180332, + "auxiliary_loss_mlp": 0.0103318, + "balance_loss_clip": 1.04787874, + "balance_loss_mlp": 1.0232259, + "epoch": 0.22293031924487464, + "flos": 15340536403200.0, + "grad_norm": 1.9116969706402882, + "language_loss": 0.80571324, + "learning_rate": 3.622235978138768e-06, + "loss": 0.82784832, + "num_input_tokens_seen": 39446905, + "step": 1854, + "time_per_iteration": 2.901883840560913 + }, + { + "auxiliary_loss_clip": 0.01202204, + "auxiliary_loss_mlp": 0.01047992, + "balance_loss_clip": 1.05892026, + "balance_loss_mlp": 1.03768587, + "epoch": 0.22305056213551375, + "flos": 22564649773440.0, + "grad_norm": 2.269705615131243, + "language_loss": 0.8105129, + "learning_rate": 3.621780248173877e-06, + "loss": 0.83301485, + "num_input_tokens_seen": 39465105, + "step": 1855, + "time_per_iteration": 2.756514310836792 + }, + { + "auxiliary_loss_clip": 0.01109246, + "auxiliary_loss_mlp": 0.01004402, + "balance_loss_clip": 1.03139794, + "balance_loss_mlp": 1.00063503, + "epoch": 0.22317080502615283, + "flos": 64880419887360.0, + "grad_norm": 0.8238125145072752, + "language_loss": 0.61082268, + "learning_rate": 3.6213242721862125e-06, + "loss": 0.6319592, + "num_input_tokens_seen": 39523560, + "step": 1856, + "time_per_iteration": 3.3362207412719727 + }, + { + "auxiliary_loss_clip": 0.01181548, + "auxiliary_loss_mlp": 0.01036039, + "balance_loss_clip": 1.05358124, + "balance_loss_mlp": 1.02572751, + "epoch": 0.2232910479167919, + "flos": 25775997310080.0, + "grad_norm": 1.4915313565230224, + "language_loss": 0.75309408, + "learning_rate": 3.620868050244945e-06, + "loss": 0.77526999, + "num_input_tokens_seen": 39544040, + "step": 1857, + "time_per_iteration": 2.808650493621826 + }, + { + "auxiliary_loss_clip": 0.01186528, + "auxiliary_loss_mlp": 0.01037753, + "balance_loss_clip": 1.05327535, + "balance_loss_mlp": 1.02589726, + "epoch": 0.22341129080743102, + "flos": 23251799928960.0, + "grad_norm": 2.0683377237225327, + "language_loss": 0.77658558, + "learning_rate": 3.6204115824192817e-06, + "loss": 0.79882836, + "num_input_tokens_seen": 39561515, + "step": 1858, + "time_per_iteration": 2.8167004585266113 + }, + { + "auxiliary_loss_clip": 0.01183865, + "auxiliary_loss_mlp": 0.01035102, + "balance_loss_clip": 1.052091, + "balance_loss_mlp": 1.02364624, + "epoch": 0.2235315336980701, + "flos": 21214552250880.0, + "grad_norm": 2.5401690304686873, + "language_loss": 0.76775074, + "learning_rate": 3.619954868778471e-06, + "loss": 0.78994048, + "num_input_tokens_seen": 39578210, + "step": 1859, + "time_per_iteration": 2.7463600635528564 + }, + { + "auxiliary_loss_clip": 0.01189284, + "auxiliary_loss_mlp": 0.01041625, + "balance_loss_clip": 1.0518291, + "balance_loss_mlp": 1.03083634, + "epoch": 0.2236517765887092, + "flos": 19901945548800.0, + "grad_norm": 1.9222970467969922, + "language_loss": 0.82974434, + "learning_rate": 3.6194979093917944e-06, + "loss": 0.85205346, + "num_input_tokens_seen": 39597625, + "step": 1860, + "time_per_iteration": 2.760664701461792 + }, + { + "auxiliary_loss_clip": 0.01185961, + "auxiliary_loss_mlp": 0.01039077, + "balance_loss_clip": 1.05298257, + "balance_loss_mlp": 1.02809739, + "epoch": 0.22377201947934827, + "flos": 23214847812480.0, + "grad_norm": 1.9144724046355772, + "language_loss": 0.8670069, + "learning_rate": 3.6190407043285724e-06, + "loss": 0.88925731, + "num_input_tokens_seen": 39615360, + "step": 1861, + "time_per_iteration": 2.7836456298828125 + }, + { + "auxiliary_loss_clip": 0.0121697, + "auxiliary_loss_mlp": 0.01036742, + "balance_loss_clip": 1.06070948, + "balance_loss_mlp": 1.0256021, + "epoch": 0.22389226236998738, + "flos": 26794244056320.0, + "grad_norm": 3.546003033473767, + "language_loss": 0.76090008, + "learning_rate": 3.618583253658163e-06, + "loss": 0.78343719, + "num_input_tokens_seen": 39635460, + "step": 1862, + "time_per_iteration": 2.66896653175354 + }, + { + "auxiliary_loss_clip": 0.01174208, + "auxiliary_loss_mlp": 0.00887982, + "balance_loss_clip": 1.05565345, + "balance_loss_mlp": 1.00045848, + "epoch": 0.22401250526062647, + "flos": 24170359455360.0, + "grad_norm": 1.8343839250696758, + "language_loss": 0.86182469, + "learning_rate": 3.618125557449961e-06, + "loss": 0.88244665, + "num_input_tokens_seen": 39653515, + "step": 1863, + "time_per_iteration": 3.7961363792419434 + }, + { + "auxiliary_loss_clip": 0.01200839, + "auxiliary_loss_mlp": 0.01040846, + "balance_loss_clip": 1.05832875, + "balance_loss_mlp": 1.02891886, + "epoch": 0.22413274815126555, + "flos": 16759761649920.0, + "grad_norm": 3.7003883748157236, + "language_loss": 0.83126342, + "learning_rate": 3.6176676157733983e-06, + "loss": 0.85368025, + "num_input_tokens_seen": 39668525, + "step": 1864, + "time_per_iteration": 4.448374271392822 + }, + { + "auxiliary_loss_clip": 0.01172364, + "auxiliary_loss_mlp": 0.01042933, + "balance_loss_clip": 1.04910576, + "balance_loss_mlp": 1.03054738, + "epoch": 0.22425299104190466, + "flos": 21360205900800.0, + "grad_norm": 2.103857589269552, + "language_loss": 0.76169336, + "learning_rate": 3.6172094286979443e-06, + "loss": 0.78384632, + "num_input_tokens_seen": 39685895, + "step": 1865, + "time_per_iteration": 2.8178417682647705 + }, + { + "auxiliary_loss_clip": 0.01191443, + "auxiliary_loss_mlp": 0.01035248, + "balance_loss_clip": 1.05302238, + "balance_loss_mlp": 1.02461493, + "epoch": 0.22437323393254374, + "flos": 32165547108480.0, + "grad_norm": 1.4772684928669964, + "language_loss": 0.81373394, + "learning_rate": 3.6167509962931064e-06, + "loss": 0.83600086, + "num_input_tokens_seen": 39711595, + "step": 1866, + "time_per_iteration": 2.8747663497924805 + }, + { + "auxiliary_loss_clip": 0.01181246, + "auxiliary_loss_mlp": 0.01041833, + "balance_loss_clip": 1.05640709, + "balance_loss_mlp": 1.03052032, + "epoch": 0.22449347682318282, + "flos": 18002809664640.0, + "grad_norm": 2.5514291704835794, + "language_loss": 0.77034622, + "learning_rate": 3.6162923186284276e-06, + "loss": 0.79257703, + "num_input_tokens_seen": 39727555, + "step": 1867, + "time_per_iteration": 2.7880234718322754 + }, + { + "auxiliary_loss_clip": 0.01188071, + "auxiliary_loss_mlp": 0.01037021, + "balance_loss_clip": 1.05021238, + "balance_loss_mlp": 1.02531469, + "epoch": 0.2246137197138219, + "flos": 18697286194560.0, + "grad_norm": 2.0009974099978134, + "language_loss": 0.85834002, + "learning_rate": 3.6158333957734888e-06, + "loss": 0.88059098, + "num_input_tokens_seen": 39746145, + "step": 1868, + "time_per_iteration": 2.7813682556152344 + }, + { + "auxiliary_loss_clip": 0.01186974, + "auxiliary_loss_mlp": 0.01043742, + "balance_loss_clip": 1.05230772, + "balance_loss_mlp": 1.0329119, + "epoch": 0.22473396260446102, + "flos": 15590653781760.0, + "grad_norm": 2.5892685616135576, + "language_loss": 0.82516372, + "learning_rate": 3.6153742277979088e-06, + "loss": 0.84747088, + "num_input_tokens_seen": 39763575, + "step": 1869, + "time_per_iteration": 2.799872636795044 + }, + { + "auxiliary_loss_clip": 0.01191468, + "auxiliary_loss_mlp": 0.01041027, + "balance_loss_clip": 1.0540626, + "balance_loss_mlp": 1.02992296, + "epoch": 0.2248542054951001, + "flos": 14465501182080.0, + "grad_norm": 3.721270352353109, + "language_loss": 0.78561246, + "learning_rate": 3.6149148147713434e-06, + "loss": 0.80793738, + "num_input_tokens_seen": 39781810, + "step": 1870, + "time_per_iteration": 2.7590208053588867 + }, + { + "auxiliary_loss_clip": 0.01211349, + "auxiliary_loss_mlp": 0.01044421, + "balance_loss_clip": 1.06365061, + "balance_loss_mlp": 1.03396583, + "epoch": 0.22497444838573918, + "flos": 19243882431360.0, + "grad_norm": 1.9522236834005238, + "language_loss": 0.86462158, + "learning_rate": 3.614455156763484e-06, + "loss": 0.88717926, + "num_input_tokens_seen": 39800115, + "step": 1871, + "time_per_iteration": 2.7822957038879395 + }, + { + "auxiliary_loss_clip": 0.01171012, + "auxiliary_loss_mlp": 0.01036891, + "balance_loss_clip": 1.05041313, + "balance_loss_mlp": 1.02544713, + "epoch": 0.2250946912763783, + "flos": 16910299549440.0, + "grad_norm": 2.0005373639838253, + "language_loss": 0.7112332, + "learning_rate": 3.613995253844061e-06, + "loss": 0.73331225, + "num_input_tokens_seen": 39817795, + "step": 1872, + "time_per_iteration": 2.790781259536743 + }, + { + "auxiliary_loss_clip": 0.01201067, + "auxiliary_loss_mlp": 0.01035492, + "balance_loss_clip": 1.0586195, + "balance_loss_mlp": 1.023839, + "epoch": 0.22521493416701738, + "flos": 24681368292480.0, + "grad_norm": 1.8132145493044756, + "language_loss": 0.80990404, + "learning_rate": 3.6135351060828414e-06, + "loss": 0.83226961, + "num_input_tokens_seen": 39838270, + "step": 1873, + "time_per_iteration": 2.7476086616516113 + }, + { + "auxiliary_loss_clip": 0.01217237, + "auxiliary_loss_mlp": 0.01038502, + "balance_loss_clip": 1.05993831, + "balance_loss_mlp": 1.02695107, + "epoch": 0.22533517705765646, + "flos": 17821963664640.0, + "grad_norm": 2.416000735330934, + "language_loss": 0.69192863, + "learning_rate": 3.6130747135496285e-06, + "loss": 0.71448612, + "num_input_tokens_seen": 39857270, + "step": 1874, + "time_per_iteration": 2.6321332454681396 + }, + { + "auxiliary_loss_clip": 0.0121398, + "auxiliary_loss_mlp": 0.01038147, + "balance_loss_clip": 1.05960393, + "balance_loss_mlp": 1.02647042, + "epoch": 0.22545541994829554, + "flos": 33691390899840.0, + "grad_norm": 1.6924738241474337, + "language_loss": 0.65914047, + "learning_rate": 3.6126140763142646e-06, + "loss": 0.68166173, + "num_input_tokens_seen": 39882300, + "step": 1875, + "time_per_iteration": 2.780090808868408 + }, + { + "auxiliary_loss_clip": 0.01217113, + "auxiliary_loss_mlp": 0.01039633, + "balance_loss_clip": 1.06108356, + "balance_loss_mlp": 1.02803373, + "epoch": 0.22557566283893465, + "flos": 19171594310400.0, + "grad_norm": 2.752890000890695, + "language_loss": 0.85790557, + "learning_rate": 3.6121531944466275e-06, + "loss": 0.88047302, + "num_input_tokens_seen": 39899625, + "step": 1876, + "time_per_iteration": 2.687798261642456 + }, + { + "auxiliary_loss_clip": 0.012013, + "auxiliary_loss_mlp": 0.01041177, + "balance_loss_clip": 1.05842495, + "balance_loss_mlp": 1.03041244, + "epoch": 0.22569590572957374, + "flos": 20773281669120.0, + "grad_norm": 4.2554415286884115, + "language_loss": 0.77838546, + "learning_rate": 3.611692068016633e-06, + "loss": 0.80081022, + "num_input_tokens_seen": 39915955, + "step": 1877, + "time_per_iteration": 2.6929614543914795 + }, + { + "auxiliary_loss_clip": 0.01176346, + "auxiliary_loss_mlp": 0.0104096, + "balance_loss_clip": 1.05081904, + "balance_loss_mlp": 1.02874112, + "epoch": 0.22581614862021282, + "flos": 18442715529600.0, + "grad_norm": 2.280338400532588, + "language_loss": 0.74979001, + "learning_rate": 3.611230697094233e-06, + "loss": 0.77196312, + "num_input_tokens_seen": 39932655, + "step": 1878, + "time_per_iteration": 2.723253011703491 + }, + { + "auxiliary_loss_clip": 0.01199632, + "auxiliary_loss_mlp": 0.01036769, + "balance_loss_clip": 1.05766368, + "balance_loss_mlp": 1.02582002, + "epoch": 0.22593639151085193, + "flos": 20048389297920.0, + "grad_norm": 1.8194722870292623, + "language_loss": 0.87336063, + "learning_rate": 3.6107690817494173e-06, + "loss": 0.89572465, + "num_input_tokens_seen": 39952875, + "step": 1879, + "time_per_iteration": 2.834102153778076 + }, + { + "auxiliary_loss_clip": 0.01165998, + "auxiliary_loss_mlp": 0.01039663, + "balance_loss_clip": 1.04847264, + "balance_loss_mlp": 1.02822471, + "epoch": 0.226056634401491, + "flos": 13115116350720.0, + "grad_norm": 2.29727047853007, + "language_loss": 0.70838773, + "learning_rate": 3.6103072220522117e-06, + "loss": 0.73044431, + "num_input_tokens_seen": 39968405, + "step": 1880, + "time_per_iteration": 2.8135595321655273 + }, + { + "auxiliary_loss_clip": 0.01190684, + "auxiliary_loss_mlp": 0.01033747, + "balance_loss_clip": 1.05563998, + "balance_loss_mlp": 1.022506, + "epoch": 0.2261768772921301, + "flos": 18988378012800.0, + "grad_norm": 1.838233161252353, + "language_loss": 0.91555119, + "learning_rate": 3.609845118072682e-06, + "loss": 0.93779552, + "num_input_tokens_seen": 39987075, + "step": 1881, + "time_per_iteration": 2.8485841751098633 + }, + { + "auxiliary_loss_clip": 0.01208649, + "auxiliary_loss_mlp": 0.00888461, + "balance_loss_clip": 1.05734205, + "balance_loss_mlp": 1.00049615, + "epoch": 0.2262971201827692, + "flos": 19974054101760.0, + "grad_norm": 1.7402664153717815, + "language_loss": 0.79902124, + "learning_rate": 3.6093827698809276e-06, + "loss": 0.81999236, + "num_input_tokens_seen": 40006175, + "step": 1882, + "time_per_iteration": 2.7900800704956055 + }, + { + "auxiliary_loss_clip": 0.01200725, + "auxiliary_loss_mlp": 0.01031093, + "balance_loss_clip": 1.05361688, + "balance_loss_mlp": 1.02012539, + "epoch": 0.2264173630734083, + "flos": 16654543735680.0, + "grad_norm": 2.64187514742761, + "language_loss": 0.85059261, + "learning_rate": 3.6089201775470864e-06, + "loss": 0.87291074, + "num_input_tokens_seen": 40021630, + "step": 1883, + "time_per_iteration": 2.6824653148651123 + }, + { + "auxiliary_loss_clip": 0.01172012, + "auxiliary_loss_mlp": 0.01033259, + "balance_loss_clip": 1.05188966, + "balance_loss_mlp": 1.02205896, + "epoch": 0.22653760596404737, + "flos": 24389809597440.0, + "grad_norm": 1.496391535521799, + "language_loss": 0.77597868, + "learning_rate": 3.6084573411413334e-06, + "loss": 0.79803145, + "num_input_tokens_seen": 40041025, + "step": 1884, + "time_per_iteration": 2.867852210998535 + }, + { + "auxiliary_loss_clip": 0.01187138, + "auxiliary_loss_mlp": 0.01042751, + "balance_loss_clip": 1.05672669, + "balance_loss_mlp": 1.03094935, + "epoch": 0.22665784885468646, + "flos": 18332541538560.0, + "grad_norm": 2.097867740046919, + "language_loss": 0.81350076, + "learning_rate": 3.607994260733881e-06, + "loss": 0.83579957, + "num_input_tokens_seen": 40060265, + "step": 1885, + "time_per_iteration": 2.7588071823120117 + }, + { + "auxiliary_loss_clip": 0.01191294, + "auxiliary_loss_mlp": 0.01040864, + "balance_loss_clip": 1.05310869, + "balance_loss_mlp": 1.03002799, + "epoch": 0.22677809174532557, + "flos": 24058102475520.0, + "grad_norm": 1.60366714615662, + "language_loss": 0.74768925, + "learning_rate": 3.6075309363949776e-06, + "loss": 0.77001083, + "num_input_tokens_seen": 40079435, + "step": 1886, + "time_per_iteration": 2.786679267883301 + }, + { + "auxiliary_loss_clip": 0.01214722, + "auxiliary_loss_mlp": 0.01040777, + "balance_loss_clip": 1.05949891, + "balance_loss_mlp": 1.02966058, + "epoch": 0.22689833463596465, + "flos": 20374242503040.0, + "grad_norm": 1.947642003843416, + "language_loss": 0.81427252, + "learning_rate": 3.6070673681949094e-06, + "loss": 0.83682752, + "num_input_tokens_seen": 40097800, + "step": 1887, + "time_per_iteration": 2.9154796600341797 + }, + { + "auxiliary_loss_clip": 0.01195308, + "auxiliary_loss_mlp": 0.00888685, + "balance_loss_clip": 1.05621207, + "balance_loss_mlp": 1.00045013, + "epoch": 0.22701857752660373, + "flos": 30120398438400.0, + "grad_norm": 1.7450985924858624, + "language_loss": 0.81410086, + "learning_rate": 3.606603556203999e-06, + "loss": 0.83494079, + "num_input_tokens_seen": 40122745, + "step": 1888, + "time_per_iteration": 3.8674497604370117 + }, + { + "auxiliary_loss_clip": 0.01205646, + "auxiliary_loss_mlp": 0.01037515, + "balance_loss_clip": 1.0571804, + "balance_loss_mlp": 1.02581453, + "epoch": 0.22713882041724284, + "flos": 22492182084480.0, + "grad_norm": 1.8469998407963646, + "language_loss": 0.83502758, + "learning_rate": 3.6061395004926066e-06, + "loss": 0.85745919, + "num_input_tokens_seen": 40141680, + "step": 1889, + "time_per_iteration": 4.593775033950806 + }, + { + "auxiliary_loss_clip": 0.01216542, + "auxiliary_loss_mlp": 0.01041286, + "balance_loss_clip": 1.05989718, + "balance_loss_mlp": 1.02888787, + "epoch": 0.22725906330788193, + "flos": 20521548178560.0, + "grad_norm": 2.0327495201002153, + "language_loss": 0.84830797, + "learning_rate": 3.605675201131129e-06, + "loss": 0.87088621, + "num_input_tokens_seen": 40160140, + "step": 1890, + "time_per_iteration": 3.513054847717285 + }, + { + "auxiliary_loss_clip": 0.01209838, + "auxiliary_loss_mlp": 0.01034092, + "balance_loss_clip": 1.05937195, + "balance_loss_mlp": 1.02277875, + "epoch": 0.227379306198521, + "flos": 18989922297600.0, + "grad_norm": 2.226416865117146, + "language_loss": 0.7930634, + "learning_rate": 3.60521065819e-06, + "loss": 0.81550264, + "num_input_tokens_seen": 40177450, + "step": 1891, + "time_per_iteration": 2.648678779602051 + }, + { + "auxiliary_loss_clip": 0.01195743, + "auxiliary_loss_mlp": 0.01038808, + "balance_loss_clip": 1.05371177, + "balance_loss_mlp": 1.02743506, + "epoch": 0.2274995490891601, + "flos": 21798351999360.0, + "grad_norm": 1.8778793872522586, + "language_loss": 0.87395811, + "learning_rate": 3.60474587173969e-06, + "loss": 0.89630365, + "num_input_tokens_seen": 40195935, + "step": 1892, + "time_per_iteration": 2.7384860515594482 + }, + { + "auxiliary_loss_clip": 0.0120563, + "auxiliary_loss_mlp": 0.0103781, + "balance_loss_clip": 1.0619061, + "balance_loss_mlp": 1.02553701, + "epoch": 0.2276197919797992, + "flos": 19058654972160.0, + "grad_norm": 1.8579035833484534, + "language_loss": 0.83729255, + "learning_rate": 3.6042808418507084e-06, + "loss": 0.85972703, + "num_input_tokens_seen": 40213620, + "step": 1893, + "time_per_iteration": 2.6964495182037354 + }, + { + "auxiliary_loss_clip": 0.01207295, + "auxiliary_loss_mlp": 0.01040426, + "balance_loss_clip": 1.06005204, + "balance_loss_mlp": 1.02893448, + "epoch": 0.22774003487043828, + "flos": 18806777827200.0, + "grad_norm": 2.8072576502642237, + "language_loss": 0.76826042, + "learning_rate": 3.6038155685935976e-06, + "loss": 0.79073763, + "num_input_tokens_seen": 40230190, + "step": 1894, + "time_per_iteration": 2.65779709815979 + }, + { + "auxiliary_loss_clip": 0.01204854, + "auxiliary_loss_mlp": 0.01031218, + "balance_loss_clip": 1.05918717, + "balance_loss_mlp": 1.01977968, + "epoch": 0.22786027776107737, + "flos": 23002544476800.0, + "grad_norm": 1.9237935945493476, + "language_loss": 0.70602, + "learning_rate": 3.6033500520389404e-06, + "loss": 0.72838068, + "num_input_tokens_seen": 40246860, + "step": 1895, + "time_per_iteration": 2.724534273147583 + }, + { + "auxiliary_loss_clip": 0.01084959, + "auxiliary_loss_mlp": 0.01003869, + "balance_loss_clip": 1.02440977, + "balance_loss_mlp": 0.99988741, + "epoch": 0.22798052065171648, + "flos": 66706872600960.0, + "grad_norm": 0.7961600245746834, + "language_loss": 0.64855504, + "learning_rate": 3.6028842922573553e-06, + "loss": 0.66944331, + "num_input_tokens_seen": 40311005, + "step": 1896, + "time_per_iteration": 3.418619155883789 + }, + { + "auxiliary_loss_clip": 0.01091028, + "auxiliary_loss_mlp": 0.00878202, + "balance_loss_clip": 1.02427244, + "balance_loss_mlp": 1.00047255, + "epoch": 0.22810076354235556, + "flos": 62080896758400.0, + "grad_norm": 0.8624518431948331, + "language_loss": 0.62905729, + "learning_rate": 3.602418289319497e-06, + "loss": 0.64874959, + "num_input_tokens_seen": 40369560, + "step": 1897, + "time_per_iteration": 3.3605003356933594 + }, + { + "auxiliary_loss_clip": 0.01172175, + "auxiliary_loss_mlp": 0.01039547, + "balance_loss_clip": 1.05322552, + "balance_loss_mlp": 1.02798414, + "epoch": 0.22822100643299464, + "flos": 23876358635520.0, + "grad_norm": 1.668053380197855, + "language_loss": 0.73436773, + "learning_rate": 3.601952043296059e-06, + "loss": 0.75648493, + "num_input_tokens_seen": 40389555, + "step": 1898, + "time_per_iteration": 2.9073424339294434 + }, + { + "auxiliary_loss_clip": 0.01197388, + "auxiliary_loss_mlp": 0.01036839, + "balance_loss_clip": 1.05319095, + "balance_loss_mlp": 1.02538323, + "epoch": 0.22834124932363373, + "flos": 20991331180800.0, + "grad_norm": 1.9788448984893794, + "language_loss": 0.80634081, + "learning_rate": 3.6014855542577696e-06, + "loss": 0.82868308, + "num_input_tokens_seen": 40406765, + "step": 1899, + "time_per_iteration": 2.732893228530884 + }, + { + "auxiliary_loss_clip": 0.0119386, + "auxiliary_loss_mlp": 0.01048067, + "balance_loss_clip": 1.05572629, + "balance_loss_mlp": 1.03600335, + "epoch": 0.22846149221427284, + "flos": 24901572620160.0, + "grad_norm": 1.8524495339637401, + "language_loss": 0.84287012, + "learning_rate": 3.6010188222753943e-06, + "loss": 0.86528933, + "num_input_tokens_seen": 40427535, + "step": 1900, + "time_per_iteration": 2.77614164352417 + }, + { + "auxiliary_loss_clip": 0.01093619, + "auxiliary_loss_mlp": 0.01003456, + "balance_loss_clip": 1.02619147, + "balance_loss_mlp": 0.99968928, + "epoch": 0.22858173510491192, + "flos": 56132294319360.0, + "grad_norm": 0.9036218841944029, + "language_loss": 0.64151657, + "learning_rate": 3.6005518474197372e-06, + "loss": 0.66248727, + "num_input_tokens_seen": 40479580, + "step": 1901, + "time_per_iteration": 3.185330629348755 + }, + { + "auxiliary_loss_clip": 0.0120723, + "auxiliary_loss_mlp": 0.01037464, + "balance_loss_clip": 1.06108928, + "balance_loss_mlp": 1.02577519, + "epoch": 0.228701977995551, + "flos": 24170826332160.0, + "grad_norm": 1.9551821031451866, + "language_loss": 0.78667271, + "learning_rate": 3.6000846297616373e-06, + "loss": 0.8091197, + "num_input_tokens_seen": 40497880, + "step": 1902, + "time_per_iteration": 2.7794363498687744 + }, + { + "auxiliary_loss_clip": 0.01224909, + "auxiliary_loss_mlp": 0.01033841, + "balance_loss_clip": 1.06756473, + "balance_loss_mlp": 1.02231348, + "epoch": 0.22882222088619011, + "flos": 21387892308480.0, + "grad_norm": 2.303281184914442, + "language_loss": 0.73188579, + "learning_rate": 3.5996171693719717e-06, + "loss": 0.75447327, + "num_input_tokens_seen": 40513975, + "step": 1903, + "time_per_iteration": 2.6348981857299805 + }, + { + "auxiliary_loss_clip": 0.01104334, + "auxiliary_loss_mlp": 0.01004962, + "balance_loss_clip": 1.02777898, + "balance_loss_mlp": 1.00105178, + "epoch": 0.2289424637768292, + "flos": 64589615377920.0, + "grad_norm": 0.8706346845816915, + "language_loss": 0.64801693, + "learning_rate": 3.5991494663216528e-06, + "loss": 0.66910994, + "num_input_tokens_seen": 40576960, + "step": 1904, + "time_per_iteration": 3.4264070987701416 + }, + { + "auxiliary_loss_clip": 0.012195, + "auxiliary_loss_mlp": 0.0103423, + "balance_loss_clip": 1.06216049, + "balance_loss_mlp": 1.02327466, + "epoch": 0.22906270666746828, + "flos": 22163419877760.0, + "grad_norm": 1.9793303886070561, + "language_loss": 0.8778528, + "learning_rate": 3.5986815206816314e-06, + "loss": 0.90039009, + "num_input_tokens_seen": 40595780, + "step": 1905, + "time_per_iteration": 2.8573992252349854 + }, + { + "auxiliary_loss_clip": 0.0121535, + "auxiliary_loss_mlp": 0.01037452, + "balance_loss_clip": 1.05927742, + "balance_loss_mlp": 1.02609181, + "epoch": 0.2291829495581074, + "flos": 25772334122880.0, + "grad_norm": 2.655453637123742, + "language_loss": 0.74626422, + "learning_rate": 3.598213332522895e-06, + "loss": 0.76879227, + "num_input_tokens_seen": 40615810, + "step": 1906, + "time_per_iteration": 2.7201578617095947 + }, + { + "auxiliary_loss_clip": 0.01206436, + "auxiliary_loss_mlp": 0.01038566, + "balance_loss_clip": 1.05863392, + "balance_loss_mlp": 1.02775931, + "epoch": 0.22930319244874647, + "flos": 31172760126720.0, + "grad_norm": 1.7397101162469135, + "language_loss": 0.77550089, + "learning_rate": 3.597744901916466e-06, + "loss": 0.79795092, + "num_input_tokens_seen": 40637095, + "step": 1907, + "time_per_iteration": 2.8082327842712402 + }, + { + "auxiliary_loss_clip": 0.01219738, + "auxiliary_loss_mlp": 0.01036324, + "balance_loss_clip": 1.05882764, + "balance_loss_mlp": 1.02430773, + "epoch": 0.22942343533938556, + "flos": 23254098399360.0, + "grad_norm": 2.2490728796866604, + "language_loss": 0.76382375, + "learning_rate": 3.5972762289334058e-06, + "loss": 0.78638434, + "num_input_tokens_seen": 40656725, + "step": 1908, + "time_per_iteration": 2.870388984680176 + }, + { + "auxiliary_loss_clip": 0.0116722, + "auxiliary_loss_mlp": 0.01034567, + "balance_loss_clip": 1.05534792, + "balance_loss_mlp": 1.02299798, + "epoch": 0.22954367823002464, + "flos": 14610903436800.0, + "grad_norm": 1.9615176972609059, + "language_loss": 0.84916997, + "learning_rate": 3.5968073136448116e-06, + "loss": 0.87118787, + "num_input_tokens_seen": 40674745, + "step": 1909, + "time_per_iteration": 3.002074718475342 + }, + { + "auxiliary_loss_clip": 0.01212097, + "auxiliary_loss_mlp": 0.010449, + "balance_loss_clip": 1.0590564, + "balance_loss_mlp": 1.03249061, + "epoch": 0.22966392112066375, + "flos": 16763604405120.0, + "grad_norm": 1.6678418925151939, + "language_loss": 0.91390347, + "learning_rate": 3.596338156121818e-06, + "loss": 0.93647343, + "num_input_tokens_seen": 40693630, + "step": 1910, + "time_per_iteration": 2.7152533531188965 + }, + { + "auxiliary_loss_clip": 0.010927, + "auxiliary_loss_mlp": 0.01006499, + "balance_loss_clip": 1.02556348, + "balance_loss_mlp": 1.00282788, + "epoch": 0.22978416401130283, + "flos": 67474247783040.0, + "grad_norm": 0.7470580600814121, + "language_loss": 0.59347302, + "learning_rate": 3.595868756435595e-06, + "loss": 0.61446506, + "num_input_tokens_seen": 40761310, + "step": 1911, + "time_per_iteration": 3.45927095413208 + }, + { + "auxiliary_loss_clip": 0.01192275, + "auxiliary_loss_mlp": 0.01038485, + "balance_loss_clip": 1.06140614, + "balance_loss_mlp": 1.02704096, + "epoch": 0.22990440690194192, + "flos": 19865137086720.0, + "grad_norm": 2.0144557177609594, + "language_loss": 0.80034405, + "learning_rate": 3.5953991146573504e-06, + "loss": 0.82265162, + "num_input_tokens_seen": 40779955, + "step": 1912, + "time_per_iteration": 2.830925226211548 + }, + { + "auxiliary_loss_clip": 0.01209643, + "auxiliary_loss_mlp": 0.01043612, + "balance_loss_clip": 1.05715537, + "balance_loss_mlp": 1.03120196, + "epoch": 0.23002464979258103, + "flos": 13289246507520.0, + "grad_norm": 2.6377906751627616, + "language_loss": 0.83436751, + "learning_rate": 3.5949292308583294e-06, + "loss": 0.85689998, + "num_input_tokens_seen": 40793200, + "step": 1913, + "time_per_iteration": 2.7126643657684326 + }, + { + "auxiliary_loss_clip": 0.01221964, + "auxiliary_loss_mlp": 0.01040845, + "balance_loss_clip": 1.06410909, + "balance_loss_mlp": 1.02905512, + "epoch": 0.2301448926832201, + "flos": 22163779013760.0, + "grad_norm": 1.9289579402475863, + "language_loss": 0.80459929, + "learning_rate": 3.594459105109811e-06, + "loss": 0.82722735, + "num_input_tokens_seen": 40812380, + "step": 1914, + "time_per_iteration": 2.7325642108917236 + }, + { + "auxiliary_loss_clip": 0.01210069, + "auxiliary_loss_mlp": 0.01036822, + "balance_loss_clip": 1.06104779, + "balance_loss_mlp": 1.0255506, + "epoch": 0.2302651355738592, + "flos": 20704477167360.0, + "grad_norm": 1.8274506167243696, + "language_loss": 0.81769741, + "learning_rate": 3.593988737483115e-06, + "loss": 0.84016633, + "num_input_tokens_seen": 40832320, + "step": 1915, + "time_per_iteration": 5.47285008430481 + }, + { + "auxiliary_loss_clip": 0.0119801, + "auxiliary_loss_mlp": 0.0103561, + "balance_loss_clip": 1.05961978, + "balance_loss_mlp": 1.02379608, + "epoch": 0.23038537846449827, + "flos": 18588943797120.0, + "grad_norm": 2.1899857505351084, + "language_loss": 0.77985871, + "learning_rate": 3.5935181280495947e-06, + "loss": 0.80219495, + "num_input_tokens_seen": 40850900, + "step": 1916, + "time_per_iteration": 2.7782721519470215 + }, + { + "auxiliary_loss_clip": 0.01090755, + "auxiliary_loss_mlp": 0.01000979, + "balance_loss_clip": 1.02367318, + "balance_loss_mlp": 0.99766487, + "epoch": 0.23050562135513739, + "flos": 64224260190720.0, + "grad_norm": 0.8021972690829585, + "language_loss": 0.54286385, + "learning_rate": 3.5930472768806412e-06, + "loss": 0.56378114, + "num_input_tokens_seen": 40909570, + "step": 1917, + "time_per_iteration": 3.287091016769409 + }, + { + "auxiliary_loss_clip": 0.01221713, + "auxiliary_loss_mlp": 0.01040811, + "balance_loss_clip": 1.06490231, + "balance_loss_mlp": 1.02931309, + "epoch": 0.23062586424577647, + "flos": 17313396952320.0, + "grad_norm": 1.9395304097520645, + "language_loss": 0.77310562, + "learning_rate": 3.5925761840476826e-06, + "loss": 0.79573095, + "num_input_tokens_seen": 40928180, + "step": 1918, + "time_per_iteration": 2.7255282402038574 + }, + { + "auxiliary_loss_clip": 0.01196236, + "auxiliary_loss_mlp": 0.01034413, + "balance_loss_clip": 1.06160331, + "balance_loss_mlp": 1.02282, + "epoch": 0.23074610713641555, + "flos": 27855979194240.0, + "grad_norm": 2.0918325362156844, + "language_loss": 0.81335068, + "learning_rate": 3.592104849622183e-06, + "loss": 0.83565724, + "num_input_tokens_seen": 40950435, + "step": 1919, + "time_per_iteration": 2.799891233444214 + }, + { + "auxiliary_loss_clip": 0.01169175, + "auxiliary_loss_mlp": 0.01035112, + "balance_loss_clip": 1.05427384, + "balance_loss_mlp": 1.02310205, + "epoch": 0.23086635002705466, + "flos": 28841798937600.0, + "grad_norm": 1.4077102382845963, + "language_loss": 0.73506165, + "learning_rate": 3.591633273675644e-06, + "loss": 0.75710458, + "num_input_tokens_seen": 40972670, + "step": 1920, + "time_per_iteration": 2.960385322570801 + }, + { + "auxiliary_loss_clip": 0.01098155, + "auxiliary_loss_mlp": 0.01005388, + "balance_loss_clip": 1.04705942, + "balance_loss_mlp": 1.0015732, + "epoch": 0.23098659291769374, + "flos": 62923681566720.0, + "grad_norm": 0.9320946730255644, + "language_loss": 0.58164734, + "learning_rate": 3.591161456279602e-06, + "loss": 0.60268277, + "num_input_tokens_seen": 41018215, + "step": 1921, + "time_per_iteration": 3.1842734813690186 + }, + { + "auxiliary_loss_clip": 0.01204197, + "auxiliary_loss_mlp": 0.01035814, + "balance_loss_clip": 1.05851483, + "balance_loss_mlp": 1.02455497, + "epoch": 0.23110683580833283, + "flos": 23476816679040.0, + "grad_norm": 1.5250846765764594, + "language_loss": 0.80223936, + "learning_rate": 3.590689397505633e-06, + "loss": 0.82463944, + "num_input_tokens_seen": 41039125, + "step": 1922, + "time_per_iteration": 2.8552632331848145 + }, + { + "auxiliary_loss_clip": 0.0121535, + "auxiliary_loss_mlp": 0.01038474, + "balance_loss_clip": 1.06091285, + "balance_loss_mlp": 1.02747083, + "epoch": 0.2312270786989719, + "flos": 27271066124160.0, + "grad_norm": 1.5659486480746905, + "language_loss": 0.86674094, + "learning_rate": 3.590217097425347e-06, + "loss": 0.88927919, + "num_input_tokens_seen": 41059025, + "step": 1923, + "time_per_iteration": 2.7404937744140625 + }, + { + "auxiliary_loss_clip": 0.01221748, + "auxiliary_loss_mlp": 0.01046756, + "balance_loss_clip": 1.06363833, + "balance_loss_mlp": 1.03504992, + "epoch": 0.23134732158961102, + "flos": 13261344618240.0, + "grad_norm": 2.090641246885632, + "language_loss": 0.71227688, + "learning_rate": 3.589744556110391e-06, + "loss": 0.73496187, + "num_input_tokens_seen": 41077015, + "step": 1924, + "time_per_iteration": 2.6574692726135254 + }, + { + "auxiliary_loss_clip": 0.01196082, + "auxiliary_loss_mlp": 0.01042916, + "balance_loss_clip": 1.05731773, + "balance_loss_mlp": 1.03069162, + "epoch": 0.2314675644802501, + "flos": 36977648250240.0, + "grad_norm": 1.7464654246357212, + "language_loss": 0.84397757, + "learning_rate": 3.58927177363245e-06, + "loss": 0.86636758, + "num_input_tokens_seen": 41099840, + "step": 1925, + "time_per_iteration": 2.974648952484131 + }, + { + "auxiliary_loss_clip": 0.01182784, + "auxiliary_loss_mlp": 0.0103709, + "balance_loss_clip": 1.05313027, + "balance_loss_mlp": 1.02491879, + "epoch": 0.2315878073708892, + "flos": 23842207779840.0, + "grad_norm": 2.0534428717588655, + "language_loss": 0.72587466, + "learning_rate": 3.5887987500632447e-06, + "loss": 0.7480734, + "num_input_tokens_seen": 41117845, + "step": 1926, + "time_per_iteration": 2.7427525520324707 + }, + { + "auxiliary_loss_clip": 0.01190494, + "auxiliary_loss_mlp": 0.01039146, + "balance_loss_clip": 1.05628347, + "balance_loss_mlp": 1.02839327, + "epoch": 0.2317080502615283, + "flos": 23039424766080.0, + "grad_norm": 1.7223796626035994, + "language_loss": 0.84301996, + "learning_rate": 3.5883254854745325e-06, + "loss": 0.86531639, + "num_input_tokens_seen": 41136235, + "step": 1927, + "time_per_iteration": 2.791865587234497 + }, + { + "auxiliary_loss_clip": 0.01214716, + "auxiliary_loss_mlp": 0.01037749, + "balance_loss_clip": 1.06050098, + "balance_loss_mlp": 1.02600098, + "epoch": 0.23182829315216738, + "flos": 11254656435840.0, + "grad_norm": 2.0429443081817746, + "language_loss": 0.75055563, + "learning_rate": 3.587851979938107e-06, + "loss": 0.77308023, + "num_input_tokens_seen": 41153125, + "step": 1928, + "time_per_iteration": 2.676814317703247 + }, + { + "auxiliary_loss_clip": 0.0120776, + "auxiliary_loss_mlp": 0.01032603, + "balance_loss_clip": 1.06150353, + "balance_loss_mlp": 1.02111089, + "epoch": 0.23194853604280646, + "flos": 19828939155840.0, + "grad_norm": 1.8536695379176769, + "language_loss": 0.77246827, + "learning_rate": 3.5873782335257985e-06, + "loss": 0.79487187, + "num_input_tokens_seen": 41171290, + "step": 1929, + "time_per_iteration": 2.777390480041504 + }, + { + "auxiliary_loss_clip": 0.01191377, + "auxiliary_loss_mlp": 0.01039891, + "balance_loss_clip": 1.06161594, + "balance_loss_mlp": 1.02714109, + "epoch": 0.23206877893344555, + "flos": 15305020830720.0, + "grad_norm": 1.9048319622970091, + "language_loss": 0.78325665, + "learning_rate": 3.5869042463094744e-06, + "loss": 0.80556935, + "num_input_tokens_seen": 41189005, + "step": 1930, + "time_per_iteration": 2.7942495346069336 + }, + { + "auxiliary_loss_clip": 0.01159462, + "auxiliary_loss_mlp": 0.01038058, + "balance_loss_clip": 1.05199957, + "balance_loss_mlp": 1.02590466, + "epoch": 0.23218902182408466, + "flos": 22711488572160.0, + "grad_norm": 2.0684044242711184, + "language_loss": 0.77437186, + "learning_rate": 3.586430018361038e-06, + "loss": 0.79634714, + "num_input_tokens_seen": 41208775, + "step": 1931, + "time_per_iteration": 2.9033029079437256 + }, + { + "auxiliary_loss_clip": 0.01180985, + "auxiliary_loss_mlp": 0.01039114, + "balance_loss_clip": 1.05207419, + "balance_loss_mlp": 1.02718115, + "epoch": 0.23230926471472374, + "flos": 22710734386560.0, + "grad_norm": 1.9003691937124059, + "language_loss": 0.76122403, + "learning_rate": 3.5859555497524283e-06, + "loss": 0.78342497, + "num_input_tokens_seen": 41226010, + "step": 1932, + "time_per_iteration": 2.7620182037353516 + }, + { + "auxiliary_loss_clip": 0.01209351, + "auxiliary_loss_mlp": 0.01042444, + "balance_loss_clip": 1.06264424, + "balance_loss_mlp": 1.03124428, + "epoch": 0.23242950760536282, + "flos": 20375499479040.0, + "grad_norm": 2.7544016197513517, + "language_loss": 0.9216069, + "learning_rate": 3.5854808405556237e-06, + "loss": 0.94412482, + "num_input_tokens_seen": 41245245, + "step": 1933, + "time_per_iteration": 2.9514708518981934 + }, + { + "auxiliary_loss_clip": 0.01189092, + "auxiliary_loss_mlp": 0.0103695, + "balance_loss_clip": 1.05725443, + "balance_loss_mlp": 1.02601898, + "epoch": 0.23254975049600193, + "flos": 16908324301440.0, + "grad_norm": 2.112236495453146, + "language_loss": 0.75101542, + "learning_rate": 3.5850058908426355e-06, + "loss": 0.77327585, + "num_input_tokens_seen": 41263795, + "step": 1934, + "time_per_iteration": 2.7939908504486084 + }, + { + "auxiliary_loss_clip": 0.01197348, + "auxiliary_loss_mlp": 0.01039723, + "balance_loss_clip": 1.05500138, + "balance_loss_mlp": 1.02857161, + "epoch": 0.23266999338664102, + "flos": 23294821443840.0, + "grad_norm": 1.7877128545582477, + "language_loss": 0.85542774, + "learning_rate": 3.584530700685514e-06, + "loss": 0.87779844, + "num_input_tokens_seen": 41284055, + "step": 1935, + "time_per_iteration": 2.7467970848083496 + }, + { + "auxiliary_loss_clip": 0.01193485, + "auxiliary_loss_mlp": 0.01037194, + "balance_loss_clip": 1.06144381, + "balance_loss_mlp": 1.02567291, + "epoch": 0.2327902362772801, + "flos": 19569987031680.0, + "grad_norm": 2.1357295241490633, + "language_loss": 0.88838077, + "learning_rate": 3.5840552701563448e-06, + "loss": 0.91068757, + "num_input_tokens_seen": 41300255, + "step": 1936, + "time_per_iteration": 2.8917860984802246 + }, + { + "auxiliary_loss_clip": 0.01213637, + "auxiliary_loss_mlp": 0.01035286, + "balance_loss_clip": 1.06061637, + "balance_loss_mlp": 1.0242945, + "epoch": 0.2329104791679192, + "flos": 16727514215040.0, + "grad_norm": 2.2403188575903363, + "language_loss": 0.81986594, + "learning_rate": 3.5835795993272513e-06, + "loss": 0.84235519, + "num_input_tokens_seen": 41318540, + "step": 1937, + "time_per_iteration": 2.657656192779541 + }, + { + "auxiliary_loss_clip": 0.01142959, + "auxiliary_loss_mlp": 0.01033535, + "balance_loss_clip": 1.05030715, + "balance_loss_mlp": 1.02216268, + "epoch": 0.2330307220585583, + "flos": 22163743100160.0, + "grad_norm": 2.2929853541044105, + "language_loss": 0.71287024, + "learning_rate": 3.583103688270391e-06, + "loss": 0.73463523, + "num_input_tokens_seen": 41338320, + "step": 1938, + "time_per_iteration": 3.1883082389831543 + }, + { + "auxiliary_loss_clip": 0.0118188, + "auxiliary_loss_mlp": 0.01033929, + "balance_loss_clip": 1.05537331, + "balance_loss_mlp": 1.02220511, + "epoch": 0.23315096494919738, + "flos": 19317319787520.0, + "grad_norm": 1.9478910450427274, + "language_loss": 0.89554811, + "learning_rate": 3.58262753705796e-06, + "loss": 0.91770619, + "num_input_tokens_seen": 41353210, + "step": 1939, + "time_per_iteration": 3.1435470581054688 + }, + { + "auxiliary_loss_clip": 0.01094714, + "auxiliary_loss_mlp": 0.01012353, + "balance_loss_clip": 1.02946854, + "balance_loss_mlp": 1.00889575, + "epoch": 0.23327120783983646, + "flos": 53031048946560.0, + "grad_norm": 0.7605350967571991, + "language_loss": 0.55526769, + "learning_rate": 3.5821511457621902e-06, + "loss": 0.57633835, + "num_input_tokens_seen": 41410510, + "step": 1940, + "time_per_iteration": 4.173599720001221 + }, + { + "auxiliary_loss_clip": 0.01188542, + "auxiliary_loss_mlp": 0.01038827, + "balance_loss_clip": 1.05628848, + "balance_loss_mlp": 1.02639329, + "epoch": 0.23339145073047557, + "flos": 17126984344320.0, + "grad_norm": 2.585909358782647, + "language_loss": 0.81109667, + "learning_rate": 3.5816745144553497e-06, + "loss": 0.83337033, + "num_input_tokens_seen": 41425830, + "step": 1941, + "time_per_iteration": 4.626391649246216 + }, + { + "auxiliary_loss_clip": 0.01173696, + "auxiliary_loss_mlp": 0.01040851, + "balance_loss_clip": 1.05735183, + "balance_loss_mlp": 1.02912068, + "epoch": 0.23351169362111465, + "flos": 13078918419840.0, + "grad_norm": 1.9529776473879894, + "language_loss": 0.75671738, + "learning_rate": 3.5811976432097424e-06, + "loss": 0.77886283, + "num_input_tokens_seen": 41443500, + "step": 1942, + "time_per_iteration": 2.8422555923461914 + }, + { + "auxiliary_loss_clip": 0.01210204, + "auxiliary_loss_mlp": 0.00888553, + "balance_loss_clip": 1.06456709, + "balance_loss_mlp": 1.00050306, + "epoch": 0.23363193651175373, + "flos": 15851257931520.0, + "grad_norm": 2.3344448215089075, + "language_loss": 0.84878826, + "learning_rate": 3.58072053209771e-06, + "loss": 0.86977583, + "num_input_tokens_seen": 41460055, + "step": 1943, + "time_per_iteration": 2.730499505996704 + }, + { + "auxiliary_loss_clip": 0.01187933, + "auxiliary_loss_mlp": 0.01032639, + "balance_loss_clip": 1.05527425, + "balance_loss_mlp": 1.02120066, + "epoch": 0.23375217940239285, + "flos": 21025769345280.0, + "grad_norm": 2.2136652301872095, + "language_loss": 0.79432714, + "learning_rate": 3.5802431811916296e-06, + "loss": 0.81653285, + "num_input_tokens_seen": 41476665, + "step": 1944, + "time_per_iteration": 2.7508232593536377 + }, + { + "auxiliary_loss_clip": 0.01196119, + "auxiliary_loss_mlp": 0.01034492, + "balance_loss_clip": 1.05970204, + "balance_loss_mlp": 1.02337027, + "epoch": 0.23387242229303193, + "flos": 20594698225920.0, + "grad_norm": 2.255722429832909, + "language_loss": 0.80818439, + "learning_rate": 3.579765590563916e-06, + "loss": 0.83049047, + "num_input_tokens_seen": 41496065, + "step": 1945, + "time_per_iteration": 2.6969521045684814 + }, + { + "auxiliary_loss_clip": 0.01198657, + "auxiliary_loss_mlp": 0.01037034, + "balance_loss_clip": 1.05948496, + "balance_loss_mlp": 1.02589345, + "epoch": 0.233992665183671, + "flos": 24279491952000.0, + "grad_norm": 1.8004531015353265, + "language_loss": 0.81438315, + "learning_rate": 3.579287760287017e-06, + "loss": 0.83674002, + "num_input_tokens_seen": 41516815, + "step": 1946, + "time_per_iteration": 2.719428300857544 + }, + { + "auxiliary_loss_clip": 0.0120516, + "auxiliary_loss_mlp": 0.01038514, + "balance_loss_clip": 1.06079268, + "balance_loss_mlp": 1.02787447, + "epoch": 0.2341129080743101, + "flos": 30154621121280.0, + "grad_norm": 1.825961223558505, + "language_loss": 0.72975028, + "learning_rate": 3.578809690433421e-06, + "loss": 0.75218695, + "num_input_tokens_seen": 41538525, + "step": 1947, + "time_per_iteration": 2.804957628250122 + }, + { + "auxiliary_loss_clip": 0.01226689, + "auxiliary_loss_mlp": 0.01040646, + "balance_loss_clip": 1.06846941, + "balance_loss_mlp": 1.0286237, + "epoch": 0.2342331509649492, + "flos": 22784135829120.0, + "grad_norm": 2.498047315592886, + "language_loss": 0.81287992, + "learning_rate": 3.578331381075651e-06, + "loss": 0.83555329, + "num_input_tokens_seen": 41559025, + "step": 1948, + "time_per_iteration": 2.7408177852630615 + }, + { + "auxiliary_loss_clip": 0.01205453, + "auxiliary_loss_mlp": 0.01041785, + "balance_loss_clip": 1.05780888, + "balance_loss_mlp": 1.02957225, + "epoch": 0.2343533938555883, + "flos": 23623152687360.0, + "grad_norm": 2.485630807284886, + "language_loss": 0.6954807, + "learning_rate": 3.5778528322862646e-06, + "loss": 0.71795315, + "num_input_tokens_seen": 41577845, + "step": 1949, + "time_per_iteration": 2.7624354362487793 + }, + { + "auxiliary_loss_clip": 0.01211932, + "auxiliary_loss_mlp": 0.01039119, + "balance_loss_clip": 1.06101155, + "balance_loss_mlp": 1.0280503, + "epoch": 0.23447363674622737, + "flos": 24570332375040.0, + "grad_norm": 1.996151840189072, + "language_loss": 0.86745125, + "learning_rate": 3.5773740441378585e-06, + "loss": 0.88996178, + "num_input_tokens_seen": 41598600, + "step": 1950, + "time_per_iteration": 2.7803409099578857 + }, + { + "auxiliary_loss_clip": 0.01203531, + "auxiliary_loss_mlp": 0.01038776, + "balance_loss_clip": 1.06088877, + "balance_loss_mlp": 1.02779639, + "epoch": 0.23459387963686648, + "flos": 53140322119680.0, + "grad_norm": 1.692962413865299, + "language_loss": 0.74152768, + "learning_rate": 3.5768950167030633e-06, + "loss": 0.76395071, + "num_input_tokens_seen": 41623300, + "step": 1951, + "time_per_iteration": 3.051487684249878 + }, + { + "auxiliary_loss_clip": 0.01183639, + "auxiliary_loss_mlp": 0.01042783, + "balance_loss_clip": 1.05418479, + "balance_loss_mlp": 1.03152347, + "epoch": 0.23471412252750556, + "flos": 23951412103680.0, + "grad_norm": 1.8449833731112575, + "language_loss": 0.78791273, + "learning_rate": 3.576415750054548e-06, + "loss": 0.81017697, + "num_input_tokens_seen": 41643420, + "step": 1952, + "time_per_iteration": 2.808192491531372 + }, + { + "auxiliary_loss_clip": 0.0118481, + "auxiliary_loss_mlp": 0.01042545, + "balance_loss_clip": 1.05678689, + "balance_loss_mlp": 1.03046358, + "epoch": 0.23483436541814465, + "flos": 15706573948800.0, + "grad_norm": 1.8473937438786727, + "language_loss": 0.85815495, + "learning_rate": 3.5759362442650172e-06, + "loss": 0.88042849, + "num_input_tokens_seen": 41660170, + "step": 1953, + "time_per_iteration": 2.743401527404785 + }, + { + "auxiliary_loss_clip": 0.01206852, + "auxiliary_loss_mlp": 0.01038671, + "balance_loss_clip": 1.06404078, + "balance_loss_mlp": 1.02712607, + "epoch": 0.23495460830878373, + "flos": 24936262179840.0, + "grad_norm": 2.3345672258041263, + "language_loss": 0.85160047, + "learning_rate": 3.5754564994072113e-06, + "loss": 0.87405568, + "num_input_tokens_seen": 41679010, + "step": 1954, + "time_per_iteration": 2.7296712398529053 + }, + { + "auxiliary_loss_clip": 0.01191513, + "auxiliary_loss_mlp": 0.01043428, + "balance_loss_clip": 1.05575502, + "balance_loss_mlp": 1.03157842, + "epoch": 0.23507485119942284, + "flos": 30482665056000.0, + "grad_norm": 3.13986035629729, + "language_loss": 0.59953243, + "learning_rate": 3.5749765155539067e-06, + "loss": 0.62188184, + "num_input_tokens_seen": 41699495, + "step": 1955, + "time_per_iteration": 2.8115100860595703 + }, + { + "auxiliary_loss_clip": 0.01183016, + "auxiliary_loss_mlp": 0.01039813, + "balance_loss_clip": 1.05455732, + "balance_loss_mlp": 1.02715313, + "epoch": 0.23519509409006192, + "flos": 18329129746560.0, + "grad_norm": 2.482115991217317, + "language_loss": 0.9254185, + "learning_rate": 3.574496292777917e-06, + "loss": 0.94764674, + "num_input_tokens_seen": 41717705, + "step": 1956, + "time_per_iteration": 2.8045835494995117 + }, + { + "auxiliary_loss_clip": 0.01204382, + "auxiliary_loss_mlp": 0.01040152, + "balance_loss_clip": 1.06090987, + "balance_loss_mlp": 1.02753353, + "epoch": 0.235315336980701, + "flos": 29643217234560.0, + "grad_norm": 2.0266414814598166, + "language_loss": 0.71411365, + "learning_rate": 3.574015831152092e-06, + "loss": 0.73655897, + "num_input_tokens_seen": 41738120, + "step": 1957, + "time_per_iteration": 2.7620773315429688 + }, + { + "auxiliary_loss_clip": 0.01185404, + "auxiliary_loss_mlp": 0.01045543, + "balance_loss_clip": 1.05740452, + "balance_loss_mlp": 1.03362226, + "epoch": 0.23543557987134012, + "flos": 18551704371840.0, + "grad_norm": 1.958948452182382, + "language_loss": 0.83058852, + "learning_rate": 3.573535130749316e-06, + "loss": 0.852898, + "num_input_tokens_seen": 41756070, + "step": 1958, + "time_per_iteration": 2.7610647678375244 + }, + { + "auxiliary_loss_clip": 0.01186635, + "auxiliary_loss_mlp": 0.01035763, + "balance_loss_clip": 1.05789769, + "balance_loss_mlp": 1.0251472, + "epoch": 0.2355558227619792, + "flos": 24679033908480.0, + "grad_norm": 1.6701984531034517, + "language_loss": 0.74030113, + "learning_rate": 3.5730541916425127e-06, + "loss": 0.76252508, + "num_input_tokens_seen": 41777550, + "step": 1959, + "time_per_iteration": 2.7659995555877686 + }, + { + "auxiliary_loss_clip": 0.01191206, + "auxiliary_loss_mlp": 0.01036422, + "balance_loss_clip": 1.05844367, + "balance_loss_mlp": 1.02475703, + "epoch": 0.23567606565261828, + "flos": 21944795748480.0, + "grad_norm": 1.9642789223241834, + "language_loss": 0.85941672, + "learning_rate": 3.572573013904639e-06, + "loss": 0.88169301, + "num_input_tokens_seen": 41797460, + "step": 1960, + "time_per_iteration": 2.8487513065338135 + }, + { + "auxiliary_loss_clip": 0.01218347, + "auxiliary_loss_mlp": 0.01041502, + "balance_loss_clip": 1.06303692, + "balance_loss_mlp": 1.02977753, + "epoch": 0.2357963085432574, + "flos": 13589352639360.0, + "grad_norm": 1.8901364234869176, + "language_loss": 0.92158866, + "learning_rate": 3.572091597608689e-06, + "loss": 0.9441871, + "num_input_tokens_seen": 41815585, + "step": 1961, + "time_per_iteration": 2.765601396560669 + }, + { + "auxiliary_loss_clip": 0.01199172, + "auxiliary_loss_mlp": 0.01037244, + "balance_loss_clip": 1.05787206, + "balance_loss_mlp": 1.02540648, + "epoch": 0.23591655143389648, + "flos": 22088689632000.0, + "grad_norm": 2.0252818216960202, + "language_loss": 0.73625052, + "learning_rate": 3.571609942827694e-06, + "loss": 0.75861466, + "num_input_tokens_seen": 41834700, + "step": 1962, + "time_per_iteration": 2.863180160522461 + }, + { + "auxiliary_loss_clip": 0.01193959, + "auxiliary_loss_mlp": 0.01034228, + "balance_loss_clip": 1.05671191, + "balance_loss_mlp": 1.02325499, + "epoch": 0.23603679432453556, + "flos": 17017349057280.0, + "grad_norm": 1.970120352724409, + "language_loss": 0.88750851, + "learning_rate": 3.57112804963472e-06, + "loss": 0.9097904, + "num_input_tokens_seen": 41852915, + "step": 1963, + "time_per_iteration": 2.742598056793213 + }, + { + "auxiliary_loss_clip": 0.01180134, + "auxiliary_loss_mlp": 0.01034606, + "balance_loss_clip": 1.05943382, + "balance_loss_mlp": 1.02422833, + "epoch": 0.23615703721517464, + "flos": 19171307001600.0, + "grad_norm": 1.723517119847139, + "language_loss": 0.7688607, + "learning_rate": 3.57064591810287e-06, + "loss": 0.79100811, + "num_input_tokens_seen": 41870415, + "step": 1964, + "time_per_iteration": 2.8049116134643555 + }, + { + "auxiliary_loss_clip": 0.01217097, + "auxiliary_loss_mlp": 0.00888026, + "balance_loss_clip": 1.06370425, + "balance_loss_mlp": 1.00059164, + "epoch": 0.23627728010581375, + "flos": 19098803399040.0, + "grad_norm": 2.6594020631391966, + "language_loss": 0.80831158, + "learning_rate": 3.570163548305284e-06, + "loss": 0.82936287, + "num_input_tokens_seen": 41889345, + "step": 1965, + "time_per_iteration": 2.7513349056243896 + }, + { + "auxiliary_loss_clip": 0.01196447, + "auxiliary_loss_mlp": 0.01039279, + "balance_loss_clip": 1.05932117, + "balance_loss_mlp": 1.02788901, + "epoch": 0.23639752299645284, + "flos": 14282213057280.0, + "grad_norm": 2.2056040535322454, + "language_loss": 0.70111609, + "learning_rate": 3.569680940315135e-06, + "loss": 0.72347337, + "num_input_tokens_seen": 41905745, + "step": 1966, + "time_per_iteration": 3.612006664276123 + }, + { + "auxiliary_loss_clip": 0.01190164, + "auxiliary_loss_mlp": 0.01034704, + "balance_loss_clip": 1.05581307, + "balance_loss_mlp": 1.02327752, + "epoch": 0.23651776588709192, + "flos": 22893411980160.0, + "grad_norm": 1.66989067100751, + "language_loss": 0.82143557, + "learning_rate": 3.5691980942056356e-06, + "loss": 0.8436842, + "num_input_tokens_seen": 41925115, + "step": 1967, + "time_per_iteration": 5.6547253131866455 + }, + { + "auxiliary_loss_clip": 0.01209741, + "auxiliary_loss_mlp": 0.01035463, + "balance_loss_clip": 1.05903673, + "balance_loss_mlp": 1.02436447, + "epoch": 0.23663800877773103, + "flos": 18624531196800.0, + "grad_norm": 1.8305400496685056, + "language_loss": 0.79450536, + "learning_rate": 3.5687150100500332e-06, + "loss": 0.81695747, + "num_input_tokens_seen": 41944815, + "step": 1968, + "time_per_iteration": 2.683182716369629 + }, + { + "auxiliary_loss_clip": 0.01209614, + "auxiliary_loss_mlp": 0.01043442, + "balance_loss_clip": 1.06123936, + "balance_loss_mlp": 1.0317359, + "epoch": 0.2367582516683701, + "flos": 25555828896000.0, + "grad_norm": 1.6758077194727528, + "language_loss": 0.7432313, + "learning_rate": 3.568231687921611e-06, + "loss": 0.76576185, + "num_input_tokens_seen": 41964990, + "step": 1969, + "time_per_iteration": 2.7833971977233887 + }, + { + "auxiliary_loss_clip": 0.01213365, + "auxiliary_loss_mlp": 0.01038579, + "balance_loss_clip": 1.05982614, + "balance_loss_mlp": 1.02792716, + "epoch": 0.2368784945590092, + "flos": 23295072839040.0, + "grad_norm": 1.5904440461868228, + "language_loss": 0.80328214, + "learning_rate": 3.5677481278936883e-06, + "loss": 0.82580161, + "num_input_tokens_seen": 41984570, + "step": 1970, + "time_per_iteration": 2.6751434803009033 + }, + { + "auxiliary_loss_clip": 0.0109008, + "auxiliary_loss_mlp": 0.01008292, + "balance_loss_clip": 1.02739859, + "balance_loss_mlp": 1.00431061, + "epoch": 0.23699873744964828, + "flos": 69859291875840.0, + "grad_norm": 0.8259594964500366, + "language_loss": 0.57864141, + "learning_rate": 3.5672643300396214e-06, + "loss": 0.59962511, + "num_input_tokens_seen": 42053715, + "step": 1971, + "time_per_iteration": 3.377883195877075 + }, + { + "auxiliary_loss_clip": 0.01186304, + "auxiliary_loss_mlp": 0.01041412, + "balance_loss_clip": 1.05785668, + "balance_loss_mlp": 1.03028941, + "epoch": 0.2371189803402874, + "flos": 21835052720640.0, + "grad_norm": 2.339499898912022, + "language_loss": 0.67490375, + "learning_rate": 3.566780294432802e-06, + "loss": 0.69718087, + "num_input_tokens_seen": 42070890, + "step": 1972, + "time_per_iteration": 2.839271306991577 + }, + { + "auxiliary_loss_clip": 0.01218765, + "auxiliary_loss_mlp": 0.01047041, + "balance_loss_clip": 1.06389201, + "balance_loss_mlp": 1.03543591, + "epoch": 0.23723922323092647, + "flos": 21908490076800.0, + "grad_norm": 3.0305397880745515, + "language_loss": 0.74780381, + "learning_rate": 3.566296021146657e-06, + "loss": 0.77046192, + "num_input_tokens_seen": 42090270, + "step": 1973, + "time_per_iteration": 2.6835503578186035 + }, + { + "auxiliary_loss_clip": 0.0121939, + "auxiliary_loss_mlp": 0.01037227, + "balance_loss_clip": 1.06289864, + "balance_loss_mlp": 1.02614689, + "epoch": 0.23735946612156555, + "flos": 32708803380480.0, + "grad_norm": 1.8170903680192967, + "language_loss": 0.73392254, + "learning_rate": 3.565811510254652e-06, + "loss": 0.75648868, + "num_input_tokens_seen": 42111150, + "step": 1974, + "time_per_iteration": 2.7853903770446777 + }, + { + "auxiliary_loss_clip": 0.01104465, + "auxiliary_loss_mlp": 0.01011319, + "balance_loss_clip": 1.03613746, + "balance_loss_mlp": 1.00793302, + "epoch": 0.23747970901220466, + "flos": 70546944821760.0, + "grad_norm": 0.8340303057505242, + "language_loss": 0.58238733, + "learning_rate": 3.5653267618302845e-06, + "loss": 0.60354507, + "num_input_tokens_seen": 42178730, + "step": 1975, + "time_per_iteration": 3.320068359375 + }, + { + "auxiliary_loss_clip": 0.01215923, + "auxiliary_loss_mlp": 0.01045953, + "balance_loss_clip": 1.06081414, + "balance_loss_mlp": 1.03500354, + "epoch": 0.23759995190284375, + "flos": 20849807594880.0, + "grad_norm": 1.7980575871636664, + "language_loss": 0.86138606, + "learning_rate": 3.564841775947093e-06, + "loss": 0.88400483, + "num_input_tokens_seen": 42199620, + "step": 1976, + "time_per_iteration": 2.7124314308166504 + }, + { + "auxiliary_loss_clip": 0.01183921, + "auxiliary_loss_mlp": 0.01045599, + "balance_loss_clip": 1.0542376, + "balance_loss_mlp": 1.03415489, + "epoch": 0.23772019479348283, + "flos": 32921645420160.0, + "grad_norm": 2.032171192601171, + "language_loss": 0.75685704, + "learning_rate": 3.5643565526786475e-06, + "loss": 0.77915221, + "num_input_tokens_seen": 42219560, + "step": 1977, + "time_per_iteration": 2.8688645362854004 + }, + { + "auxiliary_loss_clip": 0.01219144, + "auxiliary_loss_mlp": 0.01036846, + "balance_loss_clip": 1.06294322, + "balance_loss_mlp": 1.02547908, + "epoch": 0.2378404376841219, + "flos": 32342765834880.0, + "grad_norm": 1.6626301733293607, + "language_loss": 0.77271163, + "learning_rate": 3.5638710920985574e-06, + "loss": 0.79527158, + "num_input_tokens_seen": 42241020, + "step": 1978, + "time_per_iteration": 2.7018513679504395 + }, + { + "auxiliary_loss_clip": 0.01210801, + "auxiliary_loss_mlp": 0.00889185, + "balance_loss_clip": 1.05758715, + "balance_loss_mlp": 1.00054967, + "epoch": 0.23796068057476102, + "flos": 22997624313600.0, + "grad_norm": 2.2826226286525118, + "language_loss": 0.8232094, + "learning_rate": 3.5633853942804655e-06, + "loss": 0.84420919, + "num_input_tokens_seen": 42259345, + "step": 1979, + "time_per_iteration": 2.725795030593872 + }, + { + "auxiliary_loss_clip": 0.0118236, + "auxiliary_loss_mlp": 0.01045828, + "balance_loss_clip": 1.0513618, + "balance_loss_mlp": 1.03364444, + "epoch": 0.2380809234654001, + "flos": 13480938414720.0, + "grad_norm": 1.8858889315684817, + "language_loss": 0.76931286, + "learning_rate": 3.5628994592980527e-06, + "loss": 0.79159474, + "num_input_tokens_seen": 42277250, + "step": 1980, + "time_per_iteration": 2.857174873352051 + }, + { + "auxiliary_loss_clip": 0.01217693, + "auxiliary_loss_mlp": 0.01039752, + "balance_loss_clip": 1.06184363, + "balance_loss_mlp": 1.02806973, + "epoch": 0.2382011663560392, + "flos": 16871803148160.0, + "grad_norm": 1.6284571828017884, + "language_loss": 0.7033844, + "learning_rate": 3.562413287225034e-06, + "loss": 0.72595882, + "num_input_tokens_seen": 42295360, + "step": 1981, + "time_per_iteration": 2.711745262145996 + }, + { + "auxiliary_loss_clip": 0.01201369, + "auxiliary_loss_mlp": 0.01041968, + "balance_loss_clip": 1.06011927, + "balance_loss_mlp": 1.0308156, + "epoch": 0.2383214092466783, + "flos": 18441135331200.0, + "grad_norm": 2.6368974342862024, + "language_loss": 0.89357579, + "learning_rate": 3.5619268781351623e-06, + "loss": 0.91600913, + "num_input_tokens_seen": 42313430, + "step": 1982, + "time_per_iteration": 2.8342654705047607 + }, + { + "auxiliary_loss_clip": 0.01191161, + "auxiliary_loss_mlp": 0.01033106, + "balance_loss_clip": 1.06052279, + "balance_loss_mlp": 1.02161419, + "epoch": 0.23844165213731738, + "flos": 19755717281280.0, + "grad_norm": 1.7193737817645391, + "language_loss": 0.76935041, + "learning_rate": 3.5614402321022256e-06, + "loss": 0.79159302, + "num_input_tokens_seen": 42331260, + "step": 1983, + "time_per_iteration": 2.846759796142578 + }, + { + "auxiliary_loss_clip": 0.01164553, + "auxiliary_loss_mlp": 0.01034196, + "balance_loss_clip": 1.05268967, + "balance_loss_mlp": 1.02306771, + "epoch": 0.23856189502795647, + "flos": 23367360960000.0, + "grad_norm": 1.9222272076724278, + "language_loss": 0.871589, + "learning_rate": 3.5609533492000463e-06, + "loss": 0.8935765, + "num_input_tokens_seen": 42350150, + "step": 1984, + "time_per_iteration": 2.8772501945495605 + }, + { + "auxiliary_loss_clip": 0.01191638, + "auxiliary_loss_mlp": 0.0104085, + "balance_loss_clip": 1.0606432, + "balance_loss_mlp": 1.02911353, + "epoch": 0.23868213791859555, + "flos": 23475056912640.0, + "grad_norm": 1.9300484144023782, + "language_loss": 0.78581023, + "learning_rate": 3.560466229502485e-06, + "loss": 0.80813515, + "num_input_tokens_seen": 42369495, + "step": 1985, + "time_per_iteration": 2.8807179927825928 + }, + { + "auxiliary_loss_clip": 0.01198966, + "auxiliary_loss_mlp": 0.00888719, + "balance_loss_clip": 1.06424665, + "balance_loss_mlp": 1.00056744, + "epoch": 0.23880238080923466, + "flos": 16617340224000.0, + "grad_norm": 2.1225604858068787, + "language_loss": 0.89694762, + "learning_rate": 3.5599788730834384e-06, + "loss": 0.91782445, + "num_input_tokens_seen": 42387455, + "step": 1986, + "time_per_iteration": 2.8166663646698 + }, + { + "auxiliary_loss_clip": 0.01211083, + "auxiliary_loss_mlp": 0.01038193, + "balance_loss_clip": 1.06059754, + "balance_loss_mlp": 1.02685571, + "epoch": 0.23892262369987374, + "flos": 17348409734400.0, + "grad_norm": 2.5242701317160527, + "language_loss": 0.78896379, + "learning_rate": 3.559491280016836e-06, + "loss": 0.81145656, + "num_input_tokens_seen": 42405400, + "step": 1987, + "time_per_iteration": 2.834446430206299 + }, + { + "auxiliary_loss_clip": 0.01199976, + "auxiliary_loss_mlp": 0.01037345, + "balance_loss_clip": 1.0609014, + "balance_loss_mlp": 1.02601421, + "epoch": 0.23904286659051283, + "flos": 22309899540480.0, + "grad_norm": 1.728436914195099, + "language_loss": 0.70797968, + "learning_rate": 3.5590034503766465e-06, + "loss": 0.73035288, + "num_input_tokens_seen": 42425065, + "step": 1988, + "time_per_iteration": 2.8752613067626953 + }, + { + "auxiliary_loss_clip": 0.01218029, + "auxiliary_loss_mlp": 0.0103375, + "balance_loss_clip": 1.06309664, + "balance_loss_mlp": 1.02318847, + "epoch": 0.23916310948115194, + "flos": 21178246579200.0, + "grad_norm": 2.203265976589439, + "language_loss": 0.81482911, + "learning_rate": 3.558515384236874e-06, + "loss": 0.83734685, + "num_input_tokens_seen": 42442495, + "step": 1989, + "time_per_iteration": 2.668433904647827 + }, + { + "auxiliary_loss_clip": 0.01177732, + "auxiliary_loss_mlp": 0.00888422, + "balance_loss_clip": 1.05643392, + "balance_loss_mlp": 1.00063801, + "epoch": 0.23928335237179102, + "flos": 14137349506560.0, + "grad_norm": 2.0264348357344075, + "language_loss": 0.8406173, + "learning_rate": 3.558027081671556e-06, + "loss": 0.86127889, + "num_input_tokens_seen": 42459480, + "step": 1990, + "time_per_iteration": 2.8316001892089844 + }, + { + "auxiliary_loss_clip": 0.01210122, + "auxiliary_loss_mlp": 0.01035796, + "balance_loss_clip": 1.06092608, + "balance_loss_mlp": 1.02453613, + "epoch": 0.2394035952624301, + "flos": 23769596436480.0, + "grad_norm": 2.6558738660132306, + "language_loss": 0.68877637, + "learning_rate": 3.557538542754769e-06, + "loss": 0.71123552, + "num_input_tokens_seen": 42479175, + "step": 1991, + "time_per_iteration": 2.687642812728882 + }, + { + "auxiliary_loss_clip": 0.01219113, + "auxiliary_loss_mlp": 0.01045264, + "balance_loss_clip": 1.064641, + "balance_loss_mlp": 1.03298545, + "epoch": 0.2395238381530692, + "flos": 24206198250240.0, + "grad_norm": 2.17316558962457, + "language_loss": 0.66852939, + "learning_rate": 3.557049767560623e-06, + "loss": 0.69117314, + "num_input_tokens_seen": 42498090, + "step": 1992, + "time_per_iteration": 4.45845365524292 + }, + { + "auxiliary_loss_clip": 0.01182509, + "auxiliary_loss_mlp": 0.01041098, + "balance_loss_clip": 1.0600009, + "balance_loss_mlp": 1.02961826, + "epoch": 0.2396440810437083, + "flos": 25295763450240.0, + "grad_norm": 4.388409566850196, + "language_loss": 0.85918403, + "learning_rate": 3.5565607561632655e-06, + "loss": 0.88142014, + "num_input_tokens_seen": 42516930, + "step": 1993, + "time_per_iteration": 3.954270124435425 + }, + { + "auxiliary_loss_clip": 0.01192384, + "auxiliary_loss_mlp": 0.01040231, + "balance_loss_clip": 1.05896688, + "balance_loss_mlp": 1.02859664, + "epoch": 0.23976432393434738, + "flos": 28543093436160.0, + "grad_norm": 2.4096274606926693, + "language_loss": 0.79375374, + "learning_rate": 3.5560715086368787e-06, + "loss": 0.81607985, + "num_input_tokens_seen": 42534800, + "step": 1994, + "time_per_iteration": 3.7257332801818848 + }, + { + "auxiliary_loss_clip": 0.01193109, + "auxiliary_loss_mlp": 0.01038112, + "balance_loss_clip": 1.06317723, + "balance_loss_mlp": 1.02707314, + "epoch": 0.23988456682498646, + "flos": 19494358945920.0, + "grad_norm": 2.036597896608217, + "language_loss": 0.82369846, + "learning_rate": 3.5555820250556816e-06, + "loss": 0.84601068, + "num_input_tokens_seen": 42552000, + "step": 1995, + "time_per_iteration": 2.7198634147644043 + }, + { + "auxiliary_loss_clip": 0.0119994, + "auxiliary_loss_mlp": 0.01035342, + "balance_loss_clip": 1.06046653, + "balance_loss_mlp": 1.02323639, + "epoch": 0.24000480971562557, + "flos": 20266331068800.0, + "grad_norm": 2.4492480439057664, + "language_loss": 0.69383878, + "learning_rate": 3.5550923054939278e-06, + "loss": 0.71619159, + "num_input_tokens_seen": 42571455, + "step": 1996, + "time_per_iteration": 2.7988617420196533 + }, + { + "auxiliary_loss_clip": 0.0116821, + "auxiliary_loss_mlp": 0.01039201, + "balance_loss_clip": 1.05408096, + "balance_loss_mlp": 1.02757215, + "epoch": 0.24012505260626466, + "flos": 25443176866560.0, + "grad_norm": 1.7944766147440208, + "language_loss": 0.74431682, + "learning_rate": 3.5546023500259083e-06, + "loss": 0.76639092, + "num_input_tokens_seen": 42592550, + "step": 1997, + "time_per_iteration": 2.848118305206299 + }, + { + "auxiliary_loss_clip": 0.01182527, + "auxiliary_loss_mlp": 0.01044732, + "balance_loss_clip": 1.05649877, + "balance_loss_mlp": 1.03357983, + "epoch": 0.24024529549690374, + "flos": 15553342529280.0, + "grad_norm": 1.9498309877014217, + "language_loss": 0.80739808, + "learning_rate": 3.5541121587259477e-06, + "loss": 0.82967073, + "num_input_tokens_seen": 42610385, + "step": 1998, + "time_per_iteration": 2.8387954235076904 + }, + { + "auxiliary_loss_clip": 0.01098727, + "auxiliary_loss_mlp": 0.010128, + "balance_loss_clip": 1.02782869, + "balance_loss_mlp": 1.00898552, + "epoch": 0.24036553838754285, + "flos": 57122351867520.0, + "grad_norm": 0.8306025670447913, + "language_loss": 0.57841599, + "learning_rate": 3.553621731668408e-06, + "loss": 0.59953129, + "num_input_tokens_seen": 42673595, + "step": 1999, + "time_per_iteration": 3.285933494567871 + }, + { + "auxiliary_loss_clip": 0.01198708, + "auxiliary_loss_mlp": 0.01047174, + "balance_loss_clip": 1.0572772, + "balance_loss_mlp": 1.03517556, + "epoch": 0.24048578127818193, + "flos": 24969946158720.0, + "grad_norm": 1.6937528645277333, + "language_loss": 0.83223826, + "learning_rate": 3.553131068927688e-06, + "loss": 0.85469711, + "num_input_tokens_seen": 42692000, + "step": 2000, + "time_per_iteration": 2.7545456886291504 + }, + { + "auxiliary_loss_clip": 0.01188724, + "auxiliary_loss_mlp": 0.01041905, + "balance_loss_clip": 1.06125546, + "balance_loss_mlp": 1.03134882, + "epoch": 0.24060602416882101, + "flos": 23330947547520.0, + "grad_norm": 2.046459041731497, + "language_loss": 0.80521196, + "learning_rate": 3.552640170578219e-06, + "loss": 0.82751822, + "num_input_tokens_seen": 42712250, + "step": 2001, + "time_per_iteration": 2.858257532119751 + }, + { + "auxiliary_loss_clip": 0.01196363, + "auxiliary_loss_mlp": 0.01037265, + "balance_loss_clip": 1.05897999, + "balance_loss_mlp": 1.026178, + "epoch": 0.2407262670594601, + "flos": 14173260128640.0, + "grad_norm": 2.193159545382057, + "language_loss": 0.77781481, + "learning_rate": 3.5521490366944703e-06, + "loss": 0.80015105, + "num_input_tokens_seen": 42729900, + "step": 2002, + "time_per_iteration": 2.772122383117676 + }, + { + "auxiliary_loss_clip": 0.01182312, + "auxiliary_loss_mlp": 0.01037083, + "balance_loss_clip": 1.05563223, + "balance_loss_mlp": 1.02631831, + "epoch": 0.2408465099500992, + "flos": 13663113217920.0, + "grad_norm": 5.891241569999727, + "language_loss": 0.80199659, + "learning_rate": 3.5516576673509474e-06, + "loss": 0.8241905, + "num_input_tokens_seen": 42747900, + "step": 2003, + "time_per_iteration": 2.8767852783203125 + }, + { + "auxiliary_loss_clip": 0.01217147, + "auxiliary_loss_mlp": 0.01039553, + "balance_loss_clip": 1.06264615, + "balance_loss_mlp": 1.02821589, + "epoch": 0.2409667528407383, + "flos": 31248029076480.0, + "grad_norm": 1.881624700859451, + "language_loss": 0.86656785, + "learning_rate": 3.5511660626221896e-06, + "loss": 0.88913482, + "num_input_tokens_seen": 42768540, + "step": 2004, + "time_per_iteration": 2.7213926315307617 + }, + { + "auxiliary_loss_clip": 0.01196604, + "auxiliary_loss_mlp": 0.00889126, + "balance_loss_clip": 1.06167948, + "balance_loss_mlp": 1.0005753, + "epoch": 0.24108699573137737, + "flos": 22199941031040.0, + "grad_norm": 2.1756502307479364, + "language_loss": 0.89277667, + "learning_rate": 3.5506742225827744e-06, + "loss": 0.91363394, + "num_input_tokens_seen": 42785395, + "step": 2005, + "time_per_iteration": 2.7939095497131348 + }, + { + "auxiliary_loss_clip": 0.01188581, + "auxiliary_loss_mlp": 0.01039063, + "balance_loss_clip": 1.05987918, + "balance_loss_mlp": 1.02791739, + "epoch": 0.24120723862201648, + "flos": 26103035664000.0, + "grad_norm": 2.02077469756017, + "language_loss": 0.90560508, + "learning_rate": 3.5501821473073116e-06, + "loss": 0.92788148, + "num_input_tokens_seen": 42801980, + "step": 2006, + "time_per_iteration": 2.8145453929901123 + }, + { + "auxiliary_loss_clip": 0.01180079, + "auxiliary_loss_mlp": 0.01045775, + "balance_loss_clip": 1.05905259, + "balance_loss_mlp": 1.03359747, + "epoch": 0.24132748151265557, + "flos": 18624926246400.0, + "grad_norm": 2.2763396642833253, + "language_loss": 0.8635999, + "learning_rate": 3.54968983687045e-06, + "loss": 0.88585842, + "num_input_tokens_seen": 42818850, + "step": 2007, + "time_per_iteration": 2.8444814682006836 + }, + { + "auxiliary_loss_clip": 0.01201109, + "auxiliary_loss_mlp": 0.01042535, + "balance_loss_clip": 1.06154132, + "balance_loss_mlp": 1.03098404, + "epoch": 0.24144772440329465, + "flos": 15267673664640.0, + "grad_norm": 2.321477098244696, + "language_loss": 0.89563465, + "learning_rate": 3.549197291346872e-06, + "loss": 0.91807109, + "num_input_tokens_seen": 42835375, + "step": 2008, + "time_per_iteration": 2.791738271713257 + }, + { + "auxiliary_loss_clip": 0.01205186, + "auxiliary_loss_mlp": 0.01035048, + "balance_loss_clip": 1.0577662, + "balance_loss_mlp": 1.02329397, + "epoch": 0.24156796729393373, + "flos": 24024274842240.0, + "grad_norm": 2.689829098053332, + "language_loss": 0.79264522, + "learning_rate": 3.548704510811297e-06, + "loss": 0.81504756, + "num_input_tokens_seen": 42854570, + "step": 2009, + "time_per_iteration": 2.7461180686950684 + }, + { + "auxiliary_loss_clip": 0.0118125, + "auxiliary_loss_mlp": 0.01037417, + "balance_loss_clip": 1.05623722, + "balance_loss_mlp": 1.0258652, + "epoch": 0.24168821018457284, + "flos": 26286790665600.0, + "grad_norm": 2.77821663544308, + "language_loss": 0.74551666, + "learning_rate": 3.5482114953384787e-06, + "loss": 0.76770329, + "num_input_tokens_seen": 42873800, + "step": 2010, + "time_per_iteration": 2.912898302078247 + }, + { + "auxiliary_loss_clip": 0.01209387, + "auxiliary_loss_mlp": 0.01040481, + "balance_loss_clip": 1.06110954, + "balance_loss_mlp": 1.02953148, + "epoch": 0.24180845307521193, + "flos": 18223193560320.0, + "grad_norm": 1.9880159114331022, + "language_loss": 0.84073257, + "learning_rate": 3.5477182450032077e-06, + "loss": 0.86323118, + "num_input_tokens_seen": 42892400, + "step": 2011, + "time_per_iteration": 2.669579267501831 + }, + { + "auxiliary_loss_clip": 0.01205214, + "auxiliary_loss_mlp": 0.01039318, + "balance_loss_clip": 1.0608902, + "balance_loss_mlp": 1.02796912, + "epoch": 0.241928695965851, + "flos": 20449260057600.0, + "grad_norm": 1.965961273777642, + "language_loss": 0.83313823, + "learning_rate": 3.5472247598803097e-06, + "loss": 0.85558355, + "num_input_tokens_seen": 42911745, + "step": 2012, + "time_per_iteration": 2.810469150543213 + }, + { + "auxiliary_loss_clip": 0.01217095, + "auxiliary_loss_mlp": 0.01037771, + "balance_loss_clip": 1.0619669, + "balance_loss_mlp": 1.02608275, + "epoch": 0.24204893885649012, + "flos": 25556475340800.0, + "grad_norm": 2.0898357541370594, + "language_loss": 0.85198003, + "learning_rate": 3.546731040044645e-06, + "loss": 0.87452865, + "num_input_tokens_seen": 42926915, + "step": 2013, + "time_per_iteration": 2.731583833694458 + }, + { + "auxiliary_loss_clip": 0.01215247, + "auxiliary_loss_mlp": 0.01036808, + "balance_loss_clip": 1.06218362, + "balance_loss_mlp": 1.02593052, + "epoch": 0.2421691817471292, + "flos": 30660207004800.0, + "grad_norm": 1.7780292763986205, + "language_loss": 0.75112873, + "learning_rate": 3.546237085571112e-06, + "loss": 0.77364922, + "num_input_tokens_seen": 42945350, + "step": 2014, + "time_per_iteration": 2.805727958679199 + }, + { + "auxiliary_loss_clip": 0.01207136, + "auxiliary_loss_mlp": 0.01035501, + "balance_loss_clip": 1.06142426, + "balance_loss_mlp": 1.0241642, + "epoch": 0.24228942463776829, + "flos": 21945011230080.0, + "grad_norm": 3.316471277867622, + "language_loss": 0.72588801, + "learning_rate": 3.5457428965346425e-06, + "loss": 0.74831438, + "num_input_tokens_seen": 42964290, + "step": 2015, + "time_per_iteration": 2.769805669784546 + }, + { + "auxiliary_loss_clip": 0.01162542, + "auxiliary_loss_mlp": 0.01042665, + "balance_loss_clip": 1.05299687, + "balance_loss_mlp": 1.03163791, + "epoch": 0.2424096675284074, + "flos": 33984493879680.0, + "grad_norm": 1.567015112908097, + "language_loss": 0.74978852, + "learning_rate": 3.545248473010205e-06, + "loss": 0.77184057, + "num_input_tokens_seen": 42987095, + "step": 2016, + "time_per_iteration": 3.016305923461914 + }, + { + "auxiliary_loss_clip": 0.01218126, + "auxiliary_loss_mlp": 0.00889054, + "balance_loss_clip": 1.06222296, + "balance_loss_mlp": 1.00059915, + "epoch": 0.24252991041904648, + "flos": 21653416621440.0, + "grad_norm": 1.8752220689825256, + "language_loss": 0.87920231, + "learning_rate": 3.544753815072802e-06, + "loss": 0.90027416, + "num_input_tokens_seen": 43005750, + "step": 2017, + "time_per_iteration": 2.692795515060425 + }, + { + "auxiliary_loss_clip": 0.01147935, + "auxiliary_loss_mlp": 0.01045765, + "balance_loss_clip": 1.0491693, + "balance_loss_mlp": 1.03413582, + "epoch": 0.24265015330968556, + "flos": 21870065502720.0, + "grad_norm": 1.8671428795553757, + "language_loss": 0.88587397, + "learning_rate": 3.544258922797474e-06, + "loss": 0.90781093, + "num_input_tokens_seen": 43023870, + "step": 2018, + "time_per_iteration": 4.713507175445557 + }, + { + "auxiliary_loss_clip": 0.01215528, + "auxiliary_loss_mlp": 0.01040044, + "balance_loss_clip": 1.06311584, + "balance_loss_mlp": 1.02969074, + "epoch": 0.24277039620032465, + "flos": 25628260671360.0, + "grad_norm": 1.6027838939421177, + "language_loss": 0.78095937, + "learning_rate": 3.543763796259295e-06, + "loss": 0.8035152, + "num_input_tokens_seen": 43043825, + "step": 2019, + "time_per_iteration": 2.711106061935425 + }, + { + "auxiliary_loss_clip": 0.01205141, + "auxiliary_loss_mlp": 0.01035481, + "balance_loss_clip": 1.05882955, + "balance_loss_mlp": 1.02430522, + "epoch": 0.24289063909096376, + "flos": 26286575184000.0, + "grad_norm": 1.788646735135168, + "language_loss": 0.90871269, + "learning_rate": 3.5432684355333754e-06, + "loss": 0.93111891, + "num_input_tokens_seen": 43062480, + "step": 2020, + "time_per_iteration": 4.703521251678467 + }, + { + "auxiliary_loss_clip": 0.01203282, + "auxiliary_loss_mlp": 0.01036305, + "balance_loss_clip": 1.05837548, + "balance_loss_mlp": 1.02515888, + "epoch": 0.24301088198160284, + "flos": 25075056332160.0, + "grad_norm": 1.8619995812002574, + "language_loss": 0.76904017, + "learning_rate": 3.5427728406948613e-06, + "loss": 0.79143608, + "num_input_tokens_seen": 43081595, + "step": 2021, + "time_per_iteration": 2.7574615478515625 + }, + { + "auxiliary_loss_clip": 0.01089687, + "auxiliary_loss_mlp": 0.01004212, + "balance_loss_clip": 1.02135587, + "balance_loss_mlp": 1.00044537, + "epoch": 0.24313112487224192, + "flos": 69900948673920.0, + "grad_norm": 0.757768173987224, + "language_loss": 0.57920182, + "learning_rate": 3.542277011818934e-06, + "loss": 0.60014081, + "num_input_tokens_seen": 43145430, + "step": 2022, + "time_per_iteration": 3.4646048545837402 + }, + { + "auxiliary_loss_clip": 0.01194759, + "auxiliary_loss_mlp": 0.01032868, + "balance_loss_clip": 1.05972564, + "balance_loss_mlp": 1.02252054, + "epoch": 0.24325136776288103, + "flos": 40662334235520.0, + "grad_norm": 2.2504950466897595, + "language_loss": 0.74402422, + "learning_rate": 3.5417809489808104e-06, + "loss": 0.7663005, + "num_input_tokens_seen": 43167040, + "step": 2023, + "time_per_iteration": 2.9063069820404053 + }, + { + "auxiliary_loss_clip": 0.01208512, + "auxiliary_loss_mlp": 0.0104168, + "balance_loss_clip": 1.06152463, + "balance_loss_mlp": 1.03104663, + "epoch": 0.24337161065352012, + "flos": 25046400257280.0, + "grad_norm": 1.7930768937256434, + "language_loss": 0.72413111, + "learning_rate": 3.5412846522557422e-06, + "loss": 0.74663305, + "num_input_tokens_seen": 43187930, + "step": 2024, + "time_per_iteration": 2.766716957092285 + }, + { + "auxiliary_loss_clip": 0.01221195, + "auxiliary_loss_mlp": 0.0104176, + "balance_loss_clip": 1.06582868, + "balance_loss_mlp": 1.03088272, + "epoch": 0.2434918535441592, + "flos": 18661160090880.0, + "grad_norm": 2.4856742811866326, + "language_loss": 0.74303186, + "learning_rate": 3.540788121719018e-06, + "loss": 0.76566136, + "num_input_tokens_seen": 43206350, + "step": 2025, + "time_per_iteration": 2.6772572994232178 + }, + { + "auxiliary_loss_clip": 0.01177744, + "auxiliary_loss_mlp": 0.01042757, + "balance_loss_clip": 1.05846357, + "balance_loss_mlp": 1.03138435, + "epoch": 0.24361209643479828, + "flos": 23915142345600.0, + "grad_norm": 1.8919480013720484, + "language_loss": 0.82095724, + "learning_rate": 3.5402913574459604e-06, + "loss": 0.84316224, + "num_input_tokens_seen": 43226255, + "step": 2026, + "time_per_iteration": 2.7535195350646973 + }, + { + "auxiliary_loss_clip": 0.01152084, + "auxiliary_loss_mlp": 0.01033645, + "balance_loss_clip": 1.04454637, + "balance_loss_mlp": 1.02285707, + "epoch": 0.2437323393254374, + "flos": 28657505232000.0, + "grad_norm": 3.3292001793631307, + "language_loss": 0.86152864, + "learning_rate": 3.5397943595119297e-06, + "loss": 0.8833859, + "num_input_tokens_seen": 43247675, + "step": 2027, + "time_per_iteration": 2.9216554164886475 + }, + { + "auxiliary_loss_clip": 0.01189703, + "auxiliary_loss_mlp": 0.01042828, + "balance_loss_clip": 1.05989552, + "balance_loss_mlp": 1.03108013, + "epoch": 0.24385258221607647, + "flos": 23550325862400.0, + "grad_norm": 2.45221763892447, + "language_loss": 0.78031737, + "learning_rate": 3.5392971279923177e-06, + "loss": 0.8026427, + "num_input_tokens_seen": 43265895, + "step": 2028, + "time_per_iteration": 2.7813069820404053 + }, + { + "auxiliary_loss_clip": 0.01177811, + "auxiliary_loss_mlp": 0.01043198, + "balance_loss_clip": 1.05272484, + "balance_loss_mlp": 1.03131902, + "epoch": 0.24397282510671556, + "flos": 25336091445120.0, + "grad_norm": 5.364874194459594, + "language_loss": 0.83110654, + "learning_rate": 3.5387996629625557e-06, + "loss": 0.85331666, + "num_input_tokens_seen": 43283485, + "step": 2029, + "time_per_iteration": 2.9208083152770996 + }, + { + "auxiliary_loss_clip": 0.01106242, + "auxiliary_loss_mlp": 0.01002853, + "balance_loss_clip": 1.0255692, + "balance_loss_mlp": 0.99932444, + "epoch": 0.24409306799735467, + "flos": 65187421430400.0, + "grad_norm": 0.7987823640882806, + "language_loss": 0.55015606, + "learning_rate": 3.5383019644981083e-06, + "loss": 0.57124704, + "num_input_tokens_seen": 43347180, + "step": 2030, + "time_per_iteration": 3.2856996059417725 + }, + { + "auxiliary_loss_clip": 0.01195566, + "auxiliary_loss_mlp": 0.01042405, + "balance_loss_clip": 1.05947328, + "balance_loss_mlp": 1.03073406, + "epoch": 0.24421331088799375, + "flos": 19537093152000.0, + "grad_norm": 2.2216961342157933, + "language_loss": 0.72811067, + "learning_rate": 3.5378040326744763e-06, + "loss": 0.75049031, + "num_input_tokens_seen": 43366665, + "step": 2031, + "time_per_iteration": 2.7575671672821045 + }, + { + "auxiliary_loss_clip": 0.01185087, + "auxiliary_loss_mlp": 0.01038624, + "balance_loss_clip": 1.05855036, + "balance_loss_mlp": 1.02822256, + "epoch": 0.24433355377863283, + "flos": 21068575378560.0, + "grad_norm": 2.031204839698082, + "language_loss": 0.85636169, + "learning_rate": 3.5373058675671946e-06, + "loss": 0.87859881, + "num_input_tokens_seen": 43384670, + "step": 2032, + "time_per_iteration": 2.8089330196380615 + }, + { + "auxiliary_loss_clip": 0.01165207, + "auxiliary_loss_mlp": 0.01050533, + "balance_loss_clip": 1.05288959, + "balance_loss_mlp": 1.03907061, + "epoch": 0.24445379666927192, + "flos": 22637189289600.0, + "grad_norm": 2.1258886327292106, + "language_loss": 0.72213286, + "learning_rate": 3.536807469251836e-06, + "loss": 0.74429029, + "num_input_tokens_seen": 43403825, + "step": 2033, + "time_per_iteration": 2.9296000003814697 + }, + { + "auxiliary_loss_clip": 0.01187538, + "auxiliary_loss_mlp": 0.01035414, + "balance_loss_clip": 1.05461156, + "balance_loss_mlp": 1.02400541, + "epoch": 0.24457403955991103, + "flos": 21251612108160.0, + "grad_norm": 1.9254327248001797, + "language_loss": 0.82548511, + "learning_rate": 3.5363088378040055e-06, + "loss": 0.84771466, + "num_input_tokens_seen": 43422715, + "step": 2034, + "time_per_iteration": 2.765515089035034 + }, + { + "auxiliary_loss_clip": 0.01104843, + "auxiliary_loss_mlp": 0.00878217, + "balance_loss_clip": 1.02415919, + "balance_loss_mlp": 1.00146329, + "epoch": 0.2446942824505501, + "flos": 66997820764800.0, + "grad_norm": 0.7797747522534348, + "language_loss": 0.6443224, + "learning_rate": 3.5358099732993463e-06, + "loss": 0.66415298, + "num_input_tokens_seen": 43481825, + "step": 2035, + "time_per_iteration": 3.1714930534362793 + }, + { + "auxiliary_loss_clip": 0.01200529, + "auxiliary_loss_mlp": 0.01040703, + "balance_loss_clip": 1.05882919, + "balance_loss_mlp": 1.02969408, + "epoch": 0.2448145253411892, + "flos": 20411122792320.0, + "grad_norm": 2.0123371300876753, + "language_loss": 0.89613819, + "learning_rate": 3.535310875813535e-06, + "loss": 0.91855049, + "num_input_tokens_seen": 43500220, + "step": 2036, + "time_per_iteration": 2.7814316749572754 + }, + { + "auxiliary_loss_clip": 0.0120338, + "auxiliary_loss_mlp": 0.01030947, + "balance_loss_clip": 1.06008053, + "balance_loss_mlp": 1.01993227, + "epoch": 0.2449347682318283, + "flos": 28804739080320.0, + "grad_norm": 1.6236953384578345, + "language_loss": 0.81941241, + "learning_rate": 3.5348115454222843e-06, + "loss": 0.84175569, + "num_input_tokens_seen": 43522805, + "step": 2037, + "time_per_iteration": 2.9206693172454834 + }, + { + "auxiliary_loss_clip": 0.01190833, + "auxiliary_loss_mlp": 0.01033355, + "balance_loss_clip": 1.05537176, + "balance_loss_mlp": 1.02201235, + "epoch": 0.2450550111224674, + "flos": 22528990546560.0, + "grad_norm": 1.8071906822264705, + "language_loss": 0.86311364, + "learning_rate": 3.5343119822013425e-06, + "loss": 0.88535553, + "num_input_tokens_seen": 43541915, + "step": 2038, + "time_per_iteration": 2.747255802154541 + }, + { + "auxiliary_loss_clip": 0.01209384, + "auxiliary_loss_mlp": 0.01040378, + "balance_loss_clip": 1.05958307, + "balance_loss_mlp": 1.02850461, + "epoch": 0.24517525401310647, + "flos": 21759137326080.0, + "grad_norm": 1.9320909317555526, + "language_loss": 0.77939487, + "learning_rate": 3.533812186226493e-06, + "loss": 0.80189252, + "num_input_tokens_seen": 43562625, + "step": 2039, + "time_per_iteration": 2.721553087234497 + }, + { + "auxiliary_loss_clip": 0.0121229, + "auxiliary_loss_mlp": 0.01033942, + "balance_loss_clip": 1.05979919, + "balance_loss_mlp": 1.02357709, + "epoch": 0.24529549690374555, + "flos": 25043311687680.0, + "grad_norm": 1.8051747890401306, + "language_loss": 0.75765848, + "learning_rate": 3.5333121575735545e-06, + "loss": 0.78012085, + "num_input_tokens_seen": 43582265, + "step": 2040, + "time_per_iteration": 2.6949539184570312 + }, + { + "auxiliary_loss_clip": 0.01191257, + "auxiliary_loss_mlp": 0.01036698, + "balance_loss_clip": 1.05764461, + "balance_loss_mlp": 1.02627873, + "epoch": 0.24541573979438466, + "flos": 32123638915200.0, + "grad_norm": 1.8501274533042356, + "language_loss": 0.75509351, + "learning_rate": 3.532811896318381e-06, + "loss": 0.77737308, + "num_input_tokens_seen": 43604335, + "step": 2041, + "time_per_iteration": 2.854401111602783 + }, + { + "auxiliary_loss_clip": 0.01186396, + "auxiliary_loss_mlp": 0.01033302, + "balance_loss_clip": 1.05513, + "balance_loss_mlp": 1.02118492, + "epoch": 0.24553598268502375, + "flos": 31357556622720.0, + "grad_norm": 2.2379511252801745, + "language_loss": 0.81962037, + "learning_rate": 3.5323114025368615e-06, + "loss": 0.84181738, + "num_input_tokens_seen": 43619400, + "step": 2042, + "time_per_iteration": 2.8804688453674316 + }, + { + "auxiliary_loss_clip": 0.01198839, + "auxiliary_loss_mlp": 0.01034469, + "balance_loss_clip": 1.05572772, + "balance_loss_mlp": 1.02371669, + "epoch": 0.24565622557566283, + "flos": 14027462824320.0, + "grad_norm": 2.113719425846467, + "language_loss": 0.81890428, + "learning_rate": 3.53181067630492e-06, + "loss": 0.84123743, + "num_input_tokens_seen": 43636870, + "step": 2043, + "time_per_iteration": 2.697124481201172 + }, + { + "auxiliary_loss_clip": 0.01185562, + "auxiliary_loss_mlp": 0.01037963, + "balance_loss_clip": 1.05593824, + "balance_loss_mlp": 1.02717447, + "epoch": 0.24577646846630194, + "flos": 16581465515520.0, + "grad_norm": 1.9761518032446057, + "language_loss": 0.7630055, + "learning_rate": 3.5313097176985175e-06, + "loss": 0.78524077, + "num_input_tokens_seen": 43655180, + "step": 2044, + "time_per_iteration": 3.7099215984344482 + }, + { + "auxiliary_loss_clip": 0.01203646, + "auxiliary_loss_mlp": 0.01041337, + "balance_loss_clip": 1.05945957, + "balance_loss_mlp": 1.03045332, + "epoch": 0.24589671135694102, + "flos": 18807424272000.0, + "grad_norm": 1.7486311071045442, + "language_loss": 0.8089844, + "learning_rate": 3.5308085267936482e-06, + "loss": 0.83143413, + "num_input_tokens_seen": 43672895, + "step": 2045, + "time_per_iteration": 5.391521692276001 + }, + { + "auxiliary_loss_clip": 0.01166684, + "auxiliary_loss_mlp": 0.00888289, + "balance_loss_clip": 1.05693865, + "balance_loss_mlp": 1.0006876, + "epoch": 0.2460169542475801, + "flos": 19938538529280.0, + "grad_norm": 3.001838304089692, + "language_loss": 0.89942718, + "learning_rate": 3.530307103666342e-06, + "loss": 0.91997695, + "num_input_tokens_seen": 43691975, + "step": 2046, + "time_per_iteration": 2.957620143890381 + }, + { + "auxiliary_loss_clip": 0.01188401, + "auxiliary_loss_mlp": 0.0103613, + "balance_loss_clip": 1.05759978, + "balance_loss_mlp": 1.02512693, + "epoch": 0.24613719713821922, + "flos": 24171221381760.0, + "grad_norm": 1.764837957423877, + "language_loss": 0.80123675, + "learning_rate": 3.5298054483926658e-06, + "loss": 0.82348204, + "num_input_tokens_seen": 43712670, + "step": 2047, + "time_per_iteration": 2.919177770614624 + }, + { + "auxiliary_loss_clip": 0.01210212, + "auxiliary_loss_mlp": 0.01034273, + "balance_loss_clip": 1.05958176, + "balance_loss_mlp": 1.02272129, + "epoch": 0.2462574400288583, + "flos": 30221055325440.0, + "grad_norm": 1.9870472779374688, + "language_loss": 0.82918507, + "learning_rate": 3.5293035610487187e-06, + "loss": 0.85162991, + "num_input_tokens_seen": 43732035, + "step": 2048, + "time_per_iteration": 2.816668748855591 + }, + { + "auxiliary_loss_clip": 0.01086223, + "auxiliary_loss_mlp": 0.01004787, + "balance_loss_clip": 1.02113509, + "balance_loss_mlp": 1.00101984, + "epoch": 0.24637768291949738, + "flos": 68943030819840.0, + "grad_norm": 0.7204162890128545, + "language_loss": 0.61992908, + "learning_rate": 3.5288014417106374e-06, + "loss": 0.64083916, + "num_input_tokens_seen": 43798055, + "step": 2049, + "time_per_iteration": 3.3877716064453125 + }, + { + "auxiliary_loss_clip": 0.01184307, + "auxiliary_loss_mlp": 0.01034653, + "balance_loss_clip": 1.05900192, + "balance_loss_mlp": 1.02393591, + "epoch": 0.24649792581013646, + "flos": 34383999922560.0, + "grad_norm": 1.7636554668093432, + "language_loss": 0.75124991, + "learning_rate": 3.528299090454593e-06, + "loss": 0.77343953, + "num_input_tokens_seen": 43818590, + "step": 2050, + "time_per_iteration": 2.9644086360931396 + }, + { + "auxiliary_loss_clip": 0.01203084, + "auxiliary_loss_mlp": 0.01033778, + "balance_loss_clip": 1.05616879, + "balance_loss_mlp": 1.02214885, + "epoch": 0.24661816870077558, + "flos": 19680448331520.0, + "grad_norm": 2.878249781310749, + "language_loss": 0.8287468, + "learning_rate": 3.527796507356792e-06, + "loss": 0.85111535, + "num_input_tokens_seen": 43832480, + "step": 2051, + "time_per_iteration": 2.7613790035247803 + }, + { + "auxiliary_loss_clip": 0.01207734, + "auxiliary_loss_mlp": 0.01034341, + "balance_loss_clip": 1.05964816, + "balance_loss_mlp": 1.02380323, + "epoch": 0.24673841159141466, + "flos": 20002279213440.0, + "grad_norm": 2.378921690909703, + "language_loss": 0.90420741, + "learning_rate": 3.527293692493475e-06, + "loss": 0.92662823, + "num_input_tokens_seen": 43848345, + "step": 2052, + "time_per_iteration": 2.681765556335449 + }, + { + "auxiliary_loss_clip": 0.01206308, + "auxiliary_loss_mlp": 0.01036877, + "balance_loss_clip": 1.0585674, + "balance_loss_mlp": 1.02510512, + "epoch": 0.24685865448205374, + "flos": 21646593037440.0, + "grad_norm": 2.524978712097453, + "language_loss": 0.73731661, + "learning_rate": 3.52679064594092e-06, + "loss": 0.75974846, + "num_input_tokens_seen": 43865685, + "step": 2053, + "time_per_iteration": 2.7507431507110596 + }, + { + "auxiliary_loss_clip": 0.01156978, + "auxiliary_loss_mlp": 0.01033739, + "balance_loss_clip": 1.04388666, + "balance_loss_mlp": 1.02318931, + "epoch": 0.24697889737269285, + "flos": 17960470508160.0, + "grad_norm": 2.01189078455523, + "language_loss": 0.74850547, + "learning_rate": 3.5262873677754375e-06, + "loss": 0.77041256, + "num_input_tokens_seen": 43883690, + "step": 2054, + "time_per_iteration": 2.934753656387329 + }, + { + "auxiliary_loss_clip": 0.01213367, + "auxiliary_loss_mlp": 0.01040768, + "balance_loss_clip": 1.06091201, + "balance_loss_mlp": 1.03019953, + "epoch": 0.24709914026333193, + "flos": 27344611221120.0, + "grad_norm": 1.7337707012999226, + "language_loss": 0.80431575, + "learning_rate": 3.5257838580733745e-06, + "loss": 0.82685709, + "num_input_tokens_seen": 43903295, + "step": 2055, + "time_per_iteration": 2.8699285984039307 + }, + { + "auxiliary_loss_clip": 0.01204755, + "auxiliary_loss_mlp": 0.01040072, + "balance_loss_clip": 1.05742669, + "balance_loss_mlp": 1.02949214, + "epoch": 0.24721938315397102, + "flos": 19275519335040.0, + "grad_norm": 2.043257942211775, + "language_loss": 0.87263334, + "learning_rate": 3.5252801169111138e-06, + "loss": 0.89508152, + "num_input_tokens_seen": 43920960, + "step": 2056, + "time_per_iteration": 2.6829214096069336 + }, + { + "auxiliary_loss_clip": 0.01190862, + "auxiliary_loss_mlp": 0.01037327, + "balance_loss_clip": 1.0593009, + "balance_loss_mlp": 1.02708125, + "epoch": 0.2473396260446101, + "flos": 23185796688000.0, + "grad_norm": 1.8686727407304824, + "language_loss": 0.80136466, + "learning_rate": 3.524776144365072e-06, + "loss": 0.82364655, + "num_input_tokens_seen": 43939415, + "step": 2057, + "time_per_iteration": 2.7887585163116455 + }, + { + "auxiliary_loss_clip": 0.01189005, + "auxiliary_loss_mlp": 0.01038503, + "balance_loss_clip": 1.06246197, + "balance_loss_mlp": 1.0278579, + "epoch": 0.2474598689352492, + "flos": 21142443697920.0, + "grad_norm": 1.6322046797787906, + "language_loss": 0.78914976, + "learning_rate": 3.5242719405117016e-06, + "loss": 0.81142485, + "num_input_tokens_seen": 43959220, + "step": 2058, + "time_per_iteration": 2.9049863815307617 + }, + { + "auxiliary_loss_clip": 0.01194294, + "auxiliary_loss_mlp": 0.00888801, + "balance_loss_clip": 1.05870521, + "balance_loss_mlp": 1.00098646, + "epoch": 0.2475801118258883, + "flos": 21648352803840.0, + "grad_norm": 3.0417557632416656, + "language_loss": 0.75072289, + "learning_rate": 3.5237675054274893e-06, + "loss": 0.77155387, + "num_input_tokens_seen": 43978420, + "step": 2059, + "time_per_iteration": 2.764500379562378 + }, + { + "auxiliary_loss_clip": 0.01202639, + "auxiliary_loss_mlp": 0.01043027, + "balance_loss_clip": 1.05833364, + "balance_loss_mlp": 1.03150558, + "epoch": 0.24770035471652738, + "flos": 22674500542080.0, + "grad_norm": 1.887886980674119, + "language_loss": 0.80266213, + "learning_rate": 3.5232628391889584e-06, + "loss": 0.82511878, + "num_input_tokens_seen": 43996710, + "step": 2060, + "time_per_iteration": 2.8088138103485107 + }, + { + "auxiliary_loss_clip": 0.01168713, + "auxiliary_loss_mlp": 0.01032444, + "balance_loss_clip": 1.05351841, + "balance_loss_mlp": 1.02198935, + "epoch": 0.2478205976071665, + "flos": 22163814927360.0, + "grad_norm": 2.573391163332704, + "language_loss": 0.63895208, + "learning_rate": 3.522757941872666e-06, + "loss": 0.66096365, + "num_input_tokens_seen": 44014865, + "step": 2061, + "time_per_iteration": 2.9163222312927246 + }, + { + "auxiliary_loss_clip": 0.01217799, + "auxiliary_loss_mlp": 0.0088814, + "balance_loss_clip": 1.06699669, + "balance_loss_mlp": 1.00108981, + "epoch": 0.24794084049780557, + "flos": 24973106555520.0, + "grad_norm": 1.4990216313384315, + "language_loss": 0.82398748, + "learning_rate": 3.5222528135552042e-06, + "loss": 0.84504688, + "num_input_tokens_seen": 44036325, + "step": 2062, + "time_per_iteration": 2.7809715270996094 + }, + { + "auxiliary_loss_clip": 0.01202114, + "auxiliary_loss_mlp": 0.01043046, + "balance_loss_clip": 1.06229126, + "balance_loss_mlp": 1.03296626, + "epoch": 0.24806108338844465, + "flos": 18296379521280.0, + "grad_norm": 1.8854880793920026, + "language_loss": 0.80945939, + "learning_rate": 3.521747454313201e-06, + "loss": 0.83191097, + "num_input_tokens_seen": 44055005, + "step": 2063, + "time_per_iteration": 2.6962130069732666 + }, + { + "auxiliary_loss_clip": 0.01172229, + "auxiliary_loss_mlp": 0.01041415, + "balance_loss_clip": 1.05038369, + "balance_loss_mlp": 1.03056121, + "epoch": 0.24818132627908374, + "flos": 19282163351040.0, + "grad_norm": 2.069357776717143, + "language_loss": 0.66433203, + "learning_rate": 3.521241864223319e-06, + "loss": 0.68646842, + "num_input_tokens_seen": 44073965, + "step": 2064, + "time_per_iteration": 2.8479580879211426 + }, + { + "auxiliary_loss_clip": 0.01093735, + "auxiliary_loss_mlp": 0.01012126, + "balance_loss_clip": 1.02404046, + "balance_loss_mlp": 1.00840712, + "epoch": 0.24830156916972285, + "flos": 70285837881600.0, + "grad_norm": 0.7963752387029385, + "language_loss": 0.61960101, + "learning_rate": 3.5207360433622552e-06, + "loss": 0.64065969, + "num_input_tokens_seen": 44135965, + "step": 2065, + "time_per_iteration": 3.4197874069213867 + }, + { + "auxiliary_loss_clip": 0.01183171, + "auxiliary_loss_mlp": 0.01033698, + "balance_loss_clip": 1.0544101, + "balance_loss_mlp": 1.02355373, + "epoch": 0.24842181206036193, + "flos": 40409128287360.0, + "grad_norm": 1.7505582085078382, + "language_loss": 0.74547976, + "learning_rate": 3.5202299918067437e-06, + "loss": 0.76764846, + "num_input_tokens_seen": 44159560, + "step": 2066, + "time_per_iteration": 2.9508683681488037 + }, + { + "auxiliary_loss_clip": 0.01201765, + "auxiliary_loss_mlp": 0.01035293, + "balance_loss_clip": 1.05865812, + "balance_loss_mlp": 1.02507687, + "epoch": 0.248542054951001, + "flos": 20082432412800.0, + "grad_norm": 2.323153324302952, + "language_loss": 0.69199145, + "learning_rate": 3.519723709633551e-06, + "loss": 0.71436203, + "num_input_tokens_seen": 44178320, + "step": 2067, + "time_per_iteration": 2.7696101665496826 + }, + { + "auxiliary_loss_clip": 0.01182963, + "auxiliary_loss_mlp": 0.01034282, + "balance_loss_clip": 1.05585349, + "balance_loss_mlp": 1.02349317, + "epoch": 0.24866229784164012, + "flos": 23513948363520.0, + "grad_norm": 1.830697609229505, + "language_loss": 0.83571619, + "learning_rate": 3.519217196919479e-06, + "loss": 0.85788864, + "num_input_tokens_seen": 44197305, + "step": 2068, + "time_per_iteration": 2.724606513977051 + }, + { + "auxiliary_loss_clip": 0.01197529, + "auxiliary_loss_mlp": 0.01037929, + "balance_loss_clip": 1.06150723, + "balance_loss_mlp": 1.02802324, + "epoch": 0.2487825407322792, + "flos": 19865101173120.0, + "grad_norm": 2.0422917681884676, + "language_loss": 0.73403406, + "learning_rate": 3.518710453741367e-06, + "loss": 0.7563886, + "num_input_tokens_seen": 44216505, + "step": 2069, + "time_per_iteration": 2.8379950523376465 + }, + { + "auxiliary_loss_clip": 0.01184322, + "auxiliary_loss_mlp": 0.00889141, + "balance_loss_clip": 1.05657649, + "balance_loss_mlp": 1.0008198, + "epoch": 0.2489027836229183, + "flos": 22017622573440.0, + "grad_norm": 2.064417774537924, + "language_loss": 0.67770922, + "learning_rate": 3.518203480176086e-06, + "loss": 0.69844389, + "num_input_tokens_seen": 44235435, + "step": 2070, + "time_per_iteration": 3.8184728622436523 + }, + { + "auxiliary_loss_clip": 0.01148278, + "auxiliary_loss_mlp": 0.0103461, + "balance_loss_clip": 1.04657269, + "balance_loss_mlp": 1.02395225, + "epoch": 0.2490230265135574, + "flos": 23294354567040.0, + "grad_norm": 2.423860512437102, + "language_loss": 0.81015158, + "learning_rate": 3.517696276300545e-06, + "loss": 0.83198047, + "num_input_tokens_seen": 44256975, + "step": 2071, + "time_per_iteration": 5.674143314361572 + }, + { + "auxiliary_loss_clip": 0.01204941, + "auxiliary_loss_mlp": 0.01038627, + "balance_loss_clip": 1.06232154, + "balance_loss_mlp": 1.02699232, + "epoch": 0.24914326940419648, + "flos": 19826784339840.0, + "grad_norm": 2.581665931112931, + "language_loss": 0.69345403, + "learning_rate": 3.517188842191685e-06, + "loss": 0.71588969, + "num_input_tokens_seen": 44275125, + "step": 2072, + "time_per_iteration": 2.729124069213867 + }, + { + "auxiliary_loss_clip": 0.01197339, + "auxiliary_loss_mlp": 0.01039557, + "balance_loss_clip": 1.05562007, + "balance_loss_mlp": 1.02802396, + "epoch": 0.24926351229483557, + "flos": 20229271211520.0, + "grad_norm": 1.636096459773356, + "language_loss": 0.73928976, + "learning_rate": 3.5166811779264837e-06, + "loss": 0.76165873, + "num_input_tokens_seen": 44295445, + "step": 2073, + "time_per_iteration": 2.773346185684204 + }, + { + "auxiliary_loss_clip": 0.01209562, + "auxiliary_loss_mlp": 0.0103466, + "balance_loss_clip": 1.0576508, + "balance_loss_mlp": 1.02423513, + "epoch": 0.24938375518547465, + "flos": 23294570048640.0, + "grad_norm": 3.1348621596268624, + "language_loss": 0.78104907, + "learning_rate": 3.5161732835819545e-06, + "loss": 0.80349123, + "num_input_tokens_seen": 44314755, + "step": 2074, + "time_per_iteration": 2.6696701049804688 + }, + { + "auxiliary_loss_clip": 0.01212697, + "auxiliary_loss_mlp": 0.01043211, + "balance_loss_clip": 1.06210709, + "balance_loss_mlp": 1.03231502, + "epoch": 0.24950399807611376, + "flos": 17311673099520.0, + "grad_norm": 2.315948011345503, + "language_loss": 0.83403522, + "learning_rate": 3.515665159235143e-06, + "loss": 0.85659432, + "num_input_tokens_seen": 44333640, + "step": 2075, + "time_per_iteration": 2.6980934143066406 + }, + { + "auxiliary_loss_clip": 0.01182534, + "auxiliary_loss_mlp": 0.01034934, + "balance_loss_clip": 1.04956019, + "balance_loss_mlp": 1.02512896, + "epoch": 0.24962424096675284, + "flos": 19024863252480.0, + "grad_norm": 1.8430733778275175, + "language_loss": 0.7503711, + "learning_rate": 3.5151568049631318e-06, + "loss": 0.77254575, + "num_input_tokens_seen": 44352355, + "step": 2076, + "time_per_iteration": 2.7294626235961914 + }, + { + "auxiliary_loss_clip": 0.01211127, + "auxiliary_loss_mlp": 0.010349, + "balance_loss_clip": 1.05994892, + "balance_loss_mlp": 1.0240047, + "epoch": 0.24974448385739192, + "flos": 33398790710400.0, + "grad_norm": 1.6930014799138073, + "language_loss": 0.80200243, + "learning_rate": 3.5146482208430385e-06, + "loss": 0.82446277, + "num_input_tokens_seen": 44374185, + "step": 2077, + "time_per_iteration": 2.782015085220337 + }, + { + "auxiliary_loss_clip": 0.01148082, + "auxiliary_loss_mlp": 0.01038637, + "balance_loss_clip": 1.04604876, + "balance_loss_mlp": 1.02701354, + "epoch": 0.24986472674803104, + "flos": 30007279532160.0, + "grad_norm": 2.047479310516589, + "language_loss": 0.68113399, + "learning_rate": 3.514139406952014e-06, + "loss": 0.70300114, + "num_input_tokens_seen": 44396210, + "step": 2078, + "time_per_iteration": 2.93456768989563 + }, + { + "auxiliary_loss_clip": 0.01201482, + "auxiliary_loss_mlp": 0.01037403, + "balance_loss_clip": 1.05980253, + "balance_loss_mlp": 1.02733541, + "epoch": 0.24998496963867012, + "flos": 26613074833920.0, + "grad_norm": 1.9268061292065661, + "language_loss": 0.83824718, + "learning_rate": 3.5136303633672454e-06, + "loss": 0.860636, + "num_input_tokens_seen": 44416340, + "step": 2079, + "time_per_iteration": 2.7218947410583496 + }, + { + "auxiliary_loss_clip": 0.01191804, + "auxiliary_loss_mlp": 0.0088836, + "balance_loss_clip": 1.05879307, + "balance_loss_mlp": 1.00067973, + "epoch": 0.25010521252930923, + "flos": 23553989049600.0, + "grad_norm": 1.5582512711102405, + "language_loss": 0.74309593, + "learning_rate": 3.5131210901659544e-06, + "loss": 0.76389754, + "num_input_tokens_seen": 44438095, + "step": 2080, + "time_per_iteration": 2.8333592414855957 + }, + { + "auxiliary_loss_clip": 0.01176111, + "auxiliary_loss_mlp": 0.01036159, + "balance_loss_clip": 1.05231154, + "balance_loss_mlp": 1.02578759, + "epoch": 0.2502254554199483, + "flos": 23441193365760.0, + "grad_norm": 2.442952423174635, + "language_loss": 0.82098061, + "learning_rate": 3.5126115874253967e-06, + "loss": 0.84310329, + "num_input_tokens_seen": 44457650, + "step": 2081, + "time_per_iteration": 2.8970699310302734 + }, + { + "auxiliary_loss_clip": 0.01179462, + "auxiliary_loss_mlp": 0.01039002, + "balance_loss_clip": 1.05626798, + "balance_loss_mlp": 1.02836263, + "epoch": 0.2503456983105874, + "flos": 28761681651840.0, + "grad_norm": 1.7885981393852852, + "language_loss": 0.80830085, + "learning_rate": 3.5121018552228644e-06, + "loss": 0.83048546, + "num_input_tokens_seen": 44476155, + "step": 2082, + "time_per_iteration": 2.895120859146118 + }, + { + "auxiliary_loss_clip": 0.01177991, + "auxiliary_loss_mlp": 0.01037856, + "balance_loss_clip": 1.05208313, + "balance_loss_mlp": 1.02717471, + "epoch": 0.2504659412012265, + "flos": 18770256673920.0, + "grad_norm": 2.9012062993317373, + "language_loss": 0.76051813, + "learning_rate": 3.5115918936356827e-06, + "loss": 0.78267658, + "num_input_tokens_seen": 44492910, + "step": 2083, + "time_per_iteration": 2.715549945831299 + }, + { + "auxiliary_loss_clip": 0.01160331, + "auxiliary_loss_mlp": 0.01034147, + "balance_loss_clip": 1.05105495, + "balance_loss_mlp": 1.02447963, + "epoch": 0.25058618409186556, + "flos": 16873383346560.0, + "grad_norm": 1.781711625634229, + "language_loss": 0.78548646, + "learning_rate": 3.5110817027412123e-06, + "loss": 0.80743122, + "num_input_tokens_seen": 44512000, + "step": 2084, + "time_per_iteration": 2.773200511932373 + }, + { + "auxiliary_loss_clip": 0.01169107, + "auxiliary_loss_mlp": 0.01032391, + "balance_loss_clip": 1.0484494, + "balance_loss_mlp": 1.02193069, + "epoch": 0.25070642698250467, + "flos": 24425540651520.0, + "grad_norm": 2.3572293576930137, + "language_loss": 0.68764508, + "learning_rate": 3.5105712826168493e-06, + "loss": 0.70966005, + "num_input_tokens_seen": 44531650, + "step": 2085, + "time_per_iteration": 2.8172390460968018 + }, + { + "auxiliary_loss_clip": 0.01193744, + "auxiliary_loss_mlp": 0.00887379, + "balance_loss_clip": 1.05469906, + "balance_loss_mlp": 1.00073624, + "epoch": 0.2508266698731437, + "flos": 20260944028800.0, + "grad_norm": 1.8963098681359638, + "language_loss": 0.70982385, + "learning_rate": 3.5100606333400235e-06, + "loss": 0.73063511, + "num_input_tokens_seen": 44548785, + "step": 2086, + "time_per_iteration": 2.7176918983459473 + }, + { + "auxiliary_loss_clip": 0.01199758, + "auxiliary_loss_mlp": 0.0103612, + "balance_loss_clip": 1.0575918, + "balance_loss_mlp": 1.02557051, + "epoch": 0.25094691276378284, + "flos": 19245318975360.0, + "grad_norm": 3.4464981763610005, + "language_loss": 0.77067828, + "learning_rate": 3.5095497549882006e-06, + "loss": 0.79303712, + "num_input_tokens_seen": 44567230, + "step": 2087, + "time_per_iteration": 2.7558584213256836 + }, + { + "auxiliary_loss_clip": 0.01199927, + "auxiliary_loss_mlp": 0.01036758, + "balance_loss_clip": 1.0599401, + "balance_loss_mlp": 1.0264287, + "epoch": 0.25106715565442195, + "flos": 26943237671040.0, + "grad_norm": 1.8614484085816911, + "language_loss": 0.7232542, + "learning_rate": 3.50903864763888e-06, + "loss": 0.74562103, + "num_input_tokens_seen": 44588020, + "step": 2088, + "time_per_iteration": 2.7244765758514404 + }, + { + "auxiliary_loss_clip": 0.01200287, + "auxiliary_loss_mlp": 0.01032952, + "balance_loss_clip": 1.05558944, + "balance_loss_mlp": 1.02247334, + "epoch": 0.251187398545061, + "flos": 48359570572800.0, + "grad_norm": 2.315687927459775, + "language_loss": 0.75992858, + "learning_rate": 3.5085273113695965e-06, + "loss": 0.78226101, + "num_input_tokens_seen": 44612590, + "step": 2089, + "time_per_iteration": 2.9230401515960693 + }, + { + "auxiliary_loss_clip": 0.01207241, + "auxiliary_loss_mlp": 0.01032947, + "balance_loss_clip": 1.05835652, + "balance_loss_mlp": 1.02279067, + "epoch": 0.2513076414357001, + "flos": 27016100409600.0, + "grad_norm": 3.6161321166585934, + "language_loss": 0.78667164, + "learning_rate": 3.508015746257919e-06, + "loss": 0.80907351, + "num_input_tokens_seen": 44631630, + "step": 2090, + "time_per_iteration": 2.718987226486206 + }, + { + "auxiliary_loss_clip": 0.01181097, + "auxiliary_loss_mlp": 0.01035597, + "balance_loss_clip": 1.05321085, + "balance_loss_mlp": 1.02471328, + "epoch": 0.2514278843263392, + "flos": 19463619882240.0, + "grad_norm": 2.0475355177110033, + "language_loss": 0.8370384, + "learning_rate": 3.5075039523814518e-06, + "loss": 0.85920531, + "num_input_tokens_seen": 44650820, + "step": 2091, + "time_per_iteration": 3.0043740272521973 + }, + { + "auxiliary_loss_clip": 0.01202281, + "auxiliary_loss_mlp": 0.01031398, + "balance_loss_clip": 1.05440021, + "balance_loss_mlp": 1.02104473, + "epoch": 0.2515481272169783, + "flos": 16866092885760.0, + "grad_norm": 2.8419219410696526, + "language_loss": 0.81893688, + "learning_rate": 3.506991929817834e-06, + "loss": 0.84127367, + "num_input_tokens_seen": 44667540, + "step": 2092, + "time_per_iteration": 2.6931064128875732 + }, + { + "auxiliary_loss_clip": 0.01205853, + "auxiliary_loss_mlp": 0.01035811, + "balance_loss_clip": 1.06096566, + "balance_loss_mlp": 1.02601779, + "epoch": 0.2516683701076174, + "flos": 23732464752000.0, + "grad_norm": 1.747845694236141, + "language_loss": 0.82758796, + "learning_rate": 3.506479678644738e-06, + "loss": 0.85000467, + "num_input_tokens_seen": 44687935, + "step": 2093, + "time_per_iteration": 2.7333126068115234 + }, + { + "auxiliary_loss_clip": 0.01157737, + "auxiliary_loss_mlp": 0.01034696, + "balance_loss_clip": 1.04937696, + "balance_loss_mlp": 1.02402687, + "epoch": 0.2517886129982565, + "flos": 27635954434560.0, + "grad_norm": 2.8428099348980225, + "language_loss": 0.74294615, + "learning_rate": 3.505967198939873e-06, + "loss": 0.76487052, + "num_input_tokens_seen": 44704975, + "step": 2094, + "time_per_iteration": 2.9884390830993652 + }, + { + "auxiliary_loss_clip": 0.0118615, + "auxiliary_loss_mlp": 0.01027668, + "balance_loss_clip": 1.05356002, + "balance_loss_mlp": 1.01789856, + "epoch": 0.25190885588889556, + "flos": 38104596529920.0, + "grad_norm": 1.8759897651667659, + "language_loss": 0.78169948, + "learning_rate": 3.5054544907809813e-06, + "loss": 0.80383772, + "num_input_tokens_seen": 44725475, + "step": 2095, + "time_per_iteration": 3.024867057800293 + }, + { + "auxiliary_loss_clip": 0.01185983, + "auxiliary_loss_mlp": 0.0088749, + "balance_loss_clip": 1.05743718, + "balance_loss_mlp": 1.00071883, + "epoch": 0.25202909877953467, + "flos": 22269894768000.0, + "grad_norm": 1.9445312639485373, + "language_loss": 0.80277115, + "learning_rate": 3.50494155424584e-06, + "loss": 0.82350588, + "num_input_tokens_seen": 44744380, + "step": 2096, + "time_per_iteration": 4.608578681945801 + }, + { + "auxiliary_loss_clip": 0.01202499, + "auxiliary_loss_mlp": 0.01031867, + "balance_loss_clip": 1.05710888, + "balance_loss_mlp": 1.02113819, + "epoch": 0.2521493416701738, + "flos": 21761759018880.0, + "grad_norm": 1.6223612508837275, + "language_loss": 0.83058345, + "learning_rate": 3.504428389412262e-06, + "loss": 0.85292715, + "num_input_tokens_seen": 44765190, + "step": 2097, + "time_per_iteration": 4.640692472457886 + }, + { + "auxiliary_loss_clip": 0.01194266, + "auxiliary_loss_mlp": 0.01035289, + "balance_loss_clip": 1.05403435, + "balance_loss_mlp": 1.02451289, + "epoch": 0.25226958456081283, + "flos": 27746738956800.0, + "grad_norm": 1.9635855097019135, + "language_loss": 0.73040664, + "learning_rate": 3.5039149963580927e-06, + "loss": 0.75270224, + "num_input_tokens_seen": 44785210, + "step": 2098, + "time_per_iteration": 2.8380515575408936 + }, + { + "auxiliary_loss_clip": 0.01185056, + "auxiliary_loss_mlp": 0.01029421, + "balance_loss_clip": 1.05749846, + "balance_loss_mlp": 1.01943159, + "epoch": 0.25238982745145194, + "flos": 30732171903360.0, + "grad_norm": 2.2904513323263442, + "language_loss": 0.7030859, + "learning_rate": 3.503401375161215e-06, + "loss": 0.72523069, + "num_input_tokens_seen": 44804955, + "step": 2099, + "time_per_iteration": 2.8426599502563477 + }, + { + "auxiliary_loss_clip": 0.01203622, + "auxiliary_loss_mlp": 0.01034151, + "balance_loss_clip": 1.05654287, + "balance_loss_mlp": 1.02451324, + "epoch": 0.252510070342091, + "flos": 20266331068800.0, + "grad_norm": 1.5523617758488601, + "language_loss": 0.83738995, + "learning_rate": 3.502887525899544e-06, + "loss": 0.85976762, + "num_input_tokens_seen": 44823935, + "step": 2100, + "time_per_iteration": 2.760249614715576 + }, + { + "auxiliary_loss_clip": 0.0118811, + "auxiliary_loss_mlp": 0.01036084, + "balance_loss_clip": 1.05536103, + "balance_loss_mlp": 1.02559984, + "epoch": 0.2526303132327301, + "flos": 22747399194240.0, + "grad_norm": 2.0032941057361, + "language_loss": 0.82866901, + "learning_rate": 3.50237344865103e-06, + "loss": 0.85091096, + "num_input_tokens_seen": 44844935, + "step": 2101, + "time_per_iteration": 2.878667116165161 + }, + { + "auxiliary_loss_clip": 0.0120809, + "auxiliary_loss_mlp": 0.01038004, + "balance_loss_clip": 1.05901766, + "balance_loss_mlp": 1.02754891, + "epoch": 0.2527505561233692, + "flos": 30263466309120.0, + "grad_norm": 3.0843054291471432, + "language_loss": 0.76478493, + "learning_rate": 3.501859143493658e-06, + "loss": 0.78724587, + "num_input_tokens_seen": 44865565, + "step": 2102, + "time_per_iteration": 2.862783670425415 + }, + { + "auxiliary_loss_clip": 0.0110997, + "auxiliary_loss_mlp": 0.01003307, + "balance_loss_clip": 1.03152466, + "balance_loss_mlp": 1.00008821, + "epoch": 0.2528707990140083, + "flos": 58492917164160.0, + "grad_norm": 0.9135554661207916, + "language_loss": 0.6052931, + "learning_rate": 3.5013446105054488e-06, + "loss": 0.62642592, + "num_input_tokens_seen": 44918485, + "step": 2103, + "time_per_iteration": 2.97418212890625 + }, + { + "auxiliary_loss_clip": 0.01165133, + "auxiliary_loss_mlp": 0.01036073, + "balance_loss_clip": 1.05228889, + "balance_loss_mlp": 1.02604115, + "epoch": 0.2529910419046474, + "flos": 24645134448000.0, + "grad_norm": 2.0847139954207092, + "language_loss": 0.75061488, + "learning_rate": 3.5008298497644555e-06, + "loss": 0.77262688, + "num_input_tokens_seen": 44937530, + "step": 2104, + "time_per_iteration": 2.8909051418304443 + }, + { + "auxiliary_loss_clip": 0.01181559, + "auxiliary_loss_mlp": 0.01040282, + "balance_loss_clip": 1.05566621, + "balance_loss_mlp": 1.03031039, + "epoch": 0.2531112847952865, + "flos": 23842135952640.0, + "grad_norm": 1.5808376954797287, + "language_loss": 0.88101286, + "learning_rate": 3.500314861348767e-06, + "loss": 0.90323126, + "num_input_tokens_seen": 44958165, + "step": 2105, + "time_per_iteration": 2.91648006439209 + }, + { + "auxiliary_loss_clip": 0.01166299, + "auxiliary_loss_mlp": 0.01039212, + "balance_loss_clip": 1.05317116, + "balance_loss_mlp": 1.02760708, + "epoch": 0.25323152768592555, + "flos": 16143822207360.0, + "grad_norm": 2.031692981884069, + "language_loss": 0.77287889, + "learning_rate": 3.499799645336507e-06, + "loss": 0.79493403, + "num_input_tokens_seen": 44975060, + "step": 2106, + "time_per_iteration": 2.7459194660186768 + }, + { + "auxiliary_loss_clip": 0.01199421, + "auxiliary_loss_mlp": 0.01036489, + "balance_loss_clip": 1.05885923, + "balance_loss_mlp": 1.02646923, + "epoch": 0.25335177057656466, + "flos": 28405161210240.0, + "grad_norm": 1.3608978563649927, + "language_loss": 0.87036276, + "learning_rate": 3.4992842018058336e-06, + "loss": 0.89272189, + "num_input_tokens_seen": 44997960, + "step": 2107, + "time_per_iteration": 2.8499867916107178 + }, + { + "auxiliary_loss_clip": 0.01182544, + "auxiliary_loss_mlp": 0.01033945, + "balance_loss_clip": 1.05350614, + "balance_loss_mlp": 1.02402067, + "epoch": 0.25347201346720377, + "flos": 18799666934400.0, + "grad_norm": 2.0684297742193505, + "language_loss": 0.88352889, + "learning_rate": 3.4987685308349384e-06, + "loss": 0.90569377, + "num_input_tokens_seen": 45015690, + "step": 2108, + "time_per_iteration": 2.740708351135254 + }, + { + "auxiliary_loss_clip": 0.01174099, + "auxiliary_loss_mlp": 0.01036803, + "balance_loss_clip": 1.04951596, + "balance_loss_mlp": 1.02686727, + "epoch": 0.2535922563578428, + "flos": 15815490963840.0, + "grad_norm": 2.2482672885277193, + "language_loss": 0.61703736, + "learning_rate": 3.4982526325020497e-06, + "loss": 0.63914639, + "num_input_tokens_seen": 45032660, + "step": 2109, + "time_per_iteration": 2.9473934173583984 + }, + { + "auxiliary_loss_clip": 0.01189245, + "auxiliary_loss_mlp": 0.01035941, + "balance_loss_clip": 1.05656779, + "balance_loss_mlp": 1.02551031, + "epoch": 0.25371249924848194, + "flos": 16318922031360.0, + "grad_norm": 2.2862626609487076, + "language_loss": 0.82432294, + "learning_rate": 3.4977365068854273e-06, + "loss": 0.84657478, + "num_input_tokens_seen": 45048280, + "step": 2110, + "time_per_iteration": 2.829479217529297 + }, + { + "auxiliary_loss_clip": 0.01177156, + "auxiliary_loss_mlp": 0.01041861, + "balance_loss_clip": 1.05250549, + "balance_loss_mlp": 1.03212166, + "epoch": 0.25383274213912105, + "flos": 21761615364480.0, + "grad_norm": 1.7766143018381897, + "language_loss": 0.73736334, + "learning_rate": 3.4972201540633676e-06, + "loss": 0.75955349, + "num_input_tokens_seen": 45067635, + "step": 2111, + "time_per_iteration": 2.8367908000946045 + }, + { + "auxiliary_loss_clip": 0.01176178, + "auxiliary_loss_mlp": 0.01037327, + "balance_loss_clip": 1.05199337, + "balance_loss_mlp": 1.0269084, + "epoch": 0.2539529850297601, + "flos": 21396870708480.0, + "grad_norm": 3.4767794027176264, + "language_loss": 0.85149384, + "learning_rate": 3.4967035741142008e-06, + "loss": 0.87362891, + "num_input_tokens_seen": 45086455, + "step": 2112, + "time_per_iteration": 2.8126487731933594 + }, + { + "auxiliary_loss_clip": 0.0118073, + "auxiliary_loss_mlp": 0.01035943, + "balance_loss_clip": 1.06255078, + "balance_loss_mlp": 1.02677608, + "epoch": 0.2540732279203992, + "flos": 25228467319680.0, + "grad_norm": 1.775505671343882, + "language_loss": 0.81848562, + "learning_rate": 3.4961867671162917e-06, + "loss": 0.84065241, + "num_input_tokens_seen": 45106385, + "step": 2113, + "time_per_iteration": 2.8547096252441406 + }, + { + "auxiliary_loss_clip": 0.01207036, + "auxiliary_loss_mlp": 0.01042189, + "balance_loss_clip": 1.05663133, + "balance_loss_mlp": 1.03153181, + "epoch": 0.2541934708110383, + "flos": 19427386037760.0, + "grad_norm": 2.691407668535482, + "language_loss": 0.77533883, + "learning_rate": 3.4956697331480402e-06, + "loss": 0.79783106, + "num_input_tokens_seen": 45124955, + "step": 2114, + "time_per_iteration": 2.785231590270996 + }, + { + "auxiliary_loss_clip": 0.0118056, + "auxiliary_loss_mlp": 0.0103884, + "balance_loss_clip": 1.05170047, + "balance_loss_mlp": 1.02844477, + "epoch": 0.2543137137016774, + "flos": 23949436855680.0, + "grad_norm": 1.5118468876315456, + "language_loss": 0.79941821, + "learning_rate": 3.495152472287879e-06, + "loss": 0.82161212, + "num_input_tokens_seen": 45145665, + "step": 2115, + "time_per_iteration": 2.8137354850769043 + }, + { + "auxiliary_loss_clip": 0.01179332, + "auxiliary_loss_mlp": 0.010334, + "balance_loss_clip": 1.05451691, + "balance_loss_mlp": 1.02293372, + "epoch": 0.2544339565923165, + "flos": 25593283802880.0, + "grad_norm": 3.406695991699817, + "language_loss": 0.73800737, + "learning_rate": 3.4946349846142766e-06, + "loss": 0.7601347, + "num_input_tokens_seen": 45164805, + "step": 2116, + "time_per_iteration": 2.8903212547302246 + }, + { + "auxiliary_loss_clip": 0.01204103, + "auxiliary_loss_mlp": 0.01035347, + "balance_loss_clip": 1.05659986, + "balance_loss_mlp": 1.02434385, + "epoch": 0.25455419948295555, + "flos": 21689470897920.0, + "grad_norm": 2.250395665552127, + "language_loss": 0.75645232, + "learning_rate": 3.4941172702057353e-06, + "loss": 0.77884686, + "num_input_tokens_seen": 45184865, + "step": 2117, + "time_per_iteration": 2.6072182655334473 + }, + { + "auxiliary_loss_clip": 0.01183113, + "auxiliary_loss_mlp": 0.01035748, + "balance_loss_clip": 1.05470181, + "balance_loss_mlp": 1.02600884, + "epoch": 0.25467444237359466, + "flos": 26250341339520.0, + "grad_norm": 1.8081966576323152, + "language_loss": 0.80466998, + "learning_rate": 3.4935993291407924e-06, + "loss": 0.82685858, + "num_input_tokens_seen": 45203690, + "step": 2118, + "time_per_iteration": 2.8001608848571777 + }, + { + "auxiliary_loss_clip": 0.01183102, + "auxiliary_loss_mlp": 0.01035716, + "balance_loss_clip": 1.05443811, + "balance_loss_mlp": 1.02496386, + "epoch": 0.25479468526423377, + "flos": 26979686997120.0, + "grad_norm": 2.361136874701846, + "language_loss": 0.71346962, + "learning_rate": 3.4930811614980183e-06, + "loss": 0.73565781, + "num_input_tokens_seen": 45225385, + "step": 2119, + "time_per_iteration": 2.8636038303375244 + }, + { + "auxiliary_loss_clip": 0.01189122, + "auxiliary_loss_mlp": 0.0103519, + "balance_loss_clip": 1.05532718, + "balance_loss_mlp": 1.02521849, + "epoch": 0.2549149281548728, + "flos": 23475811098240.0, + "grad_norm": 1.6444051765422296, + "language_loss": 0.7905243, + "learning_rate": 3.4925627673560198e-06, + "loss": 0.81276739, + "num_input_tokens_seen": 45246045, + "step": 2120, + "time_per_iteration": 2.9126980304718018 + }, + { + "auxiliary_loss_clip": 0.01176684, + "auxiliary_loss_mlp": 0.0104049, + "balance_loss_clip": 1.05410421, + "balance_loss_mlp": 1.0305835, + "epoch": 0.25503517104551193, + "flos": 25812302981760.0, + "grad_norm": 1.72347164842057, + "language_loss": 0.88744724, + "learning_rate": 3.4920441467934357e-06, + "loss": 0.90961897, + "num_input_tokens_seen": 45266560, + "step": 2121, + "time_per_iteration": 2.835952043533325 + }, + { + "auxiliary_loss_clip": 0.01166181, + "auxiliary_loss_mlp": 0.01033332, + "balance_loss_clip": 1.05257428, + "balance_loss_mlp": 1.0239923, + "epoch": 0.25515541393615104, + "flos": 26645106787200.0, + "grad_norm": 2.3136543014005198, + "language_loss": 0.82931471, + "learning_rate": 3.491525299888941e-06, + "loss": 0.85130978, + "num_input_tokens_seen": 45285405, + "step": 2122, + "time_per_iteration": 3.815303325653076 + }, + { + "auxiliary_loss_clip": 0.01089773, + "auxiliary_loss_mlp": 0.00878096, + "balance_loss_clip": 1.03547168, + "balance_loss_mlp": 1.00181711, + "epoch": 0.2552756568267901, + "flos": 65955945847680.0, + "grad_norm": 0.9120876734823652, + "language_loss": 0.62671155, + "learning_rate": 3.491006226721244e-06, + "loss": 0.64639026, + "num_input_tokens_seen": 45349615, + "step": 2123, + "time_per_iteration": 4.399839639663696 + }, + { + "auxiliary_loss_clip": 0.01188521, + "auxiliary_loss_mlp": 0.00887387, + "balance_loss_clip": 1.05648208, + "balance_loss_mlp": 1.00090384, + "epoch": 0.2553958997174292, + "flos": 17931096161280.0, + "grad_norm": 2.0130583284044206, + "language_loss": 0.7805956, + "learning_rate": 3.4904869273690882e-06, + "loss": 0.80135465, + "num_input_tokens_seen": 45367505, + "step": 2124, + "time_per_iteration": 3.7184152603149414 + }, + { + "auxiliary_loss_clip": 0.01196637, + "auxiliary_loss_mlp": 0.01031977, + "balance_loss_clip": 1.05542135, + "balance_loss_mlp": 1.02177906, + "epoch": 0.2555161426080683, + "flos": 23367791923200.0, + "grad_norm": 1.83068559703262, + "language_loss": 0.88760608, + "learning_rate": 3.489967401911251e-06, + "loss": 0.90989226, + "num_input_tokens_seen": 45386805, + "step": 2125, + "time_per_iteration": 2.825329542160034 + }, + { + "auxiliary_loss_clip": 0.01209409, + "auxiliary_loss_mlp": 0.01035015, + "balance_loss_clip": 1.05914831, + "balance_loss_mlp": 1.02360082, + "epoch": 0.2556363854987074, + "flos": 40625130723840.0, + "grad_norm": 1.710374382453914, + "language_loss": 0.69511056, + "learning_rate": 3.4894476504265428e-06, + "loss": 0.71755481, + "num_input_tokens_seen": 45411045, + "step": 2126, + "time_per_iteration": 2.877767562866211 + }, + { + "auxiliary_loss_clip": 0.01096117, + "auxiliary_loss_mlp": 0.01007016, + "balance_loss_clip": 1.02695727, + "balance_loss_mlp": 1.00384486, + "epoch": 0.2557566283893465, + "flos": 68019443389440.0, + "grad_norm": 0.7422200484734834, + "language_loss": 0.54445994, + "learning_rate": 3.4889276729938104e-06, + "loss": 0.56549126, + "num_input_tokens_seen": 45469575, + "step": 2127, + "time_per_iteration": 3.2394821643829346 + }, + { + "auxiliary_loss_clip": 0.01181271, + "auxiliary_loss_mlp": 0.01026593, + "balance_loss_clip": 1.05418921, + "balance_loss_mlp": 1.01638889, + "epoch": 0.2558768712799856, + "flos": 22635645004800.0, + "grad_norm": 2.443370979639861, + "language_loss": 0.80653548, + "learning_rate": 3.488407469691934e-06, + "loss": 0.82861412, + "num_input_tokens_seen": 45490270, + "step": 2128, + "time_per_iteration": 2.767239809036255 + }, + { + "auxiliary_loss_clip": 0.01182429, + "auxiliary_loss_mlp": 0.01033359, + "balance_loss_clip": 1.05280364, + "balance_loss_mlp": 1.02239132, + "epoch": 0.25599711417062465, + "flos": 26396354125440.0, + "grad_norm": 2.3121767122900216, + "language_loss": 0.8098098, + "learning_rate": 3.487887040599828e-06, + "loss": 0.83196765, + "num_input_tokens_seen": 45510070, + "step": 2129, + "time_per_iteration": 2.7794079780578613 + }, + { + "auxiliary_loss_clip": 0.01208853, + "auxiliary_loss_mlp": 0.01033763, + "balance_loss_clip": 1.05957365, + "balance_loss_mlp": 1.02352262, + "epoch": 0.25611735706126376, + "flos": 22852042490880.0, + "grad_norm": 2.3669700604978683, + "language_loss": 0.76340228, + "learning_rate": 3.4873663857964407e-06, + "loss": 0.78582847, + "num_input_tokens_seen": 45527285, + "step": 2130, + "time_per_iteration": 2.6176869869232178 + }, + { + "auxiliary_loss_clip": 0.01166164, + "auxiliary_loss_mlp": 0.01034801, + "balance_loss_clip": 1.0542438, + "balance_loss_mlp": 1.0252459, + "epoch": 0.2562375999519028, + "flos": 23367863750400.0, + "grad_norm": 1.8405390925869098, + "language_loss": 0.66782731, + "learning_rate": 3.4868455053607556e-06, + "loss": 0.68983698, + "num_input_tokens_seen": 45546900, + "step": 2131, + "time_per_iteration": 2.852343797683716 + }, + { + "auxiliary_loss_clip": 0.01195728, + "auxiliary_loss_mlp": 0.01037822, + "balance_loss_clip": 1.05320191, + "balance_loss_mlp": 1.02721286, + "epoch": 0.2563578428425419, + "flos": 22856962654080.0, + "grad_norm": 2.2827924384518865, + "language_loss": 0.71974301, + "learning_rate": 3.486324399371789e-06, + "loss": 0.74207848, + "num_input_tokens_seen": 45566200, + "step": 2132, + "time_per_iteration": 2.9270715713500977 + }, + { + "auxiliary_loss_clip": 0.0117245, + "auxiliary_loss_mlp": 0.01036698, + "balance_loss_clip": 1.05437505, + "balance_loss_mlp": 1.02698874, + "epoch": 0.25647808573318104, + "flos": 21653883498240.0, + "grad_norm": 1.8152507028414229, + "language_loss": 0.7895155, + "learning_rate": 3.485803067908593e-06, + "loss": 0.81160694, + "num_input_tokens_seen": 45585710, + "step": 2133, + "time_per_iteration": 2.8502306938171387 + }, + { + "auxiliary_loss_clip": 0.0112937, + "auxiliary_loss_mlp": 0.01035132, + "balance_loss_clip": 1.04049015, + "balance_loss_mlp": 1.02508831, + "epoch": 0.2565983286238201, + "flos": 33730569659520.0, + "grad_norm": 1.9503443574791863, + "language_loss": 0.79875302, + "learning_rate": 3.485281511050253e-06, + "loss": 0.82039809, + "num_input_tokens_seen": 45607845, + "step": 2134, + "time_per_iteration": 3.125462770462036 + }, + { + "auxiliary_loss_clip": 0.01196508, + "auxiliary_loss_mlp": 0.01036453, + "balance_loss_clip": 1.05591965, + "balance_loss_mlp": 1.02674294, + "epoch": 0.2567185715144592, + "flos": 16216002587520.0, + "grad_norm": 2.850272910986279, + "language_loss": 0.89917934, + "learning_rate": 3.484759728875889e-06, + "loss": 0.92150891, + "num_input_tokens_seen": 45623210, + "step": 2135, + "time_per_iteration": 2.6594982147216797 + }, + { + "auxiliary_loss_clip": 0.0115559, + "auxiliary_loss_mlp": 0.01038111, + "balance_loss_clip": 1.0505209, + "balance_loss_mlp": 1.02813339, + "epoch": 0.2568388144050983, + "flos": 17458475984640.0, + "grad_norm": 1.7955446199006135, + "language_loss": 0.80845344, + "learning_rate": 3.4842377214646543e-06, + "loss": 0.83039057, + "num_input_tokens_seen": 45641505, + "step": 2136, + "time_per_iteration": 2.855919599533081 + }, + { + "auxiliary_loss_clip": 0.01201558, + "auxiliary_loss_mlp": 0.01033586, + "balance_loss_clip": 1.0556699, + "balance_loss_mlp": 1.0238173, + "epoch": 0.25695905729573737, + "flos": 20887442069760.0, + "grad_norm": 2.0146294186855664, + "language_loss": 0.66632921, + "learning_rate": 3.483715488895737e-06, + "loss": 0.68868065, + "num_input_tokens_seen": 45661835, + "step": 2137, + "time_per_iteration": 2.6803464889526367 + }, + { + "auxiliary_loss_clip": 0.01162942, + "auxiliary_loss_mlp": 0.01031645, + "balance_loss_clip": 1.04820848, + "balance_loss_mlp": 1.02158928, + "epoch": 0.2570793001863765, + "flos": 24717278914560.0, + "grad_norm": 1.9824086459910981, + "language_loss": 0.78647518, + "learning_rate": 3.48319303124836e-06, + "loss": 0.80842102, + "num_input_tokens_seen": 45682215, + "step": 2138, + "time_per_iteration": 3.0319132804870605 + }, + { + "auxiliary_loss_clip": 0.01184571, + "auxiliary_loss_mlp": 0.01029972, + "balance_loss_clip": 1.05832148, + "balance_loss_mlp": 1.02064443, + "epoch": 0.2571995430770156, + "flos": 26906896085760.0, + "grad_norm": 2.019019650773251, + "language_loss": 0.67065579, + "learning_rate": 3.4826703486017798e-06, + "loss": 0.69280124, + "num_input_tokens_seen": 45701840, + "step": 2139, + "time_per_iteration": 2.817509412765503 + }, + { + "auxiliary_loss_clip": 0.01195387, + "auxiliary_loss_mlp": 0.01033937, + "balance_loss_clip": 1.05911088, + "balance_loss_mlp": 1.02377403, + "epoch": 0.25731978596765465, + "flos": 19792561656960.0, + "grad_norm": 1.7066549273679985, + "language_loss": 0.76518697, + "learning_rate": 3.4821474410352867e-06, + "loss": 0.78748018, + "num_input_tokens_seen": 45720500, + "step": 2140, + "time_per_iteration": 2.7240548133850098 + }, + { + "auxiliary_loss_clip": 0.0108635, + "auxiliary_loss_mlp": 0.01019142, + "balance_loss_clip": 1.03318703, + "balance_loss_mlp": 1.01556611, + "epoch": 0.25744002885829376, + "flos": 70564970471040.0, + "grad_norm": 0.9016273524097971, + "language_loss": 0.62656099, + "learning_rate": 3.481624308628205e-06, + "loss": 0.64761591, + "num_input_tokens_seen": 45781870, + "step": 2141, + "time_per_iteration": 3.564199209213257 + }, + { + "auxiliary_loss_clip": 0.01183251, + "auxiliary_loss_mlp": 0.01033856, + "balance_loss_clip": 1.05425382, + "balance_loss_mlp": 1.02357996, + "epoch": 0.25756027174893287, + "flos": 18038181582720.0, + "grad_norm": 3.1984249409577417, + "language_loss": 0.99996722, + "learning_rate": 3.481100951459893e-06, + "loss": 1.02213836, + "num_input_tokens_seen": 45794890, + "step": 2142, + "time_per_iteration": 2.852309226989746 + }, + { + "auxiliary_loss_clip": 0.0119266, + "auxiliary_loss_mlp": 0.01041895, + "balance_loss_clip": 1.05552387, + "balance_loss_mlp": 1.03206611, + "epoch": 0.2576805146395719, + "flos": 22674069578880.0, + "grad_norm": 1.5614913642308634, + "language_loss": 0.78996068, + "learning_rate": 3.4805773696097453e-06, + "loss": 0.81230617, + "num_input_tokens_seen": 45815780, + "step": 2143, + "time_per_iteration": 2.7617762088775635 + }, + { + "auxiliary_loss_clip": 0.01182152, + "auxiliary_loss_mlp": 0.0103433, + "balance_loss_clip": 1.05801713, + "balance_loss_mlp": 1.02376842, + "epoch": 0.25780075753021103, + "flos": 16472225278080.0, + "grad_norm": 1.8841701813784038, + "language_loss": 0.87909615, + "learning_rate": 3.4800535631571874e-06, + "loss": 0.90126097, + "num_input_tokens_seen": 45831310, + "step": 2144, + "time_per_iteration": 2.8923115730285645 + }, + { + "auxiliary_loss_clip": 0.01189659, + "auxiliary_loss_mlp": 0.01042937, + "balance_loss_clip": 1.05438519, + "balance_loss_mlp": 1.03217816, + "epoch": 0.25792100042085014, + "flos": 22820297846400.0, + "grad_norm": 3.6893126797680047, + "language_loss": 0.76365614, + "learning_rate": 3.4795295321816804e-06, + "loss": 0.78598213, + "num_input_tokens_seen": 45850135, + "step": 2145, + "time_per_iteration": 2.771202325820923 + }, + { + "auxiliary_loss_clip": 0.01171936, + "auxiliary_loss_mlp": 0.01029532, + "balance_loss_clip": 1.05141902, + "balance_loss_mlp": 1.01865983, + "epoch": 0.2580412433114892, + "flos": 18697286194560.0, + "grad_norm": 2.3710328931742244, + "language_loss": 0.91042578, + "learning_rate": 3.47900527676272e-06, + "loss": 0.93244046, + "num_input_tokens_seen": 45868470, + "step": 2146, + "time_per_iteration": 2.807642698287964 + }, + { + "auxiliary_loss_clip": 0.01205994, + "auxiliary_loss_mlp": 0.01042334, + "balance_loss_clip": 1.0595386, + "balance_loss_mlp": 1.03164637, + "epoch": 0.2581614862021283, + "flos": 14283146810880.0, + "grad_norm": 1.789618038247677, + "language_loss": 0.88632655, + "learning_rate": 3.478480796979835e-06, + "loss": 0.90880984, + "num_input_tokens_seen": 45886355, + "step": 2147, + "time_per_iteration": 3.6193461418151855 + }, + { + "auxiliary_loss_clip": 0.01180292, + "auxiliary_loss_mlp": 0.01035532, + "balance_loss_clip": 1.05372858, + "balance_loss_mlp": 1.02489269, + "epoch": 0.25828172909276736, + "flos": 29498281856640.0, + "grad_norm": 2.4905200531795577, + "language_loss": 0.78465736, + "learning_rate": 3.4779560929125894e-06, + "loss": 0.80681556, + "num_input_tokens_seen": 45907900, + "step": 2148, + "time_per_iteration": 2.8730814456939697 + }, + { + "auxiliary_loss_clip": 0.01073819, + "auxiliary_loss_mlp": 0.01010265, + "balance_loss_clip": 1.02274466, + "balance_loss_mlp": 1.00659359, + "epoch": 0.2584019719834065, + "flos": 67114387376640.0, + "grad_norm": 0.6752972613175879, + "language_loss": 0.56948256, + "learning_rate": 3.4774311646405783e-06, + "loss": 0.59032345, + "num_input_tokens_seen": 45977805, + "step": 2149, + "time_per_iteration": 5.33935546875 + }, + { + "auxiliary_loss_clip": 0.0116591, + "auxiliary_loss_mlp": 0.0103637, + "balance_loss_clip": 1.04999685, + "balance_loss_mlp": 1.02657068, + "epoch": 0.2585222148740456, + "flos": 22893555634560.0, + "grad_norm": 1.8638793117210044, + "language_loss": 0.83226538, + "learning_rate": 3.476906012243435e-06, + "loss": 0.85428816, + "num_input_tokens_seen": 45996715, + "step": 2150, + "time_per_iteration": 3.89278244972229 + }, + { + "auxiliary_loss_clip": 0.01185012, + "auxiliary_loss_mlp": 0.01032796, + "balance_loss_clip": 1.05466938, + "balance_loss_mlp": 1.02314043, + "epoch": 0.25864245776468464, + "flos": 28909202808960.0, + "grad_norm": 1.5463051638752978, + "language_loss": 0.8157621, + "learning_rate": 3.476380635800824e-06, + "loss": 0.83794022, + "num_input_tokens_seen": 46017915, + "step": 2151, + "time_per_iteration": 2.800318479537964 + }, + { + "auxiliary_loss_clip": 0.0118243, + "auxiliary_loss_mlp": 0.01028655, + "balance_loss_clip": 1.05555677, + "balance_loss_mlp": 1.01834297, + "epoch": 0.25876270065532375, + "flos": 14793185980800.0, + "grad_norm": 2.133102979791639, + "language_loss": 0.85614741, + "learning_rate": 3.475855035392444e-06, + "loss": 0.87825835, + "num_input_tokens_seen": 46033235, + "step": 2152, + "time_per_iteration": 2.754868507385254 + }, + { + "auxiliary_loss_clip": 0.01153325, + "auxiliary_loss_mlp": 0.01036015, + "balance_loss_clip": 1.05033183, + "balance_loss_mlp": 1.02539372, + "epoch": 0.25888294354596286, + "flos": 60467821810560.0, + "grad_norm": 2.032352046573463, + "language_loss": 0.71109998, + "learning_rate": 3.475329211098029e-06, + "loss": 0.73299336, + "num_input_tokens_seen": 46056390, + "step": 2153, + "time_per_iteration": 3.2588253021240234 + }, + { + "auxiliary_loss_clip": 0.01171078, + "auxiliary_loss_mlp": 0.01035839, + "balance_loss_clip": 1.05341625, + "balance_loss_mlp": 1.02576625, + "epoch": 0.2590031864366019, + "flos": 27851166771840.0, + "grad_norm": 1.9891865183686648, + "language_loss": 0.82726693, + "learning_rate": 3.4748031629973453e-06, + "loss": 0.84933603, + "num_input_tokens_seen": 46077120, + "step": 2154, + "time_per_iteration": 2.962435245513916 + }, + { + "auxiliary_loss_clip": 0.01068295, + "auxiliary_loss_mlp": 0.0100722, + "balance_loss_clip": 1.02325988, + "balance_loss_mlp": 1.00390577, + "epoch": 0.25912342932724103, + "flos": 62422444206720.0, + "grad_norm": 0.9097624303700633, + "language_loss": 0.56576514, + "learning_rate": 3.4742768911701944e-06, + "loss": 0.58652031, + "num_input_tokens_seen": 46139815, + "step": 2155, + "time_per_iteration": 3.5011487007141113 + }, + { + "auxiliary_loss_clip": 0.01201787, + "auxiliary_loss_mlp": 0.01044516, + "balance_loss_clip": 1.05941653, + "balance_loss_mlp": 1.03357303, + "epoch": 0.25924367221788014, + "flos": 12378839368320.0, + "grad_norm": 3.417164337035994, + "language_loss": 0.71270055, + "learning_rate": 3.4737503956964113e-06, + "loss": 0.73516357, + "num_input_tokens_seen": 46152120, + "step": 2156, + "time_per_iteration": 2.761265993118286 + }, + { + "auxiliary_loss_clip": 0.01174044, + "auxiliary_loss_mlp": 0.01039062, + "balance_loss_clip": 1.05085576, + "balance_loss_mlp": 1.02716458, + "epoch": 0.2593639151085192, + "flos": 14575208296320.0, + "grad_norm": 1.8153346140215347, + "language_loss": 0.67235768, + "learning_rate": 3.473223676655865e-06, + "loss": 0.69448876, + "num_input_tokens_seen": 46170120, + "step": 2157, + "time_per_iteration": 2.833033800125122 + }, + { + "auxiliary_loss_clip": 0.01175635, + "auxiliary_loss_mlp": 0.01040101, + "balance_loss_clip": 1.04972649, + "balance_loss_mlp": 1.02915192, + "epoch": 0.2594841579991583, + "flos": 15230937029760.0, + "grad_norm": 1.9502690589368585, + "language_loss": 0.79615313, + "learning_rate": 3.472696734128459e-06, + "loss": 0.8183105, + "num_input_tokens_seen": 46187985, + "step": 2158, + "time_per_iteration": 2.9606006145477295 + }, + { + "auxiliary_loss_clip": 0.01196795, + "auxiliary_loss_mlp": 0.01038896, + "balance_loss_clip": 1.05633998, + "balance_loss_mlp": 1.02860188, + "epoch": 0.2596044008897974, + "flos": 23623583650560.0, + "grad_norm": 1.5715326059401153, + "language_loss": 0.75836742, + "learning_rate": 3.4721695681941286e-06, + "loss": 0.78072441, + "num_input_tokens_seen": 46207025, + "step": 2159, + "time_per_iteration": 2.7656166553497314 + }, + { + "auxiliary_loss_clip": 0.0118064, + "auxiliary_loss_mlp": 0.00888459, + "balance_loss_clip": 1.05202901, + "balance_loss_mlp": 1.00071216, + "epoch": 0.25972464378043647, + "flos": 13772281628160.0, + "grad_norm": 1.907113169453324, + "language_loss": 0.82515752, + "learning_rate": 3.471642178932845e-06, + "loss": 0.84584856, + "num_input_tokens_seen": 46225670, + "step": 2160, + "time_per_iteration": 2.7598960399627686 + }, + { + "auxiliary_loss_clip": 0.01187323, + "auxiliary_loss_mlp": 0.01031624, + "balance_loss_clip": 1.05348015, + "balance_loss_mlp": 1.02178967, + "epoch": 0.2598448866710756, + "flos": 19573578391680.0, + "grad_norm": 2.692441280049903, + "language_loss": 0.89717627, + "learning_rate": 3.471114566424613e-06, + "loss": 0.9193657, + "num_input_tokens_seen": 46244130, + "step": 2161, + "time_per_iteration": 2.7063777446746826 + }, + { + "auxiliary_loss_clip": 0.01181555, + "auxiliary_loss_mlp": 0.01032131, + "balance_loss_clip": 1.05440545, + "balance_loss_mlp": 1.02201641, + "epoch": 0.25996512956171464, + "flos": 21653237053440.0, + "grad_norm": 1.9294554300898952, + "language_loss": 0.7576102, + "learning_rate": 3.4705867307494715e-06, + "loss": 0.77974707, + "num_input_tokens_seen": 46263200, + "step": 2162, + "time_per_iteration": 2.7838289737701416 + }, + { + "auxiliary_loss_clip": 0.01196473, + "auxiliary_loss_mlp": 0.01037866, + "balance_loss_clip": 1.05509937, + "balance_loss_mlp": 1.02711332, + "epoch": 0.26008537245235375, + "flos": 18223480869120.0, + "grad_norm": 2.031414486565739, + "language_loss": 0.84731674, + "learning_rate": 3.470058671987492e-06, + "loss": 0.86966014, + "num_input_tokens_seen": 46281465, + "step": 2163, + "time_per_iteration": 2.6575913429260254 + }, + { + "auxiliary_loss_clip": 0.01198998, + "auxiliary_loss_mlp": 0.01032334, + "balance_loss_clip": 1.05420732, + "balance_loss_mlp": 1.02144969, + "epoch": 0.26020561534299286, + "flos": 24645385843200.0, + "grad_norm": 1.6479070510229012, + "language_loss": 0.84480792, + "learning_rate": 3.4695303902187805e-06, + "loss": 0.86712122, + "num_input_tokens_seen": 46301020, + "step": 2164, + "time_per_iteration": 2.7423832416534424 + }, + { + "auxiliary_loss_clip": 0.01168021, + "auxiliary_loss_mlp": 0.01034322, + "balance_loss_clip": 1.04780984, + "balance_loss_mlp": 1.0222342, + "epoch": 0.2603258582336319, + "flos": 25773662926080.0, + "grad_norm": 1.929693182443248, + "language_loss": 0.78852034, + "learning_rate": 3.469001885523478e-06, + "loss": 0.81054384, + "num_input_tokens_seen": 46321740, + "step": 2165, + "time_per_iteration": 2.8821065425872803 + }, + { + "auxiliary_loss_clip": 0.01206507, + "auxiliary_loss_mlp": 0.0103547, + "balance_loss_clip": 1.05842125, + "balance_loss_mlp": 1.02483034, + "epoch": 0.260446101124271, + "flos": 28766314506240.0, + "grad_norm": 1.8556526821514046, + "language_loss": 0.8127771, + "learning_rate": 3.4684731579817568e-06, + "loss": 0.83519691, + "num_input_tokens_seen": 46342730, + "step": 2166, + "time_per_iteration": 2.781846761703491 + }, + { + "auxiliary_loss_clip": 0.01151947, + "auxiliary_loss_mlp": 0.01036986, + "balance_loss_clip": 1.05078959, + "balance_loss_mlp": 1.02648997, + "epoch": 0.26056634401491013, + "flos": 25666757072640.0, + "grad_norm": 1.5119934719409078, + "language_loss": 0.76768214, + "learning_rate": 3.4679442076738247e-06, + "loss": 0.78957146, + "num_input_tokens_seen": 46362445, + "step": 2167, + "time_per_iteration": 2.9586832523345947 + }, + { + "auxiliary_loss_clip": 0.01206574, + "auxiliary_loss_mlp": 0.01034231, + "balance_loss_clip": 1.05822337, + "balance_loss_mlp": 1.02382398, + "epoch": 0.2606865869055492, + "flos": 27052765217280.0, + "grad_norm": 4.471423064116103, + "language_loss": 0.83522344, + "learning_rate": 3.4674150346799245e-06, + "loss": 0.85763144, + "num_input_tokens_seen": 46382145, + "step": 2168, + "time_per_iteration": 2.7613089084625244 + }, + { + "auxiliary_loss_clip": 0.01185793, + "auxiliary_loss_mlp": 0.01039828, + "balance_loss_clip": 1.05516839, + "balance_loss_mlp": 1.02898026, + "epoch": 0.2608068297961883, + "flos": 17712615686400.0, + "grad_norm": 2.1114486423015815, + "language_loss": 0.7982403, + "learning_rate": 3.4668856390803295e-06, + "loss": 0.82049656, + "num_input_tokens_seen": 46400025, + "step": 2169, + "time_per_iteration": 2.969308376312256 + }, + { + "auxiliary_loss_clip": 0.01183138, + "auxiliary_loss_mlp": 0.0103472, + "balance_loss_clip": 1.05297899, + "balance_loss_mlp": 1.02445054, + "epoch": 0.2609270726868274, + "flos": 18551632544640.0, + "grad_norm": 1.8939401263691533, + "language_loss": 0.90052116, + "learning_rate": 3.4663560209553495e-06, + "loss": 0.92269975, + "num_input_tokens_seen": 46418090, + "step": 2170, + "time_per_iteration": 2.726191759109497 + }, + { + "auxiliary_loss_clip": 0.01176755, + "auxiliary_loss_mlp": 0.0103729, + "balance_loss_clip": 1.05256724, + "balance_loss_mlp": 1.02685308, + "epoch": 0.26104731557746647, + "flos": 21835699165440.0, + "grad_norm": 1.7672822374072767, + "language_loss": 0.79049718, + "learning_rate": 3.4658261803853267e-06, + "loss": 0.81263757, + "num_input_tokens_seen": 46436015, + "step": 2171, + "time_per_iteration": 2.8255293369293213 + }, + { + "auxiliary_loss_clip": 0.01179969, + "auxiliary_loss_mlp": 0.01047403, + "balance_loss_clip": 1.05649567, + "balance_loss_mlp": 1.03667986, + "epoch": 0.2611675584681056, + "flos": 21689650465920.0, + "grad_norm": 1.9916150027805568, + "language_loss": 0.80640393, + "learning_rate": 3.4652961174506383e-06, + "loss": 0.82867765, + "num_input_tokens_seen": 46455885, + "step": 2172, + "time_per_iteration": 2.852436065673828 + }, + { + "auxiliary_loss_clip": 0.01088794, + "auxiliary_loss_mlp": 0.01011655, + "balance_loss_clip": 1.02167284, + "balance_loss_mlp": 1.00841224, + "epoch": 0.2612878013587447, + "flos": 71862101389440.0, + "grad_norm": 0.9803543524079263, + "language_loss": 0.58151436, + "learning_rate": 3.464765832231694e-06, + "loss": 0.60251886, + "num_input_tokens_seen": 46510050, + "step": 2173, + "time_per_iteration": 4.239098072052002 + }, + { + "auxiliary_loss_clip": 0.01199897, + "auxiliary_loss_mlp": 0.01035988, + "balance_loss_clip": 1.05941296, + "balance_loss_mlp": 1.02471113, + "epoch": 0.26140804424938374, + "flos": 20227511445120.0, + "grad_norm": 1.717147735389281, + "language_loss": 0.7052021, + "learning_rate": 3.4642353248089373e-06, + "loss": 0.72756094, + "num_input_tokens_seen": 46528810, + "step": 2174, + "time_per_iteration": 3.694974660873413 + }, + { + "auxiliary_loss_clip": 0.01180801, + "auxiliary_loss_mlp": 0.01039388, + "balance_loss_clip": 1.05412054, + "balance_loss_mlp": 1.0288558, + "epoch": 0.26152828714002285, + "flos": 25557085872000.0, + "grad_norm": 2.085075797488999, + "language_loss": 0.80460334, + "learning_rate": 3.463704595262846e-06, + "loss": 0.82680523, + "num_input_tokens_seen": 46549690, + "step": 2175, + "time_per_iteration": 4.572082996368408 + }, + { + "auxiliary_loss_clip": 0.01173566, + "auxiliary_loss_mlp": 0.01037353, + "balance_loss_clip": 1.05506635, + "balance_loss_mlp": 1.02662444, + "epoch": 0.26164853003066196, + "flos": 25446516831360.0, + "grad_norm": 1.8519380988393408, + "language_loss": 0.70814979, + "learning_rate": 3.463173643673931e-06, + "loss": 0.73025906, + "num_input_tokens_seen": 46572215, + "step": 2176, + "time_per_iteration": 2.771400213241577 + }, + { + "auxiliary_loss_clip": 0.010914, + "auxiliary_loss_mlp": 0.0100909, + "balance_loss_clip": 1.01966679, + "balance_loss_mlp": 1.00589502, + "epoch": 0.261768772921301, + "flos": 53944580568960.0, + "grad_norm": 0.9088579279967423, + "language_loss": 0.63547981, + "learning_rate": 3.4626424701227387e-06, + "loss": 0.65648472, + "num_input_tokens_seen": 46627275, + "step": 2177, + "time_per_iteration": 3.132389783859253 + }, + { + "auxiliary_loss_clip": 0.01096409, + "auxiliary_loss_mlp": 0.01004193, + "balance_loss_clip": 1.01987803, + "balance_loss_mlp": 1.00099778, + "epoch": 0.26188901581194013, + "flos": 70687606481280.0, + "grad_norm": 0.8178915828224337, + "language_loss": 0.55790639, + "learning_rate": 3.4621110746898452e-06, + "loss": 0.57891244, + "num_input_tokens_seen": 46695135, + "step": 2178, + "time_per_iteration": 3.2299842834472656 + }, + { + "auxiliary_loss_clip": 0.01197005, + "auxiliary_loss_mlp": 0.01036146, + "balance_loss_clip": 1.05630195, + "balance_loss_mlp": 1.02634144, + "epoch": 0.2620092587025792, + "flos": 21069580959360.0, + "grad_norm": 1.526132844191702, + "language_loss": 0.74529964, + "learning_rate": 3.4615794574558654e-06, + "loss": 0.76763117, + "num_input_tokens_seen": 46714145, + "step": 2179, + "time_per_iteration": 2.5583930015563965 + }, + { + "auxiliary_loss_clip": 0.01182223, + "auxiliary_loss_mlp": 0.01035479, + "balance_loss_clip": 1.0526247, + "balance_loss_mlp": 1.02492952, + "epoch": 0.2621295015932183, + "flos": 18369601395840.0, + "grad_norm": 1.9774095519819415, + "language_loss": 0.83939475, + "learning_rate": 3.4610476185014436e-06, + "loss": 0.86157173, + "num_input_tokens_seen": 46731405, + "step": 2180, + "time_per_iteration": 2.591078281402588 + }, + { + "auxiliary_loss_clip": 0.01205609, + "auxiliary_loss_mlp": 0.01034405, + "balance_loss_clip": 1.05720198, + "balance_loss_mlp": 1.02340806, + "epoch": 0.2622497444838574, + "flos": 23659997063040.0, + "grad_norm": 1.6675796975879367, + "language_loss": 0.79550815, + "learning_rate": 3.4605155579072597e-06, + "loss": 0.81790829, + "num_input_tokens_seen": 46751260, + "step": 2181, + "time_per_iteration": 2.6248362064361572 + }, + { + "auxiliary_loss_clip": 0.01158057, + "auxiliary_loss_mlp": 0.01042363, + "balance_loss_clip": 1.04962611, + "balance_loss_mlp": 1.03177726, + "epoch": 0.26236998737449646, + "flos": 22123810154880.0, + "grad_norm": 1.7841974938817606, + "language_loss": 0.71338165, + "learning_rate": 3.459983275754027e-06, + "loss": 0.73538589, + "num_input_tokens_seen": 46770155, + "step": 2182, + "time_per_iteration": 2.970925807952881 + }, + { + "auxiliary_loss_clip": 0.01207236, + "auxiliary_loss_mlp": 0.01034232, + "balance_loss_clip": 1.05806351, + "balance_loss_mlp": 1.02399194, + "epoch": 0.26249023026513557, + "flos": 17895185539200.0, + "grad_norm": 2.3057559793604603, + "language_loss": 0.79922467, + "learning_rate": 3.4594507721224918e-06, + "loss": 0.82163936, + "num_input_tokens_seen": 46788805, + "step": 2183, + "time_per_iteration": 2.7405712604522705 + }, + { + "auxiliary_loss_clip": 0.01185889, + "auxiliary_loss_mlp": 0.01036409, + "balance_loss_clip": 1.05319786, + "balance_loss_mlp": 1.02633035, + "epoch": 0.2626104731557747, + "flos": 18332936588160.0, + "grad_norm": 1.685936131532712, + "language_loss": 0.8180452, + "learning_rate": 3.4589180470934353e-06, + "loss": 0.84026819, + "num_input_tokens_seen": 46808670, + "step": 2184, + "time_per_iteration": 2.993985891342163 + }, + { + "auxiliary_loss_clip": 0.01198831, + "auxiliary_loss_mlp": 0.0103016, + "balance_loss_clip": 1.05253768, + "balance_loss_mlp": 1.01878715, + "epoch": 0.26273071604641374, + "flos": 19317714837120.0, + "grad_norm": 2.4760377446620163, + "language_loss": 0.77388632, + "learning_rate": 3.4583851007476713e-06, + "loss": 0.7961762, + "num_input_tokens_seen": 46827140, + "step": 2185, + "time_per_iteration": 2.9076952934265137 + }, + { + "auxiliary_loss_clip": 0.01177862, + "auxiliary_loss_mlp": 0.01051705, + "balance_loss_clip": 1.05336666, + "balance_loss_mlp": 1.04007649, + "epoch": 0.26285095893705285, + "flos": 18327477720960.0, + "grad_norm": 1.9246829717469005, + "language_loss": 0.68849933, + "learning_rate": 3.4578519331660464e-06, + "loss": 0.71079504, + "num_input_tokens_seen": 46844135, + "step": 2186, + "time_per_iteration": 2.778951406478882 + }, + { + "auxiliary_loss_clip": 0.01191761, + "auxiliary_loss_mlp": 0.01037848, + "balance_loss_clip": 1.05663371, + "balance_loss_mlp": 1.0279063, + "epoch": 0.26297120182769196, + "flos": 20193827466240.0, + "grad_norm": 1.7580577848747794, + "language_loss": 0.81587791, + "learning_rate": 3.4573185444294426e-06, + "loss": 0.83817399, + "num_input_tokens_seen": 46862500, + "step": 2187, + "time_per_iteration": 2.688645362854004 + }, + { + "auxiliary_loss_clip": 0.0118375, + "auxiliary_loss_mlp": 0.00888796, + "balance_loss_clip": 1.05276573, + "balance_loss_mlp": 1.00080311, + "epoch": 0.263091444718331, + "flos": 22418421505920.0, + "grad_norm": 1.7016599379098434, + "language_loss": 0.78730714, + "learning_rate": 3.456784934618774e-06, + "loss": 0.80803263, + "num_input_tokens_seen": 46883665, + "step": 2188, + "time_per_iteration": 2.7529258728027344 + }, + { + "auxiliary_loss_clip": 0.01180395, + "auxiliary_loss_mlp": 0.01041512, + "balance_loss_clip": 1.05205059, + "balance_loss_mlp": 1.03084826, + "epoch": 0.2632116876089701, + "flos": 19024827338880.0, + "grad_norm": 1.9935266525234279, + "language_loss": 0.80028605, + "learning_rate": 3.4562511038149897e-06, + "loss": 0.82250512, + "num_input_tokens_seen": 46899160, + "step": 2189, + "time_per_iteration": 2.7548766136169434 + }, + { + "auxiliary_loss_clip": 0.01062606, + "auxiliary_loss_mlp": 0.01019148, + "balance_loss_clip": 1.01603031, + "balance_loss_mlp": 1.01623976, + "epoch": 0.26333193049960923, + "flos": 67308054531840.0, + "grad_norm": 0.8587144366476209, + "language_loss": 0.57794207, + "learning_rate": 3.4557170520990705e-06, + "loss": 0.59875965, + "num_input_tokens_seen": 46959835, + "step": 2190, + "time_per_iteration": 3.485889434814453 + }, + { + "auxiliary_loss_clip": 0.0118904, + "auxiliary_loss_mlp": 0.01035738, + "balance_loss_clip": 1.05541587, + "balance_loss_mlp": 1.0260458, + "epoch": 0.2634521733902483, + "flos": 25048806468480.0, + "grad_norm": 1.4447013185797872, + "language_loss": 0.86429942, + "learning_rate": 3.4551827795520324e-06, + "loss": 0.88654715, + "num_input_tokens_seen": 46982720, + "step": 2191, + "time_per_iteration": 2.7785110473632812 + }, + { + "auxiliary_loss_clip": 0.01191643, + "auxiliary_loss_mlp": 0.01034263, + "balance_loss_clip": 1.05232549, + "balance_loss_mlp": 1.02417755, + "epoch": 0.2635724162808874, + "flos": 20594985534720.0, + "grad_norm": 1.7624628856056903, + "language_loss": 0.84964705, + "learning_rate": 3.4546482862549226e-06, + "loss": 0.87190616, + "num_input_tokens_seen": 47003035, + "step": 2192, + "time_per_iteration": 2.740710735321045 + }, + { + "auxiliary_loss_clip": 0.01169213, + "auxiliary_loss_mlp": 0.01040959, + "balance_loss_clip": 1.04985881, + "balance_loss_mlp": 1.03018892, + "epoch": 0.2636926591715265, + "flos": 19244636616960.0, + "grad_norm": 2.4961109095361644, + "language_loss": 0.79198372, + "learning_rate": 3.4541135722888253e-06, + "loss": 0.81408542, + "num_input_tokens_seen": 47019625, + "step": 2193, + "time_per_iteration": 2.7668330669403076 + }, + { + "auxiliary_loss_clip": 0.01203905, + "auxiliary_loss_mlp": 0.0103616, + "balance_loss_clip": 1.05531073, + "balance_loss_mlp": 1.02554476, + "epoch": 0.26381290206216557, + "flos": 28804882734720.0, + "grad_norm": 1.68509227870695, + "language_loss": 0.80310786, + "learning_rate": 3.453578637734854e-06, + "loss": 0.82550853, + "num_input_tokens_seen": 47040815, + "step": 2194, + "time_per_iteration": 2.7492737770080566 + }, + { + "auxiliary_loss_clip": 0.01209312, + "auxiliary_loss_mlp": 0.0103716, + "balance_loss_clip": 1.06122661, + "balance_loss_mlp": 1.02635419, + "epoch": 0.2639331449528047, + "flos": 25008909436800.0, + "grad_norm": 1.5521361578421722, + "language_loss": 0.78343922, + "learning_rate": 3.4530434826741605e-06, + "loss": 0.80590391, + "num_input_tokens_seen": 47061755, + "step": 2195, + "time_per_iteration": 2.7331957817077637 + }, + { + "auxiliary_loss_clip": 0.01180339, + "auxiliary_loss_mlp": 0.01035378, + "balance_loss_clip": 1.05578125, + "balance_loss_mlp": 1.02585328, + "epoch": 0.26405338784344373, + "flos": 46535775465600.0, + "grad_norm": 1.5903264344380286, + "language_loss": 0.68561804, + "learning_rate": 3.452508107187926e-06, + "loss": 0.70777524, + "num_input_tokens_seen": 47085130, + "step": 2196, + "time_per_iteration": 2.998866558074951 + }, + { + "auxiliary_loss_clip": 0.0115669, + "auxiliary_loss_mlp": 0.01035231, + "balance_loss_clip": 1.0494523, + "balance_loss_mlp": 1.02428794, + "epoch": 0.26417363073408284, + "flos": 21179467641600.0, + "grad_norm": 2.1850258404041547, + "language_loss": 0.77071297, + "learning_rate": 3.451972511357366e-06, + "loss": 0.79263222, + "num_input_tokens_seen": 47104675, + "step": 2197, + "time_per_iteration": 2.845510482788086 + }, + { + "auxiliary_loss_clip": 0.0118864, + "auxiliary_loss_mlp": 0.01035358, + "balance_loss_clip": 1.05595875, + "balance_loss_mlp": 1.02572536, + "epoch": 0.26429387362472195, + "flos": 22674751937280.0, + "grad_norm": 1.6655849046237474, + "language_loss": 0.85138726, + "learning_rate": 3.45143669526373e-06, + "loss": 0.87362731, + "num_input_tokens_seen": 47124435, + "step": 2198, + "time_per_iteration": 2.7346479892730713 + }, + { + "auxiliary_loss_clip": 0.01084776, + "auxiliary_loss_mlp": 0.0100303, + "balance_loss_clip": 1.01791584, + "balance_loss_mlp": 1.00001359, + "epoch": 0.264414116515361, + "flos": 67180534272000.0, + "grad_norm": 0.8024582684914424, + "language_loss": 0.6321367, + "learning_rate": 3.450900658988302e-06, + "loss": 0.65301466, + "num_input_tokens_seen": 47185985, + "step": 2199, + "time_per_iteration": 4.161242246627808 + }, + { + "auxiliary_loss_clip": 0.01173817, + "auxiliary_loss_mlp": 0.01034412, + "balance_loss_clip": 1.05253732, + "balance_loss_mlp": 1.02384448, + "epoch": 0.2645343594060001, + "flos": 25664709997440.0, + "grad_norm": 1.7925336115563844, + "language_loss": 0.77821285, + "learning_rate": 3.450364402612397e-06, + "loss": 0.80029511, + "num_input_tokens_seen": 47203140, + "step": 2200, + "time_per_iteration": 2.9941670894622803 + }, + { + "auxiliary_loss_clip": 0.01182839, + "auxiliary_loss_mlp": 0.01030627, + "balance_loss_clip": 1.05419922, + "balance_loss_mlp": 1.01970768, + "epoch": 0.26465460229663923, + "flos": 22491822948480.0, + "grad_norm": 2.375553014462766, + "language_loss": 0.84042692, + "learning_rate": 3.449827926217366e-06, + "loss": 0.86256158, + "num_input_tokens_seen": 47222575, + "step": 2201, + "time_per_iteration": 3.898146629333496 + }, + { + "auxiliary_loss_clip": 0.01188724, + "auxiliary_loss_mlp": 0.01036158, + "balance_loss_clip": 1.05248523, + "balance_loss_mlp": 1.02548873, + "epoch": 0.2647748451872783, + "flos": 29388036038400.0, + "grad_norm": 1.7917179632062774, + "language_loss": 0.80528569, + "learning_rate": 3.449291229884591e-06, + "loss": 0.82753444, + "num_input_tokens_seen": 47243815, + "step": 2202, + "time_per_iteration": 3.807861328125 + }, + { + "auxiliary_loss_clip": 0.01175175, + "auxiliary_loss_mlp": 0.01035056, + "balance_loss_clip": 1.05002224, + "balance_loss_mlp": 1.02339745, + "epoch": 0.2648950880779174, + "flos": 26797799502720.0, + "grad_norm": 1.7828168772610662, + "language_loss": 0.86749995, + "learning_rate": 3.4487543136954887e-06, + "loss": 0.8896023, + "num_input_tokens_seen": 47263435, + "step": 2203, + "time_per_iteration": 2.873185873031616 + }, + { + "auxiliary_loss_clip": 0.01174193, + "auxiliary_loss_mlp": 0.01036182, + "balance_loss_clip": 1.05276275, + "balance_loss_mlp": 1.0257926, + "epoch": 0.2650153309685565, + "flos": 28841008838400.0, + "grad_norm": 1.7519471765288064, + "language_loss": 0.91163552, + "learning_rate": 3.448217177731509e-06, + "loss": 0.93373924, + "num_input_tokens_seen": 47283920, + "step": 2204, + "time_per_iteration": 2.9245359897613525 + }, + { + "auxiliary_loss_clip": 0.01180334, + "auxiliary_loss_mlp": 0.01038624, + "balance_loss_clip": 1.05666935, + "balance_loss_mlp": 1.0285567, + "epoch": 0.26513557385919556, + "flos": 20303247271680.0, + "grad_norm": 1.9741711154726935, + "language_loss": 0.77749693, + "learning_rate": 3.4476798220741348e-06, + "loss": 0.79968649, + "num_input_tokens_seen": 47302800, + "step": 2205, + "time_per_iteration": 2.911980390548706 + }, + { + "auxiliary_loss_clip": 0.01205883, + "auxiliary_loss_mlp": 0.01031249, + "balance_loss_clip": 1.06016731, + "balance_loss_mlp": 1.02129555, + "epoch": 0.26525581674983467, + "flos": 17676274101120.0, + "grad_norm": 1.6908005260517378, + "language_loss": 0.7808364, + "learning_rate": 3.4471422468048826e-06, + "loss": 0.80320776, + "num_input_tokens_seen": 47321525, + "step": 2206, + "time_per_iteration": 2.79296875 + }, + { + "auxiliary_loss_clip": 0.01189501, + "auxiliary_loss_mlp": 0.01036766, + "balance_loss_clip": 1.05758798, + "balance_loss_mlp": 1.02544665, + "epoch": 0.2653760596404738, + "flos": 26833746038400.0, + "grad_norm": 2.4332066698601653, + "language_loss": 0.73361218, + "learning_rate": 3.4466044520053022e-06, + "loss": 0.75587487, + "num_input_tokens_seen": 47340530, + "step": 2207, + "time_per_iteration": 2.873422384262085 + }, + { + "auxiliary_loss_clip": 0.01170833, + "auxiliary_loss_mlp": 0.01033038, + "balance_loss_clip": 1.04932559, + "balance_loss_mlp": 1.02295327, + "epoch": 0.26549630253111284, + "flos": 22782160581120.0, + "grad_norm": 1.8762239875803872, + "language_loss": 0.60182095, + "learning_rate": 3.446066437756977e-06, + "loss": 0.62385964, + "num_input_tokens_seen": 47359735, + "step": 2208, + "time_per_iteration": 2.8770506381988525 + }, + { + "auxiliary_loss_clip": 0.01183856, + "auxiliary_loss_mlp": 0.01040913, + "balance_loss_clip": 1.0553937, + "balance_loss_mlp": 1.02995813, + "epoch": 0.26561654542175195, + "flos": 23550002640000.0, + "grad_norm": 2.00461442343525, + "language_loss": 0.75103062, + "learning_rate": 3.4455282041415224e-06, + "loss": 0.7732783, + "num_input_tokens_seen": 47378945, + "step": 2209, + "time_per_iteration": 2.8452348709106445 + }, + { + "auxiliary_loss_clip": 0.01176125, + "auxiliary_loss_mlp": 0.01033904, + "balance_loss_clip": 1.05249238, + "balance_loss_mlp": 1.02319276, + "epoch": 0.265736788312391, + "flos": 26906680604160.0, + "grad_norm": 1.9782195582822661, + "language_loss": 0.87009948, + "learning_rate": 3.4449897512405894e-06, + "loss": 0.89219975, + "num_input_tokens_seen": 47398095, + "step": 2210, + "time_per_iteration": 2.9344582557678223 + }, + { + "auxiliary_loss_clip": 0.01147722, + "auxiliary_loss_mlp": 0.00888491, + "balance_loss_clip": 1.04566193, + "balance_loss_mlp": 1.0008738, + "epoch": 0.2658570312030301, + "flos": 23477139901440.0, + "grad_norm": 2.7997829310820865, + "language_loss": 0.75342214, + "learning_rate": 3.444451079135859e-06, + "loss": 0.77378428, + "num_input_tokens_seen": 47417605, + "step": 2211, + "time_per_iteration": 3.0225844383239746 + }, + { + "auxiliary_loss_clip": 0.01155935, + "auxiliary_loss_mlp": 0.00889403, + "balance_loss_clip": 1.04837072, + "balance_loss_mlp": 1.00074244, + "epoch": 0.2659772740936692, + "flos": 21866402315520.0, + "grad_norm": 1.9403118932463528, + "language_loss": 0.73995227, + "learning_rate": 3.4439121879090493e-06, + "loss": 0.76040566, + "num_input_tokens_seen": 47435385, + "step": 2212, + "time_per_iteration": 2.965657949447632 + }, + { + "auxiliary_loss_clip": 0.01187409, + "auxiliary_loss_mlp": 0.01040692, + "balance_loss_clip": 1.05462098, + "balance_loss_mlp": 1.02999294, + "epoch": 0.2660975169843083, + "flos": 19793100360960.0, + "grad_norm": 1.8331158464046473, + "language_loss": 0.83217382, + "learning_rate": 3.4433730776419082e-06, + "loss": 0.85445487, + "num_input_tokens_seen": 47454310, + "step": 2213, + "time_per_iteration": 2.8042855262756348 + }, + { + "auxiliary_loss_clip": 0.01194917, + "auxiliary_loss_mlp": 0.00889729, + "balance_loss_clip": 1.05320859, + "balance_loss_mlp": 1.00105929, + "epoch": 0.2662177598749474, + "flos": 29018981750400.0, + "grad_norm": 2.1698293002156466, + "language_loss": 0.80360413, + "learning_rate": 3.4428337484162183e-06, + "loss": 0.82445061, + "num_input_tokens_seen": 47475120, + "step": 2214, + "time_per_iteration": 2.8690123558044434 + }, + { + "auxiliary_loss_clip": 0.01180932, + "auxiliary_loss_mlp": 0.01040106, + "balance_loss_clip": 1.05207372, + "balance_loss_mlp": 1.02919841, + "epoch": 0.2663380027655865, + "flos": 21762549118080.0, + "grad_norm": 2.2508762510765696, + "language_loss": 0.8427304, + "learning_rate": 3.442294200313797e-06, + "loss": 0.86494088, + "num_input_tokens_seen": 47493150, + "step": 2215, + "time_per_iteration": 2.77116322517395 + }, + { + "auxiliary_loss_clip": 0.01094597, + "auxiliary_loss_mlp": 0.01004239, + "balance_loss_clip": 1.01854825, + "balance_loss_mlp": 1.00121069, + "epoch": 0.26645824565622556, + "flos": 66980333819520.0, + "grad_norm": 0.7687139689833956, + "language_loss": 0.527349, + "learning_rate": 3.4417544334164916e-06, + "loss": 0.54833734, + "num_input_tokens_seen": 47557295, + "step": 2216, + "time_per_iteration": 3.331724166870117 + }, + { + "auxiliary_loss_clip": 0.01174526, + "auxiliary_loss_mlp": 0.01043801, + "balance_loss_clip": 1.05502808, + "balance_loss_mlp": 1.03295851, + "epoch": 0.26657848854686467, + "flos": 25264198373760.0, + "grad_norm": 1.6801399942482522, + "language_loss": 0.77494234, + "learning_rate": 3.4412144478061854e-06, + "loss": 0.79712558, + "num_input_tokens_seen": 47579705, + "step": 2217, + "time_per_iteration": 2.9049057960510254 + }, + { + "auxiliary_loss_clip": 0.01130637, + "auxiliary_loss_mlp": 0.0104409, + "balance_loss_clip": 1.04364669, + "balance_loss_mlp": 1.03229392, + "epoch": 0.2666987314375038, + "flos": 23696769611520.0, + "grad_norm": 1.897438919917314, + "language_loss": 0.75460446, + "learning_rate": 3.4406742435647925e-06, + "loss": 0.77635175, + "num_input_tokens_seen": 47599770, + "step": 2218, + "time_per_iteration": 3.0053365230560303 + }, + { + "auxiliary_loss_clip": 0.01197127, + "auxiliary_loss_mlp": 0.01035755, + "balance_loss_clip": 1.05968773, + "balance_loss_mlp": 1.02609897, + "epoch": 0.26681897432814283, + "flos": 27048958375680.0, + "grad_norm": 2.1720581245908117, + "language_loss": 0.78664011, + "learning_rate": 3.440133820774263e-06, + "loss": 0.8089689, + "num_input_tokens_seen": 47619580, + "step": 2219, + "time_per_iteration": 2.7547719478607178 + }, + { + "auxiliary_loss_clip": 0.01192212, + "auxiliary_loss_mlp": 0.01035653, + "balance_loss_clip": 1.05854487, + "balance_loss_mlp": 1.02413774, + "epoch": 0.26693921721878194, + "flos": 28985944216320.0, + "grad_norm": 2.1059191471093226, + "language_loss": 0.82254022, + "learning_rate": 3.439593179516578e-06, + "loss": 0.84481883, + "num_input_tokens_seen": 47639490, + "step": 2220, + "time_per_iteration": 2.8829398155212402 + }, + { + "auxiliary_loss_clip": 0.01191725, + "auxiliary_loss_mlp": 0.01041583, + "balance_loss_clip": 1.05636215, + "balance_loss_mlp": 1.03012729, + "epoch": 0.26705946010942105, + "flos": 21507834798720.0, + "grad_norm": 1.7418253904466034, + "language_loss": 0.81124973, + "learning_rate": 3.4390523198737524e-06, + "loss": 0.83358282, + "num_input_tokens_seen": 47658650, + "step": 2221, + "time_per_iteration": 2.8642964363098145 + }, + { + "auxiliary_loss_clip": 0.0120835, + "auxiliary_loss_mlp": 0.00888759, + "balance_loss_clip": 1.05897927, + "balance_loss_mlp": 1.00097871, + "epoch": 0.2671797030000601, + "flos": 21471277731840.0, + "grad_norm": 1.6191820213590322, + "language_loss": 0.73387772, + "learning_rate": 3.4385112419278333e-06, + "loss": 0.75484878, + "num_input_tokens_seen": 47679875, + "step": 2222, + "time_per_iteration": 2.931201696395874 + }, + { + "auxiliary_loss_clip": 0.01090987, + "auxiliary_loss_mlp": 0.0100584, + "balance_loss_clip": 1.01861191, + "balance_loss_mlp": 1.00259781, + "epoch": 0.2672999458906992, + "flos": 64189929767040.0, + "grad_norm": 0.7864926740424641, + "language_loss": 0.64833426, + "learning_rate": 3.4379699457609033e-06, + "loss": 0.66930258, + "num_input_tokens_seen": 47737700, + "step": 2223, + "time_per_iteration": 3.1607587337493896 + }, + { + "auxiliary_loss_clip": 0.01179162, + "auxiliary_loss_mlp": 0.01033325, + "balance_loss_clip": 1.05399108, + "balance_loss_mlp": 1.02251863, + "epoch": 0.26742018878133833, + "flos": 16909042573440.0, + "grad_norm": 1.8895820762861222, + "language_loss": 0.90585816, + "learning_rate": 3.4374284314550755e-06, + "loss": 0.92798305, + "num_input_tokens_seen": 47756740, + "step": 2224, + "time_per_iteration": 2.763561725616455 + }, + { + "auxiliary_loss_clip": 0.01205619, + "auxiliary_loss_mlp": 0.01035693, + "balance_loss_clip": 1.05839264, + "balance_loss_mlp": 1.02482748, + "epoch": 0.2675404316719774, + "flos": 20667560964480.0, + "grad_norm": 2.042040706459028, + "language_loss": 0.80858016, + "learning_rate": 3.436886699092498e-06, + "loss": 0.83099329, + "num_input_tokens_seen": 47775255, + "step": 2225, + "time_per_iteration": 3.651859760284424 + }, + { + "auxiliary_loss_clip": 0.01209107, + "auxiliary_loss_mlp": 0.01040844, + "balance_loss_clip": 1.05820131, + "balance_loss_mlp": 1.02965641, + "epoch": 0.2676606745626165, + "flos": 17485013157120.0, + "grad_norm": 2.593374410646912, + "language_loss": 0.71928406, + "learning_rate": 3.4363447487553502e-06, + "loss": 0.74178362, + "num_input_tokens_seen": 47788570, + "step": 2226, + "time_per_iteration": 3.6000468730926514 + }, + { + "auxiliary_loss_clip": 0.01186093, + "auxiliary_loss_mlp": 0.01044462, + "balance_loss_clip": 1.05760455, + "balance_loss_mlp": 1.03347683, + "epoch": 0.26778091745325555, + "flos": 27852675143040.0, + "grad_norm": 1.9794159564285971, + "language_loss": 0.7793963, + "learning_rate": 3.4358025805258455e-06, + "loss": 0.8017019, + "num_input_tokens_seen": 47808275, + "step": 2227, + "time_per_iteration": 4.676478624343872 + }, + { + "auxiliary_loss_clip": 0.01170455, + "auxiliary_loss_mlp": 0.01034101, + "balance_loss_clip": 1.05048358, + "balance_loss_mlp": 1.02290702, + "epoch": 0.26790116034389466, + "flos": 20955995176320.0, + "grad_norm": 2.02714480802006, + "language_loss": 0.83467925, + "learning_rate": 3.435260194486232e-06, + "loss": 0.8567248, + "num_input_tokens_seen": 47826245, + "step": 2228, + "time_per_iteration": 2.9601552486419678 + }, + { + "auxiliary_loss_clip": 0.01185218, + "auxiliary_loss_mlp": 0.01035824, + "balance_loss_clip": 1.05451024, + "balance_loss_mlp": 1.02448702, + "epoch": 0.2680214032345338, + "flos": 18040659621120.0, + "grad_norm": 2.4352710153556005, + "language_loss": 0.8259939, + "learning_rate": 3.4347175907187875e-06, + "loss": 0.84820426, + "num_input_tokens_seen": 47843235, + "step": 2229, + "time_per_iteration": 2.823746919631958 + }, + { + "auxiliary_loss_clip": 0.01197533, + "auxiliary_loss_mlp": 0.01043579, + "balance_loss_clip": 1.05800545, + "balance_loss_mlp": 1.03274286, + "epoch": 0.26814164612517283, + "flos": 22419427086720.0, + "grad_norm": 1.8107905169686564, + "language_loss": 0.88374817, + "learning_rate": 3.4341747693058254e-06, + "loss": 0.90615934, + "num_input_tokens_seen": 47861710, + "step": 2230, + "time_per_iteration": 2.79581356048584 + }, + { + "auxiliary_loss_clip": 0.01127407, + "auxiliary_loss_mlp": 0.0104302, + "balance_loss_clip": 1.0448482, + "balance_loss_mlp": 1.03165364, + "epoch": 0.26826188901581194, + "flos": 35627371159680.0, + "grad_norm": 1.6856658861810043, + "language_loss": 0.77178872, + "learning_rate": 3.4336317303296916e-06, + "loss": 0.79349303, + "num_input_tokens_seen": 47882685, + "step": 2231, + "time_per_iteration": 3.3272321224212646 + }, + { + "auxiliary_loss_clip": 0.01193008, + "auxiliary_loss_mlp": 0.01031548, + "balance_loss_clip": 1.05604124, + "balance_loss_mlp": 1.02152205, + "epoch": 0.26838213190645105, + "flos": 17639788861440.0, + "grad_norm": 2.153454781169121, + "language_loss": 0.75835675, + "learning_rate": 3.4330884738727635e-06, + "loss": 0.78060234, + "num_input_tokens_seen": 47900860, + "step": 2232, + "time_per_iteration": 3.1244235038757324 + }, + { + "auxiliary_loss_clip": 0.01159995, + "auxiliary_loss_mlp": 0.0104308, + "balance_loss_clip": 1.05238724, + "balance_loss_mlp": 1.03233314, + "epoch": 0.2685023747970901, + "flos": 22674823764480.0, + "grad_norm": 1.8239480532200567, + "language_loss": 0.70652443, + "learning_rate": 3.4325450000174535e-06, + "loss": 0.72855514, + "num_input_tokens_seen": 47917500, + "step": 2233, + "time_per_iteration": 2.956653356552124 + }, + { + "auxiliary_loss_clip": 0.01156016, + "auxiliary_loss_mlp": 0.01039053, + "balance_loss_clip": 1.04989719, + "balance_loss_mlp": 1.02790129, + "epoch": 0.2686226176877292, + "flos": 20120533764480.0, + "grad_norm": 1.8773464044465478, + "language_loss": 0.74803829, + "learning_rate": 3.4320013088462067e-06, + "loss": 0.76998889, + "num_input_tokens_seen": 47934860, + "step": 2234, + "time_per_iteration": 2.783102512359619 + }, + { + "auxiliary_loss_clip": 0.01175798, + "auxiliary_loss_mlp": 0.01037925, + "balance_loss_clip": 1.04936624, + "balance_loss_mlp": 1.02781618, + "epoch": 0.2687428605783683, + "flos": 21872040750720.0, + "grad_norm": 1.41622771844822, + "language_loss": 0.81581521, + "learning_rate": 3.431457400441499e-06, + "loss": 0.83795238, + "num_input_tokens_seen": 47955255, + "step": 2235, + "time_per_iteration": 2.8559629917144775 + }, + { + "auxiliary_loss_clip": 0.01050024, + "auxiliary_loss_mlp": 0.01004557, + "balance_loss_clip": 1.01814485, + "balance_loss_mlp": 1.00138652, + "epoch": 0.2688631034690074, + "flos": 69943320766080.0, + "grad_norm": 0.9097838901775058, + "language_loss": 0.60869968, + "learning_rate": 3.4309132748858424e-06, + "loss": 0.62924552, + "num_input_tokens_seen": 48016245, + "step": 2236, + "time_per_iteration": 3.481004238128662 + }, + { + "auxiliary_loss_clip": 0.01193568, + "auxiliary_loss_mlp": 0.01029937, + "balance_loss_clip": 1.05878758, + "balance_loss_mlp": 1.01992917, + "epoch": 0.2689833463596465, + "flos": 22856639431680.0, + "grad_norm": 1.6216447961880405, + "language_loss": 0.83944762, + "learning_rate": 3.430368932261779e-06, + "loss": 0.86168271, + "num_input_tokens_seen": 48036600, + "step": 2237, + "time_per_iteration": 2.7614293098449707 + }, + { + "auxiliary_loss_clip": 0.01185381, + "auxiliary_loss_mlp": 0.01041089, + "balance_loss_clip": 1.05655682, + "balance_loss_mlp": 1.03044319, + "epoch": 0.2691035892502856, + "flos": 17200242132480.0, + "grad_norm": 2.4178875878108865, + "language_loss": 0.75100565, + "learning_rate": 3.429824372651886e-06, + "loss": 0.77327037, + "num_input_tokens_seen": 48054750, + "step": 2238, + "time_per_iteration": 2.754791736602783 + }, + { + "auxiliary_loss_clip": 0.01175784, + "auxiliary_loss_mlp": 0.01043209, + "balance_loss_clip": 1.05501246, + "balance_loss_mlp": 1.03240871, + "epoch": 0.26922383214092466, + "flos": 17747484814080.0, + "grad_norm": 1.841760002679146, + "language_loss": 0.83785129, + "learning_rate": 3.4292795961387732e-06, + "loss": 0.8600412, + "num_input_tokens_seen": 48072650, + "step": 2239, + "time_per_iteration": 2.817113161087036 + }, + { + "auxiliary_loss_clip": 0.01204962, + "auxiliary_loss_mlp": 0.01042201, + "balance_loss_clip": 1.05803871, + "balance_loss_mlp": 1.03101921, + "epoch": 0.26934407503156377, + "flos": 16173376122240.0, + "grad_norm": 2.261013912503632, + "language_loss": 0.88010144, + "learning_rate": 3.4287346028050818e-06, + "loss": 0.90257311, + "num_input_tokens_seen": 48088720, + "step": 2240, + "time_per_iteration": 2.682469367980957 + }, + { + "auxiliary_loss_clip": 0.01182091, + "auxiliary_loss_mlp": 0.01028171, + "balance_loss_clip": 1.05403781, + "balance_loss_mlp": 1.0183605, + "epoch": 0.2694643179222028, + "flos": 23732895715200.0, + "grad_norm": 1.4581189933873349, + "language_loss": 0.79660439, + "learning_rate": 3.4281893927334866e-06, + "loss": 0.81870699, + "num_input_tokens_seen": 48108630, + "step": 2241, + "time_per_iteration": 2.7721168994903564 + }, + { + "auxiliary_loss_clip": 0.01195004, + "auxiliary_loss_mlp": 0.01036034, + "balance_loss_clip": 1.05747569, + "balance_loss_mlp": 1.02601409, + "epoch": 0.26958456081284193, + "flos": 24718140840960.0, + "grad_norm": 2.0521668528427144, + "language_loss": 0.75323379, + "learning_rate": 3.4276439660066963e-06, + "loss": 0.77554411, + "num_input_tokens_seen": 48128330, + "step": 2242, + "time_per_iteration": 2.809239149093628 + }, + { + "auxiliary_loss_clip": 0.01202884, + "auxiliary_loss_mlp": 0.01034595, + "balance_loss_clip": 1.05813241, + "balance_loss_mlp": 1.02444446, + "epoch": 0.26970480370348104, + "flos": 18112588606080.0, + "grad_norm": 1.9377430440786803, + "language_loss": 0.84192681, + "learning_rate": 3.427098322707452e-06, + "loss": 0.86430162, + "num_input_tokens_seen": 48144295, + "step": 2243, + "time_per_iteration": 2.6652519702911377 + }, + { + "auxiliary_loss_clip": 0.01199538, + "auxiliary_loss_mlp": 0.01035278, + "balance_loss_clip": 1.06464815, + "balance_loss_mlp": 1.02537751, + "epoch": 0.2698250465941201, + "flos": 10816546250880.0, + "grad_norm": 2.123018786364472, + "language_loss": 0.89677215, + "learning_rate": 3.426552462918526e-06, + "loss": 0.91912031, + "num_input_tokens_seen": 48162230, + "step": 2244, + "time_per_iteration": 2.8362109661102295 + }, + { + "auxiliary_loss_clip": 0.01202915, + "auxiliary_loss_mlp": 0.01036713, + "balance_loss_clip": 1.05920291, + "balance_loss_mlp": 1.02674103, + "epoch": 0.2699452894847592, + "flos": 17308117653120.0, + "grad_norm": 2.172914077754276, + "language_loss": 0.72372234, + "learning_rate": 3.426006386722726e-06, + "loss": 0.74611861, + "num_input_tokens_seen": 48180290, + "step": 2245, + "time_per_iteration": 2.7968528270721436 + }, + { + "auxiliary_loss_clip": 0.01181778, + "auxiliary_loss_mlp": 0.01035952, + "balance_loss_clip": 1.05877757, + "balance_loss_mlp": 1.02502036, + "epoch": 0.2700655323753983, + "flos": 18078150441600.0, + "grad_norm": 1.749504722131175, + "language_loss": 0.92188823, + "learning_rate": 3.4254600942028914e-06, + "loss": 0.94406545, + "num_input_tokens_seen": 48198165, + "step": 2246, + "time_per_iteration": 2.958556890487671 + }, + { + "auxiliary_loss_clip": 0.01178609, + "auxiliary_loss_mlp": 0.01036098, + "balance_loss_clip": 1.05621791, + "balance_loss_mlp": 1.02609015, + "epoch": 0.2701857752660374, + "flos": 18186636493440.0, + "grad_norm": 3.6310578270531275, + "language_loss": 0.82340425, + "learning_rate": 3.424913585441893e-06, + "loss": 0.84555131, + "num_input_tokens_seen": 48216000, + "step": 2247, + "time_per_iteration": 2.870924472808838 + }, + { + "auxiliary_loss_clip": 0.01191569, + "auxiliary_loss_mlp": 0.01032174, + "balance_loss_clip": 1.05716264, + "balance_loss_mlp": 1.02249408, + "epoch": 0.2703060181566765, + "flos": 16319496648960.0, + "grad_norm": 2.5382837019648554, + "language_loss": 0.87384832, + "learning_rate": 3.4243668605226374e-06, + "loss": 0.8960858, + "num_input_tokens_seen": 48233025, + "step": 2248, + "time_per_iteration": 2.7406558990478516 + }, + { + "auxiliary_loss_clip": 0.01174488, + "auxiliary_loss_mlp": 0.01039871, + "balance_loss_clip": 1.05407453, + "balance_loss_mlp": 1.02938056, + "epoch": 0.2704262610473156, + "flos": 19572357329280.0, + "grad_norm": 2.1979204614125147, + "language_loss": 0.82563853, + "learning_rate": 3.423819919528061e-06, + "loss": 0.84778208, + "num_input_tokens_seen": 48251110, + "step": 2249, + "time_per_iteration": 2.908931255340576 + }, + { + "auxiliary_loss_clip": 0.01168588, + "auxiliary_loss_mlp": 0.01030861, + "balance_loss_clip": 1.04861474, + "balance_loss_mlp": 1.02046049, + "epoch": 0.27054650393795465, + "flos": 20740746925440.0, + "grad_norm": 1.8033068092077074, + "language_loss": 0.78206289, + "learning_rate": 3.4232727625411355e-06, + "loss": 0.80405736, + "num_input_tokens_seen": 48270215, + "step": 2250, + "time_per_iteration": 2.971139907836914 + }, + { + "auxiliary_loss_clip": 0.01147512, + "auxiliary_loss_mlp": 0.01036925, + "balance_loss_clip": 1.04959917, + "balance_loss_mlp": 1.02685761, + "epoch": 0.27066674682859376, + "flos": 18658322916480.0, + "grad_norm": 1.6777511312310267, + "language_loss": 0.86314261, + "learning_rate": 3.4227253896448626e-06, + "loss": 0.884987, + "num_input_tokens_seen": 48288075, + "step": 2251, + "time_per_iteration": 4.168755531311035 + }, + { + "auxiliary_loss_clip": 0.01200115, + "auxiliary_loss_mlp": 0.01035703, + "balance_loss_clip": 1.05552149, + "balance_loss_mlp": 1.0261426, + "epoch": 0.2707869897192329, + "flos": 23002759958400.0, + "grad_norm": 2.107295976430965, + "language_loss": 0.82199562, + "learning_rate": 3.42217780092228e-06, + "loss": 0.8443538, + "num_input_tokens_seen": 48306415, + "step": 2252, + "time_per_iteration": 3.9531800746917725 + }, + { + "auxiliary_loss_clip": 0.01079578, + "auxiliary_loss_mlp": 0.01003514, + "balance_loss_clip": 1.01757836, + "balance_loss_mlp": 1.00060546, + "epoch": 0.27090723260987193, + "flos": 58323240293760.0, + "grad_norm": 0.7982760062183004, + "language_loss": 0.60383248, + "learning_rate": 3.421629996456456e-06, + "loss": 0.62466341, + "num_input_tokens_seen": 48365035, + "step": 2253, + "time_per_iteration": 4.360921859741211 + }, + { + "auxiliary_loss_clip": 0.01191013, + "auxiliary_loss_mlp": 0.01042705, + "balance_loss_clip": 1.05464864, + "balance_loss_mlp": 1.03233981, + "epoch": 0.27102747550051104, + "flos": 11984540797440.0, + "grad_norm": 2.1791524547595973, + "language_loss": 0.82642066, + "learning_rate": 3.421081976330491e-06, + "loss": 0.8487578, + "num_input_tokens_seen": 48383550, + "step": 2254, + "time_per_iteration": 2.7408645153045654 + }, + { + "auxiliary_loss_clip": 0.0117894, + "auxiliary_loss_mlp": 0.0103749, + "balance_loss_clip": 1.05096769, + "balance_loss_mlp": 1.02715516, + "epoch": 0.27114771839115015, + "flos": 19900401264000.0, + "grad_norm": 1.7744379712728142, + "language_loss": 0.88225746, + "learning_rate": 3.4205337406275207e-06, + "loss": 0.90442181, + "num_input_tokens_seen": 48403670, + "step": 2255, + "time_per_iteration": 2.895141363143921 + }, + { + "auxiliary_loss_clip": 0.0120115, + "auxiliary_loss_mlp": 0.01031251, + "balance_loss_clip": 1.05608892, + "balance_loss_mlp": 1.02118373, + "epoch": 0.2712679612817892, + "flos": 18331966920960.0, + "grad_norm": 2.835605436742648, + "language_loss": 0.74883163, + "learning_rate": 3.4199852894307114e-06, + "loss": 0.77115566, + "num_input_tokens_seen": 48420420, + "step": 2256, + "time_per_iteration": 2.737362861633301 + }, + { + "auxiliary_loss_clip": 0.01157187, + "auxiliary_loss_mlp": 0.01031204, + "balance_loss_clip": 1.05396485, + "balance_loss_mlp": 1.02060676, + "epoch": 0.2713882041724283, + "flos": 24460302038400.0, + "grad_norm": 2.3247552890721055, + "language_loss": 0.78994346, + "learning_rate": 3.419436622823262e-06, + "loss": 0.81182742, + "num_input_tokens_seen": 48441140, + "step": 2257, + "time_per_iteration": 2.9913747310638428 + }, + { + "auxiliary_loss_clip": 0.01180329, + "auxiliary_loss_mlp": 0.01034653, + "balance_loss_clip": 1.05421495, + "balance_loss_mlp": 1.02445507, + "epoch": 0.27150844706306737, + "flos": 23039317025280.0, + "grad_norm": 1.5445140843611957, + "language_loss": 0.74324155, + "learning_rate": 3.4188877408884063e-06, + "loss": 0.76539141, + "num_input_tokens_seen": 48461845, + "step": 2258, + "time_per_iteration": 2.862292528152466 + }, + { + "auxiliary_loss_clip": 0.01173053, + "auxiliary_loss_mlp": 0.01032522, + "balance_loss_clip": 1.05111027, + "balance_loss_mlp": 1.0221802, + "epoch": 0.2716286899537065, + "flos": 22563644192640.0, + "grad_norm": 2.4950033090756096, + "language_loss": 0.65676534, + "learning_rate": 3.4183386437094088e-06, + "loss": 0.67882109, + "num_input_tokens_seen": 48478510, + "step": 2259, + "time_per_iteration": 2.8780531883239746 + }, + { + "auxiliary_loss_clip": 0.01182526, + "auxiliary_loss_mlp": 0.01028365, + "balance_loss_clip": 1.05312777, + "balance_loss_mlp": 1.01833344, + "epoch": 0.2717489328443456, + "flos": 13115044523520.0, + "grad_norm": 2.210775754743297, + "language_loss": 0.82184762, + "learning_rate": 3.417789331369565e-06, + "loss": 0.84395659, + "num_input_tokens_seen": 48494300, + "step": 2260, + "time_per_iteration": 2.780642032623291 + }, + { + "auxiliary_loss_clip": 0.01207763, + "auxiliary_loss_mlp": 0.01038351, + "balance_loss_clip": 1.06075013, + "balance_loss_mlp": 1.02777696, + "epoch": 0.27186917573498465, + "flos": 29278688060160.0, + "grad_norm": 1.9698027757223637, + "language_loss": 0.91246367, + "learning_rate": 3.4172398039522088e-06, + "loss": 0.93492478, + "num_input_tokens_seen": 48515585, + "step": 2261, + "time_per_iteration": 2.8227970600128174 + }, + { + "auxiliary_loss_clip": 0.01192145, + "auxiliary_loss_mlp": 0.0103409, + "balance_loss_clip": 1.05428874, + "balance_loss_mlp": 1.02399898, + "epoch": 0.27198941862562376, + "flos": 26032220000640.0, + "grad_norm": 1.6910958413194113, + "language_loss": 0.79631084, + "learning_rate": 3.4166900615407e-06, + "loss": 0.81857318, + "num_input_tokens_seen": 48533500, + "step": 2262, + "time_per_iteration": 2.7978591918945312 + }, + { + "auxiliary_loss_clip": 0.01191486, + "auxiliary_loss_mlp": 0.01033363, + "balance_loss_clip": 1.05712128, + "balance_loss_mlp": 1.02323651, + "epoch": 0.27210966151626287, + "flos": 32780983760640.0, + "grad_norm": 2.2709074556057187, + "language_loss": 0.75044239, + "learning_rate": 3.416140104218436e-06, + "loss": 0.77269083, + "num_input_tokens_seen": 48552865, + "step": 2263, + "time_per_iteration": 2.845167875289917 + }, + { + "auxiliary_loss_clip": 0.01080781, + "auxiliary_loss_mlp": 0.00876905, + "balance_loss_clip": 1.01735973, + "balance_loss_mlp": 1.0009234, + "epoch": 0.2722299044069019, + "flos": 65471043219840.0, + "grad_norm": 0.8420233781425583, + "language_loss": 0.69699979, + "learning_rate": 3.4155899320688437e-06, + "loss": 0.7165767, + "num_input_tokens_seen": 48618940, + "step": 2264, + "time_per_iteration": 3.418057918548584 + }, + { + "auxiliary_loss_clip": 0.01157619, + "auxiliary_loss_mlp": 0.01040437, + "balance_loss_clip": 1.05335474, + "balance_loss_mlp": 1.03032231, + "epoch": 0.27235014729754103, + "flos": 15334143782400.0, + "grad_norm": 2.1589060877045942, + "language_loss": 0.74050653, + "learning_rate": 3.415039545175384e-06, + "loss": 0.76248705, + "num_input_tokens_seen": 48634665, + "step": 2265, + "time_per_iteration": 2.943946361541748 + }, + { + "auxiliary_loss_clip": 0.01194137, + "auxiliary_loss_mlp": 0.01032719, + "balance_loss_clip": 1.05437922, + "balance_loss_mlp": 1.02323556, + "epoch": 0.27247039018818014, + "flos": 21872363973120.0, + "grad_norm": 2.153234257570348, + "language_loss": 0.65340281, + "learning_rate": 3.414488943621551e-06, + "loss": 0.67567134, + "num_input_tokens_seen": 48653330, + "step": 2266, + "time_per_iteration": 2.761946439743042 + }, + { + "auxiliary_loss_clip": 0.01190764, + "auxiliary_loss_mlp": 0.01035855, + "balance_loss_clip": 1.05781531, + "balance_loss_mlp": 1.02561486, + "epoch": 0.2725906330788192, + "flos": 18695490514560.0, + "grad_norm": 1.789850906557245, + "language_loss": 0.73389745, + "learning_rate": 3.41393812749087e-06, + "loss": 0.75616366, + "num_input_tokens_seen": 48671375, + "step": 2267, + "time_per_iteration": 2.741211175918579 + }, + { + "auxiliary_loss_clip": 0.01180401, + "auxiliary_loss_mlp": 0.01035544, + "balance_loss_clip": 1.05507922, + "balance_loss_mlp": 1.02579904, + "epoch": 0.2727108759694583, + "flos": 17886099398400.0, + "grad_norm": 3.1945389556424764, + "language_loss": 0.71945059, + "learning_rate": 3.4133870968668984e-06, + "loss": 0.74160999, + "num_input_tokens_seen": 48686175, + "step": 2268, + "time_per_iteration": 2.755796432495117 + }, + { + "auxiliary_loss_clip": 0.01181449, + "auxiliary_loss_mlp": 0.01030789, + "balance_loss_clip": 1.05265605, + "balance_loss_mlp": 1.0200125, + "epoch": 0.2728311188600974, + "flos": 24461666755200.0, + "grad_norm": 1.7746181888177381, + "language_loss": 0.78856087, + "learning_rate": 3.412835851833229e-06, + "loss": 0.81068325, + "num_input_tokens_seen": 48708370, + "step": 2269, + "time_per_iteration": 2.825782537460327 + }, + { + "auxiliary_loss_clip": 0.01191603, + "auxiliary_loss_mlp": 0.01039955, + "balance_loss_clip": 1.05835319, + "balance_loss_mlp": 1.0299418, + "epoch": 0.2729513617507365, + "flos": 30993314757120.0, + "grad_norm": 1.7501708898973647, + "language_loss": 0.78096586, + "learning_rate": 3.4122843924734834e-06, + "loss": 0.80328143, + "num_input_tokens_seen": 48730670, + "step": 2270, + "time_per_iteration": 2.8141837120056152 + }, + { + "auxiliary_loss_clip": 0.011754, + "auxiliary_loss_mlp": 0.01037897, + "balance_loss_clip": 1.05108678, + "balance_loss_mlp": 1.02725196, + "epoch": 0.2730716046413756, + "flos": 19094637421440.0, + "grad_norm": 1.9172423343051286, + "language_loss": 0.88422096, + "learning_rate": 3.411732718871319e-06, + "loss": 0.90635389, + "num_input_tokens_seen": 48746510, + "step": 2271, + "time_per_iteration": 2.705366373062134 + }, + { + "auxiliary_loss_clip": 0.01201387, + "auxiliary_loss_mlp": 0.01032197, + "balance_loss_clip": 1.06028867, + "balance_loss_mlp": 1.02275586, + "epoch": 0.27319184753201464, + "flos": 26944566474240.0, + "grad_norm": 1.6669780853592395, + "language_loss": 0.78736567, + "learning_rate": 3.4111808311104227e-06, + "loss": 0.80970144, + "num_input_tokens_seen": 48768825, + "step": 2272, + "time_per_iteration": 2.7569901943206787 + }, + { + "auxiliary_loss_clip": 0.01187755, + "auxiliary_loss_mlp": 0.01041781, + "balance_loss_clip": 1.05290592, + "balance_loss_mlp": 1.03217256, + "epoch": 0.27331209042265375, + "flos": 31759828012800.0, + "grad_norm": 2.0068998640547657, + "language_loss": 0.69458294, + "learning_rate": 3.410628729274517e-06, + "loss": 0.71687829, + "num_input_tokens_seen": 48790345, + "step": 2273, + "time_per_iteration": 2.8896753787994385 + }, + { + "auxiliary_loss_clip": 0.01180113, + "auxiliary_loss_mlp": 0.00887867, + "balance_loss_clip": 1.05675697, + "balance_loss_mlp": 1.00055993, + "epoch": 0.27343233331329286, + "flos": 25739081107200.0, + "grad_norm": 1.8075401039750612, + "language_loss": 0.82551229, + "learning_rate": 3.4100764134473546e-06, + "loss": 0.84619206, + "num_input_tokens_seen": 48809630, + "step": 2274, + "time_per_iteration": 2.8427765369415283 + }, + { + "auxiliary_loss_clip": 0.01202364, + "auxiliary_loss_mlp": 0.01037989, + "balance_loss_clip": 1.05874419, + "balance_loss_mlp": 1.0279038, + "epoch": 0.2735525762039319, + "flos": 24389414547840.0, + "grad_norm": 2.155264345614224, + "language_loss": 0.84852928, + "learning_rate": 3.4095238837127215e-06, + "loss": 0.87093282, + "num_input_tokens_seen": 48828770, + "step": 2275, + "time_per_iteration": 2.7715916633605957 + }, + { + "auxiliary_loss_clip": 0.01168195, + "auxiliary_loss_mlp": 0.01031879, + "balance_loss_clip": 1.0511806, + "balance_loss_mlp": 1.0220325, + "epoch": 0.27367281909457103, + "flos": 14465357527680.0, + "grad_norm": 2.3272915741629805, + "language_loss": 0.79543877, + "learning_rate": 3.4089711401544355e-06, + "loss": 0.81743956, + "num_input_tokens_seen": 48846365, + "step": 2276, + "time_per_iteration": 2.876896858215332 + }, + { + "auxiliary_loss_clip": 0.01191003, + "auxiliary_loss_mlp": 0.01032859, + "balance_loss_clip": 1.05471146, + "balance_loss_mlp": 1.02323937, + "epoch": 0.27379306198521014, + "flos": 23476996247040.0, + "grad_norm": 2.575062194064457, + "language_loss": 0.6741699, + "learning_rate": 3.4084181828563486e-06, + "loss": 0.69640851, + "num_input_tokens_seen": 48863085, + "step": 2277, + "time_per_iteration": 3.6260714530944824 + }, + { + "auxiliary_loss_clip": 0.01157451, + "auxiliary_loss_mlp": 0.01034383, + "balance_loss_clip": 1.04929161, + "balance_loss_mlp": 1.02388072, + "epoch": 0.2739133048758492, + "flos": 17458152762240.0, + "grad_norm": 2.2182934678103745, + "language_loss": 0.70573139, + "learning_rate": 3.4078650119023428e-06, + "loss": 0.72764981, + "num_input_tokens_seen": 48881400, + "step": 2278, + "time_per_iteration": 2.8954148292541504 + }, + { + "auxiliary_loss_clip": 0.01154409, + "auxiliary_loss_mlp": 0.010434, + "balance_loss_clip": 1.04617953, + "balance_loss_mlp": 1.03281987, + "epoch": 0.2740335477664883, + "flos": 19273113123840.0, + "grad_norm": 2.0657068527498437, + "language_loss": 0.74138689, + "learning_rate": 3.4073116273763337e-06, + "loss": 0.76336503, + "num_input_tokens_seen": 48895845, + "step": 2279, + "time_per_iteration": 4.789755582809448 + }, + { + "auxiliary_loss_clip": 0.01183877, + "auxiliary_loss_mlp": 0.01033911, + "balance_loss_clip": 1.05151987, + "balance_loss_mlp": 1.02319396, + "epoch": 0.2741537906571274, + "flos": 26104723603200.0, + "grad_norm": 1.6914578167178071, + "language_loss": 0.80892122, + "learning_rate": 3.40675802936227e-06, + "loss": 0.83109915, + "num_input_tokens_seen": 48916630, + "step": 2280, + "time_per_iteration": 2.8848845958709717 + }, + { + "auxiliary_loss_clip": 0.01175135, + "auxiliary_loss_mlp": 0.01029255, + "balance_loss_clip": 1.05266082, + "balance_loss_mlp": 1.01917553, + "epoch": 0.27427403354776647, + "flos": 34164190644480.0, + "grad_norm": 1.9773831624083107, + "language_loss": 0.71653962, + "learning_rate": 3.4062042179441318e-06, + "loss": 0.73858351, + "num_input_tokens_seen": 48937100, + "step": 2281, + "time_per_iteration": 2.917630672454834 + }, + { + "auxiliary_loss_clip": 0.01188565, + "auxiliary_loss_mlp": 0.01032285, + "balance_loss_clip": 1.05756211, + "balance_loss_mlp": 1.02324891, + "epoch": 0.2743942764384056, + "flos": 18766988536320.0, + "grad_norm": 1.9936650572252237, + "language_loss": 0.80901963, + "learning_rate": 3.4056501932059314e-06, + "loss": 0.83122808, + "num_input_tokens_seen": 48955175, + "step": 2282, + "time_per_iteration": 2.8153207302093506 + }, + { + "auxiliary_loss_clip": 0.0109715, + "auxiliary_loss_mlp": 0.0100436, + "balance_loss_clip": 1.02176595, + "balance_loss_mlp": 1.00165403, + "epoch": 0.2745145193290447, + "flos": 64904048058240.0, + "grad_norm": 0.7755394545275611, + "language_loss": 0.58128572, + "learning_rate": 3.405095955231715e-06, + "loss": 0.60230088, + "num_input_tokens_seen": 49006830, + "step": 2283, + "time_per_iteration": 3.1988377571105957 + }, + { + "auxiliary_loss_clip": 0.01195571, + "auxiliary_loss_mlp": 0.01026698, + "balance_loss_clip": 1.05608463, + "balance_loss_mlp": 1.017102, + "epoch": 0.27463476221968375, + "flos": 16136926796160.0, + "grad_norm": 2.231058863209745, + "language_loss": 0.94271678, + "learning_rate": 3.4045415041055585e-06, + "loss": 0.96493948, + "num_input_tokens_seen": 49022470, + "step": 2284, + "time_per_iteration": 2.7139573097229004 + }, + { + "auxiliary_loss_clip": 0.01187185, + "auxiliary_loss_mlp": 0.01035236, + "balance_loss_clip": 1.05736208, + "balance_loss_mlp": 1.02484703, + "epoch": 0.27475500511032286, + "flos": 10376712213120.0, + "grad_norm": 2.3283540221384786, + "language_loss": 0.78415871, + "learning_rate": 3.4039868399115728e-06, + "loss": 0.80638289, + "num_input_tokens_seen": 49037110, + "step": 2285, + "time_per_iteration": 2.898486852645874 + }, + { + "auxiliary_loss_clip": 0.01152698, + "auxiliary_loss_mlp": 0.01033369, + "balance_loss_clip": 1.0497086, + "balance_loss_mlp": 1.02319479, + "epoch": 0.27487524800096197, + "flos": 17311062568320.0, + "grad_norm": 1.7193605392932787, + "language_loss": 0.80171072, + "learning_rate": 3.4034319627339003e-06, + "loss": 0.82357138, + "num_input_tokens_seen": 49053975, + "step": 2286, + "time_per_iteration": 2.9502007961273193 + }, + { + "auxiliary_loss_clip": 0.01183424, + "auxiliary_loss_mlp": 0.01034865, + "balance_loss_clip": 1.05577469, + "balance_loss_mlp": 1.02527499, + "epoch": 0.274995490891601, + "flos": 27120205002240.0, + "grad_norm": 3.320496571150844, + "language_loss": 0.69785762, + "learning_rate": 3.402876872656715e-06, + "loss": 0.72004044, + "num_input_tokens_seen": 49072295, + "step": 2287, + "time_per_iteration": 2.8507769107818604 + }, + { + "auxiliary_loss_clip": 0.01179392, + "auxiliary_loss_mlp": 0.01037505, + "balance_loss_clip": 1.05597436, + "balance_loss_mlp": 1.02716398, + "epoch": 0.27511573378224013, + "flos": 23436093634560.0, + "grad_norm": 1.7857371070545804, + "language_loss": 0.89943671, + "learning_rate": 3.402321569764223e-06, + "loss": 0.92160571, + "num_input_tokens_seen": 49091600, + "step": 2288, + "time_per_iteration": 2.8428475856781006 + }, + { + "auxiliary_loss_clip": 0.01162481, + "auxiliary_loss_mlp": 0.00887419, + "balance_loss_clip": 1.04888654, + "balance_loss_mlp": 1.00069749, + "epoch": 0.2752359766728792, + "flos": 16722019434240.0, + "grad_norm": 2.905020126619075, + "language_loss": 0.83480674, + "learning_rate": 3.4017660541406635e-06, + "loss": 0.85530573, + "num_input_tokens_seen": 49107665, + "step": 2289, + "time_per_iteration": 2.8772833347320557 + }, + { + "auxiliary_loss_clip": 0.01186266, + "auxiliary_loss_mlp": 0.01029663, + "balance_loss_clip": 1.05176198, + "balance_loss_mlp": 1.02022767, + "epoch": 0.2753562195635183, + "flos": 25297738698240.0, + "grad_norm": 1.7463800836191798, + "language_loss": 0.74228966, + "learning_rate": 3.4012103258703092e-06, + "loss": 0.764449, + "num_input_tokens_seen": 49126420, + "step": 2290, + "time_per_iteration": 2.8720831871032715 + }, + { + "auxiliary_loss_clip": 0.01168761, + "auxiliary_loss_mlp": 0.0103546, + "balance_loss_clip": 1.04997396, + "balance_loss_mlp": 1.02585161, + "epoch": 0.2754764624541574, + "flos": 27338972785920.0, + "grad_norm": 2.0870473675932044, + "language_loss": 0.83154583, + "learning_rate": 3.4006543850374616e-06, + "loss": 0.8535881, + "num_input_tokens_seen": 49141470, + "step": 2291, + "time_per_iteration": 2.7881317138671875 + }, + { + "auxiliary_loss_clip": 0.01193318, + "auxiliary_loss_mlp": 0.01039367, + "balance_loss_clip": 1.05529189, + "balance_loss_mlp": 1.02983665, + "epoch": 0.27559670534479647, + "flos": 17238379397760.0, + "grad_norm": 2.0707220315632218, + "language_loss": 0.74866599, + "learning_rate": 3.400098231726458e-06, + "loss": 0.77099288, + "num_input_tokens_seen": 49158570, + "step": 2292, + "time_per_iteration": 2.7703704833984375 + }, + { + "auxiliary_loss_clip": 0.01174481, + "auxiliary_loss_mlp": 0.01030136, + "balance_loss_clip": 1.04957342, + "balance_loss_mlp": 1.01908529, + "epoch": 0.2757169482354356, + "flos": 21939085486080.0, + "grad_norm": 2.557629958222165, + "language_loss": 0.87100053, + "learning_rate": 3.3995418660216657e-06, + "loss": 0.89304674, + "num_input_tokens_seen": 49176025, + "step": 2293, + "time_per_iteration": 2.8181087970733643 + }, + { + "auxiliary_loss_clip": 0.01205032, + "auxiliary_loss_mlp": 0.01034541, + "balance_loss_clip": 1.05720687, + "balance_loss_mlp": 1.02385378, + "epoch": 0.2758371911260747, + "flos": 20850669521280.0, + "grad_norm": 2.282413711092182, + "language_loss": 0.80029082, + "learning_rate": 3.3989852880074848e-06, + "loss": 0.82268655, + "num_input_tokens_seen": 49197455, + "step": 2294, + "time_per_iteration": 2.7452738285064697 + }, + { + "auxiliary_loss_clip": 0.01091488, + "auxiliary_loss_mlp": 0.01009315, + "balance_loss_clip": 1.03358579, + "balance_loss_mlp": 1.00616741, + "epoch": 0.27595743401671374, + "flos": 69269063592960.0, + "grad_norm": 0.747706937801353, + "language_loss": 0.6059252, + "learning_rate": 3.398428497768348e-06, + "loss": 0.62693322, + "num_input_tokens_seen": 49262625, + "step": 2295, + "time_per_iteration": 3.4253973960876465 + }, + { + "auxiliary_loss_clip": 0.01177172, + "auxiliary_loss_mlp": 0.01033305, + "balance_loss_clip": 1.04944348, + "balance_loss_mlp": 1.02320254, + "epoch": 0.27607767690735285, + "flos": 21215019127680.0, + "grad_norm": 1.7367660426880163, + "language_loss": 0.71898568, + "learning_rate": 3.3978714953887205e-06, + "loss": 0.74109054, + "num_input_tokens_seen": 49282380, + "step": 2296, + "time_per_iteration": 2.902225971221924 + }, + { + "auxiliary_loss_clip": 0.0114885, + "auxiliary_loss_mlp": 0.01032427, + "balance_loss_clip": 1.04567027, + "balance_loss_mlp": 1.02313471, + "epoch": 0.27619791979799196, + "flos": 24825334003200.0, + "grad_norm": 1.6484161314862162, + "language_loss": 0.86049169, + "learning_rate": 3.397314280953098e-06, + "loss": 0.88230443, + "num_input_tokens_seen": 49303205, + "step": 2297, + "time_per_iteration": 2.9919631481170654 + }, + { + "auxiliary_loss_clip": 0.01174201, + "auxiliary_loss_mlp": 0.01030919, + "balance_loss_clip": 1.05177903, + "balance_loss_mlp": 1.02139449, + "epoch": 0.276318162688631, + "flos": 24753548672640.0, + "grad_norm": 1.9037176451353413, + "language_loss": 0.80654836, + "learning_rate": 3.3967568545460108e-06, + "loss": 0.82859957, + "num_input_tokens_seen": 49322745, + "step": 2298, + "time_per_iteration": 2.7488744258880615 + }, + { + "auxiliary_loss_clip": 0.01188066, + "auxiliary_loss_mlp": 0.01036378, + "balance_loss_clip": 1.05663443, + "balance_loss_mlp": 1.02635217, + "epoch": 0.27643840557927013, + "flos": 18150007599360.0, + "grad_norm": 1.9710604728660546, + "language_loss": 0.80554688, + "learning_rate": 3.3961992162520185e-06, + "loss": 0.82779133, + "num_input_tokens_seen": 49341370, + "step": 2299, + "time_per_iteration": 2.7504308223724365 + }, + { + "auxiliary_loss_clip": 0.01193714, + "auxiliary_loss_mlp": 0.01037688, + "balance_loss_clip": 1.0564698, + "balance_loss_mlp": 1.02856827, + "epoch": 0.27655864846990924, + "flos": 24823933372800.0, + "grad_norm": 2.034274302016185, + "language_loss": 0.71995008, + "learning_rate": 3.3956413661557156e-06, + "loss": 0.74226415, + "num_input_tokens_seen": 49361545, + "step": 2300, + "time_per_iteration": 2.7884528636932373 + }, + { + "auxiliary_loss_clip": 0.01177183, + "auxiliary_loss_mlp": 0.01033794, + "balance_loss_clip": 1.05122995, + "balance_loss_mlp": 1.02398312, + "epoch": 0.2766788913605483, + "flos": 20266582464000.0, + "grad_norm": 1.9945469409940746, + "language_loss": 0.66371244, + "learning_rate": 3.3950833043417273e-06, + "loss": 0.68582225, + "num_input_tokens_seen": 49379690, + "step": 2301, + "time_per_iteration": 2.8943634033203125 + }, + { + "auxiliary_loss_clip": 0.01194962, + "auxiliary_loss_mlp": 0.01033502, + "balance_loss_clip": 1.05929625, + "balance_loss_mlp": 1.02309489, + "epoch": 0.2767991342511874, + "flos": 21470272151040.0, + "grad_norm": 2.267452436002702, + "language_loss": 0.73083591, + "learning_rate": 3.3945250308947105e-06, + "loss": 0.75312054, + "num_input_tokens_seen": 49395995, + "step": 2302, + "time_per_iteration": 2.7981204986572266 + }, + { + "auxiliary_loss_clip": 0.01092462, + "auxiliary_loss_mlp": 0.01005745, + "balance_loss_clip": 1.02364326, + "balance_loss_mlp": 1.00301552, + "epoch": 0.2769193771418265, + "flos": 66002627571840.0, + "grad_norm": 1.2359286990669234, + "language_loss": 0.68266237, + "learning_rate": 3.3939665458993556e-06, + "loss": 0.70364445, + "num_input_tokens_seen": 49450415, + "step": 2303, + "time_per_iteration": 4.164350986480713 + }, + { + "auxiliary_loss_clip": 0.01178506, + "auxiliary_loss_mlp": 0.01032998, + "balance_loss_clip": 1.05241311, + "balance_loss_mlp": 1.02312708, + "epoch": 0.27703962003246557, + "flos": 20704441253760.0, + "grad_norm": 1.816951771096578, + "language_loss": 0.77037984, + "learning_rate": 3.3934078494403843e-06, + "loss": 0.79249489, + "num_input_tokens_seen": 49469990, + "step": 2304, + "time_per_iteration": 3.853444814682007 + }, + { + "auxiliary_loss_clip": 0.01137138, + "auxiliary_loss_mlp": 0.00888398, + "balance_loss_clip": 1.0483861, + "balance_loss_mlp": 1.00069499, + "epoch": 0.2771598629231047, + "flos": 22929897219840.0, + "grad_norm": 1.7237438094452735, + "language_loss": 0.81206143, + "learning_rate": 3.3928489416025495e-06, + "loss": 0.83231676, + "num_input_tokens_seen": 49490835, + "step": 2305, + "time_per_iteration": 3.95326566696167 + }, + { + "auxiliary_loss_clip": 0.01183435, + "auxiliary_loss_mlp": 0.01031799, + "balance_loss_clip": 1.05574393, + "balance_loss_mlp": 1.02014041, + "epoch": 0.27728010581374374, + "flos": 18369457741440.0, + "grad_norm": 2.150614960123558, + "language_loss": 0.7910617, + "learning_rate": 3.392289822470638e-06, + "loss": 0.81321406, + "num_input_tokens_seen": 49508815, + "step": 2306, + "time_per_iteration": 2.805544137954712 + }, + { + "auxiliary_loss_clip": 0.01179043, + "auxiliary_loss_mlp": 0.01034121, + "balance_loss_clip": 1.05336595, + "balance_loss_mlp": 1.02439356, + "epoch": 0.27740034870438285, + "flos": 19427637432960.0, + "grad_norm": 2.2831191016216525, + "language_loss": 0.75916541, + "learning_rate": 3.3917304921294674e-06, + "loss": 0.78129709, + "num_input_tokens_seen": 49526980, + "step": 2307, + "time_per_iteration": 2.7846779823303223 + }, + { + "auxiliary_loss_clip": 0.01193737, + "auxiliary_loss_mlp": 0.01033137, + "balance_loss_clip": 1.05610156, + "balance_loss_mlp": 1.02231932, + "epoch": 0.27752059159502196, + "flos": 21614776565760.0, + "grad_norm": 1.776413302070655, + "language_loss": 0.80903429, + "learning_rate": 3.3911709506638876e-06, + "loss": 0.831303, + "num_input_tokens_seen": 49546290, + "step": 2308, + "time_per_iteration": 2.824108839035034 + }, + { + "auxiliary_loss_clip": 0.01164925, + "auxiliary_loss_mlp": 0.00888137, + "balance_loss_clip": 1.04912841, + "balance_loss_mlp": 1.00071573, + "epoch": 0.277640834485661, + "flos": 26608011016320.0, + "grad_norm": 2.141545653450933, + "language_loss": 0.81178385, + "learning_rate": 3.390611198158781e-06, + "loss": 0.83231449, + "num_input_tokens_seen": 49564165, + "step": 2309, + "time_per_iteration": 2.7954938411712646 + }, + { + "auxiliary_loss_clip": 0.01209191, + "auxiliary_loss_mlp": 0.01035783, + "balance_loss_clip": 1.06139147, + "balance_loss_mlp": 1.0250361, + "epoch": 0.2777610773763001, + "flos": 19492814661120.0, + "grad_norm": 2.117903080230022, + "language_loss": 0.90090704, + "learning_rate": 3.3900512346990612e-06, + "loss": 0.92335677, + "num_input_tokens_seen": 49580155, + "step": 2310, + "time_per_iteration": 2.720612049102783 + }, + { + "auxiliary_loss_clip": 0.01161618, + "auxiliary_loss_mlp": 0.01031187, + "balance_loss_clip": 1.04790688, + "balance_loss_mlp": 1.02079153, + "epoch": 0.27788132026693924, + "flos": 38290650001920.0, + "grad_norm": 1.6308034444439747, + "language_loss": 0.65199757, + "learning_rate": 3.389491060369674e-06, + "loss": 0.67392558, + "num_input_tokens_seen": 49605830, + "step": 2311, + "time_per_iteration": 2.976897954940796 + }, + { + "auxiliary_loss_clip": 0.01157066, + "auxiliary_loss_mlp": 0.01032301, + "balance_loss_clip": 1.05014586, + "balance_loss_mlp": 1.02284765, + "epoch": 0.2780015631575783, + "flos": 22382546797440.0, + "grad_norm": 1.8500583709782228, + "language_loss": 0.89666593, + "learning_rate": 3.388930675255598e-06, + "loss": 0.91855961, + "num_input_tokens_seen": 49625680, + "step": 2312, + "time_per_iteration": 2.9140965938568115 + }, + { + "auxiliary_loss_clip": 0.01185024, + "auxiliary_loss_mlp": 0.01040259, + "balance_loss_clip": 1.05247378, + "balance_loss_mlp": 1.03018582, + "epoch": 0.2781218060482174, + "flos": 12203200840320.0, + "grad_norm": 3.5223297353194463, + "language_loss": 0.79241729, + "learning_rate": 3.388370079441843e-06, + "loss": 0.81467015, + "num_input_tokens_seen": 49641195, + "step": 2313, + "time_per_iteration": 2.9199414253234863 + }, + { + "auxiliary_loss_clip": 0.01172935, + "auxiliary_loss_mlp": 0.01036097, + "balance_loss_clip": 1.05512393, + "balance_loss_mlp": 1.02646542, + "epoch": 0.2782420489388565, + "flos": 18107632529280.0, + "grad_norm": 1.8656608006528164, + "language_loss": 0.93048501, + "learning_rate": 3.3878092730134505e-06, + "loss": 0.95257533, + "num_input_tokens_seen": 49659180, + "step": 2314, + "time_per_iteration": 2.806062936782837 + }, + { + "auxiliary_loss_clip": 0.01186813, + "auxiliary_loss_mlp": 0.01036021, + "balance_loss_clip": 1.05507326, + "balance_loss_mlp": 1.025334, + "epoch": 0.27836229182949557, + "flos": 18514752255360.0, + "grad_norm": 1.8447603271000892, + "language_loss": 0.80482918, + "learning_rate": 3.3872482560554947e-06, + "loss": 0.82705748, + "num_input_tokens_seen": 49677955, + "step": 2315, + "time_per_iteration": 2.7664566040039062 + }, + { + "auxiliary_loss_clip": 0.01089121, + "auxiliary_loss_mlp": 0.01005988, + "balance_loss_clip": 1.0200491, + "balance_loss_mlp": 1.00346076, + "epoch": 0.2784825347201347, + "flos": 67079230940160.0, + "grad_norm": 0.7986298194694748, + "language_loss": 0.56948495, + "learning_rate": 3.386687028653082e-06, + "loss": 0.59043604, + "num_input_tokens_seen": 49740800, + "step": 2316, + "time_per_iteration": 3.337078332901001 + }, + { + "auxiliary_loss_clip": 0.01163275, + "auxiliary_loss_mlp": 0.0103941, + "balance_loss_clip": 1.05205023, + "balance_loss_mlp": 1.02902651, + "epoch": 0.2786027776107738, + "flos": 22631119891200.0, + "grad_norm": 1.6926031672504032, + "language_loss": 0.85057914, + "learning_rate": 3.386125590891349e-06, + "loss": 0.87260592, + "num_input_tokens_seen": 49757675, + "step": 2317, + "time_per_iteration": 2.900259494781494 + }, + { + "auxiliary_loss_clip": 0.0117081, + "auxiliary_loss_mlp": 0.01036894, + "balance_loss_clip": 1.05105829, + "balance_loss_mlp": 1.02735162, + "epoch": 0.27872302050141284, + "flos": 15778826156160.0, + "grad_norm": 2.0551125177628555, + "language_loss": 0.82598102, + "learning_rate": 3.3855639428554657e-06, + "loss": 0.8480581, + "num_input_tokens_seen": 49775205, + "step": 2318, + "time_per_iteration": 2.7637436389923096 + }, + { + "auxiliary_loss_clip": 0.01164448, + "auxiliary_loss_mlp": 0.01033219, + "balance_loss_clip": 1.05250049, + "balance_loss_mlp": 1.02367616, + "epoch": 0.27884326339205195, + "flos": 22126970551680.0, + "grad_norm": 1.654744713069215, + "language_loss": 0.80039823, + "learning_rate": 3.385002084630635e-06, + "loss": 0.82237488, + "num_input_tokens_seen": 49794175, + "step": 2319, + "time_per_iteration": 2.8072614669799805 + }, + { + "auxiliary_loss_clip": 0.0119757, + "auxiliary_loss_mlp": 0.01034455, + "balance_loss_clip": 1.05657697, + "balance_loss_mlp": 1.0236423, + "epoch": 0.278963506282691, + "flos": 20558715776640.0, + "grad_norm": 2.923436878725077, + "language_loss": 0.84959179, + "learning_rate": 3.384440016302088e-06, + "loss": 0.87191206, + "num_input_tokens_seen": 49812850, + "step": 2320, + "time_per_iteration": 2.819578170776367 + }, + { + "auxiliary_loss_clip": 0.01190792, + "auxiliary_loss_mlp": 0.01033546, + "balance_loss_clip": 1.05469131, + "balance_loss_mlp": 1.02386653, + "epoch": 0.2790837491733301, + "flos": 21942928241280.0, + "grad_norm": 2.016039593188802, + "language_loss": 0.62358105, + "learning_rate": 3.3838777379550923e-06, + "loss": 0.64582443, + "num_input_tokens_seen": 49832295, + "step": 2321, + "time_per_iteration": 2.7246932983398438 + }, + { + "auxiliary_loss_clip": 0.01181151, + "auxiliary_loss_mlp": 0.01033093, + "balance_loss_clip": 1.05204904, + "balance_loss_mlp": 1.02304947, + "epoch": 0.27920399206396923, + "flos": 26286790665600.0, + "grad_norm": 2.0441646602391916, + "language_loss": 0.78579748, + "learning_rate": 3.383315249674944e-06, + "loss": 0.80793983, + "num_input_tokens_seen": 49850860, + "step": 2322, + "time_per_iteration": 2.8373234272003174 + }, + { + "auxiliary_loss_clip": 0.01174973, + "auxiliary_loss_mlp": 0.01031767, + "balance_loss_clip": 1.05116439, + "balance_loss_mlp": 1.0211693, + "epoch": 0.2793242349546083, + "flos": 25400981364480.0, + "grad_norm": 2.112247366833447, + "language_loss": 0.86150503, + "learning_rate": 3.3827525515469715e-06, + "loss": 0.88357246, + "num_input_tokens_seen": 49865765, + "step": 2323, + "time_per_iteration": 2.8711655139923096 + }, + { + "auxiliary_loss_clip": 0.01163469, + "auxiliary_loss_mlp": 0.01031305, + "balance_loss_clip": 1.04915071, + "balance_loss_mlp": 1.02087975, + "epoch": 0.2794444778452474, + "flos": 20850346298880.0, + "grad_norm": 2.6872103159837972, + "language_loss": 0.7142694, + "learning_rate": 3.3821896436565367e-06, + "loss": 0.73621714, + "num_input_tokens_seen": 49885425, + "step": 2324, + "time_per_iteration": 2.907316207885742 + }, + { + "auxiliary_loss_clip": 0.01195845, + "auxiliary_loss_mlp": 0.01031683, + "balance_loss_clip": 1.05857801, + "balance_loss_mlp": 1.02142537, + "epoch": 0.2795647207358865, + "flos": 21576244250880.0, + "grad_norm": 1.6824447312689366, + "language_loss": 0.70201856, + "learning_rate": 3.381626526089032e-06, + "loss": 0.72429383, + "num_input_tokens_seen": 49904990, + "step": 2325, + "time_per_iteration": 2.6875646114349365 + }, + { + "auxiliary_loss_clip": 0.01173932, + "auxiliary_loss_mlp": 0.01031287, + "balance_loss_clip": 1.04806042, + "balance_loss_mlp": 1.01992023, + "epoch": 0.27968496362652556, + "flos": 21471744608640.0, + "grad_norm": 1.9416342191562357, + "language_loss": 0.78870463, + "learning_rate": 3.3810631989298815e-06, + "loss": 0.8107568, + "num_input_tokens_seen": 49924600, + "step": 2326, + "time_per_iteration": 2.808196544647217 + }, + { + "auxiliary_loss_clip": 0.01165735, + "auxiliary_loss_mlp": 0.01039866, + "balance_loss_clip": 1.05214131, + "balance_loss_mlp": 1.02992392, + "epoch": 0.2798052065171647, + "flos": 23258695340160.0, + "grad_norm": 2.077684101304015, + "language_loss": 0.84169555, + "learning_rate": 3.3804996622645423e-06, + "loss": 0.86375153, + "num_input_tokens_seen": 49942600, + "step": 2327, + "time_per_iteration": 2.871277332305908 + }, + { + "auxiliary_loss_clip": 0.01202095, + "auxiliary_loss_mlp": 0.01031833, + "balance_loss_clip": 1.05679703, + "balance_loss_mlp": 1.02242124, + "epoch": 0.2799254494078038, + "flos": 21539328048000.0, + "grad_norm": 1.8757186074436554, + "language_loss": 0.89620739, + "learning_rate": 3.3799359161785015e-06, + "loss": 0.91854668, + "num_input_tokens_seen": 49962250, + "step": 2328, + "time_per_iteration": 2.854884147644043 + }, + { + "auxiliary_loss_clip": 0.01188496, + "auxiliary_loss_mlp": 0.01029156, + "balance_loss_clip": 1.05489135, + "balance_loss_mlp": 1.0189817, + "epoch": 0.28004569229844284, + "flos": 26393912000640.0, + "grad_norm": 1.9569444203329698, + "language_loss": 0.85639256, + "learning_rate": 3.3793719607572798e-06, + "loss": 0.87856907, + "num_input_tokens_seen": 49983215, + "step": 2329, + "time_per_iteration": 2.7627997398376465 + }, + { + "auxiliary_loss_clip": 0.01166163, + "auxiliary_loss_mlp": 0.01033445, + "balance_loss_clip": 1.04861546, + "balance_loss_mlp": 1.02310979, + "epoch": 0.28016593518908195, + "flos": 33547676584320.0, + "grad_norm": 2.2307205065386477, + "language_loss": 0.77092856, + "learning_rate": 3.378807796086428e-06, + "loss": 0.7929247, + "num_input_tokens_seen": 50006075, + "step": 2330, + "time_per_iteration": 6.492247104644775 + }, + { + "auxiliary_loss_clip": 0.0120182, + "auxiliary_loss_mlp": 0.01035031, + "balance_loss_clip": 1.05800605, + "balance_loss_mlp": 1.02519667, + "epoch": 0.28028617807972106, + "flos": 15340823712000.0, + "grad_norm": 2.1877094424212986, + "language_loss": 0.77209973, + "learning_rate": 3.37824342225153e-06, + "loss": 0.79446828, + "num_input_tokens_seen": 50022495, + "step": 2331, + "time_per_iteration": 2.6412930488586426 + }, + { + "auxiliary_loss_clip": 0.01161246, + "auxiliary_loss_mlp": 0.01034673, + "balance_loss_clip": 1.05333853, + "balance_loss_mlp": 1.02566123, + "epoch": 0.2804064209703601, + "flos": 25520277409920.0, + "grad_norm": 1.7887476404831697, + "language_loss": 0.77845615, + "learning_rate": 3.3776788393382006e-06, + "loss": 0.8004154, + "num_input_tokens_seen": 50041975, + "step": 2332, + "time_per_iteration": 2.886688470840454 + }, + { + "auxiliary_loss_clip": 0.01200957, + "auxiliary_loss_mlp": 0.01029228, + "balance_loss_clip": 1.05639887, + "balance_loss_mlp": 1.01891661, + "epoch": 0.2805266638609992, + "flos": 29351766280320.0, + "grad_norm": 2.4861062771752853, + "language_loss": 0.76296258, + "learning_rate": 3.3771140474320872e-06, + "loss": 0.78526437, + "num_input_tokens_seen": 50061925, + "step": 2333, + "time_per_iteration": 2.7314844131469727 + }, + { + "auxiliary_loss_clip": 0.01178138, + "auxiliary_loss_mlp": 0.01039477, + "balance_loss_clip": 1.05307376, + "balance_loss_mlp": 1.02916479, + "epoch": 0.28064690675163834, + "flos": 21463735875840.0, + "grad_norm": 1.9151959148865383, + "language_loss": 0.79584849, + "learning_rate": 3.3765490466188664e-06, + "loss": 0.81802464, + "num_input_tokens_seen": 50079325, + "step": 2334, + "time_per_iteration": 2.7900500297546387 + }, + { + "auxiliary_loss_clip": 0.0116513, + "auxiliary_loss_mlp": 0.0102996, + "balance_loss_clip": 1.04891682, + "balance_loss_mlp": 1.01974404, + "epoch": 0.2807671496422774, + "flos": 20995640812800.0, + "grad_norm": 2.4029318328572393, + "language_loss": 0.73833156, + "learning_rate": 3.3759838369842508e-06, + "loss": 0.76028252, + "num_input_tokens_seen": 50097400, + "step": 2335, + "time_per_iteration": 2.85864520072937 + }, + { + "auxiliary_loss_clip": 0.01170151, + "auxiliary_loss_mlp": 0.01031513, + "balance_loss_clip": 1.05265415, + "balance_loss_mlp": 1.02167833, + "epoch": 0.2808873925329165, + "flos": 21506577822720.0, + "grad_norm": 1.8947904145791399, + "language_loss": 0.72869003, + "learning_rate": 3.375418418613981e-06, + "loss": 0.75070667, + "num_input_tokens_seen": 50116425, + "step": 2336, + "time_per_iteration": 2.8013017177581787 + }, + { + "auxiliary_loss_clip": 0.01181412, + "auxiliary_loss_mlp": 0.01032097, + "balance_loss_clip": 1.0535779, + "balance_loss_mlp": 1.022084, + "epoch": 0.28100763542355556, + "flos": 16070815814400.0, + "grad_norm": 2.5667750043907285, + "language_loss": 0.83537829, + "learning_rate": 3.374852791593831e-06, + "loss": 0.85751337, + "num_input_tokens_seen": 50132625, + "step": 2337, + "time_per_iteration": 2.8416128158569336 + }, + { + "auxiliary_loss_clip": 0.01168215, + "auxiliary_loss_mlp": 0.01035179, + "balance_loss_clip": 1.04901087, + "balance_loss_mlp": 1.02475441, + "epoch": 0.28112787831419467, + "flos": 19062605468160.0, + "grad_norm": 3.9181156923491285, + "language_loss": 0.53976196, + "learning_rate": 3.374286956009605e-06, + "loss": 0.56179589, + "num_input_tokens_seen": 50151190, + "step": 2338, + "time_per_iteration": 2.9114954471588135 + }, + { + "auxiliary_loss_clip": 0.0119164, + "auxiliary_loss_mlp": 0.01026719, + "balance_loss_clip": 1.05930662, + "balance_loss_mlp": 1.01723647, + "epoch": 0.2812481212048338, + "flos": 12823629482880.0, + "grad_norm": 2.365194897983439, + "language_loss": 0.7519002, + "learning_rate": 3.3737209119471405e-06, + "loss": 0.77408373, + "num_input_tokens_seen": 50167700, + "step": 2339, + "time_per_iteration": 2.895132541656494 + }, + { + "auxiliary_loss_clip": 0.0120033, + "auxiliary_loss_mlp": 0.01035362, + "balance_loss_clip": 1.05907393, + "balance_loss_mlp": 1.02440083, + "epoch": 0.28136836409547283, + "flos": 15633064765440.0, + "grad_norm": 3.161451750979912, + "language_loss": 0.63753211, + "learning_rate": 3.373154659492306e-06, + "loss": 0.65988898, + "num_input_tokens_seen": 50185840, + "step": 2340, + "time_per_iteration": 2.719524383544922 + }, + { + "auxiliary_loss_clip": 0.0118373, + "auxiliary_loss_mlp": 0.01032172, + "balance_loss_clip": 1.0549314, + "balance_loss_mlp": 1.02256978, + "epoch": 0.28148860698611194, + "flos": 19933726106880.0, + "grad_norm": 1.7769457945889588, + "language_loss": 0.85351545, + "learning_rate": 3.3725881987310016e-06, + "loss": 0.87567443, + "num_input_tokens_seen": 50203375, + "step": 2341, + "time_per_iteration": 2.753028154373169 + }, + { + "auxiliary_loss_clip": 0.01178814, + "auxiliary_loss_mlp": 0.01028172, + "balance_loss_clip": 1.0523138, + "balance_loss_mlp": 1.01842666, + "epoch": 0.28160884987675106, + "flos": 17457219008640.0, + "grad_norm": 1.8227620046362467, + "language_loss": 0.87369078, + "learning_rate": 3.372021529749159e-06, + "loss": 0.89576066, + "num_input_tokens_seen": 50222435, + "step": 2342, + "time_per_iteration": 2.838634967803955 + }, + { + "auxiliary_loss_clip": 0.0115329, + "auxiliary_loss_mlp": 0.01031425, + "balance_loss_clip": 1.05184007, + "balance_loss_mlp": 1.0218761, + "epoch": 0.2817290927673901, + "flos": 16834743290880.0, + "grad_norm": 2.041452073535978, + "language_loss": 0.91948235, + "learning_rate": 3.3714546526327405e-06, + "loss": 0.94132948, + "num_input_tokens_seen": 50240435, + "step": 2343, + "time_per_iteration": 2.9826271533966064 + }, + { + "auxiliary_loss_clip": 0.01177696, + "auxiliary_loss_mlp": 0.01034013, + "balance_loss_clip": 1.05176818, + "balance_loss_mlp": 1.0241425, + "epoch": 0.2818493356580292, + "flos": 15414081500160.0, + "grad_norm": 2.2416633281319633, + "language_loss": 0.88288456, + "learning_rate": 3.3708875674677423e-06, + "loss": 0.90500164, + "num_input_tokens_seen": 50258410, + "step": 2344, + "time_per_iteration": 2.7811262607574463 + }, + { + "auxiliary_loss_clip": 0.01189198, + "auxiliary_loss_mlp": 0.01037591, + "balance_loss_clip": 1.05687344, + "balance_loss_mlp": 1.02780366, + "epoch": 0.28196957854866833, + "flos": 20412451595520.0, + "grad_norm": 2.024261609822484, + "language_loss": 0.83631152, + "learning_rate": 3.37032027434019e-06, + "loss": 0.85857946, + "num_input_tokens_seen": 50277930, + "step": 2345, + "time_per_iteration": 2.7244198322296143 + }, + { + "auxiliary_loss_clip": 0.01200303, + "auxiliary_loss_mlp": 0.01033961, + "balance_loss_clip": 1.05553222, + "balance_loss_mlp": 1.02308285, + "epoch": 0.2820898214393074, + "flos": 19973120348160.0, + "grad_norm": 2.454392147255283, + "language_loss": 0.83245534, + "learning_rate": 3.369752773336141e-06, + "loss": 0.85479796, + "num_input_tokens_seen": 50297410, + "step": 2346, + "time_per_iteration": 2.7479398250579834 + }, + { + "auxiliary_loss_clip": 0.01183041, + "auxiliary_loss_mlp": 0.01035892, + "balance_loss_clip": 1.05347061, + "balance_loss_mlp": 1.02534175, + "epoch": 0.2822100643299465, + "flos": 22528308188160.0, + "grad_norm": 1.6540466534005873, + "language_loss": 0.78366983, + "learning_rate": 3.3691850645416864e-06, + "loss": 0.80585921, + "num_input_tokens_seen": 50317120, + "step": 2347, + "time_per_iteration": 2.802212953567505 + }, + { + "auxiliary_loss_clip": 0.01194921, + "auxiliary_loss_mlp": 0.01033339, + "balance_loss_clip": 1.05486238, + "balance_loss_mlp": 1.02343869, + "epoch": 0.2823303072205856, + "flos": 11546682007680.0, + "grad_norm": 2.4650008162449417, + "language_loss": 0.83084631, + "learning_rate": 3.368617148042945e-06, + "loss": 0.85312891, + "num_input_tokens_seen": 50334790, + "step": 2348, + "time_per_iteration": 2.755768299102783 + }, + { + "auxiliary_loss_clip": 0.01171607, + "auxiliary_loss_mlp": 0.01035956, + "balance_loss_clip": 1.04748249, + "balance_loss_mlp": 1.02596056, + "epoch": 0.28245055011122466, + "flos": 18259894281600.0, + "grad_norm": 1.8423500189286137, + "language_loss": 0.84487629, + "learning_rate": 3.368049023926071e-06, + "loss": 0.86695194, + "num_input_tokens_seen": 50353785, + "step": 2349, + "time_per_iteration": 2.782592535018921 + }, + { + "auxiliary_loss_clip": 0.01193449, + "auxiliary_loss_mlp": 0.01034119, + "balance_loss_clip": 1.05982506, + "balance_loss_mlp": 1.0248028, + "epoch": 0.2825707930018638, + "flos": 24608110504320.0, + "grad_norm": 1.5254533285225278, + "language_loss": 0.83652055, + "learning_rate": 3.3674806922772476e-06, + "loss": 0.85879624, + "num_input_tokens_seen": 50374670, + "step": 2350, + "time_per_iteration": 2.7107274532318115 + }, + { + "auxiliary_loss_clip": 0.01174842, + "auxiliary_loss_mlp": 0.01030639, + "balance_loss_clip": 1.04990387, + "balance_loss_mlp": 1.02083385, + "epoch": 0.28269103589250283, + "flos": 25226994862080.0, + "grad_norm": 1.9407340030697215, + "language_loss": 0.74645889, + "learning_rate": 3.3669121531826904e-06, + "loss": 0.76851368, + "num_input_tokens_seen": 50395650, + "step": 2351, + "time_per_iteration": 2.8930182456970215 + }, + { + "auxiliary_loss_clip": 0.01167176, + "auxiliary_loss_mlp": 0.01030419, + "balance_loss_clip": 1.05675483, + "balance_loss_mlp": 1.02033961, + "epoch": 0.28281127878314194, + "flos": 19281552819840.0, + "grad_norm": 1.7960832876183708, + "language_loss": 0.82930005, + "learning_rate": 3.366343406728647e-06, + "loss": 0.85127604, + "num_input_tokens_seen": 50415100, + "step": 2352, + "time_per_iteration": 2.779897689819336 + }, + { + "auxiliary_loss_clip": 0.01184972, + "auxiliary_loss_mlp": 0.01032779, + "balance_loss_clip": 1.05275202, + "balance_loss_mlp": 1.02289665, + "epoch": 0.28293152167378105, + "flos": 23878405710720.0, + "grad_norm": 2.049666497860184, + "language_loss": 0.68533438, + "learning_rate": 3.3657744530013946e-06, + "loss": 0.7075119, + "num_input_tokens_seen": 50434335, + "step": 2353, + "time_per_iteration": 2.7854995727539062 + }, + { + "auxiliary_loss_clip": 0.01195869, + "auxiliary_loss_mlp": 0.0103927, + "balance_loss_clip": 1.05639791, + "balance_loss_mlp": 1.02951837, + "epoch": 0.2830517645644201, + "flos": 43866965928960.0, + "grad_norm": 1.9072156537091485, + "language_loss": 0.71108103, + "learning_rate": 3.3652052920872437e-06, + "loss": 0.73343241, + "num_input_tokens_seen": 50457200, + "step": 2354, + "time_per_iteration": 2.878664970397949 + }, + { + "auxiliary_loss_clip": 0.01183339, + "auxiliary_loss_mlp": 0.01040528, + "balance_loss_clip": 1.05387688, + "balance_loss_mlp": 1.02959073, + "epoch": 0.2831720074550592, + "flos": 26651750803200.0, + "grad_norm": 2.2634959734430984, + "language_loss": 0.85391402, + "learning_rate": 3.3646359240725355e-06, + "loss": 0.87615269, + "num_input_tokens_seen": 50476390, + "step": 2355, + "time_per_iteration": 3.723482370376587 + }, + { + "auxiliary_loss_clip": 0.01183904, + "auxiliary_loss_mlp": 0.00888227, + "balance_loss_clip": 1.05241311, + "balance_loss_mlp": 1.00068545, + "epoch": 0.2832922503456983, + "flos": 31029979564800.0, + "grad_norm": 1.9696995514758786, + "language_loss": 0.67929506, + "learning_rate": 3.364066349043643e-06, + "loss": 0.70001638, + "num_input_tokens_seen": 50497595, + "step": 2356, + "time_per_iteration": 4.713066101074219 + }, + { + "auxiliary_loss_clip": 0.01176018, + "auxiliary_loss_mlp": 0.01030508, + "balance_loss_clip": 1.05153561, + "balance_loss_mlp": 1.0209775, + "epoch": 0.2834124932363374, + "flos": 20405699838720.0, + "grad_norm": 2.139265709714287, + "language_loss": 0.82809877, + "learning_rate": 3.363496567086969e-06, + "loss": 0.85016406, + "num_input_tokens_seen": 50514690, + "step": 2357, + "time_per_iteration": 3.8053956031799316 + }, + { + "auxiliary_loss_clip": 0.01201417, + "auxiliary_loss_mlp": 0.01033738, + "balance_loss_clip": 1.05788493, + "balance_loss_mlp": 1.0241003, + "epoch": 0.2835327361269765, + "flos": 39384848056320.0, + "grad_norm": 1.8007833939745825, + "language_loss": 0.7541945, + "learning_rate": 3.3629265782889506e-06, + "loss": 0.77654612, + "num_input_tokens_seen": 50536515, + "step": 2358, + "time_per_iteration": 2.8377089500427246 + }, + { + "auxiliary_loss_clip": 0.01164634, + "auxiliary_loss_mlp": 0.01041364, + "balance_loss_clip": 1.04815841, + "balance_loss_mlp": 1.03117752, + "epoch": 0.2836529790176156, + "flos": 30261598801920.0, + "grad_norm": 1.832632081543022, + "language_loss": 0.71736729, + "learning_rate": 3.362356382736054e-06, + "loss": 0.73942733, + "num_input_tokens_seen": 50557120, + "step": 2359, + "time_per_iteration": 2.9969723224639893 + }, + { + "auxiliary_loss_clip": 0.01169033, + "auxiliary_loss_mlp": 0.01034792, + "balance_loss_clip": 1.04732418, + "balance_loss_mlp": 1.02495742, + "epoch": 0.28377322190825466, + "flos": 12677796264960.0, + "grad_norm": 2.219537036981725, + "language_loss": 0.90929282, + "learning_rate": 3.361785980514777e-06, + "loss": 0.93133104, + "num_input_tokens_seen": 50573320, + "step": 2360, + "time_per_iteration": 2.858271598815918 + }, + { + "auxiliary_loss_clip": 0.01146095, + "auxiliary_loss_mlp": 0.01038293, + "balance_loss_clip": 1.04929638, + "balance_loss_mlp": 1.02835739, + "epoch": 0.28389346479889377, + "flos": 18296666830080.0, + "grad_norm": 2.0385434596560397, + "language_loss": 0.76530707, + "learning_rate": 3.361215371711649e-06, + "loss": 0.78715092, + "num_input_tokens_seen": 50592415, + "step": 2361, + "time_per_iteration": 2.8702187538146973 + }, + { + "auxiliary_loss_clip": 0.01161595, + "auxiliary_loss_mlp": 0.01029703, + "balance_loss_clip": 1.04959607, + "balance_loss_mlp": 1.01992202, + "epoch": 0.2840137076895329, + "flos": 20406992728320.0, + "grad_norm": 1.8329069804620153, + "language_loss": 0.83745563, + "learning_rate": 3.3606445564132326e-06, + "loss": 0.85936856, + "num_input_tokens_seen": 50609710, + "step": 2362, + "time_per_iteration": 2.8153059482574463 + }, + { + "auxiliary_loss_clip": 0.01203468, + "auxiliary_loss_mlp": 0.00886702, + "balance_loss_clip": 1.05936384, + "balance_loss_mlp": 1.00072551, + "epoch": 0.28413395058017193, + "flos": 20048030161920.0, + "grad_norm": 1.9989192400407958, + "language_loss": 0.82540846, + "learning_rate": 3.360073534706118e-06, + "loss": 0.84631014, + "num_input_tokens_seen": 50626865, + "step": 2363, + "time_per_iteration": 2.6749610900878906 + }, + { + "auxiliary_loss_clip": 0.01183647, + "auxiliary_loss_mlp": 0.01036052, + "balance_loss_clip": 1.05575562, + "balance_loss_mlp": 1.02586007, + "epoch": 0.28425419347081105, + "flos": 37663613256960.0, + "grad_norm": 2.044917945064327, + "language_loss": 0.76066452, + "learning_rate": 3.35950230667693e-06, + "loss": 0.78286147, + "num_input_tokens_seen": 50648560, + "step": 2364, + "time_per_iteration": 2.9758331775665283 + }, + { + "auxiliary_loss_clip": 0.01192171, + "auxiliary_loss_mlp": 0.01040488, + "balance_loss_clip": 1.05498123, + "balance_loss_mlp": 1.03075433, + "epoch": 0.28437443636145016, + "flos": 13845072539520.0, + "grad_norm": 2.1271620211184445, + "language_loss": 0.86294764, + "learning_rate": 3.358930872412323e-06, + "loss": 0.88527423, + "num_input_tokens_seen": 50665725, + "step": 2365, + "time_per_iteration": 2.6757149696350098 + }, + { + "auxiliary_loss_clip": 0.01186561, + "auxiliary_loss_mlp": 0.01027376, + "balance_loss_clip": 1.05483818, + "balance_loss_mlp": 1.01810813, + "epoch": 0.2844946792520892, + "flos": 22747794243840.0, + "grad_norm": 1.5354200194160237, + "language_loss": 0.80899787, + "learning_rate": 3.3583592319989825e-06, + "loss": 0.8311373, + "num_input_tokens_seen": 50685095, + "step": 2366, + "time_per_iteration": 2.712099552154541 + }, + { + "auxiliary_loss_clip": 0.01198301, + "auxiliary_loss_mlp": 0.01039969, + "balance_loss_clip": 1.0566864, + "balance_loss_mlp": 1.02922249, + "epoch": 0.2846149221427283, + "flos": 32415987709440.0, + "grad_norm": 2.1465722919966272, + "language_loss": 0.68763256, + "learning_rate": 3.357787385523627e-06, + "loss": 0.7100153, + "num_input_tokens_seen": 50706500, + "step": 2367, + "time_per_iteration": 2.7870259284973145 + }, + { + "auxiliary_loss_clip": 0.01162515, + "auxiliary_loss_mlp": 0.01038284, + "balance_loss_clip": 1.05115867, + "balance_loss_mlp": 1.02800798, + "epoch": 0.2847351650333674, + "flos": 28475976873600.0, + "grad_norm": 1.8017467899857844, + "language_loss": 0.82972658, + "learning_rate": 3.3572153330730048e-06, + "loss": 0.85173458, + "num_input_tokens_seen": 50727595, + "step": 2368, + "time_per_iteration": 3.054910659790039 + }, + { + "auxiliary_loss_clip": 0.0107462, + "auxiliary_loss_mlp": 0.01004703, + "balance_loss_clip": 1.01910269, + "balance_loss_mlp": 1.00205636, + "epoch": 0.2848554079240065, + "flos": 55753399704960.0, + "grad_norm": 0.828640913697246, + "language_loss": 0.64736998, + "learning_rate": 3.3566430747338956e-06, + "loss": 0.66816318, + "num_input_tokens_seen": 50782800, + "step": 2369, + "time_per_iteration": 3.2412943840026855 + }, + { + "auxiliary_loss_clip": 0.01192799, + "auxiliary_loss_mlp": 0.01037825, + "balance_loss_clip": 1.05534244, + "balance_loss_mlp": 1.02781177, + "epoch": 0.2849756508146456, + "flos": 11836875985920.0, + "grad_norm": 2.275161025117664, + "language_loss": 0.86896724, + "learning_rate": 3.35607061059311e-06, + "loss": 0.8912735, + "num_input_tokens_seen": 50797730, + "step": 2370, + "time_per_iteration": 2.755455255508423 + }, + { + "auxiliary_loss_clip": 0.01200002, + "auxiliary_loss_mlp": 0.01040047, + "balance_loss_clip": 1.05844438, + "balance_loss_mlp": 1.02978921, + "epoch": 0.28509589370528465, + "flos": 25155209531520.0, + "grad_norm": 1.7397090139326332, + "language_loss": 0.74734312, + "learning_rate": 3.3554979407374917e-06, + "loss": 0.76974368, + "num_input_tokens_seen": 50819840, + "step": 2371, + "time_per_iteration": 2.785982847213745 + }, + { + "auxiliary_loss_clip": 0.01187922, + "auxiliary_loss_mlp": 0.01031399, + "balance_loss_clip": 1.05154955, + "balance_loss_mlp": 1.02206445, + "epoch": 0.28521613659592376, + "flos": 19974808287360.0, + "grad_norm": 1.4654977505020517, + "language_loss": 0.73661327, + "learning_rate": 3.3549250652539134e-06, + "loss": 0.75880647, + "num_input_tokens_seen": 50838935, + "step": 2372, + "time_per_iteration": 2.745028018951416 + }, + { + "auxiliary_loss_clip": 0.01178696, + "auxiliary_loss_mlp": 0.01038763, + "balance_loss_clip": 1.04945743, + "balance_loss_mlp": 1.02878535, + "epoch": 0.2853363794865629, + "flos": 23367971491200.0, + "grad_norm": 1.8265522040719708, + "language_loss": 0.81748056, + "learning_rate": 3.3543519842292794e-06, + "loss": 0.8396551, + "num_input_tokens_seen": 50858590, + "step": 2373, + "time_per_iteration": 2.866617441177368 + }, + { + "auxiliary_loss_clip": 0.01201632, + "auxiliary_loss_mlp": 0.00887723, + "balance_loss_clip": 1.05737305, + "balance_loss_mlp": 1.00069284, + "epoch": 0.28545662237720193, + "flos": 19861940776320.0, + "grad_norm": 1.6646683432625884, + "language_loss": 0.83709252, + "learning_rate": 3.353778697750527e-06, + "loss": 0.85798609, + "num_input_tokens_seen": 50876995, + "step": 2374, + "time_per_iteration": 2.6391899585723877 + }, + { + "auxiliary_loss_clip": 0.01172661, + "auxiliary_loss_mlp": 0.01036586, + "balance_loss_clip": 1.05114031, + "balance_loss_mlp": 1.02703166, + "epoch": 0.28557686526784104, + "flos": 23879016241920.0, + "grad_norm": 1.5099202279768265, + "language_loss": 0.89335161, + "learning_rate": 3.353205205904622e-06, + "loss": 0.91544408, + "num_input_tokens_seen": 50896105, + "step": 2375, + "time_per_iteration": 2.7480828762054443 + }, + { + "auxiliary_loss_clip": 0.01179565, + "auxiliary_loss_mlp": 0.01028898, + "balance_loss_clip": 1.05398464, + "balance_loss_mlp": 1.01871181, + "epoch": 0.28569710815848015, + "flos": 44890384233600.0, + "grad_norm": 1.7105396781202842, + "language_loss": 0.71602547, + "learning_rate": 3.3526315087785637e-06, + "loss": 0.73811007, + "num_input_tokens_seen": 50917220, + "step": 2376, + "time_per_iteration": 2.9549076557159424 + }, + { + "auxiliary_loss_clip": 0.01141402, + "auxiliary_loss_mlp": 0.01034468, + "balance_loss_clip": 1.04722571, + "balance_loss_mlp": 1.02489591, + "epoch": 0.2858173510491192, + "flos": 26829759628800.0, + "grad_norm": 1.764735257997451, + "language_loss": 0.80988866, + "learning_rate": 3.3520576064593805e-06, + "loss": 0.83164734, + "num_input_tokens_seen": 50937175, + "step": 2377, + "time_per_iteration": 2.82082200050354 + }, + { + "auxiliary_loss_clip": 0.01196897, + "auxiliary_loss_mlp": 0.01028237, + "balance_loss_clip": 1.0579505, + "balance_loss_mlp": 1.01893318, + "epoch": 0.2859375939397583, + "flos": 23148916398720.0, + "grad_norm": 1.6820208790597164, + "language_loss": 0.8186956, + "learning_rate": 3.3514834990341337e-06, + "loss": 0.84094697, + "num_input_tokens_seen": 50957500, + "step": 2378, + "time_per_iteration": 2.6913914680480957 + }, + { + "auxiliary_loss_clip": 0.01191456, + "auxiliary_loss_mlp": 0.01036888, + "balance_loss_clip": 1.05842519, + "balance_loss_mlp": 1.02731538, + "epoch": 0.2860578368303974, + "flos": 12129799397760.0, + "grad_norm": 2.929222563802783, + "language_loss": 0.92669851, + "learning_rate": 3.3509091865899144e-06, + "loss": 0.94898188, + "num_input_tokens_seen": 50972690, + "step": 2379, + "time_per_iteration": 2.686417579650879 + }, + { + "auxiliary_loss_clip": 0.01203404, + "auxiliary_loss_mlp": 0.01032132, + "balance_loss_clip": 1.05826986, + "balance_loss_mlp": 1.0225234, + "epoch": 0.2861780797210365, + "flos": 19938035738880.0, + "grad_norm": 1.8447008608756874, + "language_loss": 0.70489395, + "learning_rate": 3.350334669213846e-06, + "loss": 0.72724938, + "num_input_tokens_seen": 50990095, + "step": 2380, + "time_per_iteration": 2.661118984222412 + }, + { + "auxiliary_loss_clip": 0.01192593, + "auxiliary_loss_mlp": 0.01036583, + "balance_loss_clip": 1.05894971, + "balance_loss_mlp": 1.02722502, + "epoch": 0.2862983226116756, + "flos": 27563127609600.0, + "grad_norm": 1.7792600479543668, + "language_loss": 0.75707513, + "learning_rate": 3.3497599469930816e-06, + "loss": 0.77936691, + "num_input_tokens_seen": 51008305, + "step": 2381, + "time_per_iteration": 3.6507680416107178 + }, + { + "auxiliary_loss_clip": 0.01204425, + "auxiliary_loss_mlp": 0.01037556, + "balance_loss_clip": 1.05770397, + "balance_loss_mlp": 1.02756667, + "epoch": 0.28641856550231465, + "flos": 22053964158720.0, + "grad_norm": 2.212714454869537, + "language_loss": 0.83029938, + "learning_rate": 3.349185020014807e-06, + "loss": 0.85271919, + "num_input_tokens_seen": 51025570, + "step": 2382, + "time_per_iteration": 2.746058940887451 + }, + { + "auxiliary_loss_clip": 0.01193914, + "auxiliary_loss_mlp": 0.01040061, + "balance_loss_clip": 1.0552541, + "balance_loss_mlp": 1.02984428, + "epoch": 0.28653880839295376, + "flos": 22378775869440.0, + "grad_norm": 3.313074110734814, + "language_loss": 0.74514186, + "learning_rate": 3.348609888366237e-06, + "loss": 0.76748157, + "num_input_tokens_seen": 51044585, + "step": 2383, + "time_per_iteration": 4.581822156906128 + }, + { + "auxiliary_loss_clip": 0.01148098, + "auxiliary_loss_mlp": 0.01028671, + "balance_loss_clip": 1.04756403, + "balance_loss_mlp": 1.01912284, + "epoch": 0.28665905128359287, + "flos": 23367971491200.0, + "grad_norm": 2.1659335966488773, + "language_loss": 0.63022196, + "learning_rate": 3.348034552134619e-06, + "loss": 0.6519897, + "num_input_tokens_seen": 51063990, + "step": 2384, + "time_per_iteration": 2.9134602546691895 + }, + { + "auxiliary_loss_clip": 0.01158182, + "auxiliary_loss_mlp": 0.01030733, + "balance_loss_clip": 1.0525893, + "balance_loss_mlp": 1.02091014, + "epoch": 0.2867792941742319, + "flos": 20881695893760.0, + "grad_norm": 2.6124861254628327, + "language_loss": 0.8460592, + "learning_rate": 3.3474590114072316e-06, + "loss": 0.86794835, + "num_input_tokens_seen": 51081990, + "step": 2385, + "time_per_iteration": 2.808224678039551 + }, + { + "auxiliary_loss_clip": 0.01168607, + "auxiliary_loss_mlp": 0.01038413, + "balance_loss_clip": 1.05413508, + "balance_loss_mlp": 1.02772593, + "epoch": 0.28689953706487104, + "flos": 20664005518080.0, + "grad_norm": 1.7379578367930535, + "language_loss": 0.83120525, + "learning_rate": 3.3468832662713836e-06, + "loss": 0.85327542, + "num_input_tokens_seen": 51100235, + "step": 2386, + "time_per_iteration": 2.893521547317505 + }, + { + "auxiliary_loss_clip": 0.01172212, + "auxiliary_loss_mlp": 0.0103514, + "balance_loss_clip": 1.05510473, + "balance_loss_mlp": 1.02510881, + "epoch": 0.28701977995551015, + "flos": 12675533708160.0, + "grad_norm": 2.41792490043455, + "language_loss": 0.84123218, + "learning_rate": 3.346307316814415e-06, + "loss": 0.86330569, + "num_input_tokens_seen": 51115405, + "step": 2387, + "time_per_iteration": 2.723097801208496 + }, + { + "auxiliary_loss_clip": 0.01195461, + "auxiliary_loss_mlp": 0.01038251, + "balance_loss_clip": 1.05901706, + "balance_loss_mlp": 1.02761745, + "epoch": 0.2871400228461492, + "flos": 21252366293760.0, + "grad_norm": 3.49341224210569, + "language_loss": 0.75950623, + "learning_rate": 3.3457311631236965e-06, + "loss": 0.7818433, + "num_input_tokens_seen": 51136390, + "step": 2388, + "time_per_iteration": 2.7854998111724854 + }, + { + "auxiliary_loss_clip": 0.01171387, + "auxiliary_loss_mlp": 0.01032904, + "balance_loss_clip": 1.05269647, + "balance_loss_mlp": 1.02269411, + "epoch": 0.2872602657367883, + "flos": 25119262995840.0, + "grad_norm": 2.0093310495599175, + "language_loss": 0.84090304, + "learning_rate": 3.345154805286631e-06, + "loss": 0.86294591, + "num_input_tokens_seen": 51156650, + "step": 2389, + "time_per_iteration": 2.8003153800964355 + }, + { + "auxiliary_loss_clip": 0.01184441, + "auxiliary_loss_mlp": 0.01041278, + "balance_loss_clip": 1.05246782, + "balance_loss_mlp": 1.03094852, + "epoch": 0.2873805086274274, + "flos": 16646606830080.0, + "grad_norm": 2.5666978722924356, + "language_loss": 0.76219201, + "learning_rate": 3.344578243390651e-06, + "loss": 0.78444922, + "num_input_tokens_seen": 51172210, + "step": 2390, + "time_per_iteration": 2.674891233444214 + }, + { + "auxiliary_loss_clip": 0.01179696, + "auxiliary_loss_mlp": 0.01030891, + "balance_loss_clip": 1.05489719, + "balance_loss_mlp": 1.02174735, + "epoch": 0.2875007515180665, + "flos": 17420123237760.0, + "grad_norm": 2.54633734607309, + "language_loss": 0.78813446, + "learning_rate": 3.3440014775232206e-06, + "loss": 0.81024039, + "num_input_tokens_seen": 51190265, + "step": 2391, + "time_per_iteration": 2.7203309535980225 + }, + { + "auxiliary_loss_clip": 0.01172242, + "auxiliary_loss_mlp": 0.01028529, + "balance_loss_clip": 1.05189753, + "balance_loss_mlp": 1.01865804, + "epoch": 0.2876209944087056, + "flos": 23434190213760.0, + "grad_norm": 2.013179589621678, + "language_loss": 0.71223104, + "learning_rate": 3.343424507771834e-06, + "loss": 0.73423874, + "num_input_tokens_seen": 51208475, + "step": 2392, + "time_per_iteration": 2.8355820178985596 + }, + { + "auxiliary_loss_clip": 0.01176569, + "auxiliary_loss_mlp": 0.01038656, + "balance_loss_clip": 1.05658364, + "balance_loss_mlp": 1.02867854, + "epoch": 0.2877412372993447, + "flos": 13735509079680.0, + "grad_norm": 2.876031319920336, + "language_loss": 0.8660332, + "learning_rate": 3.342847334224018e-06, + "loss": 0.88818544, + "num_input_tokens_seen": 51225875, + "step": 2393, + "time_per_iteration": 2.8816473484039307 + }, + { + "auxiliary_loss_clip": 0.01092539, + "auxiliary_loss_mlp": 0.01005946, + "balance_loss_clip": 1.02356958, + "balance_loss_mlp": 1.00326383, + "epoch": 0.28786148018998375, + "flos": 58079695104000.0, + "grad_norm": 0.9520870312267882, + "language_loss": 0.62434053, + "learning_rate": 3.342269956967329e-06, + "loss": 0.64532542, + "num_input_tokens_seen": 51287780, + "step": 2394, + "time_per_iteration": 3.384432077407837 + }, + { + "auxiliary_loss_clip": 0.01200306, + "auxiliary_loss_mlp": 0.01038576, + "balance_loss_clip": 1.05853879, + "balance_loss_mlp": 1.02691126, + "epoch": 0.28798172308062286, + "flos": 23435052140160.0, + "grad_norm": 2.886705402911799, + "language_loss": 0.71843433, + "learning_rate": 3.341692376089355e-06, + "loss": 0.74082315, + "num_input_tokens_seen": 51303335, + "step": 2395, + "time_per_iteration": 2.76397705078125 + }, + { + "auxiliary_loss_clip": 0.0118768, + "auxiliary_loss_mlp": 0.0102863, + "balance_loss_clip": 1.05478096, + "balance_loss_mlp": 1.01896834, + "epoch": 0.288101965971262, + "flos": 25110033200640.0, + "grad_norm": 3.1244315745908167, + "language_loss": 0.83652127, + "learning_rate": 3.3411145916777146e-06, + "loss": 0.8586843, + "num_input_tokens_seen": 51317495, + "step": 2396, + "time_per_iteration": 2.645157814025879 + }, + { + "auxiliary_loss_clip": 0.01176033, + "auxiliary_loss_mlp": 0.0103622, + "balance_loss_clip": 1.05182934, + "balance_loss_mlp": 1.02519906, + "epoch": 0.28822220886190103, + "flos": 16252559654400.0, + "grad_norm": 2.144938990190706, + "language_loss": 0.90653133, + "learning_rate": 3.3405366038200566e-06, + "loss": 0.92865378, + "num_input_tokens_seen": 51336430, + "step": 2397, + "time_per_iteration": 2.7460944652557373 + }, + { + "auxiliary_loss_clip": 0.01189332, + "auxiliary_loss_mlp": 0.01041221, + "balance_loss_clip": 1.05941141, + "balance_loss_mlp": 1.03102827, + "epoch": 0.28834245175254014, + "flos": 24535642815360.0, + "grad_norm": 2.0694034521637157, + "language_loss": 0.84951156, + "learning_rate": 3.3399584126040617e-06, + "loss": 0.87181711, + "num_input_tokens_seen": 51355930, + "step": 2398, + "time_per_iteration": 2.7187130451202393 + }, + { + "auxiliary_loss_clip": 0.01207132, + "auxiliary_loss_mlp": 0.00888109, + "balance_loss_clip": 1.06042838, + "balance_loss_mlp": 1.00063276, + "epoch": 0.2884626946431792, + "flos": 24571445696640.0, + "grad_norm": 1.8808757226572466, + "language_loss": 0.90682858, + "learning_rate": 3.339380018117441e-06, + "loss": 0.92778099, + "num_input_tokens_seen": 51376765, + "step": 2399, + "time_per_iteration": 2.760491371154785 + }, + { + "auxiliary_loss_clip": 0.01188496, + "auxiliary_loss_mlp": 0.01032769, + "balance_loss_clip": 1.05736411, + "balance_loss_mlp": 1.02286255, + "epoch": 0.2885829375338183, + "flos": 16544657053440.0, + "grad_norm": 2.2723633908521523, + "language_loss": 0.7797873, + "learning_rate": 3.3388014204479366e-06, + "loss": 0.80199993, + "num_input_tokens_seen": 51394570, + "step": 2400, + "time_per_iteration": 2.737736225128174 + }, + { + "auxiliary_loss_clip": 0.01207332, + "auxiliary_loss_mlp": 0.01034127, + "balance_loss_clip": 1.05875707, + "balance_loss_mlp": 1.02349901, + "epoch": 0.2887031804244574, + "flos": 24061226958720.0, + "grad_norm": 2.227032473989409, + "language_loss": 0.91331613, + "learning_rate": 3.338222619683321e-06, + "loss": 0.9357307, + "num_input_tokens_seen": 51414535, + "step": 2401, + "time_per_iteration": 2.721278429031372 + }, + { + "auxiliary_loss_clip": 0.01188301, + "auxiliary_loss_mlp": 0.01035395, + "balance_loss_clip": 1.05709839, + "balance_loss_mlp": 1.02483261, + "epoch": 0.2888234233150965, + "flos": 23330696152320.0, + "grad_norm": 1.9826276028294245, + "language_loss": 0.73103088, + "learning_rate": 3.337643615911398e-06, + "loss": 0.75326777, + "num_input_tokens_seen": 51434160, + "step": 2402, + "time_per_iteration": 2.8227803707122803 + }, + { + "auxiliary_loss_clip": 0.01199446, + "auxiliary_loss_mlp": 0.01038692, + "balance_loss_clip": 1.05897307, + "balance_loss_mlp": 1.02840424, + "epoch": 0.2889436662057356, + "flos": 22272767856000.0, + "grad_norm": 1.9136246117771158, + "language_loss": 0.78737134, + "learning_rate": 3.3370644092200026e-06, + "loss": 0.8097527, + "num_input_tokens_seen": 51451435, + "step": 2403, + "time_per_iteration": 2.725635290145874 + }, + { + "auxiliary_loss_clip": 0.0116158, + "auxiliary_loss_mlp": 0.01041321, + "balance_loss_clip": 1.05057836, + "balance_loss_mlp": 1.03072333, + "epoch": 0.2890639090963747, + "flos": 21616931381760.0, + "grad_norm": 2.4618107704661423, + "language_loss": 0.78391445, + "learning_rate": 3.3364849996969985e-06, + "loss": 0.80594349, + "num_input_tokens_seen": 51471455, + "step": 2404, + "time_per_iteration": 2.891219139099121 + }, + { + "auxiliary_loss_clip": 0.01191413, + "auxiliary_loss_mlp": 0.01035489, + "balance_loss_clip": 1.05737627, + "balance_loss_mlp": 1.02605999, + "epoch": 0.28918415198701375, + "flos": 28585540333440.0, + "grad_norm": 1.872326979089486, + "language_loss": 0.85319579, + "learning_rate": 3.335905387430283e-06, + "loss": 0.8754648, + "num_input_tokens_seen": 51492890, + "step": 2405, + "time_per_iteration": 2.8243532180786133 + }, + { + "auxiliary_loss_clip": 0.01189454, + "auxiliary_loss_mlp": 0.01038838, + "balance_loss_clip": 1.05437219, + "balance_loss_mlp": 1.02852046, + "epoch": 0.28930439487765286, + "flos": 21944688007680.0, + "grad_norm": 1.7425096879796231, + "language_loss": 0.82712269, + "learning_rate": 3.335325572507782e-06, + "loss": 0.84940559, + "num_input_tokens_seen": 51513390, + "step": 2406, + "time_per_iteration": 2.8193283081054688 + }, + { + "auxiliary_loss_clip": 0.01208972, + "auxiliary_loss_mlp": 0.00888037, + "balance_loss_clip": 1.06389558, + "balance_loss_mlp": 1.00073087, + "epoch": 0.28942463776829197, + "flos": 19281911955840.0, + "grad_norm": 1.7336093627561096, + "language_loss": 0.74049419, + "learning_rate": 3.3347455550174537e-06, + "loss": 0.76146424, + "num_input_tokens_seen": 51532730, + "step": 2407, + "time_per_iteration": 3.6782917976379395 + }, + { + "auxiliary_loss_clip": 0.01171343, + "auxiliary_loss_mlp": 0.0103383, + "balance_loss_clip": 1.05079675, + "balance_loss_mlp": 1.022457, + "epoch": 0.289544880658931, + "flos": 14645700737280.0, + "grad_norm": 1.9874609254815645, + "language_loss": 0.68433899, + "learning_rate": 3.3341653350472864e-06, + "loss": 0.70639074, + "num_input_tokens_seen": 51549560, + "step": 2408, + "time_per_iteration": 3.7413110733032227 + }, + { + "auxiliary_loss_clip": 0.01217812, + "auxiliary_loss_mlp": 0.01045269, + "balance_loss_clip": 1.06226075, + "balance_loss_mlp": 1.03322887, + "epoch": 0.28966512354957014, + "flos": 28621881918720.0, + "grad_norm": 2.252556462418721, + "language_loss": 0.69325805, + "learning_rate": 3.333584912685298e-06, + "loss": 0.71588886, + "num_input_tokens_seen": 51568180, + "step": 2409, + "time_per_iteration": 3.597308874130249 + }, + { + "auxiliary_loss_clip": 0.01079736, + "auxiliary_loss_mlp": 0.0101129, + "balance_loss_clip": 1.02672601, + "balance_loss_mlp": 1.00865602, + "epoch": 0.28978536644020925, + "flos": 64711784511360.0, + "grad_norm": 0.8660344344321993, + "language_loss": 0.55570751, + "learning_rate": 3.3330042880195385e-06, + "loss": 0.57661778, + "num_input_tokens_seen": 51622530, + "step": 2410, + "time_per_iteration": 3.291290521621704 + }, + { + "auxiliary_loss_clip": 0.01186098, + "auxiliary_loss_mlp": 0.01034607, + "balance_loss_clip": 1.05535662, + "balance_loss_mlp": 1.02422941, + "epoch": 0.2899056093308483, + "flos": 18624638937600.0, + "grad_norm": 1.7166540268333859, + "language_loss": 0.78590775, + "learning_rate": 3.3324234611380888e-06, + "loss": 0.80811483, + "num_input_tokens_seen": 51641260, + "step": 2411, + "time_per_iteration": 2.7634074687957764 + }, + { + "auxiliary_loss_clip": 0.01169889, + "auxiliary_loss_mlp": 0.0103907, + "balance_loss_clip": 1.0546484, + "balance_loss_mlp": 1.02875853, + "epoch": 0.2900258522214874, + "flos": 22893735202560.0, + "grad_norm": 1.597428815147081, + "language_loss": 0.81952602, + "learning_rate": 3.3318424321290596e-06, + "loss": 0.84161562, + "num_input_tokens_seen": 51660975, + "step": 2412, + "time_per_iteration": 2.8411495685577393 + }, + { + "auxiliary_loss_clip": 0.01083084, + "auxiliary_loss_mlp": 0.01003979, + "balance_loss_clip": 1.02836609, + "balance_loss_mlp": 1.00133216, + "epoch": 0.2901460951121265, + "flos": 71106036013440.0, + "grad_norm": 0.8246955562395023, + "language_loss": 0.59921467, + "learning_rate": 3.3312612010805917e-06, + "loss": 0.62008536, + "num_input_tokens_seen": 51720550, + "step": 2413, + "time_per_iteration": 3.3843348026275635 + }, + { + "auxiliary_loss_clip": 0.0117664, + "auxiliary_loss_mlp": 0.0103807, + "balance_loss_clip": 1.05479074, + "balance_loss_mlp": 1.02747881, + "epoch": 0.2902663380027656, + "flos": 32160986081280.0, + "grad_norm": 1.9874210879537118, + "language_loss": 0.70686859, + "learning_rate": 3.330679768080858e-06, + "loss": 0.72901565, + "num_input_tokens_seen": 51744435, + "step": 2414, + "time_per_iteration": 2.863725185394287 + }, + { + "auxiliary_loss_clip": 0.0119636, + "auxiliary_loss_mlp": 0.01036425, + "balance_loss_clip": 1.06143534, + "balance_loss_mlp": 1.02614307, + "epoch": 0.2903865808934047, + "flos": 29351658539520.0, + "grad_norm": 2.1750751143777727, + "language_loss": 0.83261806, + "learning_rate": 3.3300981332180627e-06, + "loss": 0.8549459, + "num_input_tokens_seen": 51763640, + "step": 2415, + "time_per_iteration": 2.8998613357543945 + }, + { + "auxiliary_loss_clip": 0.01182358, + "auxiliary_loss_mlp": 0.01034089, + "balance_loss_clip": 1.05657005, + "balance_loss_mlp": 1.0234735, + "epoch": 0.29050682378404374, + "flos": 17089026647040.0, + "grad_norm": 1.8777734959801333, + "language_loss": 0.80095536, + "learning_rate": 3.3295162965804373e-06, + "loss": 0.82311988, + "num_input_tokens_seen": 51782135, + "step": 2416, + "time_per_iteration": 2.7689552307128906 + }, + { + "auxiliary_loss_clip": 0.01174844, + "auxiliary_loss_mlp": 0.01033288, + "balance_loss_clip": 1.05836463, + "balance_loss_mlp": 1.02327478, + "epoch": 0.29062706667468285, + "flos": 17858233422720.0, + "grad_norm": 7.038318192672889, + "language_loss": 0.78061712, + "learning_rate": 3.328934258256247e-06, + "loss": 0.80269843, + "num_input_tokens_seen": 51800200, + "step": 2417, + "time_per_iteration": 2.853278875350952 + }, + { + "auxiliary_loss_clip": 0.01192296, + "auxiliary_loss_mlp": 0.01034759, + "balance_loss_clip": 1.05424857, + "balance_loss_mlp": 1.02423286, + "epoch": 0.29074730956532197, + "flos": 24279815174400.0, + "grad_norm": 1.9301045139944264, + "language_loss": 0.66861659, + "learning_rate": 3.3283520183337856e-06, + "loss": 0.69088715, + "num_input_tokens_seen": 51819905, + "step": 2418, + "time_per_iteration": 2.7054054737091064 + }, + { + "auxiliary_loss_clip": 0.01180597, + "auxiliary_loss_mlp": 0.01044769, + "balance_loss_clip": 1.05337036, + "balance_loss_mlp": 1.03475547, + "epoch": 0.290867552455961, + "flos": 22340961826560.0, + "grad_norm": 2.1542230209299733, + "language_loss": 0.688761, + "learning_rate": 3.3277695769013797e-06, + "loss": 0.71101463, + "num_input_tokens_seen": 51839350, + "step": 2419, + "time_per_iteration": 2.8581433296203613 + }, + { + "auxiliary_loss_clip": 0.01197626, + "auxiliary_loss_mlp": 0.01042913, + "balance_loss_clip": 1.05948925, + "balance_loss_mlp": 1.03189206, + "epoch": 0.29098779534660013, + "flos": 23186155824000.0, + "grad_norm": 2.104080954772852, + "language_loss": 0.77490127, + "learning_rate": 3.327186934047385e-06, + "loss": 0.79730666, + "num_input_tokens_seen": 51858045, + "step": 2420, + "time_per_iteration": 2.657254457473755 + }, + { + "auxiliary_loss_clip": 0.01172892, + "auxiliary_loss_mlp": 0.01031224, + "balance_loss_clip": 1.04998672, + "balance_loss_mlp": 1.02040553, + "epoch": 0.29110803823723924, + "flos": 15304194817920.0, + "grad_norm": 2.536097566914947, + "language_loss": 0.65713722, + "learning_rate": 3.3266040898601877e-06, + "loss": 0.67917836, + "num_input_tokens_seen": 51875880, + "step": 2421, + "time_per_iteration": 2.734609603881836 + }, + { + "auxiliary_loss_clip": 0.0115404, + "auxiliary_loss_mlp": 0.01038927, + "balance_loss_clip": 1.04647875, + "balance_loss_mlp": 1.02853179, + "epoch": 0.2912282811278783, + "flos": 22595352923520.0, + "grad_norm": 1.889680596201566, + "language_loss": 0.78211051, + "learning_rate": 3.3260210444282045e-06, + "loss": 0.80404019, + "num_input_tokens_seen": 51893835, + "step": 2422, + "time_per_iteration": 2.813363552093506 + }, + { + "auxiliary_loss_clip": 0.01190019, + "auxiliary_loss_mlp": 0.01037343, + "balance_loss_clip": 1.05796885, + "balance_loss_mlp": 1.02675676, + "epoch": 0.2913485240185174, + "flos": 24497900599680.0, + "grad_norm": 2.0755512189472753, + "language_loss": 0.73196226, + "learning_rate": 3.325437797839883e-06, + "loss": 0.75423592, + "num_input_tokens_seen": 51912205, + "step": 2423, + "time_per_iteration": 2.67997670173645 + }, + { + "auxiliary_loss_clip": 0.01205384, + "auxiliary_loss_mlp": 0.0103768, + "balance_loss_clip": 1.05812502, + "balance_loss_mlp": 1.02748191, + "epoch": 0.2914687669091565, + "flos": 17931024334080.0, + "grad_norm": 2.2916875175881315, + "language_loss": 0.74780405, + "learning_rate": 3.3248543501837015e-06, + "loss": 0.7702347, + "num_input_tokens_seen": 51929410, + "step": 2424, + "time_per_iteration": 2.660665512084961 + }, + { + "auxiliary_loss_clip": 0.01170497, + "auxiliary_loss_mlp": 0.01036489, + "balance_loss_clip": 1.05484748, + "balance_loss_mlp": 1.02594519, + "epoch": 0.2915890097997956, + "flos": 22529313768960.0, + "grad_norm": 1.8625329001809678, + "language_loss": 0.77457225, + "learning_rate": 3.3242707015481684e-06, + "loss": 0.79664212, + "num_input_tokens_seen": 51949345, + "step": 2425, + "time_per_iteration": 2.847644805908203 + }, + { + "auxiliary_loss_clip": 0.01182543, + "auxiliary_loss_mlp": 0.01035787, + "balance_loss_clip": 1.05231476, + "balance_loss_mlp": 1.02557039, + "epoch": 0.2917092526904347, + "flos": 13845216193920.0, + "grad_norm": 1.7862080040017987, + "language_loss": 0.80789667, + "learning_rate": 3.323686852021823e-06, + "loss": 0.83007997, + "num_input_tokens_seen": 51966855, + "step": 2426, + "time_per_iteration": 2.783569097518921 + }, + { + "auxiliary_loss_clip": 0.01178627, + "auxiliary_loss_mlp": 0.01037457, + "balance_loss_clip": 1.05116177, + "balance_loss_mlp": 1.02719927, + "epoch": 0.2918294955810738, + "flos": 22674859678080.0, + "grad_norm": 1.8894762018572133, + "language_loss": 0.79427814, + "learning_rate": 3.323102801693235e-06, + "loss": 0.81643903, + "num_input_tokens_seen": 51985620, + "step": 2427, + "time_per_iteration": 2.7656638622283936 + }, + { + "auxiliary_loss_clip": 0.01187143, + "auxiliary_loss_mlp": 0.01026418, + "balance_loss_clip": 1.05573797, + "balance_loss_mlp": 1.01611233, + "epoch": 0.29194973847171285, + "flos": 23438284364160.0, + "grad_norm": 1.9621679176893823, + "language_loss": 0.80689168, + "learning_rate": 3.322518550651003e-06, + "loss": 0.8290273, + "num_input_tokens_seen": 52004930, + "step": 2428, + "time_per_iteration": 2.7425708770751953 + }, + { + "auxiliary_loss_clip": 0.0119241, + "auxiliary_loss_mlp": 0.0103743, + "balance_loss_clip": 1.05601645, + "balance_loss_mlp": 1.0265466, + "epoch": 0.29206998136235196, + "flos": 21909064694400.0, + "grad_norm": 1.9026107207663954, + "language_loss": 0.81574678, + "learning_rate": 3.3219340989837586e-06, + "loss": 0.83804512, + "num_input_tokens_seen": 52024920, + "step": 2429, + "time_per_iteration": 2.7486298084259033 + }, + { + "auxiliary_loss_clip": 0.0118602, + "auxiliary_loss_mlp": 0.01038954, + "balance_loss_clip": 1.05665755, + "balance_loss_mlp": 1.02860665, + "epoch": 0.292190224252991, + "flos": 23215925220480.0, + "grad_norm": 1.8377007315215381, + "language_loss": 0.80159593, + "learning_rate": 3.3213494467801625e-06, + "loss": 0.82384562, + "num_input_tokens_seen": 52044095, + "step": 2430, + "time_per_iteration": 2.762493371963501 + }, + { + "auxiliary_loss_clip": 0.01142516, + "auxiliary_loss_mlp": 0.01030402, + "balance_loss_clip": 1.04382277, + "balance_loss_mlp": 1.01968503, + "epoch": 0.2923104671436301, + "flos": 20740818752640.0, + "grad_norm": 2.692137165697415, + "language_loss": 0.71298414, + "learning_rate": 3.3207645941289063e-06, + "loss": 0.73471332, + "num_input_tokens_seen": 52062440, + "step": 2431, + "time_per_iteration": 3.050419569015503 + }, + { + "auxiliary_loss_clip": 0.01195113, + "auxiliary_loss_mlp": 0.00888004, + "balance_loss_clip": 1.05865872, + "balance_loss_mlp": 1.00058079, + "epoch": 0.29243071003426924, + "flos": 35809114999680.0, + "grad_norm": 1.7982260841262392, + "language_loss": 0.80101943, + "learning_rate": 3.320179541118711e-06, + "loss": 0.82185066, + "num_input_tokens_seen": 52084940, + "step": 2432, + "time_per_iteration": 3.3006625175476074 + }, + { + "auxiliary_loss_clip": 0.01089677, + "auxiliary_loss_mlp": 0.01005223, + "balance_loss_clip": 1.02048635, + "balance_loss_mlp": 1.00257611, + "epoch": 0.2925509529249083, + "flos": 58081598524800.0, + "grad_norm": 0.9967631245931414, + "language_loss": 0.60276127, + "learning_rate": 3.3195942878383293e-06, + "loss": 0.62371016, + "num_input_tokens_seen": 52141040, + "step": 2433, + "time_per_iteration": 3.4141247272491455 + }, + { + "auxiliary_loss_clip": 0.01197306, + "auxiliary_loss_mlp": 0.01033745, + "balance_loss_clip": 1.05877805, + "balance_loss_mlp": 1.02330184, + "epoch": 0.2926711958155474, + "flos": 21397122103680.0, + "grad_norm": 1.823177988841059, + "language_loss": 0.78009528, + "learning_rate": 3.319008834376543e-06, + "loss": 0.80240577, + "num_input_tokens_seen": 52160730, + "step": 2434, + "time_per_iteration": 4.668346881866455 + }, + { + "auxiliary_loss_clip": 0.01180453, + "auxiliary_loss_mlp": 0.01034968, + "balance_loss_clip": 1.05247772, + "balance_loss_mlp": 1.02428722, + "epoch": 0.2927914387061865, + "flos": 23185796688000.0, + "grad_norm": 2.133651888229114, + "language_loss": 0.8828032, + "learning_rate": 3.3184231808221654e-06, + "loss": 0.90495741, + "num_input_tokens_seen": 52175055, + "step": 2435, + "time_per_iteration": 4.643437623977661 + }, + { + "auxiliary_loss_clip": 0.01180155, + "auxiliary_loss_mlp": 0.01035837, + "balance_loss_clip": 1.05683005, + "balance_loss_mlp": 1.02563226, + "epoch": 0.29291168159682557, + "flos": 22455553190400.0, + "grad_norm": 5.051028661950565, + "language_loss": 0.63030124, + "learning_rate": 3.3178373272640394e-06, + "loss": 0.65246117, + "num_input_tokens_seen": 52194150, + "step": 2436, + "time_per_iteration": 2.785315990447998 + }, + { + "auxiliary_loss_clip": 0.01202947, + "auxiliary_loss_mlp": 0.01040891, + "balance_loss_clip": 1.05913949, + "balance_loss_mlp": 1.03066278, + "epoch": 0.2930319244874647, + "flos": 21170632896000.0, + "grad_norm": 2.2471854092526793, + "language_loss": 0.85322309, + "learning_rate": 3.3172512737910387e-06, + "loss": 0.87566143, + "num_input_tokens_seen": 52211660, + "step": 2437, + "time_per_iteration": 2.7478291988372803 + }, + { + "auxiliary_loss_clip": 0.01192852, + "auxiliary_loss_mlp": 0.01037538, + "balance_loss_clip": 1.05470657, + "balance_loss_mlp": 1.02780473, + "epoch": 0.2931521673781038, + "flos": 31357843931520.0, + "grad_norm": 2.263763933272973, + "language_loss": 0.88828701, + "learning_rate": 3.3166650204920674e-06, + "loss": 0.91059089, + "num_input_tokens_seen": 52232830, + "step": 2438, + "time_per_iteration": 2.7851474285125732 + }, + { + "auxiliary_loss_clip": 0.01195948, + "auxiliary_loss_mlp": 0.01038204, + "balance_loss_clip": 1.05878305, + "balance_loss_mlp": 1.02701616, + "epoch": 0.29327241026874284, + "flos": 24200990778240.0, + "grad_norm": 1.5550845977349477, + "language_loss": 0.81693041, + "learning_rate": 3.316078567456059e-06, + "loss": 0.83927196, + "num_input_tokens_seen": 52250670, + "step": 2439, + "time_per_iteration": 2.6779227256774902 + }, + { + "auxiliary_loss_clip": 0.01164688, + "auxiliary_loss_mlp": 0.01038181, + "balance_loss_clip": 1.05557323, + "balance_loss_mlp": 1.02785182, + "epoch": 0.29339265315938196, + "flos": 24242611662720.0, + "grad_norm": 1.608184951533296, + "language_loss": 0.75931942, + "learning_rate": 3.3154919147719786e-06, + "loss": 0.78134811, + "num_input_tokens_seen": 52271685, + "step": 2440, + "time_per_iteration": 3.024357557296753 + }, + { + "auxiliary_loss_clip": 0.01195114, + "auxiliary_loss_mlp": 0.01036947, + "balance_loss_clip": 1.05822992, + "balance_loss_mlp": 1.02646828, + "epoch": 0.29351289605002107, + "flos": 16946641134720.0, + "grad_norm": 1.9989006720598448, + "language_loss": 0.86238313, + "learning_rate": 3.31490506252882e-06, + "loss": 0.88470376, + "num_input_tokens_seen": 52291065, + "step": 2441, + "time_per_iteration": 2.76651930809021 + }, + { + "auxiliary_loss_clip": 0.01165593, + "auxiliary_loss_mlp": 0.0102634, + "balance_loss_clip": 1.04870498, + "balance_loss_mlp": 1.01665998, + "epoch": 0.2936331389406601, + "flos": 19829082810240.0, + "grad_norm": 2.4118867525046883, + "language_loss": 0.84085631, + "learning_rate": 3.31431801081561e-06, + "loss": 0.86277562, + "num_input_tokens_seen": 52310000, + "step": 2442, + "time_per_iteration": 2.793430805206299 + }, + { + "auxiliary_loss_clip": 0.01078151, + "auxiliary_loss_mlp": 0.01001439, + "balance_loss_clip": 1.02198029, + "balance_loss_mlp": 0.99906683, + "epoch": 0.29375338183129923, + "flos": 71416844398080.0, + "grad_norm": 0.8956149876449672, + "language_loss": 0.67950237, + "learning_rate": 3.313730759721402e-06, + "loss": 0.70029831, + "num_input_tokens_seen": 52372930, + "step": 2443, + "time_per_iteration": 3.4149553775787354 + }, + { + "auxiliary_loss_clip": 0.01184072, + "auxiliary_loss_mlp": 0.01034889, + "balance_loss_clip": 1.05887341, + "balance_loss_mlp": 1.02511942, + "epoch": 0.29387362472193834, + "flos": 22054502862720.0, + "grad_norm": 2.006344048821129, + "language_loss": 0.8615666, + "learning_rate": 3.313143309335282e-06, + "loss": 0.88375616, + "num_input_tokens_seen": 52391420, + "step": 2444, + "time_per_iteration": 2.8241233825683594 + }, + { + "auxiliary_loss_clip": 0.01173608, + "auxiliary_loss_mlp": 0.01035103, + "balance_loss_clip": 1.05514717, + "balance_loss_mlp": 1.02488685, + "epoch": 0.2939938676125774, + "flos": 22966418373120.0, + "grad_norm": 1.863763921235404, + "language_loss": 0.84225082, + "learning_rate": 3.3125556597463665e-06, + "loss": 0.86433792, + "num_input_tokens_seen": 52410725, + "step": 2445, + "time_per_iteration": 2.797306537628174 + }, + { + "auxiliary_loss_clip": 0.01194949, + "auxiliary_loss_mlp": 0.01034846, + "balance_loss_clip": 1.06005859, + "balance_loss_mlp": 1.02497578, + "epoch": 0.2941141105032165, + "flos": 31358705857920.0, + "grad_norm": 1.606998205698301, + "language_loss": 0.66065246, + "learning_rate": 3.311967811043801e-06, + "loss": 0.68295038, + "num_input_tokens_seen": 52432645, + "step": 2446, + "time_per_iteration": 2.8451340198516846 + }, + { + "auxiliary_loss_clip": 0.01192027, + "auxiliary_loss_mlp": 0.01039224, + "balance_loss_clip": 1.05832839, + "balance_loss_mlp": 1.02861404, + "epoch": 0.29423435339385556, + "flos": 23222138273280.0, + "grad_norm": 1.9813323250128205, + "language_loss": 0.81449771, + "learning_rate": 3.3113797633167617e-06, + "loss": 0.83681029, + "num_input_tokens_seen": 52450940, + "step": 2447, + "time_per_iteration": 2.735443592071533 + }, + { + "auxiliary_loss_clip": 0.01202555, + "auxiliary_loss_mlp": 0.01028869, + "balance_loss_clip": 1.05776501, + "balance_loss_mlp": 1.01853418, + "epoch": 0.2943545962844947, + "flos": 26864054138880.0, + "grad_norm": 3.5834338574478486, + "language_loss": 0.69632727, + "learning_rate": 3.310791516654455e-06, + "loss": 0.71864152, + "num_input_tokens_seen": 52468000, + "step": 2448, + "time_per_iteration": 2.6816608905792236 + }, + { + "auxiliary_loss_clip": 0.01181865, + "auxiliary_loss_mlp": 0.01036257, + "balance_loss_clip": 1.05435681, + "balance_loss_mlp": 1.02557003, + "epoch": 0.2944748391751338, + "flos": 20231677422720.0, + "grad_norm": 1.8780295181218383, + "language_loss": 0.79337978, + "learning_rate": 3.3102030711461177e-06, + "loss": 0.815561, + "num_input_tokens_seen": 52487575, + "step": 2449, + "time_per_iteration": 2.8022549152374268 + }, + { + "auxiliary_loss_clip": 0.01176742, + "auxiliary_loss_mlp": 0.01033486, + "balance_loss_clip": 1.05408204, + "balance_loss_mlp": 1.02293015, + "epoch": 0.29459508206577284, + "flos": 15960965045760.0, + "grad_norm": 1.8468822353851813, + "language_loss": 0.68237203, + "learning_rate": 3.3096144268810156e-06, + "loss": 0.70447433, + "num_input_tokens_seen": 52506335, + "step": 2450, + "time_per_iteration": 2.782745599746704 + }, + { + "auxiliary_loss_clip": 0.01186019, + "auxiliary_loss_mlp": 0.01030558, + "balance_loss_clip": 1.05510712, + "balance_loss_mlp": 1.02075315, + "epoch": 0.29471532495641195, + "flos": 20412882558720.0, + "grad_norm": 1.878346956746974, + "language_loss": 0.73065281, + "learning_rate": 3.3090255839484462e-06, + "loss": 0.75281858, + "num_input_tokens_seen": 52524330, + "step": 2451, + "time_per_iteration": 2.86303973197937 + }, + { + "auxiliary_loss_clip": 0.01186374, + "auxiliary_loss_mlp": 0.01033904, + "balance_loss_clip": 1.05376625, + "balance_loss_mlp": 1.02352118, + "epoch": 0.29483556784705106, + "flos": 20376576887040.0, + "grad_norm": 1.9572701073517262, + "language_loss": 0.85744017, + "learning_rate": 3.3084365424377366e-06, + "loss": 0.87964296, + "num_input_tokens_seen": 52543095, + "step": 2452, + "time_per_iteration": 2.7274181842803955 + }, + { + "auxiliary_loss_clip": 0.01073912, + "auxiliary_loss_mlp": 0.01023737, + "balance_loss_clip": 1.02843463, + "balance_loss_mlp": 1.02042317, + "epoch": 0.2949558107376901, + "flos": 68555660595840.0, + "grad_norm": 0.7400999678228314, + "language_loss": 0.56005174, + "learning_rate": 3.307847302438245e-06, + "loss": 0.58102822, + "num_input_tokens_seen": 52597075, + "step": 2453, + "time_per_iteration": 3.2684779167175293 + }, + { + "auxiliary_loss_clip": 0.01150037, + "auxiliary_loss_mlp": 0.01036549, + "balance_loss_clip": 1.04856801, + "balance_loss_mlp": 1.02595711, + "epoch": 0.2950760536283292, + "flos": 16107085572480.0, + "grad_norm": 2.222473433329535, + "language_loss": 0.78106201, + "learning_rate": 3.3072578640393562e-06, + "loss": 0.80292785, + "num_input_tokens_seen": 52614410, + "step": 2454, + "time_per_iteration": 2.8150830268859863 + }, + { + "auxiliary_loss_clip": 0.01182326, + "auxiliary_loss_mlp": 0.01036398, + "balance_loss_clip": 1.05480599, + "balance_loss_mlp": 1.02597928, + "epoch": 0.29519629651896834, + "flos": 20483626394880.0, + "grad_norm": 2.655775991709739, + "language_loss": 0.79263359, + "learning_rate": 3.3066682273304886e-06, + "loss": 0.81482089, + "num_input_tokens_seen": 52632055, + "step": 2455, + "time_per_iteration": 2.7839152812957764 + }, + { + "auxiliary_loss_clip": 0.01199846, + "auxiliary_loss_mlp": 0.00888704, + "balance_loss_clip": 1.05714273, + "balance_loss_mlp": 1.00075722, + "epoch": 0.2953165394096074, + "flos": 18916484941440.0, + "grad_norm": 2.706841223615375, + "language_loss": 0.78934848, + "learning_rate": 3.3060783924010904e-06, + "loss": 0.81023395, + "num_input_tokens_seen": 52649980, + "step": 2456, + "time_per_iteration": 2.819626569747925 + }, + { + "auxiliary_loss_clip": 0.01175107, + "auxiliary_loss_mlp": 0.01030483, + "balance_loss_clip": 1.05443895, + "balance_loss_mlp": 1.02042818, + "epoch": 0.2954367823002465, + "flos": 20624467622400.0, + "grad_norm": 2.2155580875223695, + "language_loss": 0.85245758, + "learning_rate": 3.3054883593406387e-06, + "loss": 0.87451345, + "num_input_tokens_seen": 52664730, + "step": 2457, + "time_per_iteration": 2.866448163986206 + }, + { + "auxiliary_loss_clip": 0.01182732, + "auxiliary_loss_mlp": 0.01030989, + "balance_loss_clip": 1.05418968, + "balance_loss_mlp": 1.02048075, + "epoch": 0.2955570251908856, + "flos": 31175525473920.0, + "grad_norm": 2.5011733849537885, + "language_loss": 0.65181911, + "learning_rate": 3.3048981282386404e-06, + "loss": 0.67395627, + "num_input_tokens_seen": 52686040, + "step": 2458, + "time_per_iteration": 2.9184587001800537 + }, + { + "auxiliary_loss_clip": 0.01163022, + "auxiliary_loss_mlp": 0.01039463, + "balance_loss_clip": 1.05303836, + "balance_loss_mlp": 1.02880621, + "epoch": 0.29567726808152467, + "flos": 21650328051840.0, + "grad_norm": 1.9648250201450332, + "language_loss": 0.82658434, + "learning_rate": 3.304307699184634e-06, + "loss": 0.84860921, + "num_input_tokens_seen": 52704630, + "step": 2459, + "time_per_iteration": 3.8216819763183594 + }, + { + "auxiliary_loss_clip": 0.01182563, + "auxiliary_loss_mlp": 0.01033477, + "balance_loss_clip": 1.05591321, + "balance_loss_mlp": 1.02401161, + "epoch": 0.2957975109721638, + "flos": 24243868638720.0, + "grad_norm": 1.802141626898443, + "language_loss": 0.78546238, + "learning_rate": 3.3037170722681866e-06, + "loss": 0.80762279, + "num_input_tokens_seen": 52725465, + "step": 2460, + "time_per_iteration": 4.745901584625244 + }, + { + "auxiliary_loss_clip": 0.01166658, + "auxiliary_loss_mlp": 0.01032044, + "balance_loss_clip": 1.05207872, + "balance_loss_mlp": 1.02236986, + "epoch": 0.29591775386280283, + "flos": 13479717352320.0, + "grad_norm": 6.389280250412514, + "language_loss": 0.68164599, + "learning_rate": 3.3031262475788956e-06, + "loss": 0.70363307, + "num_input_tokens_seen": 52742405, + "step": 2461, + "time_per_iteration": 3.668882369995117 + }, + { + "auxiliary_loss_clip": 0.01179164, + "auxiliary_loss_mlp": 0.01029697, + "balance_loss_clip": 1.05324125, + "balance_loss_mlp": 1.01941502, + "epoch": 0.29603799675344195, + "flos": 17749783284480.0, + "grad_norm": 2.265615668573457, + "language_loss": 0.73171079, + "learning_rate": 3.3025352252063897e-06, + "loss": 0.75379932, + "num_input_tokens_seen": 52761100, + "step": 2462, + "time_per_iteration": 2.8425557613372803 + }, + { + "auxiliary_loss_clip": 0.01191879, + "auxiliary_loss_mlp": 0.01034492, + "balance_loss_clip": 1.0588268, + "balance_loss_mlp": 1.02434123, + "epoch": 0.29615823964408106, + "flos": 22783920347520.0, + "grad_norm": 1.6580759782565506, + "language_loss": 0.75035191, + "learning_rate": 3.3019440052403252e-06, + "loss": 0.77261555, + "num_input_tokens_seen": 52780965, + "step": 2463, + "time_per_iteration": 2.7624170780181885 + }, + { + "auxiliary_loss_clip": 0.01184562, + "auxiliary_loss_mlp": 0.01037853, + "balance_loss_clip": 1.0544982, + "balance_loss_mlp": 1.02792883, + "epoch": 0.2962784825347201, + "flos": 23514199758720.0, + "grad_norm": 2.6628264262087438, + "language_loss": 0.70917678, + "learning_rate": 3.30135258777039e-06, + "loss": 0.73140097, + "num_input_tokens_seen": 52800335, + "step": 2464, + "time_per_iteration": 2.784592866897583 + }, + { + "auxiliary_loss_clip": 0.01195713, + "auxiliary_loss_mlp": 0.00888289, + "balance_loss_clip": 1.05371785, + "balance_loss_mlp": 1.00082159, + "epoch": 0.2963987254253592, + "flos": 16362769559040.0, + "grad_norm": 2.067749463804313, + "language_loss": 0.70684516, + "learning_rate": 3.3007609728863024e-06, + "loss": 0.72768521, + "num_input_tokens_seen": 52818425, + "step": 2465, + "time_per_iteration": 2.7631337642669678 + }, + { + "auxiliary_loss_clip": 0.01145745, + "auxiliary_loss_mlp": 0.01033036, + "balance_loss_clip": 1.05209398, + "balance_loss_mlp": 1.02268267, + "epoch": 0.29651896831599833, + "flos": 33472263980160.0, + "grad_norm": 1.7879362451674943, + "language_loss": 0.729491, + "learning_rate": 3.300169160677809e-06, + "loss": 0.75127876, + "num_input_tokens_seen": 52842340, + "step": 2466, + "time_per_iteration": 3.008002281188965 + }, + { + "auxiliary_loss_clip": 0.01182067, + "auxiliary_loss_mlp": 0.01040549, + "balance_loss_clip": 1.05530047, + "balance_loss_mlp": 1.02940845, + "epoch": 0.2966392112066374, + "flos": 23805363404160.0, + "grad_norm": 2.5197268153494075, + "language_loss": 0.78223091, + "learning_rate": 3.2995771512346878e-06, + "loss": 0.80445707, + "num_input_tokens_seen": 52860690, + "step": 2467, + "time_per_iteration": 2.8692872524261475 + }, + { + "auxiliary_loss_clip": 0.01207057, + "auxiliary_loss_mlp": 0.00888291, + "balance_loss_clip": 1.05909348, + "balance_loss_mlp": 1.00089717, + "epoch": 0.2967594540972765, + "flos": 19938466702080.0, + "grad_norm": 1.9885166377621761, + "language_loss": 0.72584045, + "learning_rate": 3.298984944646746e-06, + "loss": 0.74679393, + "num_input_tokens_seen": 52879370, + "step": 2468, + "time_per_iteration": 2.712650775909424 + }, + { + "auxiliary_loss_clip": 0.01195117, + "auxiliary_loss_mlp": 0.00887605, + "balance_loss_clip": 1.05703425, + "balance_loss_mlp": 1.00067544, + "epoch": 0.2968796969879156, + "flos": 23732823888000.0, + "grad_norm": 1.8426892507722865, + "language_loss": 0.81641412, + "learning_rate": 3.298392541003822e-06, + "loss": 0.83724129, + "num_input_tokens_seen": 52898775, + "step": 2469, + "time_per_iteration": 2.7391703128814697 + }, + { + "auxiliary_loss_clip": 0.01184575, + "auxiliary_loss_mlp": 0.01033837, + "balance_loss_clip": 1.05711257, + "balance_loss_mlp": 1.02378166, + "epoch": 0.29699993987855466, + "flos": 22893699288960.0, + "grad_norm": 1.5411293318198893, + "language_loss": 0.89773607, + "learning_rate": 3.2977999403957806e-06, + "loss": 0.91992021, + "num_input_tokens_seen": 52917535, + "step": 2470, + "time_per_iteration": 2.770214796066284 + }, + { + "auxiliary_loss_clip": 0.01204458, + "auxiliary_loss_mlp": 0.0103152, + "balance_loss_clip": 1.05879998, + "balance_loss_mlp": 1.0219655, + "epoch": 0.2971201827691938, + "flos": 33832555349760.0, + "grad_norm": 2.776864359194087, + "language_loss": 0.67132992, + "learning_rate": 3.2972071429125207e-06, + "loss": 0.6936897, + "num_input_tokens_seen": 52938755, + "step": 2471, + "time_per_iteration": 2.760385751724243 + }, + { + "auxiliary_loss_clip": 0.01168842, + "auxiliary_loss_mlp": 0.01036228, + "balance_loss_clip": 1.05196655, + "balance_loss_mlp": 1.02606499, + "epoch": 0.2972404256598329, + "flos": 22054359208320.0, + "grad_norm": 1.956248345288739, + "language_loss": 0.88106936, + "learning_rate": 3.2966141486439682e-06, + "loss": 0.90312004, + "num_input_tokens_seen": 52957945, + "step": 2472, + "time_per_iteration": 2.800309419631958 + }, + { + "auxiliary_loss_clip": 0.01154205, + "auxiliary_loss_mlp": 0.01038204, + "balance_loss_clip": 1.04616952, + "balance_loss_mlp": 1.02745748, + "epoch": 0.29736066855047194, + "flos": 31978595796480.0, + "grad_norm": 1.9878284619481708, + "language_loss": 0.63977677, + "learning_rate": 3.29602095768008e-06, + "loss": 0.66170084, + "num_input_tokens_seen": 52978460, + "step": 2473, + "time_per_iteration": 3.009261131286621 + }, + { + "auxiliary_loss_clip": 0.01172899, + "auxiliary_loss_mlp": 0.01035875, + "balance_loss_clip": 1.05231953, + "balance_loss_mlp": 1.02580738, + "epoch": 0.29748091144111105, + "flos": 33510401245440.0, + "grad_norm": 1.8842783237804932, + "language_loss": 0.63828629, + "learning_rate": 3.2954275701108437e-06, + "loss": 0.66037405, + "num_input_tokens_seen": 52999640, + "step": 2474, + "time_per_iteration": 2.891608476638794 + }, + { + "auxiliary_loss_clip": 0.01154318, + "auxiliary_loss_mlp": 0.01032559, + "balance_loss_clip": 1.04805136, + "balance_loss_mlp": 1.02267027, + "epoch": 0.29760115433175016, + "flos": 41283373409280.0, + "grad_norm": 1.6763564398573219, + "language_loss": 0.68655241, + "learning_rate": 3.294833986026275e-06, + "loss": 0.70842123, + "num_input_tokens_seen": 53022880, + "step": 2475, + "time_per_iteration": 2.9089455604553223 + }, + { + "auxiliary_loss_clip": 0.01170148, + "auxiliary_loss_mlp": 0.0103167, + "balance_loss_clip": 1.05443645, + "balance_loss_mlp": 1.02255034, + "epoch": 0.2977213972223892, + "flos": 24493339572480.0, + "grad_norm": 1.890218219037689, + "language_loss": 0.85336137, + "learning_rate": 3.29424020551642e-06, + "loss": 0.87537956, + "num_input_tokens_seen": 53041515, + "step": 2476, + "time_per_iteration": 2.837888240814209 + }, + { + "auxiliary_loss_clip": 0.0120785, + "auxiliary_loss_mlp": 0.01041287, + "balance_loss_clip": 1.05957794, + "balance_loss_mlp": 1.03048706, + "epoch": 0.2978416401130283, + "flos": 21285116519040.0, + "grad_norm": 3.0811768270807853, + "language_loss": 0.72030497, + "learning_rate": 3.2936462286713546e-06, + "loss": 0.74279636, + "num_input_tokens_seen": 53059865, + "step": 2477, + "time_per_iteration": 2.6792216300964355 + }, + { + "auxiliary_loss_clip": 0.01197014, + "auxiliary_loss_mlp": 0.01033108, + "balance_loss_clip": 1.05894375, + "balance_loss_mlp": 1.02250981, + "epoch": 0.2979618830036674, + "flos": 25772154554880.0, + "grad_norm": 1.970497661933394, + "language_loss": 0.77416354, + "learning_rate": 3.2930520555811846e-06, + "loss": 0.7964648, + "num_input_tokens_seen": 53079490, + "step": 2478, + "time_per_iteration": 2.840254306793213 + }, + { + "auxiliary_loss_clip": 0.01116857, + "auxiliary_loss_mlp": 0.00888726, + "balance_loss_clip": 1.04411387, + "balance_loss_mlp": 1.00075305, + "epoch": 0.2980821258943065, + "flos": 23476996247040.0, + "grad_norm": 2.0090134453835056, + "language_loss": 0.79783404, + "learning_rate": 3.292457686336046e-06, + "loss": 0.81788981, + "num_input_tokens_seen": 53098810, + "step": 2479, + "time_per_iteration": 3.097951889038086 + }, + { + "auxiliary_loss_clip": 0.01084121, + "auxiliary_loss_mlp": 0.01004494, + "balance_loss_clip": 1.01962817, + "balance_loss_mlp": 1.00195527, + "epoch": 0.2982023687849456, + "flos": 69752314195200.0, + "grad_norm": 0.8551577571849556, + "language_loss": 0.61237395, + "learning_rate": 3.291863121026105e-06, + "loss": 0.63326013, + "num_input_tokens_seen": 53162590, + "step": 2480, + "time_per_iteration": 3.8199455738067627 + }, + { + "auxiliary_loss_clip": 0.01194527, + "auxiliary_loss_mlp": 0.01032619, + "balance_loss_clip": 1.05928636, + "balance_loss_mlp": 1.02274823, + "epoch": 0.29832261167558466, + "flos": 29825930741760.0, + "grad_norm": 2.076474719399041, + "language_loss": 0.76859248, + "learning_rate": 3.2912683597415547e-06, + "loss": 0.79086387, + "num_input_tokens_seen": 53186675, + "step": 2481, + "time_per_iteration": 2.89684796333313 + }, + { + "auxiliary_loss_clip": 0.01175669, + "auxiliary_loss_mlp": 0.01042004, + "balance_loss_clip": 1.05432642, + "balance_loss_mlp": 1.03188944, + "epoch": 0.29844285456622377, + "flos": 33910158683520.0, + "grad_norm": 1.8535258279999154, + "language_loss": 0.78208166, + "learning_rate": 3.2906734025726213e-06, + "loss": 0.80425841, + "num_input_tokens_seen": 53205940, + "step": 2482, + "time_per_iteration": 2.920675277709961 + }, + { + "auxiliary_loss_clip": 0.01201938, + "auxiliary_loss_mlp": 0.01039994, + "balance_loss_clip": 1.05971599, + "balance_loss_mlp": 1.03036165, + "epoch": 0.2985630974568629, + "flos": 23876933253120.0, + "grad_norm": 1.8644291743540775, + "language_loss": 0.87737501, + "learning_rate": 3.290078249609559e-06, + "loss": 0.89979434, + "num_input_tokens_seen": 53225360, + "step": 2483, + "time_per_iteration": 2.7820498943328857 + }, + { + "auxiliary_loss_clip": 0.01191303, + "auxiliary_loss_mlp": 0.01039213, + "balance_loss_clip": 1.05912089, + "balance_loss_mlp": 1.02955151, + "epoch": 0.29868334034750194, + "flos": 21799106184960.0, + "grad_norm": 1.8617014540122603, + "language_loss": 0.8781569, + "learning_rate": 3.2894829009426514e-06, + "loss": 0.90046203, + "num_input_tokens_seen": 53243195, + "step": 2484, + "time_per_iteration": 2.7308454513549805 + }, + { + "auxiliary_loss_clip": 0.01189782, + "auxiliary_loss_mlp": 0.01038635, + "balance_loss_clip": 1.05721521, + "balance_loss_mlp": 1.02925324, + "epoch": 0.29880358323814105, + "flos": 25666649331840.0, + "grad_norm": 1.7960857770223442, + "language_loss": 0.77537668, + "learning_rate": 3.288887356662213e-06, + "loss": 0.79766095, + "num_input_tokens_seen": 53264530, + "step": 2485, + "time_per_iteration": 3.658796548843384 + }, + { + "auxiliary_loss_clip": 0.01088844, + "auxiliary_loss_mlp": 0.0100251, + "balance_loss_clip": 1.0243516, + "balance_loss_mlp": 0.99991113, + "epoch": 0.29892382612878016, + "flos": 71005846003200.0, + "grad_norm": 0.7707588410893392, + "language_loss": 0.59718227, + "learning_rate": 3.288291616858588e-06, + "loss": 0.61809587, + "num_input_tokens_seen": 53319920, + "step": 2486, + "time_per_iteration": 5.06678032875061 + }, + { + "auxiliary_loss_clip": 0.01159318, + "auxiliary_loss_mlp": 0.01042216, + "balance_loss_clip": 1.05447304, + "balance_loss_mlp": 1.0324465, + "epoch": 0.2990440690194192, + "flos": 25481134563840.0, + "grad_norm": 2.7668214261119273, + "language_loss": 0.76672578, + "learning_rate": 3.287695681622149e-06, + "loss": 0.78874111, + "num_input_tokens_seen": 53339270, + "step": 2487, + "time_per_iteration": 3.831754446029663 + }, + { + "auxiliary_loss_clip": 0.01190226, + "auxiliary_loss_mlp": 0.01032931, + "balance_loss_clip": 1.05784249, + "balance_loss_mlp": 1.02266717, + "epoch": 0.2991643119100583, + "flos": 23732357011200.0, + "grad_norm": 1.8035427526038519, + "language_loss": 0.81054711, + "learning_rate": 3.2870995510432982e-06, + "loss": 0.83277863, + "num_input_tokens_seen": 53357750, + "step": 2488, + "time_per_iteration": 2.817228317260742 + }, + { + "auxiliary_loss_clip": 0.01187331, + "auxiliary_loss_mlp": 0.01033294, + "balance_loss_clip": 1.0570085, + "balance_loss_mlp": 1.02370381, + "epoch": 0.29928455480069743, + "flos": 27417545786880.0, + "grad_norm": 1.9123776379765909, + "language_loss": 0.76643425, + "learning_rate": 3.2865032252124697e-06, + "loss": 0.7886405, + "num_input_tokens_seen": 53378265, + "step": 2489, + "time_per_iteration": 2.7509021759033203 + }, + { + "auxiliary_loss_clip": 0.01180111, + "auxiliary_loss_mlp": 0.01034327, + "balance_loss_clip": 1.05300736, + "balance_loss_mlp": 1.02487946, + "epoch": 0.2994047976913365, + "flos": 33692935184640.0, + "grad_norm": 1.9012885936835575, + "language_loss": 0.77724636, + "learning_rate": 3.2859067042201243e-06, + "loss": 0.79939073, + "num_input_tokens_seen": 53400305, + "step": 2490, + "time_per_iteration": 2.888284206390381 + }, + { + "auxiliary_loss_clip": 0.01127634, + "auxiliary_loss_mlp": 0.01027077, + "balance_loss_clip": 1.04496992, + "balance_loss_mlp": 1.01690793, + "epoch": 0.2995250405819756, + "flos": 16763963541120.0, + "grad_norm": 2.210786394765492, + "language_loss": 0.7817781, + "learning_rate": 3.2853099881567544e-06, + "loss": 0.80332524, + "num_input_tokens_seen": 53418705, + "step": 2491, + "time_per_iteration": 2.7823853492736816 + }, + { + "auxiliary_loss_clip": 0.01199199, + "auxiliary_loss_mlp": 0.01037821, + "balance_loss_clip": 1.05816865, + "balance_loss_mlp": 1.02831459, + "epoch": 0.29964528347261465, + "flos": 22963976248320.0, + "grad_norm": 2.1593675882287418, + "language_loss": 0.7907604, + "learning_rate": 3.284713077112881e-06, + "loss": 0.81313062, + "num_input_tokens_seen": 53438135, + "step": 2492, + "time_per_iteration": 2.767272710800171 + }, + { + "auxiliary_loss_clip": 0.01182078, + "auxiliary_loss_mlp": 0.01034988, + "balance_loss_clip": 1.05910349, + "balance_loss_mlp": 1.02496839, + "epoch": 0.29976552636325376, + "flos": 16938021870720.0, + "grad_norm": 2.8713270048478927, + "language_loss": 0.86166209, + "learning_rate": 3.284115971179056e-06, + "loss": 0.88383269, + "num_input_tokens_seen": 53452165, + "step": 2493, + "time_per_iteration": 2.750324249267578 + }, + { + "auxiliary_loss_clip": 0.01157516, + "auxiliary_loss_mlp": 0.01037956, + "balance_loss_clip": 1.05345082, + "balance_loss_mlp": 1.02802002, + "epoch": 0.2998857692538929, + "flos": 17056455989760.0, + "grad_norm": 1.7830416732684193, + "language_loss": 0.78761178, + "learning_rate": 3.283518670445859e-06, + "loss": 0.8095665, + "num_input_tokens_seen": 53470075, + "step": 2494, + "time_per_iteration": 2.9475271701812744 + }, + { + "auxiliary_loss_clip": 0.01077104, + "auxiliary_loss_mlp": 0.00877587, + "balance_loss_clip": 1.02325439, + "balance_loss_mlp": 1.00147212, + "epoch": 0.30000601214453193, + "flos": 68831528025600.0, + "grad_norm": 0.684237050670669, + "language_loss": 0.54297876, + "learning_rate": 3.2829211750038995e-06, + "loss": 0.56252575, + "num_input_tokens_seen": 53538705, + "step": 2495, + "time_per_iteration": 3.372258424758911 + }, + { + "auxiliary_loss_clip": 0.01169599, + "auxiliary_loss_mlp": 0.01031029, + "balance_loss_clip": 1.05326056, + "balance_loss_mlp": 1.02055657, + "epoch": 0.30012625503517104, + "flos": 17603267708160.0, + "grad_norm": 1.9432891267769725, + "language_loss": 0.88895237, + "learning_rate": 3.2823234849438183e-06, + "loss": 0.91095859, + "num_input_tokens_seen": 53556740, + "step": 2496, + "time_per_iteration": 2.7606124877929688 + }, + { + "auxiliary_loss_clip": 0.01186144, + "auxiliary_loss_mlp": 0.01030478, + "balance_loss_clip": 1.05586529, + "balance_loss_mlp": 1.02082205, + "epoch": 0.30024649792581015, + "flos": 21252581775360.0, + "grad_norm": 2.0728214406540717, + "language_loss": 0.75801164, + "learning_rate": 3.2817256003562836e-06, + "loss": 0.78017789, + "num_input_tokens_seen": 53577115, + "step": 2497, + "time_per_iteration": 2.7761435508728027 + }, + { + "auxiliary_loss_clip": 0.01155142, + "auxiliary_loss_mlp": 0.01039163, + "balance_loss_clip": 1.04989052, + "balance_loss_mlp": 1.02931011, + "epoch": 0.3003667408164492, + "flos": 23003262748800.0, + "grad_norm": 1.8655838315056235, + "language_loss": 0.65590578, + "learning_rate": 3.281127521331995e-06, + "loss": 0.67784882, + "num_input_tokens_seen": 53598295, + "step": 2498, + "time_per_iteration": 2.8849594593048096 + }, + { + "auxiliary_loss_clip": 0.01093807, + "auxiliary_loss_mlp": 0.01005182, + "balance_loss_clip": 1.02167296, + "balance_loss_mlp": 1.00265491, + "epoch": 0.3004869837070883, + "flos": 64232340750720.0, + "grad_norm": 0.8771859601805, + "language_loss": 0.60735071, + "learning_rate": 3.2805292479616798e-06, + "loss": 0.6283406, + "num_input_tokens_seen": 53657160, + "step": 2499, + "time_per_iteration": 3.130146026611328 + }, + { + "auxiliary_loss_clip": 0.01186184, + "auxiliary_loss_mlp": 0.01034121, + "balance_loss_clip": 1.05566621, + "balance_loss_mlp": 1.02393484, + "epoch": 0.30060722659772743, + "flos": 26248653400320.0, + "grad_norm": 2.6588227688214765, + "language_loss": 0.91638541, + "learning_rate": 3.2799307803360955e-06, + "loss": 0.9385885, + "num_input_tokens_seen": 53673090, + "step": 2500, + "time_per_iteration": 2.829051971435547 + }, + { + "auxiliary_loss_clip": 0.01201728, + "auxiliary_loss_mlp": 0.01038796, + "balance_loss_clip": 1.05896425, + "balance_loss_mlp": 1.02843058, + "epoch": 0.3007274694883665, + "flos": 24970879912320.0, + "grad_norm": 1.4857542075221937, + "language_loss": 0.81440908, + "learning_rate": 3.27933211854603e-06, + "loss": 0.83681428, + "num_input_tokens_seen": 53692145, + "step": 2501, + "time_per_iteration": 2.684290885925293 + }, + { + "auxiliary_loss_clip": 0.01187149, + "auxiliary_loss_mlp": 0.01032531, + "balance_loss_clip": 1.0600369, + "balance_loss_mlp": 1.02189136, + "epoch": 0.3008477123790056, + "flos": 17055845458560.0, + "grad_norm": 3.564724925804034, + "language_loss": 0.87033963, + "learning_rate": 3.278733262682299e-06, + "loss": 0.8925364, + "num_input_tokens_seen": 53710000, + "step": 2502, + "time_per_iteration": 2.744797945022583 + }, + { + "auxiliary_loss_clip": 0.01202292, + "auxiliary_loss_mlp": 0.01033404, + "balance_loss_clip": 1.05770826, + "balance_loss_mlp": 1.02335477, + "epoch": 0.3009679552696447, + "flos": 21506398254720.0, + "grad_norm": 2.071172772181786, + "language_loss": 0.82668924, + "learning_rate": 3.2781342128357484e-06, + "loss": 0.84904623, + "num_input_tokens_seen": 53729355, + "step": 2503, + "time_per_iteration": 2.6276071071624756 + }, + { + "auxiliary_loss_clip": 0.01172699, + "auxiliary_loss_mlp": 0.0103491, + "balance_loss_clip": 1.05343199, + "balance_loss_mlp": 1.02483058, + "epoch": 0.30108819816028376, + "flos": 21134004001920.0, + "grad_norm": 2.328979492988549, + "language_loss": 0.80257648, + "learning_rate": 3.2775349690972547e-06, + "loss": 0.82465255, + "num_input_tokens_seen": 53743505, + "step": 2504, + "time_per_iteration": 2.824666976928711 + }, + { + "auxiliary_loss_clip": 0.01085148, + "auxiliary_loss_mlp": 0.01002522, + "balance_loss_clip": 1.02475548, + "balance_loss_mlp": 1.00017345, + "epoch": 0.30120844105092287, + "flos": 71126434938240.0, + "grad_norm": 0.763239970761932, + "language_loss": 0.51808828, + "learning_rate": 3.276935531557722e-06, + "loss": 0.53896499, + "num_input_tokens_seen": 53808725, + "step": 2505, + "time_per_iteration": 3.3937244415283203 + }, + { + "auxiliary_loss_clip": 0.01165028, + "auxiliary_loss_mlp": 0.01032505, + "balance_loss_clip": 1.05131972, + "balance_loss_mlp": 1.02234864, + "epoch": 0.301328683941562, + "flos": 20264571302400.0, + "grad_norm": 2.061476360300978, + "language_loss": 0.79091394, + "learning_rate": 3.2763359003080837e-06, + "loss": 0.81288922, + "num_input_tokens_seen": 53825680, + "step": 2506, + "time_per_iteration": 2.8560118675231934 + }, + { + "auxiliary_loss_clip": 0.01081059, + "auxiliary_loss_mlp": 0.01002317, + "balance_loss_clip": 1.02137184, + "balance_loss_mlp": 0.99986118, + "epoch": 0.30144892683220104, + "flos": 70648212240000.0, + "grad_norm": 0.8468468755485787, + "language_loss": 0.62500191, + "learning_rate": 3.2757360754393047e-06, + "loss": 0.64583576, + "num_input_tokens_seen": 53889750, + "step": 2507, + "time_per_iteration": 3.4089574813842773 + }, + { + "auxiliary_loss_clip": 0.01191472, + "auxiliary_loss_mlp": 0.010377, + "balance_loss_clip": 1.05727935, + "balance_loss_mlp": 1.02723932, + "epoch": 0.30156916972284015, + "flos": 22820549241600.0, + "grad_norm": 3.2226733922918407, + "language_loss": 0.64258868, + "learning_rate": 3.2751360570423767e-06, + "loss": 0.66488039, + "num_input_tokens_seen": 53908135, + "step": 2508, + "time_per_iteration": 2.740131139755249 + }, + { + "auxiliary_loss_clip": 0.01181925, + "auxiliary_loss_mlp": 0.01037471, + "balance_loss_clip": 1.05608487, + "balance_loss_mlp": 1.02683175, + "epoch": 0.3016894126134792, + "flos": 29899188529920.0, + "grad_norm": 2.03865342309334, + "language_loss": 0.75920725, + "learning_rate": 3.2745358452083236e-06, + "loss": 0.78140128, + "num_input_tokens_seen": 53931035, + "step": 2509, + "time_per_iteration": 2.942697286605835 + }, + { + "auxiliary_loss_clip": 0.01192197, + "auxiliary_loss_mlp": 0.01037347, + "balance_loss_clip": 1.05780029, + "balance_loss_mlp": 1.02827525, + "epoch": 0.3018096555041183, + "flos": 21546331200000.0, + "grad_norm": 1.4774684562880191, + "language_loss": 0.82487017, + "learning_rate": 3.2739354400281955e-06, + "loss": 0.84716558, + "num_input_tokens_seen": 53952255, + "step": 2510, + "time_per_iteration": 2.8180480003356934 + }, + { + "auxiliary_loss_clip": 0.01075343, + "auxiliary_loss_mlp": 0.00877357, + "balance_loss_clip": 1.01887465, + "balance_loss_mlp": 1.00161576, + "epoch": 0.3019298983947574, + "flos": 59136294597120.0, + "grad_norm": 0.8689970989730943, + "language_loss": 0.63688278, + "learning_rate": 3.2733348415930744e-06, + "loss": 0.65640974, + "num_input_tokens_seen": 54014125, + "step": 2511, + "time_per_iteration": 4.320324897766113 + }, + { + "auxiliary_loss_clip": 0.01170128, + "auxiliary_loss_mlp": 0.01041561, + "balance_loss_clip": 1.05532742, + "balance_loss_mlp": 1.03174973, + "epoch": 0.3020501412853965, + "flos": 34423070941440.0, + "grad_norm": 1.8441862513998548, + "language_loss": 0.80510187, + "learning_rate": 3.27273404999407e-06, + "loss": 0.82721877, + "num_input_tokens_seen": 54036345, + "step": 2512, + "time_per_iteration": 4.789566278457642 + }, + { + "auxiliary_loss_clip": 0.01081712, + "auxiliary_loss_mlp": 0.01010037, + "balance_loss_clip": 1.02084041, + "balance_loss_mlp": 1.00746238, + "epoch": 0.3021703841760356, + "flos": 71008288128000.0, + "grad_norm": 0.8128695063190898, + "language_loss": 0.60480285, + "learning_rate": 3.272133065322322e-06, + "loss": 0.62572026, + "num_input_tokens_seen": 54094615, + "step": 2513, + "time_per_iteration": 4.198628187179565 + }, + { + "auxiliary_loss_clip": 0.01200955, + "auxiliary_loss_mlp": 0.01038669, + "balance_loss_clip": 1.05750132, + "balance_loss_mlp": 1.02888179, + "epoch": 0.3022906270666747, + "flos": 21510528318720.0, + "grad_norm": 1.734303039730276, + "language_loss": 0.79437959, + "learning_rate": 3.271531887669e-06, + "loss": 0.8167758, + "num_input_tokens_seen": 54114675, + "step": 2514, + "time_per_iteration": 2.7354822158813477 + }, + { + "auxiliary_loss_clip": 0.01168757, + "auxiliary_loss_mlp": 0.01038869, + "balance_loss_clip": 1.05159426, + "balance_loss_mlp": 1.02805638, + "epoch": 0.30241086995731375, + "flos": 31132001168640.0, + "grad_norm": 2.5846287190336765, + "language_loss": 0.63390863, + "learning_rate": 3.2709305171253015e-06, + "loss": 0.65598488, + "num_input_tokens_seen": 54134795, + "step": 2515, + "time_per_iteration": 2.870568037033081 + }, + { + "auxiliary_loss_clip": 0.01192996, + "auxiliary_loss_mlp": 0.01035667, + "balance_loss_clip": 1.06009626, + "balance_loss_mlp": 1.02558827, + "epoch": 0.30253111284795287, + "flos": 23511542152320.0, + "grad_norm": 2.034801740366613, + "language_loss": 0.78024995, + "learning_rate": 3.2703289537824536e-06, + "loss": 0.80253655, + "num_input_tokens_seen": 54154595, + "step": 2516, + "time_per_iteration": 2.7289304733276367 + }, + { + "auxiliary_loss_clip": 0.01168663, + "auxiliary_loss_mlp": 0.0104274, + "balance_loss_clip": 1.05534673, + "balance_loss_mlp": 1.0313499, + "epoch": 0.302651355738592, + "flos": 18725367651840.0, + "grad_norm": 2.7032391920812198, + "language_loss": 0.78403556, + "learning_rate": 3.269727197731714e-06, + "loss": 0.8061496, + "num_input_tokens_seen": 54167360, + "step": 2517, + "time_per_iteration": 2.7612862586975098 + }, + { + "auxiliary_loss_clip": 0.01156766, + "auxiliary_loss_mlp": 0.01035914, + "balance_loss_clip": 1.05247521, + "balance_loss_mlp": 1.02601361, + "epoch": 0.30277159862923103, + "flos": 22418888382720.0, + "grad_norm": 1.6855920026549178, + "language_loss": 0.77964878, + "learning_rate": 3.269125249064367e-06, + "loss": 0.80157554, + "num_input_tokens_seen": 54187055, + "step": 2518, + "time_per_iteration": 2.8757119178771973 + }, + { + "auxiliary_loss_clip": 0.0120389, + "auxiliary_loss_mlp": 0.01035408, + "balance_loss_clip": 1.05802417, + "balance_loss_mlp": 1.02455449, + "epoch": 0.30289184151987014, + "flos": 22273126992000.0, + "grad_norm": 1.9016875703446479, + "language_loss": 0.83010244, + "learning_rate": 3.2685231078717297e-06, + "loss": 0.85249549, + "num_input_tokens_seen": 54207245, + "step": 2519, + "time_per_iteration": 2.7372074127197266 + }, + { + "auxiliary_loss_clip": 0.01165474, + "auxiliary_loss_mlp": 0.00888765, + "balance_loss_clip": 1.05587912, + "balance_loss_mlp": 1.00113535, + "epoch": 0.30301208441050925, + "flos": 25225594231680.0, + "grad_norm": 1.8973465306767991, + "language_loss": 0.75246465, + "learning_rate": 3.267920774245145e-06, + "loss": 0.77300709, + "num_input_tokens_seen": 54226650, + "step": 2520, + "time_per_iteration": 2.8015384674072266 + }, + { + "auxiliary_loss_clip": 0.01195043, + "auxiliary_loss_mlp": 0.01034235, + "balance_loss_clip": 1.05809069, + "balance_loss_mlp": 1.02361381, + "epoch": 0.3031323273011483, + "flos": 23039245198080.0, + "grad_norm": 2.433825844283437, + "language_loss": 0.84524179, + "learning_rate": 3.2673182482759876e-06, + "loss": 0.86753458, + "num_input_tokens_seen": 54245765, + "step": 2521, + "time_per_iteration": 2.707892894744873 + }, + { + "auxiliary_loss_clip": 0.01191505, + "auxiliary_loss_mlp": 0.01035779, + "balance_loss_clip": 1.05755341, + "balance_loss_mlp": 1.02580142, + "epoch": 0.3032525701917874, + "flos": 18876695650560.0, + "grad_norm": 1.9824350225380067, + "language_loss": 0.65985537, + "learning_rate": 3.266715530055659e-06, + "loss": 0.68212819, + "num_input_tokens_seen": 54263915, + "step": 2522, + "time_per_iteration": 2.7522506713867188 + }, + { + "auxiliary_loss_clip": 0.01180579, + "auxiliary_loss_mlp": 0.01031188, + "balance_loss_clip": 1.05207253, + "balance_loss_mlp": 1.02151465, + "epoch": 0.30337281308242653, + "flos": 17782641250560.0, + "grad_norm": 1.6183120815298129, + "language_loss": 0.80467653, + "learning_rate": 3.2661126196755927e-06, + "loss": 0.82679415, + "num_input_tokens_seen": 54283025, + "step": 2523, + "time_per_iteration": 2.7148492336273193 + }, + { + "auxiliary_loss_clip": 0.01089401, + "auxiliary_loss_mlp": 0.01007195, + "balance_loss_clip": 1.01745903, + "balance_loss_mlp": 1.00464392, + "epoch": 0.3034930559730656, + "flos": 57824298426240.0, + "grad_norm": 0.778523341313711, + "language_loss": 0.55977452, + "learning_rate": 3.265509517227248e-06, + "loss": 0.58074045, + "num_input_tokens_seen": 54339840, + "step": 2524, + "time_per_iteration": 3.2235026359558105 + }, + { + "auxiliary_loss_clip": 0.0118154, + "auxiliary_loss_mlp": 0.01035366, + "balance_loss_clip": 1.05303383, + "balance_loss_mlp": 1.02447581, + "epoch": 0.3036132988637047, + "flos": 14755587419520.0, + "grad_norm": 1.6721403309193288, + "language_loss": 0.81172895, + "learning_rate": 3.264906222802115e-06, + "loss": 0.83389801, + "num_input_tokens_seen": 54357690, + "step": 2525, + "time_per_iteration": 2.8870656490325928 + }, + { + "auxiliary_loss_clip": 0.01203989, + "auxiliary_loss_mlp": 0.01033464, + "balance_loss_clip": 1.05893636, + "balance_loss_mlp": 1.02343225, + "epoch": 0.30373354175434375, + "flos": 21033203460480.0, + "grad_norm": 2.169683403320508, + "language_loss": 0.78224701, + "learning_rate": 3.264302736491715e-06, + "loss": 0.80462158, + "num_input_tokens_seen": 54377810, + "step": 2526, + "time_per_iteration": 2.73519229888916 + }, + { + "auxiliary_loss_clip": 0.01188217, + "auxiliary_loss_mlp": 0.01031577, + "balance_loss_clip": 1.05919766, + "balance_loss_mlp": 1.02225518, + "epoch": 0.30385378464498286, + "flos": 21143233797120.0, + "grad_norm": 1.8660848574858602, + "language_loss": 0.87378001, + "learning_rate": 3.263699058387594e-06, + "loss": 0.89597797, + "num_input_tokens_seen": 54395245, + "step": 2527, + "time_per_iteration": 2.669107437133789 + }, + { + "auxiliary_loss_clip": 0.01166814, + "auxiliary_loss_mlp": 0.01029024, + "balance_loss_clip": 1.0486412, + "balance_loss_mlp": 1.02008331, + "epoch": 0.30397402753562197, + "flos": 20629244131200.0, + "grad_norm": 2.091447807532189, + "language_loss": 0.90350205, + "learning_rate": 3.2630951885813315e-06, + "loss": 0.9254604, + "num_input_tokens_seen": 54412640, + "step": 2528, + "time_per_iteration": 2.7258832454681396 + }, + { + "auxiliary_loss_clip": 0.01181941, + "auxiliary_loss_mlp": 0.01028411, + "balance_loss_clip": 1.05206037, + "balance_loss_mlp": 1.01861858, + "epoch": 0.304094270426261, + "flos": 15085678429440.0, + "grad_norm": 2.1186599511815456, + "language_loss": 0.78287333, + "learning_rate": 3.262491127164533e-06, + "loss": 0.80497688, + "num_input_tokens_seen": 54431455, + "step": 2529, + "time_per_iteration": 2.708787202835083 + }, + { + "auxiliary_loss_clip": 0.01189039, + "auxiliary_loss_mlp": 0.0088838, + "balance_loss_clip": 1.05510187, + "balance_loss_mlp": 1.00111282, + "epoch": 0.30421451331690014, + "flos": 13845216193920.0, + "grad_norm": 2.6867051634634698, + "language_loss": 0.80151498, + "learning_rate": 3.2618868742288337e-06, + "loss": 0.82228923, + "num_input_tokens_seen": 54448380, + "step": 2530, + "time_per_iteration": 2.696791648864746 + }, + { + "auxiliary_loss_clip": 0.01190841, + "auxiliary_loss_mlp": 0.01031287, + "balance_loss_clip": 1.05609047, + "balance_loss_mlp": 1.02160764, + "epoch": 0.30433475620753925, + "flos": 17384212615680.0, + "grad_norm": 1.907027975182721, + "language_loss": 0.72178876, + "learning_rate": 3.261282429865899e-06, + "loss": 0.74400997, + "num_input_tokens_seen": 54466385, + "step": 2531, + "time_per_iteration": 2.8145558834075928 + }, + { + "auxiliary_loss_clip": 0.01187696, + "auxiliary_loss_mlp": 0.00886942, + "balance_loss_clip": 1.05670965, + "balance_loss_mlp": 1.00109291, + "epoch": 0.3044549990981783, + "flos": 18916951818240.0, + "grad_norm": 1.6248968498698284, + "language_loss": 0.72332704, + "learning_rate": 3.2606777941674225e-06, + "loss": 0.74407345, + "num_input_tokens_seen": 54485040, + "step": 2532, + "time_per_iteration": 2.7708749771118164 + }, + { + "auxiliary_loss_clip": 0.01156764, + "auxiliary_loss_mlp": 0.01034793, + "balance_loss_clip": 1.0522902, + "balance_loss_mlp": 1.02556634, + "epoch": 0.3045752419888174, + "flos": 21068431724160.0, + "grad_norm": 2.3664304708947963, + "language_loss": 0.8447479, + "learning_rate": 3.2600729672251276e-06, + "loss": 0.86666346, + "num_input_tokens_seen": 54502755, + "step": 2533, + "time_per_iteration": 2.8044331073760986 + }, + { + "auxiliary_loss_clip": 0.01204089, + "auxiliary_loss_mlp": 0.00887451, + "balance_loss_clip": 1.06158876, + "balance_loss_mlp": 1.00096476, + "epoch": 0.3046954848794565, + "flos": 29096405516160.0, + "grad_norm": 1.869338958047633, + "language_loss": 0.65463495, + "learning_rate": 3.259467949130765e-06, + "loss": 0.67555034, + "num_input_tokens_seen": 54524165, + "step": 2534, + "time_per_iteration": 2.7773241996765137 + }, + { + "auxiliary_loss_clip": 0.01182755, + "auxiliary_loss_mlp": 0.01031611, + "balance_loss_clip": 1.05560398, + "balance_loss_mlp": 1.02219903, + "epoch": 0.3048157277700956, + "flos": 20295346279680.0, + "grad_norm": 2.687654503402954, + "language_loss": 0.82239401, + "learning_rate": 3.2588627399761164e-06, + "loss": 0.84453773, + "num_input_tokens_seen": 54540160, + "step": 2535, + "time_per_iteration": 2.6870157718658447 + }, + { + "auxiliary_loss_clip": 0.01182611, + "auxiliary_loss_mlp": 0.01034642, + "balance_loss_clip": 1.05723274, + "balance_loss_mlp": 1.02463436, + "epoch": 0.3049359706607347, + "flos": 22739929165440.0, + "grad_norm": 1.7100708193054666, + "language_loss": 0.70828617, + "learning_rate": 3.2582573398529903e-06, + "loss": 0.73045862, + "num_input_tokens_seen": 54557515, + "step": 2536, + "time_per_iteration": 2.7856595516204834 + }, + { + "auxiliary_loss_clip": 0.01172354, + "auxiliary_loss_mlp": 0.01040618, + "balance_loss_clip": 1.05165267, + "balance_loss_mlp": 1.03036582, + "epoch": 0.3050562135513738, + "flos": 18434634969600.0, + "grad_norm": 2.701976503397064, + "language_loss": 0.74733937, + "learning_rate": 3.2576517488532265e-06, + "loss": 0.76946914, + "num_input_tokens_seen": 54573865, + "step": 2537, + "time_per_iteration": 3.7820990085601807 + }, + { + "auxiliary_loss_clip": 0.01190152, + "auxiliary_loss_mlp": 0.01035019, + "balance_loss_clip": 1.05487585, + "balance_loss_mlp": 1.02515471, + "epoch": 0.30517645644201286, + "flos": 20370327920640.0, + "grad_norm": 2.0810921012021404, + "language_loss": 0.87337971, + "learning_rate": 3.257045967068692e-06, + "loss": 0.89563143, + "num_input_tokens_seen": 54593120, + "step": 2538, + "time_per_iteration": 3.6413540840148926 + }, + { + "auxiliary_loss_clip": 0.01202211, + "auxiliary_loss_mlp": 0.01032011, + "balance_loss_clip": 1.05748069, + "balance_loss_mlp": 1.02270114, + "epoch": 0.30529669933265197, + "flos": 21945118970880.0, + "grad_norm": 1.90287269671311, + "language_loss": 0.82165062, + "learning_rate": 3.2564399945912848e-06, + "loss": 0.84399283, + "num_input_tokens_seen": 54612910, + "step": 2539, + "time_per_iteration": 3.6491756439208984 + }, + { + "auxiliary_loss_clip": 0.01166772, + "auxiliary_loss_mlp": 0.01034438, + "balance_loss_clip": 1.05144024, + "balance_loss_mlp": 1.0252527, + "epoch": 0.305416942223291, + "flos": 21835411856640.0, + "grad_norm": 2.5487604254324046, + "language_loss": 0.82241714, + "learning_rate": 3.2558338315129287e-06, + "loss": 0.84442925, + "num_input_tokens_seen": 54631055, + "step": 2540, + "time_per_iteration": 2.9285855293273926 + }, + { + "auxiliary_loss_clip": 0.01185946, + "auxiliary_loss_mlp": 0.01036669, + "balance_loss_clip": 1.05634224, + "balance_loss_mlp": 1.02689362, + "epoch": 0.30553718511393013, + "flos": 33911810709120.0, + "grad_norm": 3.0028069458770763, + "language_loss": 0.75694191, + "learning_rate": 3.2552274779255785e-06, + "loss": 0.77916807, + "num_input_tokens_seen": 54651985, + "step": 2541, + "time_per_iteration": 2.93129301071167 + }, + { + "auxiliary_loss_clip": 0.01189648, + "auxiliary_loss_mlp": 0.01041947, + "balance_loss_clip": 1.05614996, + "balance_loss_mlp": 1.03291678, + "epoch": 0.30565742800456924, + "flos": 22268530051200.0, + "grad_norm": 2.091200796364848, + "language_loss": 0.77061766, + "learning_rate": 3.2546209339212184e-06, + "loss": 0.79293358, + "num_input_tokens_seen": 54671005, + "step": 2542, + "time_per_iteration": 2.8490700721740723 + }, + { + "auxiliary_loss_clip": 0.01182281, + "auxiliary_loss_mlp": 0.01037567, + "balance_loss_clip": 1.05337775, + "balance_loss_mlp": 1.02773869, + "epoch": 0.3057776708952083, + "flos": 22565044823040.0, + "grad_norm": 1.5568716472089614, + "language_loss": 0.775419, + "learning_rate": 3.25401419959186e-06, + "loss": 0.7976175, + "num_input_tokens_seen": 54691615, + "step": 2543, + "time_per_iteration": 2.748659372329712 + }, + { + "auxiliary_loss_clip": 0.01194713, + "auxiliary_loss_mlp": 0.01030492, + "balance_loss_clip": 1.06024432, + "balance_loss_mlp": 1.02057981, + "epoch": 0.3058979137858474, + "flos": 21799213925760.0, + "grad_norm": 2.3482061204545355, + "language_loss": 0.76539099, + "learning_rate": 3.253407275029545e-06, + "loss": 0.78764307, + "num_input_tokens_seen": 54710520, + "step": 2544, + "time_per_iteration": 2.8073925971984863 + }, + { + "auxiliary_loss_clip": 0.01176469, + "auxiliary_loss_mlp": 0.01034913, + "balance_loss_clip": 1.05523872, + "balance_loss_mlp": 1.0246191, + "epoch": 0.3060181566764865, + "flos": 26979435601920.0, + "grad_norm": 1.741041113139269, + "language_loss": 0.80000472, + "learning_rate": 3.2528001603263425e-06, + "loss": 0.82211852, + "num_input_tokens_seen": 54732590, + "step": 2545, + "time_per_iteration": 2.8522984981536865 + }, + { + "auxiliary_loss_clip": 0.01190768, + "auxiliary_loss_mlp": 0.01037689, + "balance_loss_clip": 1.0583353, + "balance_loss_mlp": 1.02766347, + "epoch": 0.3061383995671256, + "flos": 19865101173120.0, + "grad_norm": 2.3487320062369226, + "language_loss": 0.81377411, + "learning_rate": 3.2521928555743514e-06, + "loss": 0.83605868, + "num_input_tokens_seen": 54749935, + "step": 2546, + "time_per_iteration": 2.712864398956299 + }, + { + "auxiliary_loss_clip": 0.0117626, + "auxiliary_loss_mlp": 0.00888038, + "balance_loss_clip": 1.05225313, + "balance_loss_mlp": 1.00084949, + "epoch": 0.3062586424577647, + "flos": 22127509255680.0, + "grad_norm": 1.931365567952416, + "language_loss": 0.67572552, + "learning_rate": 3.2515853608657e-06, + "loss": 0.69636846, + "num_input_tokens_seen": 54767935, + "step": 2547, + "time_per_iteration": 2.787050485610962 + }, + { + "auxiliary_loss_clip": 0.01188213, + "auxiliary_loss_mlp": 0.01037196, + "balance_loss_clip": 1.05647016, + "balance_loss_mlp": 1.02628803, + "epoch": 0.3063788853484038, + "flos": 20845497962880.0, + "grad_norm": 2.3035802356051014, + "language_loss": 0.74321246, + "learning_rate": 3.250977676292545e-06, + "loss": 0.76546645, + "num_input_tokens_seen": 54786175, + "step": 2548, + "time_per_iteration": 2.8499956130981445 + }, + { + "auxiliary_loss_clip": 0.01185677, + "auxiliary_loss_mlp": 0.01030374, + "balance_loss_clip": 1.05632472, + "balance_loss_mlp": 1.02089071, + "epoch": 0.30649912823904285, + "flos": 16209717707520.0, + "grad_norm": 2.3968298430955755, + "language_loss": 0.7940349, + "learning_rate": 3.2503698019470712e-06, + "loss": 0.81619543, + "num_input_tokens_seen": 54801945, + "step": 2549, + "time_per_iteration": 2.858934164047241 + }, + { + "auxiliary_loss_clip": 0.01190216, + "auxiliary_loss_mlp": 0.01039704, + "balance_loss_clip": 1.05409992, + "balance_loss_mlp": 1.02883244, + "epoch": 0.30661937112968196, + "flos": 18617815353600.0, + "grad_norm": 2.0671614827732596, + "language_loss": 0.78056484, + "learning_rate": 3.249761737921492e-06, + "loss": 0.80286413, + "num_input_tokens_seen": 54818475, + "step": 2550, + "time_per_iteration": 2.8098392486572266 + }, + { + "auxiliary_loss_clip": 0.01180771, + "auxiliary_loss_mlp": 0.01037614, + "balance_loss_clip": 1.05663133, + "balance_loss_mlp": 1.02782083, + "epoch": 0.30673961402032107, + "flos": 31390809638400.0, + "grad_norm": 2.075602810559226, + "language_loss": 0.74143779, + "learning_rate": 3.249153484308051e-06, + "loss": 0.76362169, + "num_input_tokens_seen": 54837090, + "step": 2551, + "time_per_iteration": 2.8878912925720215 + }, + { + "auxiliary_loss_clip": 0.011509, + "auxiliary_loss_mlp": 0.01035163, + "balance_loss_clip": 1.05036521, + "balance_loss_mlp": 1.02407622, + "epoch": 0.3068598569109601, + "flos": 20229809915520.0, + "grad_norm": 3.4122494118962443, + "language_loss": 0.77801764, + "learning_rate": 3.2485450411990194e-06, + "loss": 0.79987824, + "num_input_tokens_seen": 54856445, + "step": 2552, + "time_per_iteration": 2.9699409008026123 + }, + { + "auxiliary_loss_clip": 0.0120315, + "auxiliary_loss_mlp": 0.01035311, + "balance_loss_clip": 1.05790615, + "balance_loss_mlp": 1.02492762, + "epoch": 0.30698009980159924, + "flos": 29601991399680.0, + "grad_norm": 2.1210705284718707, + "language_loss": 0.82052499, + "learning_rate": 3.2479364086866983e-06, + "loss": 0.84290957, + "num_input_tokens_seen": 54876700, + "step": 2553, + "time_per_iteration": 2.831045389175415 + }, + { + "auxiliary_loss_clip": 0.01180693, + "auxiliary_loss_mlp": 0.00887624, + "balance_loss_clip": 1.05532932, + "balance_loss_mlp": 1.00081074, + "epoch": 0.30710034269223835, + "flos": 23842423261440.0, + "grad_norm": 1.7561505759650275, + "language_loss": 0.8128311, + "learning_rate": 3.247327586863416e-06, + "loss": 0.83351433, + "num_input_tokens_seen": 54897580, + "step": 2554, + "time_per_iteration": 2.7845051288604736 + }, + { + "auxiliary_loss_clip": 0.01177827, + "auxiliary_loss_mlp": 0.01033057, + "balance_loss_clip": 1.05408406, + "balance_loss_mlp": 1.02235818, + "epoch": 0.3072205855828774, + "flos": 25884986152320.0, + "grad_norm": 2.5065966378778133, + "language_loss": 0.76625568, + "learning_rate": 3.2467185758215304e-06, + "loss": 0.78836453, + "num_input_tokens_seen": 54917320, + "step": 2555, + "time_per_iteration": 2.962054967880249 + }, + { + "auxiliary_loss_clip": 0.01176393, + "auxiliary_loss_mlp": 0.00887427, + "balance_loss_clip": 1.0570488, + "balance_loss_mlp": 1.00082183, + "epoch": 0.3073408284735165, + "flos": 22236390357120.0, + "grad_norm": 2.521210314186108, + "language_loss": 0.85819119, + "learning_rate": 3.246109375653428e-06, + "loss": 0.87882942, + "num_input_tokens_seen": 54934085, + "step": 2556, + "time_per_iteration": 2.9635403156280518 + }, + { + "auxiliary_loss_clip": 0.01201466, + "auxiliary_loss_mlp": 0.01036363, + "balance_loss_clip": 1.05747485, + "balance_loss_mlp": 1.02600956, + "epoch": 0.30746107136415557, + "flos": 19500284689920.0, + "grad_norm": 2.625769814195685, + "language_loss": 0.78318202, + "learning_rate": 3.2454999864515243e-06, + "loss": 0.80556035, + "num_input_tokens_seen": 54953460, + "step": 2557, + "time_per_iteration": 2.6702070236206055 + }, + { + "auxiliary_loss_clip": 0.01178385, + "auxiliary_loss_mlp": 0.00887983, + "balance_loss_clip": 1.0536474, + "balance_loss_mlp": 1.00071692, + "epoch": 0.3075813142547947, + "flos": 21724806902400.0, + "grad_norm": 1.695222506554329, + "language_loss": 0.69480491, + "learning_rate": 3.244890408308263e-06, + "loss": 0.71546853, + "num_input_tokens_seen": 54974165, + "step": 2558, + "time_per_iteration": 2.8951833248138428 + }, + { + "auxiliary_loss_clip": 0.01161637, + "auxiliary_loss_mlp": 0.01030102, + "balance_loss_clip": 1.0511167, + "balance_loss_mlp": 1.02074373, + "epoch": 0.3077015571454338, + "flos": 24097963593600.0, + "grad_norm": 2.09855237900295, + "language_loss": 0.61166006, + "learning_rate": 3.2442806413161165e-06, + "loss": 0.63357747, + "num_input_tokens_seen": 54993810, + "step": 2559, + "time_per_iteration": 2.9220244884490967 + }, + { + "auxiliary_loss_clip": 0.0116606, + "auxiliary_loss_mlp": 0.01038702, + "balance_loss_clip": 1.05225027, + "balance_loss_mlp": 1.02878428, + "epoch": 0.30782180003607285, + "flos": 18405476104320.0, + "grad_norm": 2.100655278907413, + "language_loss": 0.75923169, + "learning_rate": 3.243670685567586e-06, + "loss": 0.78127933, + "num_input_tokens_seen": 55011210, + "step": 2560, + "time_per_iteration": 2.9458258152008057 + }, + { + "auxiliary_loss_clip": 0.01182812, + "auxiliary_loss_mlp": 0.00887604, + "balance_loss_clip": 1.05616558, + "balance_loss_mlp": 1.00076699, + "epoch": 0.30794204292671196, + "flos": 23878549365120.0, + "grad_norm": 2.323969935291671, + "language_loss": 0.79763758, + "learning_rate": 3.2430605411552012e-06, + "loss": 0.81834173, + "num_input_tokens_seen": 55031325, + "step": 2561, + "time_per_iteration": 2.803269624710083 + }, + { + "auxiliary_loss_clip": 0.0107611, + "auxiliary_loss_mlp": 0.01002594, + "balance_loss_clip": 1.01843357, + "balance_loss_mlp": 1.00003052, + "epoch": 0.30806228581735107, + "flos": 67927800816000.0, + "grad_norm": 0.8889118264924284, + "language_loss": 0.70652223, + "learning_rate": 3.2424502081715205e-06, + "loss": 0.72730923, + "num_input_tokens_seen": 55094440, + "step": 2562, + "time_per_iteration": 3.399617910385132 + }, + { + "auxiliary_loss_clip": 0.01183052, + "auxiliary_loss_mlp": 0.01033115, + "balance_loss_clip": 1.05517328, + "balance_loss_mlp": 1.02266669, + "epoch": 0.3081825287079901, + "flos": 23843213360640.0, + "grad_norm": 1.747661709264974, + "language_loss": 0.77978915, + "learning_rate": 3.241839686709132e-06, + "loss": 0.80195081, + "num_input_tokens_seen": 55115375, + "step": 2563, + "time_per_iteration": 4.677973747253418 + }, + { + "auxiliary_loss_clip": 0.01187048, + "auxiliary_loss_mlp": 0.01034806, + "balance_loss_clip": 1.05166566, + "balance_loss_mlp": 1.02423239, + "epoch": 0.30830277159862923, + "flos": 16209969102720.0, + "grad_norm": 2.4207186876762092, + "language_loss": 0.82461083, + "learning_rate": 3.2412289768606495e-06, + "loss": 0.8468293, + "num_input_tokens_seen": 55131945, + "step": 2564, + "time_per_iteration": 2.8817498683929443 + }, + { + "auxiliary_loss_clip": 0.01194825, + "auxiliary_loss_mlp": 0.01035998, + "balance_loss_clip": 1.05744469, + "balance_loss_mlp": 1.02598417, + "epoch": 0.30842301448926834, + "flos": 29349503723520.0, + "grad_norm": 1.6901749303857003, + "language_loss": 0.82325041, + "learning_rate": 3.240618078718718e-06, + "loss": 0.84555864, + "num_input_tokens_seen": 55153405, + "step": 2565, + "time_per_iteration": 4.644137382507324 + }, + { + "auxiliary_loss_clip": 0.01166087, + "auxiliary_loss_mlp": 0.01040439, + "balance_loss_clip": 1.04879785, + "balance_loss_mlp": 1.02969217, + "epoch": 0.3085432573799074, + "flos": 21945190798080.0, + "grad_norm": 1.8773142575673927, + "language_loss": 0.74295092, + "learning_rate": 3.240006992376011e-06, + "loss": 0.7650162, + "num_input_tokens_seen": 55173030, + "step": 2566, + "time_per_iteration": 2.8297364711761475 + }, + { + "auxiliary_loss_clip": 0.01189098, + "auxiliary_loss_mlp": 0.01033225, + "balance_loss_clip": 1.05798125, + "balance_loss_mlp": 1.02427197, + "epoch": 0.3086635002705465, + "flos": 22054718344320.0, + "grad_norm": 2.2747031503032753, + "language_loss": 0.76126206, + "learning_rate": 3.2393957179252284e-06, + "loss": 0.78348529, + "num_input_tokens_seen": 55189565, + "step": 2567, + "time_per_iteration": 2.7708613872528076 + }, + { + "auxiliary_loss_clip": 0.01203737, + "auxiliary_loss_mlp": 0.0103095, + "balance_loss_clip": 1.05988264, + "balance_loss_mlp": 1.02086484, + "epoch": 0.3087837431611856, + "flos": 32665925520000.0, + "grad_norm": 1.8528437231411334, + "language_loss": 0.80807096, + "learning_rate": 3.2387842554591016e-06, + "loss": 0.83041781, + "num_input_tokens_seen": 55210380, + "step": 2568, + "time_per_iteration": 2.7396678924560547 + }, + { + "auxiliary_loss_clip": 0.01206371, + "auxiliary_loss_mlp": 0.01036307, + "balance_loss_clip": 1.06147707, + "balance_loss_mlp": 1.02601373, + "epoch": 0.3089039860518247, + "flos": 17599245384960.0, + "grad_norm": 2.5932820770687885, + "language_loss": 0.87840164, + "learning_rate": 3.238172605070388e-06, + "loss": 0.90082848, + "num_input_tokens_seen": 55225795, + "step": 2569, + "time_per_iteration": 2.6814208030700684 + }, + { + "auxiliary_loss_clip": 0.01190655, + "auxiliary_loss_mlp": 0.00888628, + "balance_loss_clip": 1.05642176, + "balance_loss_mlp": 1.00093031, + "epoch": 0.3090242289424638, + "flos": 14383839611520.0, + "grad_norm": 2.2166721980164175, + "language_loss": 0.79005611, + "learning_rate": 3.2375607668518745e-06, + "loss": 0.81084895, + "num_input_tokens_seen": 55238830, + "step": 2570, + "time_per_iteration": 2.7882585525512695 + }, + { + "auxiliary_loss_clip": 0.01172164, + "auxiliary_loss_mlp": 0.01040655, + "balance_loss_clip": 1.05329037, + "balance_loss_mlp": 1.0304389, + "epoch": 0.30914447183310284, + "flos": 16068625084800.0, + "grad_norm": 2.096846289644233, + "language_loss": 0.89803827, + "learning_rate": 3.236948740896377e-06, + "loss": 0.92016643, + "num_input_tokens_seen": 55253630, + "step": 2571, + "time_per_iteration": 2.811228036880493 + }, + { + "auxiliary_loss_clip": 0.01193396, + "auxiliary_loss_mlp": 0.01035949, + "balance_loss_clip": 1.05812395, + "balance_loss_mlp": 1.02533305, + "epoch": 0.30926471472374195, + "flos": 32230221546240.0, + "grad_norm": 1.6881112729241834, + "language_loss": 0.84257388, + "learning_rate": 3.2363365272967384e-06, + "loss": 0.86486739, + "num_input_tokens_seen": 55276200, + "step": 2572, + "time_per_iteration": 2.8890931606292725 + }, + { + "auxiliary_loss_clip": 0.01191001, + "auxiliary_loss_mlp": 0.01039113, + "balance_loss_clip": 1.05946863, + "balance_loss_mlp": 1.0290513, + "epoch": 0.30938495761438106, + "flos": 20370722970240.0, + "grad_norm": 2.0903880187237274, + "language_loss": 0.81417435, + "learning_rate": 3.235724126145832e-06, + "loss": 0.83647549, + "num_input_tokens_seen": 55292235, + "step": 2573, + "time_per_iteration": 2.774661064147949 + }, + { + "auxiliary_loss_clip": 0.0118078, + "auxiliary_loss_mlp": 0.0103193, + "balance_loss_clip": 1.05216098, + "balance_loss_mlp": 1.02230382, + "epoch": 0.3095052005050201, + "flos": 24061155131520.0, + "grad_norm": 1.4644280341855906, + "language_loss": 0.77634948, + "learning_rate": 3.235111537536558e-06, + "loss": 0.79847658, + "num_input_tokens_seen": 55313050, + "step": 2574, + "time_per_iteration": 2.903165340423584 + }, + { + "auxiliary_loss_clip": 0.01191692, + "auxiliary_loss_mlp": 0.01035037, + "balance_loss_clip": 1.0564816, + "balance_loss_mlp": 1.0252912, + "epoch": 0.30962544339565923, + "flos": 23401547729280.0, + "grad_norm": 1.8111773214977482, + "language_loss": 0.82735509, + "learning_rate": 3.2344987615618456e-06, + "loss": 0.84962237, + "num_input_tokens_seen": 55332885, + "step": 2575, + "time_per_iteration": 2.8125784397125244 + }, + { + "auxiliary_loss_clip": 0.01171808, + "auxiliary_loss_mlp": 0.01025701, + "balance_loss_clip": 1.05635643, + "balance_loss_mlp": 1.01610518, + "epoch": 0.30974568628629834, + "flos": 33799984692480.0, + "grad_norm": 1.7454756281575532, + "language_loss": 0.78518742, + "learning_rate": 3.2338857983146533e-06, + "loss": 0.80716252, + "num_input_tokens_seen": 55354385, + "step": 2576, + "time_per_iteration": 2.869326114654541 + }, + { + "auxiliary_loss_clip": 0.01174411, + "auxiliary_loss_mlp": 0.01033532, + "balance_loss_clip": 1.05519319, + "balance_loss_mlp": 1.0240972, + "epoch": 0.3098659291769374, + "flos": 20229594433920.0, + "grad_norm": 1.7868564105078917, + "language_loss": 0.7636106, + "learning_rate": 3.233272647887966e-06, + "loss": 0.78569013, + "num_input_tokens_seen": 55373275, + "step": 2577, + "time_per_iteration": 2.7486209869384766 + }, + { + "auxiliary_loss_clip": 0.0120412, + "auxiliary_loss_mlp": 0.01035375, + "balance_loss_clip": 1.05967212, + "balance_loss_mlp": 1.02475381, + "epoch": 0.3099861720675765, + "flos": 24748556682240.0, + "grad_norm": 1.5888025148447469, + "language_loss": 0.90165186, + "learning_rate": 3.2326593103747985e-06, + "loss": 0.92404675, + "num_input_tokens_seen": 55392290, + "step": 2578, + "time_per_iteration": 2.82242751121521 + }, + { + "auxiliary_loss_clip": 0.01192463, + "auxiliary_loss_mlp": 0.01033247, + "balance_loss_clip": 1.05974531, + "balance_loss_mlp": 1.02340591, + "epoch": 0.3101064149582156, + "flos": 11765485704960.0, + "grad_norm": 1.9022659828185229, + "language_loss": 0.84850204, + "learning_rate": 3.2320457858681936e-06, + "loss": 0.87075919, + "num_input_tokens_seen": 55410680, + "step": 2579, + "time_per_iteration": 2.714792251586914 + }, + { + "auxiliary_loss_clip": 0.01178246, + "auxiliary_loss_mlp": 0.01035852, + "balance_loss_clip": 1.05260563, + "balance_loss_mlp": 1.02622592, + "epoch": 0.31022665784885467, + "flos": 23033247626880.0, + "grad_norm": 2.3037684343538967, + "language_loss": 0.84632969, + "learning_rate": 3.2314320744612228e-06, + "loss": 0.86847067, + "num_input_tokens_seen": 55425980, + "step": 2580, + "time_per_iteration": 2.7110588550567627 + }, + { + "auxiliary_loss_clip": 0.01189248, + "auxiliary_loss_mlp": 0.01033181, + "balance_loss_clip": 1.05621707, + "balance_loss_mlp": 1.02366781, + "epoch": 0.3103469007394938, + "flos": 16289188548480.0, + "grad_norm": 1.6496853443673571, + "language_loss": 0.76643902, + "learning_rate": 3.2308181762469854e-06, + "loss": 0.78866333, + "num_input_tokens_seen": 55443925, + "step": 2581, + "time_per_iteration": 2.802579641342163 + }, + { + "auxiliary_loss_clip": 0.01204886, + "auxiliary_loss_mlp": 0.01038457, + "balance_loss_clip": 1.05843115, + "balance_loss_mlp": 1.02845514, + "epoch": 0.3104671436301329, + "flos": 30515271626880.0, + "grad_norm": 2.446491729797657, + "language_loss": 0.78135252, + "learning_rate": 3.230204091318609e-06, + "loss": 0.80378598, + "num_input_tokens_seen": 55464465, + "step": 2582, + "time_per_iteration": 2.722780227661133 + }, + { + "auxiliary_loss_clip": 0.0119877, + "auxiliary_loss_mlp": 0.00887552, + "balance_loss_clip": 1.05641675, + "balance_loss_mlp": 1.0007782, + "epoch": 0.31058738652077195, + "flos": 20047240062720.0, + "grad_norm": 1.741104216849308, + "language_loss": 0.84297502, + "learning_rate": 3.2295898197692503e-06, + "loss": 0.86383826, + "num_input_tokens_seen": 55483425, + "step": 2583, + "time_per_iteration": 2.712407350540161 + }, + { + "auxiliary_loss_clip": 0.01200474, + "auxiliary_loss_mlp": 0.01032647, + "balance_loss_clip": 1.05795169, + "balance_loss_mlp": 1.0232172, + "epoch": 0.31070762941141106, + "flos": 28074639237120.0, + "grad_norm": 1.609175893984434, + "language_loss": 0.78694707, + "learning_rate": 3.228975361692094e-06, + "loss": 0.80927825, + "num_input_tokens_seen": 55504445, + "step": 2584, + "time_per_iteration": 2.661468744277954 + }, + { + "auxiliary_loss_clip": 0.01194059, + "auxiliary_loss_mlp": 0.00888451, + "balance_loss_clip": 1.0544858, + "balance_loss_mlp": 1.00070643, + "epoch": 0.31082787230205017, + "flos": 20521907314560.0, + "grad_norm": 2.908287601863902, + "language_loss": 0.80094254, + "learning_rate": 3.228360717180352e-06, + "loss": 0.82176763, + "num_input_tokens_seen": 55521970, + "step": 2585, + "time_per_iteration": 2.7500901222229004 + }, + { + "auxiliary_loss_clip": 0.01088416, + "auxiliary_loss_mlp": 0.00877055, + "balance_loss_clip": 1.01690698, + "balance_loss_mlp": 1.00103605, + "epoch": 0.3109481151926892, + "flos": 62445928723200.0, + "grad_norm": 0.8144099061399143, + "language_loss": 0.59392542, + "learning_rate": 3.227745886327266e-06, + "loss": 0.61358011, + "num_input_tokens_seen": 55580665, + "step": 2586, + "time_per_iteration": 3.2149245738983154 + }, + { + "auxiliary_loss_clip": 0.01087705, + "auxiliary_loss_mlp": 0.01006532, + "balance_loss_clip": 1.01607704, + "balance_loss_mlp": 1.00411224, + "epoch": 0.31106835808332833, + "flos": 44746744723200.0, + "grad_norm": 0.8098389790800525, + "language_loss": 0.55895293, + "learning_rate": 3.227130869226105e-06, + "loss": 0.57989538, + "num_input_tokens_seen": 55637825, + "step": 2587, + "time_per_iteration": 3.2470486164093018 + }, + { + "auxiliary_loss_clip": 0.01189517, + "auxiliary_loss_mlp": 0.01036362, + "balance_loss_clip": 1.05455732, + "balance_loss_mlp": 1.02679586, + "epoch": 0.3111886009739674, + "flos": 23403056100480.0, + "grad_norm": 2.5848545519154764, + "language_loss": 0.82310796, + "learning_rate": 3.226515665970167e-06, + "loss": 0.84536672, + "num_input_tokens_seen": 55655365, + "step": 2588, + "time_per_iteration": 2.7593483924865723 + }, + { + "auxiliary_loss_clip": 0.01186949, + "auxiliary_loss_mlp": 0.01033761, + "balance_loss_clip": 1.05362344, + "balance_loss_mlp": 1.0230267, + "epoch": 0.3113088438646065, + "flos": 17530728192000.0, + "grad_norm": 5.9222118855404355, + "language_loss": 0.8615371, + "learning_rate": 3.225900276652777e-06, + "loss": 0.88374424, + "num_input_tokens_seen": 55672140, + "step": 2589, + "time_per_iteration": 3.660444974899292 + }, + { + "auxiliary_loss_clip": 0.01184699, + "auxiliary_loss_mlp": 0.01030611, + "balance_loss_clip": 1.05266237, + "balance_loss_mlp": 1.02113366, + "epoch": 0.3114290867552456, + "flos": 28365802882560.0, + "grad_norm": 1.710101639830641, + "language_loss": 0.75698221, + "learning_rate": 3.2252847013672906e-06, + "loss": 0.77913535, + "num_input_tokens_seen": 55694800, + "step": 2590, + "time_per_iteration": 4.826667785644531 + }, + { + "auxiliary_loss_clip": 0.01161728, + "auxiliary_loss_mlp": 0.01030383, + "balance_loss_clip": 1.04960251, + "balance_loss_mlp": 1.02078664, + "epoch": 0.31154932964588467, + "flos": 27379157126400.0, + "grad_norm": 2.0478843776224096, + "language_loss": 0.75966692, + "learning_rate": 3.224668940207089e-06, + "loss": 0.78158808, + "num_input_tokens_seen": 55713785, + "step": 2591, + "time_per_iteration": 3.870251417160034 + }, + { + "auxiliary_loss_clip": 0.01151105, + "auxiliary_loss_mlp": 0.01035237, + "balance_loss_clip": 1.04714012, + "balance_loss_mlp": 1.02556324, + "epoch": 0.3116695725365238, + "flos": 26541864120960.0, + "grad_norm": 2.077048187672076, + "language_loss": 0.86939883, + "learning_rate": 3.2240529932655828e-06, + "loss": 0.89126229, + "num_input_tokens_seen": 55733050, + "step": 2592, + "time_per_iteration": 2.9058401584625244 + }, + { + "auxiliary_loss_clip": 0.01178449, + "auxiliary_loss_mlp": 0.01039491, + "balance_loss_clip": 1.05603158, + "balance_loss_mlp": 1.02936971, + "epoch": 0.3117898154271629, + "flos": 21177600134400.0, + "grad_norm": 2.548366482787148, + "language_loss": 0.8853389, + "learning_rate": 3.223436860636211e-06, + "loss": 0.90751827, + "num_input_tokens_seen": 55748685, + "step": 2593, + "time_per_iteration": 2.8496146202087402 + }, + { + "auxiliary_loss_clip": 0.01197554, + "auxiliary_loss_mlp": 0.0103061, + "balance_loss_clip": 1.05540752, + "balance_loss_mlp": 1.02122247, + "epoch": 0.31191005831780194, + "flos": 27272430840960.0, + "grad_norm": 1.5728452303468807, + "language_loss": 0.73996705, + "learning_rate": 3.2228205424124403e-06, + "loss": 0.76224869, + "num_input_tokens_seen": 55771840, + "step": 2594, + "time_per_iteration": 2.7435314655303955 + }, + { + "auxiliary_loss_clip": 0.01167606, + "auxiliary_loss_mlp": 0.01032362, + "balance_loss_clip": 1.05222213, + "balance_loss_mlp": 1.02249122, + "epoch": 0.31203030120844105, + "flos": 12963501043200.0, + "grad_norm": 2.2686400442218404, + "language_loss": 0.75064695, + "learning_rate": 3.222204038687765e-06, + "loss": 0.77264667, + "num_input_tokens_seen": 55784975, + "step": 2595, + "time_per_iteration": 2.6878631114959717 + }, + { + "auxiliary_loss_clip": 0.01183602, + "auxiliary_loss_mlp": 0.01030549, + "balance_loss_clip": 1.05363786, + "balance_loss_mlp": 1.02097666, + "epoch": 0.31215054409908016, + "flos": 27562014288000.0, + "grad_norm": 2.8268979640071645, + "language_loss": 0.88232923, + "learning_rate": 3.221587349555709e-06, + "loss": 0.90447068, + "num_input_tokens_seen": 55805235, + "step": 2596, + "time_per_iteration": 2.7556581497192383 + }, + { + "auxiliary_loss_clip": 0.01180572, + "auxiliary_loss_mlp": 0.01032574, + "balance_loss_clip": 1.05094922, + "balance_loss_mlp": 1.02319789, + "epoch": 0.3122707869897192, + "flos": 21506326427520.0, + "grad_norm": 1.7424836425444667, + "language_loss": 0.69347203, + "learning_rate": 3.2209704751098236e-06, + "loss": 0.71560347, + "num_input_tokens_seen": 55824265, + "step": 2597, + "time_per_iteration": 2.7734508514404297 + }, + { + "auxiliary_loss_clip": 0.01180272, + "auxiliary_loss_mlp": 0.01035496, + "balance_loss_clip": 1.05266213, + "balance_loss_mlp": 1.0254705, + "epoch": 0.31239102988035833, + "flos": 15187017674880.0, + "grad_norm": 2.059712545536322, + "language_loss": 0.82884467, + "learning_rate": 3.2203534154436875e-06, + "loss": 0.85100234, + "num_input_tokens_seen": 55838620, + "step": 2598, + "time_per_iteration": 2.757305383682251 + }, + { + "auxiliary_loss_clip": 0.01145067, + "auxiliary_loss_mlp": 0.0103941, + "balance_loss_clip": 1.04614437, + "balance_loss_mlp": 1.02861547, + "epoch": 0.31251127277099744, + "flos": 22053712763520.0, + "grad_norm": 1.8214435502827786, + "language_loss": 0.75532347, + "learning_rate": 3.2197361706509084e-06, + "loss": 0.77716821, + "num_input_tokens_seen": 55859375, + "step": 2599, + "time_per_iteration": 2.8993406295776367 + }, + { + "auxiliary_loss_clip": 0.01201149, + "auxiliary_loss_mlp": 0.01038743, + "balance_loss_clip": 1.05632281, + "balance_loss_mlp": 1.02799678, + "epoch": 0.3126315156616365, + "flos": 15193984913280.0, + "grad_norm": 2.5270195966722446, + "language_loss": 0.83964634, + "learning_rate": 3.2191187408251228e-06, + "loss": 0.86204523, + "num_input_tokens_seen": 55876535, + "step": 2600, + "time_per_iteration": 2.5754613876342773 + }, + { + "auxiliary_loss_clip": 0.01192151, + "auxiliary_loss_mlp": 0.01044561, + "balance_loss_clip": 1.05266356, + "balance_loss_mlp": 1.03291392, + "epoch": 0.3127517585522756, + "flos": 18145338831360.0, + "grad_norm": 2.202268856153455, + "language_loss": 0.78376472, + "learning_rate": 3.218501126059993e-06, + "loss": 0.80613184, + "num_input_tokens_seen": 55891930, + "step": 2601, + "time_per_iteration": 2.770200729370117 + }, + { + "auxiliary_loss_clip": 0.01189455, + "auxiliary_loss_mlp": 0.01031732, + "balance_loss_clip": 1.05057049, + "balance_loss_mlp": 1.02178967, + "epoch": 0.31287200144291466, + "flos": 21908633731200.0, + "grad_norm": 1.7682779259202897, + "language_loss": 0.81229663, + "learning_rate": 3.2178833264492116e-06, + "loss": 0.83450842, + "num_input_tokens_seen": 55910635, + "step": 2602, + "time_per_iteration": 2.716465950012207 + }, + { + "auxiliary_loss_clip": 0.01198497, + "auxiliary_loss_mlp": 0.01034645, + "balance_loss_clip": 1.05736232, + "balance_loss_mlp": 1.02364159, + "epoch": 0.31299224433355377, + "flos": 29896997800320.0, + "grad_norm": 1.7867568091785941, + "language_loss": 0.75867355, + "learning_rate": 3.217265342086498e-06, + "loss": 0.78100491, + "num_input_tokens_seen": 55931125, + "step": 2603, + "time_per_iteration": 2.82026743888855 + }, + { + "auxiliary_loss_clip": 0.01173141, + "auxiliary_loss_mlp": 0.00888229, + "balance_loss_clip": 1.05347514, + "balance_loss_mlp": 1.00078321, + "epoch": 0.3131124872241929, + "flos": 11655886331520.0, + "grad_norm": 3.063381826608995, + "language_loss": 0.73383558, + "learning_rate": 3.216647173065599e-06, + "loss": 0.75444925, + "num_input_tokens_seen": 55946590, + "step": 2604, + "time_per_iteration": 2.7426564693450928 + }, + { + "auxiliary_loss_clip": 0.01177926, + "auxiliary_loss_mlp": 0.01033338, + "balance_loss_clip": 1.05633545, + "balance_loss_mlp": 1.02303863, + "epoch": 0.31323273011483194, + "flos": 49848785470080.0, + "grad_norm": 1.6998040406233499, + "language_loss": 0.73855388, + "learning_rate": 3.216028819480292e-06, + "loss": 0.76066661, + "num_input_tokens_seen": 55967930, + "step": 2605, + "time_per_iteration": 2.974273681640625 + }, + { + "auxiliary_loss_clip": 0.0116565, + "auxiliary_loss_mlp": 0.01032619, + "balance_loss_clip": 1.05082631, + "balance_loss_mlp": 1.02333307, + "epoch": 0.31335297300547105, + "flos": 22601278667520.0, + "grad_norm": 3.2280863888043476, + "language_loss": 0.75298107, + "learning_rate": 3.2154102814243793e-06, + "loss": 0.77496374, + "num_input_tokens_seen": 55987070, + "step": 2606, + "time_per_iteration": 2.7959465980529785 + }, + { + "auxiliary_loss_clip": 0.01174782, + "auxiliary_loss_mlp": 0.01044841, + "balance_loss_clip": 1.05364382, + "balance_loss_mlp": 1.03474426, + "epoch": 0.31347321589611016, + "flos": 34710858708480.0, + "grad_norm": 2.785727210572916, + "language_loss": 0.66834986, + "learning_rate": 3.2147915589916937e-06, + "loss": 0.69054615, + "num_input_tokens_seen": 56008630, + "step": 2607, + "time_per_iteration": 2.947983741760254 + }, + { + "auxiliary_loss_clip": 0.01173153, + "auxiliary_loss_mlp": 0.01041878, + "balance_loss_clip": 1.05023932, + "balance_loss_mlp": 1.03216231, + "epoch": 0.3135934587867492, + "flos": 19755789108480.0, + "grad_norm": 2.5876225612190917, + "language_loss": 0.82548141, + "learning_rate": 3.2141726522760938e-06, + "loss": 0.84763169, + "num_input_tokens_seen": 56026690, + "step": 2608, + "time_per_iteration": 2.9023995399475098 + }, + { + "auxiliary_loss_clip": 0.01078034, + "auxiliary_loss_mlp": 0.01006212, + "balance_loss_clip": 1.01522446, + "balance_loss_mlp": 1.0039115, + "epoch": 0.3137137016773883, + "flos": 65815535583360.0, + "grad_norm": 0.7012533011769247, + "language_loss": 0.52703834, + "learning_rate": 3.213553561371469e-06, + "loss": 0.54788083, + "num_input_tokens_seen": 56090425, + "step": 2609, + "time_per_iteration": 3.3393075466156006 + }, + { + "auxiliary_loss_clip": 0.01161956, + "auxiliary_loss_mlp": 0.01035986, + "balance_loss_clip": 1.05244517, + "balance_loss_mlp": 1.02696204, + "epoch": 0.31383394456802743, + "flos": 16252739222400.0, + "grad_norm": 2.179233223259636, + "language_loss": 0.95508301, + "learning_rate": 3.212934286371733e-06, + "loss": 0.9770624, + "num_input_tokens_seen": 56107135, + "step": 2610, + "time_per_iteration": 2.866266965866089 + }, + { + "auxiliary_loss_clip": 0.01189518, + "auxiliary_loss_mlp": 0.01040543, + "balance_loss_clip": 1.05786896, + "balance_loss_mlp": 1.03035617, + "epoch": 0.3139541874586665, + "flos": 38795517613440.0, + "grad_norm": 2.47604259538517, + "language_loss": 0.83449072, + "learning_rate": 3.2123148273708304e-06, + "loss": 0.85679126, + "num_input_tokens_seen": 56127325, + "step": 2611, + "time_per_iteration": 2.8576815128326416 + }, + { + "auxiliary_loss_clip": 0.01200887, + "auxiliary_loss_mlp": 0.01034306, + "balance_loss_clip": 1.05710292, + "balance_loss_mlp": 1.02506113, + "epoch": 0.3140744303493056, + "flos": 25046328430080.0, + "grad_norm": 1.873990945547629, + "language_loss": 0.77143836, + "learning_rate": 3.211695184462733e-06, + "loss": 0.79379028, + "num_input_tokens_seen": 56148500, + "step": 2612, + "time_per_iteration": 2.8631298542022705 + }, + { + "auxiliary_loss_clip": 0.01065829, + "auxiliary_loss_mlp": 0.01006135, + "balance_loss_clip": 1.0125711, + "balance_loss_mlp": 1.00372744, + "epoch": 0.3141946732399447, + "flos": 72504254782080.0, + "grad_norm": 0.8962632791218461, + "language_loss": 0.60502303, + "learning_rate": 3.2110753577414383e-06, + "loss": 0.62574267, + "num_input_tokens_seen": 56210080, + "step": 2613, + "time_per_iteration": 3.264585018157959 + }, + { + "auxiliary_loss_clip": 0.01180157, + "auxiliary_loss_mlp": 0.01034498, + "balance_loss_clip": 1.0501976, + "balance_loss_mlp": 1.02418029, + "epoch": 0.31431491613058377, + "flos": 19239788280960.0, + "grad_norm": 2.2735615383172023, + "language_loss": 0.78829777, + "learning_rate": 3.2104553473009757e-06, + "loss": 0.8104443, + "num_input_tokens_seen": 56228200, + "step": 2614, + "time_per_iteration": 2.6879537105560303 + }, + { + "auxiliary_loss_clip": 0.01155157, + "auxiliary_loss_mlp": 0.01030666, + "balance_loss_clip": 1.04903769, + "balance_loss_mlp": 1.02163649, + "epoch": 0.3144351590212229, + "flos": 36210596290560.0, + "grad_norm": 1.8259818832436707, + "language_loss": 0.67953998, + "learning_rate": 3.209835153235399e-06, + "loss": 0.70139819, + "num_input_tokens_seen": 56249755, + "step": 2615, + "time_per_iteration": 3.9585626125335693 + }, + { + "auxiliary_loss_clip": 0.01156544, + "auxiliary_loss_mlp": 0.01032775, + "balance_loss_clip": 1.04647541, + "balance_loss_mlp": 1.02331567, + "epoch": 0.314555401911862, + "flos": 18551740285440.0, + "grad_norm": 1.849816249699355, + "language_loss": 0.67504621, + "learning_rate": 3.2092147756387916e-06, + "loss": 0.69693941, + "num_input_tokens_seen": 56270080, + "step": 2616, + "time_per_iteration": 4.602299213409424 + }, + { + "auxiliary_loss_clip": 0.01170075, + "auxiliary_loss_mlp": 0.01031577, + "balance_loss_clip": 1.0492444, + "balance_loss_mlp": 1.02173615, + "epoch": 0.31467564480250104, + "flos": 16362877299840.0, + "grad_norm": 2.0500667514374706, + "language_loss": 0.83578134, + "learning_rate": 3.208594214605264e-06, + "loss": 0.85779786, + "num_input_tokens_seen": 56288625, + "step": 2617, + "time_per_iteration": 3.6655466556549072 + }, + { + "auxiliary_loss_clip": 0.01166391, + "auxiliary_loss_mlp": 0.01039661, + "balance_loss_clip": 1.04949117, + "balance_loss_mlp": 1.03006458, + "epoch": 0.31479588769314015, + "flos": 21652375127040.0, + "grad_norm": 1.944989775717395, + "language_loss": 0.77118731, + "learning_rate": 3.2079734702289553e-06, + "loss": 0.79324782, + "num_input_tokens_seen": 56307520, + "step": 2618, + "time_per_iteration": 2.7836365699768066 + }, + { + "auxiliary_loss_clip": 0.01075803, + "auxiliary_loss_mlp": 0.00876619, + "balance_loss_clip": 1.01382899, + "balance_loss_mlp": 1.00087154, + "epoch": 0.3149161305837792, + "flos": 66051072040320.0, + "grad_norm": 0.810721315790425, + "language_loss": 0.60401034, + "learning_rate": 3.207352542604031e-06, + "loss": 0.62353456, + "num_input_tokens_seen": 56369855, + "step": 2619, + "time_per_iteration": 3.400986671447754 + }, + { + "auxiliary_loss_clip": 0.01157728, + "auxiliary_loss_mlp": 0.01030672, + "balance_loss_clip": 1.04875255, + "balance_loss_mlp": 1.0217309, + "epoch": 0.3150363734744183, + "flos": 28987201192320.0, + "grad_norm": 1.6888445130333796, + "language_loss": 0.78322554, + "learning_rate": 3.2067314318246864e-06, + "loss": 0.80510956, + "num_input_tokens_seen": 56390570, + "step": 2620, + "time_per_iteration": 2.8725426197052 + }, + { + "auxiliary_loss_clip": 0.01169269, + "auxiliary_loss_mlp": 0.01039751, + "balance_loss_clip": 1.05399704, + "balance_loss_mlp": 1.02996397, + "epoch": 0.31515661636505743, + "flos": 27636600879360.0, + "grad_norm": 2.1803913161071002, + "language_loss": 0.7803427, + "learning_rate": 3.206110137985143e-06, + "loss": 0.80243289, + "num_input_tokens_seen": 56410775, + "step": 2621, + "time_per_iteration": 2.8504748344421387 + }, + { + "auxiliary_loss_clip": 0.01160141, + "auxiliary_loss_mlp": 0.01031966, + "balance_loss_clip": 1.05021584, + "balance_loss_mlp": 1.02169037, + "epoch": 0.3152768592556965, + "flos": 24605632465920.0, + "grad_norm": 1.976333476388531, + "language_loss": 0.92352867, + "learning_rate": 3.2054886611796505e-06, + "loss": 0.94544983, + "num_input_tokens_seen": 56429770, + "step": 2622, + "time_per_iteration": 2.8249216079711914 + }, + { + "auxiliary_loss_clip": 0.01084585, + "auxiliary_loss_mlp": 0.01009025, + "balance_loss_clip": 1.01458859, + "balance_loss_mlp": 1.00662899, + "epoch": 0.3153971021463356, + "flos": 68476908026880.0, + "grad_norm": 0.8882545642425641, + "language_loss": 0.63513875, + "learning_rate": 3.204867001502487e-06, + "loss": 0.65607488, + "num_input_tokens_seen": 56488425, + "step": 2623, + "time_per_iteration": 3.223560094833374 + }, + { + "auxiliary_loss_clip": 0.01199415, + "auxiliary_loss_mlp": 0.01033499, + "balance_loss_clip": 1.05620599, + "balance_loss_mlp": 1.02397442, + "epoch": 0.3155173450369747, + "flos": 25593714766080.0, + "grad_norm": 1.758510099854591, + "language_loss": 0.80521947, + "learning_rate": 3.2042451590479567e-06, + "loss": 0.82754862, + "num_input_tokens_seen": 56508940, + "step": 2624, + "time_per_iteration": 2.745025396347046 + }, + { + "auxiliary_loss_clip": 0.01193912, + "auxiliary_loss_mlp": 0.01037107, + "balance_loss_clip": 1.05434513, + "balance_loss_mlp": 1.02811265, + "epoch": 0.31563758792761376, + "flos": 24309333175680.0, + "grad_norm": 1.7436526367900655, + "language_loss": 0.86868805, + "learning_rate": 3.203623133910394e-06, + "loss": 0.89099824, + "num_input_tokens_seen": 56527245, + "step": 2625, + "time_per_iteration": 2.6994996070861816 + }, + { + "auxiliary_loss_clip": 0.01150961, + "auxiliary_loss_mlp": 0.01033683, + "balance_loss_clip": 1.04899406, + "balance_loss_mlp": 1.02406931, + "epoch": 0.31575783081825287, + "flos": 31903865550720.0, + "grad_norm": 2.3783255769449734, + "language_loss": 0.76806819, + "learning_rate": 3.203000926184158e-06, + "loss": 0.78991461, + "num_input_tokens_seen": 56546170, + "step": 2626, + "time_per_iteration": 2.968712091445923 + }, + { + "auxiliary_loss_clip": 0.01196735, + "auxiliary_loss_mlp": 0.01029594, + "balance_loss_clip": 1.05612493, + "balance_loss_mlp": 1.0207901, + "epoch": 0.315878073708892, + "flos": 30810960385920.0, + "grad_norm": 1.7345616983222858, + "language_loss": 0.77670318, + "learning_rate": 3.202378535963639e-06, + "loss": 0.79896641, + "num_input_tokens_seen": 56567085, + "step": 2627, + "time_per_iteration": 2.696852207183838 + }, + { + "auxiliary_loss_clip": 0.01168568, + "auxiliary_loss_mlp": 0.00887984, + "balance_loss_clip": 1.04963326, + "balance_loss_mlp": 1.0007689, + "epoch": 0.31599831659953104, + "flos": 22200264253440.0, + "grad_norm": 1.715706534945293, + "language_loss": 0.83924592, + "learning_rate": 3.2017559633432516e-06, + "loss": 0.85981143, + "num_input_tokens_seen": 56586715, + "step": 2628, + "time_per_iteration": 2.8338866233825684 + }, + { + "auxiliary_loss_clip": 0.01186366, + "auxiliary_loss_mlp": 0.01033581, + "balance_loss_clip": 1.05259001, + "balance_loss_mlp": 1.02425337, + "epoch": 0.31611855949017015, + "flos": 25593463370880.0, + "grad_norm": 1.8043158767302185, + "language_loss": 0.66772437, + "learning_rate": 3.2011332084174398e-06, + "loss": 0.68992388, + "num_input_tokens_seen": 56607585, + "step": 2629, + "time_per_iteration": 2.7799601554870605 + }, + { + "auxiliary_loss_clip": 0.01184198, + "auxiliary_loss_mlp": 0.01035875, + "balance_loss_clip": 1.05296695, + "balance_loss_mlp": 1.02624297, + "epoch": 0.31623880238080926, + "flos": 20594087694720.0, + "grad_norm": 1.4710808804695648, + "language_loss": 0.89316219, + "learning_rate": 3.2005102712806756e-06, + "loss": 0.91536289, + "num_input_tokens_seen": 56626415, + "step": 2630, + "time_per_iteration": 2.678339719772339 + }, + { + "auxiliary_loss_clip": 0.01191857, + "auxiliary_loss_mlp": 0.01035696, + "balance_loss_clip": 1.0544771, + "balance_loss_mlp": 1.02600408, + "epoch": 0.3163590452714483, + "flos": 12784917600000.0, + "grad_norm": 2.1238080236527277, + "language_loss": 0.72941297, + "learning_rate": 3.1998871520274575e-06, + "loss": 0.75168848, + "num_input_tokens_seen": 56641750, + "step": 2631, + "time_per_iteration": 2.6807379722595215 + }, + { + "auxiliary_loss_clip": 0.01177751, + "auxiliary_loss_mlp": 0.01033152, + "balance_loss_clip": 1.04916978, + "balance_loss_mlp": 1.02408648, + "epoch": 0.3164792881620874, + "flos": 23041292273280.0, + "grad_norm": 1.866035129407178, + "language_loss": 0.84646297, + "learning_rate": 3.199263850752312e-06, + "loss": 0.868572, + "num_input_tokens_seen": 56662585, + "step": 2632, + "time_per_iteration": 2.7221460342407227 + }, + { + "auxiliary_loss_clip": 0.01189405, + "auxiliary_loss_mlp": 0.01033645, + "balance_loss_clip": 1.05410552, + "balance_loss_mlp": 1.02360189, + "epoch": 0.31659953105272653, + "flos": 18296271780480.0, + "grad_norm": 2.153526909666415, + "language_loss": 0.85886115, + "learning_rate": 3.198640367549795e-06, + "loss": 0.88109165, + "num_input_tokens_seen": 56681480, + "step": 2633, + "time_per_iteration": 2.7584877014160156 + }, + { + "auxiliary_loss_clip": 0.01189009, + "auxiliary_loss_mlp": 0.00887213, + "balance_loss_clip": 1.05519581, + "balance_loss_mlp": 1.00082338, + "epoch": 0.3167197739433656, + "flos": 25703421880320.0, + "grad_norm": 1.674976948257288, + "language_loss": 0.85767305, + "learning_rate": 3.198016702514487e-06, + "loss": 0.87843525, + "num_input_tokens_seen": 56701760, + "step": 2634, + "time_per_iteration": 2.72622013092041 + }, + { + "auxiliary_loss_clip": 0.01194962, + "auxiliary_loss_mlp": 0.01027914, + "balance_loss_clip": 1.05556226, + "balance_loss_mlp": 1.01954508, + "epoch": 0.3168400168340047, + "flos": 23546016230400.0, + "grad_norm": 1.8654476938442355, + "language_loss": 0.84549975, + "learning_rate": 3.1973928557409972e-06, + "loss": 0.86772859, + "num_input_tokens_seen": 56719800, + "step": 2635, + "time_per_iteration": 2.7336459159851074 + }, + { + "auxiliary_loss_clip": 0.01193621, + "auxiliary_loss_mlp": 0.01037529, + "balance_loss_clip": 1.05548763, + "balance_loss_mlp": 1.02864206, + "epoch": 0.31696025972464376, + "flos": 28366449327360.0, + "grad_norm": 1.8119867829619702, + "language_loss": 0.71348256, + "learning_rate": 3.1967688273239636e-06, + "loss": 0.73579407, + "num_input_tokens_seen": 56739605, + "step": 2636, + "time_per_iteration": 2.659630298614502 + }, + { + "auxiliary_loss_clip": 0.01164464, + "auxiliary_loss_mlp": 0.01029315, + "balance_loss_clip": 1.05031323, + "balance_loss_mlp": 1.0202136, + "epoch": 0.31708050261528287, + "flos": 16399111144320.0, + "grad_norm": 1.6267613920934765, + "language_loss": 0.82068729, + "learning_rate": 3.1961446173580503e-06, + "loss": 0.84262514, + "num_input_tokens_seen": 56756545, + "step": 2637, + "time_per_iteration": 2.796236753463745 + }, + { + "auxiliary_loss_clip": 0.01174959, + "auxiliary_loss_mlp": 0.0103186, + "balance_loss_clip": 1.0540545, + "balance_loss_mlp": 1.02153635, + "epoch": 0.317200745505922, + "flos": 26212347728640.0, + "grad_norm": 1.819107565773775, + "language_loss": 0.77349186, + "learning_rate": 3.1955202259379502e-06, + "loss": 0.79556, + "num_input_tokens_seen": 56778275, + "step": 2638, + "time_per_iteration": 2.8248846530914307 + }, + { + "auxiliary_loss_clip": 0.01182664, + "auxiliary_loss_mlp": 0.0103425, + "balance_loss_clip": 1.05164146, + "balance_loss_mlp": 1.02460003, + "epoch": 0.31732098839656103, + "flos": 31350876693120.0, + "grad_norm": 1.7456023167512136, + "language_loss": 0.83121276, + "learning_rate": 3.194895653158381e-06, + "loss": 0.85338187, + "num_input_tokens_seen": 56797215, + "step": 2639, + "time_per_iteration": 2.8043365478515625 + }, + { + "auxiliary_loss_clip": 0.0108374, + "auxiliary_loss_mlp": 0.01003066, + "balance_loss_clip": 1.01450753, + "balance_loss_mlp": 1.00060987, + "epoch": 0.31744123128720014, + "flos": 58989024835200.0, + "grad_norm": 0.7698522915851669, + "language_loss": 0.55485117, + "learning_rate": 3.194270899114093e-06, + "loss": 0.57571912, + "num_input_tokens_seen": 56863010, + "step": 2640, + "time_per_iteration": 3.3766491413116455 + }, + { + "auxiliary_loss_clip": 0.01192028, + "auxiliary_loss_mlp": 0.01035456, + "balance_loss_clip": 1.05456972, + "balance_loss_mlp": 1.02550232, + "epoch": 0.31756147417783925, + "flos": 17417573372160.0, + "grad_norm": 1.8856639296928874, + "language_loss": 0.81897783, + "learning_rate": 3.193645963899858e-06, + "loss": 0.84125268, + "num_input_tokens_seen": 56880625, + "step": 2641, + "time_per_iteration": 3.565368890762329 + }, + { + "auxiliary_loss_clip": 0.0117358, + "auxiliary_loss_mlp": 0.0102784, + "balance_loss_clip": 1.05245471, + "balance_loss_mlp": 1.01821971, + "epoch": 0.3176817170684783, + "flos": 25481673267840.0, + "grad_norm": 2.3557484659055663, + "language_loss": 0.84147966, + "learning_rate": 3.193020847610479e-06, + "loss": 0.8634938, + "num_input_tokens_seen": 56900945, + "step": 2642, + "time_per_iteration": 4.494927644729614 + }, + { + "auxiliary_loss_clip": 0.01169903, + "auxiliary_loss_mlp": 0.0103701, + "balance_loss_clip": 1.05243564, + "balance_loss_mlp": 1.02725911, + "epoch": 0.3178019599591174, + "flos": 24972603765120.0, + "grad_norm": 1.9795624573152915, + "language_loss": 0.70818973, + "learning_rate": 3.192395550340787e-06, + "loss": 0.73025888, + "num_input_tokens_seen": 56918895, + "step": 2643, + "time_per_iteration": 3.666747570037842 + }, + { + "auxiliary_loss_clip": 0.01183317, + "auxiliary_loss_mlp": 0.01033961, + "balance_loss_clip": 1.05345654, + "balance_loss_mlp": 1.02450812, + "epoch": 0.31792220284975653, + "flos": 12422220019200.0, + "grad_norm": 1.7882936688632387, + "language_loss": 0.76571941, + "learning_rate": 3.191770072185638e-06, + "loss": 0.78789222, + "num_input_tokens_seen": 56935890, + "step": 2644, + "time_per_iteration": 2.7008914947509766 + }, + { + "auxiliary_loss_clip": 0.01184552, + "auxiliary_loss_mlp": 0.01031824, + "balance_loss_clip": 1.05410886, + "balance_loss_mlp": 1.0226686, + "epoch": 0.3180424457403956, + "flos": 15485759089920.0, + "grad_norm": 2.264432374901562, + "language_loss": 0.7300781, + "learning_rate": 3.191144413239916e-06, + "loss": 0.75224185, + "num_input_tokens_seen": 56952460, + "step": 2645, + "time_per_iteration": 2.6445188522338867 + }, + { + "auxiliary_loss_clip": 0.01178668, + "auxiliary_loss_mlp": 0.01036648, + "balance_loss_clip": 1.0536356, + "balance_loss_mlp": 1.02632451, + "epoch": 0.3181626886310347, + "flos": 26174964648960.0, + "grad_norm": 2.0072277178326767, + "language_loss": 0.88643783, + "learning_rate": 3.190518573598534e-06, + "loss": 0.90859103, + "num_input_tokens_seen": 56969065, + "step": 2646, + "time_per_iteration": 2.7704555988311768 + }, + { + "auxiliary_loss_clip": 0.01172751, + "auxiliary_loss_mlp": 0.01036604, + "balance_loss_clip": 1.05082083, + "balance_loss_mlp": 1.02691269, + "epoch": 0.3182829315216738, + "flos": 25483109811840.0, + "grad_norm": 1.497835347793428, + "language_loss": 0.77356493, + "learning_rate": 3.1898925533564308e-06, + "loss": 0.79565847, + "num_input_tokens_seen": 56990535, + "step": 2647, + "time_per_iteration": 2.810001850128174 + }, + { + "auxiliary_loss_clip": 0.01154013, + "auxiliary_loss_mlp": 0.01031908, + "balance_loss_clip": 1.04824376, + "balance_loss_mlp": 1.02253222, + "epoch": 0.31840317441231286, + "flos": 18113701927680.0, + "grad_norm": 2.2885759189577843, + "language_loss": 0.63582867, + "learning_rate": 3.1892663526085733e-06, + "loss": 0.6576879, + "num_input_tokens_seen": 57008910, + "step": 2648, + "time_per_iteration": 2.7608635425567627 + }, + { + "auxiliary_loss_clip": 0.01083137, + "auxiliary_loss_mlp": 0.0100437, + "balance_loss_clip": 1.01402104, + "balance_loss_mlp": 1.00189078, + "epoch": 0.31852341730295197, + "flos": 64741948957440.0, + "grad_norm": 0.7502765536538415, + "language_loss": 0.56913126, + "learning_rate": 3.188639971449956e-06, + "loss": 0.59000635, + "num_input_tokens_seen": 57074960, + "step": 2649, + "time_per_iteration": 3.1841578483581543 + }, + { + "auxiliary_loss_clip": 0.01198033, + "auxiliary_loss_mlp": 0.01038741, + "balance_loss_clip": 1.05647647, + "balance_loss_mlp": 1.02872181, + "epoch": 0.318643660193591, + "flos": 20668135582080.0, + "grad_norm": 2.506363551834449, + "language_loss": 0.72486305, + "learning_rate": 3.1880134099756e-06, + "loss": 0.74723083, + "num_input_tokens_seen": 57094595, + "step": 2650, + "time_per_iteration": 2.591174840927124 + }, + { + "auxiliary_loss_clip": 0.01182682, + "auxiliary_loss_mlp": 0.01029416, + "balance_loss_clip": 1.05237103, + "balance_loss_mlp": 1.02002811, + "epoch": 0.31876390308423014, + "flos": 26943345411840.0, + "grad_norm": 3.5602636123733924, + "language_loss": 0.69737256, + "learning_rate": 3.1873866682805535e-06, + "loss": 0.71949357, + "num_input_tokens_seen": 57115290, + "step": 2651, + "time_per_iteration": 2.7650890350341797 + }, + { + "auxiliary_loss_clip": 0.01178679, + "auxiliary_loss_mlp": 0.01031282, + "balance_loss_clip": 1.0519346, + "balance_loss_mlp": 1.02198982, + "epoch": 0.31888414597486925, + "flos": 18041916597120.0, + "grad_norm": 1.7920992534208209, + "language_loss": 0.88276702, + "learning_rate": 3.186759746459894e-06, + "loss": 0.90486664, + "num_input_tokens_seen": 57134400, + "step": 2652, + "time_per_iteration": 2.6988253593444824 + }, + { + "auxiliary_loss_clip": 0.01175349, + "auxiliary_loss_mlp": 0.01033057, + "balance_loss_clip": 1.0531621, + "balance_loss_mlp": 1.02346039, + "epoch": 0.3190043888655083, + "flos": 25149319701120.0, + "grad_norm": 2.367636134824524, + "language_loss": 0.79498065, + "learning_rate": 3.1861326446087246e-06, + "loss": 0.8170647, + "num_input_tokens_seen": 57153140, + "step": 2653, + "time_per_iteration": 2.8049654960632324 + }, + { + "auxiliary_loss_clip": 0.01187662, + "auxiliary_loss_mlp": 0.01031101, + "balance_loss_clip": 1.05323684, + "balance_loss_mlp": 1.02106392, + "epoch": 0.3191246317561474, + "flos": 22053892331520.0, + "grad_norm": 2.0816219049012634, + "language_loss": 0.71715218, + "learning_rate": 3.1855053628221763e-06, + "loss": 0.73933977, + "num_input_tokens_seen": 57172395, + "step": 2654, + "time_per_iteration": 2.668783664703369 + }, + { + "auxiliary_loss_clip": 0.01158363, + "auxiliary_loss_mlp": 0.01032304, + "balance_loss_clip": 1.04758334, + "balance_loss_mlp": 1.02252281, + "epoch": 0.3192448746467865, + "flos": 14901815687040.0, + "grad_norm": 2.557081140923948, + "language_loss": 0.89381087, + "learning_rate": 3.184877901195407e-06, + "loss": 0.9157176, + "num_input_tokens_seen": 57189090, + "step": 2655, + "time_per_iteration": 2.735255241394043 + }, + { + "auxiliary_loss_clip": 0.01069529, + "auxiliary_loss_mlp": 0.01008654, + "balance_loss_clip": 1.01725173, + "balance_loss_mlp": 1.00543559, + "epoch": 0.3193651175374256, + "flos": 67234832657280.0, + "grad_norm": 0.8002456610000528, + "language_loss": 0.6290403, + "learning_rate": 3.184250259823602e-06, + "loss": 0.64982218, + "num_input_tokens_seen": 57251620, + "step": 2656, + "time_per_iteration": 3.3309144973754883 + }, + { + "auxiliary_loss_clip": 0.0116986, + "auxiliary_loss_mlp": 0.01030547, + "balance_loss_clip": 1.05142832, + "balance_loss_mlp": 1.02119493, + "epoch": 0.3194853604280647, + "flos": 12233077977600.0, + "grad_norm": 2.121870596336363, + "language_loss": 0.81878728, + "learning_rate": 3.183622438801974e-06, + "loss": 0.84079134, + "num_input_tokens_seen": 57266910, + "step": 2657, + "time_per_iteration": 2.7478833198547363 + }, + { + "auxiliary_loss_clip": 0.01197403, + "auxiliary_loss_mlp": 0.01035429, + "balance_loss_clip": 1.05813479, + "balance_loss_mlp": 1.02601767, + "epoch": 0.3196056033187038, + "flos": 14939917038720.0, + "grad_norm": 1.9095127693042928, + "language_loss": 0.75819081, + "learning_rate": 3.1829944382257637e-06, + "loss": 0.78051913, + "num_input_tokens_seen": 57285040, + "step": 2658, + "time_per_iteration": 2.6365485191345215 + }, + { + "auxiliary_loss_clip": 0.01183889, + "auxiliary_loss_mlp": 0.01032062, + "balance_loss_clip": 1.05487323, + "balance_loss_mlp": 1.02319908, + "epoch": 0.31972584620934286, + "flos": 23768878164480.0, + "grad_norm": 2.081710340516371, + "language_loss": 0.81362855, + "learning_rate": 3.1823662581902373e-06, + "loss": 0.83578807, + "num_input_tokens_seen": 57302725, + "step": 2659, + "time_per_iteration": 2.719757080078125 + }, + { + "auxiliary_loss_clip": 0.01150394, + "auxiliary_loss_mlp": 0.0103608, + "balance_loss_clip": 1.04320288, + "balance_loss_mlp": 1.02585208, + "epoch": 0.31984608909998197, + "flos": 21251540280960.0, + "grad_norm": 3.1459883311496073, + "language_loss": 0.74187768, + "learning_rate": 3.1817378987906896e-06, + "loss": 0.76374239, + "num_input_tokens_seen": 57322230, + "step": 2660, + "time_per_iteration": 2.8359227180480957 + }, + { + "auxiliary_loss_clip": 0.01152252, + "auxiliary_loss_mlp": 0.01035729, + "balance_loss_clip": 1.0516665, + "balance_loss_mlp": 1.02629316, + "epoch": 0.3199663319906211, + "flos": 18296235866880.0, + "grad_norm": 2.0518526123940526, + "language_loss": 0.7976104, + "learning_rate": 3.181109360122442e-06, + "loss": 0.81949019, + "num_input_tokens_seen": 57339820, + "step": 2661, + "time_per_iteration": 2.8078625202178955 + }, + { + "auxiliary_loss_clip": 0.01163278, + "auxiliary_loss_mlp": 0.01036081, + "balance_loss_clip": 1.05083609, + "balance_loss_mlp": 1.02618027, + "epoch": 0.32008657488126013, + "flos": 18733627779840.0, + "grad_norm": 2.8797712326939355, + "language_loss": 0.77868128, + "learning_rate": 3.1804806422808445e-06, + "loss": 0.8006748, + "num_input_tokens_seen": 57356955, + "step": 2662, + "time_per_iteration": 2.7576301097869873 + }, + { + "auxiliary_loss_clip": 0.01167643, + "auxiliary_loss_mlp": 0.01031543, + "balance_loss_clip": 1.05073166, + "balance_loss_mlp": 1.02130306, + "epoch": 0.32020681777189924, + "flos": 20595344670720.0, + "grad_norm": 1.6270932018473179, + "language_loss": 0.72890824, + "learning_rate": 3.1798517453612714e-06, + "loss": 0.75090009, + "num_input_tokens_seen": 57376760, + "step": 2663, + "time_per_iteration": 2.713576555252075 + }, + { + "auxiliary_loss_clip": 0.01184918, + "auxiliary_loss_mlp": 0.01027281, + "balance_loss_clip": 1.05745471, + "balance_loss_mlp": 1.01807201, + "epoch": 0.32032706066253835, + "flos": 35261692750080.0, + "grad_norm": 1.8098558770980158, + "language_loss": 0.75205314, + "learning_rate": 3.1792226694591265e-06, + "loss": 0.77417517, + "num_input_tokens_seen": 57398145, + "step": 2664, + "time_per_iteration": 2.8276901245117188 + }, + { + "auxiliary_loss_clip": 0.01168101, + "auxiliary_loss_mlp": 0.01032906, + "balance_loss_clip": 1.0530324, + "balance_loss_mlp": 1.0236913, + "epoch": 0.3204473035531774, + "flos": 15304230731520.0, + "grad_norm": 1.8805543877154371, + "language_loss": 0.80416751, + "learning_rate": 3.178593414669841e-06, + "loss": 0.8261776, + "num_input_tokens_seen": 57416730, + "step": 2665, + "time_per_iteration": 2.803339958190918 + }, + { + "auxiliary_loss_clip": 0.01190088, + "auxiliary_loss_mlp": 0.01039104, + "balance_loss_clip": 1.05608225, + "balance_loss_mlp": 1.02901924, + "epoch": 0.3205675464438165, + "flos": 24462564595200.0, + "grad_norm": 2.0206099630551577, + "language_loss": 0.70529997, + "learning_rate": 3.1779639810888707e-06, + "loss": 0.72759181, + "num_input_tokens_seen": 57436325, + "step": 2666, + "time_per_iteration": 2.732357978820801 + }, + { + "auxiliary_loss_clip": 0.01185358, + "auxiliary_loss_mlp": 0.01033944, + "balance_loss_clip": 1.05651915, + "balance_loss_mlp": 1.02331007, + "epoch": 0.3206877893344556, + "flos": 22456235548800.0, + "grad_norm": 1.7406983020373339, + "language_loss": 0.75828338, + "learning_rate": 3.1773343688117013e-06, + "loss": 0.78047639, + "num_input_tokens_seen": 57457235, + "step": 2667, + "time_per_iteration": 3.6825027465820312 + }, + { + "auxiliary_loss_clip": 0.01181338, + "auxiliary_loss_mlp": 0.00887224, + "balance_loss_clip": 1.05376017, + "balance_loss_mlp": 1.00073123, + "epoch": 0.3208080322250947, + "flos": 20412236113920.0, + "grad_norm": 2.222287038063942, + "language_loss": 0.84071243, + "learning_rate": 3.1767045779338445e-06, + "loss": 0.8613981, + "num_input_tokens_seen": 57474895, + "step": 2668, + "time_per_iteration": 3.6446428298950195 + }, + { + "auxiliary_loss_clip": 0.01185373, + "auxiliary_loss_mlp": 0.01027079, + "balance_loss_clip": 1.05248272, + "balance_loss_mlp": 1.01783431, + "epoch": 0.3209282751157338, + "flos": 21762118154880.0, + "grad_norm": 1.987040276603609, + "language_loss": 0.9183768, + "learning_rate": 3.176074608550839e-06, + "loss": 0.94050121, + "num_input_tokens_seen": 57490715, + "step": 2669, + "time_per_iteration": 3.6279196739196777 + }, + { + "auxiliary_loss_clip": 0.01149159, + "auxiliary_loss_mlp": 0.01031625, + "balance_loss_clip": 1.04892147, + "balance_loss_mlp": 1.02228487, + "epoch": 0.32104851800637285, + "flos": 22055041566720.0, + "grad_norm": 2.301063407611934, + "language_loss": 0.82384425, + "learning_rate": 3.17544446075825e-06, + "loss": 0.8456521, + "num_input_tokens_seen": 57509880, + "step": 2670, + "time_per_iteration": 2.9188060760498047 + }, + { + "auxiliary_loss_clip": 0.01183393, + "auxiliary_loss_mlp": 0.01036765, + "balance_loss_clip": 1.05480552, + "balance_loss_mlp": 1.02715087, + "epoch": 0.32116876089701196, + "flos": 37012301896320.0, + "grad_norm": 1.5613601587338435, + "language_loss": 0.71196407, + "learning_rate": 3.174814134651671e-06, + "loss": 0.73416567, + "num_input_tokens_seen": 57532430, + "step": 2671, + "time_per_iteration": 2.8774051666259766 + }, + { + "auxiliary_loss_clip": 0.01195578, + "auxiliary_loss_mlp": 0.01028188, + "balance_loss_clip": 1.05841136, + "balance_loss_mlp": 1.01868129, + "epoch": 0.3212890037876511, + "flos": 21979233912960.0, + "grad_norm": 1.7244486656312608, + "language_loss": 0.80374366, + "learning_rate": 3.1741836303267215e-06, + "loss": 0.82598132, + "num_input_tokens_seen": 57551965, + "step": 2672, + "time_per_iteration": 2.670433521270752 + }, + { + "auxiliary_loss_clip": 0.01194464, + "auxiliary_loss_mlp": 0.01033137, + "balance_loss_clip": 1.05656791, + "balance_loss_mlp": 1.02364206, + "epoch": 0.32140924667829013, + "flos": 10342345875840.0, + "grad_norm": 1.902249544240216, + "language_loss": 0.75351536, + "learning_rate": 3.1735529478790496e-06, + "loss": 0.77579141, + "num_input_tokens_seen": 57569955, + "step": 2673, + "time_per_iteration": 2.6047818660736084 + }, + { + "auxiliary_loss_clip": 0.01189207, + "auxiliary_loss_mlp": 0.0103371, + "balance_loss_clip": 1.05717599, + "balance_loss_mlp": 1.0240953, + "epoch": 0.32152948956892924, + "flos": 50798910072960.0, + "grad_norm": 1.815243021024029, + "language_loss": 0.7938655, + "learning_rate": 3.172922087404328e-06, + "loss": 0.81609464, + "num_input_tokens_seen": 57592215, + "step": 2674, + "time_per_iteration": 2.9056127071380615 + }, + { + "auxiliary_loss_clip": 0.0108866, + "auxiliary_loss_mlp": 0.01002818, + "balance_loss_clip": 1.02045274, + "balance_loss_mlp": 1.0004456, + "epoch": 0.32164973245956835, + "flos": 63863250549120.0, + "grad_norm": 0.771756893240344, + "language_loss": 0.55278206, + "learning_rate": 3.1722910489982586e-06, + "loss": 0.57369685, + "num_input_tokens_seen": 57652575, + "step": 2675, + "time_per_iteration": 3.2736055850982666 + }, + { + "auxiliary_loss_clip": 0.01176147, + "auxiliary_loss_mlp": 0.0103171, + "balance_loss_clip": 1.05371344, + "balance_loss_mlp": 1.02185786, + "epoch": 0.3217699753502074, + "flos": 23513948363520.0, + "grad_norm": 1.7678067319385642, + "language_loss": 0.80119848, + "learning_rate": 3.1716598327565694e-06, + "loss": 0.82327706, + "num_input_tokens_seen": 57672215, + "step": 2676, + "time_per_iteration": 2.759598970413208 + }, + { + "auxiliary_loss_clip": 0.01198144, + "auxiliary_loss_mlp": 0.01028494, + "balance_loss_clip": 1.0593338, + "balance_loss_mlp": 1.01898694, + "epoch": 0.3218902182408465, + "flos": 19062533640960.0, + "grad_norm": 1.4655088545987651, + "language_loss": 0.84251624, + "learning_rate": 3.171028438775015e-06, + "loss": 0.86478257, + "num_input_tokens_seen": 57691410, + "step": 2677, + "time_per_iteration": 2.6726250648498535 + }, + { + "auxiliary_loss_clip": 0.01195351, + "auxiliary_loss_mlp": 0.01034744, + "balance_loss_clip": 1.05697811, + "balance_loss_mlp": 1.0256778, + "epoch": 0.3220104611314856, + "flos": 20375571306240.0, + "grad_norm": 1.798916214480224, + "language_loss": 0.84359998, + "learning_rate": 3.170396867149377e-06, + "loss": 0.86590099, + "num_input_tokens_seen": 57709415, + "step": 2678, + "time_per_iteration": 2.7261414527893066 + }, + { + "auxiliary_loss_clip": 0.01148898, + "auxiliary_loss_mlp": 0.01030995, + "balance_loss_clip": 1.04816532, + "balance_loss_mlp": 1.02099359, + "epoch": 0.3221307040221247, + "flos": 20117014231680.0, + "grad_norm": 1.7294148287351334, + "language_loss": 0.86519092, + "learning_rate": 3.1697651179754653e-06, + "loss": 0.88698983, + "num_input_tokens_seen": 57728075, + "step": 2679, + "time_per_iteration": 2.773019790649414 + }, + { + "auxiliary_loss_clip": 0.01170262, + "auxiliary_loss_mlp": 0.01034347, + "balance_loss_clip": 1.05624986, + "balance_loss_mlp": 1.02479231, + "epoch": 0.3222509469127638, + "flos": 23987789602560.0, + "grad_norm": 1.7765411356510834, + "language_loss": 0.73408043, + "learning_rate": 3.1691331913491153e-06, + "loss": 0.75612652, + "num_input_tokens_seen": 57750645, + "step": 2680, + "time_per_iteration": 2.8908238410949707 + }, + { + "auxiliary_loss_clip": 0.01196287, + "auxiliary_loss_mlp": 0.01030777, + "balance_loss_clip": 1.05637991, + "balance_loss_mlp": 1.02098393, + "epoch": 0.32237118980340285, + "flos": 17675735397120.0, + "grad_norm": 2.288243791684845, + "language_loss": 0.8445816, + "learning_rate": 3.1685010873661898e-06, + "loss": 0.86685222, + "num_input_tokens_seen": 57769820, + "step": 2681, + "time_per_iteration": 2.6967711448669434 + }, + { + "auxiliary_loss_clip": 0.01183998, + "auxiliary_loss_mlp": 0.01032728, + "balance_loss_clip": 1.05461073, + "balance_loss_mlp": 1.02264845, + "epoch": 0.32249143269404196, + "flos": 23147982645120.0, + "grad_norm": 1.826801677746064, + "language_loss": 0.79943353, + "learning_rate": 3.167868806122578e-06, + "loss": 0.82160079, + "num_input_tokens_seen": 57788870, + "step": 2682, + "time_per_iteration": 2.794891357421875 + }, + { + "auxiliary_loss_clip": 0.01181264, + "auxiliary_loss_mlp": 0.0102816, + "balance_loss_clip": 1.05466938, + "balance_loss_mlp": 1.01777649, + "epoch": 0.32261167558468107, + "flos": 24422308427520.0, + "grad_norm": 1.9587316259426832, + "language_loss": 0.66302562, + "learning_rate": 3.1672363477141968e-06, + "loss": 0.68511987, + "num_input_tokens_seen": 57808165, + "step": 2683, + "time_per_iteration": 2.8130507469177246 + }, + { + "auxiliary_loss_clip": 0.0117939, + "auxiliary_loss_mlp": 0.01036479, + "balance_loss_clip": 1.05334711, + "balance_loss_mlp": 1.02644742, + "epoch": 0.3227319184753201, + "flos": 30367175852160.0, + "grad_norm": 2.003836124987335, + "language_loss": 0.85186255, + "learning_rate": 3.1666037122369903e-06, + "loss": 0.87402123, + "num_input_tokens_seen": 57828825, + "step": 2684, + "time_per_iteration": 2.8294146060943604 + }, + { + "auxiliary_loss_clip": 0.01183594, + "auxiliary_loss_mlp": 0.01033432, + "balance_loss_clip": 1.05209863, + "balance_loss_mlp": 1.02378786, + "epoch": 0.32285216136595923, + "flos": 16946174257920.0, + "grad_norm": 2.055347249304115, + "language_loss": 0.86872745, + "learning_rate": 3.165970899786928e-06, + "loss": 0.89089769, + "num_input_tokens_seen": 57846740, + "step": 2685, + "time_per_iteration": 2.7578682899475098 + }, + { + "auxiliary_loss_clip": 0.01172372, + "auxiliary_loss_mlp": 0.01036061, + "balance_loss_clip": 1.05363274, + "balance_loss_mlp": 1.0259819, + "epoch": 0.32297240425659834, + "flos": 21981532383360.0, + "grad_norm": 1.6895095179028738, + "language_loss": 0.7555846, + "learning_rate": 3.1653379104600067e-06, + "loss": 0.77766895, + "num_input_tokens_seen": 57866885, + "step": 2686, + "time_per_iteration": 2.8192694187164307 + }, + { + "auxiliary_loss_clip": 0.01186655, + "auxiliary_loss_mlp": 0.01031982, + "balance_loss_clip": 1.05491614, + "balance_loss_mlp": 1.02214766, + "epoch": 0.3230926471472374, + "flos": 22748045639040.0, + "grad_norm": 1.4350246880288424, + "language_loss": 0.69346315, + "learning_rate": 3.164704744352251e-06, + "loss": 0.7156496, + "num_input_tokens_seen": 57887690, + "step": 2687, + "time_per_iteration": 2.6824042797088623 + }, + { + "auxiliary_loss_clip": 0.01185319, + "auxiliary_loss_mlp": 0.0103506, + "balance_loss_clip": 1.05449986, + "balance_loss_mlp": 1.02598262, + "epoch": 0.3232128900378765, + "flos": 16942977947520.0, + "grad_norm": 1.7403346909088366, + "language_loss": 0.80818665, + "learning_rate": 3.164071401559713e-06, + "loss": 0.83039045, + "num_input_tokens_seen": 57905090, + "step": 2688, + "time_per_iteration": 2.6669204235076904 + }, + { + "auxiliary_loss_clip": 0.01177735, + "auxiliary_loss_mlp": 0.01031901, + "balance_loss_clip": 1.05300927, + "balance_loss_mlp": 1.02261424, + "epoch": 0.3233331329285156, + "flos": 24023736138240.0, + "grad_norm": 4.631357603795658, + "language_loss": 0.71320277, + "learning_rate": 3.1634378821784674e-06, + "loss": 0.73529911, + "num_input_tokens_seen": 57925305, + "step": 2689, + "time_per_iteration": 2.738490581512451 + }, + { + "auxiliary_loss_clip": 0.01170344, + "auxiliary_loss_mlp": 0.01035799, + "balance_loss_clip": 1.05593181, + "balance_loss_mlp": 1.02610159, + "epoch": 0.3234533758191547, + "flos": 18113845582080.0, + "grad_norm": 2.1107814678736565, + "language_loss": 0.74060392, + "learning_rate": 3.1628041863046208e-06, + "loss": 0.76266533, + "num_input_tokens_seen": 57942720, + "step": 2690, + "time_per_iteration": 2.781107187271118 + }, + { + "auxiliary_loss_clip": 0.01202214, + "auxiliary_loss_mlp": 0.01032265, + "balance_loss_clip": 1.05673862, + "balance_loss_mlp": 1.02135777, + "epoch": 0.3235736187097938, + "flos": 16946138344320.0, + "grad_norm": 2.381979473460659, + "language_loss": 0.91207492, + "learning_rate": 3.162170314034304e-06, + "loss": 0.93441975, + "num_input_tokens_seen": 57960135, + "step": 2691, + "time_per_iteration": 2.5900778770446777 + }, + { + "auxiliary_loss_clip": 0.01199627, + "auxiliary_loss_mlp": 0.01038611, + "balance_loss_clip": 1.05618215, + "balance_loss_mlp": 1.02849066, + "epoch": 0.3236938616004329, + "flos": 22127150119680.0, + "grad_norm": 1.5168819402926683, + "language_loss": 0.81044865, + "learning_rate": 3.1615362654636738e-06, + "loss": 0.83283103, + "num_input_tokens_seen": 57980875, + "step": 2692, + "time_per_iteration": 2.8122670650482178 + }, + { + "auxiliary_loss_clip": 0.01163431, + "auxiliary_loss_mlp": 0.01032715, + "balance_loss_clip": 1.05432105, + "balance_loss_mlp": 1.02399528, + "epoch": 0.32381410449107195, + "flos": 17164618819200.0, + "grad_norm": 1.657777155382316, + "language_loss": 0.86863434, + "learning_rate": 3.1609020406889163e-06, + "loss": 0.89059579, + "num_input_tokens_seen": 57998310, + "step": 2693, + "time_per_iteration": 3.8234899044036865 + }, + { + "auxiliary_loss_clip": 0.01179394, + "auxiliary_loss_mlp": 0.010322, + "balance_loss_clip": 1.05274057, + "balance_loss_mlp": 1.02257967, + "epoch": 0.32393434738171106, + "flos": 16578125550720.0, + "grad_norm": 3.5389042304389684, + "language_loss": 0.84843183, + "learning_rate": 3.1602676398062416e-06, + "loss": 0.87054783, + "num_input_tokens_seen": 58017220, + "step": 2694, + "time_per_iteration": 3.6058549880981445 + }, + { + "auxiliary_loss_clip": 0.01187286, + "auxiliary_loss_mlp": 0.01031409, + "balance_loss_clip": 1.05688441, + "balance_loss_mlp": 1.02083564, + "epoch": 0.3240545902723502, + "flos": 25483612602240.0, + "grad_norm": 2.8119330656575996, + "language_loss": 0.61278772, + "learning_rate": 3.1596330629118886e-06, + "loss": 0.6349746, + "num_input_tokens_seen": 58037190, + "step": 2695, + "time_per_iteration": 3.692068099975586 + }, + { + "auxiliary_loss_clip": 0.0114925, + "auxiliary_loss_mlp": 0.01032474, + "balance_loss_clip": 1.04754162, + "balance_loss_mlp": 1.0226512, + "epoch": 0.32417483316298923, + "flos": 35845851634560.0, + "grad_norm": 2.278591501857804, + "language_loss": 0.73222464, + "learning_rate": 3.1589983101021223e-06, + "loss": 0.75404191, + "num_input_tokens_seen": 58055820, + "step": 2696, + "time_per_iteration": 2.9698760509490967 + }, + { + "auxiliary_loss_clip": 0.01178628, + "auxiliary_loss_mlp": 0.01032893, + "balance_loss_clip": 1.05442023, + "balance_loss_mlp": 1.02348697, + "epoch": 0.32429507605362834, + "flos": 30080501406720.0, + "grad_norm": 2.0003787873576466, + "language_loss": 0.84777415, + "learning_rate": 3.1583633814732337e-06, + "loss": 0.86988938, + "num_input_tokens_seen": 58075340, + "step": 2697, + "time_per_iteration": 2.918724775314331 + }, + { + "auxiliary_loss_clip": 0.0119548, + "auxiliary_loss_mlp": 0.01036974, + "balance_loss_clip": 1.05485773, + "balance_loss_mlp": 1.02749658, + "epoch": 0.3244153189442674, + "flos": 18223265387520.0, + "grad_norm": 2.3870567091654893, + "language_loss": 0.71437997, + "learning_rate": 3.157728277121541e-06, + "loss": 0.73670447, + "num_input_tokens_seen": 58093515, + "step": 2698, + "time_per_iteration": 2.8153979778289795 + }, + { + "auxiliary_loss_clip": 0.01198077, + "auxiliary_loss_mlp": 0.01033855, + "balance_loss_clip": 1.05659485, + "balance_loss_mlp": 1.02337623, + "epoch": 0.3245355618349065, + "flos": 17710317216000.0, + "grad_norm": 2.2004926540004153, + "language_loss": 0.78609776, + "learning_rate": 3.1570929971433897e-06, + "loss": 0.80841702, + "num_input_tokens_seen": 58109300, + "step": 2699, + "time_per_iteration": 2.524951934814453 + }, + { + "auxiliary_loss_clip": 0.01189704, + "auxiliary_loss_mlp": 0.01034515, + "balance_loss_clip": 1.05758953, + "balance_loss_mlp": 1.02472758, + "epoch": 0.3246558047255456, + "flos": 23440798316160.0, + "grad_norm": 1.8350907437431405, + "language_loss": 0.83844519, + "learning_rate": 3.1564575416351504e-06, + "loss": 0.86068738, + "num_input_tokens_seen": 58128000, + "step": 2700, + "time_per_iteration": 2.6117124557495117 + }, + { + "auxiliary_loss_clip": 0.01199034, + "auxiliary_loss_mlp": 0.0103397, + "balance_loss_clip": 1.05823445, + "balance_loss_mlp": 1.0237062, + "epoch": 0.32477604761618467, + "flos": 21760861178880.0, + "grad_norm": 2.352510390737096, + "language_loss": 0.74164623, + "learning_rate": 3.155821910693221e-06, + "loss": 0.76397622, + "num_input_tokens_seen": 58147415, + "step": 2701, + "time_per_iteration": 2.5086781978607178 + }, + { + "auxiliary_loss_clip": 0.01174924, + "auxiliary_loss_mlp": 0.01037223, + "balance_loss_clip": 1.05192971, + "balance_loss_mlp": 1.02693534, + "epoch": 0.3248962905068238, + "flos": 19828328624640.0, + "grad_norm": 1.5890202629590022, + "language_loss": 0.85763466, + "learning_rate": 3.1551861044140275e-06, + "loss": 0.87975609, + "num_input_tokens_seen": 58167050, + "step": 2702, + "time_per_iteration": 2.608898639678955 + }, + { + "auxiliary_loss_clip": 0.0115512, + "auxiliary_loss_mlp": 0.01031208, + "balance_loss_clip": 1.05224943, + "balance_loss_mlp": 1.02089047, + "epoch": 0.3250165333974629, + "flos": 23948215793280.0, + "grad_norm": 1.6066518562955578, + "language_loss": 0.77403891, + "learning_rate": 3.15455012289402e-06, + "loss": 0.79590213, + "num_input_tokens_seen": 58186695, + "step": 2703, + "time_per_iteration": 2.652590751647949 + }, + { + "auxiliary_loss_clip": 0.01192097, + "auxiliary_loss_mlp": 0.01035748, + "balance_loss_clip": 1.06094694, + "balance_loss_mlp": 1.02618754, + "epoch": 0.32513677628810195, + "flos": 23989333887360.0, + "grad_norm": 1.7160409238085121, + "language_loss": 0.84297264, + "learning_rate": 3.153913966229677e-06, + "loss": 0.86525106, + "num_input_tokens_seen": 58205815, + "step": 2704, + "time_per_iteration": 2.5891852378845215 + }, + { + "auxiliary_loss_clip": 0.01081187, + "auxiliary_loss_mlp": 0.01003231, + "balance_loss_clip": 1.01910114, + "balance_loss_mlp": 1.00078726, + "epoch": 0.32525701917874106, + "flos": 70655790009600.0, + "grad_norm": 0.6475670358548073, + "language_loss": 0.50299329, + "learning_rate": 3.1532776345175027e-06, + "loss": 0.52383745, + "num_input_tokens_seen": 58270960, + "step": 2705, + "time_per_iteration": 3.3028149604797363 + }, + { + "auxiliary_loss_clip": 0.01195486, + "auxiliary_loss_mlp": 0.01026808, + "balance_loss_clip": 1.05698097, + "balance_loss_mlp": 1.01765311, + "epoch": 0.32537726206938017, + "flos": 19682639061120.0, + "grad_norm": 2.077542428136417, + "language_loss": 0.78702462, + "learning_rate": 3.1526411278540285e-06, + "loss": 0.80924749, + "num_input_tokens_seen": 58289390, + "step": 2706, + "time_per_iteration": 2.745781421661377 + }, + { + "auxiliary_loss_clip": 0.01180311, + "auxiliary_loss_mlp": 0.01033147, + "balance_loss_clip": 1.04991579, + "balance_loss_mlp": 1.02264476, + "epoch": 0.3254975049600192, + "flos": 28760999293440.0, + "grad_norm": 2.368633887429326, + "language_loss": 0.81582665, + "learning_rate": 3.1520044463358116e-06, + "loss": 0.83796126, + "num_input_tokens_seen": 58306120, + "step": 2707, + "time_per_iteration": 2.880046844482422 + }, + { + "auxiliary_loss_clip": 0.01185743, + "auxiliary_loss_mlp": 0.01034338, + "balance_loss_clip": 1.05621171, + "balance_loss_mlp": 1.02461684, + "epoch": 0.32561774785065833, + "flos": 18877378008960.0, + "grad_norm": 1.698268957095872, + "language_loss": 0.80461597, + "learning_rate": 3.151367590059436e-06, + "loss": 0.8268168, + "num_input_tokens_seen": 58324545, + "step": 2708, + "time_per_iteration": 2.8384950160980225 + }, + { + "auxiliary_loss_clip": 0.01199612, + "auxiliary_loss_mlp": 0.00888227, + "balance_loss_clip": 1.05811357, + "balance_loss_mlp": 1.00069737, + "epoch": 0.32573799074129745, + "flos": 23112107936640.0, + "grad_norm": 2.0155640605029395, + "language_loss": 0.86959487, + "learning_rate": 3.1507305591215117e-06, + "loss": 0.89047325, + "num_input_tokens_seen": 58342455, + "step": 2709, + "time_per_iteration": 2.8052918910980225 + }, + { + "auxiliary_loss_clip": 0.01078767, + "auxiliary_loss_mlp": 0.01004528, + "balance_loss_clip": 1.01712179, + "balance_loss_mlp": 1.00200033, + "epoch": 0.3258582336319365, + "flos": 71237650423680.0, + "grad_norm": 0.6722813520677285, + "language_loss": 0.5573222, + "learning_rate": 3.150093353618677e-06, + "loss": 0.57815516, + "num_input_tokens_seen": 58407185, + "step": 2710, + "time_per_iteration": 3.3485066890716553 + }, + { + "auxiliary_loss_clip": 0.0119273, + "auxiliary_loss_mlp": 0.01038037, + "balance_loss_clip": 1.05438125, + "balance_loss_mlp": 1.02888751, + "epoch": 0.3259784765225756, + "flos": 22456020067200.0, + "grad_norm": 2.8270615928277905, + "language_loss": 0.88482559, + "learning_rate": 3.149455973647596e-06, + "loss": 0.90713322, + "num_input_tokens_seen": 58425245, + "step": 2711, + "time_per_iteration": 2.6325082778930664 + }, + { + "auxiliary_loss_clip": 0.01156929, + "auxiliary_loss_mlp": 0.01032315, + "balance_loss_clip": 1.04597294, + "balance_loss_mlp": 1.02216411, + "epoch": 0.32609871941321467, + "flos": 20484811543680.0, + "grad_norm": 1.7804176870834092, + "language_loss": 0.77067125, + "learning_rate": 3.1488184193049563e-06, + "loss": 0.79256368, + "num_input_tokens_seen": 58444780, + "step": 2712, + "time_per_iteration": 2.7099409103393555 + }, + { + "auxiliary_loss_clip": 0.01194209, + "auxiliary_loss_mlp": 0.0103436, + "balance_loss_clip": 1.05549264, + "balance_loss_mlp": 1.02496672, + "epoch": 0.3262189623038538, + "flos": 22416805393920.0, + "grad_norm": 1.543856383985829, + "language_loss": 0.72243786, + "learning_rate": 3.1481806906874767e-06, + "loss": 0.7447235, + "num_input_tokens_seen": 58466090, + "step": 2713, + "time_per_iteration": 2.7653915882110596 + }, + { + "auxiliary_loss_clip": 0.01195089, + "auxiliary_loss_mlp": 0.01030244, + "balance_loss_clip": 1.0561043, + "balance_loss_mlp": 1.02073157, + "epoch": 0.3263392051944929, + "flos": 20923496346240.0, + "grad_norm": 1.5393016716519707, + "language_loss": 0.87774062, + "learning_rate": 3.147542787891899e-06, + "loss": 0.8999939, + "num_input_tokens_seen": 58485435, + "step": 2714, + "time_per_iteration": 2.835423231124878 + }, + { + "auxiliary_loss_clip": 0.01178606, + "auxiliary_loss_mlp": 0.01031008, + "balance_loss_clip": 1.05588794, + "balance_loss_mlp": 1.02132821, + "epoch": 0.32645944808513194, + "flos": 24025172682240.0, + "grad_norm": 1.7251822317140928, + "language_loss": 0.75569057, + "learning_rate": 3.1469047110149926e-06, + "loss": 0.77778667, + "num_input_tokens_seen": 58504175, + "step": 2715, + "time_per_iteration": 2.8278870582580566 + }, + { + "auxiliary_loss_clip": 0.01152177, + "auxiliary_loss_mlp": 0.01037723, + "balance_loss_clip": 1.05077529, + "balance_loss_mlp": 1.02744102, + "epoch": 0.32657969097577105, + "flos": 21032413361280.0, + "grad_norm": 1.8549196044470233, + "language_loss": 0.84963936, + "learning_rate": 3.146266460153554e-06, + "loss": 0.87153834, + "num_input_tokens_seen": 58523885, + "step": 2716, + "time_per_iteration": 3.0020110607147217 + }, + { + "auxiliary_loss_clip": 0.01175745, + "auxiliary_loss_mlp": 0.00888292, + "balance_loss_clip": 1.05280352, + "balance_loss_mlp": 1.00064397, + "epoch": 0.32669993386641016, + "flos": 22710267509760.0, + "grad_norm": 1.8229472016170107, + "language_loss": 0.8014009, + "learning_rate": 3.145628035404404e-06, + "loss": 0.82204127, + "num_input_tokens_seen": 58543085, + "step": 2717, + "time_per_iteration": 2.8599467277526855 + }, + { + "auxiliary_loss_clip": 0.01077037, + "auxiliary_loss_mlp": 0.01006038, + "balance_loss_clip": 1.01613891, + "balance_loss_mlp": 1.00359392, + "epoch": 0.3268201767570492, + "flos": 72105718406400.0, + "grad_norm": 0.8540031728220171, + "language_loss": 0.57435828, + "learning_rate": 3.1449894368643922e-06, + "loss": 0.59518903, + "num_input_tokens_seen": 58605400, + "step": 2718, + "time_per_iteration": 3.468944787979126 + }, + { + "auxiliary_loss_clip": 0.01169047, + "auxiliary_loss_mlp": 0.01040733, + "balance_loss_clip": 1.05644703, + "balance_loss_mlp": 1.03115499, + "epoch": 0.32694041964768833, + "flos": 24535175938560.0, + "grad_norm": 1.514245596033196, + "language_loss": 0.71656191, + "learning_rate": 3.1443506646303934e-06, + "loss": 0.73865974, + "num_input_tokens_seen": 58626700, + "step": 2719, + "time_per_iteration": 3.7926011085510254 + }, + { + "auxiliary_loss_clip": 0.01188349, + "auxiliary_loss_mlp": 0.01032057, + "balance_loss_clip": 1.05335188, + "balance_loss_mlp": 1.02176332, + "epoch": 0.32706066253832744, + "flos": 33183003755520.0, + "grad_norm": 1.9144114618232944, + "language_loss": 0.66684443, + "learning_rate": 3.1437117187993086e-06, + "loss": 0.68904847, + "num_input_tokens_seen": 58649020, + "step": 2720, + "time_per_iteration": 3.713992118835449 + }, + { + "auxiliary_loss_clip": 0.01161226, + "auxiliary_loss_mlp": 0.01035309, + "balance_loss_clip": 1.04868746, + "balance_loss_mlp": 1.02566504, + "epoch": 0.3271809054289665, + "flos": 24061622008320.0, + "grad_norm": 1.4727380043724019, + "language_loss": 0.80095565, + "learning_rate": 3.143072599468065e-06, + "loss": 0.82292098, + "num_input_tokens_seen": 58668845, + "step": 2721, + "time_per_iteration": 3.7783424854278564 + }, + { + "auxiliary_loss_clip": 0.01177293, + "auxiliary_loss_mlp": 0.01037846, + "balance_loss_clip": 1.05584764, + "balance_loss_mlp": 1.02800512, + "epoch": 0.3273011483196056, + "flos": 38253769712640.0, + "grad_norm": 1.561130019755049, + "language_loss": 0.75704849, + "learning_rate": 3.1424333067336174e-06, + "loss": 0.77919984, + "num_input_tokens_seen": 58691610, + "step": 2722, + "time_per_iteration": 2.9390370845794678 + }, + { + "auxiliary_loss_clip": 0.01192306, + "auxiliary_loss_mlp": 0.01039011, + "balance_loss_clip": 1.05399919, + "balance_loss_mlp": 1.0294503, + "epoch": 0.3274213912102447, + "flos": 29054389582080.0, + "grad_norm": 1.7125238504573637, + "language_loss": 0.78289044, + "learning_rate": 3.141793840692945e-06, + "loss": 0.80520368, + "num_input_tokens_seen": 58712360, + "step": 2723, + "time_per_iteration": 2.924943208694458 + }, + { + "auxiliary_loss_clip": 0.0116943, + "auxiliary_loss_mlp": 0.01033966, + "balance_loss_clip": 1.05171156, + "balance_loss_mlp": 1.02419686, + "epoch": 0.32754163410088377, + "flos": 29133249891840.0, + "grad_norm": 1.9197532505139943, + "language_loss": 0.61465621, + "learning_rate": 3.1411542014430553e-06, + "loss": 0.63669014, + "num_input_tokens_seen": 58733440, + "step": 2724, + "time_per_iteration": 2.8476502895355225 + }, + { + "auxiliary_loss_clip": 0.01162999, + "auxiliary_loss_mlp": 0.01034939, + "balance_loss_clip": 1.04557967, + "balance_loss_mlp": 1.02578926, + "epoch": 0.3276618769915229, + "flos": 20631075724800.0, + "grad_norm": 1.7111767407012455, + "language_loss": 0.82105595, + "learning_rate": 3.1405143890809804e-06, + "loss": 0.84303534, + "num_input_tokens_seen": 58752735, + "step": 2725, + "time_per_iteration": 2.8801276683807373 + }, + { + "auxiliary_loss_clip": 0.01174065, + "auxiliary_loss_mlp": 0.01031667, + "balance_loss_clip": 1.05334663, + "balance_loss_mlp": 1.02209425, + "epoch": 0.327782119882162, + "flos": 18657425076480.0, + "grad_norm": 2.1625507319050192, + "language_loss": 0.7003125, + "learning_rate": 3.1398744037037796e-06, + "loss": 0.72236979, + "num_input_tokens_seen": 58772070, + "step": 2726, + "time_per_iteration": 2.8265063762664795 + }, + { + "auxiliary_loss_clip": 0.01177275, + "auxiliary_loss_mlp": 0.01032733, + "balance_loss_clip": 1.05398846, + "balance_loss_mlp": 1.02255809, + "epoch": 0.32790236277280105, + "flos": 21795802133760.0, + "grad_norm": 2.5455923543640733, + "language_loss": 0.8436833, + "learning_rate": 3.139234245408538e-06, + "loss": 0.86578339, + "num_input_tokens_seen": 58790950, + "step": 2727, + "time_per_iteration": 2.7859346866607666 + }, + { + "auxiliary_loss_clip": 0.01169189, + "auxiliary_loss_mlp": 0.0088703, + "balance_loss_clip": 1.05325842, + "balance_loss_mlp": 1.00059402, + "epoch": 0.32802260566344016, + "flos": 23331414424320.0, + "grad_norm": 1.3118925486754698, + "language_loss": 0.76315439, + "learning_rate": 3.1385939142923666e-06, + "loss": 0.7837165, + "num_input_tokens_seen": 58813340, + "step": 2728, + "time_per_iteration": 2.8419995307922363 + }, + { + "auxiliary_loss_clip": 0.01178181, + "auxiliary_loss_mlp": 0.01032499, + "balance_loss_clip": 1.05136442, + "balance_loss_mlp": 1.02275372, + "epoch": 0.3281428485540792, + "flos": 24206988349440.0, + "grad_norm": 2.402762071909289, + "language_loss": 0.78553188, + "learning_rate": 3.137953410452405e-06, + "loss": 0.80763876, + "num_input_tokens_seen": 58833610, + "step": 2729, + "time_per_iteration": 2.831181526184082 + }, + { + "auxiliary_loss_clip": 0.01171666, + "auxiliary_loss_mlp": 0.01032869, + "balance_loss_clip": 1.04923725, + "balance_loss_mlp": 1.02302825, + "epoch": 0.3282630914447183, + "flos": 34128962380800.0, + "grad_norm": 1.5324000489087088, + "language_loss": 0.74778599, + "learning_rate": 3.1373127339858146e-06, + "loss": 0.76983136, + "num_input_tokens_seen": 58856210, + "step": 2730, + "time_per_iteration": 3.0563175678253174 + }, + { + "auxiliary_loss_clip": 0.01163673, + "auxiliary_loss_mlp": 0.01042728, + "balance_loss_clip": 1.05071306, + "balance_loss_mlp": 1.0336082, + "epoch": 0.32838333433535744, + "flos": 27600726170880.0, + "grad_norm": 2.0375673385313133, + "language_loss": 0.74854946, + "learning_rate": 3.136671884989787e-06, + "loss": 0.77061349, + "num_input_tokens_seen": 58876120, + "step": 2731, + "time_per_iteration": 2.9733145236968994 + }, + { + "auxiliary_loss_clip": 0.01150371, + "auxiliary_loss_mlp": 0.01037029, + "balance_loss_clip": 1.04962504, + "balance_loss_mlp": 1.0269562, + "epoch": 0.3285035772259965, + "flos": 12349500935040.0, + "grad_norm": 3.8969467652076974, + "language_loss": 0.87310857, + "learning_rate": 3.1360308635615383e-06, + "loss": 0.89498258, + "num_input_tokens_seen": 58894660, + "step": 2732, + "time_per_iteration": 2.9672162532806396 + }, + { + "auxiliary_loss_clip": 0.01186045, + "auxiliary_loss_mlp": 0.01036717, + "balance_loss_clip": 1.05497384, + "balance_loss_mlp": 1.02638745, + "epoch": 0.3286238201166356, + "flos": 24316084932480.0, + "grad_norm": 2.0775910766153176, + "language_loss": 0.78785819, + "learning_rate": 3.135389669798311e-06, + "loss": 0.81008577, + "num_input_tokens_seen": 58912720, + "step": 2733, + "time_per_iteration": 2.782071113586426 + }, + { + "auxiliary_loss_clip": 0.01187402, + "auxiliary_loss_mlp": 0.00887233, + "balance_loss_clip": 1.05431402, + "balance_loss_mlp": 1.00053036, + "epoch": 0.3287440630072747, + "flos": 21392812471680.0, + "grad_norm": 1.963347170607713, + "language_loss": 0.80092776, + "learning_rate": 3.134748303797373e-06, + "loss": 0.82167411, + "num_input_tokens_seen": 58930090, + "step": 2734, + "time_per_iteration": 2.7949767112731934 + }, + { + "auxiliary_loss_clip": 0.01156253, + "auxiliary_loss_mlp": 0.01036325, + "balance_loss_clip": 1.0496552, + "balance_loss_mlp": 1.02530456, + "epoch": 0.32886430589791377, + "flos": 23732536579200.0, + "grad_norm": 2.3624521449760407, + "language_loss": 0.80987072, + "learning_rate": 3.1341067656560203e-06, + "loss": 0.83179653, + "num_input_tokens_seen": 58947935, + "step": 2735, + "time_per_iteration": 2.8872318267822266 + }, + { + "auxiliary_loss_clip": 0.01186739, + "auxiliary_loss_mlp": 0.01032745, + "balance_loss_clip": 1.05403948, + "balance_loss_mlp": 1.02273178, + "epoch": 0.3289845487885529, + "flos": 22418708814720.0, + "grad_norm": 2.39367016241699, + "language_loss": 0.86709517, + "learning_rate": 3.133465055471572e-06, + "loss": 0.88928998, + "num_input_tokens_seen": 58967720, + "step": 2736, + "time_per_iteration": 2.8203482627868652 + }, + { + "auxiliary_loss_clip": 0.01161091, + "auxiliary_loss_mlp": 0.01034155, + "balance_loss_clip": 1.05047607, + "balance_loss_mlp": 1.02436781, + "epoch": 0.329104791679192, + "flos": 19682603147520.0, + "grad_norm": 2.2009945672093854, + "language_loss": 0.66171122, + "learning_rate": 3.1328231733413767e-06, + "loss": 0.68366361, + "num_input_tokens_seen": 58984360, + "step": 2737, + "time_per_iteration": 2.921755790710449 + }, + { + "auxiliary_loss_clip": 0.01182988, + "auxiliary_loss_mlp": 0.01038952, + "balance_loss_clip": 1.05370092, + "balance_loss_mlp": 1.02845001, + "epoch": 0.32922503456983104, + "flos": 15997234803840.0, + "grad_norm": 1.9626432523148, + "language_loss": 0.90617073, + "learning_rate": 3.1321811193628067e-06, + "loss": 0.92839015, + "num_input_tokens_seen": 59002505, + "step": 2738, + "time_per_iteration": 2.7418928146362305 + }, + { + "auxiliary_loss_clip": 0.01189558, + "auxiliary_loss_mlp": 0.00888363, + "balance_loss_clip": 1.05673075, + "balance_loss_mlp": 1.00047708, + "epoch": 0.32934527746047015, + "flos": 26834069260800.0, + "grad_norm": 1.7944642371523585, + "language_loss": 0.70181525, + "learning_rate": 3.131538893633261e-06, + "loss": 0.7225945, + "num_input_tokens_seen": 59022065, + "step": 2739, + "time_per_iteration": 2.8034169673919678 + }, + { + "auxiliary_loss_clip": 0.01199748, + "auxiliary_loss_mlp": 0.01036343, + "balance_loss_clip": 1.05834818, + "balance_loss_mlp": 1.02652001, + "epoch": 0.32946552035110926, + "flos": 23403774372480.0, + "grad_norm": 2.0920738102847602, + "language_loss": 0.77985138, + "learning_rate": 3.130896496250165e-06, + "loss": 0.80221224, + "num_input_tokens_seen": 59041890, + "step": 2740, + "time_per_iteration": 2.7477078437805176 + }, + { + "auxiliary_loss_clip": 0.0119681, + "auxiliary_loss_mlp": 0.01035906, + "balance_loss_clip": 1.05421913, + "balance_loss_mlp": 1.02598178, + "epoch": 0.3295857632417483, + "flos": 14172470029440.0, + "grad_norm": 2.0006998168866286, + "language_loss": 0.86961728, + "learning_rate": 3.1302539273109693e-06, + "loss": 0.89194447, + "num_input_tokens_seen": 59058715, + "step": 2741, + "time_per_iteration": 2.6452386379241943 + }, + { + "auxiliary_loss_clip": 0.01172817, + "auxiliary_loss_mlp": 0.0104361, + "balance_loss_clip": 1.0533936, + "balance_loss_mlp": 1.03323913, + "epoch": 0.32970600613238743, + "flos": 22196708807040.0, + "grad_norm": 1.6538126648050973, + "language_loss": 0.80563265, + "learning_rate": 3.1296111869131513e-06, + "loss": 0.82779694, + "num_input_tokens_seen": 59076140, + "step": 2742, + "time_per_iteration": 2.7769997119903564 + }, + { + "auxiliary_loss_clip": 0.01195235, + "auxiliary_loss_mlp": 0.0103801, + "balance_loss_clip": 1.05441952, + "balance_loss_mlp": 1.0279429, + "epoch": 0.32982624902302654, + "flos": 22053784590720.0, + "grad_norm": 3.578733311337422, + "language_loss": 0.85745049, + "learning_rate": 3.1289682751542153e-06, + "loss": 0.87978292, + "num_input_tokens_seen": 59095700, + "step": 2743, + "time_per_iteration": 2.6927120685577393 + }, + { + "auxiliary_loss_clip": 0.0118397, + "auxiliary_loss_mlp": 0.01033728, + "balance_loss_clip": 1.05222964, + "balance_loss_mlp": 1.02457845, + "epoch": 0.3299464919136656, + "flos": 18661626967680.0, + "grad_norm": 1.7970306494352568, + "language_loss": 0.71029413, + "learning_rate": 3.1283251921316883e-06, + "loss": 0.73247111, + "num_input_tokens_seen": 59113445, + "step": 2744, + "time_per_iteration": 2.7302699089050293 + }, + { + "auxiliary_loss_clip": 0.011535, + "auxiliary_loss_mlp": 0.01032146, + "balance_loss_clip": 1.04961348, + "balance_loss_mlp": 1.02253723, + "epoch": 0.3300667348043047, + "flos": 13407357404160.0, + "grad_norm": 1.9605070116568077, + "language_loss": 0.8091017, + "learning_rate": 3.1276819379431277e-06, + "loss": 0.83095813, + "num_input_tokens_seen": 59131535, + "step": 2745, + "time_per_iteration": 3.8550052642822266 + }, + { + "auxiliary_loss_clip": 0.01187269, + "auxiliary_loss_mlp": 0.0088777, + "balance_loss_clip": 1.05524254, + "balance_loss_mlp": 1.00067043, + "epoch": 0.33018697769494376, + "flos": 15742556398080.0, + "grad_norm": 2.2124412959595112, + "language_loss": 0.7526232, + "learning_rate": 3.1270385126861134e-06, + "loss": 0.7733736, + "num_input_tokens_seen": 59149520, + "step": 2746, + "time_per_iteration": 2.790252447128296 + }, + { + "auxiliary_loss_clip": 0.01200891, + "auxiliary_loss_mlp": 0.01033875, + "balance_loss_clip": 1.05736542, + "balance_loss_mlp": 1.02367711, + "epoch": 0.3303072205855829, + "flos": 18258601392000.0, + "grad_norm": 1.699017108541932, + "language_loss": 0.81883758, + "learning_rate": 3.1263949164582533e-06, + "loss": 0.84118521, + "num_input_tokens_seen": 59169170, + "step": 2747, + "time_per_iteration": 3.6024177074432373 + }, + { + "auxiliary_loss_clip": 0.01196874, + "auxiliary_loss_mlp": 0.0103421, + "balance_loss_clip": 1.05423379, + "balance_loss_mlp": 1.02410102, + "epoch": 0.330427463476222, + "flos": 17749424148480.0, + "grad_norm": 2.5299888547386264, + "language_loss": 0.77993667, + "learning_rate": 3.1257511493571797e-06, + "loss": 0.80224752, + "num_input_tokens_seen": 59187675, + "step": 2748, + "time_per_iteration": 2.7551538944244385 + }, + { + "auxiliary_loss_clip": 0.0116728, + "auxiliary_loss_mlp": 0.010342, + "balance_loss_clip": 1.05084991, + "balance_loss_mlp": 1.0250504, + "epoch": 0.33054770636686104, + "flos": 27162580072320.0, + "grad_norm": 1.7275401279659195, + "language_loss": 0.78279841, + "learning_rate": 3.125107211480552e-06, + "loss": 0.80481321, + "num_input_tokens_seen": 59207610, + "step": 2749, + "time_per_iteration": 2.9180405139923096 + }, + { + "auxiliary_loss_clip": 0.01144814, + "auxiliary_loss_mlp": 0.01032438, + "balance_loss_clip": 1.04740262, + "balance_loss_mlp": 1.02293742, + "epoch": 0.33066794925750015, + "flos": 20117193799680.0, + "grad_norm": 1.6297524555594514, + "language_loss": 0.79996228, + "learning_rate": 3.124463102926054e-06, + "loss": 0.82173485, + "num_input_tokens_seen": 59226945, + "step": 2750, + "time_per_iteration": 2.7908775806427 + }, + { + "auxiliary_loss_clip": 0.01077989, + "auxiliary_loss_mlp": 0.01008744, + "balance_loss_clip": 1.02101827, + "balance_loss_mlp": 1.00625229, + "epoch": 0.33078819214813926, + "flos": 70642609718400.0, + "grad_norm": 0.7838102469556554, + "language_loss": 0.61516047, + "learning_rate": 3.1238188237913984e-06, + "loss": 0.63602781, + "num_input_tokens_seen": 59291485, + "step": 2751, + "time_per_iteration": 3.3665688037872314 + }, + { + "auxiliary_loss_clip": 0.01204724, + "auxiliary_loss_mlp": 0.01034538, + "balance_loss_clip": 1.05873168, + "balance_loss_mlp": 1.02438748, + "epoch": 0.3309084350387783, + "flos": 21141940907520.0, + "grad_norm": 1.9575974196311459, + "language_loss": 0.76239252, + "learning_rate": 3.1231743741743202e-06, + "loss": 0.78478515, + "num_input_tokens_seen": 59310990, + "step": 2752, + "time_per_iteration": 2.7871856689453125 + }, + { + "auxiliary_loss_clip": 0.01183205, + "auxiliary_loss_mlp": 0.01039487, + "balance_loss_clip": 1.05165339, + "balance_loss_mlp": 1.02902615, + "epoch": 0.3310286779294174, + "flos": 14209350318720.0, + "grad_norm": 2.3040605778577232, + "language_loss": 0.83469534, + "learning_rate": 3.122529754172582e-06, + "loss": 0.85692221, + "num_input_tokens_seen": 59327875, + "step": 2753, + "time_per_iteration": 2.699115753173828 + }, + { + "auxiliary_loss_clip": 0.01189402, + "auxiliary_loss_mlp": 0.01033419, + "balance_loss_clip": 1.0559845, + "balance_loss_mlp": 1.02280307, + "epoch": 0.33114892082005654, + "flos": 20778130005120.0, + "grad_norm": 3.063801837071909, + "language_loss": 0.722197, + "learning_rate": 3.1218849638839736e-06, + "loss": 0.74442518, + "num_input_tokens_seen": 59347135, + "step": 2754, + "time_per_iteration": 2.848964214324951 + }, + { + "auxiliary_loss_clip": 0.01153645, + "auxiliary_loss_mlp": 0.01044461, + "balance_loss_clip": 1.04422402, + "balance_loss_mlp": 1.03419709, + "epoch": 0.3312691637106956, + "flos": 17090750499840.0, + "grad_norm": 1.716091820962325, + "language_loss": 0.78606474, + "learning_rate": 3.121240003406307e-06, + "loss": 0.80804574, + "num_input_tokens_seen": 59365985, + "step": 2755, + "time_per_iteration": 2.7774291038513184 + }, + { + "auxiliary_loss_clip": 0.0117274, + "auxiliary_loss_mlp": 0.01036461, + "balance_loss_clip": 1.05354965, + "balance_loss_mlp": 1.02631021, + "epoch": 0.3313894066013347, + "flos": 29456230008960.0, + "grad_norm": 2.1712826097414664, + "language_loss": 0.72563869, + "learning_rate": 3.120594872837425e-06, + "loss": 0.74773067, + "num_input_tokens_seen": 59384655, + "step": 2756, + "time_per_iteration": 2.895941972732544 + }, + { + "auxiliary_loss_clip": 0.01079773, + "auxiliary_loss_mlp": 0.00876737, + "balance_loss_clip": 1.01944804, + "balance_loss_mlp": 1.00099206, + "epoch": 0.3315096494919738, + "flos": 61419242280960.0, + "grad_norm": 0.8989708376623503, + "language_loss": 0.62395972, + "learning_rate": 3.1199495722751906e-06, + "loss": 0.64352483, + "num_input_tokens_seen": 59444185, + "step": 2757, + "time_per_iteration": 3.2901461124420166 + }, + { + "auxiliary_loss_clip": 0.01159238, + "auxiliary_loss_mlp": 0.01035293, + "balance_loss_clip": 1.04868197, + "balance_loss_mlp": 1.0258162, + "epoch": 0.33162989238261287, + "flos": 21653057485440.0, + "grad_norm": 1.7149437595358603, + "language_loss": 0.83934706, + "learning_rate": 3.1193041018174972e-06, + "loss": 0.86129236, + "num_input_tokens_seen": 59464900, + "step": 2758, + "time_per_iteration": 2.926406145095825 + }, + { + "auxiliary_loss_clip": 0.01192175, + "auxiliary_loss_mlp": 0.01037147, + "balance_loss_clip": 1.05723953, + "balance_loss_mlp": 1.02745581, + "epoch": 0.331750135273252, + "flos": 22674787850880.0, + "grad_norm": 2.0058776271589607, + "language_loss": 0.94860327, + "learning_rate": 3.118658461562261e-06, + "loss": 0.97089648, + "num_input_tokens_seen": 59481000, + "step": 2759, + "time_per_iteration": 2.7715256214141846 + }, + { + "auxiliary_loss_clip": 0.01179554, + "auxiliary_loss_mlp": 0.01035178, + "balance_loss_clip": 1.05702829, + "balance_loss_mlp": 1.02545071, + "epoch": 0.33187037816389103, + "flos": 22746896403840.0, + "grad_norm": 1.6852267386759203, + "language_loss": 0.85004818, + "learning_rate": 3.118012651607426e-06, + "loss": 0.8721956, + "num_input_tokens_seen": 59502605, + "step": 2760, + "time_per_iteration": 2.893397331237793 + }, + { + "auxiliary_loss_clip": 0.01198633, + "auxiliary_loss_mlp": 0.01044974, + "balance_loss_clip": 1.05642498, + "balance_loss_mlp": 1.03504407, + "epoch": 0.33199062105453014, + "flos": 19203769918080.0, + "grad_norm": 2.1263103747248517, + "language_loss": 0.83726537, + "learning_rate": 3.1173666720509603e-06, + "loss": 0.85970145, + "num_input_tokens_seen": 59519540, + "step": 2761, + "time_per_iteration": 2.7866768836975098 + }, + { + "auxiliary_loss_clip": 0.0118216, + "auxiliary_loss_mlp": 0.01038999, + "balance_loss_clip": 1.05384004, + "balance_loss_mlp": 1.0289979, + "epoch": 0.33211086394516925, + "flos": 31577006764800.0, + "grad_norm": 1.735919130796087, + "language_loss": 0.67929268, + "learning_rate": 3.116720522990859e-06, + "loss": 0.70150423, + "num_input_tokens_seen": 59540415, + "step": 2762, + "time_per_iteration": 2.816880702972412 + }, + { + "auxiliary_loss_clip": 0.01149417, + "auxiliary_loss_mlp": 0.01038816, + "balance_loss_clip": 1.04944003, + "balance_loss_mlp": 1.02841449, + "epoch": 0.3322311068358083, + "flos": 17932496791680.0, + "grad_norm": 1.8458378677066083, + "language_loss": 0.61872768, + "learning_rate": 3.116074204525142e-06, + "loss": 0.64060998, + "num_input_tokens_seen": 59558590, + "step": 2763, + "time_per_iteration": 2.982665777206421 + }, + { + "auxiliary_loss_clip": 0.01180973, + "auxiliary_loss_mlp": 0.01035221, + "balance_loss_clip": 1.05404115, + "balance_loss_mlp": 1.02471256, + "epoch": 0.3323513497264474, + "flos": 32269831269120.0, + "grad_norm": 2.2493930315425614, + "language_loss": 0.83624434, + "learning_rate": 3.1154277167518553e-06, + "loss": 0.85840631, + "num_input_tokens_seen": 59580205, + "step": 2764, + "time_per_iteration": 2.8246684074401855 + }, + { + "auxiliary_loss_clip": 0.01067553, + "auxiliary_loss_mlp": 0.01011903, + "balance_loss_clip": 1.01554513, + "balance_loss_mlp": 1.00931668, + "epoch": 0.33247159261708653, + "flos": 52668674588160.0, + "grad_norm": 0.7871559844087673, + "language_loss": 0.59463239, + "learning_rate": 3.114781059769072e-06, + "loss": 0.61542702, + "num_input_tokens_seen": 59631530, + "step": 2765, + "time_per_iteration": 3.1413466930389404 + }, + { + "auxiliary_loss_clip": 0.01175499, + "auxiliary_loss_mlp": 0.01040484, + "balance_loss_clip": 1.05283237, + "balance_loss_mlp": 1.03052413, + "epoch": 0.3325918355077256, + "flos": 27125232906240.0, + "grad_norm": 2.664756907882488, + "language_loss": 0.6818046, + "learning_rate": 3.1141342336748874e-06, + "loss": 0.70396447, + "num_input_tokens_seen": 59651090, + "step": 2766, + "time_per_iteration": 2.8163304328918457 + }, + { + "auxiliary_loss_clip": 0.01180988, + "auxiliary_loss_mlp": 0.01035398, + "balance_loss_clip": 1.05277669, + "balance_loss_mlp": 1.02703512, + "epoch": 0.3327120783983647, + "flos": 23664414435840.0, + "grad_norm": 1.441582279689662, + "language_loss": 0.82040167, + "learning_rate": 3.1134872385674253e-06, + "loss": 0.84256554, + "num_input_tokens_seen": 59675245, + "step": 2767, + "time_per_iteration": 2.7297518253326416 + }, + { + "auxiliary_loss_clip": 0.01180312, + "auxiliary_loss_mlp": 0.01032868, + "balance_loss_clip": 1.04951882, + "balance_loss_mlp": 1.0232594, + "epoch": 0.3328323212890038, + "flos": 19171378828800.0, + "grad_norm": 1.7673430586158159, + "language_loss": 0.85832161, + "learning_rate": 3.1128400745448353e-06, + "loss": 0.88045341, + "num_input_tokens_seen": 59694625, + "step": 2768, + "time_per_iteration": 2.821606397628784 + }, + { + "auxiliary_loss_clip": 0.01190952, + "auxiliary_loss_mlp": 0.01036656, + "balance_loss_clip": 1.05503249, + "balance_loss_mlp": 1.02685738, + "epoch": 0.33295256417964286, + "flos": 37706347463040.0, + "grad_norm": 2.091491587914513, + "language_loss": 0.62783182, + "learning_rate": 3.11219274170529e-06, + "loss": 0.65010792, + "num_input_tokens_seen": 59716435, + "step": 2769, + "time_per_iteration": 2.8691718578338623 + }, + { + "auxiliary_loss_clip": 0.01170325, + "auxiliary_loss_mlp": 0.01030952, + "balance_loss_clip": 1.0511843, + "balance_loss_mlp": 1.02173078, + "epoch": 0.333072807070282, + "flos": 26505989412480.0, + "grad_norm": 2.3310333989772127, + "language_loss": 0.81850779, + "learning_rate": 3.1115452401469903e-06, + "loss": 0.8405205, + "num_input_tokens_seen": 59736835, + "step": 2770, + "time_per_iteration": 4.982961893081665 + }, + { + "auxiliary_loss_clip": 0.01148013, + "auxiliary_loss_mlp": 0.01040643, + "balance_loss_clip": 1.04567039, + "balance_loss_mlp": 1.03168464, + "epoch": 0.3331930499609211, + "flos": 21430913823360.0, + "grad_norm": 1.7701285996049738, + "language_loss": 0.86349273, + "learning_rate": 3.1108975699681613e-06, + "loss": 0.88537931, + "num_input_tokens_seen": 59754230, + "step": 2771, + "time_per_iteration": 3.8135313987731934 + }, + { + "auxiliary_loss_clip": 0.01163959, + "auxiliary_loss_mlp": 0.010341, + "balance_loss_clip": 1.0518105, + "balance_loss_mlp": 1.02470076, + "epoch": 0.33331329285156014, + "flos": 20659947281280.0, + "grad_norm": 1.7574246919273573, + "language_loss": 0.72066236, + "learning_rate": 3.1102497312670542e-06, + "loss": 0.74264294, + "num_input_tokens_seen": 59772235, + "step": 2772, + "time_per_iteration": 2.86429762840271 + }, + { + "auxiliary_loss_clip": 0.01163364, + "auxiliary_loss_mlp": 0.01034382, + "balance_loss_clip": 1.05006945, + "balance_loss_mlp": 1.02579355, + "epoch": 0.33343353574219925, + "flos": 28001596930560.0, + "grad_norm": 3.2416511171705875, + "language_loss": 0.80316341, + "learning_rate": 3.109601724141946e-06, + "loss": 0.82514083, + "num_input_tokens_seen": 59791230, + "step": 2773, + "time_per_iteration": 3.7937042713165283 + }, + { + "auxiliary_loss_clip": 0.01171483, + "auxiliary_loss_mlp": 0.01033566, + "balance_loss_clip": 1.05144572, + "balance_loss_mlp": 1.02391624, + "epoch": 0.33355377863283836, + "flos": 23764963582080.0, + "grad_norm": 1.6770974405293295, + "language_loss": 0.6804499, + "learning_rate": 3.108953548691138e-06, + "loss": 0.7025004, + "num_input_tokens_seen": 59811315, + "step": 2774, + "time_per_iteration": 2.803434133529663 + }, + { + "auxiliary_loss_clip": 0.01198801, + "auxiliary_loss_mlp": 0.01039593, + "balance_loss_clip": 1.05660558, + "balance_loss_mlp": 1.02962708, + "epoch": 0.3336740215234774, + "flos": 37779677078400.0, + "grad_norm": 2.1003870293513547, + "language_loss": 0.72537178, + "learning_rate": 3.108305205012959e-06, + "loss": 0.74775571, + "num_input_tokens_seen": 59832010, + "step": 2775, + "time_per_iteration": 2.77980375289917 + }, + { + "auxiliary_loss_clip": 0.01177532, + "auxiliary_loss_mlp": 0.01033953, + "balance_loss_clip": 1.0554142, + "balance_loss_mlp": 1.02476168, + "epoch": 0.3337942644141165, + "flos": 25519056347520.0, + "grad_norm": 1.8723457001198642, + "language_loss": 0.87081873, + "learning_rate": 3.107656693205761e-06, + "loss": 0.89293355, + "num_input_tokens_seen": 59851450, + "step": 2776, + "time_per_iteration": 2.8362858295440674 + }, + { + "auxiliary_loss_clip": 0.01201328, + "auxiliary_loss_mlp": 0.01037165, + "balance_loss_clip": 1.0570116, + "balance_loss_mlp": 1.02614415, + "epoch": 0.3339145073047556, + "flos": 25989844930560.0, + "grad_norm": 2.5061184922081807, + "language_loss": 0.70523328, + "learning_rate": 3.107008013367924e-06, + "loss": 0.72761822, + "num_input_tokens_seen": 59870245, + "step": 2777, + "time_per_iteration": 2.792036294937134 + }, + { + "auxiliary_loss_clip": 0.011638, + "auxiliary_loss_mlp": 0.01037364, + "balance_loss_clip": 1.04928613, + "balance_loss_mlp": 1.0275352, + "epoch": 0.3340347501953947, + "flos": 19062569554560.0, + "grad_norm": 2.688505142941716, + "language_loss": 0.86660111, + "learning_rate": 3.1063591655978507e-06, + "loss": 0.88861275, + "num_input_tokens_seen": 59886195, + "step": 2778, + "time_per_iteration": 2.8429064750671387 + }, + { + "auxiliary_loss_clip": 0.01143018, + "auxiliary_loss_mlp": 0.01028515, + "balance_loss_clip": 1.04442608, + "balance_loss_mlp": 1.01868057, + "epoch": 0.3341549930860338, + "flos": 18109715518080.0, + "grad_norm": 1.8002645985693182, + "language_loss": 0.79758435, + "learning_rate": 3.105710149993972e-06, + "loss": 0.8192997, + "num_input_tokens_seen": 59905525, + "step": 2779, + "time_per_iteration": 2.9319331645965576 + }, + { + "auxiliary_loss_clip": 0.01197064, + "auxiliary_loss_mlp": 0.0103349, + "balance_loss_clip": 1.05555379, + "balance_loss_mlp": 1.02342248, + "epoch": 0.33427523597667286, + "flos": 22674967418880.0, + "grad_norm": 1.7166459198749, + "language_loss": 0.85386145, + "learning_rate": 3.1050609666547427e-06, + "loss": 0.87616694, + "num_input_tokens_seen": 59925085, + "step": 2780, + "time_per_iteration": 2.7720813751220703 + }, + { + "auxiliary_loss_clip": 0.01173488, + "auxiliary_loss_mlp": 0.01036309, + "balance_loss_clip": 1.05358446, + "balance_loss_mlp": 1.0262363, + "epoch": 0.33439547886731197, + "flos": 22638338524800.0, + "grad_norm": 2.047317264239943, + "language_loss": 0.77144527, + "learning_rate": 3.104411615678644e-06, + "loss": 0.79354322, + "num_input_tokens_seen": 59943935, + "step": 2781, + "time_per_iteration": 2.92915415763855 + }, + { + "auxiliary_loss_clip": 0.01169086, + "auxiliary_loss_mlp": 0.01031754, + "balance_loss_clip": 1.04983878, + "balance_loss_mlp": 1.02171028, + "epoch": 0.3345157217579511, + "flos": 24096383395200.0, + "grad_norm": 5.63783247355776, + "language_loss": 0.73598814, + "learning_rate": 3.1037620971641803e-06, + "loss": 0.75799656, + "num_input_tokens_seen": 59963725, + "step": 2782, + "time_per_iteration": 2.7963736057281494 + }, + { + "auxiliary_loss_clip": 0.01198587, + "auxiliary_loss_mlp": 0.0102871, + "balance_loss_clip": 1.05551147, + "balance_loss_mlp": 1.01798737, + "epoch": 0.33463596464859013, + "flos": 18989491334400.0, + "grad_norm": 2.5177810988712332, + "language_loss": 0.64825523, + "learning_rate": 3.1031124112098844e-06, + "loss": 0.67052823, + "num_input_tokens_seen": 59981935, + "step": 2783, + "time_per_iteration": 2.7265303134918213 + }, + { + "auxiliary_loss_clip": 0.01180128, + "auxiliary_loss_mlp": 0.01038577, + "balance_loss_clip": 1.0536648, + "balance_loss_mlp": 1.02913606, + "epoch": 0.33475620753922924, + "flos": 20375607219840.0, + "grad_norm": 1.9485964841336207, + "language_loss": 0.72005767, + "learning_rate": 3.1024625579143127e-06, + "loss": 0.74224472, + "num_input_tokens_seen": 59999455, + "step": 2784, + "time_per_iteration": 2.8737473487854004 + }, + { + "auxiliary_loss_clip": 0.01196076, + "auxiliary_loss_mlp": 0.01032807, + "balance_loss_clip": 1.05494964, + "balance_loss_mlp": 1.02267992, + "epoch": 0.33487645042986836, + "flos": 18182578256640.0, + "grad_norm": 2.484288675535159, + "language_loss": 0.72979581, + "learning_rate": 3.101812537376048e-06, + "loss": 0.75208467, + "num_input_tokens_seen": 60018475, + "step": 2785, + "time_per_iteration": 2.706300973892212 + }, + { + "auxiliary_loss_clip": 0.01166712, + "auxiliary_loss_mlp": 0.00887795, + "balance_loss_clip": 1.04900384, + "balance_loss_mlp": 1.00079989, + "epoch": 0.3349966933205074, + "flos": 25848824135040.0, + "grad_norm": 2.013549558636616, + "language_loss": 0.84262621, + "learning_rate": 3.1011623496936973e-06, + "loss": 0.86317122, + "num_input_tokens_seen": 60036770, + "step": 2786, + "time_per_iteration": 2.9188270568847656 + }, + { + "auxiliary_loss_clip": 0.01195887, + "auxiliary_loss_mlp": 0.01032449, + "balance_loss_clip": 1.055902, + "balance_loss_mlp": 1.02270317, + "epoch": 0.3351169362111465, + "flos": 28111447699200.0, + "grad_norm": 1.7644294071062552, + "language_loss": 0.69488418, + "learning_rate": 3.100511994965893e-06, + "loss": 0.7171675, + "num_input_tokens_seen": 60056725, + "step": 2787, + "time_per_iteration": 2.7043097019195557 + }, + { + "auxiliary_loss_clip": 0.01180132, + "auxiliary_loss_mlp": 0.01036362, + "balance_loss_clip": 1.05170727, + "balance_loss_mlp": 1.02692616, + "epoch": 0.33523717910178563, + "flos": 22673315393280.0, + "grad_norm": 1.6820292613262637, + "language_loss": 0.84519386, + "learning_rate": 3.0998614732912947e-06, + "loss": 0.86735874, + "num_input_tokens_seen": 60076100, + "step": 2788, + "time_per_iteration": 2.7046799659729004 + }, + { + "auxiliary_loss_clip": 0.01185829, + "auxiliary_loss_mlp": 0.01034999, + "balance_loss_clip": 1.05566001, + "balance_loss_mlp": 1.02482998, + "epoch": 0.3353574219924247, + "flos": 15669801400320.0, + "grad_norm": 2.1226066902791265, + "language_loss": 0.68467689, + "learning_rate": 3.0992107847685855e-06, + "loss": 0.70688522, + "num_input_tokens_seen": 60093815, + "step": 2789, + "time_per_iteration": 2.6724331378936768 + }, + { + "auxiliary_loss_clip": 0.01179645, + "auxiliary_loss_mlp": 0.01036617, + "balance_loss_clip": 1.05739737, + "balance_loss_mlp": 1.0266093, + "epoch": 0.3354776648830638, + "flos": 24790644443520.0, + "grad_norm": 1.8500085777314945, + "language_loss": 0.79090905, + "learning_rate": 3.0985599294964736e-06, + "loss": 0.81307167, + "num_input_tokens_seen": 60113370, + "step": 2790, + "time_per_iteration": 2.7802577018737793 + }, + { + "auxiliary_loss_clip": 0.01177537, + "auxiliary_loss_mlp": 0.01039893, + "balance_loss_clip": 1.05073607, + "balance_loss_mlp": 1.03025532, + "epoch": 0.33559790777370285, + "flos": 28694852398080.0, + "grad_norm": 1.851821618381537, + "language_loss": 0.7007522, + "learning_rate": 3.097908907573695e-06, + "loss": 0.7229265, + "num_input_tokens_seen": 60131350, + "step": 2791, + "time_per_iteration": 2.8434441089630127 + }, + { + "auxiliary_loss_clip": 0.01151064, + "auxiliary_loss_mlp": 0.01039347, + "balance_loss_clip": 1.0515244, + "balance_loss_mlp": 1.02981007, + "epoch": 0.33571815066434196, + "flos": 22235779825920.0, + "grad_norm": 1.8258870483880436, + "language_loss": 0.89950705, + "learning_rate": 3.0972577190990067e-06, + "loss": 0.9214111, + "num_input_tokens_seen": 60149830, + "step": 2792, + "time_per_iteration": 2.933647871017456 + }, + { + "auxiliary_loss_clip": 0.01172606, + "auxiliary_loss_mlp": 0.01033796, + "balance_loss_clip": 1.05226946, + "balance_loss_mlp": 1.02423584, + "epoch": 0.3358383935549811, + "flos": 23842279607040.0, + "grad_norm": 1.8092527675807863, + "language_loss": 0.80329227, + "learning_rate": 3.096606364171196e-06, + "loss": 0.8253563, + "num_input_tokens_seen": 60169620, + "step": 2793, + "time_per_iteration": 2.9880080223083496 + }, + { + "auxiliary_loss_clip": 0.01153778, + "auxiliary_loss_mlp": 0.01044047, + "balance_loss_clip": 1.0491333, + "balance_loss_mlp": 1.03388476, + "epoch": 0.33595863644562013, + "flos": 22267308988800.0, + "grad_norm": 2.158791970285542, + "language_loss": 0.84970033, + "learning_rate": 3.0959548428890703e-06, + "loss": 0.87167859, + "num_input_tokens_seen": 60188490, + "step": 2794, + "time_per_iteration": 2.936429023742676 + }, + { + "auxiliary_loss_clip": 0.01185772, + "auxiliary_loss_mlp": 0.01030877, + "balance_loss_clip": 1.05669379, + "balance_loss_mlp": 1.02132225, + "epoch": 0.33607887933625924, + "flos": 20119779578880.0, + "grad_norm": 1.684916867820381, + "language_loss": 0.84314275, + "learning_rate": 3.095303155351468e-06, + "loss": 0.86530924, + "num_input_tokens_seen": 60208695, + "step": 2795, + "time_per_iteration": 2.7764596939086914 + }, + { + "auxiliary_loss_clip": 0.01148806, + "auxiliary_loss_mlp": 0.0103184, + "balance_loss_clip": 1.05055559, + "balance_loss_mlp": 1.02198124, + "epoch": 0.33619912222689835, + "flos": 19318109886720.0, + "grad_norm": 2.7647021718616753, + "language_loss": 0.79591542, + "learning_rate": 3.0946513016572464e-06, + "loss": 0.8177219, + "num_input_tokens_seen": 60227600, + "step": 2796, + "time_per_iteration": 3.8647115230560303 + }, + { + "auxiliary_loss_clip": 0.01187469, + "auxiliary_loss_mlp": 0.01036511, + "balance_loss_clip": 1.05418491, + "balance_loss_mlp": 1.02656913, + "epoch": 0.3363193651175374, + "flos": 16800664262400.0, + "grad_norm": 1.8145840090062322, + "language_loss": 0.7680037, + "learning_rate": 3.0939992819052938e-06, + "loss": 0.79024351, + "num_input_tokens_seen": 60245110, + "step": 2797, + "time_per_iteration": 3.6798760890960693 + }, + { + "auxiliary_loss_clip": 0.01180596, + "auxiliary_loss_mlp": 0.01042495, + "balance_loss_clip": 1.05597448, + "balance_loss_mlp": 1.03226089, + "epoch": 0.3364396080081765, + "flos": 23550289948800.0, + "grad_norm": 1.9944416803227853, + "language_loss": 0.81106031, + "learning_rate": 3.0933470961945193e-06, + "loss": 0.83329129, + "num_input_tokens_seen": 60263405, + "step": 2798, + "time_per_iteration": 2.8132402896881104 + }, + { + "auxiliary_loss_clip": 0.011768, + "auxiliary_loss_mlp": 0.01031214, + "balance_loss_clip": 1.05625224, + "balance_loss_mlp": 1.02217233, + "epoch": 0.3365598508988156, + "flos": 28037902602240.0, + "grad_norm": 1.7175273106698938, + "language_loss": 0.68019587, + "learning_rate": 3.0926947446238597e-06, + "loss": 0.70227599, + "num_input_tokens_seen": 60282975, + "step": 2799, + "time_per_iteration": 3.679548978805542 + }, + { + "auxiliary_loss_clip": 0.01191403, + "auxiliary_loss_mlp": 0.01038488, + "balance_loss_clip": 1.05205131, + "balance_loss_mlp": 1.02709186, + "epoch": 0.3366800937894547, + "flos": 16982767238400.0, + "grad_norm": 1.9945888511922663, + "language_loss": 0.82576102, + "learning_rate": 3.092042227292276e-06, + "loss": 0.84806001, + "num_input_tokens_seen": 60299810, + "step": 2800, + "time_per_iteration": 2.6593005657196045 + }, + { + "auxiliary_loss_clip": 0.01193825, + "auxiliary_loss_mlp": 0.0104809, + "balance_loss_clip": 1.05580497, + "balance_loss_mlp": 1.03905976, + "epoch": 0.3368003366800938, + "flos": 23915321913600.0, + "grad_norm": 2.8620599207622317, + "language_loss": 0.87762475, + "learning_rate": 3.0913895442987557e-06, + "loss": 0.90004385, + "num_input_tokens_seen": 60320775, + "step": 2801, + "time_per_iteration": 2.750929832458496 + }, + { + "auxiliary_loss_clip": 0.011673, + "auxiliary_loss_mlp": 0.00888067, + "balance_loss_clip": 1.05222797, + "balance_loss_mlp": 1.00102091, + "epoch": 0.3369205795707329, + "flos": 24791219061120.0, + "grad_norm": 1.6050169315235605, + "language_loss": 0.85803866, + "learning_rate": 3.090736695742308e-06, + "loss": 0.87859231, + "num_input_tokens_seen": 60341905, + "step": 2802, + "time_per_iteration": 2.763538360595703 + }, + { + "auxiliary_loss_clip": 0.01149988, + "auxiliary_loss_mlp": 0.01033349, + "balance_loss_clip": 1.04895616, + "balance_loss_mlp": 1.02403867, + "epoch": 0.33704082246137196, + "flos": 17931096161280.0, + "grad_norm": 2.379171077655687, + "language_loss": 0.5243125, + "learning_rate": 3.0900836817219713e-06, + "loss": 0.54614592, + "num_input_tokens_seen": 60358335, + "step": 2803, + "time_per_iteration": 2.857938766479492 + }, + { + "auxiliary_loss_clip": 0.01195647, + "auxiliary_loss_mlp": 0.01031191, + "balance_loss_clip": 1.0542028, + "balance_loss_mlp": 1.0211004, + "epoch": 0.33716106535201107, + "flos": 21286517149440.0, + "grad_norm": 1.7798431999291595, + "language_loss": 0.83764184, + "learning_rate": 3.089430502336807e-06, + "loss": 0.85991025, + "num_input_tokens_seen": 60378305, + "step": 2804, + "time_per_iteration": 2.7023346424102783 + }, + { + "auxiliary_loss_clip": 0.01192589, + "auxiliary_loss_mlp": 0.01035427, + "balance_loss_clip": 1.05598772, + "balance_loss_mlp": 1.02512193, + "epoch": 0.3372813082426502, + "flos": 18402962152320.0, + "grad_norm": 2.598381670184453, + "language_loss": 0.90155178, + "learning_rate": 3.088777157685902e-06, + "loss": 0.92383194, + "num_input_tokens_seen": 60393895, + "step": 2805, + "time_per_iteration": 2.7703640460968018 + }, + { + "auxiliary_loss_clip": 0.01175196, + "auxiliary_loss_mlp": 0.01027786, + "balance_loss_clip": 1.05443335, + "balance_loss_mlp": 1.01821971, + "epoch": 0.33740155113328923, + "flos": 17201391367680.0, + "grad_norm": 2.2474719342914766, + "language_loss": 0.85996014, + "learning_rate": 3.088123647868367e-06, + "loss": 0.88198996, + "num_input_tokens_seen": 60410445, + "step": 2806, + "time_per_iteration": 2.7621347904205322 + }, + { + "auxiliary_loss_clip": 0.01191696, + "auxiliary_loss_mlp": 0.01028499, + "balance_loss_clip": 1.05482793, + "balance_loss_mlp": 1.01862812, + "epoch": 0.33752179402392835, + "flos": 29058950609280.0, + "grad_norm": 2.353928955909648, + "language_loss": 0.81772941, + "learning_rate": 3.0874699729833405e-06, + "loss": 0.83993137, + "num_input_tokens_seen": 60431815, + "step": 2807, + "time_per_iteration": 2.7805240154266357 + }, + { + "auxiliary_loss_clip": 0.01171309, + "auxiliary_loss_mlp": 0.01026996, + "balance_loss_clip": 1.05160773, + "balance_loss_mlp": 1.01788211, + "epoch": 0.3376420369145674, + "flos": 25080730680960.0, + "grad_norm": 1.7423704652734497, + "language_loss": 0.79812193, + "learning_rate": 3.086816133129983e-06, + "loss": 0.82010496, + "num_input_tokens_seen": 60452075, + "step": 2808, + "time_per_iteration": 2.8193747997283936 + }, + { + "auxiliary_loss_clip": 0.0119757, + "auxiliary_loss_mlp": 0.0102813, + "balance_loss_clip": 1.05783701, + "balance_loss_mlp": 1.01948142, + "epoch": 0.3377622798052065, + "flos": 27490624007040.0, + "grad_norm": 2.0672605201314562, + "language_loss": 0.76345205, + "learning_rate": 3.0861621284074826e-06, + "loss": 0.78570902, + "num_input_tokens_seen": 60472600, + "step": 2809, + "time_per_iteration": 2.7091174125671387 + }, + { + "auxiliary_loss_clip": 0.01182512, + "auxiliary_loss_mlp": 0.0102928, + "balance_loss_clip": 1.05451608, + "balance_loss_mlp": 1.01999331, + "epoch": 0.3378825226958456, + "flos": 21975211589760.0, + "grad_norm": 1.6792782300608695, + "language_loss": 0.72986776, + "learning_rate": 3.085507958915051e-06, + "loss": 0.75198567, + "num_input_tokens_seen": 60491030, + "step": 2810, + "time_per_iteration": 2.794631004333496 + }, + { + "auxiliary_loss_clip": 0.01175702, + "auxiliary_loss_mlp": 0.01033349, + "balance_loss_clip": 1.05533552, + "balance_loss_mlp": 1.02331114, + "epoch": 0.3380027655864847, + "flos": 42523189200000.0, + "grad_norm": 1.885242417651558, + "language_loss": 0.7135545, + "learning_rate": 3.084853624751925e-06, + "loss": 0.73564506, + "num_input_tokens_seen": 60512615, + "step": 2811, + "time_per_iteration": 2.9417245388031006 + }, + { + "auxiliary_loss_clip": 0.01169305, + "auxiliary_loss_mlp": 0.01036703, + "balance_loss_clip": 1.05384803, + "balance_loss_mlp": 1.02636182, + "epoch": 0.3381230084771238, + "flos": 26725080418560.0, + "grad_norm": 2.5947526674834918, + "language_loss": 0.85789579, + "learning_rate": 3.0841991260173668e-06, + "loss": 0.87995589, + "num_input_tokens_seen": 60532520, + "step": 2812, + "time_per_iteration": 2.885035753250122 + }, + { + "auxiliary_loss_clip": 0.01197895, + "auxiliary_loss_mlp": 0.01033727, + "balance_loss_clip": 1.05627394, + "balance_loss_mlp": 1.02410734, + "epoch": 0.3382432513677629, + "flos": 22710375250560.0, + "grad_norm": 2.461913132836572, + "language_loss": 0.80428046, + "learning_rate": 3.0835444628106634e-06, + "loss": 0.82659668, + "num_input_tokens_seen": 60551500, + "step": 2813, + "time_per_iteration": 2.688716411590576 + }, + { + "auxiliary_loss_clip": 0.01194924, + "auxiliary_loss_mlp": 0.00888452, + "balance_loss_clip": 1.05545664, + "balance_loss_mlp": 1.00075388, + "epoch": 0.33836349425840195, + "flos": 22122409524480.0, + "grad_norm": 1.831559253140231, + "language_loss": 0.827766, + "learning_rate": 3.082889635231126e-06, + "loss": 0.84859979, + "num_input_tokens_seen": 60570160, + "step": 2814, + "time_per_iteration": 2.7549805641174316 + }, + { + "auxiliary_loss_clip": 0.01178903, + "auxiliary_loss_mlp": 0.01037046, + "balance_loss_clip": 1.05187106, + "balance_loss_mlp": 1.02635908, + "epoch": 0.33848373714904106, + "flos": 27308090067840.0, + "grad_norm": 2.2119583537126974, + "language_loss": 0.76711696, + "learning_rate": 3.0822346433780925e-06, + "loss": 0.78927648, + "num_input_tokens_seen": 60590885, + "step": 2815, + "time_per_iteration": 2.818363904953003 + }, + { + "auxiliary_loss_clip": 0.01187906, + "auxiliary_loss_mlp": 0.01035552, + "balance_loss_clip": 1.05310428, + "balance_loss_mlp": 1.02542484, + "epoch": 0.3386039800396802, + "flos": 25848716394240.0, + "grad_norm": 2.1498385027536004, + "language_loss": 0.86850971, + "learning_rate": 3.0815794873509237e-06, + "loss": 0.89074433, + "num_input_tokens_seen": 60609170, + "step": 2816, + "time_per_iteration": 2.7392027378082275 + }, + { + "auxiliary_loss_clip": 0.01194852, + "auxiliary_loss_mlp": 0.01035763, + "balance_loss_clip": 1.0542798, + "balance_loss_mlp": 1.02603543, + "epoch": 0.33872422293031923, + "flos": 18880646146560.0, + "grad_norm": 1.7211706286533144, + "language_loss": 0.72524321, + "learning_rate": 3.0809241672490066e-06, + "loss": 0.7475493, + "num_input_tokens_seen": 60627340, + "step": 2817, + "time_per_iteration": 2.624037504196167 + }, + { + "auxiliary_loss_clip": 0.01178086, + "auxiliary_loss_mlp": 0.01029244, + "balance_loss_clip": 1.05408597, + "balance_loss_mlp": 1.01956367, + "epoch": 0.33884446582095834, + "flos": 23146977064320.0, + "grad_norm": 1.916915927012102, + "language_loss": 0.84996092, + "learning_rate": 3.080268683171753e-06, + "loss": 0.87203419, + "num_input_tokens_seen": 60647630, + "step": 2818, + "time_per_iteration": 2.7547779083251953 + }, + { + "auxiliary_loss_clip": 0.01185909, + "auxiliary_loss_mlp": 0.01037585, + "balance_loss_clip": 1.05151284, + "balance_loss_mlp": 1.02780962, + "epoch": 0.33896470871159745, + "flos": 15997342544640.0, + "grad_norm": 2.072276051171517, + "language_loss": 0.88913715, + "learning_rate": 3.0796130352185985e-06, + "loss": 0.91137207, + "num_input_tokens_seen": 60664485, + "step": 2819, + "time_per_iteration": 2.7048139572143555 + }, + { + "auxiliary_loss_clip": 0.0116167, + "auxiliary_loss_mlp": 0.00887929, + "balance_loss_clip": 1.04591155, + "balance_loss_mlp": 1.00060296, + "epoch": 0.3390849516022365, + "flos": 34495754112000.0, + "grad_norm": 1.7458931132073248, + "language_loss": 0.66748953, + "learning_rate": 3.0789572234890057e-06, + "loss": 0.68798554, + "num_input_tokens_seen": 60686125, + "step": 2820, + "time_per_iteration": 2.8214950561523438 + }, + { + "auxiliary_loss_clip": 0.01177833, + "auxiliary_loss_mlp": 0.01039688, + "balance_loss_clip": 1.05435014, + "balance_loss_mlp": 1.02980542, + "epoch": 0.3392051944928756, + "flos": 16180307447040.0, + "grad_norm": 1.9201270460766922, + "language_loss": 0.7732023, + "learning_rate": 3.0783012480824596e-06, + "loss": 0.79537755, + "num_input_tokens_seen": 60705270, + "step": 2821, + "time_per_iteration": 2.719238042831421 + }, + { + "auxiliary_loss_clip": 0.01195129, + "auxiliary_loss_mlp": 0.01034678, + "balance_loss_clip": 1.05461836, + "balance_loss_mlp": 1.02498031, + "epoch": 0.33932543738351467, + "flos": 17086656349440.0, + "grad_norm": 2.0170362660245384, + "language_loss": 0.74109507, + "learning_rate": 3.077645109098471e-06, + "loss": 0.76339316, + "num_input_tokens_seen": 60721540, + "step": 2822, + "time_per_iteration": 2.60552716255188 + }, + { + "auxiliary_loss_clip": 0.01153745, + "auxiliary_loss_mlp": 0.0103561, + "balance_loss_clip": 1.05014908, + "balance_loss_mlp": 1.02575195, + "epoch": 0.3394456802741538, + "flos": 22126970551680.0, + "grad_norm": 1.8040631480037699, + "language_loss": 0.72177237, + "learning_rate": 3.076988806636577e-06, + "loss": 0.74366593, + "num_input_tokens_seen": 60739300, + "step": 2823, + "time_per_iteration": 5.614400863647461 + }, + { + "auxiliary_loss_clip": 0.01180539, + "auxiliary_loss_mlp": 0.00888825, + "balance_loss_clip": 1.05446649, + "balance_loss_mlp": 1.00068223, + "epoch": 0.3395659231647929, + "flos": 25226887121280.0, + "grad_norm": 2.0429726677945657, + "language_loss": 0.88796639, + "learning_rate": 3.0763323407963377e-06, + "loss": 0.90866005, + "num_input_tokens_seen": 60758910, + "step": 2824, + "time_per_iteration": 2.8080508708953857 + }, + { + "auxiliary_loss_clip": 0.01186869, + "auxiliary_loss_mlp": 0.01037738, + "balance_loss_clip": 1.05400014, + "balance_loss_mlp": 1.02788544, + "epoch": 0.33968616605543195, + "flos": 29096477343360.0, + "grad_norm": 1.9135923481408115, + "language_loss": 0.80168068, + "learning_rate": 3.075675711677337e-06, + "loss": 0.82392675, + "num_input_tokens_seen": 60779005, + "step": 2825, + "time_per_iteration": 3.782728433609009 + }, + { + "auxiliary_loss_clip": 0.01173197, + "auxiliary_loss_mlp": 0.01041061, + "balance_loss_clip": 1.05470455, + "balance_loss_mlp": 1.03181612, + "epoch": 0.33980640894607106, + "flos": 21433966479360.0, + "grad_norm": 2.103255994985501, + "language_loss": 0.78421092, + "learning_rate": 3.0750189193791865e-06, + "loss": 0.80635351, + "num_input_tokens_seen": 60798590, + "step": 2826, + "time_per_iteration": 2.739323377609253 + }, + { + "auxiliary_loss_clip": 0.01181913, + "auxiliary_loss_mlp": 0.01032582, + "balance_loss_clip": 1.05087173, + "balance_loss_mlp": 1.02255607, + "epoch": 0.33992665183671017, + "flos": 32490035596800.0, + "grad_norm": 1.880337103928861, + "language_loss": 0.7037946, + "learning_rate": 3.0743619640015203e-06, + "loss": 0.72593951, + "num_input_tokens_seen": 60818840, + "step": 2827, + "time_per_iteration": 2.80336856842041 + }, + { + "auxiliary_loss_clip": 0.01179545, + "auxiliary_loss_mlp": 0.0103982, + "balance_loss_clip": 1.04856884, + "balance_loss_mlp": 1.03027153, + "epoch": 0.3400468947273492, + "flos": 17055414495360.0, + "grad_norm": 4.914591500001335, + "language_loss": 0.92200029, + "learning_rate": 3.073704845643999e-06, + "loss": 0.9441939, + "num_input_tokens_seen": 60835965, + "step": 2828, + "time_per_iteration": 2.7734506130218506 + }, + { + "auxiliary_loss_clip": 0.01187366, + "auxiliary_loss_mlp": 0.01036959, + "balance_loss_clip": 1.05116463, + "balance_loss_mlp": 1.02640879, + "epoch": 0.34016713761798834, + "flos": 16872988296960.0, + "grad_norm": 4.052577216985805, + "language_loss": 0.77565628, + "learning_rate": 3.0730475644063063e-06, + "loss": 0.7978996, + "num_input_tokens_seen": 60851065, + "step": 2829, + "time_per_iteration": 2.7961392402648926 + }, + { + "auxiliary_loss_clip": 0.0116629, + "auxiliary_loss_mlp": 0.00887109, + "balance_loss_clip": 1.04731584, + "balance_loss_mlp": 1.00055528, + "epoch": 0.34028738050862745, + "flos": 21907161273600.0, + "grad_norm": 1.7297117559606217, + "language_loss": 0.64840323, + "learning_rate": 3.072390120388151e-06, + "loss": 0.66893727, + "num_input_tokens_seen": 60869390, + "step": 2830, + "time_per_iteration": 2.792546272277832 + }, + { + "auxiliary_loss_clip": 0.01189162, + "auxiliary_loss_mlp": 0.0104105, + "balance_loss_clip": 1.05371761, + "balance_loss_mlp": 1.03109574, + "epoch": 0.3404076233992665, + "flos": 22746034477440.0, + "grad_norm": 2.792046957016488, + "language_loss": 0.71478236, + "learning_rate": 3.071732513689267e-06, + "loss": 0.73708445, + "num_input_tokens_seen": 60887925, + "step": 2831, + "time_per_iteration": 2.7766072750091553 + }, + { + "auxiliary_loss_clip": 0.01188378, + "auxiliary_loss_mlp": 0.01036662, + "balance_loss_clip": 1.05668139, + "balance_loss_mlp": 1.0271492, + "epoch": 0.3405278662899056, + "flos": 17052361839360.0, + "grad_norm": 10.21971781172778, + "language_loss": 0.67288268, + "learning_rate": 3.0710747444094134e-06, + "loss": 0.69513309, + "num_input_tokens_seen": 60905955, + "step": 2832, + "time_per_iteration": 2.72282338142395 + }, + { + "auxiliary_loss_clip": 0.01184344, + "auxiliary_loss_mlp": 0.01034481, + "balance_loss_clip": 1.05839324, + "balance_loss_mlp": 1.02397311, + "epoch": 0.3406481091805447, + "flos": 42813131783040.0, + "grad_norm": 1.8655265270277428, + "language_loss": 0.65139306, + "learning_rate": 3.070416812648372e-06, + "loss": 0.6735813, + "num_input_tokens_seen": 60929405, + "step": 2833, + "time_per_iteration": 2.9322428703308105 + }, + { + "auxiliary_loss_clip": 0.01161317, + "auxiliary_loss_mlp": 0.01034772, + "balance_loss_clip": 1.04561758, + "balance_loss_mlp": 1.0250268, + "epoch": 0.3407683520711838, + "flos": 26761457917440.0, + "grad_norm": 2.4161098860533237, + "language_loss": 0.65127742, + "learning_rate": 3.069758718505951e-06, + "loss": 0.67323828, + "num_input_tokens_seen": 60951145, + "step": 2834, + "time_per_iteration": 2.904170036315918 + }, + { + "auxiliary_loss_clip": 0.01197719, + "auxiliary_loss_mlp": 0.01037208, + "balance_loss_clip": 1.05698824, + "balance_loss_mlp": 1.0274682, + "epoch": 0.3408885949618229, + "flos": 28767643309440.0, + "grad_norm": 1.6049700301363237, + "language_loss": 0.79718542, + "learning_rate": 3.0691004620819836e-06, + "loss": 0.81953466, + "num_input_tokens_seen": 60971275, + "step": 2835, + "time_per_iteration": 2.7413694858551025 + }, + { + "auxiliary_loss_clip": 0.01057323, + "auxiliary_loss_mlp": 0.01009721, + "balance_loss_clip": 1.01586914, + "balance_loss_mlp": 1.00726533, + "epoch": 0.341008837852462, + "flos": 63576252881280.0, + "grad_norm": 0.8121407884533034, + "language_loss": 0.60178423, + "learning_rate": 3.0684420434763254e-06, + "loss": 0.62245464, + "num_input_tokens_seen": 61037460, + "step": 2836, + "time_per_iteration": 3.514082431793213 + }, + { + "auxiliary_loss_clip": 0.01159862, + "auxiliary_loss_mlp": 0.01043354, + "balance_loss_clip": 1.05215073, + "balance_loss_mlp": 1.03431761, + "epoch": 0.34112908074310105, + "flos": 20812173120000.0, + "grad_norm": 1.7791575689944217, + "language_loss": 0.76814342, + "learning_rate": 3.06778346278886e-06, + "loss": 0.79017556, + "num_input_tokens_seen": 61056295, + "step": 2837, + "time_per_iteration": 3.004384756088257 + }, + { + "auxiliary_loss_clip": 0.01196608, + "auxiliary_loss_mlp": 0.0103534, + "balance_loss_clip": 1.05642867, + "balance_loss_mlp": 1.0249207, + "epoch": 0.34124932363374016, + "flos": 24976446520320.0, + "grad_norm": 1.790318695837147, + "language_loss": 0.78892815, + "learning_rate": 3.0671247201194906e-06, + "loss": 0.81124759, + "num_input_tokens_seen": 61078430, + "step": 2838, + "time_per_iteration": 2.67861270904541 + }, + { + "auxiliary_loss_clip": 0.01169235, + "auxiliary_loss_mlp": 0.01032957, + "balance_loss_clip": 1.05045462, + "balance_loss_mlp": 1.02245474, + "epoch": 0.3413695665243792, + "flos": 28402970480640.0, + "grad_norm": 1.8966649512301523, + "language_loss": 0.75242436, + "learning_rate": 3.066465815568151e-06, + "loss": 0.77444625, + "num_input_tokens_seen": 61099260, + "step": 2839, + "time_per_iteration": 2.8338024616241455 + }, + { + "auxiliary_loss_clip": 0.01188131, + "auxiliary_loss_mlp": 0.01033326, + "balance_loss_clip": 1.05260444, + "balance_loss_mlp": 1.02300858, + "epoch": 0.34148980941501833, + "flos": 25302012416640.0, + "grad_norm": 1.6267831593348718, + "language_loss": 0.68697298, + "learning_rate": 3.0658067492347947e-06, + "loss": 0.70918751, + "num_input_tokens_seen": 61121900, + "step": 2840, + "time_per_iteration": 2.7456624507904053 + }, + { + "auxiliary_loss_clip": 0.01131728, + "auxiliary_loss_mlp": 0.01030513, + "balance_loss_clip": 1.04743075, + "balance_loss_mlp": 1.0208931, + "epoch": 0.34161005230565744, + "flos": 17530081747200.0, + "grad_norm": 2.2478364638214097, + "language_loss": 0.66580367, + "learning_rate": 3.065147521219402e-06, + "loss": 0.68742609, + "num_input_tokens_seen": 61141155, + "step": 2841, + "time_per_iteration": 2.8937458992004395 + }, + { + "auxiliary_loss_clip": 0.01168051, + "auxiliary_loss_mlp": 0.01029591, + "balance_loss_clip": 1.05357552, + "balance_loss_mlp": 1.01977444, + "epoch": 0.3417302951962965, + "flos": 43650101566080.0, + "grad_norm": 1.4973566494153108, + "language_loss": 0.74661595, + "learning_rate": 3.064488131621977e-06, + "loss": 0.76859236, + "num_input_tokens_seen": 61164480, + "step": 2842, + "time_per_iteration": 2.961362361907959 + }, + { + "auxiliary_loss_clip": 0.01179826, + "auxiliary_loss_mlp": 0.0103573, + "balance_loss_clip": 1.05218887, + "balance_loss_mlp": 1.02588868, + "epoch": 0.3418505380869356, + "flos": 30882207012480.0, + "grad_norm": 1.8878548920889975, + "language_loss": 0.73963487, + "learning_rate": 3.063828580542549e-06, + "loss": 0.76179051, + "num_input_tokens_seen": 61185675, + "step": 2843, + "time_per_iteration": 2.894780397415161 + }, + { + "auxiliary_loss_clip": 0.01175811, + "auxiliary_loss_mlp": 0.01038734, + "balance_loss_clip": 1.05269337, + "balance_loss_mlp": 1.0298295, + "epoch": 0.3419707809775747, + "flos": 19463871277440.0, + "grad_norm": 1.8166088490109322, + "language_loss": 0.72844481, + "learning_rate": 3.0631688680811706e-06, + "loss": 0.75059026, + "num_input_tokens_seen": 61205300, + "step": 2844, + "time_per_iteration": 2.746452569961548 + }, + { + "auxiliary_loss_clip": 0.01198864, + "auxiliary_loss_mlp": 0.0103705, + "balance_loss_clip": 1.05651355, + "balance_loss_mlp": 1.02682185, + "epoch": 0.3420910238682138, + "flos": 28727818104960.0, + "grad_norm": 2.2881471699676577, + "language_loss": 0.75670111, + "learning_rate": 3.062508994337921e-06, + "loss": 0.77906024, + "num_input_tokens_seen": 61224905, + "step": 2845, + "time_per_iteration": 2.7111337184906006 + }, + { + "auxiliary_loss_clip": 0.0118826, + "auxiliary_loss_mlp": 0.01036287, + "balance_loss_clip": 1.05380702, + "balance_loss_mlp": 1.02562332, + "epoch": 0.3422112667588529, + "flos": 21397265758080.0, + "grad_norm": 1.9642966794532013, + "language_loss": 0.79156709, + "learning_rate": 3.0618489594129013e-06, + "loss": 0.81381255, + "num_input_tokens_seen": 61243045, + "step": 2846, + "time_per_iteration": 2.794020652770996 + }, + { + "auxiliary_loss_clip": 0.01171643, + "auxiliary_loss_mlp": 0.01032889, + "balance_loss_clip": 1.0525794, + "balance_loss_mlp": 1.02344799, + "epoch": 0.342331509649492, + "flos": 13881450038400.0, + "grad_norm": 5.6728090123317365, + "language_loss": 0.7129128, + "learning_rate": 3.061188763406239e-06, + "loss": 0.73495817, + "num_input_tokens_seen": 61259190, + "step": 2847, + "time_per_iteration": 2.7916839122772217 + }, + { + "auxiliary_loss_clip": 0.01173781, + "auxiliary_loss_mlp": 0.01035939, + "balance_loss_clip": 1.05248582, + "balance_loss_mlp": 1.02507925, + "epoch": 0.34245175254013105, + "flos": 28621450955520.0, + "grad_norm": 2.4379090989433783, + "language_loss": 0.81869459, + "learning_rate": 3.060528406418085e-06, + "loss": 0.84079182, + "num_input_tokens_seen": 61279040, + "step": 2848, + "time_per_iteration": 2.778071403503418 + }, + { + "auxiliary_loss_clip": 0.0117176, + "auxiliary_loss_mlp": 0.01037052, + "balance_loss_clip": 1.05280554, + "balance_loss_mlp": 1.02764678, + "epoch": 0.34257199543077016, + "flos": 34127058960000.0, + "grad_norm": 1.4791569896127612, + "language_loss": 0.61362791, + "learning_rate": 3.0598678885486145e-06, + "loss": 0.63571608, + "num_input_tokens_seen": 61301580, + "step": 2849, + "time_per_iteration": 5.579224109649658 + }, + { + "auxiliary_loss_clip": 0.01167049, + "auxiliary_loss_mlp": 0.00888212, + "balance_loss_clip": 1.04884005, + "balance_loss_mlp": 1.00069726, + "epoch": 0.34269223832140927, + "flos": 19974018188160.0, + "grad_norm": 1.6634900831000716, + "language_loss": 0.74306971, + "learning_rate": 3.0592072098980282e-06, + "loss": 0.76362228, + "num_input_tokens_seen": 61321240, + "step": 2850, + "time_per_iteration": 3.918177366256714 + }, + { + "auxiliary_loss_clip": 0.01171138, + "auxiliary_loss_mlp": 0.01038893, + "balance_loss_clip": 1.05275178, + "balance_loss_mlp": 1.02940965, + "epoch": 0.3428124812120483, + "flos": 27235658292480.0, + "grad_norm": 1.7074928469657324, + "language_loss": 0.73081362, + "learning_rate": 3.0585463705665514e-06, + "loss": 0.75291395, + "num_input_tokens_seen": 61341615, + "step": 2851, + "time_per_iteration": 2.8473598957061768 + }, + { + "auxiliary_loss_clip": 0.01167784, + "auxiliary_loss_mlp": 0.01037825, + "balance_loss_clip": 1.05171585, + "balance_loss_mlp": 1.02713144, + "epoch": 0.34293272410268744, + "flos": 24570871079040.0, + "grad_norm": 2.2012196053342516, + "language_loss": 0.70511383, + "learning_rate": 3.0578853706544304e-06, + "loss": 0.72716987, + "num_input_tokens_seen": 61359005, + "step": 2852, + "time_per_iteration": 2.969778537750244 + }, + { + "auxiliary_loss_clip": 0.01168491, + "auxiliary_loss_mlp": 0.00888353, + "balance_loss_clip": 1.050102, + "balance_loss_mlp": 1.0006361, + "epoch": 0.34305296699332655, + "flos": 21506865131520.0, + "grad_norm": 3.108772175316598, + "language_loss": 0.65388936, + "learning_rate": 3.0572242102619404e-06, + "loss": 0.67445779, + "num_input_tokens_seen": 61376160, + "step": 2853, + "time_per_iteration": 2.7991037368774414 + }, + { + "auxiliary_loss_clip": 0.01177643, + "auxiliary_loss_mlp": 0.0103151, + "balance_loss_clip": 1.05587161, + "balance_loss_mlp": 1.02208638, + "epoch": 0.3431732098839656, + "flos": 24056665931520.0, + "grad_norm": 2.041046032768501, + "language_loss": 0.80659038, + "learning_rate": 3.0565628894893784e-06, + "loss": 0.82868189, + "num_input_tokens_seen": 61396795, + "step": 2854, + "time_per_iteration": 2.9983675479888916 + }, + { + "auxiliary_loss_clip": 0.01179509, + "auxiliary_loss_mlp": 0.01032277, + "balance_loss_clip": 1.05386591, + "balance_loss_mlp": 1.02193546, + "epoch": 0.3432934527746047, + "flos": 16800879744000.0, + "grad_norm": 1.9688146894954925, + "language_loss": 0.74669439, + "learning_rate": 3.0559014084370655e-06, + "loss": 0.76881224, + "num_input_tokens_seen": 61415320, + "step": 2855, + "time_per_iteration": 2.741197347640991 + }, + { + "auxiliary_loss_clip": 0.01185503, + "auxiliary_loss_mlp": 0.01033571, + "balance_loss_clip": 1.05554557, + "balance_loss_mlp": 1.02376628, + "epoch": 0.34341369566524377, + "flos": 23439720908160.0, + "grad_norm": 1.6366244214041201, + "language_loss": 0.78654319, + "learning_rate": 3.055239767205349e-06, + "loss": 0.80873388, + "num_input_tokens_seen": 61437070, + "step": 2856, + "time_per_iteration": 2.754823684692383 + }, + { + "auxiliary_loss_clip": 0.01187993, + "auxiliary_loss_mlp": 0.01034298, + "balance_loss_clip": 1.05990243, + "balance_loss_mlp": 1.02521968, + "epoch": 0.3435339385558829, + "flos": 17267466435840.0, + "grad_norm": 2.175153723708083, + "language_loss": 0.78632003, + "learning_rate": 3.054577965894599e-06, + "loss": 0.80854297, + "num_input_tokens_seen": 61453215, + "step": 2857, + "time_per_iteration": 2.758969306945801 + }, + { + "auxiliary_loss_clip": 0.01187295, + "auxiliary_loss_mlp": 0.01035239, + "balance_loss_clip": 1.05941689, + "balance_loss_mlp": 1.02492785, + "epoch": 0.343654181446522, + "flos": 22199366413440.0, + "grad_norm": 1.5813377638552943, + "language_loss": 0.70282608, + "learning_rate": 3.0539160046052094e-06, + "loss": 0.7250514, + "num_input_tokens_seen": 61472915, + "step": 2858, + "time_per_iteration": 2.8756558895111084 + }, + { + "auxiliary_loss_clip": 0.01172783, + "auxiliary_loss_mlp": 0.0103374, + "balance_loss_clip": 1.05243027, + "balance_loss_mlp": 1.02236211, + "epoch": 0.34377442433716104, + "flos": 19901801894400.0, + "grad_norm": 2.06369500846918, + "language_loss": 0.70457351, + "learning_rate": 3.0532538834376003e-06, + "loss": 0.72663873, + "num_input_tokens_seen": 61492475, + "step": 2859, + "time_per_iteration": 2.7874598503112793 + }, + { + "auxiliary_loss_clip": 0.011983, + "auxiliary_loss_mlp": 0.01028393, + "balance_loss_clip": 1.05923319, + "balance_loss_mlp": 1.01893353, + "epoch": 0.34389466722780015, + "flos": 22197678474240.0, + "grad_norm": 1.8976779380620514, + "language_loss": 0.78324604, + "learning_rate": 3.0525916024922143e-06, + "loss": 0.80551302, + "num_input_tokens_seen": 61511660, + "step": 2860, + "time_per_iteration": 2.6830995082855225 + }, + { + "auxiliary_loss_clip": 0.01178375, + "auxiliary_loss_mlp": 0.01037459, + "balance_loss_clip": 1.05341387, + "balance_loss_mlp": 1.02722526, + "epoch": 0.34401491011843927, + "flos": 18624567110400.0, + "grad_norm": 3.0387660779283987, + "language_loss": 0.84590483, + "learning_rate": 3.0519291618695193e-06, + "loss": 0.86806327, + "num_input_tokens_seen": 61529060, + "step": 2861, + "time_per_iteration": 2.7528769969940186 + }, + { + "auxiliary_loss_clip": 0.01154075, + "auxiliary_loss_mlp": 0.01035041, + "balance_loss_clip": 1.04527342, + "balance_loss_mlp": 1.02577293, + "epoch": 0.3441351530090783, + "flos": 17858197509120.0, + "grad_norm": 1.5064917673464888, + "language_loss": 0.75393051, + "learning_rate": 3.0512665616700065e-06, + "loss": 0.77582163, + "num_input_tokens_seen": 61548125, + "step": 2862, + "time_per_iteration": 2.763111114501953 + }, + { + "auxiliary_loss_clip": 0.01150866, + "auxiliary_loss_mlp": 0.0103145, + "balance_loss_clip": 1.04767692, + "balance_loss_mlp": 1.02119184, + "epoch": 0.34425539589971743, + "flos": 23112754381440.0, + "grad_norm": 1.8563925406963675, + "language_loss": 0.89037311, + "learning_rate": 3.0506038019941933e-06, + "loss": 0.91219634, + "num_input_tokens_seen": 61568135, + "step": 2863, + "time_per_iteration": 2.8817174434661865 + }, + { + "auxiliary_loss_clip": 0.01171445, + "auxiliary_loss_mlp": 0.01034057, + "balance_loss_clip": 1.05670059, + "balance_loss_mlp": 1.02373385, + "epoch": 0.34437563879035654, + "flos": 21907699977600.0, + "grad_norm": 2.320680749416784, + "language_loss": 0.67330128, + "learning_rate": 3.049940882942617e-06, + "loss": 0.69535637, + "num_input_tokens_seen": 61586920, + "step": 2864, + "time_per_iteration": 2.8087069988250732 + }, + { + "auxiliary_loss_clip": 0.01198544, + "auxiliary_loss_mlp": 0.01036621, + "balance_loss_clip": 1.05629349, + "balance_loss_mlp": 1.02679253, + "epoch": 0.3444958816809956, + "flos": 23076915586560.0, + "grad_norm": 1.7535943847412045, + "language_loss": 0.79968739, + "learning_rate": 3.0492778046158448e-06, + "loss": 0.82203901, + "num_input_tokens_seen": 61608340, + "step": 2865, + "time_per_iteration": 2.717729330062866 + }, + { + "auxiliary_loss_clip": 0.01191149, + "auxiliary_loss_mlp": 0.01043388, + "balance_loss_clip": 1.05896604, + "balance_loss_mlp": 1.03294492, + "epoch": 0.3446161245716347, + "flos": 21908633731200.0, + "grad_norm": 2.426951076211621, + "language_loss": 0.76882195, + "learning_rate": 3.0486145671144633e-06, + "loss": 0.79116732, + "num_input_tokens_seen": 61628130, + "step": 2866, + "time_per_iteration": 2.7792911529541016 + }, + { + "auxiliary_loss_clip": 0.0113698, + "auxiliary_loss_mlp": 0.01036001, + "balance_loss_clip": 1.04994154, + "balance_loss_mlp": 1.02638078, + "epoch": 0.3447363674622738, + "flos": 25112834461440.0, + "grad_norm": 2.551564995354402, + "language_loss": 0.77208674, + "learning_rate": 3.047951170539086e-06, + "loss": 0.79381645, + "num_input_tokens_seen": 61647755, + "step": 2867, + "time_per_iteration": 2.909548282623291 + }, + { + "auxiliary_loss_clip": 0.01164697, + "auxiliary_loss_mlp": 0.01029302, + "balance_loss_clip": 1.05431747, + "balance_loss_mlp": 1.02129698, + "epoch": 0.3448566103529129, + "flos": 11984684451840.0, + "grad_norm": 1.953207850775061, + "language_loss": 0.84366673, + "learning_rate": 3.047287614990349e-06, + "loss": 0.86560667, + "num_input_tokens_seen": 61665675, + "step": 2868, + "time_per_iteration": 2.846679449081421 + }, + { + "auxiliary_loss_clip": 0.01173649, + "auxiliary_loss_mlp": 0.01037879, + "balance_loss_clip": 1.05473018, + "balance_loss_mlp": 1.0272572, + "epoch": 0.344976853243552, + "flos": 40187882465280.0, + "grad_norm": 2.3195911042711717, + "language_loss": 0.61649203, + "learning_rate": 3.046623900568914e-06, + "loss": 0.63860726, + "num_input_tokens_seen": 61688240, + "step": 2869, + "time_per_iteration": 2.874044895172119 + }, + { + "auxiliary_loss_clip": 0.01176625, + "auxiliary_loss_mlp": 0.01031661, + "balance_loss_clip": 1.05322683, + "balance_loss_mlp": 1.0217011, + "epoch": 0.34509709613419104, + "flos": 28723652127360.0, + "grad_norm": 2.9959834121587314, + "language_loss": 0.70020032, + "learning_rate": 3.045960027375465e-06, + "loss": 0.72228312, + "num_input_tokens_seen": 61706075, + "step": 2870, + "time_per_iteration": 2.853135347366333 + }, + { + "auxiliary_loss_clip": 0.01190136, + "auxiliary_loss_mlp": 0.01039064, + "balance_loss_clip": 1.05254424, + "balance_loss_mlp": 1.02884781, + "epoch": 0.34521733902483015, + "flos": 29967597982080.0, + "grad_norm": 3.1053860882279816, + "language_loss": 0.82345527, + "learning_rate": 3.045295995510711e-06, + "loss": 0.84574729, + "num_input_tokens_seen": 61723045, + "step": 2871, + "time_per_iteration": 2.797058582305908 + }, + { + "auxiliary_loss_clip": 0.01175402, + "auxiliary_loss_mlp": 0.01032762, + "balance_loss_clip": 1.05435693, + "balance_loss_mlp": 1.02391708, + "epoch": 0.34533758191546926, + "flos": 27923059843200.0, + "grad_norm": 2.262419793867014, + "language_loss": 0.73472816, + "learning_rate": 3.0446318050753865e-06, + "loss": 0.75680977, + "num_input_tokens_seen": 61743525, + "step": 2872, + "time_per_iteration": 2.827738046646118 + }, + { + "auxiliary_loss_clip": 0.0117668, + "auxiliary_loss_mlp": 0.01031797, + "balance_loss_clip": 1.05064988, + "balance_loss_mlp": 1.02277851, + "epoch": 0.3454578248061083, + "flos": 27125879351040.0, + "grad_norm": 2.238990051594134, + "language_loss": 0.77742195, + "learning_rate": 3.0439674561702474e-06, + "loss": 0.79950678, + "num_input_tokens_seen": 61763025, + "step": 2873, + "time_per_iteration": 2.7618942260742188 + }, + { + "auxiliary_loss_clip": 0.01187648, + "auxiliary_loss_mlp": 0.01033124, + "balance_loss_clip": 1.05640042, + "balance_loss_mlp": 1.02428448, + "epoch": 0.3455780676967474, + "flos": 19024899166080.0, + "grad_norm": 2.7674189856637983, + "language_loss": 0.8810612, + "learning_rate": 3.043302948896076e-06, + "loss": 0.90326893, + "num_input_tokens_seen": 61781630, + "step": 2874, + "time_per_iteration": 3.614102840423584 + }, + { + "auxiliary_loss_clip": 0.01148102, + "auxiliary_loss_mlp": 0.01029729, + "balance_loss_clip": 1.05058146, + "balance_loss_mlp": 1.02026403, + "epoch": 0.34569831058738654, + "flos": 34496005507200.0, + "grad_norm": 2.0664864537041354, + "language_loss": 0.60869485, + "learning_rate": 3.0426382833536756e-06, + "loss": 0.63047314, + "num_input_tokens_seen": 61804985, + "step": 2875, + "time_per_iteration": 4.8805553913116455 + }, + { + "auxiliary_loss_clip": 0.01158385, + "auxiliary_loss_mlp": 0.01035016, + "balance_loss_clip": 1.04629374, + "balance_loss_mlp": 1.02557468, + "epoch": 0.3458185534780256, + "flos": 31138681098240.0, + "grad_norm": 2.069853959050341, + "language_loss": 0.78105891, + "learning_rate": 3.041973459643877e-06, + "loss": 0.80299294, + "num_input_tokens_seen": 61824440, + "step": 2876, + "time_per_iteration": 2.8830795288085938 + }, + { + "auxiliary_loss_clip": 0.01148177, + "auxiliary_loss_mlp": 0.01026538, + "balance_loss_clip": 1.04534674, + "balance_loss_mlp": 1.01744187, + "epoch": 0.3459387963686647, + "flos": 32452508862720.0, + "grad_norm": 2.643252352127033, + "language_loss": 0.66845608, + "learning_rate": 3.0413084778675334e-06, + "loss": 0.69020319, + "num_input_tokens_seen": 61845690, + "step": 2877, + "time_per_iteration": 4.039004802703857 + }, + { + "auxiliary_loss_clip": 0.01167081, + "auxiliary_loss_mlp": 0.00887053, + "balance_loss_clip": 1.04846454, + "balance_loss_mlp": 1.00051832, + "epoch": 0.3460590392593038, + "flos": 24675658030080.0, + "grad_norm": 1.8455018939651429, + "language_loss": 0.83938217, + "learning_rate": 3.0406433381255214e-06, + "loss": 0.85992348, + "num_input_tokens_seen": 61863725, + "step": 2878, + "time_per_iteration": 2.8169138431549072 + }, + { + "auxiliary_loss_clip": 0.0118203, + "auxiliary_loss_mlp": 0.01038237, + "balance_loss_clip": 1.05458832, + "balance_loss_mlp": 1.02898645, + "epoch": 0.34617928214994287, + "flos": 18807316531200.0, + "grad_norm": 2.2997147249109307, + "language_loss": 0.82158482, + "learning_rate": 3.0399780405187425e-06, + "loss": 0.84378755, + "num_input_tokens_seen": 61882720, + "step": 2879, + "time_per_iteration": 2.714296579360962 + }, + { + "auxiliary_loss_clip": 0.01179111, + "auxiliary_loss_mlp": 0.01033523, + "balance_loss_clip": 1.05098224, + "balance_loss_mlp": 1.02421284, + "epoch": 0.346299525040582, + "flos": 24857653265280.0, + "grad_norm": 1.8461504493131542, + "language_loss": 0.79047942, + "learning_rate": 3.0393125851481216e-06, + "loss": 0.8126058, + "num_input_tokens_seen": 61902595, + "step": 2880, + "time_per_iteration": 2.7497806549072266 + }, + { + "auxiliary_loss_clip": 0.01159872, + "auxiliary_loss_mlp": 0.01040729, + "balance_loss_clip": 1.05153418, + "balance_loss_mlp": 1.03124046, + "epoch": 0.3464197679312211, + "flos": 16434914025600.0, + "grad_norm": 2.2001266641092867, + "language_loss": 0.86765921, + "learning_rate": 3.038646972114608e-06, + "loss": 0.88966525, + "num_input_tokens_seen": 61918920, + "step": 2881, + "time_per_iteration": 2.8096468448638916 + }, + { + "auxiliary_loss_clip": 0.01158323, + "auxiliary_loss_mlp": 0.01039854, + "balance_loss_clip": 1.05088854, + "balance_loss_mlp": 1.03091955, + "epoch": 0.34654001082186014, + "flos": 22382474970240.0, + "grad_norm": 1.819462251815696, + "language_loss": 0.66966265, + "learning_rate": 3.037981201519174e-06, + "loss": 0.69164437, + "num_input_tokens_seen": 61939520, + "step": 2882, + "time_per_iteration": 2.838613986968994 + }, + { + "auxiliary_loss_clip": 0.01182122, + "auxiliary_loss_mlp": 0.01041033, + "balance_loss_clip": 1.05476439, + "balance_loss_mlp": 1.03182399, + "epoch": 0.34666025371249926, + "flos": 19573901614080.0, + "grad_norm": 2.350931444840637, + "language_loss": 0.7148757, + "learning_rate": 3.0373152734628175e-06, + "loss": 0.73710716, + "num_input_tokens_seen": 61957800, + "step": 2883, + "time_per_iteration": 2.739712715148926 + }, + { + "auxiliary_loss_clip": 0.01175453, + "auxiliary_loss_mlp": 0.0103499, + "balance_loss_clip": 1.04953325, + "balance_loss_mlp": 1.02580523, + "epoch": 0.34678049660313837, + "flos": 15267637751040.0, + "grad_norm": 1.9200718599644295, + "language_loss": 0.75771165, + "learning_rate": 3.0366491880465584e-06, + "loss": 0.77981603, + "num_input_tokens_seen": 61975820, + "step": 2884, + "time_per_iteration": 2.7161190509796143 + }, + { + "auxiliary_loss_clip": 0.01201087, + "auxiliary_loss_mlp": 0.01031175, + "balance_loss_clip": 1.05990028, + "balance_loss_mlp": 1.02140641, + "epoch": 0.3469007394937774, + "flos": 21181550630400.0, + "grad_norm": 1.6291959915841934, + "language_loss": 0.82049024, + "learning_rate": 3.035982945371443e-06, + "loss": 0.8428129, + "num_input_tokens_seen": 61997515, + "step": 2885, + "time_per_iteration": 2.667698621749878 + }, + { + "auxiliary_loss_clip": 0.01184009, + "auxiliary_loss_mlp": 0.01032569, + "balance_loss_clip": 1.05428314, + "balance_loss_mlp": 1.02342546, + "epoch": 0.34702098238441653, + "flos": 22375471818240.0, + "grad_norm": 2.2041518104491957, + "language_loss": 0.85590911, + "learning_rate": 3.035316545538537e-06, + "loss": 0.87807488, + "num_input_tokens_seen": 62016310, + "step": 2886, + "time_per_iteration": 2.8332786560058594 + }, + { + "auxiliary_loss_clip": 0.01171538, + "auxiliary_loss_mlp": 0.01029829, + "balance_loss_clip": 1.05590868, + "balance_loss_mlp": 1.02124, + "epoch": 0.3471412252750556, + "flos": 22929430343040.0, + "grad_norm": 1.8853876429781093, + "language_loss": 0.79231465, + "learning_rate": 3.034649988648935e-06, + "loss": 0.81432831, + "num_input_tokens_seen": 62036075, + "step": 2887, + "time_per_iteration": 2.8558642864227295 + }, + { + "auxiliary_loss_clip": 0.01177033, + "auxiliary_loss_mlp": 0.01034355, + "balance_loss_clip": 1.05212986, + "balance_loss_mlp": 1.02578366, + "epoch": 0.3472614681656947, + "flos": 21324259365120.0, + "grad_norm": 1.6867400955215266, + "language_loss": 0.8051101, + "learning_rate": 3.033983274803752e-06, + "loss": 0.82722402, + "num_input_tokens_seen": 62055865, + "step": 2888, + "time_per_iteration": 2.8228347301483154 + }, + { + "auxiliary_loss_clip": 0.01167993, + "auxiliary_loss_mlp": 0.01033497, + "balance_loss_clip": 1.04943085, + "balance_loss_mlp": 1.02389479, + "epoch": 0.3473817110563338, + "flos": 23475739271040.0, + "grad_norm": 2.134006128296025, + "language_loss": 0.72783875, + "learning_rate": 3.0333164041041283e-06, + "loss": 0.74985367, + "num_input_tokens_seen": 62072180, + "step": 2889, + "time_per_iteration": 2.82100510597229 + }, + { + "auxiliary_loss_clip": 0.01148876, + "auxiliary_loss_mlp": 0.01031476, + "balance_loss_clip": 1.04792047, + "balance_loss_mlp": 1.02264857, + "epoch": 0.34750195394697286, + "flos": 22346025644160.0, + "grad_norm": 1.9242464902964191, + "language_loss": 0.71923745, + "learning_rate": 3.032649376651228e-06, + "loss": 0.74104095, + "num_input_tokens_seen": 62091600, + "step": 2890, + "time_per_iteration": 3.040524482727051 + }, + { + "auxiliary_loss_clip": 0.01165588, + "auxiliary_loss_mlp": 0.01028143, + "balance_loss_clip": 1.05054617, + "balance_loss_mlp": 1.01874316, + "epoch": 0.347622196837612, + "flos": 29095004885760.0, + "grad_norm": 1.5494201496716578, + "language_loss": 0.75782508, + "learning_rate": 3.031982192546238e-06, + "loss": 0.77976239, + "num_input_tokens_seen": 62114695, + "step": 2891, + "time_per_iteration": 3.0228633880615234 + }, + { + "auxiliary_loss_clip": 0.01186836, + "auxiliary_loss_mlp": 0.01034135, + "balance_loss_clip": 1.05501497, + "balance_loss_mlp": 1.02528977, + "epoch": 0.3477424397282511, + "flos": 22455732758400.0, + "grad_norm": 2.1232885204807603, + "language_loss": 0.94836456, + "learning_rate": 3.0313148518903696e-06, + "loss": 0.97057426, + "num_input_tokens_seen": 62134520, + "step": 2892, + "time_per_iteration": 2.717322826385498 + }, + { + "auxiliary_loss_clip": 0.0117603, + "auxiliary_loss_mlp": 0.01029335, + "balance_loss_clip": 1.05320823, + "balance_loss_mlp": 1.02105641, + "epoch": 0.34786268261889014, + "flos": 15778790242560.0, + "grad_norm": 3.564943502804203, + "language_loss": 0.80951571, + "learning_rate": 3.030647354784859e-06, + "loss": 0.83156931, + "num_input_tokens_seen": 62151560, + "step": 2893, + "time_per_iteration": 2.827242612838745 + }, + { + "auxiliary_loss_clip": 0.0116405, + "auxiliary_loss_mlp": 0.01034128, + "balance_loss_clip": 1.05108917, + "balance_loss_mlp": 1.02546716, + "epoch": 0.34798292550952925, + "flos": 20777627214720.0, + "grad_norm": 1.8180339348091556, + "language_loss": 0.77636099, + "learning_rate": 3.029979701330964e-06, + "loss": 0.79834276, + "num_input_tokens_seen": 62170985, + "step": 2894, + "time_per_iteration": 2.898146867752075 + }, + { + "auxiliary_loss_clip": 0.01177236, + "auxiliary_loss_mlp": 0.01033163, + "balance_loss_clip": 1.05185759, + "balance_loss_mlp": 1.02391779, + "epoch": 0.34810316840016836, + "flos": 19937820257280.0, + "grad_norm": 4.279071632973726, + "language_loss": 0.79808843, + "learning_rate": 3.029311891629966e-06, + "loss": 0.82019246, + "num_input_tokens_seen": 62189440, + "step": 2895, + "time_per_iteration": 2.77127742767334 + }, + { + "auxiliary_loss_clip": 0.01172808, + "auxiliary_loss_mlp": 0.01030461, + "balance_loss_clip": 1.054317, + "balance_loss_mlp": 1.0215615, + "epoch": 0.3482234112908074, + "flos": 23623296341760.0, + "grad_norm": 1.8233376396128642, + "language_loss": 0.74624693, + "learning_rate": 3.0286439257831744e-06, + "loss": 0.76827967, + "num_input_tokens_seen": 62208910, + "step": 2896, + "time_per_iteration": 2.7741830348968506 + }, + { + "auxiliary_loss_clip": 0.0119582, + "auxiliary_loss_mlp": 0.01040039, + "balance_loss_clip": 1.05520129, + "balance_loss_mlp": 1.03012037, + "epoch": 0.3483436541814465, + "flos": 23986712194560.0, + "grad_norm": 1.9340471027266235, + "language_loss": 0.71528447, + "learning_rate": 3.0279758038919156e-06, + "loss": 0.737643, + "num_input_tokens_seen": 62227135, + "step": 2897, + "time_per_iteration": 2.7236313819885254 + }, + { + "auxiliary_loss_clip": 0.0118474, + "auxiliary_loss_mlp": 0.01035278, + "balance_loss_clip": 1.05357075, + "balance_loss_mlp": 1.02618814, + "epoch": 0.34846389707208564, + "flos": 22638338524800.0, + "grad_norm": 4.0707405743941365, + "language_loss": 0.78484201, + "learning_rate": 3.0273075260575455e-06, + "loss": 0.80704212, + "num_input_tokens_seen": 62246035, + "step": 2898, + "time_per_iteration": 2.739084243774414 + }, + { + "auxiliary_loss_clip": 0.01175184, + "auxiliary_loss_mlp": 0.01030468, + "balance_loss_clip": 1.05073249, + "balance_loss_mlp": 1.02060366, + "epoch": 0.3485841399627247, + "flos": 21792857218560.0, + "grad_norm": 2.1124077371603898, + "language_loss": 0.80834305, + "learning_rate": 3.0266390923814396e-06, + "loss": 0.83039951, + "num_input_tokens_seen": 62264095, + "step": 2899, + "time_per_iteration": 2.8498361110687256 + }, + { + "auxiliary_loss_clip": 0.01176053, + "auxiliary_loss_mlp": 0.01036716, + "balance_loss_clip": 1.0554558, + "balance_loss_mlp": 1.02758479, + "epoch": 0.3487043828533638, + "flos": 17019036996480.0, + "grad_norm": 1.87767241214296, + "language_loss": 0.81965244, + "learning_rate": 3.0259705029650008e-06, + "loss": 0.84178019, + "num_input_tokens_seen": 62282025, + "step": 2900, + "time_per_iteration": 4.6522181034088135 + }, + { + "auxiliary_loss_clip": 0.01183099, + "auxiliary_loss_mlp": 0.01028666, + "balance_loss_clip": 1.05149722, + "balance_loss_mlp": 1.01950443, + "epoch": 0.34882462574400286, + "flos": 22601135013120.0, + "grad_norm": 1.7436195003644468, + "language_loss": 0.73070645, + "learning_rate": 3.025301757909652e-06, + "loss": 0.75282419, + "num_input_tokens_seen": 62302220, + "step": 2901, + "time_per_iteration": 2.7752599716186523 + }, + { + "auxiliary_loss_clip": 0.01169221, + "auxiliary_loss_mlp": 0.00887346, + "balance_loss_clip": 1.05427384, + "balance_loss_mlp": 1.00048089, + "epoch": 0.34894486863464197, + "flos": 29861518141440.0, + "grad_norm": 1.4528283904226216, + "language_loss": 0.80296469, + "learning_rate": 3.024632857316842e-06, + "loss": 0.82353038, + "num_input_tokens_seen": 62323535, + "step": 2902, + "time_per_iteration": 4.9163267612457275 + }, + { + "auxiliary_loss_clip": 0.01185115, + "auxiliary_loss_mlp": 0.01027411, + "balance_loss_clip": 1.05522871, + "balance_loss_mlp": 1.01831543, + "epoch": 0.3490651115252811, + "flos": 22122265870080.0, + "grad_norm": 1.700814134878662, + "language_loss": 0.77047408, + "learning_rate": 3.0239638012880412e-06, + "loss": 0.79259932, + "num_input_tokens_seen": 62343430, + "step": 2903, + "time_per_iteration": 2.7458839416503906 + }, + { + "auxiliary_loss_clip": 0.01150404, + "auxiliary_loss_mlp": 0.0103322, + "balance_loss_clip": 1.0490005, + "balance_loss_mlp": 1.02340937, + "epoch": 0.34918535441592014, + "flos": 12676682943360.0, + "grad_norm": 2.437652244539698, + "language_loss": 0.81775945, + "learning_rate": 3.0232945899247466e-06, + "loss": 0.83959568, + "num_input_tokens_seen": 62360365, + "step": 2904, + "time_per_iteration": 2.9020068645477295 + }, + { + "auxiliary_loss_clip": 0.01184889, + "auxiliary_loss_mlp": 0.01036324, + "balance_loss_clip": 1.05309224, + "balance_loss_mlp": 1.02677572, + "epoch": 0.34930559730655925, + "flos": 23185617120000.0, + "grad_norm": 2.38734713235688, + "language_loss": 0.77799648, + "learning_rate": 3.022625223328476e-06, + "loss": 0.80020857, + "num_input_tokens_seen": 62382105, + "step": 2905, + "time_per_iteration": 2.8254013061523438 + }, + { + "auxiliary_loss_clip": 0.01190807, + "auxiliary_loss_mlp": 0.0103075, + "balance_loss_clip": 1.05495572, + "balance_loss_mlp": 1.02110016, + "epoch": 0.34942584019719836, + "flos": 22855023319680.0, + "grad_norm": 1.389392570388077, + "language_loss": 0.69294131, + "learning_rate": 3.0219557016007723e-06, + "loss": 0.71515691, + "num_input_tokens_seen": 62402235, + "step": 2906, + "time_per_iteration": 2.972158670425415 + }, + { + "auxiliary_loss_clip": 0.01178778, + "auxiliary_loss_mlp": 0.01028174, + "balance_loss_clip": 1.05414248, + "balance_loss_mlp": 1.01872027, + "epoch": 0.3495460830878374, + "flos": 24426043441920.0, + "grad_norm": 2.5784389709599007, + "language_loss": 0.69837838, + "learning_rate": 3.021286024843202e-06, + "loss": 0.7204479, + "num_input_tokens_seen": 62420430, + "step": 2907, + "time_per_iteration": 2.7175867557525635 + }, + { + "auxiliary_loss_clip": 0.01087748, + "auxiliary_loss_mlp": 0.0100607, + "balance_loss_clip": 1.01972735, + "balance_loss_mlp": 1.00372183, + "epoch": 0.3496663259784765, + "flos": 70008749389440.0, + "grad_norm": 1.0680742659767726, + "language_loss": 0.64759523, + "learning_rate": 3.0206161931573526e-06, + "loss": 0.66853344, + "num_input_tokens_seen": 62472980, + "step": 2908, + "time_per_iteration": 3.1608469486236572 + }, + { + "auxiliary_loss_clip": 0.01166056, + "auxiliary_loss_mlp": 0.01026682, + "balance_loss_clip": 1.0470823, + "balance_loss_mlp": 1.01800942, + "epoch": 0.34978656886911563, + "flos": 28692805322880.0, + "grad_norm": 1.6459490157319656, + "language_loss": 0.93205357, + "learning_rate": 3.0199462066448388e-06, + "loss": 0.95398092, + "num_input_tokens_seen": 62495175, + "step": 2909, + "time_per_iteration": 2.858154058456421 + }, + { + "auxiliary_loss_clip": 0.01186343, + "auxiliary_loss_mlp": 0.01030133, + "balance_loss_clip": 1.05542147, + "balance_loss_mlp": 1.0210191, + "epoch": 0.3499068117597547, + "flos": 21142156389120.0, + "grad_norm": 1.7492729097552695, + "language_loss": 0.69249308, + "learning_rate": 3.019276065407296e-06, + "loss": 0.71465778, + "num_input_tokens_seen": 62514295, + "step": 2910, + "time_per_iteration": 2.8381950855255127 + }, + { + "auxiliary_loss_clip": 0.01157506, + "auxiliary_loss_mlp": 0.01030032, + "balance_loss_clip": 1.04952431, + "balance_loss_mlp": 1.02050102, + "epoch": 0.3500270546503938, + "flos": 22782699285120.0, + "grad_norm": 1.7563035603273494, + "language_loss": 0.80766124, + "learning_rate": 3.018605769546385e-06, + "loss": 0.82953656, + "num_input_tokens_seen": 62534850, + "step": 2911, + "time_per_iteration": 2.8770558834075928 + }, + { + "auxiliary_loss_clip": 0.01180436, + "auxiliary_loss_mlp": 0.01029252, + "balance_loss_clip": 1.05005538, + "balance_loss_mlp": 1.01975083, + "epoch": 0.3501472975410329, + "flos": 22894058424960.0, + "grad_norm": 1.6799578968276359, + "language_loss": 0.79473609, + "learning_rate": 3.017935319163788e-06, + "loss": 0.8168329, + "num_input_tokens_seen": 62553810, + "step": 2912, + "time_per_iteration": 2.7681260108947754 + }, + { + "auxiliary_loss_clip": 0.01184359, + "auxiliary_loss_mlp": 0.01031019, + "balance_loss_clip": 1.05294514, + "balance_loss_mlp": 1.02114868, + "epoch": 0.35026754043167196, + "flos": 25446588658560.0, + "grad_norm": 1.7135957322748485, + "language_loss": 0.7041142, + "learning_rate": 3.017264714361213e-06, + "loss": 0.72626793, + "num_input_tokens_seen": 62573460, + "step": 2913, + "time_per_iteration": 2.7664308547973633 + }, + { + "auxiliary_loss_clip": 0.01175088, + "auxiliary_loss_mlp": 0.00887149, + "balance_loss_clip": 1.05461287, + "balance_loss_mlp": 1.00062644, + "epoch": 0.3503877833223111, + "flos": 19573757959680.0, + "grad_norm": 1.9315156460665608, + "language_loss": 0.82603639, + "learning_rate": 3.016593955240389e-06, + "loss": 0.84665877, + "num_input_tokens_seen": 62592150, + "step": 2914, + "time_per_iteration": 2.756565809249878 + }, + { + "auxiliary_loss_clip": 0.01074879, + "auxiliary_loss_mlp": 0.01002202, + "balance_loss_clip": 1.01691818, + "balance_loss_mlp": 0.99986535, + "epoch": 0.3505080262129502, + "flos": 65072075880960.0, + "grad_norm": 0.858378378832769, + "language_loss": 0.63623154, + "learning_rate": 3.015923041903071e-06, + "loss": 0.65700245, + "num_input_tokens_seen": 62658275, + "step": 2915, + "time_per_iteration": 3.3310558795928955 + }, + { + "auxiliary_loss_clip": 0.01182181, + "auxiliary_loss_mlp": 0.01032509, + "balance_loss_clip": 1.05472445, + "balance_loss_mlp": 1.0229187, + "epoch": 0.35062826910358924, + "flos": 29314562768640.0, + "grad_norm": 1.9251835148812304, + "language_loss": 0.83784503, + "learning_rate": 3.0152519744510347e-06, + "loss": 0.85999191, + "num_input_tokens_seen": 62678075, + "step": 2916, + "time_per_iteration": 2.9000773429870605 + }, + { + "auxiliary_loss_clip": 0.01165298, + "auxiliary_loss_mlp": 0.01033984, + "balance_loss_clip": 1.0512296, + "balance_loss_mlp": 1.024441, + "epoch": 0.35074851199422835, + "flos": 23987717775360.0, + "grad_norm": 1.7312701167233924, + "language_loss": 0.82805628, + "learning_rate": 3.014580752986081e-06, + "loss": 0.85004914, + "num_input_tokens_seen": 62696950, + "step": 2917, + "time_per_iteration": 2.8552608489990234 + }, + { + "auxiliary_loss_clip": 0.01158923, + "auxiliary_loss_mlp": 0.01034287, + "balance_loss_clip": 1.0502255, + "balance_loss_mlp": 1.02467263, + "epoch": 0.3508687548848674, + "flos": 15224436668160.0, + "grad_norm": 2.16044065268627, + "language_loss": 0.78716034, + "learning_rate": 3.0139093776100345e-06, + "loss": 0.8090924, + "num_input_tokens_seen": 62713540, + "step": 2918, + "time_per_iteration": 2.8669118881225586 + }, + { + "auxiliary_loss_clip": 0.01191609, + "auxiliary_loss_mlp": 0.01029869, + "balance_loss_clip": 1.05527699, + "balance_loss_mlp": 1.02136946, + "epoch": 0.3509889977755065, + "flos": 21361750185600.0, + "grad_norm": 1.6412682273619963, + "language_loss": 0.7516861, + "learning_rate": 3.013237848424741e-06, + "loss": 0.77390087, + "num_input_tokens_seen": 62732925, + "step": 2919, + "time_per_iteration": 2.794863224029541 + }, + { + "auxiliary_loss_clip": 0.01177293, + "auxiliary_loss_mlp": 0.01039433, + "balance_loss_clip": 1.05474591, + "balance_loss_mlp": 1.02993178, + "epoch": 0.35110924066614563, + "flos": 19135360465920.0, + "grad_norm": 2.038095725149016, + "language_loss": 0.75066912, + "learning_rate": 3.012566165532072e-06, + "loss": 0.77283633, + "num_input_tokens_seen": 62751715, + "step": 2920, + "time_per_iteration": 2.8103270530700684 + }, + { + "auxiliary_loss_clip": 0.0114739, + "auxiliary_loss_mlp": 0.01037754, + "balance_loss_clip": 1.04730415, + "balance_loss_mlp": 1.02833676, + "epoch": 0.3512294835567847, + "flos": 21980885938560.0, + "grad_norm": 2.281489738079655, + "language_loss": 0.75815445, + "learning_rate": 3.0118943290339207e-06, + "loss": 0.78000593, + "num_input_tokens_seen": 62771925, + "step": 2921, + "time_per_iteration": 3.0140514373779297 + }, + { + "auxiliary_loss_clip": 0.01152887, + "auxiliary_loss_mlp": 0.01039566, + "balance_loss_clip": 1.04516697, + "balance_loss_mlp": 1.0295825, + "epoch": 0.3513497264474238, + "flos": 17817294896640.0, + "grad_norm": 1.8849867347183842, + "language_loss": 0.68338233, + "learning_rate": 3.011222339032204e-06, + "loss": 0.70530689, + "num_input_tokens_seen": 62790075, + "step": 2922, + "time_per_iteration": 2.8171474933624268 + }, + { + "auxiliary_loss_clip": 0.01191594, + "auxiliary_loss_mlp": 0.01029639, + "balance_loss_clip": 1.05469728, + "balance_loss_mlp": 1.02083564, + "epoch": 0.3514699693380629, + "flos": 26943417239040.0, + "grad_norm": 1.7469760922055333, + "language_loss": 0.69774753, + "learning_rate": 3.0105501956288626e-06, + "loss": 0.71995986, + "num_input_tokens_seen": 62810545, + "step": 2923, + "time_per_iteration": 2.7518486976623535 + }, + { + "auxiliary_loss_clip": 0.01188853, + "auxiliary_loss_mlp": 0.01034916, + "balance_loss_clip": 1.05277169, + "balance_loss_mlp": 1.02478361, + "epoch": 0.35159021222870196, + "flos": 15267565923840.0, + "grad_norm": 2.000985503523762, + "language_loss": 0.72730422, + "learning_rate": 3.0098778989258602e-06, + "loss": 0.74954188, + "num_input_tokens_seen": 62829155, + "step": 2924, + "time_per_iteration": 2.7417421340942383 + }, + { + "auxiliary_loss_clip": 0.01158233, + "auxiliary_loss_mlp": 0.01035804, + "balance_loss_clip": 1.05085588, + "balance_loss_mlp": 1.02588618, + "epoch": 0.35171045511934107, + "flos": 13984154000640.0, + "grad_norm": 1.960295094899774, + "language_loss": 0.88295746, + "learning_rate": 3.009205449025183e-06, + "loss": 0.90489787, + "num_input_tokens_seen": 62845350, + "step": 2925, + "time_per_iteration": 3.7395620346069336 + }, + { + "auxiliary_loss_clip": 0.01156603, + "auxiliary_loss_mlp": 0.01038243, + "balance_loss_clip": 1.04741299, + "balance_loss_mlp": 1.02903426, + "epoch": 0.3518306980099802, + "flos": 14283434119680.0, + "grad_norm": 1.7837710747301005, + "language_loss": 0.62716734, + "learning_rate": 3.008532846028842e-06, + "loss": 0.6491158, + "num_input_tokens_seen": 62862110, + "step": 2926, + "time_per_iteration": 3.7426514625549316 + }, + { + "auxiliary_loss_clip": 0.01194394, + "auxiliary_loss_mlp": 0.01034521, + "balance_loss_clip": 1.05602288, + "balance_loss_mlp": 1.0245676, + "epoch": 0.35195094090061924, + "flos": 27052872958080.0, + "grad_norm": 2.426381776830891, + "language_loss": 0.71812832, + "learning_rate": 3.0078600900388694e-06, + "loss": 0.74041748, + "num_input_tokens_seen": 62882415, + "step": 2927, + "time_per_iteration": 2.6644370555877686 + }, + { + "auxiliary_loss_clip": 0.01153351, + "auxiliary_loss_mlp": 0.01028642, + "balance_loss_clip": 1.04684734, + "balance_loss_mlp": 1.01942122, + "epoch": 0.35207118379125835, + "flos": 25629266252160.0, + "grad_norm": 1.858239116864955, + "language_loss": 0.74050248, + "learning_rate": 3.007187181157323e-06, + "loss": 0.76232243, + "num_input_tokens_seen": 62902425, + "step": 2928, + "time_per_iteration": 3.874879837036133 + }, + { + "auxiliary_loss_clip": 0.01134121, + "auxiliary_loss_mlp": 0.01029988, + "balance_loss_clip": 1.04613924, + "balance_loss_mlp": 1.02051115, + "epoch": 0.35219142668189746, + "flos": 18004713085440.0, + "grad_norm": 2.111930358418089, + "language_loss": 0.68079579, + "learning_rate": 3.006514119486282e-06, + "loss": 0.70243692, + "num_input_tokens_seen": 62919255, + "step": 2929, + "time_per_iteration": 3.8259012699127197 + }, + { + "auxiliary_loss_clip": 0.01153666, + "auxiliary_loss_mlp": 0.01025425, + "balance_loss_clip": 1.04757142, + "balance_loss_mlp": 1.01680028, + "epoch": 0.3523116695725365, + "flos": 14028109269120.0, + "grad_norm": 1.8488489288691163, + "language_loss": 0.69881535, + "learning_rate": 3.005840905127849e-06, + "loss": 0.72060633, + "num_input_tokens_seen": 62936160, + "step": 2930, + "time_per_iteration": 2.7998414039611816 + }, + { + "auxiliary_loss_clip": 0.01193742, + "auxiliary_loss_mlp": 0.01032738, + "balance_loss_clip": 1.0574348, + "balance_loss_mlp": 1.02387488, + "epoch": 0.3524319124631756, + "flos": 21433966479360.0, + "grad_norm": 1.9949063912328482, + "language_loss": 0.86338878, + "learning_rate": 3.0051675381841516e-06, + "loss": 0.88565356, + "num_input_tokens_seen": 62953470, + "step": 2931, + "time_per_iteration": 2.7365095615386963 + }, + { + "auxiliary_loss_clip": 0.01129243, + "auxiliary_loss_mlp": 0.00887618, + "balance_loss_clip": 1.0425086, + "balance_loss_mlp": 1.00061929, + "epoch": 0.3525521553538147, + "flos": 26322773114880.0, + "grad_norm": 1.5996758385113925, + "language_loss": 0.76517153, + "learning_rate": 3.0044940187573363e-06, + "loss": 0.78534019, + "num_input_tokens_seen": 62974480, + "step": 2932, + "time_per_iteration": 2.9517643451690674 + }, + { + "auxiliary_loss_clip": 0.01181695, + "auxiliary_loss_mlp": 0.01028407, + "balance_loss_clip": 1.0507431, + "balance_loss_mlp": 1.01943099, + "epoch": 0.3526723982444538, + "flos": 21543314457600.0, + "grad_norm": 1.957730345494509, + "language_loss": 0.6527245, + "learning_rate": 3.003820346949578e-06, + "loss": 0.67482543, + "num_input_tokens_seen": 62992560, + "step": 2933, + "time_per_iteration": 2.743135452270508 + }, + { + "auxiliary_loss_clip": 0.01193817, + "auxiliary_loss_mlp": 0.01028386, + "balance_loss_clip": 1.05455637, + "balance_loss_mlp": 1.01957047, + "epoch": 0.3527926411350929, + "flos": 23733649900800.0, + "grad_norm": 2.2567280287169407, + "language_loss": 0.79515499, + "learning_rate": 3.003146522863071e-06, + "loss": 0.81737709, + "num_input_tokens_seen": 63013445, + "step": 2934, + "time_per_iteration": 2.733391761779785 + }, + { + "auxiliary_loss_clip": 0.01170104, + "auxiliary_loss_mlp": 0.01029344, + "balance_loss_clip": 1.05207992, + "balance_loss_mlp": 1.01975965, + "epoch": 0.35291288402573195, + "flos": 30445461544320.0, + "grad_norm": 2.114277951858513, + "language_loss": 0.85707986, + "learning_rate": 3.0024725466000345e-06, + "loss": 0.87907434, + "num_input_tokens_seen": 63033400, + "step": 2935, + "time_per_iteration": 2.9951047897338867 + }, + { + "auxiliary_loss_clip": 0.01185744, + "auxiliary_loss_mlp": 0.01027014, + "balance_loss_clip": 1.05733156, + "balance_loss_mlp": 1.01781654, + "epoch": 0.35303312691637107, + "flos": 23112179763840.0, + "grad_norm": 2.0008844963903085, + "language_loss": 0.78956205, + "learning_rate": 3.0017984182627087e-06, + "loss": 0.81168962, + "num_input_tokens_seen": 63052725, + "step": 2936, + "time_per_iteration": 2.8638088703155518 + }, + { + "auxiliary_loss_clip": 0.01163106, + "auxiliary_loss_mlp": 0.00886798, + "balance_loss_clip": 1.04936242, + "balance_loss_mlp": 1.00057161, + "epoch": 0.3531533698070102, + "flos": 21835699165440.0, + "grad_norm": 1.9839899789887876, + "language_loss": 0.82443386, + "learning_rate": 3.00112413795336e-06, + "loss": 0.84493291, + "num_input_tokens_seen": 63072560, + "step": 2937, + "time_per_iteration": 2.919966220855713 + }, + { + "auxiliary_loss_clip": 0.01167387, + "auxiliary_loss_mlp": 0.01023187, + "balance_loss_clip": 1.04695594, + "balance_loss_mlp": 1.01432335, + "epoch": 0.35327361269764923, + "flos": 15778969810560.0, + "grad_norm": 2.3966345560601043, + "language_loss": 0.80200207, + "learning_rate": 3.000449705774275e-06, + "loss": 0.82390779, + "num_input_tokens_seen": 63090800, + "step": 2938, + "time_per_iteration": 2.8678317070007324 + }, + { + "auxiliary_loss_clip": 0.01185294, + "auxiliary_loss_mlp": 0.01032962, + "balance_loss_clip": 1.05632377, + "balance_loss_mlp": 1.02386594, + "epoch": 0.35339385558828834, + "flos": 22090413484800.0, + "grad_norm": 1.9178669800718533, + "language_loss": 0.72120124, + "learning_rate": 2.9997751218277654e-06, + "loss": 0.74338382, + "num_input_tokens_seen": 63108955, + "step": 2939, + "time_per_iteration": 2.7444093227386475 + }, + { + "auxiliary_loss_clip": 0.01192526, + "auxiliary_loss_mlp": 0.01029279, + "balance_loss_clip": 1.05528736, + "balance_loss_mlp": 1.02005863, + "epoch": 0.35351409847892745, + "flos": 24165008328960.0, + "grad_norm": 1.9908524382910062, + "language_loss": 0.77941811, + "learning_rate": 2.999100386216166e-06, + "loss": 0.80163616, + "num_input_tokens_seen": 63127895, + "step": 2940, + "time_per_iteration": 2.7153446674346924 + }, + { + "auxiliary_loss_clip": 0.01175138, + "auxiliary_loss_mlp": 0.0102905, + "balance_loss_clip": 1.05258751, + "balance_loss_mlp": 1.01976335, + "epoch": 0.3536343413695665, + "flos": 27052298340480.0, + "grad_norm": 1.8234633271593395, + "language_loss": 0.74470675, + "learning_rate": 2.998425499041831e-06, + "loss": 0.76674861, + "num_input_tokens_seen": 63148410, + "step": 2941, + "time_per_iteration": 2.84838604927063 + }, + { + "auxiliary_loss_clip": 0.01077752, + "auxiliary_loss_mlp": 0.01002607, + "balance_loss_clip": 1.01817107, + "balance_loss_mlp": 1.00030637, + "epoch": 0.3537545842602056, + "flos": 65991066370560.0, + "grad_norm": 1.2561825815997645, + "language_loss": 0.64542186, + "learning_rate": 2.997750460407142e-06, + "loss": 0.66622543, + "num_input_tokens_seen": 63209765, + "step": 2942, + "time_per_iteration": 3.4057369232177734 + }, + { + "auxiliary_loss_clip": 0.01165641, + "auxiliary_loss_mlp": 0.01030297, + "balance_loss_clip": 1.04970956, + "balance_loss_mlp": 1.02092123, + "epoch": 0.35387482715084473, + "flos": 18436897526400.0, + "grad_norm": 2.119703688101518, + "language_loss": 0.69903547, + "learning_rate": 2.997075270414501e-06, + "loss": 0.72099483, + "num_input_tokens_seen": 63226980, + "step": 2943, + "time_per_iteration": 2.8279571533203125 + }, + { + "auxiliary_loss_clip": 0.01072538, + "auxiliary_loss_mlp": 0.01006962, + "balance_loss_clip": 1.01898837, + "balance_loss_mlp": 1.00466144, + "epoch": 0.3539950700414838, + "flos": 65588579498880.0, + "grad_norm": 0.7032996300300358, + "language_loss": 0.57717657, + "learning_rate": 2.9963999291663347e-06, + "loss": 0.59797156, + "num_input_tokens_seen": 63292760, + "step": 2944, + "time_per_iteration": 3.326476573944092 + }, + { + "auxiliary_loss_clip": 0.01153136, + "auxiliary_loss_mlp": 0.01027015, + "balance_loss_clip": 1.05213404, + "balance_loss_mlp": 1.01782954, + "epoch": 0.3541153129321229, + "flos": 20521655919360.0, + "grad_norm": 2.5206661186856203, + "language_loss": 0.73736787, + "learning_rate": 2.9957244367650915e-06, + "loss": 0.75916934, + "num_input_tokens_seen": 63309005, + "step": 2945, + "time_per_iteration": 2.852313756942749 + }, + { + "auxiliary_loss_clip": 0.01149792, + "auxiliary_loss_mlp": 0.01029288, + "balance_loss_clip": 1.05100536, + "balance_loss_mlp": 1.01993608, + "epoch": 0.354235555822762, + "flos": 19573578391680.0, + "grad_norm": 1.8829344760107065, + "language_loss": 0.83955902, + "learning_rate": 2.9950487933132425e-06, + "loss": 0.86134982, + "num_input_tokens_seen": 63326420, + "step": 2946, + "time_per_iteration": 2.840986728668213 + }, + { + "auxiliary_loss_clip": 0.01184974, + "auxiliary_loss_mlp": 0.01032354, + "balance_loss_clip": 1.05184436, + "balance_loss_mlp": 1.02320421, + "epoch": 0.35435579871340106, + "flos": 20777268078720.0, + "grad_norm": 2.0838359524512273, + "language_loss": 0.71645057, + "learning_rate": 2.994372998913283e-06, + "loss": 0.73862386, + "num_input_tokens_seen": 63344925, + "step": 2947, + "time_per_iteration": 2.720618724822998 + }, + { + "auxiliary_loss_clip": 0.01175629, + "auxiliary_loss_mlp": 0.0103418, + "balance_loss_clip": 1.05688667, + "balance_loss_mlp": 1.02493489, + "epoch": 0.35447604160404017, + "flos": 23951807153280.0, + "grad_norm": 2.5599839406820584, + "language_loss": 0.6241647, + "learning_rate": 2.99369705366773e-06, + "loss": 0.64626276, + "num_input_tokens_seen": 63365170, + "step": 2948, + "time_per_iteration": 2.816241979598999 + }, + { + "auxiliary_loss_clip": 0.011688, + "auxiliary_loss_mlp": 0.01033287, + "balance_loss_clip": 1.05260968, + "balance_loss_mlp": 1.02403665, + "epoch": 0.3545962844946792, + "flos": 23435662671360.0, + "grad_norm": 2.154188056710702, + "language_loss": 0.82473671, + "learning_rate": 2.9930209576791244e-06, + "loss": 0.84675753, + "num_input_tokens_seen": 63383645, + "step": 2949, + "time_per_iteration": 2.88793683052063 + }, + { + "auxiliary_loss_clip": 0.01179551, + "auxiliary_loss_mlp": 0.01029434, + "balance_loss_clip": 1.05280602, + "balance_loss_mlp": 1.02104163, + "epoch": 0.35471652738531834, + "flos": 22085134185600.0, + "grad_norm": 2.081546990716248, + "language_loss": 0.63632065, + "learning_rate": 2.9923447110500285e-06, + "loss": 0.65841049, + "num_input_tokens_seen": 63402390, + "step": 2950, + "time_per_iteration": 2.7423243522644043 + }, + { + "auxiliary_loss_clip": 0.01173681, + "auxiliary_loss_mlp": 0.01029358, + "balance_loss_clip": 1.05303025, + "balance_loss_mlp": 1.02059054, + "epoch": 0.35483677027595745, + "flos": 27341881787520.0, + "grad_norm": 1.475731826491101, + "language_loss": 0.75277662, + "learning_rate": 2.9916683138830295e-06, + "loss": 0.77480698, + "num_input_tokens_seen": 63423055, + "step": 2951, + "time_per_iteration": 2.7520956993103027 + }, + { + "auxiliary_loss_clip": 0.01170418, + "auxiliary_loss_mlp": 0.01029162, + "balance_loss_clip": 1.05335999, + "balance_loss_mlp": 1.02069855, + "epoch": 0.3549570131665965, + "flos": 13516166678400.0, + "grad_norm": 1.9460459620438226, + "language_loss": 0.80656195, + "learning_rate": 2.9909917662807353e-06, + "loss": 0.82855773, + "num_input_tokens_seen": 63440855, + "step": 2952, + "time_per_iteration": 4.752667427062988 + }, + { + "auxiliary_loss_clip": 0.0118037, + "auxiliary_loss_mlp": 0.01030831, + "balance_loss_clip": 1.05393457, + "balance_loss_mlp": 1.02183688, + "epoch": 0.3550772560572356, + "flos": 20887549810560.0, + "grad_norm": 5.44604449078462, + "language_loss": 0.69146705, + "learning_rate": 2.9903150683457783e-06, + "loss": 0.71357906, + "num_input_tokens_seen": 63459400, + "step": 2953, + "time_per_iteration": 2.752819061279297 + }, + { + "auxiliary_loss_clip": 0.0117184, + "auxiliary_loss_mlp": 0.01027258, + "balance_loss_clip": 1.05114627, + "balance_loss_mlp": 1.01837671, + "epoch": 0.3551974989478747, + "flos": 20194042947840.0, + "grad_norm": 2.8567349887254188, + "language_loss": 0.65055501, + "learning_rate": 2.9896382201808126e-06, + "loss": 0.67254597, + "num_input_tokens_seen": 63476800, + "step": 2954, + "time_per_iteration": 3.7933335304260254 + }, + { + "auxiliary_loss_clip": 0.01194991, + "auxiliary_loss_mlp": 0.01026554, + "balance_loss_clip": 1.05621409, + "balance_loss_mlp": 1.01783979, + "epoch": 0.3553177418385138, + "flos": 19828831415040.0, + "grad_norm": 2.1747709098562424, + "language_loss": 0.81076205, + "learning_rate": 2.988961221888516e-06, + "loss": 0.83297747, + "num_input_tokens_seen": 63493475, + "step": 2955, + "time_per_iteration": 3.55419921875 + }, + { + "auxiliary_loss_clip": 0.01152011, + "auxiliary_loss_mlp": 0.01029919, + "balance_loss_clip": 1.04646242, + "balance_loss_mlp": 1.02049518, + "epoch": 0.3554379847291529, + "flos": 14829132516480.0, + "grad_norm": 6.043747340410672, + "language_loss": 0.78751725, + "learning_rate": 2.988284073571589e-06, + "loss": 0.80933654, + "num_input_tokens_seen": 63509560, + "step": 2956, + "time_per_iteration": 2.894536018371582 + }, + { + "auxiliary_loss_clip": 0.01184856, + "auxiliary_loss_mlp": 0.00886474, + "balance_loss_clip": 1.05402792, + "balance_loss_mlp": 1.00051141, + "epoch": 0.355558227619792, + "flos": 20485350247680.0, + "grad_norm": 2.4165342325105943, + "language_loss": 0.73219705, + "learning_rate": 2.9876067753327528e-06, + "loss": 0.75291038, + "num_input_tokens_seen": 63527290, + "step": 2957, + "time_per_iteration": 2.688918352127075 + }, + { + "auxiliary_loss_clip": 0.01183214, + "auxiliary_loss_mlp": 0.01033107, + "balance_loss_clip": 1.05082476, + "balance_loss_mlp": 1.02426124, + "epoch": 0.35567847051043106, + "flos": 37663613256960.0, + "grad_norm": 1.7426059944697914, + "language_loss": 0.80394137, + "learning_rate": 2.986929327274754e-06, + "loss": 0.82610458, + "num_input_tokens_seen": 63547870, + "step": 2958, + "time_per_iteration": 2.8557519912719727 + }, + { + "auxiliary_loss_clip": 0.01182026, + "auxiliary_loss_mlp": 0.01028584, + "balance_loss_clip": 1.05481744, + "balance_loss_mlp": 1.01978624, + "epoch": 0.35579871340107017, + "flos": 26943058103040.0, + "grad_norm": 1.5611708485404503, + "language_loss": 0.78540897, + "learning_rate": 2.9862517295003617e-06, + "loss": 0.80751503, + "num_input_tokens_seen": 63568285, + "step": 2959, + "time_per_iteration": 2.7433388233184814 + }, + { + "auxiliary_loss_clip": 0.01159987, + "auxiliary_loss_mlp": 0.01028198, + "balance_loss_clip": 1.04791594, + "balance_loss_mlp": 1.01932883, + "epoch": 0.3559189562917093, + "flos": 28293335193600.0, + "grad_norm": 1.569042105055357, + "language_loss": 0.72477472, + "learning_rate": 2.9855739821123654e-06, + "loss": 0.74665654, + "num_input_tokens_seen": 63589865, + "step": 2960, + "time_per_iteration": 2.886345624923706 + }, + { + "auxiliary_loss_clip": 0.01177589, + "auxiliary_loss_mlp": 0.01036981, + "balance_loss_clip": 1.0520134, + "balance_loss_mlp": 1.02756953, + "epoch": 0.35603919918234833, + "flos": 25664063552640.0, + "grad_norm": 2.0064845804775038, + "language_loss": 0.81908119, + "learning_rate": 2.98489608521358e-06, + "loss": 0.84122694, + "num_input_tokens_seen": 63609805, + "step": 2961, + "time_per_iteration": 2.7838327884674072 + }, + { + "auxiliary_loss_clip": 0.01186531, + "auxiliary_loss_mlp": 0.00886064, + "balance_loss_clip": 1.0534873, + "balance_loss_mlp": 1.00051355, + "epoch": 0.35615944207298744, + "flos": 23000856537600.0, + "grad_norm": 1.9633204754032834, + "language_loss": 0.7931639, + "learning_rate": 2.9842180389068425e-06, + "loss": 0.81388986, + "num_input_tokens_seen": 63627115, + "step": 2962, + "time_per_iteration": 2.8194611072540283 + }, + { + "auxiliary_loss_clip": 0.01063202, + "auxiliary_loss_mlp": 0.01005824, + "balance_loss_clip": 1.02023494, + "balance_loss_mlp": 1.00327265, + "epoch": 0.35627968496362655, + "flos": 68251283723520.0, + "grad_norm": 0.7666890604904958, + "language_loss": 0.59279472, + "learning_rate": 2.98353984329501e-06, + "loss": 0.61348492, + "num_input_tokens_seen": 63691460, + "step": 2963, + "time_per_iteration": 3.3703629970550537 + }, + { + "auxiliary_loss_clip": 0.0117408, + "auxiliary_loss_mlp": 0.01031612, + "balance_loss_clip": 1.05285025, + "balance_loss_mlp": 1.02215898, + "epoch": 0.3563999278542656, + "flos": 22641714403200.0, + "grad_norm": 1.986282272836154, + "language_loss": 0.7070269, + "learning_rate": 2.982861498480965e-06, + "loss": 0.72908384, + "num_input_tokens_seen": 63713840, + "step": 2964, + "time_per_iteration": 2.8089592456817627 + }, + { + "auxiliary_loss_clip": 0.01159433, + "auxiliary_loss_mlp": 0.01033124, + "balance_loss_clip": 1.04793191, + "balance_loss_mlp": 1.02480948, + "epoch": 0.3565201707449047, + "flos": 25952533678080.0, + "grad_norm": 2.1252898755765766, + "language_loss": 0.82601428, + "learning_rate": 2.9821830045676122e-06, + "loss": 0.84793985, + "num_input_tokens_seen": 63733540, + "step": 2965, + "time_per_iteration": 3.0001771450042725 + }, + { + "auxiliary_loss_clip": 0.01195238, + "auxiliary_loss_mlp": 0.01033561, + "balance_loss_clip": 1.05666566, + "balance_loss_mlp": 1.0250138, + "epoch": 0.3566404136355438, + "flos": 28475725478400.0, + "grad_norm": 1.637893374000454, + "language_loss": 0.73257959, + "learning_rate": 2.9815043616578793e-06, + "loss": 0.75486761, + "num_input_tokens_seen": 63754335, + "step": 2966, + "time_per_iteration": 2.8565261363983154 + }, + { + "auxiliary_loss_clip": 0.0116222, + "auxiliary_loss_mlp": 0.01034548, + "balance_loss_clip": 1.05049586, + "balance_loss_mlp": 1.0254817, + "epoch": 0.3567606565261829, + "flos": 38363117690880.0, + "grad_norm": 1.9609473265101236, + "language_loss": 0.77284825, + "learning_rate": 2.9808255698547145e-06, + "loss": 0.7948159, + "num_input_tokens_seen": 63777135, + "step": 2967, + "time_per_iteration": 2.940962553024292 + }, + { + "auxiliary_loss_clip": 0.0118203, + "auxiliary_loss_mlp": 0.01030533, + "balance_loss_clip": 1.0550468, + "balance_loss_mlp": 1.02104437, + "epoch": 0.356880899416822, + "flos": 21981029592960.0, + "grad_norm": 2.2714583805908477, + "language_loss": 0.79385936, + "learning_rate": 2.9801466292610913e-06, + "loss": 0.81598502, + "num_input_tokens_seen": 63797020, + "step": 2968, + "time_per_iteration": 2.8101789951324463 + }, + { + "auxiliary_loss_clip": 0.01182723, + "auxiliary_loss_mlp": 0.0103158, + "balance_loss_clip": 1.05394912, + "balance_loss_mlp": 1.02314639, + "epoch": 0.35700114230746105, + "flos": 18989132198400.0, + "grad_norm": 2.0201883978124977, + "language_loss": 0.8150667, + "learning_rate": 2.979467539980003e-06, + "loss": 0.8372097, + "num_input_tokens_seen": 63813810, + "step": 2969, + "time_per_iteration": 2.7045416831970215 + }, + { + "auxiliary_loss_clip": 0.01182895, + "auxiliary_loss_mlp": 0.01028922, + "balance_loss_clip": 1.05326557, + "balance_loss_mlp": 1.02009487, + "epoch": 0.35712138519810016, + "flos": 19756112330880.0, + "grad_norm": 1.8324553462631072, + "language_loss": 0.76747692, + "learning_rate": 2.978788302114468e-06, + "loss": 0.78959507, + "num_input_tokens_seen": 63830925, + "step": 2970, + "time_per_iteration": 2.789745807647705 + }, + { + "auxiliary_loss_clip": 0.01183321, + "auxiliary_loss_mlp": 0.01031669, + "balance_loss_clip": 1.05506384, + "balance_loss_mlp": 1.02249002, + "epoch": 0.35724162808873927, + "flos": 35183012008320.0, + "grad_norm": 2.1487659217157558, + "language_loss": 0.83565569, + "learning_rate": 2.9781089157675255e-06, + "loss": 0.85780561, + "num_input_tokens_seen": 63849385, + "step": 2971, + "time_per_iteration": 2.8225903511047363 + }, + { + "auxiliary_loss_clip": 0.01177611, + "auxiliary_loss_mlp": 0.01033578, + "balance_loss_clip": 1.05423188, + "balance_loss_mlp": 1.0240469, + "epoch": 0.3573618709793783, + "flos": 25556726736000.0, + "grad_norm": 1.4894806381973809, + "language_loss": 0.88369405, + "learning_rate": 2.977429381042238e-06, + "loss": 0.90580595, + "num_input_tokens_seen": 63870060, + "step": 2972, + "time_per_iteration": 2.7827768325805664 + }, + { + "auxiliary_loss_clip": 0.01172816, + "auxiliary_loss_mlp": 0.01031679, + "balance_loss_clip": 1.05088496, + "balance_loss_mlp": 1.02322745, + "epoch": 0.35748211387001744, + "flos": 29132352051840.0, + "grad_norm": 2.9614592091948806, + "language_loss": 0.88599288, + "learning_rate": 2.9767496980416913e-06, + "loss": 0.9080379, + "num_input_tokens_seen": 63889355, + "step": 2973, + "time_per_iteration": 2.8235161304473877 + }, + { + "auxiliary_loss_clip": 0.01167885, + "auxiliary_loss_mlp": 0.01032704, + "balance_loss_clip": 1.04974818, + "balance_loss_mlp": 1.02318549, + "epoch": 0.35760235676065655, + "flos": 13954169122560.0, + "grad_norm": 2.482064131963867, + "language_loss": 0.80996728, + "learning_rate": 2.9760698668689914e-06, + "loss": 0.83197314, + "num_input_tokens_seen": 63905580, + "step": 2974, + "time_per_iteration": 2.8542959690093994 + }, + { + "auxiliary_loss_clip": 0.01182409, + "auxiliary_loss_mlp": 0.01028649, + "balance_loss_clip": 1.05267739, + "balance_loss_mlp": 1.01986313, + "epoch": 0.3577225996512956, + "flos": 44018688977280.0, + "grad_norm": 3.2619116548794156, + "language_loss": 0.71470386, + "learning_rate": 2.975389887627269e-06, + "loss": 0.73681444, + "num_input_tokens_seen": 63928180, + "step": 2975, + "time_per_iteration": 2.9249629974365234 + }, + { + "auxiliary_loss_clip": 0.01166551, + "auxiliary_loss_mlp": 0.01030927, + "balance_loss_clip": 1.04871416, + "balance_loss_mlp": 1.02201629, + "epoch": 0.3578428425419347, + "flos": 17055199013760.0, + "grad_norm": 2.041283228666463, + "language_loss": 0.9023028, + "learning_rate": 2.9747097604196764e-06, + "loss": 0.92427766, + "num_input_tokens_seen": 63944825, + "step": 2976, + "time_per_iteration": 2.811040163040161 + }, + { + "auxiliary_loss_clip": 0.01056263, + "auxiliary_loss_mlp": 0.01006314, + "balance_loss_clip": 1.01672828, + "balance_loss_mlp": 1.00361943, + "epoch": 0.3579630854325738, + "flos": 71676550707840.0, + "grad_norm": 0.6726957227352623, + "language_loss": 0.56717479, + "learning_rate": 2.9740294853493875e-06, + "loss": 0.58780056, + "num_input_tokens_seen": 64016385, + "step": 2977, + "time_per_iteration": 4.52514386177063 + }, + { + "auxiliary_loss_clip": 0.01158047, + "auxiliary_loss_mlp": 0.01030958, + "balance_loss_clip": 1.04924107, + "balance_loss_mlp": 1.0220294, + "epoch": 0.3580833283232129, + "flos": 25046651652480.0, + "grad_norm": 2.187148038670997, + "language_loss": 0.67568731, + "learning_rate": 2.9733490625196008e-06, + "loss": 0.6975773, + "num_input_tokens_seen": 64036245, + "step": 2978, + "time_per_iteration": 2.9815773963928223 + }, + { + "auxiliary_loss_clip": 0.01154414, + "auxiliary_loss_mlp": 0.01034472, + "balance_loss_clip": 1.04911745, + "balance_loss_mlp": 1.02551317, + "epoch": 0.358203571213852, + "flos": 13953127628160.0, + "grad_norm": 3.376897469530355, + "language_loss": 0.75858295, + "learning_rate": 2.9726684920335353e-06, + "loss": 0.78047186, + "num_input_tokens_seen": 64054110, + "step": 2979, + "time_per_iteration": 3.679504632949829 + }, + { + "auxiliary_loss_clip": 0.01194904, + "auxiliary_loss_mlp": 0.0088741, + "balance_loss_clip": 1.05453253, + "balance_loss_mlp": 1.00063777, + "epoch": 0.35832381410449105, + "flos": 20302457172480.0, + "grad_norm": 3.114737382568236, + "language_loss": 0.81608373, + "learning_rate": 2.971987773994432e-06, + "loss": 0.83690691, + "num_input_tokens_seen": 64070295, + "step": 2980, + "time_per_iteration": 3.688812732696533 + }, + { + "auxiliary_loss_clip": 0.01174014, + "auxiliary_loss_mlp": 0.01031278, + "balance_loss_clip": 1.04980016, + "balance_loss_mlp": 1.02246797, + "epoch": 0.35844405699513016, + "flos": 16983234115200.0, + "grad_norm": 1.8543564050802535, + "language_loss": 0.82809627, + "learning_rate": 2.9713069085055566e-06, + "loss": 0.85014915, + "num_input_tokens_seen": 64088605, + "step": 2981, + "time_per_iteration": 3.6802237033843994 + }, + { + "auxiliary_loss_clip": 0.01165407, + "auxiliary_loss_mlp": 0.01033679, + "balance_loss_clip": 1.05276, + "balance_loss_mlp": 1.02460694, + "epoch": 0.35856429988576927, + "flos": 23216858974080.0, + "grad_norm": 2.1058170992611456, + "language_loss": 0.79024792, + "learning_rate": 2.9706258956701958e-06, + "loss": 0.81223881, + "num_input_tokens_seen": 64108595, + "step": 2982, + "time_per_iteration": 2.8362863063812256 + }, + { + "auxiliary_loss_clip": 0.01185902, + "auxiliary_loss_mlp": 0.01030134, + "balance_loss_clip": 1.05381393, + "balance_loss_mlp": 1.02149749, + "epoch": 0.3586845427764083, + "flos": 23034576430080.0, + "grad_norm": 2.279980246107894, + "language_loss": 0.77279043, + "learning_rate": 2.9699447355916575e-06, + "loss": 0.79495078, + "num_input_tokens_seen": 64127405, + "step": 2983, + "time_per_iteration": 2.7283148765563965 + }, + { + "auxiliary_loss_clip": 0.01188538, + "auxiliary_loss_mlp": 0.00886484, + "balance_loss_clip": 1.05285501, + "balance_loss_mlp": 1.00083327, + "epoch": 0.35880478566704743, + "flos": 20010682995840.0, + "grad_norm": 2.695029737818318, + "language_loss": 0.73844492, + "learning_rate": 2.969263428373275e-06, + "loss": 0.75919521, + "num_input_tokens_seen": 64145755, + "step": 2984, + "time_per_iteration": 2.757639169692993 + }, + { + "auxiliary_loss_clip": 0.01174257, + "auxiliary_loss_mlp": 0.01031334, + "balance_loss_clip": 1.05165768, + "balance_loss_mlp": 1.02230954, + "epoch": 0.35892502855768654, + "flos": 13699095667200.0, + "grad_norm": 2.449992748489761, + "language_loss": 0.79226696, + "learning_rate": 2.9685819741184007e-06, + "loss": 0.81432283, + "num_input_tokens_seen": 64164195, + "step": 2985, + "time_per_iteration": 2.7501394748687744 + }, + { + "auxiliary_loss_clip": 0.01156192, + "auxiliary_loss_mlp": 0.01029858, + "balance_loss_clip": 1.04887962, + "balance_loss_mlp": 1.02064943, + "epoch": 0.3590452714483256, + "flos": 18114096977280.0, + "grad_norm": 2.5815360751046756, + "language_loss": 0.68577731, + "learning_rate": 2.967900372930411e-06, + "loss": 0.70763779, + "num_input_tokens_seen": 64182705, + "step": 2986, + "time_per_iteration": 2.8191545009613037 + }, + { + "auxiliary_loss_clip": 0.01167079, + "auxiliary_loss_mlp": 0.01031792, + "balance_loss_clip": 1.04996753, + "balance_loss_mlp": 1.02199268, + "epoch": 0.3591655143389647, + "flos": 17749352321280.0, + "grad_norm": 2.202604304401094, + "language_loss": 0.78902864, + "learning_rate": 2.9672186249127046e-06, + "loss": 0.81101739, + "num_input_tokens_seen": 64202170, + "step": 2987, + "time_per_iteration": 2.706946849822998 + }, + { + "auxiliary_loss_clip": 0.01169381, + "auxiliary_loss_mlp": 0.01030937, + "balance_loss_clip": 1.05196524, + "balance_loss_mlp": 1.02241349, + "epoch": 0.3592857572296038, + "flos": 25224409082880.0, + "grad_norm": 2.2852440087053583, + "language_loss": 0.78915656, + "learning_rate": 2.9665367301687014e-06, + "loss": 0.81115973, + "num_input_tokens_seen": 64220415, + "step": 2988, + "time_per_iteration": 2.788571357727051 + }, + { + "auxiliary_loss_clip": 0.01164112, + "auxiliary_loss_mlp": 0.01032829, + "balance_loss_clip": 1.04923916, + "balance_loss_mlp": 1.02363181, + "epoch": 0.3594060001202429, + "flos": 29384408764800.0, + "grad_norm": 2.1107706749153055, + "language_loss": 0.76308024, + "learning_rate": 2.965854688801845e-06, + "loss": 0.78504968, + "num_input_tokens_seen": 64242475, + "step": 2989, + "time_per_iteration": 2.7965283393859863 + }, + { + "auxiliary_loss_clip": 0.01178162, + "auxiliary_loss_mlp": 0.01029358, + "balance_loss_clip": 1.04853547, + "balance_loss_mlp": 1.02066755, + "epoch": 0.359526243010882, + "flos": 17052900543360.0, + "grad_norm": 1.9888347767651395, + "language_loss": 0.76535559, + "learning_rate": 2.9651725009156005e-06, + "loss": 0.78743082, + "num_input_tokens_seen": 64260220, + "step": 2990, + "time_per_iteration": 2.8083925247192383 + }, + { + "auxiliary_loss_clip": 0.01162427, + "auxiliary_loss_mlp": 0.01028602, + "balance_loss_clip": 1.0485394, + "balance_loss_mlp": 1.01921439, + "epoch": 0.3596464859015211, + "flos": 22965089569920.0, + "grad_norm": 1.6071728366154923, + "language_loss": 0.74446112, + "learning_rate": 2.964490166613454e-06, + "loss": 0.76637143, + "num_input_tokens_seen": 64280145, + "step": 2991, + "time_per_iteration": 2.8848657608032227 + }, + { + "auxiliary_loss_clip": 0.01085053, + "auxiliary_loss_mlp": 0.01002625, + "balance_loss_clip": 1.01887584, + "balance_loss_mlp": 1.00050342, + "epoch": 0.35976672879216015, + "flos": 54739462590720.0, + "grad_norm": 0.7564222308601954, + "language_loss": 0.57723778, + "learning_rate": 2.963807685998917e-06, + "loss": 0.59811461, + "num_input_tokens_seen": 64336010, + "step": 2992, + "time_per_iteration": 3.0304298400878906 + }, + { + "auxiliary_loss_clip": 0.01153214, + "auxiliary_loss_mlp": 0.01030832, + "balance_loss_clip": 1.04619861, + "balance_loss_mlp": 1.02243376, + "epoch": 0.35988697168279926, + "flos": 43139020901760.0, + "grad_norm": 1.6837530816808715, + "language_loss": 0.77612031, + "learning_rate": 2.9631250591755196e-06, + "loss": 0.79796076, + "num_input_tokens_seen": 64358725, + "step": 2993, + "time_per_iteration": 3.063349962234497 + }, + { + "auxiliary_loss_clip": 0.01168362, + "auxiliary_loss_mlp": 0.01031477, + "balance_loss_clip": 1.05224168, + "balance_loss_mlp": 1.02260804, + "epoch": 0.36000721457343837, + "flos": 35845600239360.0, + "grad_norm": 1.781250391054254, + "language_loss": 0.57640016, + "learning_rate": 2.962442286246817e-06, + "loss": 0.59839857, + "num_input_tokens_seen": 64381555, + "step": 2994, + "time_per_iteration": 2.8446195125579834 + }, + { + "auxiliary_loss_clip": 0.01172026, + "auxiliary_loss_mlp": 0.01039854, + "balance_loss_clip": 1.0501368, + "balance_loss_mlp": 1.03072286, + "epoch": 0.3601274574640774, + "flos": 18291100222080.0, + "grad_norm": 1.5859386905281003, + "language_loss": 0.69626832, + "learning_rate": 2.9617593673163853e-06, + "loss": 0.71838713, + "num_input_tokens_seen": 64400375, + "step": 2995, + "time_per_iteration": 2.868333339691162 + }, + { + "auxiliary_loss_clip": 0.01171043, + "auxiliary_loss_mlp": 0.01031982, + "balance_loss_clip": 1.04723644, + "balance_loss_mlp": 1.02326739, + "epoch": 0.36024770035471654, + "flos": 13333955961600.0, + "grad_norm": 2.0653766496856223, + "language_loss": 0.77491629, + "learning_rate": 2.9610763024878216e-06, + "loss": 0.79694659, + "num_input_tokens_seen": 64415880, + "step": 2996, + "time_per_iteration": 2.723463773727417 + }, + { + "auxiliary_loss_clip": 0.011676, + "auxiliary_loss_mlp": 0.01031738, + "balance_loss_clip": 1.05114555, + "balance_loss_mlp": 1.02244532, + "epoch": 0.3603679432453556, + "flos": 20267013427200.0, + "grad_norm": 1.7383148176348207, + "language_loss": 0.91585439, + "learning_rate": 2.960393091864747e-06, + "loss": 0.93784773, + "num_input_tokens_seen": 64434260, + "step": 2997, + "time_per_iteration": 2.8244972229003906 + }, + { + "auxiliary_loss_clip": 0.01172369, + "auxiliary_loss_mlp": 0.01034521, + "balance_loss_clip": 1.05197525, + "balance_loss_mlp": 1.02597344, + "epoch": 0.3604881861359947, + "flos": 22451135817600.0, + "grad_norm": 1.7278921596886196, + "language_loss": 0.75003201, + "learning_rate": 2.959709735550804e-06, + "loss": 0.77210093, + "num_input_tokens_seen": 64453855, + "step": 2998, + "time_per_iteration": 2.801133871078491 + }, + { + "auxiliary_loss_clip": 0.01155906, + "auxiliary_loss_mlp": 0.01029507, + "balance_loss_clip": 1.04737902, + "balance_loss_mlp": 1.02096581, + "epoch": 0.3606084290266338, + "flos": 22054251467520.0, + "grad_norm": 2.585155270271945, + "language_loss": 0.75695038, + "learning_rate": 2.9590262336496575e-06, + "loss": 0.77880454, + "num_input_tokens_seen": 64473585, + "step": 2999, + "time_per_iteration": 2.9129624366760254 + }, + { + "auxiliary_loss_clip": 0.01156859, + "auxiliary_loss_mlp": 0.01034208, + "balance_loss_clip": 1.05100417, + "balance_loss_mlp": 1.0252614, + "epoch": 0.36072867191727287, + "flos": 15632921111040.0, + "grad_norm": 1.9931165447243462, + "language_loss": 0.85271364, + "learning_rate": 2.9583425862649936e-06, + "loss": 0.87462431, + "num_input_tokens_seen": 64491720, + "step": 3000, + "time_per_iteration": 2.8818414211273193 + }, + { + "auxiliary_loss_clip": 0.01194347, + "auxiliary_loss_mlp": 0.01025467, + "balance_loss_clip": 1.05632603, + "balance_loss_mlp": 1.01633537, + "epoch": 0.360848914807912, + "flos": 19677000625920.0, + "grad_norm": 1.964472816512517, + "language_loss": 0.73983115, + "learning_rate": 2.9576587935005215e-06, + "loss": 0.76202929, + "num_input_tokens_seen": 64509800, + "step": 3001, + "time_per_iteration": 2.7712953090667725 + }, + { + "auxiliary_loss_clip": 0.01183236, + "auxiliary_loss_mlp": 0.010286, + "balance_loss_clip": 1.05157936, + "balance_loss_mlp": 1.01908731, + "epoch": 0.3609691576985511, + "flos": 18877808972160.0, + "grad_norm": 3.018296996881954, + "language_loss": 0.72018397, + "learning_rate": 2.9569748554599713e-06, + "loss": 0.7423023, + "num_input_tokens_seen": 64525410, + "step": 3002, + "time_per_iteration": 2.863353967666626 + }, + { + "auxiliary_loss_clip": 0.01170001, + "auxiliary_loss_mlp": 0.01027294, + "balance_loss_clip": 1.05055523, + "balance_loss_mlp": 1.01886594, + "epoch": 0.36108940058919015, + "flos": 42224088648960.0, + "grad_norm": 1.991324632879271, + "language_loss": 0.73078287, + "learning_rate": 2.956290772247097e-06, + "loss": 0.75275582, + "num_input_tokens_seen": 64544085, + "step": 3003, + "time_per_iteration": 3.8272488117218018 + }, + { + "auxiliary_loss_clip": 0.01145159, + "auxiliary_loss_mlp": 0.01028684, + "balance_loss_clip": 1.04821515, + "balance_loss_mlp": 1.02042854, + "epoch": 0.36120964347982926, + "flos": 23185150243200.0, + "grad_norm": 2.85571652741967, + "language_loss": 0.73179698, + "learning_rate": 2.9556065439656724e-06, + "loss": 0.75353539, + "num_input_tokens_seen": 64563135, + "step": 3004, + "time_per_iteration": 2.8754937648773193 + }, + { + "auxiliary_loss_clip": 0.01132016, + "auxiliary_loss_mlp": 0.01029385, + "balance_loss_clip": 1.04163051, + "balance_loss_mlp": 1.02074194, + "epoch": 0.36132988637046837, + "flos": 18113055482880.0, + "grad_norm": 1.6599093968323742, + "language_loss": 0.81716514, + "learning_rate": 2.9549221707194952e-06, + "loss": 0.83877915, + "num_input_tokens_seen": 64581985, + "step": 3005, + "time_per_iteration": 3.8500773906707764 + }, + { + "auxiliary_loss_clip": 0.01182652, + "auxiliary_loss_mlp": 0.01032039, + "balance_loss_clip": 1.05321074, + "balance_loss_mlp": 1.02296114, + "epoch": 0.3614501292611074, + "flos": 27813101333760.0, + "grad_norm": 2.174938884700504, + "language_loss": 0.72699028, + "learning_rate": 2.954237652612384e-06, + "loss": 0.74913716, + "num_input_tokens_seen": 64601035, + "step": 3006, + "time_per_iteration": 3.7876572608947754 + }, + { + "auxiliary_loss_clip": 0.01166063, + "auxiliary_loss_mlp": 0.01023787, + "balance_loss_clip": 1.04984462, + "balance_loss_mlp": 1.01583552, + "epoch": 0.36157037215174653, + "flos": 22634926732800.0, + "grad_norm": 2.0239455029159394, + "language_loss": 0.84592265, + "learning_rate": 2.9535529897481796e-06, + "loss": 0.86782116, + "num_input_tokens_seen": 64618580, + "step": 3007, + "time_per_iteration": 3.7188973426818848 + }, + { + "auxiliary_loss_clip": 0.01188475, + "auxiliary_loss_mlp": 0.01030704, + "balance_loss_clip": 1.0515976, + "balance_loss_mlp": 1.0212208, + "epoch": 0.36169061504238564, + "flos": 12600839376000.0, + "grad_norm": 3.460253338925196, + "language_loss": 0.76913047, + "learning_rate": 2.9528681822307446e-06, + "loss": 0.79132223, + "num_input_tokens_seen": 64635430, + "step": 3008, + "time_per_iteration": 2.773197889328003 + }, + { + "auxiliary_loss_clip": 0.01178209, + "auxiliary_loss_mlp": 0.00885932, + "balance_loss_clip": 1.05566883, + "balance_loss_mlp": 1.00076938, + "epoch": 0.3618108579330247, + "flos": 26684644682880.0, + "grad_norm": 3.270131924986329, + "language_loss": 0.82244909, + "learning_rate": 2.952183230163964e-06, + "loss": 0.84309047, + "num_input_tokens_seen": 64655005, + "step": 3009, + "time_per_iteration": 2.8222768306732178 + }, + { + "auxiliary_loss_clip": 0.0115102, + "auxiliary_loss_mlp": 0.01028966, + "balance_loss_clip": 1.04451442, + "balance_loss_mlp": 1.02081847, + "epoch": 0.3619311008236638, + "flos": 22817029708800.0, + "grad_norm": 2.0920437988647205, + "language_loss": 0.73131859, + "learning_rate": 2.9514981336517448e-06, + "loss": 0.75311852, + "num_input_tokens_seen": 64674775, + "step": 3010, + "time_per_iteration": 2.932278633117676 + }, + { + "auxiliary_loss_clip": 0.0118236, + "auxiliary_loss_mlp": 0.01030164, + "balance_loss_clip": 1.05513167, + "balance_loss_mlp": 1.02140796, + "epoch": 0.36205134371430286, + "flos": 25919603884800.0, + "grad_norm": 1.8913840382888083, + "language_loss": 0.81236398, + "learning_rate": 2.950812892798015e-06, + "loss": 0.83448923, + "num_input_tokens_seen": 64695670, + "step": 3011, + "time_per_iteration": 2.8823482990264893 + }, + { + "auxiliary_loss_clip": 0.01142899, + "auxiliary_loss_mlp": 0.00885961, + "balance_loss_clip": 1.04688048, + "balance_loss_mlp": 1.00084901, + "epoch": 0.362171586604942, + "flos": 26139592730880.0, + "grad_norm": 1.6608387257624868, + "language_loss": 0.87166333, + "learning_rate": 2.9501275077067256e-06, + "loss": 0.89195192, + "num_input_tokens_seen": 64716290, + "step": 3012, + "time_per_iteration": 2.963076591491699 + }, + { + "auxiliary_loss_clip": 0.01123792, + "auxiliary_loss_mlp": 0.01031968, + "balance_loss_clip": 1.04112506, + "balance_loss_mlp": 1.0233078, + "epoch": 0.3622918294955811, + "flos": 28074208273920.0, + "grad_norm": 1.4657918596136514, + "language_loss": 0.88438213, + "learning_rate": 2.949441978481848e-06, + "loss": 0.90593976, + "num_input_tokens_seen": 64737190, + "step": 3013, + "time_per_iteration": 3.085933208465576 + }, + { + "auxiliary_loss_clip": 0.0116379, + "auxiliary_loss_mlp": 0.01028583, + "balance_loss_clip": 1.04956412, + "balance_loss_mlp": 1.01957655, + "epoch": 0.36241207238622014, + "flos": 19828005402240.0, + "grad_norm": 2.0166985993336484, + "language_loss": 0.79910898, + "learning_rate": 2.9487563052273778e-06, + "loss": 0.8210327, + "num_input_tokens_seen": 64753950, + "step": 3014, + "time_per_iteration": 2.9506964683532715 + }, + { + "auxiliary_loss_clip": 0.01173973, + "auxiliary_loss_mlp": 0.0102965, + "balance_loss_clip": 1.05312717, + "balance_loss_mlp": 1.0210669, + "epoch": 0.36253231527685925, + "flos": 21397158017280.0, + "grad_norm": 1.9393476157241352, + "language_loss": 0.85497439, + "learning_rate": 2.94807048804733e-06, + "loss": 0.87701058, + "num_input_tokens_seen": 64773570, + "step": 3015, + "time_per_iteration": 2.898310661315918 + }, + { + "auxiliary_loss_clip": 0.01164219, + "auxiliary_loss_mlp": 0.01032147, + "balance_loss_clip": 1.04914165, + "balance_loss_mlp": 1.02284837, + "epoch": 0.36265255816749836, + "flos": 18362885552640.0, + "grad_norm": 2.338361751279346, + "language_loss": 0.90204155, + "learning_rate": 2.9473845270457434e-06, + "loss": 0.92400521, + "num_input_tokens_seen": 64790385, + "step": 3016, + "time_per_iteration": 2.8255085945129395 + }, + { + "auxiliary_loss_clip": 0.01165289, + "auxiliary_loss_mlp": 0.01027399, + "balance_loss_clip": 1.05211139, + "balance_loss_mlp": 1.01912642, + "epoch": 0.3627728010581374, + "flos": 18660046769280.0, + "grad_norm": 2.875275026634928, + "language_loss": 0.69869447, + "learning_rate": 2.946698422326677e-06, + "loss": 0.72062135, + "num_input_tokens_seen": 64807845, + "step": 3017, + "time_per_iteration": 2.9012343883514404 + }, + { + "auxiliary_loss_clip": 0.0114972, + "auxiliary_loss_mlp": 0.01029254, + "balance_loss_clip": 1.04485631, + "balance_loss_mlp": 1.02043891, + "epoch": 0.36289304394877653, + "flos": 27524272072320.0, + "grad_norm": 2.1485033864192773, + "language_loss": 0.79640663, + "learning_rate": 2.946012173994213e-06, + "loss": 0.81819642, + "num_input_tokens_seen": 64827630, + "step": 3018, + "time_per_iteration": 3.0887956619262695 + }, + { + "auxiliary_loss_clip": 0.01176927, + "auxiliary_loss_mlp": 0.01030758, + "balance_loss_clip": 1.05540693, + "balance_loss_mlp": 1.02231216, + "epoch": 0.36301328683941564, + "flos": 34533244932480.0, + "grad_norm": 1.4184300015185698, + "language_loss": 0.67529786, + "learning_rate": 2.945325782152454e-06, + "loss": 0.6973747, + "num_input_tokens_seen": 64850665, + "step": 3019, + "time_per_iteration": 2.8205461502075195 + }, + { + "auxiliary_loss_clip": 0.01173236, + "auxiliary_loss_mlp": 0.01028611, + "balance_loss_clip": 1.04937303, + "balance_loss_mlp": 1.0200938, + "epoch": 0.3631335297300547, + "flos": 19025976574080.0, + "grad_norm": 2.052207112437658, + "language_loss": 0.78891748, + "learning_rate": 2.9446392469055257e-06, + "loss": 0.81093591, + "num_input_tokens_seen": 64868700, + "step": 3020, + "time_per_iteration": 2.809573173522949 + }, + { + "auxiliary_loss_clip": 0.01158174, + "auxiliary_loss_mlp": 0.01028625, + "balance_loss_clip": 1.05255711, + "balance_loss_mlp": 1.0204947, + "epoch": 0.3632537726206938, + "flos": 19536769929600.0, + "grad_norm": 1.8344299084535092, + "language_loss": 0.7972275, + "learning_rate": 2.9439525683575745e-06, + "loss": 0.81909549, + "num_input_tokens_seen": 64887620, + "step": 3021, + "time_per_iteration": 2.8172998428344727 + }, + { + "auxiliary_loss_clip": 0.01194529, + "auxiliary_loss_mlp": 0.01035523, + "balance_loss_clip": 1.05656922, + "balance_loss_mlp": 1.02662373, + "epoch": 0.3633740155113329, + "flos": 21068611292160.0, + "grad_norm": 2.322823775013044, + "language_loss": 0.75308198, + "learning_rate": 2.9432657466127694e-06, + "loss": 0.77538246, + "num_input_tokens_seen": 64907190, + "step": 3022, + "time_per_iteration": 2.6790244579315186 + }, + { + "auxiliary_loss_clip": 0.01155321, + "auxiliary_loss_mlp": 0.01038819, + "balance_loss_clip": 1.0546515, + "balance_loss_mlp": 1.03002715, + "epoch": 0.36349425840197197, + "flos": 20298722158080.0, + "grad_norm": 1.5843377379308639, + "language_loss": 0.75987208, + "learning_rate": 2.9425787817753007e-06, + "loss": 0.7818135, + "num_input_tokens_seen": 64925850, + "step": 3023, + "time_per_iteration": 2.861985206604004 + }, + { + "auxiliary_loss_clip": 0.01158933, + "auxiliary_loss_mlp": 0.01030721, + "balance_loss_clip": 1.04937148, + "balance_loss_mlp": 1.02231073, + "epoch": 0.3636145012926111, + "flos": 29716762331520.0, + "grad_norm": 3.533527226018886, + "language_loss": 0.71844023, + "learning_rate": 2.94189167394938e-06, + "loss": 0.74033678, + "num_input_tokens_seen": 64948285, + "step": 3024, + "time_per_iteration": 2.997490882873535 + }, + { + "auxiliary_loss_clip": 0.01192431, + "auxiliary_loss_mlp": 0.01030471, + "balance_loss_clip": 1.05723321, + "balance_loss_mlp": 1.02127981, + "epoch": 0.3637347441832502, + "flos": 21431847576960.0, + "grad_norm": 1.8740304075896803, + "language_loss": 0.80865467, + "learning_rate": 2.941204423239241e-06, + "loss": 0.83088374, + "num_input_tokens_seen": 64967160, + "step": 3025, + "time_per_iteration": 2.7004973888397217 + }, + { + "auxiliary_loss_clip": 0.0117863, + "auxiliary_loss_mlp": 0.01037211, + "balance_loss_clip": 1.05395699, + "balance_loss_mlp": 1.02866995, + "epoch": 0.36385498707388925, + "flos": 29533941083520.0, + "grad_norm": 1.889813344501146, + "language_loss": 0.76399404, + "learning_rate": 2.9405170297491395e-06, + "loss": 0.78615242, + "num_input_tokens_seen": 64987155, + "step": 3026, + "time_per_iteration": 2.7082643508911133 + }, + { + "auxiliary_loss_clip": 0.01130839, + "auxiliary_loss_mlp": 0.00886021, + "balance_loss_clip": 1.05004048, + "balance_loss_mlp": 1.00065827, + "epoch": 0.36397522996452836, + "flos": 22236569925120.0, + "grad_norm": 2.168593163669292, + "language_loss": 0.79957074, + "learning_rate": 2.939829493583353e-06, + "loss": 0.81973934, + "num_input_tokens_seen": 65003800, + "step": 3027, + "time_per_iteration": 2.8899166584014893 + }, + { + "auxiliary_loss_clip": 0.0115434, + "auxiliary_loss_mlp": 0.01027459, + "balance_loss_clip": 1.04636145, + "balance_loss_mlp": 1.01895952, + "epoch": 0.3640954728551674, + "flos": 21506505995520.0, + "grad_norm": 2.6754490134007405, + "language_loss": 0.83117032, + "learning_rate": 2.939141814846179e-06, + "loss": 0.85298824, + "num_input_tokens_seen": 65021215, + "step": 3028, + "time_per_iteration": 2.8618295192718506 + }, + { + "auxiliary_loss_clip": 0.01169617, + "auxiliary_loss_mlp": 0.01028958, + "balance_loss_clip": 1.05031276, + "balance_loss_mlp": 1.02001739, + "epoch": 0.3642157157458065, + "flos": 17712867081600.0, + "grad_norm": 1.561993998150019, + "language_loss": 0.82566833, + "learning_rate": 2.938453993641938e-06, + "loss": 0.8476541, + "num_input_tokens_seen": 65039590, + "step": 3029, + "time_per_iteration": 3.014601945877075 + }, + { + "auxiliary_loss_clip": 0.01173379, + "auxiliary_loss_mlp": 0.01031933, + "balance_loss_clip": 1.05466795, + "balance_loss_mlp": 1.02259898, + "epoch": 0.36433595863644563, + "flos": 17639537466240.0, + "grad_norm": 2.2181262632478242, + "language_loss": 0.70468175, + "learning_rate": 2.937766030074973e-06, + "loss": 0.72673488, + "num_input_tokens_seen": 65056845, + "step": 3030, + "time_per_iteration": 4.69536018371582 + }, + { + "auxiliary_loss_clip": 0.011624, + "auxiliary_loss_mlp": 0.01030297, + "balance_loss_clip": 1.04886794, + "balance_loss_mlp": 1.02152336, + "epoch": 0.3644562015270847, + "flos": 26833279161600.0, + "grad_norm": 1.7653185139989958, + "language_loss": 0.82393062, + "learning_rate": 2.937077924249646e-06, + "loss": 0.8458575, + "num_input_tokens_seen": 65079435, + "step": 3031, + "time_per_iteration": 3.7081692218780518 + }, + { + "auxiliary_loss_clip": 0.01176538, + "auxiliary_loss_mlp": 0.01036369, + "balance_loss_clip": 1.05174279, + "balance_loss_mlp": 1.02790487, + "epoch": 0.3645764444177238, + "flos": 14282715847680.0, + "grad_norm": 2.5019025227614557, + "language_loss": 0.74970192, + "learning_rate": 2.9363896762703443e-06, + "loss": 0.77183092, + "num_input_tokens_seen": 65096500, + "step": 3032, + "time_per_iteration": 2.805006742477417 + }, + { + "auxiliary_loss_clip": 0.01188603, + "auxiliary_loss_mlp": 0.0102782, + "balance_loss_clip": 1.05329645, + "balance_loss_mlp": 1.01868236, + "epoch": 0.3646966873083629, + "flos": 20667489137280.0, + "grad_norm": 1.687492310446557, + "language_loss": 0.84239846, + "learning_rate": 2.9357012862414725e-06, + "loss": 0.86456269, + "num_input_tokens_seen": 65115860, + "step": 3033, + "time_per_iteration": 3.7059967517852783 + }, + { + "auxiliary_loss_clip": 0.01180983, + "auxiliary_loss_mlp": 0.01033627, + "balance_loss_clip": 1.05382574, + "balance_loss_mlp": 1.02502, + "epoch": 0.36481693019900197, + "flos": 27782613665280.0, + "grad_norm": 2.3174992028883783, + "language_loss": 0.71352673, + "learning_rate": 2.9350127542674593e-06, + "loss": 0.73567283, + "num_input_tokens_seen": 65138070, + "step": 3034, + "time_per_iteration": 2.7632806301116943 + }, + { + "auxiliary_loss_clip": 0.01173357, + "auxiliary_loss_mlp": 0.01034916, + "balance_loss_clip": 1.05189216, + "balance_loss_mlp": 1.02682221, + "epoch": 0.3649371730896411, + "flos": 19712588025600.0, + "grad_norm": 2.1264478618416094, + "language_loss": 0.76386762, + "learning_rate": 2.934324080452755e-06, + "loss": 0.78595036, + "num_input_tokens_seen": 65155860, + "step": 3035, + "time_per_iteration": 2.7591984272003174 + }, + { + "auxiliary_loss_clip": 0.01149645, + "auxiliary_loss_mlp": 0.00886911, + "balance_loss_clip": 1.04586506, + "balance_loss_mlp": 1.00076818, + "epoch": 0.3650574159802802, + "flos": 24750496016640.0, + "grad_norm": 1.4617897161766218, + "language_loss": 0.78397757, + "learning_rate": 2.9336352649018307e-06, + "loss": 0.8043431, + "num_input_tokens_seen": 65175930, + "step": 3036, + "time_per_iteration": 2.916267156600952 + }, + { + "auxiliary_loss_clip": 0.01173284, + "auxiliary_loss_mlp": 0.01035307, + "balance_loss_clip": 1.05388618, + "balance_loss_mlp": 1.02609229, + "epoch": 0.36517765887091924, + "flos": 32853487363200.0, + "grad_norm": 1.7032766872718963, + "language_loss": 0.7020846, + "learning_rate": 2.9329463077191783e-06, + "loss": 0.72417051, + "num_input_tokens_seen": 65199305, + "step": 3037, + "time_per_iteration": 2.8765838146209717 + }, + { + "auxiliary_loss_clip": 0.01149628, + "auxiliary_loss_mlp": 0.01031139, + "balance_loss_clip": 1.04846239, + "balance_loss_mlp": 1.0211966, + "epoch": 0.36529790176155835, + "flos": 20120318282880.0, + "grad_norm": 2.218150149957531, + "language_loss": 0.6428858, + "learning_rate": 2.9322572090093135e-06, + "loss": 0.66469347, + "num_input_tokens_seen": 65218010, + "step": 3038, + "time_per_iteration": 2.88620924949646 + }, + { + "auxiliary_loss_clip": 0.01147865, + "auxiliary_loss_mlp": 0.01026287, + "balance_loss_clip": 1.04425764, + "balance_loss_mlp": 1.01772785, + "epoch": 0.36541814465219746, + "flos": 17639573379840.0, + "grad_norm": 2.7827335304091982, + "language_loss": 0.76488304, + "learning_rate": 2.9315679688767713e-06, + "loss": 0.78662455, + "num_input_tokens_seen": 65236020, + "step": 3039, + "time_per_iteration": 2.7832610607147217 + }, + { + "auxiliary_loss_clip": 0.01165266, + "auxiliary_loss_mlp": 0.01031916, + "balance_loss_clip": 1.04884672, + "balance_loss_mlp": 1.0233326, + "epoch": 0.3655383875428365, + "flos": 22674356887680.0, + "grad_norm": 1.5518002235322927, + "language_loss": 0.66428769, + "learning_rate": 2.9308785874261085e-06, + "loss": 0.68625951, + "num_input_tokens_seen": 65256210, + "step": 3040, + "time_per_iteration": 2.8393588066101074 + }, + { + "auxiliary_loss_clip": 0.01193747, + "auxiliary_loss_mlp": 0.01034472, + "balance_loss_clip": 1.05711091, + "balance_loss_mlp": 1.02588868, + "epoch": 0.36565863043347563, + "flos": 21981173247360.0, + "grad_norm": 1.7026264763433276, + "language_loss": 0.81639761, + "learning_rate": 2.9301890647619045e-06, + "loss": 0.83867979, + "num_input_tokens_seen": 65275505, + "step": 3041, + "time_per_iteration": 2.7324721813201904 + }, + { + "auxiliary_loss_clip": 0.01177001, + "auxiliary_loss_mlp": 0.01024698, + "balance_loss_clip": 1.0539608, + "balance_loss_mlp": 1.01566172, + "epoch": 0.36577887332411474, + "flos": 24827632473600.0, + "grad_norm": 4.442520167228817, + "language_loss": 0.80235255, + "learning_rate": 2.929499400988759e-06, + "loss": 0.82436955, + "num_input_tokens_seen": 65296665, + "step": 3042, + "time_per_iteration": 2.7621512413024902 + }, + { + "auxiliary_loss_clip": 0.01182652, + "auxiliary_loss_mlp": 0.01029367, + "balance_loss_clip": 1.05743575, + "balance_loss_mlp": 1.02064109, + "epoch": 0.3658991162147538, + "flos": 28293191539200.0, + "grad_norm": 1.671278339668907, + "language_loss": 0.65187562, + "learning_rate": 2.9288095962112927e-06, + "loss": 0.67399579, + "num_input_tokens_seen": 65317370, + "step": 3043, + "time_per_iteration": 2.7929909229278564 + }, + { + "auxiliary_loss_clip": 0.01189363, + "auxiliary_loss_mlp": 0.01039001, + "balance_loss_clip": 1.05433297, + "balance_loss_mlp": 1.0299952, + "epoch": 0.3660193591053929, + "flos": 17785550252160.0, + "grad_norm": 1.8567512900933758, + "language_loss": 0.85264277, + "learning_rate": 2.9281196505341503e-06, + "loss": 0.87492639, + "num_input_tokens_seen": 65334540, + "step": 3044, + "time_per_iteration": 2.6350436210632324 + }, + { + "auxiliary_loss_clip": 0.01146972, + "auxiliary_loss_mlp": 0.00885677, + "balance_loss_clip": 1.04861236, + "balance_loss_mlp": 1.00047839, + "epoch": 0.36613960199603196, + "flos": 10342776839040.0, + "grad_norm": 2.6034171320295547, + "language_loss": 0.78792208, + "learning_rate": 2.9274295640619946e-06, + "loss": 0.80824852, + "num_input_tokens_seen": 65351670, + "step": 3045, + "time_per_iteration": 2.8289802074432373 + }, + { + "auxiliary_loss_clip": 0.01157301, + "auxiliary_loss_mlp": 0.01031016, + "balance_loss_clip": 1.04606473, + "balance_loss_mlp": 1.0223434, + "epoch": 0.36625984488667107, + "flos": 19755609540480.0, + "grad_norm": 5.0501404577272, + "language_loss": 0.78513926, + "learning_rate": 2.9267393368995103e-06, + "loss": 0.80702245, + "num_input_tokens_seen": 65370900, + "step": 3046, + "time_per_iteration": 2.7933292388916016 + }, + { + "auxiliary_loss_clip": 0.01193137, + "auxiliary_loss_mlp": 0.01032614, + "balance_loss_clip": 1.05634558, + "balance_loss_mlp": 1.02329218, + "epoch": 0.3663800877773102, + "flos": 17674262939520.0, + "grad_norm": 2.7075261062878524, + "language_loss": 0.74308741, + "learning_rate": 2.926048969151407e-06, + "loss": 0.76534492, + "num_input_tokens_seen": 65388185, + "step": 3047, + "time_per_iteration": 2.764324188232422 + }, + { + "auxiliary_loss_clip": 0.01146421, + "auxiliary_loss_mlp": 0.01029695, + "balance_loss_clip": 1.05165195, + "balance_loss_mlp": 1.02068901, + "epoch": 0.36650033066794924, + "flos": 20303606407680.0, + "grad_norm": 1.91489006892252, + "language_loss": 0.6872893, + "learning_rate": 2.92535846092241e-06, + "loss": 0.70905048, + "num_input_tokens_seen": 65407200, + "step": 3048, + "time_per_iteration": 2.9477858543395996 + }, + { + "auxiliary_loss_clip": 0.01174444, + "auxiliary_loss_mlp": 0.01031962, + "balance_loss_clip": 1.05322361, + "balance_loss_mlp": 1.02274728, + "epoch": 0.36662057355858835, + "flos": 24716237420160.0, + "grad_norm": 1.7568165852756261, + "language_loss": 0.82887602, + "learning_rate": 2.9246678123172704e-06, + "loss": 0.85094011, + "num_input_tokens_seen": 65427290, + "step": 3049, + "time_per_iteration": 2.8567562103271484 + }, + { + "auxiliary_loss_clip": 0.01194398, + "auxiliary_loss_mlp": 0.01029771, + "balance_loss_clip": 1.05637908, + "balance_loss_mlp": 1.0201385, + "epoch": 0.36674081644922746, + "flos": 12385267902720.0, + "grad_norm": 2.0896561469083523, + "language_loss": 0.74440819, + "learning_rate": 2.9239770234407596e-06, + "loss": 0.76664984, + "num_input_tokens_seen": 65445595, + "step": 3050, + "time_per_iteration": 2.8456685543060303 + }, + { + "auxiliary_loss_clip": 0.01183685, + "auxiliary_loss_mlp": 0.01029272, + "balance_loss_clip": 1.05379462, + "balance_loss_mlp": 1.02022374, + "epoch": 0.3668610593398665, + "flos": 21105922544640.0, + "grad_norm": 1.7029033066146695, + "language_loss": 0.68099451, + "learning_rate": 2.9232860943976686e-06, + "loss": 0.70312405, + "num_input_tokens_seen": 65466330, + "step": 3051, + "time_per_iteration": 2.730511426925659 + }, + { + "auxiliary_loss_clip": 0.01169256, + "auxiliary_loss_mlp": 0.01032029, + "balance_loss_clip": 1.05219817, + "balance_loss_mlp": 1.02351117, + "epoch": 0.3669813022305056, + "flos": 26758082039040.0, + "grad_norm": 1.7003564406823208, + "language_loss": 0.8403933, + "learning_rate": 2.9225950252928115e-06, + "loss": 0.86240613, + "num_input_tokens_seen": 65487180, + "step": 3052, + "time_per_iteration": 2.8823001384735107 + }, + { + "auxiliary_loss_clip": 0.01181463, + "auxiliary_loss_mlp": 0.01043074, + "balance_loss_clip": 1.05578876, + "balance_loss_mlp": 1.03359687, + "epoch": 0.36710154512114473, + "flos": 19099521671040.0, + "grad_norm": 2.031776409816997, + "language_loss": 0.81613117, + "learning_rate": 2.9219038162310217e-06, + "loss": 0.83837652, + "num_input_tokens_seen": 65505380, + "step": 3053, + "time_per_iteration": 2.71860933303833 + }, + { + "auxiliary_loss_clip": 0.01129762, + "auxiliary_loss_mlp": 0.00886926, + "balance_loss_clip": 1.04511809, + "balance_loss_mlp": 1.00059628, + "epoch": 0.3672217880117838, + "flos": 20812029465600.0, + "grad_norm": 2.063909859310612, + "language_loss": 0.82858521, + "learning_rate": 2.921212467317157e-06, + "loss": 0.84875214, + "num_input_tokens_seen": 65524825, + "step": 3054, + "time_per_iteration": 2.990034818649292 + }, + { + "auxiliary_loss_clip": 0.01157232, + "auxiliary_loss_mlp": 0.01031819, + "balance_loss_clip": 1.04718673, + "balance_loss_mlp": 1.0219543, + "epoch": 0.3673420309024229, + "flos": 13590394133760.0, + "grad_norm": 2.120800999058893, + "language_loss": 0.79724336, + "learning_rate": 2.920520978656093e-06, + "loss": 0.81913382, + "num_input_tokens_seen": 65541790, + "step": 3055, + "time_per_iteration": 2.7932820320129395 + }, + { + "auxiliary_loss_clip": 0.01191618, + "auxiliary_loss_mlp": 0.00886729, + "balance_loss_clip": 1.05609035, + "balance_loss_mlp": 1.00072432, + "epoch": 0.367462273793062, + "flos": 28986877969920.0, + "grad_norm": 2.1764920414939137, + "language_loss": 0.76893771, + "learning_rate": 2.919829350352729e-06, + "loss": 0.78972125, + "num_input_tokens_seen": 65563395, + "step": 3056, + "time_per_iteration": 4.639745473861694 + }, + { + "auxiliary_loss_clip": 0.01101292, + "auxiliary_loss_mlp": 0.01010889, + "balance_loss_clip": 1.03337514, + "balance_loss_mlp": 1.00854015, + "epoch": 0.36758251668370107, + "flos": 62643148346880.0, + "grad_norm": 0.7616878650056172, + "language_loss": 0.60006011, + "learning_rate": 2.919137582511983e-06, + "loss": 0.62118185, + "num_input_tokens_seen": 65619835, + "step": 3057, + "time_per_iteration": 4.05775785446167 + }, + { + "auxiliary_loss_clip": 0.0116978, + "auxiliary_loss_mlp": 0.01033367, + "balance_loss_clip": 1.05616665, + "balance_loss_mlp": 1.02379477, + "epoch": 0.3677027595743402, + "flos": 12713886455040.0, + "grad_norm": 1.851189824295175, + "language_loss": 0.63707387, + "learning_rate": 2.918445675238797e-06, + "loss": 0.65910524, + "num_input_tokens_seen": 65636760, + "step": 3058, + "time_per_iteration": 2.8148930072784424 + }, + { + "auxiliary_loss_clip": 0.01191461, + "auxiliary_loss_mlp": 0.01031056, + "balance_loss_clip": 1.05391955, + "balance_loss_mlp": 1.02144194, + "epoch": 0.36782300246497923, + "flos": 25046579825280.0, + "grad_norm": 1.6897686720893168, + "language_loss": 0.69641608, + "learning_rate": 2.917753628638132e-06, + "loss": 0.71864128, + "num_input_tokens_seen": 65657065, + "step": 3059, + "time_per_iteration": 3.845958709716797 + }, + { + "auxiliary_loss_clip": 0.01175134, + "auxiliary_loss_mlp": 0.01039279, + "balance_loss_clip": 1.05606937, + "balance_loss_mlp": 1.02983165, + "epoch": 0.36794324535561834, + "flos": 17419512706560.0, + "grad_norm": 2.6563365782514015, + "language_loss": 0.7040832, + "learning_rate": 2.9170614428149716e-06, + "loss": 0.72622728, + "num_input_tokens_seen": 65675400, + "step": 3060, + "time_per_iteration": 2.7991621494293213 + }, + { + "auxiliary_loss_clip": 0.01160463, + "auxiliary_loss_mlp": 0.01031973, + "balance_loss_clip": 1.05231786, + "balance_loss_mlp": 1.02282357, + "epoch": 0.36806348824625745, + "flos": 24089128848000.0, + "grad_norm": 2.241964750179042, + "language_loss": 0.86729348, + "learning_rate": 2.9163691178743195e-06, + "loss": 0.88921785, + "num_input_tokens_seen": 65694050, + "step": 3061, + "time_per_iteration": 2.848478317260742 + }, + { + "auxiliary_loss_clip": 0.01177236, + "auxiliary_loss_mlp": 0.01028804, + "balance_loss_clip": 1.05241573, + "balance_loss_mlp": 1.02000093, + "epoch": 0.3681837311368965, + "flos": 20521871400960.0, + "grad_norm": 2.123775447533073, + "language_loss": 0.77410597, + "learning_rate": 2.9156766539212006e-06, + "loss": 0.79616636, + "num_input_tokens_seen": 65711695, + "step": 3062, + "time_per_iteration": 2.771404981613159 + }, + { + "auxiliary_loss_clip": 0.01188816, + "auxiliary_loss_mlp": 0.01034269, + "balance_loss_clip": 1.05491805, + "balance_loss_mlp": 1.02497649, + "epoch": 0.3683039740275356, + "flos": 21466644877440.0, + "grad_norm": 1.976818722348202, + "language_loss": 0.71708292, + "learning_rate": 2.9149840510606614e-06, + "loss": 0.73931372, + "num_input_tokens_seen": 65730350, + "step": 3063, + "time_per_iteration": 2.7719132900238037 + }, + { + "auxiliary_loss_clip": 0.0109002, + "auxiliary_loss_mlp": 0.00877337, + "balance_loss_clip": 1.0296731, + "balance_loss_mlp": 1.00101209, + "epoch": 0.36842421691817473, + "flos": 70380999987840.0, + "grad_norm": 1.0289782500681113, + "language_loss": 0.64154506, + "learning_rate": 2.914291309397769e-06, + "loss": 0.66121864, + "num_input_tokens_seen": 65787820, + "step": 3064, + "time_per_iteration": 3.3729875087738037 + }, + { + "auxiliary_loss_clip": 0.01134366, + "auxiliary_loss_mlp": 0.0102936, + "balance_loss_clip": 1.0477252, + "balance_loss_mlp": 1.01923287, + "epoch": 0.3685444598088138, + "flos": 23331378510720.0, + "grad_norm": 2.010252037890688, + "language_loss": 0.78294384, + "learning_rate": 2.9135984290376117e-06, + "loss": 0.80458117, + "num_input_tokens_seen": 65806685, + "step": 3065, + "time_per_iteration": 2.914425849914551 + }, + { + "auxiliary_loss_clip": 0.01142935, + "auxiliary_loss_mlp": 0.01026914, + "balance_loss_clip": 1.04827189, + "balance_loss_mlp": 1.01761556, + "epoch": 0.3686647026994529, + "flos": 23070271570560.0, + "grad_norm": 1.650772055399447, + "language_loss": 0.82737994, + "learning_rate": 2.9129054100853e-06, + "loss": 0.84907848, + "num_input_tokens_seen": 65825525, + "step": 3066, + "time_per_iteration": 2.933457136154175 + }, + { + "auxiliary_loss_clip": 0.01175982, + "auxiliary_loss_mlp": 0.01027859, + "balance_loss_clip": 1.0525434, + "balance_loss_mlp": 1.01834655, + "epoch": 0.368784945590092, + "flos": 25119909440640.0, + "grad_norm": 1.7202029179106262, + "language_loss": 0.75848323, + "learning_rate": 2.912212252645963e-06, + "loss": 0.78052163, + "num_input_tokens_seen": 65848110, + "step": 3067, + "time_per_iteration": 2.8136696815490723 + }, + { + "auxiliary_loss_clip": 0.01192686, + "auxiliary_loss_mlp": 0.01032354, + "balance_loss_clip": 1.05705869, + "balance_loss_mlp": 1.02180958, + "epoch": 0.36890518848073106, + "flos": 18442284566400.0, + "grad_norm": 2.9690452711468267, + "language_loss": 0.76766205, + "learning_rate": 2.9115189568247523e-06, + "loss": 0.78991246, + "num_input_tokens_seen": 65865670, + "step": 3068, + "time_per_iteration": 2.8350343704223633 + }, + { + "auxiliary_loss_clip": 0.01147042, + "auxiliary_loss_mlp": 0.0103045, + "balance_loss_clip": 1.05508256, + "balance_loss_mlp": 1.02044821, + "epoch": 0.36902543137137017, + "flos": 16362446336640.0, + "grad_norm": 2.090761381538553, + "language_loss": 0.9224866, + "learning_rate": 2.910825522726841e-06, + "loss": 0.94426155, + "num_input_tokens_seen": 65883195, + "step": 3069, + "time_per_iteration": 2.901122570037842 + }, + { + "auxiliary_loss_clip": 0.01147748, + "auxiliary_loss_mlp": 0.01030973, + "balance_loss_clip": 1.04664111, + "balance_loss_mlp": 1.0220325, + "epoch": 0.3691456742620093, + "flos": 12275596702080.0, + "grad_norm": 2.1049700026152753, + "language_loss": 0.77087867, + "learning_rate": 2.9101319504574215e-06, + "loss": 0.79266584, + "num_input_tokens_seen": 65899635, + "step": 3070, + "time_per_iteration": 2.8514652252197266 + }, + { + "auxiliary_loss_clip": 0.0117488, + "auxiliary_loss_mlp": 0.01036994, + "balance_loss_clip": 1.05072606, + "balance_loss_mlp": 1.02722466, + "epoch": 0.36926591715264834, + "flos": 17786412178560.0, + "grad_norm": 1.6535969577174314, + "language_loss": 0.76133096, + "learning_rate": 2.909438240121709e-06, + "loss": 0.78344977, + "num_input_tokens_seen": 65919910, + "step": 3071, + "time_per_iteration": 2.7686409950256348 + }, + { + "auxiliary_loss_clip": 0.0116922, + "auxiliary_loss_mlp": 0.01032892, + "balance_loss_clip": 1.05298638, + "balance_loss_mlp": 1.02413583, + "epoch": 0.36938616004328745, + "flos": 28948309741440.0, + "grad_norm": 1.8764377727218888, + "language_loss": 0.70485532, + "learning_rate": 2.908744391824939e-06, + "loss": 0.7268765, + "num_input_tokens_seen": 65940930, + "step": 3072, + "time_per_iteration": 2.831220865249634 + }, + { + "auxiliary_loss_clip": 0.01146547, + "auxiliary_loss_mlp": 0.01031771, + "balance_loss_clip": 1.04679894, + "balance_loss_mlp": 1.02246666, + "epoch": 0.36950640293392656, + "flos": 29205394358400.0, + "grad_norm": 2.0636548960691177, + "language_loss": 0.79496682, + "learning_rate": 2.908050405672367e-06, + "loss": 0.81674999, + "num_input_tokens_seen": 65960475, + "step": 3073, + "time_per_iteration": 2.93109130859375 + }, + { + "auxiliary_loss_clip": 0.01180399, + "auxiliary_loss_mlp": 0.01034935, + "balance_loss_clip": 1.05084944, + "balance_loss_mlp": 1.0257616, + "epoch": 0.3696266458245656, + "flos": 24827776128000.0, + "grad_norm": 1.849767298427024, + "language_loss": 0.79140449, + "learning_rate": 2.9073562817692703e-06, + "loss": 0.81355774, + "num_input_tokens_seen": 65979160, + "step": 3074, + "time_per_iteration": 2.7971041202545166 + }, + { + "auxiliary_loss_clip": 0.0107622, + "auxiliary_loss_mlp": 0.01023268, + "balance_loss_clip": 1.02873731, + "balance_loss_mlp": 1.02090728, + "epoch": 0.3697468887152047, + "flos": 59887257264000.0, + "grad_norm": 0.725690641093544, + "language_loss": 0.56489313, + "learning_rate": 2.9066620202209468e-06, + "loss": 0.58588803, + "num_input_tokens_seen": 66041650, + "step": 3075, + "time_per_iteration": 3.3554484844207764 + }, + { + "auxiliary_loss_clip": 0.01155579, + "auxiliary_loss_mlp": 0.01032016, + "balance_loss_clip": 1.04969192, + "balance_loss_mlp": 1.02272987, + "epoch": 0.3698671316058438, + "flos": 26137581569280.0, + "grad_norm": 2.185931275811541, + "language_loss": 0.78016925, + "learning_rate": 2.905967621132716e-06, + "loss": 0.80204517, + "num_input_tokens_seen": 66059260, + "step": 3076, + "time_per_iteration": 2.816030979156494 + }, + { + "auxiliary_loss_clip": 0.01179007, + "auxiliary_loss_mlp": 0.01035741, + "balance_loss_clip": 1.05381668, + "balance_loss_mlp": 1.02534604, + "epoch": 0.3699873744964829, + "flos": 24607464059520.0, + "grad_norm": 2.257467408756529, + "language_loss": 0.75253987, + "learning_rate": 2.9052730846099172e-06, + "loss": 0.77468741, + "num_input_tokens_seen": 66080605, + "step": 3077, + "time_per_iteration": 2.8304550647735596 + }, + { + "auxiliary_loss_clip": 0.01081562, + "auxiliary_loss_mlp": 0.01004128, + "balance_loss_clip": 1.02693486, + "balance_loss_mlp": 1.00142217, + "epoch": 0.370107617387122, + "flos": 64885340050560.0, + "grad_norm": 0.8660390462995263, + "language_loss": 0.60913932, + "learning_rate": 2.9045784107579123e-06, + "loss": 0.62999618, + "num_input_tokens_seen": 66140710, + "step": 3078, + "time_per_iteration": 3.3609447479248047 + }, + { + "auxiliary_loss_clip": 0.01195711, + "auxiliary_loss_mlp": 0.01033182, + "balance_loss_clip": 1.05715191, + "balance_loss_mlp": 1.02384758, + "epoch": 0.37022786027776106, + "flos": 15961683317760.0, + "grad_norm": 1.7548033945617825, + "language_loss": 0.6687175, + "learning_rate": 2.9038835996820807e-06, + "loss": 0.69100642, + "num_input_tokens_seen": 66158320, + "step": 3079, + "time_per_iteration": 2.6894941329956055 + }, + { + "auxiliary_loss_clip": 0.01167366, + "auxiliary_loss_mlp": 0.0103673, + "balance_loss_clip": 1.04776311, + "balance_loss_mlp": 1.02753329, + "epoch": 0.37034810316840017, + "flos": 18546927863040.0, + "grad_norm": 1.9876344891312085, + "language_loss": 0.79651755, + "learning_rate": 2.903188651487826e-06, + "loss": 0.81855857, + "num_input_tokens_seen": 66176875, + "step": 3080, + "time_per_iteration": 2.813837766647339 + }, + { + "auxiliary_loss_clip": 0.01183318, + "auxiliary_loss_mlp": 0.0103018, + "balance_loss_clip": 1.05288589, + "balance_loss_mlp": 1.0202204, + "epoch": 0.3704683460590393, + "flos": 17821927751040.0, + "grad_norm": 2.3205081874019435, + "language_loss": 0.86342931, + "learning_rate": 2.902493566280571e-06, + "loss": 0.88556433, + "num_input_tokens_seen": 66194980, + "step": 3081, + "time_per_iteration": 2.6825151443481445 + }, + { + "auxiliary_loss_clip": 0.01172119, + "auxiliary_loss_mlp": 0.01036384, + "balance_loss_clip": 1.05238163, + "balance_loss_mlp": 1.02628088, + "epoch": 0.37058858894967833, + "flos": 14134081368960.0, + "grad_norm": 1.9361240965367201, + "language_loss": 0.81033939, + "learning_rate": 2.9017983441657595e-06, + "loss": 0.8324244, + "num_input_tokens_seen": 66212310, + "step": 3082, + "time_per_iteration": 4.540497064590454 + }, + { + "auxiliary_loss_clip": 0.01155178, + "auxiliary_loss_mlp": 0.01028422, + "balance_loss_clip": 1.04679918, + "balance_loss_mlp": 1.01905179, + "epoch": 0.37070883184031744, + "flos": 13954492344960.0, + "grad_norm": 2.026894906830146, + "language_loss": 0.75312173, + "learning_rate": 2.9011029852488564e-06, + "loss": 0.77495766, + "num_input_tokens_seen": 66229545, + "step": 3083, + "time_per_iteration": 3.8472821712493896 + }, + { + "auxiliary_loss_clip": 0.01091498, + "auxiliary_loss_mlp": 0.01004665, + "balance_loss_clip": 1.02516556, + "balance_loss_mlp": 1.00213766, + "epoch": 0.37082907473095655, + "flos": 52315419306240.0, + "grad_norm": 0.9857843525339903, + "language_loss": 0.62427914, + "learning_rate": 2.9004074896353465e-06, + "loss": 0.64524078, + "num_input_tokens_seen": 66283545, + "step": 3084, + "time_per_iteration": 3.109938621520996 + }, + { + "auxiliary_loss_clip": 0.01196383, + "auxiliary_loss_mlp": 0.01030914, + "balance_loss_clip": 1.06064415, + "balance_loss_mlp": 1.02242637, + "epoch": 0.3709493176215956, + "flos": 15998096730240.0, + "grad_norm": 1.7964112015348384, + "language_loss": 0.81250083, + "learning_rate": 2.8997118574307362e-06, + "loss": 0.83477378, + "num_input_tokens_seen": 66300500, + "step": 3085, + "time_per_iteration": 3.596538543701172 + }, + { + "auxiliary_loss_clip": 0.01172535, + "auxiliary_loss_mlp": 0.01029969, + "balance_loss_clip": 1.05473042, + "balance_loss_mlp": 1.02117145, + "epoch": 0.3710695605122347, + "flos": 20959837931520.0, + "grad_norm": 1.96512913509197, + "language_loss": 0.73981929, + "learning_rate": 2.899016088740553e-06, + "loss": 0.76184428, + "num_input_tokens_seen": 66318610, + "step": 3086, + "time_per_iteration": 2.8337185382843018 + }, + { + "auxiliary_loss_clip": 0.01155646, + "auxiliary_loss_mlp": 0.01032229, + "balance_loss_clip": 1.05244184, + "balance_loss_mlp": 1.02304411, + "epoch": 0.37118980340287383, + "flos": 14355578586240.0, + "grad_norm": 2.4049765522504876, + "language_loss": 0.79182434, + "learning_rate": 2.898320183670344e-06, + "loss": 0.81370306, + "num_input_tokens_seen": 66336025, + "step": 3087, + "time_per_iteration": 2.795633316040039 + }, + { + "auxiliary_loss_clip": 0.01151774, + "auxiliary_loss_mlp": 0.01028721, + "balance_loss_clip": 1.05250025, + "balance_loss_mlp": 1.01967263, + "epoch": 0.3713100462935129, + "flos": 25885381201920.0, + "grad_norm": 1.7516383955819013, + "language_loss": 0.89214969, + "learning_rate": 2.8976241423256767e-06, + "loss": 0.91395462, + "num_input_tokens_seen": 66356120, + "step": 3088, + "time_per_iteration": 2.8676671981811523 + }, + { + "auxiliary_loss_clip": 0.01168863, + "auxiliary_loss_mlp": 0.01034132, + "balance_loss_clip": 1.0520072, + "balance_loss_mlp": 1.02473831, + "epoch": 0.371430289184152, + "flos": 30518934814080.0, + "grad_norm": 1.8777377907368942, + "language_loss": 0.68142307, + "learning_rate": 2.896927964812142e-06, + "loss": 0.70345306, + "num_input_tokens_seen": 66376685, + "step": 3089, + "time_per_iteration": 2.8942885398864746 + }, + { + "auxiliary_loss_clip": 0.01173683, + "auxiliary_loss_mlp": 0.01034467, + "balance_loss_clip": 1.05647993, + "balance_loss_mlp": 1.025455, + "epoch": 0.37155053207479105, + "flos": 15742233175680.0, + "grad_norm": 2.2656182011646644, + "language_loss": 0.75197339, + "learning_rate": 2.8962316512353465e-06, + "loss": 0.77405488, + "num_input_tokens_seen": 66394230, + "step": 3090, + "time_per_iteration": 2.7256340980529785 + }, + { + "auxiliary_loss_clip": 0.0113967, + "auxiliary_loss_mlp": 0.01029438, + "balance_loss_clip": 1.04536784, + "balance_loss_mlp": 1.02002668, + "epoch": 0.37167077496543016, + "flos": 23404061681280.0, + "grad_norm": 1.6230205635605193, + "language_loss": 0.74632823, + "learning_rate": 2.8955352017009233e-06, + "loss": 0.76801932, + "num_input_tokens_seen": 66413475, + "step": 3091, + "time_per_iteration": 2.935837507247925 + }, + { + "auxiliary_loss_clip": 0.01171981, + "auxiliary_loss_mlp": 0.01029515, + "balance_loss_clip": 1.05577493, + "balance_loss_mlp": 1.02053869, + "epoch": 0.3717910178560693, + "flos": 22088653718400.0, + "grad_norm": 2.3634050591073943, + "language_loss": 0.77132195, + "learning_rate": 2.8948386163145212e-06, + "loss": 0.79333693, + "num_input_tokens_seen": 66432685, + "step": 3092, + "time_per_iteration": 2.76114559173584 + }, + { + "auxiliary_loss_clip": 0.01189145, + "auxiliary_loss_mlp": 0.01034067, + "balance_loss_clip": 1.05674934, + "balance_loss_mlp": 1.02515662, + "epoch": 0.3719112607467083, + "flos": 26939969533440.0, + "grad_norm": 2.0500370502102316, + "language_loss": 0.79351211, + "learning_rate": 2.8941418951818135e-06, + "loss": 0.81574416, + "num_input_tokens_seen": 66452245, + "step": 3093, + "time_per_iteration": 2.856114387512207 + }, + { + "auxiliary_loss_clip": 0.01167544, + "auxiliary_loss_mlp": 0.01033409, + "balance_loss_clip": 1.05259705, + "balance_loss_mlp": 1.02428317, + "epoch": 0.37203150363734744, + "flos": 12166500119040.0, + "grad_norm": 2.2643826573443864, + "language_loss": 0.70559275, + "learning_rate": 2.8934450384084903e-06, + "loss": 0.72760224, + "num_input_tokens_seen": 66469760, + "step": 3094, + "time_per_iteration": 2.77544903755188 + }, + { + "auxiliary_loss_clip": 0.01165544, + "auxiliary_loss_mlp": 0.01027907, + "balance_loss_clip": 1.05084562, + "balance_loss_mlp": 1.01840031, + "epoch": 0.37215174652798655, + "flos": 23697595624320.0, + "grad_norm": 2.03870920352394, + "language_loss": 0.69880426, + "learning_rate": 2.8927480461002653e-06, + "loss": 0.72073877, + "num_input_tokens_seen": 66489730, + "step": 3095, + "time_per_iteration": 2.857208251953125 + }, + { + "auxiliary_loss_clip": 0.01174041, + "auxiliary_loss_mlp": 0.01033523, + "balance_loss_clip": 1.05144811, + "balance_loss_mlp": 1.0232476, + "epoch": 0.3722719894186256, + "flos": 17887751424000.0, + "grad_norm": 2.2175913000629897, + "language_loss": 0.86794472, + "learning_rate": 2.892050918362872e-06, + "loss": 0.89002037, + "num_input_tokens_seen": 66504785, + "step": 3096, + "time_per_iteration": 2.728837728500366 + }, + { + "auxiliary_loss_clip": 0.01046097, + "auxiliary_loss_mlp": 0.01008323, + "balance_loss_clip": 1.02050793, + "balance_loss_mlp": 1.00570095, + "epoch": 0.3723922323092647, + "flos": 62419891363200.0, + "grad_norm": 0.8446147692096431, + "language_loss": 0.55884874, + "learning_rate": 2.8913536553020626e-06, + "loss": 0.57939291, + "num_input_tokens_seen": 66558840, + "step": 3097, + "time_per_iteration": 3.5931553840637207 + }, + { + "auxiliary_loss_clip": 0.01143311, + "auxiliary_loss_mlp": 0.01034412, + "balance_loss_clip": 1.04665399, + "balance_loss_mlp": 1.02556086, + "epoch": 0.3725124751999038, + "flos": 23039747988480.0, + "grad_norm": 1.7773986077346642, + "language_loss": 0.84578133, + "learning_rate": 2.8906562570236137e-06, + "loss": 0.8675586, + "num_input_tokens_seen": 66576750, + "step": 3098, + "time_per_iteration": 3.3449196815490723 + }, + { + "auxiliary_loss_clip": 0.01140408, + "auxiliary_loss_mlp": 0.01032023, + "balance_loss_clip": 1.045645, + "balance_loss_mlp": 1.02311194, + "epoch": 0.3726327180905429, + "flos": 20920551431040.0, + "grad_norm": 1.827808519986905, + "language_loss": 0.76755482, + "learning_rate": 2.889958723633318e-06, + "loss": 0.7892791, + "num_input_tokens_seen": 66595690, + "step": 3099, + "time_per_iteration": 2.958414316177368 + }, + { + "auxiliary_loss_clip": 0.01165617, + "auxiliary_loss_mlp": 0.01029148, + "balance_loss_clip": 1.05488157, + "balance_loss_mlp": 1.0204699, + "epoch": 0.372752960981182, + "flos": 30592156688640.0, + "grad_norm": 2.1077957889440007, + "language_loss": 0.73922199, + "learning_rate": 2.889261055236992e-06, + "loss": 0.76116961, + "num_input_tokens_seen": 66617905, + "step": 3100, + "time_per_iteration": 2.8463120460510254 + }, + { + "auxiliary_loss_clip": 0.01172869, + "auxiliary_loss_mlp": 0.0103701, + "balance_loss_clip": 1.05663133, + "balance_loss_mlp": 1.02783084, + "epoch": 0.3728732038718211, + "flos": 25116749043840.0, + "grad_norm": 1.877423839558077, + "language_loss": 0.82720768, + "learning_rate": 2.8885632519404704e-06, + "loss": 0.84930646, + "num_input_tokens_seen": 66638175, + "step": 3101, + "time_per_iteration": 2.771576404571533 + }, + { + "auxiliary_loss_clip": 0.01175002, + "auxiliary_loss_mlp": 0.01031092, + "balance_loss_clip": 1.05610299, + "balance_loss_mlp": 1.02167463, + "epoch": 0.37299344676246016, + "flos": 25302048330240.0, + "grad_norm": 2.9180511762685595, + "language_loss": 0.75845933, + "learning_rate": 2.8878653138496107e-06, + "loss": 0.7805202, + "num_input_tokens_seen": 66658670, + "step": 3102, + "time_per_iteration": 2.7260735034942627 + }, + { + "auxiliary_loss_clip": 0.01139604, + "auxiliary_loss_mlp": 0.01038177, + "balance_loss_clip": 1.04403853, + "balance_loss_mlp": 1.02843809, + "epoch": 0.37311368965309927, + "flos": 23842531002240.0, + "grad_norm": 2.620124830790239, + "language_loss": 0.7628023, + "learning_rate": 2.8871672410702878e-06, + "loss": 0.78458011, + "num_input_tokens_seen": 66676030, + "step": 3103, + "time_per_iteration": 2.897782802581787 + }, + { + "auxiliary_loss_clip": 0.01176047, + "auxiliary_loss_mlp": 0.01036229, + "balance_loss_clip": 1.05297089, + "balance_loss_mlp": 1.02616215, + "epoch": 0.3732339325437384, + "flos": 25811943845760.0, + "grad_norm": 1.8664674773310383, + "language_loss": 0.82011998, + "learning_rate": 2.8864690337084008e-06, + "loss": 0.84224272, + "num_input_tokens_seen": 66695305, + "step": 3104, + "time_per_iteration": 2.7968149185180664 + }, + { + "auxiliary_loss_clip": 0.01177599, + "auxiliary_loss_mlp": 0.01037291, + "balance_loss_clip": 1.05265522, + "balance_loss_mlp": 1.02761078, + "epoch": 0.37335417543437743, + "flos": 26208433146240.0, + "grad_norm": 1.7067817946788746, + "language_loss": 0.78186911, + "learning_rate": 2.885770691869866e-06, + "loss": 0.80401802, + "num_input_tokens_seen": 66716185, + "step": 3105, + "time_per_iteration": 2.7684578895568848 + }, + { + "auxiliary_loss_clip": 0.01178623, + "auxiliary_loss_mlp": 0.01033353, + "balance_loss_clip": 1.05377078, + "balance_loss_mlp": 1.02443027, + "epoch": 0.37347441832501654, + "flos": 24023879792640.0, + "grad_norm": 1.9720149735542996, + "language_loss": 0.74346584, + "learning_rate": 2.8850722156606207e-06, + "loss": 0.76558566, + "num_input_tokens_seen": 66734575, + "step": 3106, + "time_per_iteration": 2.7168314456939697 + }, + { + "auxiliary_loss_clip": 0.01178953, + "auxiliary_loss_mlp": 0.01030652, + "balance_loss_clip": 1.05401707, + "balance_loss_mlp": 1.02144909, + "epoch": 0.3735946612156556, + "flos": 19714922409600.0, + "grad_norm": 1.576840729714123, + "language_loss": 0.66877222, + "learning_rate": 2.8843736051866252e-06, + "loss": 0.69086832, + "num_input_tokens_seen": 66753500, + "step": 3107, + "time_per_iteration": 2.7791895866394043 + }, + { + "auxiliary_loss_clip": 0.01153806, + "auxiliary_loss_mlp": 0.00887235, + "balance_loss_clip": 1.05271769, + "balance_loss_mlp": 1.00074399, + "epoch": 0.3737149041062947, + "flos": 23039604334080.0, + "grad_norm": 1.7337985542350458, + "language_loss": 0.69659519, + "learning_rate": 2.8836748605538557e-06, + "loss": 0.71700555, + "num_input_tokens_seen": 66775140, + "step": 3108, + "time_per_iteration": 4.819289684295654 + }, + { + "auxiliary_loss_clip": 0.01178193, + "auxiliary_loss_mlp": 0.01036534, + "balance_loss_clip": 1.05285168, + "balance_loss_mlp": 1.02668691, + "epoch": 0.3738351469969338, + "flos": 34678108483200.0, + "grad_norm": 2.0811090614386165, + "language_loss": 0.63559246, + "learning_rate": 2.882975981868313e-06, + "loss": 0.65773976, + "num_input_tokens_seen": 66795525, + "step": 3109, + "time_per_iteration": 3.8163111209869385 + }, + { + "auxiliary_loss_clip": 0.01185952, + "auxiliary_loss_mlp": 0.01037693, + "balance_loss_clip": 1.0570333, + "balance_loss_mlp": 1.02777433, + "epoch": 0.3739553898875729, + "flos": 43507967448960.0, + "grad_norm": 1.949733034667924, + "language_loss": 0.69037229, + "learning_rate": 2.882276969236016e-06, + "loss": 0.71260875, + "num_input_tokens_seen": 66816885, + "step": 3110, + "time_per_iteration": 2.8821871280670166 + }, + { + "auxiliary_loss_clip": 0.01172038, + "auxiliary_loss_mlp": 0.01037675, + "balance_loss_clip": 1.05295062, + "balance_loss_mlp": 1.02750647, + "epoch": 0.374075632778212, + "flos": 12856487448960.0, + "grad_norm": 2.109719687129259, + "language_loss": 0.7644757, + "learning_rate": 2.881577822763005e-06, + "loss": 0.78657281, + "num_input_tokens_seen": 66834835, + "step": 3111, + "time_per_iteration": 3.723738193511963 + }, + { + "auxiliary_loss_clip": 0.01183136, + "auxiliary_loss_mlp": 0.01033418, + "balance_loss_clip": 1.0539639, + "balance_loss_mlp": 1.0239408, + "epoch": 0.3741958756688511, + "flos": 26024031699840.0, + "grad_norm": 1.8002113583514405, + "language_loss": 0.87194121, + "learning_rate": 2.880878542555338e-06, + "loss": 0.89410681, + "num_input_tokens_seen": 66852600, + "step": 3112, + "time_per_iteration": 2.756394863128662 + }, + { + "auxiliary_loss_clip": 0.01198394, + "auxiliary_loss_mlp": 0.01033257, + "balance_loss_clip": 1.05824757, + "balance_loss_mlp": 1.02392256, + "epoch": 0.37431611855949015, + "flos": 21433894652160.0, + "grad_norm": 2.2282848950987337, + "language_loss": 0.80399793, + "learning_rate": 2.8801791287190976e-06, + "loss": 0.82631445, + "num_input_tokens_seen": 66870595, + "step": 3113, + "time_per_iteration": 2.7080206871032715 + }, + { + "auxiliary_loss_clip": 0.01187926, + "auxiliary_loss_mlp": 0.0103294, + "balance_loss_clip": 1.05605626, + "balance_loss_mlp": 1.02360034, + "epoch": 0.37443636145012926, + "flos": 24207096090240.0, + "grad_norm": 4.719897608406108, + "language_loss": 0.86192501, + "learning_rate": 2.8794795813603817e-06, + "loss": 0.8841337, + "num_input_tokens_seen": 66886060, + "step": 3114, + "time_per_iteration": 2.768018960952759 + }, + { + "auxiliary_loss_clip": 0.01189922, + "auxiliary_loss_mlp": 0.01036516, + "balance_loss_clip": 1.05708981, + "balance_loss_mlp": 1.02731895, + "epoch": 0.3745566043407684, + "flos": 15378601841280.0, + "grad_norm": 1.7827382253099946, + "language_loss": 0.81458473, + "learning_rate": 2.878779900585314e-06, + "loss": 0.83684915, + "num_input_tokens_seen": 66903900, + "step": 3115, + "time_per_iteration": 2.684617042541504 + }, + { + "auxiliary_loss_clip": 0.01181126, + "auxiliary_loss_mlp": 0.01034297, + "balance_loss_clip": 1.05481064, + "balance_loss_mlp": 1.02514172, + "epoch": 0.37467684723140743, + "flos": 24608218245120.0, + "grad_norm": 1.5651515852169073, + "language_loss": 0.74922657, + "learning_rate": 2.8780800865000336e-06, + "loss": 0.77138078, + "num_input_tokens_seen": 66925210, + "step": 3116, + "time_per_iteration": 2.827683210372925 + }, + { + "auxiliary_loss_clip": 0.01083976, + "auxiliary_loss_mlp": 0.01006817, + "balance_loss_clip": 1.02371371, + "balance_loss_mlp": 1.0044446, + "epoch": 0.37479709012204654, + "flos": 64377491610240.0, + "grad_norm": 0.9756576170104376, + "language_loss": 0.59222245, + "learning_rate": 2.877380139210702e-06, + "loss": 0.61313039, + "num_input_tokens_seen": 66983880, + "step": 3117, + "time_per_iteration": 3.244863510131836 + }, + { + "auxiliary_loss_clip": 0.01172252, + "auxiliary_loss_mlp": 0.01033879, + "balance_loss_clip": 1.05646527, + "balance_loss_mlp": 1.02438426, + "epoch": 0.37491733301268565, + "flos": 23803962773760.0, + "grad_norm": 1.8794855817607523, + "language_loss": 0.7670486, + "learning_rate": 2.876680058823501e-06, + "loss": 0.78910989, + "num_input_tokens_seen": 67004280, + "step": 3118, + "time_per_iteration": 2.9063990116119385 + }, + { + "auxiliary_loss_clip": 0.01167265, + "auxiliary_loss_mlp": 0.01033645, + "balance_loss_clip": 1.05257726, + "balance_loss_mlp": 1.02405453, + "epoch": 0.3750375759033247, + "flos": 32160950167680.0, + "grad_norm": 2.3599587012211543, + "language_loss": 0.66097134, + "learning_rate": 2.8759798454446314e-06, + "loss": 0.68298042, + "num_input_tokens_seen": 67027445, + "step": 3119, + "time_per_iteration": 2.961740255355835 + }, + { + "auxiliary_loss_clip": 0.01188135, + "auxiliary_loss_mlp": 0.01034026, + "balance_loss_clip": 1.05601501, + "balance_loss_mlp": 1.02540684, + "epoch": 0.3751578187939638, + "flos": 23367791923200.0, + "grad_norm": 1.8897302654259103, + "language_loss": 0.8109442, + "learning_rate": 2.8752794991803173e-06, + "loss": 0.83316576, + "num_input_tokens_seen": 67045130, + "step": 3120, + "time_per_iteration": 2.821626663208008 + }, + { + "auxiliary_loss_clip": 0.01170774, + "auxiliary_loss_mlp": 0.01029955, + "balance_loss_clip": 1.05372608, + "balance_loss_mlp": 1.02127075, + "epoch": 0.37527806168460287, + "flos": 14605731878400.0, + "grad_norm": 2.7634757062856936, + "language_loss": 0.74333918, + "learning_rate": 2.8745790201367976e-06, + "loss": 0.76534647, + "num_input_tokens_seen": 67060885, + "step": 3121, + "time_per_iteration": 2.7542076110839844 + }, + { + "auxiliary_loss_clip": 0.01194633, + "auxiliary_loss_mlp": 0.01034447, + "balance_loss_clip": 1.05448711, + "balance_loss_mlp": 1.02449942, + "epoch": 0.375398304575242, + "flos": 26390823431040.0, + "grad_norm": 1.9387096959478662, + "language_loss": 0.83915287, + "learning_rate": 2.8738784084203373e-06, + "loss": 0.8614437, + "num_input_tokens_seen": 67080960, + "step": 3122, + "time_per_iteration": 2.7178330421447754 + }, + { + "auxiliary_loss_clip": 0.01165539, + "auxiliary_loss_mlp": 0.01031477, + "balance_loss_clip": 1.04886961, + "balance_loss_mlp": 1.0224762, + "epoch": 0.3755185474658811, + "flos": 22236605838720.0, + "grad_norm": 1.6766478060260372, + "language_loss": 0.78737903, + "learning_rate": 2.873177664137216e-06, + "loss": 0.80934918, + "num_input_tokens_seen": 67101890, + "step": 3123, + "time_per_iteration": 2.80271577835083 + }, + { + "auxiliary_loss_clip": 0.01160107, + "auxiliary_loss_mlp": 0.01025439, + "balance_loss_clip": 1.05213559, + "balance_loss_mlp": 1.01656961, + "epoch": 0.37563879035652015, + "flos": 30812935633920.0, + "grad_norm": 1.7151848417385223, + "language_loss": 0.69411886, + "learning_rate": 2.8724767873937384e-06, + "loss": 0.71597433, + "num_input_tokens_seen": 67126010, + "step": 3124, + "time_per_iteration": 2.8532826900482178 + }, + { + "auxiliary_loss_clip": 0.01174606, + "auxiliary_loss_mlp": 0.01031001, + "balance_loss_clip": 1.05470622, + "balance_loss_mlp": 1.02257323, + "epoch": 0.37575903324715926, + "flos": 20773533064320.0, + "grad_norm": 2.3175438627959357, + "language_loss": 0.87241447, + "learning_rate": 2.871775778296225e-06, + "loss": 0.89447051, + "num_input_tokens_seen": 67143100, + "step": 3125, + "time_per_iteration": 2.76786470413208 + }, + { + "auxiliary_loss_clip": 0.01185226, + "auxiliary_loss_mlp": 0.01033286, + "balance_loss_clip": 1.0566622, + "balance_loss_mlp": 1.0233922, + "epoch": 0.37587927613779837, + "flos": 18697681244160.0, + "grad_norm": 2.0447255396371284, + "language_loss": 0.785689, + "learning_rate": 2.8710746369510196e-06, + "loss": 0.80787408, + "num_input_tokens_seen": 67161085, + "step": 3126, + "time_per_iteration": 2.627800703048706 + }, + { + "auxiliary_loss_clip": 0.01169225, + "auxiliary_loss_mlp": 0.0103841, + "balance_loss_clip": 1.05310655, + "balance_loss_mlp": 1.02939749, + "epoch": 0.3759995190284374, + "flos": 13624796384640.0, + "grad_norm": 2.3691382482441212, + "language_loss": 0.8326174, + "learning_rate": 2.8703733634644846e-06, + "loss": 0.85469377, + "num_input_tokens_seen": 67175840, + "step": 3127, + "time_per_iteration": 2.7402892112731934 + }, + { + "auxiliary_loss_clip": 0.01192797, + "auxiliary_loss_mlp": 0.01036504, + "balance_loss_clip": 1.05767345, + "balance_loss_mlp": 1.02808809, + "epoch": 0.37611976191907653, + "flos": 20484847457280.0, + "grad_norm": 2.196500161653979, + "language_loss": 0.78965294, + "learning_rate": 2.869671957943002e-06, + "loss": 0.81194597, + "num_input_tokens_seen": 67194995, + "step": 3128, + "time_per_iteration": 2.6618897914886475 + }, + { + "auxiliary_loss_clip": 0.01167734, + "auxiliary_loss_mlp": 0.01027741, + "balance_loss_clip": 1.05774021, + "balance_loss_mlp": 1.01900351, + "epoch": 0.37624000480971564, + "flos": 21141797253120.0, + "grad_norm": 1.915227254507733, + "language_loss": 0.74295187, + "learning_rate": 2.8689704204929747e-06, + "loss": 0.76490664, + "num_input_tokens_seen": 67214175, + "step": 3129, + "time_per_iteration": 2.7214431762695312 + }, + { + "auxiliary_loss_clip": 0.01192673, + "auxiliary_loss_mlp": 0.01027857, + "balance_loss_clip": 1.05628967, + "balance_loss_mlp": 1.01895237, + "epoch": 0.3763602477003547, + "flos": 22564470205440.0, + "grad_norm": 2.6377825144717093, + "language_loss": 0.81170058, + "learning_rate": 2.8682687512208253e-06, + "loss": 0.83390594, + "num_input_tokens_seen": 67233185, + "step": 3130, + "time_per_iteration": 2.7020974159240723 + }, + { + "auxiliary_loss_clip": 0.01192131, + "auxiliary_loss_mlp": 0.01037443, + "balance_loss_clip": 1.05663919, + "balance_loss_mlp": 1.02798986, + "epoch": 0.3764804905909938, + "flos": 27526857851520.0, + "grad_norm": 2.0461775872977226, + "language_loss": 0.80679643, + "learning_rate": 2.8675669502329972e-06, + "loss": 0.82909214, + "num_input_tokens_seen": 67254715, + "step": 3131, + "time_per_iteration": 2.7068512439727783 + }, + { + "auxiliary_loss_clip": 0.01184949, + "auxiliary_loss_mlp": 0.00886449, + "balance_loss_clip": 1.0562942, + "balance_loss_mlp": 1.00072956, + "epoch": 0.3766007334816329, + "flos": 22528092706560.0, + "grad_norm": 2.3587567690301436, + "language_loss": 0.85812247, + "learning_rate": 2.866865017635952e-06, + "loss": 0.87883645, + "num_input_tokens_seen": 67272535, + "step": 3132, + "time_per_iteration": 2.6868882179260254 + }, + { + "auxiliary_loss_clip": 0.01160195, + "auxiliary_loss_mlp": 0.01030796, + "balance_loss_clip": 1.05374861, + "balance_loss_mlp": 1.0218972, + "epoch": 0.376720976372272, + "flos": 25957166532480.0, + "grad_norm": 1.6759121901847662, + "language_loss": 0.7979998, + "learning_rate": 2.866162953536174e-06, + "loss": 0.81990969, + "num_input_tokens_seen": 67293505, + "step": 3133, + "time_per_iteration": 2.817434787750244 + }, + { + "auxiliary_loss_clip": 0.01171409, + "auxiliary_loss_mlp": 0.00886427, + "balance_loss_clip": 1.05084324, + "balance_loss_mlp": 1.00063837, + "epoch": 0.3768412192629111, + "flos": 18041162411520.0, + "grad_norm": 1.6392461230681086, + "language_loss": 0.74875593, + "learning_rate": 2.8654607580401634e-06, + "loss": 0.76933432, + "num_input_tokens_seen": 67313240, + "step": 3134, + "time_per_iteration": 4.689310789108276 + }, + { + "auxiliary_loss_clip": 0.01081645, + "auxiliary_loss_mlp": 0.01003323, + "balance_loss_clip": 1.02277207, + "balance_loss_mlp": 1.00093901, + "epoch": 0.3769614621535502, + "flos": 62989472304000.0, + "grad_norm": 0.88550972670641, + "language_loss": 0.65262437, + "learning_rate": 2.8647584312544446e-06, + "loss": 0.67347407, + "num_input_tokens_seen": 67378445, + "step": 3135, + "time_per_iteration": 4.270604848861694 + }, + { + "auxiliary_loss_clip": 0.01157919, + "auxiliary_loss_mlp": 0.00886511, + "balance_loss_clip": 1.0492233, + "balance_loss_mlp": 1.00073183, + "epoch": 0.37708170504418925, + "flos": 23661685002240.0, + "grad_norm": 1.3784466415144458, + "language_loss": 0.85268414, + "learning_rate": 2.864055973285559e-06, + "loss": 0.87312841, + "num_input_tokens_seen": 67400445, + "step": 3136, + "time_per_iteration": 2.9324581623077393 + }, + { + "auxiliary_loss_clip": 0.01160181, + "auxiliary_loss_mlp": 0.01032507, + "balance_loss_clip": 1.04920125, + "balance_loss_mlp": 1.02360797, + "epoch": 0.37720194793482836, + "flos": 24423170353920.0, + "grad_norm": 2.0572589095245433, + "language_loss": 0.86414182, + "learning_rate": 2.8633533842400698e-06, + "loss": 0.8860687, + "num_input_tokens_seen": 67420645, + "step": 3137, + "time_per_iteration": 3.718168258666992 + }, + { + "auxiliary_loss_clip": 0.01186816, + "auxiliary_loss_mlp": 0.00887328, + "balance_loss_clip": 1.05728316, + "balance_loss_mlp": 1.00073004, + "epoch": 0.3773221908254674, + "flos": 20996502739200.0, + "grad_norm": 1.8199097037704939, + "language_loss": 0.77292937, + "learning_rate": 2.862650664224558e-06, + "loss": 0.79367077, + "num_input_tokens_seen": 67439495, + "step": 3138, + "time_per_iteration": 2.8241641521453857 + }, + { + "auxiliary_loss_clip": 0.01180277, + "auxiliary_loss_mlp": 0.01030451, + "balance_loss_clip": 1.05727172, + "balance_loss_mlp": 1.02201104, + "epoch": 0.37744243371610653, + "flos": 37631724958080.0, + "grad_norm": 1.408439361411264, + "language_loss": 0.69852614, + "learning_rate": 2.861947813345627e-06, + "loss": 0.72063351, + "num_input_tokens_seen": 67462195, + "step": 3139, + "time_per_iteration": 2.8955183029174805 + }, + { + "auxiliary_loss_clip": 0.01195504, + "auxiliary_loss_mlp": 0.00886856, + "balance_loss_clip": 1.0573566, + "balance_loss_mlp": 1.00065601, + "epoch": 0.37756267660674564, + "flos": 26140526484480.0, + "grad_norm": 11.607165645586614, + "language_loss": 0.72531301, + "learning_rate": 2.8612448317098974e-06, + "loss": 0.74613667, + "num_input_tokens_seen": 67482530, + "step": 3140, + "time_per_iteration": 2.8064470291137695 + }, + { + "auxiliary_loss_clip": 0.01166666, + "auxiliary_loss_mlp": 0.00887303, + "balance_loss_clip": 1.05080557, + "balance_loss_mlp": 1.00083256, + "epoch": 0.3776829194973847, + "flos": 19427888828160.0, + "grad_norm": 2.094587430409655, + "language_loss": 0.82894373, + "learning_rate": 2.8605417194240114e-06, + "loss": 0.84948343, + "num_input_tokens_seen": 67500890, + "step": 3141, + "time_per_iteration": 2.8072149753570557 + }, + { + "auxiliary_loss_clip": 0.01178134, + "auxiliary_loss_mlp": 0.01031816, + "balance_loss_clip": 1.05378294, + "balance_loss_mlp": 1.02310205, + "epoch": 0.3778031623880238, + "flos": 17382309194880.0, + "grad_norm": 2.0178475903261854, + "language_loss": 0.7858938, + "learning_rate": 2.8598384765946315e-06, + "loss": 0.80799323, + "num_input_tokens_seen": 67519545, + "step": 3142, + "time_per_iteration": 2.77449893951416 + }, + { + "auxiliary_loss_clip": 0.01189994, + "auxiliary_loss_mlp": 0.01026129, + "balance_loss_clip": 1.05376792, + "balance_loss_mlp": 1.0170629, + "epoch": 0.3779234052786629, + "flos": 27125843437440.0, + "grad_norm": 1.8620794304529589, + "language_loss": 0.71242929, + "learning_rate": 2.8591351033284377e-06, + "loss": 0.73459053, + "num_input_tokens_seen": 67539275, + "step": 3143, + "time_per_iteration": 2.784519672393799 + }, + { + "auxiliary_loss_clip": 0.01182364, + "auxiliary_loss_mlp": 0.0103178, + "balance_loss_clip": 1.05269122, + "balance_loss_mlp": 1.02300048, + "epoch": 0.37804364816930197, + "flos": 19682639061120.0, + "grad_norm": 2.094954077152784, + "language_loss": 0.84082448, + "learning_rate": 2.8584315997321325e-06, + "loss": 0.86296594, + "num_input_tokens_seen": 67558280, + "step": 3144, + "time_per_iteration": 2.774106979370117 + }, + { + "auxiliary_loss_clip": 0.01189897, + "auxiliary_loss_mlp": 0.01041339, + "balance_loss_clip": 1.05376172, + "balance_loss_mlp": 1.03196311, + "epoch": 0.3781638910599411, + "flos": 22702905221760.0, + "grad_norm": 4.0458708382801785, + "language_loss": 0.78420126, + "learning_rate": 2.8577279659124356e-06, + "loss": 0.80651367, + "num_input_tokens_seen": 67575955, + "step": 3145, + "time_per_iteration": 2.7217612266540527 + }, + { + "auxiliary_loss_clip": 0.01177015, + "auxiliary_loss_mlp": 0.01028372, + "balance_loss_clip": 1.05043733, + "balance_loss_mlp": 1.01949143, + "epoch": 0.3782841339505802, + "flos": 14647604158080.0, + "grad_norm": 1.747507687692277, + "language_loss": 0.83602655, + "learning_rate": 2.857024201976089e-06, + "loss": 0.85808051, + "num_input_tokens_seen": 67593515, + "step": 3146, + "time_per_iteration": 2.6875994205474854 + }, + { + "auxiliary_loss_clip": 0.01171635, + "auxiliary_loss_mlp": 0.01036042, + "balance_loss_clip": 1.05296803, + "balance_loss_mlp": 1.02646303, + "epoch": 0.37840437684121925, + "flos": 32818223185920.0, + "grad_norm": 2.5863817705948398, + "language_loss": 0.72808224, + "learning_rate": 2.8563203080298516e-06, + "loss": 0.75015903, + "num_input_tokens_seen": 67614290, + "step": 3147, + "time_per_iteration": 2.9569876194000244 + }, + { + "auxiliary_loss_clip": 0.01169735, + "auxiliary_loss_mlp": 0.00887232, + "balance_loss_clip": 1.05296564, + "balance_loss_mlp": 1.00063634, + "epoch": 0.37852461973185836, + "flos": 18369206346240.0, + "grad_norm": 2.524759103285165, + "language_loss": 0.88984108, + "learning_rate": 2.855616284180505e-06, + "loss": 0.9104107, + "num_input_tokens_seen": 67631340, + "step": 3148, + "time_per_iteration": 2.778135061264038 + }, + { + "auxiliary_loss_clip": 0.01086379, + "auxiliary_loss_mlp": 0.01006399, + "balance_loss_clip": 1.02389693, + "balance_loss_mlp": 1.00390708, + "epoch": 0.37864486262249747, + "flos": 59500680117120.0, + "grad_norm": 0.8742175991272886, + "language_loss": 0.66069889, + "learning_rate": 2.8549121305348477e-06, + "loss": 0.68162668, + "num_input_tokens_seen": 67691125, + "step": 3149, + "time_per_iteration": 3.2240612506866455 + }, + { + "auxiliary_loss_clip": 0.01181224, + "auxiliary_loss_mlp": 0.01032621, + "balance_loss_clip": 1.05222344, + "balance_loss_mlp": 1.02389455, + "epoch": 0.3787651055131365, + "flos": 23363015414400.0, + "grad_norm": 2.293430241950504, + "language_loss": 0.83540928, + "learning_rate": 2.8542078471997006e-06, + "loss": 0.8575477, + "num_input_tokens_seen": 67708740, + "step": 3150, + "time_per_iteration": 2.7336699962615967 + }, + { + "auxiliary_loss_clip": 0.01181958, + "auxiliary_loss_mlp": 0.01035657, + "balance_loss_clip": 1.05154181, + "balance_loss_mlp": 1.0269134, + "epoch": 0.37888534840377563, + "flos": 24601394661120.0, + "grad_norm": 1.6943168987516661, + "language_loss": 0.7561878, + "learning_rate": 2.8535034342819013e-06, + "loss": 0.77836394, + "num_input_tokens_seen": 67726150, + "step": 3151, + "time_per_iteration": 2.7788944244384766 + }, + { + "auxiliary_loss_clip": 0.01187248, + "auxiliary_loss_mlp": 0.01035967, + "balance_loss_clip": 1.05343831, + "balance_loss_mlp": 1.02721155, + "epoch": 0.37900559129441475, + "flos": 23986891762560.0, + "grad_norm": 1.6119558794588775, + "language_loss": 0.72459984, + "learning_rate": 2.85279889188831e-06, + "loss": 0.74683195, + "num_input_tokens_seen": 67746525, + "step": 3152, + "time_per_iteration": 2.754934787750244 + }, + { + "auxiliary_loss_clip": 0.01158993, + "auxiliary_loss_mlp": 0.01035034, + "balance_loss_clip": 1.0465169, + "balance_loss_mlp": 1.02571213, + "epoch": 0.3791258341850538, + "flos": 24644667571200.0, + "grad_norm": 1.6848926886327522, + "language_loss": 0.81164652, + "learning_rate": 2.852094220125805e-06, + "loss": 0.83358681, + "num_input_tokens_seen": 67766035, + "step": 3153, + "time_per_iteration": 3.032672882080078 + }, + { + "auxiliary_loss_clip": 0.01183808, + "auxiliary_loss_mlp": 0.01036686, + "balance_loss_clip": 1.05571127, + "balance_loss_mlp": 1.02726293, + "epoch": 0.3792460770756929, + "flos": 17420841509760.0, + "grad_norm": 2.2809267421933113, + "language_loss": 0.7098068, + "learning_rate": 2.8513894191012846e-06, + "loss": 0.73201174, + "num_input_tokens_seen": 67785015, + "step": 3154, + "time_per_iteration": 2.7488291263580322 + }, + { + "auxiliary_loss_clip": 0.01192332, + "auxiliary_loss_mlp": 0.0103237, + "balance_loss_clip": 1.05536807, + "balance_loss_mlp": 1.02386427, + "epoch": 0.37936631996633197, + "flos": 24206557386240.0, + "grad_norm": 1.5748064718928492, + "language_loss": 0.78908932, + "learning_rate": 2.8506844889216664e-06, + "loss": 0.81133634, + "num_input_tokens_seen": 67804400, + "step": 3155, + "time_per_iteration": 2.741234302520752 + }, + { + "auxiliary_loss_clip": 0.01081099, + "auxiliary_loss_mlp": 0.01005652, + "balance_loss_clip": 1.02666342, + "balance_loss_mlp": 1.00286233, + "epoch": 0.3794865628569711, + "flos": 70297114752000.0, + "grad_norm": 0.8617660284325658, + "language_loss": 0.6285708, + "learning_rate": 2.849979429693887e-06, + "loss": 0.64943826, + "num_input_tokens_seen": 67865385, + "step": 3156, + "time_per_iteration": 3.376011610031128 + }, + { + "auxiliary_loss_clip": 0.01190488, + "auxiliary_loss_mlp": 0.01033629, + "balance_loss_clip": 1.05488813, + "balance_loss_mlp": 1.02428889, + "epoch": 0.3796068057476102, + "flos": 15779364860160.0, + "grad_norm": 2.137403097640614, + "language_loss": 0.74367166, + "learning_rate": 2.8492742415249042e-06, + "loss": 0.76591283, + "num_input_tokens_seen": 67883030, + "step": 3157, + "time_per_iteration": 2.689849615097046 + }, + { + "auxiliary_loss_clip": 0.01188564, + "auxiliary_loss_mlp": 0.01031143, + "balance_loss_clip": 1.05341339, + "balance_loss_mlp": 1.02168417, + "epoch": 0.37972704863824924, + "flos": 25191694771200.0, + "grad_norm": 1.5846437349397176, + "language_loss": 0.76350939, + "learning_rate": 2.848568924521694e-06, + "loss": 0.78570646, + "num_input_tokens_seen": 67903810, + "step": 3158, + "time_per_iteration": 2.689192533493042 + }, + { + "auxiliary_loss_clip": 0.01172626, + "auxiliary_loss_mlp": 0.01032519, + "balance_loss_clip": 1.04883194, + "balance_loss_mlp": 1.02280307, + "epoch": 0.37984729152888835, + "flos": 26210372480640.0, + "grad_norm": 2.006123563272714, + "language_loss": 0.73666525, + "learning_rate": 2.8478634787912526e-06, + "loss": 0.75871664, + "num_input_tokens_seen": 67921865, + "step": 3159, + "time_per_iteration": 2.765493392944336 + }, + { + "auxiliary_loss_clip": 0.01181556, + "auxiliary_loss_mlp": 0.01035816, + "balance_loss_clip": 1.05283237, + "balance_loss_mlp": 1.02674985, + "epoch": 0.37996753441952746, + "flos": 25629302165760.0, + "grad_norm": 2.18962829975436, + "language_loss": 0.76316595, + "learning_rate": 2.847157904440596e-06, + "loss": 0.78533965, + "num_input_tokens_seen": 67941595, + "step": 3160, + "time_per_iteration": 3.735621929168701 + }, + { + "auxiliary_loss_clip": 0.01180186, + "auxiliary_loss_mlp": 0.01034638, + "balance_loss_clip": 1.05375886, + "balance_loss_mlp": 1.02574444, + "epoch": 0.3800877773101665, + "flos": 20118414862080.0, + "grad_norm": 1.7156288340625543, + "language_loss": 0.73816311, + "learning_rate": 2.846452201576759e-06, + "loss": 0.76031137, + "num_input_tokens_seen": 67960970, + "step": 3161, + "time_per_iteration": 3.6004135608673096 + }, + { + "auxiliary_loss_clip": 0.01076375, + "auxiliary_loss_mlp": 0.01002324, + "balance_loss_clip": 1.01984549, + "balance_loss_mlp": 1.00016618, + "epoch": 0.38020802020080563, + "flos": 63053608037760.0, + "grad_norm": 0.855215917188423, + "language_loss": 0.62802094, + "learning_rate": 2.845746370306795e-06, + "loss": 0.64880794, + "num_input_tokens_seen": 68026160, + "step": 3162, + "time_per_iteration": 3.3732473850250244 + }, + { + "auxiliary_loss_clip": 0.01181357, + "auxiliary_loss_mlp": 0.01032852, + "balance_loss_clip": 1.05267942, + "balance_loss_mlp": 1.02362537, + "epoch": 0.38032826309144474, + "flos": 21288420570240.0, + "grad_norm": 1.9006528341523667, + "language_loss": 0.78260136, + "learning_rate": 2.84504041073778e-06, + "loss": 0.80474341, + "num_input_tokens_seen": 68044575, + "step": 3163, + "time_per_iteration": 3.6548750400543213 + }, + { + "auxiliary_loss_clip": 0.01163044, + "auxiliary_loss_mlp": 0.01039005, + "balance_loss_clip": 1.05120361, + "balance_loss_mlp": 1.0295403, + "epoch": 0.3804485059820838, + "flos": 18954119416320.0, + "grad_norm": 1.767739511062379, + "language_loss": 0.79029727, + "learning_rate": 2.844334322976806e-06, + "loss": 0.81231773, + "num_input_tokens_seen": 68064790, + "step": 3164, + "time_per_iteration": 2.7371127605438232 + }, + { + "auxiliary_loss_clip": 0.01158185, + "auxiliary_loss_mlp": 0.01028558, + "balance_loss_clip": 1.04889631, + "balance_loss_mlp": 1.01921821, + "epoch": 0.3805687488727229, + "flos": 21833759831040.0, + "grad_norm": 1.6643697542659952, + "language_loss": 0.83442521, + "learning_rate": 2.8436281071309866e-06, + "loss": 0.85629261, + "num_input_tokens_seen": 68083330, + "step": 3165, + "time_per_iteration": 2.894789457321167 + }, + { + "auxiliary_loss_clip": 0.01065028, + "auxiliary_loss_mlp": 0.01007259, + "balance_loss_clip": 1.01724756, + "balance_loss_mlp": 1.00479162, + "epoch": 0.380688991763362, + "flos": 58546209968640.0, + "grad_norm": 0.7436687869958173, + "language_loss": 0.53001761, + "learning_rate": 2.842921763307455e-06, + "loss": 0.55074048, + "num_input_tokens_seen": 68146140, + "step": 3166, + "time_per_iteration": 3.3993446826934814 + }, + { + "auxiliary_loss_clip": 0.01164393, + "auxiliary_loss_mlp": 0.0102877, + "balance_loss_clip": 1.04744828, + "balance_loss_mlp": 1.02038324, + "epoch": 0.38080923465400107, + "flos": 23799509487360.0, + "grad_norm": 1.776125359187705, + "language_loss": 0.82662308, + "learning_rate": 2.842215291613361e-06, + "loss": 0.84855473, + "num_input_tokens_seen": 68164520, + "step": 3167, + "time_per_iteration": 2.789640188217163 + }, + { + "auxiliary_loss_clip": 0.01042064, + "auxiliary_loss_mlp": 0.01004808, + "balance_loss_clip": 1.01890612, + "balance_loss_mlp": 1.00218534, + "epoch": 0.3809294775446402, + "flos": 54969866380800.0, + "grad_norm": 0.8185682223355285, + "language_loss": 0.59265697, + "learning_rate": 2.8415086921558774e-06, + "loss": 0.61312568, + "num_input_tokens_seen": 68227945, + "step": 3168, + "time_per_iteration": 3.6084651947021484 + }, + { + "auxiliary_loss_clip": 0.01156599, + "auxiliary_loss_mlp": 0.01025663, + "balance_loss_clip": 1.04528618, + "balance_loss_mlp": 1.01711595, + "epoch": 0.38104972043527924, + "flos": 24643697904000.0, + "grad_norm": 1.5633537390332632, + "language_loss": 0.78691411, + "learning_rate": 2.840801965042194e-06, + "loss": 0.80873674, + "num_input_tokens_seen": 68247405, + "step": 3169, + "time_per_iteration": 3.0065083503723145 + }, + { + "auxiliary_loss_clip": 0.01161683, + "auxiliary_loss_mlp": 0.01030828, + "balance_loss_clip": 1.04708338, + "balance_loss_mlp": 1.02200687, + "epoch": 0.38116996332591835, + "flos": 22856783086080.0, + "grad_norm": 1.7815912325560588, + "language_loss": 0.83563691, + "learning_rate": 2.840095110379521e-06, + "loss": 0.85756207, + "num_input_tokens_seen": 68266925, + "step": 3170, + "time_per_iteration": 2.8443219661712646 + }, + { + "auxiliary_loss_clip": 0.01056845, + "auxiliary_loss_mlp": 0.01006744, + "balance_loss_clip": 1.0196979, + "balance_loss_mlp": 1.00435972, + "epoch": 0.38129020621655746, + "flos": 60836160804480.0, + "grad_norm": 0.7821001883323218, + "language_loss": 0.53869772, + "learning_rate": 2.8393881282750884e-06, + "loss": 0.55933368, + "num_input_tokens_seen": 68329755, + "step": 3171, + "time_per_iteration": 3.3207976818084717 + }, + { + "auxiliary_loss_clip": 0.01173042, + "auxiliary_loss_mlp": 0.01025852, + "balance_loss_clip": 1.05275512, + "balance_loss_mlp": 1.01697135, + "epoch": 0.3814104491071965, + "flos": 21648101408640.0, + "grad_norm": 2.0533319671304286, + "language_loss": 0.78593791, + "learning_rate": 2.838681018836144e-06, + "loss": 0.80792689, + "num_input_tokens_seen": 68347075, + "step": 3172, + "time_per_iteration": 2.81002140045166 + }, + { + "auxiliary_loss_clip": 0.01163665, + "auxiliary_loss_mlp": 0.00886383, + "balance_loss_clip": 1.04882514, + "balance_loss_mlp": 1.00085366, + "epoch": 0.3815306919978356, + "flos": 19099090707840.0, + "grad_norm": 2.5299617850675045, + "language_loss": 0.78391516, + "learning_rate": 2.837973782169955e-06, + "loss": 0.8044157, + "num_input_tokens_seen": 68365450, + "step": 3173, + "time_per_iteration": 2.777172803878784 + }, + { + "auxiliary_loss_clip": 0.01085374, + "auxiliary_loss_mlp": 0.01003293, + "balance_loss_clip": 1.01958132, + "balance_loss_mlp": 1.00117135, + "epoch": 0.38165093488847474, + "flos": 67067918156160.0, + "grad_norm": 0.8345850819597067, + "language_loss": 0.59244668, + "learning_rate": 2.8372664183838096e-06, + "loss": 0.61333334, + "num_input_tokens_seen": 68428470, + "step": 3174, + "time_per_iteration": 3.304480791091919 + }, + { + "auxiliary_loss_clip": 0.01189888, + "auxiliary_loss_mlp": 0.01031582, + "balance_loss_clip": 1.05410361, + "balance_loss_mlp": 1.02282, + "epoch": 0.3817711777791138, + "flos": 22341105480960.0, + "grad_norm": 2.0516595977180865, + "language_loss": 0.67909789, + "learning_rate": 2.836558927585015e-06, + "loss": 0.70131266, + "num_input_tokens_seen": 68445440, + "step": 3175, + "time_per_iteration": 2.703110456466675 + }, + { + "auxiliary_loss_clip": 0.01182794, + "auxiliary_loss_mlp": 0.01027555, + "balance_loss_clip": 1.05252576, + "balance_loss_mlp": 1.01896024, + "epoch": 0.3818914206697529, + "flos": 22820621068800.0, + "grad_norm": 1.8727171432912166, + "language_loss": 0.82106656, + "learning_rate": 2.8358513098808957e-06, + "loss": 0.84317005, + "num_input_tokens_seen": 68465755, + "step": 3176, + "time_per_iteration": 2.7954514026641846 + }, + { + "auxiliary_loss_clip": 0.01141386, + "auxiliary_loss_mlp": 0.01027908, + "balance_loss_clip": 1.04817915, + "balance_loss_mlp": 1.01917577, + "epoch": 0.382011663560392, + "flos": 24386074583040.0, + "grad_norm": 2.272485209449576, + "language_loss": 0.77049738, + "learning_rate": 2.835143565378798e-06, + "loss": 0.79219031, + "num_input_tokens_seen": 68486220, + "step": 3177, + "time_per_iteration": 2.8099617958068848 + }, + { + "auxiliary_loss_clip": 0.01136567, + "auxiliary_loss_mlp": 0.01031571, + "balance_loss_clip": 1.04564822, + "balance_loss_mlp": 1.0226779, + "epoch": 0.38213190645103107, + "flos": 21981568296960.0, + "grad_norm": 2.15421969115284, + "language_loss": 0.78691983, + "learning_rate": 2.8344356941860847e-06, + "loss": 0.80860114, + "num_input_tokens_seen": 68505850, + "step": 3178, + "time_per_iteration": 2.935560703277588 + }, + { + "auxiliary_loss_clip": 0.01157343, + "auxiliary_loss_mlp": 0.01036834, + "balance_loss_clip": 1.05008817, + "balance_loss_mlp": 1.02718449, + "epoch": 0.3822521493416702, + "flos": 35516945773440.0, + "grad_norm": 2.3681512383602272, + "language_loss": 0.66184354, + "learning_rate": 2.8337276964101403e-06, + "loss": 0.68378532, + "num_input_tokens_seen": 68526290, + "step": 3179, + "time_per_iteration": 2.8816568851470947 + }, + { + "auxiliary_loss_clip": 0.01183396, + "auxiliary_loss_mlp": 0.01027669, + "balance_loss_clip": 1.05212951, + "balance_loss_mlp": 1.0187583, + "epoch": 0.3823723922323093, + "flos": 21069904181760.0, + "grad_norm": 1.7785523507613574, + "language_loss": 0.76257896, + "learning_rate": 2.833019572158367e-06, + "loss": 0.78468961, + "num_input_tokens_seen": 68544725, + "step": 3180, + "time_per_iteration": 2.7630012035369873 + }, + { + "auxiliary_loss_clip": 0.01173201, + "auxiliary_loss_mlp": 0.01029287, + "balance_loss_clip": 1.05311739, + "balance_loss_mlp": 1.01988745, + "epoch": 0.38249263512294834, + "flos": 19789149864960.0, + "grad_norm": 1.7926495026660982, + "language_loss": 0.80006242, + "learning_rate": 2.8323113215381872e-06, + "loss": 0.82208729, + "num_input_tokens_seen": 68563070, + "step": 3181, + "time_per_iteration": 2.8680272102355957 + }, + { + "auxiliary_loss_clip": 0.01160023, + "auxiliary_loss_mlp": 0.01025972, + "balance_loss_clip": 1.04974222, + "balance_loss_mlp": 1.01596427, + "epoch": 0.38261287801358745, + "flos": 21433930565760.0, + "grad_norm": 1.923252057709925, + "language_loss": 0.76378369, + "learning_rate": 2.831602944657042e-06, + "loss": 0.78564358, + "num_input_tokens_seen": 68581150, + "step": 3182, + "time_per_iteration": 2.83957576751709 + }, + { + "auxiliary_loss_clip": 0.0117652, + "auxiliary_loss_mlp": 0.01030501, + "balance_loss_clip": 1.05146003, + "balance_loss_mlp": 1.02145863, + "epoch": 0.38273312090422656, + "flos": 21981568296960.0, + "grad_norm": 2.2653328468749954, + "language_loss": 0.7377075, + "learning_rate": 2.830894441622391e-06, + "loss": 0.75977778, + "num_input_tokens_seen": 68597800, + "step": 3183, + "time_per_iteration": 2.806718587875366 + }, + { + "auxiliary_loss_clip": 0.01157341, + "auxiliary_loss_mlp": 0.00886844, + "balance_loss_clip": 1.04478455, + "balance_loss_mlp": 1.00086999, + "epoch": 0.3828533637948656, + "flos": 24790895838720.0, + "grad_norm": 1.8589986024951908, + "language_loss": 0.79504704, + "learning_rate": 2.8301858125417134e-06, + "loss": 0.81548887, + "num_input_tokens_seen": 68617640, + "step": 3184, + "time_per_iteration": 2.868279218673706 + }, + { + "auxiliary_loss_clip": 0.01170061, + "auxiliary_loss_mlp": 0.01024446, + "balance_loss_clip": 1.05230141, + "balance_loss_mlp": 1.01603007, + "epoch": 0.38297360668550473, + "flos": 22455445449600.0, + "grad_norm": 1.8588750467369086, + "language_loss": 0.73745406, + "learning_rate": 2.8294770575225082e-06, + "loss": 0.75939906, + "num_input_tokens_seen": 68637770, + "step": 3185, + "time_per_iteration": 4.013945817947388 + }, + { + "auxiliary_loss_clip": 0.01183133, + "auxiliary_loss_mlp": 0.01032991, + "balance_loss_clip": 1.05492723, + "balance_loss_mlp": 1.0239073, + "epoch": 0.3830938495761438, + "flos": 24896903852160.0, + "grad_norm": 1.939088926696426, + "language_loss": 0.83872235, + "learning_rate": 2.828768176672293e-06, + "loss": 0.86088359, + "num_input_tokens_seen": 68656885, + "step": 3186, + "time_per_iteration": 3.7162039279937744 + }, + { + "auxiliary_loss_clip": 0.01159739, + "auxiliary_loss_mlp": 0.0103415, + "balance_loss_clip": 1.04822314, + "balance_loss_mlp": 1.02479768, + "epoch": 0.3832140924667829, + "flos": 33036236784000.0, + "grad_norm": 1.6411956350461077, + "language_loss": 0.71645594, + "learning_rate": 2.8280591700986044e-06, + "loss": 0.73839486, + "num_input_tokens_seen": 68678750, + "step": 3187, + "time_per_iteration": 3.8469526767730713 + }, + { + "auxiliary_loss_clip": 0.0117431, + "auxiliary_loss_mlp": 0.01032712, + "balance_loss_clip": 1.0493536, + "balance_loss_mlp": 1.02315176, + "epoch": 0.383334335357422, + "flos": 31903721896320.0, + "grad_norm": 1.8261805058940095, + "language_loss": 0.74845409, + "learning_rate": 2.827350037908999e-06, + "loss": 0.77052432, + "num_input_tokens_seen": 68698190, + "step": 3188, + "time_per_iteration": 2.8818907737731934 + }, + { + "auxiliary_loss_clip": 0.01167865, + "auxiliary_loss_mlp": 0.01033714, + "balance_loss_clip": 1.0496695, + "balance_loss_mlp": 1.02368855, + "epoch": 0.38345457824806106, + "flos": 19791915212160.0, + "grad_norm": 2.206404770899599, + "language_loss": 0.78650272, + "learning_rate": 2.8266407802110496e-06, + "loss": 0.80851853, + "num_input_tokens_seen": 68716445, + "step": 3189, + "time_per_iteration": 3.7412185668945312 + }, + { + "auxiliary_loss_clip": 0.01138789, + "auxiliary_loss_mlp": 0.0103251, + "balance_loss_clip": 1.04310644, + "balance_loss_mlp": 1.02339661, + "epoch": 0.3835748211387002, + "flos": 22419391173120.0, + "grad_norm": 1.9020082514572363, + "language_loss": 0.75365281, + "learning_rate": 2.8259313971123515e-06, + "loss": 0.77536583, + "num_input_tokens_seen": 68737565, + "step": 3190, + "time_per_iteration": 2.862539291381836 + }, + { + "auxiliary_loss_clip": 0.01180467, + "auxiliary_loss_mlp": 0.01028226, + "balance_loss_clip": 1.05563593, + "balance_loss_mlp": 1.02010751, + "epoch": 0.3836950640293393, + "flos": 25118436983040.0, + "grad_norm": 1.7730931582376666, + "language_loss": 0.78238392, + "learning_rate": 2.8252218887205166e-06, + "loss": 0.8044709, + "num_input_tokens_seen": 68758255, + "step": 3191, + "time_per_iteration": 2.8131868839263916 + }, + { + "auxiliary_loss_clip": 0.0114606, + "auxiliary_loss_mlp": 0.01028074, + "balance_loss_clip": 1.05049849, + "balance_loss_mlp": 1.01869249, + "epoch": 0.38381530691997834, + "flos": 21799213925760.0, + "grad_norm": 2.04498024969194, + "language_loss": 0.81187266, + "learning_rate": 2.824512255143178e-06, + "loss": 0.83361399, + "num_input_tokens_seen": 68777490, + "step": 3192, + "time_per_iteration": 2.8834049701690674 + }, + { + "auxiliary_loss_clip": 0.01166928, + "auxiliary_loss_mlp": 0.01032675, + "balance_loss_clip": 1.0504837, + "balance_loss_mlp": 1.02316248, + "epoch": 0.38393554981061745, + "flos": 21252689516160.0, + "grad_norm": 1.6591038072767779, + "language_loss": 0.79272026, + "learning_rate": 2.8238024964879855e-06, + "loss": 0.81471628, + "num_input_tokens_seen": 68798385, + "step": 3193, + "time_per_iteration": 2.8716495037078857 + }, + { + "auxiliary_loss_clip": 0.01194628, + "auxiliary_loss_mlp": 0.01028987, + "balance_loss_clip": 1.05515003, + "balance_loss_mlp": 1.02045119, + "epoch": 0.38405579270125656, + "flos": 17019360218880.0, + "grad_norm": 2.0480004622475967, + "language_loss": 0.76681548, + "learning_rate": 2.8230926128626095e-06, + "loss": 0.78905165, + "num_input_tokens_seen": 68816880, + "step": 3194, + "time_per_iteration": 2.7376463413238525 + }, + { + "auxiliary_loss_clip": 0.01163055, + "auxiliary_loss_mlp": 0.01033242, + "balance_loss_clip": 1.04756951, + "balance_loss_mlp": 1.02397323, + "epoch": 0.3841760355918956, + "flos": 21835375943040.0, + "grad_norm": 1.9512585020179583, + "language_loss": 0.79171354, + "learning_rate": 2.822382604374738e-06, + "loss": 0.81367654, + "num_input_tokens_seen": 68835805, + "step": 3195, + "time_per_iteration": 2.889362335205078 + }, + { + "auxiliary_loss_clip": 0.01173232, + "auxiliary_loss_mlp": 0.01036797, + "balance_loss_clip": 1.05662239, + "balance_loss_mlp": 1.02780318, + "epoch": 0.3842962784825347, + "flos": 25915114684800.0, + "grad_norm": 2.1430256205930296, + "language_loss": 0.65622878, + "learning_rate": 2.8216724711320793e-06, + "loss": 0.67832899, + "num_input_tokens_seen": 68854930, + "step": 3196, + "time_per_iteration": 2.87270188331604 + }, + { + "auxiliary_loss_clip": 0.01189473, + "auxiliary_loss_mlp": 0.00886053, + "balance_loss_clip": 1.05344772, + "balance_loss_mlp": 1.00094306, + "epoch": 0.38441652137317384, + "flos": 25337492075520.0, + "grad_norm": 1.5655043327346458, + "language_loss": 0.79787493, + "learning_rate": 2.820962213242361e-06, + "loss": 0.8186301, + "num_input_tokens_seen": 68874260, + "step": 3197, + "time_per_iteration": 2.756181001663208 + }, + { + "auxiliary_loss_clip": 0.01176502, + "auxiliary_loss_mlp": 0.01033347, + "balance_loss_clip": 1.05452514, + "balance_loss_mlp": 1.02503777, + "epoch": 0.3845367642638129, + "flos": 18113486446080.0, + "grad_norm": 1.9883852537309545, + "language_loss": 0.84260815, + "learning_rate": 2.8202518308133264e-06, + "loss": 0.86470664, + "num_input_tokens_seen": 68891535, + "step": 3198, + "time_per_iteration": 2.8175208568573 + }, + { + "auxiliary_loss_clip": 0.01192541, + "auxiliary_loss_mlp": 0.0103, + "balance_loss_clip": 1.05297852, + "balance_loss_mlp": 1.02083266, + "epoch": 0.384657007154452, + "flos": 25228395492480.0, + "grad_norm": 1.8066303905757561, + "language_loss": 0.73348439, + "learning_rate": 2.8195413239527426e-06, + "loss": 0.75570977, + "num_input_tokens_seen": 68911275, + "step": 3199, + "time_per_iteration": 2.7383873462677 + }, + { + "auxiliary_loss_clip": 0.01175738, + "auxiliary_loss_mlp": 0.01034956, + "balance_loss_clip": 1.05027664, + "balance_loss_mlp": 1.02598011, + "epoch": 0.38477725004509106, + "flos": 19865855358720.0, + "grad_norm": 2.5698111097105656, + "language_loss": 0.80755532, + "learning_rate": 2.8188306927683906e-06, + "loss": 0.82966226, + "num_input_tokens_seen": 68930745, + "step": 3200, + "time_per_iteration": 2.7854244709014893 + }, + { + "auxiliary_loss_clip": 0.01172934, + "auxiliary_loss_mlp": 0.01028336, + "balance_loss_clip": 1.05330229, + "balance_loss_mlp": 1.01950812, + "epoch": 0.38489749293573017, + "flos": 18259391491200.0, + "grad_norm": 2.0020254457723765, + "language_loss": 0.75050098, + "learning_rate": 2.818119937368074e-06, + "loss": 0.77251369, + "num_input_tokens_seen": 68949380, + "step": 3201, + "time_per_iteration": 2.8611674308776855 + }, + { + "auxiliary_loss_clip": 0.01186868, + "auxiliary_loss_mlp": 0.01031031, + "balance_loss_clip": 1.05342603, + "balance_loss_mlp": 1.02117848, + "epoch": 0.3850177358263693, + "flos": 24389163152640.0, + "grad_norm": 1.9188667784855493, + "language_loss": 0.66098905, + "learning_rate": 2.817409057859613e-06, + "loss": 0.68316805, + "num_input_tokens_seen": 68968370, + "step": 3202, + "time_per_iteration": 2.803774833679199 + }, + { + "auxiliary_loss_clip": 0.01148365, + "auxiliary_loss_mlp": 0.01036706, + "balance_loss_clip": 1.04919028, + "balance_loss_mlp": 1.02691853, + "epoch": 0.38513797871700833, + "flos": 17671533505920.0, + "grad_norm": 2.020003193240035, + "language_loss": 0.78920269, + "learning_rate": 2.8166980543508482e-06, + "loss": 0.81105345, + "num_input_tokens_seen": 68984260, + "step": 3203, + "time_per_iteration": 2.920121431350708 + }, + { + "auxiliary_loss_clip": 0.01194405, + "auxiliary_loss_mlp": 0.0103275, + "balance_loss_clip": 1.05674911, + "balance_loss_mlp": 1.02382731, + "epoch": 0.38525822160764744, + "flos": 25739583897600.0, + "grad_norm": 2.0784523185187065, + "language_loss": 0.7975443, + "learning_rate": 2.815986926949638e-06, + "loss": 0.81981587, + "num_input_tokens_seen": 69002760, + "step": 3204, + "time_per_iteration": 2.739306926727295 + }, + { + "auxiliary_loss_clip": 0.01180932, + "auxiliary_loss_mlp": 0.01022932, + "balance_loss_clip": 1.05379057, + "balance_loss_mlp": 1.01420546, + "epoch": 0.38537846449828655, + "flos": 20193647898240.0, + "grad_norm": 2.0811368432512447, + "language_loss": 0.80155945, + "learning_rate": 2.8152756757638597e-06, + "loss": 0.82359815, + "num_input_tokens_seen": 69021260, + "step": 3205, + "time_per_iteration": 2.7646753787994385 + }, + { + "auxiliary_loss_clip": 0.01179639, + "auxiliary_loss_mlp": 0.01031604, + "balance_loss_clip": 1.05285907, + "balance_loss_mlp": 1.02237713, + "epoch": 0.3854987073889256, + "flos": 23039352938880.0, + "grad_norm": 2.1914067214491046, + "language_loss": 0.84491765, + "learning_rate": 2.8145643009014093e-06, + "loss": 0.86703008, + "num_input_tokens_seen": 69039755, + "step": 3206, + "time_per_iteration": 2.7809505462646484 + }, + { + "auxiliary_loss_clip": 0.01181943, + "auxiliary_loss_mlp": 0.0103438, + "balance_loss_clip": 1.05305934, + "balance_loss_mlp": 1.02520049, + "epoch": 0.3856189502795647, + "flos": 20190631155840.0, + "grad_norm": 1.8051983236228792, + "language_loss": 0.7897262, + "learning_rate": 2.813852802470202e-06, + "loss": 0.81188941, + "num_input_tokens_seen": 69057650, + "step": 3207, + "time_per_iteration": 2.7314329147338867 + }, + { + "auxiliary_loss_clip": 0.01169247, + "auxiliary_loss_mlp": 0.01031557, + "balance_loss_clip": 1.05198801, + "balance_loss_mlp": 1.0225929, + "epoch": 0.38573919317020383, + "flos": 25702631781120.0, + "grad_norm": 2.148987380817577, + "language_loss": 0.7193327, + "learning_rate": 2.8131411805781717e-06, + "loss": 0.7413407, + "num_input_tokens_seen": 69077775, + "step": 3208, + "time_per_iteration": 2.7385034561157227 + }, + { + "auxiliary_loss_clip": 0.01178036, + "auxiliary_loss_mlp": 0.01039277, + "balance_loss_clip": 1.05905795, + "balance_loss_mlp": 1.02966857, + "epoch": 0.3858594360608429, + "flos": 29821405628160.0, + "grad_norm": 4.075984717218951, + "language_loss": 0.63987303, + "learning_rate": 2.8124294353332707e-06, + "loss": 0.66204613, + "num_input_tokens_seen": 69096450, + "step": 3209, + "time_per_iteration": 2.8505353927612305 + }, + { + "auxiliary_loss_clip": 0.01166025, + "auxiliary_loss_mlp": 0.01037163, + "balance_loss_clip": 1.05006099, + "balance_loss_mlp": 1.02855647, + "epoch": 0.385979678951482, + "flos": 24790428961920.0, + "grad_norm": 1.6203948451884067, + "language_loss": 0.77139211, + "learning_rate": 2.8117175668434713e-06, + "loss": 0.79342401, + "num_input_tokens_seen": 69116110, + "step": 3210, + "time_per_iteration": 2.9388837814331055 + }, + { + "auxiliary_loss_clip": 0.01191896, + "auxiliary_loss_mlp": 0.01031029, + "balance_loss_clip": 1.05399692, + "balance_loss_mlp": 1.02183223, + "epoch": 0.3860999218421211, + "flos": 21287881866240.0, + "grad_norm": 2.3111775192466424, + "language_loss": 0.70497888, + "learning_rate": 2.811005575216762e-06, + "loss": 0.72720814, + "num_input_tokens_seen": 69134825, + "step": 3211, + "time_per_iteration": 3.6852128505706787 + }, + { + "auxiliary_loss_clip": 0.0115445, + "auxiliary_loss_mlp": 0.0103289, + "balance_loss_clip": 1.04906201, + "balance_loss_mlp": 1.02395582, + "epoch": 0.38622016473276016, + "flos": 24536720223360.0, + "grad_norm": 1.4056269493225495, + "language_loss": 0.7887553, + "learning_rate": 2.8102934605611513e-06, + "loss": 0.81062877, + "num_input_tokens_seen": 69156460, + "step": 3212, + "time_per_iteration": 3.8079135417938232 + }, + { + "auxiliary_loss_clip": 0.01180986, + "auxiliary_loss_mlp": 0.01031711, + "balance_loss_clip": 1.05636001, + "balance_loss_mlp": 1.02231169, + "epoch": 0.3863404076233993, + "flos": 20558212986240.0, + "grad_norm": 1.9198321019745508, + "language_loss": 0.67092991, + "learning_rate": 2.8095812229846665e-06, + "loss": 0.69305694, + "num_input_tokens_seen": 69176420, + "step": 3213, + "time_per_iteration": 3.7953577041625977 + }, + { + "auxiliary_loss_clip": 0.01174271, + "auxiliary_loss_mlp": 0.01027197, + "balance_loss_clip": 1.05288231, + "balance_loss_mlp": 1.01764202, + "epoch": 0.3864606505140384, + "flos": 22346277039360.0, + "grad_norm": 3.070085082094941, + "language_loss": 0.69038814, + "learning_rate": 2.808868862595355e-06, + "loss": 0.71240282, + "num_input_tokens_seen": 69196665, + "step": 3214, + "time_per_iteration": 2.7724461555480957 + }, + { + "auxiliary_loss_clip": 0.01184882, + "auxiliary_loss_mlp": 0.01028711, + "balance_loss_clip": 1.05422127, + "balance_loss_mlp": 1.02004457, + "epoch": 0.38658089340467744, + "flos": 25703601448320.0, + "grad_norm": 1.7971639777608113, + "language_loss": 0.79467261, + "learning_rate": 2.8081563795012795e-06, + "loss": 0.81680858, + "num_input_tokens_seen": 69216290, + "step": 3215, + "time_per_iteration": 3.71067476272583 + }, + { + "auxiliary_loss_clip": 0.01179813, + "auxiliary_loss_mlp": 0.01028828, + "balance_loss_clip": 1.05239654, + "balance_loss_mlp": 1.01981544, + "epoch": 0.38670113629531655, + "flos": 33802534558080.0, + "grad_norm": 1.7626055747818552, + "language_loss": 0.73933667, + "learning_rate": 2.807443773810524e-06, + "loss": 0.76142305, + "num_input_tokens_seen": 69237550, + "step": 3216, + "time_per_iteration": 2.9367499351501465 + }, + { + "auxiliary_loss_clip": 0.01164747, + "auxiliary_loss_mlp": 0.01032325, + "balance_loss_clip": 1.05363214, + "balance_loss_mlp": 1.02319348, + "epoch": 0.3868213791859556, + "flos": 23331522165120.0, + "grad_norm": 2.3662771454328717, + "language_loss": 0.89367723, + "learning_rate": 2.80673104563119e-06, + "loss": 0.91564798, + "num_input_tokens_seen": 69258175, + "step": 3217, + "time_per_iteration": 2.9513306617736816 + }, + { + "auxiliary_loss_clip": 0.0117921, + "auxiliary_loss_mlp": 0.01026525, + "balance_loss_clip": 1.05421877, + "balance_loss_mlp": 1.01818633, + "epoch": 0.3869416220765947, + "flos": 18441530380800.0, + "grad_norm": 2.6879099368517543, + "language_loss": 0.78830004, + "learning_rate": 2.8060181950713976e-06, + "loss": 0.81035739, + "num_input_tokens_seen": 69274965, + "step": 3218, + "time_per_iteration": 2.7653496265411377 + }, + { + "auxiliary_loss_clip": 0.0115918, + "auxiliary_loss_mlp": 0.01030315, + "balance_loss_clip": 1.0487951, + "balance_loss_mlp": 1.02033734, + "epoch": 0.3870618649672338, + "flos": 15632992938240.0, + "grad_norm": 2.035007129503408, + "language_loss": 0.8094722, + "learning_rate": 2.805305222239286e-06, + "loss": 0.83136714, + "num_input_tokens_seen": 69292220, + "step": 3219, + "time_per_iteration": 2.8497884273529053 + }, + { + "auxiliary_loss_clip": 0.01167484, + "auxiliary_loss_mlp": 0.0103006, + "balance_loss_clip": 1.04944229, + "balance_loss_mlp": 1.02096415, + "epoch": 0.3871821078578729, + "flos": 23513804709120.0, + "grad_norm": 2.027648981305526, + "language_loss": 0.73775959, + "learning_rate": 2.8045921272430118e-06, + "loss": 0.75973499, + "num_input_tokens_seen": 69311900, + "step": 3220, + "time_per_iteration": 2.8763530254364014 + }, + { + "auxiliary_loss_clip": 0.01187359, + "auxiliary_loss_mlp": 0.01029131, + "balance_loss_clip": 1.05322683, + "balance_loss_mlp": 1.01991582, + "epoch": 0.387302350748512, + "flos": 17778259791360.0, + "grad_norm": 3.4925835550486966, + "language_loss": 0.77054125, + "learning_rate": 2.803878910190753e-06, + "loss": 0.79270613, + "num_input_tokens_seen": 69328820, + "step": 3221, + "time_per_iteration": 2.7377614974975586 + }, + { + "auxiliary_loss_clip": 0.01188133, + "auxiliary_loss_mlp": 0.01037894, + "balance_loss_clip": 1.05479717, + "balance_loss_mlp": 1.02915049, + "epoch": 0.3874225936391511, + "flos": 11503409097600.0, + "grad_norm": 2.610379570303601, + "language_loss": 0.81920743, + "learning_rate": 2.8031655711907017e-06, + "loss": 0.84146774, + "num_input_tokens_seen": 69342525, + "step": 3222, + "time_per_iteration": 2.6502974033355713 + }, + { + "auxiliary_loss_clip": 0.01184423, + "auxiliary_loss_mlp": 0.01029872, + "balance_loss_clip": 1.05517817, + "balance_loss_mlp": 1.02065086, + "epoch": 0.38754283652979016, + "flos": 21945154884480.0, + "grad_norm": 2.31350058557062, + "language_loss": 0.80447096, + "learning_rate": 2.8024521103510723e-06, + "loss": 0.8266139, + "num_input_tokens_seen": 69359295, + "step": 3223, + "time_per_iteration": 2.7544445991516113 + }, + { + "auxiliary_loss_clip": 0.01181451, + "auxiliary_loss_mlp": 0.01030898, + "balance_loss_clip": 1.05083179, + "balance_loss_mlp": 1.02175462, + "epoch": 0.38766307942042927, + "flos": 21175984022400.0, + "grad_norm": 2.1459258492986777, + "language_loss": 0.75100887, + "learning_rate": 2.8017385277800952e-06, + "loss": 0.77313232, + "num_input_tokens_seen": 69377650, + "step": 3224, + "time_per_iteration": 2.7013099193573 + }, + { + "auxiliary_loss_clip": 0.01169762, + "auxiliary_loss_mlp": 0.01035507, + "balance_loss_clip": 1.05284202, + "balance_loss_mlp": 1.02581513, + "epoch": 0.3877833223110684, + "flos": 27417294391680.0, + "grad_norm": 2.0496595706197835, + "language_loss": 0.74666291, + "learning_rate": 2.8010248235860213e-06, + "loss": 0.76871562, + "num_input_tokens_seen": 69397765, + "step": 3225, + "time_per_iteration": 2.9356706142425537 + }, + { + "auxiliary_loss_clip": 0.0107714, + "auxiliary_loss_mlp": 0.00876649, + "balance_loss_clip": 1.02112889, + "balance_loss_mlp": 1.00100923, + "epoch": 0.38790356520170743, + "flos": 64500019879680.0, + "grad_norm": 0.830476879603956, + "language_loss": 0.62853032, + "learning_rate": 2.8003109978771192e-06, + "loss": 0.64806819, + "num_input_tokens_seen": 69458930, + "step": 3226, + "time_per_iteration": 3.407334327697754 + }, + { + "auxiliary_loss_clip": 0.01152672, + "auxiliary_loss_mlp": 0.01030141, + "balance_loss_clip": 1.04613233, + "balance_loss_mlp": 1.02118826, + "epoch": 0.38802380809234654, + "flos": 22345415112960.0, + "grad_norm": 1.9702187286483133, + "language_loss": 0.78873527, + "learning_rate": 2.799597050761674e-06, + "loss": 0.81056345, + "num_input_tokens_seen": 69475135, + "step": 3227, + "time_per_iteration": 2.8147754669189453 + }, + { + "auxiliary_loss_clip": 0.01192282, + "auxiliary_loss_mlp": 0.01030825, + "balance_loss_clip": 1.0551914, + "balance_loss_mlp": 1.0205493, + "epoch": 0.38814405098298566, + "flos": 25261361199360.0, + "grad_norm": 1.98581118402129, + "language_loss": 0.79310298, + "learning_rate": 2.7988829823479924e-06, + "loss": 0.81533408, + "num_input_tokens_seen": 69493525, + "step": 3228, + "time_per_iteration": 2.901244640350342 + }, + { + "auxiliary_loss_clip": 0.01168102, + "auxiliary_loss_mlp": 0.01034461, + "balance_loss_clip": 1.05149794, + "balance_loss_mlp": 1.02594924, + "epoch": 0.3882642938736247, + "flos": 18841180078080.0, + "grad_norm": 2.0177708559488954, + "language_loss": 0.64145815, + "learning_rate": 2.7981687927443976e-06, + "loss": 0.6634838, + "num_input_tokens_seen": 69510325, + "step": 3229, + "time_per_iteration": 2.70526385307312 + }, + { + "auxiliary_loss_clip": 0.01181739, + "auxiliary_loss_mlp": 0.01032572, + "balance_loss_clip": 1.05073762, + "balance_loss_mlp": 1.02404213, + "epoch": 0.3883845367642638, + "flos": 21652806090240.0, + "grad_norm": 1.7720557673657988, + "language_loss": 0.85749316, + "learning_rate": 2.797454482059231e-06, + "loss": 0.87963629, + "num_input_tokens_seen": 69530480, + "step": 3230, + "time_per_iteration": 2.7428793907165527 + }, + { + "auxiliary_loss_clip": 0.01194996, + "auxiliary_loss_mlp": 0.01026651, + "balance_loss_clip": 1.05638409, + "balance_loss_mlp": 1.01676202, + "epoch": 0.3885047796549029, + "flos": 20557530627840.0, + "grad_norm": 1.6355718323991888, + "language_loss": 0.84469783, + "learning_rate": 2.7967400504008537e-06, + "loss": 0.86691427, + "num_input_tokens_seen": 69549780, + "step": 3231, + "time_per_iteration": 2.687309503555298 + }, + { + "auxiliary_loss_clip": 0.01055509, + "auxiliary_loss_mlp": 0.01005267, + "balance_loss_clip": 1.0179112, + "balance_loss_mlp": 1.00272739, + "epoch": 0.388625022545542, + "flos": 64325491695360.0, + "grad_norm": 0.8047075874754216, + "language_loss": 0.57474685, + "learning_rate": 2.7960254978776456e-06, + "loss": 0.59535462, + "num_input_tokens_seen": 69611870, + "step": 3232, + "time_per_iteration": 3.496769428253174 + }, + { + "auxiliary_loss_clip": 0.01193652, + "auxiliary_loss_mlp": 0.01035926, + "balance_loss_clip": 1.05444002, + "balance_loss_mlp": 1.02646673, + "epoch": 0.3887452654361811, + "flos": 18113881495680.0, + "grad_norm": 2.0112245284195254, + "language_loss": 0.81530297, + "learning_rate": 2.7953108245980006e-06, + "loss": 0.8375988, + "num_input_tokens_seen": 69630385, + "step": 3233, + "time_per_iteration": 2.647439956665039 + }, + { + "auxiliary_loss_clip": 0.01169329, + "auxiliary_loss_mlp": 0.01031842, + "balance_loss_clip": 1.05300474, + "balance_loss_mlp": 1.02249002, + "epoch": 0.38886550832682015, + "flos": 24975261371520.0, + "grad_norm": 1.4934687868490844, + "language_loss": 0.7368387, + "learning_rate": 2.7945960306703365e-06, + "loss": 0.7588504, + "num_input_tokens_seen": 69653370, + "step": 3234, + "time_per_iteration": 2.752582550048828 + }, + { + "auxiliary_loss_clip": 0.01187337, + "auxiliary_loss_mlp": 0.01032742, + "balance_loss_clip": 1.05430424, + "balance_loss_mlp": 1.02383733, + "epoch": 0.38898575121745926, + "flos": 27199496275200.0, + "grad_norm": 1.8110961109773138, + "language_loss": 0.65719342, + "learning_rate": 2.7938811162030865e-06, + "loss": 0.67939425, + "num_input_tokens_seen": 69673635, + "step": 3235, + "time_per_iteration": 2.618619918823242 + }, + { + "auxiliary_loss_clip": 0.0118111, + "auxiliary_loss_mlp": 0.0103117, + "balance_loss_clip": 1.05421972, + "balance_loss_mlp": 1.02216995, + "epoch": 0.3891059941080984, + "flos": 28763728727040.0, + "grad_norm": 1.8598362067472332, + "language_loss": 0.82597923, + "learning_rate": 2.793166081304702e-06, + "loss": 0.84810209, + "num_input_tokens_seen": 69694130, + "step": 3236, + "time_per_iteration": 2.6342012882232666 + }, + { + "auxiliary_loss_clip": 0.01170517, + "auxiliary_loss_mlp": 0.01031453, + "balance_loss_clip": 1.05157566, + "balance_loss_mlp": 1.02219057, + "epoch": 0.38922623699873743, + "flos": 22893447893760.0, + "grad_norm": 1.9219487093175407, + "language_loss": 0.8215062, + "learning_rate": 2.7924509260836543e-06, + "loss": 0.84352589, + "num_input_tokens_seen": 69713255, + "step": 3237, + "time_per_iteration": 3.5322227478027344 + }, + { + "auxiliary_loss_clip": 0.01161731, + "auxiliary_loss_mlp": 0.01031283, + "balance_loss_clip": 1.04980981, + "balance_loss_mlp": 1.02167416, + "epoch": 0.38934647988937654, + "flos": 19792418002560.0, + "grad_norm": 1.9076650139458164, + "language_loss": 0.68559909, + "learning_rate": 2.791735650648431e-06, + "loss": 0.70752925, + "num_input_tokens_seen": 69732375, + "step": 3238, + "time_per_iteration": 3.3570375442504883 + }, + { + "auxiliary_loss_clip": 0.01169731, + "auxiliary_loss_mlp": 0.01028045, + "balance_loss_clip": 1.05296338, + "balance_loss_mlp": 1.01936102, + "epoch": 0.38946672278001565, + "flos": 19202081978880.0, + "grad_norm": 1.7951860797085777, + "language_loss": 0.74361587, + "learning_rate": 2.791020255107538e-06, + "loss": 0.76559359, + "num_input_tokens_seen": 69749745, + "step": 3239, + "time_per_iteration": 3.451078176498413 + }, + { + "auxiliary_loss_clip": 0.01158565, + "auxiliary_loss_mlp": 0.01032775, + "balance_loss_clip": 1.04739952, + "balance_loss_mlp": 1.02435303, + "epoch": 0.3895869656706547, + "flos": 24936477661440.0, + "grad_norm": 1.5870686969893864, + "language_loss": 0.80860734, + "learning_rate": 2.7903047395695023e-06, + "loss": 0.83052075, + "num_input_tokens_seen": 69769645, + "step": 3240, + "time_per_iteration": 2.82265567779541 + }, + { + "auxiliary_loss_clip": 0.01184039, + "auxiliary_loss_mlp": 0.00887103, + "balance_loss_clip": 1.05669785, + "balance_loss_mlp": 1.00061953, + "epoch": 0.3897072085612938, + "flos": 24133622820480.0, + "grad_norm": 2.26391240908342, + "language_loss": 0.89879274, + "learning_rate": 2.789589104142865e-06, + "loss": 0.91950417, + "num_input_tokens_seen": 69787270, + "step": 3241, + "time_per_iteration": 3.7235796451568604 + }, + { + "auxiliary_loss_clip": 0.01164059, + "auxiliary_loss_mlp": 0.01031921, + "balance_loss_clip": 1.05120921, + "balance_loss_mlp": 1.02240765, + "epoch": 0.3898274514519329, + "flos": 17166342672000.0, + "grad_norm": 3.7277782305573046, + "language_loss": 0.76312447, + "learning_rate": 2.7888733489361895e-06, + "loss": 0.78508425, + "num_input_tokens_seen": 69805685, + "step": 3242, + "time_per_iteration": 2.889133930206299 + }, + { + "auxiliary_loss_clip": 0.01085731, + "auxiliary_loss_mlp": 0.01008095, + "balance_loss_clip": 1.02122319, + "balance_loss_mlp": 1.00584149, + "epoch": 0.389947694342572, + "flos": 66074807952000.0, + "grad_norm": 0.727609910936887, + "language_loss": 0.5877586, + "learning_rate": 2.788157474058054e-06, + "loss": 0.60869682, + "num_input_tokens_seen": 69867960, + "step": 3243, + "time_per_iteration": 3.4901580810546875 + }, + { + "auxiliary_loss_clip": 0.01188343, + "auxiliary_loss_mlp": 0.01030257, + "balance_loss_clip": 1.05304265, + "balance_loss_mlp": 1.02112567, + "epoch": 0.3900679372332111, + "flos": 25740912700800.0, + "grad_norm": 1.4954387296836933, + "language_loss": 0.69769335, + "learning_rate": 2.7874414796170555e-06, + "loss": 0.71987939, + "num_input_tokens_seen": 69889450, + "step": 3244, + "time_per_iteration": 2.7226550579071045 + }, + { + "auxiliary_loss_clip": 0.01180793, + "auxiliary_loss_mlp": 0.01035731, + "balance_loss_clip": 1.05362368, + "balance_loss_mlp": 1.02588415, + "epoch": 0.3901881801238502, + "flos": 11801611808640.0, + "grad_norm": 2.442515747271709, + "language_loss": 0.83983052, + "learning_rate": 2.7867253657218113e-06, + "loss": 0.86199582, + "num_input_tokens_seen": 69903340, + "step": 3245, + "time_per_iteration": 2.624901533126831 + }, + { + "auxiliary_loss_clip": 0.01169978, + "auxiliary_loss_mlp": 0.00887353, + "balance_loss_clip": 1.05010462, + "balance_loss_mlp": 1.00049114, + "epoch": 0.39030842301448926, + "flos": 27308951994240.0, + "grad_norm": 1.728203448170267, + "language_loss": 0.73002815, + "learning_rate": 2.7860091324809544e-06, + "loss": 0.75060141, + "num_input_tokens_seen": 69924400, + "step": 3246, + "time_per_iteration": 2.711000442504883 + }, + { + "auxiliary_loss_clip": 0.01181677, + "auxiliary_loss_mlp": 0.01033171, + "balance_loss_clip": 1.05648673, + "balance_loss_mlp": 1.02483261, + "epoch": 0.39042866590512837, + "flos": 27163334257920.0, + "grad_norm": 1.7005031524125684, + "language_loss": 0.81392682, + "learning_rate": 2.7852927800031377e-06, + "loss": 0.83607537, + "num_input_tokens_seen": 69944565, + "step": 3247, + "time_per_iteration": 2.773606300354004 + }, + { + "auxiliary_loss_clip": 0.01175152, + "auxiliary_loss_mlp": 0.01038312, + "balance_loss_clip": 1.0520643, + "balance_loss_mlp": 1.02968121, + "epoch": 0.3905489087957674, + "flos": 29716115886720.0, + "grad_norm": 2.3626425862521114, + "language_loss": 0.82643962, + "learning_rate": 2.7845763083970298e-06, + "loss": 0.84857422, + "num_input_tokens_seen": 69964965, + "step": 3248, + "time_per_iteration": 2.7770891189575195 + }, + { + "auxiliary_loss_clip": 0.01175517, + "auxiliary_loss_mlp": 0.01030033, + "balance_loss_clip": 1.05153155, + "balance_loss_mlp": 1.0208658, + "epoch": 0.39066915168640653, + "flos": 24498618871680.0, + "grad_norm": 1.8601587047462147, + "language_loss": 0.82395065, + "learning_rate": 2.7838597177713205e-06, + "loss": 0.84600616, + "num_input_tokens_seen": 69986055, + "step": 3249, + "time_per_iteration": 2.805469274520874 + }, + { + "auxiliary_loss_clip": 0.01133893, + "auxiliary_loss_mlp": 0.01037657, + "balance_loss_clip": 1.04903436, + "balance_loss_mlp": 1.02883518, + "epoch": 0.39078939457704565, + "flos": 20558572122240.0, + "grad_norm": 1.6599131662745656, + "language_loss": 0.73468792, + "learning_rate": 2.7831430082347143e-06, + "loss": 0.75640345, + "num_input_tokens_seen": 70005260, + "step": 3250, + "time_per_iteration": 2.898655891418457 + }, + { + "auxiliary_loss_clip": 0.01182579, + "auxiliary_loss_mlp": 0.00886585, + "balance_loss_clip": 1.05368316, + "balance_loss_mlp": 1.00060534, + "epoch": 0.3909096374676847, + "flos": 22783417557120.0, + "grad_norm": 2.457567547208366, + "language_loss": 0.82174641, + "learning_rate": 2.7824261798959373e-06, + "loss": 0.84243804, + "num_input_tokens_seen": 70023440, + "step": 3251, + "time_per_iteration": 2.6749846935272217 + }, + { + "auxiliary_loss_clip": 0.01173377, + "auxiliary_loss_mlp": 0.01036639, + "balance_loss_clip": 1.04839969, + "balance_loss_mlp": 1.02705431, + "epoch": 0.3910298803583238, + "flos": 23003119094400.0, + "grad_norm": 1.7616383324990625, + "language_loss": 0.79405421, + "learning_rate": 2.78170923286373e-06, + "loss": 0.81615436, + "num_input_tokens_seen": 70043040, + "step": 3252, + "time_per_iteration": 2.7481586933135986 + }, + { + "auxiliary_loss_clip": 0.01127087, + "auxiliary_loss_mlp": 0.01036044, + "balance_loss_clip": 1.04656231, + "balance_loss_mlp": 1.02586961, + "epoch": 0.3911501232489629, + "flos": 24316264500480.0, + "grad_norm": 3.9807470204005724, + "language_loss": 0.84150863, + "learning_rate": 2.780992167246854e-06, + "loss": 0.86313987, + "num_input_tokens_seen": 70060565, + "step": 3253, + "time_per_iteration": 2.9583113193511963 + }, + { + "auxiliary_loss_clip": 0.01076997, + "auxiliary_loss_mlp": 0.01001823, + "balance_loss_clip": 1.020082, + "balance_loss_mlp": 0.99958152, + "epoch": 0.391270366139602, + "flos": 60869054684160.0, + "grad_norm": 0.9878863301008038, + "language_loss": 0.72154838, + "learning_rate": 2.7802749831540883e-06, + "loss": 0.74233657, + "num_input_tokens_seen": 70119465, + "step": 3254, + "time_per_iteration": 3.5588858127593994 + }, + { + "auxiliary_loss_clip": 0.01156593, + "auxiliary_loss_mlp": 0.01034792, + "balance_loss_clip": 1.04999828, + "balance_loss_mlp": 1.02682877, + "epoch": 0.3913906090302411, + "flos": 21543494025600.0, + "grad_norm": 2.113593902034873, + "language_loss": 0.81639934, + "learning_rate": 2.7795576806942268e-06, + "loss": 0.83831322, + "num_input_tokens_seen": 70138270, + "step": 3255, + "time_per_iteration": 2.81564998626709 + }, + { + "auxiliary_loss_clip": 0.01064608, + "auxiliary_loss_mlp": 0.01004972, + "balance_loss_clip": 1.02054954, + "balance_loss_mlp": 1.0026598, + "epoch": 0.3915108519208802, + "flos": 49839953702400.0, + "grad_norm": 0.7583172134104122, + "language_loss": 0.54876858, + "learning_rate": 2.778840259976085e-06, + "loss": 0.56946439, + "num_input_tokens_seen": 70193500, + "step": 3256, + "time_per_iteration": 3.3174078464508057 + }, + { + "auxiliary_loss_clip": 0.01187406, + "auxiliary_loss_mlp": 0.0103432, + "balance_loss_clip": 1.05603969, + "balance_loss_mlp": 1.02474177, + "epoch": 0.39163109481151925, + "flos": 16506447960960.0, + "grad_norm": 2.051912478237572, + "language_loss": 0.76650739, + "learning_rate": 2.778122721108495e-06, + "loss": 0.78872466, + "num_input_tokens_seen": 70211730, + "step": 3257, + "time_per_iteration": 2.740621328353882 + }, + { + "auxiliary_loss_clip": 0.0118006, + "auxiliary_loss_mlp": 0.01030354, + "balance_loss_clip": 1.05389333, + "balance_loss_mlp": 1.02226615, + "epoch": 0.39175133770215836, + "flos": 26067484177920.0, + "grad_norm": 1.9659483044953372, + "language_loss": 0.88243592, + "learning_rate": 2.7774050642003076e-06, + "loss": 0.90454006, + "num_input_tokens_seen": 70232540, + "step": 3258, + "time_per_iteration": 2.7803401947021484 + }, + { + "auxiliary_loss_clip": 0.01196185, + "auxiliary_loss_mlp": 0.0103177, + "balance_loss_clip": 1.0564642, + "balance_loss_mlp": 1.02298391, + "epoch": 0.3918715805927975, + "flos": 21872076664320.0, + "grad_norm": 2.5027450277422005, + "language_loss": 0.9344542, + "learning_rate": 2.7766872893603896e-06, + "loss": 0.9567337, + "num_input_tokens_seen": 70252515, + "step": 3259, + "time_per_iteration": 2.785273790359497 + }, + { + "auxiliary_loss_clip": 0.01183223, + "auxiliary_loss_mlp": 0.01036473, + "balance_loss_clip": 1.05393362, + "balance_loss_mlp": 1.02735317, + "epoch": 0.39199182348343653, + "flos": 20376181837440.0, + "grad_norm": 1.6657398249105428, + "language_loss": 0.73310447, + "learning_rate": 2.7759693966976275e-06, + "loss": 0.75530148, + "num_input_tokens_seen": 70271020, + "step": 3260, + "time_per_iteration": 2.752459764480591 + }, + { + "auxiliary_loss_clip": 0.01161653, + "auxiliary_loss_mlp": 0.010369, + "balance_loss_clip": 1.04846394, + "balance_loss_mlp": 1.02723169, + "epoch": 0.39211206637407564, + "flos": 21683545153920.0, + "grad_norm": 1.8291657559224388, + "language_loss": 0.85281146, + "learning_rate": 2.7752513863209242e-06, + "loss": 0.87479699, + "num_input_tokens_seen": 70289600, + "step": 3261, + "time_per_iteration": 2.9565117359161377 + }, + { + "auxiliary_loss_clip": 0.01168173, + "auxiliary_loss_mlp": 0.00886055, + "balance_loss_clip": 1.05314088, + "balance_loss_mlp": 1.00058985, + "epoch": 0.39223230926471475, + "flos": 21066276908160.0, + "grad_norm": 1.766704804214104, + "language_loss": 0.84232402, + "learning_rate": 2.774533258339203e-06, + "loss": 0.86286628, + "num_input_tokens_seen": 70307060, + "step": 3262, + "time_per_iteration": 2.751420736312866 + }, + { + "auxiliary_loss_clip": 0.01150399, + "auxiliary_loss_mlp": 0.01033912, + "balance_loss_clip": 1.04611385, + "balance_loss_mlp": 1.0247035, + "epoch": 0.3923525521553538, + "flos": 17603016312960.0, + "grad_norm": 2.074567871099886, + "language_loss": 0.79656625, + "learning_rate": 2.7738150128614014e-06, + "loss": 0.81840932, + "num_input_tokens_seen": 70324465, + "step": 3263, + "time_per_iteration": 3.784501075744629 + }, + { + "auxiliary_loss_clip": 0.01150475, + "auxiliary_loss_mlp": 0.01033921, + "balance_loss_clip": 1.0474447, + "balance_loss_mlp": 1.02409256, + "epoch": 0.3924727950459929, + "flos": 20558284813440.0, + "grad_norm": 1.86310547455492, + "language_loss": 0.89713931, + "learning_rate": 2.7730966499964777e-06, + "loss": 0.91898328, + "num_input_tokens_seen": 70341415, + "step": 3264, + "time_per_iteration": 3.6838927268981934 + }, + { + "auxiliary_loss_clip": 0.01192505, + "auxiliary_loss_mlp": 0.01033581, + "balance_loss_clip": 1.05349159, + "balance_loss_mlp": 1.02385974, + "epoch": 0.39259303793663197, + "flos": 16216110328320.0, + "grad_norm": 2.43862737037489, + "language_loss": 0.80764008, + "learning_rate": 2.772378169853408e-06, + "loss": 0.82990092, + "num_input_tokens_seen": 70358985, + "step": 3265, + "time_per_iteration": 3.510918617248535 + }, + { + "auxiliary_loss_clip": 0.01164392, + "auxiliary_loss_mlp": 0.01031642, + "balance_loss_clip": 1.05199504, + "balance_loss_mlp": 1.02286792, + "epoch": 0.3927132808272711, + "flos": 16797001075200.0, + "grad_norm": 1.774406321538981, + "language_loss": 0.74031806, + "learning_rate": 2.771659572541183e-06, + "loss": 0.76227838, + "num_input_tokens_seen": 70376915, + "step": 3266, + "time_per_iteration": 2.750753402709961 + }, + { + "auxiliary_loss_clip": 0.01185837, + "auxiliary_loss_mlp": 0.01034398, + "balance_loss_clip": 1.05674863, + "balance_loss_mlp": 1.02571952, + "epoch": 0.3928335237179102, + "flos": 20267228908800.0, + "grad_norm": 2.1597808781796113, + "language_loss": 0.86505514, + "learning_rate": 2.7709408581688143e-06, + "loss": 0.88725752, + "num_input_tokens_seen": 70396900, + "step": 3267, + "time_per_iteration": 3.7087910175323486 + }, + { + "auxiliary_loss_clip": 0.01166668, + "auxiliary_loss_mlp": 0.01033288, + "balance_loss_clip": 1.05186081, + "balance_loss_mlp": 1.02402508, + "epoch": 0.39295376660854925, + "flos": 24973250209920.0, + "grad_norm": 1.6173342649179359, + "language_loss": 0.87710017, + "learning_rate": 2.7702220268453307e-06, + "loss": 0.89909971, + "num_input_tokens_seen": 70417260, + "step": 3268, + "time_per_iteration": 2.909445285797119 + }, + { + "auxiliary_loss_clip": 0.0117888, + "auxiliary_loss_mlp": 0.01034745, + "balance_loss_clip": 1.05350041, + "balance_loss_mlp": 1.02521455, + "epoch": 0.39307400949918836, + "flos": 18697788984960.0, + "grad_norm": 2.1193642346384913, + "language_loss": 0.845155, + "learning_rate": 2.7695030786797785e-06, + "loss": 0.86729127, + "num_input_tokens_seen": 70433155, + "step": 3269, + "time_per_iteration": 2.723356246948242 + }, + { + "auxiliary_loss_clip": 0.01149449, + "auxiliary_loss_mlp": 0.0103112, + "balance_loss_clip": 1.04826248, + "balance_loss_mlp": 1.02199435, + "epoch": 0.39319425238982747, + "flos": 22415476590720.0, + "grad_norm": 2.184217248303627, + "language_loss": 0.74894553, + "learning_rate": 2.7687840137812206e-06, + "loss": 0.77075124, + "num_input_tokens_seen": 70451240, + "step": 3270, + "time_per_iteration": 2.8683080673217773 + }, + { + "auxiliary_loss_clip": 0.01070517, + "auxiliary_loss_mlp": 0.0100422, + "balance_loss_clip": 1.01579452, + "balance_loss_mlp": 1.00207448, + "epoch": 0.3933144952804665, + "flos": 66192954762240.0, + "grad_norm": 0.7957944034355638, + "language_loss": 0.62101448, + "learning_rate": 2.7680648322587395e-06, + "loss": 0.64176184, + "num_input_tokens_seen": 70516115, + "step": 3271, + "time_per_iteration": 3.4101157188415527 + }, + { + "auxiliary_loss_clip": 0.01194163, + "auxiliary_loss_mlp": 0.01035501, + "balance_loss_clip": 1.05741727, + "balance_loss_mlp": 1.02656043, + "epoch": 0.39343473817110564, + "flos": 15487159720320.0, + "grad_norm": 2.1973401196513507, + "language_loss": 0.81213152, + "learning_rate": 2.7673455342214334e-06, + "loss": 0.83442813, + "num_input_tokens_seen": 70533105, + "step": 3272, + "time_per_iteration": 2.716644048690796 + }, + { + "auxiliary_loss_clip": 0.0118546, + "auxiliary_loss_mlp": 0.01031001, + "balance_loss_clip": 1.05637407, + "balance_loss_mlp": 1.02211452, + "epoch": 0.39355498106174475, + "flos": 21324905809920.0, + "grad_norm": 1.9568562832951895, + "language_loss": 0.75609863, + "learning_rate": 2.7666261197784198e-06, + "loss": 0.77826321, + "num_input_tokens_seen": 70551920, + "step": 3273, + "time_per_iteration": 2.886422634124756 + }, + { + "auxiliary_loss_clip": 0.01171753, + "auxiliary_loss_mlp": 0.01033493, + "balance_loss_clip": 1.05519152, + "balance_loss_mlp": 1.02535391, + "epoch": 0.3936752239523838, + "flos": 13296357400320.0, + "grad_norm": 2.0684957353570357, + "language_loss": 0.76754606, + "learning_rate": 2.7659065890388336e-06, + "loss": 0.78959858, + "num_input_tokens_seen": 70567920, + "step": 3274, + "time_per_iteration": 2.7090229988098145 + }, + { + "auxiliary_loss_clip": 0.01171076, + "auxiliary_loss_mlp": 0.01037806, + "balance_loss_clip": 1.04937887, + "balance_loss_mlp": 1.02898502, + "epoch": 0.3937954668430229, + "flos": 16800161472000.0, + "grad_norm": 1.8025281220629545, + "language_loss": 0.85335404, + "learning_rate": 2.7651869421118266e-06, + "loss": 0.87544286, + "num_input_tokens_seen": 70584530, + "step": 3275, + "time_per_iteration": 2.8285694122314453 + }, + { + "auxiliary_loss_clip": 0.01188356, + "auxiliary_loss_mlp": 0.01027136, + "balance_loss_clip": 1.05854011, + "balance_loss_mlp": 1.01793289, + "epoch": 0.393915709733662, + "flos": 21064229832960.0, + "grad_norm": 1.6485350599193878, + "language_loss": 0.83038831, + "learning_rate": 2.76446717910657e-06, + "loss": 0.85254318, + "num_input_tokens_seen": 70605235, + "step": 3276, + "time_per_iteration": 2.6937663555145264 + }, + { + "auxiliary_loss_clip": 0.01178586, + "auxiliary_loss_mlp": 0.01034277, + "balance_loss_clip": 1.05242991, + "balance_loss_mlp": 1.02556884, + "epoch": 0.3940359526243011, + "flos": 17165265264000.0, + "grad_norm": 2.4262824966036454, + "language_loss": 0.77374732, + "learning_rate": 2.763747300132249e-06, + "loss": 0.79587597, + "num_input_tokens_seen": 70622675, + "step": 3277, + "time_per_iteration": 2.720689058303833 + }, + { + "auxiliary_loss_clip": 0.01192626, + "auxiliary_loss_mlp": 0.01034938, + "balance_loss_clip": 1.05611825, + "balance_loss_mlp": 1.02572894, + "epoch": 0.3941561955149402, + "flos": 20995856294400.0, + "grad_norm": 1.6765504893773202, + "language_loss": 0.86477655, + "learning_rate": 2.7630273052980704e-06, + "loss": 0.88705218, + "num_input_tokens_seen": 70643265, + "step": 3278, + "time_per_iteration": 2.680241584777832 + }, + { + "auxiliary_loss_clip": 0.01160895, + "auxiliary_loss_mlp": 0.01032667, + "balance_loss_clip": 1.05026698, + "balance_loss_mlp": 1.02403045, + "epoch": 0.39427643840557924, + "flos": 18843406721280.0, + "grad_norm": 2.3676385294624267, + "language_loss": 0.6702435, + "learning_rate": 2.7623071947132554e-06, + "loss": 0.69217908, + "num_input_tokens_seen": 70660295, + "step": 3279, + "time_per_iteration": 2.749743700027466 + }, + { + "auxiliary_loss_clip": 0.01177337, + "auxiliary_loss_mlp": 0.01039235, + "balance_loss_clip": 1.05118072, + "balance_loss_mlp": 1.02981174, + "epoch": 0.39439668129621835, + "flos": 23258659426560.0, + "grad_norm": 1.901301109461397, + "language_loss": 0.78762054, + "learning_rate": 2.7615869684870458e-06, + "loss": 0.80978626, + "num_input_tokens_seen": 70679605, + "step": 3280, + "time_per_iteration": 2.7544102668762207 + }, + { + "auxiliary_loss_clip": 0.01181811, + "auxiliary_loss_mlp": 0.01030544, + "balance_loss_clip": 1.0558573, + "balance_loss_mlp": 1.02164483, + "epoch": 0.39451692418685746, + "flos": 26652289507200.0, + "grad_norm": 1.6353463331748945, + "language_loss": 0.84599793, + "learning_rate": 2.7608666267286986e-06, + "loss": 0.86812139, + "num_input_tokens_seen": 70699835, + "step": 3281, + "time_per_iteration": 2.7640483379364014 + }, + { + "auxiliary_loss_clip": 0.01139108, + "auxiliary_loss_mlp": 0.01035856, + "balance_loss_clip": 1.04454446, + "balance_loss_mlp": 1.02626538, + "epoch": 0.3946371670774965, + "flos": 18258709132800.0, + "grad_norm": 2.1439921635475954, + "language_loss": 0.86642241, + "learning_rate": 2.760146169547489e-06, + "loss": 0.88817203, + "num_input_tokens_seen": 70716600, + "step": 3282, + "time_per_iteration": 2.9103097915649414 + }, + { + "auxiliary_loss_clip": 0.01178497, + "auxiliary_loss_mlp": 0.01042399, + "balance_loss_clip": 1.05865288, + "balance_loss_mlp": 1.03327918, + "epoch": 0.39475740996813563, + "flos": 24206126423040.0, + "grad_norm": 1.6289473091074533, + "language_loss": 0.76259029, + "learning_rate": 2.75942559705271e-06, + "loss": 0.78479922, + "num_input_tokens_seen": 70736335, + "step": 3283, + "time_per_iteration": 2.8276233673095703 + }, + { + "auxiliary_loss_clip": 0.01181095, + "auxiliary_loss_mlp": 0.01031611, + "balance_loss_clip": 1.05452454, + "balance_loss_mlp": 1.0227536, + "epoch": 0.39487765285877474, + "flos": 19317858491520.0, + "grad_norm": 2.3158060791221993, + "language_loss": 0.89407754, + "learning_rate": 2.7587049093536713e-06, + "loss": 0.91620463, + "num_input_tokens_seen": 70752665, + "step": 3284, + "time_per_iteration": 2.8313119411468506 + }, + { + "auxiliary_loss_clip": 0.01183755, + "auxiliary_loss_mlp": 0.01033713, + "balance_loss_clip": 1.05367827, + "balance_loss_mlp": 1.02535629, + "epoch": 0.3949978957494138, + "flos": 17311744926720.0, + "grad_norm": 2.8935188673635643, + "language_loss": 0.80643588, + "learning_rate": 2.757984106559701e-06, + "loss": 0.82861054, + "num_input_tokens_seen": 70771650, + "step": 3285, + "time_per_iteration": 2.7452924251556396 + }, + { + "auxiliary_loss_clip": 0.01166803, + "auxiliary_loss_mlp": 0.01034463, + "balance_loss_clip": 1.05313563, + "balance_loss_mlp": 1.02556968, + "epoch": 0.3951181386400529, + "flos": 36317861280000.0, + "grad_norm": 2.1694337994762, + "language_loss": 0.71046978, + "learning_rate": 2.7572631887801446e-06, + "loss": 0.73248237, + "num_input_tokens_seen": 70793275, + "step": 3286, + "time_per_iteration": 2.929070472717285 + }, + { + "auxiliary_loss_clip": 0.01181541, + "auxiliary_loss_mlp": 0.0102884, + "balance_loss_clip": 1.05372465, + "balance_loss_mlp": 1.01930332, + "epoch": 0.395238381530692, + "flos": 23110348170240.0, + "grad_norm": 1.868763144887186, + "language_loss": 0.76736915, + "learning_rate": 2.7565421561243654e-06, + "loss": 0.789473, + "num_input_tokens_seen": 70811440, + "step": 3287, + "time_per_iteration": 2.708850860595703 + }, + { + "auxiliary_loss_clip": 0.01154449, + "auxiliary_loss_mlp": 0.01040129, + "balance_loss_clip": 1.04772043, + "balance_loss_mlp": 1.03043127, + "epoch": 0.3953586244213311, + "flos": 24347614095360.0, + "grad_norm": 2.355888553937089, + "language_loss": 0.82304454, + "learning_rate": 2.7558210087017413e-06, + "loss": 0.84499025, + "num_input_tokens_seen": 70831375, + "step": 3288, + "time_per_iteration": 3.011720895767212 + }, + { + "auxiliary_loss_clip": 0.0115599, + "auxiliary_loss_mlp": 0.01031964, + "balance_loss_clip": 1.05330348, + "balance_loss_mlp": 1.02265954, + "epoch": 0.3954788673119702, + "flos": 23440080044160.0, + "grad_norm": 2.1459987465140697, + "language_loss": 0.73364633, + "learning_rate": 2.7550997466216724e-06, + "loss": 0.75552589, + "num_input_tokens_seen": 70849170, + "step": 3289, + "time_per_iteration": 4.8606767654418945 + }, + { + "auxiliary_loss_clip": 0.01169848, + "auxiliary_loss_mlp": 0.0103251, + "balance_loss_clip": 1.05598927, + "balance_loss_mlp": 1.02324712, + "epoch": 0.3955991102026093, + "flos": 17494063384320.0, + "grad_norm": 2.296562272770316, + "language_loss": 0.80958265, + "learning_rate": 2.7543783699935714e-06, + "loss": 0.83160621, + "num_input_tokens_seen": 70867200, + "step": 3290, + "time_per_iteration": 2.8117778301239014 + }, + { + "auxiliary_loss_clip": 0.01179034, + "auxiliary_loss_mlp": 0.01027221, + "balance_loss_clip": 1.05582333, + "balance_loss_mlp": 1.01832843, + "epoch": 0.39571935309324835, + "flos": 18221326053120.0, + "grad_norm": 2.3849884622413953, + "language_loss": 0.85917956, + "learning_rate": 2.753656878926872e-06, + "loss": 0.8812421, + "num_input_tokens_seen": 70883080, + "step": 3291, + "time_per_iteration": 3.7592625617980957 + }, + { + "auxiliary_loss_clip": 0.0116191, + "auxiliary_loss_mlp": 0.01028011, + "balance_loss_clip": 1.04907537, + "balance_loss_mlp": 1.01879025, + "epoch": 0.39583959598388746, + "flos": 17748813617280.0, + "grad_norm": 2.6862722731844366, + "language_loss": 0.74302506, + "learning_rate": 2.752935273531023e-06, + "loss": 0.76492423, + "num_input_tokens_seen": 70901230, + "step": 3292, + "time_per_iteration": 2.882376194000244 + }, + { + "auxiliary_loss_clip": 0.01181794, + "auxiliary_loss_mlp": 0.01028507, + "balance_loss_clip": 1.0536623, + "balance_loss_mlp": 1.018803, + "epoch": 0.39595983887452657, + "flos": 19352368483200.0, + "grad_norm": 1.9495175690030817, + "language_loss": 0.78723538, + "learning_rate": 2.752213553915492e-06, + "loss": 0.80933839, + "num_input_tokens_seen": 70919585, + "step": 3293, + "time_per_iteration": 3.8290138244628906 + }, + { + "auxiliary_loss_clip": 0.01068, + "auxiliary_loss_mlp": 0.01002674, + "balance_loss_clip": 1.02085245, + "balance_loss_mlp": 1.00052834, + "epoch": 0.3960800817651656, + "flos": 60682282940160.0, + "grad_norm": 0.8122381750402315, + "language_loss": 0.66049248, + "learning_rate": 2.751491720189762e-06, + "loss": 0.68119925, + "num_input_tokens_seen": 70977695, + "step": 3294, + "time_per_iteration": 3.3645827770233154 + }, + { + "auxiliary_loss_clip": 0.0117436, + "auxiliary_loss_mlp": 0.0088635, + "balance_loss_clip": 1.05544877, + "balance_loss_mlp": 1.00074339, + "epoch": 0.39620032465580474, + "flos": 16836718538880.0, + "grad_norm": 2.0938202937126253, + "language_loss": 0.92006242, + "learning_rate": 2.7507697724633364e-06, + "loss": 0.94066954, + "num_input_tokens_seen": 70994455, + "step": 3295, + "time_per_iteration": 2.8954079151153564 + }, + { + "auxiliary_loss_clip": 0.01061599, + "auxiliary_loss_mlp": 0.01014519, + "balance_loss_clip": 1.0222218, + "balance_loss_mlp": 1.01182461, + "epoch": 0.3963205675464438, + "flos": 69071445941760.0, + "grad_norm": 0.7787179604996881, + "language_loss": 0.54678476, + "learning_rate": 2.7500477108457327e-06, + "loss": 0.56754589, + "num_input_tokens_seen": 71046465, + "step": 3296, + "time_per_iteration": 3.238037347793579 + }, + { + "auxiliary_loss_clip": 0.01178346, + "auxiliary_loss_mlp": 0.01032005, + "balance_loss_clip": 1.0530144, + "balance_loss_mlp": 1.0236305, + "epoch": 0.3964408104370829, + "flos": 25667439431040.0, + "grad_norm": 1.9444884702074503, + "language_loss": 0.80966687, + "learning_rate": 2.7493255354464877e-06, + "loss": 0.83177042, + "num_input_tokens_seen": 71064275, + "step": 3297, + "time_per_iteration": 2.8550164699554443 + }, + { + "auxiliary_loss_clip": 0.01096091, + "auxiliary_loss_mlp": 0.01029366, + "balance_loss_clip": 1.03794098, + "balance_loss_mlp": 1.02083027, + "epoch": 0.396561053327722, + "flos": 24277480790400.0, + "grad_norm": 1.9103756001648233, + "language_loss": 0.76166552, + "learning_rate": 2.748603246375156e-06, + "loss": 0.78292012, + "num_input_tokens_seen": 71082290, + "step": 3298, + "time_per_iteration": 3.223796844482422 + }, + { + "auxiliary_loss_clip": 0.01191797, + "auxiliary_loss_mlp": 0.01040099, + "balance_loss_clip": 1.05646229, + "balance_loss_mlp": 1.03093743, + "epoch": 0.39668129621836107, + "flos": 20522302364160.0, + "grad_norm": 3.4668350517475393, + "language_loss": 0.69595307, + "learning_rate": 2.7478808437413055e-06, + "loss": 0.71827197, + "num_input_tokens_seen": 71101700, + "step": 3299, + "time_per_iteration": 2.8617782592773438 + }, + { + "auxiliary_loss_clip": 0.01154603, + "auxiliary_loss_mlp": 0.01026351, + "balance_loss_clip": 1.05444121, + "balance_loss_mlp": 1.01749992, + "epoch": 0.3968015391090002, + "flos": 27052585649280.0, + "grad_norm": 1.9893294973418334, + "language_loss": 0.65796649, + "learning_rate": 2.7471583276545263e-06, + "loss": 0.67977601, + "num_input_tokens_seen": 71122360, + "step": 3300, + "time_per_iteration": 2.8600852489471436 + }, + { + "auxiliary_loss_clip": 0.01169731, + "auxiliary_loss_mlp": 0.01025729, + "balance_loss_clip": 1.05046999, + "balance_loss_mlp": 1.01660919, + "epoch": 0.3969217819996393, + "flos": 12531819392640.0, + "grad_norm": 1.936072476523248, + "language_loss": 0.70532811, + "learning_rate": 2.7464356982244224e-06, + "loss": 0.7272827, + "num_input_tokens_seen": 71140360, + "step": 3301, + "time_per_iteration": 2.7506062984466553 + }, + { + "auxiliary_loss_clip": 0.01070382, + "auxiliary_loss_mlp": 0.01003179, + "balance_loss_clip": 1.01942325, + "balance_loss_mlp": 1.00065136, + "epoch": 0.39704202489027834, + "flos": 66241399230720.0, + "grad_norm": 0.7749885647511232, + "language_loss": 0.61736834, + "learning_rate": 2.745712955560617e-06, + "loss": 0.63810396, + "num_input_tokens_seen": 71196565, + "step": 3302, + "time_per_iteration": 3.2498726844787598 + }, + { + "auxiliary_loss_clip": 0.01143514, + "auxiliary_loss_mlp": 0.01037369, + "balance_loss_clip": 1.0509131, + "balance_loss_mlp": 1.02805924, + "epoch": 0.39716226778091746, + "flos": 16982982720000.0, + "grad_norm": 2.3056658624721673, + "language_loss": 0.76750785, + "learning_rate": 2.7449900997727496e-06, + "loss": 0.78931665, + "num_input_tokens_seen": 71214675, + "step": 3303, + "time_per_iteration": 2.975576639175415 + }, + { + "auxiliary_loss_clip": 0.01164902, + "auxiliary_loss_mlp": 0.01029435, + "balance_loss_clip": 1.05271482, + "balance_loss_mlp": 1.02090526, + "epoch": 0.39728251067155657, + "flos": 23477139901440.0, + "grad_norm": 1.7687668137587513, + "language_loss": 0.84258795, + "learning_rate": 2.744267130970476e-06, + "loss": 0.86453128, + "num_input_tokens_seen": 71234400, + "step": 3304, + "time_per_iteration": 2.785675048828125 + }, + { + "auxiliary_loss_clip": 0.01164361, + "auxiliary_loss_mlp": 0.01036626, + "balance_loss_clip": 1.05155957, + "balance_loss_mlp": 1.0276556, + "epoch": 0.3974027535621956, + "flos": 20704441253760.0, + "grad_norm": 1.7925971431807315, + "language_loss": 0.7736721, + "learning_rate": 2.7435440492634697e-06, + "loss": 0.79568195, + "num_input_tokens_seen": 71253725, + "step": 3305, + "time_per_iteration": 2.7789723873138428 + }, + { + "auxiliary_loss_clip": 0.01173091, + "auxiliary_loss_mlp": 0.01038556, + "balance_loss_clip": 1.05244493, + "balance_loss_mlp": 1.02844715, + "epoch": 0.39752299645283473, + "flos": 21543278544000.0, + "grad_norm": 2.31484612775031, + "language_loss": 0.67365646, + "learning_rate": 2.7428208547614228e-06, + "loss": 0.69577295, + "num_input_tokens_seen": 71273220, + "step": 3306, + "time_per_iteration": 2.7786238193511963 + }, + { + "auxiliary_loss_clip": 0.01182728, + "auxiliary_loss_mlp": 0.01032573, + "balance_loss_clip": 1.05481291, + "balance_loss_mlp": 1.02346563, + "epoch": 0.39764323934347384, + "flos": 19208295031680.0, + "grad_norm": 1.770075884497149, + "language_loss": 0.76935613, + "learning_rate": 2.742097547574043e-06, + "loss": 0.79150915, + "num_input_tokens_seen": 71291445, + "step": 3307, + "time_per_iteration": 2.7536685466766357 + }, + { + "auxiliary_loss_clip": 0.01177201, + "auxiliary_loss_mlp": 0.00887353, + "balance_loss_clip": 1.05242836, + "balance_loss_mlp": 1.00061417, + "epoch": 0.3977634822341129, + "flos": 20850202644480.0, + "grad_norm": 1.7859674345263254, + "language_loss": 0.77676356, + "learning_rate": 2.7413741278110544e-06, + "loss": 0.79740906, + "num_input_tokens_seen": 71310135, + "step": 3308, + "time_per_iteration": 2.7513813972473145 + }, + { + "auxiliary_loss_clip": 0.01174822, + "auxiliary_loss_mlp": 0.01036328, + "balance_loss_clip": 1.05449438, + "balance_loss_mlp": 1.02662444, + "epoch": 0.397883725124752, + "flos": 39786042038400.0, + "grad_norm": 2.229626355023296, + "language_loss": 0.69097447, + "learning_rate": 2.7406505955822016e-06, + "loss": 0.71308601, + "num_input_tokens_seen": 71331160, + "step": 3309, + "time_per_iteration": 2.9130353927612305 + }, + { + "auxiliary_loss_clip": 0.01168686, + "auxiliary_loss_mlp": 0.01033188, + "balance_loss_clip": 1.04911709, + "balance_loss_mlp": 1.02375829, + "epoch": 0.39800396801539106, + "flos": 17379507934080.0, + "grad_norm": 2.306203696949049, + "language_loss": 0.66126579, + "learning_rate": 2.7399269509972415e-06, + "loss": 0.68328452, + "num_input_tokens_seen": 71345315, + "step": 3310, + "time_per_iteration": 2.7412660121917725 + }, + { + "auxiliary_loss_clip": 0.01162283, + "auxiliary_loss_mlp": 0.01036213, + "balance_loss_clip": 1.04595089, + "balance_loss_mlp": 1.0259428, + "epoch": 0.3981242109060302, + "flos": 19202764337280.0, + "grad_norm": 2.202214701839998, + "language_loss": 0.84866643, + "learning_rate": 2.7392031941659514e-06, + "loss": 0.87065142, + "num_input_tokens_seen": 71363160, + "step": 3311, + "time_per_iteration": 2.737619638442993 + }, + { + "auxiliary_loss_clip": 0.01172359, + "auxiliary_loss_mlp": 0.01037307, + "balance_loss_clip": 1.05510783, + "balance_loss_mlp": 1.02814019, + "epoch": 0.3982444537966693, + "flos": 24565124903040.0, + "grad_norm": 1.6934385791760471, + "language_loss": 0.85948485, + "learning_rate": 2.7384793251981244e-06, + "loss": 0.88158154, + "num_input_tokens_seen": 71382145, + "step": 3312, + "time_per_iteration": 2.865192413330078 + }, + { + "auxiliary_loss_clip": 0.01188688, + "auxiliary_loss_mlp": 0.0103831, + "balance_loss_clip": 1.05450296, + "balance_loss_mlp": 1.02934551, + "epoch": 0.39836469668730834, + "flos": 26213856099840.0, + "grad_norm": 2.3031132651632733, + "language_loss": 0.80970764, + "learning_rate": 2.737755344203571e-06, + "loss": 0.83197761, + "num_input_tokens_seen": 71402095, + "step": 3313, + "time_per_iteration": 2.8613038063049316 + }, + { + "auxiliary_loss_clip": 0.01183145, + "auxiliary_loss_mlp": 0.01036893, + "balance_loss_clip": 1.05597508, + "balance_loss_mlp": 1.02749944, + "epoch": 0.39848493957794745, + "flos": 27636134002560.0, + "grad_norm": 1.5367741150771885, + "language_loss": 0.79784983, + "learning_rate": 2.7370312512921186e-06, + "loss": 0.82005024, + "num_input_tokens_seen": 71423875, + "step": 3314, + "time_per_iteration": 2.918980121612549 + }, + { + "auxiliary_loss_clip": 0.01175632, + "auxiliary_loss_mlp": 0.01033453, + "balance_loss_clip": 1.05089688, + "balance_loss_mlp": 1.02380908, + "epoch": 0.39860518246858656, + "flos": 12239326944000.0, + "grad_norm": 2.27027996060095, + "language_loss": 0.76858318, + "learning_rate": 2.736307046573611e-06, + "loss": 0.79067403, + "num_input_tokens_seen": 71439745, + "step": 3315, + "time_per_iteration": 4.61094069480896 + }, + { + "auxiliary_loss_clip": 0.01188923, + "auxiliary_loss_mlp": 0.01035412, + "balance_loss_clip": 1.0540905, + "balance_loss_mlp": 1.02703714, + "epoch": 0.3987254253592256, + "flos": 22379135005440.0, + "grad_norm": 1.635119012199631, + "language_loss": 0.81532145, + "learning_rate": 2.73558273015791e-06, + "loss": 0.83756483, + "num_input_tokens_seen": 71459575, + "step": 3316, + "time_per_iteration": 2.59334397315979 + }, + { + "auxiliary_loss_clip": 0.01193766, + "auxiliary_loss_mlp": 0.01033478, + "balance_loss_clip": 1.05673099, + "balance_loss_mlp": 1.02395344, + "epoch": 0.3988456682498647, + "flos": 23514020190720.0, + "grad_norm": 2.0041006346933483, + "language_loss": 0.70466602, + "learning_rate": 2.734858302154894e-06, + "loss": 0.72693849, + "num_input_tokens_seen": 71481075, + "step": 3317, + "time_per_iteration": 3.5658199787139893 + }, + { + "auxiliary_loss_clip": 0.01163347, + "auxiliary_loss_mlp": 0.01030366, + "balance_loss_clip": 1.0502243, + "balance_loss_mlp": 1.02093637, + "epoch": 0.39896591114050384, + "flos": 19208761908480.0, + "grad_norm": 1.9803836264620465, + "language_loss": 0.75971556, + "learning_rate": 2.734133762674457e-06, + "loss": 0.78165269, + "num_input_tokens_seen": 71500665, + "step": 3318, + "time_per_iteration": 2.698803424835205 + }, + { + "auxiliary_loss_clip": 0.01171463, + "auxiliary_loss_mlp": 0.01039324, + "balance_loss_clip": 1.05216277, + "balance_loss_mlp": 1.03016257, + "epoch": 0.3990861540311429, + "flos": 28401031146240.0, + "grad_norm": 2.115404993821301, + "language_loss": 0.70547748, + "learning_rate": 2.7334091118265124e-06, + "loss": 0.72758532, + "num_input_tokens_seen": 71522560, + "step": 3319, + "time_per_iteration": 3.6676838397979736 + }, + { + "auxiliary_loss_clip": 0.0107852, + "auxiliary_loss_mlp": 0.01005669, + "balance_loss_clip": 1.01899219, + "balance_loss_mlp": 1.00358295, + "epoch": 0.399206396921782, + "flos": 61758563086080.0, + "grad_norm": 0.6752007412597825, + "language_loss": 0.57866347, + "learning_rate": 2.732684349720989e-06, + "loss": 0.59950542, + "num_input_tokens_seen": 71590520, + "step": 3320, + "time_per_iteration": 3.28867769241333 + }, + { + "auxiliary_loss_clip": 0.01161546, + "auxiliary_loss_mlp": 0.01028738, + "balance_loss_clip": 1.05033135, + "balance_loss_mlp": 1.01967216, + "epoch": 0.3993266398124211, + "flos": 28074567409920.0, + "grad_norm": 1.7246733434972452, + "language_loss": 0.75681251, + "learning_rate": 2.7319594764678318e-06, + "loss": 0.77871537, + "num_input_tokens_seen": 71612620, + "step": 3321, + "time_per_iteration": 2.87203311920166 + }, + { + "auxiliary_loss_clip": 0.01154487, + "auxiliary_loss_mlp": 0.01027395, + "balance_loss_clip": 1.05118251, + "balance_loss_mlp": 1.01798308, + "epoch": 0.39944688270306017, + "flos": 23225083188480.0, + "grad_norm": 1.75551787659225, + "language_loss": 0.83225632, + "learning_rate": 2.7312344921770044e-06, + "loss": 0.85407507, + "num_input_tokens_seen": 71634320, + "step": 3322, + "time_per_iteration": 2.8532464504241943 + }, + { + "auxiliary_loss_clip": 0.0116982, + "auxiliary_loss_mlp": 0.01032241, + "balance_loss_clip": 1.0488472, + "balance_loss_mlp": 1.02323449, + "epoch": 0.3995671255936993, + "flos": 19390433921280.0, + "grad_norm": 1.909600510185098, + "language_loss": 0.78437328, + "learning_rate": 2.7305093969584857e-06, + "loss": 0.80639386, + "num_input_tokens_seen": 71653145, + "step": 3323, + "time_per_iteration": 2.991741895675659 + }, + { + "auxiliary_loss_clip": 0.01172376, + "auxiliary_loss_mlp": 0.01030453, + "balance_loss_clip": 1.04935563, + "balance_loss_mlp": 1.02241206, + "epoch": 0.3996873684843384, + "flos": 23842638743040.0, + "grad_norm": 3.313588997574351, + "language_loss": 0.79786092, + "learning_rate": 2.729784190922272e-06, + "loss": 0.81988919, + "num_input_tokens_seen": 71674580, + "step": 3324, + "time_per_iteration": 2.8619277477264404 + }, + { + "auxiliary_loss_clip": 0.01067722, + "auxiliary_loss_mlp": 0.01002199, + "balance_loss_clip": 1.01855397, + "balance_loss_mlp": 0.99994582, + "epoch": 0.39980761137497745, + "flos": 66576877280640.0, + "grad_norm": 0.98191165308964, + "language_loss": 0.57230645, + "learning_rate": 2.729058874178378e-06, + "loss": 0.59300566, + "num_input_tokens_seen": 71745260, + "step": 3325, + "time_per_iteration": 3.426706075668335 + }, + { + "auxiliary_loss_clip": 0.01173279, + "auxiliary_loss_mlp": 0.01031729, + "balance_loss_clip": 1.05073571, + "balance_loss_mlp": 1.02198327, + "epoch": 0.39992785426561656, + "flos": 28549162834560.0, + "grad_norm": 1.886904489878882, + "language_loss": 0.69160229, + "learning_rate": 2.7283334468368315e-06, + "loss": 0.71365237, + "num_input_tokens_seen": 71766540, + "step": 3326, + "time_per_iteration": 2.7830607891082764 + }, + { + "auxiliary_loss_clip": 0.01116696, + "auxiliary_loss_mlp": 0.01034162, + "balance_loss_clip": 1.03989244, + "balance_loss_mlp": 1.02470303, + "epoch": 0.4000480971562556, + "flos": 15049408671360.0, + "grad_norm": 1.8812930009216327, + "language_loss": 0.72729605, + "learning_rate": 2.72760790900768e-06, + "loss": 0.74880469, + "num_input_tokens_seen": 71783125, + "step": 3327, + "time_per_iteration": 3.047955274581909 + }, + { + "auxiliary_loss_clip": 0.01194654, + "auxiliary_loss_mlp": 0.01037221, + "balance_loss_clip": 1.05860353, + "balance_loss_mlp": 1.02828026, + "epoch": 0.4001683400468947, + "flos": 23915609222400.0, + "grad_norm": 1.9060170949593889, + "language_loss": 0.7865746, + "learning_rate": 2.7268822608009875e-06, + "loss": 0.80889338, + "num_input_tokens_seen": 71802500, + "step": 3328, + "time_per_iteration": 3.727001190185547 + }, + { + "auxiliary_loss_clip": 0.01166835, + "auxiliary_loss_mlp": 0.01033002, + "balance_loss_clip": 1.05042505, + "balance_loss_mlp": 1.02372718, + "epoch": 0.40028858293753383, + "flos": 24352677912960.0, + "grad_norm": 1.7786416864689854, + "language_loss": 0.78403395, + "learning_rate": 2.726156502326834e-06, + "loss": 0.8060323, + "num_input_tokens_seen": 71823800, + "step": 3329, + "time_per_iteration": 285.2562131881714 + }, + { + "auxiliary_loss_clip": 0.0104357, + "auxiliary_loss_mlp": 0.01007332, + "balance_loss_clip": 1.01270974, + "balance_loss_mlp": 1.0044831, + "epoch": 0.4004088258281729, + "flos": 66787025800320.0, + "grad_norm": 0.6977156315860853, + "language_loss": 0.60275853, + "learning_rate": 2.725430633695316e-06, + "loss": 0.62326753, + "num_input_tokens_seen": 71886880, + "step": 3330, + "time_per_iteration": 3.5546634197235107 + }, + { + "auxiliary_loss_clip": 0.01080891, + "auxiliary_loss_mlp": 0.0100266, + "balance_loss_clip": 1.01656175, + "balance_loss_mlp": 1.00064528, + "epoch": 0.400529068718812, + "flos": 58598386473600.0, + "grad_norm": 0.8881912493030969, + "language_loss": 0.57971036, + "learning_rate": 2.7247046550165485e-06, + "loss": 0.60054582, + "num_input_tokens_seen": 71939005, + "step": 3331, + "time_per_iteration": 3.150181770324707 + }, + { + "auxiliary_loss_clip": 0.01192982, + "auxiliary_loss_mlp": 0.01036841, + "balance_loss_clip": 1.05587792, + "balance_loss_mlp": 1.02704191, + "epoch": 0.4006493116094511, + "flos": 25377460934400.0, + "grad_norm": 1.611195094350839, + "language_loss": 0.76160878, + "learning_rate": 2.7239785664006606e-06, + "loss": 0.783907, + "num_input_tokens_seen": 71962545, + "step": 3332, + "time_per_iteration": 2.82342791557312 + }, + { + "auxiliary_loss_clip": 0.01076597, + "auxiliary_loss_mlp": 0.01001975, + "balance_loss_clip": 1.01640821, + "balance_loss_mlp": 0.9999842, + "epoch": 0.40076955450009016, + "flos": 60280729822080.0, + "grad_norm": 0.798342830096355, + "language_loss": 0.61815917, + "learning_rate": 2.7232523679578002e-06, + "loss": 0.63894486, + "num_input_tokens_seen": 72025625, + "step": 3333, + "time_per_iteration": 3.433861017227173 + }, + { + "auxiliary_loss_clip": 0.0117898, + "auxiliary_loss_mlp": 0.0102916, + "balance_loss_clip": 1.05463421, + "balance_loss_mlp": 1.02061272, + "epoch": 0.4008897973907293, + "flos": 16617268396800.0, + "grad_norm": 2.084080908902394, + "language_loss": 0.788077, + "learning_rate": 2.7225260597981295e-06, + "loss": 0.81015837, + "num_input_tokens_seen": 72043330, + "step": 3334, + "time_per_iteration": 2.7948801517486572 + }, + { + "auxiliary_loss_clip": 0.01156474, + "auxiliary_loss_mlp": 0.00887422, + "balance_loss_clip": 1.04893839, + "balance_loss_mlp": 1.00061977, + "epoch": 0.4010100402813684, + "flos": 15377344865280.0, + "grad_norm": 2.5010718679751593, + "language_loss": 0.7835803, + "learning_rate": 2.721799642031831e-06, + "loss": 0.80401933, + "num_input_tokens_seen": 72059500, + "step": 3335, + "time_per_iteration": 2.7567367553710938 + }, + { + "auxiliary_loss_clip": 0.0117324, + "auxiliary_loss_mlp": 0.01030502, + "balance_loss_clip": 1.04843378, + "balance_loss_mlp": 1.02115571, + "epoch": 0.40113028317200744, + "flos": 13298835438720.0, + "grad_norm": 2.192399163382691, + "language_loss": 0.77924049, + "learning_rate": 2.721073114769101e-06, + "loss": 0.80127788, + "num_input_tokens_seen": 72077175, + "step": 3336, + "time_per_iteration": 2.7434422969818115 + }, + { + "auxiliary_loss_clip": 0.01159056, + "auxiliary_loss_mlp": 0.01035271, + "balance_loss_clip": 1.05101085, + "balance_loss_mlp": 1.0261929, + "epoch": 0.40125052606264655, + "flos": 20668027841280.0, + "grad_norm": 1.7846825144742013, + "language_loss": 0.75342834, + "learning_rate": 2.7203464781201523e-06, + "loss": 0.77537155, + "num_input_tokens_seen": 72096490, + "step": 3337, + "time_per_iteration": 2.793673038482666 + }, + { + "auxiliary_loss_clip": 0.01193368, + "auxiliary_loss_mlp": 0.01034039, + "balance_loss_clip": 1.05620885, + "balance_loss_mlp": 1.02463984, + "epoch": 0.40137076895328566, + "flos": 24607679541120.0, + "grad_norm": 1.934666754015008, + "language_loss": 0.77924061, + "learning_rate": 2.719619732195215e-06, + "loss": 0.80151469, + "num_input_tokens_seen": 72118130, + "step": 3338, + "time_per_iteration": 2.7082972526550293 + }, + { + "auxiliary_loss_clip": 0.01161314, + "auxiliary_loss_mlp": 0.01032959, + "balance_loss_clip": 1.05049956, + "balance_loss_mlp": 1.02383351, + "epoch": 0.4014910118439247, + "flos": 24206593299840.0, + "grad_norm": 1.5550548964249742, + "language_loss": 0.72513783, + "learning_rate": 2.7188928771045377e-06, + "loss": 0.74708056, + "num_input_tokens_seen": 72139450, + "step": 3339, + "time_per_iteration": 2.8166000843048096 + }, + { + "auxiliary_loss_clip": 0.01155729, + "auxiliary_loss_mlp": 0.01029287, + "balance_loss_clip": 1.04762161, + "balance_loss_mlp": 1.01975632, + "epoch": 0.4016112547345638, + "flos": 26725080418560.0, + "grad_norm": 1.6486240949352269, + "language_loss": 0.79498923, + "learning_rate": 2.7181659129583815e-06, + "loss": 0.81683934, + "num_input_tokens_seen": 72159040, + "step": 3340, + "time_per_iteration": 4.0037291049957275 + }, + { + "auxiliary_loss_clip": 0.01162601, + "auxiliary_loss_mlp": 0.01028444, + "balance_loss_clip": 1.04670358, + "balance_loss_mlp": 1.0200932, + "epoch": 0.4017314976252029, + "flos": 21288025520640.0, + "grad_norm": 1.7184736764356154, + "language_loss": 0.76130319, + "learning_rate": 2.7174388398670276e-06, + "loss": 0.78321362, + "num_input_tokens_seen": 72178220, + "step": 3341, + "time_per_iteration": 3.6799275875091553 + }, + { + "auxiliary_loss_clip": 0.01190575, + "auxiliary_loss_mlp": 0.01034091, + "balance_loss_clip": 1.05202818, + "balance_loss_mlp": 1.02463794, + "epoch": 0.401851740515842, + "flos": 25484690010240.0, + "grad_norm": 2.100696605663739, + "language_loss": 0.92024374, + "learning_rate": 2.716711657940773e-06, + "loss": 0.9424904, + "num_input_tokens_seen": 72199230, + "step": 3342, + "time_per_iteration": 3.6076531410217285 + }, + { + "auxiliary_loss_clip": 0.01059091, + "auxiliary_loss_mlp": 0.01005721, + "balance_loss_clip": 1.01772308, + "balance_loss_mlp": 1.00355184, + "epoch": 0.4019719834064811, + "flos": 55395334978560.0, + "grad_norm": 0.9031279528746762, + "language_loss": 0.56473649, + "learning_rate": 2.7159843672899284e-06, + "loss": 0.58538461, + "num_input_tokens_seen": 72263430, + "step": 3343, + "time_per_iteration": 3.4715166091918945 + }, + { + "auxiliary_loss_clip": 0.01182774, + "auxiliary_loss_mlp": 0.01036691, + "balance_loss_clip": 1.05554914, + "balance_loss_mlp": 1.02771473, + "epoch": 0.40209222629712016, + "flos": 18180100218240.0, + "grad_norm": 1.7809175186275596, + "language_loss": 0.81028694, + "learning_rate": 2.715256968024825e-06, + "loss": 0.83248162, + "num_input_tokens_seen": 72280505, + "step": 3344, + "time_per_iteration": 3.718878746032715 + }, + { + "auxiliary_loss_clip": 0.01177112, + "auxiliary_loss_mlp": 0.01028384, + "balance_loss_clip": 1.05403984, + "balance_loss_mlp": 1.01888895, + "epoch": 0.40221246918775927, + "flos": 25961009287680.0, + "grad_norm": 2.170473020583165, + "language_loss": 0.82202131, + "learning_rate": 2.7145294602558083e-06, + "loss": 0.84407628, + "num_input_tokens_seen": 72301215, + "step": 3345, + "time_per_iteration": 2.8137059211730957 + }, + { + "auxiliary_loss_clip": 0.01182494, + "auxiliary_loss_mlp": 0.01032487, + "balance_loss_clip": 1.0534091, + "balance_loss_mlp": 1.02249765, + "epoch": 0.4023327120783984, + "flos": 33838912056960.0, + "grad_norm": 1.9074886737690195, + "language_loss": 0.70510745, + "learning_rate": 2.713801844093241e-06, + "loss": 0.72725725, + "num_input_tokens_seen": 72322365, + "step": 3346, + "time_per_iteration": 2.829503059387207 + }, + { + "auxiliary_loss_clip": 0.0118111, + "auxiliary_loss_mlp": 0.0103073, + "balance_loss_clip": 1.05222857, + "balance_loss_mlp": 1.02192056, + "epoch": 0.40245295496903744, + "flos": 26900252069760.0, + "grad_norm": 2.3138675773514885, + "language_loss": 0.88260841, + "learning_rate": 2.7130741196475014e-06, + "loss": 0.9047268, + "num_input_tokens_seen": 72340495, + "step": 3347, + "time_per_iteration": 2.722287893295288 + }, + { + "auxiliary_loss_clip": 0.0117041, + "auxiliary_loss_mlp": 0.01034485, + "balance_loss_clip": 1.05288124, + "balance_loss_mlp": 1.02459073, + "epoch": 0.40257319785967655, + "flos": 36902738436480.0, + "grad_norm": 1.862686252699284, + "language_loss": 0.79180795, + "learning_rate": 2.7123462870289848e-06, + "loss": 0.8138569, + "num_input_tokens_seen": 72360545, + "step": 3348, + "time_per_iteration": 2.8865020275115967 + }, + { + "auxiliary_loss_clip": 0.01171119, + "auxiliary_loss_mlp": 0.01040944, + "balance_loss_clip": 1.04748106, + "balance_loss_mlp": 1.03158617, + "epoch": 0.40269344075031566, + "flos": 24353180703360.0, + "grad_norm": 1.5301435783791846, + "language_loss": 0.81166774, + "learning_rate": 2.711618346348102e-06, + "loss": 0.83378839, + "num_input_tokens_seen": 72381070, + "step": 3349, + "time_per_iteration": 2.835740804672241 + }, + { + "auxiliary_loss_clip": 0.01163726, + "auxiliary_loss_mlp": 0.01036118, + "balance_loss_clip": 1.05081224, + "balance_loss_mlp": 1.0275408, + "epoch": 0.4028136836409547, + "flos": 14389657614720.0, + "grad_norm": 1.9365500455483073, + "language_loss": 0.63668334, + "learning_rate": 2.7108902977152825e-06, + "loss": 0.65868175, + "num_input_tokens_seen": 72398970, + "step": 3350, + "time_per_iteration": 2.789454936981201 + }, + { + "auxiliary_loss_clip": 0.01176962, + "auxiliary_loss_mlp": 0.01035074, + "balance_loss_clip": 1.0496906, + "balance_loss_mlp": 1.02568018, + "epoch": 0.4029339265315938, + "flos": 26136037284480.0, + "grad_norm": 2.02739266863548, + "language_loss": 0.74367797, + "learning_rate": 2.7101621412409704e-06, + "loss": 0.76579833, + "num_input_tokens_seen": 72418455, + "step": 3351, + "time_per_iteration": 2.7575700283050537 + }, + { + "auxiliary_loss_clip": 0.01191278, + "auxiliary_loss_mlp": 0.01028239, + "balance_loss_clip": 1.05356443, + "balance_loss_mlp": 1.01919651, + "epoch": 0.40305416942223293, + "flos": 23256325042560.0, + "grad_norm": 2.6445495491812543, + "language_loss": 0.86250913, + "learning_rate": 2.7094338770356256e-06, + "loss": 0.88470423, + "num_input_tokens_seen": 72437540, + "step": 3352, + "time_per_iteration": 2.70176100730896 + }, + { + "auxiliary_loss_clip": 0.01168821, + "auxiliary_loss_mlp": 0.01028816, + "balance_loss_clip": 1.05178666, + "balance_loss_mlp": 1.01945233, + "epoch": 0.403174412312872, + "flos": 27089645506560.0, + "grad_norm": 2.016239749800799, + "language_loss": 0.64191389, + "learning_rate": 2.708705505209726e-06, + "loss": 0.66389024, + "num_input_tokens_seen": 72458315, + "step": 3353, + "time_per_iteration": 2.7960104942321777 + }, + { + "auxiliary_loss_clip": 0.01143117, + "auxiliary_loss_mlp": 0.01037825, + "balance_loss_clip": 1.04489434, + "balance_loss_mlp": 1.02801418, + "epoch": 0.4032946552035111, + "flos": 21756336065280.0, + "grad_norm": 3.615223629399354, + "language_loss": 0.92051733, + "learning_rate": 2.7079770258737646e-06, + "loss": 0.94232678, + "num_input_tokens_seen": 72476225, + "step": 3354, + "time_per_iteration": 2.829927682876587 + }, + { + "auxiliary_loss_clip": 0.01157211, + "auxiliary_loss_mlp": 0.01032715, + "balance_loss_clip": 1.04811454, + "balance_loss_mlp": 1.02375603, + "epoch": 0.4034148980941502, + "flos": 17343956448000.0, + "grad_norm": 3.5933226348974943, + "language_loss": 0.75142646, + "learning_rate": 2.707248439138251e-06, + "loss": 0.77332574, + "num_input_tokens_seen": 72492460, + "step": 3355, + "time_per_iteration": 2.8069841861724854 + }, + { + "auxiliary_loss_clip": 0.01169542, + "auxiliary_loss_mlp": 0.01035187, + "balance_loss_clip": 1.05622482, + "balance_loss_mlp": 1.0261395, + "epoch": 0.40353514098478926, + "flos": 22017838055040.0, + "grad_norm": 1.7778310944059241, + "language_loss": 0.65445989, + "learning_rate": 2.7065197451137114e-06, + "loss": 0.67650712, + "num_input_tokens_seen": 72513840, + "step": 3356, + "time_per_iteration": 2.7406508922576904 + }, + { + "auxiliary_loss_clip": 0.01170408, + "auxiliary_loss_mlp": 0.01032584, + "balance_loss_clip": 1.05199504, + "balance_loss_mlp": 1.0237323, + "epoch": 0.4036553838754284, + "flos": 14246446089600.0, + "grad_norm": 1.9253179838476993, + "language_loss": 0.66990244, + "learning_rate": 2.7057909439106894e-06, + "loss": 0.69193232, + "num_input_tokens_seen": 72531695, + "step": 3357, + "time_per_iteration": 2.755373239517212 + }, + { + "auxiliary_loss_clip": 0.01172115, + "auxiliary_loss_mlp": 0.00886309, + "balance_loss_clip": 1.05064106, + "balance_loss_mlp": 1.00063038, + "epoch": 0.40377562676606743, + "flos": 24790644443520.0, + "grad_norm": 1.9249945866907103, + "language_loss": 0.78211725, + "learning_rate": 2.7050620356397417e-06, + "loss": 0.80270141, + "num_input_tokens_seen": 72550645, + "step": 3358, + "time_per_iteration": 2.7508645057678223 + }, + { + "auxiliary_loss_clip": 0.01187467, + "auxiliary_loss_mlp": 0.01031665, + "balance_loss_clip": 1.05545878, + "balance_loss_mlp": 1.02300453, + "epoch": 0.40389586965670654, + "flos": 24061226958720.0, + "grad_norm": 1.8546296632349655, + "language_loss": 0.71972263, + "learning_rate": 2.7043330204114437e-06, + "loss": 0.74191403, + "num_input_tokens_seen": 72569355, + "step": 3359, + "time_per_iteration": 2.7173972129821777 + }, + { + "auxiliary_loss_clip": 0.01181917, + "auxiliary_loss_mlp": 0.01030678, + "balance_loss_clip": 1.05049336, + "balance_loss_mlp": 1.02198172, + "epoch": 0.40401611254734565, + "flos": 16399613934720.0, + "grad_norm": 1.8504198858353227, + "language_loss": 0.85590172, + "learning_rate": 2.7036038983363862e-06, + "loss": 0.87802762, + "num_input_tokens_seen": 72585960, + "step": 3360, + "time_per_iteration": 2.668347120285034 + }, + { + "auxiliary_loss_clip": 0.01174973, + "auxiliary_loss_mlp": 0.01034305, + "balance_loss_clip": 1.05276251, + "balance_loss_mlp": 1.02635407, + "epoch": 0.4041363554379847, + "flos": 23988220565760.0, + "grad_norm": 1.725198263058033, + "language_loss": 0.84112716, + "learning_rate": 2.702874669525177e-06, + "loss": 0.86321998, + "num_input_tokens_seen": 72604440, + "step": 3361, + "time_per_iteration": 2.749617576599121 + }, + { + "auxiliary_loss_clip": 0.01160425, + "auxiliary_loss_mlp": 0.01037562, + "balance_loss_clip": 1.05321026, + "balance_loss_mlp": 1.02905607, + "epoch": 0.4042565983286238, + "flos": 28401964899840.0, + "grad_norm": 2.3020146658014355, + "language_loss": 0.69491696, + "learning_rate": 2.7021453340884394e-06, + "loss": 0.71689689, + "num_input_tokens_seen": 72622165, + "step": 3362, + "time_per_iteration": 2.8407673835754395 + }, + { + "auxiliary_loss_clip": 0.01161927, + "auxiliary_loss_mlp": 0.00886256, + "balance_loss_clip": 1.05064702, + "balance_loss_mlp": 1.00065494, + "epoch": 0.40437684121926293, + "flos": 17710963660800.0, + "grad_norm": 3.321014275533484, + "language_loss": 0.73342198, + "learning_rate": 2.7014158921368125e-06, + "loss": 0.75390381, + "num_input_tokens_seen": 72640490, + "step": 3363, + "time_per_iteration": 2.754610538482666 + }, + { + "auxiliary_loss_clip": 0.01190114, + "auxiliary_loss_mlp": 0.01036388, + "balance_loss_clip": 1.05529404, + "balance_loss_mlp": 1.02772176, + "epoch": 0.404497084109902, + "flos": 24018959629440.0, + "grad_norm": 2.299466675676439, + "language_loss": 0.85812998, + "learning_rate": 2.700686343780953e-06, + "loss": 0.88039505, + "num_input_tokens_seen": 72660360, + "step": 3364, + "time_per_iteration": 2.663113832473755 + }, + { + "auxiliary_loss_clip": 0.01171624, + "auxiliary_loss_mlp": 0.0102998, + "balance_loss_clip": 1.05103135, + "balance_loss_mlp": 1.02140903, + "epoch": 0.4046173270005411, + "flos": 22929861306240.0, + "grad_norm": 1.5890545451373403, + "language_loss": 0.88445878, + "learning_rate": 2.699956689131532e-06, + "loss": 0.90647483, + "num_input_tokens_seen": 72680345, + "step": 3365, + "time_per_iteration": 2.9407765865325928 + }, + { + "auxiliary_loss_clip": 0.01172042, + "auxiliary_loss_mlp": 0.01034998, + "balance_loss_clip": 1.0540278, + "balance_loss_mlp": 1.02569973, + "epoch": 0.4047375698911802, + "flos": 20668135582080.0, + "grad_norm": 2.2005641278156, + "language_loss": 0.84568352, + "learning_rate": 2.699226928299238e-06, + "loss": 0.86775392, + "num_input_tokens_seen": 72698365, + "step": 3366, + "time_per_iteration": 4.63777756690979 + }, + { + "auxiliary_loss_clip": 0.01183751, + "auxiliary_loss_mlp": 0.01028983, + "balance_loss_clip": 1.05479968, + "balance_loss_mlp": 1.02110958, + "epoch": 0.40485781278181926, + "flos": 28912865996160.0, + "grad_norm": 2.393129096995118, + "language_loss": 0.79015553, + "learning_rate": 2.698497061394774e-06, + "loss": 0.8122828, + "num_input_tokens_seen": 72716850, + "step": 3367, + "time_per_iteration": 2.7767508029937744 + }, + { + "auxiliary_loss_clip": 0.01169111, + "auxiliary_loss_mlp": 0.00886714, + "balance_loss_clip": 1.05443287, + "balance_loss_mlp": 1.00065446, + "epoch": 0.40497805567245837, + "flos": 23148377694720.0, + "grad_norm": 1.4672833821207858, + "language_loss": 0.80369484, + "learning_rate": 2.6977670885288627e-06, + "loss": 0.82425308, + "num_input_tokens_seen": 72738250, + "step": 3368, + "time_per_iteration": 3.7116827964782715 + }, + { + "auxiliary_loss_clip": 0.01160283, + "auxiliary_loss_mlp": 0.01028901, + "balance_loss_clip": 1.04951894, + "balance_loss_mlp": 1.01987696, + "epoch": 0.4050982985630975, + "flos": 16289404030080.0, + "grad_norm": 1.952743239287165, + "language_loss": 0.74941856, + "learning_rate": 2.6970370098122378e-06, + "loss": 0.77131045, + "num_input_tokens_seen": 72755235, + "step": 3369, + "time_per_iteration": 2.829939603805542 + }, + { + "auxiliary_loss_clip": 0.01189233, + "auxiliary_loss_mlp": 0.01032293, + "balance_loss_clip": 1.05386448, + "balance_loss_mlp": 1.02338183, + "epoch": 0.40521854145373654, + "flos": 34459484353920.0, + "grad_norm": 1.672607285843276, + "language_loss": 0.86648697, + "learning_rate": 2.6963068253556535e-06, + "loss": 0.88870221, + "num_input_tokens_seen": 72776620, + "step": 3370, + "time_per_iteration": 3.721069574356079 + }, + { + "auxiliary_loss_clip": 0.01185798, + "auxiliary_loss_mlp": 0.01042947, + "balance_loss_clip": 1.05348349, + "balance_loss_mlp": 1.03333235, + "epoch": 0.40533878434437565, + "flos": 25331099454720.0, + "grad_norm": 2.0525447862754236, + "language_loss": 0.85741556, + "learning_rate": 2.6955765352698763e-06, + "loss": 0.87970304, + "num_input_tokens_seen": 72796765, + "step": 3371, + "time_per_iteration": 2.6825098991394043 + }, + { + "auxiliary_loss_clip": 0.0119372, + "auxiliary_loss_mlp": 0.01029069, + "balance_loss_clip": 1.05527008, + "balance_loss_mlp": 1.0199796, + "epoch": 0.40545902723501476, + "flos": 15012061505280.0, + "grad_norm": 1.8943890093061495, + "language_loss": 0.73476553, + "learning_rate": 2.6948461396656923e-06, + "loss": 0.75699341, + "num_input_tokens_seen": 72814175, + "step": 3372, + "time_per_iteration": 2.7215466499328613 + }, + { + "auxiliary_loss_clip": 0.01188088, + "auxiliary_loss_mlp": 0.01031167, + "balance_loss_clip": 1.0569247, + "balance_loss_mlp": 1.02170825, + "epoch": 0.4055792701256538, + "flos": 25521103422720.0, + "grad_norm": 2.1868675446751435, + "language_loss": 0.74779844, + "learning_rate": 2.6941156386539013e-06, + "loss": 0.76999098, + "num_input_tokens_seen": 72834125, + "step": 3373, + "time_per_iteration": 2.7569847106933594 + }, + { + "auxiliary_loss_clip": 0.01167769, + "auxiliary_loss_mlp": 0.01037501, + "balance_loss_clip": 1.05477738, + "balance_loss_mlp": 1.02844691, + "epoch": 0.4056995130162929, + "flos": 19574583972480.0, + "grad_norm": 2.644918698481804, + "language_loss": 0.81018263, + "learning_rate": 2.6933850323453203e-06, + "loss": 0.83223534, + "num_input_tokens_seen": 72852570, + "step": 3374, + "time_per_iteration": 2.7394051551818848 + }, + { + "auxiliary_loss_clip": 0.01193188, + "auxiliary_loss_mlp": 0.01040791, + "balance_loss_clip": 1.05812681, + "balance_loss_mlp": 1.03226805, + "epoch": 0.405819755906932, + "flos": 15413794191360.0, + "grad_norm": 2.0790319623248408, + "language_loss": 0.74766058, + "learning_rate": 2.6926543208507806e-06, + "loss": 0.77000034, + "num_input_tokens_seen": 72871250, + "step": 3375, + "time_per_iteration": 2.605595111846924 + }, + { + "auxiliary_loss_clip": 0.01182686, + "auxiliary_loss_mlp": 0.01035873, + "balance_loss_clip": 1.05600882, + "balance_loss_mlp": 1.0271883, + "epoch": 0.4059399987975711, + "flos": 21433930565760.0, + "grad_norm": 2.4276406136599458, + "language_loss": 0.80510169, + "learning_rate": 2.6919235042811316e-06, + "loss": 0.82728726, + "num_input_tokens_seen": 72890035, + "step": 3376, + "time_per_iteration": 2.7484116554260254 + }, + { + "auxiliary_loss_clip": 0.01160299, + "auxiliary_loss_mlp": 0.01036774, + "balance_loss_clip": 1.05023861, + "balance_loss_mlp": 1.02732611, + "epoch": 0.4060602416882102, + "flos": 25556942217600.0, + "grad_norm": 2.026788671239633, + "language_loss": 0.76331836, + "learning_rate": 2.691192582747237e-06, + "loss": 0.78528905, + "num_input_tokens_seen": 72909665, + "step": 3377, + "time_per_iteration": 2.8647072315216064 + }, + { + "auxiliary_loss_clip": 0.01192196, + "auxiliary_loss_mlp": 0.01033813, + "balance_loss_clip": 1.05678463, + "balance_loss_mlp": 1.02453232, + "epoch": 0.40618048457884925, + "flos": 23766759262080.0, + "grad_norm": 1.7559793315819252, + "language_loss": 0.73801982, + "learning_rate": 2.6904615563599765e-06, + "loss": 0.76027989, + "num_input_tokens_seen": 72929465, + "step": 3378, + "time_per_iteration": 2.6550705432891846 + }, + { + "auxiliary_loss_clip": 0.01159975, + "auxiliary_loss_mlp": 0.01032967, + "balance_loss_clip": 1.05093384, + "balance_loss_mlp": 1.02448487, + "epoch": 0.40630072746948837, + "flos": 17639681120640.0, + "grad_norm": 1.6298257415516522, + "language_loss": 0.82747573, + "learning_rate": 2.6897304252302477e-06, + "loss": 0.84940517, + "num_input_tokens_seen": 72946785, + "step": 3379, + "time_per_iteration": 2.737846612930298 + }, + { + "auxiliary_loss_clip": 0.01061111, + "auxiliary_loss_mlp": 0.01007733, + "balance_loss_clip": 1.01894903, + "balance_loss_mlp": 1.00540817, + "epoch": 0.4064209703601275, + "flos": 60836053063680.0, + "grad_norm": 0.7876085662760178, + "language_loss": 0.54802912, + "learning_rate": 2.688999189468962e-06, + "loss": 0.5687176, + "num_input_tokens_seen": 73003215, + "step": 3380, + "time_per_iteration": 3.237478733062744 + }, + { + "auxiliary_loss_clip": 0.01180074, + "auxiliary_loss_mlp": 0.01035054, + "balance_loss_clip": 1.05510867, + "balance_loss_mlp": 1.02589881, + "epoch": 0.40654121325076653, + "flos": 24024346669440.0, + "grad_norm": 2.3465446850494636, + "language_loss": 0.76155758, + "learning_rate": 2.6882678491870464e-06, + "loss": 0.78370881, + "num_input_tokens_seen": 73023650, + "step": 3381, + "time_per_iteration": 2.7569797039031982 + }, + { + "auxiliary_loss_clip": 0.01185625, + "auxiliary_loss_mlp": 0.01030533, + "balance_loss_clip": 1.05609226, + "balance_loss_mlp": 1.02171135, + "epoch": 0.40666145614140564, + "flos": 27344252085120.0, + "grad_norm": 1.7081002952085822, + "language_loss": 0.7144053, + "learning_rate": 2.6875364044954453e-06, + "loss": 0.7365669, + "num_input_tokens_seen": 73043880, + "step": 3382, + "time_per_iteration": 2.750858783721924 + }, + { + "auxiliary_loss_clip": 0.01165858, + "auxiliary_loss_mlp": 0.01028366, + "balance_loss_clip": 1.04680037, + "balance_loss_mlp": 1.0198307, + "epoch": 0.40678169903204475, + "flos": 26176724415360.0, + "grad_norm": 1.6865455834676513, + "language_loss": 0.82239735, + "learning_rate": 2.6868048555051185e-06, + "loss": 0.84433961, + "num_input_tokens_seen": 73065410, + "step": 3383, + "time_per_iteration": 2.8466949462890625 + }, + { + "auxiliary_loss_clip": 0.01177912, + "auxiliary_loss_mlp": 0.0103143, + "balance_loss_clip": 1.05108881, + "balance_loss_mlp": 1.02240539, + "epoch": 0.4069019419226838, + "flos": 28622420622720.0, + "grad_norm": 2.369720418570247, + "language_loss": 0.86060107, + "learning_rate": 2.686073202327041e-06, + "loss": 0.88269448, + "num_input_tokens_seen": 73084410, + "step": 3384, + "time_per_iteration": 2.775296449661255 + }, + { + "auxiliary_loss_clip": 0.01159369, + "auxiliary_loss_mlp": 0.0103759, + "balance_loss_clip": 1.04751503, + "balance_loss_mlp": 1.02885151, + "epoch": 0.4070221848133229, + "flos": 25229006023680.0, + "grad_norm": 1.7843812928786518, + "language_loss": 0.73327971, + "learning_rate": 2.6853414450722043e-06, + "loss": 0.75524938, + "num_input_tokens_seen": 73104075, + "step": 3385, + "time_per_iteration": 2.7868385314941406 + }, + { + "auxiliary_loss_clip": 0.01179402, + "auxiliary_loss_mlp": 0.01032012, + "balance_loss_clip": 1.05422831, + "balance_loss_mlp": 1.02276158, + "epoch": 0.40714242770396203, + "flos": 18405224709120.0, + "grad_norm": 1.8390202165663412, + "language_loss": 0.84933871, + "learning_rate": 2.684609583851616e-06, + "loss": 0.87145281, + "num_input_tokens_seen": 73122250, + "step": 3386, + "time_per_iteration": 2.6553092002868652 + }, + { + "auxiliary_loss_clip": 0.01149254, + "auxiliary_loss_mlp": 0.01029571, + "balance_loss_clip": 1.04757547, + "balance_loss_mlp": 1.02063644, + "epoch": 0.4072626705946011, + "flos": 30228920403840.0, + "grad_norm": 2.6129705100553515, + "language_loss": 0.80814838, + "learning_rate": 2.683877618776297e-06, + "loss": 0.82993662, + "num_input_tokens_seen": 73144505, + "step": 3387, + "time_per_iteration": 2.882655382156372 + }, + { + "auxiliary_loss_clip": 0.01163478, + "auxiliary_loss_mlp": 0.01039503, + "balance_loss_clip": 1.04803967, + "balance_loss_mlp": 1.03005517, + "epoch": 0.4073829134852402, + "flos": 21834549930240.0, + "grad_norm": 13.039551408051341, + "language_loss": 0.73602605, + "learning_rate": 2.6831455499572876e-06, + "loss": 0.75805587, + "num_input_tokens_seen": 73162440, + "step": 3388, + "time_per_iteration": 2.8203930854797363 + }, + { + "auxiliary_loss_clip": 0.01188404, + "auxiliary_loss_mlp": 0.01026277, + "balance_loss_clip": 1.05281186, + "balance_loss_mlp": 1.01662743, + "epoch": 0.40750315637587925, + "flos": 25260211964160.0, + "grad_norm": 1.967895942755641, + "language_loss": 0.77426112, + "learning_rate": 2.682413377505641e-06, + "loss": 0.79640788, + "num_input_tokens_seen": 73181245, + "step": 3389, + "time_per_iteration": 2.7130439281463623 + }, + { + "auxiliary_loss_clip": 0.01180106, + "auxiliary_loss_mlp": 0.01031415, + "balance_loss_clip": 1.05179191, + "balance_loss_mlp": 1.02187824, + "epoch": 0.40762339926651836, + "flos": 19712767593600.0, + "grad_norm": 3.3083852612777047, + "language_loss": 0.76566195, + "learning_rate": 2.6816811015324284e-06, + "loss": 0.78777719, + "num_input_tokens_seen": 73199295, + "step": 3390, + "time_per_iteration": 2.7589094638824463 + }, + { + "auxiliary_loss_clip": 0.01080997, + "auxiliary_loss_mlp": 0.0100186, + "balance_loss_clip": 1.018116, + "balance_loss_mlp": 0.99966663, + "epoch": 0.40774364215715747, + "flos": 71449307314560.0, + "grad_norm": 0.7320542456698453, + "language_loss": 0.56685781, + "learning_rate": 2.6809487221487343e-06, + "loss": 0.58768642, + "num_input_tokens_seen": 73258780, + "step": 3391, + "time_per_iteration": 3.187483310699463 + }, + { + "auxiliary_loss_clip": 0.01168645, + "auxiliary_loss_mlp": 0.01033189, + "balance_loss_clip": 1.0479548, + "balance_loss_mlp": 1.02390218, + "epoch": 0.4078638850477965, + "flos": 15084134144640.0, + "grad_norm": 2.6408733200803365, + "language_loss": 0.8174262, + "learning_rate": 2.6802162394656605e-06, + "loss": 0.83944452, + "num_input_tokens_seen": 73275490, + "step": 3392, + "time_per_iteration": 4.623152494430542 + }, + { + "auxiliary_loss_clip": 0.01166289, + "auxiliary_loss_mlp": 0.01032198, + "balance_loss_clip": 1.04787385, + "balance_loss_mlp": 1.02241671, + "epoch": 0.40798412793843564, + "flos": 23842890138240.0, + "grad_norm": 2.1158987594901815, + "language_loss": 0.71917045, + "learning_rate": 2.679483653594324e-06, + "loss": 0.74115539, + "num_input_tokens_seen": 73297260, + "step": 3393, + "time_per_iteration": 2.8471217155456543 + }, + { + "auxiliary_loss_clip": 0.01182622, + "auxiliary_loss_mlp": 0.01029131, + "balance_loss_clip": 1.05373061, + "balance_loss_mlp": 1.02067304, + "epoch": 0.40810437082907475, + "flos": 21065774117760.0, + "grad_norm": 2.6451551197317786, + "language_loss": 0.7687012, + "learning_rate": 2.678750964645857e-06, + "loss": 0.79081869, + "num_input_tokens_seen": 73316340, + "step": 3394, + "time_per_iteration": 3.708184242248535 + }, + { + "auxiliary_loss_clip": 0.01181849, + "auxiliary_loss_mlp": 0.0103287, + "balance_loss_clip": 1.05679059, + "balance_loss_mlp": 1.02397084, + "epoch": 0.4082246137197138, + "flos": 11321377948800.0, + "grad_norm": 3.5425061573082592, + "language_loss": 0.82852364, + "learning_rate": 2.6780181727314094e-06, + "loss": 0.85067081, + "num_input_tokens_seen": 73331245, + "step": 3395, + "time_per_iteration": 2.703597068786621 + }, + { + "auxiliary_loss_clip": 0.01165971, + "auxiliary_loss_mlp": 0.00887032, + "balance_loss_clip": 1.05222869, + "balance_loss_mlp": 1.00071084, + "epoch": 0.4083448566103529, + "flos": 19062569554560.0, + "grad_norm": 2.1547526909912644, + "language_loss": 0.77624404, + "learning_rate": 2.6772852779621435e-06, + "loss": 0.79677409, + "num_input_tokens_seen": 73349105, + "step": 3396, + "time_per_iteration": 4.0210793018341064 + }, + { + "auxiliary_loss_clip": 0.01173122, + "auxiliary_loss_mlp": 0.00885673, + "balance_loss_clip": 1.05409813, + "balance_loss_mlp": 1.00060749, + "epoch": 0.408465099500992, + "flos": 23550254035200.0, + "grad_norm": 1.8970325660370622, + "language_loss": 0.86727726, + "learning_rate": 2.676552280449239e-06, + "loss": 0.88786525, + "num_input_tokens_seen": 73368990, + "step": 3397, + "time_per_iteration": 2.722996234893799 + }, + { + "auxiliary_loss_clip": 0.01171542, + "auxiliary_loss_mlp": 0.01031384, + "balance_loss_clip": 1.05092609, + "balance_loss_mlp": 1.02163267, + "epoch": 0.4085853423916311, + "flos": 12750012558720.0, + "grad_norm": 2.9622929824440014, + "language_loss": 0.75974852, + "learning_rate": 2.6758191803038917e-06, + "loss": 0.78177774, + "num_input_tokens_seen": 73387485, + "step": 3398, + "time_per_iteration": 2.712153434753418 + }, + { + "auxiliary_loss_clip": 0.01135113, + "auxiliary_loss_mlp": 0.01033983, + "balance_loss_clip": 1.04823124, + "balance_loss_mlp": 1.02482736, + "epoch": 0.4087055852822702, + "flos": 24353072962560.0, + "grad_norm": 1.8569092790239534, + "language_loss": 0.8274529, + "learning_rate": 2.6750859776373125e-06, + "loss": 0.84914386, + "num_input_tokens_seen": 73406940, + "step": 3399, + "time_per_iteration": 2.8703739643096924 + }, + { + "auxiliary_loss_clip": 0.01046785, + "auxiliary_loss_mlp": 0.01001775, + "balance_loss_clip": 1.01285768, + "balance_loss_mlp": 0.99921244, + "epoch": 0.4088258281729093, + "flos": 66387950720640.0, + "grad_norm": 0.8207734899736304, + "language_loss": 0.60366338, + "learning_rate": 2.674352672560727e-06, + "loss": 0.62414896, + "num_input_tokens_seen": 73468385, + "step": 3400, + "time_per_iteration": 3.3831562995910645 + }, + { + "auxiliary_loss_clip": 0.011546, + "auxiliary_loss_mlp": 0.01034336, + "balance_loss_clip": 1.04646742, + "balance_loss_mlp": 1.02503777, + "epoch": 0.40894607106354836, + "flos": 20449260057600.0, + "grad_norm": 1.9612015927466897, + "language_loss": 0.76930368, + "learning_rate": 2.673619265185377e-06, + "loss": 0.79119301, + "num_input_tokens_seen": 73488225, + "step": 3401, + "time_per_iteration": 2.7824769020080566 + }, + { + "auxiliary_loss_clip": 0.01183674, + "auxiliary_loss_mlp": 0.01038474, + "balance_loss_clip": 1.05347264, + "balance_loss_mlp": 1.02875805, + "epoch": 0.40906631395418747, + "flos": 27053627143680.0, + "grad_norm": 1.7224227109599475, + "language_loss": 0.78237259, + "learning_rate": 2.672885755622521e-06, + "loss": 0.80459404, + "num_input_tokens_seen": 73510640, + "step": 3402, + "time_per_iteration": 2.7451858520507812 + }, + { + "auxiliary_loss_clip": 0.01149261, + "auxiliary_loss_mlp": 0.01033493, + "balance_loss_clip": 1.04728568, + "balance_loss_mlp": 1.02420616, + "epoch": 0.4091865568448266, + "flos": 25484151306240.0, + "grad_norm": 1.887425015170985, + "language_loss": 0.70069313, + "learning_rate": 2.67215214398343e-06, + "loss": 0.72252071, + "num_input_tokens_seen": 73530655, + "step": 3403, + "time_per_iteration": 2.92669415473938 + }, + { + "auxiliary_loss_clip": 0.011531, + "auxiliary_loss_mlp": 0.01032405, + "balance_loss_clip": 1.04635727, + "balance_loss_mlp": 1.02265334, + "epoch": 0.40930679973546563, + "flos": 28657864368000.0, + "grad_norm": 2.580440622635244, + "language_loss": 0.77763796, + "learning_rate": 2.671418430379393e-06, + "loss": 0.79949296, + "num_input_tokens_seen": 73549340, + "step": 3404, + "time_per_iteration": 2.9015562534332275 + }, + { + "auxiliary_loss_clip": 0.0118968, + "auxiliary_loss_mlp": 0.01030229, + "balance_loss_clip": 1.0534364, + "balance_loss_mlp": 1.02144349, + "epoch": 0.40942704262610474, + "flos": 20886292834560.0, + "grad_norm": 4.504159340579704, + "language_loss": 0.83119386, + "learning_rate": 2.670684614921715e-06, + "loss": 0.85339296, + "num_input_tokens_seen": 73568315, + "step": 3405, + "time_per_iteration": 2.6983158588409424 + }, + { + "auxiliary_loss_clip": 0.01171645, + "auxiliary_loss_mlp": 0.01034033, + "balance_loss_clip": 1.04948115, + "balance_loss_mlp": 1.02457976, + "epoch": 0.4095472855167438, + "flos": 21618080616960.0, + "grad_norm": 2.104900293306033, + "language_loss": 0.68800735, + "learning_rate": 2.6699506977217128e-06, + "loss": 0.71006411, + "num_input_tokens_seen": 73588490, + "step": 3406, + "time_per_iteration": 2.855894088745117 + }, + { + "auxiliary_loss_clip": 0.01178132, + "auxiliary_loss_mlp": 0.0103047, + "balance_loss_clip": 1.05550921, + "balance_loss_mlp": 1.02186942, + "epoch": 0.4096675284073829, + "flos": 27926112499200.0, + "grad_norm": 2.200123320138174, + "language_loss": 0.70336998, + "learning_rate": 2.6692166788907233e-06, + "loss": 0.725456, + "num_input_tokens_seen": 73608685, + "step": 3407, + "time_per_iteration": 2.7875845432281494 + }, + { + "auxiliary_loss_clip": 0.01172717, + "auxiliary_loss_mlp": 0.01034432, + "balance_loss_clip": 1.05049002, + "balance_loss_mlp": 1.02483547, + "epoch": 0.409787771298022, + "flos": 19206607092480.0, + "grad_norm": 1.8712620058358278, + "language_loss": 0.76926684, + "learning_rate": 2.6684825585400957e-06, + "loss": 0.79133838, + "num_input_tokens_seen": 73627630, + "step": 3408, + "time_per_iteration": 2.8182172775268555 + }, + { + "auxiliary_loss_clip": 0.01058186, + "auxiliary_loss_mlp": 0.01002864, + "balance_loss_clip": 1.01339889, + "balance_loss_mlp": 1.00053978, + "epoch": 0.4099080141886611, + "flos": 59269234832640.0, + "grad_norm": 0.8160061416632566, + "language_loss": 0.65162539, + "learning_rate": 2.6677483367811947e-06, + "loss": 0.67223585, + "num_input_tokens_seen": 73687670, + "step": 3409, + "time_per_iteration": 3.4580254554748535 + }, + { + "auxiliary_loss_clip": 0.01181595, + "auxiliary_loss_mlp": 0.0102837, + "balance_loss_clip": 1.05122352, + "balance_loss_mlp": 1.01958406, + "epoch": 0.4100282570793002, + "flos": 21906443001600.0, + "grad_norm": 1.651109256615169, + "language_loss": 0.75769019, + "learning_rate": 2.6670140137254028e-06, + "loss": 0.77978981, + "num_input_tokens_seen": 73707145, + "step": 3410, + "time_per_iteration": 2.7895846366882324 + }, + { + "auxiliary_loss_clip": 0.01145138, + "auxiliary_loss_mlp": 0.01033763, + "balance_loss_clip": 1.04742706, + "balance_loss_mlp": 1.02462542, + "epoch": 0.4101484999699393, + "flos": 18551596631040.0, + "grad_norm": 2.412996903758167, + "language_loss": 0.89566439, + "learning_rate": 2.666279589484115e-06, + "loss": 0.91745341, + "num_input_tokens_seen": 73725045, + "step": 3411, + "time_per_iteration": 2.860830783843994 + }, + { + "auxiliary_loss_clip": 0.01147063, + "auxiliary_loss_mlp": 0.0103085, + "balance_loss_clip": 1.04436553, + "balance_loss_mlp": 1.02162969, + "epoch": 0.41026874286057835, + "flos": 19094529680640.0, + "grad_norm": 1.7913750130402366, + "language_loss": 0.81183589, + "learning_rate": 2.6655450641687435e-06, + "loss": 0.83361506, + "num_input_tokens_seen": 73742610, + "step": 3412, + "time_per_iteration": 2.9913318157196045 + }, + { + "auxiliary_loss_clip": 0.01187519, + "auxiliary_loss_mlp": 0.01039352, + "balance_loss_clip": 1.05391788, + "balance_loss_mlp": 1.03075755, + "epoch": 0.41038898575121746, + "flos": 31209568588800.0, + "grad_norm": 1.720993330341102, + "language_loss": 0.69401622, + "learning_rate": 2.664810437890715e-06, + "loss": 0.71628487, + "num_input_tokens_seen": 73764280, + "step": 3413, + "time_per_iteration": 2.748931407928467 + }, + { + "auxiliary_loss_clip": 0.01131773, + "auxiliary_loss_mlp": 0.01032988, + "balance_loss_clip": 1.05041194, + "balance_loss_mlp": 1.02491784, + "epoch": 0.41050922864185657, + "flos": 14355865895040.0, + "grad_norm": 2.0353880003428846, + "language_loss": 0.79739541, + "learning_rate": 2.6640757107614714e-06, + "loss": 0.81904304, + "num_input_tokens_seen": 73782375, + "step": 3414, + "time_per_iteration": 2.8774561882019043 + }, + { + "auxiliary_loss_clip": 0.01150622, + "auxiliary_loss_mlp": 0.01027935, + "balance_loss_clip": 1.04770589, + "balance_loss_mlp": 1.0186249, + "epoch": 0.4106294715324956, + "flos": 30956290813440.0, + "grad_norm": 2.1853025283206824, + "language_loss": 0.68590617, + "learning_rate": 2.6633408828924697e-06, + "loss": 0.70769179, + "num_input_tokens_seen": 73801240, + "step": 3415, + "time_per_iteration": 2.916193962097168 + }, + { + "auxiliary_loss_clip": 0.01164846, + "auxiliary_loss_mlp": 0.01035857, + "balance_loss_clip": 1.04873347, + "balance_loss_mlp": 1.0273335, + "epoch": 0.41074971442313474, + "flos": 24457321209600.0, + "grad_norm": 1.6570857217623152, + "language_loss": 0.69533664, + "learning_rate": 2.662605954395185e-06, + "loss": 0.71734369, + "num_input_tokens_seen": 73821200, + "step": 3416, + "time_per_iteration": 2.771695852279663 + }, + { + "auxiliary_loss_clip": 0.01180654, + "auxiliary_loss_mlp": 0.01030106, + "balance_loss_clip": 1.04966712, + "balance_loss_mlp": 1.02086103, + "epoch": 0.41086995731377385, + "flos": 21542991235200.0, + "grad_norm": 1.6858458811969474, + "language_loss": 0.83351743, + "learning_rate": 2.6618709253811027e-06, + "loss": 0.85562503, + "num_input_tokens_seen": 73840655, + "step": 3417, + "time_per_iteration": 2.753847599029541 + }, + { + "auxiliary_loss_clip": 0.01181674, + "auxiliary_loss_mlp": 0.01027886, + "balance_loss_clip": 1.05256581, + "balance_loss_mlp": 1.0200007, + "epoch": 0.4109902002044129, + "flos": 20702753314560.0, + "grad_norm": 1.824853248294846, + "language_loss": 0.87515444, + "learning_rate": 2.6611357959617277e-06, + "loss": 0.89725006, + "num_input_tokens_seen": 73860275, + "step": 3418, + "time_per_iteration": 3.546323299407959 + }, + { + "auxiliary_loss_clip": 0.01153421, + "auxiliary_loss_mlp": 0.01032606, + "balance_loss_clip": 1.04859626, + "balance_loss_mlp": 1.02318215, + "epoch": 0.411110443095052, + "flos": 18179992477440.0, + "grad_norm": 1.9687099726877146, + "language_loss": 0.90937519, + "learning_rate": 2.660400566248578e-06, + "loss": 0.93123543, + "num_input_tokens_seen": 73878400, + "step": 3419, + "time_per_iteration": 3.7194931507110596 + }, + { + "auxiliary_loss_clip": 0.01155797, + "auxiliary_loss_mlp": 0.01024948, + "balance_loss_clip": 1.04656422, + "balance_loss_mlp": 1.015113, + "epoch": 0.41123068598569107, + "flos": 14575244209920.0, + "grad_norm": 2.2871808660291912, + "language_loss": 0.66915745, + "learning_rate": 2.6596652363531876e-06, + "loss": 0.69096494, + "num_input_tokens_seen": 73894275, + "step": 3420, + "time_per_iteration": 3.823421001434326 + }, + { + "auxiliary_loss_clip": 0.01186836, + "auxiliary_loss_mlp": 0.01031897, + "balance_loss_clip": 1.0538336, + "balance_loss_mlp": 1.0238179, + "epoch": 0.4113509288763302, + "flos": 21177995184000.0, + "grad_norm": 2.5623162496332217, + "language_loss": 0.78135586, + "learning_rate": 2.6589298063871055e-06, + "loss": 0.80354321, + "num_input_tokens_seen": 73914450, + "step": 3421, + "time_per_iteration": 3.670707941055298 + }, + { + "auxiliary_loss_clip": 0.01189869, + "auxiliary_loss_mlp": 0.01033229, + "balance_loss_clip": 1.05516386, + "balance_loss_mlp": 1.02460396, + "epoch": 0.4114711717669693, + "flos": 18442212739200.0, + "grad_norm": 1.8525364219551756, + "language_loss": 0.69709694, + "learning_rate": 2.658194276461895e-06, + "loss": 0.71932793, + "num_input_tokens_seen": 73932375, + "step": 3422, + "time_per_iteration": 2.654918909072876 + }, + { + "auxiliary_loss_clip": 0.01164845, + "auxiliary_loss_mlp": 0.01031427, + "balance_loss_clip": 1.04526353, + "balance_loss_mlp": 1.02175903, + "epoch": 0.41159141465760835, + "flos": 27233395735680.0, + "grad_norm": 2.0060332482854726, + "language_loss": 0.6718756, + "learning_rate": 2.6574586466891368e-06, + "loss": 0.69383836, + "num_input_tokens_seen": 73952850, + "step": 3423, + "time_per_iteration": 2.8632009029388428 + }, + { + "auxiliary_loss_clip": 0.01167239, + "auxiliary_loss_mlp": 0.00885885, + "balance_loss_clip": 1.04900289, + "balance_loss_mlp": 1.00044072, + "epoch": 0.41171165754824746, + "flos": 20006876154240.0, + "grad_norm": 4.191459515421442, + "language_loss": 0.64479119, + "learning_rate": 2.6567229171804247e-06, + "loss": 0.66532242, + "num_input_tokens_seen": 73970735, + "step": 3424, + "time_per_iteration": 2.8254005908966064 + }, + { + "auxiliary_loss_clip": 0.01163134, + "auxiliary_loss_mlp": 0.01032868, + "balance_loss_clip": 1.04439235, + "balance_loss_mlp": 1.02346814, + "epoch": 0.41183190043888657, + "flos": 18004318035840.0, + "grad_norm": 3.0148072162006168, + "language_loss": 0.88005006, + "learning_rate": 2.655987088047368e-06, + "loss": 0.90201008, + "num_input_tokens_seen": 73989080, + "step": 3425, + "time_per_iteration": 2.7822389602661133 + }, + { + "auxiliary_loss_clip": 0.01162176, + "auxiliary_loss_mlp": 0.0103538, + "balance_loss_clip": 1.04761207, + "balance_loss_mlp": 1.02598047, + "epoch": 0.4119521433295256, + "flos": 27163370171520.0, + "grad_norm": 1.9886575130028545, + "language_loss": 0.7865696, + "learning_rate": 2.6552511594015912e-06, + "loss": 0.80854517, + "num_input_tokens_seen": 74009470, + "step": 3426, + "time_per_iteration": 2.848170042037964 + }, + { + "auxiliary_loss_clip": 0.01163335, + "auxiliary_loss_mlp": 0.01030405, + "balance_loss_clip": 1.04462779, + "balance_loss_mlp": 1.02070725, + "epoch": 0.41207238622016473, + "flos": 15122020014720.0, + "grad_norm": 1.955436387396674, + "language_loss": 0.85384911, + "learning_rate": 2.654515131354735e-06, + "loss": 0.87578654, + "num_input_tokens_seen": 74027735, + "step": 3427, + "time_per_iteration": 2.7728724479675293 + }, + { + "auxiliary_loss_clip": 0.01160285, + "auxiliary_loss_mlp": 0.0103467, + "balance_loss_clip": 1.05036783, + "balance_loss_mlp": 1.02610469, + "epoch": 0.41219262911080384, + "flos": 27052872958080.0, + "grad_norm": 4.003569990046205, + "language_loss": 0.8507483, + "learning_rate": 2.653779004018453e-06, + "loss": 0.87269783, + "num_input_tokens_seen": 74048300, + "step": 3428, + "time_per_iteration": 2.8254172801971436 + }, + { + "auxiliary_loss_clip": 0.01160048, + "auxiliary_loss_mlp": 0.01032788, + "balance_loss_clip": 1.04766846, + "balance_loss_mlp": 1.02319753, + "epoch": 0.4123128720014429, + "flos": 24686360282880.0, + "grad_norm": 1.9700948822723432, + "language_loss": 0.82860529, + "learning_rate": 2.653042777504417e-06, + "loss": 0.85053366, + "num_input_tokens_seen": 74070890, + "step": 3429, + "time_per_iteration": 2.857930898666382 + }, + { + "auxiliary_loss_clip": 0.01173886, + "auxiliary_loss_mlp": 0.01029914, + "balance_loss_clip": 1.05113316, + "balance_loss_mlp": 1.02120566, + "epoch": 0.412433114892082, + "flos": 26244774731520.0, + "grad_norm": 2.7723080225092684, + "language_loss": 0.79608291, + "learning_rate": 2.6523064519243105e-06, + "loss": 0.8181209, + "num_input_tokens_seen": 74090460, + "step": 3430, + "time_per_iteration": 2.8189878463745117 + }, + { + "auxiliary_loss_clip": 0.01178864, + "auxiliary_loss_mlp": 0.01032357, + "balance_loss_clip": 1.05196214, + "balance_loss_mlp": 1.02332735, + "epoch": 0.4125533577827211, + "flos": 21361031913600.0, + "grad_norm": 2.1552374438358606, + "language_loss": 0.78794777, + "learning_rate": 2.6515700273898333e-06, + "loss": 0.81006002, + "num_input_tokens_seen": 74108335, + "step": 3431, + "time_per_iteration": 2.7777364253997803 + }, + { + "auxiliary_loss_clip": 0.01154277, + "auxiliary_loss_mlp": 0.01030911, + "balance_loss_clip": 1.04770279, + "balance_loss_mlp": 1.02259028, + "epoch": 0.4126736006733602, + "flos": 26067556005120.0, + "grad_norm": 2.5032878117930477, + "language_loss": 0.69234699, + "learning_rate": 2.6508335040127018e-06, + "loss": 0.71419883, + "num_input_tokens_seen": 74128030, + "step": 3432, + "time_per_iteration": 2.780583143234253 + }, + { + "auxiliary_loss_clip": 0.01181105, + "auxiliary_loss_mlp": 0.01037187, + "balance_loss_clip": 1.05277765, + "balance_loss_mlp": 1.02892613, + "epoch": 0.4127938435639993, + "flos": 25666146541440.0, + "grad_norm": 1.637104582852654, + "language_loss": 0.76965559, + "learning_rate": 2.6500968819046446e-06, + "loss": 0.79183853, + "num_input_tokens_seen": 74148330, + "step": 3433, + "time_per_iteration": 2.8080694675445557 + }, + { + "auxiliary_loss_clip": 0.01145472, + "auxiliary_loss_mlp": 0.01029028, + "balance_loss_clip": 1.04406309, + "balance_loss_mlp": 1.02037382, + "epoch": 0.4129140864546384, + "flos": 17995914253440.0, + "grad_norm": 3.89810554854623, + "language_loss": 0.5897423, + "learning_rate": 2.649360161177408e-06, + "loss": 0.61148739, + "num_input_tokens_seen": 74163390, + "step": 3434, + "time_per_iteration": 2.755941867828369 + }, + { + "auxiliary_loss_clip": 0.01184791, + "auxiliary_loss_mlp": 0.01030598, + "balance_loss_clip": 1.05229282, + "balance_loss_mlp": 1.02215755, + "epoch": 0.41303432934527745, + "flos": 23732895715200.0, + "grad_norm": 2.049977741420584, + "language_loss": 0.73405033, + "learning_rate": 2.6486233419427504e-06, + "loss": 0.75620425, + "num_input_tokens_seen": 74183205, + "step": 3435, + "time_per_iteration": 2.739520311355591 + }, + { + "auxiliary_loss_clip": 0.01152937, + "auxiliary_loss_mlp": 0.01030917, + "balance_loss_clip": 1.05022657, + "balance_loss_mlp": 1.02173805, + "epoch": 0.41315457223591656, + "flos": 19755286318080.0, + "grad_norm": 2.798555628880798, + "language_loss": 0.75452423, + "learning_rate": 2.6478864243124484e-06, + "loss": 0.77636272, + "num_input_tokens_seen": 74202870, + "step": 3436, + "time_per_iteration": 2.7923126220703125 + }, + { + "auxiliary_loss_clip": 0.01177702, + "auxiliary_loss_mlp": 0.01029551, + "balance_loss_clip": 1.04887629, + "balance_loss_mlp": 1.02104545, + "epoch": 0.4132748151265556, + "flos": 20923316778240.0, + "grad_norm": 2.09654970951164, + "language_loss": 0.85205197, + "learning_rate": 2.6471494083982903e-06, + "loss": 0.87412453, + "num_input_tokens_seen": 74222255, + "step": 3437, + "time_per_iteration": 2.8091976642608643 + }, + { + "auxiliary_loss_clip": 0.01162405, + "auxiliary_loss_mlp": 0.01031566, + "balance_loss_clip": 1.04793429, + "balance_loss_mlp": 1.02170706, + "epoch": 0.4133950580171947, + "flos": 32232520016640.0, + "grad_norm": 1.86298020164411, + "language_loss": 0.74578679, + "learning_rate": 2.6464122943120818e-06, + "loss": 0.76772648, + "num_input_tokens_seen": 74242480, + "step": 3438, + "time_per_iteration": 2.8326473236083984 + }, + { + "auxiliary_loss_clip": 0.01158543, + "auxiliary_loss_mlp": 0.0103952, + "balance_loss_clip": 1.05150056, + "balance_loss_mlp": 1.02989972, + "epoch": 0.41351530090783384, + "flos": 23292487059840.0, + "grad_norm": 3.2687742714125023, + "language_loss": 0.82182109, + "learning_rate": 2.645675082165642e-06, + "loss": 0.84380174, + "num_input_tokens_seen": 74258690, + "step": 3439, + "time_per_iteration": 2.833967685699463 + }, + { + "auxiliary_loss_clip": 0.01170641, + "auxiliary_loss_mlp": 0.01040462, + "balance_loss_clip": 1.0537796, + "balance_loss_mlp": 1.03096092, + "epoch": 0.4136355437984729, + "flos": 25593571111680.0, + "grad_norm": 2.050566058912175, + "language_loss": 0.75416923, + "learning_rate": 2.644937772070806e-06, + "loss": 0.77628034, + "num_input_tokens_seen": 74277135, + "step": 3440, + "time_per_iteration": 2.782886505126953 + }, + { + "auxiliary_loss_clip": 0.01188641, + "auxiliary_loss_mlp": 0.01030259, + "balance_loss_clip": 1.05443275, + "balance_loss_mlp": 1.02134812, + "epoch": 0.413755786689112, + "flos": 19828615933440.0, + "grad_norm": 2.76011040781316, + "language_loss": 0.83727527, + "learning_rate": 2.6442003641394225e-06, + "loss": 0.85946417, + "num_input_tokens_seen": 74294730, + "step": 3441, + "time_per_iteration": 2.6686666011810303 + }, + { + "auxiliary_loss_clip": 0.01164791, + "auxiliary_loss_mlp": 0.01030182, + "balance_loss_clip": 1.0480094, + "balance_loss_mlp": 1.02105069, + "epoch": 0.4138760295797511, + "flos": 26870446759680.0, + "grad_norm": 1.4267537094093696, + "language_loss": 0.838956, + "learning_rate": 2.643462858483356e-06, + "loss": 0.86090571, + "num_input_tokens_seen": 74315015, + "step": 3442, + "time_per_iteration": 2.85701584815979 + }, + { + "auxiliary_loss_clip": 0.01145004, + "auxiliary_loss_mlp": 0.01041711, + "balance_loss_clip": 1.04631341, + "balance_loss_mlp": 1.03143525, + "epoch": 0.41399627247039017, + "flos": 16399254798720.0, + "grad_norm": 1.8844799246836654, + "language_loss": 0.72726572, + "learning_rate": 2.6427252552144856e-06, + "loss": 0.74913281, + "num_input_tokens_seen": 74333665, + "step": 3443, + "time_per_iteration": 3.73665189743042 + }, + { + "auxiliary_loss_clip": 0.01188354, + "auxiliary_loss_mlp": 0.01035254, + "balance_loss_clip": 1.05346799, + "balance_loss_mlp": 1.02531242, + "epoch": 0.4141165153610293, + "flos": 22930220442240.0, + "grad_norm": 21.277376202241992, + "language_loss": 0.74908018, + "learning_rate": 2.6419875544447044e-06, + "loss": 0.77131629, + "num_input_tokens_seen": 74355065, + "step": 3444, + "time_per_iteration": 3.6622204780578613 + }, + { + "auxiliary_loss_clip": 0.01192405, + "auxiliary_loss_mlp": 0.01030089, + "balance_loss_clip": 1.0556761, + "balance_loss_mlp": 1.02080274, + "epoch": 0.4142367582516684, + "flos": 25192556697600.0, + "grad_norm": 1.5702795566404832, + "language_loss": 0.71436596, + "learning_rate": 2.6412497562859218e-06, + "loss": 0.73659098, + "num_input_tokens_seen": 74376345, + "step": 3445, + "time_per_iteration": 2.7265729904174805 + }, + { + "auxiliary_loss_clip": 0.01185987, + "auxiliary_loss_mlp": 0.01036503, + "balance_loss_clip": 1.05429125, + "balance_loss_mlp": 1.02715135, + "epoch": 0.41435700114230745, + "flos": 21690476478720.0, + "grad_norm": 4.658681480120656, + "language_loss": 0.75863367, + "learning_rate": 2.6405118608500617e-06, + "loss": 0.78085858, + "num_input_tokens_seen": 74395170, + "step": 3446, + "time_per_iteration": 3.6320621967315674 + }, + { + "auxiliary_loss_clip": 0.01155272, + "auxiliary_loss_mlp": 0.01029961, + "balance_loss_clip": 1.05373621, + "balance_loss_mlp": 1.02138352, + "epoch": 0.41447724403294656, + "flos": 25995160143360.0, + "grad_norm": 1.8489262546891014, + "language_loss": 0.81370378, + "learning_rate": 2.6397738682490613e-06, + "loss": 0.83555609, + "num_input_tokens_seen": 74416070, + "step": 3447, + "time_per_iteration": 3.8783206939697266 + }, + { + "auxiliary_loss_clip": 0.01189125, + "auxiliary_loss_mlp": 0.01036308, + "balance_loss_clip": 1.05413294, + "balance_loss_mlp": 1.02752256, + "epoch": 0.41459748692358567, + "flos": 18259678800000.0, + "grad_norm": 1.6137155939184196, + "language_loss": 0.75267291, + "learning_rate": 2.6390357785948734e-06, + "loss": 0.7749272, + "num_input_tokens_seen": 74433185, + "step": 3448, + "time_per_iteration": 2.728381633758545 + }, + { + "auxiliary_loss_clip": 0.01180366, + "auxiliary_loss_mlp": 0.01030998, + "balance_loss_clip": 1.05402875, + "balance_loss_mlp": 1.02166343, + "epoch": 0.4147177298142247, + "flos": 24168456034560.0, + "grad_norm": 2.3950583122219786, + "language_loss": 0.80206907, + "learning_rate": 2.6382975919994667e-06, + "loss": 0.82418275, + "num_input_tokens_seen": 74453760, + "step": 3449, + "time_per_iteration": 2.738464117050171 + }, + { + "auxiliary_loss_clip": 0.01170515, + "auxiliary_loss_mlp": 0.01026046, + "balance_loss_clip": 1.04959345, + "balance_loss_mlp": 1.01798177, + "epoch": 0.41483797270486383, + "flos": 20084659056000.0, + "grad_norm": 1.5838060590604006, + "language_loss": 0.72885191, + "learning_rate": 2.637559308574822e-06, + "loss": 0.75081754, + "num_input_tokens_seen": 74473505, + "step": 3450, + "time_per_iteration": 2.7968626022338867 + }, + { + "auxiliary_loss_clip": 0.01189406, + "auxiliary_loss_mlp": 0.01037147, + "balance_loss_clip": 1.05505395, + "balance_loss_mlp": 1.02782464, + "epoch": 0.4149582155955029, + "flos": 30081040110720.0, + "grad_norm": 2.166965794298013, + "language_loss": 0.71093059, + "learning_rate": 2.6368209284329376e-06, + "loss": 0.73319614, + "num_input_tokens_seen": 74494135, + "step": 3451, + "time_per_iteration": 2.7575838565826416 + }, + { + "auxiliary_loss_clip": 0.01179281, + "auxiliary_loss_mlp": 0.01029269, + "balance_loss_clip": 1.05123997, + "balance_loss_mlp": 1.01973796, + "epoch": 0.415078458486142, + "flos": 16764394504320.0, + "grad_norm": 2.09784836052553, + "language_loss": 0.75477779, + "learning_rate": 2.636082451685825e-06, + "loss": 0.77686328, + "num_input_tokens_seen": 74512335, + "step": 3452, + "time_per_iteration": 2.834059000015259 + }, + { + "auxiliary_loss_clip": 0.01173382, + "auxiliary_loss_mlp": 0.01032048, + "balance_loss_clip": 1.05411053, + "balance_loss_mlp": 1.02383399, + "epoch": 0.4151987013767811, + "flos": 26033692458240.0, + "grad_norm": 1.708108125881818, + "language_loss": 0.86023188, + "learning_rate": 2.6353438784455094e-06, + "loss": 0.88228607, + "num_input_tokens_seen": 74535620, + "step": 3453, + "time_per_iteration": 2.8099701404571533 + }, + { + "auxiliary_loss_clip": 0.01165159, + "auxiliary_loss_mlp": 0.01039236, + "balance_loss_clip": 1.05075431, + "balance_loss_mlp": 1.0299139, + "epoch": 0.41531894426742016, + "flos": 24608002763520.0, + "grad_norm": 3.1874464035748256, + "language_loss": 0.71546698, + "learning_rate": 2.6346052088240326e-06, + "loss": 0.73751092, + "num_input_tokens_seen": 74555140, + "step": 3454, + "time_per_iteration": 2.7786896228790283 + }, + { + "auxiliary_loss_clip": 0.01171065, + "auxiliary_loss_mlp": 0.01034054, + "balance_loss_clip": 1.05038834, + "balance_loss_mlp": 1.02523255, + "epoch": 0.4154391871580593, + "flos": 14975791747200.0, + "grad_norm": 1.9981639300885865, + "language_loss": 0.76805747, + "learning_rate": 2.63386644293345e-06, + "loss": 0.79010868, + "num_input_tokens_seen": 74571485, + "step": 3455, + "time_per_iteration": 2.8689825534820557 + }, + { + "auxiliary_loss_clip": 0.01155772, + "auxiliary_loss_mlp": 0.0102943, + "balance_loss_clip": 1.04601884, + "balance_loss_mlp": 1.02089465, + "epoch": 0.4155594300486984, + "flos": 14647173194880.0, + "grad_norm": 2.9327238249589183, + "language_loss": 0.82784808, + "learning_rate": 2.633127580885833e-06, + "loss": 0.84970009, + "num_input_tokens_seen": 74585985, + "step": 3456, + "time_per_iteration": 2.82867169380188 + }, + { + "auxiliary_loss_clip": 0.01189462, + "auxiliary_loss_mlp": 0.01035059, + "balance_loss_clip": 1.05667353, + "balance_loss_mlp": 1.02584457, + "epoch": 0.41567967293933744, + "flos": 29497276275840.0, + "grad_norm": 2.0852023836528355, + "language_loss": 0.64810658, + "learning_rate": 2.632388622793265e-06, + "loss": 0.6703518, + "num_input_tokens_seen": 74605140, + "step": 3457, + "time_per_iteration": 2.710434675216675 + }, + { + "auxiliary_loss_clip": 0.01179865, + "auxiliary_loss_mlp": 0.01031024, + "balance_loss_clip": 1.05397785, + "balance_loss_mlp": 1.022614, + "epoch": 0.41579991582997655, + "flos": 19238387650560.0, + "grad_norm": 5.075800335348857, + "language_loss": 0.68111086, + "learning_rate": 2.6316495687678457e-06, + "loss": 0.70321977, + "num_input_tokens_seen": 74623790, + "step": 3458, + "time_per_iteration": 2.75117826461792 + }, + { + "auxiliary_loss_clip": 0.01141795, + "auxiliary_loss_mlp": 0.0102811, + "balance_loss_clip": 1.04567313, + "balance_loss_mlp": 1.01887131, + "epoch": 0.41592015872061566, + "flos": 24462061804800.0, + "grad_norm": 2.6316136846460765, + "language_loss": 0.76886058, + "learning_rate": 2.6309104189216887e-06, + "loss": 0.79055965, + "num_input_tokens_seen": 74641355, + "step": 3459, + "time_per_iteration": 2.88014817237854 + }, + { + "auxiliary_loss_clip": 0.01146594, + "auxiliary_loss_mlp": 0.00886965, + "balance_loss_clip": 1.04624438, + "balance_loss_mlp": 1.00048912, + "epoch": 0.4160404016112547, + "flos": 20775651966720.0, + "grad_norm": 2.325432067037071, + "language_loss": 0.74755466, + "learning_rate": 2.630171173366923e-06, + "loss": 0.76789027, + "num_input_tokens_seen": 74657155, + "step": 3460, + "time_per_iteration": 2.828230142593384 + }, + { + "auxiliary_loss_clip": 0.01148856, + "auxiliary_loss_mlp": 0.01030014, + "balance_loss_clip": 1.04610384, + "balance_loss_mlp": 1.02090693, + "epoch": 0.41616064450189383, + "flos": 13916462820480.0, + "grad_norm": 2.425527109402251, + "language_loss": 0.74314845, + "learning_rate": 2.629431832215691e-06, + "loss": 0.76493716, + "num_input_tokens_seen": 74671960, + "step": 3461, + "time_per_iteration": 2.821237564086914 + }, + { + "auxiliary_loss_clip": 0.0116545, + "auxiliary_loss_mlp": 0.0102637, + "balance_loss_clip": 1.05140066, + "balance_loss_mlp": 1.01733959, + "epoch": 0.41628088739253294, + "flos": 20010826650240.0, + "grad_norm": 2.1466754286831904, + "language_loss": 0.87380755, + "learning_rate": 2.628692395580151e-06, + "loss": 0.89572573, + "num_input_tokens_seen": 74692050, + "step": 3462, + "time_per_iteration": 2.7949678897857666 + }, + { + "auxiliary_loss_clip": 0.01117371, + "auxiliary_loss_mlp": 0.01034055, + "balance_loss_clip": 1.03981566, + "balance_loss_mlp": 1.02464294, + "epoch": 0.416401130283172, + "flos": 29168801377920.0, + "grad_norm": 2.432176715628492, + "language_loss": 0.79015326, + "learning_rate": 2.6279528635724747e-06, + "loss": 0.8116675, + "num_input_tokens_seen": 74712205, + "step": 3463, + "time_per_iteration": 2.9356062412261963 + }, + { + "auxiliary_loss_clip": 0.01178697, + "auxiliary_loss_mlp": 0.01035063, + "balance_loss_clip": 1.05054486, + "balance_loss_mlp": 1.02525842, + "epoch": 0.4165213731738111, + "flos": 16246813478400.0, + "grad_norm": 2.5671274125124146, + "language_loss": 0.77965319, + "learning_rate": 2.627213236304848e-06, + "loss": 0.80179083, + "num_input_tokens_seen": 74729005, + "step": 3464, + "time_per_iteration": 2.8654723167419434 + }, + { + "auxiliary_loss_clip": 0.01181103, + "auxiliary_loss_mlp": 0.0103156, + "balance_loss_clip": 1.05242956, + "balance_loss_mlp": 1.02264857, + "epoch": 0.4166416160644502, + "flos": 33765438787200.0, + "grad_norm": 2.199780349374333, + "language_loss": 0.70910889, + "learning_rate": 2.626473513889472e-06, + "loss": 0.73123556, + "num_input_tokens_seen": 74751385, + "step": 3465, + "time_per_iteration": 2.827399492263794 + }, + { + "auxiliary_loss_clip": 0.01171817, + "auxiliary_loss_mlp": 0.01029841, + "balance_loss_clip": 1.05207729, + "balance_loss_mlp": 1.02071559, + "epoch": 0.41676185895508927, + "flos": 20917498775040.0, + "grad_norm": 1.785948247242991, + "language_loss": 0.82617164, + "learning_rate": 2.625733696438562e-06, + "loss": 0.84818816, + "num_input_tokens_seen": 74768890, + "step": 3466, + "time_per_iteration": 2.7626454830169678 + }, + { + "auxiliary_loss_clip": 0.01167952, + "auxiliary_loss_mlp": 0.01039969, + "balance_loss_clip": 1.05162776, + "balance_loss_mlp": 1.03076053, + "epoch": 0.4168821018457284, + "flos": 18406122549120.0, + "grad_norm": 1.889046486747462, + "language_loss": 0.75080192, + "learning_rate": 2.6249937840643476e-06, + "loss": 0.77288109, + "num_input_tokens_seen": 74787195, + "step": 3467, + "time_per_iteration": 2.766397714614868 + }, + { + "auxiliary_loss_clip": 0.01190803, + "auxiliary_loss_mlp": 0.00885934, + "balance_loss_clip": 1.05671585, + "balance_loss_mlp": 1.00059009, + "epoch": 0.41700234473636744, + "flos": 18698399516160.0, + "grad_norm": 2.2789440592670442, + "language_loss": 0.66792685, + "learning_rate": 2.6242537768790733e-06, + "loss": 0.68869424, + "num_input_tokens_seen": 74806350, + "step": 3468, + "time_per_iteration": 2.77502703666687 + }, + { + "auxiliary_loss_clip": 0.01177527, + "auxiliary_loss_mlp": 0.01034843, + "balance_loss_clip": 1.05271339, + "balance_loss_mlp": 1.02521706, + "epoch": 0.41712258762700655, + "flos": 31033283616000.0, + "grad_norm": 1.928153472579287, + "language_loss": 0.68505406, + "learning_rate": 2.6235136749949975e-06, + "loss": 0.70717782, + "num_input_tokens_seen": 74829800, + "step": 3469, + "time_per_iteration": 4.701595306396484 + }, + { + "auxiliary_loss_clip": 0.01189327, + "auxiliary_loss_mlp": 0.01036413, + "balance_loss_clip": 1.05521393, + "balance_loss_mlp": 1.02739525, + "epoch": 0.41724283051764566, + "flos": 35914763877120.0, + "grad_norm": 2.1646404036451554, + "language_loss": 0.61410314, + "learning_rate": 2.6227734785243924e-06, + "loss": 0.63636053, + "num_input_tokens_seen": 74849760, + "step": 3470, + "time_per_iteration": 2.798039436340332 + }, + { + "auxiliary_loss_clip": 0.01133805, + "auxiliary_loss_mlp": 0.01032123, + "balance_loss_clip": 1.04453456, + "balance_loss_mlp": 1.0229913, + "epoch": 0.4173630734082847, + "flos": 25333649320320.0, + "grad_norm": 2.037055944610711, + "language_loss": 0.78975809, + "learning_rate": 2.6220331875795466e-06, + "loss": 0.81141728, + "num_input_tokens_seen": 74869110, + "step": 3471, + "time_per_iteration": 2.9979169368743896 + }, + { + "auxiliary_loss_clip": 0.01173785, + "auxiliary_loss_mlp": 0.01031154, + "balance_loss_clip": 1.05080855, + "balance_loss_mlp": 1.02155185, + "epoch": 0.4174833162989238, + "flos": 26685398868480.0, + "grad_norm": 1.7316756220574139, + "language_loss": 0.74963105, + "learning_rate": 2.62129280227276e-06, + "loss": 0.77168047, + "num_input_tokens_seen": 74889110, + "step": 3472, + "time_per_iteration": 3.6959900856018066 + }, + { + "auxiliary_loss_clip": 0.01184854, + "auxiliary_loss_mlp": 0.01041055, + "balance_loss_clip": 1.0543505, + "balance_loss_mlp": 1.03242445, + "epoch": 0.41760355918956293, + "flos": 74739584010240.0, + "grad_norm": 2.865922322208941, + "language_loss": 0.68474042, + "learning_rate": 2.62055232271635e-06, + "loss": 0.70699954, + "num_input_tokens_seen": 74916260, + "step": 3473, + "time_per_iteration": 4.011932373046875 + }, + { + "auxiliary_loss_clip": 0.0114746, + "auxiliary_loss_mlp": 0.01035608, + "balance_loss_clip": 1.04524088, + "balance_loss_mlp": 1.02673292, + "epoch": 0.417723802080202, + "flos": 14317513148160.0, + "grad_norm": 1.995051965074562, + "language_loss": 0.8772825, + "learning_rate": 2.619811749022646e-06, + "loss": 0.89911318, + "num_input_tokens_seen": 74931570, + "step": 3474, + "time_per_iteration": 2.8595476150512695 + }, + { + "auxiliary_loss_clip": 0.01184617, + "auxiliary_loss_mlp": 0.01029655, + "balance_loss_clip": 1.05504501, + "balance_loss_mlp": 1.0199753, + "epoch": 0.4178440449708411, + "flos": 14643797316480.0, + "grad_norm": 2.236512416549094, + "language_loss": 0.71793067, + "learning_rate": 2.6190710813039917e-06, + "loss": 0.74007338, + "num_input_tokens_seen": 74944695, + "step": 3475, + "time_per_iteration": 2.6677064895629883 + }, + { + "auxiliary_loss_clip": 0.01146703, + "auxiliary_loss_mlp": 0.00887378, + "balance_loss_clip": 1.0452739, + "balance_loss_mlp": 1.00038123, + "epoch": 0.4179642878614802, + "flos": 21507296094720.0, + "grad_norm": 2.5950520810991606, + "language_loss": 0.8362661, + "learning_rate": 2.618330319672747e-06, + "loss": 0.85660696, + "num_input_tokens_seen": 74964115, + "step": 3476, + "time_per_iteration": 2.8968892097473145 + }, + { + "auxiliary_loss_clip": 0.01190156, + "auxiliary_loss_mlp": 0.01039815, + "balance_loss_clip": 1.0550673, + "balance_loss_mlp": 1.03070092, + "epoch": 0.41808453075211927, + "flos": 18441997257600.0, + "grad_norm": 3.8329686779078647, + "language_loss": 0.91797167, + "learning_rate": 2.617589464241284e-06, + "loss": 0.94027138, + "num_input_tokens_seen": 74978515, + "step": 3477, + "time_per_iteration": 2.6962430477142334 + }, + { + "auxiliary_loss_clip": 0.01160923, + "auxiliary_loss_mlp": 0.01029249, + "balance_loss_clip": 1.04962945, + "balance_loss_mlp": 1.02107739, + "epoch": 0.4182047736427584, + "flos": 20301020628480.0, + "grad_norm": 1.988720867219546, + "language_loss": 0.74547815, + "learning_rate": 2.6168485151219914e-06, + "loss": 0.76737988, + "num_input_tokens_seen": 74998135, + "step": 3478, + "time_per_iteration": 2.925189733505249 + }, + { + "auxiliary_loss_clip": 0.01182879, + "auxiliary_loss_mlp": 0.01033865, + "balance_loss_clip": 1.05586958, + "balance_loss_mlp": 1.02450109, + "epoch": 0.4183250165333975, + "flos": 18876623823360.0, + "grad_norm": 4.039270032104503, + "language_loss": 0.71569645, + "learning_rate": 2.616107472427269e-06, + "loss": 0.7378639, + "num_input_tokens_seen": 75012830, + "step": 3479, + "time_per_iteration": 2.733841896057129 + }, + { + "auxiliary_loss_clip": 0.01183345, + "auxiliary_loss_mlp": 0.01031152, + "balance_loss_clip": 1.05129409, + "balance_loss_mlp": 1.02127528, + "epoch": 0.41844525942403654, + "flos": 17740050698880.0, + "grad_norm": 2.295784684263393, + "language_loss": 0.76306415, + "learning_rate": 2.615366336269533e-06, + "loss": 0.78520918, + "num_input_tokens_seen": 75026495, + "step": 3480, + "time_per_iteration": 2.713268280029297 + }, + { + "auxiliary_loss_clip": 0.01194893, + "auxiliary_loss_mlp": 0.0103886, + "balance_loss_clip": 1.05636621, + "balance_loss_mlp": 1.02938282, + "epoch": 0.41856550231467565, + "flos": 18361377181440.0, + "grad_norm": 3.2097517722966735, + "language_loss": 0.80561018, + "learning_rate": 2.6146251067612126e-06, + "loss": 0.82794768, + "num_input_tokens_seen": 75041970, + "step": 3481, + "time_per_iteration": 2.6654818058013916 + }, + { + "auxiliary_loss_clip": 0.01176737, + "auxiliary_loss_mlp": 0.01030866, + "balance_loss_clip": 1.05428052, + "balance_loss_mlp": 1.02196121, + "epoch": 0.41868574520531476, + "flos": 22781801445120.0, + "grad_norm": 1.5905387725988678, + "language_loss": 0.8223629, + "learning_rate": 2.6138837840147525e-06, + "loss": 0.84443891, + "num_input_tokens_seen": 75061005, + "step": 3482, + "time_per_iteration": 2.8027100563049316 + }, + { + "auxiliary_loss_clip": 0.01157433, + "auxiliary_loss_mlp": 0.01034859, + "balance_loss_clip": 1.04914892, + "balance_loss_mlp": 1.02604914, + "epoch": 0.4188059880959538, + "flos": 13699167494400.0, + "grad_norm": 2.1606053163369316, + "language_loss": 0.7635892, + "learning_rate": 2.6131423681426103e-06, + "loss": 0.78551209, + "num_input_tokens_seen": 75076920, + "step": 3483, + "time_per_iteration": 2.7787322998046875 + }, + { + "auxiliary_loss_clip": 0.01189685, + "auxiliary_loss_mlp": 0.01028812, + "balance_loss_clip": 1.05605817, + "balance_loss_mlp": 1.02023149, + "epoch": 0.41892623098659293, + "flos": 37818281220480.0, + "grad_norm": 1.6353792203790831, + "language_loss": 0.72745669, + "learning_rate": 2.6124008592572587e-06, + "loss": 0.74964166, + "num_input_tokens_seen": 75100905, + "step": 3484, + "time_per_iteration": 2.7984697818756104 + }, + { + "auxiliary_loss_clip": 0.01192829, + "auxiliary_loss_mlp": 0.01029343, + "balance_loss_clip": 1.05490828, + "balance_loss_mlp": 1.01968718, + "epoch": 0.419046473877232, + "flos": 23258874908160.0, + "grad_norm": 2.1835602728814516, + "language_loss": 0.81610203, + "learning_rate": 2.6116592574711835e-06, + "loss": 0.83832377, + "num_input_tokens_seen": 75119205, + "step": 3485, + "time_per_iteration": 2.744903087615967 + }, + { + "auxiliary_loss_clip": 0.01194092, + "auxiliary_loss_mlp": 0.01034629, + "balance_loss_clip": 1.05700517, + "balance_loss_mlp": 1.02478194, + "epoch": 0.4191667167678711, + "flos": 20741034234240.0, + "grad_norm": 1.9453242184106272, + "language_loss": 0.8397699, + "learning_rate": 2.6109175628968853e-06, + "loss": 0.86205709, + "num_input_tokens_seen": 75138970, + "step": 3486, + "time_per_iteration": 2.763359546661377 + }, + { + "auxiliary_loss_clip": 0.01168501, + "auxiliary_loss_mlp": 0.01032076, + "balance_loss_clip": 1.04903412, + "balance_loss_mlp": 1.02308726, + "epoch": 0.4192869596585102, + "flos": 23586416052480.0, + "grad_norm": 2.6190897372663127, + "language_loss": 0.8304438, + "learning_rate": 2.610175775646878e-06, + "loss": 0.85244954, + "num_input_tokens_seen": 75157550, + "step": 3487, + "time_per_iteration": 2.754126787185669 + }, + { + "auxiliary_loss_clip": 0.01166236, + "auxiliary_loss_mlp": 0.01029293, + "balance_loss_clip": 1.04892159, + "balance_loss_mlp": 1.01988173, + "epoch": 0.41940720254914926, + "flos": 25081269384960.0, + "grad_norm": 1.9382044615263538, + "language_loss": 0.73169446, + "learning_rate": 2.6094338958336907e-06, + "loss": 0.75364971, + "num_input_tokens_seen": 75176220, + "step": 3488, + "time_per_iteration": 2.731708526611328 + }, + { + "auxiliary_loss_clip": 0.01168547, + "auxiliary_loss_mlp": 0.01025867, + "balance_loss_clip": 1.05140817, + "balance_loss_mlp": 1.01736736, + "epoch": 0.41952744543978837, + "flos": 15554132628480.0, + "grad_norm": 2.3193744917715318, + "language_loss": 0.82499504, + "learning_rate": 2.608691923569867e-06, + "loss": 0.84693915, + "num_input_tokens_seen": 75193095, + "step": 3489, + "time_per_iteration": 2.7317147254943848 + }, + { + "auxiliary_loss_clip": 0.01185411, + "auxiliary_loss_mlp": 0.01037168, + "balance_loss_clip": 1.05757618, + "balance_loss_mlp": 1.02809048, + "epoch": 0.4196476883304275, + "flos": 24644775312000.0, + "grad_norm": 1.9517286944296572, + "language_loss": 0.76026785, + "learning_rate": 2.6079498589679616e-06, + "loss": 0.78249365, + "num_input_tokens_seen": 75214185, + "step": 3490, + "time_per_iteration": 2.8000471591949463 + }, + { + "auxiliary_loss_clip": 0.01136626, + "auxiliary_loss_mlp": 0.01040118, + "balance_loss_clip": 1.04528511, + "balance_loss_mlp": 1.02950215, + "epoch": 0.41976793122106654, + "flos": 24531333183360.0, + "grad_norm": 1.8003847277952345, + "language_loss": 0.7638973, + "learning_rate": 2.6072077021405465e-06, + "loss": 0.78566468, + "num_input_tokens_seen": 75233020, + "step": 3491, + "time_per_iteration": 3.0541625022888184 + }, + { + "auxiliary_loss_clip": 0.01165492, + "auxiliary_loss_mlp": 0.01036117, + "balance_loss_clip": 1.04830801, + "balance_loss_mlp": 1.02783167, + "epoch": 0.41988817411170565, + "flos": 21175301664000.0, + "grad_norm": 2.403215651876499, + "language_loss": 0.69297767, + "learning_rate": 2.6064654532002054e-06, + "loss": 0.71499372, + "num_input_tokens_seen": 75252030, + "step": 3492, + "time_per_iteration": 2.8564975261688232 + }, + { + "auxiliary_loss_clip": 0.01189885, + "auxiliary_loss_mlp": 0.01033051, + "balance_loss_clip": 1.0548296, + "balance_loss_mlp": 1.02415752, + "epoch": 0.42000841700234476, + "flos": 31649402626560.0, + "grad_norm": 1.5249270306526923, + "language_loss": 0.75593829, + "learning_rate": 2.6057231122595375e-06, + "loss": 0.77816761, + "num_input_tokens_seen": 75273340, + "step": 3493, + "time_per_iteration": 2.7833285331726074 + }, + { + "auxiliary_loss_clip": 0.01169798, + "auxiliary_loss_mlp": 0.01031952, + "balance_loss_clip": 1.04867554, + "balance_loss_mlp": 1.02337468, + "epoch": 0.4201286598929838, + "flos": 21281525159040.0, + "grad_norm": 1.5602849620372212, + "language_loss": 0.72913432, + "learning_rate": 2.604980679431154e-06, + "loss": 0.7511518, + "num_input_tokens_seen": 75291580, + "step": 3494, + "time_per_iteration": 2.7500853538513184 + }, + { + "auxiliary_loss_clip": 0.0118007, + "auxiliary_loss_mlp": 0.01032965, + "balance_loss_clip": 1.0496161, + "balance_loss_mlp": 1.02423322, + "epoch": 0.4202489027836229, + "flos": 18546532813440.0, + "grad_norm": 2.245721286727286, + "language_loss": 0.74832809, + "learning_rate": 2.604238154827684e-06, + "loss": 0.77045846, + "num_input_tokens_seen": 75308205, + "step": 3495, + "time_per_iteration": 4.648188829421997 + }, + { + "auxiliary_loss_clip": 0.01181275, + "auxiliary_loss_mlp": 0.01027294, + "balance_loss_clip": 1.05316043, + "balance_loss_mlp": 1.0190382, + "epoch": 0.42036914567426203, + "flos": 19317643009920.0, + "grad_norm": 1.8985352276070433, + "language_loss": 0.72748172, + "learning_rate": 2.6034955385617656e-06, + "loss": 0.74956739, + "num_input_tokens_seen": 75326535, + "step": 3496, + "time_per_iteration": 2.75374436378479 + }, + { + "auxiliary_loss_clip": 0.0106375, + "auxiliary_loss_mlp": 0.01003183, + "balance_loss_clip": 1.0175966, + "balance_loss_mlp": 1.00122821, + "epoch": 0.4204893885649011, + "flos": 67842942935040.0, + "grad_norm": 0.7196440976374074, + "language_loss": 0.61634916, + "learning_rate": 2.6027528307460544e-06, + "loss": 0.63701844, + "num_input_tokens_seen": 75390540, + "step": 3497, + "time_per_iteration": 3.441068410873413 + }, + { + "auxiliary_loss_clip": 0.0119047, + "auxiliary_loss_mlp": 0.01030122, + "balance_loss_clip": 1.0551219, + "balance_loss_mlp": 1.02103841, + "epoch": 0.4206096314555402, + "flos": 21908777385600.0, + "grad_norm": 1.7833879889316957, + "language_loss": 0.86210376, + "learning_rate": 2.602010031493217e-06, + "loss": 0.88430965, + "num_input_tokens_seen": 75408770, + "step": 3498, + "time_per_iteration": 2.7160723209381104 + }, + { + "auxiliary_loss_clip": 0.01153455, + "auxiliary_loss_mlp": 0.01030239, + "balance_loss_clip": 1.04834461, + "balance_loss_mlp": 1.02173352, + "epoch": 0.42072987434617926, + "flos": 29278185269760.0, + "grad_norm": 2.11156154511677, + "language_loss": 0.87004411, + "learning_rate": 2.6012671409159367e-06, + "loss": 0.89188111, + "num_input_tokens_seen": 75430105, + "step": 3499, + "time_per_iteration": 4.573685884475708 + }, + { + "auxiliary_loss_clip": 0.01163687, + "auxiliary_loss_mlp": 0.01031181, + "balance_loss_clip": 1.05091119, + "balance_loss_mlp": 1.02204347, + "epoch": 0.42085011723681837, + "flos": 27600726170880.0, + "grad_norm": 1.7314399592930296, + "language_loss": 0.81363618, + "learning_rate": 2.6005241591269097e-06, + "loss": 0.83558494, + "num_input_tokens_seen": 75449475, + "step": 3500, + "time_per_iteration": 2.7956936359405518 + }, + { + "auxiliary_loss_clip": 0.01156615, + "auxiliary_loss_mlp": 0.01028647, + "balance_loss_clip": 1.05106902, + "balance_loss_mlp": 1.02084458, + "epoch": 0.4209703601274575, + "flos": 27818632028160.0, + "grad_norm": 1.6701189991969674, + "language_loss": 0.80030215, + "learning_rate": 2.5997810862388454e-06, + "loss": 0.82215476, + "num_input_tokens_seen": 75469315, + "step": 3501, + "time_per_iteration": 2.921941041946411 + }, + { + "auxiliary_loss_clip": 0.01168104, + "auxiliary_loss_mlp": 0.0103281, + "balance_loss_clip": 1.04780102, + "balance_loss_mlp": 1.02388668, + "epoch": 0.42109060301809653, + "flos": 27525529048320.0, + "grad_norm": 2.3831837512930982, + "language_loss": 0.75860441, + "learning_rate": 2.599037922364467e-06, + "loss": 0.78061354, + "num_input_tokens_seen": 75488215, + "step": 3502, + "time_per_iteration": 2.9040775299072266 + }, + { + "auxiliary_loss_clip": 0.01154162, + "auxiliary_loss_mlp": 0.01032176, + "balance_loss_clip": 1.05055761, + "balance_loss_mlp": 1.02317595, + "epoch": 0.42121084590873564, + "flos": 29314275459840.0, + "grad_norm": 2.3601221000789887, + "language_loss": 0.75057364, + "learning_rate": 2.5982946676165112e-06, + "loss": 0.77243704, + "num_input_tokens_seen": 75507985, + "step": 3503, + "time_per_iteration": 2.846322774887085 + }, + { + "auxiliary_loss_clip": 0.01060543, + "auxiliary_loss_mlp": 0.01006544, + "balance_loss_clip": 1.02132583, + "balance_loss_mlp": 1.00427949, + "epoch": 0.42133108879937475, + "flos": 67398835178880.0, + "grad_norm": 0.7280004763637116, + "language_loss": 0.57657874, + "learning_rate": 2.5975513221077313e-06, + "loss": 0.59724963, + "num_input_tokens_seen": 75571955, + "step": 3504, + "time_per_iteration": 3.4029343128204346 + }, + { + "auxiliary_loss_clip": 0.01157494, + "auxiliary_loss_mlp": 0.01029097, + "balance_loss_clip": 1.0477674, + "balance_loss_mlp": 1.0198108, + "epoch": 0.4214513316900138, + "flos": 23106038538240.0, + "grad_norm": 3.074105576369714, + "language_loss": 0.88686442, + "learning_rate": 2.5968078859508897e-06, + "loss": 0.90873033, + "num_input_tokens_seen": 75589155, + "step": 3505, + "time_per_iteration": 2.7531416416168213 + }, + { + "auxiliary_loss_clip": 0.01175884, + "auxiliary_loss_mlp": 0.01030274, + "balance_loss_clip": 1.04971671, + "balance_loss_mlp": 1.02063012, + "epoch": 0.4215715745806529, + "flos": 15336190857600.0, + "grad_norm": 2.0335735136519175, + "language_loss": 0.79784763, + "learning_rate": 2.5960643592587673e-06, + "loss": 0.81990921, + "num_input_tokens_seen": 75606565, + "step": 3506, + "time_per_iteration": 2.6979784965515137 + }, + { + "auxiliary_loss_clip": 0.01154641, + "auxiliary_loss_mlp": 0.0102955, + "balance_loss_clip": 1.04507613, + "balance_loss_mlp": 1.0206989, + "epoch": 0.42169181747129203, + "flos": 22127257860480.0, + "grad_norm": 1.8798185290956082, + "language_loss": 0.81140542, + "learning_rate": 2.5953207421441553e-06, + "loss": 0.8332473, + "num_input_tokens_seen": 75625165, + "step": 3507, + "time_per_iteration": 2.831937789916992 + }, + { + "auxiliary_loss_clip": 0.01159033, + "auxiliary_loss_mlp": 0.01032285, + "balance_loss_clip": 1.04881978, + "balance_loss_mlp": 1.02305865, + "epoch": 0.4218120603619311, + "flos": 22630724841600.0, + "grad_norm": 2.3247877633733105, + "language_loss": 0.75167996, + "learning_rate": 2.5945770347198603e-06, + "loss": 0.77359319, + "num_input_tokens_seen": 75643320, + "step": 3508, + "time_per_iteration": 2.7117087841033936 + }, + { + "auxiliary_loss_clip": 0.01157084, + "auxiliary_loss_mlp": 0.01034142, + "balance_loss_clip": 1.04539561, + "balance_loss_mlp": 1.02499318, + "epoch": 0.4219323032525702, + "flos": 19682818629120.0, + "grad_norm": 1.7695921248120001, + "language_loss": 0.82111096, + "learning_rate": 2.593833237098701e-06, + "loss": 0.8430233, + "num_input_tokens_seen": 75660920, + "step": 3509, + "time_per_iteration": 2.726672649383545 + }, + { + "auxiliary_loss_clip": 0.01175543, + "auxiliary_loss_mlp": 0.01034472, + "balance_loss_clip": 1.04814744, + "balance_loss_mlp": 1.02497673, + "epoch": 0.4220525461432093, + "flos": 30190747224960.0, + "grad_norm": 2.165032047932347, + "language_loss": 0.62601852, + "learning_rate": 2.593089349393512e-06, + "loss": 0.64811862, + "num_input_tokens_seen": 75681410, + "step": 3510, + "time_per_iteration": 2.7267308235168457 + }, + { + "auxiliary_loss_clip": 0.01175704, + "auxiliary_loss_mlp": 0.01029832, + "balance_loss_clip": 1.05410051, + "balance_loss_mlp": 1.02076006, + "epoch": 0.42217278903384836, + "flos": 24315941278080.0, + "grad_norm": 2.0014139802444566, + "language_loss": 0.83388019, + "learning_rate": 2.592345371717141e-06, + "loss": 0.85593551, + "num_input_tokens_seen": 75700940, + "step": 3511, + "time_per_iteration": 2.6987662315368652 + }, + { + "auxiliary_loss_clip": 0.0117926, + "auxiliary_loss_mlp": 0.01030377, + "balance_loss_clip": 1.05682957, + "balance_loss_mlp": 1.02166891, + "epoch": 0.42229303192448747, + "flos": 17092474352640.0, + "grad_norm": 2.3546199638669427, + "language_loss": 0.71731722, + "learning_rate": 2.591601304182448e-06, + "loss": 0.73941356, + "num_input_tokens_seen": 75718910, + "step": 3512, + "time_per_iteration": 2.682993173599243 + }, + { + "auxiliary_loss_clip": 0.01170054, + "auxiliary_loss_mlp": 0.01035998, + "balance_loss_clip": 1.05345225, + "balance_loss_mlp": 1.02757001, + "epoch": 0.4224132748151266, + "flos": 22784530878720.0, + "grad_norm": 1.7889983018960471, + "language_loss": 0.79260731, + "learning_rate": 2.5908571469023067e-06, + "loss": 0.81466782, + "num_input_tokens_seen": 75738395, + "step": 3513, + "time_per_iteration": 2.783238172531128 + }, + { + "auxiliary_loss_clip": 0.01184456, + "auxiliary_loss_mlp": 0.01030863, + "balance_loss_clip": 1.05211258, + "balance_loss_mlp": 1.02166009, + "epoch": 0.42253351770576564, + "flos": 17819090576640.0, + "grad_norm": 2.571819010127853, + "language_loss": 0.75613815, + "learning_rate": 2.5901128999896067e-06, + "loss": 0.77829134, + "num_input_tokens_seen": 75753825, + "step": 3514, + "time_per_iteration": 2.771399736404419 + }, + { + "auxiliary_loss_clip": 0.01174855, + "auxiliary_loss_mlp": 0.01032866, + "balance_loss_clip": 1.05212772, + "balance_loss_mlp": 1.0240798, + "epoch": 0.42265376059640475, + "flos": 28512390286080.0, + "grad_norm": 1.7913316913631825, + "language_loss": 0.68118441, + "learning_rate": 2.5893685635572487e-06, + "loss": 0.70326161, + "num_input_tokens_seen": 75774675, + "step": 3515, + "time_per_iteration": 2.7590527534484863 + }, + { + "auxiliary_loss_clip": 0.01164773, + "auxiliary_loss_mlp": 0.01028688, + "balance_loss_clip": 1.05092168, + "balance_loss_mlp": 1.0192349, + "epoch": 0.4227740034870438, + "flos": 16253349753600.0, + "grad_norm": 2.007273960612908, + "language_loss": 0.69468439, + "learning_rate": 2.5886241377181483e-06, + "loss": 0.71661901, + "num_input_tokens_seen": 75793545, + "step": 3516, + "time_per_iteration": 2.7488222122192383 + }, + { + "auxiliary_loss_clip": 0.01182045, + "auxiliary_loss_mlp": 0.0102917, + "balance_loss_clip": 1.05348063, + "balance_loss_mlp": 1.02006209, + "epoch": 0.4228942463776829, + "flos": 25295691623040.0, + "grad_norm": 4.323800780397552, + "language_loss": 0.81540495, + "learning_rate": 2.587879622585234e-06, + "loss": 0.83751708, + "num_input_tokens_seen": 75812145, + "step": 3517, + "time_per_iteration": 2.7772672176361084 + }, + { + "auxiliary_loss_clip": 0.01179046, + "auxiliary_loss_mlp": 0.01037878, + "balance_loss_clip": 1.05432177, + "balance_loss_mlp": 1.02874041, + "epoch": 0.423014489268322, + "flos": 26395779507840.0, + "grad_norm": 2.045063824854926, + "language_loss": 0.76151186, + "learning_rate": 2.5871350182714486e-06, + "loss": 0.78368109, + "num_input_tokens_seen": 75833025, + "step": 3518, + "time_per_iteration": 2.8239216804504395 + }, + { + "auxiliary_loss_clip": 0.01183996, + "auxiliary_loss_mlp": 0.01033341, + "balance_loss_clip": 1.05164969, + "balance_loss_mlp": 1.02451921, + "epoch": 0.4231347321589611, + "flos": 17274002711040.0, + "grad_norm": 2.051534164724179, + "language_loss": 0.80378914, + "learning_rate": 2.586390324889748e-06, + "loss": 0.82596248, + "num_input_tokens_seen": 75848925, + "step": 3519, + "time_per_iteration": 2.6079325675964355 + }, + { + "auxiliary_loss_clip": 0.0117535, + "auxiliary_loss_mlp": 0.0103111, + "balance_loss_clip": 1.05088186, + "balance_loss_mlp": 1.02254486, + "epoch": 0.4232549750496002, + "flos": 22999635475200.0, + "grad_norm": 1.9050827178641134, + "language_loss": 0.6705519, + "learning_rate": 2.5856455425531003e-06, + "loss": 0.69261646, + "num_input_tokens_seen": 75870400, + "step": 3520, + "time_per_iteration": 2.7924792766571045 + }, + { + "auxiliary_loss_clip": 0.01176628, + "auxiliary_loss_mlp": 0.01033713, + "balance_loss_clip": 1.05213487, + "balance_loss_mlp": 1.02524304, + "epoch": 0.4233752179402393, + "flos": 21248343970560.0, + "grad_norm": 1.751733833850333, + "language_loss": 0.80918515, + "learning_rate": 2.5849006713744902e-06, + "loss": 0.83128858, + "num_input_tokens_seen": 75889195, + "step": 3521, + "time_per_iteration": 3.6313743591308594 + }, + { + "auxiliary_loss_clip": 0.01166388, + "auxiliary_loss_mlp": 0.01032337, + "balance_loss_clip": 1.05081904, + "balance_loss_mlp": 1.0238378, + "epoch": 0.42349546083087836, + "flos": 20704297599360.0, + "grad_norm": 2.1840594705162295, + "language_loss": 0.7255193, + "learning_rate": 2.5841557114669135e-06, + "loss": 0.7475065, + "num_input_tokens_seen": 75906055, + "step": 3522, + "time_per_iteration": 2.7564291954040527 + }, + { + "auxiliary_loss_clip": 0.01191476, + "auxiliary_loss_mlp": 0.0103777, + "balance_loss_clip": 1.05334711, + "balance_loss_mlp": 1.02820349, + "epoch": 0.42361570372151747, + "flos": 18585065128320.0, + "grad_norm": 2.936635857450122, + "language_loss": 0.67495173, + "learning_rate": 2.58341066294338e-06, + "loss": 0.69724423, + "num_input_tokens_seen": 75922720, + "step": 3523, + "time_per_iteration": 2.5820865631103516 + }, + { + "auxiliary_loss_clip": 0.01153731, + "auxiliary_loss_mlp": 0.00886278, + "balance_loss_clip": 1.04876637, + "balance_loss_mlp": 1.00054026, + "epoch": 0.4237359466121566, + "flos": 20959478795520.0, + "grad_norm": 2.4387786674404475, + "language_loss": 0.85270405, + "learning_rate": 2.5826655259169124e-06, + "loss": 0.8731041, + "num_input_tokens_seen": 75941375, + "step": 3524, + "time_per_iteration": 2.779019355773926 + }, + { + "auxiliary_loss_clip": 0.01190379, + "auxiliary_loss_mlp": 0.01032109, + "balance_loss_clip": 1.05675578, + "balance_loss_mlp": 1.02342463, + "epoch": 0.42385618950279563, + "flos": 18038181582720.0, + "grad_norm": 1.7484942315236236, + "language_loss": 0.90273142, + "learning_rate": 2.5819203005005475e-06, + "loss": 0.92495632, + "num_input_tokens_seen": 75958710, + "step": 3525, + "time_per_iteration": 4.390966176986694 + }, + { + "auxiliary_loss_clip": 0.01165872, + "auxiliary_loss_mlp": 0.0103224, + "balance_loss_clip": 1.05207515, + "balance_loss_mlp": 1.023669, + "epoch": 0.42397643239343474, + "flos": 23769129559680.0, + "grad_norm": 1.5891385069608026, + "language_loss": 0.78471529, + "learning_rate": 2.581174986807336e-06, + "loss": 0.80669636, + "num_input_tokens_seen": 75978945, + "step": 3526, + "time_per_iteration": 2.7733333110809326 + }, + { + "auxiliary_loss_clip": 0.01173669, + "auxiliary_loss_mlp": 0.00886435, + "balance_loss_clip": 1.0528717, + "balance_loss_mlp": 1.00060701, + "epoch": 0.42409667528407385, + "flos": 16545088016640.0, + "grad_norm": 2.662158535055365, + "language_loss": 0.91285592, + "learning_rate": 2.580429584950341e-06, + "loss": 0.93345702, + "num_input_tokens_seen": 75994695, + "step": 3527, + "time_per_iteration": 2.675616979598999 + }, + { + "auxiliary_loss_clip": 0.01162186, + "auxiliary_loss_mlp": 0.01032146, + "balance_loss_clip": 1.05167103, + "balance_loss_mlp": 1.02266264, + "epoch": 0.4242169181747129, + "flos": 16034186920320.0, + "grad_norm": 2.4349634447854243, + "language_loss": 0.66865271, + "learning_rate": 2.5796840950426397e-06, + "loss": 0.6905961, + "num_input_tokens_seen": 76011780, + "step": 3528, + "time_per_iteration": 2.7815935611724854 + }, + { + "auxiliary_loss_clip": 0.01168881, + "auxiliary_loss_mlp": 0.01030639, + "balance_loss_clip": 1.05001128, + "balance_loss_mlp": 1.02131641, + "epoch": 0.424337161065352, + "flos": 20084012611200.0, + "grad_norm": 2.0731495086175245, + "language_loss": 0.6556775, + "learning_rate": 2.578938517197322e-06, + "loss": 0.67767274, + "num_input_tokens_seen": 76029875, + "step": 3529, + "time_per_iteration": 2.750697612762451 + }, + { + "auxiliary_loss_clip": 0.01158994, + "auxiliary_loss_mlp": 0.01032929, + "balance_loss_clip": 1.04861081, + "balance_loss_mlp": 1.02416086, + "epoch": 0.4244574039559911, + "flos": 23878369797120.0, + "grad_norm": 3.058961156092645, + "language_loss": 0.62469387, + "learning_rate": 2.5781928515274916e-06, + "loss": 0.64661306, + "num_input_tokens_seen": 76048595, + "step": 3530, + "time_per_iteration": 2.78716778755188 + }, + { + "auxiliary_loss_clip": 0.01182661, + "auxiliary_loss_mlp": 0.0103016, + "balance_loss_clip": 1.05554533, + "balance_loss_mlp": 1.02132058, + "epoch": 0.4245776468466302, + "flos": 17565920542080.0, + "grad_norm": 2.0532337409670443, + "language_loss": 0.67761427, + "learning_rate": 2.577447098146265e-06, + "loss": 0.6997425, + "num_input_tokens_seen": 76065770, + "step": 3531, + "time_per_iteration": 2.6528873443603516 + }, + { + "auxiliary_loss_clip": 0.01165878, + "auxiliary_loss_mlp": 0.01035765, + "balance_loss_clip": 1.05454791, + "balance_loss_mlp": 1.02724171, + "epoch": 0.4246978897372693, + "flos": 27776256958080.0, + "grad_norm": 1.6880487497902688, + "language_loss": 0.78838098, + "learning_rate": 2.5767012571667724e-06, + "loss": 0.81039739, + "num_input_tokens_seen": 76085250, + "step": 3532, + "time_per_iteration": 2.8689022064208984 + }, + { + "auxiliary_loss_clip": 0.0117999, + "auxiliary_loss_mlp": 0.01038531, + "balance_loss_clip": 1.05116439, + "balance_loss_mlp": 1.02897024, + "epoch": 0.42481813262790835, + "flos": 15596615439360.0, + "grad_norm": 1.9898218869420081, + "language_loss": 0.68139136, + "learning_rate": 2.5759553287021587e-06, + "loss": 0.70357656, + "num_input_tokens_seen": 76103580, + "step": 3533, + "time_per_iteration": 2.709036350250244 + }, + { + "auxiliary_loss_clip": 0.01167008, + "auxiliary_loss_mlp": 0.0102641, + "balance_loss_clip": 1.05152512, + "balance_loss_mlp": 1.01692724, + "epoch": 0.42493837551854746, + "flos": 23951088881280.0, + "grad_norm": 1.8138723682313307, + "language_loss": 0.7729193, + "learning_rate": 2.5752093128655786e-06, + "loss": 0.79485357, + "num_input_tokens_seen": 76121825, + "step": 3534, + "time_per_iteration": 2.8356244564056396 + }, + { + "auxiliary_loss_clip": 0.01160621, + "auxiliary_loss_mlp": 0.01033999, + "balance_loss_clip": 1.04921579, + "balance_loss_mlp": 1.02513623, + "epoch": 0.4250586184091866, + "flos": 20813466009600.0, + "grad_norm": 1.8063111325305927, + "language_loss": 0.73569202, + "learning_rate": 2.574463209770204e-06, + "loss": 0.75763822, + "num_input_tokens_seen": 76141140, + "step": 3535, + "time_per_iteration": 2.7576944828033447 + }, + { + "auxiliary_loss_clip": 0.01153052, + "auxiliary_loss_mlp": 0.01030356, + "balance_loss_clip": 1.04505169, + "balance_loss_mlp": 1.02080679, + "epoch": 0.42517886129982563, + "flos": 30371018607360.0, + "grad_norm": 1.74370231091406, + "language_loss": 0.79179156, + "learning_rate": 2.5737170195292165e-06, + "loss": 0.81362563, + "num_input_tokens_seen": 76164475, + "step": 3536, + "time_per_iteration": 2.9261579513549805 + }, + { + "auxiliary_loss_clip": 0.01157073, + "auxiliary_loss_mlp": 0.01030171, + "balance_loss_clip": 1.04696321, + "balance_loss_mlp": 1.02107537, + "epoch": 0.42529910419046474, + "flos": 20080636732800.0, + "grad_norm": 6.674973979856779, + "language_loss": 0.77921104, + "learning_rate": 2.572970742255814e-06, + "loss": 0.80108345, + "num_input_tokens_seen": 76182965, + "step": 3537, + "time_per_iteration": 2.9438700675964355 + }, + { + "auxiliary_loss_clip": 0.01177678, + "auxiliary_loss_mlp": 0.0103321, + "balance_loss_clip": 1.05379248, + "balance_loss_mlp": 1.02448356, + "epoch": 0.42541934708110385, + "flos": 22632448694400.0, + "grad_norm": 2.2448348351580103, + "language_loss": 0.81244612, + "learning_rate": 2.5722243780632046e-06, + "loss": 0.83455503, + "num_input_tokens_seen": 76201230, + "step": 3538, + "time_per_iteration": 2.622850179672241 + }, + { + "auxiliary_loss_clip": 0.01056379, + "auxiliary_loss_mlp": 0.0100404, + "balance_loss_clip": 1.01623929, + "balance_loss_mlp": 1.00192976, + "epoch": 0.4255395899717429, + "flos": 66200676186240.0, + "grad_norm": 0.7528623685861033, + "language_loss": 0.6047219, + "learning_rate": 2.5714779270646125e-06, + "loss": 0.62532604, + "num_input_tokens_seen": 76262000, + "step": 3539, + "time_per_iteration": 3.297869920730591 + }, + { + "auxiliary_loss_clip": 0.01170779, + "auxiliary_loss_mlp": 0.00886776, + "balance_loss_clip": 1.05264783, + "balance_loss_mlp": 1.00041163, + "epoch": 0.425659832862382, + "flos": 17931814433280.0, + "grad_norm": 2.9225171284076374, + "language_loss": 0.77995062, + "learning_rate": 2.5707313893732735e-06, + "loss": 0.80052614, + "num_input_tokens_seen": 76280540, + "step": 3540, + "time_per_iteration": 2.9342873096466064 + }, + { + "auxiliary_loss_clip": 0.01121812, + "auxiliary_loss_mlp": 0.01031845, + "balance_loss_clip": 1.04008245, + "balance_loss_mlp": 1.02233768, + "epoch": 0.4257800757530211, + "flos": 24022550989440.0, + "grad_norm": 1.8135074192343972, + "language_loss": 0.77238214, + "learning_rate": 2.5699847651024364e-06, + "loss": 0.79391873, + "num_input_tokens_seen": 76301180, + "step": 3541, + "time_per_iteration": 2.9882123470306396 + }, + { + "auxiliary_loss_clip": 0.01175378, + "auxiliary_loss_mlp": 0.01033293, + "balance_loss_clip": 1.055269, + "balance_loss_mlp": 1.02437067, + "epoch": 0.4259003186436602, + "flos": 23696015425920.0, + "grad_norm": 2.0213861049022976, + "language_loss": 0.77130979, + "learning_rate": 2.5692380543653627e-06, + "loss": 0.79339653, + "num_input_tokens_seen": 76319335, + "step": 3542, + "time_per_iteration": 2.8152341842651367 + }, + { + "auxiliary_loss_clip": 0.01181542, + "auxiliary_loss_mlp": 0.00886322, + "balance_loss_clip": 1.05267882, + "balance_loss_mlp": 1.00061035, + "epoch": 0.4260205615342993, + "flos": 15259772672640.0, + "grad_norm": 1.9474444382526979, + "language_loss": 0.69717765, + "learning_rate": 2.5684912572753293e-06, + "loss": 0.71785629, + "num_input_tokens_seen": 76335010, + "step": 3543, + "time_per_iteration": 2.715024709701538 + }, + { + "auxiliary_loss_clip": 0.01182277, + "auxiliary_loss_mlp": 0.01030279, + "balance_loss_clip": 1.05026245, + "balance_loss_mlp": 1.02189827, + "epoch": 0.4261408044249384, + "flos": 30665306736000.0, + "grad_norm": 1.7785927504103305, + "language_loss": 0.83747995, + "learning_rate": 2.5677443739456245e-06, + "loss": 0.85960549, + "num_input_tokens_seen": 76356670, + "step": 3544, + "time_per_iteration": 2.8806066513061523 + }, + { + "auxiliary_loss_clip": 0.01168961, + "auxiliary_loss_mlp": 0.01028221, + "balance_loss_clip": 1.0517149, + "balance_loss_mlp": 1.01960802, + "epoch": 0.42626104731557746, + "flos": 23257905240960.0, + "grad_norm": 2.361496334031647, + "language_loss": 0.7924304, + "learning_rate": 2.5669974044895495e-06, + "loss": 0.81440222, + "num_input_tokens_seen": 76373065, + "step": 3545, + "time_per_iteration": 2.842928171157837 + }, + { + "auxiliary_loss_clip": 0.01165477, + "auxiliary_loss_mlp": 0.01031118, + "balance_loss_clip": 1.04749906, + "balance_loss_mlp": 1.02229619, + "epoch": 0.42638129020621657, + "flos": 25884770670720.0, + "grad_norm": 2.047642702846912, + "language_loss": 0.79193473, + "learning_rate": 2.5662503490204187e-06, + "loss": 0.81390065, + "num_input_tokens_seen": 76393230, + "step": 3546, + "time_per_iteration": 2.9292867183685303 + }, + { + "auxiliary_loss_clip": 0.01165419, + "auxiliary_loss_mlp": 0.01031362, + "balance_loss_clip": 1.04914761, + "balance_loss_mlp": 1.02257609, + "epoch": 0.4265015330968556, + "flos": 26502362138880.0, + "grad_norm": 2.415242595468842, + "language_loss": 0.75893188, + "learning_rate": 2.5655032076515603e-06, + "loss": 0.78089964, + "num_input_tokens_seen": 76412555, + "step": 3547, + "time_per_iteration": 3.9207496643066406 + }, + { + "auxiliary_loss_clip": 0.01164457, + "auxiliary_loss_mlp": 0.01029738, + "balance_loss_clip": 1.04980016, + "balance_loss_mlp": 1.0211612, + "epoch": 0.42662177598749473, + "flos": 24389522288640.0, + "grad_norm": 2.0450243468509233, + "language_loss": 0.81876385, + "learning_rate": 2.5647559804963155e-06, + "loss": 0.84070575, + "num_input_tokens_seen": 76432485, + "step": 3548, + "time_per_iteration": 2.79469895362854 + }, + { + "auxiliary_loss_clip": 0.01158028, + "auxiliary_loss_mlp": 0.01038184, + "balance_loss_clip": 1.04965365, + "balance_loss_mlp": 1.02933872, + "epoch": 0.42674201887813384, + "flos": 23148629089920.0, + "grad_norm": 1.8518910834547824, + "language_loss": 0.78848994, + "learning_rate": 2.5640086676680364e-06, + "loss": 0.81045198, + "num_input_tokens_seen": 76453980, + "step": 3549, + "time_per_iteration": 2.9421162605285645 + }, + { + "auxiliary_loss_clip": 0.01175646, + "auxiliary_loss_mlp": 0.01032825, + "balance_loss_clip": 1.05098248, + "balance_loss_mlp": 1.02371168, + "epoch": 0.4268622617687729, + "flos": 21689614552320.0, + "grad_norm": 2.252609544162803, + "language_loss": 0.80836993, + "learning_rate": 2.5632612692800923e-06, + "loss": 0.83045471, + "num_input_tokens_seen": 76473045, + "step": 3550, + "time_per_iteration": 3.658278226852417 + }, + { + "auxiliary_loss_clip": 0.01162541, + "auxiliary_loss_mlp": 0.01036871, + "balance_loss_clip": 1.05004621, + "balance_loss_mlp": 1.02702463, + "epoch": 0.426982504659412, + "flos": 23440151871360.0, + "grad_norm": 2.0954322730042936, + "language_loss": 0.75021368, + "learning_rate": 2.5625137854458603e-06, + "loss": 0.77220774, + "num_input_tokens_seen": 76492060, + "step": 3551, + "time_per_iteration": 3.7087888717651367 + }, + { + "auxiliary_loss_clip": 0.01168965, + "auxiliary_loss_mlp": 0.01022588, + "balance_loss_clip": 1.0492487, + "balance_loss_mlp": 1.01486921, + "epoch": 0.4271027475500511, + "flos": 18916556768640.0, + "grad_norm": 1.7829354301496116, + "language_loss": 0.79837829, + "learning_rate": 2.561766216278735e-06, + "loss": 0.82029378, + "num_input_tokens_seen": 76509655, + "step": 3552, + "time_per_iteration": 2.7316153049468994 + }, + { + "auxiliary_loss_clip": 0.0114856, + "auxiliary_loss_mlp": 0.0103518, + "balance_loss_clip": 1.05047429, + "balance_loss_mlp": 1.02623928, + "epoch": 0.4272229904406902, + "flos": 26870554500480.0, + "grad_norm": 1.8180976869585892, + "language_loss": 0.81354564, + "learning_rate": 2.561018561892121e-06, + "loss": 0.83538306, + "num_input_tokens_seen": 76528795, + "step": 3553, + "time_per_iteration": 2.850489854812622 + }, + { + "auxiliary_loss_clip": 0.01164229, + "auxiliary_loss_mlp": 0.01034392, + "balance_loss_clip": 1.04659843, + "balance_loss_mlp": 1.02538562, + "epoch": 0.4273432333313293, + "flos": 23951376190080.0, + "grad_norm": 2.3916122944232803, + "language_loss": 0.7649163, + "learning_rate": 2.5602708223994363e-06, + "loss": 0.78690249, + "num_input_tokens_seen": 76550660, + "step": 3554, + "time_per_iteration": 2.883277177810669 + }, + { + "auxiliary_loss_clip": 0.01156992, + "auxiliary_loss_mlp": 0.01029365, + "balance_loss_clip": 1.04421031, + "balance_loss_mlp": 1.0206089, + "epoch": 0.4274634762219684, + "flos": 29570354496000.0, + "grad_norm": 2.7618677459007768, + "language_loss": 0.6775732, + "learning_rate": 2.559522997914115e-06, + "loss": 0.69943678, + "num_input_tokens_seen": 76570240, + "step": 3555, + "time_per_iteration": 2.8315088748931885 + }, + { + "auxiliary_loss_clip": 0.0118515, + "auxiliary_loss_mlp": 0.01031605, + "balance_loss_clip": 1.05465424, + "balance_loss_mlp": 1.02277756, + "epoch": 0.42758371911260745, + "flos": 21434146047360.0, + "grad_norm": 2.1209631973029217, + "language_loss": 0.84571147, + "learning_rate": 2.558775088549599e-06, + "loss": 0.86787891, + "num_input_tokens_seen": 76589820, + "step": 3556, + "time_per_iteration": 2.7648065090179443 + }, + { + "auxiliary_loss_clip": 0.01181041, + "auxiliary_loss_mlp": 0.01036129, + "balance_loss_clip": 1.05254221, + "balance_loss_mlp": 1.02637196, + "epoch": 0.42770396200324656, + "flos": 14752822072320.0, + "grad_norm": 2.9300266240943262, + "language_loss": 0.66456652, + "learning_rate": 2.5580270944193467e-06, + "loss": 0.68673825, + "num_input_tokens_seen": 76606640, + "step": 3557, + "time_per_iteration": 2.7224061489105225 + }, + { + "auxiliary_loss_clip": 0.01079247, + "auxiliary_loss_mlp": 0.01006916, + "balance_loss_clip": 1.01664329, + "balance_loss_mlp": 1.00505602, + "epoch": 0.4278242048938857, + "flos": 70654712601600.0, + "grad_norm": 0.7555539051995476, + "language_loss": 0.55460024, + "learning_rate": 2.557279015636827e-06, + "loss": 0.57546186, + "num_input_tokens_seen": 76667050, + "step": 3558, + "time_per_iteration": 3.266235113143921 + }, + { + "auxiliary_loss_clip": 0.01071094, + "auxiliary_loss_mlp": 0.0100899, + "balance_loss_clip": 1.01774716, + "balance_loss_mlp": 1.00716615, + "epoch": 0.42794444778452473, + "flos": 69366165033600.0, + "grad_norm": 0.7691693204448745, + "language_loss": 0.61252832, + "learning_rate": 2.5565308523155245e-06, + "loss": 0.63332915, + "num_input_tokens_seen": 76726650, + "step": 3559, + "time_per_iteration": 3.3057987689971924 + }, + { + "auxiliary_loss_clip": 0.0114441, + "auxiliary_loss_mlp": 0.01035715, + "balance_loss_clip": 1.04859245, + "balance_loss_mlp": 1.02659583, + "epoch": 0.42806469067516384, + "flos": 18215328481920.0, + "grad_norm": 2.786930038072982, + "language_loss": 0.8177973, + "learning_rate": 2.5557826045689336e-06, + "loss": 0.8395986, + "num_input_tokens_seen": 76742890, + "step": 3560, + "time_per_iteration": 2.842531681060791 + }, + { + "auxiliary_loss_clip": 0.01059962, + "auxiliary_loss_mlp": 0.01006762, + "balance_loss_clip": 1.01982737, + "balance_loss_mlp": 1.00449753, + "epoch": 0.4281849335658029, + "flos": 54535814432640.0, + "grad_norm": 0.8380825302706487, + "language_loss": 0.5885272, + "learning_rate": 2.5550342725105643e-06, + "loss": 0.6091944, + "num_input_tokens_seen": 76801055, + "step": 3561, + "time_per_iteration": 3.3410439491271973 + }, + { + "auxiliary_loss_clip": 0.01182001, + "auxiliary_loss_mlp": 0.01040495, + "balance_loss_clip": 1.05742836, + "balance_loss_mlp": 1.03206706, + "epoch": 0.428305176456442, + "flos": 17274828723840.0, + "grad_norm": 1.6406012392914366, + "language_loss": 0.81057703, + "learning_rate": 2.554285856253937e-06, + "loss": 0.83280206, + "num_input_tokens_seen": 76819890, + "step": 3562, + "time_per_iteration": 2.7547857761383057 + }, + { + "auxiliary_loss_clip": 0.01167731, + "auxiliary_loss_mlp": 0.01038005, + "balance_loss_clip": 1.05398107, + "balance_loss_mlp": 1.02900457, + "epoch": 0.4284254193470811, + "flos": 26359509749760.0, + "grad_norm": 1.9241496050927898, + "language_loss": 0.77395928, + "learning_rate": 2.5535373559125855e-06, + "loss": 0.79601657, + "num_input_tokens_seen": 76840255, + "step": 3563, + "time_per_iteration": 2.8492984771728516 + }, + { + "auxiliary_loss_clip": 0.01130029, + "auxiliary_loss_mlp": 0.01039616, + "balance_loss_clip": 1.04385722, + "balance_loss_mlp": 1.03021038, + "epoch": 0.42854566223772017, + "flos": 29714248379520.0, + "grad_norm": 1.5707047717534404, + "language_loss": 0.81780273, + "learning_rate": 2.552788771600057e-06, + "loss": 0.83949918, + "num_input_tokens_seen": 76860565, + "step": 3564, + "time_per_iteration": 3.081812858581543 + }, + { + "auxiliary_loss_clip": 0.01166602, + "auxiliary_loss_mlp": 0.01032661, + "balance_loss_clip": 1.05435658, + "balance_loss_mlp": 1.02437544, + "epoch": 0.4286659051283593, + "flos": 22018161277440.0, + "grad_norm": 3.593676685361582, + "language_loss": 0.82237327, + "learning_rate": 2.5520401034299118e-06, + "loss": 0.84436584, + "num_input_tokens_seen": 76878325, + "step": 3565, + "time_per_iteration": 2.7909114360809326 + }, + { + "auxiliary_loss_clip": 0.01180034, + "auxiliary_loss_mlp": 0.01033867, + "balance_loss_clip": 1.05125403, + "balance_loss_mlp": 1.02370465, + "epoch": 0.4287861480189984, + "flos": 13334422838400.0, + "grad_norm": 2.1526215836032847, + "language_loss": 0.88180345, + "learning_rate": 2.551291351515722e-06, + "loss": 0.90394247, + "num_input_tokens_seen": 76895340, + "step": 3566, + "time_per_iteration": 2.7008790969848633 + }, + { + "auxiliary_loss_clip": 0.01154075, + "auxiliary_loss_mlp": 0.00886765, + "balance_loss_clip": 1.04530525, + "balance_loss_mlp": 1.00069809, + "epoch": 0.42890639090963745, + "flos": 26651535321600.0, + "grad_norm": 1.7346639876875516, + "language_loss": 0.85780722, + "learning_rate": 2.5505425159710726e-06, + "loss": 0.87821567, + "num_input_tokens_seen": 76915150, + "step": 3567, + "time_per_iteration": 2.8690295219421387 + }, + { + "auxiliary_loss_clip": 0.01176451, + "auxiliary_loss_mlp": 0.00887361, + "balance_loss_clip": 1.05032635, + "balance_loss_mlp": 1.00065267, + "epoch": 0.42902663380027656, + "flos": 24055768091520.0, + "grad_norm": 4.029650826128024, + "language_loss": 0.83025658, + "learning_rate": 2.549793596909561e-06, + "loss": 0.85089475, + "num_input_tokens_seen": 76933770, + "step": 3568, + "time_per_iteration": 2.7944681644439697 + }, + { + "auxiliary_loss_clip": 0.0116366, + "auxiliary_loss_mlp": 0.010327, + "balance_loss_clip": 1.05029798, + "balance_loss_mlp": 1.02340734, + "epoch": 0.42914687669091567, + "flos": 15632561975040.0, + "grad_norm": 2.9379576248842043, + "language_loss": 0.66177988, + "learning_rate": 2.5490445944447976e-06, + "loss": 0.68374348, + "num_input_tokens_seen": 76952265, + "step": 3569, + "time_per_iteration": 2.7590737342834473 + }, + { + "auxiliary_loss_clip": 0.0117948, + "auxiliary_loss_mlp": 0.01032117, + "balance_loss_clip": 1.05352843, + "balance_loss_mlp": 1.02327728, + "epoch": 0.4292671195815547, + "flos": 31467802440960.0, + "grad_norm": 2.576615637919129, + "language_loss": 0.65044355, + "learning_rate": 2.548295508690406e-06, + "loss": 0.6725595, + "num_input_tokens_seen": 76973560, + "step": 3570, + "time_per_iteration": 2.7456321716308594 + }, + { + "auxiliary_loss_clip": 0.01179762, + "auxiliary_loss_mlp": 0.01024509, + "balance_loss_clip": 1.05048943, + "balance_loss_mlp": 1.01551473, + "epoch": 0.42938736247219383, + "flos": 30257756046720.0, + "grad_norm": 1.7216485818062972, + "language_loss": 0.76345468, + "learning_rate": 2.5475463397600217e-06, + "loss": 0.78549743, + "num_input_tokens_seen": 76993640, + "step": 3571, + "time_per_iteration": 2.7770204544067383 + }, + { + "auxiliary_loss_clip": 0.01193274, + "auxiliary_loss_mlp": 0.01037084, + "balance_loss_clip": 1.05689359, + "balance_loss_mlp": 1.02778578, + "epoch": 0.42950760536283294, + "flos": 29349683291520.0, + "grad_norm": 1.8697528935667391, + "language_loss": 0.77285278, + "learning_rate": 2.546797087767293e-06, + "loss": 0.79515636, + "num_input_tokens_seen": 77013765, + "step": 3572, + "time_per_iteration": 2.686558485031128 + }, + { + "auxiliary_loss_clip": 0.01151339, + "auxiliary_loss_mlp": 0.0103141, + "balance_loss_clip": 1.05049491, + "balance_loss_mlp": 1.02268362, + "epoch": 0.429627848253472, + "flos": 26869943969280.0, + "grad_norm": 1.7422926959537846, + "language_loss": 0.87395233, + "learning_rate": 2.546047752825881e-06, + "loss": 0.89577979, + "num_input_tokens_seen": 77034370, + "step": 3573, + "time_per_iteration": 4.750330448150635 + }, + { + "auxiliary_loss_clip": 0.01156374, + "auxiliary_loss_mlp": 0.01027283, + "balance_loss_clip": 1.05089116, + "balance_loss_mlp": 1.01843786, + "epoch": 0.4297480911441111, + "flos": 13881270470400.0, + "grad_norm": 2.3292169861151586, + "language_loss": 0.92764735, + "learning_rate": 2.5452983350494595e-06, + "loss": 0.94948399, + "num_input_tokens_seen": 77049925, + "step": 3574, + "time_per_iteration": 2.8072078227996826 + }, + { + "auxiliary_loss_clip": 0.01179803, + "auxiliary_loss_mlp": 0.00887128, + "balance_loss_clip": 1.05372167, + "balance_loss_mlp": 1.00082815, + "epoch": 0.4298683340347502, + "flos": 20741141975040.0, + "grad_norm": 2.1583612737975657, + "language_loss": 0.65518415, + "learning_rate": 2.544548834551713e-06, + "loss": 0.67585349, + "num_input_tokens_seen": 77068930, + "step": 3575, + "time_per_iteration": 2.6766040325164795 + }, + { + "auxiliary_loss_clip": 0.0115616, + "auxiliary_loss_mlp": 0.008866, + "balance_loss_clip": 1.04815304, + "balance_loss_mlp": 1.00071573, + "epoch": 0.4299885769253893, + "flos": 20882126856960.0, + "grad_norm": 12.935839924716527, + "language_loss": 0.94276857, + "learning_rate": 2.5437992514463424e-06, + "loss": 0.96319622, + "num_input_tokens_seen": 77082255, + "step": 3576, + "time_per_iteration": 3.722825288772583 + }, + { + "auxiliary_loss_clip": 0.01179048, + "auxiliary_loss_mlp": 0.01033898, + "balance_loss_clip": 1.05426693, + "balance_loss_mlp": 1.02467752, + "epoch": 0.4301088198160284, + "flos": 25484618183040.0, + "grad_norm": 1.7655783956257107, + "language_loss": 0.88004088, + "learning_rate": 2.5430495858470565e-06, + "loss": 0.9021703, + "num_input_tokens_seen": 77101725, + "step": 3577, + "time_per_iteration": 2.7538561820983887 + }, + { + "auxiliary_loss_clip": 0.01178221, + "auxiliary_loss_mlp": 0.01028962, + "balance_loss_clip": 1.05369163, + "balance_loss_mlp": 1.02014029, + "epoch": 0.43022906270666744, + "flos": 18259427404800.0, + "grad_norm": 2.350279641177825, + "language_loss": 0.77063137, + "learning_rate": 2.54229983786758e-06, + "loss": 0.79270321, + "num_input_tokens_seen": 77119670, + "step": 3578, + "time_per_iteration": 2.809007167816162 + }, + { + "auxiliary_loss_clip": 0.01169398, + "auxiliary_loss_mlp": 0.01034354, + "balance_loss_clip": 1.04923105, + "balance_loss_mlp": 1.02503777, + "epoch": 0.43034930559730655, + "flos": 23399536567680.0, + "grad_norm": 2.661838731720664, + "language_loss": 0.84836352, + "learning_rate": 2.541550007621651e-06, + "loss": 0.87040102, + "num_input_tokens_seen": 77138160, + "step": 3579, + "time_per_iteration": 2.914649486541748 + }, + { + "auxiliary_loss_clip": 0.01177374, + "auxiliary_loss_mlp": 0.01034054, + "balance_loss_clip": 1.05338109, + "balance_loss_mlp": 1.02578676, + "epoch": 0.43046954848794566, + "flos": 28184382264960.0, + "grad_norm": 1.8676512435844816, + "language_loss": 0.80160207, + "learning_rate": 2.5408000952230156e-06, + "loss": 0.82371628, + "num_input_tokens_seen": 77156950, + "step": 3580, + "time_per_iteration": 2.755371332168579 + }, + { + "auxiliary_loss_clip": 0.01165608, + "auxiliary_loss_mlp": 0.01035942, + "balance_loss_clip": 1.04854774, + "balance_loss_mlp": 1.0263164, + "epoch": 0.4305897913785847, + "flos": 28580476515840.0, + "grad_norm": 2.141426997559273, + "language_loss": 0.90249854, + "learning_rate": 2.5400501007854357e-06, + "loss": 0.92451406, + "num_input_tokens_seen": 77176395, + "step": 3581, + "time_per_iteration": 2.926438808441162 + }, + { + "auxiliary_loss_clip": 0.01147265, + "auxiliary_loss_mlp": 0.010312, + "balance_loss_clip": 1.04486597, + "balance_loss_mlp": 1.02242005, + "epoch": 0.43071003426922383, + "flos": 20448721353600.0, + "grad_norm": 1.9825711947447535, + "language_loss": 0.75326741, + "learning_rate": 2.539300024422685e-06, + "loss": 0.77505207, + "num_input_tokens_seen": 77194340, + "step": 3582, + "time_per_iteration": 2.8609373569488525 + }, + { + "auxiliary_loss_clip": 0.01053458, + "auxiliary_loss_mlp": 0.01008815, + "balance_loss_clip": 1.0142324, + "balance_loss_mlp": 1.00695503, + "epoch": 0.43083027715986294, + "flos": 51997969883520.0, + "grad_norm": 0.7904947369732707, + "language_loss": 0.60936725, + "learning_rate": 2.538549866248549e-06, + "loss": 0.62998998, + "num_input_tokens_seen": 77249320, + "step": 3583, + "time_per_iteration": 3.2347166538238525 + }, + { + "auxiliary_loss_clip": 0.01179812, + "auxiliary_loss_mlp": 0.01036056, + "balance_loss_clip": 1.05301476, + "balance_loss_mlp": 1.02667379, + "epoch": 0.430950520050502, + "flos": 16690885320960.0, + "grad_norm": 1.9115189368735892, + "language_loss": 0.80930579, + "learning_rate": 2.5377996263768274e-06, + "loss": 0.83146447, + "num_input_tokens_seen": 77267400, + "step": 3584, + "time_per_iteration": 2.7551705837249756 + }, + { + "auxiliary_loss_clip": 0.01173423, + "auxiliary_loss_mlp": 0.0103604, + "balance_loss_clip": 1.04901958, + "balance_loss_mlp": 1.02698588, + "epoch": 0.4310707629411411, + "flos": 24608433726720.0, + "grad_norm": 1.7815451933172861, + "language_loss": 0.68763244, + "learning_rate": 2.5370493049213293e-06, + "loss": 0.70972705, + "num_input_tokens_seen": 77287045, + "step": 3585, + "time_per_iteration": 2.7320852279663086 + }, + { + "auxiliary_loss_clip": 0.01114065, + "auxiliary_loss_mlp": 0.01033492, + "balance_loss_clip": 1.0430088, + "balance_loss_mlp": 1.02398562, + "epoch": 0.4311910058317802, + "flos": 26432983019520.0, + "grad_norm": 1.9990024180838069, + "language_loss": 0.7969178, + "learning_rate": 2.536298901995878e-06, + "loss": 0.81839341, + "num_input_tokens_seen": 77306255, + "step": 3586, + "time_per_iteration": 3.1285481452941895 + }, + { + "auxiliary_loss_clip": 0.01171949, + "auxiliary_loss_mlp": 0.01032294, + "balance_loss_clip": 1.05275035, + "balance_loss_mlp": 1.0231092, + "epoch": 0.43131124872241927, + "flos": 25155891889920.0, + "grad_norm": 1.6734656273185482, + "language_loss": 0.79848349, + "learning_rate": 2.535548417714311e-06, + "loss": 0.82052588, + "num_input_tokens_seen": 77325555, + "step": 3587, + "time_per_iteration": 2.976797342300415 + }, + { + "auxiliary_loss_clip": 0.01178516, + "auxiliary_loss_mlp": 0.01029555, + "balance_loss_clip": 1.05067015, + "balance_loss_mlp": 1.02081656, + "epoch": 0.4314314916130584, + "flos": 21614812479360.0, + "grad_norm": 1.5702332946444382, + "language_loss": 0.87055814, + "learning_rate": 2.534797852190474e-06, + "loss": 0.8926388, + "num_input_tokens_seen": 77345735, + "step": 3588, + "time_per_iteration": 2.7887141704559326 + }, + { + "auxiliary_loss_clip": 0.01173543, + "auxiliary_loss_mlp": 0.01033912, + "balance_loss_clip": 1.04909849, + "balance_loss_mlp": 1.02531683, + "epoch": 0.4315517345036975, + "flos": 19275016544640.0, + "grad_norm": 1.9329339225247986, + "language_loss": 0.81333619, + "learning_rate": 2.5340472055382283e-06, + "loss": 0.83541077, + "num_input_tokens_seen": 77361765, + "step": 3589, + "time_per_iteration": 2.700199842453003 + }, + { + "auxiliary_loss_clip": 0.01155086, + "auxiliary_loss_mlp": 0.01024099, + "balance_loss_clip": 1.04503894, + "balance_loss_mlp": 1.01504827, + "epoch": 0.43167197739433655, + "flos": 24273853516800.0, + "grad_norm": 2.0165147923562405, + "language_loss": 0.80934983, + "learning_rate": 2.5332964778714468e-06, + "loss": 0.83114171, + "num_input_tokens_seen": 77378950, + "step": 3590, + "time_per_iteration": 2.9247231483459473 + }, + { + "auxiliary_loss_clip": 0.01160234, + "auxiliary_loss_mlp": 0.01031724, + "balance_loss_clip": 1.05135596, + "balance_loss_mlp": 1.02297974, + "epoch": 0.43179222028497566, + "flos": 16867816738560.0, + "grad_norm": 1.7250053389838675, + "language_loss": 0.66357636, + "learning_rate": 2.5325456693040123e-06, + "loss": 0.68549591, + "num_input_tokens_seen": 77396145, + "step": 3591, + "time_per_iteration": 2.713038921356201 + }, + { + "auxiliary_loss_clip": 0.01179622, + "auxiliary_loss_mlp": 0.01029604, + "balance_loss_clip": 1.05027175, + "balance_loss_mlp": 1.02009749, + "epoch": 0.43191246317561477, + "flos": 17639214243840.0, + "grad_norm": 2.4889567488801454, + "language_loss": 0.74867499, + "learning_rate": 2.531794779949824e-06, + "loss": 0.77076721, + "num_input_tokens_seen": 77414045, + "step": 3592, + "time_per_iteration": 2.766301155090332 + }, + { + "auxiliary_loss_clip": 0.01147876, + "auxiliary_loss_mlp": 0.01029188, + "balance_loss_clip": 1.04580021, + "balance_loss_mlp": 1.01983595, + "epoch": 0.4320327060662538, + "flos": 23878800760320.0, + "grad_norm": 1.7935586083971142, + "language_loss": 0.88043463, + "learning_rate": 2.5310438099227903e-06, + "loss": 0.90220523, + "num_input_tokens_seen": 77431310, + "step": 3593, + "time_per_iteration": 2.7800395488739014 + }, + { + "auxiliary_loss_clip": 0.0106969, + "auxiliary_loss_mlp": 0.01002334, + "balance_loss_clip": 1.01315641, + "balance_loss_mlp": 1.00046277, + "epoch": 0.43215294895689293, + "flos": 66394917959040.0, + "grad_norm": 0.8046665607615362, + "language_loss": 0.53382289, + "learning_rate": 2.530292759336833e-06, + "loss": 0.55454314, + "num_input_tokens_seen": 77492045, + "step": 3594, + "time_per_iteration": 3.339245080947876 + }, + { + "auxiliary_loss_clip": 0.0116109, + "auxiliary_loss_mlp": 0.01031972, + "balance_loss_clip": 1.04886961, + "balance_loss_mlp": 1.02219081, + "epoch": 0.432273191847532, + "flos": 20594267262720.0, + "grad_norm": 2.206898239031923, + "language_loss": 0.6890372, + "learning_rate": 2.5295416283058855e-06, + "loss": 0.71096778, + "num_input_tokens_seen": 77510910, + "step": 3595, + "time_per_iteration": 2.766303777694702 + }, + { + "auxiliary_loss_clip": 0.01164507, + "auxiliary_loss_mlp": 0.00886136, + "balance_loss_clip": 1.05100322, + "balance_loss_mlp": 1.00074029, + "epoch": 0.4323934347381711, + "flos": 19282127437440.0, + "grad_norm": 1.5910341543949469, + "language_loss": 0.66120636, + "learning_rate": 2.5287904169438943e-06, + "loss": 0.68171281, + "num_input_tokens_seen": 77530115, + "step": 3596, + "time_per_iteration": 2.758958101272583 + }, + { + "auxiliary_loss_clip": 0.01137349, + "auxiliary_loss_mlp": 0.01039136, + "balance_loss_clip": 1.04530025, + "balance_loss_mlp": 1.03021932, + "epoch": 0.4325136776288102, + "flos": 21726315273600.0, + "grad_norm": 2.889456140316249, + "language_loss": 0.64327079, + "learning_rate": 2.528039125364817e-06, + "loss": 0.66503567, + "num_input_tokens_seen": 77548920, + "step": 3597, + "time_per_iteration": 2.873344659805298 + }, + { + "auxiliary_loss_clip": 0.01161782, + "auxiliary_loss_mlp": 0.01028806, + "balance_loss_clip": 1.0501529, + "balance_loss_mlp": 1.01930547, + "epoch": 0.43263392051944927, + "flos": 22340746344960.0, + "grad_norm": 2.015547928655815, + "language_loss": 0.75734431, + "learning_rate": 2.5272877536826246e-06, + "loss": 0.7792502, + "num_input_tokens_seen": 77567715, + "step": 3598, + "time_per_iteration": 2.8812038898468018 + }, + { + "auxiliary_loss_clip": 0.01148586, + "auxiliary_loss_mlp": 0.01030759, + "balance_loss_clip": 1.04288983, + "balance_loss_mlp": 1.02138352, + "epoch": 0.4327541634100884, + "flos": 29168406328320.0, + "grad_norm": 4.200837755887452, + "language_loss": 0.70331961, + "learning_rate": 2.5265363020112986e-06, + "loss": 0.72511303, + "num_input_tokens_seen": 77588035, + "step": 3599, + "time_per_iteration": 4.811551332473755 + }, + { + "auxiliary_loss_clip": 0.01176695, + "auxiliary_loss_mlp": 0.01031632, + "balance_loss_clip": 1.05238152, + "balance_loss_mlp": 1.02286434, + "epoch": 0.4328744063007275, + "flos": 26067448264320.0, + "grad_norm": 2.4126066293605484, + "language_loss": 0.83940506, + "learning_rate": 2.5257847704648344e-06, + "loss": 0.8614884, + "num_input_tokens_seen": 77609265, + "step": 3600, + "time_per_iteration": 2.758683681488037 + }, + { + "auxiliary_loss_clip": 0.01185875, + "auxiliary_loss_mlp": 0.01028498, + "balance_loss_clip": 1.0538063, + "balance_loss_mlp": 1.01976037, + "epoch": 0.43299464919136654, + "flos": 16581357774720.0, + "grad_norm": 1.7520152496325527, + "language_loss": 0.75401902, + "learning_rate": 2.525033159157239e-06, + "loss": 0.7761628, + "num_input_tokens_seen": 77625580, + "step": 3601, + "time_per_iteration": 3.606712818145752 + }, + { + "auxiliary_loss_clip": 0.01173312, + "auxiliary_loss_mlp": 0.01030669, + "balance_loss_clip": 1.05038905, + "balance_loss_mlp": 1.02054262, + "epoch": 0.43311489208200565, + "flos": 16107265140480.0, + "grad_norm": 1.83685213419354, + "language_loss": 0.77027583, + "learning_rate": 2.52428146820253e-06, + "loss": 0.79231572, + "num_input_tokens_seen": 77643835, + "step": 3602, + "time_per_iteration": 3.646965980529785 + }, + { + "auxiliary_loss_clip": 0.01159051, + "auxiliary_loss_mlp": 0.01037049, + "balance_loss_clip": 1.05025887, + "balance_loss_mlp": 1.02801251, + "epoch": 0.43323513497264476, + "flos": 22930220442240.0, + "grad_norm": 1.7424745276869222, + "language_loss": 0.81734979, + "learning_rate": 2.52352969771474e-06, + "loss": 0.83931077, + "num_input_tokens_seen": 77663060, + "step": 3603, + "time_per_iteration": 2.7782723903656006 + }, + { + "auxiliary_loss_clip": 0.01170899, + "auxiliary_loss_mlp": 0.01031113, + "balance_loss_clip": 1.05265999, + "balance_loss_mlp": 1.02280414, + "epoch": 0.4333553778632838, + "flos": 25299031587840.0, + "grad_norm": 2.0414565705179704, + "language_loss": 0.88469541, + "learning_rate": 2.5227778478079106e-06, + "loss": 0.90671551, + "num_input_tokens_seen": 77682470, + "step": 3604, + "time_per_iteration": 2.797863721847534 + }, + { + "auxiliary_loss_clip": 0.01172678, + "auxiliary_loss_mlp": 0.01035231, + "balance_loss_clip": 1.04964876, + "balance_loss_mlp": 1.02714241, + "epoch": 0.43347562075392293, + "flos": 19387165783680.0, + "grad_norm": 2.0941625204761154, + "language_loss": 0.76732606, + "learning_rate": 2.522025918596098e-06, + "loss": 0.78940517, + "num_input_tokens_seen": 77700770, + "step": 3605, + "time_per_iteration": 2.657595634460449 + }, + { + "auxiliary_loss_clip": 0.01172882, + "auxiliary_loss_mlp": 0.01027288, + "balance_loss_clip": 1.05051208, + "balance_loss_mlp": 1.01854992, + "epoch": 0.43359586364456204, + "flos": 26325969425280.0, + "grad_norm": 1.8768680431108977, + "language_loss": 0.65321565, + "learning_rate": 2.521273910193368e-06, + "loss": 0.67521739, + "num_input_tokens_seen": 77723950, + "step": 3606, + "time_per_iteration": 2.78843092918396 + }, + { + "auxiliary_loss_clip": 0.01184137, + "auxiliary_loss_mlp": 0.01033328, + "balance_loss_clip": 1.05363071, + "balance_loss_mlp": 1.0246619, + "epoch": 0.4337161065352011, + "flos": 15989261984640.0, + "grad_norm": 2.3038787738328472, + "language_loss": 0.87054801, + "learning_rate": 2.5205218227138006e-06, + "loss": 0.89272267, + "num_input_tokens_seen": 77736905, + "step": 3607, + "time_per_iteration": 2.7461442947387695 + }, + { + "auxiliary_loss_clip": 0.01186889, + "auxiliary_loss_mlp": 0.01034497, + "balance_loss_clip": 1.05289197, + "balance_loss_mlp": 1.0254612, + "epoch": 0.4338363494258402, + "flos": 20224710184320.0, + "grad_norm": 2.695136189233081, + "language_loss": 0.79267257, + "learning_rate": 2.519769656271486e-06, + "loss": 0.81488645, + "num_input_tokens_seen": 77754325, + "step": 3608, + "time_per_iteration": 2.6993865966796875 + }, + { + "auxiliary_loss_clip": 0.01139369, + "auxiliary_loss_mlp": 0.01034516, + "balance_loss_clip": 1.04565108, + "balance_loss_mlp": 1.02581334, + "epoch": 0.43395659231647926, + "flos": 20083904870400.0, + "grad_norm": 2.3782515516453255, + "language_loss": 0.67650604, + "learning_rate": 2.5190174109805285e-06, + "loss": 0.69824493, + "num_input_tokens_seen": 77774150, + "step": 3609, + "time_per_iteration": 2.945587635040283 + }, + { + "auxiliary_loss_clip": 0.0115702, + "auxiliary_loss_mlp": 0.01033616, + "balance_loss_clip": 1.04765785, + "balance_loss_mlp": 1.02461016, + "epoch": 0.43407683520711837, + "flos": 19901801894400.0, + "grad_norm": 3.6793579622928814, + "language_loss": 0.64257306, + "learning_rate": 2.518265086955042e-06, + "loss": 0.66447943, + "num_input_tokens_seen": 77791870, + "step": 3610, + "time_per_iteration": 2.8205220699310303 + }, + { + "auxiliary_loss_clip": 0.01187694, + "auxiliary_loss_mlp": 0.01036187, + "balance_loss_clip": 1.05366504, + "balance_loss_mlp": 1.02740097, + "epoch": 0.4341970780977575, + "flos": 23108732058240.0, + "grad_norm": 1.7829830735065906, + "language_loss": 0.83894205, + "learning_rate": 2.5175126843091534e-06, + "loss": 0.8611809, + "num_input_tokens_seen": 77811240, + "step": 3611, + "time_per_iteration": 2.793617010116577 + }, + { + "auxiliary_loss_clip": 0.01167853, + "auxiliary_loss_mlp": 0.01031436, + "balance_loss_clip": 1.05241275, + "balance_loss_mlp": 1.02269232, + "epoch": 0.43431732098839654, + "flos": 37408288406400.0, + "grad_norm": 2.2355581211931654, + "language_loss": 0.75297439, + "learning_rate": 2.5167602031570034e-06, + "loss": 0.77496725, + "num_input_tokens_seen": 77831425, + "step": 3612, + "time_per_iteration": 2.929471254348755 + }, + { + "auxiliary_loss_clip": 0.01187513, + "auxiliary_loss_mlp": 0.01029856, + "balance_loss_clip": 1.05414498, + "balance_loss_mlp": 1.02105236, + "epoch": 0.43443756387903565, + "flos": 31868206323840.0, + "grad_norm": 1.7174625840530529, + "language_loss": 0.73211718, + "learning_rate": 2.51600764361274e-06, + "loss": 0.75429082, + "num_input_tokens_seen": 77852950, + "step": 3613, + "time_per_iteration": 2.848262310028076 + }, + { + "auxiliary_loss_clip": 0.01188228, + "auxiliary_loss_mlp": 0.01030623, + "balance_loss_clip": 1.0539906, + "balance_loss_mlp": 1.02157509, + "epoch": 0.43455780676967476, + "flos": 23477139901440.0, + "grad_norm": 8.181426985318057, + "language_loss": 0.78815901, + "learning_rate": 2.5152550057905283e-06, + "loss": 0.81034744, + "num_input_tokens_seen": 77872840, + "step": 3614, + "time_per_iteration": 2.713541030883789 + }, + { + "auxiliary_loss_clip": 0.01179055, + "auxiliary_loss_mlp": 0.00886655, + "balance_loss_clip": 1.0539279, + "balance_loss_mlp": 1.00066566, + "epoch": 0.4346780496603138, + "flos": 24207060176640.0, + "grad_norm": 1.9953535109269827, + "language_loss": 0.77098209, + "learning_rate": 2.5145022898045415e-06, + "loss": 0.79163921, + "num_input_tokens_seen": 77892025, + "step": 3615, + "time_per_iteration": 2.868665933609009 + }, + { + "auxiliary_loss_clip": 0.01164737, + "auxiliary_loss_mlp": 0.01028392, + "balance_loss_clip": 1.0471468, + "balance_loss_mlp": 1.01929069, + "epoch": 0.4347982925509529, + "flos": 17092366611840.0, + "grad_norm": 2.823988887806791, + "language_loss": 0.89414072, + "learning_rate": 2.5137494957689664e-06, + "loss": 0.91607207, + "num_input_tokens_seen": 77907635, + "step": 3616, + "time_per_iteration": 2.798534631729126 + }, + { + "auxiliary_loss_clip": 0.01061357, + "auxiliary_loss_mlp": 0.01006613, + "balance_loss_clip": 1.01465464, + "balance_loss_mlp": 1.00466979, + "epoch": 0.43491853544159204, + "flos": 60945544696320.0, + "grad_norm": 0.7670342291650204, + "language_loss": 0.57412446, + "learning_rate": 2.5129966237980016e-06, + "loss": 0.59480417, + "num_input_tokens_seen": 77970630, + "step": 3617, + "time_per_iteration": 3.287076234817505 + }, + { + "auxiliary_loss_clip": 0.01158067, + "auxiliary_loss_mlp": 0.01030932, + "balance_loss_clip": 1.04680049, + "balance_loss_mlp": 1.02193165, + "epoch": 0.4350387783322311, + "flos": 21944652094080.0, + "grad_norm": 1.829870999755027, + "language_loss": 0.7803812, + "learning_rate": 2.512243674005857e-06, + "loss": 0.80227113, + "num_input_tokens_seen": 77989995, + "step": 3618, + "time_per_iteration": 2.7743446826934814 + }, + { + "auxiliary_loss_clip": 0.01133762, + "auxiliary_loss_mlp": 0.01036837, + "balance_loss_clip": 1.04307199, + "balance_loss_mlp": 1.02806282, + "epoch": 0.4351590212228702, + "flos": 25082705928960.0, + "grad_norm": 2.1504958283985216, + "language_loss": 0.86300671, + "learning_rate": 2.5114906465067537e-06, + "loss": 0.8847127, + "num_input_tokens_seen": 78010980, + "step": 3619, + "time_per_iteration": 2.8615686893463135 + }, + { + "auxiliary_loss_clip": 0.01175867, + "auxiliary_loss_mlp": 0.0103111, + "balance_loss_clip": 1.04883337, + "balance_loss_mlp": 1.02281284, + "epoch": 0.4352792641135093, + "flos": 21506541909120.0, + "grad_norm": 1.9626555127309093, + "language_loss": 0.74997729, + "learning_rate": 2.5107375414149264e-06, + "loss": 0.77204704, + "num_input_tokens_seen": 78030225, + "step": 3620, + "time_per_iteration": 2.7406651973724365 + }, + { + "auxiliary_loss_clip": 0.01137726, + "auxiliary_loss_mlp": 0.0102881, + "balance_loss_clip": 1.04120302, + "balance_loss_mlp": 1.01967895, + "epoch": 0.43539950700414837, + "flos": 16253457494400.0, + "grad_norm": 2.9531108564215627, + "language_loss": 0.71731234, + "learning_rate": 2.5099843588446197e-06, + "loss": 0.73897767, + "num_input_tokens_seen": 78048545, + "step": 3621, + "time_per_iteration": 2.787193775177002 + }, + { + "auxiliary_loss_clip": 0.0115135, + "auxiliary_loss_mlp": 0.01032724, + "balance_loss_clip": 1.04894733, + "balance_loss_mlp": 1.02423644, + "epoch": 0.4355197498947875, + "flos": 16691819074560.0, + "grad_norm": 1.9312057301650356, + "language_loss": 0.61601347, + "learning_rate": 2.509231098910091e-06, + "loss": 0.63785428, + "num_input_tokens_seen": 78068415, + "step": 3622, + "time_per_iteration": 3.079214334487915 + }, + { + "auxiliary_loss_clip": 0.01164213, + "auxiliary_loss_mlp": 0.01029122, + "balance_loss_clip": 1.05360496, + "balance_loss_mlp": 1.01985896, + "epoch": 0.4356399927854266, + "flos": 16362733645440.0, + "grad_norm": 2.59925783556509, + "language_loss": 0.74740434, + "learning_rate": 2.508477761725611e-06, + "loss": 0.76933765, + "num_input_tokens_seen": 78086690, + "step": 3623, + "time_per_iteration": 2.806856870651245 + }, + { + "auxiliary_loss_clip": 0.01182727, + "auxiliary_loss_mlp": 0.01029201, + "balance_loss_clip": 1.05319524, + "balance_loss_mlp": 1.02048683, + "epoch": 0.43576023567606564, + "flos": 17202037812480.0, + "grad_norm": 2.409966282919244, + "language_loss": 0.80688983, + "learning_rate": 2.507724347405458e-06, + "loss": 0.82900912, + "num_input_tokens_seen": 78104640, + "step": 3624, + "time_per_iteration": 2.701225757598877 + }, + { + "auxiliary_loss_clip": 0.01140366, + "auxiliary_loss_mlp": 0.01026844, + "balance_loss_clip": 1.04248548, + "balance_loss_mlp": 1.01788497, + "epoch": 0.43588047856670475, + "flos": 15917656222080.0, + "grad_norm": 1.9408870700476633, + "language_loss": 0.82350475, + "learning_rate": 2.5069708560639243e-06, + "loss": 0.84517682, + "num_input_tokens_seen": 78122550, + "step": 3625, + "time_per_iteration": 4.660166501998901 + }, + { + "auxiliary_loss_clip": 0.01157904, + "auxiliary_loss_mlp": 0.0102967, + "balance_loss_clip": 1.0492835, + "balance_loss_mlp": 1.02094364, + "epoch": 0.4360007214573438, + "flos": 23659566099840.0, + "grad_norm": 2.2684199972807715, + "language_loss": 0.61562663, + "learning_rate": 2.5062172878153158e-06, + "loss": 0.63750231, + "num_input_tokens_seen": 78141825, + "step": 3626, + "time_per_iteration": 2.8797483444213867 + }, + { + "auxiliary_loss_clip": 0.01143242, + "auxiliary_loss_mlp": 0.01039151, + "balance_loss_clip": 1.04643774, + "balance_loss_mlp": 1.03004384, + "epoch": 0.4361209643479829, + "flos": 21978767036160.0, + "grad_norm": 2.810878240133556, + "language_loss": 0.87324464, + "learning_rate": 2.505463642773947e-06, + "loss": 0.89506859, + "num_input_tokens_seen": 78161790, + "step": 3627, + "time_per_iteration": 4.78658390045166 + }, + { + "auxiliary_loss_clip": 0.0115618, + "auxiliary_loss_mlp": 0.00886644, + "balance_loss_clip": 1.04900837, + "balance_loss_mlp": 1.0005455, + "epoch": 0.43624120723862203, + "flos": 17420159151360.0, + "grad_norm": 2.2947509745337533, + "language_loss": 0.75228053, + "learning_rate": 2.504709921054146e-06, + "loss": 0.77270877, + "num_input_tokens_seen": 78178605, + "step": 3628, + "time_per_iteration": 2.897264242172241 + }, + { + "auxiliary_loss_clip": 0.01150655, + "auxiliary_loss_mlp": 0.01037279, + "balance_loss_clip": 1.04420137, + "balance_loss_mlp": 1.02843356, + "epoch": 0.4363614501292611, + "flos": 17895293280000.0, + "grad_norm": 2.150901806049675, + "language_loss": 0.83627772, + "learning_rate": 2.50395612277025e-06, + "loss": 0.8581571, + "num_input_tokens_seen": 78194460, + "step": 3629, + "time_per_iteration": 2.905754327774048 + }, + { + "auxiliary_loss_clip": 0.01170821, + "auxiliary_loss_mlp": 0.01031717, + "balance_loss_clip": 1.04944503, + "balance_loss_mlp": 1.02290785, + "epoch": 0.4364816930199002, + "flos": 20302888135680.0, + "grad_norm": 2.225249038393898, + "language_loss": 0.72274756, + "learning_rate": 2.503202248036612e-06, + "loss": 0.74477291, + "num_input_tokens_seen": 78213315, + "step": 3630, + "time_per_iteration": 2.784484386444092 + }, + { + "auxiliary_loss_clip": 0.01183727, + "auxiliary_loss_mlp": 0.01037087, + "balance_loss_clip": 1.05241871, + "balance_loss_mlp": 1.02827156, + "epoch": 0.4366019359105393, + "flos": 24061334699520.0, + "grad_norm": 2.647432012930312, + "language_loss": 0.73154509, + "learning_rate": 2.5024482969675927e-06, + "loss": 0.75375319, + "num_input_tokens_seen": 78233270, + "step": 3631, + "time_per_iteration": 2.8089942932128906 + }, + { + "auxiliary_loss_clip": 0.01147407, + "auxiliary_loss_mlp": 0.01028472, + "balance_loss_clip": 1.04617119, + "balance_loss_mlp": 1.02036548, + "epoch": 0.43672217880117836, + "flos": 21754109422080.0, + "grad_norm": 1.917131443728794, + "language_loss": 0.84057248, + "learning_rate": 2.501694269677566e-06, + "loss": 0.86233127, + "num_input_tokens_seen": 78251040, + "step": 3632, + "time_per_iteration": 2.9158544540405273 + }, + { + "auxiliary_loss_clip": 0.01179289, + "auxiliary_loss_mlp": 0.01025962, + "balance_loss_clip": 1.05106258, + "balance_loss_mlp": 1.01670527, + "epoch": 0.4368424216918175, + "flos": 18035200753920.0, + "grad_norm": 2.168186750854523, + "language_loss": 0.80771768, + "learning_rate": 2.500940166280918e-06, + "loss": 0.82977021, + "num_input_tokens_seen": 78269470, + "step": 3633, + "time_per_iteration": 2.873120069503784 + }, + { + "auxiliary_loss_clip": 0.01172721, + "auxiliary_loss_mlp": 0.01029874, + "balance_loss_clip": 1.04956591, + "balance_loss_mlp": 1.02117229, + "epoch": 0.4369626645824566, + "flos": 25447127362560.0, + "grad_norm": 1.9540224737796272, + "language_loss": 0.79207325, + "learning_rate": 2.500185986892045e-06, + "loss": 0.81409913, + "num_input_tokens_seen": 78288955, + "step": 3634, + "time_per_iteration": 2.8453845977783203 + }, + { + "auxiliary_loss_clip": 0.01170515, + "auxiliary_loss_mlp": 0.01033267, + "balance_loss_clip": 1.04962873, + "balance_loss_mlp": 1.02380753, + "epoch": 0.43708290747309564, + "flos": 25302694775040.0, + "grad_norm": 2.0521883574899897, + "language_loss": 0.76963937, + "learning_rate": 2.499431731625355e-06, + "loss": 0.79167724, + "num_input_tokens_seen": 78307980, + "step": 3635, + "time_per_iteration": 2.8005919456481934 + }, + { + "auxiliary_loss_clip": 0.01186933, + "auxiliary_loss_mlp": 0.01031568, + "balance_loss_clip": 1.05238724, + "balance_loss_mlp": 1.02252626, + "epoch": 0.43720315036373475, + "flos": 31575103344000.0, + "grad_norm": 1.9931790060597745, + "language_loss": 0.79247093, + "learning_rate": 2.4986774005952686e-06, + "loss": 0.8146559, + "num_input_tokens_seen": 78330355, + "step": 3636, + "time_per_iteration": 2.810739755630493 + }, + { + "auxiliary_loss_clip": 0.01172683, + "auxiliary_loss_mlp": 0.01029227, + "balance_loss_clip": 1.05207503, + "balance_loss_mlp": 1.02005386, + "epoch": 0.43732339325437386, + "flos": 23112000195840.0, + "grad_norm": 2.187046514704066, + "language_loss": 0.84212101, + "learning_rate": 2.4979229939162166e-06, + "loss": 0.86414009, + "num_input_tokens_seen": 78349135, + "step": 3637, + "time_per_iteration": 2.734360456466675 + }, + { + "auxiliary_loss_clip": 0.01169522, + "auxiliary_loss_mlp": 0.01027155, + "balance_loss_clip": 1.05042315, + "balance_loss_mlp": 1.01848245, + "epoch": 0.4374436361450129, + "flos": 27746272080000.0, + "grad_norm": 1.8914905422878587, + "language_loss": 0.80563617, + "learning_rate": 2.4971685117026433e-06, + "loss": 0.82760286, + "num_input_tokens_seen": 78368900, + "step": 3638, + "time_per_iteration": 2.8286521434783936 + }, + { + "auxiliary_loss_clip": 0.01180001, + "auxiliary_loss_mlp": 0.01028292, + "balance_loss_clip": 1.05278099, + "balance_loss_mlp": 1.01981592, + "epoch": 0.437563879035652, + "flos": 24172370616960.0, + "grad_norm": 1.6435410700014328, + "language_loss": 0.76707226, + "learning_rate": 2.4964139540690018e-06, + "loss": 0.78915519, + "num_input_tokens_seen": 78392235, + "step": 3639, + "time_per_iteration": 2.7742934226989746 + }, + { + "auxiliary_loss_clip": 0.01159162, + "auxiliary_loss_mlp": 0.01032419, + "balance_loss_clip": 1.04785943, + "balance_loss_mlp": 1.02277541, + "epoch": 0.4376841219262911, + "flos": 23477211728640.0, + "grad_norm": 1.8153425242881724, + "language_loss": 0.72829503, + "learning_rate": 2.495659321129758e-06, + "loss": 0.75021088, + "num_input_tokens_seen": 78409980, + "step": 3640, + "time_per_iteration": 2.80456280708313 + }, + { + "auxiliary_loss_clip": 0.01172765, + "auxiliary_loss_mlp": 0.01039249, + "balance_loss_clip": 1.05007684, + "balance_loss_mlp": 1.03017664, + "epoch": 0.4378043648169302, + "flos": 25447809720960.0, + "grad_norm": 1.6853767581019696, + "language_loss": 0.75084668, + "learning_rate": 2.494904612999389e-06, + "loss": 0.7729668, + "num_input_tokens_seen": 78428690, + "step": 3641, + "time_per_iteration": 2.657111883163452 + }, + { + "auxiliary_loss_clip": 0.01070546, + "auxiliary_loss_mlp": 0.01004212, + "balance_loss_clip": 1.01743412, + "balance_loss_mlp": 1.00218546, + "epoch": 0.4379246077075693, + "flos": 53914056986880.0, + "grad_norm": 0.755196556199034, + "language_loss": 0.5651468, + "learning_rate": 2.4941498297923843e-06, + "loss": 0.58589447, + "num_input_tokens_seen": 78489260, + "step": 3642, + "time_per_iteration": 3.2073516845703125 + }, + { + "auxiliary_loss_clip": 0.01172552, + "auxiliary_loss_mlp": 0.01026912, + "balance_loss_clip": 1.0515182, + "balance_loss_mlp": 1.0178225, + "epoch": 0.43804485059820836, + "flos": 20588305605120.0, + "grad_norm": 1.6916787544734089, + "language_loss": 0.69920933, + "learning_rate": 2.4933949716232424e-06, + "loss": 0.72120404, + "num_input_tokens_seen": 78506785, + "step": 3643, + "time_per_iteration": 2.7597248554229736 + }, + { + "auxiliary_loss_clip": 0.0115291, + "auxiliary_loss_mlp": 0.01031659, + "balance_loss_clip": 1.05045605, + "balance_loss_mlp": 1.02287304, + "epoch": 0.43816509348884747, + "flos": 23876214981120.0, + "grad_norm": 2.1811071798169213, + "language_loss": 0.73674667, + "learning_rate": 2.492640038606476e-06, + "loss": 0.75859237, + "num_input_tokens_seen": 78525150, + "step": 3644, + "time_per_iteration": 2.77107310295105 + }, + { + "auxiliary_loss_clip": 0.01176142, + "auxiliary_loss_mlp": 0.01028623, + "balance_loss_clip": 1.05061197, + "balance_loss_mlp": 1.01955771, + "epoch": 0.4382853363794866, + "flos": 14684448533760.0, + "grad_norm": 1.89185133842533, + "language_loss": 0.78786886, + "learning_rate": 2.491885030856608e-06, + "loss": 0.8099165, + "num_input_tokens_seen": 78543245, + "step": 3645, + "time_per_iteration": 2.699516534805298 + }, + { + "auxiliary_loss_clip": 0.01165644, + "auxiliary_loss_mlp": 0.01035163, + "balance_loss_clip": 1.0490551, + "balance_loss_mlp": 1.02604961, + "epoch": 0.43840557927012563, + "flos": 17165301177600.0, + "grad_norm": 2.1104879833074723, + "language_loss": 0.82555389, + "learning_rate": 2.4911299484881713e-06, + "loss": 0.84756196, + "num_input_tokens_seen": 78560775, + "step": 3646, + "time_per_iteration": 2.721155881881714 + }, + { + "auxiliary_loss_clip": 0.01160127, + "auxiliary_loss_mlp": 0.0103339, + "balance_loss_clip": 1.0473206, + "balance_loss_mlp": 1.02482474, + "epoch": 0.43852582216076474, + "flos": 19390685316480.0, + "grad_norm": 1.7604590347414135, + "language_loss": 0.80957472, + "learning_rate": 2.490374791615712e-06, + "loss": 0.83150983, + "num_input_tokens_seen": 78580800, + "step": 3647, + "time_per_iteration": 2.788037061691284 + }, + { + "auxiliary_loss_clip": 0.01189167, + "auxiliary_loss_mlp": 0.00887153, + "balance_loss_clip": 1.0545311, + "balance_loss_mlp": 1.00069177, + "epoch": 0.43864606505140386, + "flos": 18075133699200.0, + "grad_norm": 2.969267034653321, + "language_loss": 0.77435058, + "learning_rate": 2.4896195603537867e-06, + "loss": 0.7951138, + "num_input_tokens_seen": 78595410, + "step": 3648, + "time_per_iteration": 2.6964781284332275 + }, + { + "auxiliary_loss_clip": 0.01145065, + "auxiliary_loss_mlp": 0.01034959, + "balance_loss_clip": 1.05043125, + "balance_loss_mlp": 1.02602458, + "epoch": 0.4387663079420429, + "flos": 19644896845440.0, + "grad_norm": 2.5671727069100014, + "language_loss": 0.74093187, + "learning_rate": 2.488864254816964e-06, + "loss": 0.76273215, + "num_input_tokens_seen": 78614100, + "step": 3649, + "time_per_iteration": 2.808204412460327 + }, + { + "auxiliary_loss_clip": 0.01180901, + "auxiliary_loss_mlp": 0.01034677, + "balance_loss_clip": 1.05509543, + "balance_loss_mlp": 1.02534342, + "epoch": 0.438886550832682, + "flos": 19719339782400.0, + "grad_norm": 2.2615341837327727, + "language_loss": 0.6788063, + "learning_rate": 2.4881088751198218e-06, + "loss": 0.70096213, + "num_input_tokens_seen": 78632260, + "step": 3650, + "time_per_iteration": 2.6486103534698486 + }, + { + "auxiliary_loss_clip": 0.01165225, + "auxiliary_loss_mlp": 0.01030241, + "balance_loss_clip": 1.04723132, + "balance_loss_mlp": 1.02087748, + "epoch": 0.43900679372332113, + "flos": 14536675981440.0, + "grad_norm": 3.347949983408385, + "language_loss": 0.64604568, + "learning_rate": 2.4873534213769517e-06, + "loss": 0.66800034, + "num_input_tokens_seen": 78647490, + "step": 3651, + "time_per_iteration": 4.55400276184082 + }, + { + "auxiliary_loss_clip": 0.01150898, + "auxiliary_loss_mlp": 0.01035312, + "balance_loss_clip": 1.0498544, + "balance_loss_mlp": 1.02691925, + "epoch": 0.4391270366139602, + "flos": 24056234968320.0, + "grad_norm": 1.6524492964686557, + "language_loss": 0.71563333, + "learning_rate": 2.4865978937029547e-06, + "loss": 0.73749542, + "num_input_tokens_seen": 78666470, + "step": 3652, + "time_per_iteration": 2.7937676906585693 + }, + { + "auxiliary_loss_clip": 0.01137071, + "auxiliary_loss_mlp": 0.01034291, + "balance_loss_clip": 1.04606676, + "balance_loss_mlp": 1.02572548, + "epoch": 0.4392472795045993, + "flos": 31538510363520.0, + "grad_norm": 1.622971410739856, + "language_loss": 0.66049331, + "learning_rate": 2.485842292212445e-06, + "loss": 0.68220693, + "num_input_tokens_seen": 78687685, + "step": 3653, + "time_per_iteration": 3.854316234588623 + }, + { + "auxiliary_loss_clip": 0.01191029, + "auxiliary_loss_mlp": 0.01029754, + "balance_loss_clip": 1.05659401, + "balance_loss_mlp": 1.02068806, + "epoch": 0.4393675223952384, + "flos": 14866300114560.0, + "grad_norm": 1.842062497425909, + "language_loss": 0.80472636, + "learning_rate": 2.485086617020045e-06, + "loss": 0.82693422, + "num_input_tokens_seen": 78706180, + "step": 3654, + "time_per_iteration": 2.6424126625061035 + }, + { + "auxiliary_loss_clip": 0.01159486, + "auxiliary_loss_mlp": 0.01028902, + "balance_loss_clip": 1.04717243, + "balance_loss_mlp": 1.01976418, + "epoch": 0.43948776528587746, + "flos": 14825900292480.0, + "grad_norm": 2.123732667396018, + "language_loss": 0.8220011, + "learning_rate": 2.4843308682403903e-06, + "loss": 0.84388494, + "num_input_tokens_seen": 78723095, + "step": 3655, + "time_per_iteration": 2.894885301589966 + }, + { + "auxiliary_loss_clip": 0.0118423, + "auxiliary_loss_mlp": 0.01029541, + "balance_loss_clip": 1.05200481, + "balance_loss_mlp": 1.02123249, + "epoch": 0.4396080081765166, + "flos": 13914523486080.0, + "grad_norm": 1.7359696007215288, + "language_loss": 0.82663512, + "learning_rate": 2.4835750459881294e-06, + "loss": 0.84877288, + "num_input_tokens_seen": 78739720, + "step": 3656, + "time_per_iteration": 2.700632095336914 + }, + { + "auxiliary_loss_clip": 0.01158755, + "auxiliary_loss_mlp": 0.01030654, + "balance_loss_clip": 1.04701602, + "balance_loss_mlp": 1.02208269, + "epoch": 0.43972825106715563, + "flos": 18222978078720.0, + "grad_norm": 2.94276525349212, + "language_loss": 0.81955862, + "learning_rate": 2.4828191503779177e-06, + "loss": 0.84145272, + "num_input_tokens_seen": 78757820, + "step": 3657, + "time_per_iteration": 2.82377552986145 + }, + { + "auxiliary_loss_clip": 0.01154265, + "auxiliary_loss_mlp": 0.01027167, + "balance_loss_clip": 1.04558992, + "balance_loss_mlp": 1.01839328, + "epoch": 0.43984849395779474, + "flos": 16873239692160.0, + "grad_norm": 1.9983813351932727, + "language_loss": 0.89564073, + "learning_rate": 2.482063181524425e-06, + "loss": 0.91745508, + "num_input_tokens_seen": 78773720, + "step": 3658, + "time_per_iteration": 2.9506142139434814 + }, + { + "auxiliary_loss_clip": 0.01189778, + "auxiliary_loss_mlp": 0.01037421, + "balance_loss_clip": 1.05476809, + "balance_loss_mlp": 1.02828395, + "epoch": 0.43996873684843385, + "flos": 18691504104960.0, + "grad_norm": 2.245998734063297, + "language_loss": 0.81132871, + "learning_rate": 2.4813071395423307e-06, + "loss": 0.8336007, + "num_input_tokens_seen": 78791285, + "step": 3659, + "time_per_iteration": 2.7935147285461426 + }, + { + "auxiliary_loss_clip": 0.011759, + "auxiliary_loss_mlp": 0.01030167, + "balance_loss_clip": 1.0490036, + "balance_loss_mlp": 1.02085721, + "epoch": 0.4400889797390729, + "flos": 23653460787840.0, + "grad_norm": 1.7772849899543768, + "language_loss": 0.64451528, + "learning_rate": 2.4805510245463263e-06, + "loss": 0.66657597, + "num_input_tokens_seen": 78811440, + "step": 3660, + "time_per_iteration": 2.794468641281128 + }, + { + "auxiliary_loss_clip": 0.01176853, + "auxiliary_loss_mlp": 0.01039633, + "balance_loss_clip": 1.04998291, + "balance_loss_mlp": 1.03068018, + "epoch": 0.440209222629712, + "flos": 23149203707520.0, + "grad_norm": 2.093762841765101, + "language_loss": 0.60316509, + "learning_rate": 2.4797948366511137e-06, + "loss": 0.62532997, + "num_input_tokens_seen": 78831150, + "step": 3661, + "time_per_iteration": 2.762967109680176 + }, + { + "auxiliary_loss_clip": 0.01152511, + "auxiliary_loss_mlp": 0.01038018, + "balance_loss_clip": 1.04405177, + "balance_loss_mlp": 1.02885067, + "epoch": 0.4403294655203511, + "flos": 24823394668800.0, + "grad_norm": 1.9175012992619405, + "language_loss": 0.75885427, + "learning_rate": 2.4790385759714055e-06, + "loss": 0.78075957, + "num_input_tokens_seen": 78850215, + "step": 3662, + "time_per_iteration": 2.976367950439453 + }, + { + "auxiliary_loss_clip": 0.01173103, + "auxiliary_loss_mlp": 0.01032117, + "balance_loss_clip": 1.05293298, + "balance_loss_mlp": 1.02370071, + "epoch": 0.4404497084109902, + "flos": 22565080736640.0, + "grad_norm": 1.6849937379171336, + "language_loss": 0.71379536, + "learning_rate": 2.478282242621926e-06, + "loss": 0.73584759, + "num_input_tokens_seen": 78870675, + "step": 3663, + "time_per_iteration": 2.7654402256011963 + }, + { + "auxiliary_loss_clip": 0.0106138, + "auxiliary_loss_mlp": 0.01005971, + "balance_loss_clip": 1.01804709, + "balance_loss_mlp": 1.00402784, + "epoch": 0.4405699513016293, + "flos": 64967073448320.0, + "grad_norm": 0.8417696659152991, + "language_loss": 0.59522492, + "learning_rate": 2.477525836717411e-06, + "loss": 0.61589843, + "num_input_tokens_seen": 78938440, + "step": 3664, + "time_per_iteration": 3.545836925506592 + }, + { + "auxiliary_loss_clip": 0.01177185, + "auxiliary_loss_mlp": 0.01028177, + "balance_loss_clip": 1.05069232, + "balance_loss_mlp": 1.01904607, + "epoch": 0.4406901941922684, + "flos": 35661952978560.0, + "grad_norm": 2.7300669076189217, + "language_loss": 0.79983521, + "learning_rate": 2.476769358372606e-06, + "loss": 0.82188886, + "num_input_tokens_seen": 78960090, + "step": 3665, + "time_per_iteration": 2.9435367584228516 + }, + { + "auxiliary_loss_clip": 0.01156901, + "auxiliary_loss_mlp": 0.01029253, + "balance_loss_clip": 1.05077231, + "balance_loss_mlp": 1.02089667, + "epoch": 0.44081043708290746, + "flos": 18040767361920.0, + "grad_norm": 2.0429140461693303, + "language_loss": 0.74956572, + "learning_rate": 2.4760128077022683e-06, + "loss": 0.77142721, + "num_input_tokens_seen": 78978225, + "step": 3666, + "time_per_iteration": 2.802755355834961 + }, + { + "auxiliary_loss_clip": 0.01135416, + "auxiliary_loss_mlp": 0.01032561, + "balance_loss_clip": 1.04596829, + "balance_loss_mlp": 1.02388871, + "epoch": 0.44093067997354657, + "flos": 30153507799680.0, + "grad_norm": 1.595162585616692, + "language_loss": 0.68625104, + "learning_rate": 2.4752561848211672e-06, + "loss": 0.70793086, + "num_input_tokens_seen": 79000625, + "step": 3667, + "time_per_iteration": 2.939840316772461 + }, + { + "auxiliary_loss_clip": 0.01176764, + "auxiliary_loss_mlp": 0.01028764, + "balance_loss_clip": 1.05622351, + "balance_loss_mlp": 1.02077675, + "epoch": 0.4410509228641857, + "flos": 23255068066560.0, + "grad_norm": 2.68595075527433, + "language_loss": 0.71496594, + "learning_rate": 2.4744994898440797e-06, + "loss": 0.73702121, + "num_input_tokens_seen": 79019415, + "step": 3668, + "time_per_iteration": 2.7291641235351562 + }, + { + "auxiliary_loss_clip": 0.01163172, + "auxiliary_loss_mlp": 0.01024932, + "balance_loss_clip": 1.05084872, + "balance_loss_mlp": 1.01550901, + "epoch": 0.44117116575482473, + "flos": 19500571998720.0, + "grad_norm": 2.0582199201830718, + "language_loss": 0.83572924, + "learning_rate": 2.473742722885797e-06, + "loss": 0.85761029, + "num_input_tokens_seen": 79038435, + "step": 3669, + "time_per_iteration": 2.9204440116882324 + }, + { + "auxiliary_loss_clip": 0.01178959, + "auxiliary_loss_mlp": 0.00887568, + "balance_loss_clip": 1.05556011, + "balance_loss_mlp": 1.00078046, + "epoch": 0.44129140864546385, + "flos": 27053124353280.0, + "grad_norm": 2.425895893820238, + "language_loss": 0.65299118, + "learning_rate": 2.4729858840611197e-06, + "loss": 0.6736564, + "num_input_tokens_seen": 79057345, + "step": 3670, + "time_per_iteration": 2.848947525024414 + }, + { + "auxiliary_loss_clip": 0.01185493, + "auxiliary_loss_mlp": 0.01029866, + "balance_loss_clip": 1.05341065, + "balance_loss_mlp": 1.02148604, + "epoch": 0.4414116515361029, + "flos": 26102101910400.0, + "grad_norm": 2.202360339098736, + "language_loss": 0.72447026, + "learning_rate": 2.4722289734848605e-06, + "loss": 0.74662387, + "num_input_tokens_seen": 79077810, + "step": 3671, + "time_per_iteration": 2.7937190532684326 + }, + { + "auxiliary_loss_clip": 0.0115624, + "auxiliary_loss_mlp": 0.01031207, + "balance_loss_clip": 1.05247712, + "balance_loss_mlp": 1.02247453, + "epoch": 0.441531894426742, + "flos": 21906083865600.0, + "grad_norm": 2.2603674388206243, + "language_loss": 0.77521861, + "learning_rate": 2.471471991271841e-06, + "loss": 0.79709303, + "num_input_tokens_seen": 79094935, + "step": 3672, + "time_per_iteration": 2.843658924102783 + }, + { + "auxiliary_loss_clip": 0.01169838, + "auxiliary_loss_mlp": 0.010301, + "balance_loss_clip": 1.04963541, + "balance_loss_mlp": 1.02049768, + "epoch": 0.4416521373173811, + "flos": 23437099215360.0, + "grad_norm": 1.977919508853521, + "language_loss": 0.79362541, + "learning_rate": 2.470714937536896e-06, + "loss": 0.81562477, + "num_input_tokens_seen": 79113660, + "step": 3673, + "time_per_iteration": 2.749636650085449 + }, + { + "auxiliary_loss_clip": 0.01144464, + "auxiliary_loss_mlp": 0.01034344, + "balance_loss_clip": 1.04890299, + "balance_loss_mlp": 1.02551663, + "epoch": 0.4417723802080202, + "flos": 20334345471360.0, + "grad_norm": 2.233609516206399, + "language_loss": 0.70607877, + "learning_rate": 2.469957812394868e-06, + "loss": 0.72786677, + "num_input_tokens_seen": 79132470, + "step": 3674, + "time_per_iteration": 2.892954111099243 + }, + { + "auxiliary_loss_clip": 0.01187065, + "auxiliary_loss_mlp": 0.01028497, + "balance_loss_clip": 1.05520201, + "balance_loss_mlp": 1.01956832, + "epoch": 0.4418926230986593, + "flos": 18880682060160.0, + "grad_norm": 2.1403825636904634, + "language_loss": 0.75851887, + "learning_rate": 2.4692006159606148e-06, + "loss": 0.78067446, + "num_input_tokens_seen": 79150000, + "step": 3675, + "time_per_iteration": 2.5864157676696777 + }, + { + "auxiliary_loss_clip": 0.01185046, + "auxiliary_loss_mlp": 0.01028074, + "balance_loss_clip": 1.05192637, + "balance_loss_mlp": 1.01918101, + "epoch": 0.4420128659892984, + "flos": 19464409981440.0, + "grad_norm": 2.0342827610084013, + "language_loss": 0.78903389, + "learning_rate": 2.468443348349e-06, + "loss": 0.81116509, + "num_input_tokens_seen": 79167875, + "step": 3676, + "time_per_iteration": 3.6711199283599854 + }, + { + "auxiliary_loss_clip": 0.01145317, + "auxiliary_loss_mlp": 0.01033119, + "balance_loss_clip": 1.04587078, + "balance_loss_mlp": 1.02392232, + "epoch": 0.44213310887993745, + "flos": 17894359526400.0, + "grad_norm": 3.9105020340211825, + "language_loss": 0.81976414, + "learning_rate": 2.467686009674902e-06, + "loss": 0.8415485, + "num_input_tokens_seen": 79182325, + "step": 3677, + "time_per_iteration": 3.8199992179870605 + }, + { + "auxiliary_loss_clip": 0.01169871, + "auxiliary_loss_mlp": 0.01028196, + "balance_loss_clip": 1.04834533, + "balance_loss_mlp": 1.01855767, + "epoch": 0.44225335177057656, + "flos": 19204667758080.0, + "grad_norm": 1.8829100615731569, + "language_loss": 0.84974229, + "learning_rate": 2.466928600053209e-06, + "loss": 0.871723, + "num_input_tokens_seen": 79197630, + "step": 3678, + "time_per_iteration": 2.7530922889709473 + }, + { + "auxiliary_loss_clip": 0.01161646, + "auxiliary_loss_mlp": 0.01029733, + "balance_loss_clip": 1.04785132, + "balance_loss_mlp": 1.02068532, + "epoch": 0.4423735946612157, + "flos": 23471321898240.0, + "grad_norm": 2.636956824399565, + "language_loss": 0.71162075, + "learning_rate": 2.466171119598818e-06, + "loss": 0.73353457, + "num_input_tokens_seen": 79217600, + "step": 3679, + "time_per_iteration": 3.7261834144592285 + }, + { + "auxiliary_loss_clip": 0.01182326, + "auxiliary_loss_mlp": 0.01032044, + "balance_loss_clip": 1.05055475, + "balance_loss_mlp": 1.02271605, + "epoch": 0.44249383755185473, + "flos": 26685398868480.0, + "grad_norm": 1.8564927588334756, + "language_loss": 0.77027822, + "learning_rate": 2.465413568426639e-06, + "loss": 0.79242194, + "num_input_tokens_seen": 79238550, + "step": 3680, + "time_per_iteration": 2.824237108230591 + }, + { + "auxiliary_loss_clip": 0.01166935, + "auxiliary_loss_mlp": 0.01030691, + "balance_loss_clip": 1.04808807, + "balance_loss_mlp": 1.02224493, + "epoch": 0.44261408044249384, + "flos": 23147659422720.0, + "grad_norm": 1.8013982891219937, + "language_loss": 0.8137995, + "learning_rate": 2.464655946651591e-06, + "loss": 0.83577579, + "num_input_tokens_seen": 79257555, + "step": 3681, + "time_per_iteration": 2.782808303833008 + }, + { + "auxiliary_loss_clip": 0.0117689, + "auxiliary_loss_mlp": 0.01029578, + "balance_loss_clip": 1.05126762, + "balance_loss_mlp": 1.02032721, + "epoch": 0.44273432333313295, + "flos": 24462564595200.0, + "grad_norm": 2.0592209306080638, + "language_loss": 0.80873048, + "learning_rate": 2.4638982543886065e-06, + "loss": 0.83079517, + "num_input_tokens_seen": 79277595, + "step": 3682, + "time_per_iteration": 2.8458213806152344 + }, + { + "auxiliary_loss_clip": 0.011835, + "auxiliary_loss_mlp": 0.01035787, + "balance_loss_clip": 1.05560875, + "balance_loss_mlp": 1.02564216, + "epoch": 0.442854566223772, + "flos": 17528932512000.0, + "grad_norm": 2.2183658096788452, + "language_loss": 0.87329876, + "learning_rate": 2.4631404917526254e-06, + "loss": 0.8954916, + "num_input_tokens_seen": 79294550, + "step": 3683, + "time_per_iteration": 2.645026922225952 + }, + { + "auxiliary_loss_clip": 0.01169636, + "auxiliary_loss_mlp": 0.01029843, + "balance_loss_clip": 1.04977858, + "balance_loss_mlp": 1.02045584, + "epoch": 0.4429748091144111, + "flos": 24896293320960.0, + "grad_norm": 1.7713763147702328, + "language_loss": 0.79048908, + "learning_rate": 2.4623826588586e-06, + "loss": 0.81248385, + "num_input_tokens_seen": 79314820, + "step": 3684, + "time_per_iteration": 2.780388355255127 + }, + { + "auxiliary_loss_clip": 0.01162757, + "auxiliary_loss_mlp": 0.01034306, + "balance_loss_clip": 1.04745579, + "balance_loss_mlp": 1.0246501, + "epoch": 0.4430950520050502, + "flos": 21614704738560.0, + "grad_norm": 1.555669253980653, + "language_loss": 0.82756877, + "learning_rate": 2.461624755821492e-06, + "loss": 0.8495394, + "num_input_tokens_seen": 79334300, + "step": 3685, + "time_per_iteration": 3.1185169219970703 + }, + { + "auxiliary_loss_clip": 0.01156699, + "auxiliary_loss_mlp": 0.01032415, + "balance_loss_clip": 1.0506053, + "balance_loss_mlp": 1.02382624, + "epoch": 0.4432152948956893, + "flos": 24572271709440.0, + "grad_norm": 2.6720790908450254, + "language_loss": 0.7671057, + "learning_rate": 2.4608667827562763e-06, + "loss": 0.78899682, + "num_input_tokens_seen": 79353630, + "step": 3686, + "time_per_iteration": 2.884392738342285 + }, + { + "auxiliary_loss_clip": 0.0118308, + "auxiliary_loss_mlp": 0.01035851, + "balance_loss_clip": 1.05381012, + "balance_loss_mlp": 1.02651691, + "epoch": 0.4433355377863284, + "flos": 21762261809280.0, + "grad_norm": 1.9483790153578617, + "language_loss": 0.90196645, + "learning_rate": 2.460108739777936e-06, + "loss": 0.92415577, + "num_input_tokens_seen": 79372765, + "step": 3687, + "time_per_iteration": 2.7073841094970703 + }, + { + "auxiliary_loss_clip": 0.01165517, + "auxiliary_loss_mlp": 0.01037804, + "balance_loss_clip": 1.05010915, + "balance_loss_mlp": 1.02869678, + "epoch": 0.44345578067696745, + "flos": 20084479488000.0, + "grad_norm": 1.538042153281284, + "language_loss": 0.7615304, + "learning_rate": 2.4593506270014656e-06, + "loss": 0.78356361, + "num_input_tokens_seen": 79391735, + "step": 3688, + "time_per_iteration": 2.703139543533325 + }, + { + "auxiliary_loss_clip": 0.01165797, + "auxiliary_loss_mlp": 0.01034742, + "balance_loss_clip": 1.04674149, + "balance_loss_mlp": 1.02557516, + "epoch": 0.44357602356760656, + "flos": 24169497528960.0, + "grad_norm": 1.818807379089097, + "language_loss": 0.82274395, + "learning_rate": 2.45859244454187e-06, + "loss": 0.84474933, + "num_input_tokens_seen": 79411525, + "step": 3689, + "time_per_iteration": 2.815704584121704 + }, + { + "auxiliary_loss_clip": 0.01173037, + "auxiliary_loss_mlp": 0.01035665, + "balance_loss_clip": 1.05153465, + "balance_loss_mlp": 1.02724862, + "epoch": 0.44369626645824567, + "flos": 22707717644160.0, + "grad_norm": 1.9489350736997766, + "language_loss": 0.66664398, + "learning_rate": 2.4578341925141655e-06, + "loss": 0.68873096, + "num_input_tokens_seen": 79430740, + "step": 3690, + "time_per_iteration": 2.812504529953003 + }, + { + "auxiliary_loss_clip": 0.01182763, + "auxiliary_loss_mlp": 0.01028371, + "balance_loss_clip": 1.05097055, + "balance_loss_mlp": 1.01888764, + "epoch": 0.4438165093488847, + "flos": 38030225420160.0, + "grad_norm": 1.954771350099959, + "language_loss": 0.72434944, + "learning_rate": 2.457075871033378e-06, + "loss": 0.7464608, + "num_input_tokens_seen": 79452615, + "step": 3691, + "time_per_iteration": 2.95064640045166 + }, + { + "auxiliary_loss_clip": 0.01152803, + "auxiliary_loss_mlp": 0.01033067, + "balance_loss_clip": 1.0469749, + "balance_loss_mlp": 1.02361345, + "epoch": 0.44393675223952384, + "flos": 15523213996800.0, + "grad_norm": 2.0324710601578624, + "language_loss": 0.88625592, + "learning_rate": 2.4563174802145445e-06, + "loss": 0.90811467, + "num_input_tokens_seen": 79469865, + "step": 3692, + "time_per_iteration": 2.7969565391540527 + }, + { + "auxiliary_loss_clip": 0.01069501, + "auxiliary_loss_mlp": 0.01004736, + "balance_loss_clip": 1.01988435, + "balance_loss_mlp": 1.00256681, + "epoch": 0.44405699513016295, + "flos": 64574893779840.0, + "grad_norm": 0.6375113714266561, + "language_loss": 0.48584312, + "learning_rate": 2.455559020172712e-06, + "loss": 0.50658548, + "num_input_tokens_seen": 79537220, + "step": 3693, + "time_per_iteration": 3.373257875442505 + }, + { + "auxiliary_loss_clip": 0.01148702, + "auxiliary_loss_mlp": 0.01039604, + "balance_loss_clip": 1.04938078, + "balance_loss_mlp": 1.03028762, + "epoch": 0.444177238020802, + "flos": 23987394552960.0, + "grad_norm": 2.2930986891683154, + "language_loss": 0.89732504, + "learning_rate": 2.4548004910229385e-06, + "loss": 0.91920805, + "num_input_tokens_seen": 79554795, + "step": 3694, + "time_per_iteration": 2.71921443939209 + }, + { + "auxiliary_loss_clip": 0.01178629, + "auxiliary_loss_mlp": 0.00886906, + "balance_loss_clip": 1.05284917, + "balance_loss_mlp": 1.0008589, + "epoch": 0.4442974809114411, + "flos": 22563069575040.0, + "grad_norm": 1.743898940197062, + "language_loss": 0.87038791, + "learning_rate": 2.4540418928802913e-06, + "loss": 0.89104331, + "num_input_tokens_seen": 79573530, + "step": 3695, + "time_per_iteration": 2.5829660892486572 + }, + { + "auxiliary_loss_clip": 0.01167938, + "auxiliary_loss_mlp": 0.01033855, + "balance_loss_clip": 1.05122924, + "balance_loss_mlp": 1.02428246, + "epoch": 0.4444177238020802, + "flos": 17675699483520.0, + "grad_norm": 2.436809272618502, + "language_loss": 0.65776587, + "learning_rate": 2.4532832258598506e-06, + "loss": 0.67978382, + "num_input_tokens_seen": 79591360, + "step": 3696, + "time_per_iteration": 2.62404727935791 + }, + { + "auxiliary_loss_clip": 0.01184024, + "auxiliary_loss_mlp": 0.01031531, + "balance_loss_clip": 1.05192554, + "balance_loss_mlp": 1.02272737, + "epoch": 0.4445379666927193, + "flos": 28621594609920.0, + "grad_norm": 1.7152942979390007, + "language_loss": 0.80674428, + "learning_rate": 2.4525244900767047e-06, + "loss": 0.8288998, + "num_input_tokens_seen": 79612175, + "step": 3697, + "time_per_iteration": 2.569221019744873 + }, + { + "auxiliary_loss_clip": 0.01067133, + "auxiliary_loss_mlp": 0.01003853, + "balance_loss_clip": 1.01763582, + "balance_loss_mlp": 1.00187421, + "epoch": 0.4446582095833584, + "flos": 70487370115200.0, + "grad_norm": 0.7721946109055748, + "language_loss": 0.60462868, + "learning_rate": 2.4517656856459536e-06, + "loss": 0.62533855, + "num_input_tokens_seen": 79678020, + "step": 3698, + "time_per_iteration": 3.2649669647216797 + }, + { + "auxiliary_loss_clip": 0.01176873, + "auxiliary_loss_mlp": 0.01037344, + "balance_loss_clip": 1.04991114, + "balance_loss_mlp": 1.02847445, + "epoch": 0.4447784524739975, + "flos": 26505199313280.0, + "grad_norm": 2.2640894232830964, + "language_loss": 0.68010712, + "learning_rate": 2.4510068126827073e-06, + "loss": 0.70224929, + "num_input_tokens_seen": 79699020, + "step": 3699, + "time_per_iteration": 2.6472980976104736 + }, + { + "auxiliary_loss_clip": 0.01164787, + "auxiliary_loss_mlp": 0.01028648, + "balance_loss_clip": 1.04891121, + "balance_loss_mlp": 1.02030981, + "epoch": 0.44489869536463655, + "flos": 11656209553920.0, + "grad_norm": 6.183815262086186, + "language_loss": 0.8160665, + "learning_rate": 2.450247871302086e-06, + "loss": 0.83800089, + "num_input_tokens_seen": 79716795, + "step": 3700, + "time_per_iteration": 2.782902956008911 + }, + { + "auxiliary_loss_clip": 0.01176592, + "auxiliary_loss_mlp": 0.01034057, + "balance_loss_clip": 1.05107713, + "balance_loss_mlp": 1.02515197, + "epoch": 0.44501893825527566, + "flos": 20448469958400.0, + "grad_norm": 2.0561955394448583, + "language_loss": 0.83065271, + "learning_rate": 2.44948886161922e-06, + "loss": 0.85275924, + "num_input_tokens_seen": 79735810, + "step": 3701, + "time_per_iteration": 2.7697267532348633 + }, + { + "auxiliary_loss_clip": 0.01181241, + "auxiliary_loss_mlp": 0.01029449, + "balance_loss_clip": 1.05309296, + "balance_loss_mlp": 1.0209136, + "epoch": 0.4451391811459148, + "flos": 18261079430400.0, + "grad_norm": 1.7731845162085254, + "language_loss": 0.8487944, + "learning_rate": 2.4487297837492524e-06, + "loss": 0.87090129, + "num_input_tokens_seen": 79754975, + "step": 3702, + "time_per_iteration": 3.8445703983306885 + }, + { + "auxiliary_loss_clip": 0.01158231, + "auxiliary_loss_mlp": 0.01036773, + "balance_loss_clip": 1.0483402, + "balance_loss_mlp": 1.02767146, + "epoch": 0.44525942403655383, + "flos": 16910155895040.0, + "grad_norm": 1.8874128798989478, + "language_loss": 0.62085301, + "learning_rate": 2.4479706378073323e-06, + "loss": 0.64280307, + "num_input_tokens_seen": 79773515, + "step": 3703, + "time_per_iteration": 3.9103376865386963 + }, + { + "auxiliary_loss_clip": 0.01147682, + "auxiliary_loss_mlp": 0.01030462, + "balance_loss_clip": 1.0456214, + "balance_loss_mlp": 1.02193904, + "epoch": 0.44537966692719294, + "flos": 23258838994560.0, + "grad_norm": 1.6264829022503908, + "language_loss": 0.83995712, + "learning_rate": 2.447211423908623e-06, + "loss": 0.86173856, + "num_input_tokens_seen": 79793560, + "step": 3704, + "time_per_iteration": 3.0393199920654297 + }, + { + "auxiliary_loss_clip": 0.01178312, + "auxiliary_loss_mlp": 0.01031993, + "balance_loss_clip": 1.05034518, + "balance_loss_mlp": 1.02333856, + "epoch": 0.445499909817832, + "flos": 21724160457600.0, + "grad_norm": 2.1609134730667217, + "language_loss": 0.75317526, + "learning_rate": 2.4464521421682966e-06, + "loss": 0.77527833, + "num_input_tokens_seen": 79811150, + "step": 3705, + "time_per_iteration": 4.605870246887207 + }, + { + "auxiliary_loss_clip": 0.0117205, + "auxiliary_loss_mlp": 0.01029665, + "balance_loss_clip": 1.052441, + "balance_loss_mlp": 1.02067041, + "epoch": 0.4456201527084711, + "flos": 23987969170560.0, + "grad_norm": 2.094929078914078, + "language_loss": 0.87496459, + "learning_rate": 2.4456927927015345e-06, + "loss": 0.89698172, + "num_input_tokens_seen": 79832190, + "step": 3706, + "time_per_iteration": 2.818493127822876 + }, + { + "auxiliary_loss_clip": 0.01175135, + "auxiliary_loss_mlp": 0.0103333, + "balance_loss_clip": 1.05382454, + "balance_loss_mlp": 1.02325058, + "epoch": 0.4457403955991102, + "flos": 18807065136000.0, + "grad_norm": 3.169357660808316, + "language_loss": 0.76616138, + "learning_rate": 2.4449333756235307e-06, + "loss": 0.78824604, + "num_input_tokens_seen": 79848905, + "step": 3707, + "time_per_iteration": 2.857517957687378 + }, + { + "auxiliary_loss_clip": 0.01179386, + "auxiliary_loss_mlp": 0.0103561, + "balance_loss_clip": 1.05411863, + "balance_loss_mlp": 1.02696705, + "epoch": 0.4458606384897493, + "flos": 19207756327680.0, + "grad_norm": 2.7296409551515946, + "language_loss": 0.7898615, + "learning_rate": 2.4441738910494876e-06, + "loss": 0.81201148, + "num_input_tokens_seen": 79863640, + "step": 3708, + "time_per_iteration": 2.7426159381866455 + }, + { + "auxiliary_loss_clip": 0.01173997, + "auxiliary_loss_mlp": 0.01035916, + "balance_loss_clip": 1.05001163, + "balance_loss_mlp": 1.02644491, + "epoch": 0.4459808813803884, + "flos": 21361283308800.0, + "grad_norm": 1.9216719895600354, + "language_loss": 0.82026893, + "learning_rate": 2.4434143390946176e-06, + "loss": 0.84236807, + "num_input_tokens_seen": 79882450, + "step": 3709, + "time_per_iteration": 2.8081140518188477 + }, + { + "auxiliary_loss_clip": 0.01157867, + "auxiliary_loss_mlp": 0.0103669, + "balance_loss_clip": 1.05070198, + "balance_loss_mlp": 1.02761269, + "epoch": 0.4461011242710275, + "flos": 23288967527040.0, + "grad_norm": 1.9907217286894805, + "language_loss": 0.85690892, + "learning_rate": 2.4426547198741457e-06, + "loss": 0.87885451, + "num_input_tokens_seen": 79900655, + "step": 3710, + "time_per_iteration": 2.809128999710083 + }, + { + "auxiliary_loss_clip": 0.01154922, + "auxiliary_loss_mlp": 0.01039477, + "balance_loss_clip": 1.05355549, + "balance_loss_mlp": 1.0300597, + "epoch": 0.44622136716166655, + "flos": 20193001453440.0, + "grad_norm": 2.0604937731540347, + "language_loss": 0.74528217, + "learning_rate": 2.441895033503305e-06, + "loss": 0.76722616, + "num_input_tokens_seen": 79918575, + "step": 3711, + "time_per_iteration": 2.8839809894561768 + }, + { + "auxiliary_loss_clip": 0.0117833, + "auxiliary_loss_mlp": 0.01035726, + "balance_loss_clip": 1.0534972, + "balance_loss_mlp": 1.02587342, + "epoch": 0.44634161005230566, + "flos": 21283033530240.0, + "grad_norm": 1.7685095543845029, + "language_loss": 0.82007587, + "learning_rate": 2.4411352800973375e-06, + "loss": 0.84221649, + "num_input_tokens_seen": 79937010, + "step": 3712, + "time_per_iteration": 2.776895523071289 + }, + { + "auxiliary_loss_clip": 0.01152877, + "auxiliary_loss_mlp": 0.01031672, + "balance_loss_clip": 1.0462749, + "balance_loss_mlp": 1.0214262, + "epoch": 0.44646185294294477, + "flos": 22929358515840.0, + "grad_norm": 2.4252372683652044, + "language_loss": 0.74741495, + "learning_rate": 2.4403754597715005e-06, + "loss": 0.76926041, + "num_input_tokens_seen": 79956455, + "step": 3713, + "time_per_iteration": 2.849032402038574 + }, + { + "auxiliary_loss_clip": 0.01171594, + "auxiliary_loss_mlp": 0.01030337, + "balance_loss_clip": 1.0486306, + "balance_loss_mlp": 1.02023399, + "epoch": 0.4465820958335838, + "flos": 22637692080000.0, + "grad_norm": 2.688391293929975, + "language_loss": 0.92971504, + "learning_rate": 2.4396155726410553e-06, + "loss": 0.95173436, + "num_input_tokens_seen": 79975065, + "step": 3714, + "time_per_iteration": 2.870570659637451 + }, + { + "auxiliary_loss_clip": 0.0117989, + "auxiliary_loss_mlp": 0.01033156, + "balance_loss_clip": 1.05239153, + "balance_loss_mlp": 1.02383351, + "epoch": 0.44670233872422294, + "flos": 22672525294080.0, + "grad_norm": 6.636499486830214, + "language_loss": 0.90785956, + "learning_rate": 2.438855618821278e-06, + "loss": 0.92999005, + "num_input_tokens_seen": 79990865, + "step": 3715, + "time_per_iteration": 2.738234758377075 + }, + { + "auxiliary_loss_clip": 0.01169778, + "auxiliary_loss_mlp": 0.01027074, + "balance_loss_clip": 1.04793286, + "balance_loss_mlp": 1.01781726, + "epoch": 0.44682258161486205, + "flos": 23582178247680.0, + "grad_norm": 1.827736214254627, + "language_loss": 0.6739881, + "learning_rate": 2.4380955984274517e-06, + "loss": 0.69595659, + "num_input_tokens_seen": 80009520, + "step": 3716, + "time_per_iteration": 2.862159490585327 + }, + { + "auxiliary_loss_clip": 0.01178411, + "auxiliary_loss_mlp": 0.01034985, + "balance_loss_clip": 1.05091166, + "balance_loss_mlp": 1.0262711, + "epoch": 0.4469428245055011, + "flos": 26501356558080.0, + "grad_norm": 1.8960940214094453, + "language_loss": 0.77180409, + "learning_rate": 2.4373355115748716e-06, + "loss": 0.79393804, + "num_input_tokens_seen": 80030350, + "step": 3717, + "time_per_iteration": 2.817458391189575 + }, + { + "auxiliary_loss_clip": 0.01159228, + "auxiliary_loss_mlp": 0.01033551, + "balance_loss_clip": 1.04934931, + "balance_loss_mlp": 1.024122, + "epoch": 0.4470630673961402, + "flos": 21504925797120.0, + "grad_norm": 1.7682349677059042, + "language_loss": 0.72305167, + "learning_rate": 2.436575358378842e-06, + "loss": 0.74497944, + "num_input_tokens_seen": 80049840, + "step": 3718, + "time_per_iteration": 2.890357732772827 + }, + { + "auxiliary_loss_clip": 0.01174069, + "auxiliary_loss_mlp": 0.01031949, + "balance_loss_clip": 1.0508064, + "balance_loss_mlp": 1.02136326, + "epoch": 0.44718331028677927, + "flos": 16173986653440.0, + "grad_norm": 3.1623724837101257, + "language_loss": 0.82984054, + "learning_rate": 2.4358151389546782e-06, + "loss": 0.8519007, + "num_input_tokens_seen": 80066525, + "step": 3719, + "time_per_iteration": 2.7536542415618896 + }, + { + "auxiliary_loss_clip": 0.01188089, + "auxiliary_loss_mlp": 0.01028972, + "balance_loss_clip": 1.05458772, + "balance_loss_mlp": 1.01988792, + "epoch": 0.4473035531774184, + "flos": 19681238430720.0, + "grad_norm": 2.312551411715573, + "language_loss": 0.76006293, + "learning_rate": 2.4350548534177035e-06, + "loss": 0.78223354, + "num_input_tokens_seen": 80083355, + "step": 3720, + "time_per_iteration": 2.6497762203216553 + }, + { + "auxiliary_loss_clip": 0.01159762, + "auxiliary_loss_mlp": 0.01031126, + "balance_loss_clip": 1.05041409, + "balance_loss_mlp": 1.0226624, + "epoch": 0.4474237960680575, + "flos": 41427590515200.0, + "grad_norm": 1.7405025484569971, + "language_loss": 0.66514456, + "learning_rate": 2.434294501883254e-06, + "loss": 0.6870535, + "num_input_tokens_seen": 80106450, + "step": 3721, + "time_per_iteration": 2.999135971069336 + }, + { + "auxiliary_loss_clip": 0.01158995, + "auxiliary_loss_mlp": 0.01036608, + "balance_loss_clip": 1.04737282, + "balance_loss_mlp": 1.02751863, + "epoch": 0.44754403895869654, + "flos": 22891328991360.0, + "grad_norm": 1.6658284130291023, + "language_loss": 0.6556654, + "learning_rate": 2.433534084466674e-06, + "loss": 0.67762148, + "num_input_tokens_seen": 80125670, + "step": 3722, + "time_per_iteration": 2.7946698665618896 + }, + { + "auxiliary_loss_clip": 0.01185114, + "auxiliary_loss_mlp": 0.01032538, + "balance_loss_clip": 1.05282569, + "balance_loss_mlp": 1.023103, + "epoch": 0.44766428184933565, + "flos": 25630271832960.0, + "grad_norm": 1.5883681172504427, + "language_loss": 0.70944536, + "learning_rate": 2.4327736012833178e-06, + "loss": 0.73162192, + "num_input_tokens_seen": 80147390, + "step": 3723, + "time_per_iteration": 2.857670307159424 + }, + { + "auxiliary_loss_clip": 0.01175516, + "auxiliary_loss_mlp": 0.0102821, + "balance_loss_clip": 1.05093074, + "balance_loss_mlp": 1.01904905, + "epoch": 0.44778452473997477, + "flos": 20448972748800.0, + "grad_norm": 3.74689622278217, + "language_loss": 0.76610094, + "learning_rate": 2.4320130524485506e-06, + "loss": 0.78813821, + "num_input_tokens_seen": 80166185, + "step": 3724, + "time_per_iteration": 2.7237818241119385 + }, + { + "auxiliary_loss_clip": 0.01160333, + "auxiliary_loss_mlp": 0.01029919, + "balance_loss_clip": 1.05287528, + "balance_loss_mlp": 1.02209902, + "epoch": 0.4479047676306138, + "flos": 21975462984960.0, + "grad_norm": 1.7792522362558592, + "language_loss": 0.79552895, + "learning_rate": 2.431252438077746e-06, + "loss": 0.81743145, + "num_input_tokens_seen": 80185685, + "step": 3725, + "time_per_iteration": 2.820723295211792 + }, + { + "auxiliary_loss_clip": 0.01181103, + "auxiliary_loss_mlp": 0.00886626, + "balance_loss_clip": 1.05128121, + "balance_loss_mlp": 1.00088668, + "epoch": 0.44802501052125293, + "flos": 21467219495040.0, + "grad_norm": 3.7810895688921744, + "language_loss": 0.76994944, + "learning_rate": 2.4304917582862906e-06, + "loss": 0.7906267, + "num_input_tokens_seen": 80204865, + "step": 3726, + "time_per_iteration": 2.7929611206054688 + }, + { + "auxiliary_loss_clip": 0.01189078, + "auxiliary_loss_mlp": 0.0103349, + "balance_loss_clip": 1.0541085, + "balance_loss_mlp": 1.02489543, + "epoch": 0.44814525341189204, + "flos": 22126970551680.0, + "grad_norm": 2.306018656496641, + "language_loss": 0.87594867, + "learning_rate": 2.4297310131895774e-06, + "loss": 0.89817435, + "num_input_tokens_seen": 80223410, + "step": 3727, + "time_per_iteration": 2.675248146057129 + }, + { + "auxiliary_loss_clip": 0.01175289, + "auxiliary_loss_mlp": 0.01033489, + "balance_loss_clip": 1.04977655, + "balance_loss_mlp": 1.02383888, + "epoch": 0.4482654963025311, + "flos": 16653933204480.0, + "grad_norm": 2.33225911331968, + "language_loss": 0.74543428, + "learning_rate": 2.4289702029030113e-06, + "loss": 0.7675221, + "num_input_tokens_seen": 80240880, + "step": 3728, + "time_per_iteration": 4.611654758453369 + }, + { + "auxiliary_loss_clip": 0.01183093, + "auxiliary_loss_mlp": 0.01033857, + "balance_loss_clip": 1.05776215, + "balance_loss_mlp": 1.02423131, + "epoch": 0.4483857391931702, + "flos": 18841251905280.0, + "grad_norm": 1.783545423181804, + "language_loss": 0.82798851, + "learning_rate": 2.4282093275420057e-06, + "loss": 0.85015804, + "num_input_tokens_seen": 80259910, + "step": 3729, + "time_per_iteration": 2.772049903869629 + }, + { + "auxiliary_loss_clip": 0.01178223, + "auxiliary_loss_mlp": 0.01029021, + "balance_loss_clip": 1.05250561, + "balance_loss_mlp": 1.01973474, + "epoch": 0.4485059820838093, + "flos": 20372590477440.0, + "grad_norm": 2.5203977359580354, + "language_loss": 0.70714742, + "learning_rate": 2.4274483872219863e-06, + "loss": 0.72921985, + "num_input_tokens_seen": 80277270, + "step": 3730, + "time_per_iteration": 2.738006830215454 + }, + { + "auxiliary_loss_clip": 0.01174405, + "auxiliary_loss_mlp": 0.0103636, + "balance_loss_clip": 1.05055976, + "balance_loss_mlp": 1.02752066, + "epoch": 0.4486262249744484, + "flos": 20047742853120.0, + "grad_norm": 1.7075530601635966, + "language_loss": 0.93851829, + "learning_rate": 2.426687382058386e-06, + "loss": 0.96062595, + "num_input_tokens_seen": 80295550, + "step": 3731, + "time_per_iteration": 4.542062759399414 + }, + { + "auxiliary_loss_clip": 0.0106826, + "auxiliary_loss_mlp": 0.01002065, + "balance_loss_clip": 1.01771736, + "balance_loss_mlp": 1.00012231, + "epoch": 0.4487464678650875, + "flos": 64595684776320.0, + "grad_norm": 0.8678680954885414, + "language_loss": 0.59840119, + "learning_rate": 2.425926312166649e-06, + "loss": 0.61910439, + "num_input_tokens_seen": 80348425, + "step": 3732, + "time_per_iteration": 3.2371304035186768 + }, + { + "auxiliary_loss_clip": 0.01173624, + "auxiliary_loss_mlp": 0.01025872, + "balance_loss_clip": 1.05283046, + "balance_loss_mlp": 1.01621008, + "epoch": 0.4488667107557266, + "flos": 20769798049920.0, + "grad_norm": 2.3408637297675825, + "language_loss": 0.73231214, + "learning_rate": 2.42516517766223e-06, + "loss": 0.75430709, + "num_input_tokens_seen": 80366505, + "step": 3733, + "time_per_iteration": 2.898951530456543 + }, + { + "auxiliary_loss_clip": 0.01189549, + "auxiliary_loss_mlp": 0.01031306, + "balance_loss_clip": 1.05672681, + "balance_loss_mlp": 1.02233529, + "epoch": 0.44898695364636565, + "flos": 23951735326080.0, + "grad_norm": 2.158572770316615, + "language_loss": 0.67648268, + "learning_rate": 2.4244039786605907e-06, + "loss": 0.69869125, + "num_input_tokens_seen": 80387510, + "step": 3734, + "time_per_iteration": 2.8516852855682373 + }, + { + "auxiliary_loss_clip": 0.01149944, + "auxiliary_loss_mlp": 0.01034438, + "balance_loss_clip": 1.04729879, + "balance_loss_mlp": 1.02520537, + "epoch": 0.44910719653700476, + "flos": 18624351628800.0, + "grad_norm": 2.3705627331337347, + "language_loss": 0.82037014, + "learning_rate": 2.4236427152772055e-06, + "loss": 0.84221399, + "num_input_tokens_seen": 80405915, + "step": 3735, + "time_per_iteration": 2.8889286518096924 + }, + { + "auxiliary_loss_clip": 0.0104661, + "auxiliary_loss_mlp": 0.01000532, + "balance_loss_clip": 1.01359391, + "balance_loss_mlp": 0.99843359, + "epoch": 0.4492274394276438, + "flos": 57033435749760.0, + "grad_norm": 0.8227514685983041, + "language_loss": 0.57337737, + "learning_rate": 2.422881387627557e-06, + "loss": 0.59384882, + "num_input_tokens_seen": 80458365, + "step": 3736, + "time_per_iteration": 3.1397416591644287 + }, + { + "auxiliary_loss_clip": 0.01165193, + "auxiliary_loss_mlp": 0.01030182, + "balance_loss_clip": 1.05059326, + "balance_loss_mlp": 1.02153313, + "epoch": 0.4493476823182829, + "flos": 23254888498560.0, + "grad_norm": 1.9276038796107413, + "language_loss": 0.77050263, + "learning_rate": 2.422119995827139e-06, + "loss": 0.79245639, + "num_input_tokens_seen": 80478490, + "step": 3737, + "time_per_iteration": 2.7568795680999756 + }, + { + "auxiliary_loss_clip": 0.01175097, + "auxiliary_loss_mlp": 0.01032394, + "balance_loss_clip": 1.05159187, + "balance_loss_mlp": 1.02305365, + "epoch": 0.44946792520892204, + "flos": 15815131827840.0, + "grad_norm": 2.6492954173897845, + "language_loss": 0.73956281, + "learning_rate": 2.4213585399914528e-06, + "loss": 0.76163769, + "num_input_tokens_seen": 80495695, + "step": 3738, + "time_per_iteration": 2.705338716506958 + }, + { + "auxiliary_loss_clip": 0.01176324, + "auxiliary_loss_mlp": 0.01030142, + "balance_loss_clip": 1.05268562, + "balance_loss_mlp": 1.02163661, + "epoch": 0.4495881680995611, + "flos": 19610063631360.0, + "grad_norm": 1.8210782279757003, + "language_loss": 0.85262156, + "learning_rate": 2.4205970202360113e-06, + "loss": 0.87468612, + "num_input_tokens_seen": 80515260, + "step": 3739, + "time_per_iteration": 2.7089476585388184 + }, + { + "auxiliary_loss_clip": 0.01134106, + "auxiliary_loss_mlp": 0.010278, + "balance_loss_clip": 1.04470539, + "balance_loss_mlp": 1.01850724, + "epoch": 0.4497084109902002, + "flos": 26031465815040.0, + "grad_norm": 2.229271326536355, + "language_loss": 0.78063166, + "learning_rate": 2.4198354366763354e-06, + "loss": 0.80225074, + "num_input_tokens_seen": 80533900, + "step": 3740, + "time_per_iteration": 2.8887383937835693 + }, + { + "auxiliary_loss_clip": 0.01171413, + "auxiliary_loss_mlp": 0.01032804, + "balance_loss_clip": 1.05112576, + "balance_loss_mlp": 1.02378583, + "epoch": 0.4498286538808393, + "flos": 14793688771200.0, + "grad_norm": 2.5835151455779246, + "language_loss": 0.78537965, + "learning_rate": 2.4190737894279587e-06, + "loss": 0.8074218, + "num_input_tokens_seen": 80551270, + "step": 3741, + "time_per_iteration": 2.8170390129089355 + }, + { + "auxiliary_loss_clip": 0.01142102, + "auxiliary_loss_mlp": 0.01032597, + "balance_loss_clip": 1.04045606, + "balance_loss_mlp": 1.02400804, + "epoch": 0.44994889677147837, + "flos": 15450171690240.0, + "grad_norm": 2.091170496449841, + "language_loss": 0.80420429, + "learning_rate": 2.4183120786064203e-06, + "loss": 0.82595128, + "num_input_tokens_seen": 80568145, + "step": 3742, + "time_per_iteration": 2.812593460083008 + }, + { + "auxiliary_loss_clip": 0.0117707, + "auxiliary_loss_mlp": 0.00886585, + "balance_loss_clip": 1.05104494, + "balance_loss_mlp": 1.0009892, + "epoch": 0.4500691396621175, + "flos": 21798316085760.0, + "grad_norm": 2.876292875031944, + "language_loss": 0.85431796, + "learning_rate": 2.417550304327273e-06, + "loss": 0.87495458, + "num_input_tokens_seen": 80586185, + "step": 3743, + "time_per_iteration": 2.7713100910186768 + }, + { + "auxiliary_loss_clip": 0.01190498, + "auxiliary_loss_mlp": 0.01034302, + "balance_loss_clip": 1.05501938, + "balance_loss_mlp": 1.02468181, + "epoch": 0.4501893825527566, + "flos": 32382016421760.0, + "grad_norm": 1.660786606842158, + "language_loss": 0.75693429, + "learning_rate": 2.4167884667060763e-06, + "loss": 0.77918226, + "num_input_tokens_seen": 80608895, + "step": 3744, + "time_per_iteration": 2.798916816711426 + }, + { + "auxiliary_loss_clip": 0.01168608, + "auxiliary_loss_mlp": 0.01034121, + "balance_loss_clip": 1.05027103, + "balance_loss_mlp": 1.02380335, + "epoch": 0.45030962544339564, + "flos": 16544944362240.0, + "grad_norm": 5.382420966938116, + "language_loss": 0.87162632, + "learning_rate": 2.4160265658584e-06, + "loss": 0.89365363, + "num_input_tokens_seen": 80623785, + "step": 3745, + "time_per_iteration": 2.818399429321289 + }, + { + "auxiliary_loss_clip": 0.01178929, + "auxiliary_loss_mlp": 0.0103199, + "balance_loss_clip": 1.05148292, + "balance_loss_mlp": 1.02320397, + "epoch": 0.45042986833403476, + "flos": 19573039687680.0, + "grad_norm": 2.625050046085839, + "language_loss": 0.68317252, + "learning_rate": 2.4152646018998253e-06, + "loss": 0.70528173, + "num_input_tokens_seen": 80642735, + "step": 3746, + "time_per_iteration": 2.738619089126587 + }, + { + "auxiliary_loss_clip": 0.01175345, + "auxiliary_loss_mlp": 0.01033121, + "balance_loss_clip": 1.05068362, + "balance_loss_mlp": 1.02378118, + "epoch": 0.45055011122467387, + "flos": 23112467072640.0, + "grad_norm": 3.8782797814492045, + "language_loss": 0.71888697, + "learning_rate": 2.4145025749459403e-06, + "loss": 0.74097162, + "num_input_tokens_seen": 80663760, + "step": 3747, + "time_per_iteration": 2.807790756225586 + }, + { + "auxiliary_loss_clip": 0.01129281, + "auxiliary_loss_mlp": 0.01031185, + "balance_loss_clip": 1.04831791, + "balance_loss_mlp": 1.02199996, + "epoch": 0.4506703541153129, + "flos": 19934623946880.0, + "grad_norm": 2.090336606058608, + "language_loss": 0.69608641, + "learning_rate": 2.413740485112344e-06, + "loss": 0.71769106, + "num_input_tokens_seen": 80682100, + "step": 3748, + "time_per_iteration": 2.904109477996826 + }, + { + "auxiliary_loss_clip": 0.01162964, + "auxiliary_loss_mlp": 0.01028566, + "balance_loss_clip": 1.0529753, + "balance_loss_mlp": 1.02008426, + "epoch": 0.45079059700595203, + "flos": 19499530504320.0, + "grad_norm": 1.7042930793534965, + "language_loss": 0.82246566, + "learning_rate": 2.412978332514646e-06, + "loss": 0.84438097, + "num_input_tokens_seen": 80700880, + "step": 3749, + "time_per_iteration": 2.7230522632598877 + }, + { + "auxiliary_loss_clip": 0.01170799, + "auxiliary_loss_mlp": 0.01029092, + "balance_loss_clip": 1.05076957, + "balance_loss_mlp": 1.01962113, + "epoch": 0.4509108398965911, + "flos": 27636313570560.0, + "grad_norm": 2.0533602813094873, + "language_loss": 0.72040105, + "learning_rate": 2.4122161172684623e-06, + "loss": 0.74239993, + "num_input_tokens_seen": 80721675, + "step": 3750, + "time_per_iteration": 2.775582790374756 + }, + { + "auxiliary_loss_clip": 0.0116374, + "auxiliary_loss_mlp": 0.01031971, + "balance_loss_clip": 1.04880655, + "balance_loss_mlp": 1.02273178, + "epoch": 0.4510310827872302, + "flos": 20995712640000.0, + "grad_norm": 3.0187868252681658, + "language_loss": 0.84080654, + "learning_rate": 2.4114538394894216e-06, + "loss": 0.86276364, + "num_input_tokens_seen": 80739315, + "step": 3751, + "time_per_iteration": 2.743037462234497 + }, + { + "auxiliary_loss_clip": 0.01163198, + "auxiliary_loss_mlp": 0.01029665, + "balance_loss_clip": 1.04583561, + "balance_loss_mlp": 1.02038455, + "epoch": 0.4511513256778693, + "flos": 16216684945920.0, + "grad_norm": 1.9256341491022138, + "language_loss": 0.82677019, + "learning_rate": 2.410691499293161e-06, + "loss": 0.84869885, + "num_input_tokens_seen": 80757470, + "step": 3752, + "time_per_iteration": 2.8282713890075684 + }, + { + "auxiliary_loss_clip": 0.01173154, + "auxiliary_loss_mlp": 0.0102727, + "balance_loss_clip": 1.04787326, + "balance_loss_mlp": 1.01834142, + "epoch": 0.45127156856850836, + "flos": 25186702780800.0, + "grad_norm": 1.700702482978745, + "language_loss": 0.73762041, + "learning_rate": 2.409929096795326e-06, + "loss": 0.75962466, + "num_input_tokens_seen": 80777840, + "step": 3753, + "time_per_iteration": 2.699491262435913 + }, + { + "auxiliary_loss_clip": 0.01175831, + "auxiliary_loss_mlp": 0.01030059, + "balance_loss_clip": 1.04852021, + "balance_loss_mlp": 1.02089763, + "epoch": 0.4513918114591475, + "flos": 20412523422720.0, + "grad_norm": 2.079554802619984, + "language_loss": 0.79409945, + "learning_rate": 2.409166632111573e-06, + "loss": 0.81615829, + "num_input_tokens_seen": 80795975, + "step": 3754, + "time_per_iteration": 3.58056640625 + }, + { + "auxiliary_loss_clip": 0.0118741, + "auxiliary_loss_mlp": 0.01030213, + "balance_loss_clip": 1.05390418, + "balance_loss_mlp": 1.02059317, + "epoch": 0.4515120543497866, + "flos": 26648482665600.0, + "grad_norm": 1.8744777464430302, + "language_loss": 0.80658281, + "learning_rate": 2.4084041053575674e-06, + "loss": 0.82875907, + "num_input_tokens_seen": 80815395, + "step": 3755, + "time_per_iteration": 3.8397860527038574 + }, + { + "auxiliary_loss_clip": 0.01169349, + "auxiliary_loss_mlp": 0.0102645, + "balance_loss_clip": 1.05249977, + "balance_loss_mlp": 1.01716912, + "epoch": 0.45163229724042564, + "flos": 20595093275520.0, + "grad_norm": 2.275255765596135, + "language_loss": 0.725797, + "learning_rate": 2.4076415166489834e-06, + "loss": 0.74775499, + "num_input_tokens_seen": 80834805, + "step": 3756, + "time_per_iteration": 3.7322421073913574 + }, + { + "auxiliary_loss_clip": 0.01149304, + "auxiliary_loss_mlp": 0.01029421, + "balance_loss_clip": 1.04714429, + "balance_loss_mlp": 1.02096343, + "epoch": 0.45175254013106475, + "flos": 21689004021120.0, + "grad_norm": 1.5775810609251981, + "language_loss": 0.79079241, + "learning_rate": 2.406878866101506e-06, + "loss": 0.81257963, + "num_input_tokens_seen": 80853770, + "step": 3757, + "time_per_iteration": 3.8650736808776855 + }, + { + "auxiliary_loss_clip": 0.01187626, + "auxiliary_loss_mlp": 0.010275, + "balance_loss_clip": 1.0547502, + "balance_loss_mlp": 1.01804709, + "epoch": 0.45187278302170386, + "flos": 18878850466560.0, + "grad_norm": 2.3956902946750622, + "language_loss": 0.78041542, + "learning_rate": 2.4061161538308273e-06, + "loss": 0.80256665, + "num_input_tokens_seen": 80870615, + "step": 3758, + "time_per_iteration": 2.6751043796539307 + }, + { + "auxiliary_loss_clip": 0.01176217, + "auxiliary_loss_mlp": 0.01028804, + "balance_loss_clip": 1.05116081, + "balance_loss_mlp": 1.01977432, + "epoch": 0.4519930259123429, + "flos": 18582479349120.0, + "grad_norm": 1.906002078731051, + "language_loss": 0.88816845, + "learning_rate": 2.4053533799526523e-06, + "loss": 0.91021866, + "num_input_tokens_seen": 80886335, + "step": 3759, + "time_per_iteration": 2.623234510421753 + }, + { + "auxiliary_loss_clip": 0.01163833, + "auxiliary_loss_mlp": 0.01026133, + "balance_loss_clip": 1.05151176, + "balance_loss_mlp": 1.01771724, + "epoch": 0.452113268802982, + "flos": 25192377129600.0, + "grad_norm": 1.8616076019147867, + "language_loss": 0.86206698, + "learning_rate": 2.404590544582691e-06, + "loss": 0.88396662, + "num_input_tokens_seen": 80904570, + "step": 3760, + "time_per_iteration": 2.7731521129608154 + }, + { + "auxiliary_loss_clip": 0.01151162, + "auxiliary_loss_mlp": 0.0103336, + "balance_loss_clip": 1.04323769, + "balance_loss_mlp": 1.02400851, + "epoch": 0.45223351169362114, + "flos": 39378922312320.0, + "grad_norm": 1.9511713370885408, + "language_loss": 0.81015193, + "learning_rate": 2.403827647836666e-06, + "loss": 0.83199728, + "num_input_tokens_seen": 80925125, + "step": 3761, + "time_per_iteration": 3.095215082168579 + }, + { + "auxiliary_loss_clip": 0.01186341, + "auxiliary_loss_mlp": 0.01029131, + "balance_loss_clip": 1.05193436, + "balance_loss_mlp": 1.01999903, + "epoch": 0.4523537545842602, + "flos": 21582169994880.0, + "grad_norm": 4.4105290479914245, + "language_loss": 0.69403499, + "learning_rate": 2.4030646898303075e-06, + "loss": 0.71618974, + "num_input_tokens_seen": 80946615, + "step": 3762, + "time_per_iteration": 2.7630255222320557 + }, + { + "auxiliary_loss_clip": 0.01171745, + "auxiliary_loss_mlp": 0.01028244, + "balance_loss_clip": 1.05018127, + "balance_loss_mlp": 1.01930308, + "epoch": 0.4524739974748993, + "flos": 28439527547520.0, + "grad_norm": 2.4036453066605845, + "language_loss": 0.81982005, + "learning_rate": 2.4023016706793566e-06, + "loss": 0.84181988, + "num_input_tokens_seen": 80966410, + "step": 3763, + "time_per_iteration": 2.8309834003448486 + }, + { + "auxiliary_loss_clip": 0.01068832, + "auxiliary_loss_mlp": 0.01012657, + "balance_loss_clip": 1.02016783, + "balance_loss_mlp": 1.01051128, + "epoch": 0.4525942403655384, + "flos": 61556492148480.0, + "grad_norm": 0.7699459290183882, + "language_loss": 0.56849289, + "learning_rate": 2.401538590499561e-06, + "loss": 0.58930779, + "num_input_tokens_seen": 81026865, + "step": 3764, + "time_per_iteration": 3.366948366165161 + }, + { + "auxiliary_loss_clip": 0.01180654, + "auxiliary_loss_mlp": 0.00887295, + "balance_loss_clip": 1.05377805, + "balance_loss_mlp": 1.0009855, + "epoch": 0.45271448325617747, + "flos": 27529838680320.0, + "grad_norm": 1.8045992016294647, + "language_loss": 0.71962559, + "learning_rate": 2.400775449406682e-06, + "loss": 0.74030507, + "num_input_tokens_seen": 81050060, + "step": 3765, + "time_per_iteration": 2.7945642471313477 + }, + { + "auxiliary_loss_clip": 0.01175554, + "auxiliary_loss_mlp": 0.01031589, + "balance_loss_clip": 1.05003452, + "balance_loss_mlp": 1.02279711, + "epoch": 0.4528347261468166, + "flos": 22452608275200.0, + "grad_norm": 1.7893439241179068, + "language_loss": 0.72778261, + "learning_rate": 2.400012247516485e-06, + "loss": 0.74985403, + "num_input_tokens_seen": 81070625, + "step": 3766, + "time_per_iteration": 2.718069076538086 + }, + { + "auxiliary_loss_clip": 0.01165201, + "auxiliary_loss_mlp": 0.01030946, + "balance_loss_clip": 1.05177724, + "balance_loss_mlp": 1.02165389, + "epoch": 0.45295496903745563, + "flos": 21103875469440.0, + "grad_norm": 1.7030650294668381, + "language_loss": 0.9043861, + "learning_rate": 2.3992489849447484e-06, + "loss": 0.92634761, + "num_input_tokens_seen": 81089080, + "step": 3767, + "time_per_iteration": 2.8454740047454834 + }, + { + "auxiliary_loss_clip": 0.01162448, + "auxiliary_loss_mlp": 0.01031212, + "balance_loss_clip": 1.05182767, + "balance_loss_mlp": 1.02203882, + "epoch": 0.45307521192809475, + "flos": 23221168606080.0, + "grad_norm": 1.5786043114841604, + "language_loss": 0.78689599, + "learning_rate": 2.3984856618072584e-06, + "loss": 0.80883265, + "num_input_tokens_seen": 81109115, + "step": 3768, + "time_per_iteration": 2.847014904022217 + }, + { + "auxiliary_loss_clip": 0.01163071, + "auxiliary_loss_mlp": 0.01030524, + "balance_loss_clip": 1.05006719, + "balance_loss_mlp": 1.0224117, + "epoch": 0.45319545481873386, + "flos": 15560094286080.0, + "grad_norm": 2.0046816895132618, + "language_loss": 0.74116099, + "learning_rate": 2.3977222782198098e-06, + "loss": 0.76309693, + "num_input_tokens_seen": 81127750, + "step": 3769, + "time_per_iteration": 2.802067756652832 + }, + { + "auxiliary_loss_clip": 0.01147098, + "auxiliary_loss_mlp": 0.01032062, + "balance_loss_clip": 1.04611897, + "balance_loss_mlp": 1.02234626, + "epoch": 0.4533156977093729, + "flos": 21944759834880.0, + "grad_norm": 1.6141359820792065, + "language_loss": 0.75179946, + "learning_rate": 2.3969588342982077e-06, + "loss": 0.77359104, + "num_input_tokens_seen": 81147125, + "step": 3770, + "time_per_iteration": 2.8065011501312256 + }, + { + "auxiliary_loss_clip": 0.01174915, + "auxiliary_loss_mlp": 0.01025469, + "balance_loss_clip": 1.05310464, + "balance_loss_mlp": 1.01651669, + "epoch": 0.453435940600012, + "flos": 24242180699520.0, + "grad_norm": 1.4969465582520298, + "language_loss": 0.72623217, + "learning_rate": 2.396195330158267e-06, + "loss": 0.748236, + "num_input_tokens_seen": 81167015, + "step": 3771, + "time_per_iteration": 2.6783320903778076 + }, + { + "auxiliary_loss_clip": 0.01188348, + "auxiliary_loss_mlp": 0.01030098, + "balance_loss_clip": 1.0540942, + "balance_loss_mlp": 1.02103257, + "epoch": 0.45355618349065113, + "flos": 23440367352960.0, + "grad_norm": 1.8751118346388866, + "language_loss": 0.79594302, + "learning_rate": 2.3954317659158094e-06, + "loss": 0.81812757, + "num_input_tokens_seen": 81187350, + "step": 3772, + "time_per_iteration": 2.6889266967773438 + }, + { + "auxiliary_loss_clip": 0.01077375, + "auxiliary_loss_mlp": 0.01004129, + "balance_loss_clip": 1.01644611, + "balance_loss_mlp": 1.00238848, + "epoch": 0.4536764263812902, + "flos": 66903161448960.0, + "grad_norm": 0.8918643565661012, + "language_loss": 0.56926221, + "learning_rate": 2.394668141686667e-06, + "loss": 0.59007728, + "num_input_tokens_seen": 81249315, + "step": 3773, + "time_per_iteration": 3.221230983734131 + }, + { + "auxiliary_loss_clip": 0.01169417, + "auxiliary_loss_mlp": 0.0103091, + "balance_loss_clip": 1.04678214, + "balance_loss_mlp": 1.02212405, + "epoch": 0.4537966692719293, + "flos": 42739766254080.0, + "grad_norm": 2.212096989513087, + "language_loss": 0.69561779, + "learning_rate": 2.3939044575866813e-06, + "loss": 0.71762109, + "num_input_tokens_seen": 81272065, + "step": 3774, + "time_per_iteration": 2.8923134803771973 + }, + { + "auxiliary_loss_clip": 0.01159994, + "auxiliary_loss_mlp": 0.00886433, + "balance_loss_clip": 1.04711998, + "balance_loss_mlp": 1.00099027, + "epoch": 0.4539169121625684, + "flos": 35549480517120.0, + "grad_norm": 2.8992285125065766, + "language_loss": 0.75587845, + "learning_rate": 2.3931407137317024e-06, + "loss": 0.77634269, + "num_input_tokens_seen": 81292220, + "step": 3775, + "time_per_iteration": 2.94199800491333 + }, + { + "auxiliary_loss_clip": 0.01149599, + "auxiliary_loss_mlp": 0.01029023, + "balance_loss_clip": 1.04443789, + "balance_loss_mlp": 1.02035677, + "epoch": 0.45403715505320746, + "flos": 18514716341760.0, + "grad_norm": 2.070423571245379, + "language_loss": 0.84758878, + "learning_rate": 2.3923769102375907e-06, + "loss": 0.86937499, + "num_input_tokens_seen": 81311085, + "step": 3776, + "time_per_iteration": 2.8096113204956055 + }, + { + "auxiliary_loss_clip": 0.01153258, + "auxiliary_loss_mlp": 0.01033818, + "balance_loss_clip": 1.04578567, + "balance_loss_mlp": 1.02466297, + "epoch": 0.4541573979438466, + "flos": 25045825639680.0, + "grad_norm": 1.9450853087639028, + "language_loss": 0.78471619, + "learning_rate": 2.391613047220213e-06, + "loss": 0.80658698, + "num_input_tokens_seen": 81330985, + "step": 3777, + "time_per_iteration": 2.9281790256500244 + }, + { + "auxiliary_loss_clip": 0.01150582, + "auxiliary_loss_mlp": 0.01030464, + "balance_loss_clip": 1.04702783, + "balance_loss_mlp": 1.02144635, + "epoch": 0.4542776408344857, + "flos": 18332397884160.0, + "grad_norm": 1.930399277381073, + "language_loss": 0.78755683, + "learning_rate": 2.390849124795447e-06, + "loss": 0.80936724, + "num_input_tokens_seen": 81346985, + "step": 3778, + "time_per_iteration": 2.8887948989868164 + }, + { + "auxiliary_loss_clip": 0.01187712, + "auxiliary_loss_mlp": 0.01028816, + "balance_loss_clip": 1.05353045, + "balance_loss_mlp": 1.02026272, + "epoch": 0.45439788372512474, + "flos": 20701173116160.0, + "grad_norm": 2.0363342582232304, + "language_loss": 0.84011811, + "learning_rate": 2.3900851430791804e-06, + "loss": 0.86228341, + "num_input_tokens_seen": 81365005, + "step": 3779, + "time_per_iteration": 2.789259195327759 + }, + { + "auxiliary_loss_clip": 0.01190533, + "auxiliary_loss_mlp": 0.01038048, + "balance_loss_clip": 1.05502629, + "balance_loss_mlp": 1.02866077, + "epoch": 0.45451812661576385, + "flos": 22309432663680.0, + "grad_norm": 1.9023872971734588, + "language_loss": 0.84396601, + "learning_rate": 2.389321102187307e-06, + "loss": 0.86625183, + "num_input_tokens_seen": 81383785, + "step": 3780, + "time_per_iteration": 3.5996639728546143 + }, + { + "auxiliary_loss_clip": 0.01165421, + "auxiliary_loss_mlp": 0.00887173, + "balance_loss_clip": 1.04985046, + "balance_loss_mlp": 1.00084472, + "epoch": 0.4546383695064029, + "flos": 21763303303680.0, + "grad_norm": 5.019796303953384, + "language_loss": 0.81438887, + "learning_rate": 2.3885570022357326e-06, + "loss": 0.8349148, + "num_input_tokens_seen": 81402915, + "step": 3781, + "time_per_iteration": 3.759517192840576 + }, + { + "auxiliary_loss_clip": 0.01058005, + "auxiliary_loss_mlp": 0.0100251, + "balance_loss_clip": 1.01446366, + "balance_loss_mlp": 1.00044811, + "epoch": 0.454758612397042, + "flos": 64242755694720.0, + "grad_norm": 0.803668797031234, + "language_loss": 0.60884637, + "learning_rate": 2.38779284334037e-06, + "loss": 0.62945151, + "num_input_tokens_seen": 81467890, + "step": 3782, + "time_per_iteration": 3.398904323577881 + }, + { + "auxiliary_loss_clip": 0.01133369, + "auxiliary_loss_mlp": 0.01036436, + "balance_loss_clip": 1.04411197, + "balance_loss_mlp": 1.02727532, + "epoch": 0.4548788552876811, + "flos": 27304175485440.0, + "grad_norm": 2.157965750390785, + "language_loss": 0.78879368, + "learning_rate": 2.387028625617141e-06, + "loss": 0.81049174, + "num_input_tokens_seen": 81487105, + "step": 3783, + "time_per_iteration": 4.701916694641113 + }, + { + "auxiliary_loss_clip": 0.01152173, + "auxiliary_loss_mlp": 0.01028781, + "balance_loss_clip": 1.04508579, + "balance_loss_mlp": 1.02022767, + "epoch": 0.4549990981783202, + "flos": 22857142222080.0, + "grad_norm": 2.383947200243635, + "language_loss": 0.8453849, + "learning_rate": 2.3862643491819766e-06, + "loss": 0.86719441, + "num_input_tokens_seen": 81505670, + "step": 3784, + "time_per_iteration": 2.7351555824279785 + }, + { + "auxiliary_loss_clip": 0.01172417, + "auxiliary_loss_mlp": 0.01035534, + "balance_loss_clip": 1.04823208, + "balance_loss_mlp": 1.02677226, + "epoch": 0.4551193410689593, + "flos": 23258587599360.0, + "grad_norm": 1.7630168520596456, + "language_loss": 0.83982539, + "learning_rate": 2.3855000141508186e-06, + "loss": 0.86190486, + "num_input_tokens_seen": 81525825, + "step": 3785, + "time_per_iteration": 2.7688047885894775 + }, + { + "auxiliary_loss_clip": 0.01169943, + "auxiliary_loss_mlp": 0.01037117, + "balance_loss_clip": 1.05310702, + "balance_loss_mlp": 1.0280869, + "epoch": 0.4552395839595984, + "flos": 20777519473920.0, + "grad_norm": 2.6789640476978076, + "language_loss": 0.83766174, + "learning_rate": 2.3847356206396143e-06, + "loss": 0.85973233, + "num_input_tokens_seen": 81543135, + "step": 3786, + "time_per_iteration": 2.7495691776275635 + }, + { + "auxiliary_loss_clip": 0.01186133, + "auxiliary_loss_mlp": 0.01031937, + "balance_loss_clip": 1.05383563, + "balance_loss_mlp": 1.02271605, + "epoch": 0.45535982685023746, + "flos": 23257510191360.0, + "grad_norm": 1.5749255054770577, + "language_loss": 0.78368717, + "learning_rate": 2.3839711687643227e-06, + "loss": 0.80586791, + "num_input_tokens_seen": 81564360, + "step": 3787, + "time_per_iteration": 2.740816831588745 + }, + { + "auxiliary_loss_clip": 0.01175776, + "auxiliary_loss_mlp": 0.01033061, + "balance_loss_clip": 1.05199671, + "balance_loss_mlp": 1.02287507, + "epoch": 0.45548006974087657, + "flos": 19646117907840.0, + "grad_norm": 1.9576294457659467, + "language_loss": 0.74043214, + "learning_rate": 2.38320665864091e-06, + "loss": 0.76252055, + "num_input_tokens_seen": 81583710, + "step": 3788, + "time_per_iteration": 2.7949934005737305 + }, + { + "auxiliary_loss_clip": 0.01137218, + "auxiliary_loss_mlp": 0.01031758, + "balance_loss_clip": 1.04239869, + "balance_loss_mlp": 1.02225149, + "epoch": 0.4556003126315157, + "flos": 20047778766720.0, + "grad_norm": 1.872316467380244, + "language_loss": 0.81676316, + "learning_rate": 2.3824420903853516e-06, + "loss": 0.83845294, + "num_input_tokens_seen": 81602175, + "step": 3789, + "time_per_iteration": 2.920114040374756 + }, + { + "auxiliary_loss_clip": 0.01176298, + "auxiliary_loss_mlp": 0.01026308, + "balance_loss_clip": 1.05235779, + "balance_loss_mlp": 1.01777887, + "epoch": 0.45572055552215474, + "flos": 22959738443520.0, + "grad_norm": 2.198821849764585, + "language_loss": 0.82007247, + "learning_rate": 2.3816774641136324e-06, + "loss": 0.84209847, + "num_input_tokens_seen": 81619430, + "step": 3790, + "time_per_iteration": 2.7593135833740234 + }, + { + "auxiliary_loss_clip": 0.01175127, + "auxiliary_loss_mlp": 0.00886672, + "balance_loss_clip": 1.05154347, + "balance_loss_mlp": 1.00085521, + "epoch": 0.45584079841279385, + "flos": 33109925535360.0, + "grad_norm": 1.7172750356969115, + "language_loss": 0.71196473, + "learning_rate": 2.380912779941745e-06, + "loss": 0.73258269, + "num_input_tokens_seen": 81642550, + "step": 3791, + "time_per_iteration": 2.784818172454834 + }, + { + "auxiliary_loss_clip": 0.01176688, + "auxiliary_loss_mlp": 0.01039218, + "balance_loss_clip": 1.04872465, + "balance_loss_mlp": 1.02980042, + "epoch": 0.45596104130343296, + "flos": 27272179445760.0, + "grad_norm": 2.0274142223453406, + "language_loss": 0.82960725, + "learning_rate": 2.3801480379856918e-06, + "loss": 0.85176635, + "num_input_tokens_seen": 81664260, + "step": 3792, + "time_per_iteration": 2.795057535171509 + }, + { + "auxiliary_loss_clip": 0.01168378, + "auxiliary_loss_mlp": 0.01034708, + "balance_loss_clip": 1.05152762, + "balance_loss_mlp": 1.02527833, + "epoch": 0.456081284194072, + "flos": 21579799697280.0, + "grad_norm": 1.675126124959895, + "language_loss": 0.836254, + "learning_rate": 2.379383238361484e-06, + "loss": 0.85828483, + "num_input_tokens_seen": 81683620, + "step": 3793, + "time_per_iteration": 2.8275508880615234 + }, + { + "auxiliary_loss_clip": 0.01176694, + "auxiliary_loss_mlp": 0.01031993, + "balance_loss_clip": 1.05192113, + "balance_loss_mlp": 1.02273655, + "epoch": 0.4562015270847111, + "flos": 35918822113920.0, + "grad_norm": 2.323958254308023, + "language_loss": 0.79610974, + "learning_rate": 2.3786183811851407e-06, + "loss": 0.81819665, + "num_input_tokens_seen": 81704325, + "step": 3794, + "time_per_iteration": 2.8427698612213135 + }, + { + "auxiliary_loss_clip": 0.01188885, + "auxiliary_loss_mlp": 0.0103402, + "balance_loss_clip": 1.05490732, + "balance_loss_mlp": 1.02521062, + "epoch": 0.45632176997535023, + "flos": 13589783602560.0, + "grad_norm": 1.793879838815351, + "language_loss": 0.80033076, + "learning_rate": 2.3778534665726892e-06, + "loss": 0.82255983, + "num_input_tokens_seen": 81721155, + "step": 3795, + "time_per_iteration": 2.693119764328003 + }, + { + "auxiliary_loss_clip": 0.01167014, + "auxiliary_loss_mlp": 0.01029278, + "balance_loss_clip": 1.05049813, + "balance_loss_mlp": 1.02047479, + "epoch": 0.4564420128659893, + "flos": 32635401937920.0, + "grad_norm": 1.9151694910799237, + "language_loss": 0.72833455, + "learning_rate": 2.377088494640168e-06, + "loss": 0.75029743, + "num_input_tokens_seen": 81742905, + "step": 3796, + "time_per_iteration": 2.7753453254699707 + }, + { + "auxiliary_loss_clip": 0.01175516, + "auxiliary_loss_mlp": 0.01033592, + "balance_loss_clip": 1.05251682, + "balance_loss_mlp": 1.02458537, + "epoch": 0.4565622557566284, + "flos": 20377690208640.0, + "grad_norm": 2.2297742052730496, + "language_loss": 0.78365982, + "learning_rate": 2.3763234655036216e-06, + "loss": 0.80575091, + "num_input_tokens_seen": 81762105, + "step": 3797, + "time_per_iteration": 2.811321973800659 + }, + { + "auxiliary_loss_clip": 0.01156487, + "auxiliary_loss_mlp": 0.01030257, + "balance_loss_clip": 1.04534864, + "balance_loss_mlp": 1.02133369, + "epoch": 0.45668249864726745, + "flos": 25374372364800.0, + "grad_norm": 1.936586936125705, + "language_loss": 0.86757827, + "learning_rate": 2.3755583792791046e-06, + "loss": 0.88944566, + "num_input_tokens_seen": 81781975, + "step": 3798, + "time_per_iteration": 2.987597942352295 + }, + { + "auxiliary_loss_clip": 0.01178631, + "auxiliary_loss_mlp": 0.01027898, + "balance_loss_clip": 1.05245018, + "balance_loss_mlp": 1.0189991, + "epoch": 0.45680274153790656, + "flos": 15559806977280.0, + "grad_norm": 1.989041004203725, + "language_loss": 0.74057364, + "learning_rate": 2.3747932360826803e-06, + "loss": 0.76263893, + "num_input_tokens_seen": 81798905, + "step": 3799, + "time_per_iteration": 2.773224353790283 + }, + { + "auxiliary_loss_clip": 0.01171673, + "auxiliary_loss_mlp": 0.01033572, + "balance_loss_clip": 1.0501982, + "balance_loss_mlp": 1.02411246, + "epoch": 0.4569229844285457, + "flos": 19792884879360.0, + "grad_norm": 1.9884837342314823, + "language_loss": 0.82343292, + "learning_rate": 2.3740280360304205e-06, + "loss": 0.84548545, + "num_input_tokens_seen": 81816630, + "step": 3800, + "time_per_iteration": 2.674813985824585 + }, + { + "auxiliary_loss_clip": 0.01159415, + "auxiliary_loss_mlp": 0.01029529, + "balance_loss_clip": 1.05248737, + "balance_loss_mlp": 1.01980114, + "epoch": 0.45704322731918473, + "flos": 24093941270400.0, + "grad_norm": 1.7383193754838457, + "language_loss": 0.67951059, + "learning_rate": 2.3732627792384038e-06, + "loss": 0.70140004, + "num_input_tokens_seen": 81837700, + "step": 3801, + "time_per_iteration": 2.955920934677124 + }, + { + "auxiliary_loss_clip": 0.0118695, + "auxiliary_loss_mlp": 0.01026209, + "balance_loss_clip": 1.05307651, + "balance_loss_mlp": 1.0167613, + "epoch": 0.45716347020982384, + "flos": 31317803245440.0, + "grad_norm": 1.8990636319602154, + "language_loss": 0.74970239, + "learning_rate": 2.3724974658227207e-06, + "loss": 0.77183402, + "num_input_tokens_seen": 81858490, + "step": 3802, + "time_per_iteration": 2.783156633377075 + }, + { + "auxiliary_loss_clip": 0.01165513, + "auxiliary_loss_mlp": 0.00887238, + "balance_loss_clip": 1.05023837, + "balance_loss_mlp": 1.00086534, + "epoch": 0.45728371310046295, + "flos": 26501392471680.0, + "grad_norm": 2.095824038280992, + "language_loss": 0.71164143, + "learning_rate": 2.3717320958994687e-06, + "loss": 0.73216903, + "num_input_tokens_seen": 81876050, + "step": 3803, + "time_per_iteration": 2.7153964042663574 + }, + { + "auxiliary_loss_clip": 0.0115536, + "auxiliary_loss_mlp": 0.0102854, + "balance_loss_clip": 1.04290962, + "balance_loss_mlp": 1.01955175, + "epoch": 0.457403955991102, + "flos": 17929408222080.0, + "grad_norm": 2.0251492437848437, + "language_loss": 0.70448506, + "learning_rate": 2.3709666695847534e-06, + "loss": 0.72632408, + "num_input_tokens_seen": 81894230, + "step": 3804, + "time_per_iteration": 2.7638299465179443 + }, + { + "auxiliary_loss_clip": 0.01140433, + "auxiliary_loss_mlp": 0.01029452, + "balance_loss_clip": 1.04518008, + "balance_loss_mlp": 1.02051711, + "epoch": 0.4575241988817411, + "flos": 42230660837760.0, + "grad_norm": 1.7947639902630619, + "language_loss": 0.7026577, + "learning_rate": 2.370201186994689e-06, + "loss": 0.72435653, + "num_input_tokens_seen": 81917915, + "step": 3805, + "time_per_iteration": 2.982043981552124 + }, + { + "auxiliary_loss_clip": 0.01159545, + "auxiliary_loss_mlp": 0.01028054, + "balance_loss_clip": 1.04997575, + "balance_loss_mlp": 1.01950049, + "epoch": 0.45764444177238023, + "flos": 30117309868800.0, + "grad_norm": 1.9150257785387135, + "language_loss": 0.6986928, + "learning_rate": 2.369435648245399e-06, + "loss": 0.72056878, + "num_input_tokens_seen": 81938130, + "step": 3806, + "time_per_iteration": 3.7540714740753174 + }, + { + "auxiliary_loss_clip": 0.0116655, + "auxiliary_loss_mlp": 0.01037107, + "balance_loss_clip": 1.04965961, + "balance_loss_mlp": 1.02783275, + "epoch": 0.4577646846630193, + "flos": 24060293205120.0, + "grad_norm": 1.7436054528547, + "language_loss": 0.85327941, + "learning_rate": 2.368670053453015e-06, + "loss": 0.87531602, + "num_input_tokens_seen": 81959820, + "step": 3807, + "time_per_iteration": 3.7954649925231934 + }, + { + "auxiliary_loss_clip": 0.01185922, + "auxiliary_loss_mlp": 0.01033135, + "balance_loss_clip": 1.05596471, + "balance_loss_mlp": 1.02412868, + "epoch": 0.4578849275536584, + "flos": 17418578952960.0, + "grad_norm": 2.316353767188003, + "language_loss": 0.73777825, + "learning_rate": 2.3679044027336757e-06, + "loss": 0.75996876, + "num_input_tokens_seen": 81975710, + "step": 3808, + "time_per_iteration": 2.7042222023010254 + }, + { + "auxiliary_loss_clip": 0.01188829, + "auxiliary_loss_mlp": 0.01032882, + "balance_loss_clip": 1.0538938, + "balance_loss_mlp": 1.02352428, + "epoch": 0.4580051704442975, + "flos": 13510169107200.0, + "grad_norm": 3.038146693155694, + "language_loss": 0.68936408, + "learning_rate": 2.3671386962035326e-06, + "loss": 0.71158123, + "num_input_tokens_seen": 81993180, + "step": 3809, + "time_per_iteration": 3.6217331886291504 + }, + { + "auxiliary_loss_clip": 0.01179761, + "auxiliary_loss_mlp": 0.01034392, + "balance_loss_clip": 1.05230153, + "balance_loss_mlp": 1.02452791, + "epoch": 0.45812541333493656, + "flos": 18037606965120.0, + "grad_norm": 2.3264832761027465, + "language_loss": 0.68636, + "learning_rate": 2.3663729339787405e-06, + "loss": 0.70850158, + "num_input_tokens_seen": 82010115, + "step": 3810, + "time_per_iteration": 2.756568431854248 + }, + { + "auxiliary_loss_clip": 0.01188976, + "auxiliary_loss_mlp": 0.01033416, + "balance_loss_clip": 1.05495572, + "balance_loss_mlp": 1.0240283, + "epoch": 0.45824565622557567, + "flos": 20222196232320.0, + "grad_norm": 2.5756242015258906, + "language_loss": 0.73324734, + "learning_rate": 2.365607116175466e-06, + "loss": 0.75547123, + "num_input_tokens_seen": 82025540, + "step": 3811, + "time_per_iteration": 2.6350011825561523 + }, + { + "auxiliary_loss_clip": 0.01188176, + "auxiliary_loss_mlp": 0.01040466, + "balance_loss_clip": 1.05436254, + "balance_loss_mlp": 1.03185868, + "epoch": 0.4583658991162148, + "flos": 19864885691520.0, + "grad_norm": 2.862127751969107, + "language_loss": 0.66644394, + "learning_rate": 2.3648412429098825e-06, + "loss": 0.68873036, + "num_input_tokens_seen": 82043890, + "step": 3812, + "time_per_iteration": 2.629045248031616 + }, + { + "auxiliary_loss_clip": 0.01153941, + "auxiliary_loss_mlp": 0.01048001, + "balance_loss_clip": 1.04967916, + "balance_loss_mlp": 1.03700423, + "epoch": 0.45848614200685384, + "flos": 21029935322880.0, + "grad_norm": 1.8585311413574066, + "language_loss": 0.81984687, + "learning_rate": 2.364075314298172e-06, + "loss": 0.84186625, + "num_input_tokens_seen": 82061345, + "step": 3813, + "time_per_iteration": 2.7618391513824463 + }, + { + "auxiliary_loss_clip": 0.01184058, + "auxiliary_loss_mlp": 0.00887448, + "balance_loss_clip": 1.05394864, + "balance_loss_mlp": 1.00091624, + "epoch": 0.45860638489749295, + "flos": 21069293650560.0, + "grad_norm": 1.8578870290933678, + "language_loss": 0.70139444, + "learning_rate": 2.3633093304565267e-06, + "loss": 0.72210956, + "num_input_tokens_seen": 82080400, + "step": 3814, + "time_per_iteration": 2.6980783939361572 + }, + { + "auxiliary_loss_clip": 0.01194405, + "auxiliary_loss_mlp": 0.0104026, + "balance_loss_clip": 1.05721211, + "balance_loss_mlp": 1.03011584, + "epoch": 0.458726627788132, + "flos": 26833889692800.0, + "grad_norm": 1.939471078677411, + "language_loss": 0.62979722, + "learning_rate": 2.3625432915011443e-06, + "loss": 0.65214396, + "num_input_tokens_seen": 82102310, + "step": 3815, + "time_per_iteration": 2.731071710586548 + }, + { + "auxiliary_loss_clip": 0.01162415, + "auxiliary_loss_mlp": 0.01035114, + "balance_loss_clip": 1.05005002, + "balance_loss_mlp": 1.02543426, + "epoch": 0.4588468706787711, + "flos": 24097927680000.0, + "grad_norm": 1.672798923653623, + "language_loss": 0.65562868, + "learning_rate": 2.3617771975482334e-06, + "loss": 0.6776039, + "num_input_tokens_seen": 82121140, + "step": 3816, + "time_per_iteration": 2.7591779232025146 + }, + { + "auxiliary_loss_clip": 0.01145663, + "auxiliary_loss_mlp": 0.01037681, + "balance_loss_clip": 1.04881418, + "balance_loss_mlp": 1.02881229, + "epoch": 0.4589671135694102, + "flos": 17889331622400.0, + "grad_norm": 1.7821773712759335, + "language_loss": 0.74710226, + "learning_rate": 2.3610110487140083e-06, + "loss": 0.76893574, + "num_input_tokens_seen": 82139575, + "step": 3817, + "time_per_iteration": 2.9011330604553223 + }, + { + "auxiliary_loss_clip": 0.01168606, + "auxiliary_loss_mlp": 0.01031495, + "balance_loss_clip": 1.05235684, + "balance_loss_mlp": 1.02291822, + "epoch": 0.4590873564600493, + "flos": 25626967781760.0, + "grad_norm": 1.8600604518299397, + "language_loss": 0.80648303, + "learning_rate": 2.360244845114695e-06, + "loss": 0.82848406, + "num_input_tokens_seen": 82159195, + "step": 3818, + "time_per_iteration": 2.8443992137908936 + }, + { + "auxiliary_loss_clip": 0.01163433, + "auxiliary_loss_mlp": 0.01033128, + "balance_loss_clip": 1.0528264, + "balance_loss_mlp": 1.02379405, + "epoch": 0.4592075993506884, + "flos": 18514788168960.0, + "grad_norm": 2.0779764091042185, + "language_loss": 0.68640888, + "learning_rate": 2.3594785868665245e-06, + "loss": 0.7083745, + "num_input_tokens_seen": 82175500, + "step": 3819, + "time_per_iteration": 2.7890734672546387 + }, + { + "auxiliary_loss_clip": 0.01162207, + "auxiliary_loss_mlp": 0.00887193, + "balance_loss_clip": 1.05064678, + "balance_loss_mlp": 1.00094247, + "epoch": 0.4593278422413275, + "flos": 20631111638400.0, + "grad_norm": 1.9901376565570723, + "language_loss": 0.80644733, + "learning_rate": 2.3587122740857386e-06, + "loss": 0.82694125, + "num_input_tokens_seen": 82192600, + "step": 3820, + "time_per_iteration": 2.821000337600708 + }, + { + "auxiliary_loss_clip": 0.0117628, + "auxiliary_loss_mlp": 0.0103182, + "balance_loss_clip": 1.05143452, + "balance_loss_mlp": 1.02284908, + "epoch": 0.45944808513196655, + "flos": 21358517961600.0, + "grad_norm": 1.8355834243232225, + "language_loss": 0.78339428, + "learning_rate": 2.357945906888586e-06, + "loss": 0.80547523, + "num_input_tokens_seen": 82212040, + "step": 3821, + "time_per_iteration": 2.793837785720825 + }, + { + "auxiliary_loss_clip": 0.01181661, + "auxiliary_loss_mlp": 0.01036386, + "balance_loss_clip": 1.05374277, + "balance_loss_mlp": 1.02594328, + "epoch": 0.45956832802260567, + "flos": 21427789340160.0, + "grad_norm": 3.396025558020037, + "language_loss": 0.80355859, + "learning_rate": 2.357179485391324e-06, + "loss": 0.82573903, + "num_input_tokens_seen": 82229895, + "step": 3822, + "time_per_iteration": 2.7168850898742676 + }, + { + "auxiliary_loss_clip": 0.01184291, + "auxiliary_loss_mlp": 0.01033268, + "balance_loss_clip": 1.05333853, + "balance_loss_mlp": 1.02402973, + "epoch": 0.4596885709132448, + "flos": 22382654538240.0, + "grad_norm": 1.8147582230015096, + "language_loss": 0.86226857, + "learning_rate": 2.3564130097102173e-06, + "loss": 0.88444424, + "num_input_tokens_seen": 82249550, + "step": 3823, + "time_per_iteration": 2.667584180831909 + }, + { + "auxiliary_loss_clip": 0.01157858, + "auxiliary_loss_mlp": 0.01037193, + "balance_loss_clip": 1.05102682, + "balance_loss_mlp": 1.02716756, + "epoch": 0.45980881380388383, + "flos": 28981957806720.0, + "grad_norm": 1.9881380273534206, + "language_loss": 0.75121909, + "learning_rate": 2.355646479961541e-06, + "loss": 0.77316964, + "num_input_tokens_seen": 82268860, + "step": 3824, + "time_per_iteration": 2.753678560256958 + }, + { + "auxiliary_loss_clip": 0.01188058, + "auxiliary_loss_mlp": 0.01036217, + "balance_loss_clip": 1.0534842, + "balance_loss_mlp": 1.02617979, + "epoch": 0.45992905669452294, + "flos": 33396599980800.0, + "grad_norm": 1.9255714092595146, + "language_loss": 0.71220767, + "learning_rate": 2.354879896261576e-06, + "loss": 0.73445046, + "num_input_tokens_seen": 82289070, + "step": 3825, + "time_per_iteration": 2.713684320449829 + }, + { + "auxiliary_loss_clip": 0.01156578, + "auxiliary_loss_mlp": 0.01033727, + "balance_loss_clip": 1.05272412, + "balance_loss_mlp": 1.02419055, + "epoch": 0.46004929958516205, + "flos": 36318184502400.0, + "grad_norm": 1.7040193771803775, + "language_loss": 0.57339394, + "learning_rate": 2.3541132587266133e-06, + "loss": 0.59529698, + "num_input_tokens_seen": 82311790, + "step": 3826, + "time_per_iteration": 2.8578341007232666 + }, + { + "auxiliary_loss_clip": 0.01165001, + "auxiliary_loss_mlp": 0.01035355, + "balance_loss_clip": 1.05138326, + "balance_loss_mlp": 1.02541864, + "epoch": 0.4601695424758011, + "flos": 17238451224960.0, + "grad_norm": 2.1128829299430585, + "language_loss": 0.69069839, + "learning_rate": 2.3533465674729515e-06, + "loss": 0.71270192, + "num_input_tokens_seen": 82329020, + "step": 3827, + "time_per_iteration": 2.7874467372894287 + }, + { + "auxiliary_loss_clip": 0.01191966, + "auxiliary_loss_mlp": 0.01031529, + "balance_loss_clip": 1.05666709, + "balance_loss_mlp": 1.02149796, + "epoch": 0.4602897853664402, + "flos": 15888425529600.0, + "grad_norm": 4.782320377563321, + "language_loss": 0.73168504, + "learning_rate": 2.352579822616895e-06, + "loss": 0.75391996, + "num_input_tokens_seen": 82346455, + "step": 3828, + "time_per_iteration": 2.583173990249634 + }, + { + "auxiliary_loss_clip": 0.01167191, + "auxiliary_loss_mlp": 0.01031591, + "balance_loss_clip": 1.05181801, + "balance_loss_mlp": 1.0223465, + "epoch": 0.4604100282570793, + "flos": 25412617370880.0, + "grad_norm": 1.6694614895669755, + "language_loss": 0.77733576, + "learning_rate": 2.351813024274761e-06, + "loss": 0.79932368, + "num_input_tokens_seen": 82367810, + "step": 3829, + "time_per_iteration": 2.7607262134552 + }, + { + "auxiliary_loss_clip": 0.01164189, + "auxiliary_loss_mlp": 0.01037717, + "balance_loss_clip": 1.05243695, + "balance_loss_mlp": 1.02787042, + "epoch": 0.4605302711477184, + "flos": 27630711048960.0, + "grad_norm": 2.4954894611269163, + "language_loss": 0.73531502, + "learning_rate": 2.3510461725628693e-06, + "loss": 0.75733411, + "num_input_tokens_seen": 82388275, + "step": 3830, + "time_per_iteration": 2.83840274810791 + }, + { + "auxiliary_loss_clip": 0.01162566, + "auxiliary_loss_mlp": 0.01032754, + "balance_loss_clip": 1.04994106, + "balance_loss_mlp": 1.02383757, + "epoch": 0.4606505140383575, + "flos": 23839657914240.0, + "grad_norm": 2.2298735060077584, + "language_loss": 0.70548922, + "learning_rate": 2.350279267597554e-06, + "loss": 0.72744238, + "num_input_tokens_seen": 82408915, + "step": 3831, + "time_per_iteration": 2.801860809326172 + }, + { + "auxiliary_loss_clip": 0.01179665, + "auxiliary_loss_mlp": 0.01033514, + "balance_loss_clip": 1.05472362, + "balance_loss_mlp": 1.02286875, + "epoch": 0.46077075692899655, + "flos": 16107013745280.0, + "grad_norm": 2.2140803669981977, + "language_loss": 0.83060801, + "learning_rate": 2.3495123094951515e-06, + "loss": 0.85273975, + "num_input_tokens_seen": 82427260, + "step": 3832, + "time_per_iteration": 3.5638227462768555 + }, + { + "auxiliary_loss_clip": 0.01164983, + "auxiliary_loss_mlp": 0.01027249, + "balance_loss_clip": 1.0526911, + "balance_loss_mlp": 1.01777148, + "epoch": 0.46089099981963566, + "flos": 48798147634560.0, + "grad_norm": 3.2402936526405566, + "language_loss": 0.75997126, + "learning_rate": 2.34874529837201e-06, + "loss": 0.78189361, + "num_input_tokens_seen": 82450805, + "step": 3833, + "time_per_iteration": 3.9846086502075195 + }, + { + "auxiliary_loss_clip": 0.01134055, + "auxiliary_loss_mlp": 0.01039021, + "balance_loss_clip": 1.04407191, + "balance_loss_mlp": 1.0297823, + "epoch": 0.46101124271027477, + "flos": 19099234362240.0, + "grad_norm": 1.8574414775983659, + "language_loss": 0.78901041, + "learning_rate": 2.347978234344483e-06, + "loss": 0.81074119, + "num_input_tokens_seen": 82467010, + "step": 3834, + "time_per_iteration": 2.918489456176758 + }, + { + "auxiliary_loss_clip": 0.01185647, + "auxiliary_loss_mlp": 0.01036076, + "balance_loss_clip": 1.05587244, + "balance_loss_mlp": 1.02613401, + "epoch": 0.4611314856009138, + "flos": 39347931853440.0, + "grad_norm": 1.7188603744032396, + "language_loss": 0.68809384, + "learning_rate": 2.347211117528935e-06, + "loss": 0.71031106, + "num_input_tokens_seen": 82489310, + "step": 3835, + "time_per_iteration": 4.756485939025879 + }, + { + "auxiliary_loss_clip": 0.01170525, + "auxiliary_loss_mlp": 0.01033487, + "balance_loss_clip": 1.05517817, + "balance_loss_mlp": 1.02380669, + "epoch": 0.46125172849155294, + "flos": 20810772489600.0, + "grad_norm": 1.7708678799891675, + "language_loss": 0.71662688, + "learning_rate": 2.3464439480417374e-06, + "loss": 0.73866701, + "num_input_tokens_seen": 82508830, + "step": 3836, + "time_per_iteration": 2.838866949081421 + }, + { + "auxiliary_loss_clip": 0.01184298, + "auxiliary_loss_mlp": 0.01027154, + "balance_loss_clip": 1.05387735, + "balance_loss_mlp": 1.01722956, + "epoch": 0.46137197138219205, + "flos": 17930808852480.0, + "grad_norm": 5.996796893296448, + "language_loss": 0.76690775, + "learning_rate": 2.3456767259992676e-06, + "loss": 0.78902233, + "num_input_tokens_seen": 82526475, + "step": 3837, + "time_per_iteration": 2.6842050552368164 + }, + { + "auxiliary_loss_clip": 0.01190287, + "auxiliary_loss_mlp": 0.00888274, + "balance_loss_clip": 1.05331087, + "balance_loss_mlp": 1.00081921, + "epoch": 0.4614922142728311, + "flos": 16836610798080.0, + "grad_norm": 2.052330635318865, + "language_loss": 0.88896513, + "learning_rate": 2.3449094515179135e-06, + "loss": 0.90975076, + "num_input_tokens_seen": 82543935, + "step": 3838, + "time_per_iteration": 2.7433583736419678 + }, + { + "auxiliary_loss_clip": 0.01174557, + "auxiliary_loss_mlp": 0.01028591, + "balance_loss_clip": 1.05165088, + "balance_loss_mlp": 1.01962638, + "epoch": 0.4616124571634702, + "flos": 26614906427520.0, + "grad_norm": 1.9115114099027908, + "language_loss": 0.81737578, + "learning_rate": 2.34414212471407e-06, + "loss": 0.83940727, + "num_input_tokens_seen": 82563730, + "step": 3839, + "time_per_iteration": 2.780806303024292 + }, + { + "auxiliary_loss_clip": 0.01191233, + "auxiliary_loss_mlp": 0.01038056, + "balance_loss_clip": 1.05815625, + "balance_loss_mlp": 1.02795935, + "epoch": 0.4617327000541093, + "flos": 20340127560960.0, + "grad_norm": 1.9034398552020506, + "language_loss": 0.72864413, + "learning_rate": 2.3433747457041394e-06, + "loss": 0.7509371, + "num_input_tokens_seen": 82582435, + "step": 3840, + "time_per_iteration": 2.7219033241271973 + }, + { + "auxiliary_loss_clip": 0.01157893, + "auxiliary_loss_mlp": 0.01029461, + "balance_loss_clip": 1.05047464, + "balance_loss_mlp": 1.0194298, + "epoch": 0.4618529429447484, + "flos": 29570749545600.0, + "grad_norm": 1.8339441315791187, + "language_loss": 0.85137403, + "learning_rate": 2.342607314604533e-06, + "loss": 0.87324762, + "num_input_tokens_seen": 82602185, + "step": 3841, + "time_per_iteration": 2.8350613117218018 + }, + { + "auxiliary_loss_clip": 0.0118033, + "auxiliary_loss_mlp": 0.01038895, + "balance_loss_clip": 1.05633926, + "balance_loss_mlp": 1.02869642, + "epoch": 0.4619731858353875, + "flos": 19787030962560.0, + "grad_norm": 1.8092538658828525, + "language_loss": 0.842507, + "learning_rate": 2.3418398315316694e-06, + "loss": 0.86469924, + "num_input_tokens_seen": 82620005, + "step": 3842, + "time_per_iteration": 2.7198610305786133 + }, + { + "auxiliary_loss_clip": 0.01192742, + "auxiliary_loss_mlp": 0.01034459, + "balance_loss_clip": 1.05785143, + "balance_loss_mlp": 1.02461767, + "epoch": 0.4620934287260266, + "flos": 18951138587520.0, + "grad_norm": 3.32907199850081, + "language_loss": 0.78533322, + "learning_rate": 2.3410722966019755e-06, + "loss": 0.80760527, + "num_input_tokens_seen": 82635120, + "step": 3843, + "time_per_iteration": 2.6766700744628906 + }, + { + "auxiliary_loss_clip": 0.01180613, + "auxiliary_loss_mlp": 0.01032121, + "balance_loss_clip": 1.05387902, + "balance_loss_mlp": 1.0220722, + "epoch": 0.46221367161666566, + "flos": 37341674634240.0, + "grad_norm": 1.8251173654891208, + "language_loss": 0.65489399, + "learning_rate": 2.3403047099318848e-06, + "loss": 0.67702127, + "num_input_tokens_seen": 82659190, + "step": 3844, + "time_per_iteration": 2.7694172859191895 + }, + { + "auxiliary_loss_clip": 0.0114407, + "auxiliary_loss_mlp": 0.01031502, + "balance_loss_clip": 1.04558194, + "balance_loss_mlp": 1.02143526, + "epoch": 0.46233391450730477, + "flos": 14428549065600.0, + "grad_norm": 2.4466384272783155, + "language_loss": 0.75341463, + "learning_rate": 2.3395370716378405e-06, + "loss": 0.77517039, + "num_input_tokens_seen": 82676635, + "step": 3845, + "time_per_iteration": 2.755678176879883 + }, + { + "auxiliary_loss_clip": 0.0118315, + "auxiliary_loss_mlp": 0.01031762, + "balance_loss_clip": 1.05296111, + "balance_loss_mlp": 1.02217746, + "epoch": 0.4624541573979438, + "flos": 22493044010880.0, + "grad_norm": 2.195544697154001, + "language_loss": 0.72263384, + "learning_rate": 2.338769381836292e-06, + "loss": 0.74478292, + "num_input_tokens_seen": 82696245, + "step": 3846, + "time_per_iteration": 2.656709909439087 + }, + { + "auxiliary_loss_clip": 0.01157616, + "auxiliary_loss_mlp": 0.01035707, + "balance_loss_clip": 1.05256116, + "balance_loss_mlp": 1.02532434, + "epoch": 0.46257440028858293, + "flos": 14465070218880.0, + "grad_norm": 2.2484491606308588, + "language_loss": 0.73394656, + "learning_rate": 2.3380016406436984e-06, + "loss": 0.75587976, + "num_input_tokens_seen": 82713725, + "step": 3847, + "time_per_iteration": 2.794726610183716 + }, + { + "auxiliary_loss_clip": 0.01144268, + "auxiliary_loss_mlp": 0.01038718, + "balance_loss_clip": 1.04847741, + "balance_loss_mlp": 1.02848375, + "epoch": 0.46269464317922204, + "flos": 23332204523520.0, + "grad_norm": 1.8717080860410473, + "language_loss": 0.81286287, + "learning_rate": 2.337233848176524e-06, + "loss": 0.83469272, + "num_input_tokens_seen": 82731495, + "step": 3848, + "time_per_iteration": 2.7868993282318115 + }, + { + "auxiliary_loss_clip": 0.01140693, + "auxiliary_loss_mlp": 0.01032088, + "balance_loss_clip": 1.04767132, + "balance_loss_mlp": 1.02245295, + "epoch": 0.4628148860698611, + "flos": 18552027594240.0, + "grad_norm": 1.8383288862014089, + "language_loss": 0.83247793, + "learning_rate": 2.3364660045512435e-06, + "loss": 0.85420573, + "num_input_tokens_seen": 82750255, + "step": 3849, + "time_per_iteration": 2.8708999156951904 + }, + { + "auxiliary_loss_clip": 0.01056445, + "auxiliary_loss_mlp": 0.01003419, + "balance_loss_clip": 1.01322031, + "balance_loss_mlp": 1.00147557, + "epoch": 0.4629351289605002, + "flos": 70667569670400.0, + "grad_norm": 0.810245310429002, + "language_loss": 0.58256185, + "learning_rate": 2.335698109884337e-06, + "loss": 0.6031605, + "num_input_tokens_seen": 82815460, + "step": 3850, + "time_per_iteration": 3.449373722076416 + }, + { + "auxiliary_loss_clip": 0.01050431, + "auxiliary_loss_mlp": 0.01008241, + "balance_loss_clip": 1.01802993, + "balance_loss_mlp": 1.00596416, + "epoch": 0.4630553718511393, + "flos": 59687200465920.0, + "grad_norm": 0.7923885875327772, + "language_loss": 0.5984782, + "learning_rate": 2.334930164292294e-06, + "loss": 0.61906493, + "num_input_tokens_seen": 82878010, + "step": 3851, + "time_per_iteration": 3.4933865070343018 + }, + { + "auxiliary_loss_clip": 0.01141364, + "auxiliary_loss_mlp": 0.01033977, + "balance_loss_clip": 1.0447067, + "balance_loss_mlp": 1.02450538, + "epoch": 0.4631756147417784, + "flos": 15960605909760.0, + "grad_norm": 3.0202831593515547, + "language_loss": 0.79902655, + "learning_rate": 2.334162167891612e-06, + "loss": 0.82078004, + "num_input_tokens_seen": 82895275, + "step": 3852, + "time_per_iteration": 2.8361520767211914 + }, + { + "auxiliary_loss_clip": 0.01173796, + "auxiliary_loss_mlp": 0.01034448, + "balance_loss_clip": 1.05140626, + "balance_loss_mlp": 1.02432775, + "epoch": 0.4632958576324175, + "flos": 16472907636480.0, + "grad_norm": 2.120530721663271, + "language_loss": 0.74911702, + "learning_rate": 2.333394120798795e-06, + "loss": 0.77119946, + "num_input_tokens_seen": 82914010, + "step": 3853, + "time_per_iteration": 2.8036742210388184 + }, + { + "auxiliary_loss_clip": 0.01169594, + "auxiliary_loss_mlp": 0.01027728, + "balance_loss_clip": 1.04964221, + "balance_loss_mlp": 1.01785159, + "epoch": 0.4634161005230566, + "flos": 22346492520960.0, + "grad_norm": 2.186205282819442, + "language_loss": 0.7171849, + "learning_rate": 2.3326260231303545e-06, + "loss": 0.73915815, + "num_input_tokens_seen": 82932610, + "step": 3854, + "time_per_iteration": 2.7513532638549805 + }, + { + "auxiliary_loss_clip": 0.01188151, + "auxiliary_loss_mlp": 0.0103178, + "balance_loss_clip": 1.05508018, + "balance_loss_mlp": 1.02244568, + "epoch": 0.46353634341369565, + "flos": 15742233175680.0, + "grad_norm": 1.6981719796106363, + "language_loss": 0.86621809, + "learning_rate": 2.331857875002811e-06, + "loss": 0.88841736, + "num_input_tokens_seen": 82951210, + "step": 3855, + "time_per_iteration": 2.6319167613983154 + }, + { + "auxiliary_loss_clip": 0.01171313, + "auxiliary_loss_mlp": 0.01034868, + "balance_loss_clip": 1.05409145, + "balance_loss_mlp": 1.02552223, + "epoch": 0.46365658630433476, + "flos": 28329820433280.0, + "grad_norm": 1.6269420923716005, + "language_loss": 0.76176178, + "learning_rate": 2.3310896765326916e-06, + "loss": 0.78382361, + "num_input_tokens_seen": 82972210, + "step": 3856, + "time_per_iteration": 2.828310966491699 + }, + { + "auxiliary_loss_clip": 0.01154967, + "auxiliary_loss_mlp": 0.01040895, + "balance_loss_clip": 1.05000782, + "balance_loss_mlp": 1.03145993, + "epoch": 0.46377682919497387, + "flos": 24608074590720.0, + "grad_norm": 1.6363444408808094, + "language_loss": 0.84095466, + "learning_rate": 2.330321427836531e-06, + "loss": 0.86291325, + "num_input_tokens_seen": 82994080, + "step": 3857, + "time_per_iteration": 2.802182912826538 + }, + { + "auxiliary_loss_clip": 0.01175606, + "auxiliary_loss_mlp": 0.0103462, + "balance_loss_clip": 1.05257452, + "balance_loss_mlp": 1.02454066, + "epoch": 0.4638970720856129, + "flos": 19060953442560.0, + "grad_norm": 1.7147379582157747, + "language_loss": 0.82837147, + "learning_rate": 2.3295531290308733e-06, + "loss": 0.8504737, + "num_input_tokens_seen": 83012230, + "step": 3858, + "time_per_iteration": 3.648102045059204 + }, + { + "auxiliary_loss_clip": 0.0119414, + "auxiliary_loss_mlp": 0.00887978, + "balance_loss_clip": 1.05700731, + "balance_loss_mlp": 1.00071049, + "epoch": 0.46401731497625204, + "flos": 18471012468480.0, + "grad_norm": 2.334983082139885, + "language_loss": 0.75335634, + "learning_rate": 2.3287847802322678e-06, + "loss": 0.77417755, + "num_input_tokens_seen": 83027800, + "step": 3859, + "time_per_iteration": 3.584498167037964 + }, + { + "auxiliary_loss_clip": 0.01174335, + "auxiliary_loss_mlp": 0.01040074, + "balance_loss_clip": 1.0562638, + "balance_loss_mlp": 1.02911913, + "epoch": 0.4641375578668911, + "flos": 26067053214720.0, + "grad_norm": 1.9055540982095205, + "language_loss": 0.84095633, + "learning_rate": 2.3280163815572723e-06, + "loss": 0.86310041, + "num_input_tokens_seen": 83048395, + "step": 3860, + "time_per_iteration": 3.6193110942840576 + }, + { + "auxiliary_loss_clip": 0.01160916, + "auxiliary_loss_mlp": 0.01030809, + "balance_loss_clip": 1.0496068, + "balance_loss_mlp": 1.02128434, + "epoch": 0.4642578007575302, + "flos": 19570382081280.0, + "grad_norm": 2.9320270841796594, + "language_loss": 0.77307576, + "learning_rate": 2.3272479331224522e-06, + "loss": 0.79499298, + "num_input_tokens_seen": 83065825, + "step": 3861, + "time_per_iteration": 3.6228158473968506 + }, + { + "auxiliary_loss_clip": 0.01193263, + "auxiliary_loss_mlp": 0.01037926, + "balance_loss_clip": 1.05647326, + "balance_loss_mlp": 1.02881265, + "epoch": 0.4643780436481693, + "flos": 28186249772160.0, + "grad_norm": 2.6193139532236827, + "language_loss": 0.77884305, + "learning_rate": 2.3264794350443817e-06, + "loss": 0.80115491, + "num_input_tokens_seen": 83087920, + "step": 3862, + "time_per_iteration": 2.699296712875366 + }, + { + "auxiliary_loss_clip": 0.01178636, + "auxiliary_loss_mlp": 0.01036029, + "balance_loss_clip": 1.05013144, + "balance_loss_mlp": 1.02674866, + "epoch": 0.46449828653880837, + "flos": 25375270204800.0, + "grad_norm": 1.9213648292199883, + "language_loss": 0.78700477, + "learning_rate": 2.3257108874396396e-06, + "loss": 0.80915141, + "num_input_tokens_seen": 83109015, + "step": 3863, + "time_per_iteration": 2.7659835815429688 + }, + { + "auxiliary_loss_clip": 0.01168133, + "auxiliary_loss_mlp": 0.01034792, + "balance_loss_clip": 1.05023074, + "balance_loss_mlp": 1.0257616, + "epoch": 0.4646185294294475, + "flos": 16034330574720.0, + "grad_norm": 2.4951745029908445, + "language_loss": 0.7364974, + "learning_rate": 2.3249422904248152e-06, + "loss": 0.75852668, + "num_input_tokens_seen": 83127450, + "step": 3864, + "time_per_iteration": 2.6980738639831543 + }, + { + "auxiliary_loss_clip": 0.01182365, + "auxiliary_loss_mlp": 0.01029596, + "balance_loss_clip": 1.05298364, + "balance_loss_mlp": 1.02053654, + "epoch": 0.4647387723200866, + "flos": 26363101109760.0, + "grad_norm": 2.1952486623345164, + "language_loss": 0.87407905, + "learning_rate": 2.324173644116504e-06, + "loss": 0.89619863, + "num_input_tokens_seen": 83150300, + "step": 3865, + "time_per_iteration": 2.835822582244873 + }, + { + "auxiliary_loss_clip": 0.01177323, + "auxiliary_loss_mlp": 0.01033295, + "balance_loss_clip": 1.05513966, + "balance_loss_mlp": 1.02451551, + "epoch": 0.46485901521072565, + "flos": 27160209774720.0, + "grad_norm": 1.69221206186422, + "language_loss": 0.81214011, + "learning_rate": 2.3234049486313087e-06, + "loss": 0.83424628, + "num_input_tokens_seen": 83171750, + "step": 3866, + "time_per_iteration": 2.7721173763275146 + }, + { + "auxiliary_loss_clip": 0.01178049, + "auxiliary_loss_mlp": 0.01027818, + "balance_loss_clip": 1.05329418, + "balance_loss_mlp": 1.01898444, + "epoch": 0.46497925810136476, + "flos": 24279851088000.0, + "grad_norm": 1.9893132914886726, + "language_loss": 0.7579577, + "learning_rate": 2.322636204085839e-06, + "loss": 0.78001636, + "num_input_tokens_seen": 83191820, + "step": 3867, + "time_per_iteration": 2.8037192821502686 + }, + { + "auxiliary_loss_clip": 0.01157253, + "auxiliary_loss_mlp": 0.01039726, + "balance_loss_clip": 1.04502153, + "balance_loss_mlp": 1.02975965, + "epoch": 0.46509950099200387, + "flos": 16253134272000.0, + "grad_norm": 2.3491216221060736, + "language_loss": 0.78318059, + "learning_rate": 2.3218674105967143e-06, + "loss": 0.80515039, + "num_input_tokens_seen": 83210085, + "step": 3868, + "time_per_iteration": 2.7257816791534424 + }, + { + "auxiliary_loss_clip": 0.01160624, + "auxiliary_loss_mlp": 0.01027771, + "balance_loss_clip": 1.0484767, + "balance_loss_mlp": 1.01846719, + "epoch": 0.4652197438826429, + "flos": 23442270773760.0, + "grad_norm": 1.833006929777914, + "language_loss": 0.83389032, + "learning_rate": 2.3210985682805593e-06, + "loss": 0.85577434, + "num_input_tokens_seen": 83231865, + "step": 3869, + "time_per_iteration": 2.761937379837036 + }, + { + "auxiliary_loss_clip": 0.01191448, + "auxiliary_loss_mlp": 0.01031811, + "balance_loss_clip": 1.05690134, + "balance_loss_mlp": 1.02179098, + "epoch": 0.46533998677328203, + "flos": 16216397637120.0, + "grad_norm": 4.211216847743187, + "language_loss": 0.68072379, + "learning_rate": 2.320329677254007e-06, + "loss": 0.70295638, + "num_input_tokens_seen": 83249195, + "step": 3870, + "time_per_iteration": 2.623664379119873 + }, + { + "auxiliary_loss_clip": 0.01190261, + "auxiliary_loss_mlp": 0.01030243, + "balance_loss_clip": 1.05544066, + "balance_loss_mlp": 1.02068257, + "epoch": 0.46546022966392114, + "flos": 21141869080320.0, + "grad_norm": 3.3610908951226204, + "language_loss": 0.7239483, + "learning_rate": 2.319560737633697e-06, + "loss": 0.74615335, + "num_input_tokens_seen": 83267915, + "step": 3871, + "time_per_iteration": 2.705091714859009 + }, + { + "auxiliary_loss_clip": 0.01165862, + "auxiliary_loss_mlp": 0.01033728, + "balance_loss_clip": 1.04998529, + "balance_loss_mlp": 1.02372682, + "epoch": 0.4655804725545602, + "flos": 41171942442240.0, + "grad_norm": 1.6980325362310116, + "language_loss": 0.68022287, + "learning_rate": 2.3187917495362775e-06, + "loss": 0.70221883, + "num_input_tokens_seen": 83292325, + "step": 3872, + "time_per_iteration": 3.0103871822357178 + }, + { + "auxiliary_loss_clip": 0.01147195, + "auxiliary_loss_mlp": 0.01034155, + "balance_loss_clip": 1.04795504, + "balance_loss_mlp": 1.02495825, + "epoch": 0.4657007154451993, + "flos": 19570956698880.0, + "grad_norm": 2.524706226886504, + "language_loss": 0.76787978, + "learning_rate": 2.318022713078403e-06, + "loss": 0.7896933, + "num_input_tokens_seen": 83306905, + "step": 3873, + "time_per_iteration": 2.7738096714019775 + }, + { + "auxiliary_loss_clip": 0.01168207, + "auxiliary_loss_mlp": 0.01032331, + "balance_loss_clip": 1.0494169, + "balance_loss_mlp": 1.02285349, + "epoch": 0.4658209583358384, + "flos": 15517826956800.0, + "grad_norm": 2.0315586588476355, + "language_loss": 0.84966558, + "learning_rate": 2.3172536283767354e-06, + "loss": 0.87167096, + "num_input_tokens_seen": 83320665, + "step": 3874, + "time_per_iteration": 2.8621740341186523 + }, + { + "auxiliary_loss_clip": 0.01155704, + "auxiliary_loss_mlp": 0.01037655, + "balance_loss_clip": 1.05045843, + "balance_loss_mlp": 1.02857685, + "epoch": 0.4659412012264775, + "flos": 14903180403840.0, + "grad_norm": 2.24657729872201, + "language_loss": 0.80995089, + "learning_rate": 2.3164844955479447e-06, + "loss": 0.8318845, + "num_input_tokens_seen": 83336475, + "step": 3875, + "time_per_iteration": 2.867797374725342 + }, + { + "auxiliary_loss_clip": 0.01154715, + "auxiliary_loss_mlp": 0.01034752, + "balance_loss_clip": 1.04991269, + "balance_loss_mlp": 1.02479243, + "epoch": 0.4660614441171166, + "flos": 24425612478720.0, + "grad_norm": 1.5520565316023147, + "language_loss": 0.70542043, + "learning_rate": 2.3157153147087082e-06, + "loss": 0.72731513, + "num_input_tokens_seen": 83358365, + "step": 3876, + "time_per_iteration": 2.882195234298706 + }, + { + "auxiliary_loss_clip": 0.01152, + "auxiliary_loss_mlp": 0.01032426, + "balance_loss_clip": 1.05044663, + "balance_loss_mlp": 1.02381921, + "epoch": 0.46618168700775564, + "flos": 22091095843200.0, + "grad_norm": 8.83900735825454, + "language_loss": 0.83180833, + "learning_rate": 2.314946085975709e-06, + "loss": 0.8536526, + "num_input_tokens_seen": 83377345, + "step": 3877, + "time_per_iteration": 2.8693408966064453 + }, + { + "auxiliary_loss_clip": 0.01146916, + "auxiliary_loss_mlp": 0.01031064, + "balance_loss_clip": 1.04927945, + "balance_loss_mlp": 1.02266526, + "epoch": 0.46630192989839475, + "flos": 26176975810560.0, + "grad_norm": 1.6704768594704913, + "language_loss": 0.82149327, + "learning_rate": 2.3141768094656393e-06, + "loss": 0.84327304, + "num_input_tokens_seen": 83395920, + "step": 3878, + "time_per_iteration": 2.831439733505249 + }, + { + "auxiliary_loss_clip": 0.01137572, + "auxiliary_loss_mlp": 0.01031445, + "balance_loss_clip": 1.04504204, + "balance_loss_mlp": 1.02203941, + "epoch": 0.46642217278903386, + "flos": 11509622150400.0, + "grad_norm": 2.597667180044179, + "language_loss": 0.83070123, + "learning_rate": 2.3134074852951966e-06, + "loss": 0.85239136, + "num_input_tokens_seen": 83412510, + "step": 3879, + "time_per_iteration": 2.9393131732940674 + }, + { + "auxiliary_loss_clip": 0.01140008, + "auxiliary_loss_mlp": 0.01034706, + "balance_loss_clip": 1.04353333, + "balance_loss_mlp": 1.02447224, + "epoch": 0.4665424156796729, + "flos": 32306819299200.0, + "grad_norm": 1.7936963880552645, + "language_loss": 0.77764058, + "learning_rate": 2.312638113581088e-06, + "loss": 0.79938769, + "num_input_tokens_seen": 83432995, + "step": 3880, + "time_per_iteration": 2.9988536834716797 + }, + { + "auxiliary_loss_clip": 0.01175601, + "auxiliary_loss_mlp": 0.01035029, + "balance_loss_clip": 1.04917979, + "balance_loss_mlp": 1.02590895, + "epoch": 0.46666265857031203, + "flos": 18436179254400.0, + "grad_norm": 2.357573755324563, + "language_loss": 0.78072888, + "learning_rate": 2.311868694440027e-06, + "loss": 0.80283517, + "num_input_tokens_seen": 83447415, + "step": 3881, + "time_per_iteration": 2.746164560317993 + }, + { + "auxiliary_loss_clip": 0.01077105, + "auxiliary_loss_mlp": 0.0100298, + "balance_loss_clip": 1.0159682, + "balance_loss_mlp": 1.00123942, + "epoch": 0.46678290146095114, + "flos": 68438989221120.0, + "grad_norm": 0.7298602436053768, + "language_loss": 0.62472272, + "learning_rate": 2.3110992279887323e-06, + "loss": 0.64552355, + "num_input_tokens_seen": 83519340, + "step": 3882, + "time_per_iteration": 3.4217782020568848 + }, + { + "auxiliary_loss_clip": 0.01162562, + "auxiliary_loss_mlp": 0.01032627, + "balance_loss_clip": 1.05002713, + "balance_loss_mlp": 1.0232867, + "epoch": 0.4669031443515902, + "flos": 17712507945600.0, + "grad_norm": 3.434987380178413, + "language_loss": 0.84796417, + "learning_rate": 2.310329714343932e-06, + "loss": 0.86991596, + "num_input_tokens_seen": 83535490, + "step": 3883, + "time_per_iteration": 2.8504457473754883 + }, + { + "auxiliary_loss_clip": 0.01164075, + "auxiliary_loss_mlp": 0.01035416, + "balance_loss_clip": 1.05000055, + "balance_loss_mlp": 1.02629638, + "epoch": 0.4670233872422293, + "flos": 23947748916480.0, + "grad_norm": 2.596231135815362, + "language_loss": 0.81772745, + "learning_rate": 2.309560153622361e-06, + "loss": 0.83972234, + "num_input_tokens_seen": 83552400, + "step": 3884, + "time_per_iteration": 3.6359829902648926 + }, + { + "auxiliary_loss_clip": 0.01158728, + "auxiliary_loss_mlp": 0.01037246, + "balance_loss_clip": 1.0519594, + "balance_loss_mlp": 1.0273571, + "epoch": 0.4671436301328684, + "flos": 28111268131200.0, + "grad_norm": 2.0189282761161262, + "language_loss": 0.74325001, + "learning_rate": 2.3087905459407602e-06, + "loss": 0.76520979, + "num_input_tokens_seen": 83571340, + "step": 3885, + "time_per_iteration": 3.8034892082214355 + }, + { + "auxiliary_loss_clip": 0.0106659, + "auxiliary_loss_mlp": 0.01003562, + "balance_loss_clip": 1.01502979, + "balance_loss_mlp": 1.00180936, + "epoch": 0.46726387302350747, + "flos": 69369684566400.0, + "grad_norm": 0.8064252948881919, + "language_loss": 0.62899601, + "learning_rate": 2.3080208914158795e-06, + "loss": 0.64969748, + "num_input_tokens_seen": 83634340, + "step": 3886, + "time_per_iteration": 4.143473386764526 + }, + { + "auxiliary_loss_clip": 0.01169138, + "auxiliary_loss_mlp": 0.01032414, + "balance_loss_clip": 1.05374265, + "balance_loss_mlp": 1.02273417, + "epoch": 0.4673841159141466, + "flos": 25519666878720.0, + "grad_norm": 1.9800722243001578, + "language_loss": 0.72348815, + "learning_rate": 2.3072511901644753e-06, + "loss": 0.74550366, + "num_input_tokens_seen": 83653410, + "step": 3887, + "time_per_iteration": 3.6493191719055176 + }, + { + "auxiliary_loss_clip": 0.01186004, + "auxiliary_loss_mlp": 0.01032834, + "balance_loss_clip": 1.05500293, + "balance_loss_mlp": 1.02459657, + "epoch": 0.4675043588047857, + "flos": 24499265316480.0, + "grad_norm": 1.8477752578987725, + "language_loss": 0.8069517, + "learning_rate": 2.306481442303309e-06, + "loss": 0.82914007, + "num_input_tokens_seen": 83672985, + "step": 3888, + "time_per_iteration": 2.715444803237915 + }, + { + "auxiliary_loss_clip": 0.01181112, + "auxiliary_loss_mlp": 0.01033671, + "balance_loss_clip": 1.05280256, + "balance_loss_mlp": 1.02440202, + "epoch": 0.46762460169542475, + "flos": 20960771685120.0, + "grad_norm": 1.743151362896458, + "language_loss": 0.73266673, + "learning_rate": 2.3057116479491515e-06, + "loss": 0.75481462, + "num_input_tokens_seen": 83692395, + "step": 3889, + "time_per_iteration": 2.7206146717071533 + }, + { + "auxiliary_loss_clip": 0.01172757, + "auxiliary_loss_mlp": 0.0103226, + "balance_loss_clip": 1.04849696, + "balance_loss_mlp": 1.0225029, + "epoch": 0.46774484458606386, + "flos": 19171666137600.0, + "grad_norm": 2.446029158913468, + "language_loss": 0.76394677, + "learning_rate": 2.30494180721878e-06, + "loss": 0.78599691, + "num_input_tokens_seen": 83709735, + "step": 3890, + "time_per_iteration": 2.7164556980133057 + }, + { + "auxiliary_loss_clip": 0.01174965, + "auxiliary_loss_mlp": 0.01039695, + "balance_loss_clip": 1.05121171, + "balance_loss_mlp": 1.03061152, + "epoch": 0.4678650874767029, + "flos": 17967689141760.0, + "grad_norm": 2.547054380133713, + "language_loss": 0.89894056, + "learning_rate": 2.3041719202289794e-06, + "loss": 0.92108715, + "num_input_tokens_seen": 83725910, + "step": 3891, + "time_per_iteration": 2.646850824356079 + }, + { + "auxiliary_loss_clip": 0.01178475, + "auxiliary_loss_mlp": 0.01031701, + "balance_loss_clip": 1.05421209, + "balance_loss_mlp": 1.02302909, + "epoch": 0.467985330367342, + "flos": 21360816432000.0, + "grad_norm": 1.7118215189549009, + "language_loss": 0.80433428, + "learning_rate": 2.30340198709654e-06, + "loss": 0.82643604, + "num_input_tokens_seen": 83745745, + "step": 3892, + "time_per_iteration": 2.7656679153442383 + }, + { + "auxiliary_loss_clip": 0.01170478, + "auxiliary_loss_mlp": 0.01035119, + "balance_loss_clip": 1.04894876, + "balance_loss_mlp": 1.02659512, + "epoch": 0.46810557325798113, + "flos": 20521835487360.0, + "grad_norm": 2.8060368921058108, + "language_loss": 0.74384665, + "learning_rate": 2.3026320079382605e-06, + "loss": 0.76590264, + "num_input_tokens_seen": 83762680, + "step": 3893, + "time_per_iteration": 2.7424216270446777 + }, + { + "auxiliary_loss_clip": 0.01186747, + "auxiliary_loss_mlp": 0.01030813, + "balance_loss_clip": 1.05553579, + "balance_loss_mlp": 1.02158022, + "epoch": 0.4682258161486202, + "flos": 30117848572800.0, + "grad_norm": 3.461545430667907, + "language_loss": 0.76511312, + "learning_rate": 2.3018619828709454e-06, + "loss": 0.78728867, + "num_input_tokens_seen": 83784220, + "step": 3894, + "time_per_iteration": 2.796741008758545 + }, + { + "auxiliary_loss_clip": 0.01173997, + "auxiliary_loss_mlp": 0.00886518, + "balance_loss_clip": 1.05351818, + "balance_loss_mlp": 1.00081134, + "epoch": 0.4683460590392593, + "flos": 25293357239040.0, + "grad_norm": 1.7585430947808616, + "language_loss": 0.8202917, + "learning_rate": 2.3010919120114084e-06, + "loss": 0.84089684, + "num_input_tokens_seen": 83800750, + "step": 3895, + "time_per_iteration": 2.7740097045898438 + }, + { + "auxiliary_loss_clip": 0.01169278, + "auxiliary_loss_mlp": 0.01031663, + "balance_loss_clip": 1.0463351, + "balance_loss_mlp": 1.02181697, + "epoch": 0.4684663019298984, + "flos": 15368330551680.0, + "grad_norm": 2.13717406817541, + "language_loss": 0.66103399, + "learning_rate": 2.3003217954764672e-06, + "loss": 0.68304342, + "num_input_tokens_seen": 83815455, + "step": 3896, + "time_per_iteration": 2.7954962253570557 + }, + { + "auxiliary_loss_clip": 0.01177771, + "auxiliary_loss_mlp": 0.01029375, + "balance_loss_clip": 1.04907072, + "balance_loss_mlp": 1.02045178, + "epoch": 0.46858654482053747, + "flos": 27778842737280.0, + "grad_norm": 1.7826378487367407, + "language_loss": 0.79171509, + "learning_rate": 2.299551633382949e-06, + "loss": 0.81378657, + "num_input_tokens_seen": 83835765, + "step": 3897, + "time_per_iteration": 2.8302648067474365 + }, + { + "auxiliary_loss_clip": 0.01159547, + "auxiliary_loss_mlp": 0.01033206, + "balance_loss_clip": 1.04772055, + "balance_loss_mlp": 1.02347255, + "epoch": 0.4687067877111766, + "flos": 18040623707520.0, + "grad_norm": 1.9977827065796665, + "language_loss": 0.85467386, + "learning_rate": 2.2987814258476854e-06, + "loss": 0.87660134, + "num_input_tokens_seen": 83853565, + "step": 3898, + "time_per_iteration": 2.7895915508270264 + }, + { + "auxiliary_loss_clip": 0.01150148, + "auxiliary_loss_mlp": 0.0103269, + "balance_loss_clip": 1.04491937, + "balance_loss_mlp": 1.02363563, + "epoch": 0.4688270306018157, + "flos": 16977380198400.0, + "grad_norm": 2.9739078138604422, + "language_loss": 0.67704284, + "learning_rate": 2.2980111729875177e-06, + "loss": 0.69887125, + "num_input_tokens_seen": 83869815, + "step": 3899, + "time_per_iteration": 2.837193727493286 + }, + { + "auxiliary_loss_clip": 0.01163459, + "auxiliary_loss_mlp": 0.0103385, + "balance_loss_clip": 1.05362546, + "balance_loss_mlp": 1.02488577, + "epoch": 0.46894727349245474, + "flos": 17821640442240.0, + "grad_norm": 1.6813425002398483, + "language_loss": 0.8233912, + "learning_rate": 2.2972408749192917e-06, + "loss": 0.84536427, + "num_input_tokens_seen": 83887545, + "step": 3900, + "time_per_iteration": 2.7144861221313477 + }, + { + "auxiliary_loss_clip": 0.01175821, + "auxiliary_loss_mlp": 0.00886376, + "balance_loss_clip": 1.05429435, + "balance_loss_mlp": 1.00077569, + "epoch": 0.46906751638309385, + "flos": 21471349559040.0, + "grad_norm": 1.821541376439088, + "language_loss": 0.66641212, + "learning_rate": 2.296470531759861e-06, + "loss": 0.68703413, + "num_input_tokens_seen": 83905645, + "step": 3901, + "time_per_iteration": 2.768643379211426 + }, + { + "auxiliary_loss_clip": 0.01150311, + "auxiliary_loss_mlp": 0.01032622, + "balance_loss_clip": 1.04623628, + "balance_loss_mlp": 1.02346694, + "epoch": 0.46918775927373296, + "flos": 20337829090560.0, + "grad_norm": 2.0650510907110347, + "language_loss": 0.79481244, + "learning_rate": 2.2957001436260866e-06, + "loss": 0.81664181, + "num_input_tokens_seen": 83922705, + "step": 3902, + "time_per_iteration": 2.799476146697998 + }, + { + "auxiliary_loss_clip": 0.01163222, + "auxiliary_loss_mlp": 0.01036679, + "balance_loss_clip": 1.05054927, + "balance_loss_mlp": 1.02773809, + "epoch": 0.469308002164372, + "flos": 18403249461120.0, + "grad_norm": 1.6088424605741491, + "language_loss": 0.73212373, + "learning_rate": 2.294929710634836e-06, + "loss": 0.75412279, + "num_input_tokens_seen": 83940795, + "step": 3903, + "time_per_iteration": 2.7919321060180664 + }, + { + "auxiliary_loss_clip": 0.01175338, + "auxiliary_loss_mlp": 0.01031784, + "balance_loss_clip": 1.04986072, + "balance_loss_mlp": 1.02262282, + "epoch": 0.46942824505501113, + "flos": 37962067363200.0, + "grad_norm": 1.842516752913391, + "language_loss": 0.61149567, + "learning_rate": 2.2941592329029823e-06, + "loss": 0.63356686, + "num_input_tokens_seen": 83961900, + "step": 3904, + "time_per_iteration": 2.850949287414551 + }, + { + "auxiliary_loss_clip": 0.01171805, + "auxiliary_loss_mlp": 0.0103944, + "balance_loss_clip": 1.05032957, + "balance_loss_mlp": 1.03042793, + "epoch": 0.46954848794565024, + "flos": 21872507627520.0, + "grad_norm": 1.9815089873963097, + "language_loss": 0.78997409, + "learning_rate": 2.2933887105474067e-06, + "loss": 0.81208652, + "num_input_tokens_seen": 83980075, + "step": 3905, + "time_per_iteration": 2.710878610610962 + }, + { + "auxiliary_loss_clip": 0.01172527, + "auxiliary_loss_mlp": 0.0103588, + "balance_loss_clip": 1.05268764, + "balance_loss_mlp": 1.02779782, + "epoch": 0.4696687308362893, + "flos": 22016545165440.0, + "grad_norm": 1.5671447987766287, + "language_loss": 0.81489331, + "learning_rate": 2.2926181436849974e-06, + "loss": 0.83697736, + "num_input_tokens_seen": 83999430, + "step": 3906, + "time_per_iteration": 2.7135491371154785 + }, + { + "auxiliary_loss_clip": 0.01175939, + "auxiliary_loss_mlp": 0.01032014, + "balance_loss_clip": 1.0536176, + "balance_loss_mlp": 1.02350843, + "epoch": 0.4697889737269284, + "flos": 21613663244160.0, + "grad_norm": 1.8034211885501306, + "language_loss": 0.72426939, + "learning_rate": 2.2918475324326478e-06, + "loss": 0.74634886, + "num_input_tokens_seen": 84019150, + "step": 3907, + "time_per_iteration": 2.7468886375427246 + }, + { + "auxiliary_loss_clip": 0.01180057, + "auxiliary_loss_mlp": 0.00887537, + "balance_loss_clip": 1.05340099, + "balance_loss_mlp": 1.00077093, + "epoch": 0.46990921661756746, + "flos": 25228323665280.0, + "grad_norm": 2.124291932282582, + "language_loss": 0.91124028, + "learning_rate": 2.2910768769072603e-06, + "loss": 0.93191624, + "num_input_tokens_seen": 84037930, + "step": 3908, + "time_per_iteration": 2.7879204750061035 + }, + { + "auxiliary_loss_clip": 0.01171808, + "auxiliary_loss_mlp": 0.0102654, + "balance_loss_clip": 1.05010819, + "balance_loss_mlp": 1.01783204, + "epoch": 0.47002945950820657, + "flos": 13844031045120.0, + "grad_norm": 1.8908502907953997, + "language_loss": 0.7588768, + "learning_rate": 2.2903061772257417e-06, + "loss": 0.78086019, + "num_input_tokens_seen": 84055915, + "step": 3909, + "time_per_iteration": 2.7010269165039062 + }, + { + "auxiliary_loss_clip": 0.01178376, + "auxiliary_loss_mlp": 0.01034592, + "balance_loss_clip": 1.05483747, + "balance_loss_mlp": 1.0249126, + "epoch": 0.4701497023988457, + "flos": 26247001374720.0, + "grad_norm": 1.5646081460103327, + "language_loss": 0.7838279, + "learning_rate": 2.289535433505007e-06, + "loss": 0.80595762, + "num_input_tokens_seen": 84077270, + "step": 3910, + "time_per_iteration": 3.7461681365966797 + }, + { + "auxiliary_loss_clip": 0.01164647, + "auxiliary_loss_mlp": 0.01035819, + "balance_loss_clip": 1.05070877, + "balance_loss_mlp": 1.02710485, + "epoch": 0.47026994528948474, + "flos": 25629517647360.0, + "grad_norm": 1.8122033630393581, + "language_loss": 0.63953054, + "learning_rate": 2.2887646458619767e-06, + "loss": 0.6615352, + "num_input_tokens_seen": 84098635, + "step": 3911, + "time_per_iteration": 3.7989954948425293 + }, + { + "auxiliary_loss_clip": 0.01162911, + "auxiliary_loss_mlp": 0.01036836, + "balance_loss_clip": 1.05082583, + "balance_loss_mlp": 1.02740073, + "epoch": 0.47039018818012385, + "flos": 20554406144640.0, + "grad_norm": 1.7699037020215678, + "language_loss": 0.76191884, + "learning_rate": 2.2879938144135797e-06, + "loss": 0.78391629, + "num_input_tokens_seen": 84114740, + "step": 3912, + "time_per_iteration": 3.7184817790985107 + }, + { + "auxiliary_loss_clip": 0.01152345, + "auxiliary_loss_mlp": 0.00886258, + "balance_loss_clip": 1.0442704, + "balance_loss_mlp": 1.00077116, + "epoch": 0.47051043107076296, + "flos": 21577249831680.0, + "grad_norm": 1.5614349607981746, + "language_loss": 0.75218797, + "learning_rate": 2.2872229392767496e-06, + "loss": 0.77257395, + "num_input_tokens_seen": 84134845, + "step": 3913, + "time_per_iteration": 4.011695623397827 + }, + { + "auxiliary_loss_clip": 0.0117976, + "auxiliary_loss_mlp": 0.01034761, + "balance_loss_clip": 1.05198669, + "balance_loss_mlp": 1.02577209, + "epoch": 0.470630673961402, + "flos": 18953185662720.0, + "grad_norm": 1.5675496822316213, + "language_loss": 0.74473464, + "learning_rate": 2.286452020568428e-06, + "loss": 0.76687986, + "num_input_tokens_seen": 84152920, + "step": 3914, + "time_per_iteration": 2.7830722332000732 + }, + { + "auxiliary_loss_clip": 0.01191123, + "auxiliary_loss_mlp": 0.01031203, + "balance_loss_clip": 1.05450916, + "balance_loss_mlp": 1.02160704, + "epoch": 0.4707509168520411, + "flos": 19938969492480.0, + "grad_norm": 1.6156341429514187, + "language_loss": 0.73026156, + "learning_rate": 2.2856810584055637e-06, + "loss": 0.7524848, + "num_input_tokens_seen": 84170455, + "step": 3915, + "time_per_iteration": 2.6787917613983154 + }, + { + "auxiliary_loss_clip": 0.01177385, + "auxiliary_loss_mlp": 0.01039848, + "balance_loss_clip": 1.05215001, + "balance_loss_mlp": 1.03108072, + "epoch": 0.47087115974268023, + "flos": 40118754741120.0, + "grad_norm": 1.6307509937587017, + "language_loss": 0.68000823, + "learning_rate": 2.2849100529051085e-06, + "loss": 0.70218062, + "num_input_tokens_seen": 84197390, + "step": 3916, + "time_per_iteration": 3.012949228286743 + }, + { + "auxiliary_loss_clip": 0.01182863, + "auxiliary_loss_mlp": 0.01037242, + "balance_loss_clip": 1.05276179, + "balance_loss_mlp": 1.02847958, + "epoch": 0.4709914026333193, + "flos": 13552723745280.0, + "grad_norm": 2.2127964296451004, + "language_loss": 0.79752803, + "learning_rate": 2.284139004184026e-06, + "loss": 0.81972903, + "num_input_tokens_seen": 84214620, + "step": 3917, + "time_per_iteration": 2.689970016479492 + }, + { + "auxiliary_loss_clip": 0.01187273, + "auxiliary_loss_mlp": 0.01032616, + "balance_loss_clip": 1.05406737, + "balance_loss_mlp": 1.02391052, + "epoch": 0.4711116455239584, + "flos": 19974628719360.0, + "grad_norm": 2.7797277189079854, + "language_loss": 0.74465656, + "learning_rate": 2.2833679123592814e-06, + "loss": 0.76685548, + "num_input_tokens_seen": 84231880, + "step": 3918, + "time_per_iteration": 2.751000165939331 + }, + { + "auxiliary_loss_clip": 0.01166322, + "auxiliary_loss_mlp": 0.01030196, + "balance_loss_clip": 1.05189276, + "balance_loss_mlp": 1.02083254, + "epoch": 0.4712318884145975, + "flos": 32124824064000.0, + "grad_norm": 1.6553206095206578, + "language_loss": 0.63150352, + "learning_rate": 2.2825967775478508e-06, + "loss": 0.65346867, + "num_input_tokens_seen": 84252980, + "step": 3919, + "time_per_iteration": 2.8694374561309814 + }, + { + "auxiliary_loss_clip": 0.01186625, + "auxiliary_loss_mlp": 0.01027528, + "balance_loss_clip": 1.05316544, + "balance_loss_mlp": 1.01846182, + "epoch": 0.47135213130523657, + "flos": 20047850593920.0, + "grad_norm": 2.224844475109051, + "language_loss": 0.83647692, + "learning_rate": 2.2818255998667135e-06, + "loss": 0.8586185, + "num_input_tokens_seen": 84271490, + "step": 3920, + "time_per_iteration": 2.627056360244751 + }, + { + "auxiliary_loss_clip": 0.01176833, + "auxiliary_loss_mlp": 0.01030265, + "balance_loss_clip": 1.05395675, + "balance_loss_mlp": 1.02112746, + "epoch": 0.4714723741958757, + "flos": 19426990988160.0, + "grad_norm": 1.612540074020252, + "language_loss": 0.79189575, + "learning_rate": 2.2810543794328566e-06, + "loss": 0.81396675, + "num_input_tokens_seen": 84290525, + "step": 3921, + "time_per_iteration": 2.6928272247314453 + }, + { + "auxiliary_loss_clip": 0.01179389, + "auxiliary_loss_mlp": 0.01031827, + "balance_loss_clip": 1.05268764, + "balance_loss_mlp": 1.02262437, + "epoch": 0.4715926170865148, + "flos": 20373883367040.0, + "grad_norm": 1.8185915974405888, + "language_loss": 0.82318974, + "learning_rate": 2.2802831163632735e-06, + "loss": 0.84530187, + "num_input_tokens_seen": 84309245, + "step": 3922, + "time_per_iteration": 2.7349653244018555 + }, + { + "auxiliary_loss_clip": 0.01139301, + "auxiliary_loss_mlp": 0.01035357, + "balance_loss_clip": 1.04718661, + "balance_loss_mlp": 1.02537286, + "epoch": 0.47171285997715384, + "flos": 22672884430080.0, + "grad_norm": 1.9513790041544776, + "language_loss": 0.74495864, + "learning_rate": 2.279511810774965e-06, + "loss": 0.76670521, + "num_input_tokens_seen": 84330775, + "step": 3923, + "time_per_iteration": 2.9403700828552246 + }, + { + "auxiliary_loss_clip": 0.01184161, + "auxiliary_loss_mlp": 0.01023688, + "balance_loss_clip": 1.0514822, + "balance_loss_mlp": 1.01493788, + "epoch": 0.47183310286779295, + "flos": 21105419754240.0, + "grad_norm": 1.9398564277809154, + "language_loss": 0.71738714, + "learning_rate": 2.2787404627849364e-06, + "loss": 0.73946559, + "num_input_tokens_seen": 84349985, + "step": 3924, + "time_per_iteration": 2.6647284030914307 + }, + { + "auxiliary_loss_clip": 0.01163225, + "auxiliary_loss_mlp": 0.01028887, + "balance_loss_clip": 1.04857278, + "balance_loss_mlp": 1.0200541, + "epoch": 0.471953345758432, + "flos": 21726566668800.0, + "grad_norm": 1.724732851167327, + "language_loss": 0.78984618, + "learning_rate": 2.277969072510202e-06, + "loss": 0.81176734, + "num_input_tokens_seen": 84368965, + "step": 3925, + "time_per_iteration": 2.8823330402374268 + }, + { + "auxiliary_loss_clip": 0.01166498, + "auxiliary_loss_mlp": 0.01026006, + "balance_loss_clip": 1.05189967, + "balance_loss_mlp": 1.01759601, + "epoch": 0.4720735886490711, + "flos": 19861078849920.0, + "grad_norm": 1.5806564121407018, + "language_loss": 0.8134982, + "learning_rate": 2.2771976400677803e-06, + "loss": 0.83542323, + "num_input_tokens_seen": 84387795, + "step": 3926, + "time_per_iteration": 2.8273091316223145 + }, + { + "auxiliary_loss_clip": 0.01136567, + "auxiliary_loss_mlp": 0.01032996, + "balance_loss_clip": 1.04603326, + "balance_loss_mlp": 1.02445459, + "epoch": 0.47219383153971023, + "flos": 19171809792000.0, + "grad_norm": 1.7308131601528576, + "language_loss": 0.78657889, + "learning_rate": 2.2764261655746965e-06, + "loss": 0.80827451, + "num_input_tokens_seen": 84405290, + "step": 3927, + "time_per_iteration": 2.9650816917419434 + }, + { + "auxiliary_loss_clip": 0.01158162, + "auxiliary_loss_mlp": 0.01030576, + "balance_loss_clip": 1.05211473, + "balance_loss_mlp": 1.02157593, + "epoch": 0.4723140744303493, + "flos": 23224005780480.0, + "grad_norm": 1.7248174928983369, + "language_loss": 0.75856209, + "learning_rate": 2.2756546491479832e-06, + "loss": 0.78044951, + "num_input_tokens_seen": 84426205, + "step": 3928, + "time_per_iteration": 2.932588815689087 + }, + { + "auxiliary_loss_clip": 0.01187117, + "auxiliary_loss_mlp": 0.00886925, + "balance_loss_clip": 1.05335736, + "balance_loss_mlp": 1.00074959, + "epoch": 0.4724343173209884, + "flos": 18223265387520.0, + "grad_norm": 3.2017643318127025, + "language_loss": 0.80334604, + "learning_rate": 2.274883090904679e-06, + "loss": 0.82408643, + "num_input_tokens_seen": 84443970, + "step": 3929, + "time_per_iteration": 2.7198257446289062 + }, + { + "auxiliary_loss_clip": 0.01191508, + "auxiliary_loss_mlp": 0.01028203, + "balance_loss_clip": 1.05657911, + "balance_loss_mlp": 1.01978123, + "epoch": 0.4725545602116275, + "flos": 21251037490560.0, + "grad_norm": 2.2994604434417174, + "language_loss": 0.67833728, + "learning_rate": 2.2741114909618283e-06, + "loss": 0.7005344, + "num_input_tokens_seen": 84459865, + "step": 3930, + "time_per_iteration": 2.7505669593811035 + }, + { + "auxiliary_loss_clip": 0.01157975, + "auxiliary_loss_mlp": 0.01031975, + "balance_loss_clip": 1.04984426, + "balance_loss_mlp": 1.02332008, + "epoch": 0.47267480310226656, + "flos": 21434002392960.0, + "grad_norm": 2.599188196010833, + "language_loss": 0.72322148, + "learning_rate": 2.2733398494364828e-06, + "loss": 0.74512094, + "num_input_tokens_seen": 84479110, + "step": 3931, + "time_per_iteration": 2.8006296157836914 + }, + { + "auxiliary_loss_clip": 0.01158192, + "auxiliary_loss_mlp": 0.01035475, + "balance_loss_clip": 1.05123615, + "balance_loss_mlp": 1.02658772, + "epoch": 0.47279504599290567, + "flos": 18770508069120.0, + "grad_norm": 1.9817182422005948, + "language_loss": 0.84379536, + "learning_rate": 2.272568166445699e-06, + "loss": 0.86573195, + "num_input_tokens_seen": 84497675, + "step": 3932, + "time_per_iteration": 2.8619441986083984 + }, + { + "auxiliary_loss_clip": 0.01175982, + "auxiliary_loss_mlp": 0.01029081, + "balance_loss_clip": 1.05158806, + "balance_loss_mlp": 1.01975822, + "epoch": 0.4729152888835448, + "flos": 21105742976640.0, + "grad_norm": 3.0084438798784423, + "language_loss": 0.64667338, + "learning_rate": 2.271796442106541e-06, + "loss": 0.668724, + "num_input_tokens_seen": 84517030, + "step": 3933, + "time_per_iteration": 2.814423084259033 + }, + { + "auxiliary_loss_clip": 0.0106021, + "auxiliary_loss_mlp": 0.01002376, + "balance_loss_clip": 1.01895571, + "balance_loss_mlp": 1.00064754, + "epoch": 0.47303553177418384, + "flos": 70201877840640.0, + "grad_norm": 0.7981665546982292, + "language_loss": 0.56469303, + "learning_rate": 2.271024676536079e-06, + "loss": 0.58531892, + "num_input_tokens_seen": 84577290, + "step": 3934, + "time_per_iteration": 3.634107828140259 + }, + { + "auxiliary_loss_clip": 0.01172262, + "auxiliary_loss_mlp": 0.01031797, + "balance_loss_clip": 1.05372322, + "balance_loss_mlp": 1.02276134, + "epoch": 0.47315577466482295, + "flos": 22455122227200.0, + "grad_norm": 1.8962441931541496, + "language_loss": 0.73019528, + "learning_rate": 2.2702528698513894e-06, + "loss": 0.75223589, + "num_input_tokens_seen": 84598415, + "step": 3935, + "time_per_iteration": 2.985069751739502 + }, + { + "auxiliary_loss_clip": 0.01165771, + "auxiliary_loss_mlp": 0.01033128, + "balance_loss_clip": 1.04712224, + "balance_loss_mlp": 1.02425313, + "epoch": 0.47327601755546206, + "flos": 24352857480960.0, + "grad_norm": 1.8556842789439245, + "language_loss": 0.78691095, + "learning_rate": 2.269481022169554e-06, + "loss": 0.80889994, + "num_input_tokens_seen": 84617010, + "step": 3936, + "time_per_iteration": 3.840916872024536 + }, + { + "auxiliary_loss_clip": 0.0117685, + "auxiliary_loss_mlp": 0.01033269, + "balance_loss_clip": 1.05256915, + "balance_loss_mlp": 1.02336836, + "epoch": 0.4733962604461011, + "flos": 22926772736640.0, + "grad_norm": 1.8861269236313039, + "language_loss": 0.80659628, + "learning_rate": 2.2687091336076614e-06, + "loss": 0.82869756, + "num_input_tokens_seen": 84636350, + "step": 3937, + "time_per_iteration": 3.857588291168213 + }, + { + "auxiliary_loss_clip": 0.01174971, + "auxiliary_loss_mlp": 0.01032993, + "balance_loss_clip": 1.05343604, + "balance_loss_mlp": 1.02346253, + "epoch": 0.4735165033367402, + "flos": 18327369980160.0, + "grad_norm": 2.3897375809436583, + "language_loss": 0.79989159, + "learning_rate": 2.267937204282807e-06, + "loss": 0.8219713, + "num_input_tokens_seen": 84653490, + "step": 3938, + "time_per_iteration": 3.6185269355773926 + }, + { + "auxiliary_loss_clip": 0.01184178, + "auxiliary_loss_mlp": 0.01033311, + "balance_loss_clip": 1.05475402, + "balance_loss_mlp": 1.02366662, + "epoch": 0.4736367462273793, + "flos": 23037018554880.0, + "grad_norm": 2.022589678283732, + "language_loss": 0.7863788, + "learning_rate": 2.2671652343120926e-06, + "loss": 0.80855364, + "num_input_tokens_seen": 84673965, + "step": 3939, + "time_per_iteration": 3.7112247943878174 + }, + { + "auxiliary_loss_clip": 0.01184107, + "auxiliary_loss_mlp": 0.01032802, + "balance_loss_clip": 1.05276966, + "balance_loss_mlp": 1.0234977, + "epoch": 0.4737569891180184, + "flos": 25374336451200.0, + "grad_norm": 2.147920533654836, + "language_loss": 0.8063311, + "learning_rate": 2.2663932238126236e-06, + "loss": 0.82850021, + "num_input_tokens_seen": 84692525, + "step": 3940, + "time_per_iteration": 2.7964463233947754 + }, + { + "auxiliary_loss_clip": 0.01173859, + "auxiliary_loss_mlp": 0.01024881, + "balance_loss_clip": 1.04951334, + "balance_loss_mlp": 1.01562417, + "epoch": 0.4738772320086575, + "flos": 25849326925440.0, + "grad_norm": 1.438476348002166, + "language_loss": 0.79972374, + "learning_rate": 2.265621172901515e-06, + "loss": 0.82171118, + "num_input_tokens_seen": 84715640, + "step": 3941, + "time_per_iteration": 2.909008026123047 + }, + { + "auxiliary_loss_clip": 0.01190919, + "auxiliary_loss_mlp": 0.01037831, + "balance_loss_clip": 1.05767202, + "balance_loss_mlp": 1.02884305, + "epoch": 0.47399747489929656, + "flos": 27564420499200.0, + "grad_norm": 3.292771515024799, + "language_loss": 0.71402538, + "learning_rate": 2.2648490816958854e-06, + "loss": 0.73631287, + "num_input_tokens_seen": 84736635, + "step": 3942, + "time_per_iteration": 2.79439377784729 + }, + { + "auxiliary_loss_clip": 0.01176389, + "auxiliary_loss_mlp": 0.01036784, + "balance_loss_clip": 1.04991865, + "balance_loss_mlp": 1.02741992, + "epoch": 0.47411771778993567, + "flos": 24863650836480.0, + "grad_norm": 2.203197098763463, + "language_loss": 0.7300477, + "learning_rate": 2.264076950312861e-06, + "loss": 0.75217944, + "num_input_tokens_seen": 84755445, + "step": 3943, + "time_per_iteration": 2.7819697856903076 + }, + { + "auxiliary_loss_clip": 0.01172997, + "auxiliary_loss_mlp": 0.01031118, + "balance_loss_clip": 1.05288649, + "balance_loss_mlp": 1.02171779, + "epoch": 0.4742379606805748, + "flos": 22748009725440.0, + "grad_norm": 1.8360815782923234, + "language_loss": 0.82487524, + "learning_rate": 2.2633047788695727e-06, + "loss": 0.84691638, + "num_input_tokens_seen": 84775750, + "step": 3944, + "time_per_iteration": 2.9887146949768066 + }, + { + "auxiliary_loss_clip": 0.01169697, + "auxiliary_loss_mlp": 0.01027022, + "balance_loss_clip": 1.05497384, + "balance_loss_mlp": 1.01870084, + "epoch": 0.47435820357121383, + "flos": 19681130689920.0, + "grad_norm": 1.8417142189715272, + "language_loss": 0.63838011, + "learning_rate": 2.262532567483159e-06, + "loss": 0.66034728, + "num_input_tokens_seen": 84794310, + "step": 3945, + "time_per_iteration": 2.8435091972351074 + }, + { + "auxiliary_loss_clip": 0.01191154, + "auxiliary_loss_mlp": 0.00886759, + "balance_loss_clip": 1.05781198, + "balance_loss_mlp": 1.00088274, + "epoch": 0.47447844646185294, + "flos": 25228718714880.0, + "grad_norm": 3.304886120174973, + "language_loss": 0.80404329, + "learning_rate": 2.2617603162707635e-06, + "loss": 0.82482243, + "num_input_tokens_seen": 84814720, + "step": 3946, + "time_per_iteration": 2.9248428344726562 + }, + { + "auxiliary_loss_clip": 0.01186522, + "auxiliary_loss_mlp": 0.01038815, + "balance_loss_clip": 1.05554378, + "balance_loss_mlp": 1.0300585, + "epoch": 0.47459868935249205, + "flos": 24570619683840.0, + "grad_norm": 1.6782573386761324, + "language_loss": 0.82569087, + "learning_rate": 2.2609880253495363e-06, + "loss": 0.84794426, + "num_input_tokens_seen": 84834355, + "step": 3947, + "time_per_iteration": 2.7176122665405273 + }, + { + "auxiliary_loss_clip": 0.01162893, + "auxiliary_loss_mlp": 0.01030329, + "balance_loss_clip": 1.04866481, + "balance_loss_mlp": 1.02095342, + "epoch": 0.4747189322431311, + "flos": 20558500295040.0, + "grad_norm": 1.9025615708159247, + "language_loss": 0.86342609, + "learning_rate": 2.260215694836633e-06, + "loss": 0.88535833, + "num_input_tokens_seen": 84853530, + "step": 3948, + "time_per_iteration": 2.848010301589966 + }, + { + "auxiliary_loss_clip": 0.01149908, + "auxiliary_loss_mlp": 0.00887478, + "balance_loss_clip": 1.04605246, + "balance_loss_mlp": 1.00078297, + "epoch": 0.4748391751337702, + "flos": 25995231970560.0, + "grad_norm": 2.5886405237491847, + "language_loss": 0.64732504, + "learning_rate": 2.2594433248492157e-06, + "loss": 0.66769892, + "num_input_tokens_seen": 84872505, + "step": 3949, + "time_per_iteration": 3.044700860977173 + }, + { + "auxiliary_loss_clip": 0.01182059, + "auxiliary_loss_mlp": 0.01035864, + "balance_loss_clip": 1.05290651, + "balance_loss_mlp": 1.02641654, + "epoch": 0.47495941802440933, + "flos": 22821052032000.0, + "grad_norm": 1.879028970542357, + "language_loss": 0.80004621, + "learning_rate": 2.2586709155044527e-06, + "loss": 0.82222533, + "num_input_tokens_seen": 84893105, + "step": 3950, + "time_per_iteration": 2.8305442333221436 + }, + { + "auxiliary_loss_clip": 0.01190269, + "auxiliary_loss_mlp": 0.01030284, + "balance_loss_clip": 1.05551779, + "balance_loss_mlp": 1.02100992, + "epoch": 0.4750796609150484, + "flos": 27891782075520.0, + "grad_norm": 1.62161546177805, + "language_loss": 0.76209497, + "learning_rate": 2.2578984669195167e-06, + "loss": 0.78430045, + "num_input_tokens_seen": 84914070, + "step": 3951, + "time_per_iteration": 2.7637579441070557 + }, + { + "auxiliary_loss_clip": 0.01173767, + "auxiliary_loss_mlp": 0.01032799, + "balance_loss_clip": 1.04890919, + "balance_loss_mlp": 1.02377462, + "epoch": 0.4751999038056875, + "flos": 35660085471360.0, + "grad_norm": 2.1326465320466763, + "language_loss": 0.67711073, + "learning_rate": 2.2571259792115887e-06, + "loss": 0.69917643, + "num_input_tokens_seen": 84935290, + "step": 3952, + "time_per_iteration": 2.9062139987945557 + }, + { + "auxiliary_loss_clip": 0.01171246, + "auxiliary_loss_mlp": 0.01023326, + "balance_loss_clip": 1.05076671, + "balance_loss_mlp": 1.01469505, + "epoch": 0.4753201466963266, + "flos": 22090880361600.0, + "grad_norm": 1.6694854516188795, + "language_loss": 0.78960973, + "learning_rate": 2.2563534524978544e-06, + "loss": 0.81155539, + "num_input_tokens_seen": 84952760, + "step": 3953, + "time_per_iteration": 2.8234503269195557 + }, + { + "auxiliary_loss_clip": 0.01153803, + "auxiliary_loss_mlp": 0.01032723, + "balance_loss_clip": 1.05196488, + "balance_loss_mlp": 1.02400851, + "epoch": 0.47544038958696566, + "flos": 30190854965760.0, + "grad_norm": 1.6675974759547987, + "language_loss": 0.70683575, + "learning_rate": 2.2555808868955052e-06, + "loss": 0.728701, + "num_input_tokens_seen": 84974890, + "step": 3954, + "time_per_iteration": 2.8721024990081787 + }, + { + "auxiliary_loss_clip": 0.01150046, + "auxiliary_loss_mlp": 0.01036565, + "balance_loss_clip": 1.04887652, + "balance_loss_mlp": 1.02684987, + "epoch": 0.47556063247760477, + "flos": 23472219738240.0, + "grad_norm": 2.2344921200190213, + "language_loss": 0.74033928, + "learning_rate": 2.254808282521738e-06, + "loss": 0.76220536, + "num_input_tokens_seen": 84993640, + "step": 3955, + "time_per_iteration": 2.8490099906921387 + }, + { + "auxiliary_loss_clip": 0.01162852, + "auxiliary_loss_mlp": 0.0088703, + "balance_loss_clip": 1.05032003, + "balance_loss_mlp": 1.00091732, + "epoch": 0.4756808753682438, + "flos": 25155209531520.0, + "grad_norm": 3.609132610769487, + "language_loss": 0.80956614, + "learning_rate": 2.2540356394937573e-06, + "loss": 0.83006489, + "num_input_tokens_seen": 85012340, + "step": 3956, + "time_per_iteration": 2.826805353164673 + }, + { + "auxiliary_loss_clip": 0.01164249, + "auxiliary_loss_mlp": 0.01032146, + "balance_loss_clip": 1.05039454, + "balance_loss_mlp": 1.02293777, + "epoch": 0.47580111825888294, + "flos": 15669729573120.0, + "grad_norm": 1.9682849905660904, + "language_loss": 0.83944958, + "learning_rate": 2.253262957928772e-06, + "loss": 0.86141354, + "num_input_tokens_seen": 85029225, + "step": 3957, + "time_per_iteration": 2.781696319580078 + }, + { + "auxiliary_loss_clip": 0.01162312, + "auxiliary_loss_mlp": 0.01033322, + "balance_loss_clip": 1.04847121, + "balance_loss_mlp": 1.0243814, + "epoch": 0.47592136114952205, + "flos": 17636556637440.0, + "grad_norm": 1.9177752358890319, + "language_loss": 0.71871829, + "learning_rate": 2.2524902379439976e-06, + "loss": 0.74067461, + "num_input_tokens_seen": 85047895, + "step": 3958, + "time_per_iteration": 2.753509521484375 + }, + { + "auxiliary_loss_clip": 0.01044318, + "auxiliary_loss_mlp": 0.01003806, + "balance_loss_clip": 1.0190897, + "balance_loss_mlp": 1.00158906, + "epoch": 0.4760416040401611, + "flos": 61417159292160.0, + "grad_norm": 0.7405320647663809, + "language_loss": 0.63692439, + "learning_rate": 2.251717479656655e-06, + "loss": 0.65740561, + "num_input_tokens_seen": 85112690, + "step": 3959, + "time_per_iteration": 3.4463515281677246 + }, + { + "auxiliary_loss_clip": 0.01192543, + "auxiliary_loss_mlp": 0.01035176, + "balance_loss_clip": 1.05605173, + "balance_loss_mlp": 1.02603257, + "epoch": 0.4761618469308002, + "flos": 18405871153920.0, + "grad_norm": 1.905940128501852, + "language_loss": 0.76038641, + "learning_rate": 2.2509446831839704e-06, + "loss": 0.78266358, + "num_input_tokens_seen": 85132130, + "step": 3960, + "time_per_iteration": 2.6971702575683594 + }, + { + "auxiliary_loss_clip": 0.01169643, + "auxiliary_loss_mlp": 0.01035491, + "balance_loss_clip": 1.04905677, + "balance_loss_mlp": 1.02683079, + "epoch": 0.4762820898214393, + "flos": 18040911016320.0, + "grad_norm": 2.185101324030392, + "language_loss": 0.82031369, + "learning_rate": 2.250171848643177e-06, + "loss": 0.84236509, + "num_input_tokens_seen": 85149420, + "step": 3961, + "time_per_iteration": 2.775639772415161 + }, + { + "auxiliary_loss_clip": 0.01165345, + "auxiliary_loss_mlp": 0.0103025, + "balance_loss_clip": 1.05360782, + "balance_loss_mlp": 1.02177382, + "epoch": 0.4764023327120784, + "flos": 19318253541120.0, + "grad_norm": 2.3572898380297773, + "language_loss": 0.86135817, + "learning_rate": 2.249398976151513e-06, + "loss": 0.88331407, + "num_input_tokens_seen": 85166970, + "step": 3962, + "time_per_iteration": 3.735322952270508 + }, + { + "auxiliary_loss_clip": 0.01187267, + "auxiliary_loss_mlp": 0.01034839, + "balance_loss_clip": 1.05576515, + "balance_loss_mlp": 1.02566552, + "epoch": 0.4765225756027175, + "flos": 22747255539840.0, + "grad_norm": 2.480810848457852, + "language_loss": 0.78684878, + "learning_rate": 2.248626065826223e-06, + "loss": 0.80906987, + "num_input_tokens_seen": 85185175, + "step": 3963, + "time_per_iteration": 3.6077778339385986 + }, + { + "auxiliary_loss_clip": 0.01075272, + "auxiliary_loss_mlp": 0.01008501, + "balance_loss_clip": 1.01422286, + "balance_loss_mlp": 1.00685585, + "epoch": 0.4766428184933566, + "flos": 65933392106880.0, + "grad_norm": 0.7637129219487346, + "language_loss": 0.62609887, + "learning_rate": 2.2478531177845564e-06, + "loss": 0.64693654, + "num_input_tokens_seen": 85246170, + "step": 3964, + "time_per_iteration": 3.21397066116333 + }, + { + "auxiliary_loss_clip": 0.0117277, + "auxiliary_loss_mlp": 0.01036628, + "balance_loss_clip": 1.05340695, + "balance_loss_mlp": 1.02777672, + "epoch": 0.47676306138399566, + "flos": 24136495908480.0, + "grad_norm": 2.031516974833207, + "language_loss": 0.84995651, + "learning_rate": 2.247080132143769e-06, + "loss": 0.87205052, + "num_input_tokens_seen": 85268525, + "step": 3965, + "time_per_iteration": 3.7983341217041016 + }, + { + "auxiliary_loss_clip": 0.01154247, + "auxiliary_loss_mlp": 0.01027311, + "balance_loss_clip": 1.04491889, + "balance_loss_mlp": 1.01773286, + "epoch": 0.47688330427463477, + "flos": 12604322995200.0, + "grad_norm": 3.0214496460439344, + "language_loss": 0.68649286, + "learning_rate": 2.246307109021121e-06, + "loss": 0.70830846, + "num_input_tokens_seen": 85285930, + "step": 3966, + "time_per_iteration": 2.8527021408081055 + }, + { + "auxiliary_loss_clip": 0.01162091, + "auxiliary_loss_mlp": 0.01032285, + "balance_loss_clip": 1.0474174, + "balance_loss_mlp": 1.02340353, + "epoch": 0.4770035471652739, + "flos": 21390585828480.0, + "grad_norm": 1.6694127074125262, + "language_loss": 0.8263551, + "learning_rate": 2.2455340485338817e-06, + "loss": 0.84829879, + "num_input_tokens_seen": 85303565, + "step": 3967, + "time_per_iteration": 2.7139089107513428 + }, + { + "auxiliary_loss_clip": 0.01177774, + "auxiliary_loss_mlp": 0.01033479, + "balance_loss_clip": 1.05163789, + "balance_loss_mlp": 1.02468157, + "epoch": 0.47712379005591293, + "flos": 25156251025920.0, + "grad_norm": 2.2496937714272787, + "language_loss": 0.67414308, + "learning_rate": 2.244760950799322e-06, + "loss": 0.69625556, + "num_input_tokens_seen": 85321835, + "step": 3968, + "time_per_iteration": 2.840972423553467 + }, + { + "auxiliary_loss_clip": 0.01144617, + "auxiliary_loss_mlp": 0.0103182, + "balance_loss_clip": 1.04812396, + "balance_loss_mlp": 1.02330232, + "epoch": 0.47724403294655204, + "flos": 22054323294720.0, + "grad_norm": 1.910746171203056, + "language_loss": 0.72486901, + "learning_rate": 2.2439878159347203e-06, + "loss": 0.74663341, + "num_input_tokens_seen": 85341260, + "step": 3969, + "time_per_iteration": 2.8145883083343506 + }, + { + "auxiliary_loss_clip": 0.01074551, + "auxiliary_loss_mlp": 0.01002955, + "balance_loss_clip": 1.01362944, + "balance_loss_mlp": 1.00132167, + "epoch": 0.4773642758371911, + "flos": 70229387658240.0, + "grad_norm": 0.7274434462277913, + "language_loss": 0.55252564, + "learning_rate": 2.2432146440573616e-06, + "loss": 0.57330072, + "num_input_tokens_seen": 85407220, + "step": 3970, + "time_per_iteration": 3.351283311843872 + }, + { + "auxiliary_loss_clip": 0.01160895, + "auxiliary_loss_mlp": 0.0102669, + "balance_loss_clip": 1.05073929, + "balance_loss_mlp": 1.01754689, + "epoch": 0.4774845187278302, + "flos": 23548602009600.0, + "grad_norm": 1.9460527497980489, + "language_loss": 0.66560453, + "learning_rate": 2.242441435284534e-06, + "loss": 0.68748039, + "num_input_tokens_seen": 85426095, + "step": 3971, + "time_per_iteration": 2.8904051780700684 + }, + { + "auxiliary_loss_clip": 0.01178958, + "auxiliary_loss_mlp": 0.01039669, + "balance_loss_clip": 1.05291414, + "balance_loss_mlp": 1.03057981, + "epoch": 0.4776047616184693, + "flos": 23075371301760.0, + "grad_norm": 5.310939919685394, + "language_loss": 0.85284674, + "learning_rate": 2.2416681897335337e-06, + "loss": 0.87503302, + "num_input_tokens_seen": 85444245, + "step": 3972, + "time_per_iteration": 2.8246896266937256 + }, + { + "auxiliary_loss_clip": 0.0114331, + "auxiliary_loss_mlp": 0.01029316, + "balance_loss_clip": 1.0472573, + "balance_loss_mlp": 1.01993418, + "epoch": 0.4777250045091084, + "flos": 31898119374720.0, + "grad_norm": 2.599691332246262, + "language_loss": 0.67146093, + "learning_rate": 2.240894907521661e-06, + "loss": 0.69318718, + "num_input_tokens_seen": 85463325, + "step": 3973, + "time_per_iteration": 3.0386407375335693 + }, + { + "auxiliary_loss_clip": 0.01166173, + "auxiliary_loss_mlp": 0.01034518, + "balance_loss_clip": 1.04969192, + "balance_loss_mlp": 1.02505279, + "epoch": 0.4778452473997475, + "flos": 24278163148800.0, + "grad_norm": 2.513967873188656, + "language_loss": 0.63819712, + "learning_rate": 2.240121588766223e-06, + "loss": 0.66020405, + "num_input_tokens_seen": 85483375, + "step": 3974, + "time_per_iteration": 2.733797788619995 + }, + { + "auxiliary_loss_clip": 0.01159772, + "auxiliary_loss_mlp": 0.01032007, + "balance_loss_clip": 1.04786062, + "balance_loss_mlp": 1.02346551, + "epoch": 0.4779654902903866, + "flos": 31575031516800.0, + "grad_norm": 2.8511867828471322, + "language_loss": 0.71360177, + "learning_rate": 2.239348233584531e-06, + "loss": 0.73551953, + "num_input_tokens_seen": 85504230, + "step": 3975, + "time_per_iteration": 2.842496633529663 + }, + { + "auxiliary_loss_clip": 0.0117846, + "auxiliary_loss_mlp": 0.01042547, + "balance_loss_clip": 1.05302167, + "balance_loss_mlp": 1.03413093, + "epoch": 0.47808573318102565, + "flos": 19500428344320.0, + "grad_norm": 1.9504696341273384, + "language_loss": 0.81080854, + "learning_rate": 2.2385748420939013e-06, + "loss": 0.83301866, + "num_input_tokens_seen": 85523425, + "step": 3976, + "time_per_iteration": 2.784543514251709 + }, + { + "auxiliary_loss_clip": 0.01186422, + "auxiliary_loss_mlp": 0.01034152, + "balance_loss_clip": 1.05648422, + "balance_loss_mlp": 1.02526522, + "epoch": 0.47820597607166476, + "flos": 22601135013120.0, + "grad_norm": 6.059278505163013, + "language_loss": 0.7233001, + "learning_rate": 2.2378014144116583e-06, + "loss": 0.74550581, + "num_input_tokens_seen": 85542235, + "step": 3977, + "time_per_iteration": 2.778083086013794 + }, + { + "auxiliary_loss_clip": 0.01189091, + "auxiliary_loss_mlp": 0.01033148, + "balance_loss_clip": 1.05419946, + "balance_loss_mlp": 1.02384341, + "epoch": 0.4783262189623039, + "flos": 23003011353600.0, + "grad_norm": 1.820833843908276, + "language_loss": 0.79470217, + "learning_rate": 2.23702795065513e-06, + "loss": 0.81692457, + "num_input_tokens_seen": 85561815, + "step": 3978, + "time_per_iteration": 2.809473752975464 + }, + { + "auxiliary_loss_clip": 0.01069088, + "auxiliary_loss_mlp": 0.01002063, + "balance_loss_clip": 1.01196456, + "balance_loss_mlp": 1.00026321, + "epoch": 0.47844646185294293, + "flos": 49772801226240.0, + "grad_norm": 0.9833530929449408, + "language_loss": 0.67385256, + "learning_rate": 2.2362544509416493e-06, + "loss": 0.6945641, + "num_input_tokens_seen": 85613930, + "step": 3979, + "time_per_iteration": 3.1515302658081055 + }, + { + "auxiliary_loss_clip": 0.01157991, + "auxiliary_loss_mlp": 0.01031075, + "balance_loss_clip": 1.04769588, + "balance_loss_mlp": 1.02251005, + "epoch": 0.47856670474358204, + "flos": 20229558520320.0, + "grad_norm": 2.5388443866968506, + "language_loss": 0.82367814, + "learning_rate": 2.2354809153885572e-06, + "loss": 0.84556878, + "num_input_tokens_seen": 85631000, + "step": 3980, + "time_per_iteration": 2.8117032051086426 + }, + { + "auxiliary_loss_clip": 0.01174996, + "auxiliary_loss_mlp": 0.01031852, + "balance_loss_clip": 1.05076265, + "balance_loss_mlp": 1.02285814, + "epoch": 0.47868694763422115, + "flos": 20990936131200.0, + "grad_norm": 2.9257605728101415, + "language_loss": 0.83096242, + "learning_rate": 2.234707344113197e-06, + "loss": 0.85303092, + "num_input_tokens_seen": 85649095, + "step": 3981, + "time_per_iteration": 2.731417655944824 + }, + { + "auxiliary_loss_clip": 0.0118582, + "auxiliary_loss_mlp": 0.01030146, + "balance_loss_clip": 1.05522811, + "balance_loss_mlp": 1.02192056, + "epoch": 0.4788071905248602, + "flos": 19026551191680.0, + "grad_norm": 2.6911668366947405, + "language_loss": 0.77468091, + "learning_rate": 2.233933737232919e-06, + "loss": 0.79684055, + "num_input_tokens_seen": 85666875, + "step": 3982, + "time_per_iteration": 2.6147003173828125 + }, + { + "auxiliary_loss_clip": 0.01140026, + "auxiliary_loss_mlp": 0.00886476, + "balance_loss_clip": 1.04510117, + "balance_loss_mlp": 1.00090158, + "epoch": 0.4789274334154993, + "flos": 23002221254400.0, + "grad_norm": 1.56487948093822, + "language_loss": 0.78282094, + "learning_rate": 2.2331600948650793e-06, + "loss": 0.80308592, + "num_input_tokens_seen": 85687020, + "step": 3983, + "time_per_iteration": 3.1201462745666504 + }, + { + "auxiliary_loss_clip": 0.01147345, + "auxiliary_loss_mlp": 0.00887849, + "balance_loss_clip": 1.05029798, + "balance_loss_mlp": 1.00083697, + "epoch": 0.4790476763061384, + "flos": 23075586783360.0, + "grad_norm": 1.473713284127022, + "language_loss": 0.80063534, + "learning_rate": 2.2323864171270386e-06, + "loss": 0.82098728, + "num_input_tokens_seen": 85708290, + "step": 3984, + "time_per_iteration": 2.8009910583496094 + }, + { + "auxiliary_loss_clip": 0.01158604, + "auxiliary_loss_mlp": 0.01043102, + "balance_loss_clip": 1.0475719, + "balance_loss_mlp": 1.03388166, + "epoch": 0.4791679191967775, + "flos": 21179288073600.0, + "grad_norm": 2.03151821220825, + "language_loss": 0.72600973, + "learning_rate": 2.231612704136164e-06, + "loss": 0.74802685, + "num_input_tokens_seen": 85728660, + "step": 3985, + "time_per_iteration": 2.873305320739746 + }, + { + "auxiliary_loss_clip": 0.01171488, + "auxiliary_loss_mlp": 0.01032694, + "balance_loss_clip": 1.04933071, + "balance_loss_mlp": 1.02330041, + "epoch": 0.4792881620874166, + "flos": 22301495758080.0, + "grad_norm": 7.902007670593239, + "language_loss": 0.75203037, + "learning_rate": 2.2308389560098253e-06, + "loss": 0.77407223, + "num_input_tokens_seen": 85745035, + "step": 3986, + "time_per_iteration": 2.670205593109131 + }, + { + "auxiliary_loss_clip": 0.01155884, + "auxiliary_loss_mlp": 0.01038084, + "balance_loss_clip": 1.05144739, + "balance_loss_mlp": 1.02909017, + "epoch": 0.47940840497805565, + "flos": 17420877423360.0, + "grad_norm": 1.8992857454414858, + "language_loss": 0.76908863, + "learning_rate": 2.2300651728654008e-06, + "loss": 0.79102832, + "num_input_tokens_seen": 85760295, + "step": 3987, + "time_per_iteration": 2.7310516834259033 + }, + { + "auxiliary_loss_clip": 0.01062785, + "auxiliary_loss_mlp": 0.0087698, + "balance_loss_clip": 1.01244521, + "balance_loss_mlp": 1.00126386, + "epoch": 0.47952864786869476, + "flos": 65358175708800.0, + "grad_norm": 0.7235723524640287, + "language_loss": 0.60199481, + "learning_rate": 2.229291354820272e-06, + "loss": 0.62139249, + "num_input_tokens_seen": 85821305, + "step": 3988, + "time_per_iteration": 4.2578957080841064 + }, + { + "auxiliary_loss_clip": 0.01174827, + "auxiliary_loss_mlp": 0.01034443, + "balance_loss_clip": 1.05034971, + "balance_loss_mlp": 1.02578211, + "epoch": 0.47964889075933387, + "flos": 16799802336000.0, + "grad_norm": 2.0806230222499122, + "language_loss": 0.75460649, + "learning_rate": 2.228517501991828e-06, + "loss": 0.77669919, + "num_input_tokens_seen": 85840105, + "step": 3989, + "time_per_iteration": 3.7195630073547363 + }, + { + "auxiliary_loss_clip": 0.01057513, + "auxiliary_loss_mlp": 0.01008794, + "balance_loss_clip": 1.01229548, + "balance_loss_mlp": 1.00718486, + "epoch": 0.4797691336499729, + "flos": 70079244808320.0, + "grad_norm": 0.8134257697918618, + "language_loss": 0.61097449, + "learning_rate": 2.22774361449746e-06, + "loss": 0.63163757, + "num_input_tokens_seen": 85896585, + "step": 3990, + "time_per_iteration": 4.371600866317749 + }, + { + "auxiliary_loss_clip": 0.01138844, + "auxiliary_loss_mlp": 0.01030826, + "balance_loss_clip": 1.0483799, + "balance_loss_mlp": 1.02201605, + "epoch": 0.47988937654061203, + "flos": 18953329317120.0, + "grad_norm": 2.5972023046839166, + "language_loss": 0.70780754, + "learning_rate": 2.2269696924545668e-06, + "loss": 0.72950429, + "num_input_tokens_seen": 85914415, + "step": 3991, + "time_per_iteration": 3.94411563873291 + }, + { + "auxiliary_loss_clip": 0.01161985, + "auxiliary_loss_mlp": 0.01027927, + "balance_loss_clip": 1.05315781, + "balance_loss_mlp": 1.01930833, + "epoch": 0.48000961943125114, + "flos": 14461981649280.0, + "grad_norm": 3.136417293916722, + "language_loss": 0.77395904, + "learning_rate": 2.2261957359805523e-06, + "loss": 0.79585814, + "num_input_tokens_seen": 85931650, + "step": 3992, + "time_per_iteration": 2.870213031768799 + }, + { + "auxiliary_loss_clip": 0.01188624, + "auxiliary_loss_mlp": 0.01032006, + "balance_loss_clip": 1.05600023, + "balance_loss_mlp": 1.02296424, + "epoch": 0.4801298623218902, + "flos": 27051149105280.0, + "grad_norm": 1.853523496183992, + "language_loss": 0.74100798, + "learning_rate": 2.225421745192823e-06, + "loss": 0.76321429, + "num_input_tokens_seen": 85951805, + "step": 3993, + "time_per_iteration": 2.8639330863952637 + }, + { + "auxiliary_loss_clip": 0.01175152, + "auxiliary_loss_mlp": 0.01034229, + "balance_loss_clip": 1.05291891, + "balance_loss_mlp": 1.02558625, + "epoch": 0.4802501052125293, + "flos": 26355236031360.0, + "grad_norm": 1.9117831116680635, + "language_loss": 0.78335559, + "learning_rate": 2.2246477202087955e-06, + "loss": 0.80544937, + "num_input_tokens_seen": 85972485, + "step": 3994, + "time_per_iteration": 2.77546763420105 + }, + { + "auxiliary_loss_clip": 0.01170457, + "auxiliary_loss_mlp": 0.01029734, + "balance_loss_clip": 1.05218315, + "balance_loss_mlp": 1.02127612, + "epoch": 0.4803703481031684, + "flos": 20993916960000.0, + "grad_norm": 1.9093060239272752, + "language_loss": 0.83038729, + "learning_rate": 2.223873661145887e-06, + "loss": 0.85238922, + "num_input_tokens_seen": 85992540, + "step": 3995, + "time_per_iteration": 2.732778310775757 + }, + { + "auxiliary_loss_clip": 0.01160872, + "auxiliary_loss_mlp": 0.00886289, + "balance_loss_clip": 1.05145466, + "balance_loss_mlp": 1.00087643, + "epoch": 0.4804905909938075, + "flos": 20703722981760.0, + "grad_norm": 1.951086087834632, + "language_loss": 0.71160662, + "learning_rate": 2.2230995681215226e-06, + "loss": 0.73207819, + "num_input_tokens_seen": 86012065, + "step": 3996, + "time_per_iteration": 2.797656774520874 + }, + { + "auxiliary_loss_clip": 0.01156003, + "auxiliary_loss_mlp": 0.01030223, + "balance_loss_clip": 1.05075336, + "balance_loss_mlp": 1.02093697, + "epoch": 0.4806108338844466, + "flos": 16654831044480.0, + "grad_norm": 2.0575682325783253, + "language_loss": 0.78048539, + "learning_rate": 2.2223254412531305e-06, + "loss": 0.80234766, + "num_input_tokens_seen": 86029435, + "step": 3997, + "time_per_iteration": 2.7727267742156982 + }, + { + "auxiliary_loss_clip": 0.01152221, + "auxiliary_loss_mlp": 0.01025106, + "balance_loss_clip": 1.04519248, + "balance_loss_mlp": 1.01668942, + "epoch": 0.4807310767750857, + "flos": 20011329440640.0, + "grad_norm": 1.8159336257879992, + "language_loss": 0.82199764, + "learning_rate": 2.221551280658146e-06, + "loss": 0.84377092, + "num_input_tokens_seen": 86048495, + "step": 3998, + "time_per_iteration": 2.729114294052124 + }, + { + "auxiliary_loss_clip": 0.01144751, + "auxiliary_loss_mlp": 0.01032454, + "balance_loss_clip": 1.04809856, + "balance_loss_mlp": 1.02391541, + "epoch": 0.48085131966572475, + "flos": 23185257984000.0, + "grad_norm": 1.852905411729814, + "language_loss": 0.74354875, + "learning_rate": 2.2207770864540085e-06, + "loss": 0.76532078, + "num_input_tokens_seen": 86067470, + "step": 3999, + "time_per_iteration": 2.8843255043029785 + }, + { + "auxiliary_loss_clip": 0.01160422, + "auxiliary_loss_mlp": 0.01031814, + "balance_loss_clip": 1.04999542, + "balance_loss_mlp": 1.02341008, + "epoch": 0.48097156255636386, + "flos": 20558643949440.0, + "grad_norm": 1.965753115912133, + "language_loss": 0.7298615, + "learning_rate": 2.220002858758162e-06, + "loss": 0.75178385, + "num_input_tokens_seen": 86085460, + "step": 4000, + "time_per_iteration": 2.747058629989624 + }, + { + "auxiliary_loss_clip": 0.01069988, + "auxiliary_loss_mlp": 0.01003441, + "balance_loss_clip": 1.01357484, + "balance_loss_mlp": 1.00154543, + "epoch": 0.481091805447003, + "flos": 70511608817280.0, + "grad_norm": 0.8863614169983313, + "language_loss": 0.60870427, + "learning_rate": 2.2192285976880573e-06, + "loss": 0.62943846, + "num_input_tokens_seen": 86149715, + "step": 4001, + "time_per_iteration": 3.2785842418670654 + }, + { + "auxiliary_loss_clip": 0.01157181, + "auxiliary_loss_mlp": 0.00886401, + "balance_loss_clip": 1.04718328, + "balance_loss_mlp": 1.00087857, + "epoch": 0.48121204833764203, + "flos": 36428214839040.0, + "grad_norm": 1.6759216170870845, + "language_loss": 0.80681181, + "learning_rate": 2.2184543033611485e-06, + "loss": 0.82724762, + "num_input_tokens_seen": 86170795, + "step": 4002, + "time_per_iteration": 2.978567123413086 + }, + { + "auxiliary_loss_clip": 0.01177805, + "auxiliary_loss_mlp": 0.01026484, + "balance_loss_clip": 1.05152881, + "balance_loss_mlp": 1.0178293, + "epoch": 0.48133229122828114, + "flos": 27490264871040.0, + "grad_norm": 2.306708645475334, + "language_loss": 0.82252753, + "learning_rate": 2.2176799758948957e-06, + "loss": 0.84457046, + "num_input_tokens_seen": 86190955, + "step": 4003, + "time_per_iteration": 2.7860445976257324 + }, + { + "auxiliary_loss_clip": 0.01162276, + "auxiliary_loss_mlp": 0.01040633, + "balance_loss_clip": 1.05111313, + "balance_loss_mlp": 1.03122783, + "epoch": 0.4814525341189202, + "flos": 43072802179200.0, + "grad_norm": 1.9231192525682765, + "language_loss": 0.73374522, + "learning_rate": 2.2169056154067635e-06, + "loss": 0.75577426, + "num_input_tokens_seen": 86214875, + "step": 4004, + "time_per_iteration": 2.9956486225128174 + }, + { + "auxiliary_loss_clip": 0.01176553, + "auxiliary_loss_mlp": 0.00886887, + "balance_loss_clip": 1.05240679, + "balance_loss_mlp": 1.00084782, + "epoch": 0.4815727770095593, + "flos": 24236901400320.0, + "grad_norm": 1.8934138935486529, + "language_loss": 0.82365948, + "learning_rate": 2.216131222014222e-06, + "loss": 0.84429389, + "num_input_tokens_seen": 86232950, + "step": 4005, + "time_per_iteration": 2.7505319118499756 + }, + { + "auxiliary_loss_clip": 0.01152212, + "auxiliary_loss_mlp": 0.01040158, + "balance_loss_clip": 1.04948282, + "balance_loss_mlp": 1.03069305, + "epoch": 0.4816930199001984, + "flos": 18113630100480.0, + "grad_norm": 2.0115046028093273, + "language_loss": 0.80164456, + "learning_rate": 2.2153567958347455e-06, + "loss": 0.82356828, + "num_input_tokens_seen": 86249160, + "step": 4006, + "time_per_iteration": 2.8665921688079834 + }, + { + "auxiliary_loss_clip": 0.01167124, + "auxiliary_loss_mlp": 0.01036427, + "balance_loss_clip": 1.0532248, + "balance_loss_mlp": 1.02736747, + "epoch": 0.48181326279083747, + "flos": 17274720983040.0, + "grad_norm": 2.029946226812303, + "language_loss": 0.7966156, + "learning_rate": 2.214582336985815e-06, + "loss": 0.81865108, + "num_input_tokens_seen": 86267060, + "step": 4007, + "time_per_iteration": 2.796322822570801 + }, + { + "auxiliary_loss_clip": 0.01158033, + "auxiliary_loss_mlp": 0.01037208, + "balance_loss_clip": 1.04829323, + "balance_loss_mlp": 1.02739096, + "epoch": 0.4819335056814766, + "flos": 14903252231040.0, + "grad_norm": 2.0962444552336157, + "language_loss": 0.6625725, + "learning_rate": 2.2138078455849142e-06, + "loss": 0.68452495, + "num_input_tokens_seen": 86285055, + "step": 4008, + "time_per_iteration": 2.7885525226593018 + }, + { + "auxiliary_loss_clip": 0.01176187, + "auxiliary_loss_mlp": 0.0103435, + "balance_loss_clip": 1.05299425, + "balance_loss_mlp": 1.02533793, + "epoch": 0.4820537485721157, + "flos": 19244888012160.0, + "grad_norm": 1.948586971719126, + "language_loss": 0.78360391, + "learning_rate": 2.2130333217495334e-06, + "loss": 0.80570924, + "num_input_tokens_seen": 86304225, + "step": 4009, + "time_per_iteration": 2.7193729877471924 + }, + { + "auxiliary_loss_clip": 0.01161953, + "auxiliary_loss_mlp": 0.01034109, + "balance_loss_clip": 1.04895139, + "balance_loss_mlp": 1.02497745, + "epoch": 0.48217399146275475, + "flos": 16033791870720.0, + "grad_norm": 2.7149756957300015, + "language_loss": 0.67507148, + "learning_rate": 2.2122587655971665e-06, + "loss": 0.69703209, + "num_input_tokens_seen": 86319170, + "step": 4010, + "time_per_iteration": 2.786285638809204 + }, + { + "auxiliary_loss_clip": 0.01166403, + "auxiliary_loss_mlp": 0.01031775, + "balance_loss_clip": 1.05128777, + "balance_loss_mlp": 1.02360296, + "epoch": 0.48229423435339386, + "flos": 24134197438080.0, + "grad_norm": 1.9432501898385333, + "language_loss": 0.64007652, + "learning_rate": 2.211484177245314e-06, + "loss": 0.66205823, + "num_input_tokens_seen": 86338760, + "step": 4011, + "time_per_iteration": 2.762874126434326 + }, + { + "auxiliary_loss_clip": 0.01188181, + "auxiliary_loss_mlp": 0.01029013, + "balance_loss_clip": 1.05467546, + "balance_loss_mlp": 1.0195713, + "epoch": 0.48241447724403297, + "flos": 23805435231360.0, + "grad_norm": 2.2079234389963167, + "language_loss": 0.72054642, + "learning_rate": 2.21070955681148e-06, + "loss": 0.74271834, + "num_input_tokens_seen": 86357865, + "step": 4012, + "time_per_iteration": 2.688971519470215 + }, + { + "auxiliary_loss_clip": 0.01146494, + "auxiliary_loss_mlp": 0.01033381, + "balance_loss_clip": 1.04692054, + "balance_loss_mlp": 1.02473271, + "epoch": 0.482534720134672, + "flos": 23110312256640.0, + "grad_norm": 2.8190312632390677, + "language_loss": 0.78101027, + "learning_rate": 2.209934904413174e-06, + "loss": 0.80280906, + "num_input_tokens_seen": 86379470, + "step": 4013, + "time_per_iteration": 2.8153109550476074 + }, + { + "auxiliary_loss_clip": 0.01133196, + "auxiliary_loss_mlp": 0.01031903, + "balance_loss_clip": 1.0385139, + "balance_loss_mlp": 1.0230279, + "epoch": 0.48265496302531113, + "flos": 20923819568640.0, + "grad_norm": 2.0981005909073414, + "language_loss": 0.71881115, + "learning_rate": 2.2091602201679095e-06, + "loss": 0.74046218, + "num_input_tokens_seen": 86399080, + "step": 4014, + "time_per_iteration": 3.8278214931488037 + }, + { + "auxiliary_loss_clip": 0.01161979, + "auxiliary_loss_mlp": 0.0102787, + "balance_loss_clip": 1.05020118, + "balance_loss_mlp": 1.01886344, + "epoch": 0.48277520591595025, + "flos": 15231152511360.0, + "grad_norm": 2.3573895552716744, + "language_loss": 0.83369792, + "learning_rate": 2.208385504193206e-06, + "loss": 0.85559642, + "num_input_tokens_seen": 86416580, + "step": 4015, + "time_per_iteration": 3.7735540866851807 + }, + { + "auxiliary_loss_clip": 0.01183665, + "auxiliary_loss_mlp": 0.01029752, + "balance_loss_clip": 1.05070305, + "balance_loss_mlp": 1.02087653, + "epoch": 0.4828954488065893, + "flos": 17858664385920.0, + "grad_norm": 4.635717646099356, + "language_loss": 0.81232738, + "learning_rate": 2.2076107566065873e-06, + "loss": 0.83446163, + "num_input_tokens_seen": 86434365, + "step": 4016, + "time_per_iteration": 3.5425093173980713 + }, + { + "auxiliary_loss_clip": 0.01177265, + "auxiliary_loss_mlp": 0.01027834, + "balance_loss_clip": 1.05521464, + "balance_loss_mlp": 1.01911426, + "epoch": 0.4830156916972284, + "flos": 32087405070720.0, + "grad_norm": 2.4431780183052902, + "language_loss": 0.75616276, + "learning_rate": 2.2068359775255816e-06, + "loss": 0.7782138, + "num_input_tokens_seen": 86452675, + "step": 4017, + "time_per_iteration": 3.647271156311035 + }, + { + "auxiliary_loss_clip": 0.01143302, + "auxiliary_loss_mlp": 0.01026509, + "balance_loss_clip": 1.04670572, + "balance_loss_mlp": 1.01809835, + "epoch": 0.48313593458786747, + "flos": 21871717528320.0, + "grad_norm": 4.761049854825634, + "language_loss": 0.78511703, + "learning_rate": 2.206061167067723e-06, + "loss": 0.80681515, + "num_input_tokens_seen": 86470785, + "step": 4018, + "time_per_iteration": 2.813225507736206 + }, + { + "auxiliary_loss_clip": 0.01152232, + "auxiliary_loss_mlp": 0.0103561, + "balance_loss_clip": 1.04532564, + "balance_loss_mlp": 1.02609694, + "epoch": 0.4832561774785066, + "flos": 22601206840320.0, + "grad_norm": 2.455419971397479, + "language_loss": 0.79628366, + "learning_rate": 2.205286325350549e-06, + "loss": 0.81816208, + "num_input_tokens_seen": 86489850, + "step": 4019, + "time_per_iteration": 2.8171796798706055 + }, + { + "auxiliary_loss_clip": 0.01146925, + "auxiliary_loss_mlp": 0.01031838, + "balance_loss_clip": 1.04804456, + "balance_loss_mlp": 1.02282572, + "epoch": 0.4833764203691457, + "flos": 13437342282240.0, + "grad_norm": 4.199475330143622, + "language_loss": 0.72532213, + "learning_rate": 2.204511452491603e-06, + "loss": 0.74710971, + "num_input_tokens_seen": 86506475, + "step": 4020, + "time_per_iteration": 2.7797939777374268 + }, + { + "auxiliary_loss_clip": 0.01182927, + "auxiliary_loss_mlp": 0.01029202, + "balance_loss_clip": 1.05469656, + "balance_loss_mlp": 1.02051783, + "epoch": 0.48349666325978474, + "flos": 44128036955520.0, + "grad_norm": 1.681968071773964, + "language_loss": 0.75066978, + "learning_rate": 2.2037365486084316e-06, + "loss": 0.77279103, + "num_input_tokens_seen": 86529715, + "step": 4021, + "time_per_iteration": 2.855086326599121 + }, + { + "auxiliary_loss_clip": 0.01158771, + "auxiliary_loss_mlp": 0.0103449, + "balance_loss_clip": 1.04642057, + "balance_loss_mlp": 1.02546585, + "epoch": 0.48361690615042385, + "flos": 26028377245440.0, + "grad_norm": 2.8150424457311396, + "language_loss": 0.77938682, + "learning_rate": 2.2029616138185886e-06, + "loss": 0.80131942, + "num_input_tokens_seen": 86548715, + "step": 4022, + "time_per_iteration": 2.905099868774414 + }, + { + "auxiliary_loss_clip": 0.01154747, + "auxiliary_loss_mlp": 0.01031082, + "balance_loss_clip": 1.05259335, + "balance_loss_mlp": 1.02211761, + "epoch": 0.48373714904106296, + "flos": 22273306560000.0, + "grad_norm": 1.5889059446627691, + "language_loss": 0.8275916, + "learning_rate": 2.202186648239629e-06, + "loss": 0.84944987, + "num_input_tokens_seen": 86568650, + "step": 4023, + "time_per_iteration": 2.8941972255706787 + }, + { + "auxiliary_loss_clip": 0.01174148, + "auxiliary_loss_mlp": 0.01028009, + "balance_loss_clip": 1.05195081, + "balance_loss_mlp": 1.01949108, + "epoch": 0.483857391931702, + "flos": 28292293699200.0, + "grad_norm": 1.8606324899997408, + "language_loss": 0.71845549, + "learning_rate": 2.201411651989117e-06, + "loss": 0.74047709, + "num_input_tokens_seen": 86590630, + "step": 4024, + "time_per_iteration": 2.7610347270965576 + }, + { + "auxiliary_loss_clip": 0.01166467, + "auxiliary_loss_mlp": 0.00886513, + "balance_loss_clip": 1.0527029, + "balance_loss_mlp": 1.0009501, + "epoch": 0.48397763482234113, + "flos": 27418048577280.0, + "grad_norm": 2.028596719082519, + "language_loss": 0.78100556, + "learning_rate": 2.2006366251846167e-06, + "loss": 0.80153537, + "num_input_tokens_seen": 86611270, + "step": 4025, + "time_per_iteration": 2.8971822261810303 + }, + { + "auxiliary_loss_clip": 0.01167344, + "auxiliary_loss_mlp": 0.01029008, + "balance_loss_clip": 1.05335069, + "balance_loss_mlp": 1.02105093, + "epoch": 0.48409787771298024, + "flos": 16797252470400.0, + "grad_norm": 2.0970536111213565, + "language_loss": 0.75677299, + "learning_rate": 2.1998615679436997e-06, + "loss": 0.77873647, + "num_input_tokens_seen": 86628810, + "step": 4026, + "time_per_iteration": 2.7913856506347656 + }, + { + "auxiliary_loss_clip": 0.01173405, + "auxiliary_loss_mlp": 0.01033465, + "balance_loss_clip": 1.04968953, + "balance_loss_mlp": 1.02481043, + "epoch": 0.4842181206036193, + "flos": 25083496028160.0, + "grad_norm": 2.134575301936082, + "language_loss": 0.76831836, + "learning_rate": 2.199086480383942e-06, + "loss": 0.79038703, + "num_input_tokens_seen": 86648185, + "step": 4027, + "time_per_iteration": 2.757251739501953 + }, + { + "auxiliary_loss_clip": 0.01179533, + "auxiliary_loss_mlp": 0.01040466, + "balance_loss_clip": 1.05172634, + "balance_loss_mlp": 1.03062487, + "epoch": 0.4843383634942584, + "flos": 30372311496960.0, + "grad_norm": 3.412935317388037, + "language_loss": 0.67429996, + "learning_rate": 2.1983113626229234e-06, + "loss": 0.69649994, + "num_input_tokens_seen": 86667435, + "step": 4028, + "time_per_iteration": 2.839872121810913 + }, + { + "auxiliary_loss_clip": 0.01147733, + "auxiliary_loss_mlp": 0.0088703, + "balance_loss_clip": 1.04430616, + "balance_loss_mlp": 1.00105274, + "epoch": 0.4844586063848975, + "flos": 20413564917120.0, + "grad_norm": 1.5688445518952867, + "language_loss": 0.7851752, + "learning_rate": 2.1975362147782293e-06, + "loss": 0.8055228, + "num_input_tokens_seen": 86686630, + "step": 4029, + "time_per_iteration": 2.7861344814300537 + }, + { + "auxiliary_loss_clip": 0.01057514, + "auxiliary_loss_mlp": 0.01006987, + "balance_loss_clip": 1.01423347, + "balance_loss_mlp": 1.00484109, + "epoch": 0.48457884927553657, + "flos": 70303722854400.0, + "grad_norm": 0.6919379610979686, + "language_loss": 0.54138899, + "learning_rate": 2.196761036967448e-06, + "loss": 0.56203401, + "num_input_tokens_seen": 86754595, + "step": 4030, + "time_per_iteration": 3.4851391315460205 + }, + { + "auxiliary_loss_clip": 0.01171811, + "auxiliary_loss_mlp": 0.01032318, + "balance_loss_clip": 1.05142546, + "balance_loss_mlp": 1.02387822, + "epoch": 0.4846990921661757, + "flos": 19934516206080.0, + "grad_norm": 1.657494279826689, + "language_loss": 0.77489382, + "learning_rate": 2.1959858293081743e-06, + "loss": 0.79693508, + "num_input_tokens_seen": 86773730, + "step": 4031, + "time_per_iteration": 2.754838228225708 + }, + { + "auxiliary_loss_clip": 0.01154768, + "auxiliary_loss_mlp": 0.01026797, + "balance_loss_clip": 1.05114031, + "balance_loss_mlp": 1.01830888, + "epoch": 0.4848193350568148, + "flos": 23075945919360.0, + "grad_norm": 2.2345765432744007, + "language_loss": 0.76198882, + "learning_rate": 2.1952105919180056e-06, + "loss": 0.78380442, + "num_input_tokens_seen": 86792985, + "step": 4032, + "time_per_iteration": 2.8942482471466064 + }, + { + "auxiliary_loss_clip": 0.01164689, + "auxiliary_loss_mlp": 0.01026079, + "balance_loss_clip": 1.05242622, + "balance_loss_mlp": 1.01690578, + "epoch": 0.48493957794745385, + "flos": 22455481363200.0, + "grad_norm": 2.220370729319379, + "language_loss": 0.67749298, + "learning_rate": 2.1944353249145456e-06, + "loss": 0.69940066, + "num_input_tokens_seen": 86812095, + "step": 4033, + "time_per_iteration": 2.848848342895508 + }, + { + "auxiliary_loss_clip": 0.01187361, + "auxiliary_loss_mlp": 0.01026127, + "balance_loss_clip": 1.05650413, + "balance_loss_mlp": 1.01759124, + "epoch": 0.48505982083809296, + "flos": 25046112948480.0, + "grad_norm": 1.639240010215323, + "language_loss": 0.74569017, + "learning_rate": 2.193660028415401e-06, + "loss": 0.76782501, + "num_input_tokens_seen": 86832875, + "step": 4034, + "time_per_iteration": 2.803210973739624 + }, + { + "auxiliary_loss_clip": 0.01157262, + "auxiliary_loss_mlp": 0.01030397, + "balance_loss_clip": 1.04823971, + "balance_loss_mlp": 1.02108645, + "epoch": 0.485180063728732, + "flos": 26761386090240.0, + "grad_norm": 1.8131510680292624, + "language_loss": 0.82047945, + "learning_rate": 2.1928847025381852e-06, + "loss": 0.84235603, + "num_input_tokens_seen": 86853480, + "step": 4035, + "time_per_iteration": 2.902674674987793 + }, + { + "auxiliary_loss_clip": 0.01173312, + "auxiliary_loss_mlp": 0.01029766, + "balance_loss_clip": 1.04682064, + "balance_loss_mlp": 1.02025294, + "epoch": 0.4853003066193711, + "flos": 24059143969920.0, + "grad_norm": 1.8177965070856694, + "language_loss": 0.84099376, + "learning_rate": 2.192109347400512e-06, + "loss": 0.86302459, + "num_input_tokens_seen": 86873695, + "step": 4036, + "time_per_iteration": 2.8133814334869385 + }, + { + "auxiliary_loss_clip": 0.01167984, + "auxiliary_loss_mlp": 0.01037239, + "balance_loss_clip": 1.04968715, + "balance_loss_mlp": 1.02759528, + "epoch": 0.48542054951001024, + "flos": 23076376882560.0, + "grad_norm": 1.6806890238597794, + "language_loss": 0.79054517, + "learning_rate": 2.191333963120004e-06, + "loss": 0.81259745, + "num_input_tokens_seen": 86892675, + "step": 4037, + "time_per_iteration": 2.9014053344726562 + }, + { + "auxiliary_loss_clip": 0.01163967, + "auxiliary_loss_mlp": 0.01034436, + "balance_loss_clip": 1.04906225, + "balance_loss_mlp": 1.02547181, + "epoch": 0.4855407924006493, + "flos": 25664889565440.0, + "grad_norm": 2.5932389271837915, + "language_loss": 0.70086491, + "learning_rate": 2.190558549814286e-06, + "loss": 0.72284901, + "num_input_tokens_seen": 86912835, + "step": 4038, + "time_per_iteration": 2.90752911567688 + }, + { + "auxiliary_loss_clip": 0.01162585, + "auxiliary_loss_mlp": 0.01042283, + "balance_loss_clip": 1.04799497, + "balance_loss_mlp": 1.03311014, + "epoch": 0.4856610352912884, + "flos": 23987933256960.0, + "grad_norm": 1.6303999180848703, + "language_loss": 0.79369128, + "learning_rate": 2.1897831076009872e-06, + "loss": 0.81573993, + "num_input_tokens_seen": 86932475, + "step": 4039, + "time_per_iteration": 2.8795318603515625 + }, + { + "auxiliary_loss_clip": 0.01174208, + "auxiliary_loss_mlp": 0.01037142, + "balance_loss_clip": 1.05016518, + "balance_loss_mlp": 1.02787983, + "epoch": 0.4857812781819275, + "flos": 24096814358400.0, + "grad_norm": 1.755768061563721, + "language_loss": 0.79790282, + "learning_rate": 2.1890076365977426e-06, + "loss": 0.82001638, + "num_input_tokens_seen": 86952300, + "step": 4040, + "time_per_iteration": 3.7083351612091064 + }, + { + "auxiliary_loss_clip": 0.01061689, + "auxiliary_loss_mlp": 0.01001433, + "balance_loss_clip": 1.01632059, + "balance_loss_mlp": 0.99975187, + "epoch": 0.48590152107256657, + "flos": 56266635185280.0, + "grad_norm": 0.8507216488951206, + "language_loss": 0.52764344, + "learning_rate": 2.188232136922189e-06, + "loss": 0.54827464, + "num_input_tokens_seen": 87010420, + "step": 4041, + "time_per_iteration": 4.954216241836548 + }, + { + "auxiliary_loss_clip": 0.01125084, + "auxiliary_loss_mlp": 0.01030491, + "balance_loss_clip": 1.04316115, + "balance_loss_mlp": 1.02093005, + "epoch": 0.4860217639632057, + "flos": 20046988667520.0, + "grad_norm": 2.299390645659022, + "language_loss": 0.76145256, + "learning_rate": 2.187456608691971e-06, + "loss": 0.78300822, + "num_input_tokens_seen": 87029295, + "step": 4042, + "time_per_iteration": 2.855698347091675 + }, + { + "auxiliary_loss_clip": 0.01157305, + "auxiliary_loss_mlp": 0.01029127, + "balance_loss_clip": 1.0492425, + "balance_loss_mlp": 1.02015018, + "epoch": 0.4861420068538448, + "flos": 17822143232640.0, + "grad_norm": 2.0101425421661965, + "language_loss": 0.87235224, + "learning_rate": 2.1866810520247334e-06, + "loss": 0.89421654, + "num_input_tokens_seen": 87048165, + "step": 4043, + "time_per_iteration": 3.7479848861694336 + }, + { + "auxiliary_loss_clip": 0.01177615, + "auxiliary_loss_mlp": 0.0103031, + "balance_loss_clip": 1.04980826, + "balance_loss_mlp": 1.02138102, + "epoch": 0.48626224974448384, + "flos": 26250125857920.0, + "grad_norm": 2.0249780468179046, + "language_loss": 0.64905328, + "learning_rate": 2.185905467038129e-06, + "loss": 0.67113256, + "num_input_tokens_seen": 87067070, + "step": 4044, + "time_per_iteration": 2.7663156986236572 + }, + { + "auxiliary_loss_clip": 0.0118381, + "auxiliary_loss_mlp": 0.01030167, + "balance_loss_clip": 1.05491722, + "balance_loss_mlp": 1.02159572, + "epoch": 0.48638249263512295, + "flos": 22054502862720.0, + "grad_norm": 1.7841627959634359, + "language_loss": 0.77636582, + "learning_rate": 2.1851298538498127e-06, + "loss": 0.7985056, + "num_input_tokens_seen": 87086785, + "step": 4045, + "time_per_iteration": 2.6696643829345703 + }, + { + "auxiliary_loss_clip": 0.01182279, + "auxiliary_loss_mlp": 0.00887305, + "balance_loss_clip": 1.05392635, + "balance_loss_mlp": 1.0009644, + "epoch": 0.48650273552576206, + "flos": 25119945354240.0, + "grad_norm": 1.9537240894477292, + "language_loss": 0.80059665, + "learning_rate": 2.184354212577446e-06, + "loss": 0.82129252, + "num_input_tokens_seen": 87107090, + "step": 4046, + "time_per_iteration": 2.840752363204956 + }, + { + "auxiliary_loss_clip": 0.01188703, + "auxiliary_loss_mlp": 0.01033478, + "balance_loss_clip": 1.05382991, + "balance_loss_mlp": 1.02406693, + "epoch": 0.4866229784164011, + "flos": 17456931699840.0, + "grad_norm": 3.208418661774963, + "language_loss": 0.62967062, + "learning_rate": 2.1835785433386907e-06, + "loss": 0.65189242, + "num_input_tokens_seen": 87125905, + "step": 4047, + "time_per_iteration": 2.6205697059631348 + }, + { + "auxiliary_loss_clip": 0.01147752, + "auxiliary_loss_mlp": 0.01042278, + "balance_loss_clip": 1.0489049, + "balance_loss_mlp": 1.03236032, + "epoch": 0.48674322130704023, + "flos": 23331127115520.0, + "grad_norm": 6.592821144819724, + "language_loss": 0.6556356, + "learning_rate": 2.182802846251216e-06, + "loss": 0.67753589, + "num_input_tokens_seen": 87146175, + "step": 4048, + "time_per_iteration": 2.801934242248535 + }, + { + "auxiliary_loss_clip": 0.0116145, + "auxiliary_loss_mlp": 0.01032541, + "balance_loss_clip": 1.04795742, + "balance_loss_mlp": 1.02398205, + "epoch": 0.4868634641976793, + "flos": 28804344030720.0, + "grad_norm": 1.9220251381489168, + "language_loss": 0.7200827, + "learning_rate": 2.182027121432696e-06, + "loss": 0.74202257, + "num_input_tokens_seen": 87166800, + "step": 4049, + "time_per_iteration": 2.7905473709106445 + }, + { + "auxiliary_loss_clip": 0.01188505, + "auxiliary_loss_mlp": 0.01037171, + "balance_loss_clip": 1.05286717, + "balance_loss_mlp": 1.0279088, + "epoch": 0.4869837070883184, + "flos": 19025976574080.0, + "grad_norm": 2.1564297967727075, + "language_loss": 0.81956542, + "learning_rate": 2.1812513690008054e-06, + "loss": 0.84182221, + "num_input_tokens_seen": 87185920, + "step": 4050, + "time_per_iteration": 2.6915838718414307 + }, + { + "auxiliary_loss_clip": 0.01183001, + "auxiliary_loss_mlp": 0.01032481, + "balance_loss_clip": 1.0517894, + "balance_loss_mlp": 1.02381456, + "epoch": 0.4871039499789575, + "flos": 15121409483520.0, + "grad_norm": 2.172262847996705, + "language_loss": 0.79895699, + "learning_rate": 2.180475589073227e-06, + "loss": 0.8211118, + "num_input_tokens_seen": 87203620, + "step": 4051, + "time_per_iteration": 2.7621991634368896 + }, + { + "auxiliary_loss_clip": 0.01163987, + "auxiliary_loss_mlp": 0.0102995, + "balance_loss_clip": 1.04720783, + "balance_loss_mlp": 1.02119398, + "epoch": 0.48722419286959656, + "flos": 26174066808960.0, + "grad_norm": 1.9127286556642804, + "language_loss": 0.73516548, + "learning_rate": 2.1796997817676456e-06, + "loss": 0.75710481, + "num_input_tokens_seen": 87224630, + "step": 4052, + "time_per_iteration": 2.7209346294403076 + }, + { + "auxiliary_loss_clip": 0.01177228, + "auxiliary_loss_mlp": 0.0088635, + "balance_loss_clip": 1.05151963, + "balance_loss_mlp": 1.00104201, + "epoch": 0.4873444357602357, + "flos": 24026142349440.0, + "grad_norm": 1.6123009578686756, + "language_loss": 0.67095965, + "learning_rate": 2.1789239472017494e-06, + "loss": 0.69159544, + "num_input_tokens_seen": 87246280, + "step": 4053, + "time_per_iteration": 2.7544233798980713 + }, + { + "auxiliary_loss_clip": 0.01152715, + "auxiliary_loss_mlp": 0.01032251, + "balance_loss_clip": 1.04682648, + "balance_loss_mlp": 1.02305984, + "epoch": 0.4874646786508748, + "flos": 22820441500800.0, + "grad_norm": 2.7833608244696753, + "language_loss": 0.72833943, + "learning_rate": 2.1781480854932326e-06, + "loss": 0.75018913, + "num_input_tokens_seen": 87266045, + "step": 4054, + "time_per_iteration": 2.8689870834350586 + }, + { + "auxiliary_loss_clip": 0.01143277, + "auxiliary_loss_mlp": 0.01031791, + "balance_loss_clip": 1.04804873, + "balance_loss_mlp": 1.02346432, + "epoch": 0.48758492154151384, + "flos": 21287594557440.0, + "grad_norm": 2.0063576687150833, + "language_loss": 0.79468989, + "learning_rate": 2.1773721967597933e-06, + "loss": 0.81644058, + "num_input_tokens_seen": 87284495, + "step": 4055, + "time_per_iteration": 2.8558294773101807 + }, + { + "auxiliary_loss_clip": 0.0105309, + "auxiliary_loss_mlp": 0.01004792, + "balance_loss_clip": 1.01184452, + "balance_loss_mlp": 1.003052, + "epoch": 0.48770516443215295, + "flos": 62244109180800.0, + "grad_norm": 0.8473541207066745, + "language_loss": 0.57397819, + "learning_rate": 2.1765962811191322e-06, + "loss": 0.59455693, + "num_input_tokens_seen": 87338960, + "step": 4056, + "time_per_iteration": 3.2098259925842285 + }, + { + "auxiliary_loss_clip": 0.01042248, + "auxiliary_loss_mlp": 0.01001981, + "balance_loss_clip": 1.01613569, + "balance_loss_mlp": 1.00009727, + "epoch": 0.48782540732279206, + "flos": 66133451882880.0, + "grad_norm": 0.8286670989831822, + "language_loss": 0.62044573, + "learning_rate": 2.1758203386889566e-06, + "loss": 0.64088798, + "num_input_tokens_seen": 87401730, + "step": 4057, + "time_per_iteration": 3.442580461502075 + }, + { + "auxiliary_loss_clip": 0.01156385, + "auxiliary_loss_mlp": 0.00887756, + "balance_loss_clip": 1.04694486, + "balance_loss_mlp": 1.00104141, + "epoch": 0.4879456502134311, + "flos": 14607922608000.0, + "grad_norm": 1.9628107476060428, + "language_loss": 0.84708792, + "learning_rate": 2.1750443695869746e-06, + "loss": 0.86752933, + "num_input_tokens_seen": 87417300, + "step": 4058, + "time_per_iteration": 2.8277931213378906 + }, + { + "auxiliary_loss_clip": 0.01176113, + "auxiliary_loss_mlp": 0.01032248, + "balance_loss_clip": 1.04962325, + "balance_loss_mlp": 1.02328372, + "epoch": 0.4880658931040702, + "flos": 19500464257920.0, + "grad_norm": 2.1752087657715147, + "language_loss": 0.8572858, + "learning_rate": 2.174268373930901e-06, + "loss": 0.87936938, + "num_input_tokens_seen": 87434815, + "step": 4059, + "time_per_iteration": 2.69077467918396 + }, + { + "auxiliary_loss_clip": 0.01151655, + "auxiliary_loss_mlp": 0.00887373, + "balance_loss_clip": 1.05186522, + "balance_loss_mlp": 1.00105667, + "epoch": 0.48818613599470934, + "flos": 16723060928640.0, + "grad_norm": 1.8427696428264424, + "language_loss": 0.79951066, + "learning_rate": 2.1734923518384537e-06, + "loss": 0.81990093, + "num_input_tokens_seen": 87451420, + "step": 4060, + "time_per_iteration": 2.805091619491577 + }, + { + "auxiliary_loss_clip": 0.0114144, + "auxiliary_loss_mlp": 0.01031468, + "balance_loss_clip": 1.04886436, + "balance_loss_mlp": 1.02333188, + "epoch": 0.4883063788853484, + "flos": 26756932803840.0, + "grad_norm": 1.7918329154053196, + "language_loss": 0.82635766, + "learning_rate": 2.1727163034273547e-06, + "loss": 0.84808671, + "num_input_tokens_seen": 87469585, + "step": 4061, + "time_per_iteration": 2.802351236343384 + }, + { + "auxiliary_loss_clip": 0.01176817, + "auxiliary_loss_mlp": 0.01028139, + "balance_loss_clip": 1.04964399, + "balance_loss_mlp": 1.01887083, + "epoch": 0.4884266217759875, + "flos": 16763388923520.0, + "grad_norm": 2.293885733087813, + "language_loss": 0.79154253, + "learning_rate": 2.17194022881533e-06, + "loss": 0.81359208, + "num_input_tokens_seen": 87485675, + "step": 4062, + "time_per_iteration": 2.7216620445251465 + }, + { + "auxiliary_loss_clip": 0.01172994, + "auxiliary_loss_mlp": 0.01035849, + "balance_loss_clip": 1.0512563, + "balance_loss_mlp": 1.02630639, + "epoch": 0.4885468646666266, + "flos": 24207132003840.0, + "grad_norm": 1.681043333372181, + "language_loss": 0.6741339, + "learning_rate": 2.1711641281201092e-06, + "loss": 0.69622231, + "num_input_tokens_seen": 87505605, + "step": 4063, + "time_per_iteration": 2.748762607574463 + }, + { + "auxiliary_loss_clip": 0.01176583, + "auxiliary_loss_mlp": 0.01030565, + "balance_loss_clip": 1.0538094, + "balance_loss_mlp": 1.02173162, + "epoch": 0.48866710755726567, + "flos": 14610795696000.0, + "grad_norm": 2.462014474224222, + "language_loss": 0.79380739, + "learning_rate": 2.1703880014594264e-06, + "loss": 0.81587887, + "num_input_tokens_seen": 87523195, + "step": 4064, + "time_per_iteration": 2.7233197689056396 + }, + { + "auxiliary_loss_clip": 0.01141374, + "auxiliary_loss_mlp": 0.01029633, + "balance_loss_clip": 1.04881954, + "balance_loss_mlp": 1.02135408, + "epoch": 0.4887873504479048, + "flos": 28804451771520.0, + "grad_norm": 2.2611557461013336, + "language_loss": 0.73433042, + "learning_rate": 2.1696118489510182e-06, + "loss": 0.75604045, + "num_input_tokens_seen": 87544125, + "step": 4065, + "time_per_iteration": 2.87996244430542 + }, + { + "auxiliary_loss_clip": 0.01164334, + "auxiliary_loss_mlp": 0.00888219, + "balance_loss_clip": 1.05069947, + "balance_loss_mlp": 1.00102949, + "epoch": 0.48890759333854383, + "flos": 22784387224320.0, + "grad_norm": 1.8224257952035294, + "language_loss": 0.72423482, + "learning_rate": 2.1688356707126286e-06, + "loss": 0.74476045, + "num_input_tokens_seen": 87563745, + "step": 4066, + "time_per_iteration": 3.7852885723114014 + }, + { + "auxiliary_loss_clip": 0.01151145, + "auxiliary_loss_mlp": 0.01032114, + "balance_loss_clip": 1.04802823, + "balance_loss_mlp": 1.02254748, + "epoch": 0.48902783622918294, + "flos": 17786088956160.0, + "grad_norm": 2.3272104802643665, + "language_loss": 0.69512987, + "learning_rate": 2.168059466862001e-06, + "loss": 0.71696252, + "num_input_tokens_seen": 87581895, + "step": 4067, + "time_per_iteration": 4.86472487449646 + }, + { + "auxiliary_loss_clip": 0.01165379, + "auxiliary_loss_mlp": 0.01030193, + "balance_loss_clip": 1.04567075, + "balance_loss_mlp": 1.02140725, + "epoch": 0.48914807911982205, + "flos": 22310294590080.0, + "grad_norm": 1.8982152331519342, + "language_loss": 0.81496936, + "learning_rate": 2.167283237516887e-06, + "loss": 0.83692509, + "num_input_tokens_seen": 87600170, + "step": 4068, + "time_per_iteration": 2.8725032806396484 + }, + { + "auxiliary_loss_clip": 0.0116821, + "auxiliary_loss_mlp": 0.01033488, + "balance_loss_clip": 1.04898119, + "balance_loss_mlp": 1.02435064, + "epoch": 0.4892683220104611, + "flos": 16363020954240.0, + "grad_norm": 1.79988159989597, + "language_loss": 0.74743873, + "learning_rate": 2.1665069827950383e-06, + "loss": 0.76945567, + "num_input_tokens_seen": 87617455, + "step": 4069, + "time_per_iteration": 3.6509366035461426 + }, + { + "auxiliary_loss_clip": 0.01163309, + "auxiliary_loss_mlp": 0.01025608, + "balance_loss_clip": 1.0494082, + "balance_loss_mlp": 1.01694715, + "epoch": 0.4893885649011002, + "flos": 15739144606080.0, + "grad_norm": 1.8349445191520548, + "language_loss": 0.86596107, + "learning_rate": 2.1657307028142126e-06, + "loss": 0.88785028, + "num_input_tokens_seen": 87634995, + "step": 4070, + "time_per_iteration": 2.7657034397125244 + }, + { + "auxiliary_loss_clip": 0.01167437, + "auxiliary_loss_mlp": 0.01028765, + "balance_loss_clip": 1.05097735, + "balance_loss_mlp": 1.0194366, + "epoch": 0.48950880779173933, + "flos": 28581984887040.0, + "grad_norm": 2.3827728157632335, + "language_loss": 0.6762355, + "learning_rate": 2.164954397692171e-06, + "loss": 0.69819748, + "num_input_tokens_seen": 87654420, + "step": 4071, + "time_per_iteration": 2.860412359237671 + }, + { + "auxiliary_loss_clip": 0.01065728, + "auxiliary_loss_mlp": 0.01007924, + "balance_loss_clip": 1.01480055, + "balance_loss_mlp": 1.00595689, + "epoch": 0.4896290506823784, + "flos": 66186310746240.0, + "grad_norm": 1.1088852817935944, + "language_loss": 0.77301967, + "learning_rate": 2.164178067546678e-06, + "loss": 0.79375619, + "num_input_tokens_seen": 87713585, + "step": 4072, + "time_per_iteration": 3.36031436920166 + }, + { + "auxiliary_loss_clip": 0.0117393, + "auxiliary_loss_mlp": 0.01029756, + "balance_loss_clip": 1.04934382, + "balance_loss_mlp": 1.02109563, + "epoch": 0.4897492935730175, + "flos": 12531065207040.0, + "grad_norm": 1.8158938916823402, + "language_loss": 0.90526557, + "learning_rate": 2.163401712495504e-06, + "loss": 0.92730242, + "num_input_tokens_seen": 87731280, + "step": 4073, + "time_per_iteration": 2.7899067401885986 + }, + { + "auxiliary_loss_clip": 0.01144239, + "auxiliary_loss_mlp": 0.0103282, + "balance_loss_clip": 1.04815435, + "balance_loss_mlp": 1.02305698, + "epoch": 0.4898695364636566, + "flos": 23476816679040.0, + "grad_norm": 1.6492688226210213, + "language_loss": 0.7913413, + "learning_rate": 2.1626253326564194e-06, + "loss": 0.8131119, + "num_input_tokens_seen": 87750230, + "step": 4074, + "time_per_iteration": 3.0537946224212646 + }, + { + "auxiliary_loss_clip": 0.01165016, + "auxiliary_loss_mlp": 0.01034515, + "balance_loss_clip": 1.04800832, + "balance_loss_mlp": 1.0247401, + "epoch": 0.48998977935429566, + "flos": 27160209774720.0, + "grad_norm": 1.5892899845780635, + "language_loss": 0.77009141, + "learning_rate": 2.161848928147201e-06, + "loss": 0.79208672, + "num_input_tokens_seen": 87770500, + "step": 4075, + "time_per_iteration": 2.8480064868927 + }, + { + "auxiliary_loss_clip": 0.01178058, + "auxiliary_loss_mlp": 0.01035313, + "balance_loss_clip": 1.05335224, + "balance_loss_mlp": 1.02606893, + "epoch": 0.4901100222449348, + "flos": 20339588856960.0, + "grad_norm": 2.0467447977240747, + "language_loss": 0.80458391, + "learning_rate": 2.161072499085629e-06, + "loss": 0.82671762, + "num_input_tokens_seen": 87789495, + "step": 4076, + "time_per_iteration": 2.788243055343628 + }, + { + "auxiliary_loss_clip": 0.0116103, + "auxiliary_loss_mlp": 0.01030746, + "balance_loss_clip": 1.05035877, + "balance_loss_mlp": 1.02217484, + "epoch": 0.4902302651355739, + "flos": 30446359384320.0, + "grad_norm": 1.9054895595989987, + "language_loss": 0.8319959, + "learning_rate": 2.160296045589487e-06, + "loss": 0.85391366, + "num_input_tokens_seen": 87812955, + "step": 4077, + "time_per_iteration": 2.9210598468780518 + }, + { + "auxiliary_loss_clip": 0.01176284, + "auxiliary_loss_mlp": 0.01031607, + "balance_loss_clip": 1.05235577, + "balance_loss_mlp": 1.02236795, + "epoch": 0.49035050802621294, + "flos": 19174180089600.0, + "grad_norm": 1.7166148898073639, + "language_loss": 0.69945633, + "learning_rate": 2.159519567776562e-06, + "loss": 0.72153527, + "num_input_tokens_seen": 87832605, + "step": 4078, + "time_per_iteration": 2.690678119659424 + }, + { + "auxiliary_loss_clip": 0.0114509, + "auxiliary_loss_mlp": 0.01036161, + "balance_loss_clip": 1.04357421, + "balance_loss_mlp": 1.02756047, + "epoch": 0.49047075091685205, + "flos": 22228489365120.0, + "grad_norm": 2.6094244021150903, + "language_loss": 0.70619881, + "learning_rate": 2.1587430657646463e-06, + "loss": 0.72801137, + "num_input_tokens_seen": 87846040, + "step": 4079, + "time_per_iteration": 2.7791495323181152 + }, + { + "auxiliary_loss_clip": 0.01164205, + "auxiliary_loss_mlp": 0.01029318, + "balance_loss_clip": 1.05078244, + "balance_loss_mlp": 1.02034736, + "epoch": 0.4905909938074911, + "flos": 20156516213760.0, + "grad_norm": 2.181739717485846, + "language_loss": 0.77914989, + "learning_rate": 2.157966539671533e-06, + "loss": 0.80108517, + "num_input_tokens_seen": 87865680, + "step": 4080, + "time_per_iteration": 2.7745091915130615 + }, + { + "auxiliary_loss_clip": 0.01160307, + "auxiliary_loss_mlp": 0.01035669, + "balance_loss_clip": 1.0480262, + "balance_loss_mlp": 1.02746165, + "epoch": 0.4907112366981302, + "flos": 17202217380480.0, + "grad_norm": 1.9320516564318397, + "language_loss": 0.67293203, + "learning_rate": 2.157189989615021e-06, + "loss": 0.69489181, + "num_input_tokens_seen": 87884270, + "step": 4081, + "time_per_iteration": 2.7879035472869873 + }, + { + "auxiliary_loss_clip": 0.01175309, + "auxiliary_loss_mlp": 0.00888255, + "balance_loss_clip": 1.04852152, + "balance_loss_mlp": 1.00084448, + "epoch": 0.4908314795887693, + "flos": 21688968107520.0, + "grad_norm": 3.112926634662026, + "language_loss": 0.75119978, + "learning_rate": 2.156413415712913e-06, + "loss": 0.77183545, + "num_input_tokens_seen": 87906320, + "step": 4082, + "time_per_iteration": 2.8113725185394287 + }, + { + "auxiliary_loss_clip": 0.01172064, + "auxiliary_loss_mlp": 0.0088781, + "balance_loss_clip": 1.05083346, + "balance_loss_mlp": 1.00090063, + "epoch": 0.4909517224794084, + "flos": 26213676531840.0, + "grad_norm": 1.767074407350905, + "language_loss": 0.78666186, + "learning_rate": 2.155636818083014e-06, + "loss": 0.80726057, + "num_input_tokens_seen": 87927690, + "step": 4083, + "time_per_iteration": 2.915217638015747 + }, + { + "auxiliary_loss_clip": 0.01163797, + "auxiliary_loss_mlp": 0.01034043, + "balance_loss_clip": 1.05108154, + "balance_loss_mlp": 1.02584696, + "epoch": 0.4910719653700475, + "flos": 23148377694720.0, + "grad_norm": 2.1751809895195544, + "language_loss": 0.84524846, + "learning_rate": 2.154860196843134e-06, + "loss": 0.86722684, + "num_input_tokens_seen": 87946885, + "step": 4084, + "time_per_iteration": 2.8124349117279053 + }, + { + "auxiliary_loss_clip": 0.01186234, + "auxiliary_loss_mlp": 0.01033272, + "balance_loss_clip": 1.05217576, + "balance_loss_mlp": 1.02425396, + "epoch": 0.4911922082606866, + "flos": 23331845387520.0, + "grad_norm": 1.699393736142223, + "language_loss": 0.76605636, + "learning_rate": 2.154083552111085e-06, + "loss": 0.7882514, + "num_input_tokens_seen": 87966055, + "step": 4085, + "time_per_iteration": 2.7009382247924805 + }, + { + "auxiliary_loss_clip": 0.01185495, + "auxiliary_loss_mlp": 0.01033649, + "balance_loss_clip": 1.05031323, + "balance_loss_mlp": 1.02445245, + "epoch": 0.49131245115132566, + "flos": 29203239542400.0, + "grad_norm": 1.955354948077384, + "language_loss": 0.81521201, + "learning_rate": 2.1533068840046834e-06, + "loss": 0.83740348, + "num_input_tokens_seen": 87986320, + "step": 4086, + "time_per_iteration": 2.6877357959747314 + }, + { + "auxiliary_loss_clip": 0.01162802, + "auxiliary_loss_mlp": 0.00887701, + "balance_loss_clip": 1.04972541, + "balance_loss_mlp": 1.00091338, + "epoch": 0.49143269404196477, + "flos": 20147465986560.0, + "grad_norm": 2.9330858153796897, + "language_loss": 0.62347966, + "learning_rate": 2.152530192641749e-06, + "loss": 0.64398468, + "num_input_tokens_seen": 88001230, + "step": 4087, + "time_per_iteration": 2.7438342571258545 + }, + { + "auxiliary_loss_clip": 0.01176545, + "auxiliary_loss_mlp": 0.01034307, + "balance_loss_clip": 1.05274296, + "balance_loss_mlp": 1.02485955, + "epoch": 0.4915529369326039, + "flos": 24389809597440.0, + "grad_norm": 1.8630377421878273, + "language_loss": 0.7254824, + "learning_rate": 2.1517534781401068e-06, + "loss": 0.7475909, + "num_input_tokens_seen": 88019110, + "step": 4088, + "time_per_iteration": 2.816204071044922 + }, + { + "auxiliary_loss_clip": 0.01173818, + "auxiliary_loss_mlp": 0.01033565, + "balance_loss_clip": 1.05014765, + "balance_loss_mlp": 1.02451062, + "epoch": 0.49167317982324293, + "flos": 10524305197440.0, + "grad_norm": 2.2238818360875503, + "language_loss": 0.69390571, + "learning_rate": 2.150976740617581e-06, + "loss": 0.71597952, + "num_input_tokens_seen": 88035670, + "step": 4089, + "time_per_iteration": 2.8257365226745605 + }, + { + "auxiliary_loss_clip": 0.01169624, + "auxiliary_loss_mlp": 0.01026632, + "balance_loss_clip": 1.04988623, + "balance_loss_mlp": 1.01798892, + "epoch": 0.49179342271388204, + "flos": 25593427457280.0, + "grad_norm": 1.7855509464273378, + "language_loss": 0.71221769, + "learning_rate": 2.150199980192006e-06, + "loss": 0.73418021, + "num_input_tokens_seen": 88054790, + "step": 4090, + "time_per_iteration": 2.8372414112091064 + }, + { + "auxiliary_loss_clip": 0.0116181, + "auxiliary_loss_mlp": 0.01035304, + "balance_loss_clip": 1.04804194, + "balance_loss_mlp": 1.02667892, + "epoch": 0.49191366560452116, + "flos": 21102043875840.0, + "grad_norm": 1.7926267995003664, + "language_loss": 0.81129813, + "learning_rate": 2.1494231969812114e-06, + "loss": 0.83326936, + "num_input_tokens_seen": 88073780, + "step": 4091, + "time_per_iteration": 2.7416441440582275 + }, + { + "auxiliary_loss_clip": 0.01162968, + "auxiliary_loss_mlp": 0.01037092, + "balance_loss_clip": 1.05157912, + "balance_loss_mlp": 1.02856243, + "epoch": 0.4920339084951602, + "flos": 26067520091520.0, + "grad_norm": 2.52810953490227, + "language_loss": 0.80995339, + "learning_rate": 2.1486463911030372e-06, + "loss": 0.83195394, + "num_input_tokens_seen": 88094430, + "step": 4092, + "time_per_iteration": 3.791412591934204 + }, + { + "auxiliary_loss_clip": 0.01163662, + "auxiliary_loss_mlp": 0.01031103, + "balance_loss_clip": 1.04712772, + "balance_loss_mlp": 1.02258003, + "epoch": 0.4921541513857993, + "flos": 25081269384960.0, + "grad_norm": 2.1998201058946245, + "language_loss": 0.74482131, + "learning_rate": 2.147869562675324e-06, + "loss": 0.76676899, + "num_input_tokens_seen": 88113400, + "step": 4093, + "time_per_iteration": 4.693979263305664 + }, + { + "auxiliary_loss_clip": 0.01177237, + "auxiliary_loss_mlp": 0.0103142, + "balance_loss_clip": 1.05329573, + "balance_loss_mlp": 1.02243757, + "epoch": 0.49227439427643843, + "flos": 24389809597440.0, + "grad_norm": 2.727150781994266, + "language_loss": 0.72579253, + "learning_rate": 2.147092711815915e-06, + "loss": 0.74787909, + "num_input_tokens_seen": 88132750, + "step": 4094, + "time_per_iteration": 2.6791255474090576 + }, + { + "auxiliary_loss_clip": 0.01151937, + "auxiliary_loss_mlp": 0.0103228, + "balance_loss_clip": 1.04825091, + "balance_loss_mlp": 1.02348244, + "epoch": 0.4923946371670775, + "flos": 11363753018880.0, + "grad_norm": 2.497387360694899, + "language_loss": 0.86564863, + "learning_rate": 2.1463158386426593e-06, + "loss": 0.88749081, + "num_input_tokens_seen": 88150560, + "step": 4095, + "time_per_iteration": 3.660087823867798 + }, + { + "auxiliary_loss_clip": 0.01174629, + "auxiliary_loss_mlp": 0.01029068, + "balance_loss_clip": 1.05312061, + "balance_loss_mlp": 1.02038383, + "epoch": 0.4925148800577166, + "flos": 30445964334720.0, + "grad_norm": 14.502426162131203, + "language_loss": 0.77384198, + "learning_rate": 2.145538943273407e-06, + "loss": 0.79587889, + "num_input_tokens_seen": 88170835, + "step": 4096, + "time_per_iteration": 2.8776662349700928 + }, + { + "auxiliary_loss_clip": 0.01188051, + "auxiliary_loss_mlp": 0.01031111, + "balance_loss_clip": 1.05483854, + "balance_loss_mlp": 1.02274871, + "epoch": 0.49263512294835565, + "flos": 20850454039680.0, + "grad_norm": 1.7282025399515906, + "language_loss": 0.71627599, + "learning_rate": 2.144762025826013e-06, + "loss": 0.73846763, + "num_input_tokens_seen": 88189925, + "step": 4097, + "time_per_iteration": 2.7778193950653076 + }, + { + "auxiliary_loss_clip": 0.01178766, + "auxiliary_loss_mlp": 0.0102686, + "balance_loss_clip": 1.05102324, + "balance_loss_mlp": 1.01781249, + "epoch": 0.49275536583899476, + "flos": 23767477534080.0, + "grad_norm": 2.2247062180104913, + "language_loss": 0.86723965, + "learning_rate": 2.143985086418334e-06, + "loss": 0.88929594, + "num_input_tokens_seen": 88205105, + "step": 4098, + "time_per_iteration": 2.6987881660461426 + }, + { + "auxiliary_loss_clip": 0.01167826, + "auxiliary_loss_mlp": 0.01029495, + "balance_loss_clip": 1.04986525, + "balance_loss_mlp": 1.0210253, + "epoch": 0.4928756087296339, + "flos": 22273522041600.0, + "grad_norm": 1.3647892634146308, + "language_loss": 0.76846766, + "learning_rate": 2.1432081251682324e-06, + "loss": 0.79044092, + "num_input_tokens_seen": 88225475, + "step": 4099, + "time_per_iteration": 2.75787615776062 + }, + { + "auxiliary_loss_clip": 0.0117589, + "auxiliary_loss_mlp": 0.01035722, + "balance_loss_clip": 1.05636823, + "balance_loss_mlp": 1.02690041, + "epoch": 0.49299585162027293, + "flos": 19645471463040.0, + "grad_norm": 1.8414919161678405, + "language_loss": 0.87109292, + "learning_rate": 2.142431142193572e-06, + "loss": 0.89320904, + "num_input_tokens_seen": 88243255, + "step": 4100, + "time_per_iteration": 2.7944350242614746 + }, + { + "auxiliary_loss_clip": 0.01186249, + "auxiliary_loss_mlp": 0.01031791, + "balance_loss_clip": 1.05520511, + "balance_loss_mlp": 1.02298188, + "epoch": 0.49311609451091204, + "flos": 38837138497920.0, + "grad_norm": 2.4670125374605516, + "language_loss": 0.71690714, + "learning_rate": 2.1416541376122207e-06, + "loss": 0.73908758, + "num_input_tokens_seen": 88263435, + "step": 4101, + "time_per_iteration": 2.7842581272125244 + }, + { + "auxiliary_loss_clip": 0.01183391, + "auxiliary_loss_mlp": 0.0103535, + "balance_loss_clip": 1.05010808, + "balance_loss_mlp": 1.02580166, + "epoch": 0.49323633740155115, + "flos": 28329102161280.0, + "grad_norm": 1.8735540658489531, + "language_loss": 0.73310554, + "learning_rate": 2.1408771115420496e-06, + "loss": 0.75529289, + "num_input_tokens_seen": 88283295, + "step": 4102, + "time_per_iteration": 2.697486162185669 + }, + { + "auxiliary_loss_clip": 0.01152193, + "auxiliary_loss_mlp": 0.01030582, + "balance_loss_clip": 1.05218697, + "balance_loss_mlp": 1.02204108, + "epoch": 0.4933565802921902, + "flos": 21135584200320.0, + "grad_norm": 2.003142219279998, + "language_loss": 0.6479193, + "learning_rate": 2.140100064100932e-06, + "loss": 0.66974699, + "num_input_tokens_seen": 88299270, + "step": 4103, + "time_per_iteration": 2.7783288955688477 + }, + { + "auxiliary_loss_clip": 0.01174378, + "auxiliary_loss_mlp": 0.01031859, + "balance_loss_clip": 1.05285692, + "balance_loss_mlp": 1.02284074, + "epoch": 0.4934768231828293, + "flos": 18039007595520.0, + "grad_norm": 2.0168845440016896, + "language_loss": 0.76100957, + "learning_rate": 2.139322995406746e-06, + "loss": 0.78307188, + "num_input_tokens_seen": 88316905, + "step": 4104, + "time_per_iteration": 2.740114450454712 + }, + { + "auxiliary_loss_clip": 0.01188051, + "auxiliary_loss_mlp": 0.01033538, + "balance_loss_clip": 1.05499601, + "balance_loss_mlp": 1.02440655, + "epoch": 0.4935970660734684, + "flos": 23469957181440.0, + "grad_norm": 1.9414334271286753, + "language_loss": 0.79593104, + "learning_rate": 2.1385459055773727e-06, + "loss": 0.81814694, + "num_input_tokens_seen": 88335095, + "step": 4105, + "time_per_iteration": 2.7206544876098633 + }, + { + "auxiliary_loss_clip": 0.01133583, + "auxiliary_loss_mlp": 0.00886276, + "balance_loss_clip": 1.04450393, + "balance_loss_mlp": 1.00100899, + "epoch": 0.4937173089641075, + "flos": 64479258840960.0, + "grad_norm": 2.4491652509140285, + "language_loss": 0.73593146, + "learning_rate": 2.137768794730696e-06, + "loss": 0.7561301, + "num_input_tokens_seen": 88358545, + "step": 4106, + "time_per_iteration": 3.2231638431549072 + }, + { + "auxiliary_loss_clip": 0.01174876, + "auxiliary_loss_mlp": 0.01032405, + "balance_loss_clip": 1.05466485, + "balance_loss_mlp": 1.02267158, + "epoch": 0.4938375518547466, + "flos": 22346025644160.0, + "grad_norm": 2.589967788960554, + "language_loss": 0.80424201, + "learning_rate": 2.1369916629846026e-06, + "loss": 0.82631487, + "num_input_tokens_seen": 88378295, + "step": 4107, + "time_per_iteration": 2.7587625980377197 + }, + { + "auxiliary_loss_clip": 0.01167507, + "auxiliary_loss_mlp": 0.01034062, + "balance_loss_clip": 1.04954052, + "balance_loss_mlp": 1.02504992, + "epoch": 0.4939577947453857, + "flos": 17858700299520.0, + "grad_norm": 1.845190986827956, + "language_loss": 0.75061542, + "learning_rate": 2.136214510456983e-06, + "loss": 0.77263111, + "num_input_tokens_seen": 88396750, + "step": 4108, + "time_per_iteration": 2.7535552978515625 + }, + { + "auxiliary_loss_clip": 0.01053625, + "auxiliary_loss_mlp": 0.00877264, + "balance_loss_clip": 1.01282477, + "balance_loss_mlp": 1.00136089, + "epoch": 0.49407803763602476, + "flos": 70066746875520.0, + "grad_norm": 0.8855190712278455, + "language_loss": 0.63160932, + "learning_rate": 2.1354373372657296e-06, + "loss": 0.65091825, + "num_input_tokens_seen": 88455190, + "step": 4109, + "time_per_iteration": 3.540813684463501 + }, + { + "auxiliary_loss_clip": 0.01185899, + "auxiliary_loss_mlp": 0.01036239, + "balance_loss_clip": 1.05448508, + "balance_loss_mlp": 1.02784705, + "epoch": 0.49419828052666387, + "flos": 24317485562880.0, + "grad_norm": 1.4945768572034643, + "language_loss": 0.70737547, + "learning_rate": 2.1346601435287404e-06, + "loss": 0.72959679, + "num_input_tokens_seen": 88477460, + "step": 4110, + "time_per_iteration": 2.67692494392395 + }, + { + "auxiliary_loss_clip": 0.01163611, + "auxiliary_loss_mlp": 0.01028995, + "balance_loss_clip": 1.04785073, + "balance_loss_mlp": 1.02073944, + "epoch": 0.494318523417303, + "flos": 29386060790400.0, + "grad_norm": 1.7181378280614907, + "language_loss": 0.8010453, + "learning_rate": 2.1338829293639144e-06, + "loss": 0.82297134, + "num_input_tokens_seen": 88497820, + "step": 4111, + "time_per_iteration": 2.867802143096924 + }, + { + "auxiliary_loss_clip": 0.01143163, + "auxiliary_loss_mlp": 0.01032038, + "balance_loss_clip": 1.0463376, + "balance_loss_mlp": 1.02320433, + "epoch": 0.49443876630794203, + "flos": 15268284195840.0, + "grad_norm": 2.17189979811455, + "language_loss": 0.83138913, + "learning_rate": 2.1331056948891547e-06, + "loss": 0.85314119, + "num_input_tokens_seen": 88514920, + "step": 4112, + "time_per_iteration": 2.786996603012085 + }, + { + "auxiliary_loss_clip": 0.01159188, + "auxiliary_loss_mlp": 0.01028494, + "balance_loss_clip": 1.04863286, + "balance_loss_mlp": 1.01984584, + "epoch": 0.49455900919858115, + "flos": 12347453859840.0, + "grad_norm": 2.534795478639953, + "language_loss": 0.76139164, + "learning_rate": 2.1323284402223666e-06, + "loss": 0.78326845, + "num_input_tokens_seen": 88530910, + "step": 4113, + "time_per_iteration": 2.6969189643859863 + }, + { + "auxiliary_loss_clip": 0.01186691, + "auxiliary_loss_mlp": 0.00885473, + "balance_loss_clip": 1.05768359, + "balance_loss_mlp": 1.00100207, + "epoch": 0.4946792520892202, + "flos": 22779610715520.0, + "grad_norm": 1.760909916173635, + "language_loss": 0.87861013, + "learning_rate": 2.1315511654814597e-06, + "loss": 0.89933175, + "num_input_tokens_seen": 88549320, + "step": 4114, + "time_per_iteration": 2.6706323623657227 + }, + { + "auxiliary_loss_clip": 0.01157421, + "auxiliary_loss_mlp": 0.01032799, + "balance_loss_clip": 1.04958642, + "balance_loss_mlp": 1.02457929, + "epoch": 0.4947994949798593, + "flos": 23148126299520.0, + "grad_norm": 3.5847813038980587, + "language_loss": 0.7799691, + "learning_rate": 2.1307738707843456e-06, + "loss": 0.8018713, + "num_input_tokens_seen": 88568985, + "step": 4115, + "time_per_iteration": 2.841510057449341 + }, + { + "auxiliary_loss_clip": 0.01182766, + "auxiliary_loss_mlp": 0.01030819, + "balance_loss_clip": 1.05365491, + "balance_loss_mlp": 1.02129447, + "epoch": 0.4949197378704984, + "flos": 23659997063040.0, + "grad_norm": 2.6175976369513076, + "language_loss": 0.68802941, + "learning_rate": 2.1299965562489385e-06, + "loss": 0.71016532, + "num_input_tokens_seen": 88588790, + "step": 4116, + "time_per_iteration": 2.7596585750579834 + }, + { + "auxiliary_loss_clip": 0.01172239, + "auxiliary_loss_mlp": 0.01029644, + "balance_loss_clip": 1.04866195, + "balance_loss_mlp": 1.02081621, + "epoch": 0.4950399807611375, + "flos": 26911493026560.0, + "grad_norm": 1.6082323239957168, + "language_loss": 0.7885561, + "learning_rate": 2.129219221993158e-06, + "loss": 0.81057495, + "num_input_tokens_seen": 88613575, + "step": 4117, + "time_per_iteration": 2.8290457725524902 + }, + { + "auxiliary_loss_clip": 0.01056608, + "auxiliary_loss_mlp": 0.01000667, + "balance_loss_clip": 1.01683927, + "balance_loss_mlp": 0.99866468, + "epoch": 0.4951602236517766, + "flos": 67315270187520.0, + "grad_norm": 0.7882818328858866, + "language_loss": 0.59980124, + "learning_rate": 2.128441868134924e-06, + "loss": 0.62037396, + "num_input_tokens_seen": 88675510, + "step": 4118, + "time_per_iteration": 4.345280170440674 + }, + { + "auxiliary_loss_clip": 0.01157417, + "auxiliary_loss_mlp": 0.01029832, + "balance_loss_clip": 1.04725456, + "balance_loss_mlp": 1.02120173, + "epoch": 0.4952804665424157, + "flos": 19901442758400.0, + "grad_norm": 2.1199479284749967, + "language_loss": 0.82758993, + "learning_rate": 2.1276644947921606e-06, + "loss": 0.84946239, + "num_input_tokens_seen": 88694425, + "step": 4119, + "time_per_iteration": 3.9110355377197266 + }, + { + "auxiliary_loss_clip": 0.01173883, + "auxiliary_loss_mlp": 0.01031409, + "balance_loss_clip": 1.05026221, + "balance_loss_mlp": 1.02222991, + "epoch": 0.49540070943305475, + "flos": 18806813740800.0, + "grad_norm": 2.981890512300296, + "language_loss": 0.82399315, + "learning_rate": 2.126887102082795e-06, + "loss": 0.84604603, + "num_input_tokens_seen": 88714450, + "step": 4120, + "time_per_iteration": 2.7601542472839355 + }, + { + "auxiliary_loss_clip": 0.01155205, + "auxiliary_loss_mlp": 0.01033592, + "balance_loss_clip": 1.04633093, + "balance_loss_mlp": 1.02500868, + "epoch": 0.49552095232369386, + "flos": 24934179191040.0, + "grad_norm": 1.9158336942246346, + "language_loss": 0.7052446, + "learning_rate": 2.126109690124757e-06, + "loss": 0.72713256, + "num_input_tokens_seen": 88735265, + "step": 4121, + "time_per_iteration": 3.8007524013519287 + }, + { + "auxiliary_loss_clip": 0.01148451, + "auxiliary_loss_mlp": 0.01032972, + "balance_loss_clip": 1.04726434, + "balance_loss_mlp": 1.02466309, + "epoch": 0.495641195214333, + "flos": 22857249962880.0, + "grad_norm": 1.8006296437139622, + "language_loss": 0.7132777, + "learning_rate": 2.1253322590359786e-06, + "loss": 0.73509192, + "num_input_tokens_seen": 88754600, + "step": 4122, + "time_per_iteration": 2.8794548511505127 + }, + { + "auxiliary_loss_clip": 0.01170388, + "auxiliary_loss_mlp": 0.01035064, + "balance_loss_clip": 1.04777026, + "balance_loss_mlp": 1.02595615, + "epoch": 0.49576143810497203, + "flos": 25769748343680.0, + "grad_norm": 1.6426657635872919, + "language_loss": 0.7374717, + "learning_rate": 2.124554808934397e-06, + "loss": 0.75952625, + "num_input_tokens_seen": 88775180, + "step": 4123, + "time_per_iteration": 2.710129976272583 + }, + { + "auxiliary_loss_clip": 0.01139383, + "auxiliary_loss_mlp": 0.01033357, + "balance_loss_clip": 1.04344511, + "balance_loss_mlp": 1.02503622, + "epoch": 0.49588168099561114, + "flos": 22128838058880.0, + "grad_norm": 1.8359208653265249, + "language_loss": 0.72838843, + "learning_rate": 2.1237773399379496e-06, + "loss": 0.75011575, + "num_input_tokens_seen": 88796145, + "step": 4124, + "time_per_iteration": 2.871079444885254 + }, + { + "auxiliary_loss_clip": 0.01164603, + "auxiliary_loss_mlp": 0.01026279, + "balance_loss_clip": 1.04785633, + "balance_loss_mlp": 1.01751077, + "epoch": 0.49600192388625025, + "flos": 24387331559040.0, + "grad_norm": 9.035408639349079, + "language_loss": 0.87027347, + "learning_rate": 2.122999852164578e-06, + "loss": 0.89218223, + "num_input_tokens_seen": 88816765, + "step": 4125, + "time_per_iteration": 2.8303005695343018 + }, + { + "auxiliary_loss_clip": 0.01141222, + "auxiliary_loss_mlp": 0.01024877, + "balance_loss_clip": 1.04925132, + "balance_loss_mlp": 1.01583517, + "epoch": 0.4961221667768893, + "flos": 22857429530880.0, + "grad_norm": 15.562711704373758, + "language_loss": 0.58594513, + "learning_rate": 2.122222345732227e-06, + "loss": 0.60760617, + "num_input_tokens_seen": 88836680, + "step": 4126, + "time_per_iteration": 2.858920097351074 + }, + { + "auxiliary_loss_clip": 0.01155977, + "auxiliary_loss_mlp": 0.01029788, + "balance_loss_clip": 1.04718387, + "balance_loss_mlp": 1.02032912, + "epoch": 0.4962424096675284, + "flos": 17858089768320.0, + "grad_norm": 2.0266802944307973, + "language_loss": 0.82886457, + "learning_rate": 2.121444820758843e-06, + "loss": 0.85072219, + "num_input_tokens_seen": 88855320, + "step": 4127, + "time_per_iteration": 2.9647562503814697 + }, + { + "auxiliary_loss_clip": 0.01147515, + "auxiliary_loss_mlp": 0.01028086, + "balance_loss_clip": 1.04830229, + "balance_loss_mlp": 1.01827455, + "epoch": 0.49636265255816747, + "flos": 21793611404160.0, + "grad_norm": 2.055536400020736, + "language_loss": 0.78860629, + "learning_rate": 2.120667277362376e-06, + "loss": 0.81036234, + "num_input_tokens_seen": 88874035, + "step": 4128, + "time_per_iteration": 2.8726346492767334 + }, + { + "auxiliary_loss_clip": 0.01186632, + "auxiliary_loss_mlp": 0.01033466, + "balance_loss_clip": 1.05464983, + "balance_loss_mlp": 1.024472, + "epoch": 0.4964828954488066, + "flos": 16358603581440.0, + "grad_norm": 2.339562133963871, + "language_loss": 0.84937233, + "learning_rate": 2.1198897156607796e-06, + "loss": 0.87157333, + "num_input_tokens_seen": 88891390, + "step": 4129, + "time_per_iteration": 2.703248977661133 + }, + { + "auxiliary_loss_clip": 0.01175146, + "auxiliary_loss_mlp": 0.01026286, + "balance_loss_clip": 1.05193448, + "balance_loss_mlp": 1.01750064, + "epoch": 0.4966031383394457, + "flos": 24711101775360.0, + "grad_norm": 2.538206820043643, + "language_loss": 0.73689973, + "learning_rate": 2.1191121357720085e-06, + "loss": 0.75891405, + "num_input_tokens_seen": 88909450, + "step": 4130, + "time_per_iteration": 2.70802640914917 + }, + { + "auxiliary_loss_clip": 0.01138481, + "auxiliary_loss_mlp": 0.01032074, + "balance_loss_clip": 1.04644275, + "balance_loss_mlp": 1.02297831, + "epoch": 0.49672338123008475, + "flos": 22930615491840.0, + "grad_norm": 2.220066406211833, + "language_loss": 0.74596262, + "learning_rate": 2.1183345378140206e-06, + "loss": 0.76766819, + "num_input_tokens_seen": 88929195, + "step": 4131, + "time_per_iteration": 2.8717825412750244 + }, + { + "auxiliary_loss_clip": 0.01068507, + "auxiliary_loss_mlp": 0.01008882, + "balance_loss_clip": 1.01281285, + "balance_loss_mlp": 1.00712955, + "epoch": 0.49684362412072386, + "flos": 65976736844160.0, + "grad_norm": 0.8619744151375953, + "language_loss": 0.62002391, + "learning_rate": 2.1175569219047783e-06, + "loss": 0.64079779, + "num_input_tokens_seen": 88990635, + "step": 4132, + "time_per_iteration": 3.3635056018829346 + }, + { + "auxiliary_loss_clip": 0.01185143, + "auxiliary_loss_mlp": 0.01031644, + "balance_loss_clip": 1.05265236, + "balance_loss_mlp": 1.0229423, + "epoch": 0.49696386701136297, + "flos": 19971288754560.0, + "grad_norm": 1.852292426551614, + "language_loss": 0.73409313, + "learning_rate": 2.1167792881622437e-06, + "loss": 0.75626099, + "num_input_tokens_seen": 89009655, + "step": 4133, + "time_per_iteration": 2.771444320678711 + }, + { + "auxiliary_loss_clip": 0.01156358, + "auxiliary_loss_mlp": 0.01027793, + "balance_loss_clip": 1.04849803, + "balance_loss_mlp": 1.01916826, + "epoch": 0.497084109902002, + "flos": 24750819239040.0, + "grad_norm": 1.8011113215207244, + "language_loss": 0.80861175, + "learning_rate": 2.116001636704384e-06, + "loss": 0.83045328, + "num_input_tokens_seen": 89030040, + "step": 4134, + "time_per_iteration": 2.954247236251831 + }, + { + "auxiliary_loss_clip": 0.01148665, + "auxiliary_loss_mlp": 0.01037874, + "balance_loss_clip": 1.04679883, + "balance_loss_mlp": 1.02893317, + "epoch": 0.49720435279264114, + "flos": 21871825269120.0, + "grad_norm": 2.9298236146714323, + "language_loss": 0.80027276, + "learning_rate": 2.1152239676491685e-06, + "loss": 0.82213813, + "num_input_tokens_seen": 89048145, + "step": 4135, + "time_per_iteration": 2.896728992462158 + }, + { + "auxiliary_loss_clip": 0.0117095, + "auxiliary_loss_mlp": 0.01028991, + "balance_loss_clip": 1.04827952, + "balance_loss_mlp": 1.02007461, + "epoch": 0.49732459568328025, + "flos": 23805794367360.0, + "grad_norm": 1.7323077957790818, + "language_loss": 0.73503882, + "learning_rate": 2.114446281114569e-06, + "loss": 0.75703824, + "num_input_tokens_seen": 89067165, + "step": 4136, + "time_per_iteration": 2.8799800872802734 + }, + { + "auxiliary_loss_clip": 0.01157668, + "auxiliary_loss_mlp": 0.01033824, + "balance_loss_clip": 1.04892027, + "balance_loss_mlp": 1.02529502, + "epoch": 0.4974448385739193, + "flos": 20047742853120.0, + "grad_norm": 1.831271548054411, + "language_loss": 0.76021159, + "learning_rate": 2.1136685772185587e-06, + "loss": 0.78212649, + "num_input_tokens_seen": 89086190, + "step": 4137, + "time_per_iteration": 2.885012626647949 + }, + { + "auxiliary_loss_clip": 0.01164023, + "auxiliary_loss_mlp": 0.00886942, + "balance_loss_clip": 1.04366255, + "balance_loss_mlp": 1.0009203, + "epoch": 0.4975650814645584, + "flos": 24821347593600.0, + "grad_norm": 1.6298828958848675, + "language_loss": 0.78095675, + "learning_rate": 2.1128908560791163e-06, + "loss": 0.80146641, + "num_input_tokens_seen": 89106020, + "step": 4138, + "time_per_iteration": 2.929577350616455 + }, + { + "auxiliary_loss_clip": 0.01183681, + "auxiliary_loss_mlp": 0.01030022, + "balance_loss_clip": 1.05204391, + "balance_loss_mlp": 1.02173114, + "epoch": 0.4976853243551975, + "flos": 19829477859840.0, + "grad_norm": 3.0609970876940724, + "language_loss": 0.78263009, + "learning_rate": 2.1121131178142203e-06, + "loss": 0.80476707, + "num_input_tokens_seen": 89125385, + "step": 4139, + "time_per_iteration": 2.7170228958129883 + }, + { + "auxiliary_loss_clip": 0.01165029, + "auxiliary_loss_mlp": 0.01031503, + "balance_loss_clip": 1.04790008, + "balance_loss_mlp": 1.02335536, + "epoch": 0.4978055672458366, + "flos": 23142990654720.0, + "grad_norm": 1.6960697609061828, + "language_loss": 0.82589328, + "learning_rate": 2.1113353625418544e-06, + "loss": 0.84785855, + "num_input_tokens_seen": 89143935, + "step": 4140, + "time_per_iteration": 2.8080978393554688 + }, + { + "auxiliary_loss_clip": 0.01170776, + "auxiliary_loss_mlp": 0.01028922, + "balance_loss_clip": 1.05386233, + "balance_loss_mlp": 1.02094138, + "epoch": 0.4979258101364757, + "flos": 15559914718080.0, + "grad_norm": 1.5983584710912848, + "language_loss": 0.78764135, + "learning_rate": 2.1105575903800017e-06, + "loss": 0.80963838, + "num_input_tokens_seen": 89162655, + "step": 4141, + "time_per_iteration": 2.757108449935913 + }, + { + "auxiliary_loss_clip": 0.01176135, + "auxiliary_loss_mlp": 0.01036362, + "balance_loss_clip": 1.05153811, + "balance_loss_mlp": 1.02789259, + "epoch": 0.4980460530271148, + "flos": 26356169784960.0, + "grad_norm": 1.8558042796404648, + "language_loss": 0.85005605, + "learning_rate": 2.1097798014466502e-06, + "loss": 0.87218106, + "num_input_tokens_seen": 89182255, + "step": 4142, + "time_per_iteration": 2.7056970596313477 + }, + { + "auxiliary_loss_clip": 0.01179717, + "auxiliary_loss_mlp": 0.01030539, + "balance_loss_clip": 1.05218232, + "balance_loss_mlp": 1.0210743, + "epoch": 0.49816629591775385, + "flos": 17274541415040.0, + "grad_norm": 2.2816010565874048, + "language_loss": 0.58932102, + "learning_rate": 2.109001995859791e-06, + "loss": 0.61142355, + "num_input_tokens_seen": 89201155, + "step": 4143, + "time_per_iteration": 2.677304744720459 + }, + { + "auxiliary_loss_clip": 0.01059732, + "auxiliary_loss_mlp": 0.01003528, + "balance_loss_clip": 1.01339912, + "balance_loss_mlp": 1.00165653, + "epoch": 0.49828653880839296, + "flos": 64930947344640.0, + "grad_norm": 0.80058663651006, + "language_loss": 0.60083365, + "learning_rate": 2.108224173737415e-06, + "loss": 0.62146628, + "num_input_tokens_seen": 89264455, + "step": 4144, + "time_per_iteration": 4.161435127258301 + }, + { + "auxiliary_loss_clip": 0.01163861, + "auxiliary_loss_mlp": 0.01035013, + "balance_loss_clip": 1.047351, + "balance_loss_mlp": 1.026227, + "epoch": 0.498406781699032, + "flos": 27484806003840.0, + "grad_norm": 1.8510557314941225, + "language_loss": 0.76304889, + "learning_rate": 2.1074463351975183e-06, + "loss": 0.78503764, + "num_input_tokens_seen": 89283340, + "step": 4145, + "time_per_iteration": 4.665031671524048 + }, + { + "auxiliary_loss_clip": 0.01154275, + "auxiliary_loss_mlp": 0.01028312, + "balance_loss_clip": 1.04765284, + "balance_loss_mlp": 1.02004504, + "epoch": 0.49852702458967113, + "flos": 31499870307840.0, + "grad_norm": 1.6858232836828286, + "language_loss": 0.71082079, + "learning_rate": 2.106668480358098e-06, + "loss": 0.7326467, + "num_input_tokens_seen": 89303565, + "step": 4146, + "time_per_iteration": 2.8254120349884033 + }, + { + "auxiliary_loss_clip": 0.01161016, + "auxiliary_loss_mlp": 0.01033456, + "balance_loss_clip": 1.04658759, + "balance_loss_mlp": 1.02394342, + "epoch": 0.49864726748031024, + "flos": 22852868503680.0, + "grad_norm": 1.7471670573050049, + "language_loss": 0.7095952, + "learning_rate": 2.105890609337154e-06, + "loss": 0.73153985, + "num_input_tokens_seen": 89322080, + "step": 4147, + "time_per_iteration": 3.6945126056671143 + }, + { + "auxiliary_loss_clip": 0.01071671, + "auxiliary_loss_mlp": 0.01002551, + "balance_loss_clip": 1.01216125, + "balance_loss_mlp": 1.00069129, + "epoch": 0.4987675103709493, + "flos": 70405708544640.0, + "grad_norm": 0.6856057191674725, + "language_loss": 0.63793671, + "learning_rate": 2.1051127222526883e-06, + "loss": 0.65867895, + "num_input_tokens_seen": 89394195, + "step": 4148, + "time_per_iteration": 3.3328213691711426 + }, + { + "auxiliary_loss_clip": 0.01173214, + "auxiliary_loss_mlp": 0.01029025, + "balance_loss_clip": 1.05263877, + "balance_loss_mlp": 1.02048993, + "epoch": 0.4988877532615884, + "flos": 28767571482240.0, + "grad_norm": 2.0636475839415804, + "language_loss": 0.80705106, + "learning_rate": 2.1043348192227067e-06, + "loss": 0.82907349, + "num_input_tokens_seen": 89414565, + "step": 4149, + "time_per_iteration": 2.8479795455932617 + }, + { + "auxiliary_loss_clip": 0.01143472, + "auxiliary_loss_mlp": 0.01032687, + "balance_loss_clip": 1.04609442, + "balance_loss_mlp": 1.02375269, + "epoch": 0.4990079961522275, + "flos": 16872700988160.0, + "grad_norm": 2.3867020935987022, + "language_loss": 0.6169976, + "learning_rate": 2.1035569003652156e-06, + "loss": 0.63875914, + "num_input_tokens_seen": 89433195, + "step": 4150, + "time_per_iteration": 2.877908706665039 + }, + { + "auxiliary_loss_clip": 0.01142437, + "auxiliary_loss_mlp": 0.01034583, + "balance_loss_clip": 1.04609644, + "balance_loss_mlp": 1.02519536, + "epoch": 0.4991282390428666, + "flos": 13291042187520.0, + "grad_norm": 2.1462804174368015, + "language_loss": 0.8169989, + "learning_rate": 2.1027789657982255e-06, + "loss": 0.83876908, + "num_input_tokens_seen": 89447410, + "step": 4151, + "time_per_iteration": 2.8632009029388428 + }, + { + "auxiliary_loss_clip": 0.01139246, + "auxiliary_loss_mlp": 0.01034732, + "balance_loss_clip": 1.04578125, + "balance_loss_mlp": 1.02597058, + "epoch": 0.4992484819335057, + "flos": 21537496454400.0, + "grad_norm": 1.8376460753507338, + "language_loss": 0.77506274, + "learning_rate": 2.1020010156397482e-06, + "loss": 0.79680252, + "num_input_tokens_seen": 89464630, + "step": 4152, + "time_per_iteration": 2.877047300338745 + }, + { + "auxiliary_loss_clip": 0.01178327, + "auxiliary_loss_mlp": 0.01030489, + "balance_loss_clip": 1.05222821, + "balance_loss_mlp": 1.02179265, + "epoch": 0.4993687248241448, + "flos": 24860095390080.0, + "grad_norm": 1.6693256008112316, + "language_loss": 0.77297795, + "learning_rate": 2.101223050007797e-06, + "loss": 0.79506612, + "num_input_tokens_seen": 89483180, + "step": 4153, + "time_per_iteration": 2.680116891860962 + }, + { + "auxiliary_loss_clip": 0.01070391, + "auxiliary_loss_mlp": 0.01002648, + "balance_loss_clip": 1.0108552, + "balance_loss_mlp": 1.00072908, + "epoch": 0.49948896771478385, + "flos": 62941602453120.0, + "grad_norm": 0.8209892171834348, + "language_loss": 0.53805089, + "learning_rate": 2.1004450690203904e-06, + "loss": 0.55878133, + "num_input_tokens_seen": 89539260, + "step": 4154, + "time_per_iteration": 3.346240520477295 + }, + { + "auxiliary_loss_clip": 0.0107039, + "auxiliary_loss_mlp": 0.01002076, + "balance_loss_clip": 1.0107013, + "balance_loss_mlp": 1.00027585, + "epoch": 0.49960921060542296, + "flos": 68284213516800.0, + "grad_norm": 0.9746333127617375, + "language_loss": 0.63331819, + "learning_rate": 2.099667072795546e-06, + "loss": 0.65404284, + "num_input_tokens_seen": 89601380, + "step": 4155, + "time_per_iteration": 3.3220295906066895 + }, + { + "auxiliary_loss_clip": 0.01174766, + "auxiliary_loss_mlp": 0.01034167, + "balance_loss_clip": 1.04991639, + "balance_loss_mlp": 1.02510071, + "epoch": 0.49972945349606207, + "flos": 23659350618240.0, + "grad_norm": 1.8863330968672634, + "language_loss": 0.79908222, + "learning_rate": 2.0988890614512864e-06, + "loss": 0.82117158, + "num_input_tokens_seen": 89621270, + "step": 4156, + "time_per_iteration": 2.7034549713134766 + }, + { + "auxiliary_loss_clip": 0.01169162, + "auxiliary_loss_mlp": 0.01030784, + "balance_loss_clip": 1.05204749, + "balance_loss_mlp": 1.02192664, + "epoch": 0.4998496963867011, + "flos": 19755825022080.0, + "grad_norm": 2.3670899627110846, + "language_loss": 0.84343433, + "learning_rate": 2.098111035105635e-06, + "loss": 0.86543381, + "num_input_tokens_seen": 89639695, + "step": 4157, + "time_per_iteration": 2.696906566619873 + }, + { + "auxiliary_loss_clip": 0.01141338, + "auxiliary_loss_mlp": 0.0103535, + "balance_loss_clip": 1.04775143, + "balance_loss_mlp": 1.02581322, + "epoch": 0.49996993927734024, + "flos": 22265728790400.0, + "grad_norm": 1.8084102243531943, + "language_loss": 0.7310611, + "learning_rate": 2.0973329938766176e-06, + "loss": 0.752828, + "num_input_tokens_seen": 89657125, + "step": 4158, + "time_per_iteration": 2.783536434173584 + }, + { + "auxiliary_loss_clip": 0.01183395, + "auxiliary_loss_mlp": 0.01030188, + "balance_loss_clip": 1.05324793, + "balance_loss_mlp": 1.02021658, + "epoch": 0.5000901821679793, + "flos": 23327212533120.0, + "grad_norm": 1.819915687389414, + "language_loss": 0.78874874, + "learning_rate": 2.0965549378822618e-06, + "loss": 0.81088454, + "num_input_tokens_seen": 89678415, + "step": 4159, + "time_per_iteration": 2.7185287475585938 + }, + { + "auxiliary_loss_clip": 0.01107326, + "auxiliary_loss_mlp": 0.01037123, + "balance_loss_clip": 1.03975844, + "balance_loss_mlp": 1.02798581, + "epoch": 0.5002104250586185, + "flos": 20339014239360.0, + "grad_norm": 2.0447640129374887, + "language_loss": 0.84176862, + "learning_rate": 2.095776867240599e-06, + "loss": 0.86321306, + "num_input_tokens_seen": 89695405, + "step": 4160, + "time_per_iteration": 3.049694299697876 + }, + { + "auxiliary_loss_clip": 0.01151876, + "auxiliary_loss_mlp": 0.0103346, + "balance_loss_clip": 1.04808152, + "balance_loss_mlp": 1.02482295, + "epoch": 0.5003306679492575, + "flos": 13991372634240.0, + "grad_norm": 1.9326619038655177, + "language_loss": 0.8236537, + "learning_rate": 2.094998782069661e-06, + "loss": 0.84550709, + "num_input_tokens_seen": 89713110, + "step": 4161, + "time_per_iteration": 2.8252034187316895 + }, + { + "auxiliary_loss_clip": 0.01183433, + "auxiliary_loss_mlp": 0.01031559, + "balance_loss_clip": 1.0516274, + "balance_loss_mlp": 1.02249312, + "epoch": 0.5004509108398966, + "flos": 27672762896640.0, + "grad_norm": 1.6507394658792751, + "language_loss": 0.75537932, + "learning_rate": 2.0942206824874845e-06, + "loss": 0.77752924, + "num_input_tokens_seen": 89735885, + "step": 4162, + "time_per_iteration": 2.7148945331573486 + }, + { + "auxiliary_loss_clip": 0.01171638, + "auxiliary_loss_mlp": 0.0103307, + "balance_loss_clip": 1.05115318, + "balance_loss_mlp": 1.02446318, + "epoch": 0.5005711537305357, + "flos": 14976186796800.0, + "grad_norm": 2.4005060912282996, + "language_loss": 0.78626722, + "learning_rate": 2.093442568612105e-06, + "loss": 0.80831432, + "num_input_tokens_seen": 89753690, + "step": 4163, + "time_per_iteration": 2.656193494796753 + }, + { + "auxiliary_loss_clip": 0.01185641, + "auxiliary_loss_mlp": 0.01030923, + "balance_loss_clip": 1.05202115, + "balance_loss_mlp": 1.02194035, + "epoch": 0.5006913966211748, + "flos": 26503259978880.0, + "grad_norm": 1.5017113617918825, + "language_loss": 0.85322416, + "learning_rate": 2.0926644405615613e-06, + "loss": 0.87538981, + "num_input_tokens_seen": 89774590, + "step": 4164, + "time_per_iteration": 2.8142831325531006 + }, + { + "auxiliary_loss_clip": 0.0115216, + "auxiliary_loss_mlp": 0.01030046, + "balance_loss_clip": 1.04668498, + "balance_loss_mlp": 1.02189255, + "epoch": 0.5008116395118138, + "flos": 20449295971200.0, + "grad_norm": 3.2139111978146895, + "language_loss": 0.80811918, + "learning_rate": 2.091886298453897e-06, + "loss": 0.82994127, + "num_input_tokens_seen": 89792775, + "step": 4165, + "time_per_iteration": 2.8485686779022217 + }, + { + "auxiliary_loss_clip": 0.01171833, + "auxiliary_loss_mlp": 0.01036081, + "balance_loss_clip": 1.04911041, + "balance_loss_mlp": 1.02717018, + "epoch": 0.500931882402453, + "flos": 21579871524480.0, + "grad_norm": 2.0691444621534534, + "language_loss": 0.73115289, + "learning_rate": 2.091108142407153e-06, + "loss": 0.753232, + "num_input_tokens_seen": 89811515, + "step": 4166, + "time_per_iteration": 2.8324553966522217 + }, + { + "auxiliary_loss_clip": 0.01056467, + "auxiliary_loss_mlp": 0.01001534, + "balance_loss_clip": 1.01335645, + "balance_loss_mlp": 0.99950773, + "epoch": 0.5010521252930921, + "flos": 57785011925760.0, + "grad_norm": 0.840908296661177, + "language_loss": 0.62388903, + "learning_rate": 2.090329972539377e-06, + "loss": 0.64446908, + "num_input_tokens_seen": 89870080, + "step": 4167, + "time_per_iteration": 3.383150339126587 + }, + { + "auxiliary_loss_clip": 0.01109358, + "auxiliary_loss_mlp": 0.01033455, + "balance_loss_clip": 1.04237163, + "balance_loss_mlp": 1.02447867, + "epoch": 0.5011723681837311, + "flos": 18625500864000.0, + "grad_norm": 1.7137470924850218, + "language_loss": 0.68659258, + "learning_rate": 2.089551788968616e-06, + "loss": 0.70802063, + "num_input_tokens_seen": 89888045, + "step": 4168, + "time_per_iteration": 2.897765636444092 + }, + { + "auxiliary_loss_clip": 0.01071736, + "auxiliary_loss_mlp": 0.01003179, + "balance_loss_clip": 1.01171672, + "balance_loss_mlp": 1.00136745, + "epoch": 0.5012926110743702, + "flos": 55883146608000.0, + "grad_norm": 0.8386561914450598, + "language_loss": 0.60810411, + "learning_rate": 2.08877359181292e-06, + "loss": 0.6288532, + "num_input_tokens_seen": 89944610, + "step": 4169, + "time_per_iteration": 4.048341989517212 + }, + { + "auxiliary_loss_clip": 0.01157087, + "auxiliary_loss_mlp": 0.010312, + "balance_loss_clip": 1.04326057, + "balance_loss_mlp": 1.02209878, + "epoch": 0.5014128539650093, + "flos": 24238266117120.0, + "grad_norm": 2.6542843297690504, + "language_loss": 0.85744429, + "learning_rate": 2.0879953811903396e-06, + "loss": 0.87932718, + "num_input_tokens_seen": 89959495, + "step": 4170, + "time_per_iteration": 2.806083917617798 + }, + { + "auxiliary_loss_clip": 0.0117403, + "auxiliary_loss_mlp": 0.01033962, + "balance_loss_clip": 1.05214167, + "balance_loss_mlp": 1.02516389, + "epoch": 0.5015330968556484, + "flos": 27527468382720.0, + "grad_norm": 1.9363650910156955, + "language_loss": 0.78773993, + "learning_rate": 2.08721715721893e-06, + "loss": 0.80981988, + "num_input_tokens_seen": 89978820, + "step": 4171, + "time_per_iteration": 3.662111282348633 + }, + { + "auxiliary_loss_clip": 0.01174256, + "auxiliary_loss_mlp": 0.01028773, + "balance_loss_clip": 1.05130768, + "balance_loss_mlp": 1.01990402, + "epoch": 0.5016533397462875, + "flos": 23800802376960.0, + "grad_norm": 2.0527021257457543, + "language_loss": 0.7730056, + "learning_rate": 2.0864389200167477e-06, + "loss": 0.7950359, + "num_input_tokens_seen": 89997075, + "step": 4172, + "time_per_iteration": 2.730860948562622 + }, + { + "auxiliary_loss_clip": 0.01176468, + "auxiliary_loss_mlp": 0.00887378, + "balance_loss_clip": 1.05131614, + "balance_loss_mlp": 1.00100756, + "epoch": 0.5017735826369266, + "flos": 25295009264640.0, + "grad_norm": 1.672261105447332, + "language_loss": 0.78627872, + "learning_rate": 2.0856606697018504e-06, + "loss": 0.80691719, + "num_input_tokens_seen": 90015085, + "step": 4173, + "time_per_iteration": 3.8000874519348145 + }, + { + "auxiliary_loss_clip": 0.01157209, + "auxiliary_loss_mlp": 0.01032868, + "balance_loss_clip": 1.04600251, + "balance_loss_mlp": 1.02352762, + "epoch": 0.5018938255275657, + "flos": 16873203778560.0, + "grad_norm": 1.9197547942835382, + "language_loss": 0.72930276, + "learning_rate": 2.084882406392297e-06, + "loss": 0.75120348, + "num_input_tokens_seen": 90033045, + "step": 4174, + "time_per_iteration": 2.8121795654296875 + }, + { + "auxiliary_loss_clip": 0.01177999, + "auxiliary_loss_mlp": 0.0103066, + "balance_loss_clip": 1.05159521, + "balance_loss_mlp": 1.02204704, + "epoch": 0.5020140684182047, + "flos": 25515429073920.0, + "grad_norm": 2.7283985635267696, + "language_loss": 0.71196735, + "learning_rate": 2.0841041302061496e-06, + "loss": 0.73405391, + "num_input_tokens_seen": 90052505, + "step": 4175, + "time_per_iteration": 2.847364664077759 + }, + { + "auxiliary_loss_clip": 0.01154091, + "auxiliary_loss_mlp": 0.01036553, + "balance_loss_clip": 1.04529238, + "balance_loss_mlp": 1.02751112, + "epoch": 0.5021343113088439, + "flos": 23659278791040.0, + "grad_norm": 1.8639496624491467, + "language_loss": 0.75738287, + "learning_rate": 2.083325841261473e-06, + "loss": 0.77928931, + "num_input_tokens_seen": 90071565, + "step": 4176, + "time_per_iteration": 2.794126510620117 + }, + { + "auxiliary_loss_clip": 0.01159018, + "auxiliary_loss_mlp": 0.01031491, + "balance_loss_clip": 1.04592633, + "balance_loss_mlp": 1.02232385, + "epoch": 0.502254554199483, + "flos": 24534673148160.0, + "grad_norm": 2.013938648566689, + "language_loss": 0.66367626, + "learning_rate": 2.0825475396763322e-06, + "loss": 0.68558133, + "num_input_tokens_seen": 90092215, + "step": 4177, + "time_per_iteration": 2.927180528640747 + }, + { + "auxiliary_loss_clip": 0.011065, + "auxiliary_loss_mlp": 0.01029152, + "balance_loss_clip": 1.03977323, + "balance_loss_mlp": 1.02013385, + "epoch": 0.502374797090122, + "flos": 34240285607040.0, + "grad_norm": 7.0646484011692925, + "language_loss": 0.65730304, + "learning_rate": 2.081769225568796e-06, + "loss": 0.67865956, + "num_input_tokens_seen": 90114665, + "step": 4178, + "time_per_iteration": 3.0581841468811035 + }, + { + "auxiliary_loss_clip": 0.01176252, + "auxiliary_loss_mlp": 0.01029085, + "balance_loss_clip": 1.04966831, + "balance_loss_mlp": 1.02030492, + "epoch": 0.5024950399807612, + "flos": 26031106679040.0, + "grad_norm": 1.5921627546701425, + "language_loss": 0.75790679, + "learning_rate": 2.0809908990569327e-06, + "loss": 0.77996016, + "num_input_tokens_seen": 90136445, + "step": 4179, + "time_per_iteration": 2.7735424041748047 + }, + { + "auxiliary_loss_clip": 0.0116293, + "auxiliary_loss_mlp": 0.01031519, + "balance_loss_clip": 1.04847634, + "balance_loss_mlp": 1.02229249, + "epoch": 0.5026152828714002, + "flos": 21252438120960.0, + "grad_norm": 2.0851010610752643, + "language_loss": 0.78950721, + "learning_rate": 2.0802125602588146e-06, + "loss": 0.81145167, + "num_input_tokens_seen": 90155710, + "step": 4180, + "time_per_iteration": 2.7562994956970215 + }, + { + "auxiliary_loss_clip": 0.01181855, + "auxiliary_loss_mlp": 0.01031768, + "balance_loss_clip": 1.05009198, + "balance_loss_mlp": 1.0232091, + "epoch": 0.5027355257620393, + "flos": 30956111245440.0, + "grad_norm": 2.3092358994969495, + "language_loss": 0.66424793, + "learning_rate": 2.0794342092925146e-06, + "loss": 0.6863842, + "num_input_tokens_seen": 90176845, + "step": 4181, + "time_per_iteration": 2.6840946674346924 + }, + { + "auxiliary_loss_clip": 0.01178095, + "auxiliary_loss_mlp": 0.01027829, + "balance_loss_clip": 1.05201042, + "balance_loss_mlp": 1.01948392, + "epoch": 0.5028557686526784, + "flos": 24791147233920.0, + "grad_norm": 2.336672899579486, + "language_loss": 0.68243641, + "learning_rate": 2.078655846276108e-06, + "loss": 0.70449567, + "num_input_tokens_seen": 90197175, + "step": 4182, + "time_per_iteration": 2.7553369998931885 + }, + { + "auxiliary_loss_clip": 0.01160068, + "auxiliary_loss_mlp": 0.01029862, + "balance_loss_clip": 1.04785168, + "balance_loss_mlp": 1.02091515, + "epoch": 0.5029760115433175, + "flos": 22966992990720.0, + "grad_norm": 1.8711480025621656, + "language_loss": 0.68759483, + "learning_rate": 2.0778774713276727e-06, + "loss": 0.70949411, + "num_input_tokens_seen": 90216650, + "step": 4183, + "time_per_iteration": 2.7805137634277344 + }, + { + "auxiliary_loss_clip": 0.01170477, + "auxiliary_loss_mlp": 0.01033433, + "balance_loss_clip": 1.04671609, + "balance_loss_mlp": 1.02359247, + "epoch": 0.5030962544339566, + "flos": 15305164485120.0, + "grad_norm": 4.614519453741359, + "language_loss": 0.67761695, + "learning_rate": 2.077099084565287e-06, + "loss": 0.69965607, + "num_input_tokens_seen": 90234055, + "step": 4184, + "time_per_iteration": 2.743133783340454 + }, + { + "auxiliary_loss_clip": 0.01159163, + "auxiliary_loss_mlp": 0.01032095, + "balance_loss_clip": 1.04626822, + "balance_loss_mlp": 1.02282023, + "epoch": 0.5032164973245957, + "flos": 24494847943680.0, + "grad_norm": 3.2585587601255717, + "language_loss": 0.64938223, + "learning_rate": 2.0763206861070313e-06, + "loss": 0.67129481, + "num_input_tokens_seen": 90253115, + "step": 4185, + "time_per_iteration": 2.7719740867614746 + }, + { + "auxiliary_loss_clip": 0.01185259, + "auxiliary_loss_mlp": 0.01032416, + "balance_loss_clip": 1.05195224, + "balance_loss_mlp": 1.02332056, + "epoch": 0.5033367402152348, + "flos": 16213452721920.0, + "grad_norm": 2.2235411354247243, + "language_loss": 0.75684881, + "learning_rate": 2.0755422760709876e-06, + "loss": 0.77902555, + "num_input_tokens_seen": 90270515, + "step": 4186, + "time_per_iteration": 2.657203197479248 + }, + { + "auxiliary_loss_clip": 0.011329, + "auxiliary_loss_mlp": 0.01034253, + "balance_loss_clip": 1.04526544, + "balance_loss_mlp": 1.02550304, + "epoch": 0.5034569831058738, + "flos": 21391375927680.0, + "grad_norm": 1.8782602503930481, + "language_loss": 0.76628637, + "learning_rate": 2.0747638545752417e-06, + "loss": 0.78795791, + "num_input_tokens_seen": 90289075, + "step": 4187, + "time_per_iteration": 2.910240650177002 + }, + { + "auxiliary_loss_clip": 0.01160455, + "auxiliary_loss_mlp": 0.01027884, + "balance_loss_clip": 1.04960227, + "balance_loss_mlp": 1.01946819, + "epoch": 0.503577225996513, + "flos": 20558751690240.0, + "grad_norm": 1.9947363365468769, + "language_loss": 0.83290327, + "learning_rate": 2.073985421737878e-06, + "loss": 0.85478663, + "num_input_tokens_seen": 90306385, + "step": 4188, + "time_per_iteration": 2.7656447887420654 + }, + { + "auxiliary_loss_clip": 0.01177568, + "auxiliary_loss_mlp": 0.01036391, + "balance_loss_clip": 1.05141687, + "balance_loss_mlp": 1.02680683, + "epoch": 0.5036974688871521, + "flos": 27229157930880.0, + "grad_norm": 2.1798265624573956, + "language_loss": 0.73305696, + "learning_rate": 2.0732069776769844e-06, + "loss": 0.75519657, + "num_input_tokens_seen": 90323795, + "step": 4189, + "time_per_iteration": 2.769932508468628 + }, + { + "auxiliary_loss_clip": 0.01186174, + "auxiliary_loss_mlp": 0.01026646, + "balance_loss_clip": 1.05327415, + "balance_loss_mlp": 1.01784253, + "epoch": 0.5038177117777911, + "flos": 20412164286720.0, + "grad_norm": 1.9482762559691653, + "language_loss": 0.73114121, + "learning_rate": 2.072428522510651e-06, + "loss": 0.75326937, + "num_input_tokens_seen": 90340360, + "step": 4190, + "time_per_iteration": 2.626476287841797 + }, + { + "auxiliary_loss_clip": 0.01150811, + "auxiliary_loss_mlp": 0.01031534, + "balance_loss_clip": 1.0488286, + "balance_loss_mlp": 1.02230716, + "epoch": 0.5039379546684303, + "flos": 21907987286400.0, + "grad_norm": 1.991067940116985, + "language_loss": 0.76155496, + "learning_rate": 2.071650056356968e-06, + "loss": 0.78337842, + "num_input_tokens_seen": 90357900, + "step": 4191, + "time_per_iteration": 2.8112375736236572 + }, + { + "auxiliary_loss_clip": 0.01184643, + "auxiliary_loss_mlp": 0.01030505, + "balance_loss_clip": 1.05178571, + "balance_loss_mlp": 1.02146339, + "epoch": 0.5040581975590693, + "flos": 20010718909440.0, + "grad_norm": 1.945822373949498, + "language_loss": 0.79549611, + "learning_rate": 2.070871579334028e-06, + "loss": 0.81764758, + "num_input_tokens_seen": 90377010, + "step": 4192, + "time_per_iteration": 2.630988359451294 + }, + { + "auxiliary_loss_clip": 0.0118178, + "auxiliary_loss_mlp": 0.0103531, + "balance_loss_clip": 1.05034924, + "balance_loss_mlp": 1.02645254, + "epoch": 0.5041784404497084, + "flos": 20959837931520.0, + "grad_norm": 1.833952980662514, + "language_loss": 0.71589315, + "learning_rate": 2.0700930915599264e-06, + "loss": 0.73806405, + "num_input_tokens_seen": 90396740, + "step": 4193, + "time_per_iteration": 2.663490056991577 + }, + { + "auxiliary_loss_clip": 0.0117988, + "auxiliary_loss_mlp": 0.01036089, + "balance_loss_clip": 1.05026817, + "balance_loss_mlp": 1.02691615, + "epoch": 0.5042986833403476, + "flos": 12495082757760.0, + "grad_norm": 2.2136923133758915, + "language_loss": 0.78422076, + "learning_rate": 2.0693145931527583e-06, + "loss": 0.80638045, + "num_input_tokens_seen": 90413220, + "step": 4194, + "time_per_iteration": 2.5792272090911865 + }, + { + "auxiliary_loss_clip": 0.01161281, + "auxiliary_loss_mlp": 0.01033484, + "balance_loss_clip": 1.04879165, + "balance_loss_mlp": 1.02432847, + "epoch": 0.5044189262309866, + "flos": 29202305788800.0, + "grad_norm": 1.643306443284753, + "language_loss": 0.77569461, + "learning_rate": 2.068536084230622e-06, + "loss": 0.79764229, + "num_input_tokens_seen": 90435085, + "step": 4195, + "time_per_iteration": 2.8790862560272217 + }, + { + "auxiliary_loss_clip": 0.01172256, + "auxiliary_loss_mlp": 0.01035853, + "balance_loss_clip": 1.0493207, + "balance_loss_mlp": 1.02575588, + "epoch": 0.5045391691216257, + "flos": 23873198238720.0, + "grad_norm": 1.8977276823738816, + "language_loss": 0.88660079, + "learning_rate": 2.067757564911616e-06, + "loss": 0.90868187, + "num_input_tokens_seen": 90453660, + "step": 4196, + "time_per_iteration": 3.588397741317749 + }, + { + "auxiliary_loss_clip": 0.0117211, + "auxiliary_loss_mlp": 0.00886871, + "balance_loss_clip": 1.04962301, + "balance_loss_mlp": 1.00101089, + "epoch": 0.5046594120122648, + "flos": 24644990793600.0, + "grad_norm": 2.1022132601582806, + "language_loss": 0.92241454, + "learning_rate": 2.0669790353138407e-06, + "loss": 0.94300437, + "num_input_tokens_seen": 90472625, + "step": 4197, + "time_per_iteration": 4.6888206005096436 + }, + { + "auxiliary_loss_clip": 0.01153646, + "auxiliary_loss_mlp": 0.00887096, + "balance_loss_clip": 1.05037034, + "balance_loss_mlp": 1.00103998, + "epoch": 0.5047796549029039, + "flos": 23362835846400.0, + "grad_norm": 2.800246837862335, + "language_loss": 0.73162287, + "learning_rate": 2.0662004955553995e-06, + "loss": 0.75203031, + "num_input_tokens_seen": 90492325, + "step": 4198, + "time_per_iteration": 2.8448972702026367 + }, + { + "auxiliary_loss_clip": 0.01158143, + "auxiliary_loss_mlp": 0.01028557, + "balance_loss_clip": 1.04651403, + "balance_loss_mlp": 1.01972365, + "epoch": 0.5048998977935429, + "flos": 17304095329920.0, + "grad_norm": 1.9572228591679752, + "language_loss": 0.76896501, + "learning_rate": 2.065421945754395e-06, + "loss": 0.79083204, + "num_input_tokens_seen": 90510055, + "step": 4199, + "time_per_iteration": 3.705711603164673 + }, + { + "auxiliary_loss_clip": 0.01146616, + "auxiliary_loss_mlp": 0.0103572, + "balance_loss_clip": 1.04864502, + "balance_loss_mlp": 1.0274291, + "epoch": 0.505020140684182, + "flos": 34856979235200.0, + "grad_norm": 1.6351557645303152, + "language_loss": 0.77812642, + "learning_rate": 2.0646433860289344e-06, + "loss": 0.79994977, + "num_input_tokens_seen": 90528980, + "step": 4200, + "time_per_iteration": 2.9471731185913086 + }, + { + "auxiliary_loss_clip": 0.01178823, + "auxiliary_loss_mlp": 0.00887162, + "balance_loss_clip": 1.05145395, + "balance_loss_mlp": 1.00098789, + "epoch": 0.5051403835748212, + "flos": 24863974058880.0, + "grad_norm": 1.9750005567404791, + "language_loss": 0.82615322, + "learning_rate": 2.0638648164971233e-06, + "loss": 0.84681308, + "num_input_tokens_seen": 90547445, + "step": 4201, + "time_per_iteration": 2.800731897354126 + }, + { + "auxiliary_loss_clip": 0.01163288, + "auxiliary_loss_mlp": 0.01026035, + "balance_loss_clip": 1.04886293, + "balance_loss_mlp": 1.01742232, + "epoch": 0.5052606264654602, + "flos": 20959694277120.0, + "grad_norm": 2.018818527899504, + "language_loss": 0.88395965, + "learning_rate": 2.06308623727707e-06, + "loss": 0.90585291, + "num_input_tokens_seen": 90567545, + "step": 4202, + "time_per_iteration": 2.7575137615203857 + }, + { + "auxiliary_loss_clip": 0.01172396, + "auxiliary_loss_mlp": 0.01038599, + "balance_loss_clip": 1.04922092, + "balance_loss_mlp": 1.02896094, + "epoch": 0.5053808693560993, + "flos": 19642382893440.0, + "grad_norm": 3.3141190917170227, + "language_loss": 0.76378644, + "learning_rate": 2.0623076484868846e-06, + "loss": 0.78589636, + "num_input_tokens_seen": 90585000, + "step": 4203, + "time_per_iteration": 2.6857213973999023 + }, + { + "auxiliary_loss_clip": 0.01058657, + "auxiliary_loss_mlp": 0.01003018, + "balance_loss_clip": 1.01660681, + "balance_loss_mlp": 1.00093162, + "epoch": 0.5055011122467384, + "flos": 67504915019520.0, + "grad_norm": 0.8354680538691587, + "language_loss": 0.60724944, + "learning_rate": 2.061529050244679e-06, + "loss": 0.62786615, + "num_input_tokens_seen": 90644745, + "step": 4204, + "time_per_iteration": 3.277947187423706 + }, + { + "auxiliary_loss_clip": 0.01156061, + "auxiliary_loss_mlp": 0.01031508, + "balance_loss_clip": 1.04809058, + "balance_loss_mlp": 1.02277625, + "epoch": 0.5056213551373775, + "flos": 16872952383360.0, + "grad_norm": 1.8809081732168074, + "language_loss": 0.74111664, + "learning_rate": 2.060750442668565e-06, + "loss": 0.76299238, + "num_input_tokens_seen": 90662500, + "step": 4205, + "time_per_iteration": 2.7664451599121094 + }, + { + "auxiliary_loss_clip": 0.01171724, + "auxiliary_loss_mlp": 0.0103117, + "balance_loss_clip": 1.04977465, + "balance_loss_mlp": 1.0227598, + "epoch": 0.5057415980280165, + "flos": 15334179696000.0, + "grad_norm": 2.447615218090069, + "language_loss": 0.64276743, + "learning_rate": 2.059971825876657e-06, + "loss": 0.66479635, + "num_input_tokens_seen": 90677010, + "step": 4206, + "time_per_iteration": 2.6644058227539062 + }, + { + "auxiliary_loss_clip": 0.01174513, + "auxiliary_loss_mlp": 0.01027321, + "balance_loss_clip": 1.04942834, + "balance_loss_mlp": 1.01856494, + "epoch": 0.5058618409186557, + "flos": 19025976574080.0, + "grad_norm": 1.884691728449054, + "language_loss": 0.76043445, + "learning_rate": 2.0591931999870713e-06, + "loss": 0.78245276, + "num_input_tokens_seen": 90695935, + "step": 4207, + "time_per_iteration": 2.668086290359497 + }, + { + "auxiliary_loss_clip": 0.01062657, + "auxiliary_loss_mlp": 0.01001093, + "balance_loss_clip": 1.01516879, + "balance_loss_mlp": 0.99907869, + "epoch": 0.5059820838092948, + "flos": 63453114080640.0, + "grad_norm": 0.8244185021955439, + "language_loss": 0.57565272, + "learning_rate": 2.0584145651179234e-06, + "loss": 0.59629023, + "num_input_tokens_seen": 90751645, + "step": 4208, + "time_per_iteration": 3.3342902660369873 + }, + { + "auxiliary_loss_clip": 0.01165146, + "auxiliary_loss_mlp": 0.0088721, + "balance_loss_clip": 1.05145907, + "balance_loss_mlp": 1.00109088, + "epoch": 0.5061023266999338, + "flos": 15441803821440.0, + "grad_norm": 2.717864529810573, + "language_loss": 0.79935563, + "learning_rate": 2.0576359213873327e-06, + "loss": 0.81987917, + "num_input_tokens_seen": 90766795, + "step": 4209, + "time_per_iteration": 2.717463254928589 + }, + { + "auxiliary_loss_clip": 0.01169475, + "auxiliary_loss_mlp": 0.01032729, + "balance_loss_clip": 1.04569554, + "balance_loss_mlp": 1.02284598, + "epoch": 0.506222569590573, + "flos": 22451063990400.0, + "grad_norm": 2.4997784024684244, + "language_loss": 0.70455325, + "learning_rate": 2.056857268913419e-06, + "loss": 0.72657526, + "num_input_tokens_seen": 90786845, + "step": 4210, + "time_per_iteration": 2.8192930221557617 + }, + { + "auxiliary_loss_clip": 0.01175203, + "auxiliary_loss_mlp": 0.01029428, + "balance_loss_clip": 1.05177796, + "balance_loss_mlp": 1.02085686, + "epoch": 0.506342812481212, + "flos": 17558665994880.0, + "grad_norm": 2.711447584674369, + "language_loss": 0.8379029, + "learning_rate": 2.056078607814303e-06, + "loss": 0.85994923, + "num_input_tokens_seen": 90802630, + "step": 4211, + "time_per_iteration": 2.695826530456543 + }, + { + "auxiliary_loss_clip": 0.01174536, + "auxiliary_loss_mlp": 0.01032277, + "balance_loss_clip": 1.05151653, + "balance_loss_mlp": 1.02338958, + "epoch": 0.5064630553718511, + "flos": 23402050519680.0, + "grad_norm": 2.1411597608307997, + "language_loss": 0.78000581, + "learning_rate": 2.055299938208106e-06, + "loss": 0.80207396, + "num_input_tokens_seen": 90823620, + "step": 4212, + "time_per_iteration": 2.729794502258301 + }, + { + "auxiliary_loss_clip": 0.01177144, + "auxiliary_loss_mlp": 0.01034455, + "balance_loss_clip": 1.0508225, + "balance_loss_mlp": 1.0255084, + "epoch": 0.5065832982624903, + "flos": 23987035416960.0, + "grad_norm": 2.331317043631751, + "language_loss": 0.86465287, + "learning_rate": 2.0545212602129526e-06, + "loss": 0.88676882, + "num_input_tokens_seen": 90843475, + "step": 4213, + "time_per_iteration": 2.7707338333129883 + }, + { + "auxiliary_loss_clip": 0.01160715, + "auxiliary_loss_mlp": 0.01031906, + "balance_loss_clip": 1.04759824, + "balance_loss_mlp": 1.02269745, + "epoch": 0.5067035411531293, + "flos": 21503058289920.0, + "grad_norm": 2.0889342716824997, + "language_loss": 0.66007817, + "learning_rate": 2.0537425739469673e-06, + "loss": 0.68200445, + "num_input_tokens_seen": 90862410, + "step": 4214, + "time_per_iteration": 2.7386107444763184 + }, + { + "auxiliary_loss_clip": 0.0106581, + "auxiliary_loss_mlp": 0.01002737, + "balance_loss_clip": 1.01063061, + "balance_loss_mlp": 1.00085342, + "epoch": 0.5068237840437684, + "flos": 65934397687680.0, + "grad_norm": 0.9614948916432962, + "language_loss": 0.5945819, + "learning_rate": 2.052963879528276e-06, + "loss": 0.61526734, + "num_input_tokens_seen": 90922280, + "step": 4215, + "time_per_iteration": 3.2922897338867188 + }, + { + "auxiliary_loss_clip": 0.01172211, + "auxiliary_loss_mlp": 0.01030245, + "balance_loss_clip": 1.04950047, + "balance_loss_mlp": 1.02108431, + "epoch": 0.5069440269344075, + "flos": 27264206626560.0, + "grad_norm": 2.7559992987857287, + "language_loss": 0.76100671, + "learning_rate": 2.052185177075007e-06, + "loss": 0.78303134, + "num_input_tokens_seen": 90941850, + "step": 4216, + "time_per_iteration": 2.7301008701324463 + }, + { + "auxiliary_loss_clip": 0.01176357, + "auxiliary_loss_mlp": 0.01037825, + "balance_loss_clip": 1.04977107, + "balance_loss_mlp": 1.02857411, + "epoch": 0.5070642698250466, + "flos": 23366319465600.0, + "grad_norm": 1.6008777267329537, + "language_loss": 0.82648355, + "learning_rate": 2.051406466705288e-06, + "loss": 0.84862536, + "num_input_tokens_seen": 90961390, + "step": 4217, + "time_per_iteration": 2.6544973850250244 + }, + { + "auxiliary_loss_clip": 0.01179999, + "auxiliary_loss_mlp": 0.01034566, + "balance_loss_clip": 1.04850388, + "balance_loss_mlp": 1.02597106, + "epoch": 0.5071845127156857, + "flos": 20340127560960.0, + "grad_norm": 1.8844982889447266, + "language_loss": 0.8078388, + "learning_rate": 2.0506277485372486e-06, + "loss": 0.82998443, + "num_input_tokens_seen": 90980215, + "step": 4218, + "time_per_iteration": 2.566300392150879 + }, + { + "auxiliary_loss_clip": 0.01166527, + "auxiliary_loss_mlp": 0.0102613, + "balance_loss_clip": 1.04735434, + "balance_loss_mlp": 1.01683772, + "epoch": 0.5073047556063248, + "flos": 12092955022080.0, + "grad_norm": 1.8782173478456545, + "language_loss": 0.66482413, + "learning_rate": 2.04984902268902e-06, + "loss": 0.68675077, + "num_input_tokens_seen": 90997415, + "step": 4219, + "time_per_iteration": 2.7336912155151367 + }, + { + "auxiliary_loss_clip": 0.01178255, + "auxiliary_loss_mlp": 0.01029645, + "balance_loss_clip": 1.04784393, + "balance_loss_mlp": 1.02002478, + "epoch": 0.5074249984969639, + "flos": 19682854542720.0, + "grad_norm": 2.4752405076441746, + "language_loss": 0.75947368, + "learning_rate": 2.0490702892787345e-06, + "loss": 0.78155267, + "num_input_tokens_seen": 91016475, + "step": 4220, + "time_per_iteration": 2.65535044670105 + }, + { + "auxiliary_loss_clip": 0.01162866, + "auxiliary_loss_mlp": 0.01025195, + "balance_loss_clip": 1.04517245, + "balance_loss_mlp": 1.01670146, + "epoch": 0.5075452413876029, + "flos": 28765703975040.0, + "grad_norm": 1.567792366459725, + "language_loss": 0.62245834, + "learning_rate": 2.0482915484245246e-06, + "loss": 0.64433897, + "num_input_tokens_seen": 91038095, + "step": 4221, + "time_per_iteration": 3.4495136737823486 + }, + { + "auxiliary_loss_clip": 0.01139096, + "auxiliary_loss_mlp": 0.0103175, + "balance_loss_clip": 1.04482901, + "balance_loss_mlp": 1.02202892, + "epoch": 0.5076654842782421, + "flos": 20339445202560.0, + "grad_norm": 2.303467148709195, + "language_loss": 0.8414799, + "learning_rate": 2.047512800244526e-06, + "loss": 0.86318833, + "num_input_tokens_seen": 91053360, + "step": 4222, + "time_per_iteration": 2.685127019882202 + }, + { + "auxiliary_loss_clip": 0.01174749, + "auxiliary_loss_mlp": 0.01036085, + "balance_loss_clip": 1.05184698, + "balance_loss_mlp": 1.02718043, + "epoch": 0.5077857271688812, + "flos": 26359653404160.0, + "grad_norm": 2.117136215098064, + "language_loss": 0.78918821, + "learning_rate": 2.046734044856873e-06, + "loss": 0.81129658, + "num_input_tokens_seen": 91072770, + "step": 4223, + "time_per_iteration": 3.446122169494629 + }, + { + "auxiliary_loss_clip": 0.01170044, + "auxiliary_loss_mlp": 0.01032015, + "balance_loss_clip": 1.04854417, + "balance_loss_mlp": 1.02333629, + "epoch": 0.5079059700595202, + "flos": 21798962530560.0, + "grad_norm": 2.0349878412764992, + "language_loss": 0.81443584, + "learning_rate": 2.045955282379702e-06, + "loss": 0.83645642, + "num_input_tokens_seen": 91091430, + "step": 4224, + "time_per_iteration": 2.585031032562256 + }, + { + "auxiliary_loss_clip": 0.0117099, + "auxiliary_loss_mlp": 0.01029964, + "balance_loss_clip": 1.04708004, + "balance_loss_mlp": 1.02148175, + "epoch": 0.5080262129501594, + "flos": 13187943175680.0, + "grad_norm": 2.5278880282076583, + "language_loss": 0.75910699, + "learning_rate": 2.045176512931152e-06, + "loss": 0.78111649, + "num_input_tokens_seen": 91106060, + "step": 4225, + "time_per_iteration": 3.2940587997436523 + }, + { + "auxiliary_loss_clip": 0.01153331, + "auxiliary_loss_mlp": 0.0102811, + "balance_loss_clip": 1.04873323, + "balance_loss_mlp": 1.01950359, + "epoch": 0.5081464558407984, + "flos": 25301473712640.0, + "grad_norm": 1.9274497697846054, + "language_loss": 0.75921959, + "learning_rate": 2.0443977366293604e-06, + "loss": 0.78103393, + "num_input_tokens_seen": 91124100, + "step": 4226, + "time_per_iteration": 2.77270245552063 + }, + { + "auxiliary_loss_clip": 0.01134937, + "auxiliary_loss_mlp": 0.01036365, + "balance_loss_clip": 1.0434227, + "balance_loss_mlp": 1.02688158, + "epoch": 0.5082666987314375, + "flos": 30951226995840.0, + "grad_norm": 2.3525402175949215, + "language_loss": 0.76822162, + "learning_rate": 2.043618953592468e-06, + "loss": 0.78993464, + "num_input_tokens_seen": 91146555, + "step": 4227, + "time_per_iteration": 3.085200071334839 + }, + { + "auxiliary_loss_clip": 0.01167107, + "auxiliary_loss_mlp": 0.01029923, + "balance_loss_clip": 1.05147636, + "balance_loss_mlp": 1.02007663, + "epoch": 0.5083869416220766, + "flos": 19682495406720.0, + "grad_norm": 1.6731025276385654, + "language_loss": 0.81163251, + "learning_rate": 2.0428401639386144e-06, + "loss": 0.83360285, + "num_input_tokens_seen": 91167120, + "step": 4228, + "time_per_iteration": 2.827313184738159 + }, + { + "auxiliary_loss_clip": 0.01053179, + "auxiliary_loss_mlp": 0.01005074, + "balance_loss_clip": 1.01009774, + "balance_loss_mlp": 1.00335753, + "epoch": 0.5085071845127157, + "flos": 71817535589760.0, + "grad_norm": 0.822540179297134, + "language_loss": 0.58148682, + "learning_rate": 2.042061367785943e-06, + "loss": 0.60206926, + "num_input_tokens_seen": 91220260, + "step": 4229, + "time_per_iteration": 3.3883793354034424 + }, + { + "auxiliary_loss_clip": 0.01153663, + "auxiliary_loss_mlp": 0.0102989, + "balance_loss_clip": 1.04334664, + "balance_loss_mlp": 1.0209074, + "epoch": 0.5086274274033548, + "flos": 35951608252800.0, + "grad_norm": 2.2157716015245037, + "language_loss": 0.75339043, + "learning_rate": 2.041282565252594e-06, + "loss": 0.775226, + "num_input_tokens_seen": 91240425, + "step": 4230, + "time_per_iteration": 3.0963528156280518 + }, + { + "auxiliary_loss_clip": 0.01152613, + "auxiliary_loss_mlp": 0.01027417, + "balance_loss_clip": 1.04515076, + "balance_loss_mlp": 1.01817214, + "epoch": 0.5087476702939938, + "flos": 23513732881920.0, + "grad_norm": 2.04013340296295, + "language_loss": 0.77287167, + "learning_rate": 2.040503756456714e-06, + "loss": 0.79467201, + "num_input_tokens_seen": 91259635, + "step": 4231, + "time_per_iteration": 2.79891300201416 + }, + { + "auxiliary_loss_clip": 0.01167118, + "auxiliary_loss_mlp": 0.01034479, + "balance_loss_clip": 1.04822516, + "balance_loss_mlp": 1.02546728, + "epoch": 0.508867913184633, + "flos": 15122091841920.0, + "grad_norm": 2.176273836190789, + "language_loss": 0.786017, + "learning_rate": 2.0397249415164456e-06, + "loss": 0.80803293, + "num_input_tokens_seen": 91276990, + "step": 4232, + "time_per_iteration": 2.729506492614746 + }, + { + "auxiliary_loss_clip": 0.01151975, + "auxiliary_loss_mlp": 0.01035469, + "balance_loss_clip": 1.04224348, + "balance_loss_mlp": 1.02659965, + "epoch": 0.508988156075272, + "flos": 25885309374720.0, + "grad_norm": 1.7614247796608316, + "language_loss": 0.80170619, + "learning_rate": 2.0389461205499354e-06, + "loss": 0.82358062, + "num_input_tokens_seen": 91296125, + "step": 4233, + "time_per_iteration": 2.7851572036743164 + }, + { + "auxiliary_loss_clip": 0.01153255, + "auxiliary_loss_mlp": 0.0103097, + "balance_loss_clip": 1.04407573, + "balance_loss_mlp": 1.02298903, + "epoch": 0.5091083989659111, + "flos": 13844857057920.0, + "grad_norm": 1.9109024925599702, + "language_loss": 0.7362054, + "learning_rate": 2.03816729367533e-06, + "loss": 0.75804764, + "num_input_tokens_seen": 91314280, + "step": 4234, + "time_per_iteration": 2.8057334423065186 + }, + { + "auxiliary_loss_clip": 0.01167358, + "auxiliary_loss_mlp": 0.01034274, + "balance_loss_clip": 1.05046034, + "balance_loss_mlp": 1.02514875, + "epoch": 0.5092286418565503, + "flos": 21104881050240.0, + "grad_norm": 2.081484095626408, + "language_loss": 0.71468341, + "learning_rate": 2.0373884610107765e-06, + "loss": 0.7366997, + "num_input_tokens_seen": 91334595, + "step": 4235, + "time_per_iteration": 2.7112514972686768 + }, + { + "auxiliary_loss_clip": 0.01175466, + "auxiliary_loss_mlp": 0.0102645, + "balance_loss_clip": 1.04708946, + "balance_loss_mlp": 1.01726508, + "epoch": 0.5093488847471893, + "flos": 18621298972800.0, + "grad_norm": 2.442996951938675, + "language_loss": 0.69608003, + "learning_rate": 2.0366096226744225e-06, + "loss": 0.71809918, + "num_input_tokens_seen": 91349790, + "step": 4236, + "time_per_iteration": 2.7412002086639404 + }, + { + "auxiliary_loss_clip": 0.01162653, + "auxiliary_loss_mlp": 0.01042371, + "balance_loss_clip": 1.04744959, + "balance_loss_mlp": 1.03337622, + "epoch": 0.5094691276378284, + "flos": 23803783205760.0, + "grad_norm": 1.761257877448553, + "language_loss": 0.7666871, + "learning_rate": 2.035830778784418e-06, + "loss": 0.78873736, + "num_input_tokens_seen": 91370465, + "step": 4237, + "time_per_iteration": 2.767164707183838 + }, + { + "auxiliary_loss_clip": 0.01163485, + "auxiliary_loss_mlp": 0.01032276, + "balance_loss_clip": 1.05332136, + "balance_loss_mlp": 1.02380013, + "epoch": 0.5095893705284675, + "flos": 17420410546560.0, + "grad_norm": 4.0744184017294245, + "language_loss": 0.79800344, + "learning_rate": 2.0350519294589134e-06, + "loss": 0.81996101, + "num_input_tokens_seen": 91388505, + "step": 4238, + "time_per_iteration": 2.7671186923980713 + }, + { + "auxiliary_loss_clip": 0.0114324, + "auxiliary_loss_mlp": 0.01028729, + "balance_loss_clip": 1.04348421, + "balance_loss_mlp": 1.01962113, + "epoch": 0.5097096134191066, + "flos": 25849362839040.0, + "grad_norm": 1.6827871804890202, + "language_loss": 0.82949501, + "learning_rate": 2.0342730748160588e-06, + "loss": 0.85121465, + "num_input_tokens_seen": 91408970, + "step": 4239, + "time_per_iteration": 2.9781389236450195 + }, + { + "auxiliary_loss_clip": 0.01163111, + "auxiliary_loss_mlp": 0.01038761, + "balance_loss_clip": 1.04715991, + "balance_loss_mlp": 1.02937949, + "epoch": 0.5098298563097456, + "flos": 27745122844800.0, + "grad_norm": 2.2614516410878105, + "language_loss": 0.70345533, + "learning_rate": 2.033494214974006e-06, + "loss": 0.72547406, + "num_input_tokens_seen": 91430115, + "step": 4240, + "time_per_iteration": 2.768043279647827 + }, + { + "auxiliary_loss_clip": 0.01151723, + "auxiliary_loss_mlp": 0.01034391, + "balance_loss_clip": 1.04636395, + "balance_loss_mlp": 1.02587318, + "epoch": 0.5099500992003848, + "flos": 21358913011200.0, + "grad_norm": 1.896905757725197, + "language_loss": 0.84132159, + "learning_rate": 2.0327153500509067e-06, + "loss": 0.86318272, + "num_input_tokens_seen": 91449140, + "step": 4241, + "time_per_iteration": 2.84678316116333 + }, + { + "auxiliary_loss_clip": 0.01163932, + "auxiliary_loss_mlp": 0.010269, + "balance_loss_clip": 1.0492599, + "balance_loss_mlp": 1.01754236, + "epoch": 0.5100703420910239, + "flos": 19865999013120.0, + "grad_norm": 2.861822950374594, + "language_loss": 0.85042953, + "learning_rate": 2.031936480164916e-06, + "loss": 0.87233782, + "num_input_tokens_seen": 91466880, + "step": 4242, + "time_per_iteration": 2.72786808013916 + }, + { + "auxiliary_loss_clip": 0.01162799, + "auxiliary_loss_mlp": 0.01031614, + "balance_loss_clip": 1.05063462, + "balance_loss_mlp": 1.02235174, + "epoch": 0.5101905849816629, + "flos": 24648797635200.0, + "grad_norm": 2.1582412999743026, + "language_loss": 0.79986376, + "learning_rate": 2.0311576054341857e-06, + "loss": 0.82180786, + "num_input_tokens_seen": 91487495, + "step": 4243, + "time_per_iteration": 2.8162736892700195 + }, + { + "auxiliary_loss_clip": 0.01184241, + "auxiliary_loss_mlp": 0.01028688, + "balance_loss_clip": 1.05360067, + "balance_loss_mlp": 1.0200212, + "epoch": 0.5103108278723021, + "flos": 22930076787840.0, + "grad_norm": 1.6959697509926357, + "language_loss": 0.62482929, + "learning_rate": 2.0303787259768715e-06, + "loss": 0.64695859, + "num_input_tokens_seen": 91508395, + "step": 4244, + "time_per_iteration": 2.6998937129974365 + }, + { + "auxiliary_loss_clip": 0.0116762, + "auxiliary_loss_mlp": 0.01033666, + "balance_loss_clip": 1.05148029, + "balance_loss_mlp": 1.02519011, + "epoch": 0.5104310707629411, + "flos": 21506613736320.0, + "grad_norm": 2.438493877484053, + "language_loss": 0.68588662, + "learning_rate": 2.0295998419111294e-06, + "loss": 0.70789945, + "num_input_tokens_seen": 91525685, + "step": 4245, + "time_per_iteration": 2.8247902393341064 + }, + { + "auxiliary_loss_clip": 0.01136478, + "auxiliary_loss_mlp": 0.01034195, + "balance_loss_clip": 1.04245555, + "balance_loss_mlp": 1.02493215, + "epoch": 0.5105513136535802, + "flos": 14903180403840.0, + "grad_norm": 2.466662028246785, + "language_loss": 0.73580039, + "learning_rate": 2.028820953355115e-06, + "loss": 0.75750709, + "num_input_tokens_seen": 91543785, + "step": 4246, + "time_per_iteration": 2.932934045791626 + }, + { + "auxiliary_loss_clip": 0.01170291, + "auxiliary_loss_mlp": 0.01032369, + "balance_loss_clip": 1.04940391, + "balance_loss_mlp": 1.02257013, + "epoch": 0.5106715565442194, + "flos": 22602212421120.0, + "grad_norm": 2.2299068382741933, + "language_loss": 0.78779042, + "learning_rate": 2.0280420604269834e-06, + "loss": 0.80981702, + "num_input_tokens_seen": 91563325, + "step": 4247, + "time_per_iteration": 2.750157356262207 + }, + { + "auxiliary_loss_clip": 0.01064243, + "auxiliary_loss_mlp": 0.01003709, + "balance_loss_clip": 1.01183116, + "balance_loss_mlp": 1.0019803, + "epoch": 0.5107917994348584, + "flos": 71027645558400.0, + "grad_norm": 0.708559396757867, + "language_loss": 0.58937818, + "learning_rate": 2.027263163244895e-06, + "loss": 0.61005771, + "num_input_tokens_seen": 91632450, + "step": 4248, + "time_per_iteration": 5.199890851974487 + }, + { + "auxiliary_loss_clip": 0.01168158, + "auxiliary_loss_mlp": 0.01025529, + "balance_loss_clip": 1.0481447, + "balance_loss_mlp": 1.01700604, + "epoch": 0.5109120423254975, + "flos": 24827416992000.0, + "grad_norm": 2.1644607099169253, + "language_loss": 0.74368507, + "learning_rate": 2.026484261927005e-06, + "loss": 0.76562202, + "num_input_tokens_seen": 91651945, + "step": 4249, + "time_per_iteration": 3.7427024841308594 + }, + { + "auxiliary_loss_clip": 0.01182145, + "auxiliary_loss_mlp": 0.01028442, + "balance_loss_clip": 1.05408144, + "balance_loss_mlp": 1.01901889, + "epoch": 0.5110322852161366, + "flos": 21247661612160.0, + "grad_norm": 2.399180888212173, + "language_loss": 0.73979008, + "learning_rate": 2.025705356591475e-06, + "loss": 0.76189601, + "num_input_tokens_seen": 91669635, + "step": 4250, + "time_per_iteration": 2.6901159286499023 + }, + { + "auxiliary_loss_clip": 0.01050732, + "auxiliary_loss_mlp": 0.00876577, + "balance_loss_clip": 1.01057601, + "balance_loss_mlp": 1.00125265, + "epoch": 0.5111525281067757, + "flos": 66457114358400.0, + "grad_norm": 0.7515462710222847, + "language_loss": 0.57910812, + "learning_rate": 2.024926447356462e-06, + "loss": 0.59838122, + "num_input_tokens_seen": 91731920, + "step": 4251, + "time_per_iteration": 4.32967472076416 + }, + { + "auxiliary_loss_clip": 0.0117502, + "auxiliary_loss_mlp": 0.01028198, + "balance_loss_clip": 1.05031002, + "balance_loss_mlp": 1.0193404, + "epoch": 0.5112727709974147, + "flos": 14866731077760.0, + "grad_norm": 2.0505052020363186, + "language_loss": 0.78848392, + "learning_rate": 2.024147534340127e-06, + "loss": 0.81051612, + "num_input_tokens_seen": 91749780, + "step": 4252, + "time_per_iteration": 2.8573107719421387 + }, + { + "auxiliary_loss_clip": 0.01153405, + "auxiliary_loss_mlp": 0.01029264, + "balance_loss_clip": 1.0439204, + "balance_loss_mlp": 1.02091908, + "epoch": 0.5113930138880539, + "flos": 21177600134400.0, + "grad_norm": 1.876606670323134, + "language_loss": 0.79718786, + "learning_rate": 2.02336861766063e-06, + "loss": 0.81901455, + "num_input_tokens_seen": 91768840, + "step": 4253, + "time_per_iteration": 2.7258431911468506 + }, + { + "auxiliary_loss_clip": 0.0118294, + "auxiliary_loss_mlp": 0.01032672, + "balance_loss_clip": 1.053473, + "balance_loss_mlp": 1.02309322, + "epoch": 0.511513256778693, + "flos": 20409111630720.0, + "grad_norm": 2.004929960087418, + "language_loss": 0.78818297, + "learning_rate": 2.0225896974361327e-06, + "loss": 0.81033909, + "num_input_tokens_seen": 91788945, + "step": 4254, + "time_per_iteration": 2.8247978687286377 + }, + { + "auxiliary_loss_clip": 0.01051329, + "auxiliary_loss_mlp": 0.01004006, + "balance_loss_clip": 1.01144505, + "balance_loss_mlp": 1.00234878, + "epoch": 0.511633499669332, + "flos": 69879975131520.0, + "grad_norm": 0.8637880485604503, + "language_loss": 0.59945422, + "learning_rate": 2.0218107737847962e-06, + "loss": 0.62000757, + "num_input_tokens_seen": 91850990, + "step": 4255, + "time_per_iteration": 3.4269773960113525 + }, + { + "auxiliary_loss_clip": 0.01181952, + "auxiliary_loss_mlp": 0.01031105, + "balance_loss_clip": 1.05193925, + "balance_loss_mlp": 1.02217627, + "epoch": 0.5117537425599712, + "flos": 24097855852800.0, + "grad_norm": 1.86823699298615, + "language_loss": 0.74599576, + "learning_rate": 2.0210318468247826e-06, + "loss": 0.76812631, + "num_input_tokens_seen": 91869960, + "step": 4256, + "time_per_iteration": 2.7036213874816895 + }, + { + "auxiliary_loss_clip": 0.01162793, + "auxiliary_loss_mlp": 0.01031816, + "balance_loss_clip": 1.04641533, + "balance_loss_mlp": 1.02359617, + "epoch": 0.5118739854506102, + "flos": 20959550622720.0, + "grad_norm": 1.809413332971353, + "language_loss": 0.81382543, + "learning_rate": 2.020252916674255e-06, + "loss": 0.83577156, + "num_input_tokens_seen": 91889075, + "step": 4257, + "time_per_iteration": 2.7641406059265137 + }, + { + "auxiliary_loss_clip": 0.01173365, + "auxiliary_loss_mlp": 0.01030267, + "balance_loss_clip": 1.04784822, + "balance_loss_mlp": 1.02186835, + "epoch": 0.5119942283412493, + "flos": 17457326749440.0, + "grad_norm": 1.8135064246996901, + "language_loss": 0.81182551, + "learning_rate": 2.019473983451375e-06, + "loss": 0.83386177, + "num_input_tokens_seen": 91907495, + "step": 4258, + "time_per_iteration": 2.827146291732788 + }, + { + "auxiliary_loss_clip": 0.01154485, + "auxiliary_loss_mlp": 0.01029977, + "balance_loss_clip": 1.04652667, + "balance_loss_mlp": 1.02136397, + "epoch": 0.5121144712318885, + "flos": 21066743784960.0, + "grad_norm": 1.8571845002195324, + "language_loss": 0.71153152, + "learning_rate": 2.0186950472743076e-06, + "loss": 0.73337615, + "num_input_tokens_seen": 91927400, + "step": 4259, + "time_per_iteration": 2.8839776515960693 + }, + { + "auxiliary_loss_clip": 0.01181157, + "auxiliary_loss_mlp": 0.01028223, + "balance_loss_clip": 1.04933774, + "balance_loss_mlp": 1.01884162, + "epoch": 0.5122347141225275, + "flos": 19860791541120.0, + "grad_norm": 1.8546694748169823, + "language_loss": 0.7398653, + "learning_rate": 2.0179161082612162e-06, + "loss": 0.7619592, + "num_input_tokens_seen": 91946790, + "step": 4260, + "time_per_iteration": 2.6948297023773193 + }, + { + "auxiliary_loss_clip": 0.01160064, + "auxiliary_loss_mlp": 0.01023249, + "balance_loss_clip": 1.04618955, + "balance_loss_mlp": 1.01460063, + "epoch": 0.5123549570131666, + "flos": 22528487756160.0, + "grad_norm": 2.1708557066929766, + "language_loss": 0.72902548, + "learning_rate": 2.017137166530266e-06, + "loss": 0.75085855, + "num_input_tokens_seen": 91966325, + "step": 4261, + "time_per_iteration": 2.818574905395508 + }, + { + "auxiliary_loss_clip": 0.01169121, + "auxiliary_loss_mlp": 0.0103651, + "balance_loss_clip": 1.0499779, + "balance_loss_mlp": 1.02739656, + "epoch": 0.5124751999038056, + "flos": 20333375804160.0, + "grad_norm": 1.9845937484842773, + "language_loss": 0.80180055, + "learning_rate": 2.0163582221996213e-06, + "loss": 0.82385683, + "num_input_tokens_seen": 91984700, + "step": 4262, + "time_per_iteration": 2.8377411365509033 + }, + { + "auxiliary_loss_clip": 0.01163782, + "auxiliary_loss_mlp": 0.01029734, + "balance_loss_clip": 1.0485816, + "balance_loss_mlp": 1.02075791, + "epoch": 0.5125954427944448, + "flos": 39785970211200.0, + "grad_norm": 2.298774023885002, + "language_loss": 0.67787427, + "learning_rate": 2.015579275387446e-06, + "loss": 0.69980943, + "num_input_tokens_seen": 92010020, + "step": 4263, + "time_per_iteration": 2.9594340324401855 + }, + { + "auxiliary_loss_clip": 0.01154891, + "auxiliary_loss_mlp": 0.01031946, + "balance_loss_clip": 1.04827178, + "balance_loss_mlp": 1.02350593, + "epoch": 0.5127156856850839, + "flos": 29205394358400.0, + "grad_norm": 1.825341858314237, + "language_loss": 0.68217504, + "learning_rate": 2.0148003262119085e-06, + "loss": 0.70404339, + "num_input_tokens_seen": 92030990, + "step": 4264, + "time_per_iteration": 2.880199670791626 + }, + { + "auxiliary_loss_clip": 0.01152922, + "auxiliary_loss_mlp": 0.01034407, + "balance_loss_clip": 1.0463891, + "balance_loss_mlp": 1.02515006, + "epoch": 0.5128359285757229, + "flos": 13553693412480.0, + "grad_norm": 2.082389532821139, + "language_loss": 0.7687974, + "learning_rate": 2.0140213747911728e-06, + "loss": 0.79067075, + "num_input_tokens_seen": 92049525, + "step": 4265, + "time_per_iteration": 2.8252248764038086 + }, + { + "auxiliary_loss_clip": 0.01153045, + "auxiliary_loss_mlp": 0.01029771, + "balance_loss_clip": 1.04805863, + "balance_loss_mlp": 1.02112222, + "epoch": 0.5129561714663621, + "flos": 25192089820800.0, + "grad_norm": 1.9745923164564676, + "language_loss": 0.80307239, + "learning_rate": 2.013242421243406e-06, + "loss": 0.82490051, + "num_input_tokens_seen": 92068430, + "step": 4266, + "time_per_iteration": 2.8758561611175537 + }, + { + "auxiliary_loss_clip": 0.01148582, + "auxiliary_loss_mlp": 0.01033006, + "balance_loss_clip": 1.04996943, + "balance_loss_mlp": 1.02498364, + "epoch": 0.5130764143570011, + "flos": 18150223080960.0, + "grad_norm": 2.0704523725054487, + "language_loss": 0.79070228, + "learning_rate": 2.012463465686774e-06, + "loss": 0.81251812, + "num_input_tokens_seen": 92088180, + "step": 4267, + "time_per_iteration": 2.832099199295044 + }, + { + "auxiliary_loss_clip": 0.0104106, + "auxiliary_loss_mlp": 0.01003994, + "balance_loss_clip": 1.01404607, + "balance_loss_mlp": 1.00182426, + "epoch": 0.5131966572476402, + "flos": 59794896418560.0, + "grad_norm": 0.7707732161349914, + "language_loss": 0.54801214, + "learning_rate": 2.0116845082394446e-06, + "loss": 0.56846267, + "num_input_tokens_seen": 92153015, + "step": 4268, + "time_per_iteration": 3.3435239791870117 + }, + { + "auxiliary_loss_clip": 0.01174193, + "auxiliary_loss_mlp": 0.01030249, + "balance_loss_clip": 1.04769731, + "balance_loss_mlp": 1.02119505, + "epoch": 0.5133169001382794, + "flos": 18515219132160.0, + "grad_norm": 2.0510792246241696, + "language_loss": 0.78806663, + "learning_rate": 2.0109055490195836e-06, + "loss": 0.81011105, + "num_input_tokens_seen": 92171470, + "step": 4269, + "time_per_iteration": 2.7910053730010986 + }, + { + "auxiliary_loss_clip": 0.01134666, + "auxiliary_loss_mlp": 0.01030204, + "balance_loss_clip": 1.03658438, + "balance_loss_mlp": 1.02126348, + "epoch": 0.5134371430289184, + "flos": 15523537219200.0, + "grad_norm": 1.8688722662498836, + "language_loss": 0.64411038, + "learning_rate": 2.0101265881453605e-06, + "loss": 0.66575909, + "num_input_tokens_seen": 92189945, + "step": 4270, + "time_per_iteration": 2.86266827583313 + }, + { + "auxiliary_loss_clip": 0.01158518, + "auxiliary_loss_mlp": 0.01030982, + "balance_loss_clip": 1.04989314, + "balance_loss_mlp": 1.02282858, + "epoch": 0.5135573859195575, + "flos": 21433786911360.0, + "grad_norm": 2.061200086020447, + "language_loss": 0.77898359, + "learning_rate": 2.009347625734941e-06, + "loss": 0.80087858, + "num_input_tokens_seen": 92209855, + "step": 4271, + "time_per_iteration": 2.8699967861175537 + }, + { + "auxiliary_loss_clip": 0.01186226, + "auxiliary_loss_mlp": 0.01039115, + "balance_loss_clip": 1.05298233, + "balance_loss_mlp": 1.02989984, + "epoch": 0.5136776288101966, + "flos": 17712651600000.0, + "grad_norm": 2.233731090641053, + "language_loss": 0.75278032, + "learning_rate": 2.0085686619064954e-06, + "loss": 0.77503371, + "num_input_tokens_seen": 92226295, + "step": 4272, + "time_per_iteration": 2.6300554275512695 + }, + { + "auxiliary_loss_clip": 0.01181741, + "auxiliary_loss_mlp": 0.01032232, + "balance_loss_clip": 1.05463386, + "balance_loss_mlp": 1.02304149, + "epoch": 0.5137978717008357, + "flos": 16581680997120.0, + "grad_norm": 2.071106807749317, + "language_loss": 0.8276, + "learning_rate": 2.00778969677819e-06, + "loss": 0.84973973, + "num_input_tokens_seen": 92243330, + "step": 4273, + "time_per_iteration": 3.7342755794525146 + }, + { + "auxiliary_loss_clip": 0.01160194, + "auxiliary_loss_mlp": 0.01032202, + "balance_loss_clip": 1.04631555, + "balance_loss_mlp": 1.02339268, + "epoch": 0.5139181145914747, + "flos": 20668243322880.0, + "grad_norm": 2.4173740095602, + "language_loss": 0.63868105, + "learning_rate": 2.0070107304681934e-06, + "loss": 0.66060507, + "num_input_tokens_seen": 92262285, + "step": 4274, + "time_per_iteration": 3.75471830368042 + }, + { + "auxiliary_loss_clip": 0.01155249, + "auxiliary_loss_mlp": 0.01030935, + "balance_loss_clip": 1.04989946, + "balance_loss_mlp": 1.02295423, + "epoch": 0.5140383574821139, + "flos": 32926996546560.0, + "grad_norm": 1.7716106627302588, + "language_loss": 0.78285289, + "learning_rate": 2.006231763094675e-06, + "loss": 0.80471468, + "num_input_tokens_seen": 92283305, + "step": 4275, + "time_per_iteration": 3.845953941345215 + }, + { + "auxiliary_loss_clip": 0.01160063, + "auxiliary_loss_mlp": 0.01027904, + "balance_loss_clip": 1.05207229, + "balance_loss_mlp": 1.01919532, + "epoch": 0.514158600372753, + "flos": 19537093152000.0, + "grad_norm": 1.9657920657017616, + "language_loss": 0.87216592, + "learning_rate": 2.0054527947758027e-06, + "loss": 0.89404559, + "num_input_tokens_seen": 92302105, + "step": 4276, + "time_per_iteration": 2.75028395652771 + }, + { + "auxiliary_loss_clip": 0.01064164, + "auxiliary_loss_mlp": 0.01004218, + "balance_loss_clip": 1.01188993, + "balance_loss_mlp": 1.00248957, + "epoch": 0.514278843263392, + "flos": 62523855279360.0, + "grad_norm": 0.7225784325456945, + "language_loss": 0.55877292, + "learning_rate": 2.004673825629746e-06, + "loss": 0.57945681, + "num_input_tokens_seen": 92362885, + "step": 4277, + "time_per_iteration": 4.335436820983887 + }, + { + "auxiliary_loss_clip": 0.01158391, + "auxiliary_loss_mlp": 0.01028431, + "balance_loss_clip": 1.04596591, + "balance_loss_mlp": 1.01945436, + "epoch": 0.5143990861540312, + "flos": 25882328545920.0, + "grad_norm": 1.539196716386072, + "language_loss": 0.72351122, + "learning_rate": 2.0038948557746744e-06, + "loss": 0.74537945, + "num_input_tokens_seen": 92384740, + "step": 4278, + "time_per_iteration": 2.865025043487549 + }, + { + "auxiliary_loss_clip": 0.01170359, + "auxiliary_loss_mlp": 0.01027312, + "balance_loss_clip": 1.04983282, + "balance_loss_mlp": 1.01842463, + "epoch": 0.5145193290446702, + "flos": 23330660238720.0, + "grad_norm": 1.7599495347813485, + "language_loss": 0.75237346, + "learning_rate": 2.0031158853287558e-06, + "loss": 0.77435017, + "num_input_tokens_seen": 92405175, + "step": 4279, + "time_per_iteration": 2.6598572731018066 + }, + { + "auxiliary_loss_clip": 0.01164288, + "auxiliary_loss_mlp": 0.01031039, + "balance_loss_clip": 1.05225158, + "balance_loss_mlp": 1.0221225, + "epoch": 0.5146395719353093, + "flos": 22856603518080.0, + "grad_norm": 1.92271151774532, + "language_loss": 0.70338607, + "learning_rate": 2.0023369144101593e-06, + "loss": 0.72533935, + "num_input_tokens_seen": 92423345, + "step": 4280, + "time_per_iteration": 2.750562906265259 + }, + { + "auxiliary_loss_clip": 0.011543, + "auxiliary_loss_mlp": 0.01027938, + "balance_loss_clip": 1.04619718, + "balance_loss_mlp": 1.01943254, + "epoch": 0.5147598148259485, + "flos": 26391577616640.0, + "grad_norm": 1.917967207772933, + "language_loss": 0.77049994, + "learning_rate": 2.0015579431370555e-06, + "loss": 0.7923224, + "num_input_tokens_seen": 92445025, + "step": 4281, + "time_per_iteration": 2.7635369300842285 + }, + { + "auxiliary_loss_clip": 0.01168081, + "auxiliary_loss_mlp": 0.01032458, + "balance_loss_clip": 1.04874897, + "balance_loss_mlp": 1.02343404, + "epoch": 0.5148800577165875, + "flos": 29965694561280.0, + "grad_norm": 2.700383601996055, + "language_loss": 0.70089793, + "learning_rate": 2.000778971627612e-06, + "loss": 0.72290331, + "num_input_tokens_seen": 92464490, + "step": 4282, + "time_per_iteration": 2.8805580139160156 + }, + { + "auxiliary_loss_clip": 0.01155393, + "auxiliary_loss_mlp": 0.01033119, + "balance_loss_clip": 1.04575968, + "balance_loss_mlp": 1.02382016, + "epoch": 0.5150003006072266, + "flos": 17931383470080.0, + "grad_norm": 2.1001287221736873, + "language_loss": 0.90523565, + "learning_rate": 2e-06, + "loss": 0.92712075, + "num_input_tokens_seen": 92482085, + "step": 4283, + "time_per_iteration": 2.834975242614746 + }, + { + "auxiliary_loss_clip": 0.01180408, + "auxiliary_loss_mlp": 0.01028421, + "balance_loss_clip": 1.05088544, + "balance_loss_mlp": 1.02015936, + "epoch": 0.5151205434978657, + "flos": 18478733892480.0, + "grad_norm": 1.640939025071156, + "language_loss": 0.85516036, + "learning_rate": 1.9992210283723878e-06, + "loss": 0.87724864, + "num_input_tokens_seen": 92499325, + "step": 4284, + "time_per_iteration": 2.729879140853882 + }, + { + "auxiliary_loss_clip": 0.01182426, + "auxiliary_loss_mlp": 0.01034378, + "balance_loss_clip": 1.05267382, + "balance_loss_mlp": 1.02593827, + "epoch": 0.5152407863885048, + "flos": 25341263003520.0, + "grad_norm": 2.337112711997568, + "language_loss": 0.79311681, + "learning_rate": 1.9984420568629448e-06, + "loss": 0.81528485, + "num_input_tokens_seen": 92522090, + "step": 4285, + "time_per_iteration": 2.749319553375244 + }, + { + "auxiliary_loss_clip": 0.01175682, + "auxiliary_loss_mlp": 0.01031241, + "balance_loss_clip": 1.05061078, + "balance_loss_mlp": 1.02227044, + "epoch": 0.5153610292791438, + "flos": 18329740277760.0, + "grad_norm": 2.0982518271226955, + "language_loss": 0.78069335, + "learning_rate": 1.9976630855898405e-06, + "loss": 0.80276251, + "num_input_tokens_seen": 92539845, + "step": 4286, + "time_per_iteration": 2.7116570472717285 + }, + { + "auxiliary_loss_clip": 0.01154562, + "auxiliary_loss_mlp": 0.01032019, + "balance_loss_clip": 1.04065895, + "balance_loss_mlp": 1.02308452, + "epoch": 0.515481272169783, + "flos": 30409945971840.0, + "grad_norm": 2.047477470923533, + "language_loss": 0.74309027, + "learning_rate": 1.9968841146712445e-06, + "loss": 0.764956, + "num_input_tokens_seen": 92559460, + "step": 4287, + "time_per_iteration": 2.846449851989746 + }, + { + "auxiliary_loss_clip": 0.01132433, + "auxiliary_loss_mlp": 0.00887088, + "balance_loss_clip": 1.04336882, + "balance_loss_mlp": 1.00101924, + "epoch": 0.5156015150604221, + "flos": 23037305863680.0, + "grad_norm": 1.5868971341441909, + "language_loss": 0.71516752, + "learning_rate": 1.996105144225326e-06, + "loss": 0.73536277, + "num_input_tokens_seen": 92579695, + "step": 4288, + "time_per_iteration": 2.9223899841308594 + }, + { + "auxiliary_loss_clip": 0.01170046, + "auxiliary_loss_mlp": 0.01030363, + "balance_loss_clip": 1.05085421, + "balance_loss_mlp": 1.02180362, + "epoch": 0.5157217579510611, + "flos": 17858556645120.0, + "grad_norm": 2.357492214547712, + "language_loss": 0.786107, + "learning_rate": 1.995326174370254e-06, + "loss": 0.80811113, + "num_input_tokens_seen": 92598795, + "step": 4289, + "time_per_iteration": 2.70418381690979 + }, + { + "auxiliary_loss_clip": 0.01167281, + "auxiliary_loss_mlp": 0.00885652, + "balance_loss_clip": 1.04710507, + "balance_loss_mlp": 1.00105977, + "epoch": 0.5158420008417003, + "flos": 19171486569600.0, + "grad_norm": 1.6171141314769342, + "language_loss": 0.72885895, + "learning_rate": 1.994547205224197e-06, + "loss": 0.74938834, + "num_input_tokens_seen": 92617700, + "step": 4290, + "time_per_iteration": 2.733555555343628 + }, + { + "auxiliary_loss_clip": 0.01158618, + "auxiliary_loss_mlp": 0.0102559, + "balance_loss_clip": 1.04861748, + "balance_loss_mlp": 1.01701832, + "epoch": 0.5159622437323393, + "flos": 22419534827520.0, + "grad_norm": 2.114188710112046, + "language_loss": 0.6763072, + "learning_rate": 1.993768236905325e-06, + "loss": 0.69814932, + "num_input_tokens_seen": 92638370, + "step": 4291, + "time_per_iteration": 2.70135498046875 + }, + { + "auxiliary_loss_clip": 0.01159342, + "auxiliary_loss_mlp": 0.01024918, + "balance_loss_clip": 1.04675412, + "balance_loss_mlp": 1.01623344, + "epoch": 0.5160824866229784, + "flos": 24603010773120.0, + "grad_norm": 2.7020068864992637, + "language_loss": 0.66002715, + "learning_rate": 1.992989269531807e-06, + "loss": 0.6818698, + "num_input_tokens_seen": 92657180, + "step": 4292, + "time_per_iteration": 2.7546722888946533 + }, + { + "auxiliary_loss_clip": 0.01159333, + "auxiliary_loss_mlp": 0.01027197, + "balance_loss_clip": 1.04519153, + "balance_loss_mlp": 1.01846457, + "epoch": 0.5162027295136175, + "flos": 18002737837440.0, + "grad_norm": 2.442780909787811, + "language_loss": 0.6767174, + "learning_rate": 1.99221030322181e-06, + "loss": 0.69858265, + "num_input_tokens_seen": 92673985, + "step": 4293, + "time_per_iteration": 2.7586772441864014 + }, + { + "auxiliary_loss_clip": 0.01164218, + "auxiliary_loss_mlp": 0.01031331, + "balance_loss_clip": 1.04748034, + "balance_loss_mlp": 1.02338612, + "epoch": 0.5163229724042566, + "flos": 27344611221120.0, + "grad_norm": 1.7336846917108721, + "language_loss": 0.80831397, + "learning_rate": 1.991431338093505e-06, + "loss": 0.83026946, + "num_input_tokens_seen": 92696340, + "step": 4294, + "time_per_iteration": 2.7947914600372314 + }, + { + "auxiliary_loss_clip": 0.01162107, + "auxiliary_loss_mlp": 0.0103726, + "balance_loss_clip": 1.05196583, + "balance_loss_mlp": 1.02923131, + "epoch": 0.5164432152948957, + "flos": 21762764599680.0, + "grad_norm": 1.7839718832342528, + "language_loss": 0.79104424, + "learning_rate": 1.9906523742650587e-06, + "loss": 0.81303793, + "num_input_tokens_seen": 92715200, + "step": 4295, + "time_per_iteration": 2.7959327697753906 + }, + { + "auxiliary_loss_clip": 0.01182337, + "auxiliary_loss_mlp": 0.01028592, + "balance_loss_clip": 1.04953539, + "balance_loss_mlp": 1.01948428, + "epoch": 0.5165634581855347, + "flos": 25550334115200.0, + "grad_norm": 2.0311146840432523, + "language_loss": 0.77637804, + "learning_rate": 1.9898734118546397e-06, + "loss": 0.79848731, + "num_input_tokens_seen": 92735150, + "step": 4296, + "time_per_iteration": 2.763256072998047 + }, + { + "auxiliary_loss_clip": 0.01129409, + "auxiliary_loss_mlp": 0.01037, + "balance_loss_clip": 1.04383898, + "balance_loss_mlp": 1.02847672, + "epoch": 0.5166837010761739, + "flos": 19901191363200.0, + "grad_norm": 1.8181185308051342, + "language_loss": 0.80162239, + "learning_rate": 1.989094450980416e-06, + "loss": 0.82328647, + "num_input_tokens_seen": 92755250, + "step": 4297, + "time_per_iteration": 2.9424681663513184 + }, + { + "auxiliary_loss_clip": 0.01172199, + "auxiliary_loss_mlp": 0.01028732, + "balance_loss_clip": 1.05059052, + "balance_loss_mlp": 1.02004743, + "epoch": 0.516803943966813, + "flos": 26646076454400.0, + "grad_norm": 2.1700451282222426, + "language_loss": 0.76758236, + "learning_rate": 1.9883154917605556e-06, + "loss": 0.78959167, + "num_input_tokens_seen": 92774460, + "step": 4298, + "time_per_iteration": 2.751816987991333 + }, + { + "auxiliary_loss_clip": 0.01180829, + "auxiliary_loss_mlp": 0.01028625, + "balance_loss_clip": 1.05106103, + "balance_loss_mlp": 1.01951194, + "epoch": 0.516924186857452, + "flos": 19682854542720.0, + "grad_norm": 2.186225876506342, + "language_loss": 0.83503997, + "learning_rate": 1.9875365343132262e-06, + "loss": 0.85713446, + "num_input_tokens_seen": 92791580, + "step": 4299, + "time_per_iteration": 3.6517269611358643 + }, + { + "auxiliary_loss_clip": 0.01172646, + "auxiliary_loss_mlp": 0.00886043, + "balance_loss_clip": 1.05248415, + "balance_loss_mlp": 1.00100088, + "epoch": 0.5170444297480912, + "flos": 15956583586560.0, + "grad_norm": 2.467456239902209, + "language_loss": 0.85071218, + "learning_rate": 1.9867575787565946e-06, + "loss": 0.87129903, + "num_input_tokens_seen": 92806240, + "step": 4300, + "time_per_iteration": 3.6018691062927246 + }, + { + "auxiliary_loss_clip": 0.01175135, + "auxiliary_loss_mlp": 0.01031855, + "balance_loss_clip": 1.05315185, + "balance_loss_mlp": 1.02238989, + "epoch": 0.5171646726387302, + "flos": 14174157968640.0, + "grad_norm": 2.473700788224015, + "language_loss": 0.86045384, + "learning_rate": 1.9859786252088275e-06, + "loss": 0.88252378, + "num_input_tokens_seen": 92823420, + "step": 4301, + "time_per_iteration": 3.769336223602295 + }, + { + "auxiliary_loss_clip": 0.01156568, + "auxiliary_loss_mlp": 0.01035399, + "balance_loss_clip": 1.04721975, + "balance_loss_mlp": 1.02599931, + "epoch": 0.5172849155293693, + "flos": 23578550974080.0, + "grad_norm": 4.031472941004752, + "language_loss": 0.66355735, + "learning_rate": 1.9851996737880914e-06, + "loss": 0.68547702, + "num_input_tokens_seen": 92838605, + "step": 4302, + "time_per_iteration": 2.8836023807525635 + }, + { + "auxiliary_loss_clip": 0.01179337, + "auxiliary_loss_mlp": 0.0103086, + "balance_loss_clip": 1.05255795, + "balance_loss_mlp": 1.02154422, + "epoch": 0.5174051584200084, + "flos": 14283541860480.0, + "grad_norm": 2.0859420600421794, + "language_loss": 0.74125677, + "learning_rate": 1.9844207246125537e-06, + "loss": 0.76335871, + "num_input_tokens_seen": 92855185, + "step": 4303, + "time_per_iteration": 3.614105224609375 + }, + { + "auxiliary_loss_clip": 0.01160705, + "auxiliary_loss_mlp": 0.01036412, + "balance_loss_clip": 1.04815602, + "balance_loss_mlp": 1.02764416, + "epoch": 0.5175254013106475, + "flos": 37889384192640.0, + "grad_norm": 1.6936244288098363, + "language_loss": 0.68661094, + "learning_rate": 1.983641777800379e-06, + "loss": 0.7085821, + "num_input_tokens_seen": 92877830, + "step": 4304, + "time_per_iteration": 2.971669912338257 + }, + { + "auxiliary_loss_clip": 0.01062277, + "auxiliary_loss_mlp": 0.01001876, + "balance_loss_clip": 1.01119971, + "balance_loss_mlp": 1.0001477, + "epoch": 0.5176456442012866, + "flos": 68549737829760.0, + "grad_norm": 0.749822707293018, + "language_loss": 0.58765149, + "learning_rate": 1.9828628334697343e-06, + "loss": 0.60829306, + "num_input_tokens_seen": 92945040, + "step": 4305, + "time_per_iteration": 3.503575563430786 + }, + { + "auxiliary_loss_clip": 0.0106465, + "auxiliary_loss_mlp": 0.01003651, + "balance_loss_clip": 1.0122335, + "balance_loss_mlp": 1.0018034, + "epoch": 0.5177658870919257, + "flos": 64084137235200.0, + "grad_norm": 0.7660761567738109, + "language_loss": 0.54658723, + "learning_rate": 1.982083891738784e-06, + "loss": 0.56727028, + "num_input_tokens_seen": 93005910, + "step": 4306, + "time_per_iteration": 3.3547558784484863 + }, + { + "auxiliary_loss_clip": 0.01158328, + "auxiliary_loss_mlp": 0.01035423, + "balance_loss_clip": 1.05242276, + "balance_loss_mlp": 1.02744746, + "epoch": 0.5178861299825648, + "flos": 26651248012800.0, + "grad_norm": 1.4870485378088614, + "language_loss": 0.82835215, + "learning_rate": 1.9813049527256923e-06, + "loss": 0.8502897, + "num_input_tokens_seen": 93026305, + "step": 4307, + "time_per_iteration": 2.8692665100097656 + }, + { + "auxiliary_loss_clip": 0.01149244, + "auxiliary_loss_mlp": 0.01032061, + "balance_loss_clip": 1.04428315, + "balance_loss_mlp": 1.02310848, + "epoch": 0.5180063728732038, + "flos": 17931886260480.0, + "grad_norm": 2.009786993859424, + "language_loss": 0.82255578, + "learning_rate": 1.9805260165486252e-06, + "loss": 0.84436882, + "num_input_tokens_seen": 93045675, + "step": 4308, + "time_per_iteration": 2.8343708515167236 + }, + { + "auxiliary_loss_clip": 0.01174675, + "auxiliary_loss_mlp": 0.01027166, + "balance_loss_clip": 1.05275583, + "balance_loss_mlp": 1.01893497, + "epoch": 0.518126615763843, + "flos": 19500895221120.0, + "grad_norm": 1.9050185049640245, + "language_loss": 0.86660498, + "learning_rate": 1.9797470833257457e-06, + "loss": 0.88862342, + "num_input_tokens_seen": 93065375, + "step": 4309, + "time_per_iteration": 2.7467029094696045 + }, + { + "auxiliary_loss_clip": 0.01174185, + "auxiliary_loss_mlp": 0.01029957, + "balance_loss_clip": 1.0532136, + "balance_loss_mlp": 1.02108812, + "epoch": 0.5182468586544821, + "flos": 20704082117760.0, + "grad_norm": 2.6949456528260196, + "language_loss": 0.77231044, + "learning_rate": 1.9789681531752177e-06, + "loss": 0.79435182, + "num_input_tokens_seen": 93085595, + "step": 4310, + "time_per_iteration": 2.6505126953125 + }, + { + "auxiliary_loss_clip": 0.01139353, + "auxiliary_loss_mlp": 0.01027477, + "balance_loss_clip": 1.04565787, + "balance_loss_mlp": 1.01977038, + "epoch": 0.5183671015451211, + "flos": 23112107936640.0, + "grad_norm": 1.4188070211439066, + "language_loss": 0.72481692, + "learning_rate": 1.978189226215204e-06, + "loss": 0.74648523, + "num_input_tokens_seen": 93106140, + "step": 4311, + "time_per_iteration": 2.7968626022338867 + }, + { + "auxiliary_loss_clip": 0.01181158, + "auxiliary_loss_mlp": 0.01026168, + "balance_loss_clip": 1.05107069, + "balance_loss_mlp": 1.01776958, + "epoch": 0.5184873444357603, + "flos": 17597090568960.0, + "grad_norm": 1.8429695699154598, + "language_loss": 0.7723304, + "learning_rate": 1.9774103025638675e-06, + "loss": 0.79440367, + "num_input_tokens_seen": 93124265, + "step": 4312, + "time_per_iteration": 2.678424596786499 + }, + { + "auxiliary_loss_clip": 0.01146731, + "auxiliary_loss_mlp": 0.0103243, + "balance_loss_clip": 1.05144238, + "balance_loss_mlp": 1.02389455, + "epoch": 0.5186075873263993, + "flos": 24936800883840.0, + "grad_norm": 1.5461570185913904, + "language_loss": 0.76592898, + "learning_rate": 1.9766313823393696e-06, + "loss": 0.78772056, + "num_input_tokens_seen": 93145130, + "step": 4313, + "time_per_iteration": 2.8930788040161133 + }, + { + "auxiliary_loss_clip": 0.01135535, + "auxiliary_loss_mlp": 0.01033025, + "balance_loss_clip": 1.04131651, + "balance_loss_mlp": 1.02350032, + "epoch": 0.5187278302170384, + "flos": 15190106244480.0, + "grad_norm": 2.1504094638520765, + "language_loss": 0.68949461, + "learning_rate": 1.975852465659873e-06, + "loss": 0.71118021, + "num_input_tokens_seen": 93161110, + "step": 4314, + "time_per_iteration": 2.8415236473083496 + }, + { + "auxiliary_loss_clip": 0.01174178, + "auxiliary_loss_mlp": 0.01031369, + "balance_loss_clip": 1.05184674, + "balance_loss_mlp": 1.02255356, + "epoch": 0.5188480731076776, + "flos": 25009412227200.0, + "grad_norm": 2.708812792846366, + "language_loss": 0.69992489, + "learning_rate": 1.9750735526435377e-06, + "loss": 0.72198033, + "num_input_tokens_seen": 93178055, + "step": 4315, + "time_per_iteration": 2.7109293937683105 + }, + { + "auxiliary_loss_clip": 0.01159811, + "auxiliary_loss_mlp": 0.01028201, + "balance_loss_clip": 1.05006003, + "balance_loss_mlp": 1.01908147, + "epoch": 0.5189683159983166, + "flos": 24790141653120.0, + "grad_norm": 2.022153407346966, + "language_loss": 0.79049349, + "learning_rate": 1.974294643408525e-06, + "loss": 0.81237364, + "num_input_tokens_seen": 93195850, + "step": 4316, + "time_per_iteration": 2.8905982971191406 + }, + { + "auxiliary_loss_clip": 0.01179072, + "auxiliary_loss_mlp": 0.01028736, + "balance_loss_clip": 1.05164516, + "balance_loss_mlp": 1.01943207, + "epoch": 0.5190885588889557, + "flos": 24754266944640.0, + "grad_norm": 2.263026839478634, + "language_loss": 0.67399913, + "learning_rate": 1.9735157380729947e-06, + "loss": 0.69607723, + "num_input_tokens_seen": 93216260, + "step": 4317, + "time_per_iteration": 2.8339149951934814 + }, + { + "auxiliary_loss_clip": 0.01162639, + "auxiliary_loss_mlp": 0.01031117, + "balance_loss_clip": 1.04689407, + "balance_loss_mlp": 1.02249229, + "epoch": 0.5192088017795948, + "flos": 24712646060160.0, + "grad_norm": 1.756996927313576, + "language_loss": 0.84332848, + "learning_rate": 1.9727368367551053e-06, + "loss": 0.86526597, + "num_input_tokens_seen": 93234810, + "step": 4318, + "time_per_iteration": 2.900073766708374 + }, + { + "auxiliary_loss_clip": 0.01153364, + "auxiliary_loss_mlp": 0.01035931, + "balance_loss_clip": 1.0471375, + "balance_loss_mlp": 1.0270679, + "epoch": 0.5193290446702339, + "flos": 27229588894080.0, + "grad_norm": 1.821200923763855, + "language_loss": 0.68789309, + "learning_rate": 1.9719579395730164e-06, + "loss": 0.70978606, + "num_input_tokens_seen": 93254185, + "step": 4319, + "time_per_iteration": 2.836610794067383 + }, + { + "auxiliary_loss_clip": 0.01184766, + "auxiliary_loss_mlp": 0.01039619, + "balance_loss_clip": 1.05414069, + "balance_loss_mlp": 1.03128624, + "epoch": 0.5194492875608729, + "flos": 11473352392320.0, + "grad_norm": 2.2101840632628216, + "language_loss": 0.93421549, + "learning_rate": 1.9711790466448854e-06, + "loss": 0.95645934, + "num_input_tokens_seen": 93268205, + "step": 4320, + "time_per_iteration": 2.9073328971862793 + }, + { + "auxiliary_loss_clip": 0.01148962, + "auxiliary_loss_mlp": 0.01036311, + "balance_loss_clip": 1.04711366, + "balance_loss_mlp": 1.02707815, + "epoch": 0.5195695304515121, + "flos": 20338906498560.0, + "grad_norm": 2.357623101826413, + "language_loss": 0.71217871, + "learning_rate": 1.9704001580888704e-06, + "loss": 0.73403138, + "num_input_tokens_seen": 93286945, + "step": 4321, + "time_per_iteration": 3.1143319606781006 + }, + { + "auxiliary_loss_clip": 0.01157864, + "auxiliary_loss_mlp": 0.00887088, + "balance_loss_clip": 1.04654098, + "balance_loss_mlp": 1.00114191, + "epoch": 0.5196897733421512, + "flos": 20048317470720.0, + "grad_norm": 1.9316058586945344, + "language_loss": 0.86553365, + "learning_rate": 1.9696212740231283e-06, + "loss": 0.88598311, + "num_input_tokens_seen": 93305595, + "step": 4322, + "time_per_iteration": 2.8904340267181396 + }, + { + "auxiliary_loss_clip": 0.01178488, + "auxiliary_loss_mlp": 0.01034649, + "balance_loss_clip": 1.05044782, + "balance_loss_mlp": 1.02491605, + "epoch": 0.5198100162327902, + "flos": 23805507058560.0, + "grad_norm": 2.0529347908882816, + "language_loss": 0.81934541, + "learning_rate": 1.9688423945658146e-06, + "loss": 0.8414768, + "num_input_tokens_seen": 93326460, + "step": 4323, + "time_per_iteration": 2.929989814758301 + }, + { + "auxiliary_loss_clip": 0.01129392, + "auxiliary_loss_mlp": 0.01032075, + "balance_loss_clip": 1.03776145, + "balance_loss_mlp": 1.02286589, + "epoch": 0.5199302591234293, + "flos": 24023951619840.0, + "grad_norm": 3.1224785141734195, + "language_loss": 0.72380459, + "learning_rate": 1.9680635198350845e-06, + "loss": 0.74541926, + "num_input_tokens_seen": 93346170, + "step": 4324, + "time_per_iteration": 3.922600030899048 + }, + { + "auxiliary_loss_clip": 0.01173756, + "auxiliary_loss_mlp": 0.01030643, + "balance_loss_clip": 1.04917681, + "balance_loss_mlp": 1.02118421, + "epoch": 0.5200505020140684, + "flos": 26359366095360.0, + "grad_norm": 1.9810209846708462, + "language_loss": 0.7262491, + "learning_rate": 1.967284649949093e-06, + "loss": 0.74829304, + "num_input_tokens_seen": 93365380, + "step": 4325, + "time_per_iteration": 2.8794193267822266 + }, + { + "auxiliary_loss_clip": 0.01148172, + "auxiliary_loss_mlp": 0.01026167, + "balance_loss_clip": 1.04453242, + "balance_loss_mlp": 1.01780999, + "epoch": 0.5201707449047075, + "flos": 39604262284800.0, + "grad_norm": 2.322858739674223, + "language_loss": 0.72220635, + "learning_rate": 1.966505785025994e-06, + "loss": 0.74394971, + "num_input_tokens_seen": 93387285, + "step": 4326, + "time_per_iteration": 4.014614582061768 + }, + { + "auxiliary_loss_clip": 0.01150795, + "auxiliary_loss_mlp": 0.01033168, + "balance_loss_clip": 1.04935765, + "balance_loss_mlp": 1.02420974, + "epoch": 0.5202909877953465, + "flos": 53682788292480.0, + "grad_norm": 2.204709578757975, + "language_loss": 0.75802845, + "learning_rate": 1.965726925183941e-06, + "loss": 0.77986813, + "num_input_tokens_seen": 93410390, + "step": 4327, + "time_per_iteration": 4.080608606338501 + }, + { + "auxiliary_loss_clip": 0.01181759, + "auxiliary_loss_mlp": 0.01030822, + "balance_loss_clip": 1.05163121, + "balance_loss_mlp": 1.02241182, + "epoch": 0.5204112306859857, + "flos": 19537021324800.0, + "grad_norm": 1.6742600823676361, + "language_loss": 0.84101284, + "learning_rate": 1.964948070541087e-06, + "loss": 0.86313862, + "num_input_tokens_seen": 93429050, + "step": 4328, + "time_per_iteration": 2.6887564659118652 + }, + { + "auxiliary_loss_clip": 0.01162879, + "auxiliary_loss_mlp": 0.01035205, + "balance_loss_clip": 1.04759693, + "balance_loss_mlp": 1.02603817, + "epoch": 0.5205314735766248, + "flos": 15304697608320.0, + "grad_norm": 2.776018842614461, + "language_loss": 0.69632936, + "learning_rate": 1.9641692212155816e-06, + "loss": 0.71831024, + "num_input_tokens_seen": 93446815, + "step": 4329, + "time_per_iteration": 2.6394994258880615 + }, + { + "auxiliary_loss_clip": 0.01141457, + "auxiliary_loss_mlp": 0.01029645, + "balance_loss_clip": 1.04906297, + "balance_loss_mlp": 1.02084112, + "epoch": 0.5206517164672638, + "flos": 59263701160320.0, + "grad_norm": 4.075629144426082, + "language_loss": 0.72733021, + "learning_rate": 1.9633903773255777e-06, + "loss": 0.74904126, + "num_input_tokens_seen": 93469130, + "step": 4330, + "time_per_iteration": 4.074136018753052 + }, + { + "auxiliary_loss_clip": 0.01179234, + "auxiliary_loss_mlp": 0.01031136, + "balance_loss_clip": 1.0496645, + "balance_loss_mlp": 1.02258277, + "epoch": 0.520771959357903, + "flos": 26871129118080.0, + "grad_norm": 1.8028625659341977, + "language_loss": 0.74665022, + "learning_rate": 1.9626115389892237e-06, + "loss": 0.76875395, + "num_input_tokens_seen": 93489920, + "step": 4331, + "time_per_iteration": 2.8009331226348877 + }, + { + "auxiliary_loss_clip": 0.01157863, + "auxiliary_loss_mlp": 0.01031216, + "balance_loss_clip": 1.05057454, + "balance_loss_mlp": 1.02251339, + "epoch": 0.520892202248542, + "flos": 26907075653760.0, + "grad_norm": 2.0900817810927634, + "language_loss": 0.85449123, + "learning_rate": 1.96183270632467e-06, + "loss": 0.87638205, + "num_input_tokens_seen": 93509770, + "step": 4332, + "time_per_iteration": 2.9069433212280273 + }, + { + "auxiliary_loss_clip": 0.0114663, + "auxiliary_loss_mlp": 0.00887715, + "balance_loss_clip": 1.04877591, + "balance_loss_mlp": 1.00107312, + "epoch": 0.5210124451391811, + "flos": 25849434666240.0, + "grad_norm": 1.6097968513766763, + "language_loss": 0.78959751, + "learning_rate": 1.9610538794500644e-06, + "loss": 0.80994093, + "num_input_tokens_seen": 93529320, + "step": 4333, + "time_per_iteration": 2.7847487926483154 + }, + { + "auxiliary_loss_clip": 0.01053362, + "auxiliary_loss_mlp": 0.01004753, + "balance_loss_clip": 1.01050794, + "balance_loss_mlp": 1.00306022, + "epoch": 0.5211326880298203, + "flos": 70553804319360.0, + "grad_norm": 0.7796141170607804, + "language_loss": 0.59442055, + "learning_rate": 1.9602750584835542e-06, + "loss": 0.6150018, + "num_input_tokens_seen": 93595255, + "step": 4334, + "time_per_iteration": 3.495204448699951 + }, + { + "auxiliary_loss_clip": 0.01158346, + "auxiliary_loss_mlp": 0.0102928, + "balance_loss_clip": 1.04644561, + "balance_loss_mlp": 1.020661, + "epoch": 0.5212529309204593, + "flos": 15628898787840.0, + "grad_norm": 2.0778869598169996, + "language_loss": 0.83008254, + "learning_rate": 1.959496243543286e-06, + "loss": 0.85195875, + "num_input_tokens_seen": 93613135, + "step": 4335, + "time_per_iteration": 2.7692103385925293 + }, + { + "auxiliary_loss_clip": 0.01176715, + "auxiliary_loss_mlp": 0.0103687, + "balance_loss_clip": 1.05517602, + "balance_loss_mlp": 1.02858543, + "epoch": 0.5213731738110984, + "flos": 26242655829120.0, + "grad_norm": 2.1273520039132157, + "language_loss": 0.79540557, + "learning_rate": 1.9587174347474057e-06, + "loss": 0.81754142, + "num_input_tokens_seen": 93629645, + "step": 4336, + "time_per_iteration": 2.7984511852264404 + }, + { + "auxiliary_loss_clip": 0.01129988, + "auxiliary_loss_mlp": 0.01034926, + "balance_loss_clip": 1.0430392, + "balance_loss_mlp": 1.02561033, + "epoch": 0.5214934167017375, + "flos": 19418407637760.0, + "grad_norm": 2.196061553970962, + "language_loss": 0.81945246, + "learning_rate": 1.9579386322140574e-06, + "loss": 0.84110165, + "num_input_tokens_seen": 93645325, + "step": 4337, + "time_per_iteration": 2.883118152618408 + }, + { + "auxiliary_loss_clip": 0.01184644, + "auxiliary_loss_mlp": 0.00886762, + "balance_loss_clip": 1.05316305, + "balance_loss_mlp": 1.00097418, + "epoch": 0.5216136595923766, + "flos": 30955788023040.0, + "grad_norm": 1.9684899065931367, + "language_loss": 0.81029379, + "learning_rate": 1.9571598360613854e-06, + "loss": 0.8310079, + "num_input_tokens_seen": 93668200, + "step": 4338, + "time_per_iteration": 2.7192487716674805 + }, + { + "auxiliary_loss_clip": 0.01147886, + "auxiliary_loss_mlp": 0.0102986, + "balance_loss_clip": 1.04440999, + "balance_loss_mlp": 1.020818, + "epoch": 0.5217339024830157, + "flos": 21945047143680.0, + "grad_norm": 4.821414056641969, + "language_loss": 0.69453382, + "learning_rate": 1.956381046407532e-06, + "loss": 0.71631122, + "num_input_tokens_seen": 93688495, + "step": 4339, + "time_per_iteration": 2.762669086456299 + }, + { + "auxiliary_loss_clip": 0.01150333, + "auxiliary_loss_mlp": 0.01033188, + "balance_loss_clip": 1.04777503, + "balance_loss_mlp": 1.02471185, + "epoch": 0.5218541453736548, + "flos": 20923209037440.0, + "grad_norm": 2.118910050468753, + "language_loss": 0.86182356, + "learning_rate": 1.9556022633706394e-06, + "loss": 0.88365877, + "num_input_tokens_seen": 93707285, + "step": 4340, + "time_per_iteration": 3.0057923793792725 + }, + { + "auxiliary_loss_clip": 0.01157361, + "auxiliary_loss_mlp": 0.01033538, + "balance_loss_clip": 1.05059779, + "balance_loss_mlp": 1.02451408, + "epoch": 0.5219743882642939, + "flos": 23951663498880.0, + "grad_norm": 1.8225224102677362, + "language_loss": 0.79997611, + "learning_rate": 1.954823487068848e-06, + "loss": 0.82188511, + "num_input_tokens_seen": 93727495, + "step": 4341, + "time_per_iteration": 2.799105644226074 + }, + { + "auxiliary_loss_clip": 0.01171641, + "auxiliary_loss_mlp": 0.01033247, + "balance_loss_clip": 1.05193484, + "balance_loss_mlp": 1.02465761, + "epoch": 0.5220946311549329, + "flos": 28799280213120.0, + "grad_norm": 1.776748994494853, + "language_loss": 0.81274617, + "learning_rate": 1.9540447176202976e-06, + "loss": 0.834795, + "num_input_tokens_seen": 93748740, + "step": 4342, + "time_per_iteration": 2.873352289199829 + }, + { + "auxiliary_loss_clip": 0.01065036, + "auxiliary_loss_mlp": 0.01002627, + "balance_loss_clip": 1.0126946, + "balance_loss_mlp": 1.00081468, + "epoch": 0.5222148740455721, + "flos": 67189369017600.0, + "grad_norm": 0.8681428485256336, + "language_loss": 0.60679513, + "learning_rate": 1.9532659551431272e-06, + "loss": 0.62747175, + "num_input_tokens_seen": 93815770, + "step": 4343, + "time_per_iteration": 3.425635814666748 + }, + { + "auxiliary_loss_clip": 0.01176956, + "auxiliary_loss_mlp": 0.01027217, + "balance_loss_clip": 1.05119276, + "balance_loss_mlp": 1.01902723, + "epoch": 0.5223351169362112, + "flos": 61856164339200.0, + "grad_norm": 1.8089925037186056, + "language_loss": 0.67504489, + "learning_rate": 1.9524871997554744e-06, + "loss": 0.69708663, + "num_input_tokens_seen": 93843530, + "step": 4344, + "time_per_iteration": 3.0930399894714355 + }, + { + "auxiliary_loss_clip": 0.01173243, + "auxiliary_loss_mlp": 0.01033183, + "balance_loss_clip": 1.05239546, + "balance_loss_mlp": 1.02380705, + "epoch": 0.5224553598268502, + "flos": 14647388676480.0, + "grad_norm": 2.074577924710356, + "language_loss": 0.80404031, + "learning_rate": 1.951708451575475e-06, + "loss": 0.82610458, + "num_input_tokens_seen": 93860595, + "step": 4345, + "time_per_iteration": 2.7009165287017822 + }, + { + "auxiliary_loss_clip": 0.0116312, + "auxiliary_loss_mlp": 0.01033328, + "balance_loss_clip": 1.04669023, + "balance_loss_mlp": 1.02475095, + "epoch": 0.5225756027174894, + "flos": 14826043946880.0, + "grad_norm": 1.9274645362319567, + "language_loss": 0.82046497, + "learning_rate": 1.9509297107212657e-06, + "loss": 0.84242952, + "num_input_tokens_seen": 93877365, + "step": 4346, + "time_per_iteration": 2.748847007751465 + }, + { + "auxiliary_loss_clip": 0.01182744, + "auxiliary_loss_mlp": 0.01030113, + "balance_loss_clip": 1.053231, + "balance_loss_mlp": 1.02160764, + "epoch": 0.5226958456081284, + "flos": 23512009029120.0, + "grad_norm": 1.5473732523766555, + "language_loss": 0.78854448, + "learning_rate": 1.95015097731098e-06, + "loss": 0.81067312, + "num_input_tokens_seen": 93896855, + "step": 4347, + "time_per_iteration": 2.7445802688598633 + }, + { + "auxiliary_loss_clip": 0.01183036, + "auxiliary_loss_mlp": 0.01039127, + "balance_loss_clip": 1.05349684, + "balance_loss_mlp": 1.0302335, + "epoch": 0.5228160884987675, + "flos": 19062928690560.0, + "grad_norm": 2.0308873061854067, + "language_loss": 0.81914878, + "learning_rate": 1.949372251462751e-06, + "loss": 0.8413704, + "num_input_tokens_seen": 93914270, + "step": 4348, + "time_per_iteration": 2.6516592502593994 + }, + { + "auxiliary_loss_clip": 0.01151893, + "auxiliary_loss_mlp": 0.00887217, + "balance_loss_clip": 1.0470798, + "balance_loss_mlp": 1.001122, + "epoch": 0.5229363313894067, + "flos": 21063224252160.0, + "grad_norm": 1.7379930658538283, + "language_loss": 0.82611012, + "learning_rate": 1.9485935332947124e-06, + "loss": 0.84650123, + "num_input_tokens_seen": 93932180, + "step": 4349, + "time_per_iteration": 2.8675436973571777 + }, + { + "auxiliary_loss_clip": 0.01160654, + "auxiliary_loss_mlp": 0.01028373, + "balance_loss_clip": 1.04935479, + "balance_loss_mlp": 1.02010036, + "epoch": 0.5230565742800457, + "flos": 14830389492480.0, + "grad_norm": 5.213349493447625, + "language_loss": 0.8367027, + "learning_rate": 1.947814822924993e-06, + "loss": 0.85859299, + "num_input_tokens_seen": 93949690, + "step": 4350, + "time_per_iteration": 3.6769859790802 + }, + { + "auxiliary_loss_clip": 0.01183343, + "auxiliary_loss_mlp": 0.01028206, + "balance_loss_clip": 1.05490303, + "balance_loss_mlp": 1.01960552, + "epoch": 0.5231768171706848, + "flos": 25813021253760.0, + "grad_norm": 1.7732983990250553, + "language_loss": 0.82680321, + "learning_rate": 1.9470361204717236e-06, + "loss": 0.84891868, + "num_input_tokens_seen": 93968830, + "step": 4351, + "time_per_iteration": 2.71561336517334 + }, + { + "auxiliary_loss_clip": 0.01154431, + "auxiliary_loss_mlp": 0.00887489, + "balance_loss_clip": 1.04672754, + "balance_loss_mlp": 1.00102997, + "epoch": 0.5232970600613239, + "flos": 22743807834240.0, + "grad_norm": 1.5098387242690516, + "language_loss": 0.80689323, + "learning_rate": 1.9462574260530326e-06, + "loss": 0.82731241, + "num_input_tokens_seen": 93989110, + "step": 4352, + "time_per_iteration": 3.923081398010254 + }, + { + "auxiliary_loss_clip": 0.0116206, + "auxiliary_loss_mlp": 0.01027793, + "balance_loss_clip": 1.0476805, + "balance_loss_mlp": 1.01933503, + "epoch": 0.523417302951963, + "flos": 17310703432320.0, + "grad_norm": 2.0102841249957284, + "language_loss": 0.81046581, + "learning_rate": 1.9454787397870472e-06, + "loss": 0.83236432, + "num_input_tokens_seen": 94006430, + "step": 4353, + "time_per_iteration": 3.6508891582489014 + }, + { + "auxiliary_loss_clip": 0.0111964, + "auxiliary_loss_mlp": 0.01035427, + "balance_loss_clip": 1.04464662, + "balance_loss_mlp": 1.02708197, + "epoch": 0.523537545842602, + "flos": 18551740285440.0, + "grad_norm": 1.9533065550312518, + "language_loss": 0.71624351, + "learning_rate": 1.944700061791894e-06, + "loss": 0.73779416, + "num_input_tokens_seen": 94024825, + "step": 4354, + "time_per_iteration": 2.8380868434906006 + }, + { + "auxiliary_loss_clip": 0.01169475, + "auxiliary_loss_mlp": 0.0103019, + "balance_loss_clip": 1.05058455, + "balance_loss_mlp": 1.02131534, + "epoch": 0.5236577887332411, + "flos": 19719267955200.0, + "grad_norm": 2.473206882930544, + "language_loss": 0.6563226, + "learning_rate": 1.943921392185698e-06, + "loss": 0.67831933, + "num_input_tokens_seen": 94043450, + "step": 4355, + "time_per_iteration": 3.6251378059387207 + }, + { + "auxiliary_loss_clip": 0.01160316, + "auxiliary_loss_mlp": 0.01028297, + "balance_loss_clip": 1.04865694, + "balance_loss_mlp": 1.01930821, + "epoch": 0.5237780316238803, + "flos": 23550218121600.0, + "grad_norm": 2.0727915920012046, + "language_loss": 0.77288079, + "learning_rate": 1.9431427310865814e-06, + "loss": 0.7947669, + "num_input_tokens_seen": 94063055, + "step": 4356, + "time_per_iteration": 2.7546634674072266 + }, + { + "auxiliary_loss_clip": 0.01135836, + "auxiliary_loss_mlp": 0.01028625, + "balance_loss_clip": 1.04515433, + "balance_loss_mlp": 1.02033389, + "epoch": 0.5238982745145193, + "flos": 22491894775680.0, + "grad_norm": 1.7030682416864333, + "language_loss": 0.78757644, + "learning_rate": 1.942364078612667e-06, + "loss": 0.80922103, + "num_input_tokens_seen": 94081785, + "step": 4357, + "time_per_iteration": 2.826838731765747 + }, + { + "auxiliary_loss_clip": 0.01152348, + "auxiliary_loss_mlp": 0.01032189, + "balance_loss_clip": 1.04557955, + "balance_loss_mlp": 1.02280068, + "epoch": 0.5240185174051584, + "flos": 27088927234560.0, + "grad_norm": 1.995362909407533, + "language_loss": 0.75538778, + "learning_rate": 1.9415854348820765e-06, + "loss": 0.77723312, + "num_input_tokens_seen": 94101635, + "step": 4358, + "time_per_iteration": 2.8714237213134766 + }, + { + "auxiliary_loss_clip": 0.01169919, + "auxiliary_loss_mlp": 0.01032859, + "balance_loss_clip": 1.04816222, + "balance_loss_mlp": 1.02423978, + "epoch": 0.5241387602957975, + "flos": 22674680110080.0, + "grad_norm": 2.38733746765796, + "language_loss": 0.68288732, + "learning_rate": 1.940806800012929e-06, + "loss": 0.70491511, + "num_input_tokens_seen": 94121705, + "step": 4359, + "time_per_iteration": 2.709883213043213 + }, + { + "auxiliary_loss_clip": 0.01143505, + "auxiliary_loss_mlp": 0.00887623, + "balance_loss_clip": 1.04754114, + "balance_loss_mlp": 1.00107241, + "epoch": 0.5242590031864366, + "flos": 40553453134080.0, + "grad_norm": 1.5960753671594077, + "language_loss": 0.63547432, + "learning_rate": 1.9400281741233432e-06, + "loss": 0.65578568, + "num_input_tokens_seen": 94146595, + "step": 4360, + "time_per_iteration": 3.02058482170105 + }, + { + "auxiliary_loss_clip": 0.01055666, + "auxiliary_loss_mlp": 0.01008522, + "balance_loss_clip": 1.01777101, + "balance_loss_mlp": 1.00662637, + "epoch": 0.5243792460770756, + "flos": 66676313105280.0, + "grad_norm": 0.6568866055289211, + "language_loss": 0.5258323, + "learning_rate": 1.939249557331435e-06, + "loss": 0.54647416, + "num_input_tokens_seen": 94212410, + "step": 4361, + "time_per_iteration": 3.3824031352996826 + }, + { + "auxiliary_loss_clip": 0.01163816, + "auxiliary_loss_mlp": 0.01032382, + "balance_loss_clip": 1.04869711, + "balance_loss_mlp": 1.02415049, + "epoch": 0.5244994889677148, + "flos": 28183663992960.0, + "grad_norm": 1.9696222155393204, + "language_loss": 0.72639894, + "learning_rate": 1.938470949755321e-06, + "loss": 0.74836087, + "num_input_tokens_seen": 94232290, + "step": 4362, + "time_per_iteration": 2.865781784057617 + }, + { + "auxiliary_loss_clip": 0.01053836, + "auxiliary_loss_mlp": 0.01005069, + "balance_loss_clip": 1.01082587, + "balance_loss_mlp": 1.00311434, + "epoch": 0.5246197318583539, + "flos": 65950379239680.0, + "grad_norm": 0.8252764611524088, + "language_loss": 0.55669439, + "learning_rate": 1.937692351513115e-06, + "loss": 0.57728338, + "num_input_tokens_seen": 94291285, + "step": 4363, + "time_per_iteration": 3.3115274906158447 + }, + { + "auxiliary_loss_clip": 0.01178699, + "auxiliary_loss_mlp": 0.01031159, + "balance_loss_clip": 1.05158067, + "balance_loss_mlp": 1.02181911, + "epoch": 0.5247399747489929, + "flos": 21033490769280.0, + "grad_norm": 1.5680904307010906, + "language_loss": 0.80553979, + "learning_rate": 1.9369137627229297e-06, + "loss": 0.82763839, + "num_input_tokens_seen": 94309685, + "step": 4364, + "time_per_iteration": 2.7509241104125977 + }, + { + "auxiliary_loss_clip": 0.01172019, + "auxiliary_loss_mlp": 0.01025896, + "balance_loss_clip": 1.05263686, + "balance_loss_mlp": 1.01749742, + "epoch": 0.5248602176396321, + "flos": 19025940660480.0, + "grad_norm": 1.8866149932812801, + "language_loss": 0.88264346, + "learning_rate": 1.936135183502877e-06, + "loss": 0.90462267, + "num_input_tokens_seen": 94326985, + "step": 4365, + "time_per_iteration": 2.7715940475463867 + }, + { + "auxiliary_loss_clip": 0.01151577, + "auxiliary_loss_mlp": 0.01040931, + "balance_loss_clip": 1.04827297, + "balance_loss_mlp": 1.0313524, + "epoch": 0.5249804605302711, + "flos": 22200084685440.0, + "grad_norm": 2.9519465379754486, + "language_loss": 0.8047204, + "learning_rate": 1.935356613971066e-06, + "loss": 0.82664549, + "num_input_tokens_seen": 94347645, + "step": 4366, + "time_per_iteration": 2.9194748401641846 + }, + { + "auxiliary_loss_clip": 0.01161441, + "auxiliary_loss_mlp": 0.0088724, + "balance_loss_clip": 1.04970193, + "balance_loss_mlp": 1.00113177, + "epoch": 0.5251007034209102, + "flos": 23805686626560.0, + "grad_norm": 1.696542237488686, + "language_loss": 0.7676909, + "learning_rate": 1.9345780542456047e-06, + "loss": 0.78817767, + "num_input_tokens_seen": 94367020, + "step": 4367, + "time_per_iteration": 2.863854169845581 + }, + { + "auxiliary_loss_clip": 0.01160458, + "auxiliary_loss_mlp": 0.0103038, + "balance_loss_clip": 1.04710817, + "balance_loss_mlp": 1.02149916, + "epoch": 0.5252209463115494, + "flos": 23294605962240.0, + "grad_norm": 3.414478216692234, + "language_loss": 0.7185123, + "learning_rate": 1.9337995044446007e-06, + "loss": 0.7404207, + "num_input_tokens_seen": 94385860, + "step": 4368, + "time_per_iteration": 2.8094587326049805 + }, + { + "auxiliary_loss_clip": 0.01175788, + "auxiliary_loss_mlp": 0.010295, + "balance_loss_clip": 1.04899096, + "balance_loss_mlp": 1.02094054, + "epoch": 0.5253411892021884, + "flos": 19828687760640.0, + "grad_norm": 2.0795039987423976, + "language_loss": 0.79881746, + "learning_rate": 1.9330209646861596e-06, + "loss": 0.82087028, + "num_input_tokens_seen": 94405010, + "step": 4369, + "time_per_iteration": 2.860090970993042 + }, + { + "auxiliary_loss_clip": 0.01161191, + "auxiliary_loss_mlp": 0.01034456, + "balance_loss_clip": 1.04883838, + "balance_loss_mlp": 1.02590322, + "epoch": 0.5254614320928275, + "flos": 24133730561280.0, + "grad_norm": 1.6168319333712355, + "language_loss": 0.77545774, + "learning_rate": 1.9322424350883843e-06, + "loss": 0.79741418, + "num_input_tokens_seen": 94426845, + "step": 4370, + "time_per_iteration": 2.845085382461548 + }, + { + "auxiliary_loss_clip": 0.01163382, + "auxiliary_loss_mlp": 0.01029107, + "balance_loss_clip": 1.04737496, + "balance_loss_mlp": 1.02001166, + "epoch": 0.5255816749834666, + "flos": 24644954880000.0, + "grad_norm": 1.7369226796413566, + "language_loss": 0.79060566, + "learning_rate": 1.931463915769379e-06, + "loss": 0.81253052, + "num_input_tokens_seen": 94446960, + "step": 4371, + "time_per_iteration": 2.771467685699463 + }, + { + "auxiliary_loss_clip": 0.01140963, + "auxiliary_loss_mlp": 0.0103188, + "balance_loss_clip": 1.0450238, + "balance_loss_mlp": 1.02336836, + "epoch": 0.5257019178741057, + "flos": 14136595320960.0, + "grad_norm": 3.502790685527615, + "language_loss": 0.73671526, + "learning_rate": 1.930685406847242e-06, + "loss": 0.75844371, + "num_input_tokens_seen": 94461535, + "step": 4372, + "time_per_iteration": 2.8793489933013916 + }, + { + "auxiliary_loss_clip": 0.01158617, + "auxiliary_loss_mlp": 0.01029016, + "balance_loss_clip": 1.04719675, + "balance_loss_mlp": 1.02055812, + "epoch": 0.5258221607647448, + "flos": 23548961145600.0, + "grad_norm": 1.5089020868033511, + "language_loss": 0.81883389, + "learning_rate": 1.9299069084400734e-06, + "loss": 0.84071022, + "num_input_tokens_seen": 94482395, + "step": 4373, + "time_per_iteration": 2.797720193862915 + }, + { + "auxiliary_loss_clip": 0.01149898, + "auxiliary_loss_mlp": 0.0102982, + "balance_loss_clip": 1.04876912, + "balance_loss_mlp": 1.02037227, + "epoch": 0.5259424036553839, + "flos": 24966103403520.0, + "grad_norm": 2.0965131708163414, + "language_loss": 0.69886076, + "learning_rate": 1.9291284206659717e-06, + "loss": 0.72065794, + "num_input_tokens_seen": 94500580, + "step": 4374, + "time_per_iteration": 2.89520263671875 + }, + { + "auxiliary_loss_clip": 0.01182454, + "auxiliary_loss_mlp": 0.01030693, + "balance_loss_clip": 1.0513978, + "balance_loss_mlp": 1.02128196, + "epoch": 0.526062646546023, + "flos": 28763908295040.0, + "grad_norm": 1.794434490521148, + "language_loss": 0.71811157, + "learning_rate": 1.928349943643032e-06, + "loss": 0.74024308, + "num_input_tokens_seen": 94519680, + "step": 4375, + "time_per_iteration": 2.6963818073272705 + }, + { + "auxiliary_loss_clip": 0.01170177, + "auxiliary_loss_mlp": 0.01032142, + "balance_loss_clip": 1.05123854, + "balance_loss_mlp": 1.02315331, + "epoch": 0.526182889436662, + "flos": 22821375254400.0, + "grad_norm": 2.12141487756221, + "language_loss": 0.81656623, + "learning_rate": 1.9275714774893493e-06, + "loss": 0.83858943, + "num_input_tokens_seen": 94539135, + "step": 4376, + "time_per_iteration": 3.6142215728759766 + }, + { + "auxiliary_loss_clip": 0.01144723, + "auxiliary_loss_mlp": 0.01034203, + "balance_loss_clip": 1.04451609, + "balance_loss_mlp": 1.02438045, + "epoch": 0.5263031323273012, + "flos": 22929466256640.0, + "grad_norm": 2.2197814491610885, + "language_loss": 0.73038441, + "learning_rate": 1.9267930223230154e-06, + "loss": 0.75217366, + "num_input_tokens_seen": 94557610, + "step": 4377, + "time_per_iteration": 2.831160306930542 + }, + { + "auxiliary_loss_clip": 0.01162557, + "auxiliary_loss_mlp": 0.01037683, + "balance_loss_clip": 1.0482893, + "balance_loss_mlp": 1.02865934, + "epoch": 0.5264233752179402, + "flos": 17748634049280.0, + "grad_norm": 2.1176868584540975, + "language_loss": 0.77798533, + "learning_rate": 1.9260145782621224e-06, + "loss": 0.79998779, + "num_input_tokens_seen": 94575390, + "step": 4378, + "time_per_iteration": 3.687978506088257 + }, + { + "auxiliary_loss_clip": 0.01161959, + "auxiliary_loss_mlp": 0.01033959, + "balance_loss_clip": 1.05272365, + "balance_loss_mlp": 1.02554858, + "epoch": 0.5265436181085793, + "flos": 24421626069120.0, + "grad_norm": 1.6519005410345602, + "language_loss": 0.87705833, + "learning_rate": 1.925236145424758e-06, + "loss": 0.89901745, + "num_input_tokens_seen": 94594210, + "step": 4379, + "time_per_iteration": 3.7619800567626953 + }, + { + "auxiliary_loss_clip": 0.01065181, + "auxiliary_loss_mlp": 0.01005165, + "balance_loss_clip": 1.01080191, + "balance_loss_mlp": 1.00340116, + "epoch": 0.5266638609992185, + "flos": 69207298156800.0, + "grad_norm": 0.7004926606013755, + "language_loss": 0.57592726, + "learning_rate": 1.924457723929012e-06, + "loss": 0.59663081, + "num_input_tokens_seen": 94665020, + "step": 4380, + "time_per_iteration": 3.444855213165283 + }, + { + "auxiliary_loss_clip": 0.01172591, + "auxiliary_loss_mlp": 0.01030591, + "balance_loss_clip": 1.05067897, + "balance_loss_mlp": 1.02162039, + "epoch": 0.5267841038898575, + "flos": 20738699850240.0, + "grad_norm": 1.4622125953869078, + "language_loss": 0.82637626, + "learning_rate": 1.9236793138929685e-06, + "loss": 0.84840804, + "num_input_tokens_seen": 94684290, + "step": 4381, + "time_per_iteration": 3.7390124797821045 + }, + { + "auxiliary_loss_clip": 0.01176311, + "auxiliary_loss_mlp": 0.01033947, + "balance_loss_clip": 1.04978514, + "balance_loss_mlp": 1.02444041, + "epoch": 0.5269043467804966, + "flos": 17234392988160.0, + "grad_norm": 2.1209497718002464, + "language_loss": 0.81246054, + "learning_rate": 1.9229009154347133e-06, + "loss": 0.83456314, + "num_input_tokens_seen": 94701880, + "step": 4382, + "time_per_iteration": 2.757504940032959 + }, + { + "auxiliary_loss_clip": 0.01128279, + "auxiliary_loss_mlp": 0.00886274, + "balance_loss_clip": 1.04234052, + "balance_loss_mlp": 1.00119233, + "epoch": 0.5270245896711357, + "flos": 18223157646720.0, + "grad_norm": 2.136799221849324, + "language_loss": 0.80519038, + "learning_rate": 1.922122528672327e-06, + "loss": 0.82533598, + "num_input_tokens_seen": 94720545, + "step": 4383, + "time_per_iteration": 3.01131272315979 + }, + { + "auxiliary_loss_clip": 0.01179557, + "auxiliary_loss_mlp": 0.01026085, + "balance_loss_clip": 1.05257523, + "balance_loss_mlp": 1.01780581, + "epoch": 0.5271448325617748, + "flos": 21287558643840.0, + "grad_norm": 2.529905243808054, + "language_loss": 0.78108543, + "learning_rate": 1.9213441537238914e-06, + "loss": 0.80314183, + "num_input_tokens_seen": 94737420, + "step": 4384, + "time_per_iteration": 2.9454455375671387 + }, + { + "auxiliary_loss_clip": 0.010465, + "auxiliary_loss_mlp": 0.01002742, + "balance_loss_clip": 1.01720977, + "balance_loss_mlp": 1.00081062, + "epoch": 0.5272650754524139, + "flos": 65495497403520.0, + "grad_norm": 0.838855839084415, + "language_loss": 0.57395816, + "learning_rate": 1.920565790707485e-06, + "loss": 0.59445059, + "num_input_tokens_seen": 94802810, + "step": 4385, + "time_per_iteration": 3.5809895992279053 + }, + { + "auxiliary_loss_clip": 0.0114636, + "auxiliary_loss_mlp": 0.01033792, + "balance_loss_clip": 1.04834843, + "balance_loss_mlp": 1.02461243, + "epoch": 0.527385318343053, + "flos": 19676426008320.0, + "grad_norm": 5.087303505125429, + "language_loss": 0.65860927, + "learning_rate": 1.9197874397411853e-06, + "loss": 0.68041074, + "num_input_tokens_seen": 94819440, + "step": 4386, + "time_per_iteration": 2.891702651977539 + }, + { + "auxiliary_loss_clip": 0.01150108, + "auxiliary_loss_mlp": 0.01027729, + "balance_loss_clip": 1.04192185, + "balance_loss_mlp": 1.01855552, + "epoch": 0.5275055612336921, + "flos": 12712018947840.0, + "grad_norm": 2.9240218167343857, + "language_loss": 0.65789419, + "learning_rate": 1.919009100943067e-06, + "loss": 0.6796726, + "num_input_tokens_seen": 94835130, + "step": 4387, + "time_per_iteration": 2.8106462955474854 + }, + { + "auxiliary_loss_clip": 0.01155908, + "auxiliary_loss_mlp": 0.01031258, + "balance_loss_clip": 1.05042768, + "balance_loss_mlp": 1.02188826, + "epoch": 0.5276258041243311, + "flos": 17749029098880.0, + "grad_norm": 2.042071927317603, + "language_loss": 0.65898287, + "learning_rate": 1.9182307744312043e-06, + "loss": 0.6808545, + "num_input_tokens_seen": 94852235, + "step": 4388, + "time_per_iteration": 2.902254343032837 + }, + { + "auxiliary_loss_clip": 0.01166607, + "auxiliary_loss_mlp": 0.01027189, + "balance_loss_clip": 1.04880929, + "balance_loss_mlp": 1.01850474, + "epoch": 0.5277460470149702, + "flos": 22710447077760.0, + "grad_norm": 1.6963148342612429, + "language_loss": 0.76328933, + "learning_rate": 1.9174524603236676e-06, + "loss": 0.7852273, + "num_input_tokens_seen": 94871185, + "step": 4389, + "time_per_iteration": 2.748986005783081 + }, + { + "auxiliary_loss_clip": 0.01162178, + "auxiliary_loss_mlp": 0.01027635, + "balance_loss_clip": 1.0492084, + "balance_loss_mlp": 1.01899838, + "epoch": 0.5278662899056094, + "flos": 19902699734400.0, + "grad_norm": 3.7453951047683973, + "language_loss": 0.7645039, + "learning_rate": 1.916674158738527e-06, + "loss": 0.78640199, + "num_input_tokens_seen": 94890090, + "step": 4390, + "time_per_iteration": 2.744598627090454 + }, + { + "auxiliary_loss_clip": 0.01154213, + "auxiliary_loss_mlp": 0.00888045, + "balance_loss_clip": 1.05157006, + "balance_loss_mlp": 1.00113845, + "epoch": 0.5279865327962484, + "flos": 18005215875840.0, + "grad_norm": 2.1369449877923197, + "language_loss": 0.60245937, + "learning_rate": 1.9158958697938506e-06, + "loss": 0.62288189, + "num_input_tokens_seen": 94908470, + "step": 4391, + "time_per_iteration": 2.7597761154174805 + }, + { + "auxiliary_loss_clip": 0.01156364, + "auxiliary_loss_mlp": 0.01033771, + "balance_loss_clip": 1.04766476, + "balance_loss_mlp": 1.02483022, + "epoch": 0.5281067756868875, + "flos": 15924443892480.0, + "grad_norm": 2.498000730446738, + "language_loss": 0.85796595, + "learning_rate": 1.9151175936077032e-06, + "loss": 0.87986732, + "num_input_tokens_seen": 94923440, + "step": 4392, + "time_per_iteration": 2.7447123527526855 + }, + { + "auxiliary_loss_clip": 0.01166468, + "auxiliary_loss_mlp": 0.01037403, + "balance_loss_clip": 1.04934001, + "balance_loss_mlp": 1.02881408, + "epoch": 0.5282270185775266, + "flos": 19426488197760.0, + "grad_norm": 1.678472827109814, + "language_loss": 0.79498065, + "learning_rate": 1.9143393302981507e-06, + "loss": 0.81701934, + "num_input_tokens_seen": 94941125, + "step": 4393, + "time_per_iteration": 2.66918683052063 + }, + { + "auxiliary_loss_clip": 0.01165885, + "auxiliary_loss_mlp": 0.01037163, + "balance_loss_clip": 1.04996789, + "balance_loss_mlp": 1.02885461, + "epoch": 0.5283472614681657, + "flos": 16399613934720.0, + "grad_norm": 1.9526694942902532, + "language_loss": 0.83382833, + "learning_rate": 1.913561079983252e-06, + "loss": 0.8558588, + "num_input_tokens_seen": 94959950, + "step": 4394, + "time_per_iteration": 2.828490972518921 + }, + { + "auxiliary_loss_clip": 0.01160547, + "auxiliary_loss_mlp": 0.01039638, + "balance_loss_clip": 1.04709673, + "balance_loss_mlp": 1.03090632, + "epoch": 0.5284675043588047, + "flos": 26760524163840.0, + "grad_norm": 2.2838891206940226, + "language_loss": 0.75269526, + "learning_rate": 1.9127828427810693e-06, + "loss": 0.77469707, + "num_input_tokens_seen": 94980515, + "step": 4395, + "time_per_iteration": 2.808870315551758 + }, + { + "auxiliary_loss_clip": 0.01161816, + "auxiliary_loss_mlp": 0.01028663, + "balance_loss_clip": 1.04708588, + "balance_loss_mlp": 1.01928782, + "epoch": 0.5285877472494439, + "flos": 19899898473600.0, + "grad_norm": 2.0018318276188, + "language_loss": 0.81019586, + "learning_rate": 1.9120046188096607e-06, + "loss": 0.83210069, + "num_input_tokens_seen": 94998560, + "step": 4396, + "time_per_iteration": 2.7615723609924316 + }, + { + "auxiliary_loss_clip": 0.01157622, + "auxiliary_loss_mlp": 0.01027973, + "balance_loss_clip": 1.04989123, + "balance_loss_mlp": 1.01855004, + "epoch": 0.528707990140083, + "flos": 20011257613440.0, + "grad_norm": 1.8891438163540184, + "language_loss": 0.74134809, + "learning_rate": 1.9112264081870804e-06, + "loss": 0.76320404, + "num_input_tokens_seen": 95016950, + "step": 4397, + "time_per_iteration": 2.7790238857269287 + }, + { + "auxiliary_loss_clip": 0.01147472, + "auxiliary_loss_mlp": 0.01028581, + "balance_loss_clip": 1.04777789, + "balance_loss_mlp": 1.01978362, + "epoch": 0.528828233030722, + "flos": 20667956014080.0, + "grad_norm": 2.2098486768739307, + "language_loss": 0.75950915, + "learning_rate": 1.9104482110313843e-06, + "loss": 0.78126967, + "num_input_tokens_seen": 95036540, + "step": 4398, + "time_per_iteration": 2.875483274459839 + }, + { + "auxiliary_loss_clip": 0.01172563, + "auxiliary_loss_mlp": 0.01035138, + "balance_loss_clip": 1.05166459, + "balance_loss_mlp": 1.02607179, + "epoch": 0.5289484759213612, + "flos": 25192448956800.0, + "grad_norm": 3.122469584519346, + "language_loss": 0.74010468, + "learning_rate": 1.909670027460623e-06, + "loss": 0.76218164, + "num_input_tokens_seen": 95053840, + "step": 4399, + "time_per_iteration": 2.836244821548462 + }, + { + "auxiliary_loss_clip": 0.01172643, + "auxiliary_loss_mlp": 0.01032677, + "balance_loss_clip": 1.05175281, + "balance_loss_mlp": 1.02342105, + "epoch": 0.5290687188120002, + "flos": 31139255715840.0, + "grad_norm": 1.8066993042840302, + "language_loss": 0.71510184, + "learning_rate": 1.908891857592847e-06, + "loss": 0.73715508, + "num_input_tokens_seen": 95074910, + "step": 4400, + "time_per_iteration": 2.8339452743530273 + }, + { + "auxiliary_loss_clip": 0.01144757, + "auxiliary_loss_mlp": 0.01032613, + "balance_loss_clip": 1.0496614, + "balance_loss_mlp": 1.02374375, + "epoch": 0.5291889617026393, + "flos": 20119851406080.0, + "grad_norm": 2.513693402187216, + "language_loss": 0.89786506, + "learning_rate": 1.9081137015461034e-06, + "loss": 0.91963875, + "num_input_tokens_seen": 95090985, + "step": 4401, + "time_per_iteration": 2.9474494457244873 + }, + { + "auxiliary_loss_clip": 0.01136547, + "auxiliary_loss_mlp": 0.01030909, + "balance_loss_clip": 1.04914236, + "balance_loss_mlp": 1.02208209, + "epoch": 0.5293092045932785, + "flos": 19643747610240.0, + "grad_norm": 1.7727502063161475, + "language_loss": 0.90288568, + "learning_rate": 1.9073355594384383e-06, + "loss": 0.92456025, + "num_input_tokens_seen": 95109225, + "step": 4402, + "time_per_iteration": 3.7841482162475586 + }, + { + "auxiliary_loss_clip": 0.0114742, + "auxiliary_loss_mlp": 0.01033205, + "balance_loss_clip": 1.04865909, + "balance_loss_mlp": 1.02404988, + "epoch": 0.5294294474839175, + "flos": 24317736958080.0, + "grad_norm": 1.8944175390557476, + "language_loss": 0.80431587, + "learning_rate": 1.906557431387895e-06, + "loss": 0.82612211, + "num_input_tokens_seen": 95128215, + "step": 4403, + "time_per_iteration": 2.856703281402588 + }, + { + "auxiliary_loss_clip": 0.011492, + "auxiliary_loss_mlp": 0.01032149, + "balance_loss_clip": 1.05178165, + "balance_loss_mlp": 1.02338147, + "epoch": 0.5295496903745566, + "flos": 18875941464960.0, + "grad_norm": 2.376971978906678, + "language_loss": 0.78915894, + "learning_rate": 1.905779317512516e-06, + "loss": 0.81097233, + "num_input_tokens_seen": 95145760, + "step": 4404, + "time_per_iteration": 3.7400612831115723 + }, + { + "auxiliary_loss_clip": 0.01171068, + "auxiliary_loss_mlp": 0.01028146, + "balance_loss_clip": 1.05185306, + "balance_loss_mlp": 1.01992631, + "epoch": 0.5296699332651957, + "flos": 20923101296640.0, + "grad_norm": 2.5660878627309334, + "language_loss": 0.8033663, + "learning_rate": 1.9050012179303385e-06, + "loss": 0.82535839, + "num_input_tokens_seen": 95164270, + "step": 4405, + "time_per_iteration": 3.6650617122650146 + }, + { + "auxiliary_loss_clip": 0.01173572, + "auxiliary_loss_mlp": 0.0104149, + "balance_loss_clip": 1.0488497, + "balance_loss_mlp": 1.03269207, + "epoch": 0.5297901761558348, + "flos": 22046745525120.0, + "grad_norm": 2.214281798841507, + "language_loss": 0.6941824, + "learning_rate": 1.904223132759401e-06, + "loss": 0.71633297, + "num_input_tokens_seen": 95182870, + "step": 4406, + "time_per_iteration": 2.6885523796081543 + }, + { + "auxiliary_loss_clip": 0.01170483, + "auxiliary_loss_mlp": 0.01031561, + "balance_loss_clip": 1.04850912, + "balance_loss_mlp": 1.02293599, + "epoch": 0.5299104190464738, + "flos": 21798495653760.0, + "grad_norm": 2.0016964556342973, + "language_loss": 0.68869114, + "learning_rate": 1.9034450621177383e-06, + "loss": 0.7107116, + "num_input_tokens_seen": 95201190, + "step": 4407, + "time_per_iteration": 3.6648504734039307 + }, + { + "auxiliary_loss_clip": 0.01175041, + "auxiliary_loss_mlp": 0.01035692, + "balance_loss_clip": 1.05392265, + "balance_loss_mlp": 1.0269835, + "epoch": 0.530030661937113, + "flos": 14720790119040.0, + "grad_norm": 2.175634273348331, + "language_loss": 0.70192051, + "learning_rate": 1.9026670061233824e-06, + "loss": 0.72402787, + "num_input_tokens_seen": 95218625, + "step": 4408, + "time_per_iteration": 2.7979989051818848 + }, + { + "auxiliary_loss_clip": 0.01160723, + "auxiliary_loss_mlp": 0.01031814, + "balance_loss_clip": 1.05069387, + "balance_loss_mlp": 1.02323723, + "epoch": 0.5301509048277521, + "flos": 21251504367360.0, + "grad_norm": 1.587252863430229, + "language_loss": 0.80258548, + "learning_rate": 1.901888964894365e-06, + "loss": 0.82451081, + "num_input_tokens_seen": 95237665, + "step": 4409, + "time_per_iteration": 2.873729705810547 + }, + { + "auxiliary_loss_clip": 0.01183486, + "auxiliary_loss_mlp": 0.01028758, + "balance_loss_clip": 1.05105686, + "balance_loss_mlp": 1.02004957, + "epoch": 0.5302711477183911, + "flos": 25957058791680.0, + "grad_norm": 2.0028111715728656, + "language_loss": 0.68059146, + "learning_rate": 1.9011109385487134e-06, + "loss": 0.70271385, + "num_input_tokens_seen": 95258915, + "step": 4410, + "time_per_iteration": 2.6923828125 + }, + { + "auxiliary_loss_clip": 0.01186905, + "auxiliary_loss_mlp": 0.0103388, + "balance_loss_clip": 1.05388641, + "balance_loss_mlp": 1.02460563, + "epoch": 0.5303913906090303, + "flos": 22273126992000.0, + "grad_norm": 2.4674896453631088, + "language_loss": 0.66181546, + "learning_rate": 1.900332927204454e-06, + "loss": 0.68402332, + "num_input_tokens_seen": 95277365, + "step": 4411, + "time_per_iteration": 2.696465015411377 + }, + { + "auxiliary_loss_clip": 0.01162806, + "auxiliary_loss_mlp": 0.01028238, + "balance_loss_clip": 1.04988623, + "balance_loss_mlp": 1.01989388, + "epoch": 0.5305116334996693, + "flos": 24936010784640.0, + "grad_norm": 1.6590440047542407, + "language_loss": 0.76928401, + "learning_rate": 1.8995549309796097e-06, + "loss": 0.7911945, + "num_input_tokens_seen": 95296670, + "step": 4412, + "time_per_iteration": 2.8483076095581055 + }, + { + "auxiliary_loss_clip": 0.01173511, + "auxiliary_loss_mlp": 0.01032888, + "balance_loss_clip": 1.0534811, + "balance_loss_mlp": 1.02391136, + "epoch": 0.5306318763903084, + "flos": 20189338266240.0, + "grad_norm": 1.6589039843834446, + "language_loss": 0.76523465, + "learning_rate": 1.8987769499922028e-06, + "loss": 0.78729862, + "num_input_tokens_seen": 95315640, + "step": 4413, + "time_per_iteration": 2.7623279094696045 + }, + { + "auxiliary_loss_clip": 0.01172063, + "auxiliary_loss_mlp": 0.00886661, + "balance_loss_clip": 1.05135179, + "balance_loss_mlp": 1.00113785, + "epoch": 0.5307521192809476, + "flos": 20266366982400.0, + "grad_norm": 2.2631701577758494, + "language_loss": 0.70629823, + "learning_rate": 1.897998984360252e-06, + "loss": 0.72688544, + "num_input_tokens_seen": 95334610, + "step": 4414, + "time_per_iteration": 2.7815115451812744 + }, + { + "auxiliary_loss_clip": 0.01156652, + "auxiliary_loss_mlp": 0.0102763, + "balance_loss_clip": 1.04692209, + "balance_loss_mlp": 1.01908255, + "epoch": 0.5308723621715866, + "flos": 28844276976000.0, + "grad_norm": 1.8741226834074973, + "language_loss": 0.78361005, + "learning_rate": 1.897221034201775e-06, + "loss": 0.80545288, + "num_input_tokens_seen": 95358350, + "step": 4415, + "time_per_iteration": 2.9019813537597656 + }, + { + "auxiliary_loss_clip": 0.01147337, + "auxiliary_loss_mlp": 0.01025973, + "balance_loss_clip": 1.04618835, + "balance_loss_mlp": 1.01740181, + "epoch": 0.5309926050622257, + "flos": 27457766040960.0, + "grad_norm": 1.4156614485353696, + "language_loss": 0.6677351, + "learning_rate": 1.8964430996347842e-06, + "loss": 0.68946815, + "num_input_tokens_seen": 95379900, + "step": 4416, + "time_per_iteration": 2.9128198623657227 + }, + { + "auxiliary_loss_clip": 0.01162362, + "auxiliary_loss_mlp": 0.01030154, + "balance_loss_clip": 1.04853213, + "balance_loss_mlp": 1.0215292, + "epoch": 0.5311128479528648, + "flos": 20514545026560.0, + "grad_norm": 1.557706181035855, + "language_loss": 0.82397294, + "learning_rate": 1.8956651807772931e-06, + "loss": 0.84589803, + "num_input_tokens_seen": 95397935, + "step": 4417, + "time_per_iteration": 2.862020492553711 + }, + { + "auxiliary_loss_clip": 0.01168858, + "auxiliary_loss_mlp": 0.010271, + "balance_loss_clip": 1.05064321, + "balance_loss_mlp": 1.01924443, + "epoch": 0.5312330908435039, + "flos": 21397660807680.0, + "grad_norm": 1.7644800254093427, + "language_loss": 0.83635974, + "learning_rate": 1.8948872777473115e-06, + "loss": 0.85831928, + "num_input_tokens_seen": 95415890, + "step": 4418, + "time_per_iteration": 2.7421181201934814 + }, + { + "auxiliary_loss_clip": 0.01164063, + "auxiliary_loss_mlp": 0.0103377, + "balance_loss_clip": 1.05084443, + "balance_loss_mlp": 1.02497792, + "epoch": 0.531353333734143, + "flos": 24717350741760.0, + "grad_norm": 1.6380681551514873, + "language_loss": 0.63783312, + "learning_rate": 1.8941093906628458e-06, + "loss": 0.6598115, + "num_input_tokens_seen": 95433675, + "step": 4419, + "time_per_iteration": 2.8797926902770996 + }, + { + "auxiliary_loss_clip": 0.01156502, + "auxiliary_loss_mlp": 0.01033425, + "balance_loss_clip": 1.04752302, + "balance_loss_mlp": 1.02468085, + "epoch": 0.531473576624782, + "flos": 30480689808000.0, + "grad_norm": 1.7614389254551759, + "language_loss": 0.71153009, + "learning_rate": 1.893331519641902e-06, + "loss": 0.73342943, + "num_input_tokens_seen": 95455820, + "step": 4420, + "time_per_iteration": 2.868339776992798 + }, + { + "auxiliary_loss_clip": 0.01140616, + "auxiliary_loss_mlp": 0.01034049, + "balance_loss_clip": 1.04362655, + "balance_loss_mlp": 1.0249418, + "epoch": 0.5315938195154212, + "flos": 23002975440000.0, + "grad_norm": 4.280759896485354, + "language_loss": 0.73926187, + "learning_rate": 1.8925536648024815e-06, + "loss": 0.7610085, + "num_input_tokens_seen": 95473240, + "step": 4421, + "time_per_iteration": 2.8119354248046875 + }, + { + "auxiliary_loss_clip": 0.01183759, + "auxiliary_loss_mlp": 0.01029901, + "balance_loss_clip": 1.05221629, + "balance_loss_mlp": 1.02138948, + "epoch": 0.5317140624060602, + "flos": 22748584343040.0, + "grad_norm": 1.94044609715366, + "language_loss": 0.76043397, + "learning_rate": 1.8917758262625849e-06, + "loss": 0.78257054, + "num_input_tokens_seen": 95493480, + "step": 4422, + "time_per_iteration": 2.6260673999786377 + }, + { + "auxiliary_loss_clip": 0.01159571, + "auxiliary_loss_mlp": 0.01035543, + "balance_loss_clip": 1.05067945, + "balance_loss_mlp": 1.02752078, + "epoch": 0.5318343052966993, + "flos": 22821087945600.0, + "grad_norm": 1.9029899714651495, + "language_loss": 0.8104614, + "learning_rate": 1.8909980041402089e-06, + "loss": 0.8324126, + "num_input_tokens_seen": 95512075, + "step": 4423, + "time_per_iteration": 2.8439738750457764 + }, + { + "auxiliary_loss_clip": 0.01164509, + "auxiliary_loss_mlp": 0.01027597, + "balance_loss_clip": 1.04783297, + "balance_loss_mlp": 1.01833427, + "epoch": 0.5319545481873384, + "flos": 13626089274240.0, + "grad_norm": 2.5989267514546905, + "language_loss": 0.65779865, + "learning_rate": 1.8902201985533494e-06, + "loss": 0.67971969, + "num_input_tokens_seen": 95529340, + "step": 4424, + "time_per_iteration": 2.7644991874694824 + }, + { + "auxiliary_loss_clip": 0.01163235, + "auxiliary_loss_mlp": 0.01027292, + "balance_loss_clip": 1.05126262, + "balance_loss_mlp": 1.01872134, + "epoch": 0.5320747910779775, + "flos": 22162522037760.0, + "grad_norm": 4.78465409802318, + "language_loss": 0.74930578, + "learning_rate": 1.8894424096199983e-06, + "loss": 0.77121103, + "num_input_tokens_seen": 95548545, + "step": 4425, + "time_per_iteration": 2.781646966934204 + }, + { + "auxiliary_loss_clip": 0.01170201, + "auxiliary_loss_mlp": 0.01028315, + "balance_loss_clip": 1.05110025, + "balance_loss_mlp": 1.02007735, + "epoch": 0.5321950339686166, + "flos": 18588081870720.0, + "grad_norm": 1.9286966519994366, + "language_loss": 0.85860419, + "learning_rate": 1.8886646374581463e-06, + "loss": 0.88058937, + "num_input_tokens_seen": 95567770, + "step": 4426, + "time_per_iteration": 2.740569829940796 + }, + { + "auxiliary_loss_clip": 0.01169921, + "auxiliary_loss_mlp": 0.01034187, + "balance_loss_clip": 1.04968441, + "balance_loss_mlp": 1.02525222, + "epoch": 0.5323152768592557, + "flos": 22856818999680.0, + "grad_norm": 1.584832596927249, + "language_loss": 0.71204567, + "learning_rate": 1.8878868821857795e-06, + "loss": 0.73408669, + "num_input_tokens_seen": 95587420, + "step": 4427, + "time_per_iteration": 2.698967933654785 + }, + { + "auxiliary_loss_clip": 0.01137033, + "auxiliary_loss_mlp": 0.01031346, + "balance_loss_clip": 1.04330873, + "balance_loss_mlp": 1.02194691, + "epoch": 0.5324355197498948, + "flos": 33948690998400.0, + "grad_norm": 2.1064059337361036, + "language_loss": 0.75077039, + "learning_rate": 1.8871091439208838e-06, + "loss": 0.7724542, + "num_input_tokens_seen": 95609030, + "step": 4428, + "time_per_iteration": 3.829662561416626 + }, + { + "auxiliary_loss_clip": 0.01139379, + "auxiliary_loss_mlp": 0.01025068, + "balance_loss_clip": 1.04618669, + "balance_loss_mlp": 1.01556647, + "epoch": 0.5325557626405338, + "flos": 23256720092160.0, + "grad_norm": 2.5330140151229754, + "language_loss": 0.77148712, + "learning_rate": 1.8863314227814414e-06, + "loss": 0.79313159, + "num_input_tokens_seen": 95627340, + "step": 4429, + "time_per_iteration": 2.902728796005249 + }, + { + "auxiliary_loss_clip": 0.01177767, + "auxiliary_loss_mlp": 0.0103814, + "balance_loss_clip": 1.05303383, + "balance_loss_mlp": 1.02853179, + "epoch": 0.532676005531173, + "flos": 26718687797760.0, + "grad_norm": 3.340603836598022, + "language_loss": 0.48697597, + "learning_rate": 1.8855537188854313e-06, + "loss": 0.50913507, + "num_input_tokens_seen": 95646315, + "step": 4430, + "time_per_iteration": 3.8020529747009277 + }, + { + "auxiliary_loss_clip": 0.0117148, + "auxiliary_loss_mlp": 0.01033281, + "balance_loss_clip": 1.04712176, + "balance_loss_mlp": 1.02520442, + "epoch": 0.5327962484218121, + "flos": 17894610921600.0, + "grad_norm": 1.9512481478911288, + "language_loss": 0.78500414, + "learning_rate": 1.8847760323508315e-06, + "loss": 0.80705178, + "num_input_tokens_seen": 95665220, + "step": 4431, + "time_per_iteration": 3.827160120010376 + }, + { + "auxiliary_loss_clip": 0.01157763, + "auxiliary_loss_mlp": 0.01028946, + "balance_loss_clip": 1.04926682, + "balance_loss_mlp": 1.0209651, + "epoch": 0.5329164913124511, + "flos": 17925385898880.0, + "grad_norm": 1.6810295636547203, + "language_loss": 0.75399452, + "learning_rate": 1.883998363295616e-06, + "loss": 0.77586162, + "num_input_tokens_seen": 95682700, + "step": 4432, + "time_per_iteration": 2.811201333999634 + }, + { + "auxiliary_loss_clip": 0.0105795, + "auxiliary_loss_mlp": 0.01009345, + "balance_loss_clip": 1.01294529, + "balance_loss_mlp": 1.00774753, + "epoch": 0.5330367342030903, + "flos": 57254178781440.0, + "grad_norm": 0.8721715584503386, + "language_loss": 0.62603486, + "learning_rate": 1.8832207118377565e-06, + "loss": 0.64670777, + "num_input_tokens_seen": 95738070, + "step": 4433, + "time_per_iteration": 4.177112579345703 + }, + { + "auxiliary_loss_clip": 0.01178318, + "auxiliary_loss_mlp": 0.01025285, + "balance_loss_clip": 1.05111539, + "balance_loss_mlp": 1.01699948, + "epoch": 0.5331569770937293, + "flos": 17420518287360.0, + "grad_norm": 1.8327313951874047, + "language_loss": 0.69271493, + "learning_rate": 1.882443078095222e-06, + "loss": 0.71475095, + "num_input_tokens_seen": 95756950, + "step": 4434, + "time_per_iteration": 2.8940322399139404 + }, + { + "auxiliary_loss_clip": 0.01047893, + "auxiliary_loss_mlp": 0.01002336, + "balance_loss_clip": 1.01309729, + "balance_loss_mlp": 1.00050056, + "epoch": 0.5332772199843684, + "flos": 56750783627520.0, + "grad_norm": 0.8819998634723445, + "language_loss": 0.66806257, + "learning_rate": 1.8816654621859794e-06, + "loss": 0.6885649, + "num_input_tokens_seen": 95816615, + "step": 4435, + "time_per_iteration": 3.37191104888916 + }, + { + "auxiliary_loss_clip": 0.01178363, + "auxiliary_loss_mlp": 0.01030265, + "balance_loss_clip": 1.05052555, + "balance_loss_mlp": 1.02161026, + "epoch": 0.5333974628750076, + "flos": 18697753071360.0, + "grad_norm": 2.1419025171543473, + "language_loss": 0.73067427, + "learning_rate": 1.8808878642279915e-06, + "loss": 0.75276059, + "num_input_tokens_seen": 95832020, + "step": 4436, + "time_per_iteration": 2.7209770679473877 + }, + { + "auxiliary_loss_clip": 0.01152162, + "auxiliary_loss_mlp": 0.01031199, + "balance_loss_clip": 1.04253089, + "balance_loss_mlp": 1.02202582, + "epoch": 0.5335177057656466, + "flos": 23805507058560.0, + "grad_norm": 3.025291864961089, + "language_loss": 0.65189952, + "learning_rate": 1.8801102843392209e-06, + "loss": 0.67373312, + "num_input_tokens_seen": 95851425, + "step": 4437, + "time_per_iteration": 2.9920172691345215 + }, + { + "auxiliary_loss_clip": 0.01150626, + "auxiliary_loss_mlp": 0.01030432, + "balance_loss_clip": 1.04630327, + "balance_loss_mlp": 1.0224514, + "epoch": 0.5336379486562857, + "flos": 25078683605760.0, + "grad_norm": 1.496157682994671, + "language_loss": 0.85163981, + "learning_rate": 1.8793327226376238e-06, + "loss": 0.8734504, + "num_input_tokens_seen": 95870745, + "step": 4438, + "time_per_iteration": 2.8320837020874023 + }, + { + "auxiliary_loss_clip": 0.01163381, + "auxiliary_loss_mlp": 0.01040166, + "balance_loss_clip": 1.04936302, + "balance_loss_mlp": 1.0316422, + "epoch": 0.5337581915469248, + "flos": 21396691140480.0, + "grad_norm": 1.7820125498106625, + "language_loss": 0.80135232, + "learning_rate": 1.8785551792411569e-06, + "loss": 0.8233878, + "num_input_tokens_seen": 95889755, + "step": 4439, + "time_per_iteration": 2.790330171585083 + }, + { + "auxiliary_loss_clip": 0.01162373, + "auxiliary_loss_mlp": 0.0102763, + "balance_loss_clip": 1.04985869, + "balance_loss_mlp": 1.01954126, + "epoch": 0.5338784344375639, + "flos": 14865905064960.0, + "grad_norm": 2.0769082752992736, + "language_loss": 0.82811487, + "learning_rate": 1.8777776542677733e-06, + "loss": 0.85001487, + "num_input_tokens_seen": 95907805, + "step": 4440, + "time_per_iteration": 2.817225694656372 + }, + { + "auxiliary_loss_clip": 0.01149591, + "auxiliary_loss_mlp": 0.01029616, + "balance_loss_clip": 1.04227984, + "balance_loss_mlp": 1.02049613, + "epoch": 0.5339986773282029, + "flos": 20813501923200.0, + "grad_norm": 1.8331595637246931, + "language_loss": 0.73169166, + "learning_rate": 1.8770001478354216e-06, + "loss": 0.75348377, + "num_input_tokens_seen": 95927480, + "step": 4441, + "time_per_iteration": 2.817392587661743 + }, + { + "auxiliary_loss_clip": 0.01168092, + "auxiliary_loss_mlp": 0.01035224, + "balance_loss_clip": 1.04905295, + "balance_loss_mlp": 1.02570498, + "epoch": 0.5341189202188421, + "flos": 17969089772160.0, + "grad_norm": 2.3316953705196815, + "language_loss": 0.83556569, + "learning_rate": 1.8762226600620504e-06, + "loss": 0.85759878, + "num_input_tokens_seen": 95946095, + "step": 4442, + "time_per_iteration": 2.770937204360962 + }, + { + "auxiliary_loss_clip": 0.0117, + "auxiliary_loss_mlp": 0.01037979, + "balance_loss_clip": 1.04813051, + "balance_loss_mlp": 1.02830553, + "epoch": 0.5342391631094812, + "flos": 11031866328960.0, + "grad_norm": 2.617103578845173, + "language_loss": 0.59209299, + "learning_rate": 1.8754451910656031e-06, + "loss": 0.61417276, + "num_input_tokens_seen": 95959995, + "step": 4443, + "time_per_iteration": 2.8849997520446777 + }, + { + "auxiliary_loss_clip": 0.01148987, + "auxiliary_loss_mlp": 0.01033542, + "balance_loss_clip": 1.045591, + "balance_loss_mlp": 1.02478063, + "epoch": 0.5343594060001202, + "flos": 15339135772800.0, + "grad_norm": 1.791780651451072, + "language_loss": 0.82747877, + "learning_rate": 1.8746677409640212e-06, + "loss": 0.84930408, + "num_input_tokens_seen": 95977095, + "step": 4444, + "time_per_iteration": 2.9023916721343994 + }, + { + "auxiliary_loss_clip": 0.01173718, + "auxiliary_loss_mlp": 0.01032931, + "balance_loss_clip": 1.05026984, + "balance_loss_mlp": 1.02370453, + "epoch": 0.5344796488907594, + "flos": 26900898514560.0, + "grad_norm": 1.7189984093498079, + "language_loss": 0.84440386, + "learning_rate": 1.8738903098752432e-06, + "loss": 0.86647034, + "num_input_tokens_seen": 95996225, + "step": 4445, + "time_per_iteration": 2.7828495502471924 + }, + { + "auxiliary_loss_clip": 0.01163294, + "auxiliary_loss_mlp": 0.01032425, + "balance_loss_clip": 1.04960823, + "balance_loss_mlp": 1.0238719, + "epoch": 0.5345998917813984, + "flos": 25411216740480.0, + "grad_norm": 2.2599970212641245, + "language_loss": 0.73406947, + "learning_rate": 1.8731128979172052e-06, + "loss": 0.75602669, + "num_input_tokens_seen": 96015425, + "step": 4446, + "time_per_iteration": 2.95418119430542 + }, + { + "auxiliary_loss_clip": 0.0115693, + "auxiliary_loss_mlp": 0.01027957, + "balance_loss_clip": 1.04731405, + "balance_loss_mlp": 1.01952267, + "epoch": 0.5347201346720375, + "flos": 32853379622400.0, + "grad_norm": 2.539656356880314, + "language_loss": 0.6715793, + "learning_rate": 1.8723355052078394e-06, + "loss": 0.69342816, + "num_input_tokens_seen": 96035460, + "step": 4447, + "time_per_iteration": 2.9691193103790283 + }, + { + "auxiliary_loss_clip": 0.01169237, + "auxiliary_loss_mlp": 0.01029174, + "balance_loss_clip": 1.04893184, + "balance_loss_mlp": 1.0203824, + "epoch": 0.5348403775626767, + "flos": 17967940536960.0, + "grad_norm": 2.2242281367228496, + "language_loss": 0.76893783, + "learning_rate": 1.8715581318650765e-06, + "loss": 0.79092199, + "num_input_tokens_seen": 96054515, + "step": 4448, + "time_per_iteration": 2.7357349395751953 + }, + { + "auxiliary_loss_clip": 0.01163313, + "auxiliary_loss_mlp": 0.01032419, + "balance_loss_clip": 1.0501734, + "balance_loss_mlp": 1.02343631, + "epoch": 0.5349606204533157, + "flos": 17603339535360.0, + "grad_norm": 2.6380482750814793, + "language_loss": 0.82059228, + "learning_rate": 1.8707807780068422e-06, + "loss": 0.84254956, + "num_input_tokens_seen": 96072330, + "step": 4449, + "time_per_iteration": 2.86620831489563 + }, + { + "auxiliary_loss_clip": 0.01163066, + "auxiliary_loss_mlp": 0.01032105, + "balance_loss_clip": 1.04910851, + "balance_loss_mlp": 1.02358723, + "epoch": 0.5350808633439548, + "flos": 29167831710720.0, + "grad_norm": 4.596850014188522, + "language_loss": 0.66312301, + "learning_rate": 1.8700034437510611e-06, + "loss": 0.68507469, + "num_input_tokens_seen": 96092425, + "step": 4450, + "time_per_iteration": 2.8684146404266357 + }, + { + "auxiliary_loss_clip": 0.01144431, + "auxiliary_loss_mlp": 0.01033284, + "balance_loss_clip": 1.04576349, + "balance_loss_mlp": 1.02308595, + "epoch": 0.5352011062345938, + "flos": 19499997381120.0, + "grad_norm": 2.098260471560275, + "language_loss": 0.81527722, + "learning_rate": 1.8692261292156549e-06, + "loss": 0.83705437, + "num_input_tokens_seen": 96111660, + "step": 4451, + "time_per_iteration": 2.891568660736084 + }, + { + "auxiliary_loss_clip": 0.011816, + "auxiliary_loss_mlp": 0.01032087, + "balance_loss_clip": 1.05441725, + "balance_loss_mlp": 1.02397501, + "epoch": 0.535321349125233, + "flos": 23477642691840.0, + "grad_norm": 2.32408338969205, + "language_loss": 0.8097406, + "learning_rate": 1.8684488345185401e-06, + "loss": 0.83187747, + "num_input_tokens_seen": 96131835, + "step": 4452, + "time_per_iteration": 2.697920560836792 + }, + { + "auxiliary_loss_clip": 0.01184117, + "auxiliary_loss_mlp": 0.01033612, + "balance_loss_clip": 1.05290723, + "balance_loss_mlp": 1.02439713, + "epoch": 0.535441592015872, + "flos": 20478059786880.0, + "grad_norm": 2.7890473937405162, + "language_loss": 0.78782165, + "learning_rate": 1.8676715597776332e-06, + "loss": 0.80999899, + "num_input_tokens_seen": 96150180, + "step": 4453, + "time_per_iteration": 2.7425971031188965 + }, + { + "auxiliary_loss_clip": 0.01132458, + "auxiliary_loss_mlp": 0.01026091, + "balance_loss_clip": 1.04322386, + "balance_loss_mlp": 1.01751399, + "epoch": 0.5355618349065111, + "flos": 19573147428480.0, + "grad_norm": 2.0688933465019455, + "language_loss": 0.76452035, + "learning_rate": 1.8668943051108455e-06, + "loss": 0.78610587, + "num_input_tokens_seen": 96167485, + "step": 4454, + "time_per_iteration": 3.726670980453491 + }, + { + "auxiliary_loss_clip": 0.01161395, + "auxiliary_loss_mlp": 0.01030051, + "balance_loss_clip": 1.04921257, + "balance_loss_mlp": 1.02093196, + "epoch": 0.5356820777971503, + "flos": 24024633978240.0, + "grad_norm": 1.9312948253498319, + "language_loss": 0.76524246, + "learning_rate": 1.8661170706360856e-06, + "loss": 0.78715694, + "num_input_tokens_seen": 96186650, + "step": 4455, + "time_per_iteration": 2.9205923080444336 + }, + { + "auxiliary_loss_clip": 0.01169162, + "auxiliary_loss_mlp": 0.01031826, + "balance_loss_clip": 1.05008543, + "balance_loss_mlp": 1.02405357, + "epoch": 0.5358023206877893, + "flos": 20884676722560.0, + "grad_norm": 1.5680276544472065, + "language_loss": 0.81489056, + "learning_rate": 1.8653398564712594e-06, + "loss": 0.83690047, + "num_input_tokens_seen": 96205595, + "step": 4456, + "time_per_iteration": 3.698071241378784 + }, + { + "auxiliary_loss_clip": 0.01170759, + "auxiliary_loss_mlp": 0.01030268, + "balance_loss_clip": 1.0521903, + "balance_loss_mlp": 1.02188814, + "epoch": 0.5359225635784284, + "flos": 22418996123520.0, + "grad_norm": 1.518746095144113, + "language_loss": 0.8243112, + "learning_rate": 1.8645626627342704e-06, + "loss": 0.84632146, + "num_input_tokens_seen": 96226360, + "step": 4457, + "time_per_iteration": 2.8066229820251465 + }, + { + "auxiliary_loss_clip": 0.0117432, + "auxiliary_loss_mlp": 0.01027776, + "balance_loss_clip": 1.04994178, + "balance_loss_mlp": 1.01931798, + "epoch": 0.5360428064690675, + "flos": 24097784025600.0, + "grad_norm": 2.0423471486374822, + "language_loss": 0.81423467, + "learning_rate": 1.8637854895430172e-06, + "loss": 0.83625567, + "num_input_tokens_seen": 96245625, + "step": 4458, + "time_per_iteration": 3.698529005050659 + }, + { + "auxiliary_loss_clip": 0.01141461, + "auxiliary_loss_mlp": 0.01034613, + "balance_loss_clip": 1.04457963, + "balance_loss_mlp": 1.0258925, + "epoch": 0.5361630493597066, + "flos": 21434505183360.0, + "grad_norm": 2.5507167213331656, + "language_loss": 0.69560838, + "learning_rate": 1.8630083370153978e-06, + "loss": 0.71736914, + "num_input_tokens_seen": 96265265, + "step": 4459, + "time_per_iteration": 3.674077033996582 + }, + { + "auxiliary_loss_clip": 0.01043349, + "auxiliary_loss_mlp": 0.01002801, + "balance_loss_clip": 1.01175237, + "balance_loss_mlp": 1.00086963, + "epoch": 0.5362832922503457, + "flos": 68888696520960.0, + "grad_norm": 0.7453432369481915, + "language_loss": 0.55336547, + "learning_rate": 1.8622312052693041e-06, + "loss": 0.57382697, + "num_input_tokens_seen": 96326445, + "step": 4460, + "time_per_iteration": 3.4651412963867188 + }, + { + "auxiliary_loss_clip": 0.01164164, + "auxiliary_loss_mlp": 0.01030695, + "balance_loss_clip": 1.04662657, + "balance_loss_mlp": 1.02179027, + "epoch": 0.5364035351409848, + "flos": 9793702563840.0, + "grad_norm": 2.5824027986658327, + "language_loss": 0.71932364, + "learning_rate": 1.8614540944226267e-06, + "loss": 0.74127221, + "num_input_tokens_seen": 96343115, + "step": 4461, + "time_per_iteration": 2.665273666381836 + }, + { + "auxiliary_loss_clip": 0.01156676, + "auxiliary_loss_mlp": 0.01029612, + "balance_loss_clip": 1.04931355, + "balance_loss_mlp": 1.02124619, + "epoch": 0.5365237780316239, + "flos": 23290080848640.0, + "grad_norm": 1.722504670615771, + "language_loss": 0.67715549, + "learning_rate": 1.8606770045932537e-06, + "loss": 0.69901836, + "num_input_tokens_seen": 96362230, + "step": 4462, + "time_per_iteration": 2.759066104888916 + }, + { + "auxiliary_loss_clip": 0.01144518, + "auxiliary_loss_mlp": 0.01034714, + "balance_loss_clip": 1.04173625, + "balance_loss_mlp": 1.02489114, + "epoch": 0.5366440209222629, + "flos": 26578133879040.0, + "grad_norm": 2.093572857680402, + "language_loss": 0.81415355, + "learning_rate": 1.859899935899068e-06, + "loss": 0.83594584, + "num_input_tokens_seen": 96382085, + "step": 4463, + "time_per_iteration": 2.7958245277404785 + }, + { + "auxiliary_loss_clip": 0.01157311, + "auxiliary_loss_mlp": 0.0102883, + "balance_loss_clip": 1.04862165, + "balance_loss_mlp": 1.02055669, + "epoch": 0.5367642638129021, + "flos": 19608052469760.0, + "grad_norm": 1.7716049011673833, + "language_loss": 0.79160297, + "learning_rate": 1.8591228884579506e-06, + "loss": 0.81346434, + "num_input_tokens_seen": 96400580, + "step": 4464, + "time_per_iteration": 2.782634973526001 + }, + { + "auxiliary_loss_clip": 0.01156255, + "auxiliary_loss_mlp": 0.01030189, + "balance_loss_clip": 1.04606497, + "balance_loss_mlp": 1.021469, + "epoch": 0.5368845067035412, + "flos": 23915214172800.0, + "grad_norm": 1.8290751762086124, + "language_loss": 0.82063377, + "learning_rate": 1.8583458623877795e-06, + "loss": 0.84249818, + "num_input_tokens_seen": 96419680, + "step": 4465, + "time_per_iteration": 2.920884847640991 + }, + { + "auxiliary_loss_clip": 0.01172399, + "auxiliary_loss_mlp": 0.01032718, + "balance_loss_clip": 1.04984593, + "balance_loss_mlp": 1.0243789, + "epoch": 0.5370047495941802, + "flos": 16873131951360.0, + "grad_norm": 1.8322477463756792, + "language_loss": 0.73298073, + "learning_rate": 1.8575688578064281e-06, + "loss": 0.75503194, + "num_input_tokens_seen": 96437805, + "step": 4466, + "time_per_iteration": 2.7288260459899902 + }, + { + "auxiliary_loss_clip": 0.01173578, + "auxiliary_loss_mlp": 0.01031758, + "balance_loss_clip": 1.05161548, + "balance_loss_mlp": 1.02353287, + "epoch": 0.5371249924848194, + "flos": 20740926493440.0, + "grad_norm": 1.9761450705020303, + "language_loss": 0.76853263, + "learning_rate": 1.8567918748317674e-06, + "loss": 0.79058605, + "num_input_tokens_seen": 96457155, + "step": 4467, + "time_per_iteration": 2.739121913909912 + }, + { + "auxiliary_loss_clip": 0.01151419, + "auxiliary_loss_mlp": 0.01030968, + "balance_loss_clip": 1.04416394, + "balance_loss_mlp": 1.02210462, + "epoch": 0.5372452353754584, + "flos": 17968120104960.0, + "grad_norm": 2.1842454747952256, + "language_loss": 0.82444328, + "learning_rate": 1.8560149135816659e-06, + "loss": 0.8462671, + "num_input_tokens_seen": 96473990, + "step": 4468, + "time_per_iteration": 2.856661319732666 + }, + { + "auxiliary_loss_clip": 0.01164636, + "auxiliary_loss_mlp": 0.01033761, + "balance_loss_clip": 1.04659629, + "balance_loss_mlp": 1.02536273, + "epoch": 0.5373654782660975, + "flos": 15377021642880.0, + "grad_norm": 2.296576367153298, + "language_loss": 0.84393972, + "learning_rate": 1.8552379741739873e-06, + "loss": 0.86592376, + "num_input_tokens_seen": 96491335, + "step": 4469, + "time_per_iteration": 2.706113576889038 + }, + { + "auxiliary_loss_clip": 0.01052938, + "auxiliary_loss_mlp": 0.00876706, + "balance_loss_clip": 1.01295972, + "balance_loss_mlp": 1.00161743, + "epoch": 0.5374857211567367, + "flos": 69000091574400.0, + "grad_norm": 0.8961283395565286, + "language_loss": 0.55674505, + "learning_rate": 1.8544610567265935e-06, + "loss": 0.57604146, + "num_input_tokens_seen": 96545275, + "step": 4470, + "time_per_iteration": 3.275766134262085 + }, + { + "auxiliary_loss_clip": 0.01164954, + "auxiliary_loss_mlp": 0.008866, + "balance_loss_clip": 1.05369353, + "balance_loss_mlp": 1.00124943, + "epoch": 0.5376059640473757, + "flos": 15085355207040.0, + "grad_norm": 2.0076193142674317, + "language_loss": 0.83293015, + "learning_rate": 1.853684161357341e-06, + "loss": 0.85344577, + "num_input_tokens_seen": 96562935, + "step": 4471, + "time_per_iteration": 2.8063981533050537 + }, + { + "auxiliary_loss_clip": 0.01167239, + "auxiliary_loss_mlp": 0.0088705, + "balance_loss_clip": 1.04922521, + "balance_loss_mlp": 1.00113475, + "epoch": 0.5377262069380148, + "flos": 19792597570560.0, + "grad_norm": 2.5111773526055643, + "language_loss": 0.76612651, + "learning_rate": 1.852907288184085e-06, + "loss": 0.78666937, + "num_input_tokens_seen": 96581820, + "step": 4472, + "time_per_iteration": 2.7290101051330566 + }, + { + "auxiliary_loss_clip": 0.01144804, + "auxiliary_loss_mlp": 0.01025276, + "balance_loss_clip": 1.04654932, + "balance_loss_mlp": 1.01635289, + "epoch": 0.5378464498286539, + "flos": 30003077640960.0, + "grad_norm": 2.5600085869091656, + "language_loss": 0.7005384, + "learning_rate": 1.8521304373246762e-06, + "loss": 0.72223926, + "num_input_tokens_seen": 96602865, + "step": 4473, + "time_per_iteration": 2.979032039642334 + }, + { + "auxiliary_loss_clip": 0.01175454, + "auxiliary_loss_mlp": 0.01031792, + "balance_loss_clip": 1.05075884, + "balance_loss_mlp": 1.02261901, + "epoch": 0.537966692719293, + "flos": 21251217058560.0, + "grad_norm": 2.4884054544988348, + "language_loss": 0.88556546, + "learning_rate": 1.8513536088969626e-06, + "loss": 0.90763789, + "num_input_tokens_seen": 96620530, + "step": 4474, + "time_per_iteration": 2.7213573455810547 + }, + { + "auxiliary_loss_clip": 0.01175053, + "auxiliary_loss_mlp": 0.01033819, + "balance_loss_clip": 1.05336308, + "balance_loss_mlp": 1.02568352, + "epoch": 0.538086935609932, + "flos": 21543170803200.0, + "grad_norm": 1.7277850304976277, + "language_loss": 0.80262566, + "learning_rate": 1.8505768030187884e-06, + "loss": 0.8247143, + "num_input_tokens_seen": 96640660, + "step": 4475, + "time_per_iteration": 2.702126979827881 + }, + { + "auxiliary_loss_clip": 0.01159008, + "auxiliary_loss_mlp": 0.01031181, + "balance_loss_clip": 1.05134511, + "balance_loss_mlp": 1.02284276, + "epoch": 0.5382071785005712, + "flos": 22747219626240.0, + "grad_norm": 1.6288578400568752, + "language_loss": 0.80110908, + "learning_rate": 1.849800019807995e-06, + "loss": 0.82301098, + "num_input_tokens_seen": 96661885, + "step": 4476, + "time_per_iteration": 2.8473470211029053 + }, + { + "auxiliary_loss_clip": 0.01151215, + "auxiliary_loss_mlp": 0.01030919, + "balance_loss_clip": 1.04606926, + "balance_loss_mlp": 1.02227652, + "epoch": 0.5383274213912103, + "flos": 24934574240640.0, + "grad_norm": 2.8917148803056367, + "language_loss": 0.71013343, + "learning_rate": 1.8490232593824186e-06, + "loss": 0.73195469, + "num_input_tokens_seen": 96678340, + "step": 4477, + "time_per_iteration": 2.86503529548645 + }, + { + "auxiliary_loss_clip": 0.01158293, + "auxiliary_loss_mlp": 0.01028637, + "balance_loss_clip": 1.04981732, + "balance_loss_mlp": 1.0204289, + "epoch": 0.5384476642818493, + "flos": 22310186849280.0, + "grad_norm": 1.5641131180376637, + "language_loss": 0.84798121, + "learning_rate": 1.8482465218598935e-06, + "loss": 0.86985058, + "num_input_tokens_seen": 96698285, + "step": 4478, + "time_per_iteration": 2.9549498558044434 + }, + { + "auxiliary_loss_clip": 0.0114988, + "auxiliary_loss_mlp": 0.01027857, + "balance_loss_clip": 1.04751003, + "balance_loss_mlp": 1.01917243, + "epoch": 0.5385679071724885, + "flos": 22711021695360.0, + "grad_norm": 2.3119163211804796, + "language_loss": 0.83097684, + "learning_rate": 1.8474698073582508e-06, + "loss": 0.85275424, + "num_input_tokens_seen": 96719655, + "step": 4479, + "time_per_iteration": 3.110300302505493 + }, + { + "auxiliary_loss_clip": 0.01158113, + "auxiliary_loss_mlp": 0.0103217, + "balance_loss_clip": 1.04762411, + "balance_loss_mlp": 1.02363515, + "epoch": 0.5386881500631275, + "flos": 15953746412160.0, + "grad_norm": 1.9143889437594332, + "language_loss": 0.8704263, + "learning_rate": 1.8466931159953166e-06, + "loss": 0.89232922, + "num_input_tokens_seen": 96736290, + "step": 4480, + "time_per_iteration": 3.802252769470215 + }, + { + "auxiliary_loss_clip": 0.01164054, + "auxiliary_loss_mlp": 0.01036691, + "balance_loss_clip": 1.05127323, + "balance_loss_mlp": 1.0279237, + "epoch": 0.5388083929537666, + "flos": 24060041809920.0, + "grad_norm": 2.0719114230561706, + "language_loss": 0.83918285, + "learning_rate": 1.8459164478889158e-06, + "loss": 0.86119026, + "num_input_tokens_seen": 96757685, + "step": 4481, + "time_per_iteration": 2.8488893508911133 + }, + { + "auxiliary_loss_clip": 0.01143191, + "auxiliary_loss_mlp": 0.01027329, + "balance_loss_clip": 1.0439651, + "balance_loss_mlp": 1.01901436, + "epoch": 0.5389286358444056, + "flos": 22236893147520.0, + "grad_norm": 1.6563515845816417, + "language_loss": 0.75740707, + "learning_rate": 1.8451398031568663e-06, + "loss": 0.77911228, + "num_input_tokens_seen": 96777310, + "step": 4482, + "time_per_iteration": 3.7845335006713867 + }, + { + "auxiliary_loss_clip": 0.01149479, + "auxiliary_loss_mlp": 0.01029583, + "balance_loss_clip": 1.04709589, + "balance_loss_mlp": 1.02066576, + "epoch": 0.5390488787350448, + "flos": 24281718595200.0, + "grad_norm": 2.5870242984162024, + "language_loss": 0.74674153, + "learning_rate": 1.844363181916986e-06, + "loss": 0.76853216, + "num_input_tokens_seen": 96798035, + "step": 4483, + "time_per_iteration": 3.9360828399658203 + }, + { + "auxiliary_loss_clip": 0.0116789, + "auxiliary_loss_mlp": 0.01030423, + "balance_loss_clip": 1.04838765, + "balance_loss_mlp": 1.0214231, + "epoch": 0.5391691216256839, + "flos": 16581393688320.0, + "grad_norm": 1.8890597813537542, + "language_loss": 0.83088642, + "learning_rate": 1.8435865842870868e-06, + "loss": 0.85286951, + "num_input_tokens_seen": 96815975, + "step": 4484, + "time_per_iteration": 2.727229118347168 + }, + { + "auxiliary_loss_clip": 0.011489, + "auxiliary_loss_mlp": 0.008864, + "balance_loss_clip": 1.0422473, + "balance_loss_mlp": 1.00105298, + "epoch": 0.5392893645163229, + "flos": 23330049707520.0, + "grad_norm": 1.8843466781498768, + "language_loss": 0.71827251, + "learning_rate": 1.8428100103849787e-06, + "loss": 0.73862553, + "num_input_tokens_seen": 96835770, + "step": 4485, + "time_per_iteration": 3.671614408493042 + }, + { + "auxiliary_loss_clip": 0.0116097, + "auxiliary_loss_mlp": 0.01032743, + "balance_loss_clip": 1.05134439, + "balance_loss_mlp": 1.02391517, + "epoch": 0.5394096074069621, + "flos": 15669801400320.0, + "grad_norm": 2.2611785680480043, + "language_loss": 0.73302197, + "learning_rate": 1.842033460328467e-06, + "loss": 0.75495911, + "num_input_tokens_seen": 96854490, + "step": 4486, + "time_per_iteration": 2.759026050567627 + }, + { + "auxiliary_loss_clip": 0.0116543, + "auxiliary_loss_mlp": 0.00886131, + "balance_loss_clip": 1.04791093, + "balance_loss_mlp": 1.00110126, + "epoch": 0.5395298502976011, + "flos": 22893447893760.0, + "grad_norm": 1.637824220373065, + "language_loss": 0.74790728, + "learning_rate": 1.8412569342353541e-06, + "loss": 0.7684229, + "num_input_tokens_seen": 96874645, + "step": 4487, + "time_per_iteration": 2.8457772731781006 + }, + { + "auxiliary_loss_clip": 0.01165596, + "auxiliary_loss_mlp": 0.0103062, + "balance_loss_clip": 1.05098486, + "balance_loss_mlp": 1.02210236, + "epoch": 0.5396500931882402, + "flos": 23842135952640.0, + "grad_norm": 1.756954084513413, + "language_loss": 0.84586871, + "learning_rate": 1.840480432223438e-06, + "loss": 0.86783093, + "num_input_tokens_seen": 96893650, + "step": 4488, + "time_per_iteration": 2.782641887664795 + }, + { + "auxiliary_loss_clip": 0.01160051, + "auxiliary_loss_mlp": 0.01036231, + "balance_loss_clip": 1.04570341, + "balance_loss_mlp": 1.02716553, + "epoch": 0.5397703360788794, + "flos": 26322988596480.0, + "grad_norm": 2.2168904445154407, + "language_loss": 0.77707875, + "learning_rate": 1.8397039544105131e-06, + "loss": 0.79904157, + "num_input_tokens_seen": 96912735, + "step": 4489, + "time_per_iteration": 2.828578472137451 + }, + { + "auxiliary_loss_clip": 0.01154987, + "auxiliary_loss_mlp": 0.0103089, + "balance_loss_clip": 1.0453192, + "balance_loss_mlp": 1.02199137, + "epoch": 0.5398905789695184, + "flos": 21214588164480.0, + "grad_norm": 1.809732979845193, + "language_loss": 0.69958115, + "learning_rate": 1.8389275009143711e-06, + "loss": 0.72143996, + "num_input_tokens_seen": 96932475, + "step": 4490, + "time_per_iteration": 2.833714246749878 + }, + { + "auxiliary_loss_clip": 0.01174936, + "auxiliary_loss_mlp": 0.01026513, + "balance_loss_clip": 1.04866683, + "balance_loss_mlp": 1.01839805, + "epoch": 0.5400108218601575, + "flos": 25080335631360.0, + "grad_norm": 1.6599312825614574, + "language_loss": 0.73346567, + "learning_rate": 1.8381510718527988e-06, + "loss": 0.75548017, + "num_input_tokens_seen": 96952085, + "step": 4491, + "time_per_iteration": 2.6571648120880127 + }, + { + "auxiliary_loss_clip": 0.01161164, + "auxiliary_loss_mlp": 0.01034696, + "balance_loss_clip": 1.04492807, + "balance_loss_mlp": 1.0259583, + "epoch": 0.5401310647507966, + "flos": 26357498588160.0, + "grad_norm": 1.7570968984530144, + "language_loss": 0.63309276, + "learning_rate": 1.8373746673435812e-06, + "loss": 0.65505141, + "num_input_tokens_seen": 96973110, + "step": 4492, + "time_per_iteration": 2.833000898361206 + }, + { + "auxiliary_loss_clip": 0.01182497, + "auxiliary_loss_mlp": 0.01035485, + "balance_loss_clip": 1.0535543, + "balance_loss_mlp": 1.02708697, + "epoch": 0.5402513076414357, + "flos": 27855332749440.0, + "grad_norm": 1.7001718571502962, + "language_loss": 0.79130995, + "learning_rate": 1.8365982875044964e-06, + "loss": 0.81348974, + "num_input_tokens_seen": 96993420, + "step": 4493, + "time_per_iteration": 2.745837450027466 + }, + { + "auxiliary_loss_clip": 0.01179978, + "auxiliary_loss_mlp": 0.00887129, + "balance_loss_clip": 1.05342174, + "balance_loss_mlp": 1.0010581, + "epoch": 0.5403715505320748, + "flos": 22893771116160.0, + "grad_norm": 8.359751143787244, + "language_loss": 0.75734824, + "learning_rate": 1.8358219324533217e-06, + "loss": 0.77801931, + "num_input_tokens_seen": 97013685, + "step": 4494, + "time_per_iteration": 2.851245641708374 + }, + { + "auxiliary_loss_clip": 0.01153194, + "auxiliary_loss_mlp": 0.01030074, + "balance_loss_clip": 1.04471564, + "balance_loss_mlp": 1.02211618, + "epoch": 0.5404917934227139, + "flos": 30224143895040.0, + "grad_norm": 1.862035744995969, + "language_loss": 0.70548612, + "learning_rate": 1.8350456023078292e-06, + "loss": 0.72731882, + "num_input_tokens_seen": 97036060, + "step": 4495, + "time_per_iteration": 2.8303043842315674 + }, + { + "auxiliary_loss_clip": 0.01184855, + "auxiliary_loss_mlp": 0.01033648, + "balance_loss_clip": 1.05235791, + "balance_loss_mlp": 1.02444494, + "epoch": 0.540612036313353, + "flos": 19938502615680.0, + "grad_norm": 2.239722137080037, + "language_loss": 0.78117442, + "learning_rate": 1.8342692971857874e-06, + "loss": 0.80335939, + "num_input_tokens_seen": 97055260, + "step": 4496, + "time_per_iteration": 2.6670031547546387 + }, + { + "auxiliary_loss_clip": 0.01156744, + "auxiliary_loss_mlp": 0.01026488, + "balance_loss_clip": 1.04731095, + "balance_loss_mlp": 1.01846564, + "epoch": 0.540732279203992, + "flos": 24279599692800.0, + "grad_norm": 2.456077036078547, + "language_loss": 0.71189523, + "learning_rate": 1.833493017204962e-06, + "loss": 0.73372757, + "num_input_tokens_seen": 97075365, + "step": 4497, + "time_per_iteration": 2.793224573135376 + }, + { + "auxiliary_loss_clip": 0.01179736, + "auxiliary_loss_mlp": 0.01025721, + "balance_loss_clip": 1.05096054, + "balance_loss_mlp": 1.01731086, + "epoch": 0.5408525220946312, + "flos": 20193216935040.0, + "grad_norm": 2.0537658889269887, + "language_loss": 0.77842438, + "learning_rate": 1.8327167624831134e-06, + "loss": 0.80047894, + "num_input_tokens_seen": 97093095, + "step": 4498, + "time_per_iteration": 2.6625735759735107 + }, + { + "auxiliary_loss_clip": 0.01182076, + "auxiliary_loss_mlp": 0.01028508, + "balance_loss_clip": 1.05414486, + "balance_loss_mlp": 1.02055073, + "epoch": 0.5409727649852702, + "flos": 24134448833280.0, + "grad_norm": 2.2846105032593433, + "language_loss": 0.7079252, + "learning_rate": 1.831940533137999e-06, + "loss": 0.73003107, + "num_input_tokens_seen": 97112000, + "step": 4499, + "time_per_iteration": 2.7361464500427246 + }, + { + "auxiliary_loss_clip": 0.01166951, + "auxiliary_loss_mlp": 0.01035131, + "balance_loss_clip": 1.05070674, + "balance_loss_mlp": 1.02663088, + "epoch": 0.5410930078759093, + "flos": 23912700220800.0, + "grad_norm": 1.7454530906440486, + "language_loss": 0.72245669, + "learning_rate": 1.8311643292873718e-06, + "loss": 0.74447751, + "num_input_tokens_seen": 97130820, + "step": 4500, + "time_per_iteration": 2.930506706237793 + }, + { + "auxiliary_loss_clip": 0.01169151, + "auxiliary_loss_mlp": 0.01026724, + "balance_loss_clip": 1.05194521, + "balance_loss_mlp": 1.01873696, + "epoch": 0.5412132507665485, + "flos": 21105132445440.0, + "grad_norm": 6.422963637483748, + "language_loss": 0.88194609, + "learning_rate": 1.8303881510489818e-06, + "loss": 0.9039048, + "num_input_tokens_seen": 97149210, + "step": 4501, + "time_per_iteration": 2.754683494567871 + }, + { + "auxiliary_loss_clip": 0.01161563, + "auxiliary_loss_mlp": 0.01028559, + "balance_loss_clip": 1.049582, + "balance_loss_mlp": 1.02082241, + "epoch": 0.5413334936571875, + "flos": 30227340205440.0, + "grad_norm": 2.3724510900921665, + "language_loss": 0.69625694, + "learning_rate": 1.829611998540574e-06, + "loss": 0.71815819, + "num_input_tokens_seen": 97170415, + "step": 4502, + "time_per_iteration": 2.8161418437957764 + }, + { + "auxiliary_loss_clip": 0.01169223, + "auxiliary_loss_mlp": 0.00886866, + "balance_loss_clip": 1.048962, + "balance_loss_mlp": 1.00104392, + "epoch": 0.5414537365478266, + "flos": 24279635606400.0, + "grad_norm": 1.701779977457509, + "language_loss": 0.79669338, + "learning_rate": 1.8288358718798914e-06, + "loss": 0.8172543, + "num_input_tokens_seen": 97189605, + "step": 4503, + "time_per_iteration": 2.83697509765625 + }, + { + "auxiliary_loss_clip": 0.01166155, + "auxiliary_loss_mlp": 0.00886699, + "balance_loss_clip": 1.05174041, + "balance_loss_mlp": 1.00113368, + "epoch": 0.5415739794384657, + "flos": 16654543735680.0, + "grad_norm": 1.9798656409330713, + "language_loss": 0.72778797, + "learning_rate": 1.8280597711846703e-06, + "loss": 0.74831647, + "num_input_tokens_seen": 97207845, + "step": 4504, + "time_per_iteration": 2.6649715900421143 + }, + { + "auxiliary_loss_clip": 0.01167935, + "auxiliary_loss_mlp": 0.01031242, + "balance_loss_clip": 1.05164289, + "balance_loss_mlp": 1.02349341, + "epoch": 0.5416942223291048, + "flos": 23185724860800.0, + "grad_norm": 2.5056297326019874, + "language_loss": 0.8329469, + "learning_rate": 1.8272836965726455e-06, + "loss": 0.85493869, + "num_input_tokens_seen": 97226780, + "step": 4505, + "time_per_iteration": 2.737598180770874 + }, + { + "auxiliary_loss_clip": 0.01128803, + "auxiliary_loss_mlp": 0.01033271, + "balance_loss_clip": 1.0409224, + "balance_loss_mlp": 1.02426457, + "epoch": 0.5418144652197439, + "flos": 20303247271680.0, + "grad_norm": 2.5397815431303994, + "language_loss": 0.78492165, + "learning_rate": 1.8265076481615461e-06, + "loss": 0.8065424, + "num_input_tokens_seen": 97246695, + "step": 4506, + "time_per_iteration": 3.7679738998413086 + }, + { + "auxiliary_loss_clip": 0.01161171, + "auxiliary_loss_mlp": 0.01030518, + "balance_loss_clip": 1.05211425, + "balance_loss_mlp": 1.0213275, + "epoch": 0.541934708110383, + "flos": 12458633431680.0, + "grad_norm": 2.6937247827613082, + "language_loss": 0.87326604, + "learning_rate": 1.8257316260690987e-06, + "loss": 0.89518297, + "num_input_tokens_seen": 97264480, + "step": 4507, + "time_per_iteration": 3.694936752319336 + }, + { + "auxiliary_loss_clip": 0.01165971, + "auxiliary_loss_mlp": 0.01032888, + "balance_loss_clip": 1.05063581, + "balance_loss_mlp": 1.0250144, + "epoch": 0.5420549510010221, + "flos": 21253802837760.0, + "grad_norm": 1.690534342559091, + "language_loss": 0.75830042, + "learning_rate": 1.8249556304130254e-06, + "loss": 0.78028893, + "num_input_tokens_seen": 97285760, + "step": 4508, + "time_per_iteration": 2.7326958179473877 + }, + { + "auxiliary_loss_clip": 0.01150501, + "auxiliary_loss_mlp": 0.01029943, + "balance_loss_clip": 1.04579949, + "balance_loss_mlp": 1.02150917, + "epoch": 0.5421751938916611, + "flos": 29490524519040.0, + "grad_norm": 2.8707133002257823, + "language_loss": 0.68573761, + "learning_rate": 1.824179661311044e-06, + "loss": 0.70754206, + "num_input_tokens_seen": 97304510, + "step": 4509, + "time_per_iteration": 2.8017654418945312 + }, + { + "auxiliary_loss_clip": 0.0114025, + "auxiliary_loss_mlp": 0.01027706, + "balance_loss_clip": 1.04110873, + "balance_loss_mlp": 1.01892018, + "epoch": 0.5422954367823003, + "flos": 18734238311040.0, + "grad_norm": 1.9300109297815036, + "language_loss": 0.80060863, + "learning_rate": 1.823403718880868e-06, + "loss": 0.82228816, + "num_input_tokens_seen": 97323270, + "step": 4510, + "time_per_iteration": 3.9443202018737793 + }, + { + "auxiliary_loss_clip": 0.01158234, + "auxiliary_loss_mlp": 0.0103194, + "balance_loss_clip": 1.04439497, + "balance_loss_mlp": 1.02345872, + "epoch": 0.5424156796729394, + "flos": 39969006940800.0, + "grad_norm": 1.5884304368039244, + "language_loss": 0.66720945, + "learning_rate": 1.822627803240207e-06, + "loss": 0.68911117, + "num_input_tokens_seen": 97345600, + "step": 4511, + "time_per_iteration": 4.000308036804199 + }, + { + "auxiliary_loss_clip": 0.01151308, + "auxiliary_loss_mlp": 0.01031562, + "balance_loss_clip": 1.04574001, + "balance_loss_mlp": 1.02330053, + "epoch": 0.5425359225635784, + "flos": 11546538353280.0, + "grad_norm": 2.259610517676277, + "language_loss": 0.85597312, + "learning_rate": 1.8218519145067675e-06, + "loss": 0.87780184, + "num_input_tokens_seen": 97361220, + "step": 4512, + "time_per_iteration": 2.8480136394500732 + }, + { + "auxiliary_loss_clip": 0.01149837, + "auxiliary_loss_mlp": 0.01026794, + "balance_loss_clip": 1.04746795, + "balance_loss_mlp": 1.01833582, + "epoch": 0.5426561654542175, + "flos": 20229702174720.0, + "grad_norm": 2.059929608904119, + "language_loss": 0.89483124, + "learning_rate": 1.8210760527982508e-06, + "loss": 0.91659755, + "num_input_tokens_seen": 97381505, + "step": 4513, + "time_per_iteration": 2.862379312515259 + }, + { + "auxiliary_loss_clip": 0.01162812, + "auxiliary_loss_mlp": 0.00885797, + "balance_loss_clip": 1.05098271, + "balance_loss_mlp": 1.00102782, + "epoch": 0.5427764083448566, + "flos": 21871681614720.0, + "grad_norm": 1.7771959673237334, + "language_loss": 0.75241059, + "learning_rate": 1.8203002182323552e-06, + "loss": 0.77289671, + "num_input_tokens_seen": 97399060, + "step": 4514, + "time_per_iteration": 2.793344259262085 + }, + { + "auxiliary_loss_clip": 0.01161958, + "auxiliary_loss_mlp": 0.01031592, + "balance_loss_clip": 1.0495522, + "balance_loss_mlp": 1.02307463, + "epoch": 0.5428966512354957, + "flos": 19640946349440.0, + "grad_norm": 2.5525903148575324, + "language_loss": 0.75430346, + "learning_rate": 1.819524410926773e-06, + "loss": 0.77623892, + "num_input_tokens_seen": 97416740, + "step": 4515, + "time_per_iteration": 2.7632737159729004 + }, + { + "auxiliary_loss_clip": 0.01132724, + "auxiliary_loss_mlp": 0.01035685, + "balance_loss_clip": 1.04651785, + "balance_loss_mlp": 1.02713799, + "epoch": 0.5430168941261347, + "flos": 22382187661440.0, + "grad_norm": 2.50100706062157, + "language_loss": 0.77018738, + "learning_rate": 1.8187486309991944e-06, + "loss": 0.79187149, + "num_input_tokens_seen": 97437620, + "step": 4516, + "time_per_iteration": 2.9232070446014404 + }, + { + "auxiliary_loss_clip": 0.01169829, + "auxiliary_loss_mlp": 0.0102811, + "balance_loss_clip": 1.05183363, + "balance_loss_mlp": 1.01922309, + "epoch": 0.5431371370167739, + "flos": 18764187275520.0, + "grad_norm": 1.7072441764996753, + "language_loss": 0.77291614, + "learning_rate": 1.817972878567304e-06, + "loss": 0.79489553, + "num_input_tokens_seen": 97456275, + "step": 4517, + "time_per_iteration": 2.8567848205566406 + }, + { + "auxiliary_loss_clip": 0.01168273, + "auxiliary_loss_mlp": 0.01036967, + "balance_loss_clip": 1.04873276, + "balance_loss_mlp": 1.02872944, + "epoch": 0.543257379907413, + "flos": 18806023641600.0, + "grad_norm": 1.8278703781527266, + "language_loss": 0.7618193, + "learning_rate": 1.8171971537487834e-06, + "loss": 0.78387171, + "num_input_tokens_seen": 97474925, + "step": 4518, + "time_per_iteration": 2.8502120971679688 + }, + { + "auxiliary_loss_clip": 0.0118326, + "auxiliary_loss_mlp": 0.01031324, + "balance_loss_clip": 1.05314589, + "balance_loss_mlp": 1.02243066, + "epoch": 0.543377622798052, + "flos": 17493381025920.0, + "grad_norm": 1.800545642252038, + "language_loss": 0.80845428, + "learning_rate": 1.8164214566613093e-06, + "loss": 0.83060014, + "num_input_tokens_seen": 97493550, + "step": 4519, + "time_per_iteration": 2.7195487022399902 + }, + { + "auxiliary_loss_clip": 0.01178567, + "auxiliary_loss_mlp": 0.01032135, + "balance_loss_clip": 1.05153358, + "balance_loss_mlp": 1.02416623, + "epoch": 0.5434978656886912, + "flos": 18989311766400.0, + "grad_norm": 3.2369028105820443, + "language_loss": 0.66223061, + "learning_rate": 1.8156457874225547e-06, + "loss": 0.68433768, + "num_input_tokens_seen": 97512010, + "step": 4520, + "time_per_iteration": 2.676459312438965 + }, + { + "auxiliary_loss_clip": 0.01155672, + "auxiliary_loss_mlp": 0.01027367, + "balance_loss_clip": 1.05151463, + "balance_loss_mlp": 1.01896858, + "epoch": 0.5436181085793302, + "flos": 17274936464640.0, + "grad_norm": 1.7631260751641942, + "language_loss": 0.8035301, + "learning_rate": 1.814870146150187e-06, + "loss": 0.82536054, + "num_input_tokens_seen": 97530120, + "step": 4521, + "time_per_iteration": 2.79710054397583 + }, + { + "auxiliary_loss_clip": 0.01162614, + "auxiliary_loss_mlp": 0.01027682, + "balance_loss_clip": 1.04820442, + "balance_loss_mlp": 1.01914072, + "epoch": 0.5437383514699693, + "flos": 19098587917440.0, + "grad_norm": 1.9283851199967699, + "language_loss": 0.78714955, + "learning_rate": 1.814094532961871e-06, + "loss": 0.80905253, + "num_input_tokens_seen": 97548695, + "step": 4522, + "time_per_iteration": 2.7154645919799805 + }, + { + "auxiliary_loss_clip": 0.01142797, + "auxiliary_loss_mlp": 0.01025349, + "balance_loss_clip": 1.0433706, + "balance_loss_mlp": 1.01693273, + "epoch": 0.5438585943606085, + "flos": 22602715211520.0, + "grad_norm": 1.9716372259765593, + "language_loss": 0.83518457, + "learning_rate": 1.8133189479752666e-06, + "loss": 0.856866, + "num_input_tokens_seen": 97567625, + "step": 4523, + "time_per_iteration": 2.9021475315093994 + }, + { + "auxiliary_loss_clip": 0.01178642, + "auxiliary_loss_mlp": 0.01030674, + "balance_loss_clip": 1.05199194, + "balance_loss_mlp": 1.02258551, + "epoch": 0.5439788372512475, + "flos": 21798495653760.0, + "grad_norm": 2.9580207272154615, + "language_loss": 0.81768358, + "learning_rate": 1.8125433913080292e-06, + "loss": 0.83977675, + "num_input_tokens_seen": 97585325, + "step": 4524, + "time_per_iteration": 2.6594629287719727 + }, + { + "auxiliary_loss_clip": 0.0109465, + "auxiliary_loss_mlp": 0.01029827, + "balance_loss_clip": 1.03830993, + "balance_loss_mlp": 1.02160716, + "epoch": 0.5440990801418866, + "flos": 16399362539520.0, + "grad_norm": 2.0170428663605646, + "language_loss": 0.8303982, + "learning_rate": 1.811767863077811e-06, + "loss": 0.85164297, + "num_input_tokens_seen": 97604275, + "step": 4525, + "time_per_iteration": 3.0961976051330566 + }, + { + "auxiliary_loss_clip": 0.011232, + "auxiliary_loss_mlp": 0.01027096, + "balance_loss_clip": 1.04438043, + "balance_loss_mlp": 1.01857209, + "epoch": 0.5442193230325257, + "flos": 21615638492160.0, + "grad_norm": 1.547670994116629, + "language_loss": 0.78225911, + "learning_rate": 1.8109923634022577e-06, + "loss": 0.80376208, + "num_input_tokens_seen": 97624300, + "step": 4526, + "time_per_iteration": 3.227540969848633 + }, + { + "auxiliary_loss_clip": 0.01182624, + "auxiliary_loss_mlp": 0.01028599, + "balance_loss_clip": 1.05247164, + "balance_loss_mlp": 1.02023685, + "epoch": 0.5443395659231648, + "flos": 15481198062720.0, + "grad_norm": 2.1265320427264, + "language_loss": 0.86518908, + "learning_rate": 1.8102168923990128e-06, + "loss": 0.88730139, + "num_input_tokens_seen": 97637845, + "step": 4527, + "time_per_iteration": 2.632024049758911 + }, + { + "auxiliary_loss_clip": 0.01173509, + "auxiliary_loss_mlp": 0.00885895, + "balance_loss_clip": 1.05251825, + "balance_loss_mlp": 1.00087869, + "epoch": 0.5444598088138038, + "flos": 18770436241920.0, + "grad_norm": 1.9493188864748516, + "language_loss": 0.79816651, + "learning_rate": 1.809441450185714e-06, + "loss": 0.81876057, + "num_input_tokens_seen": 97656330, + "step": 4528, + "time_per_iteration": 2.775696277618408 + }, + { + "auxiliary_loss_clip": 0.01164061, + "auxiliary_loss_mlp": 0.01025446, + "balance_loss_clip": 1.04550815, + "balance_loss_mlp": 1.01716137, + "epoch": 0.544580051704443, + "flos": 21142335957120.0, + "grad_norm": 2.003889052807076, + "language_loss": 0.73573971, + "learning_rate": 1.8086660368799958e-06, + "loss": 0.75763476, + "num_input_tokens_seen": 97674380, + "step": 4529, + "time_per_iteration": 2.755821466445923 + }, + { + "auxiliary_loss_clip": 0.01164038, + "auxiliary_loss_mlp": 0.01027908, + "balance_loss_clip": 1.051162, + "balance_loss_mlp": 1.01905036, + "epoch": 0.5447002945950821, + "flos": 32491508054400.0, + "grad_norm": 1.8655396872772765, + "language_loss": 0.77490985, + "learning_rate": 1.807890652599488e-06, + "loss": 0.79682934, + "num_input_tokens_seen": 97698765, + "step": 4530, + "time_per_iteration": 2.8641884326934814 + }, + { + "auxiliary_loss_clip": 0.01177118, + "auxiliary_loss_mlp": 0.01029444, + "balance_loss_clip": 1.05174613, + "balance_loss_mlp": 1.02178216, + "epoch": 0.5448205374857211, + "flos": 11798307757440.0, + "grad_norm": 2.148842487967249, + "language_loss": 0.82949865, + "learning_rate": 1.8071152974618156e-06, + "loss": 0.85156429, + "num_input_tokens_seen": 97716565, + "step": 4531, + "time_per_iteration": 2.6703906059265137 + }, + { + "auxiliary_loss_clip": 0.01152491, + "auxiliary_loss_mlp": 0.00886036, + "balance_loss_clip": 1.04513097, + "balance_loss_mlp": 1.00082588, + "epoch": 0.5449407803763603, + "flos": 24133766474880.0, + "grad_norm": 2.134836449952924, + "language_loss": 0.78272325, + "learning_rate": 1.806339971584599e-06, + "loss": 0.80310851, + "num_input_tokens_seen": 97733225, + "step": 4532, + "time_per_iteration": 3.7524659633636475 + }, + { + "auxiliary_loss_clip": 0.01180068, + "auxiliary_loss_mlp": 0.01029181, + "balance_loss_clip": 1.05181384, + "balance_loss_mlp": 1.02107513, + "epoch": 0.5450610232669993, + "flos": 23258551685760.0, + "grad_norm": 1.8059193229730326, + "language_loss": 0.85249031, + "learning_rate": 1.8055646750854546e-06, + "loss": 0.87458277, + "num_input_tokens_seen": 97752735, + "step": 4533, + "time_per_iteration": 3.5825419425964355 + }, + { + "auxiliary_loss_clip": 0.01167037, + "auxiliary_loss_mlp": 0.01030805, + "balance_loss_clip": 1.05086279, + "balance_loss_mlp": 1.02278781, + "epoch": 0.5451812661576384, + "flos": 17785083375360.0, + "grad_norm": 3.6608887430363555, + "language_loss": 0.81427425, + "learning_rate": 1.8047894080819945e-06, + "loss": 0.83625269, + "num_input_tokens_seen": 97769985, + "step": 4534, + "time_per_iteration": 2.7938601970672607 + }, + { + "auxiliary_loss_clip": 0.01075752, + "auxiliary_loss_mlp": 0.0100165, + "balance_loss_clip": 1.01813197, + "balance_loss_mlp": 1.00014782, + "epoch": 0.5453015090482776, + "flos": 71062586513280.0, + "grad_norm": 0.7351486801359699, + "language_loss": 0.63212037, + "learning_rate": 1.8040141706918258e-06, + "loss": 0.65289438, + "num_input_tokens_seen": 97831225, + "step": 4535, + "time_per_iteration": 4.258888483047485 + }, + { + "auxiliary_loss_clip": 0.01153846, + "auxiliary_loss_mlp": 0.01031172, + "balance_loss_clip": 1.0496887, + "balance_loss_mlp": 1.02308977, + "epoch": 0.5454217519389166, + "flos": 25552201622400.0, + "grad_norm": 1.9111355477205323, + "language_loss": 0.77188706, + "learning_rate": 1.8032389630325525e-06, + "loss": 0.79373717, + "num_input_tokens_seen": 97849975, + "step": 4536, + "time_per_iteration": 2.779299020767212 + }, + { + "auxiliary_loss_clip": 0.01156619, + "auxiliary_loss_mlp": 0.01029275, + "balance_loss_clip": 1.04411113, + "balance_loss_mlp": 1.02169669, + "epoch": 0.5455419948295557, + "flos": 23658345037440.0, + "grad_norm": 1.6964609934547572, + "language_loss": 0.75635314, + "learning_rate": 1.8024637852217707e-06, + "loss": 0.77821207, + "num_input_tokens_seen": 97869700, + "step": 4537, + "time_per_iteration": 2.740128517150879 + }, + { + "auxiliary_loss_clip": 0.01162032, + "auxiliary_loss_mlp": 0.01038003, + "balance_loss_clip": 1.05114329, + "balance_loss_mlp": 1.0299741, + "epoch": 0.5456622377201948, + "flos": 23403989854080.0, + "grad_norm": 1.7796960664290604, + "language_loss": 0.84570849, + "learning_rate": 1.8016886373770766e-06, + "loss": 0.86770892, + "num_input_tokens_seen": 97888215, + "step": 4538, + "time_per_iteration": 3.7292098999023438 + }, + { + "auxiliary_loss_clip": 0.01160556, + "auxiliary_loss_mlp": 0.01028704, + "balance_loss_clip": 1.04904795, + "balance_loss_mlp": 1.02082467, + "epoch": 0.5457824806108339, + "flos": 23988040997760.0, + "grad_norm": 1.5798180777064985, + "language_loss": 0.79026037, + "learning_rate": 1.8009135196160579e-06, + "loss": 0.81215298, + "num_input_tokens_seen": 97907090, + "step": 4539, + "time_per_iteration": 2.827754497528076 + }, + { + "auxiliary_loss_clip": 0.01146247, + "auxiliary_loss_mlp": 0.01031964, + "balance_loss_clip": 1.04459703, + "balance_loss_mlp": 1.02441263, + "epoch": 0.545902723501473, + "flos": 22565870835840.0, + "grad_norm": 2.165580095750666, + "language_loss": 0.8424657, + "learning_rate": 1.8001384320563e-06, + "loss": 0.8642478, + "num_input_tokens_seen": 97927345, + "step": 4540, + "time_per_iteration": 2.8422188758850098 + }, + { + "auxiliary_loss_clip": 0.01074946, + "auxiliary_loss_mlp": 0.0100162, + "balance_loss_clip": 1.01766753, + "balance_loss_mlp": 1.00017786, + "epoch": 0.5460229663921121, + "flos": 55198399685760.0, + "grad_norm": 0.8043692200716696, + "language_loss": 0.57830971, + "learning_rate": 1.7993633748153833e-06, + "loss": 0.59907538, + "num_input_tokens_seen": 97981950, + "step": 4541, + "time_per_iteration": 3.199612855911255 + }, + { + "auxiliary_loss_clip": 0.01175878, + "auxiliary_loss_mlp": 0.01039152, + "balance_loss_clip": 1.05204272, + "balance_loss_mlp": 1.03067923, + "epoch": 0.5461432092827512, + "flos": 15413866018560.0, + "grad_norm": 1.965208961753001, + "language_loss": 0.72846645, + "learning_rate": 1.7985883480108834e-06, + "loss": 0.75061679, + "num_input_tokens_seen": 97999585, + "step": 4542, + "time_per_iteration": 2.742565155029297 + }, + { + "auxiliary_loss_clip": 0.01164944, + "auxiliary_loss_mlp": 0.0102695, + "balance_loss_clip": 1.04825187, + "balance_loss_mlp": 1.01879621, + "epoch": 0.5462634521733902, + "flos": 24024921287040.0, + "grad_norm": 9.476333541820486, + "language_loss": 0.72592849, + "learning_rate": 1.797813351760371e-06, + "loss": 0.74784744, + "num_input_tokens_seen": 98021290, + "step": 4543, + "time_per_iteration": 2.7523460388183594 + }, + { + "auxiliary_loss_clip": 0.01180303, + "auxiliary_loss_mlp": 0.01032113, + "balance_loss_clip": 1.05097413, + "balance_loss_mlp": 1.02355933, + "epoch": 0.5463836950640293, + "flos": 22820944291200.0, + "grad_norm": 2.085033532884189, + "language_loss": 0.77677429, + "learning_rate": 1.7970383861814116e-06, + "loss": 0.79889846, + "num_input_tokens_seen": 98041060, + "step": 4544, + "time_per_iteration": 2.6556782722473145 + }, + { + "auxiliary_loss_clip": 0.01173506, + "auxiliary_loss_mlp": 0.01030038, + "balance_loss_clip": 1.05396271, + "balance_loss_mlp": 1.02199149, + "epoch": 0.5465039379546685, + "flos": 20448290390400.0, + "grad_norm": 1.9715740441726144, + "language_loss": 0.74001247, + "learning_rate": 1.7962634513915684e-06, + "loss": 0.76204789, + "num_input_tokens_seen": 98058410, + "step": 4545, + "time_per_iteration": 2.725921869277954 + }, + { + "auxiliary_loss_clip": 0.0118018, + "auxiliary_loss_mlp": 0.01025693, + "balance_loss_clip": 1.05227065, + "balance_loss_mlp": 1.01771235, + "epoch": 0.5466241808453075, + "flos": 17343310003200.0, + "grad_norm": 2.84622899371222, + "language_loss": 0.79598093, + "learning_rate": 1.7954885475083969e-06, + "loss": 0.81803966, + "num_input_tokens_seen": 98076080, + "step": 4546, + "time_per_iteration": 2.609161853790283 + }, + { + "auxiliary_loss_clip": 0.01187831, + "auxiliary_loss_mlp": 0.01033507, + "balance_loss_clip": 1.05676508, + "balance_loss_mlp": 1.02512062, + "epoch": 0.5467444237359466, + "flos": 21617039122560.0, + "grad_norm": 2.1007058963504837, + "language_loss": 0.72826487, + "learning_rate": 1.7947136746494513e-06, + "loss": 0.75047827, + "num_input_tokens_seen": 98096995, + "step": 4547, + "time_per_iteration": 2.6961755752563477 + }, + { + "auxiliary_loss_clip": 0.01170433, + "auxiliary_loss_mlp": 0.01026378, + "balance_loss_clip": 1.05117583, + "balance_loss_mlp": 1.0179472, + "epoch": 0.5468646666265857, + "flos": 24170467196160.0, + "grad_norm": 2.0534810858705774, + "language_loss": 0.87718642, + "learning_rate": 1.793938832932277e-06, + "loss": 0.89915454, + "num_input_tokens_seen": 98115105, + "step": 4548, + "time_per_iteration": 2.7085013389587402 + }, + { + "auxiliary_loss_clip": 0.01179314, + "auxiliary_loss_mlp": 0.01030034, + "balance_loss_clip": 1.0504899, + "balance_loss_mlp": 1.02176678, + "epoch": 0.5469849095172248, + "flos": 27527001505920.0, + "grad_norm": 1.8121771317869906, + "language_loss": 0.70013028, + "learning_rate": 1.7931640224744185e-06, + "loss": 0.72222376, + "num_input_tokens_seen": 98135655, + "step": 4549, + "time_per_iteration": 2.7073562145233154 + }, + { + "auxiliary_loss_clip": 0.01138835, + "auxiliary_loss_mlp": 0.01023661, + "balance_loss_clip": 1.04136419, + "balance_loss_mlp": 1.01599622, + "epoch": 0.5471051524078638, + "flos": 27964680727680.0, + "grad_norm": 1.6454828274294824, + "language_loss": 0.73668683, + "learning_rate": 1.7923892433934127e-06, + "loss": 0.75831175, + "num_input_tokens_seen": 98156730, + "step": 4550, + "time_per_iteration": 2.89625883102417 + }, + { + "auxiliary_loss_clip": 0.01163128, + "auxiliary_loss_mlp": 0.00886769, + "balance_loss_clip": 1.05012822, + "balance_loss_mlp": 1.00073647, + "epoch": 0.547225395298503, + "flos": 18150510389760.0, + "grad_norm": 2.2413862417003334, + "language_loss": 0.78890866, + "learning_rate": 1.7916144958067939e-06, + "loss": 0.80940765, + "num_input_tokens_seen": 98174590, + "step": 4551, + "time_per_iteration": 2.715841054916382 + }, + { + "auxiliary_loss_clip": 0.01169727, + "auxiliary_loss_mlp": 0.01029994, + "balance_loss_clip": 1.049016, + "balance_loss_mlp": 1.02150047, + "epoch": 0.5473456381891421, + "flos": 21361498790400.0, + "grad_norm": 1.6171627174431964, + "language_loss": 0.78600115, + "learning_rate": 1.7908397798320905e-06, + "loss": 0.80799836, + "num_input_tokens_seen": 98194325, + "step": 4552, + "time_per_iteration": 2.79813814163208 + }, + { + "auxiliary_loss_clip": 0.01170968, + "auxiliary_loss_mlp": 0.00886747, + "balance_loss_clip": 1.04979277, + "balance_loss_mlp": 1.00077844, + "epoch": 0.5474658810797811, + "flos": 19932145908480.0, + "grad_norm": 1.8446348825402699, + "language_loss": 0.75171489, + "learning_rate": 1.7900650955868265e-06, + "loss": 0.77229202, + "num_input_tokens_seen": 98213970, + "step": 4553, + "time_per_iteration": 2.7484052181243896 + }, + { + "auxiliary_loss_clip": 0.01171066, + "auxiliary_loss_mlp": 0.00885949, + "balance_loss_clip": 1.05211496, + "balance_loss_mlp": 1.00074553, + "epoch": 0.5475861239704203, + "flos": 50476217264640.0, + "grad_norm": 1.384053309302849, + "language_loss": 0.76549256, + "learning_rate": 1.7892904431885202e-06, + "loss": 0.78606266, + "num_input_tokens_seen": 98241145, + "step": 4554, + "time_per_iteration": 2.9704043865203857 + }, + { + "auxiliary_loss_clip": 0.01142287, + "auxiliary_loss_mlp": 0.01025984, + "balance_loss_clip": 1.04306674, + "balance_loss_mlp": 1.01841974, + "epoch": 0.5477063668610593, + "flos": 20705123612160.0, + "grad_norm": 1.6183037393101254, + "language_loss": 0.75337052, + "learning_rate": 1.788515822754686e-06, + "loss": 0.77505326, + "num_input_tokens_seen": 98261565, + "step": 4555, + "time_per_iteration": 2.824017286300659 + }, + { + "auxiliary_loss_clip": 0.01151426, + "auxiliary_loss_mlp": 0.01033718, + "balance_loss_clip": 1.04470694, + "balance_loss_mlp": 1.02472973, + "epoch": 0.5478266097516984, + "flos": 19609740408960.0, + "grad_norm": 2.244944735473245, + "language_loss": 0.78368306, + "learning_rate": 1.7877412344028335e-06, + "loss": 0.80553454, + "num_input_tokens_seen": 98281370, + "step": 4556, + "time_per_iteration": 2.842696189880371 + }, + { + "auxiliary_loss_clip": 0.01171735, + "auxiliary_loss_mlp": 0.01023545, + "balance_loss_clip": 1.04898238, + "balance_loss_mlp": 1.01546264, + "epoch": 0.5479468526423376, + "flos": 12896599962240.0, + "grad_norm": 2.153262400035492, + "language_loss": 0.77590406, + "learning_rate": 1.7869666782504668e-06, + "loss": 0.79785681, + "num_input_tokens_seen": 98297950, + "step": 4557, + "time_per_iteration": 2.7069482803344727 + }, + { + "auxiliary_loss_clip": 0.01147267, + "auxiliary_loss_mlp": 0.01026382, + "balance_loss_clip": 1.04354608, + "balance_loss_mlp": 1.01867461, + "epoch": 0.5480670955329766, + "flos": 18588800142720.0, + "grad_norm": 1.7884298124733626, + "language_loss": 0.68625176, + "learning_rate": 1.7861921544150867e-06, + "loss": 0.70798826, + "num_input_tokens_seen": 98316800, + "step": 4558, + "time_per_iteration": 4.652795791625977 + }, + { + "auxiliary_loss_clip": 0.01119016, + "auxiliary_loss_mlp": 0.00885553, + "balance_loss_clip": 1.04358077, + "balance_loss_mlp": 1.00065243, + "epoch": 0.5481873384236157, + "flos": 15954608338560.0, + "grad_norm": 1.7999041239681466, + "language_loss": 0.76061243, + "learning_rate": 1.7854176630141856e-06, + "loss": 0.78065813, + "num_input_tokens_seen": 98333935, + "step": 4559, + "time_per_iteration": 2.899749994277954 + }, + { + "auxiliary_loss_clip": 0.01184422, + "auxiliary_loss_mlp": 0.01029122, + "balance_loss_clip": 1.05414391, + "balance_loss_mlp": 1.02008629, + "epoch": 0.5483075813142548, + "flos": 22783812606720.0, + "grad_norm": 2.317838703802387, + "language_loss": 0.8428691, + "learning_rate": 1.784643204165255e-06, + "loss": 0.86500454, + "num_input_tokens_seen": 98353255, + "step": 4560, + "time_per_iteration": 2.662886142730713 + }, + { + "auxiliary_loss_clip": 0.01167813, + "auxiliary_loss_mlp": 0.01032139, + "balance_loss_clip": 1.05094981, + "balance_loss_mlp": 1.02390742, + "epoch": 0.5484278242048939, + "flos": 19317212046720.0, + "grad_norm": 2.010730282495693, + "language_loss": 0.77311331, + "learning_rate": 1.7838687779857783e-06, + "loss": 0.79511285, + "num_input_tokens_seen": 98371130, + "step": 4561, + "time_per_iteration": 3.6963112354278564 + }, + { + "auxiliary_loss_clip": 0.01151034, + "auxiliary_loss_mlp": 0.01024246, + "balance_loss_clip": 1.04505515, + "balance_loss_mlp": 1.01665235, + "epoch": 0.5485480670955329, + "flos": 22816024128000.0, + "grad_norm": 1.7467696021859422, + "language_loss": 0.64186442, + "learning_rate": 1.7830943845932366e-06, + "loss": 0.66361719, + "num_input_tokens_seen": 98390455, + "step": 4562, + "time_per_iteration": 2.8344974517822266 + }, + { + "auxiliary_loss_clip": 0.01166646, + "auxiliary_loss_mlp": 0.01029037, + "balance_loss_clip": 1.05109024, + "balance_loss_mlp": 1.02058494, + "epoch": 0.5486683099861721, + "flos": 22671304231680.0, + "grad_norm": 1.7422630090890012, + "language_loss": 0.7528249, + "learning_rate": 1.7823200241051044e-06, + "loss": 0.7747817, + "num_input_tokens_seen": 98409370, + "step": 4563, + "time_per_iteration": 2.764832019805908 + }, + { + "auxiliary_loss_clip": 0.01177835, + "auxiliary_loss_mlp": 0.01030615, + "balance_loss_clip": 1.05005646, + "balance_loss_mlp": 1.02291989, + "epoch": 0.5487885528768112, + "flos": 23149383275520.0, + "grad_norm": 2.175114767636336, + "language_loss": 0.80329734, + "learning_rate": 1.7815456966388513e-06, + "loss": 0.82538188, + "num_input_tokens_seen": 98428465, + "step": 4564, + "time_per_iteration": 3.649787425994873 + }, + { + "auxiliary_loss_clip": 0.0114445, + "auxiliary_loss_mlp": 0.01029345, + "balance_loss_clip": 1.04489708, + "balance_loss_mlp": 1.02053547, + "epoch": 0.5489087957674502, + "flos": 22053928245120.0, + "grad_norm": 2.3141004480909872, + "language_loss": 0.80853319, + "learning_rate": 1.780771402311943e-06, + "loss": 0.83027112, + "num_input_tokens_seen": 98447300, + "step": 4565, + "time_per_iteration": 2.8123364448547363 + }, + { + "auxiliary_loss_clip": 0.01163714, + "auxiliary_loss_mlp": 0.01032866, + "balance_loss_clip": 1.04973149, + "balance_loss_mlp": 1.02394342, + "epoch": 0.5490290386580894, + "flos": 24315977191680.0, + "grad_norm": 1.9176103259973505, + "language_loss": 0.784971, + "learning_rate": 1.7799971412418374e-06, + "loss": 0.8069368, + "num_input_tokens_seen": 98468695, + "step": 4566, + "time_per_iteration": 2.75978684425354 + }, + { + "auxiliary_loss_clip": 0.01143827, + "auxiliary_loss_mlp": 0.01025824, + "balance_loss_clip": 1.04595208, + "balance_loss_mlp": 1.01690698, + "epoch": 0.5491492815487284, + "flos": 18294942977280.0, + "grad_norm": 3.807046730734561, + "language_loss": 0.74197125, + "learning_rate": 1.7792229135459918e-06, + "loss": 0.76366776, + "num_input_tokens_seen": 98485345, + "step": 4567, + "time_per_iteration": 2.826345920562744 + }, + { + "auxiliary_loss_clip": 0.01043043, + "auxiliary_loss_mlp": 0.01006752, + "balance_loss_clip": 1.01911283, + "balance_loss_mlp": 1.00488079, + "epoch": 0.5492695244393675, + "flos": 64550257050240.0, + "grad_norm": 0.7372353198986485, + "language_loss": 0.61646199, + "learning_rate": 1.7784487193418538e-06, + "loss": 0.63696003, + "num_input_tokens_seen": 98543195, + "step": 4568, + "time_per_iteration": 3.2771096229553223 + }, + { + "auxiliary_loss_clip": 0.01135892, + "auxiliary_loss_mlp": 0.01029007, + "balance_loss_clip": 1.04085779, + "balance_loss_mlp": 1.01979828, + "epoch": 0.5493897673300067, + "flos": 17379579761280.0, + "grad_norm": 4.189886668771894, + "language_loss": 0.6097402, + "learning_rate": 1.7776745587468698e-06, + "loss": 0.63138914, + "num_input_tokens_seen": 98560620, + "step": 4569, + "time_per_iteration": 2.7624738216400146 + }, + { + "auxiliary_loss_clip": 0.01178293, + "auxiliary_loss_mlp": 0.01024244, + "balance_loss_clip": 1.04958963, + "balance_loss_mlp": 1.01629305, + "epoch": 0.5495100102206457, + "flos": 19901765980800.0, + "grad_norm": 2.2585547790577998, + "language_loss": 0.81620395, + "learning_rate": 1.7769004318784776e-06, + "loss": 0.83822936, + "num_input_tokens_seen": 98578265, + "step": 4570, + "time_per_iteration": 2.6807212829589844 + }, + { + "auxiliary_loss_clip": 0.01167522, + "auxiliary_loss_mlp": 0.0102819, + "balance_loss_clip": 1.04764485, + "balance_loss_mlp": 1.01997685, + "epoch": 0.5496302531112848, + "flos": 16727190992640.0, + "grad_norm": 2.14621298068466, + "language_loss": 0.805704, + "learning_rate": 1.776126338854113e-06, + "loss": 0.82766104, + "num_input_tokens_seen": 98596055, + "step": 4571, + "time_per_iteration": 2.6481828689575195 + }, + { + "auxiliary_loss_clip": 0.0116318, + "auxiliary_loss_mlp": 0.01029752, + "balance_loss_clip": 1.04931235, + "balance_loss_mlp": 1.02132356, + "epoch": 0.5497504960019239, + "flos": 24572343536640.0, + "grad_norm": 1.6034583449916762, + "language_loss": 0.84298265, + "learning_rate": 1.7753522797912044e-06, + "loss": 0.86491197, + "num_input_tokens_seen": 98616140, + "step": 4572, + "time_per_iteration": 2.7445478439331055 + }, + { + "auxiliary_loss_clip": 0.01168461, + "auxiliary_loss_mlp": 0.01030721, + "balance_loss_clip": 1.04877424, + "balance_loss_mlp": 1.02222753, + "epoch": 0.549870738892563, + "flos": 15450494912640.0, + "grad_norm": 2.6946339500420953, + "language_loss": 0.69679844, + "learning_rate": 1.7745782548071765e-06, + "loss": 0.71879029, + "num_input_tokens_seen": 98633035, + "step": 4573, + "time_per_iteration": 2.717038631439209 + }, + { + "auxiliary_loss_clip": 0.01143864, + "auxiliary_loss_mlp": 0.01031834, + "balance_loss_clip": 1.04801631, + "balance_loss_mlp": 1.02421093, + "epoch": 0.549990981783202, + "flos": 21069114082560.0, + "grad_norm": 2.0348468324854223, + "language_loss": 0.74081993, + "learning_rate": 1.7738042640194482e-06, + "loss": 0.762577, + "num_input_tokens_seen": 98652700, + "step": 4574, + "time_per_iteration": 2.769098997116089 + }, + { + "auxiliary_loss_clip": 0.01178697, + "auxiliary_loss_mlp": 0.01025767, + "balance_loss_clip": 1.04990935, + "balance_loss_mlp": 1.01713574, + "epoch": 0.5501112246738411, + "flos": 21395901041280.0, + "grad_norm": 1.736406486628853, + "language_loss": 0.70748317, + "learning_rate": 1.7730303075454335e-06, + "loss": 0.72952777, + "num_input_tokens_seen": 98671590, + "step": 4575, + "time_per_iteration": 2.7078051567077637 + }, + { + "auxiliary_loss_clip": 0.01155416, + "auxiliary_loss_mlp": 0.01028841, + "balance_loss_clip": 1.04698944, + "balance_loss_mlp": 1.02053237, + "epoch": 0.5502314675644803, + "flos": 17456931699840.0, + "grad_norm": 2.398810255111601, + "language_loss": 0.84547633, + "learning_rate": 1.7722563855025402e-06, + "loss": 0.86731899, + "num_input_tokens_seen": 98689620, + "step": 4576, + "time_per_iteration": 2.818836212158203 + }, + { + "auxiliary_loss_clip": 0.0115898, + "auxiliary_loss_mlp": 0.01033573, + "balance_loss_clip": 1.04443824, + "balance_loss_mlp": 1.02448344, + "epoch": 0.5503517104551193, + "flos": 24310410583680.0, + "grad_norm": 2.0763087267525537, + "language_loss": 0.70455456, + "learning_rate": 1.7714824980081721e-06, + "loss": 0.72648013, + "num_input_tokens_seen": 98708915, + "step": 4577, + "time_per_iteration": 2.769308090209961 + }, + { + "auxiliary_loss_clip": 0.01166155, + "auxiliary_loss_mlp": 0.0102627, + "balance_loss_clip": 1.04992557, + "balance_loss_mlp": 1.01830125, + "epoch": 0.5504719533457584, + "flos": 22419427086720.0, + "grad_norm": 1.6762283497537607, + "language_loss": 0.73937833, + "learning_rate": 1.7707086451797276e-06, + "loss": 0.76130259, + "num_input_tokens_seen": 98729790, + "step": 4578, + "time_per_iteration": 2.7213008403778076 + }, + { + "auxiliary_loss_clip": 0.01054198, + "auxiliary_loss_mlp": 0.01003981, + "balance_loss_clip": 1.01446152, + "balance_loss_mlp": 1.00247884, + "epoch": 0.5505921962363975, + "flos": 67294155968640.0, + "grad_norm": 0.701250279316403, + "language_loss": 0.5232904, + "learning_rate": 1.7699348271345993e-06, + "loss": 0.54387224, + "num_input_tokens_seen": 98792415, + "step": 4579, + "time_per_iteration": 3.2551655769348145 + }, + { + "auxiliary_loss_clip": 0.01052961, + "auxiliary_loss_mlp": 0.01001658, + "balance_loss_clip": 1.01411414, + "balance_loss_mlp": 0.99989361, + "epoch": 0.5507124391270366, + "flos": 45685125578880.0, + "grad_norm": 0.7089920675607956, + "language_loss": 0.54432714, + "learning_rate": 1.7691610439901753e-06, + "loss": 0.56487334, + "num_input_tokens_seen": 98855350, + "step": 4580, + "time_per_iteration": 3.412295341491699 + }, + { + "auxiliary_loss_clip": 0.01175036, + "auxiliary_loss_mlp": 0.01032174, + "balance_loss_clip": 1.05195379, + "balance_loss_mlp": 1.02443171, + "epoch": 0.5508326820176757, + "flos": 22273845264000.0, + "grad_norm": 1.682744734527659, + "language_loss": 0.753272, + "learning_rate": 1.7683872958638367e-06, + "loss": 0.77534413, + "num_input_tokens_seen": 98874230, + "step": 4581, + "time_per_iteration": 2.7547647953033447 + }, + { + "auxiliary_loss_clip": 0.0115673, + "auxiliary_loss_mlp": 0.01030119, + "balance_loss_clip": 1.04527926, + "balance_loss_mlp": 1.02199459, + "epoch": 0.5509529249083148, + "flos": 20012442762240.0, + "grad_norm": 1.8635078090223662, + "language_loss": 0.84400779, + "learning_rate": 1.7676135828729614e-06, + "loss": 0.86587632, + "num_input_tokens_seen": 98893940, + "step": 4582, + "time_per_iteration": 2.7467966079711914 + }, + { + "auxiliary_loss_clip": 0.01168926, + "auxiliary_loss_mlp": 0.01033674, + "balance_loss_clip": 1.05145025, + "balance_loss_mlp": 1.02572274, + "epoch": 0.5510731677989539, + "flos": 21834801325440.0, + "grad_norm": 1.797796668924549, + "language_loss": 0.82523572, + "learning_rate": 1.7668399051349205e-06, + "loss": 0.84726179, + "num_input_tokens_seen": 98913620, + "step": 4583, + "time_per_iteration": 2.719310760498047 + }, + { + "auxiliary_loss_clip": 0.01149132, + "auxiliary_loss_mlp": 0.01029514, + "balance_loss_clip": 1.04571104, + "balance_loss_mlp": 1.020859, + "epoch": 0.5511934106895929, + "flos": 21467901853440.0, + "grad_norm": 2.310561194992771, + "language_loss": 0.83467495, + "learning_rate": 1.766066262767081e-06, + "loss": 0.85646147, + "num_input_tokens_seen": 98931460, + "step": 4584, + "time_per_iteration": 3.6992714405059814 + }, + { + "auxiliary_loss_clip": 0.01153427, + "auxiliary_loss_mlp": 0.01029863, + "balance_loss_clip": 1.04652452, + "balance_loss_mlp": 1.02167344, + "epoch": 0.5513136535802321, + "flos": 21068934514560.0, + "grad_norm": 2.05848292634525, + "language_loss": 0.76751053, + "learning_rate": 1.765292655886803e-06, + "loss": 0.78934342, + "num_input_tokens_seen": 98950105, + "step": 4585, + "time_per_iteration": 2.729532480239868 + }, + { + "auxiliary_loss_clip": 0.01160876, + "auxiliary_loss_mlp": 0.01035259, + "balance_loss_clip": 1.04837048, + "balance_loss_mlp": 1.02674747, + "epoch": 0.5514338964708712, + "flos": 27815004754560.0, + "grad_norm": 1.8433597607296617, + "language_loss": 0.70933104, + "learning_rate": 1.764519084611443e-06, + "loss": 0.73129243, + "num_input_tokens_seen": 98970560, + "step": 4586, + "time_per_iteration": 2.943983554840088 + }, + { + "auxiliary_loss_clip": 0.01161926, + "auxiliary_loss_mlp": 0.01030007, + "balance_loss_clip": 1.04829526, + "balance_loss_mlp": 1.02069068, + "epoch": 0.5515541393615102, + "flos": 21908525990400.0, + "grad_norm": 1.7151580792143346, + "language_loss": 0.77684712, + "learning_rate": 1.7637455490583505e-06, + "loss": 0.79876643, + "num_input_tokens_seen": 98989885, + "step": 4587, + "time_per_iteration": 3.942770481109619 + }, + { + "auxiliary_loss_clip": 0.01172035, + "auxiliary_loss_mlp": 0.01032957, + "balance_loss_clip": 1.05072594, + "balance_loss_mlp": 1.02492273, + "epoch": 0.5516743822521494, + "flos": 20485422074880.0, + "grad_norm": 2.0630965979623213, + "language_loss": 0.77459657, + "learning_rate": 1.7629720493448701e-06, + "loss": 0.79664648, + "num_input_tokens_seen": 99007180, + "step": 4588, + "time_per_iteration": 2.846876859664917 + }, + { + "auxiliary_loss_clip": 0.01168309, + "auxiliary_loss_mlp": 0.01028206, + "balance_loss_clip": 1.04925287, + "balance_loss_mlp": 1.01975954, + "epoch": 0.5517946251427884, + "flos": 14940383915520.0, + "grad_norm": 1.7599836201095131, + "language_loss": 0.8515107, + "learning_rate": 1.7621985855883418e-06, + "loss": 0.87347591, + "num_input_tokens_seen": 99023880, + "step": 4589, + "time_per_iteration": 2.826289415359497 + }, + { + "auxiliary_loss_clip": 0.01156361, + "auxiliary_loss_mlp": 0.01026685, + "balance_loss_clip": 1.04921913, + "balance_loss_mlp": 1.01812601, + "epoch": 0.5519148680334275, + "flos": 18404865573120.0, + "grad_norm": 1.8186888299970656, + "language_loss": 0.72271723, + "learning_rate": 1.7614251579060983e-06, + "loss": 0.74454772, + "num_input_tokens_seen": 99042475, + "step": 4590, + "time_per_iteration": 3.6851983070373535 + }, + { + "auxiliary_loss_clip": 0.01149442, + "auxiliary_loss_mlp": 0.01027474, + "balance_loss_clip": 1.04857135, + "balance_loss_mlp": 1.01893854, + "epoch": 0.5520351109240667, + "flos": 25113337251840.0, + "grad_norm": 1.8435861386880936, + "language_loss": 0.84571505, + "learning_rate": 1.76065176641547e-06, + "loss": 0.86748421, + "num_input_tokens_seen": 99065185, + "step": 4591, + "time_per_iteration": 2.8390398025512695 + }, + { + "auxiliary_loss_clip": 0.01170188, + "auxiliary_loss_mlp": 0.01033878, + "balance_loss_clip": 1.04764104, + "balance_loss_mlp": 1.02600384, + "epoch": 0.5521553538147057, + "flos": 21069545045760.0, + "grad_norm": 2.457704073019488, + "language_loss": 0.7777763, + "learning_rate": 1.759878411233777e-06, + "loss": 0.79981697, + "num_input_tokens_seen": 99083645, + "step": 4592, + "time_per_iteration": 2.723822593688965 + }, + { + "auxiliary_loss_clip": 0.01167882, + "auxiliary_loss_mlp": 0.0103265, + "balance_loss_clip": 1.04878318, + "balance_loss_mlp": 1.02420998, + "epoch": 0.5522755967053448, + "flos": 18879999701760.0, + "grad_norm": 2.234577922045949, + "language_loss": 0.75887895, + "learning_rate": 1.7591050924783388e-06, + "loss": 0.78088427, + "num_input_tokens_seen": 99100835, + "step": 4593, + "time_per_iteration": 2.6995387077331543 + }, + { + "auxiliary_loss_clip": 0.01044513, + "auxiliary_loss_mlp": 0.01004025, + "balance_loss_clip": 1.0135169, + "balance_loss_mlp": 1.00238037, + "epoch": 0.5523958395959839, + "flos": 64675622494080.0, + "grad_norm": 0.8449570357102826, + "language_loss": 0.5792557, + "learning_rate": 1.7583318102664661e-06, + "loss": 0.5997411, + "num_input_tokens_seen": 99168400, + "step": 4594, + "time_per_iteration": 3.5137219429016113 + }, + { + "auxiliary_loss_clip": 0.01170816, + "auxiliary_loss_mlp": 0.01027163, + "balance_loss_clip": 1.04553652, + "balance_loss_mlp": 1.01875889, + "epoch": 0.552516082486623, + "flos": 10889732211840.0, + "grad_norm": 1.8711625988091247, + "language_loss": 0.79102606, + "learning_rate": 1.757558564715466e-06, + "loss": 0.81300581, + "num_input_tokens_seen": 99186475, + "step": 4595, + "time_per_iteration": 2.716639757156372 + }, + { + "auxiliary_loss_clip": 0.01170455, + "auxiliary_loss_mlp": 0.01030351, + "balance_loss_clip": 1.04792857, + "balance_loss_mlp": 1.02153587, + "epoch": 0.552636325377262, + "flos": 22199797376640.0, + "grad_norm": 2.6778569218299237, + "language_loss": 0.74493343, + "learning_rate": 1.7567853559426386e-06, + "loss": 0.76694155, + "num_input_tokens_seen": 99203525, + "step": 4596, + "time_per_iteration": 2.858107328414917 + }, + { + "auxiliary_loss_clip": 0.01172774, + "auxiliary_loss_mlp": 0.01030729, + "balance_loss_clip": 1.04973042, + "balance_loss_mlp": 1.02201509, + "epoch": 0.5527565682679012, + "flos": 23988184652160.0, + "grad_norm": 2.1173628853551123, + "language_loss": 0.7579186, + "learning_rate": 1.7560121840652797e-06, + "loss": 0.7799536, + "num_input_tokens_seen": 99222910, + "step": 4597, + "time_per_iteration": 2.834195137023926 + }, + { + "auxiliary_loss_clip": 0.01142642, + "auxiliary_loss_mlp": 0.01026963, + "balance_loss_clip": 1.04714298, + "balance_loss_mlp": 1.019238, + "epoch": 0.5528768111585403, + "flos": 19719267955200.0, + "grad_norm": 1.8209394134060866, + "language_loss": 0.69376892, + "learning_rate": 1.7552390492006782e-06, + "loss": 0.71546495, + "num_input_tokens_seen": 99241230, + "step": 4598, + "time_per_iteration": 2.8217356204986572 + }, + { + "auxiliary_loss_clip": 0.01150341, + "auxiliary_loss_mlp": 0.00886911, + "balance_loss_clip": 1.04436493, + "balance_loss_mlp": 1.00073385, + "epoch": 0.5529970540491793, + "flos": 26215975002240.0, + "grad_norm": 1.6683919887449008, + "language_loss": 0.64820957, + "learning_rate": 1.7544659514661184e-06, + "loss": 0.66858208, + "num_input_tokens_seen": 99264320, + "step": 4599, + "time_per_iteration": 2.96999192237854 + }, + { + "auxiliary_loss_clip": 0.01156759, + "auxiliary_loss_mlp": 0.01029298, + "balance_loss_clip": 1.04555488, + "balance_loss_mlp": 1.02090585, + "epoch": 0.5531172969398185, + "flos": 24425971614720.0, + "grad_norm": 2.211869043728845, + "language_loss": 0.79660547, + "learning_rate": 1.7536928909788786e-06, + "loss": 0.81846601, + "num_input_tokens_seen": 99283625, + "step": 4600, + "time_per_iteration": 2.8190619945526123 + }, + { + "auxiliary_loss_clip": 0.0104571, + "auxiliary_loss_mlp": 0.01002876, + "balance_loss_clip": 1.01486242, + "balance_loss_mlp": 1.00120759, + "epoch": 0.5532375398304575, + "flos": 64907316195840.0, + "grad_norm": 0.8791891122936187, + "language_loss": 0.61906648, + "learning_rate": 1.752919867856231e-06, + "loss": 0.63955235, + "num_input_tokens_seen": 99335270, + "step": 4601, + "time_per_iteration": 3.159590244293213 + }, + { + "auxiliary_loss_clip": 0.01154633, + "auxiliary_loss_mlp": 0.01030618, + "balance_loss_clip": 1.04576206, + "balance_loss_mlp": 1.02241039, + "epoch": 0.5533577827210966, + "flos": 19683105937920.0, + "grad_norm": 1.758293416351695, + "language_loss": 0.78950775, + "learning_rate": 1.7521468822154436e-06, + "loss": 0.81136024, + "num_input_tokens_seen": 99354185, + "step": 4602, + "time_per_iteration": 2.7786152362823486 + }, + { + "auxiliary_loss_clip": 0.0115333, + "auxiliary_loss_mlp": 0.01030351, + "balance_loss_clip": 1.04755247, + "balance_loss_mlp": 1.02191687, + "epoch": 0.5534780256117358, + "flos": 32306496076800.0, + "grad_norm": 1.766028691489635, + "language_loss": 0.75171459, + "learning_rate": 1.751373934173777e-06, + "loss": 0.77355134, + "num_input_tokens_seen": 99376930, + "step": 4603, + "time_per_iteration": 2.8583223819732666 + }, + { + "auxiliary_loss_clip": 0.01182343, + "auxiliary_loss_mlp": 0.01035556, + "balance_loss_clip": 1.05090451, + "balance_loss_mlp": 1.02670515, + "epoch": 0.5535982685023748, + "flos": 23222425582080.0, + "grad_norm": 1.6509272796140368, + "language_loss": 0.73127198, + "learning_rate": 1.750601023848487e-06, + "loss": 0.75345105, + "num_input_tokens_seen": 99397655, + "step": 4604, + "time_per_iteration": 2.837498664855957 + }, + { + "auxiliary_loss_clip": 0.01179554, + "auxiliary_loss_mlp": 0.00886179, + "balance_loss_clip": 1.05129409, + "balance_loss_mlp": 1.0007199, + "epoch": 0.5537185113930139, + "flos": 24352534258560.0, + "grad_norm": 2.934537060103374, + "language_loss": 0.73307228, + "learning_rate": 1.749828151356823e-06, + "loss": 0.75372958, + "num_input_tokens_seen": 99417850, + "step": 4605, + "time_per_iteration": 2.815371036529541 + }, + { + "auxiliary_loss_clip": 0.01161939, + "auxiliary_loss_mlp": 0.01030591, + "balance_loss_clip": 1.04920864, + "balance_loss_mlp": 1.02278256, + "epoch": 0.553838754283653, + "flos": 23549068886400.0, + "grad_norm": 1.5932621730027514, + "language_loss": 0.75531542, + "learning_rate": 1.7490553168160297e-06, + "loss": 0.77724075, + "num_input_tokens_seen": 99438920, + "step": 4606, + "time_per_iteration": 2.9737226963043213 + }, + { + "auxiliary_loss_clip": 0.01160028, + "auxiliary_loss_mlp": 0.01027577, + "balance_loss_clip": 1.04633856, + "balance_loss_mlp": 1.01936376, + "epoch": 0.5539589971742921, + "flos": 17275044205440.0, + "grad_norm": 2.0227393561020466, + "language_loss": 0.76309651, + "learning_rate": 1.748282520343345e-06, + "loss": 0.78497255, + "num_input_tokens_seen": 99457950, + "step": 4607, + "time_per_iteration": 2.9795782566070557 + }, + { + "auxiliary_loss_clip": 0.01175708, + "auxiliary_loss_mlp": 0.0102788, + "balance_loss_clip": 1.04853678, + "balance_loss_mlp": 1.01912427, + "epoch": 0.5540792400649311, + "flos": 27564169104000.0, + "grad_norm": 1.8742892103562343, + "language_loss": 0.78474647, + "learning_rate": 1.7475097620560023e-06, + "loss": 0.80678236, + "num_input_tokens_seen": 99478015, + "step": 4608, + "time_per_iteration": 2.9107563495635986 + }, + { + "auxiliary_loss_clip": 0.01177615, + "auxiliary_loss_mlp": 0.01029083, + "balance_loss_clip": 1.04968166, + "balance_loss_mlp": 1.02082157, + "epoch": 0.5541994829555702, + "flos": 23878657105920.0, + "grad_norm": 1.657676833008225, + "language_loss": 0.71338928, + "learning_rate": 1.746737042071228e-06, + "loss": 0.73545623, + "num_input_tokens_seen": 99496520, + "step": 4609, + "time_per_iteration": 2.6910698413848877 + }, + { + "auxiliary_loss_clip": 0.01153435, + "auxiliary_loss_mlp": 0.01022497, + "balance_loss_clip": 1.04680443, + "balance_loss_mlp": 1.01419973, + "epoch": 0.5543197258462094, + "flos": 20115721342080.0, + "grad_norm": 1.8208574740965575, + "language_loss": 0.79052687, + "learning_rate": 1.7459643605062424e-06, + "loss": 0.81228626, + "num_input_tokens_seen": 99513780, + "step": 4610, + "time_per_iteration": 4.612404108047485 + }, + { + "auxiliary_loss_clip": 0.01135928, + "auxiliary_loss_mlp": 0.01030786, + "balance_loss_clip": 1.04635382, + "balance_loss_mlp": 1.0223577, + "epoch": 0.5544399687368484, + "flos": 20916565021440.0, + "grad_norm": 1.6212393440859945, + "language_loss": 0.80524254, + "learning_rate": 1.745191717478262e-06, + "loss": 0.82690966, + "num_input_tokens_seen": 99532360, + "step": 4611, + "time_per_iteration": 2.872663736343384 + }, + { + "auxiliary_loss_clip": 0.01154917, + "auxiliary_loss_mlp": 0.0102611, + "balance_loss_clip": 1.04779947, + "balance_loss_mlp": 1.01843894, + "epoch": 0.5545602116274875, + "flos": 25518661297920.0, + "grad_norm": 2.8605651108526526, + "language_loss": 0.79506254, + "learning_rate": 1.7444191131044948e-06, + "loss": 0.81687289, + "num_input_tokens_seen": 99552635, + "step": 4612, + "time_per_iteration": 2.7869064807891846 + }, + { + "auxiliary_loss_clip": 0.01161422, + "auxiliary_loss_mlp": 0.01027085, + "balance_loss_clip": 1.05083907, + "balance_loss_mlp": 1.01876962, + "epoch": 0.5546804545181266, + "flos": 20995568985600.0, + "grad_norm": 2.1649670932578218, + "language_loss": 0.7309258, + "learning_rate": 1.7436465475021456e-06, + "loss": 0.75281084, + "num_input_tokens_seen": 99572685, + "step": 4613, + "time_per_iteration": 3.6171300411224365 + }, + { + "auxiliary_loss_clip": 0.01140626, + "auxiliary_loss_mlp": 0.01029599, + "balance_loss_clip": 1.04524612, + "balance_loss_mlp": 1.02120626, + "epoch": 0.5548006974087657, + "flos": 26833638297600.0, + "grad_norm": 2.022561771581999, + "language_loss": 0.71597296, + "learning_rate": 1.7428740207884111e-06, + "loss": 0.73767519, + "num_input_tokens_seen": 99593565, + "step": 4614, + "time_per_iteration": 2.8463809490203857 + }, + { + "auxiliary_loss_clip": 0.01138905, + "auxiliary_loss_mlp": 0.0102494, + "balance_loss_clip": 1.04587841, + "balance_loss_mlp": 1.0165, + "epoch": 0.5549209402994048, + "flos": 33656414031360.0, + "grad_norm": 2.301694061108978, + "language_loss": 0.60911208, + "learning_rate": 1.7421015330804833e-06, + "loss": 0.63075054, + "num_input_tokens_seen": 99613485, + "step": 4615, + "time_per_iteration": 4.058435440063477 + }, + { + "auxiliary_loss_clip": 0.01180381, + "auxiliary_loss_mlp": 0.01025824, + "balance_loss_clip": 1.05082178, + "balance_loss_mlp": 1.01753306, + "epoch": 0.5550411831900439, + "flos": 23769524609280.0, + "grad_norm": 3.6365218709809146, + "language_loss": 0.72541785, + "learning_rate": 1.7413290844955475e-06, + "loss": 0.74747992, + "num_input_tokens_seen": 99633515, + "step": 4616, + "time_per_iteration": 2.6503825187683105 + }, + { + "auxiliary_loss_clip": 0.01162061, + "auxiliary_loss_mlp": 0.01030849, + "balance_loss_clip": 1.04888511, + "balance_loss_mlp": 1.02277851, + "epoch": 0.555161426080683, + "flos": 21651189978240.0, + "grad_norm": 1.7478310049947063, + "language_loss": 0.78579807, + "learning_rate": 1.7405566751507843e-06, + "loss": 0.80772722, + "num_input_tokens_seen": 99651560, + "step": 4617, + "time_per_iteration": 2.742372512817383 + }, + { + "auxiliary_loss_clip": 0.01147674, + "auxiliary_loss_mlp": 0.01024124, + "balance_loss_clip": 1.04550862, + "balance_loss_mlp": 1.01602352, + "epoch": 0.555281668971322, + "flos": 49563116605440.0, + "grad_norm": 1.5651294405026799, + "language_loss": 0.6759094, + "learning_rate": 1.7397843051633668e-06, + "loss": 0.69762737, + "num_input_tokens_seen": 99674255, + "step": 4618, + "time_per_iteration": 3.0249521732330322 + }, + { + "auxiliary_loss_clip": 0.01166479, + "auxiliary_loss_mlp": 0.01024228, + "balance_loss_clip": 1.04904926, + "balance_loss_mlp": 1.01603842, + "epoch": 0.5554019118619612, + "flos": 20741608851840.0, + "grad_norm": 4.301711002027488, + "language_loss": 0.71146446, + "learning_rate": 1.739011974650464e-06, + "loss": 0.73337156, + "num_input_tokens_seen": 99693585, + "step": 4619, + "time_per_iteration": 2.839733123779297 + }, + { + "auxiliary_loss_clip": 0.01145092, + "auxiliary_loss_mlp": 0.01028919, + "balance_loss_clip": 1.04721117, + "balance_loss_mlp": 1.02025211, + "epoch": 0.5555221547526003, + "flos": 25483217552640.0, + "grad_norm": 3.0589352993716505, + "language_loss": 0.77239352, + "learning_rate": 1.7382396837292365e-06, + "loss": 0.79413366, + "num_input_tokens_seen": 99714045, + "step": 4620, + "time_per_iteration": 2.9029834270477295 + }, + { + "auxiliary_loss_clip": 0.01183729, + "auxiliary_loss_mlp": 0.0103101, + "balance_loss_clip": 1.05383909, + "balance_loss_mlp": 1.0223825, + "epoch": 0.5556423976432393, + "flos": 21762513204480.0, + "grad_norm": 1.5672853529194741, + "language_loss": 0.73362994, + "learning_rate": 1.737467432516841e-06, + "loss": 0.7557773, + "num_input_tokens_seen": 99734145, + "step": 4621, + "time_per_iteration": 2.695239782333374 + }, + { + "auxiliary_loss_clip": 0.0115964, + "auxiliary_loss_mlp": 0.01023155, + "balance_loss_clip": 1.04522634, + "balance_loss_mlp": 1.01505709, + "epoch": 0.5557626405338785, + "flos": 24900171989760.0, + "grad_norm": 2.465808655732917, + "language_loss": 0.73851717, + "learning_rate": 1.7366952211304274e-06, + "loss": 0.76034504, + "num_input_tokens_seen": 99751990, + "step": 4622, + "time_per_iteration": 2.934908151626587 + }, + { + "auxiliary_loss_clip": 0.01151643, + "auxiliary_loss_mlp": 0.01026958, + "balance_loss_clip": 1.04645872, + "balance_loss_mlp": 1.01873255, + "epoch": 0.5558828834245175, + "flos": 18697501676160.0, + "grad_norm": 2.2831998659375263, + "language_loss": 0.83882111, + "learning_rate": 1.735923049687139e-06, + "loss": 0.86060709, + "num_input_tokens_seen": 99768565, + "step": 4623, + "time_per_iteration": 2.7746734619140625 + }, + { + "auxiliary_loss_clip": 0.01155671, + "auxiliary_loss_mlp": 0.01036469, + "balance_loss_clip": 1.04681826, + "balance_loss_mlp": 1.02811801, + "epoch": 0.5560031263151566, + "flos": 27272179445760.0, + "grad_norm": 1.524261059594109, + "language_loss": 0.74104178, + "learning_rate": 1.7351509183041144e-06, + "loss": 0.76296318, + "num_input_tokens_seen": 99788895, + "step": 4624, + "time_per_iteration": 2.8123724460601807 + }, + { + "auxiliary_loss_clip": 0.01182583, + "auxiliary_loss_mlp": 0.01028854, + "balance_loss_clip": 1.05194855, + "balance_loss_mlp": 1.02101552, + "epoch": 0.5561233692057957, + "flos": 23403738458880.0, + "grad_norm": 1.703249653820291, + "language_loss": 0.71617591, + "learning_rate": 1.7343788270984852e-06, + "loss": 0.73829025, + "num_input_tokens_seen": 99808035, + "step": 4625, + "time_per_iteration": 2.662440538406372 + }, + { + "auxiliary_loss_clip": 0.01161649, + "auxiliary_loss_mlp": 0.01031644, + "balance_loss_clip": 1.05117393, + "balance_loss_mlp": 1.02309692, + "epoch": 0.5562436120964348, + "flos": 37670867804160.0, + "grad_norm": 1.7451425080528833, + "language_loss": 0.7504428, + "learning_rate": 1.7336067761873764e-06, + "loss": 0.7723757, + "num_input_tokens_seen": 99830460, + "step": 4626, + "time_per_iteration": 2.9395487308502197 + }, + { + "auxiliary_loss_clip": 0.01175576, + "auxiliary_loss_mlp": 0.01027922, + "balance_loss_clip": 1.04927516, + "balance_loss_mlp": 1.01864755, + "epoch": 0.5563638549870739, + "flos": 25155245445120.0, + "grad_norm": 2.123154199887286, + "language_loss": 0.76464677, + "learning_rate": 1.7328347656879076e-06, + "loss": 0.78668177, + "num_input_tokens_seen": 99850320, + "step": 4627, + "time_per_iteration": 2.7108774185180664 + }, + { + "auxiliary_loss_clip": 0.01148146, + "auxiliary_loss_mlp": 0.01032971, + "balance_loss_clip": 1.04388022, + "balance_loss_mlp": 1.02422726, + "epoch": 0.556484097877713, + "flos": 13581810783360.0, + "grad_norm": 8.834779471573189, + "language_loss": 0.68173289, + "learning_rate": 1.7320627957171927e-06, + "loss": 0.70354414, + "num_input_tokens_seen": 99864980, + "step": 4628, + "time_per_iteration": 2.793820858001709 + }, + { + "auxiliary_loss_clip": 0.0118271, + "auxiliary_loss_mlp": 0.01032007, + "balance_loss_clip": 1.05296421, + "balance_loss_mlp": 1.02371645, + "epoch": 0.5566043407683521, + "flos": 24681368292480.0, + "grad_norm": 1.882141610002432, + "language_loss": 0.81608093, + "learning_rate": 1.7312908663923382e-06, + "loss": 0.83822817, + "num_input_tokens_seen": 99881155, + "step": 4629, + "time_per_iteration": 2.696113109588623 + }, + { + "auxiliary_loss_clip": 0.01163912, + "auxiliary_loss_mlp": 0.01029817, + "balance_loss_clip": 1.04679251, + "balance_loss_mlp": 1.02147865, + "epoch": 0.5567245836589911, + "flos": 20588161950720.0, + "grad_norm": 2.4888273713267353, + "language_loss": 0.67536235, + "learning_rate": 1.7305189778304463e-06, + "loss": 0.69729966, + "num_input_tokens_seen": 99899330, + "step": 4630, + "time_per_iteration": 2.6732676029205322 + }, + { + "auxiliary_loss_clip": 0.01161659, + "auxiliary_loss_mlp": 0.01031009, + "balance_loss_clip": 1.05224204, + "balance_loss_mlp": 1.02254558, + "epoch": 0.5568448265496303, + "flos": 20704189858560.0, + "grad_norm": 2.22188861250338, + "language_loss": 0.80043757, + "learning_rate": 1.729747130148611e-06, + "loss": 0.82236427, + "num_input_tokens_seen": 99918525, + "step": 4631, + "time_per_iteration": 2.803189516067505 + }, + { + "auxiliary_loss_clip": 0.01159383, + "auxiliary_loss_mlp": 0.01031638, + "balance_loss_clip": 1.04916942, + "balance_loss_mlp": 1.02292943, + "epoch": 0.5569650694402694, + "flos": 25302910256640.0, + "grad_norm": 2.139654657640133, + "language_loss": 0.76795763, + "learning_rate": 1.7289753234639208e-06, + "loss": 0.78986788, + "num_input_tokens_seen": 99937500, + "step": 4632, + "time_per_iteration": 2.8328490257263184 + }, + { + "auxiliary_loss_clip": 0.01176607, + "auxiliary_loss_mlp": 0.01029811, + "balance_loss_clip": 1.05104995, + "balance_loss_mlp": 1.02137065, + "epoch": 0.5570853123309084, + "flos": 19712623939200.0, + "grad_norm": 1.8519402572745647, + "language_loss": 0.76720273, + "learning_rate": 1.7282035578934592e-06, + "loss": 0.78926694, + "num_input_tokens_seen": 99955665, + "step": 4633, + "time_per_iteration": 2.727283477783203 + }, + { + "auxiliary_loss_clip": 0.01155284, + "auxiliary_loss_mlp": 0.01030811, + "balance_loss_clip": 1.05116749, + "balance_loss_mlp": 1.02287173, + "epoch": 0.5572055552215476, + "flos": 16108091153280.0, + "grad_norm": 2.2702812518687305, + "language_loss": 0.79103643, + "learning_rate": 1.727431833554301e-06, + "loss": 0.81289732, + "num_input_tokens_seen": 99974140, + "step": 4634, + "time_per_iteration": 2.7019848823547363 + }, + { + "auxiliary_loss_clip": 0.01131027, + "auxiliary_loss_mlp": 0.01033883, + "balance_loss_clip": 1.04386282, + "balance_loss_mlp": 1.02477551, + "epoch": 0.5573257981121866, + "flos": 17128815937920.0, + "grad_norm": 1.94228494465755, + "language_loss": 0.7696892, + "learning_rate": 1.7266601505635175e-06, + "loss": 0.79133832, + "num_input_tokens_seen": 99991480, + "step": 4635, + "time_per_iteration": 3.8568034172058105 + }, + { + "auxiliary_loss_clip": 0.01171654, + "auxiliary_loss_mlp": 0.01028777, + "balance_loss_clip": 1.05254936, + "balance_loss_mlp": 1.02018785, + "epoch": 0.5574460410028257, + "flos": 18807029222400.0, + "grad_norm": 2.8195838816015093, + "language_loss": 0.7535677, + "learning_rate": 1.7258885090381717e-06, + "loss": 0.775572, + "num_input_tokens_seen": 100009520, + "step": 4636, + "time_per_iteration": 3.7239999771118164 + }, + { + "auxiliary_loss_clip": 0.01160703, + "auxiliary_loss_mlp": 0.010269, + "balance_loss_clip": 1.04518831, + "balance_loss_mlp": 1.01851916, + "epoch": 0.5575662838934649, + "flos": 29642678530560.0, + "grad_norm": 1.8497335769664254, + "language_loss": 0.78564417, + "learning_rate": 1.7251169090953213e-06, + "loss": 0.80752021, + "num_input_tokens_seen": 100029995, + "step": 4637, + "time_per_iteration": 2.8516592979431152 + }, + { + "auxiliary_loss_clip": 0.01169595, + "auxiliary_loss_mlp": 0.01034073, + "balance_loss_clip": 1.04893708, + "balance_loss_mlp": 1.02521574, + "epoch": 0.5576865267841039, + "flos": 22054466949120.0, + "grad_norm": 3.3355729018794817, + "language_loss": 0.76188916, + "learning_rate": 1.7243453508520168e-06, + "loss": 0.78392589, + "num_input_tokens_seen": 100046980, + "step": 4638, + "time_per_iteration": 2.7357821464538574 + }, + { + "auxiliary_loss_clip": 0.01160959, + "auxiliary_loss_mlp": 0.01032645, + "balance_loss_clip": 1.04560792, + "balance_loss_mlp": 1.02416277, + "epoch": 0.557806769674743, + "flos": 17196040241280.0, + "grad_norm": 2.4451109946826923, + "language_loss": 0.84880781, + "learning_rate": 1.7235738344253038e-06, + "loss": 0.87074381, + "num_input_tokens_seen": 100060610, + "step": 4639, + "time_per_iteration": 3.692586898803711 + }, + { + "auxiliary_loss_clip": 0.01168019, + "auxiliary_loss_mlp": 0.01026282, + "balance_loss_clip": 1.05116439, + "balance_loss_mlp": 1.01751375, + "epoch": 0.557927012565382, + "flos": 24712717887360.0, + "grad_norm": 3.968768139903963, + "language_loss": 0.8274132, + "learning_rate": 1.72280235993222e-06, + "loss": 0.84935617, + "num_input_tokens_seen": 100078915, + "step": 4640, + "time_per_iteration": 2.7196033000946045 + }, + { + "auxiliary_loss_clip": 0.01169219, + "auxiliary_loss_mlp": 0.00887308, + "balance_loss_clip": 1.05090165, + "balance_loss_mlp": 1.00083184, + "epoch": 0.5580472554560212, + "flos": 16983090460800.0, + "grad_norm": 2.553320718220715, + "language_loss": 0.68925804, + "learning_rate": 1.722030927489798e-06, + "loss": 0.70982337, + "num_input_tokens_seen": 100096195, + "step": 4641, + "time_per_iteration": 2.70373797416687 + }, + { + "auxiliary_loss_clip": 0.01150385, + "auxiliary_loss_mlp": 0.0102672, + "balance_loss_clip": 1.04897213, + "balance_loss_mlp": 1.0183996, + "epoch": 0.5581674983466602, + "flos": 23509100027520.0, + "grad_norm": 1.9199584676567811, + "language_loss": 0.73915154, + "learning_rate": 1.7212595372150634e-06, + "loss": 0.76092261, + "num_input_tokens_seen": 100116175, + "step": 4642, + "time_per_iteration": 3.6298751831054688 + }, + { + "auxiliary_loss_clip": 0.01181105, + "auxiliary_loss_mlp": 0.01026998, + "balance_loss_clip": 1.05242682, + "balance_loss_mlp": 1.01860023, + "epoch": 0.5582877412372993, + "flos": 13480291969920.0, + "grad_norm": 2.0679004403610284, + "language_loss": 0.7242521, + "learning_rate": 1.720488189225035e-06, + "loss": 0.74633312, + "num_input_tokens_seen": 100133875, + "step": 4643, + "time_per_iteration": 2.6505038738250732 + }, + { + "auxiliary_loss_clip": 0.01173929, + "auxiliary_loss_mlp": 0.01031095, + "balance_loss_clip": 1.05111694, + "balance_loss_mlp": 1.02199364, + "epoch": 0.5584079841279385, + "flos": 21903605827200.0, + "grad_norm": 2.629330811161015, + "language_loss": 0.79571325, + "learning_rate": 1.7197168836367265e-06, + "loss": 0.81776345, + "num_input_tokens_seen": 100150685, + "step": 4644, + "time_per_iteration": 2.716501474380493 + }, + { + "auxiliary_loss_clip": 0.01166014, + "auxiliary_loss_mlp": 0.00885699, + "balance_loss_clip": 1.04835808, + "balance_loss_mlp": 1.00084567, + "epoch": 0.5585282270185775, + "flos": 18843550375680.0, + "grad_norm": 1.70064043440485, + "language_loss": 0.81802058, + "learning_rate": 1.7189456205671433e-06, + "loss": 0.83853775, + "num_input_tokens_seen": 100169530, + "step": 4645, + "time_per_iteration": 2.6590797901153564 + }, + { + "auxiliary_loss_clip": 0.01173443, + "auxiliary_loss_mlp": 0.01034358, + "balance_loss_clip": 1.05246353, + "balance_loss_mlp": 1.02603102, + "epoch": 0.5586484699092166, + "flos": 21868449390720.0, + "grad_norm": 2.335677501001692, + "language_loss": 0.82370049, + "learning_rate": 1.7181744001332866e-06, + "loss": 0.84577852, + "num_input_tokens_seen": 100188140, + "step": 4646, + "time_per_iteration": 2.702129602432251 + }, + { + "auxiliary_loss_clip": 0.01179428, + "auxiliary_loss_mlp": 0.01029697, + "balance_loss_clip": 1.05401492, + "balance_loss_mlp": 1.02204418, + "epoch": 0.5587687127998557, + "flos": 22893232412160.0, + "grad_norm": 1.8997545734692012, + "language_loss": 0.63596481, + "learning_rate": 1.7174032224521493e-06, + "loss": 0.65805608, + "num_input_tokens_seen": 100206850, + "step": 4647, + "time_per_iteration": 2.630798816680908 + }, + { + "auxiliary_loss_clip": 0.0116934, + "auxiliary_loss_mlp": 0.0102787, + "balance_loss_clip": 1.04966426, + "balance_loss_mlp": 1.01994228, + "epoch": 0.5588889556904948, + "flos": 20303067703680.0, + "grad_norm": 1.5716535466405275, + "language_loss": 0.69384718, + "learning_rate": 1.7166320876407184e-06, + "loss": 0.71581924, + "num_input_tokens_seen": 100226270, + "step": 4648, + "time_per_iteration": 2.7499163150787354 + }, + { + "auxiliary_loss_clip": 0.01181921, + "auxiliary_loss_mlp": 0.00886791, + "balance_loss_clip": 1.05362427, + "balance_loss_mlp": 1.0008378, + "epoch": 0.5590091985811338, + "flos": 16472153450880.0, + "grad_norm": 2.0706317685236266, + "language_loss": 0.67668194, + "learning_rate": 1.7158609958159742e-06, + "loss": 0.69736904, + "num_input_tokens_seen": 100243675, + "step": 4649, + "time_per_iteration": 2.716806650161743 + }, + { + "auxiliary_loss_clip": 0.01133832, + "auxiliary_loss_mlp": 0.0103084, + "balance_loss_clip": 1.04516709, + "balance_loss_mlp": 1.02263212, + "epoch": 0.559129441471773, + "flos": 14532186781440.0, + "grad_norm": 2.175119803385222, + "language_loss": 0.78432524, + "learning_rate": 1.7150899470948911e-06, + "loss": 0.80597198, + "num_input_tokens_seen": 100258940, + "step": 4650, + "time_per_iteration": 2.8952412605285645 + }, + { + "auxiliary_loss_clip": 0.01055064, + "auxiliary_loss_mlp": 0.01002737, + "balance_loss_clip": 1.0123601, + "balance_loss_mlp": 1.00103188, + "epoch": 0.5592496843624121, + "flos": 60521009852160.0, + "grad_norm": 0.8009128935212508, + "language_loss": 0.56683755, + "learning_rate": 1.7143189415944365e-06, + "loss": 0.58741558, + "num_input_tokens_seen": 100323400, + "step": 4651, + "time_per_iteration": 3.322854518890381 + }, + { + "auxiliary_loss_clip": 0.01168692, + "auxiliary_loss_mlp": 0.01026331, + "balance_loss_clip": 1.05127931, + "balance_loss_mlp": 1.0177958, + "epoch": 0.5593699272530511, + "flos": 20886256920960.0, + "grad_norm": 1.7615931707483823, + "language_loss": 0.76343262, + "learning_rate": 1.7135479794315714e-06, + "loss": 0.78538281, + "num_input_tokens_seen": 100340355, + "step": 4652, + "time_per_iteration": 2.741766929626465 + }, + { + "auxiliary_loss_clip": 0.01147274, + "auxiliary_loss_mlp": 0.01020975, + "balance_loss_clip": 1.04710841, + "balance_loss_mlp": 1.01360822, + "epoch": 0.5594901701436903, + "flos": 12896743616640.0, + "grad_norm": 1.9517409064537559, + "language_loss": 0.79149717, + "learning_rate": 1.7127770607232502e-06, + "loss": 0.81317961, + "num_input_tokens_seen": 100358900, + "step": 4653, + "time_per_iteration": 2.769620656967163 + }, + { + "auxiliary_loss_clip": 0.01157903, + "auxiliary_loss_mlp": 0.01030998, + "balance_loss_clip": 1.04909325, + "balance_loss_mlp": 1.02323747, + "epoch": 0.5596104130343293, + "flos": 23112107936640.0, + "grad_norm": 5.255827528263875, + "language_loss": 0.799389, + "learning_rate": 1.7120061855864204e-06, + "loss": 0.82127798, + "num_input_tokens_seen": 100378910, + "step": 4654, + "time_per_iteration": 2.770777940750122 + }, + { + "auxiliary_loss_clip": 0.01169695, + "auxiliary_loss_mlp": 0.01027418, + "balance_loss_clip": 1.05166459, + "balance_loss_mlp": 1.01857853, + "epoch": 0.5597306559249684, + "flos": 25957812977280.0, + "grad_norm": 1.8313496061750654, + "language_loss": 0.70831639, + "learning_rate": 1.7112353541380233e-06, + "loss": 0.73028749, + "num_input_tokens_seen": 100398770, + "step": 4655, + "time_per_iteration": 2.7724225521087646 + }, + { + "auxiliary_loss_clip": 0.01163633, + "auxiliary_loss_mlp": 0.01029202, + "balance_loss_clip": 1.05087161, + "balance_loss_mlp": 1.0208931, + "epoch": 0.5598508988156076, + "flos": 22492289825280.0, + "grad_norm": 1.5950617824297029, + "language_loss": 0.72139835, + "learning_rate": 1.7104645664949931e-06, + "loss": 0.74332678, + "num_input_tokens_seen": 100421240, + "step": 4656, + "time_per_iteration": 2.7552428245544434 + }, + { + "auxiliary_loss_clip": 0.01158012, + "auxiliary_loss_mlp": 0.01029095, + "balance_loss_clip": 1.04604447, + "balance_loss_mlp": 1.02094102, + "epoch": 0.5599711417062466, + "flos": 23112538899840.0, + "grad_norm": 2.2207272022365543, + "language_loss": 0.71898079, + "learning_rate": 1.7096938227742584e-06, + "loss": 0.74085188, + "num_input_tokens_seen": 100442370, + "step": 4657, + "time_per_iteration": 2.7679457664489746 + }, + { + "auxiliary_loss_clip": 0.01178439, + "auxiliary_loss_mlp": 0.0102494, + "balance_loss_clip": 1.05081606, + "balance_loss_mlp": 1.01647615, + "epoch": 0.5600913845968857, + "flos": 22339345714560.0, + "grad_norm": 2.236462065967111, + "language_loss": 0.84356976, + "learning_rate": 1.70892312309274e-06, + "loss": 0.86560357, + "num_input_tokens_seen": 100460260, + "step": 4658, + "time_per_iteration": 2.6399121284484863 + }, + { + "auxiliary_loss_clip": 0.01158208, + "auxiliary_loss_mlp": 0.01030356, + "balance_loss_clip": 1.04308248, + "balance_loss_mlp": 1.02230918, + "epoch": 0.5602116274875248, + "flos": 17633791290240.0, + "grad_norm": 1.9626753894544007, + "language_loss": 0.68320012, + "learning_rate": 1.7081524675673523e-06, + "loss": 0.70508581, + "num_input_tokens_seen": 100475750, + "step": 4659, + "time_per_iteration": 2.773172378540039 + }, + { + "auxiliary_loss_clip": 0.010582, + "auxiliary_loss_mlp": 0.0100344, + "balance_loss_clip": 1.01340079, + "balance_loss_mlp": 1.00178301, + "epoch": 0.5603318703781639, + "flos": 70115945529600.0, + "grad_norm": 0.7740917938820521, + "language_loss": 0.59613967, + "learning_rate": 1.7073818563150026e-06, + "loss": 0.61675608, + "num_input_tokens_seen": 100537830, + "step": 4660, + "time_per_iteration": 3.393622398376465 + }, + { + "auxiliary_loss_clip": 0.01161789, + "auxiliary_loss_mlp": 0.01032247, + "balance_loss_clip": 1.04526281, + "balance_loss_mlp": 1.02330625, + "epoch": 0.560452113268803, + "flos": 18545850455040.0, + "grad_norm": 2.1393275275116093, + "language_loss": 0.86803454, + "learning_rate": 1.7066112894525935e-06, + "loss": 0.88997483, + "num_input_tokens_seen": 100555910, + "step": 4661, + "time_per_iteration": 3.627694606781006 + }, + { + "auxiliary_loss_clip": 0.01153168, + "auxiliary_loss_mlp": 0.01024745, + "balance_loss_clip": 1.04751921, + "balance_loss_mlp": 1.01665032, + "epoch": 0.5605723561594421, + "flos": 25264665250560.0, + "grad_norm": 2.1277406604923534, + "language_loss": 0.72560632, + "learning_rate": 1.7058407670970177e-06, + "loss": 0.74738544, + "num_input_tokens_seen": 100577385, + "step": 4662, + "time_per_iteration": 3.7623472213745117 + }, + { + "auxiliary_loss_clip": 0.01171257, + "auxiliary_loss_mlp": 0.010288, + "balance_loss_clip": 1.05201828, + "balance_loss_mlp": 1.01996017, + "epoch": 0.5606925990500812, + "flos": 20594949621120.0, + "grad_norm": 1.654731826748802, + "language_loss": 0.61163026, + "learning_rate": 1.7050702893651643e-06, + "loss": 0.63363081, + "num_input_tokens_seen": 100596965, + "step": 4663, + "time_per_iteration": 2.7545907497406006 + }, + { + "auxiliary_loss_clip": 0.01171252, + "auxiliary_loss_mlp": 0.0102674, + "balance_loss_clip": 1.05206788, + "balance_loss_mlp": 1.01836586, + "epoch": 0.5608128419407202, + "flos": 35006044677120.0, + "grad_norm": 1.9797649017197663, + "language_loss": 0.75295663, + "learning_rate": 1.7042998563739134e-06, + "loss": 0.77493656, + "num_input_tokens_seen": 100615315, + "step": 4664, + "time_per_iteration": 2.795506238937378 + }, + { + "auxiliary_loss_clip": 0.0117026, + "auxiliary_loss_mlp": 0.01034182, + "balance_loss_clip": 1.05021393, + "balance_loss_mlp": 1.02506256, + "epoch": 0.5609330848313594, + "flos": 24639819235200.0, + "grad_norm": 5.304174204151337, + "language_loss": 0.71761131, + "learning_rate": 1.703529468240139e-06, + "loss": 0.73965573, + "num_input_tokens_seen": 100634185, + "step": 4665, + "time_per_iteration": 3.746262550354004 + }, + { + "auxiliary_loss_clip": 0.01156003, + "auxiliary_loss_mlp": 0.01026013, + "balance_loss_clip": 1.04844284, + "balance_loss_mlp": 1.01790655, + "epoch": 0.5610533277219985, + "flos": 18762894385920.0, + "grad_norm": 2.0583406008608676, + "language_loss": 0.73478383, + "learning_rate": 1.7027591250807088e-06, + "loss": 0.75660396, + "num_input_tokens_seen": 100651360, + "step": 4666, + "time_per_iteration": 2.9034945964813232 + }, + { + "auxiliary_loss_clip": 0.01183636, + "auxiliary_loss_mlp": 0.01031147, + "balance_loss_clip": 1.05408299, + "balance_loss_mlp": 1.02252197, + "epoch": 0.5611735706126375, + "flos": 15012384727680.0, + "grad_norm": 2.960033161659513, + "language_loss": 0.84282702, + "learning_rate": 1.7019888270124825e-06, + "loss": 0.86497486, + "num_input_tokens_seen": 100668525, + "step": 4667, + "time_per_iteration": 2.668304204940796 + }, + { + "auxiliary_loss_clip": 0.01170687, + "auxiliary_loss_mlp": 0.01032562, + "balance_loss_clip": 1.04999781, + "balance_loss_mlp": 1.02471828, + "epoch": 0.5612938135032767, + "flos": 16468167041280.0, + "grad_norm": 1.9296565953006735, + "language_loss": 0.8197574, + "learning_rate": 1.7012185741523147e-06, + "loss": 0.8417899, + "num_input_tokens_seen": 100684850, + "step": 4668, + "time_per_iteration": 3.6235620975494385 + }, + { + "auxiliary_loss_clip": 0.01181279, + "auxiliary_loss_mlp": 0.01033589, + "balance_loss_clip": 1.05424833, + "balance_loss_mlp": 1.02553058, + "epoch": 0.5614140563939157, + "flos": 25666433850240.0, + "grad_norm": 2.274754534375001, + "language_loss": 0.62823164, + "learning_rate": 1.7004483666170514e-06, + "loss": 0.65038037, + "num_input_tokens_seen": 100705345, + "step": 4669, + "time_per_iteration": 2.7078187465667725 + }, + { + "auxiliary_loss_clip": 0.01166829, + "auxiliary_loss_mlp": 0.01028236, + "balance_loss_clip": 1.0486033, + "balance_loss_mlp": 1.02012968, + "epoch": 0.5615342992845548, + "flos": 24717566223360.0, + "grad_norm": 2.0150645303006947, + "language_loss": 0.80428696, + "learning_rate": 1.699678204523533e-06, + "loss": 0.82623756, + "num_input_tokens_seen": 100725210, + "step": 4670, + "time_per_iteration": 2.7694156169891357 + }, + { + "auxiliary_loss_clip": 0.01162638, + "auxiliary_loss_mlp": 0.01034872, + "balance_loss_clip": 1.04978967, + "balance_loss_mlp": 1.02557969, + "epoch": 0.5616545421751938, + "flos": 22015934634240.0, + "grad_norm": 9.953682716859317, + "language_loss": 0.68835759, + "learning_rate": 1.6989080879885918e-06, + "loss": 0.71033275, + "num_input_tokens_seen": 100743070, + "step": 4671, + "time_per_iteration": 2.8052501678466797 + }, + { + "auxiliary_loss_clip": 0.01052357, + "auxiliary_loss_mlp": 0.01001513, + "balance_loss_clip": 1.01452017, + "balance_loss_mlp": 0.9999159, + "epoch": 0.561774785065833, + "flos": 53760358690560.0, + "grad_norm": 0.9171005148942001, + "language_loss": 0.6105023, + "learning_rate": 1.6981380171290544e-06, + "loss": 0.63104099, + "num_input_tokens_seen": 100804095, + "step": 4672, + "time_per_iteration": 3.2917654514312744 + }, + { + "auxiliary_loss_clip": 0.011473, + "auxiliary_loss_mlp": 0.01025851, + "balance_loss_clip": 1.04394221, + "balance_loss_mlp": 1.01719069, + "epoch": 0.5618950279564721, + "flos": 19750007018880.0, + "grad_norm": 2.097079488594721, + "language_loss": 0.74342573, + "learning_rate": 1.6973679920617396e-06, + "loss": 0.76515722, + "num_input_tokens_seen": 100821630, + "step": 4673, + "time_per_iteration": 2.752387523651123 + }, + { + "auxiliary_loss_clip": 0.01160204, + "auxiliary_loss_mlp": 0.01029316, + "balance_loss_clip": 1.05115724, + "balance_loss_mlp": 1.02092683, + "epoch": 0.5620152708471111, + "flos": 16800592435200.0, + "grad_norm": 1.9904762148079, + "language_loss": 0.85474151, + "learning_rate": 1.6965980129034603e-06, + "loss": 0.87663674, + "num_input_tokens_seen": 100839015, + "step": 4674, + "time_per_iteration": 2.6867222785949707 + }, + { + "auxiliary_loss_clip": 0.01160068, + "auxiliary_loss_mlp": 0.01026767, + "balance_loss_clip": 1.04981399, + "balance_loss_mlp": 1.01822591, + "epoch": 0.5621355137377503, + "flos": 26797799502720.0, + "grad_norm": 1.5451543758236104, + "language_loss": 0.76581252, + "learning_rate": 1.6958280797710209e-06, + "loss": 0.78768086, + "num_input_tokens_seen": 100860940, + "step": 4675, + "time_per_iteration": 2.809837818145752 + }, + { + "auxiliary_loss_clip": 0.01058152, + "auxiliary_loss_mlp": 0.01000592, + "balance_loss_clip": 1.01406932, + "balance_loss_mlp": 0.99898261, + "epoch": 0.5622557566283893, + "flos": 61207046686080.0, + "grad_norm": 0.7162340615062509, + "language_loss": 0.54758346, + "learning_rate": 1.6950581927812198e-06, + "loss": 0.56817091, + "num_input_tokens_seen": 100920510, + "step": 4676, + "time_per_iteration": 3.1741669178009033 + }, + { + "auxiliary_loss_clip": 0.01167501, + "auxiliary_loss_mlp": 0.01029102, + "balance_loss_clip": 1.04775286, + "balance_loss_mlp": 1.02106178, + "epoch": 0.5623759995190284, + "flos": 26468534505600.0, + "grad_norm": 2.4466141875188705, + "language_loss": 0.79376459, + "learning_rate": 1.6942883520508486e-06, + "loss": 0.81573069, + "num_input_tokens_seen": 100939245, + "step": 4677, + "time_per_iteration": 2.7394986152648926 + }, + { + "auxiliary_loss_clip": 0.01164826, + "auxiliary_loss_mlp": 0.01030871, + "balance_loss_clip": 1.05043888, + "balance_loss_mlp": 1.02321219, + "epoch": 0.5624962424096676, + "flos": 19390900798080.0, + "grad_norm": 1.9122030989493282, + "language_loss": 0.77042282, + "learning_rate": 1.693518557696691e-06, + "loss": 0.79237974, + "num_input_tokens_seen": 100958385, + "step": 4678, + "time_per_iteration": 2.6932969093322754 + }, + { + "auxiliary_loss_clip": 0.01164957, + "auxiliary_loss_mlp": 0.0102619, + "balance_loss_clip": 1.04533696, + "balance_loss_mlp": 1.01793432, + "epoch": 0.5626164853003066, + "flos": 20667345482880.0, + "grad_norm": 1.8449898947623502, + "language_loss": 0.89013195, + "learning_rate": 1.6927488098355252e-06, + "loss": 0.91204339, + "num_input_tokens_seen": 100976015, + "step": 4679, + "time_per_iteration": 2.682868242263794 + }, + { + "auxiliary_loss_clip": 0.01046618, + "auxiliary_loss_mlp": 0.01003139, + "balance_loss_clip": 1.01335287, + "balance_loss_mlp": 1.00142276, + "epoch": 0.5627367281909457, + "flos": 62766071665920.0, + "grad_norm": 0.9076134048927894, + "language_loss": 0.63166487, + "learning_rate": 1.6919791085841201e-06, + "loss": 0.65216243, + "num_input_tokens_seen": 101033425, + "step": 4680, + "time_per_iteration": 3.3331005573272705 + }, + { + "auxiliary_loss_clip": 0.01162513, + "auxiliary_loss_mlp": 0.01026287, + "balance_loss_clip": 1.04636645, + "balance_loss_mlp": 1.01732838, + "epoch": 0.5628569710815848, + "flos": 12787144243200.0, + "grad_norm": 2.440797988665171, + "language_loss": 0.78918886, + "learning_rate": 1.6912094540592396e-06, + "loss": 0.81107676, + "num_input_tokens_seen": 101048945, + "step": 4681, + "time_per_iteration": 2.7407338619232178 + }, + { + "auxiliary_loss_clip": 0.01168971, + "auxiliary_loss_mlp": 0.01029241, + "balance_loss_clip": 1.04987991, + "balance_loss_mlp": 1.02080047, + "epoch": 0.5629772139722239, + "flos": 13762082165760.0, + "grad_norm": 4.088882856034486, + "language_loss": 0.81221265, + "learning_rate": 1.6904398463776393e-06, + "loss": 0.83419478, + "num_input_tokens_seen": 101062745, + "step": 4682, + "time_per_iteration": 2.796229124069214 + }, + { + "auxiliary_loss_clip": 0.01169889, + "auxiliary_loss_mlp": 0.01026495, + "balance_loss_clip": 1.04735756, + "balance_loss_mlp": 1.01872241, + "epoch": 0.5630974568628629, + "flos": 21467830026240.0, + "grad_norm": 1.8820653437640213, + "language_loss": 0.72538996, + "learning_rate": 1.6896702856560683e-06, + "loss": 0.74735379, + "num_input_tokens_seen": 101081840, + "step": 4683, + "time_per_iteration": 2.722327709197998 + }, + { + "auxiliary_loss_clip": 0.01147062, + "auxiliary_loss_mlp": 0.01027191, + "balance_loss_clip": 1.04263115, + "balance_loss_mlp": 1.0186882, + "epoch": 0.5632176997535021, + "flos": 14245907385600.0, + "grad_norm": 3.0135252885940895, + "language_loss": 0.69000196, + "learning_rate": 1.6889007720112677e-06, + "loss": 0.71174449, + "num_input_tokens_seen": 101099585, + "step": 4684, + "time_per_iteration": 2.801455020904541 + }, + { + "auxiliary_loss_clip": 0.01172721, + "auxiliary_loss_mlp": 0.01028704, + "balance_loss_clip": 1.05133629, + "balance_loss_mlp": 1.02084184, + "epoch": 0.5633379426441412, + "flos": 20812244947200.0, + "grad_norm": 1.6860962697089605, + "language_loss": 0.77136803, + "learning_rate": 1.6881313055599734e-06, + "loss": 0.79338223, + "num_input_tokens_seen": 101119515, + "step": 4685, + "time_per_iteration": 2.7094316482543945 + }, + { + "auxiliary_loss_clip": 0.01147208, + "auxiliary_loss_mlp": 0.01033967, + "balance_loss_clip": 1.04372644, + "balance_loss_mlp": 1.02526784, + "epoch": 0.5634581855347802, + "flos": 22600883617920.0, + "grad_norm": 2.0361606088216986, + "language_loss": 0.81797147, + "learning_rate": 1.6873618864189117e-06, + "loss": 0.83978319, + "num_input_tokens_seen": 101135285, + "step": 4686, + "time_per_iteration": 2.7709052562713623 + }, + { + "auxiliary_loss_clip": 0.01167839, + "auxiliary_loss_mlp": 0.01026209, + "balance_loss_clip": 1.04847014, + "balance_loss_mlp": 1.0180074, + "epoch": 0.5635784284254194, + "flos": 21506972872320.0, + "grad_norm": 4.300199382498471, + "language_loss": 0.7794975, + "learning_rate": 1.686592514704803e-06, + "loss": 0.80143803, + "num_input_tokens_seen": 101152680, + "step": 4687, + "time_per_iteration": 2.650932788848877 + }, + { + "auxiliary_loss_clip": 0.01156246, + "auxiliary_loss_mlp": 0.01027267, + "balance_loss_clip": 1.04896474, + "balance_loss_mlp": 1.01963198, + "epoch": 0.5636986713160584, + "flos": 19827466698240.0, + "grad_norm": 2.2758139824121937, + "language_loss": 0.70687664, + "learning_rate": 1.685823190534361e-06, + "loss": 0.72871178, + "num_input_tokens_seen": 101170920, + "step": 4688, + "time_per_iteration": 3.685894250869751 + }, + { + "auxiliary_loss_clip": 0.01182954, + "auxiliary_loss_mlp": 0.01028657, + "balance_loss_clip": 1.05117106, + "balance_loss_mlp": 1.01981163, + "epoch": 0.5638189142066975, + "flos": 19792453916160.0, + "grad_norm": 1.759910166508937, + "language_loss": 0.83966768, + "learning_rate": 1.6850539140242907e-06, + "loss": 0.86178374, + "num_input_tokens_seen": 101190180, + "step": 4689, + "time_per_iteration": 3.6222193241119385 + }, + { + "auxiliary_loss_clip": 0.0116447, + "auxiliary_loss_mlp": 0.01029267, + "balance_loss_clip": 1.04844654, + "balance_loss_mlp": 1.02131629, + "epoch": 0.5639391570973367, + "flos": 22893771116160.0, + "grad_norm": 1.8791389334858732, + "language_loss": 0.82101667, + "learning_rate": 1.684284685291292e-06, + "loss": 0.84295404, + "num_input_tokens_seen": 101211825, + "step": 4690, + "time_per_iteration": 3.6462223529815674 + }, + { + "auxiliary_loss_clip": 0.01179608, + "auxiliary_loss_mlp": 0.01030786, + "balance_loss_clip": 1.05153847, + "balance_loss_mlp": 1.02222681, + "epoch": 0.5640593999879757, + "flos": 23727077712000.0, + "grad_norm": 2.0650603436794635, + "language_loss": 0.81184262, + "learning_rate": 1.683515504452055e-06, + "loss": 0.83394659, + "num_input_tokens_seen": 101229200, + "step": 4691, + "time_per_iteration": 2.6384754180908203 + }, + { + "auxiliary_loss_clip": 0.0114245, + "auxiliary_loss_mlp": 0.01033472, + "balance_loss_clip": 1.04380858, + "balance_loss_mlp": 1.02499068, + "epoch": 0.5641796428786148, + "flos": 22710123855360.0, + "grad_norm": 5.977087931449844, + "language_loss": 0.66668069, + "learning_rate": 1.6827463716232648e-06, + "loss": 0.68843997, + "num_input_tokens_seen": 101249860, + "step": 4692, + "time_per_iteration": 2.7835707664489746 + }, + { + "auxiliary_loss_clip": 0.01168974, + "auxiliary_loss_mlp": 0.00885946, + "balance_loss_clip": 1.04949069, + "balance_loss_mlp": 1.00063002, + "epoch": 0.5642998857692539, + "flos": 19791987039360.0, + "grad_norm": 1.6535370279246924, + "language_loss": 0.75231612, + "learning_rate": 1.6819772869215972e-06, + "loss": 0.7728653, + "num_input_tokens_seen": 101268940, + "step": 4693, + "time_per_iteration": 3.654045581817627 + }, + { + "auxiliary_loss_clip": 0.0116777, + "auxiliary_loss_mlp": 0.01029185, + "balance_loss_clip": 1.05163968, + "balance_loss_mlp": 1.02088833, + "epoch": 0.564420128659893, + "flos": 23185904428800.0, + "grad_norm": 1.6878093483124499, + "language_loss": 0.82009095, + "learning_rate": 1.6812082504637228e-06, + "loss": 0.84206057, + "num_input_tokens_seen": 101290260, + "step": 4694, + "time_per_iteration": 2.736447334289551 + }, + { + "auxiliary_loss_clip": 0.01168082, + "auxiliary_loss_mlp": 0.01026775, + "balance_loss_clip": 1.0504564, + "balance_loss_mlp": 1.01899624, + "epoch": 0.564540371550532, + "flos": 23258264376960.0, + "grad_norm": 1.5498969777477134, + "language_loss": 0.7420696, + "learning_rate": 1.6804392623663025e-06, + "loss": 0.76401818, + "num_input_tokens_seen": 101311465, + "step": 4695, + "time_per_iteration": 2.8108911514282227 + }, + { + "auxiliary_loss_clip": 0.01162103, + "auxiliary_loss_mlp": 0.01026835, + "balance_loss_clip": 1.04798198, + "balance_loss_mlp": 1.01894689, + "epoch": 0.5646606144411712, + "flos": 25010058672000.0, + "grad_norm": 1.9277466565790962, + "language_loss": 0.78446841, + "learning_rate": 1.6796703227459935e-06, + "loss": 0.8063578, + "num_input_tokens_seen": 101329420, + "step": 4696, + "time_per_iteration": 2.818004608154297 + }, + { + "auxiliary_loss_clip": 0.01128415, + "auxiliary_loss_mlp": 0.01022732, + "balance_loss_clip": 1.04091668, + "balance_loss_mlp": 1.014817, + "epoch": 0.5647808573318103, + "flos": 36539645806080.0, + "grad_norm": 1.7365184924721637, + "language_loss": 0.75723982, + "learning_rate": 1.6789014317194407e-06, + "loss": 0.77875137, + "num_input_tokens_seen": 101350900, + "step": 4697, + "time_per_iteration": 2.9721710681915283 + }, + { + "auxiliary_loss_clip": 0.01157817, + "auxiliary_loss_mlp": 0.01027312, + "balance_loss_clip": 1.04848981, + "balance_loss_mlp": 1.01859736, + "epoch": 0.5649011002224493, + "flos": 22528451842560.0, + "grad_norm": 2.696719362909149, + "language_loss": 0.72775429, + "learning_rate": 1.6781325894032853e-06, + "loss": 0.74960566, + "num_input_tokens_seen": 101369860, + "step": 4698, + "time_per_iteration": 2.731448173522949 + }, + { + "auxiliary_loss_clip": 0.01155222, + "auxiliary_loss_mlp": 0.01030505, + "balance_loss_clip": 1.04917312, + "balance_loss_mlp": 1.02235734, + "epoch": 0.5650213431130885, + "flos": 18515147304960.0, + "grad_norm": 1.9952434450284864, + "language_loss": 0.9180041, + "learning_rate": 1.6773637959141608e-06, + "loss": 0.93986136, + "num_input_tokens_seen": 101386835, + "step": 4699, + "time_per_iteration": 2.797382354736328 + }, + { + "auxiliary_loss_clip": 0.01149642, + "auxiliary_loss_mlp": 0.01024981, + "balance_loss_clip": 1.04618526, + "balance_loss_mlp": 1.01657093, + "epoch": 0.5651415860037275, + "flos": 17526310819200.0, + "grad_norm": 2.8481611815164145, + "language_loss": 0.66283536, + "learning_rate": 1.6765950513686915e-06, + "loss": 0.68458164, + "num_input_tokens_seen": 101404945, + "step": 4700, + "time_per_iteration": 2.7682442665100098 + }, + { + "auxiliary_loss_clip": 0.01132091, + "auxiliary_loss_mlp": 0.0103297, + "balance_loss_clip": 1.03973389, + "balance_loss_mlp": 1.02498245, + "epoch": 0.5652618288943666, + "flos": 25520026014720.0, + "grad_norm": 1.9933066080972393, + "language_loss": 0.76301241, + "learning_rate": 1.675826355883496e-06, + "loss": 0.78466308, + "num_input_tokens_seen": 101424160, + "step": 4701, + "time_per_iteration": 2.9295454025268555 + }, + { + "auxiliary_loss_clip": 0.01152423, + "auxiliary_loss_mlp": 0.01025265, + "balance_loss_clip": 1.04860687, + "balance_loss_mlp": 1.01728976, + "epoch": 0.5653820717850057, + "flos": 19683105937920.0, + "grad_norm": 2.184859319220551, + "language_loss": 0.79431027, + "learning_rate": 1.6750577095751848e-06, + "loss": 0.81608713, + "num_input_tokens_seen": 101443270, + "step": 4702, + "time_per_iteration": 2.7513298988342285 + }, + { + "auxiliary_loss_clip": 0.01173525, + "auxiliary_loss_mlp": 0.01030589, + "balance_loss_clip": 1.04881644, + "balance_loss_mlp": 1.02298939, + "epoch": 0.5655023146756448, + "flos": 26979722910720.0, + "grad_norm": 1.6719572152311664, + "language_loss": 0.72970212, + "learning_rate": 1.6742891125603605e-06, + "loss": 0.75174332, + "num_input_tokens_seen": 101464175, + "step": 4703, + "time_per_iteration": 2.816478967666626 + }, + { + "auxiliary_loss_clip": 0.01168964, + "auxiliary_loss_mlp": 0.01029532, + "balance_loss_clip": 1.05067003, + "balance_loss_mlp": 1.0211277, + "epoch": 0.5656225575662839, + "flos": 27669351104640.0, + "grad_norm": 1.8150892670793415, + "language_loss": 0.7173962, + "learning_rate": 1.6735205649556185e-06, + "loss": 0.73938119, + "num_input_tokens_seen": 101484045, + "step": 4704, + "time_per_iteration": 2.796468496322632 + }, + { + "auxiliary_loss_clip": 0.01148481, + "auxiliary_loss_mlp": 0.01029384, + "balance_loss_clip": 1.0471257, + "balance_loss_mlp": 1.02154589, + "epoch": 0.5657428004569229, + "flos": 24349732997760.0, + "grad_norm": 1.6303830994094701, + "language_loss": 0.84773707, + "learning_rate": 1.6727520668775476e-06, + "loss": 0.86951578, + "num_input_tokens_seen": 101504330, + "step": 4705, + "time_per_iteration": 2.7627618312835693 + }, + { + "auxiliary_loss_clip": 0.01178679, + "auxiliary_loss_mlp": 0.01026512, + "balance_loss_clip": 1.04864061, + "balance_loss_mlp": 1.01785135, + "epoch": 0.5658630433475621, + "flos": 21944041562880.0, + "grad_norm": 3.4000707913785937, + "language_loss": 0.75159693, + "learning_rate": 1.6719836184427275e-06, + "loss": 0.77364886, + "num_input_tokens_seen": 101524635, + "step": 4706, + "time_per_iteration": 2.700962781906128 + }, + { + "auxiliary_loss_clip": 0.01153647, + "auxiliary_loss_mlp": 0.01031864, + "balance_loss_clip": 1.0453943, + "balance_loss_mlp": 1.02346563, + "epoch": 0.5659832862382012, + "flos": 30409012218240.0, + "grad_norm": 1.9394978928548174, + "language_loss": 0.64587897, + "learning_rate": 1.671215219767733e-06, + "loss": 0.66773409, + "num_input_tokens_seen": 101544095, + "step": 4707, + "time_per_iteration": 2.8615009784698486 + }, + { + "auxiliary_loss_clip": 0.01140144, + "auxiliary_loss_mlp": 0.01027692, + "balance_loss_clip": 1.0439496, + "balance_loss_mlp": 1.019526, + "epoch": 0.5661035291288402, + "flos": 13188194570880.0, + "grad_norm": 2.1714469179266205, + "language_loss": 0.76395941, + "learning_rate": 1.670446870969127e-06, + "loss": 0.7856378, + "num_input_tokens_seen": 101561760, + "step": 4708, + "time_per_iteration": 2.936176300048828 + }, + { + "auxiliary_loss_clip": 0.01153068, + "auxiliary_loss_mlp": 0.01029603, + "balance_loss_clip": 1.04777789, + "balance_loss_mlp": 1.02183628, + "epoch": 0.5662237720194794, + "flos": 16143032108160.0, + "grad_norm": 2.5879190901631484, + "language_loss": 0.79977012, + "learning_rate": 1.6696785721634685e-06, + "loss": 0.82159686, + "num_input_tokens_seen": 101576245, + "step": 4709, + "time_per_iteration": 2.7379674911499023 + }, + { + "auxiliary_loss_clip": 0.0117056, + "auxiliary_loss_mlp": 0.01032292, + "balance_loss_clip": 1.0499568, + "balance_loss_mlp": 1.0238874, + "epoch": 0.5663440149101184, + "flos": 17676848718720.0, + "grad_norm": 1.9787092949596343, + "language_loss": 0.73761046, + "learning_rate": 1.6689103234673086e-06, + "loss": 0.75963902, + "num_input_tokens_seen": 101594565, + "step": 4710, + "time_per_iteration": 2.685778856277466 + }, + { + "auxiliary_loss_clip": 0.01157285, + "auxiliary_loss_mlp": 0.01024588, + "balance_loss_clip": 1.05046391, + "balance_loss_mlp": 1.01680374, + "epoch": 0.5664642578007575, + "flos": 23368330627200.0, + "grad_norm": 2.459862374168897, + "language_loss": 0.77069712, + "learning_rate": 1.668142124997189e-06, + "loss": 0.79251587, + "num_input_tokens_seen": 101614225, + "step": 4711, + "time_per_iteration": 2.767409086227417 + }, + { + "auxiliary_loss_clip": 0.01056736, + "auxiliary_loss_mlp": 0.01005631, + "balance_loss_clip": 1.0188179, + "balance_loss_mlp": 1.00402164, + "epoch": 0.5665845006913967, + "flos": 65516470945920.0, + "grad_norm": 0.7281622593137569, + "language_loss": 0.59743011, + "learning_rate": 1.6673739768696453e-06, + "loss": 0.61805379, + "num_input_tokens_seen": 101680795, + "step": 4712, + "time_per_iteration": 3.2938430309295654 + }, + { + "auxiliary_loss_clip": 0.011627, + "auxiliary_loss_mlp": 0.01030153, + "balance_loss_clip": 1.04586947, + "balance_loss_mlp": 1.02245831, + "epoch": 0.5667047435820357, + "flos": 26140885620480.0, + "grad_norm": 1.8667135787686353, + "language_loss": 0.77704525, + "learning_rate": 1.6666058792012052e-06, + "loss": 0.7989738, + "num_input_tokens_seen": 101701680, + "step": 4713, + "time_per_iteration": 2.7523093223571777 + }, + { + "auxiliary_loss_clip": 0.01062585, + "auxiliary_loss_mlp": 0.01003548, + "balance_loss_clip": 1.01066709, + "balance_loss_mlp": 1.00198591, + "epoch": 0.5668249864726748, + "flos": 71866949725440.0, + "grad_norm": 0.8875904708066237, + "language_loss": 0.68803883, + "learning_rate": 1.6658378321083878e-06, + "loss": 0.70870018, + "num_input_tokens_seen": 101766010, + "step": 4714, + "time_per_iteration": 4.233440160751343 + }, + { + "auxiliary_loss_clip": 0.01129208, + "auxiliary_loss_mlp": 0.01026816, + "balance_loss_clip": 1.04358387, + "balance_loss_mlp": 1.01930034, + "epoch": 0.5669452293633139, + "flos": 22195667312640.0, + "grad_norm": 1.7119027394493522, + "language_loss": 0.82685375, + "learning_rate": 1.6650698357077055e-06, + "loss": 0.84841406, + "num_input_tokens_seen": 101783055, + "step": 4715, + "time_per_iteration": 3.803642988204956 + }, + { + "auxiliary_loss_clip": 0.01159553, + "auxiliary_loss_mlp": 0.01029751, + "balance_loss_clip": 1.04525852, + "balance_loss_mlp": 1.02194858, + "epoch": 0.567065472253953, + "flos": 18223193560320.0, + "grad_norm": 2.4201792092139565, + "language_loss": 0.81003988, + "learning_rate": 1.6643018901156632e-06, + "loss": 0.8319329, + "num_input_tokens_seen": 101802150, + "step": 4716, + "time_per_iteration": 3.7125325202941895 + }, + { + "auxiliary_loss_clip": 0.01164906, + "auxiliary_loss_mlp": 0.01028466, + "balance_loss_clip": 1.04847145, + "balance_loss_mlp": 1.02050877, + "epoch": 0.567185715144592, + "flos": 20371548983040.0, + "grad_norm": 2.822828441832456, + "language_loss": 0.79471636, + "learning_rate": 1.6635339954487566e-06, + "loss": 0.81665003, + "num_input_tokens_seen": 101818025, + "step": 4717, + "time_per_iteration": 2.7899508476257324 + }, + { + "auxiliary_loss_clip": 0.01162431, + "auxiliary_loss_mlp": 0.0102552, + "balance_loss_clip": 1.04789484, + "balance_loss_mlp": 1.01750886, + "epoch": 0.5673059580352312, + "flos": 23221348174080.0, + "grad_norm": 1.7619724950023772, + "language_loss": 0.82139564, + "learning_rate": 1.6627661518234765e-06, + "loss": 0.84327519, + "num_input_tokens_seen": 101837280, + "step": 4718, + "time_per_iteration": 2.7622361183166504 + }, + { + "auxiliary_loss_clip": 0.01140325, + "auxiliary_loss_mlp": 0.01030556, + "balance_loss_clip": 1.04599178, + "balance_loss_mlp": 1.02223575, + "epoch": 0.5674262009258703, + "flos": 21719599430400.0, + "grad_norm": 1.6212640612800053, + "language_loss": 0.85040617, + "learning_rate": 1.661998359356302e-06, + "loss": 0.87211502, + "num_input_tokens_seen": 101856310, + "step": 4719, + "time_per_iteration": 3.7465603351593018 + }, + { + "auxiliary_loss_clip": 0.01067834, + "auxiliary_loss_mlp": 0.01002286, + "balance_loss_clip": 1.01193166, + "balance_loss_mlp": 1.00070047, + "epoch": 0.5675464438165093, + "flos": 67470369114240.0, + "grad_norm": 0.7477454220038398, + "language_loss": 0.5581677, + "learning_rate": 1.6612306181637077e-06, + "loss": 0.57886899, + "num_input_tokens_seen": 101915635, + "step": 4720, + "time_per_iteration": 3.202880620956421 + }, + { + "auxiliary_loss_clip": 0.0114693, + "auxiliary_loss_mlp": 0.01029249, + "balance_loss_clip": 1.04405785, + "balance_loss_mlp": 1.02121711, + "epoch": 0.5676666867071485, + "flos": 18879173688960.0, + "grad_norm": 2.802110506238086, + "language_loss": 0.65670055, + "learning_rate": 1.6604629283621598e-06, + "loss": 0.67846233, + "num_input_tokens_seen": 101933565, + "step": 4721, + "time_per_iteration": 2.821607828140259 + }, + { + "auxiliary_loss_clip": 0.01181023, + "auxiliary_loss_mlp": 0.01027531, + "balance_loss_clip": 1.051965, + "balance_loss_mlp": 1.01900768, + "epoch": 0.5677869295977875, + "flos": 33546778744320.0, + "grad_norm": 1.7112617876832195, + "language_loss": 0.73984182, + "learning_rate": 1.6596952900681152e-06, + "loss": 0.76192737, + "num_input_tokens_seen": 101954325, + "step": 4722, + "time_per_iteration": 2.707671880722046 + }, + { + "auxiliary_loss_clip": 0.01128976, + "auxiliary_loss_mlp": 0.01032869, + "balance_loss_clip": 1.0460484, + "balance_loss_mlp": 1.02456594, + "epoch": 0.5679071724884266, + "flos": 28037256157440.0, + "grad_norm": 2.393057931095984, + "language_loss": 0.82020694, + "learning_rate": 1.658927703398025e-06, + "loss": 0.84182537, + "num_input_tokens_seen": 101974390, + "step": 4723, + "time_per_iteration": 2.7753190994262695 + }, + { + "auxiliary_loss_clip": 0.01138388, + "auxiliary_loss_mlp": 0.01031737, + "balance_loss_clip": 1.0391891, + "balance_loss_mlp": 1.02312994, + "epoch": 0.5680274153790658, + "flos": 23550110380800.0, + "grad_norm": 2.6230003162996995, + "language_loss": 0.77996802, + "learning_rate": 1.6581601684683309e-06, + "loss": 0.80166936, + "num_input_tokens_seen": 101994815, + "step": 4724, + "time_per_iteration": 2.845424175262451 + }, + { + "auxiliary_loss_clip": 0.01171977, + "auxiliary_loss_mlp": 0.0102803, + "balance_loss_clip": 1.05134821, + "balance_loss_mlp": 1.01944101, + "epoch": 0.5681476582697048, + "flos": 22455158140800.0, + "grad_norm": 3.794516458269059, + "language_loss": 0.68880117, + "learning_rate": 1.6573926853954674e-06, + "loss": 0.71080118, + "num_input_tokens_seen": 102012400, + "step": 4725, + "time_per_iteration": 2.772545099258423 + }, + { + "auxiliary_loss_clip": 0.01150867, + "auxiliary_loss_mlp": 0.01033237, + "balance_loss_clip": 1.04320812, + "balance_loss_mlp": 1.02553344, + "epoch": 0.5682679011603439, + "flos": 19536913584000.0, + "grad_norm": 1.8087667962423235, + "language_loss": 0.83052784, + "learning_rate": 1.6566252542958608e-06, + "loss": 0.85236889, + "num_input_tokens_seen": 102031900, + "step": 4726, + "time_per_iteration": 2.7793147563934326 + }, + { + "auxiliary_loss_clip": 0.01141429, + "auxiliary_loss_mlp": 0.01034328, + "balance_loss_clip": 1.04467666, + "balance_loss_mlp": 1.0265286, + "epoch": 0.568388144050983, + "flos": 28765488493440.0, + "grad_norm": 2.095525782339197, + "language_loss": 0.78404915, + "learning_rate": 1.6558578752859305e-06, + "loss": 0.80580664, + "num_input_tokens_seen": 102050860, + "step": 4727, + "time_per_iteration": 2.7665765285491943 + }, + { + "auxiliary_loss_clip": 0.01146687, + "auxiliary_loss_mlp": 0.01034872, + "balance_loss_clip": 1.0439893, + "balance_loss_mlp": 1.02684331, + "epoch": 0.5685083869416221, + "flos": 21209452519680.0, + "grad_norm": 1.9894869794638204, + "language_loss": 0.78414202, + "learning_rate": 1.6550905484820865e-06, + "loss": 0.80595767, + "num_input_tokens_seen": 102069320, + "step": 4728, + "time_per_iteration": 2.7887847423553467 + }, + { + "auxiliary_loss_clip": 0.01178789, + "auxiliary_loss_mlp": 0.01031251, + "balance_loss_clip": 1.04991877, + "balance_loss_mlp": 1.02241158, + "epoch": 0.5686286298322611, + "flos": 24827021942400.0, + "grad_norm": 1.993798945834439, + "language_loss": 0.78448653, + "learning_rate": 1.6543232740007328e-06, + "loss": 0.80658692, + "num_input_tokens_seen": 102086435, + "step": 4729, + "time_per_iteration": 2.6684319972991943 + }, + { + "auxiliary_loss_clip": 0.01171574, + "auxiliary_loss_mlp": 0.01032487, + "balance_loss_clip": 1.05088425, + "balance_loss_mlp": 1.02379036, + "epoch": 0.5687488727229003, + "flos": 26615121909120.0, + "grad_norm": 2.9987486717096696, + "language_loss": 0.67346883, + "learning_rate": 1.653556051958263e-06, + "loss": 0.69550943, + "num_input_tokens_seen": 102106115, + "step": 4730, + "time_per_iteration": 2.7391560077667236 + }, + { + "auxiliary_loss_clip": 0.01112326, + "auxiliary_loss_mlp": 0.01032042, + "balance_loss_clip": 1.04038906, + "balance_loss_mlp": 1.02433538, + "epoch": 0.5688691156135394, + "flos": 20808725414400.0, + "grad_norm": 1.7554426634940214, + "language_loss": 0.7383641, + "learning_rate": 1.6527888824710642e-06, + "loss": 0.75980777, + "num_input_tokens_seen": 102125715, + "step": 4731, + "time_per_iteration": 2.979375123977661 + }, + { + "auxiliary_loss_clip": 0.01141879, + "auxiliary_loss_mlp": 0.01029552, + "balance_loss_clip": 1.04286408, + "balance_loss_mlp": 1.02115965, + "epoch": 0.5689893585041784, + "flos": 25880963829120.0, + "grad_norm": 2.2724316112382694, + "language_loss": 0.76839268, + "learning_rate": 1.6520217656555166e-06, + "loss": 0.79010701, + "num_input_tokens_seen": 102145005, + "step": 4732, + "time_per_iteration": 2.767655849456787 + }, + { + "auxiliary_loss_clip": 0.01145995, + "auxiliary_loss_mlp": 0.01029343, + "balance_loss_clip": 1.0435257, + "balance_loss_mlp": 1.02160072, + "epoch": 0.5691096013948175, + "flos": 23477463123840.0, + "grad_norm": 1.5325611067924418, + "language_loss": 0.70663595, + "learning_rate": 1.65125470162799e-06, + "loss": 0.72838938, + "num_input_tokens_seen": 102165360, + "step": 4733, + "time_per_iteration": 2.767587900161743 + }, + { + "auxiliary_loss_clip": 0.01154797, + "auxiliary_loss_mlp": 0.01030623, + "balance_loss_clip": 1.04616332, + "balance_loss_mlp": 1.02168798, + "epoch": 0.5692298442854566, + "flos": 18075600576000.0, + "grad_norm": 2.1763209746083794, + "language_loss": 0.697896, + "learning_rate": 1.6504876905048485e-06, + "loss": 0.71975023, + "num_input_tokens_seen": 102182320, + "step": 4734, + "time_per_iteration": 2.729081392288208 + }, + { + "auxiliary_loss_clip": 0.01176691, + "auxiliary_loss_mlp": 0.01029817, + "balance_loss_clip": 1.05080378, + "balance_loss_mlp": 1.02215183, + "epoch": 0.5693500871760957, + "flos": 23039317025280.0, + "grad_norm": 1.5908216529625796, + "language_loss": 0.72099209, + "learning_rate": 1.6497207324024464e-06, + "loss": 0.74305719, + "num_input_tokens_seen": 102201220, + "step": 4735, + "time_per_iteration": 2.862170696258545 + }, + { + "auxiliary_loss_clip": 0.01158968, + "auxiliary_loss_mlp": 0.01031116, + "balance_loss_clip": 1.0473156, + "balance_loss_mlp": 1.02264631, + "epoch": 0.5694703300667348, + "flos": 18989670902400.0, + "grad_norm": 1.9159779428274526, + "language_loss": 0.82738411, + "learning_rate": 1.6489538274371305e-06, + "loss": 0.84928501, + "num_input_tokens_seen": 102219825, + "step": 4736, + "time_per_iteration": 2.8549463748931885 + }, + { + "auxiliary_loss_clip": 0.01164158, + "auxiliary_loss_mlp": 0.01026273, + "balance_loss_clip": 1.05119956, + "balance_loss_mlp": 1.01818514, + "epoch": 0.5695905729573739, + "flos": 21908705558400.0, + "grad_norm": 2.0498632157115577, + "language_loss": 0.8303237, + "learning_rate": 1.6481869757252396e-06, + "loss": 0.85222799, + "num_input_tokens_seen": 102238160, + "step": 4737, + "time_per_iteration": 2.757920742034912 + }, + { + "auxiliary_loss_clip": 0.01167528, + "auxiliary_loss_mlp": 0.01029716, + "balance_loss_clip": 1.04902911, + "balance_loss_mlp": 1.02172852, + "epoch": 0.569710815848013, + "flos": 28476659232000.0, + "grad_norm": 1.8906724654789733, + "language_loss": 0.71779031, + "learning_rate": 1.647420177383105e-06, + "loss": 0.73976272, + "num_input_tokens_seen": 102261030, + "step": 4738, + "time_per_iteration": 2.8389885425567627 + }, + { + "auxiliary_loss_clip": 0.01166268, + "auxiliary_loss_mlp": 0.0102707, + "balance_loss_clip": 1.05220926, + "balance_loss_mlp": 1.01938152, + "epoch": 0.569831058738652, + "flos": 28366162018560.0, + "grad_norm": 2.3384316815684962, + "language_loss": 0.72693598, + "learning_rate": 1.646653432527049e-06, + "loss": 0.74886942, + "num_input_tokens_seen": 102281670, + "step": 4739, + "time_per_iteration": 2.773073196411133 + }, + { + "auxiliary_loss_clip": 0.01154076, + "auxiliary_loss_mlp": 0.01034152, + "balance_loss_clip": 1.05003452, + "balance_loss_mlp": 1.02583098, + "epoch": 0.5699513016292912, + "flos": 25849973370240.0, + "grad_norm": 1.4834753650391677, + "language_loss": 0.74200046, + "learning_rate": 1.645886741273387e-06, + "loss": 0.76388276, + "num_input_tokens_seen": 102303485, + "step": 4740, + "time_per_iteration": 3.7868595123291016 + }, + { + "auxiliary_loss_clip": 0.01145806, + "auxiliary_loss_mlp": 0.01029577, + "balance_loss_clip": 1.04978347, + "balance_loss_mlp": 1.02104783, + "epoch": 0.5700715445199303, + "flos": 18037858360320.0, + "grad_norm": 2.2803348425758885, + "language_loss": 0.74153173, + "learning_rate": 1.645120103738424e-06, + "loss": 0.76328552, + "num_input_tokens_seen": 102320995, + "step": 4741, + "time_per_iteration": 3.715160608291626 + }, + { + "auxiliary_loss_clip": 0.01156338, + "auxiliary_loss_mlp": 0.00885622, + "balance_loss_clip": 1.04612398, + "balance_loss_mlp": 1.00067592, + "epoch": 0.5701917874105693, + "flos": 11473352392320.0, + "grad_norm": 2.269533723650647, + "language_loss": 0.83801842, + "learning_rate": 1.6443535200384591e-06, + "loss": 0.85843801, + "num_input_tokens_seen": 102339170, + "step": 4742, + "time_per_iteration": 3.5508406162261963 + }, + { + "auxiliary_loss_clip": 0.01179924, + "auxiliary_loss_mlp": 0.01030322, + "balance_loss_clip": 1.05249631, + "balance_loss_mlp": 1.02216268, + "epoch": 0.5703120303012085, + "flos": 21761759018880.0, + "grad_norm": 1.9354561922203017, + "language_loss": 0.70873606, + "learning_rate": 1.6435869902897827e-06, + "loss": 0.73083854, + "num_input_tokens_seen": 102357750, + "step": 4743, + "time_per_iteration": 2.592972993850708 + }, + { + "auxiliary_loss_clip": 0.01051609, + "auxiliary_loss_mlp": 0.01010353, + "balance_loss_clip": 1.01401043, + "balance_loss_mlp": 1.00860023, + "epoch": 0.5704322731918475, + "flos": 56746258513920.0, + "grad_norm": 1.4280515549159003, + "language_loss": 0.61979198, + "learning_rate": 1.6428205146086764e-06, + "loss": 0.64041162, + "num_input_tokens_seen": 102419730, + "step": 4744, + "time_per_iteration": 3.298006296157837 + }, + { + "auxiliary_loss_clip": 0.01166932, + "auxiliary_loss_mlp": 0.01031569, + "balance_loss_clip": 1.04901314, + "balance_loss_mlp": 1.02343035, + "epoch": 0.5705525160824866, + "flos": 20741141975040.0, + "grad_norm": 1.5950343890479088, + "language_loss": 0.70305324, + "learning_rate": 1.6420540931114142e-06, + "loss": 0.72503817, + "num_input_tokens_seen": 102440320, + "step": 4745, + "time_per_iteration": 3.481031894683838 + }, + { + "auxiliary_loss_clip": 0.01161829, + "auxiliary_loss_mlp": 0.01032693, + "balance_loss_clip": 1.04939687, + "balance_loss_mlp": 1.02459896, + "epoch": 0.5706727589731257, + "flos": 18771262254720.0, + "grad_norm": 1.5706414006120242, + "language_loss": 0.7883321, + "learning_rate": 1.6412877259142616e-06, + "loss": 0.81027728, + "num_input_tokens_seen": 102460240, + "step": 4746, + "time_per_iteration": 2.611687660217285 + }, + { + "auxiliary_loss_clip": 0.01156931, + "auxiliary_loss_mlp": 0.01027143, + "balance_loss_clip": 1.04834163, + "balance_loss_mlp": 1.01860738, + "epoch": 0.5707930018637648, + "flos": 27634733372160.0, + "grad_norm": 2.1535502846990306, + "language_loss": 0.73824084, + "learning_rate": 1.6405214131334757e-06, + "loss": 0.76008153, + "num_input_tokens_seen": 102478765, + "step": 4747, + "time_per_iteration": 2.648634433746338 + }, + { + "auxiliary_loss_clip": 0.01137249, + "auxiliary_loss_mlp": 0.01028041, + "balance_loss_clip": 1.04681015, + "balance_loss_mlp": 1.02016771, + "epoch": 0.5709132447544039, + "flos": 27597673514880.0, + "grad_norm": 1.7487224578891172, + "language_loss": 0.7956214, + "learning_rate": 1.6397551548853052e-06, + "loss": 0.81727421, + "num_input_tokens_seen": 102496930, + "step": 4748, + "time_per_iteration": 2.7130026817321777 + }, + { + "auxiliary_loss_clip": 0.01159289, + "auxiliary_loss_mlp": 0.01030949, + "balance_loss_clip": 1.0489043, + "balance_loss_mlp": 1.02231848, + "epoch": 0.571033487645043, + "flos": 21686095019520.0, + "grad_norm": 1.8887098359789052, + "language_loss": 0.70855731, + "learning_rate": 1.6389889512859917e-06, + "loss": 0.73045969, + "num_input_tokens_seen": 102516590, + "step": 4749, + "time_per_iteration": 2.6081600189208984 + }, + { + "auxiliary_loss_clip": 0.01057951, + "auxiliary_loss_mlp": 0.01009496, + "balance_loss_clip": 1.01481187, + "balance_loss_mlp": 1.00774336, + "epoch": 0.5711537305356821, + "flos": 70181445980160.0, + "grad_norm": 0.8119790382428581, + "language_loss": 0.60414588, + "learning_rate": 1.638222802451767e-06, + "loss": 0.62482035, + "num_input_tokens_seen": 102578070, + "step": 4750, + "time_per_iteration": 3.3712778091430664 + }, + { + "auxiliary_loss_clip": 0.01160721, + "auxiliary_loss_mlp": 0.0103089, + "balance_loss_clip": 1.04916716, + "balance_loss_mlp": 1.0234102, + "epoch": 0.5712739734263211, + "flos": 24717494396160.0, + "grad_norm": 2.0886605842015866, + "language_loss": 0.75022602, + "learning_rate": 1.6374567084988561e-06, + "loss": 0.77214205, + "num_input_tokens_seen": 102599255, + "step": 4751, + "time_per_iteration": 2.7870094776153564 + }, + { + "auxiliary_loss_clip": 0.01160029, + "auxiliary_loss_mlp": 0.01032795, + "balance_loss_clip": 1.05192351, + "balance_loss_mlp": 1.02379489, + "epoch": 0.5713942163169603, + "flos": 26578169792640.0, + "grad_norm": 1.834265365102851, + "language_loss": 0.76544923, + "learning_rate": 1.6366906695434738e-06, + "loss": 0.78737748, + "num_input_tokens_seen": 102621775, + "step": 4752, + "time_per_iteration": 2.8441498279571533 + }, + { + "auxiliary_loss_clip": 0.01170042, + "auxiliary_loss_mlp": 0.01032436, + "balance_loss_clip": 1.05138206, + "balance_loss_mlp": 1.02453256, + "epoch": 0.5715144592075994, + "flos": 21142443697920.0, + "grad_norm": 1.866931487482019, + "language_loss": 0.86141706, + "learning_rate": 1.6359246857018275e-06, + "loss": 0.88344181, + "num_input_tokens_seen": 102639305, + "step": 4753, + "time_per_iteration": 2.887681245803833 + }, + { + "auxiliary_loss_clip": 0.01131718, + "auxiliary_loss_mlp": 0.01027457, + "balance_loss_clip": 1.0433656, + "balance_loss_mlp": 1.01934481, + "epoch": 0.5716347020982384, + "flos": 23330265189120.0, + "grad_norm": 2.0080361665214825, + "language_loss": 0.78214097, + "learning_rate": 1.6351587570901178e-06, + "loss": 0.80373275, + "num_input_tokens_seen": 102659430, + "step": 4754, + "time_per_iteration": 2.9226882457733154 + }, + { + "auxiliary_loss_clip": 0.01147612, + "auxiliary_loss_mlp": 0.01032461, + "balance_loss_clip": 1.04668641, + "balance_loss_mlp": 1.02364564, + "epoch": 0.5717549449888776, + "flos": 17009555806080.0, + "grad_norm": 2.555411166263876, + "language_loss": 0.75805789, + "learning_rate": 1.634392883824534e-06, + "loss": 0.77985859, + "num_input_tokens_seen": 102671430, + "step": 4755, + "time_per_iteration": 2.6887662410736084 + }, + { + "auxiliary_loss_clip": 0.0113768, + "auxiliary_loss_mlp": 0.01033471, + "balance_loss_clip": 1.0441854, + "balance_loss_mlp": 1.02525139, + "epoch": 0.5718751878795166, + "flos": 35518130922240.0, + "grad_norm": 1.8496380368493976, + "language_loss": 0.68009365, + "learning_rate": 1.6336270660212595e-06, + "loss": 0.70180517, + "num_input_tokens_seen": 102693025, + "step": 4756, + "time_per_iteration": 2.835901975631714 + }, + { + "auxiliary_loss_clip": 0.01158496, + "auxiliary_loss_mlp": 0.01036866, + "balance_loss_clip": 1.05142772, + "balance_loss_mlp": 1.028265, + "epoch": 0.5719954307701557, + "flos": 38613989255040.0, + "grad_norm": 1.967030520228842, + "language_loss": 0.65925843, + "learning_rate": 1.6328613037964676e-06, + "loss": 0.68121207, + "num_input_tokens_seen": 102716090, + "step": 4757, + "time_per_iteration": 2.884227752685547 + }, + { + "auxiliary_loss_clip": 0.01168718, + "auxiliary_loss_mlp": 0.01034243, + "balance_loss_clip": 1.04895687, + "balance_loss_mlp": 1.02554095, + "epoch": 0.5721156736607949, + "flos": 20631111638400.0, + "grad_norm": 2.99878311620997, + "language_loss": 0.67831534, + "learning_rate": 1.6320955972663241e-06, + "loss": 0.70034504, + "num_input_tokens_seen": 102735685, + "step": 4758, + "time_per_iteration": 2.620090961456299 + }, + { + "auxiliary_loss_clip": 0.01168232, + "auxiliary_loss_mlp": 0.01028461, + "balance_loss_clip": 1.04817235, + "balance_loss_mlp": 1.02005088, + "epoch": 0.5722359165514339, + "flos": 37415076076800.0, + "grad_norm": 1.638199152451621, + "language_loss": 0.65470779, + "learning_rate": 1.6313299465469857e-06, + "loss": 0.67667472, + "num_input_tokens_seen": 102758415, + "step": 4759, + "time_per_iteration": 2.7587170600891113 + }, + { + "auxiliary_loss_clip": 0.01167577, + "auxiliary_loss_mlp": 0.01031526, + "balance_loss_clip": 1.05042291, + "balance_loss_mlp": 1.02303219, + "epoch": 0.572356159442073, + "flos": 21972877205760.0, + "grad_norm": 2.9703457597775897, + "language_loss": 0.79537827, + "learning_rate": 1.6305643517546014e-06, + "loss": 0.81736922, + "num_input_tokens_seen": 102773795, + "step": 4760, + "time_per_iteration": 2.749729633331299 + }, + { + "auxiliary_loss_clip": 0.01178613, + "auxiliary_loss_mlp": 0.01032549, + "balance_loss_clip": 1.0518254, + "balance_loss_mlp": 1.02452588, + "epoch": 0.5724764023327121, + "flos": 19135540033920.0, + "grad_norm": 1.7479298987382381, + "language_loss": 0.84270537, + "learning_rate": 1.629798813005311e-06, + "loss": 0.8648169, + "num_input_tokens_seen": 102793515, + "step": 4761, + "time_per_iteration": 2.708768606185913 + }, + { + "auxiliary_loss_clip": 0.01143963, + "auxiliary_loss_mlp": 0.0102893, + "balance_loss_clip": 1.04802239, + "balance_loss_mlp": 1.02119935, + "epoch": 0.5725966452233512, + "flos": 22819759142400.0, + "grad_norm": 1.9483793825748827, + "language_loss": 0.70503867, + "learning_rate": 1.6290333304152473e-06, + "loss": 0.72676754, + "num_input_tokens_seen": 102813390, + "step": 4762, + "time_per_iteration": 2.9925010204315186 + }, + { + "auxiliary_loss_clip": 0.01156578, + "auxiliary_loss_mlp": 0.01026335, + "balance_loss_clip": 1.05161154, + "balance_loss_mlp": 1.01800823, + "epoch": 0.5727168881139902, + "flos": 41496610498560.0, + "grad_norm": 1.793997967721978, + "language_loss": 0.56642854, + "learning_rate": 1.6282679041005314e-06, + "loss": 0.58825767, + "num_input_tokens_seen": 102838980, + "step": 4763, + "time_per_iteration": 2.955249071121216 + }, + { + "auxiliary_loss_clip": 0.01153806, + "auxiliary_loss_mlp": 0.01029087, + "balance_loss_clip": 1.04541445, + "balance_loss_mlp": 1.02070069, + "epoch": 0.5728371310046293, + "flos": 14647675985280.0, + "grad_norm": 2.4672694226814667, + "language_loss": 0.8704046, + "learning_rate": 1.6275025341772789e-06, + "loss": 0.89223355, + "num_input_tokens_seen": 102855285, + "step": 4764, + "time_per_iteration": 2.89270281791687 + }, + { + "auxiliary_loss_clip": 0.01160126, + "auxiliary_loss_mlp": 0.01031476, + "balance_loss_clip": 1.04695117, + "balance_loss_mlp": 1.02281547, + "epoch": 0.5729573738952685, + "flos": 21506613736320.0, + "grad_norm": 2.6222566317468132, + "language_loss": 0.8147676, + "learning_rate": 1.626737220761596e-06, + "loss": 0.83668363, + "num_input_tokens_seen": 102872750, + "step": 4765, + "time_per_iteration": 2.7656102180480957 + }, + { + "auxiliary_loss_clip": 0.01168151, + "auxiliary_loss_mlp": 0.01035062, + "balance_loss_clip": 1.05019963, + "balance_loss_mlp": 1.0266695, + "epoch": 0.5730776167859075, + "flos": 23621680229760.0, + "grad_norm": 2.452542620183796, + "language_loss": 0.78788114, + "learning_rate": 1.62597196396958e-06, + "loss": 0.80991328, + "num_input_tokens_seen": 102890920, + "step": 4766, + "time_per_iteration": 3.7312045097351074 + }, + { + "auxiliary_loss_clip": 0.01168572, + "auxiliary_loss_mlp": 0.01027258, + "balance_loss_clip": 1.05070865, + "balance_loss_mlp": 1.01868105, + "epoch": 0.5731978596765466, + "flos": 25739224761600.0, + "grad_norm": 1.6977117237627646, + "language_loss": 0.85720038, + "learning_rate": 1.6252067639173197e-06, + "loss": 0.87915862, + "num_input_tokens_seen": 102912830, + "step": 4767, + "time_per_iteration": 3.811121702194214 + }, + { + "auxiliary_loss_clip": 0.01170217, + "auxiliary_loss_mlp": 0.01030514, + "balance_loss_clip": 1.04996729, + "balance_loss_mlp": 1.02198446, + "epoch": 0.5733181025671857, + "flos": 26359509749760.0, + "grad_norm": 6.100470849589001, + "language_loss": 0.69702232, + "learning_rate": 1.6244416207208956e-06, + "loss": 0.71902966, + "num_input_tokens_seen": 102933765, + "step": 4768, + "time_per_iteration": 3.9056215286254883 + }, + { + "auxiliary_loss_clip": 0.01150936, + "auxiliary_loss_mlp": 0.01030895, + "balance_loss_clip": 1.04713607, + "balance_loss_mlp": 1.02312827, + "epoch": 0.5734383454578248, + "flos": 29423874833280.0, + "grad_norm": 8.041491057136966, + "language_loss": 0.7382043, + "learning_rate": 1.6236765344963787e-06, + "loss": 0.76002258, + "num_input_tokens_seen": 102955025, + "step": 4769, + "time_per_iteration": 2.892503023147583 + }, + { + "auxiliary_loss_clip": 0.01160561, + "auxiliary_loss_mlp": 0.01026385, + "balance_loss_clip": 1.04902267, + "balance_loss_mlp": 1.01791477, + "epoch": 0.5735585883484638, + "flos": 34969954487040.0, + "grad_norm": 2.259420089825551, + "language_loss": 0.69216698, + "learning_rate": 1.6229115053598322e-06, + "loss": 0.71403646, + "num_input_tokens_seen": 102976780, + "step": 4770, + "time_per_iteration": 3.832111120223999 + }, + { + "auxiliary_loss_clip": 0.01171621, + "auxiliary_loss_mlp": 0.01028959, + "balance_loss_clip": 1.0531863, + "balance_loss_mlp": 1.02064979, + "epoch": 0.573678831239103, + "flos": 18770759464320.0, + "grad_norm": 1.900292802379095, + "language_loss": 0.7259115, + "learning_rate": 1.6221465334273108e-06, + "loss": 0.74791729, + "num_input_tokens_seen": 102995990, + "step": 4771, + "time_per_iteration": 2.7634284496307373 + }, + { + "auxiliary_loss_clip": 0.01155489, + "auxiliary_loss_mlp": 0.01027246, + "balance_loss_clip": 1.04799306, + "balance_loss_mlp": 1.01855016, + "epoch": 0.5737990741297421, + "flos": 25702883176320.0, + "grad_norm": 2.6128221903275297, + "language_loss": 0.61676359, + "learning_rate": 1.6213816188148593e-06, + "loss": 0.63859093, + "num_input_tokens_seen": 103014695, + "step": 4772, + "time_per_iteration": 2.852609872817993 + }, + { + "auxiliary_loss_clip": 0.01150704, + "auxiliary_loss_mlp": 0.0102639, + "balance_loss_clip": 1.04978824, + "balance_loss_mlp": 1.01758647, + "epoch": 0.5739193170203811, + "flos": 27269234530560.0, + "grad_norm": 2.089709643393088, + "language_loss": 0.77015245, + "learning_rate": 1.6206167616385162e-06, + "loss": 0.79192328, + "num_input_tokens_seen": 103035760, + "step": 4773, + "time_per_iteration": 2.8684656620025635 + }, + { + "auxiliary_loss_clip": 0.01169353, + "auxiliary_loss_mlp": 0.01032716, + "balance_loss_clip": 1.05094433, + "balance_loss_mlp": 1.02317953, + "epoch": 0.5740395599110203, + "flos": 12239721993600.0, + "grad_norm": 2.080350376692058, + "language_loss": 0.73858327, + "learning_rate": 1.6198519620143078e-06, + "loss": 0.76060396, + "num_input_tokens_seen": 103052915, + "step": 4774, + "time_per_iteration": 2.758601427078247 + }, + { + "auxiliary_loss_clip": 0.01158191, + "auxiliary_loss_mlp": 0.01029755, + "balance_loss_clip": 1.05094719, + "balance_loss_mlp": 1.02162445, + "epoch": 0.5741598028016593, + "flos": 25921399564800.0, + "grad_norm": 1.6078210003218947, + "language_loss": 0.78082371, + "learning_rate": 1.6190872200582546e-06, + "loss": 0.80270314, + "num_input_tokens_seen": 103074655, + "step": 4775, + "time_per_iteration": 2.9424664974212646 + }, + { + "auxiliary_loss_clip": 0.01155728, + "auxiliary_loss_mlp": 0.0088597, + "balance_loss_clip": 1.04786587, + "balance_loss_mlp": 1.0007621, + "epoch": 0.5742800456922984, + "flos": 19244133826560.0, + "grad_norm": 2.1391274457182394, + "language_loss": 0.77982742, + "learning_rate": 1.6183225358863676e-06, + "loss": 0.80024439, + "num_input_tokens_seen": 103091550, + "step": 4776, + "time_per_iteration": 2.8157622814178467 + }, + { + "auxiliary_loss_clip": 0.01153645, + "auxiliary_loss_mlp": 0.01030688, + "balance_loss_clip": 1.04645443, + "balance_loss_mlp": 1.02152658, + "epoch": 0.5744002885829376, + "flos": 30920487932160.0, + "grad_norm": 2.5540360118951093, + "language_loss": 0.71967316, + "learning_rate": 1.617557909614648e-06, + "loss": 0.74151647, + "num_input_tokens_seen": 103110985, + "step": 4777, + "time_per_iteration": 2.867934226989746 + }, + { + "auxiliary_loss_clip": 0.01150198, + "auxiliary_loss_mlp": 0.01028957, + "balance_loss_clip": 1.04711103, + "balance_loss_mlp": 1.02040339, + "epoch": 0.5745205314735766, + "flos": 23840017050240.0, + "grad_norm": 1.8924135595581748, + "language_loss": 0.86064506, + "learning_rate": 1.6167933413590899e-06, + "loss": 0.88243657, + "num_input_tokens_seen": 103129890, + "step": 4778, + "time_per_iteration": 2.9201717376708984 + }, + { + "auxiliary_loss_clip": 0.01169242, + "auxiliary_loss_mlp": 0.01029451, + "balance_loss_clip": 1.04936361, + "balance_loss_mlp": 1.02020013, + "epoch": 0.5746407743642157, + "flos": 12311902373760.0, + "grad_norm": 1.9962122459063665, + "language_loss": 0.90451968, + "learning_rate": 1.6160288312356773e-06, + "loss": 0.92650664, + "num_input_tokens_seen": 103147020, + "step": 4779, + "time_per_iteration": 2.850296974182129 + }, + { + "auxiliary_loss_clip": 0.01172964, + "auxiliary_loss_mlp": 0.01027255, + "balance_loss_clip": 1.04925442, + "balance_loss_mlp": 1.01860607, + "epoch": 0.5747610172548548, + "flos": 24133658734080.0, + "grad_norm": 4.313591699544792, + "language_loss": 0.81598008, + "learning_rate": 1.6152643793603857e-06, + "loss": 0.83798224, + "num_input_tokens_seen": 103167370, + "step": 4780, + "time_per_iteration": 2.8044331073760986 + }, + { + "auxiliary_loss_clip": 0.0118002, + "auxiliary_loss_mlp": 0.01032311, + "balance_loss_clip": 1.05250168, + "balance_loss_mlp": 1.02423453, + "epoch": 0.5748812601454939, + "flos": 25408451393280.0, + "grad_norm": 1.7608914130422462, + "language_loss": 0.87560868, + "learning_rate": 1.6144999858491815e-06, + "loss": 0.89773202, + "num_input_tokens_seen": 103186000, + "step": 4781, + "time_per_iteration": 2.7660627365112305 + }, + { + "auxiliary_loss_clip": 0.01164554, + "auxiliary_loss_mlp": 0.01032555, + "balance_loss_clip": 1.04774141, + "balance_loss_mlp": 1.02371609, + "epoch": 0.575001503036133, + "flos": 30624942827520.0, + "grad_norm": 1.8958204592169765, + "language_loss": 0.85807818, + "learning_rate": 1.6137356508180232e-06, + "loss": 0.88004929, + "num_input_tokens_seen": 103207710, + "step": 4782, + "time_per_iteration": 2.8486790657043457 + }, + { + "auxiliary_loss_clip": 0.01178202, + "auxiliary_loss_mlp": 0.00886599, + "balance_loss_clip": 1.04940534, + "balance_loss_mlp": 1.00080264, + "epoch": 0.5751217459267721, + "flos": 21726566668800.0, + "grad_norm": 1.8513347923494257, + "language_loss": 0.81141019, + "learning_rate": 1.6129713743828593e-06, + "loss": 0.83205819, + "num_input_tokens_seen": 103226720, + "step": 4783, + "time_per_iteration": 2.6948659420013428 + }, + { + "auxiliary_loss_clip": 0.01156443, + "auxiliary_loss_mlp": 0.01032688, + "balance_loss_clip": 1.0440706, + "balance_loss_mlp": 1.02424812, + "epoch": 0.5752419888174112, + "flos": 21651620941440.0, + "grad_norm": 1.5381125822408575, + "language_loss": 0.75628948, + "learning_rate": 1.6122071566596306e-06, + "loss": 0.77818078, + "num_input_tokens_seen": 103246995, + "step": 4784, + "time_per_iteration": 2.7716329097747803 + }, + { + "auxiliary_loss_clip": 0.0117114, + "auxiliary_loss_mlp": 0.01033208, + "balance_loss_clip": 1.05036521, + "balance_loss_mlp": 1.02481556, + "epoch": 0.5753622317080502, + "flos": 17775997234560.0, + "grad_norm": 2.238539904119028, + "language_loss": 0.83219552, + "learning_rate": 1.6114429977642674e-06, + "loss": 0.85423899, + "num_input_tokens_seen": 103261500, + "step": 4785, + "time_per_iteration": 2.660249948501587 + }, + { + "auxiliary_loss_clip": 0.01172089, + "auxiliary_loss_mlp": 0.01032159, + "balance_loss_clip": 1.05306649, + "balance_loss_mlp": 1.02396369, + "epoch": 0.5754824745986894, + "flos": 19789616741760.0, + "grad_norm": 1.7058312821953818, + "language_loss": 0.73555946, + "learning_rate": 1.6106788978126926e-06, + "loss": 0.75760198, + "num_input_tokens_seen": 103280475, + "step": 4786, + "time_per_iteration": 2.6927571296691895 + }, + { + "auxiliary_loss_clip": 0.01139416, + "auxiliary_loss_mlp": 0.01031323, + "balance_loss_clip": 1.04488981, + "balance_loss_mlp": 1.02206635, + "epoch": 0.5756027174893285, + "flos": 30985665160320.0, + "grad_norm": 6.663262395940318, + "language_loss": 0.79151338, + "learning_rate": 1.6099148569208196e-06, + "loss": 0.81322074, + "num_input_tokens_seen": 103297695, + "step": 4787, + "time_per_iteration": 2.9423587322235107 + }, + { + "auxiliary_loss_clip": 0.01162819, + "auxiliary_loss_mlp": 0.01028029, + "balance_loss_clip": 1.05254459, + "balance_loss_mlp": 1.01939785, + "epoch": 0.5757229603799675, + "flos": 28546864364160.0, + "grad_norm": 2.8322807150368505, + "language_loss": 0.62657332, + "learning_rate": 1.6091508752045523e-06, + "loss": 0.64848179, + "num_input_tokens_seen": 103318575, + "step": 4788, + "time_per_iteration": 2.7747058868408203 + }, + { + "auxiliary_loss_clip": 0.01142902, + "auxiliary_loss_mlp": 0.01030391, + "balance_loss_clip": 1.04345179, + "balance_loss_mlp": 1.02210605, + "epoch": 0.5758432032706067, + "flos": 22999024944000.0, + "grad_norm": 1.5362905457788454, + "language_loss": 0.86458004, + "learning_rate": 1.608386952779787e-06, + "loss": 0.88631296, + "num_input_tokens_seen": 103337945, + "step": 4789, + "time_per_iteration": 2.8072736263275146 + }, + { + "auxiliary_loss_clip": 0.01166435, + "auxiliary_loss_mlp": 0.010366, + "balance_loss_clip": 1.05001104, + "balance_loss_mlp": 1.02823138, + "epoch": 0.5759634461612457, + "flos": 25739727552000.0, + "grad_norm": 1.8597309819163743, + "language_loss": 0.74910218, + "learning_rate": 1.6076230897624098e-06, + "loss": 0.77113247, + "num_input_tokens_seen": 103360150, + "step": 4790, + "time_per_iteration": 2.7657933235168457 + }, + { + "auxiliary_loss_clip": 0.01172229, + "auxiliary_loss_mlp": 0.01024464, + "balance_loss_clip": 1.04868698, + "balance_loss_mlp": 1.01604748, + "epoch": 0.5760836890518848, + "flos": 30591761639040.0, + "grad_norm": 2.816657560899159, + "language_loss": 0.77587873, + "learning_rate": 1.6068592862682974e-06, + "loss": 0.7978456, + "num_input_tokens_seen": 103378305, + "step": 4791, + "time_per_iteration": 3.658707618713379 + }, + { + "auxiliary_loss_clip": 0.01164919, + "auxiliary_loss_mlp": 0.01027212, + "balance_loss_clip": 1.05104995, + "balance_loss_mlp": 1.0187782, + "epoch": 0.576203931942524, + "flos": 36538963447680.0, + "grad_norm": 1.9456476772622082, + "language_loss": 0.73959607, + "learning_rate": 1.6060955424133187e-06, + "loss": 0.76151741, + "num_input_tokens_seen": 103399230, + "step": 4792, + "time_per_iteration": 3.7972476482391357 + }, + { + "auxiliary_loss_clip": 0.01169702, + "auxiliary_loss_mlp": 0.01027196, + "balance_loss_clip": 1.05094063, + "balance_loss_mlp": 1.01844573, + "epoch": 0.576324174833163, + "flos": 25516937445120.0, + "grad_norm": 1.6467906349650552, + "language_loss": 0.89513755, + "learning_rate": 1.6053318583133332e-06, + "loss": 0.91710651, + "num_input_tokens_seen": 103420100, + "step": 4793, + "time_per_iteration": 2.728074073791504 + }, + { + "auxiliary_loss_clip": 0.01167031, + "auxiliary_loss_mlp": 0.01032101, + "balance_loss_clip": 1.04927993, + "balance_loss_mlp": 1.0233928, + "epoch": 0.5764444177238021, + "flos": 25119262995840.0, + "grad_norm": 2.2182907993302696, + "language_loss": 0.74991351, + "learning_rate": 1.6045682340841907e-06, + "loss": 0.77190483, + "num_input_tokens_seen": 103439025, + "step": 4794, + "time_per_iteration": 3.6759209632873535 + }, + { + "auxiliary_loss_clip": 0.01056572, + "auxiliary_loss_mlp": 0.00876862, + "balance_loss_clip": 1.01745868, + "balance_loss_mlp": 1.00114059, + "epoch": 0.5765646606144411, + "flos": 62212687758720.0, + "grad_norm": 0.7518748874438463, + "language_loss": 0.5800252, + "learning_rate": 1.6038046698417336e-06, + "loss": 0.59935951, + "num_input_tokens_seen": 103499920, + "step": 4795, + "time_per_iteration": 3.338759422302246 + }, + { + "auxiliary_loss_clip": 0.01170535, + "auxiliary_loss_mlp": 0.01028829, + "balance_loss_clip": 1.05105376, + "balance_loss_mlp": 1.02025771, + "epoch": 0.5766849035050803, + "flos": 25118760205440.0, + "grad_norm": 2.5103320651596874, + "language_loss": 0.68616521, + "learning_rate": 1.6030411657017919e-06, + "loss": 0.70815885, + "num_input_tokens_seen": 103519575, + "step": 4796, + "time_per_iteration": 3.762212038040161 + }, + { + "auxiliary_loss_clip": 0.01164569, + "auxiliary_loss_mlp": 0.01030199, + "balance_loss_clip": 1.05016196, + "balance_loss_mlp": 1.02231944, + "epoch": 0.5768051463957193, + "flos": 15991093578240.0, + "grad_norm": 2.059431824534267, + "language_loss": 0.84359443, + "learning_rate": 1.6022777217801903e-06, + "loss": 0.86554211, + "num_input_tokens_seen": 103536530, + "step": 4797, + "time_per_iteration": 2.730773448944092 + }, + { + "auxiliary_loss_clip": 0.0115829, + "auxiliary_loss_mlp": 0.01026946, + "balance_loss_clip": 1.05251455, + "balance_loss_mlp": 1.01826191, + "epoch": 0.5769253892863584, + "flos": 22163635359360.0, + "grad_norm": 3.6788699296807836, + "language_loss": 0.73880768, + "learning_rate": 1.601514338192742e-06, + "loss": 0.76066005, + "num_input_tokens_seen": 103556460, + "step": 4798, + "time_per_iteration": 2.879429340362549 + }, + { + "auxiliary_loss_clip": 0.01176529, + "auxiliary_loss_mlp": 0.01028242, + "balance_loss_clip": 1.05126894, + "balance_loss_mlp": 1.02080917, + "epoch": 0.5770456321769976, + "flos": 22856388036480.0, + "grad_norm": 2.2100111274489063, + "language_loss": 0.71554077, + "learning_rate": 1.6007510150552514e-06, + "loss": 0.73758841, + "num_input_tokens_seen": 103574520, + "step": 4799, + "time_per_iteration": 2.821218490600586 + }, + { + "auxiliary_loss_clip": 0.01176534, + "auxiliary_loss_mlp": 0.0103096, + "balance_loss_clip": 1.05157304, + "balance_loss_mlp": 1.02212691, + "epoch": 0.5771658750676366, + "flos": 46353672489600.0, + "grad_norm": 1.4871116324759195, + "language_loss": 0.61885262, + "learning_rate": 1.599987752483515e-06, + "loss": 0.64092755, + "num_input_tokens_seen": 103598965, + "step": 4800, + "time_per_iteration": 2.929039716720581 + }, + { + "auxiliary_loss_clip": 0.01144504, + "auxiliary_loss_mlp": 0.01030854, + "balance_loss_clip": 1.04564464, + "balance_loss_mlp": 1.0228219, + "epoch": 0.5772861179582757, + "flos": 22159972172160.0, + "grad_norm": 2.2372106913397465, + "language_loss": 0.68133336, + "learning_rate": 1.5992245505933184e-06, + "loss": 0.70308697, + "num_input_tokens_seen": 103618665, + "step": 4801, + "time_per_iteration": 2.894458293914795 + }, + { + "auxiliary_loss_clip": 0.01180677, + "auxiliary_loss_mlp": 0.01025069, + "balance_loss_clip": 1.05277002, + "balance_loss_mlp": 1.01662314, + "epoch": 0.5774063608489148, + "flos": 31248926916480.0, + "grad_norm": 2.198372535795739, + "language_loss": 0.71100217, + "learning_rate": 1.5984614095004388e-06, + "loss": 0.73305964, + "num_input_tokens_seen": 103639800, + "step": 4802, + "time_per_iteration": 2.7096312046051025 + }, + { + "auxiliary_loss_clip": 0.01163776, + "auxiliary_loss_mlp": 0.01032077, + "balance_loss_clip": 1.0477103, + "balance_loss_mlp": 1.02334523, + "epoch": 0.5775266037395539, + "flos": 22527123039360.0, + "grad_norm": 2.5864564214345878, + "language_loss": 0.80557972, + "learning_rate": 1.5976983293206438e-06, + "loss": 0.82753825, + "num_input_tokens_seen": 103655605, + "step": 4803, + "time_per_iteration": 2.6645853519439697 + }, + { + "auxiliary_loss_clip": 0.01156152, + "auxiliary_loss_mlp": 0.01031746, + "balance_loss_clip": 1.0459789, + "balance_loss_mlp": 1.02329445, + "epoch": 0.577646846630193, + "flos": 21068790860160.0, + "grad_norm": 1.7000683521654116, + "language_loss": 0.71324146, + "learning_rate": 1.5969353101696928e-06, + "loss": 0.73512042, + "num_input_tokens_seen": 103674045, + "step": 4804, + "time_per_iteration": 2.7813222408294678 + }, + { + "auxiliary_loss_clip": 0.01170195, + "auxiliary_loss_mlp": 0.01029886, + "balance_loss_clip": 1.04884112, + "balance_loss_mlp": 1.02139854, + "epoch": 0.5777670895208321, + "flos": 29714284293120.0, + "grad_norm": 2.7335093800166717, + "language_loss": 0.79616386, + "learning_rate": 1.5961723521633341e-06, + "loss": 0.81816465, + "num_input_tokens_seen": 103695285, + "step": 4805, + "time_per_iteration": 2.789475679397583 + }, + { + "auxiliary_loss_clip": 0.01149833, + "auxiliary_loss_mlp": 0.01028298, + "balance_loss_clip": 1.04754066, + "balance_loss_mlp": 1.01978028, + "epoch": 0.5778873324114712, + "flos": 19500428344320.0, + "grad_norm": 2.0690010657253435, + "language_loss": 0.90869093, + "learning_rate": 1.5954094554173097e-06, + "loss": 0.93047225, + "num_input_tokens_seen": 103713275, + "step": 4806, + "time_per_iteration": 2.7186176776885986 + }, + { + "auxiliary_loss_clip": 0.01167601, + "auxiliary_loss_mlp": 0.01030089, + "balance_loss_clip": 1.05308259, + "balance_loss_mlp": 1.02235794, + "epoch": 0.5780075753021102, + "flos": 14136846716160.0, + "grad_norm": 3.1233527632316465, + "language_loss": 0.79260081, + "learning_rate": 1.5946466200473482e-06, + "loss": 0.8145777, + "num_input_tokens_seen": 103731185, + "step": 4807, + "time_per_iteration": 2.7073614597320557 + }, + { + "auxiliary_loss_clip": 0.01164463, + "auxiliary_loss_mlp": 0.0102783, + "balance_loss_clip": 1.04759514, + "balance_loss_mlp": 1.01974177, + "epoch": 0.5781278181927494, + "flos": 15262178883840.0, + "grad_norm": 1.8405648236084755, + "language_loss": 0.83761162, + "learning_rate": 1.5938838461691723e-06, + "loss": 0.85953456, + "num_input_tokens_seen": 103748095, + "step": 4808, + "time_per_iteration": 2.713712692260742 + }, + { + "auxiliary_loss_clip": 0.0118226, + "auxiliary_loss_mlp": 0.01040178, + "balance_loss_clip": 1.05391347, + "balance_loss_mlp": 1.03153539, + "epoch": 0.5782480610833884, + "flos": 16726831856640.0, + "grad_norm": 3.1213221954624735, + "language_loss": 0.82393765, + "learning_rate": 1.593121133898494e-06, + "loss": 0.84616202, + "num_input_tokens_seen": 103765300, + "step": 4809, + "time_per_iteration": 2.6528422832489014 + }, + { + "auxiliary_loss_clip": 0.01178424, + "auxiliary_loss_mlp": 0.01037527, + "balance_loss_clip": 1.05232418, + "balance_loss_mlp": 1.02915239, + "epoch": 0.5783683039740275, + "flos": 25482140144640.0, + "grad_norm": 2.0860767096276156, + "language_loss": 0.78970492, + "learning_rate": 1.592358483351016e-06, + "loss": 0.81186444, + "num_input_tokens_seen": 103785475, + "step": 4810, + "time_per_iteration": 2.730665445327759 + }, + { + "auxiliary_loss_clip": 0.01170105, + "auxiliary_loss_mlp": 0.01030852, + "balance_loss_clip": 1.0519383, + "balance_loss_mlp": 1.02285886, + "epoch": 0.5784885468646667, + "flos": 18405835240320.0, + "grad_norm": 2.647354724029795, + "language_loss": 0.72474116, + "learning_rate": 1.5915958946424326e-06, + "loss": 0.74675071, + "num_input_tokens_seen": 103804160, + "step": 4811, + "time_per_iteration": 2.774285316467285 + }, + { + "auxiliary_loss_clip": 0.01153941, + "auxiliary_loss_mlp": 0.00886531, + "balance_loss_clip": 1.05010223, + "balance_loss_mlp": 1.00070477, + "epoch": 0.5786087897553057, + "flos": 46100717936640.0, + "grad_norm": 3.3919943760493885, + "language_loss": 0.74497306, + "learning_rate": 1.5908333678884271e-06, + "loss": 0.76537776, + "num_input_tokens_seen": 103830580, + "step": 4812, + "time_per_iteration": 3.059596538543701 + }, + { + "auxiliary_loss_clip": 0.01170952, + "auxiliary_loss_mlp": 0.01030736, + "balance_loss_clip": 1.05140591, + "balance_loss_mlp": 1.02251017, + "epoch": 0.5787290326459448, + "flos": 12385950261120.0, + "grad_norm": 1.9102048632414963, + "language_loss": 0.74131858, + "learning_rate": 1.5900709032046743e-06, + "loss": 0.76333547, + "num_input_tokens_seen": 103848655, + "step": 4813, + "time_per_iteration": 2.707719087600708 + }, + { + "auxiliary_loss_clip": 0.01157529, + "auxiliary_loss_mlp": 0.01030965, + "balance_loss_clip": 1.05135572, + "balance_loss_mlp": 1.02248371, + "epoch": 0.5788492755365839, + "flos": 23290332243840.0, + "grad_norm": 2.4357353172434824, + "language_loss": 0.77912861, + "learning_rate": 1.5893085007068391e-06, + "loss": 0.80101353, + "num_input_tokens_seen": 103866215, + "step": 4814, + "time_per_iteration": 2.799089193344116 + }, + { + "auxiliary_loss_clip": 0.01152397, + "auxiliary_loss_mlp": 0.01029023, + "balance_loss_clip": 1.04910779, + "balance_loss_mlp": 1.02074981, + "epoch": 0.578969518427223, + "flos": 24061047390720.0, + "grad_norm": 1.8345332457432157, + "language_loss": 0.70874619, + "learning_rate": 1.5885461605105786e-06, + "loss": 0.73056042, + "num_input_tokens_seen": 103887815, + "step": 4815, + "time_per_iteration": 2.8100712299346924 + }, + { + "auxiliary_loss_clip": 0.01163843, + "auxiliary_loss_mlp": 0.01024008, + "balance_loss_clip": 1.05091763, + "balance_loss_mlp": 1.01543105, + "epoch": 0.579089761317862, + "flos": 21871825269120.0, + "grad_norm": 1.9378994722970282, + "language_loss": 0.7662437, + "learning_rate": 1.5877838827315375e-06, + "loss": 0.78812218, + "num_input_tokens_seen": 103906360, + "step": 4816, + "time_per_iteration": 2.804121971130371 + }, + { + "auxiliary_loss_clip": 0.01180994, + "auxiliary_loss_mlp": 0.01034265, + "balance_loss_clip": 1.05359221, + "balance_loss_mlp": 1.02536047, + "epoch": 0.5792100042085012, + "flos": 22929681738240.0, + "grad_norm": 1.83725185844357, + "language_loss": 0.7029779, + "learning_rate": 1.587021667485355e-06, + "loss": 0.72513056, + "num_input_tokens_seen": 103925730, + "step": 4817, + "time_per_iteration": 3.5823607444763184 + }, + { + "auxiliary_loss_clip": 0.01165416, + "auxiliary_loss_mlp": 0.01028149, + "balance_loss_clip": 1.04947317, + "balance_loss_mlp": 1.02015281, + "epoch": 0.5793302470991403, + "flos": 21470056669440.0, + "grad_norm": 2.3775859301133524, + "language_loss": 0.78527099, + "learning_rate": 1.5862595148876559e-06, + "loss": 0.80720663, + "num_input_tokens_seen": 103945835, + "step": 4818, + "time_per_iteration": 3.7298057079315186 + }, + { + "auxiliary_loss_clip": 0.01143475, + "auxiliary_loss_mlp": 0.01028087, + "balance_loss_clip": 1.04568624, + "balance_loss_mlp": 1.01977861, + "epoch": 0.5794504899897793, + "flos": 12711013367040.0, + "grad_norm": 2.53416193626311, + "language_loss": 0.75792509, + "learning_rate": 1.58549742505406e-06, + "loss": 0.77964073, + "num_input_tokens_seen": 103960580, + "step": 4819, + "time_per_iteration": 2.9892725944519043 + }, + { + "auxiliary_loss_clip": 0.01181618, + "auxiliary_loss_mlp": 0.01034116, + "balance_loss_clip": 1.05342782, + "balance_loss_mlp": 1.02565193, + "epoch": 0.5795707328804185, + "flos": 14867054300160.0, + "grad_norm": 2.251351262720086, + "language_loss": 0.75509381, + "learning_rate": 1.5847353981001747e-06, + "loss": 0.77725112, + "num_input_tokens_seen": 103977760, + "step": 4820, + "time_per_iteration": 3.588331460952759 + }, + { + "auxiliary_loss_clip": 0.01152491, + "auxiliary_loss_mlp": 0.01024142, + "balance_loss_clip": 1.04494524, + "balance_loss_mlp": 1.01543427, + "epoch": 0.5796909757710575, + "flos": 36430046432640.0, + "grad_norm": 1.6893208023268538, + "language_loss": 0.69582403, + "learning_rate": 1.5839734341415993e-06, + "loss": 0.71759033, + "num_input_tokens_seen": 103999960, + "step": 4821, + "time_per_iteration": 2.856166362762451 + }, + { + "auxiliary_loss_clip": 0.01169159, + "auxiliary_loss_mlp": 0.01030712, + "balance_loss_clip": 1.05433989, + "balance_loss_mlp": 1.02274883, + "epoch": 0.5798112186616966, + "flos": 23039891642880.0, + "grad_norm": 1.9333546396656074, + "language_loss": 0.76394367, + "learning_rate": 1.5832115332939238e-06, + "loss": 0.78594238, + "num_input_tokens_seen": 104018400, + "step": 4822, + "time_per_iteration": 3.6653847694396973 + }, + { + "auxiliary_loss_clip": 0.0117013, + "auxiliary_loss_mlp": 0.01034365, + "balance_loss_clip": 1.04961824, + "balance_loss_mlp": 1.02628303, + "epoch": 0.5799314615523358, + "flos": 16652604401280.0, + "grad_norm": 1.7339217305721752, + "language_loss": 0.74631006, + "learning_rate": 1.5824496956727272e-06, + "loss": 0.76835501, + "num_input_tokens_seen": 104035605, + "step": 4823, + "time_per_iteration": 2.7528090476989746 + }, + { + "auxiliary_loss_clip": 0.01160879, + "auxiliary_loss_mlp": 0.01028154, + "balance_loss_clip": 1.04849935, + "balance_loss_mlp": 1.02034616, + "epoch": 0.5800517044429748, + "flos": 20485673470080.0, + "grad_norm": 1.736461124261528, + "language_loss": 0.72889042, + "learning_rate": 1.5816879213935797e-06, + "loss": 0.75078082, + "num_input_tokens_seen": 104054415, + "step": 4824, + "time_per_iteration": 2.873262405395508 + }, + { + "auxiliary_loss_clip": 0.01167397, + "auxiliary_loss_mlp": 0.01029253, + "balance_loss_clip": 1.05120659, + "balance_loss_mlp": 1.02152276, + "epoch": 0.5801719473336139, + "flos": 31538258968320.0, + "grad_norm": 1.5233413044232864, + "language_loss": 0.79444015, + "learning_rate": 1.5809262105720416e-06, + "loss": 0.81640661, + "num_input_tokens_seen": 104075455, + "step": 4825, + "time_per_iteration": 2.7736079692840576 + }, + { + "auxiliary_loss_clip": 0.0117986, + "auxiliary_loss_mlp": 0.01027461, + "balance_loss_clip": 1.05320048, + "balance_loss_mlp": 1.01983202, + "epoch": 0.580292190224253, + "flos": 20375966355840.0, + "grad_norm": 2.5447066306018558, + "language_loss": 0.79737484, + "learning_rate": 1.5801645633236644e-06, + "loss": 0.81944805, + "num_input_tokens_seen": 104096440, + "step": 4826, + "time_per_iteration": 2.735874652862549 + }, + { + "auxiliary_loss_clip": 0.01154958, + "auxiliary_loss_mlp": 0.01035042, + "balance_loss_clip": 1.04811835, + "balance_loss_mlp": 1.02653074, + "epoch": 0.5804124331148921, + "flos": 26615373304320.0, + "grad_norm": 1.9394019186796176, + "language_loss": 0.77051806, + "learning_rate": 1.579402979763989e-06, + "loss": 0.79241812, + "num_input_tokens_seen": 104116775, + "step": 4827, + "time_per_iteration": 2.8344502449035645 + }, + { + "auxiliary_loss_clip": 0.01136011, + "auxiliary_loss_mlp": 0.01029093, + "balance_loss_clip": 1.0455153, + "balance_loss_mlp": 1.02124953, + "epoch": 0.5805326760055312, + "flos": 13478496289920.0, + "grad_norm": 2.8665771969646014, + "language_loss": 0.81541443, + "learning_rate": 1.578641460008548e-06, + "loss": 0.83706546, + "num_input_tokens_seen": 104134510, + "step": 4828, + "time_per_iteration": 2.9655508995056152 + }, + { + "auxiliary_loss_clip": 0.01165065, + "auxiliary_loss_mlp": 0.01024745, + "balance_loss_clip": 1.04857016, + "balance_loss_mlp": 1.01728797, + "epoch": 0.5806529188961702, + "flos": 12091374823680.0, + "grad_norm": 2.0079787145520065, + "language_loss": 0.67736208, + "learning_rate": 1.5778800041728613e-06, + "loss": 0.69926012, + "num_input_tokens_seen": 104150800, + "step": 4829, + "time_per_iteration": 2.731978416442871 + }, + { + "auxiliary_loss_clip": 0.01163461, + "auxiliary_loss_mlp": 0.01023773, + "balance_loss_clip": 1.04937077, + "balance_loss_mlp": 1.01614666, + "epoch": 0.5807731617868094, + "flos": 26214107495040.0, + "grad_norm": 1.4294258990878899, + "language_loss": 0.6592344, + "learning_rate": 1.577118612372443e-06, + "loss": 0.68110669, + "num_input_tokens_seen": 104172640, + "step": 4830, + "time_per_iteration": 2.783517360687256 + }, + { + "auxiliary_loss_clip": 0.01155411, + "auxiliary_loss_mlp": 0.00886736, + "balance_loss_clip": 1.04553008, + "balance_loss_mlp": 1.00072968, + "epoch": 0.5808934046774484, + "flos": 37962139190400.0, + "grad_norm": 2.2203587519299286, + "language_loss": 0.70326239, + "learning_rate": 1.5763572847227943e-06, + "loss": 0.72368383, + "num_input_tokens_seen": 104193525, + "step": 4831, + "time_per_iteration": 3.006544828414917 + }, + { + "auxiliary_loss_clip": 0.01166272, + "auxiliary_loss_mlp": 0.01024775, + "balance_loss_clip": 1.04731131, + "balance_loss_mlp": 1.01653504, + "epoch": 0.5810136475680875, + "flos": 20485853038080.0, + "grad_norm": 1.7018471469886038, + "language_loss": 0.81289893, + "learning_rate": 1.5755960213394091e-06, + "loss": 0.83480942, + "num_input_tokens_seen": 104210625, + "step": 4832, + "time_per_iteration": 2.718691110610962 + }, + { + "auxiliary_loss_clip": 0.01147944, + "auxiliary_loss_mlp": 0.01027954, + "balance_loss_clip": 1.04750013, + "balance_loss_mlp": 1.02002025, + "epoch": 0.5811338904587267, + "flos": 17530153574400.0, + "grad_norm": 2.1404544039955136, + "language_loss": 0.77966899, + "learning_rate": 1.5748348223377703e-06, + "loss": 0.80142796, + "num_input_tokens_seen": 104228180, + "step": 4833, + "time_per_iteration": 2.7898073196411133 + }, + { + "auxiliary_loss_clip": 0.01157122, + "auxiliary_loss_mlp": 0.01033586, + "balance_loss_clip": 1.04962289, + "balance_loss_mlp": 1.02539587, + "epoch": 0.5812541333493657, + "flos": 19458017360640.0, + "grad_norm": 1.9270514295382062, + "language_loss": 0.77705944, + "learning_rate": 1.5740736878333507e-06, + "loss": 0.79896647, + "num_input_tokens_seen": 104246020, + "step": 4834, + "time_per_iteration": 2.7754480838775635 + }, + { + "auxiliary_loss_clip": 0.01161987, + "auxiliary_loss_mlp": 0.01024677, + "balance_loss_clip": 1.0475961, + "balance_loss_mlp": 1.01625466, + "epoch": 0.5813743762400048, + "flos": 20594949621120.0, + "grad_norm": 2.1728493827879154, + "language_loss": 0.77629232, + "learning_rate": 1.5733126179416143e-06, + "loss": 0.79815894, + "num_input_tokens_seen": 104260505, + "step": 4835, + "time_per_iteration": 2.7641727924346924 + }, + { + "auxiliary_loss_clip": 0.01166748, + "auxiliary_loss_mlp": 0.01028487, + "balance_loss_clip": 1.04938579, + "balance_loss_mlp": 1.0202738, + "epoch": 0.5814946191306439, + "flos": 33178227246720.0, + "grad_norm": 1.8850570407487561, + "language_loss": 0.72390366, + "learning_rate": 1.5725516127780137e-06, + "loss": 0.74585605, + "num_input_tokens_seen": 104282640, + "step": 4836, + "time_per_iteration": 2.797513723373413 + }, + { + "auxiliary_loss_clip": 0.01173034, + "auxiliary_loss_mlp": 0.01028565, + "balance_loss_clip": 1.04791403, + "balance_loss_mlp": 1.02018785, + "epoch": 0.581614862021283, + "flos": 16143283503360.0, + "grad_norm": 2.2794825543182107, + "language_loss": 0.88724786, + "learning_rate": 1.5717906724579943e-06, + "loss": 0.90926391, + "num_input_tokens_seen": 104299700, + "step": 4837, + "time_per_iteration": 2.6539671421051025 + }, + { + "auxiliary_loss_clip": 0.01158711, + "auxiliary_loss_mlp": 0.01028733, + "balance_loss_clip": 1.0485487, + "balance_loss_mlp": 1.02048063, + "epoch": 0.581735104911922, + "flos": 33802642298880.0, + "grad_norm": 2.2633105052032896, + "language_loss": 0.67993039, + "learning_rate": 1.571029797096989e-06, + "loss": 0.70180482, + "num_input_tokens_seen": 104320805, + "step": 4838, + "time_per_iteration": 2.9044477939605713 + }, + { + "auxiliary_loss_clip": 0.01175611, + "auxiliary_loss_mlp": 0.01027836, + "balance_loss_clip": 1.04940271, + "balance_loss_mlp": 1.02025461, + "epoch": 0.5818553478025612, + "flos": 23331163029120.0, + "grad_norm": 1.8890431385843738, + "language_loss": 0.78624856, + "learning_rate": 1.570268986810423e-06, + "loss": 0.80828309, + "num_input_tokens_seen": 104340700, + "step": 4839, + "time_per_iteration": 2.652283191680908 + }, + { + "auxiliary_loss_clip": 0.01154724, + "auxiliary_loss_mlp": 0.01029938, + "balance_loss_clip": 1.04665661, + "balance_loss_mlp": 1.0217185, + "epoch": 0.5819755906932003, + "flos": 20996143603200.0, + "grad_norm": 1.9782342307855527, + "language_loss": 0.75141525, + "learning_rate": 1.5695082417137096e-06, + "loss": 0.7732619, + "num_input_tokens_seen": 104358575, + "step": 4840, + "time_per_iteration": 2.783970832824707 + }, + { + "auxiliary_loss_clip": 0.01155812, + "auxiliary_loss_mlp": 0.0103849, + "balance_loss_clip": 1.04485774, + "balance_loss_mlp": 1.03064609, + "epoch": 0.5820958335838393, + "flos": 21431668008960.0, + "grad_norm": 1.6208498208720552, + "language_loss": 0.75254714, + "learning_rate": 1.5687475619222539e-06, + "loss": 0.77449012, + "num_input_tokens_seen": 104378530, + "step": 4841, + "time_per_iteration": 2.798460006713867 + }, + { + "auxiliary_loss_clip": 0.01154306, + "auxiliary_loss_mlp": 0.01026209, + "balance_loss_clip": 1.04628229, + "balance_loss_mlp": 1.0172739, + "epoch": 0.5822160764744785, + "flos": 17967473660160.0, + "grad_norm": 2.8030529944545433, + "language_loss": 0.73362279, + "learning_rate": 1.5679869475514496e-06, + "loss": 0.75542796, + "num_input_tokens_seen": 104395465, + "step": 4842, + "time_per_iteration": 2.7845113277435303 + }, + { + "auxiliary_loss_clip": 0.01169581, + "auxiliary_loss_mlp": 0.01028504, + "balance_loss_clip": 1.04940176, + "balance_loss_mlp": 1.01959872, + "epoch": 0.5823363193651175, + "flos": 23033858158080.0, + "grad_norm": 2.088701935025635, + "language_loss": 0.80935884, + "learning_rate": 1.567226398716682e-06, + "loss": 0.83133966, + "num_input_tokens_seen": 104415380, + "step": 4843, + "time_per_iteration": 3.5697851181030273 + }, + { + "auxiliary_loss_clip": 0.0116828, + "auxiliary_loss_mlp": 0.01030428, + "balance_loss_clip": 1.05072951, + "balance_loss_mlp": 1.02183867, + "epoch": 0.5824565622557566, + "flos": 32891840110080.0, + "grad_norm": 2.01185740387372, + "language_loss": 0.61847639, + "learning_rate": 1.566465915533326e-06, + "loss": 0.64046341, + "num_input_tokens_seen": 104437410, + "step": 4844, + "time_per_iteration": 3.8788437843322754 + }, + { + "auxiliary_loss_clip": 0.01164051, + "auxiliary_loss_mlp": 0.01026945, + "balance_loss_clip": 1.04797697, + "balance_loss_mlp": 1.01890385, + "epoch": 0.5825768051463958, + "flos": 22229674513920.0, + "grad_norm": 2.5465328036643307, + "language_loss": 0.88128281, + "learning_rate": 1.5657054981167458e-06, + "loss": 0.90319276, + "num_input_tokens_seen": 104456305, + "step": 4845, + "time_per_iteration": 2.7625699043273926 + }, + { + "auxiliary_loss_clip": 0.01164524, + "auxiliary_loss_mlp": 0.01029581, + "balance_loss_clip": 1.04854238, + "balance_loss_mlp": 1.02193928, + "epoch": 0.5826970480370348, + "flos": 28001561016960.0, + "grad_norm": 1.8738219429574619, + "language_loss": 0.6781106, + "learning_rate": 1.5649451465822965e-06, + "loss": 0.70005161, + "num_input_tokens_seen": 104477695, + "step": 4846, + "time_per_iteration": 3.7102222442626953 + }, + { + "auxiliary_loss_clip": 0.0113972, + "auxiliary_loss_mlp": 0.010314, + "balance_loss_clip": 1.04811978, + "balance_loss_mlp": 1.02372551, + "epoch": 0.5828172909276739, + "flos": 17858053854720.0, + "grad_norm": 2.9317505384122238, + "language_loss": 0.83872104, + "learning_rate": 1.5641848610453218e-06, + "loss": 0.86043221, + "num_input_tokens_seen": 104496355, + "step": 4847, + "time_per_iteration": 2.7480902671813965 + }, + { + "auxiliary_loss_clip": 0.01165882, + "auxiliary_loss_mlp": 0.01029472, + "balance_loss_clip": 1.04893148, + "balance_loss_mlp": 1.02085948, + "epoch": 0.582937533818313, + "flos": 19865244827520.0, + "grad_norm": 2.885776523852463, + "language_loss": 0.86229354, + "learning_rate": 1.563424641621158e-06, + "loss": 0.88424706, + "num_input_tokens_seen": 104515535, + "step": 4848, + "time_per_iteration": 3.6232826709747314 + }, + { + "auxiliary_loss_clip": 0.01164454, + "auxiliary_loss_mlp": 0.01032662, + "balance_loss_clip": 1.04908514, + "balance_loss_mlp": 1.02409077, + "epoch": 0.5830577767089521, + "flos": 26870734068480.0, + "grad_norm": 2.5104161806167355, + "language_loss": 0.70028502, + "learning_rate": 1.5626644884251282e-06, + "loss": 0.72225618, + "num_input_tokens_seen": 104535055, + "step": 4849, + "time_per_iteration": 2.766284465789795 + }, + { + "auxiliary_loss_clip": 0.01174733, + "auxiliary_loss_mlp": 0.01026931, + "balance_loss_clip": 1.04847491, + "balance_loss_mlp": 1.01853251, + "epoch": 0.5831780195995911, + "flos": 25298205575040.0, + "grad_norm": 1.5777424480681395, + "language_loss": 0.87866414, + "learning_rate": 1.5619044015725488e-06, + "loss": 0.90068078, + "num_input_tokens_seen": 104554745, + "step": 4850, + "time_per_iteration": 2.640515089035034 + }, + { + "auxiliary_loss_clip": 0.01188001, + "auxiliary_loss_mlp": 0.01032657, + "balance_loss_clip": 1.05706334, + "balance_loss_mlp": 1.02295303, + "epoch": 0.5832982624902303, + "flos": 14756988049920.0, + "grad_norm": 2.8750473882471024, + "language_loss": 0.87074709, + "learning_rate": 1.5611443811787224e-06, + "loss": 0.89295363, + "num_input_tokens_seen": 104568870, + "step": 4851, + "time_per_iteration": 2.580470323562622 + }, + { + "auxiliary_loss_clip": 0.01163022, + "auxiliary_loss_mlp": 0.01032897, + "balance_loss_clip": 1.04878604, + "balance_loss_mlp": 1.02461803, + "epoch": 0.5834185053808694, + "flos": 20444555376000.0, + "grad_norm": 2.15202153386241, + "language_loss": 0.69331586, + "learning_rate": 1.560384427358945e-06, + "loss": 0.71527505, + "num_input_tokens_seen": 104588415, + "step": 4852, + "time_per_iteration": 2.6736598014831543 + }, + { + "auxiliary_loss_clip": 0.01151177, + "auxiliary_loss_mlp": 0.01026457, + "balance_loss_clip": 1.04291272, + "balance_loss_mlp": 1.01825523, + "epoch": 0.5835387482715084, + "flos": 27200394115200.0, + "grad_norm": 1.4257287191867754, + "language_loss": 0.73246729, + "learning_rate": 1.5596245402284998e-06, + "loss": 0.75424361, + "num_input_tokens_seen": 104611940, + "step": 4853, + "time_per_iteration": 2.8231253623962402 + }, + { + "auxiliary_loss_clip": 0.01169962, + "auxiliary_loss_mlp": 0.01035374, + "balance_loss_clip": 1.05145419, + "balance_loss_mlp": 1.02730918, + "epoch": 0.5836589911621476, + "flos": 16654615562880.0, + "grad_norm": 1.6652449712338837, + "language_loss": 0.81791145, + "learning_rate": 1.5588647199026619e-06, + "loss": 0.83996481, + "num_input_tokens_seen": 104629675, + "step": 4854, + "time_per_iteration": 2.7378439903259277 + }, + { + "auxiliary_loss_clip": 0.01181325, + "auxiliary_loss_mlp": 0.01031775, + "balance_loss_clip": 1.05275476, + "balance_loss_mlp": 1.02325189, + "epoch": 0.5837792340527866, + "flos": 20446817932800.0, + "grad_norm": 2.1726497706970833, + "language_loss": 0.87703252, + "learning_rate": 1.5581049664966956e-06, + "loss": 0.89916354, + "num_input_tokens_seen": 104647435, + "step": 4855, + "time_per_iteration": 2.6590158939361572 + }, + { + "auxiliary_loss_clip": 0.01041486, + "auxiliary_loss_mlp": 0.01001217, + "balance_loss_clip": 1.01305699, + "balance_loss_mlp": 0.99930972, + "epoch": 0.5838994769434257, + "flos": 65995480765440.0, + "grad_norm": 0.9929549929383876, + "language_loss": 0.6511758, + "learning_rate": 1.5573452801258545e-06, + "loss": 0.67160279, + "num_input_tokens_seen": 104694605, + "step": 4856, + "time_per_iteration": 3.1918528079986572 + }, + { + "auxiliary_loss_clip": 0.01174348, + "auxiliary_loss_mlp": 0.0103432, + "balance_loss_clip": 1.05131602, + "balance_loss_mlp": 1.02449107, + "epoch": 0.5840197198340649, + "flos": 21470523546240.0, + "grad_norm": 2.1530788683746103, + "language_loss": 0.6335125, + "learning_rate": 1.5565856609053824e-06, + "loss": 0.65559918, + "num_input_tokens_seen": 104713400, + "step": 4857, + "time_per_iteration": 2.7385122776031494 + }, + { + "auxiliary_loss_clip": 0.01179314, + "auxiliary_loss_mlp": 0.01030151, + "balance_loss_clip": 1.05134082, + "balance_loss_mlp": 1.02122235, + "epoch": 0.5841399627247039, + "flos": 19135144984320.0, + "grad_norm": 1.9369741926993729, + "language_loss": 0.80141622, + "learning_rate": 1.5558261089505127e-06, + "loss": 0.82351089, + "num_input_tokens_seen": 104732130, + "step": 4858, + "time_per_iteration": 2.682130813598633 + }, + { + "auxiliary_loss_clip": 0.01169291, + "auxiliary_loss_mlp": 0.01032063, + "balance_loss_clip": 1.05093932, + "balance_loss_mlp": 1.02371264, + "epoch": 0.584260205615343, + "flos": 26425692558720.0, + "grad_norm": 2.4769910469855354, + "language_loss": 0.79870522, + "learning_rate": 1.5550666243764697e-06, + "loss": 0.82071877, + "num_input_tokens_seen": 104750290, + "step": 4859, + "time_per_iteration": 2.7380053997039795 + }, + { + "auxiliary_loss_clip": 0.01167524, + "auxiliary_loss_mlp": 0.01037214, + "balance_loss_clip": 1.04951251, + "balance_loss_mlp": 1.0284431, + "epoch": 0.584380448505982, + "flos": 13881809174400.0, + "grad_norm": 2.329460232902972, + "language_loss": 0.77821076, + "learning_rate": 1.554307207298465e-06, + "loss": 0.80025816, + "num_input_tokens_seen": 104768550, + "step": 4860, + "time_per_iteration": 2.671161651611328 + }, + { + "auxiliary_loss_clip": 0.01179626, + "auxiliary_loss_mlp": 0.01034372, + "balance_loss_clip": 1.05113423, + "balance_loss_mlp": 1.02589595, + "epoch": 0.5845006913966212, + "flos": 21543709507200.0, + "grad_norm": 2.315043125956106, + "language_loss": 0.7895534, + "learning_rate": 1.553547857831704e-06, + "loss": 0.81169343, + "num_input_tokens_seen": 104785060, + "step": 4861, + "time_per_iteration": 2.6274309158325195 + }, + { + "auxiliary_loss_clip": 0.01067999, + "auxiliary_loss_mlp": 0.01003643, + "balance_loss_clip": 1.01257062, + "balance_loss_mlp": 1.00205731, + "epoch": 0.5846209342872603, + "flos": 58375452712320.0, + "grad_norm": 0.8929027245448268, + "language_loss": 0.64231491, + "learning_rate": 1.5527885760913771e-06, + "loss": 0.66303134, + "num_input_tokens_seen": 104834950, + "step": 4862, + "time_per_iteration": 3.04510235786438 + }, + { + "auxiliary_loss_clip": 0.01158676, + "auxiliary_loss_mlp": 0.01025334, + "balance_loss_clip": 1.05022752, + "balance_loss_mlp": 1.01843786, + "epoch": 0.5847411771778993, + "flos": 18588045957120.0, + "grad_norm": 1.9016110938874047, + "language_loss": 0.76708853, + "learning_rate": 1.552029362192668e-06, + "loss": 0.78892863, + "num_input_tokens_seen": 104854210, + "step": 4863, + "time_per_iteration": 2.7040629386901855 + }, + { + "auxiliary_loss_clip": 0.01142835, + "auxiliary_loss_mlp": 0.01035417, + "balance_loss_clip": 1.0447644, + "balance_loss_mlp": 1.02674437, + "epoch": 0.5848614200685385, + "flos": 24240780069120.0, + "grad_norm": 2.068602725361692, + "language_loss": 0.72287327, + "learning_rate": 1.5512702162507478e-06, + "loss": 0.74465579, + "num_input_tokens_seen": 104874525, + "step": 4864, + "time_per_iteration": 2.815640926361084 + }, + { + "auxiliary_loss_clip": 0.01057798, + "auxiliary_loss_mlp": 0.01002915, + "balance_loss_clip": 1.01304901, + "balance_loss_mlp": 1.0013175, + "epoch": 0.5849816629591775, + "flos": 71660245933440.0, + "grad_norm": 1.1556048318168985, + "language_loss": 0.55857444, + "learning_rate": 1.5505111383807792e-06, + "loss": 0.57918155, + "num_input_tokens_seen": 104937195, + "step": 4865, + "time_per_iteration": 3.350529670715332 + }, + { + "auxiliary_loss_clip": 0.01140343, + "auxiliary_loss_mlp": 0.01023986, + "balance_loss_clip": 1.04334712, + "balance_loss_mlp": 1.01616001, + "epoch": 0.5851019058498166, + "flos": 23802095266560.0, + "grad_norm": 1.7753057408190012, + "language_loss": 0.80501854, + "learning_rate": 1.5497521286979138e-06, + "loss": 0.82666183, + "num_input_tokens_seen": 104957435, + "step": 4866, + "time_per_iteration": 2.837160348892212 + }, + { + "auxiliary_loss_clip": 0.01157198, + "auxiliary_loss_mlp": 0.0103054, + "balance_loss_clip": 1.04714537, + "balance_loss_mlp": 1.02119446, + "epoch": 0.5852221487404557, + "flos": 24388516707840.0, + "grad_norm": 1.864095411000157, + "language_loss": 0.74376839, + "learning_rate": 1.5489931873172927e-06, + "loss": 0.76564574, + "num_input_tokens_seen": 104978755, + "step": 4867, + "time_per_iteration": 2.918926477432251 + }, + { + "auxiliary_loss_clip": 0.0111355, + "auxiliary_loss_mlp": 0.01028259, + "balance_loss_clip": 1.03836536, + "balance_loss_mlp": 1.01974094, + "epoch": 0.5853423916310948, + "flos": 27271425260160.0, + "grad_norm": 1.6497798384038174, + "language_loss": 0.79065311, + "learning_rate": 1.5482343143540467e-06, + "loss": 0.81207114, + "num_input_tokens_seen": 105000020, + "step": 4868, + "time_per_iteration": 2.901113510131836 + }, + { + "auxiliary_loss_clip": 0.01145795, + "auxiliary_loss_mlp": 0.00885961, + "balance_loss_clip": 1.04639995, + "balance_loss_mlp": 1.00074601, + "epoch": 0.5854626345217339, + "flos": 11983786611840.0, + "grad_norm": 1.8613434256059718, + "language_loss": 0.83104885, + "learning_rate": 1.547475509923295e-06, + "loss": 0.8513664, + "num_input_tokens_seen": 105017060, + "step": 4869, + "time_per_iteration": 3.7119901180267334 + }, + { + "auxiliary_loss_clip": 0.01042604, + "auxiliary_loss_mlp": 0.01002796, + "balance_loss_clip": 1.0140543, + "balance_loss_mlp": 1.00094795, + "epoch": 0.585582877412373, + "flos": 64342335173760.0, + "grad_norm": 0.7304473710038878, + "language_loss": 0.56012082, + "learning_rate": 1.5467167741401495e-06, + "loss": 0.58057481, + "num_input_tokens_seen": 105078540, + "step": 4870, + "time_per_iteration": 3.3628201484680176 + }, + { + "auxiliary_loss_clip": 0.01156621, + "auxiliary_loss_mlp": 0.01025907, + "balance_loss_clip": 1.04526353, + "balance_loss_mlp": 1.01781905, + "epoch": 0.5857031203030121, + "flos": 17011926103680.0, + "grad_norm": 2.806904978805432, + "language_loss": 0.71485662, + "learning_rate": 1.5459581071197083e-06, + "loss": 0.73668194, + "num_input_tokens_seen": 105094200, + "step": 4871, + "time_per_iteration": 3.718803644180298 + }, + { + "auxiliary_loss_clip": 0.01173505, + "auxiliary_loss_mlp": 0.01031487, + "balance_loss_clip": 1.05334008, + "balance_loss_mlp": 1.0231477, + "epoch": 0.5858233631936511, + "flos": 20885682303360.0, + "grad_norm": 2.374995259745343, + "language_loss": 0.8343007, + "learning_rate": 1.5451995089770624e-06, + "loss": 0.8563506, + "num_input_tokens_seen": 105113985, + "step": 4872, + "time_per_iteration": 3.7668533325195312 + }, + { + "auxiliary_loss_clip": 0.01175197, + "auxiliary_loss_mlp": 0.01030069, + "balance_loss_clip": 1.05002952, + "balance_loss_mlp": 1.02187347, + "epoch": 0.5859436060842903, + "flos": 23191902000000.0, + "grad_norm": 1.3013009757129512, + "language_loss": 0.71955919, + "learning_rate": 1.5444409798272885e-06, + "loss": 0.74161184, + "num_input_tokens_seen": 105138075, + "step": 4873, + "time_per_iteration": 2.7908787727355957 + }, + { + "auxiliary_loss_clip": 0.0115268, + "auxiliary_loss_mlp": 0.0103254, + "balance_loss_clip": 1.04682517, + "balance_loss_mlp": 1.02426744, + "epoch": 0.5860638489749294, + "flos": 22492648961280.0, + "grad_norm": 2.2966895293986194, + "language_loss": 0.80363321, + "learning_rate": 1.543682519785456e-06, + "loss": 0.82548547, + "num_input_tokens_seen": 105156555, + "step": 4874, + "time_per_iteration": 3.8106348514556885 + }, + { + "auxiliary_loss_clip": 0.01157146, + "auxiliary_loss_mlp": 0.0102664, + "balance_loss_clip": 1.04576683, + "balance_loss_mlp": 1.01837301, + "epoch": 0.5861840918655684, + "flos": 17566243764480.0, + "grad_norm": 2.6418315212691486, + "language_loss": 0.80145639, + "learning_rate": 1.5429241289666219e-06, + "loss": 0.82329428, + "num_input_tokens_seen": 105174055, + "step": 4875, + "time_per_iteration": 2.7555551528930664 + }, + { + "auxiliary_loss_clip": 0.01149906, + "auxiliary_loss_mlp": 0.01031776, + "balance_loss_clip": 1.04576683, + "balance_loss_mlp": 1.02341962, + "epoch": 0.5863043347562076, + "flos": 25556152118400.0, + "grad_norm": 1.8596225188648345, + "language_loss": 0.69882679, + "learning_rate": 1.5421658074858342e-06, + "loss": 0.72064352, + "num_input_tokens_seen": 105192160, + "step": 4876, + "time_per_iteration": 2.8581488132476807 + }, + { + "auxiliary_loss_clip": 0.01157038, + "auxiliary_loss_mlp": 0.01032166, + "balance_loss_clip": 1.04876077, + "balance_loss_mlp": 1.0235498, + "epoch": 0.5864245776468466, + "flos": 20667525050880.0, + "grad_norm": 2.213549472851987, + "language_loss": 0.66390425, + "learning_rate": 1.5414075554581298e-06, + "loss": 0.68579632, + "num_input_tokens_seen": 105210205, + "step": 4877, + "time_per_iteration": 2.720141887664795 + }, + { + "auxiliary_loss_clip": 0.01181332, + "auxiliary_loss_mlp": 0.01031088, + "balance_loss_clip": 1.05287385, + "balance_loss_mlp": 1.02283895, + "epoch": 0.5865448205374857, + "flos": 28913907490560.0, + "grad_norm": 2.1831320911184604, + "language_loss": 0.78481829, + "learning_rate": 1.5406493729985348e-06, + "loss": 0.80694246, + "num_input_tokens_seen": 105229400, + "step": 4878, + "time_per_iteration": 2.7124948501586914 + }, + { + "auxiliary_loss_clip": 0.01142518, + "auxiliary_loss_mlp": 0.00886921, + "balance_loss_clip": 1.04988134, + "balance_loss_mlp": 1.00071311, + "epoch": 0.5866650634281249, + "flos": 25842575168640.0, + "grad_norm": 1.9586327295661454, + "language_loss": 0.71847421, + "learning_rate": 1.5398912602220644e-06, + "loss": 0.73876858, + "num_input_tokens_seen": 105248675, + "step": 4879, + "time_per_iteration": 2.8214683532714844 + }, + { + "auxiliary_loss_clip": 0.0114511, + "auxiliary_loss_mlp": 0.01031644, + "balance_loss_clip": 1.04624295, + "balance_loss_mlp": 1.02279925, + "epoch": 0.5867853063187639, + "flos": 17052325925760.0, + "grad_norm": 2.964645541651738, + "language_loss": 0.78548551, + "learning_rate": 1.539133217243724e-06, + "loss": 0.80725312, + "num_input_tokens_seen": 105265695, + "step": 4880, + "time_per_iteration": 2.7775678634643555 + }, + { + "auxiliary_loss_clip": 0.01159928, + "auxiliary_loss_mlp": 0.01029571, + "balance_loss_clip": 1.04959238, + "balance_loss_mlp": 1.02021313, + "epoch": 0.586905549209403, + "flos": 24645026707200.0, + "grad_norm": 1.9398861222412804, + "language_loss": 0.76018131, + "learning_rate": 1.5383752441785081e-06, + "loss": 0.78207624, + "num_input_tokens_seen": 105284920, + "step": 4881, + "time_per_iteration": 2.7837626934051514 + }, + { + "auxiliary_loss_clip": 0.01174168, + "auxiliary_loss_mlp": 0.01034788, + "balance_loss_clip": 1.05070972, + "balance_loss_mlp": 1.02620471, + "epoch": 0.5870257921000421, + "flos": 14720538723840.0, + "grad_norm": 2.222424041037795, + "language_loss": 0.85827106, + "learning_rate": 1.5376173411414003e-06, + "loss": 0.8803606, + "num_input_tokens_seen": 105302960, + "step": 4882, + "time_per_iteration": 2.7166085243225098 + }, + { + "auxiliary_loss_clip": 0.01158694, + "auxiliary_loss_mlp": 0.0103094, + "balance_loss_clip": 1.04556966, + "balance_loss_mlp": 1.02193344, + "epoch": 0.5871460349906812, + "flos": 23914998691200.0, + "grad_norm": 4.756773740843721, + "language_loss": 0.79322743, + "learning_rate": 1.5368595082473753e-06, + "loss": 0.8151238, + "num_input_tokens_seen": 105321260, + "step": 4883, + "time_per_iteration": 2.7313599586486816 + }, + { + "auxiliary_loss_clip": 0.01171332, + "auxiliary_loss_mlp": 0.0103219, + "balance_loss_clip": 1.04900789, + "balance_loss_mlp": 1.02375019, + "epoch": 0.5872662778813202, + "flos": 22164174063360.0, + "grad_norm": 2.252572987886227, + "language_loss": 0.78126788, + "learning_rate": 1.5361017456113935e-06, + "loss": 0.80330312, + "num_input_tokens_seen": 105341610, + "step": 4884, + "time_per_iteration": 2.661574125289917 + }, + { + "auxiliary_loss_clip": 0.01166984, + "auxiliary_loss_mlp": 0.01026921, + "balance_loss_clip": 1.04643738, + "balance_loss_mlp": 1.01792669, + "epoch": 0.5873865207719594, + "flos": 18441925430400.0, + "grad_norm": 2.8496409760456847, + "language_loss": 0.86132467, + "learning_rate": 1.5353440533484085e-06, + "loss": 0.88326371, + "num_input_tokens_seen": 105360465, + "step": 4885, + "time_per_iteration": 2.6656694412231445 + }, + { + "auxiliary_loss_clip": 0.01161672, + "auxiliary_loss_mlp": 0.01033523, + "balance_loss_clip": 1.04944634, + "balance_loss_mlp": 1.02497613, + "epoch": 0.5875067636625985, + "flos": 54015321427200.0, + "grad_norm": 2.067881911778048, + "language_loss": 0.66234219, + "learning_rate": 1.534586431573361e-06, + "loss": 0.6842941, + "num_input_tokens_seen": 105385405, + "step": 4886, + "time_per_iteration": 3.0566611289978027 + }, + { + "auxiliary_loss_clip": 0.0113047, + "auxiliary_loss_mlp": 0.01042173, + "balance_loss_clip": 1.04120994, + "balance_loss_mlp": 1.03216529, + "epoch": 0.5876270065532375, + "flos": 27995707100160.0, + "grad_norm": 2.1229798801989497, + "language_loss": 0.79522216, + "learning_rate": 1.5338288804011817e-06, + "loss": 0.81694853, + "num_input_tokens_seen": 105404905, + "step": 4887, + "time_per_iteration": 2.9132137298583984 + }, + { + "auxiliary_loss_clip": 0.01155927, + "auxiliary_loss_mlp": 0.01025761, + "balance_loss_clip": 1.0455296, + "balance_loss_mlp": 1.01761258, + "epoch": 0.5877472494438767, + "flos": 21361462876800.0, + "grad_norm": 2.0300186676965324, + "language_loss": 0.71188748, + "learning_rate": 1.533071399946791e-06, + "loss": 0.73370439, + "num_input_tokens_seen": 105423650, + "step": 4888, + "time_per_iteration": 2.7312252521514893 + }, + { + "auxiliary_loss_clip": 0.01153213, + "auxiliary_loss_mlp": 0.01027976, + "balance_loss_clip": 1.04594922, + "balance_loss_mlp": 1.019768, + "epoch": 0.5878674923345157, + "flos": 22383013674240.0, + "grad_norm": 2.2756508967729197, + "language_loss": 0.57520264, + "learning_rate": 1.5323139903250977e-06, + "loss": 0.59701455, + "num_input_tokens_seen": 105444255, + "step": 4889, + "time_per_iteration": 2.7974069118499756 + }, + { + "auxiliary_loss_clip": 0.01163669, + "auxiliary_loss_mlp": 0.01030345, + "balance_loss_clip": 1.05151534, + "balance_loss_mlp": 1.02163672, + "epoch": 0.5879877352251548, + "flos": 21868664872320.0, + "grad_norm": 1.7928444440083053, + "language_loss": 0.76985508, + "learning_rate": 1.5315566516510002e-06, + "loss": 0.79179525, + "num_input_tokens_seen": 105462425, + "step": 4890, + "time_per_iteration": 2.738682270050049 + }, + { + "auxiliary_loss_clip": 0.01178645, + "auxiliary_loss_mlp": 0.01027623, + "balance_loss_clip": 1.05133879, + "balance_loss_mlp": 1.01903427, + "epoch": 0.5881079781157939, + "flos": 17493811989120.0, + "grad_norm": 1.8082910457730172, + "language_loss": 0.67501116, + "learning_rate": 1.5307993840393857e-06, + "loss": 0.69707382, + "num_input_tokens_seen": 105480505, + "step": 4891, + "time_per_iteration": 2.666654109954834 + }, + { + "auxiliary_loss_clip": 0.01176749, + "auxiliary_loss_mlp": 0.01029635, + "balance_loss_clip": 1.05029464, + "balance_loss_mlp": 1.0218029, + "epoch": 0.588228221006433, + "flos": 22601853285120.0, + "grad_norm": 1.864263497178839, + "language_loss": 0.80145359, + "learning_rate": 1.530042187605132e-06, + "loss": 0.82351744, + "num_input_tokens_seen": 105499760, + "step": 4892, + "time_per_iteration": 2.6665961742401123 + }, + { + "auxiliary_loss_clip": 0.01168298, + "auxiliary_loss_mlp": 0.00885532, + "balance_loss_clip": 1.04858744, + "balance_loss_mlp": 1.00076604, + "epoch": 0.5883484638970721, + "flos": 26176939896960.0, + "grad_norm": 1.3772570179175836, + "language_loss": 0.84028465, + "learning_rate": 1.5292850624631044e-06, + "loss": 0.86082298, + "num_input_tokens_seen": 105521955, + "step": 4893, + "time_per_iteration": 2.786454200744629 + }, + { + "auxiliary_loss_clip": 0.01166542, + "auxiliary_loss_mlp": 0.01035062, + "balance_loss_clip": 1.05017328, + "balance_loss_mlp": 1.02674675, + "epoch": 0.5884687067877111, + "flos": 30443737691520.0, + "grad_norm": 2.066305381402503, + "language_loss": 0.80590546, + "learning_rate": 1.5285280087281593e-06, + "loss": 0.82792145, + "num_input_tokens_seen": 105542685, + "step": 4894, + "time_per_iteration": 2.794867753982544 + }, + { + "auxiliary_loss_clip": 0.01059365, + "auxiliary_loss_mlp": 0.01002255, + "balance_loss_clip": 1.01410508, + "balance_loss_mlp": 1.00078917, + "epoch": 0.5885889496783503, + "flos": 70507550580480.0, + "grad_norm": 0.6453355538771872, + "language_loss": 0.56632328, + "learning_rate": 1.5277710265151398e-06, + "loss": 0.58693945, + "num_input_tokens_seen": 105612165, + "step": 4895, + "time_per_iteration": 4.376126766204834 + }, + { + "auxiliary_loss_clip": 0.01170644, + "auxiliary_loss_mlp": 0.01030196, + "balance_loss_clip": 1.05000424, + "balance_loss_mlp": 1.02153563, + "epoch": 0.5887091925689893, + "flos": 19098767485440.0, + "grad_norm": 2.9616860878392393, + "language_loss": 0.76889348, + "learning_rate": 1.5270141159388803e-06, + "loss": 0.79090184, + "num_input_tokens_seen": 105629185, + "step": 4896, + "time_per_iteration": 2.711862802505493 + }, + { + "auxiliary_loss_clip": 0.01177705, + "auxiliary_loss_mlp": 0.01036222, + "balance_loss_clip": 1.05002701, + "balance_loss_mlp": 1.02778828, + "epoch": 0.5888294354596284, + "flos": 23294282739840.0, + "grad_norm": 1.7128461017927286, + "language_loss": 0.8053118, + "learning_rate": 1.526257277114203e-06, + "loss": 0.82745105, + "num_input_tokens_seen": 105650260, + "step": 4897, + "time_per_iteration": 3.6599085330963135 + }, + { + "auxiliary_loss_clip": 0.01154043, + "auxiliary_loss_mlp": 0.01037227, + "balance_loss_clip": 1.04687989, + "balance_loss_mlp": 1.0292995, + "epoch": 0.5889496783502676, + "flos": 21981532383360.0, + "grad_norm": 1.74051584843916, + "language_loss": 0.79461801, + "learning_rate": 1.5255005101559201e-06, + "loss": 0.8165307, + "num_input_tokens_seen": 105667870, + "step": 4898, + "time_per_iteration": 3.761605739593506 + }, + { + "auxiliary_loss_clip": 0.0116245, + "auxiliary_loss_mlp": 0.01027819, + "balance_loss_clip": 1.04887033, + "balance_loss_mlp": 1.02032638, + "epoch": 0.5890699212409066, + "flos": 21685233093120.0, + "grad_norm": 2.61082684790069, + "language_loss": 0.76847506, + "learning_rate": 1.524743815178833e-06, + "loss": 0.79037774, + "num_input_tokens_seen": 105685830, + "step": 4899, + "time_per_iteration": 2.742591619491577 + }, + { + "auxiliary_loss_clip": 0.01159304, + "auxiliary_loss_mlp": 0.01023674, + "balance_loss_clip": 1.04475594, + "balance_loss_mlp": 1.01613367, + "epoch": 0.5891901641315457, + "flos": 19464553635840.0, + "grad_norm": 2.4671965250750327, + "language_loss": 0.80772394, + "learning_rate": 1.5239871922977315e-06, + "loss": 0.82955372, + "num_input_tokens_seen": 105705745, + "step": 4900, + "time_per_iteration": 3.6792778968811035 + }, + { + "auxiliary_loss_clip": 0.01156957, + "auxiliary_loss_mlp": 0.01029472, + "balance_loss_clip": 1.04625988, + "balance_loss_mlp": 1.02087665, + "epoch": 0.5893104070221848, + "flos": 19609884063360.0, + "grad_norm": 2.0960884776302833, + "language_loss": 0.8950479, + "learning_rate": 1.523230641627394e-06, + "loss": 0.9169122, + "num_input_tokens_seen": 105724730, + "step": 4901, + "time_per_iteration": 2.7713539600372314 + }, + { + "auxiliary_loss_clip": 0.01139347, + "auxiliary_loss_mlp": 0.0102624, + "balance_loss_clip": 1.04002166, + "balance_loss_mlp": 1.01760912, + "epoch": 0.5894306499128239, + "flos": 29060063930880.0, + "grad_norm": 2.0973511010703034, + "language_loss": 0.73120952, + "learning_rate": 1.5224741632825888e-06, + "loss": 0.75286543, + "num_input_tokens_seen": 105744920, + "step": 4902, + "time_per_iteration": 2.8881964683532715 + }, + { + "auxiliary_loss_clip": 0.01182373, + "auxiliary_loss_mlp": 0.01035994, + "balance_loss_clip": 1.05377412, + "balance_loss_mlp": 1.0276618, + "epoch": 0.589550892803463, + "flos": 42298890721920.0, + "grad_norm": 1.6181020013218619, + "language_loss": 0.69104373, + "learning_rate": 1.521717757378074e-06, + "loss": 0.71322739, + "num_input_tokens_seen": 105765465, + "step": 4903, + "time_per_iteration": 2.779378652572632 + }, + { + "auxiliary_loss_clip": 0.01175396, + "auxiliary_loss_mlp": 0.01025947, + "balance_loss_clip": 1.05070925, + "balance_loss_mlp": 1.01747727, + "epoch": 0.5896711356941021, + "flos": 14137062197760.0, + "grad_norm": 2.0740676726806697, + "language_loss": 0.68816531, + "learning_rate": 1.5209614240285943e-06, + "loss": 0.71017873, + "num_input_tokens_seen": 105783120, + "step": 4904, + "time_per_iteration": 2.718224048614502 + }, + { + "auxiliary_loss_clip": 0.01179852, + "auxiliary_loss_mlp": 0.00886699, + "balance_loss_clip": 1.05121374, + "balance_loss_mlp": 1.00073504, + "epoch": 0.5897913785847412, + "flos": 17201355454080.0, + "grad_norm": 1.9504225849119743, + "language_loss": 0.84384203, + "learning_rate": 1.520205163348887e-06, + "loss": 0.86450756, + "num_input_tokens_seen": 105801055, + "step": 4905, + "time_per_iteration": 2.660923957824707 + }, + { + "auxiliary_loss_clip": 0.01056089, + "auxiliary_loss_mlp": 0.0100458, + "balance_loss_clip": 1.01441228, + "balance_loss_mlp": 1.00299454, + "epoch": 0.5899116214753802, + "flos": 48794164202880.0, + "grad_norm": 0.7230678399870712, + "language_loss": 0.57023549, + "learning_rate": 1.519448975453674e-06, + "loss": 0.59084225, + "num_input_tokens_seen": 105856155, + "step": 4906, + "time_per_iteration": 3.2505252361297607 + }, + { + "auxiliary_loss_clip": 0.01172342, + "auxiliary_loss_mlp": 0.00887069, + "balance_loss_clip": 1.05262327, + "balance_loss_mlp": 1.00078499, + "epoch": 0.5900318643660194, + "flos": 21103659987840.0, + "grad_norm": 1.955967218960068, + "language_loss": 0.75784397, + "learning_rate": 1.5186928604576696e-06, + "loss": 0.77843809, + "num_input_tokens_seen": 105873350, + "step": 4907, + "time_per_iteration": 2.7499232292175293 + }, + { + "auxiliary_loss_clip": 0.01162649, + "auxiliary_loss_mlp": 0.01029699, + "balance_loss_clip": 1.04914355, + "balance_loss_mlp": 1.02144325, + "epoch": 0.5901521072566585, + "flos": 21178390233600.0, + "grad_norm": 2.2865078220374837, + "language_loss": 0.76854658, + "learning_rate": 1.5179368184755752e-06, + "loss": 0.79047, + "num_input_tokens_seen": 105891435, + "step": 4908, + "time_per_iteration": 2.7941198348999023 + }, + { + "auxiliary_loss_clip": 0.01157705, + "auxiliary_loss_mlp": 0.0102658, + "balance_loss_clip": 1.04750669, + "balance_loss_mlp": 1.01838434, + "epoch": 0.5902723501472975, + "flos": 20225967160320.0, + "grad_norm": 1.607436668390969, + "language_loss": 0.82391894, + "learning_rate": 1.5171808496220821e-06, + "loss": 0.84576178, + "num_input_tokens_seen": 105910190, + "step": 4909, + "time_per_iteration": 2.7863361835479736 + }, + { + "auxiliary_loss_clip": 0.01163944, + "auxiliary_loss_mlp": 0.01030394, + "balance_loss_clip": 1.04833233, + "balance_loss_mlp": 1.022264, + "epoch": 0.5903925930379367, + "flos": 22964407211520.0, + "grad_norm": 1.7308859464773279, + "language_loss": 0.81514156, + "learning_rate": 1.5164249540118708e-06, + "loss": 0.83708489, + "num_input_tokens_seen": 105929315, + "step": 4910, + "time_per_iteration": 2.745445728302002 + }, + { + "auxiliary_loss_clip": 0.01135927, + "auxiliary_loss_mlp": 0.01027915, + "balance_loss_clip": 1.04385424, + "balance_loss_mlp": 1.01962709, + "epoch": 0.5905128359285757, + "flos": 23367720096000.0, + "grad_norm": 1.8796329171898725, + "language_loss": 0.83474553, + "learning_rate": 1.5156691317596093e-06, + "loss": 0.85638392, + "num_input_tokens_seen": 105950740, + "step": 4911, + "time_per_iteration": 2.9222376346588135 + }, + { + "auxiliary_loss_clip": 0.01170352, + "auxiliary_loss_mlp": 0.00885435, + "balance_loss_clip": 1.04938006, + "balance_loss_mlp": 1.00067496, + "epoch": 0.5906330788192148, + "flos": 28032335994240.0, + "grad_norm": 2.614814398133663, + "language_loss": 0.66669381, + "learning_rate": 1.5149133829799556e-06, + "loss": 0.68725163, + "num_input_tokens_seen": 105968735, + "step": 4912, + "time_per_iteration": 2.767876625061035 + }, + { + "auxiliary_loss_clip": 0.01161984, + "auxiliary_loss_mlp": 0.01025252, + "balance_loss_clip": 1.04948914, + "balance_loss_mlp": 1.01706266, + "epoch": 0.590753321709854, + "flos": 18477943793280.0, + "grad_norm": 2.203700428320718, + "language_loss": 0.80469543, + "learning_rate": 1.5141577077875556e-06, + "loss": 0.82656777, + "num_input_tokens_seen": 105986060, + "step": 4913, + "time_per_iteration": 2.7037513256073 + }, + { + "auxiliary_loss_clip": 0.01171657, + "auxiliary_loss_mlp": 0.01027883, + "balance_loss_clip": 1.04959369, + "balance_loss_mlp": 1.01991725, + "epoch": 0.590873564600493, + "flos": 16873706568960.0, + "grad_norm": 2.189146149397669, + "language_loss": 0.72400415, + "learning_rate": 1.5134021062970451e-06, + "loss": 0.74599957, + "num_input_tokens_seen": 106004440, + "step": 4914, + "time_per_iteration": 2.6599044799804688 + }, + { + "auxiliary_loss_clip": 0.01138017, + "auxiliary_loss_mlp": 0.01029085, + "balance_loss_clip": 1.04644966, + "balance_loss_mlp": 1.02093697, + "epoch": 0.5909938074911321, + "flos": 13516166678400.0, + "grad_norm": 1.8793331709209038, + "language_loss": 0.80798328, + "learning_rate": 1.5126465786230483e-06, + "loss": 0.82965434, + "num_input_tokens_seen": 106021215, + "step": 4915, + "time_per_iteration": 2.7695209980010986 + }, + { + "auxiliary_loss_clip": 0.01178839, + "auxiliary_loss_mlp": 0.01030307, + "balance_loss_clip": 1.05115438, + "balance_loss_mlp": 1.0218792, + "epoch": 0.5911140503817712, + "flos": 26024067613440.0, + "grad_norm": 2.1358310560376004, + "language_loss": 0.82376403, + "learning_rate": 1.5118911248801787e-06, + "loss": 0.84585547, + "num_input_tokens_seen": 106039225, + "step": 4916, + "time_per_iteration": 2.690715789794922 + }, + { + "auxiliary_loss_clip": 0.01164263, + "auxiliary_loss_mlp": 0.01028575, + "balance_loss_clip": 1.0496316, + "balance_loss_mlp": 1.02054679, + "epoch": 0.5912342932724103, + "flos": 23258731253760.0, + "grad_norm": 3.112775931284902, + "language_loss": 0.7991752, + "learning_rate": 1.5111357451830364e-06, + "loss": 0.82110357, + "num_input_tokens_seen": 106057920, + "step": 4917, + "time_per_iteration": 2.743525505065918 + }, + { + "auxiliary_loss_clip": 0.0116964, + "auxiliary_loss_mlp": 0.01027345, + "balance_loss_clip": 1.04841805, + "balance_loss_mlp": 1.01960266, + "epoch": 0.5913545361630493, + "flos": 19573039687680.0, + "grad_norm": 1.85500818823111, + "language_loss": 0.71087766, + "learning_rate": 1.5103804396462131e-06, + "loss": 0.73284757, + "num_input_tokens_seen": 106077855, + "step": 4918, + "time_per_iteration": 2.7406210899353027 + }, + { + "auxiliary_loss_clip": 0.0117317, + "auxiliary_loss_mlp": 0.01029167, + "balance_loss_clip": 1.04934621, + "balance_loss_mlp": 1.02058959, + "epoch": 0.5914747790536885, + "flos": 26213532877440.0, + "grad_norm": 1.8148091290152761, + "language_loss": 0.79608601, + "learning_rate": 1.5096252083842877e-06, + "loss": 0.81810939, + "num_input_tokens_seen": 106097065, + "step": 4919, + "time_per_iteration": 2.7427945137023926 + }, + { + "auxiliary_loss_clip": 0.01167044, + "auxiliary_loss_mlp": 0.01024798, + "balance_loss_clip": 1.04798579, + "balance_loss_mlp": 1.01611948, + "epoch": 0.5915950219443276, + "flos": 27417545786880.0, + "grad_norm": 1.7889412179904258, + "language_loss": 0.85225809, + "learning_rate": 1.5088700515118285e-06, + "loss": 0.8741765, + "num_input_tokens_seen": 106116385, + "step": 4920, + "time_per_iteration": 2.7787063121795654 + }, + { + "auxiliary_loss_clip": 0.01145452, + "auxiliary_loss_mlp": 0.01023313, + "balance_loss_clip": 1.04782546, + "balance_loss_mlp": 1.0150702, + "epoch": 0.5917152648349666, + "flos": 21907879545600.0, + "grad_norm": 1.5986689099716798, + "language_loss": 0.66611731, + "learning_rate": 1.508114969143392e-06, + "loss": 0.68780494, + "num_input_tokens_seen": 106136370, + "step": 4921, + "time_per_iteration": 3.7718117237091064 + }, + { + "auxiliary_loss_clip": 0.01161331, + "auxiliary_loss_mlp": 0.01026195, + "balance_loss_clip": 1.04657054, + "balance_loss_mlp": 1.01869369, + "epoch": 0.5918355077256057, + "flos": 28109185142400.0, + "grad_norm": 1.6064003678913092, + "language_loss": 0.77378213, + "learning_rate": 1.5073599613935238e-06, + "loss": 0.7956574, + "num_input_tokens_seen": 106158490, + "step": 4922, + "time_per_iteration": 2.7811880111694336 + }, + { + "auxiliary_loss_clip": 0.01158448, + "auxiliary_loss_mlp": 0.01028292, + "balance_loss_clip": 1.04723048, + "balance_loss_mlp": 1.0203234, + "epoch": 0.5919557506162448, + "flos": 28183807647360.0, + "grad_norm": 2.418218786681187, + "language_loss": 0.57224286, + "learning_rate": 1.5066050283767574e-06, + "loss": 0.59411031, + "num_input_tokens_seen": 106179170, + "step": 4923, + "time_per_iteration": 3.795947313308716 + }, + { + "auxiliary_loss_clip": 0.01147581, + "auxiliary_loss_mlp": 0.01033024, + "balance_loss_clip": 1.04762268, + "balance_loss_mlp": 1.02506673, + "epoch": 0.5920759935068839, + "flos": 12094355652480.0, + "grad_norm": 2.5781578566969654, + "language_loss": 0.83088827, + "learning_rate": 1.505850170207616e-06, + "loss": 0.85269433, + "num_input_tokens_seen": 106196035, + "step": 4924, + "time_per_iteration": 3.620255947113037 + }, + { + "auxiliary_loss_clip": 0.01157328, + "auxiliary_loss_mlp": 0.01026873, + "balance_loss_clip": 1.04679084, + "balance_loss_mlp": 1.01846242, + "epoch": 0.592196236397523, + "flos": 29424772673280.0, + "grad_norm": 1.966563549616266, + "language_loss": 0.77661556, + "learning_rate": 1.505095387000611e-06, + "loss": 0.79845762, + "num_input_tokens_seen": 106218335, + "step": 4925, + "time_per_iteration": 2.8361315727233887 + }, + { + "auxiliary_loss_clip": 0.01150756, + "auxiliary_loss_mlp": 0.01032988, + "balance_loss_clip": 1.04812217, + "balance_loss_mlp": 1.02511382, + "epoch": 0.5923164792881621, + "flos": 24384709866240.0, + "grad_norm": 2.1708708397892678, + "language_loss": 0.74378014, + "learning_rate": 1.504340678870242e-06, + "loss": 0.76561755, + "num_input_tokens_seen": 106236550, + "step": 4926, + "time_per_iteration": 3.7451319694519043 + }, + { + "auxiliary_loss_clip": 0.01167129, + "auxiliary_loss_mlp": 0.01022839, + "balance_loss_clip": 1.0496043, + "balance_loss_mlp": 1.01395154, + "epoch": 0.5924367221788012, + "flos": 24024238928640.0, + "grad_norm": 1.9192589927780124, + "language_loss": 0.89765269, + "learning_rate": 1.5035860459309989e-06, + "loss": 0.91955239, + "num_input_tokens_seen": 106254265, + "step": 4927, + "time_per_iteration": 2.763521909713745 + }, + { + "auxiliary_loss_clip": 0.01152916, + "auxiliary_loss_mlp": 0.01032928, + "balance_loss_clip": 1.04575658, + "balance_loss_mlp": 1.02417231, + "epoch": 0.5925569650694402, + "flos": 26870590414080.0, + "grad_norm": 1.7753637006159857, + "language_loss": 0.6338675, + "learning_rate": 1.5028314882973568e-06, + "loss": 0.65572596, + "num_input_tokens_seen": 106274670, + "step": 4928, + "time_per_iteration": 2.770689010620117 + }, + { + "auxiliary_loss_clip": 0.01162075, + "auxiliary_loss_mlp": 0.01031508, + "balance_loss_clip": 1.0510484, + "balance_loss_mlp": 1.0235393, + "epoch": 0.5926772079600794, + "flos": 22302788647680.0, + "grad_norm": 1.7714315144302497, + "language_loss": 0.84527767, + "learning_rate": 1.502077006083783e-06, + "loss": 0.86721361, + "num_input_tokens_seen": 106293330, + "step": 4929, + "time_per_iteration": 2.7252306938171387 + }, + { + "auxiliary_loss_clip": 0.01165453, + "auxiliary_loss_mlp": 0.00886202, + "balance_loss_clip": 1.05117297, + "balance_loss_mlp": 1.00074375, + "epoch": 0.5927974508507184, + "flos": 19865244827520.0, + "grad_norm": 3.1785304850503318, + "language_loss": 0.7661218, + "learning_rate": 1.5013225994047315e-06, + "loss": 0.78663838, + "num_input_tokens_seen": 106310960, + "step": 4930, + "time_per_iteration": 2.6248409748077393 + }, + { + "auxiliary_loss_clip": 0.01172764, + "auxiliary_loss_mlp": 0.00886753, + "balance_loss_clip": 1.05295205, + "balance_loss_mlp": 1.00075257, + "epoch": 0.5929176937413575, + "flos": 15776743167360.0, + "grad_norm": 1.784616881552945, + "language_loss": 0.80777085, + "learning_rate": 1.5005682683746452e-06, + "loss": 0.82836604, + "num_input_tokens_seen": 106329475, + "step": 4931, + "time_per_iteration": 2.7337257862091064 + }, + { + "auxiliary_loss_clip": 0.01169676, + "auxiliary_loss_mlp": 0.01030505, + "balance_loss_clip": 1.05235791, + "balance_loss_mlp": 1.0228219, + "epoch": 0.5930379366319967, + "flos": 17601472028160.0, + "grad_norm": 2.08754328630271, + "language_loss": 0.72544914, + "learning_rate": 1.4998140131079553e-06, + "loss": 0.74745095, + "num_input_tokens_seen": 106345565, + "step": 4932, + "time_per_iteration": 2.5971171855926514 + }, + { + "auxiliary_loss_clip": 0.01128421, + "auxiliary_loss_mlp": 0.00885931, + "balance_loss_clip": 1.04235232, + "balance_loss_mlp": 1.00079632, + "epoch": 0.5931581795226357, + "flos": 17704283731200.0, + "grad_norm": 1.863927496525187, + "language_loss": 0.73511016, + "learning_rate": 1.4990598337190821e-06, + "loss": 0.75525367, + "num_input_tokens_seen": 106361920, + "step": 4933, + "time_per_iteration": 2.815519332885742 + }, + { + "auxiliary_loss_clip": 0.01176464, + "auxiliary_loss_mlp": 0.00886207, + "balance_loss_clip": 1.04933989, + "balance_loss_mlp": 1.00076771, + "epoch": 0.5932784224132748, + "flos": 24280102483200.0, + "grad_norm": 1.6300568774442772, + "language_loss": 0.6739499, + "learning_rate": 1.4983057303224338e-06, + "loss": 0.69457656, + "num_input_tokens_seen": 106381735, + "step": 4934, + "time_per_iteration": 2.6319966316223145 + }, + { + "auxiliary_loss_clip": 0.01137177, + "auxiliary_loss_mlp": 0.01025459, + "balance_loss_clip": 1.0448339, + "balance_loss_mlp": 1.01773977, + "epoch": 0.5933986653039139, + "flos": 22926700909440.0, + "grad_norm": 1.625057411796001, + "language_loss": 0.87803859, + "learning_rate": 1.4975517030324072e-06, + "loss": 0.89966494, + "num_input_tokens_seen": 106399745, + "step": 4935, + "time_per_iteration": 2.807499408721924 + }, + { + "auxiliary_loss_clip": 0.01067054, + "auxiliary_loss_mlp": 0.00876832, + "balance_loss_clip": 1.01213229, + "balance_loss_mlp": 1.00119197, + "epoch": 0.593518908194553, + "flos": 71121730256640.0, + "grad_norm": 0.784588159177682, + "language_loss": 0.61836004, + "learning_rate": 1.4967977519633882e-06, + "loss": 0.63779891, + "num_input_tokens_seen": 106457205, + "step": 4936, + "time_per_iteration": 3.311959981918335 + }, + { + "auxiliary_loss_clip": 0.01148988, + "auxiliary_loss_mlp": 0.01032575, + "balance_loss_clip": 1.04589558, + "balance_loss_mlp": 1.02399826, + "epoch": 0.593639151085192, + "flos": 20448649526400.0, + "grad_norm": 2.2175399625419367, + "language_loss": 0.78028691, + "learning_rate": 1.4960438772297494e-06, + "loss": 0.80210257, + "num_input_tokens_seen": 106474250, + "step": 4937, + "time_per_iteration": 2.7775254249572754 + }, + { + "auxiliary_loss_clip": 0.01162042, + "auxiliary_loss_mlp": 0.01028351, + "balance_loss_clip": 1.04698253, + "balance_loss_mlp": 1.01954198, + "epoch": 0.5937593939758312, + "flos": 30883428074880.0, + "grad_norm": 2.6692656027012855, + "language_loss": 0.73647392, + "learning_rate": 1.495290078945855e-06, + "loss": 0.75837791, + "num_input_tokens_seen": 106494015, + "step": 4938, + "time_per_iteration": 2.871541738510132 + }, + { + "auxiliary_loss_clip": 0.01177356, + "auxiliary_loss_mlp": 0.0103078, + "balance_loss_clip": 1.05177486, + "balance_loss_mlp": 1.02257848, + "epoch": 0.5938796368664703, + "flos": 36898069668480.0, + "grad_norm": 1.889857992004153, + "language_loss": 0.74292123, + "learning_rate": 1.4945363572260529e-06, + "loss": 0.76500261, + "num_input_tokens_seen": 106515010, + "step": 4939, + "time_per_iteration": 2.784959316253662 + }, + { + "auxiliary_loss_clip": 0.01166785, + "auxiliary_loss_mlp": 0.01026629, + "balance_loss_clip": 1.04872251, + "balance_loss_mlp": 1.01872778, + "epoch": 0.5939998797571093, + "flos": 23842926051840.0, + "grad_norm": 4.486671182799105, + "language_loss": 0.6810286, + "learning_rate": 1.4937827121846845e-06, + "loss": 0.70296276, + "num_input_tokens_seen": 106535265, + "step": 4940, + "time_per_iteration": 2.7157368659973145 + }, + { + "auxiliary_loss_clip": 0.01143616, + "auxiliary_loss_mlp": 0.01032018, + "balance_loss_clip": 1.04829144, + "balance_loss_mlp": 1.02391183, + "epoch": 0.5941201226477485, + "flos": 25191407462400.0, + "grad_norm": 2.571336369375912, + "language_loss": 0.73388875, + "learning_rate": 1.4930291439360755e-06, + "loss": 0.75564504, + "num_input_tokens_seen": 106557830, + "step": 4941, + "time_per_iteration": 2.828613519668579 + }, + { + "auxiliary_loss_clip": 0.01168148, + "auxiliary_loss_mlp": 0.01031922, + "balance_loss_clip": 1.04929507, + "balance_loss_mlp": 1.02386975, + "epoch": 0.5942403655383875, + "flos": 22418996123520.0, + "grad_norm": 2.1207399540831835, + "language_loss": 0.79305458, + "learning_rate": 1.4922756525945427e-06, + "loss": 0.81505525, + "num_input_tokens_seen": 106577140, + "step": 4942, + "time_per_iteration": 2.7487003803253174 + }, + { + "auxiliary_loss_clip": 0.0105577, + "auxiliary_loss_mlp": 0.01002674, + "balance_loss_clip": 1.01254058, + "balance_loss_mlp": 1.00118434, + "epoch": 0.5943606084290266, + "flos": 67629310796160.0, + "grad_norm": 0.7758381961300824, + "language_loss": 0.5956791, + "learning_rate": 1.4915222382743894e-06, + "loss": 0.61626357, + "num_input_tokens_seen": 106635975, + "step": 4943, + "time_per_iteration": 3.278794527053833 + }, + { + "auxiliary_loss_clip": 0.01172629, + "auxiliary_loss_mlp": 0.01028603, + "balance_loss_clip": 1.05276608, + "balance_loss_mlp": 1.01985931, + "epoch": 0.5944808513196658, + "flos": 18223157646720.0, + "grad_norm": 2.820091844712987, + "language_loss": 0.71825707, + "learning_rate": 1.4907689010899085e-06, + "loss": 0.74026936, + "num_input_tokens_seen": 106653555, + "step": 4944, + "time_per_iteration": 2.686436653137207 + }, + { + "auxiliary_loss_clip": 0.01157293, + "auxiliary_loss_mlp": 0.01031161, + "balance_loss_clip": 1.04580748, + "balance_loss_mlp": 1.02245247, + "epoch": 0.5946010942103048, + "flos": 24790824011520.0, + "grad_norm": 3.1087769185994865, + "language_loss": 0.62201965, + "learning_rate": 1.4900156411553804e-06, + "loss": 0.64390421, + "num_input_tokens_seen": 106673385, + "step": 4945, + "time_per_iteration": 2.7736427783966064 + }, + { + "auxiliary_loss_clip": 0.01163557, + "auxiliary_loss_mlp": 0.01034337, + "balance_loss_clip": 1.05032218, + "balance_loss_mlp": 1.02627802, + "epoch": 0.5947213371009439, + "flos": 15231619388160.0, + "grad_norm": 2.154048498862568, + "language_loss": 0.85882717, + "learning_rate": 1.4892624585850739e-06, + "loss": 0.88080609, + "num_input_tokens_seen": 106691740, + "step": 4946, + "time_per_iteration": 2.732889175415039 + }, + { + "auxiliary_loss_clip": 0.0118051, + "auxiliary_loss_mlp": 0.01030431, + "balance_loss_clip": 1.05127239, + "balance_loss_mlp": 1.02179456, + "epoch": 0.594841579991583, + "flos": 25848069949440.0, + "grad_norm": 2.092408641184897, + "language_loss": 0.79508227, + "learning_rate": 1.4885093534932465e-06, + "loss": 0.81719172, + "num_input_tokens_seen": 106709705, + "step": 4947, + "time_per_iteration": 3.66465163230896 + }, + { + "auxiliary_loss_clip": 0.01160935, + "auxiliary_loss_mlp": 0.01028528, + "balance_loss_clip": 1.05116093, + "balance_loss_mlp": 1.0201056, + "epoch": 0.5949618228822221, + "flos": 23981109672960.0, + "grad_norm": 2.154117653937077, + "language_loss": 0.71808255, + "learning_rate": 1.4877563259941433e-06, + "loss": 0.73997724, + "num_input_tokens_seen": 106727560, + "step": 4948, + "time_per_iteration": 2.772179365158081 + }, + { + "auxiliary_loss_clip": 0.01178171, + "auxiliary_loss_mlp": 0.01032009, + "balance_loss_clip": 1.05306602, + "balance_loss_mlp": 1.02326512, + "epoch": 0.5950820657728612, + "flos": 40547491476480.0, + "grad_norm": 2.0743576196738194, + "language_loss": 0.67856646, + "learning_rate": 1.4870033762019988e-06, + "loss": 0.70066822, + "num_input_tokens_seen": 106747725, + "step": 4949, + "time_per_iteration": 4.66207480430603 + }, + { + "auxiliary_loss_clip": 0.01158575, + "auxiliary_loss_mlp": 0.01033472, + "balance_loss_clip": 1.04844165, + "balance_loss_mlp": 1.02436388, + "epoch": 0.5952023086635003, + "flos": 23184467884800.0, + "grad_norm": 1.725262486273879, + "language_loss": 0.73356003, + "learning_rate": 1.4862505042310334e-06, + "loss": 0.75548053, + "num_input_tokens_seen": 106767010, + "step": 4950, + "time_per_iteration": 2.7535016536712646 + }, + { + "auxiliary_loss_clip": 0.01153192, + "auxiliary_loss_mlp": 0.01029002, + "balance_loss_clip": 1.04797935, + "balance_loss_mlp": 1.02086592, + "epoch": 0.5953225515541394, + "flos": 33653289548160.0, + "grad_norm": 3.6519145057626234, + "language_loss": 0.69669729, + "learning_rate": 1.4854977101954587e-06, + "loss": 0.71851921, + "num_input_tokens_seen": 106789230, + "step": 4951, + "time_per_iteration": 2.8193700313568115 + }, + { + "auxiliary_loss_clip": 0.01167907, + "auxiliary_loss_mlp": 0.01027258, + "balance_loss_clip": 1.04559231, + "balance_loss_mlp": 1.01833487, + "epoch": 0.5954427944447784, + "flos": 24459619680000.0, + "grad_norm": 1.7702663194662447, + "language_loss": 0.86522853, + "learning_rate": 1.4847449942094716e-06, + "loss": 0.88718021, + "num_input_tokens_seen": 106808110, + "step": 4952, + "time_per_iteration": 3.6293065547943115 + }, + { + "auxiliary_loss_clip": 0.01154188, + "auxiliary_loss_mlp": 0.01029521, + "balance_loss_clip": 1.04732132, + "balance_loss_mlp": 1.02047932, + "epoch": 0.5955630373354175, + "flos": 18551848026240.0, + "grad_norm": 2.0197051438518905, + "language_loss": 0.86469877, + "learning_rate": 1.4839923563872598e-06, + "loss": 0.88653582, + "num_input_tokens_seen": 106826650, + "step": 4953, + "time_per_iteration": 2.7546563148498535 + }, + { + "auxiliary_loss_clip": 0.01150935, + "auxiliary_loss_mlp": 0.01032151, + "balance_loss_clip": 1.04922557, + "balance_loss_mlp": 1.02272129, + "epoch": 0.5956832802260567, + "flos": 19791699730560.0, + "grad_norm": 2.595738590349667, + "language_loss": 0.75701475, + "learning_rate": 1.483239796842997e-06, + "loss": 0.77884561, + "num_input_tokens_seen": 106844680, + "step": 4954, + "time_per_iteration": 2.743673324584961 + }, + { + "auxiliary_loss_clip": 0.01145033, + "auxiliary_loss_mlp": 0.01032994, + "balance_loss_clip": 1.0490675, + "balance_loss_mlp": 1.02476215, + "epoch": 0.5958035231166957, + "flos": 19750868945280.0, + "grad_norm": 2.52637493077255, + "language_loss": 0.83958685, + "learning_rate": 1.4824873156908462e-06, + "loss": 0.86136711, + "num_input_tokens_seen": 106862605, + "step": 4955, + "time_per_iteration": 2.753815174102783 + }, + { + "auxiliary_loss_clip": 0.01172133, + "auxiliary_loss_mlp": 0.00886772, + "balance_loss_clip": 1.05245781, + "balance_loss_mlp": 1.00068867, + "epoch": 0.5959237660073348, + "flos": 21652806090240.0, + "grad_norm": 3.085619164157532, + "language_loss": 0.75599337, + "learning_rate": 1.4817349130449584e-06, + "loss": 0.77658248, + "num_input_tokens_seen": 106882325, + "step": 4956, + "time_per_iteration": 2.680269956588745 + }, + { + "auxiliary_loss_clip": 0.01167288, + "auxiliary_loss_mlp": 0.01029756, + "balance_loss_clip": 1.04982269, + "balance_loss_mlp": 1.02136993, + "epoch": 0.5960440088979739, + "flos": 21171207513600.0, + "grad_norm": 1.8832523917765391, + "language_loss": 0.83156437, + "learning_rate": 1.4809825890194717e-06, + "loss": 0.85353482, + "num_input_tokens_seen": 106900995, + "step": 4957, + "time_per_iteration": 2.68813419342041 + }, + { + "auxiliary_loss_clip": 0.01153316, + "auxiliary_loss_mlp": 0.01024414, + "balance_loss_clip": 1.04540181, + "balance_loss_mlp": 1.01639724, + "epoch": 0.596164251788613, + "flos": 14757526753920.0, + "grad_norm": 1.831618735630348, + "language_loss": 0.7726782, + "learning_rate": 1.4802303437285139e-06, + "loss": 0.79445553, + "num_input_tokens_seen": 106918265, + "step": 4958, + "time_per_iteration": 2.667642831802368 + }, + { + "auxiliary_loss_clip": 0.01155163, + "auxiliary_loss_mlp": 0.01032338, + "balance_loss_clip": 1.04489851, + "balance_loss_mlp": 1.0236243, + "epoch": 0.596284494679252, + "flos": 20485924865280.0, + "grad_norm": 2.1841857773509403, + "language_loss": 0.80322891, + "learning_rate": 1.4794781772861994e-06, + "loss": 0.82510394, + "num_input_tokens_seen": 106934760, + "step": 4959, + "time_per_iteration": 2.771713972091675 + }, + { + "auxiliary_loss_clip": 0.01159545, + "auxiliary_loss_mlp": 0.0088655, + "balance_loss_clip": 1.04947579, + "balance_loss_mlp": 1.00072551, + "epoch": 0.5964047375698912, + "flos": 31212262108800.0, + "grad_norm": 2.0932569800839453, + "language_loss": 0.66608995, + "learning_rate": 1.4787260898066324e-06, + "loss": 0.68655086, + "num_input_tokens_seen": 106954760, + "step": 4960, + "time_per_iteration": 2.8292860984802246 + }, + { + "auxiliary_loss_clip": 0.01176988, + "auxiliary_loss_mlp": 0.01027307, + "balance_loss_clip": 1.05121636, + "balance_loss_mlp": 1.01919174, + "epoch": 0.5965249804605303, + "flos": 27483620855040.0, + "grad_norm": 1.9530106423092273, + "language_loss": 0.8586477, + "learning_rate": 1.4779740814039023e-06, + "loss": 0.88069063, + "num_input_tokens_seen": 106974845, + "step": 4961, + "time_per_iteration": 2.6936073303222656 + }, + { + "auxiliary_loss_clip": 0.01178836, + "auxiliary_loss_mlp": 0.01032063, + "balance_loss_clip": 1.05080581, + "balance_loss_mlp": 1.02365899, + "epoch": 0.5966452233511693, + "flos": 30773936442240.0, + "grad_norm": 1.932867499055471, + "language_loss": 0.68385196, + "learning_rate": 1.4772221521920894e-06, + "loss": 0.70596087, + "num_input_tokens_seen": 106994870, + "step": 4962, + "time_per_iteration": 2.8169641494750977 + }, + { + "auxiliary_loss_clip": 0.01154567, + "auxiliary_loss_mlp": 0.01031129, + "balance_loss_clip": 1.04724312, + "balance_loss_mlp": 1.02271259, + "epoch": 0.5967654662418085, + "flos": 25481170477440.0, + "grad_norm": 1.9461686790906367, + "language_loss": 0.73876327, + "learning_rate": 1.4764703022852598e-06, + "loss": 0.76062024, + "num_input_tokens_seen": 107015390, + "step": 4963, + "time_per_iteration": 2.869504928588867 + }, + { + "auxiliary_loss_clip": 0.01117898, + "auxiliary_loss_mlp": 0.01028724, + "balance_loss_clip": 1.0410881, + "balance_loss_mlp": 1.02068377, + "epoch": 0.5968857091324475, + "flos": 19099126621440.0, + "grad_norm": 3.434602347911533, + "language_loss": 0.76914477, + "learning_rate": 1.4757185317974696e-06, + "loss": 0.79061097, + "num_input_tokens_seen": 107033775, + "step": 4964, + "time_per_iteration": 2.8991663455963135 + }, + { + "auxiliary_loss_clip": 0.01168483, + "auxiliary_loss_mlp": 0.01028242, + "balance_loss_clip": 1.04825902, + "balance_loss_mlp": 1.01894999, + "epoch": 0.5970059520230866, + "flos": 23692711374720.0, + "grad_norm": 2.444812108818101, + "language_loss": 0.70835221, + "learning_rate": 1.474966840842761e-06, + "loss": 0.73031938, + "num_input_tokens_seen": 107053355, + "step": 4965, + "time_per_iteration": 2.682361602783203 + }, + { + "auxiliary_loss_clip": 0.0117118, + "auxiliary_loss_mlp": 0.01025805, + "balance_loss_clip": 1.05017626, + "balance_loss_mlp": 1.01758003, + "epoch": 0.5971261949137258, + "flos": 23185545292800.0, + "grad_norm": 2.159950865085037, + "language_loss": 0.86692882, + "learning_rate": 1.4742152295351655e-06, + "loss": 0.88889861, + "num_input_tokens_seen": 107072510, + "step": 4966, + "time_per_iteration": 2.6622462272644043 + }, + { + "auxiliary_loss_clip": 0.01169401, + "auxiliary_loss_mlp": 0.00886901, + "balance_loss_clip": 1.0491662, + "balance_loss_mlp": 1.00071096, + "epoch": 0.5972464378043648, + "flos": 20557710195840.0, + "grad_norm": 3.031530904347014, + "language_loss": 0.63664651, + "learning_rate": 1.4734636979887016e-06, + "loss": 0.65720952, + "num_input_tokens_seen": 107089970, + "step": 4967, + "time_per_iteration": 2.634227991104126 + }, + { + "auxiliary_loss_clip": 0.01152941, + "auxiliary_loss_mlp": 0.01030353, + "balance_loss_clip": 1.04483032, + "balance_loss_mlp": 1.02182376, + "epoch": 0.5973666806950039, + "flos": 29387030457600.0, + "grad_norm": 1.9848421430452652, + "language_loss": 0.90246791, + "learning_rate": 1.4727122463173755e-06, + "loss": 0.92430079, + "num_input_tokens_seen": 107108500, + "step": 4968, + "time_per_iteration": 2.7825801372528076 + }, + { + "auxiliary_loss_clip": 0.01158134, + "auxiliary_loss_mlp": 0.01032805, + "balance_loss_clip": 1.04924464, + "balance_loss_mlp": 1.02414405, + "epoch": 0.597486923585643, + "flos": 22273522041600.0, + "grad_norm": 1.8455255191251208, + "language_loss": 0.64461589, + "learning_rate": 1.471960874635183e-06, + "loss": 0.66652524, + "num_input_tokens_seen": 107128060, + "step": 4969, + "time_per_iteration": 2.7126924991607666 + }, + { + "auxiliary_loss_clip": 0.01156641, + "auxiliary_loss_mlp": 0.01032254, + "balance_loss_clip": 1.04746604, + "balance_loss_mlp": 1.02355182, + "epoch": 0.5976071664762821, + "flos": 13772461196160.0, + "grad_norm": 2.00535214633105, + "language_loss": 0.71502548, + "learning_rate": 1.4712095830561055e-06, + "loss": 0.7369144, + "num_input_tokens_seen": 107146550, + "step": 4970, + "time_per_iteration": 2.721151828765869 + }, + { + "auxiliary_loss_clip": 0.01155743, + "auxiliary_loss_mlp": 0.01026281, + "balance_loss_clip": 1.0439564, + "balance_loss_mlp": 1.01874125, + "epoch": 0.5977274093669211, + "flos": 19098623831040.0, + "grad_norm": 3.0792997625891725, + "language_loss": 0.8081044, + "learning_rate": 1.4704583716941147e-06, + "loss": 0.82992464, + "num_input_tokens_seen": 107165415, + "step": 4971, + "time_per_iteration": 2.764875888824463 + }, + { + "auxiliary_loss_clip": 0.01161361, + "auxiliary_loss_mlp": 0.01036299, + "balance_loss_clip": 1.0496527, + "balance_loss_mlp": 1.02747798, + "epoch": 0.5978476522575603, + "flos": 20376002269440.0, + "grad_norm": 1.6684357216109016, + "language_loss": 0.72186279, + "learning_rate": 1.4697072406631672e-06, + "loss": 0.74383938, + "num_input_tokens_seen": 107185320, + "step": 4972, + "time_per_iteration": 2.7161924839019775 + }, + { + "auxiliary_loss_clip": 0.01144279, + "auxiliary_loss_mlp": 0.01026624, + "balance_loss_clip": 1.04822993, + "balance_loss_mlp": 1.0181179, + "epoch": 0.5979678951481994, + "flos": 29023147728000.0, + "grad_norm": 1.8826776828931757, + "language_loss": 0.72614318, + "learning_rate": 1.4689561900772097e-06, + "loss": 0.74785221, + "num_input_tokens_seen": 107205380, + "step": 4973, + "time_per_iteration": 3.803173542022705 + }, + { + "auxiliary_loss_clip": 0.0115857, + "auxiliary_loss_mlp": 0.01032979, + "balance_loss_clip": 1.04729009, + "balance_loss_mlp": 1.02430677, + "epoch": 0.5980881380388384, + "flos": 17967689141760.0, + "grad_norm": 2.621569116965104, + "language_loss": 0.72467119, + "learning_rate": 1.4682052200501758e-06, + "loss": 0.74658668, + "num_input_tokens_seen": 107222585, + "step": 4974, + "time_per_iteration": 2.7424442768096924 + }, + { + "auxiliary_loss_clip": 0.01178542, + "auxiliary_loss_mlp": 0.01026436, + "balance_loss_clip": 1.05120337, + "balance_loss_mlp": 1.01855648, + "epoch": 0.5982083809294776, + "flos": 22962827013120.0, + "grad_norm": 1.8190391853730834, + "language_loss": 0.80315948, + "learning_rate": 1.4674543306959876e-06, + "loss": 0.82520926, + "num_input_tokens_seen": 107242055, + "step": 4975, + "time_per_iteration": 3.599440813064575 + }, + { + "auxiliary_loss_clip": 0.01163361, + "auxiliary_loss_mlp": 0.01029651, + "balance_loss_clip": 1.04781032, + "balance_loss_mlp": 1.02089536, + "epoch": 0.5983286238201166, + "flos": 20991941712000.0, + "grad_norm": 2.257323751531007, + "language_loss": 0.840689, + "learning_rate": 1.4667035221285535e-06, + "loss": 0.86261916, + "num_input_tokens_seen": 107259695, + "step": 4976, + "time_per_iteration": 2.713590621948242 + }, + { + "auxiliary_loss_clip": 0.01168468, + "auxiliary_loss_mlp": 0.01031425, + "balance_loss_clip": 1.05136395, + "balance_loss_mlp": 1.02235305, + "epoch": 0.5984488667107557, + "flos": 28183448511360.0, + "grad_norm": 1.7783481361743732, + "language_loss": 0.74253291, + "learning_rate": 1.4659527944617715e-06, + "loss": 0.76453185, + "num_input_tokens_seen": 107279640, + "step": 4977, + "time_per_iteration": 2.745553970336914 + }, + { + "auxiliary_loss_clip": 0.01122417, + "auxiliary_loss_mlp": 0.01032157, + "balance_loss_clip": 1.04136133, + "balance_loss_mlp": 1.02324593, + "epoch": 0.5985691096013949, + "flos": 16471794314880.0, + "grad_norm": 3.1237170577479247, + "language_loss": 0.7594198, + "learning_rate": 1.465202147809526e-06, + "loss": 0.78096557, + "num_input_tokens_seen": 107298135, + "step": 4978, + "time_per_iteration": 3.829785108566284 + }, + { + "auxiliary_loss_clip": 0.01178667, + "auxiliary_loss_mlp": 0.01028589, + "balance_loss_clip": 1.05127454, + "balance_loss_mlp": 1.02023232, + "epoch": 0.5986893524920339, + "flos": 26719046933760.0, + "grad_norm": 2.0626303060121485, + "language_loss": 0.75935763, + "learning_rate": 1.4644515822856888e-06, + "loss": 0.78143024, + "num_input_tokens_seen": 107316570, + "step": 4979, + "time_per_iteration": 2.754520893096924 + }, + { + "auxiliary_loss_clip": 0.0104986, + "auxiliary_loss_mlp": 0.01002346, + "balance_loss_clip": 1.0110997, + "balance_loss_mlp": 1.00077212, + "epoch": 0.598809595382673, + "flos": 61608061100160.0, + "grad_norm": 0.7547755653145467, + "language_loss": 0.56542099, + "learning_rate": 1.4637010980041215e-06, + "loss": 0.5859431, + "num_input_tokens_seen": 107378680, + "step": 4980, + "time_per_iteration": 3.340193033218384 + }, + { + "auxiliary_loss_clip": 0.01178128, + "auxiliary_loss_mlp": 0.0103094, + "balance_loss_clip": 1.05010605, + "balance_loss_mlp": 1.02215457, + "epoch": 0.5989298382733121, + "flos": 11801719549440.0, + "grad_norm": 5.4197342796026895, + "language_loss": 0.89974242, + "learning_rate": 1.4629506950786707e-06, + "loss": 0.92183304, + "num_input_tokens_seen": 107394860, + "step": 4981, + "time_per_iteration": 2.5974369049072266 + }, + { + "auxiliary_loss_clip": 0.01066963, + "auxiliary_loss_mlp": 0.01001037, + "balance_loss_clip": 1.01190138, + "balance_loss_mlp": 0.99954063, + "epoch": 0.5990500811639512, + "flos": 60025800021120.0, + "grad_norm": 0.8133479870410736, + "language_loss": 0.56082338, + "learning_rate": 1.4622003736231733e-06, + "loss": 0.58150339, + "num_input_tokens_seen": 107453850, + "step": 4982, + "time_per_iteration": 3.2494001388549805 + }, + { + "auxiliary_loss_clip": 0.01164996, + "auxiliary_loss_mlp": 0.01028415, + "balance_loss_clip": 1.04917252, + "balance_loss_mlp": 1.02027917, + "epoch": 0.5991703240545903, + "flos": 18222726683520.0, + "grad_norm": 1.8530024834669623, + "language_loss": 0.80735677, + "learning_rate": 1.461450133751451e-06, + "loss": 0.82929087, + "num_input_tokens_seen": 107471920, + "step": 4983, + "time_per_iteration": 2.691535711288452 + }, + { + "auxiliary_loss_clip": 0.01164535, + "auxiliary_loss_mlp": 0.01028494, + "balance_loss_clip": 1.05043912, + "balance_loss_mlp": 1.02049255, + "epoch": 0.5992905669452293, + "flos": 27709894581120.0, + "grad_norm": 2.120466414935818, + "language_loss": 0.75969976, + "learning_rate": 1.4606999755773153e-06, + "loss": 0.78163004, + "num_input_tokens_seen": 107493125, + "step": 4984, + "time_per_iteration": 2.784545660018921 + }, + { + "auxiliary_loss_clip": 0.01177129, + "auxiliary_loss_mlp": 0.0102854, + "balance_loss_clip": 1.05149341, + "balance_loss_mlp": 1.02064276, + "epoch": 0.5994108098358685, + "flos": 20449008662400.0, + "grad_norm": 2.338124055478713, + "language_loss": 0.82028198, + "learning_rate": 1.4599498992145643e-06, + "loss": 0.84233874, + "num_input_tokens_seen": 107513150, + "step": 4985, + "time_per_iteration": 2.863900899887085 + }, + { + "auxiliary_loss_clip": 0.0116458, + "auxiliary_loss_mlp": 0.00886308, + "balance_loss_clip": 1.04906499, + "balance_loss_mlp": 1.00082052, + "epoch": 0.5995310527265075, + "flos": 22269966595200.0, + "grad_norm": 1.894730769629704, + "language_loss": 0.70783687, + "learning_rate": 1.4591999047769846e-06, + "loss": 0.72834575, + "num_input_tokens_seen": 107532005, + "step": 4986, + "time_per_iteration": 2.7812936305999756 + }, + { + "auxiliary_loss_clip": 0.01126598, + "auxiliary_loss_mlp": 0.01029599, + "balance_loss_clip": 1.04193354, + "balance_loss_mlp": 1.02082491, + "epoch": 0.5996512956171466, + "flos": 18916951818240.0, + "grad_norm": 1.9366225872070133, + "language_loss": 0.75287509, + "learning_rate": 1.4584499923783486e-06, + "loss": 0.77443707, + "num_input_tokens_seen": 107550585, + "step": 4987, + "time_per_iteration": 2.839327812194824 + }, + { + "auxiliary_loss_clip": 0.01159693, + "auxiliary_loss_mlp": 0.01029589, + "balance_loss_clip": 1.04882598, + "balance_loss_mlp": 1.02163196, + "epoch": 0.5997715385077858, + "flos": 15370916330880.0, + "grad_norm": 1.691753080909907, + "language_loss": 0.7586391, + "learning_rate": 1.457700162132419e-06, + "loss": 0.78053188, + "num_input_tokens_seen": 107567575, + "step": 4988, + "time_per_iteration": 2.7313146591186523 + }, + { + "auxiliary_loss_clip": 0.01136163, + "auxiliary_loss_mlp": 0.01025037, + "balance_loss_clip": 1.04516566, + "balance_loss_mlp": 1.01729405, + "epoch": 0.5998917813984248, + "flos": 25264844818560.0, + "grad_norm": 2.004725630365285, + "language_loss": 0.72146165, + "learning_rate": 1.4569504141529433e-06, + "loss": 0.74307358, + "num_input_tokens_seen": 107585410, + "step": 4989, + "time_per_iteration": 2.8330063819885254 + }, + { + "auxiliary_loss_clip": 0.0116721, + "auxiliary_loss_mlp": 0.01032992, + "balance_loss_clip": 1.05033278, + "balance_loss_mlp": 1.02418792, + "epoch": 0.6000120242890639, + "flos": 22054502862720.0, + "grad_norm": 2.108130654869793, + "language_loss": 0.71524024, + "learning_rate": 1.456200748553658e-06, + "loss": 0.73724222, + "num_input_tokens_seen": 107603405, + "step": 4990, + "time_per_iteration": 2.6551756858825684 + }, + { + "auxiliary_loss_clip": 0.01179771, + "auxiliary_loss_mlp": 0.01031628, + "balance_loss_clip": 1.05090117, + "balance_loss_mlp": 1.02293754, + "epoch": 0.600132267179703, + "flos": 29863421562240.0, + "grad_norm": 1.5716394300640373, + "language_loss": 0.78518176, + "learning_rate": 1.455451165448287e-06, + "loss": 0.80729574, + "num_input_tokens_seen": 107626060, + "step": 4991, + "time_per_iteration": 2.661510705947876 + }, + { + "auxiliary_loss_clip": 0.01159585, + "auxiliary_loss_mlp": 0.01032383, + "balance_loss_clip": 1.04947233, + "balance_loss_mlp": 1.02331674, + "epoch": 0.6002525100703421, + "flos": 25045358762880.0, + "grad_norm": 2.5012401818674657, + "language_loss": 0.73557246, + "learning_rate": 1.4547016649505407e-06, + "loss": 0.75749213, + "num_input_tokens_seen": 107644070, + "step": 4992, + "time_per_iteration": 2.755850315093994 + }, + { + "auxiliary_loss_clip": 0.0114468, + "auxiliary_loss_mlp": 0.01030228, + "balance_loss_clip": 1.04271019, + "balance_loss_mlp": 1.02165115, + "epoch": 0.6003727529609811, + "flos": 20849592113280.0, + "grad_norm": 2.8809521678006633, + "language_loss": 0.85051376, + "learning_rate": 1.4539522471741193e-06, + "loss": 0.87226284, + "num_input_tokens_seen": 107661495, + "step": 4993, + "time_per_iteration": 297.57828521728516 + }, + { + "auxiliary_loss_clip": 0.011715, + "auxiliary_loss_mlp": 0.01032555, + "balance_loss_clip": 1.04764509, + "balance_loss_mlp": 1.02416885, + "epoch": 0.6004929958516203, + "flos": 15594604277760.0, + "grad_norm": 2.895788819454985, + "language_loss": 0.70640916, + "learning_rate": 1.4532029122327067e-06, + "loss": 0.7284497, + "num_input_tokens_seen": 107678280, + "step": 4994, + "time_per_iteration": 2.8693928718566895 + }, + { + "auxiliary_loss_clip": 0.01141682, + "auxiliary_loss_mlp": 0.01029548, + "balance_loss_clip": 1.04819381, + "balance_loss_mlp": 1.02161455, + "epoch": 0.6006132387422594, + "flos": 21763267390080.0, + "grad_norm": 3.2447258388575184, + "language_loss": 0.75685024, + "learning_rate": 1.4524536602399783e-06, + "loss": 0.77856255, + "num_input_tokens_seen": 107697370, + "step": 4995, + "time_per_iteration": 3.000623941421509 + }, + { + "auxiliary_loss_clip": 0.01158858, + "auxiliary_loss_mlp": 0.01029134, + "balance_loss_clip": 1.05297208, + "balance_loss_mlp": 1.02081597, + "epoch": 0.6007334816328984, + "flos": 22858542852480.0, + "grad_norm": 1.5592853074883175, + "language_loss": 0.7739954, + "learning_rate": 1.4517044913095938e-06, + "loss": 0.79587537, + "num_input_tokens_seen": 107717790, + "step": 4996, + "time_per_iteration": 3.0263140201568604 + }, + { + "auxiliary_loss_clip": 0.01168836, + "auxiliary_loss_mlp": 0.01028705, + "balance_loss_clip": 1.04899991, + "balance_loss_mlp": 1.02030635, + "epoch": 0.6008537245235376, + "flos": 28324577047680.0, + "grad_norm": 1.7239425379077993, + "language_loss": 0.81511331, + "learning_rate": 1.4509554055552022e-06, + "loss": 0.8370887, + "num_input_tokens_seen": 107738020, + "step": 4997, + "time_per_iteration": 2.8718788623809814 + }, + { + "auxiliary_loss_clip": 0.01156445, + "auxiliary_loss_mlp": 0.0102846, + "balance_loss_clip": 1.04833245, + "balance_loss_mlp": 1.0204854, + "epoch": 0.6009739674141766, + "flos": 20886113266560.0, + "grad_norm": 2.686137501197324, + "language_loss": 0.84053987, + "learning_rate": 1.450206403090439e-06, + "loss": 0.86238897, + "num_input_tokens_seen": 107756215, + "step": 4998, + "time_per_iteration": 2.8151004314422607 + }, + { + "auxiliary_loss_clip": 0.01168321, + "auxiliary_loss_mlp": 0.01030403, + "balance_loss_clip": 1.05175936, + "balance_loss_mlp": 1.02211237, + "epoch": 0.6010942103048157, + "flos": 20481004702080.0, + "grad_norm": 2.052827941527831, + "language_loss": 0.85900235, + "learning_rate": 1.4494574840289274e-06, + "loss": 0.88098955, + "num_input_tokens_seen": 107773330, + "step": 4999, + "time_per_iteration": 3.685547113418579 + }, + { + "auxiliary_loss_clip": 0.01175055, + "auxiliary_loss_mlp": 0.01029326, + "balance_loss_clip": 1.05082798, + "balance_loss_mlp": 1.02095163, + "epoch": 0.6012144531954549, + "flos": 23805973935360.0, + "grad_norm": 2.112373960141259, + "language_loss": 0.73740971, + "learning_rate": 1.4487086484842782e-06, + "loss": 0.75945359, + "num_input_tokens_seen": 107791975, + "step": 5000, + "time_per_iteration": 3.711864471435547 + }, + { + "auxiliary_loss_clip": 0.01176333, + "auxiliary_loss_mlp": 0.01030791, + "balance_loss_clip": 1.05015278, + "balance_loss_mlp": 1.02225018, + "epoch": 0.6013346960860939, + "flos": 18988378012800.0, + "grad_norm": 2.244273659198229, + "language_loss": 0.60133141, + "learning_rate": 1.4479598965700878e-06, + "loss": 0.62340266, + "num_input_tokens_seen": 107809240, + "step": 5001, + "time_per_iteration": 3.684809684753418 + }, + { + "auxiliary_loss_clip": 0.01147915, + "auxiliary_loss_mlp": 0.01032284, + "balance_loss_clip": 1.04532611, + "balance_loss_mlp": 1.02455342, + "epoch": 0.601454938976733, + "flos": 24025316336640.0, + "grad_norm": 2.9670195812331133, + "language_loss": 0.69085038, + "learning_rate": 1.4472112283999427e-06, + "loss": 0.71265233, + "num_input_tokens_seen": 107827895, + "step": 5002, + "time_per_iteration": 2.845445394515991 + }, + { + "auxiliary_loss_clip": 0.01163263, + "auxiliary_loss_mlp": 0.01027794, + "balance_loss_clip": 1.04954398, + "balance_loss_mlp": 1.01975286, + "epoch": 0.6015751818673721, + "flos": 26427129102720.0, + "grad_norm": 2.132155999594229, + "language_loss": 0.69155931, + "learning_rate": 1.4464626440874143e-06, + "loss": 0.71346992, + "num_input_tokens_seen": 107847010, + "step": 5003, + "time_per_iteration": 3.6282944679260254 + }, + { + "auxiliary_loss_clip": 0.01139062, + "auxiliary_loss_mlp": 0.01030781, + "balance_loss_clip": 1.04252326, + "balance_loss_mlp": 1.02207541, + "epoch": 0.6016954247580112, + "flos": 13115260005120.0, + "grad_norm": 3.0553163936946155, + "language_loss": 0.74572897, + "learning_rate": 1.4457141437460636e-06, + "loss": 0.76742744, + "num_input_tokens_seen": 107864235, + "step": 5004, + "time_per_iteration": 2.765740394592285 + }, + { + "auxiliary_loss_clip": 0.0115898, + "auxiliary_loss_mlp": 0.01033643, + "balance_loss_clip": 1.04712832, + "balance_loss_mlp": 1.02459478, + "epoch": 0.6018156676486502, + "flos": 23768447201280.0, + "grad_norm": 1.9479102448373151, + "language_loss": 0.73372519, + "learning_rate": 1.444965727489436e-06, + "loss": 0.75565141, + "num_input_tokens_seen": 107883680, + "step": 5005, + "time_per_iteration": 2.7731380462646484 + }, + { + "auxiliary_loss_clip": 0.01147967, + "auxiliary_loss_mlp": 0.0103088, + "balance_loss_clip": 1.04495454, + "balance_loss_mlp": 1.0232923, + "epoch": 0.6019359105392894, + "flos": 26469360518400.0, + "grad_norm": 1.9203916407529844, + "language_loss": 0.63076472, + "learning_rate": 1.444217395431066e-06, + "loss": 0.6525532, + "num_input_tokens_seen": 107906220, + "step": 5006, + "time_per_iteration": 2.9102604389190674 + }, + { + "auxiliary_loss_clip": 0.01049196, + "auxiliary_loss_mlp": 0.01002451, + "balance_loss_clip": 1.01684046, + "balance_loss_mlp": 1.00074637, + "epoch": 0.6020561534299285, + "flos": 69190849728000.0, + "grad_norm": 0.8080532918182334, + "language_loss": 0.55841947, + "learning_rate": 1.4434691476844755e-06, + "loss": 0.57893586, + "num_input_tokens_seen": 107967195, + "step": 5007, + "time_per_iteration": 3.275221347808838 + }, + { + "auxiliary_loss_clip": 0.01156169, + "auxiliary_loss_mlp": 0.01029911, + "balance_loss_clip": 1.0520345, + "balance_loss_mlp": 1.02175152, + "epoch": 0.6021763963205675, + "flos": 21835304115840.0, + "grad_norm": 2.480734595653296, + "language_loss": 0.66925198, + "learning_rate": 1.4427209843631729e-06, + "loss": 0.69111282, + "num_input_tokens_seen": 107984245, + "step": 5008, + "time_per_iteration": 2.8131978511810303 + }, + { + "auxiliary_loss_clip": 0.01175899, + "auxiliary_loss_mlp": 0.00886105, + "balance_loss_clip": 1.05120158, + "balance_loss_mlp": 1.00076652, + "epoch": 0.6022966392112067, + "flos": 26578636669440.0, + "grad_norm": 1.7412071193224332, + "language_loss": 0.80631864, + "learning_rate": 1.4419729055806534e-06, + "loss": 0.82693863, + "num_input_tokens_seen": 108003680, + "step": 5009, + "time_per_iteration": 2.7774317264556885 + }, + { + "auxiliary_loss_clip": 0.01155083, + "auxiliary_loss_mlp": 0.00885578, + "balance_loss_clip": 1.05033088, + "balance_loss_mlp": 1.00072324, + "epoch": 0.6024168821018457, + "flos": 20703722981760.0, + "grad_norm": 1.858984107599091, + "language_loss": 0.82053745, + "learning_rate": 1.441224911450401e-06, + "loss": 0.84094405, + "num_input_tokens_seen": 108019635, + "step": 5010, + "time_per_iteration": 2.8380744457244873 + }, + { + "auxiliary_loss_clip": 0.01172875, + "auxiliary_loss_mlp": 0.01034887, + "balance_loss_clip": 1.05075896, + "balance_loss_mlp": 1.02683461, + "epoch": 0.6025371249924848, + "flos": 24680973242880.0, + "grad_norm": 1.624065278639828, + "language_loss": 0.8244822, + "learning_rate": 1.4404770020858851e-06, + "loss": 0.84655988, + "num_input_tokens_seen": 108039120, + "step": 5011, + "time_per_iteration": 2.7514939308166504 + }, + { + "auxiliary_loss_clip": 0.01162706, + "auxiliary_loss_mlp": 0.01032036, + "balance_loss_clip": 1.04833722, + "balance_loss_mlp": 1.02376914, + "epoch": 0.602657367883124, + "flos": 25955801815680.0, + "grad_norm": 1.6031757685548382, + "language_loss": 0.8595742, + "learning_rate": 1.439729177600563e-06, + "loss": 0.88152164, + "num_input_tokens_seen": 108059615, + "step": 5012, + "time_per_iteration": 2.7401950359344482 + }, + { + "auxiliary_loss_clip": 0.01169174, + "auxiliary_loss_mlp": 0.01031764, + "balance_loss_clip": 1.0517422, + "balance_loss_mlp": 1.02318072, + "epoch": 0.602777610773763, + "flos": 16690633925760.0, + "grad_norm": 5.77628934035109, + "language_loss": 0.73226535, + "learning_rate": 1.4389814381078793e-06, + "loss": 0.75427473, + "num_input_tokens_seen": 108078855, + "step": 5013, + "time_per_iteration": 2.865781784057617 + }, + { + "auxiliary_loss_clip": 0.01101596, + "auxiliary_loss_mlp": 0.01026406, + "balance_loss_clip": 1.04111552, + "balance_loss_mlp": 1.01903868, + "epoch": 0.6028978536644021, + "flos": 13334243270400.0, + "grad_norm": 2.2524836818687346, + "language_loss": 0.79812968, + "learning_rate": 1.438233783721265e-06, + "loss": 0.81940967, + "num_input_tokens_seen": 108095020, + "step": 5014, + "time_per_iteration": 3.0697367191314697 + }, + { + "auxiliary_loss_clip": 0.01153995, + "auxiliary_loss_mlp": 0.01031538, + "balance_loss_clip": 1.05048752, + "balance_loss_mlp": 1.02359259, + "epoch": 0.6030180965550412, + "flos": 19644825018240.0, + "grad_norm": 3.10506277400508, + "language_loss": 0.77725917, + "learning_rate": 1.43748621455414e-06, + "loss": 0.79911447, + "num_input_tokens_seen": 108111455, + "step": 5015, + "time_per_iteration": 2.945124387741089 + }, + { + "auxiliary_loss_clip": 0.011569, + "auxiliary_loss_mlp": 0.01031175, + "balance_loss_clip": 1.04952621, + "balance_loss_mlp": 1.02330136, + "epoch": 0.6031383394456803, + "flos": 14458390289280.0, + "grad_norm": 2.457714726105238, + "language_loss": 0.8065275, + "learning_rate": 1.4367387307199082e-06, + "loss": 0.82840818, + "num_input_tokens_seen": 108128305, + "step": 5016, + "time_per_iteration": 2.698988199234009 + }, + { + "auxiliary_loss_clip": 0.01164469, + "auxiliary_loss_mlp": 0.01022441, + "balance_loss_clip": 1.04840088, + "balance_loss_mlp": 1.01392341, + "epoch": 0.6032585823363193, + "flos": 13917791623680.0, + "grad_norm": 2.221702613531658, + "language_loss": 0.82685167, + "learning_rate": 1.4359913323319632e-06, + "loss": 0.84872079, + "num_input_tokens_seen": 108145475, + "step": 5017, + "time_per_iteration": 2.7681078910827637 + }, + { + "auxiliary_loss_clip": 0.01120303, + "auxiliary_loss_mlp": 0.01030086, + "balance_loss_clip": 1.04077935, + "balance_loss_mlp": 1.02202141, + "epoch": 0.6033788252269584, + "flos": 24353252530560.0, + "grad_norm": 1.6688680654323842, + "language_loss": 0.77400136, + "learning_rate": 1.4352440195036847e-06, + "loss": 0.79550523, + "num_input_tokens_seen": 108165650, + "step": 5018, + "time_per_iteration": 3.2396812438964844 + }, + { + "auxiliary_loss_clip": 0.01117478, + "auxiliary_loss_mlp": 0.01029061, + "balance_loss_clip": 1.03958988, + "balance_loss_mlp": 1.02132416, + "epoch": 0.6034990681175976, + "flos": 25521247077120.0, + "grad_norm": 1.757926981811238, + "language_loss": 0.79944766, + "learning_rate": 1.4344967923484395e-06, + "loss": 0.82091308, + "num_input_tokens_seen": 108187620, + "step": 5019, + "time_per_iteration": 2.959244728088379 + }, + { + "auxiliary_loss_clip": 0.01167254, + "auxiliary_loss_mlp": 0.01032681, + "balance_loss_clip": 1.05029678, + "balance_loss_mlp": 1.02422881, + "epoch": 0.6036193110082366, + "flos": 25958387594880.0, + "grad_norm": 2.33865657952239, + "language_loss": 0.71937197, + "learning_rate": 1.433749650979581e-06, + "loss": 0.74137139, + "num_input_tokens_seen": 108207605, + "step": 5020, + "time_per_iteration": 2.770169258117676 + }, + { + "auxiliary_loss_clip": 0.01143748, + "auxiliary_loss_mlp": 0.01028871, + "balance_loss_clip": 1.04420733, + "balance_loss_mlp": 1.0211761, + "epoch": 0.6037395538988757, + "flos": 25593427457280.0, + "grad_norm": 2.3325115530886156, + "language_loss": 0.67739832, + "learning_rate": 1.433002595510451e-06, + "loss": 0.69912452, + "num_input_tokens_seen": 108226385, + "step": 5021, + "time_per_iteration": 2.824207067489624 + }, + { + "auxiliary_loss_clip": 0.01151601, + "auxiliary_loss_mlp": 0.00886213, + "balance_loss_clip": 1.04479742, + "balance_loss_mlp": 1.00069785, + "epoch": 0.6038597967895148, + "flos": 17816253402240.0, + "grad_norm": 2.185906531624197, + "language_loss": 0.71575713, + "learning_rate": 1.4322556260543757e-06, + "loss": 0.73613524, + "num_input_tokens_seen": 108242960, + "step": 5022, + "time_per_iteration": 2.706825017929077 + }, + { + "auxiliary_loss_clip": 0.0104992, + "auxiliary_loss_mlp": 0.01003417, + "balance_loss_clip": 1.01243424, + "balance_loss_mlp": 1.00177217, + "epoch": 0.6039800396801539, + "flos": 65169213235200.0, + "grad_norm": 0.9136537231860071, + "language_loss": 0.62701666, + "learning_rate": 1.4315087427246703e-06, + "loss": 0.64754999, + "num_input_tokens_seen": 108296785, + "step": 5023, + "time_per_iteration": 3.200143814086914 + }, + { + "auxiliary_loss_clip": 0.01066765, + "auxiliary_loss_mlp": 0.01001877, + "balance_loss_clip": 1.01226997, + "balance_loss_mlp": 1.00041676, + "epoch": 0.604100282570793, + "flos": 67386409073280.0, + "grad_norm": 0.8755879066882521, + "language_loss": 0.58520865, + "learning_rate": 1.4307619456346372e-06, + "loss": 0.60589504, + "num_input_tokens_seen": 108341090, + "step": 5024, + "time_per_iteration": 3.045804977416992 + }, + { + "auxiliary_loss_clip": 0.01170501, + "auxiliary_loss_mlp": 0.01027606, + "balance_loss_clip": 1.04786229, + "balance_loss_mlp": 1.01921403, + "epoch": 0.6042205254614321, + "flos": 35297495631360.0, + "grad_norm": 4.480536975447593, + "language_loss": 0.74017596, + "learning_rate": 1.430015234897564e-06, + "loss": 0.76215702, + "num_input_tokens_seen": 108364370, + "step": 5025, + "time_per_iteration": 3.8001952171325684 + }, + { + "auxiliary_loss_clip": 0.01175888, + "auxiliary_loss_mlp": 0.00886481, + "balance_loss_clip": 1.04905903, + "balance_loss_mlp": 1.00073397, + "epoch": 0.6043407683520712, + "flos": 45658262206080.0, + "grad_norm": 1.514950012728818, + "language_loss": 0.66151297, + "learning_rate": 1.4292686106267274e-06, + "loss": 0.68213665, + "num_input_tokens_seen": 108387220, + "step": 5026, + "time_per_iteration": 4.826996564865112 + }, + { + "auxiliary_loss_clip": 0.01172567, + "auxiliary_loss_mlp": 0.01031009, + "balance_loss_clip": 1.05051529, + "balance_loss_mlp": 1.02225924, + "epoch": 0.6044610112427102, + "flos": 16180020138240.0, + "grad_norm": 1.6149852269169036, + "language_loss": 0.77097583, + "learning_rate": 1.4285220729353876e-06, + "loss": 0.79301155, + "num_input_tokens_seen": 108405760, + "step": 5027, + "time_per_iteration": 2.7132272720336914 + }, + { + "auxiliary_loss_clip": 0.01154482, + "auxiliary_loss_mlp": 0.01031433, + "balance_loss_clip": 1.0435915, + "balance_loss_mlp": 1.02346992, + "epoch": 0.6045812541333494, + "flos": 13804062186240.0, + "grad_norm": 1.843579977645595, + "language_loss": 0.77619451, + "learning_rate": 1.4277756219367957e-06, + "loss": 0.79805374, + "num_input_tokens_seen": 108422785, + "step": 5028, + "time_per_iteration": 2.7219650745391846 + }, + { + "auxiliary_loss_clip": 0.01156224, + "auxiliary_loss_mlp": 0.01034565, + "balance_loss_clip": 1.04819489, + "balance_loss_mlp": 1.0250814, + "epoch": 0.6047014970239885, + "flos": 19975059682560.0, + "grad_norm": 2.2172505148846238, + "language_loss": 0.80193561, + "learning_rate": 1.4270292577441864e-06, + "loss": 0.82384354, + "num_input_tokens_seen": 108442290, + "step": 5029, + "time_per_iteration": 3.813812494277954 + }, + { + "auxiliary_loss_clip": 0.01170763, + "auxiliary_loss_mlp": 0.01026428, + "balance_loss_clip": 1.04747725, + "balance_loss_mlp": 1.01768398, + "epoch": 0.6048217399146275, + "flos": 25337097025920.0, + "grad_norm": 1.5459001575079887, + "language_loss": 0.71843314, + "learning_rate": 1.4262829804707836e-06, + "loss": 0.74040508, + "num_input_tokens_seen": 108464280, + "step": 5030, + "time_per_iteration": 2.8428497314453125 + }, + { + "auxiliary_loss_clip": 0.01168176, + "auxiliary_loss_mlp": 0.01030733, + "balance_loss_clip": 1.04729021, + "balance_loss_mlp": 1.02207851, + "epoch": 0.6049419828052667, + "flos": 26030819370240.0, + "grad_norm": 1.4294491623001007, + "language_loss": 0.69800615, + "learning_rate": 1.4255367902297958e-06, + "loss": 0.71999526, + "num_input_tokens_seen": 108485610, + "step": 5031, + "time_per_iteration": 2.7615647315979004 + }, + { + "auxiliary_loss_clip": 0.01174589, + "auxiliary_loss_mlp": 0.01027684, + "balance_loss_clip": 1.0502634, + "balance_loss_mlp": 1.01982808, + "epoch": 0.6050622256959057, + "flos": 14648106948480.0, + "grad_norm": 1.960016157857514, + "language_loss": 0.79047483, + "learning_rate": 1.4247906871344215e-06, + "loss": 0.8124975, + "num_input_tokens_seen": 108501005, + "step": 5032, + "time_per_iteration": 2.5990028381347656 + }, + { + "auxiliary_loss_clip": 0.01151981, + "auxiliary_loss_mlp": 0.01027845, + "balance_loss_clip": 1.04515839, + "balance_loss_mlp": 1.01986074, + "epoch": 0.6051824685865448, + "flos": 23331450337920.0, + "grad_norm": 1.9767233199472867, + "language_loss": 0.75245422, + "learning_rate": 1.4240446712978415e-06, + "loss": 0.77425253, + "num_input_tokens_seen": 108519990, + "step": 5033, + "time_per_iteration": 2.7497596740722656 + }, + { + "auxiliary_loss_clip": 0.01168614, + "auxiliary_loss_mlp": 0.01031079, + "balance_loss_clip": 1.0474298, + "balance_loss_mlp": 1.02254963, + "epoch": 0.605302711477184, + "flos": 27563307177600.0, + "grad_norm": 2.096038772140586, + "language_loss": 0.74309897, + "learning_rate": 1.423298742833227e-06, + "loss": 0.76509595, + "num_input_tokens_seen": 108538650, + "step": 5034, + "time_per_iteration": 2.77850079536438 + }, + { + "auxiliary_loss_clip": 0.01153772, + "auxiliary_loss_mlp": 0.01034684, + "balance_loss_clip": 1.04528093, + "balance_loss_mlp": 1.02605951, + "epoch": 0.605422954367823, + "flos": 15154698412800.0, + "grad_norm": 2.046313142837142, + "language_loss": 0.71722168, + "learning_rate": 1.4225529018537352e-06, + "loss": 0.7391063, + "num_input_tokens_seen": 108554155, + "step": 5035, + "time_per_iteration": 2.800870656967163 + }, + { + "auxiliary_loss_clip": 0.01177934, + "auxiliary_loss_mlp": 0.01032448, + "balance_loss_clip": 1.05074263, + "balance_loss_mlp": 1.02446651, + "epoch": 0.6055431972584621, + "flos": 27673912131840.0, + "grad_norm": 1.5395981048514, + "language_loss": 0.77601814, + "learning_rate": 1.4218071484725082e-06, + "loss": 0.79812205, + "num_input_tokens_seen": 108576275, + "step": 5036, + "time_per_iteration": 2.6696314811706543 + }, + { + "auxiliary_loss_clip": 0.01157178, + "auxiliary_loss_mlp": 0.01031302, + "balance_loss_clip": 1.04984999, + "balance_loss_mlp": 1.02206326, + "epoch": 0.6056634401491012, + "flos": 19387489006080.0, + "grad_norm": 2.198027937380742, + "language_loss": 0.76269245, + "learning_rate": 1.4210614828026786e-06, + "loss": 0.78457725, + "num_input_tokens_seen": 108594125, + "step": 5037, + "time_per_iteration": 2.9026496410369873 + }, + { + "auxiliary_loss_clip": 0.01174791, + "auxiliary_loss_mlp": 0.0102418, + "balance_loss_clip": 1.04816198, + "balance_loss_mlp": 1.01581192, + "epoch": 0.6057836830397403, + "flos": 24789459294720.0, + "grad_norm": 1.5136824807801343, + "language_loss": 0.74309957, + "learning_rate": 1.4203159049573605e-06, + "loss": 0.76508927, + "num_input_tokens_seen": 108615360, + "step": 5038, + "time_per_iteration": 2.7331910133361816 + }, + { + "auxiliary_loss_clip": 0.01157215, + "auxiliary_loss_mlp": 0.01031549, + "balance_loss_clip": 1.04642558, + "balance_loss_mlp": 1.0228703, + "epoch": 0.6059039259303793, + "flos": 20558248899840.0, + "grad_norm": 2.10465557447614, + "language_loss": 0.87306696, + "learning_rate": 1.4195704150496593e-06, + "loss": 0.89495456, + "num_input_tokens_seen": 108633075, + "step": 5039, + "time_per_iteration": 2.8209168910980225 + }, + { + "auxiliary_loss_clip": 0.01161413, + "auxiliary_loss_mlp": 0.01027929, + "balance_loss_clip": 1.05035162, + "balance_loss_mlp": 1.02018309, + "epoch": 0.6060241688210185, + "flos": 21069724613760.0, + "grad_norm": 1.7273643864948183, + "language_loss": 0.73938537, + "learning_rate": 1.4188250131926639e-06, + "loss": 0.76127875, + "num_input_tokens_seen": 108651875, + "step": 5040, + "time_per_iteration": 2.8421287536621094 + }, + { + "auxiliary_loss_clip": 0.01155332, + "auxiliary_loss_mlp": 0.01028196, + "balance_loss_clip": 1.04584932, + "balance_loss_mlp": 1.01962531, + "epoch": 0.6061444117116576, + "flos": 16361081619840.0, + "grad_norm": 2.5122891438784243, + "language_loss": 0.80660868, + "learning_rate": 1.4180796994994525e-06, + "loss": 0.828444, + "num_input_tokens_seen": 108669290, + "step": 5041, + "time_per_iteration": 2.7519490718841553 + }, + { + "auxiliary_loss_clip": 0.01153826, + "auxiliary_loss_mlp": 0.01028287, + "balance_loss_clip": 1.04488397, + "balance_loss_mlp": 1.02033901, + "epoch": 0.6062646546022966, + "flos": 21507296094720.0, + "grad_norm": 1.8711036580294031, + "language_loss": 0.72228885, + "learning_rate": 1.4173344740830877e-06, + "loss": 0.74410993, + "num_input_tokens_seen": 108688420, + "step": 5042, + "time_per_iteration": 2.8139336109161377 + }, + { + "auxiliary_loss_clip": 0.01147755, + "auxiliary_loss_mlp": 0.0102434, + "balance_loss_clip": 1.04667735, + "balance_loss_mlp": 1.01626325, + "epoch": 0.6063848974929358, + "flos": 38983151283840.0, + "grad_norm": 1.581651364261604, + "language_loss": 0.70467615, + "learning_rate": 1.4165893370566206e-06, + "loss": 0.7263971, + "num_input_tokens_seen": 108712175, + "step": 5043, + "time_per_iteration": 2.8704590797424316 + }, + { + "auxiliary_loss_clip": 0.01161975, + "auxiliary_loss_mlp": 0.01024575, + "balance_loss_clip": 1.04672313, + "balance_loss_mlp": 1.01581335, + "epoch": 0.6065051403835748, + "flos": 19646584784640.0, + "grad_norm": 1.6716853805728455, + "language_loss": 0.77607638, + "learning_rate": 1.4158442885330865e-06, + "loss": 0.79794192, + "num_input_tokens_seen": 108730745, + "step": 5044, + "time_per_iteration": 2.724173069000244 + }, + { + "auxiliary_loss_clip": 0.01160572, + "auxiliary_loss_mlp": 0.01028872, + "balance_loss_clip": 1.04678392, + "balance_loss_mlp": 1.02084923, + "epoch": 0.6066253832742139, + "flos": 23513086437120.0, + "grad_norm": 2.9040139538088274, + "language_loss": 0.78850412, + "learning_rate": 1.4150993286255094e-06, + "loss": 0.81039852, + "num_input_tokens_seen": 108749995, + "step": 5045, + "time_per_iteration": 2.8182973861694336 + }, + { + "auxiliary_loss_clip": 0.01172989, + "auxiliary_loss_mlp": 0.01026117, + "balance_loss_clip": 1.04851389, + "balance_loss_mlp": 1.01817203, + "epoch": 0.6067456261648531, + "flos": 19133708440320.0, + "grad_norm": 1.9912711912202825, + "language_loss": 0.79901326, + "learning_rate": 1.4143544574468993e-06, + "loss": 0.82100433, + "num_input_tokens_seen": 108768355, + "step": 5046, + "time_per_iteration": 2.712474822998047 + }, + { + "auxiliary_loss_clip": 0.01159549, + "auxiliary_loss_mlp": 0.01031865, + "balance_loss_clip": 1.04597461, + "balance_loss_mlp": 1.0234369, + "epoch": 0.6068658690554921, + "flos": 20520614424960.0, + "grad_norm": 1.8269631977971166, + "language_loss": 0.82607687, + "learning_rate": 1.4136096751102523e-06, + "loss": 0.84799099, + "num_input_tokens_seen": 108786685, + "step": 5047, + "time_per_iteration": 2.810563564300537 + }, + { + "auxiliary_loss_clip": 0.01161314, + "auxiliary_loss_mlp": 0.01029662, + "balance_loss_clip": 1.04847634, + "balance_loss_mlp": 1.02148449, + "epoch": 0.6069861119461312, + "flos": 27374560185600.0, + "grad_norm": 2.0327385422178907, + "language_loss": 0.83083522, + "learning_rate": 1.4128649817285516e-06, + "loss": 0.852745, + "num_input_tokens_seen": 108804820, + "step": 5048, + "time_per_iteration": 2.797860622406006 + }, + { + "auxiliary_loss_clip": 0.01158845, + "auxiliary_loss_mlp": 0.0102761, + "balance_loss_clip": 1.04677594, + "balance_loss_mlp": 1.02034473, + "epoch": 0.6071063548367702, + "flos": 25626500904960.0, + "grad_norm": 2.084123427094394, + "language_loss": 0.63370782, + "learning_rate": 1.412120377414766e-06, + "loss": 0.65557235, + "num_input_tokens_seen": 108825010, + "step": 5049, + "time_per_iteration": 2.764410972595215 + }, + { + "auxiliary_loss_clip": 0.01173625, + "auxiliary_loss_mlp": 0.01029956, + "balance_loss_clip": 1.05031335, + "balance_loss_mlp": 1.0217073, + "epoch": 0.6072265977274094, + "flos": 24460517520000.0, + "grad_norm": 1.697530308895461, + "language_loss": 0.71418059, + "learning_rate": 1.4113758622818522e-06, + "loss": 0.73621643, + "num_input_tokens_seen": 108845075, + "step": 5050, + "time_per_iteration": 2.6501402854919434 + }, + { + "auxiliary_loss_clip": 0.01162843, + "auxiliary_loss_mlp": 0.00885537, + "balance_loss_clip": 1.05024111, + "balance_loss_mlp": 1.00052762, + "epoch": 0.6073468406180484, + "flos": 18149253413760.0, + "grad_norm": 1.9695334586315156, + "language_loss": 0.83081537, + "learning_rate": 1.410631436442751e-06, + "loss": 0.85129917, + "num_input_tokens_seen": 108863870, + "step": 5051, + "time_per_iteration": 3.631246328353882 + }, + { + "auxiliary_loss_clip": 0.01170831, + "auxiliary_loss_mlp": 0.01030491, + "balance_loss_clip": 1.04852903, + "balance_loss_mlp": 1.02209234, + "epoch": 0.6074670835086875, + "flos": 20697617669760.0, + "grad_norm": 2.6368618957020167, + "language_loss": 0.8661167, + "learning_rate": 1.4098871000103936e-06, + "loss": 0.88812989, + "num_input_tokens_seen": 108882470, + "step": 5052, + "time_per_iteration": 3.613518238067627 + }, + { + "auxiliary_loss_clip": 0.0114909, + "auxiliary_loss_mlp": 0.01028604, + "balance_loss_clip": 1.045434, + "balance_loss_mlp": 1.02058411, + "epoch": 0.6075873263993267, + "flos": 23769955572480.0, + "grad_norm": 1.6176988455259378, + "language_loss": 0.82416463, + "learning_rate": 1.409142853097693e-06, + "loss": 0.84594154, + "num_input_tokens_seen": 108902710, + "step": 5053, + "time_per_iteration": 3.692368507385254 + }, + { + "auxiliary_loss_clip": 0.01158263, + "auxiliary_loss_mlp": 0.0103018, + "balance_loss_clip": 1.04698086, + "balance_loss_mlp": 1.02169776, + "epoch": 0.6077075692899657, + "flos": 24454484035200.0, + "grad_norm": 2.115263236497783, + "language_loss": 0.79479504, + "learning_rate": 1.408398695817553e-06, + "loss": 0.81667948, + "num_input_tokens_seen": 108919935, + "step": 5054, + "time_per_iteration": 2.8085293769836426 + }, + { + "auxiliary_loss_clip": 0.01153731, + "auxiliary_loss_mlp": 0.01037469, + "balance_loss_clip": 1.04464245, + "balance_loss_mlp": 1.02828407, + "epoch": 0.6078278121806048, + "flos": 27382102041600.0, + "grad_norm": 1.7122189321859378, + "language_loss": 0.70324928, + "learning_rate": 1.4076546282828593e-06, + "loss": 0.72516131, + "num_input_tokens_seen": 108942790, + "step": 5055, + "time_per_iteration": 3.6801178455352783 + }, + { + "auxiliary_loss_clip": 0.0115834, + "auxiliary_loss_mlp": 0.01031483, + "balance_loss_clip": 1.04309094, + "balance_loss_mlp": 1.02296531, + "epoch": 0.6079480550712439, + "flos": 38436447306240.0, + "grad_norm": 2.2288663973956395, + "language_loss": 0.65909886, + "learning_rate": 1.4069106506064874e-06, + "loss": 0.68099707, + "num_input_tokens_seen": 108964215, + "step": 5056, + "time_per_iteration": 2.910557508468628 + }, + { + "auxiliary_loss_clip": 0.01150581, + "auxiliary_loss_mlp": 0.01024505, + "balance_loss_clip": 1.04685688, + "balance_loss_mlp": 1.01670885, + "epoch": 0.608068297961883, + "flos": 25336271013120.0, + "grad_norm": 1.7873211997112233, + "language_loss": 0.7833671, + "learning_rate": 1.4061667629012989e-06, + "loss": 0.80511796, + "num_input_tokens_seen": 108984885, + "step": 5057, + "time_per_iteration": 2.838010311126709 + }, + { + "auxiliary_loss_clip": 0.01148494, + "auxiliary_loss_mlp": 0.01028302, + "balance_loss_clip": 1.04665053, + "balance_loss_mlp": 1.02056277, + "epoch": 0.608188540852522, + "flos": 24202463235840.0, + "grad_norm": 1.6028311937588429, + "language_loss": 0.83245444, + "learning_rate": 1.40542296528014e-06, + "loss": 0.85422242, + "num_input_tokens_seen": 109004545, + "step": 5058, + "time_per_iteration": 2.786062717437744 + }, + { + "auxiliary_loss_clip": 0.01164301, + "auxiliary_loss_mlp": 0.01026849, + "balance_loss_clip": 1.04556763, + "balance_loss_mlp": 1.0189569, + "epoch": 0.6083087837431612, + "flos": 21284146851840.0, + "grad_norm": 1.8871421641988717, + "language_loss": 0.7614423, + "learning_rate": 1.4046792578558452e-06, + "loss": 0.78335387, + "num_input_tokens_seen": 109022440, + "step": 5059, + "time_per_iteration": 2.6930129528045654 + }, + { + "auxiliary_loss_clip": 0.01152366, + "auxiliary_loss_mlp": 0.01031304, + "balance_loss_clip": 1.04610395, + "balance_loss_mlp": 1.02335858, + "epoch": 0.6084290266338003, + "flos": 16471435178880.0, + "grad_norm": 2.136890995507971, + "language_loss": 0.75704223, + "learning_rate": 1.4039356407412325e-06, + "loss": 0.77887893, + "num_input_tokens_seen": 109035680, + "step": 5060, + "time_per_iteration": 2.7195448875427246 + }, + { + "auxiliary_loss_clip": 0.01060884, + "auxiliary_loss_mlp": 0.01002891, + "balance_loss_clip": 1.01106143, + "balance_loss_mlp": 1.00138271, + "epoch": 0.6085492695244393, + "flos": 66443574931200.0, + "grad_norm": 0.7830517071526719, + "language_loss": 0.57123911, + "learning_rate": 1.40319211404911e-06, + "loss": 0.59187686, + "num_input_tokens_seen": 109090680, + "step": 5061, + "time_per_iteration": 3.285675287246704 + }, + { + "auxiliary_loss_clip": 0.01175498, + "auxiliary_loss_mlp": 0.01027256, + "balance_loss_clip": 1.049963, + "balance_loss_mlp": 1.01924551, + "epoch": 0.6086695124150785, + "flos": 23618986709760.0, + "grad_norm": 1.661290444283582, + "language_loss": 0.90353656, + "learning_rate": 1.4024486778922691e-06, + "loss": 0.92556411, + "num_input_tokens_seen": 109108995, + "step": 5062, + "time_per_iteration": 2.686281442642212 + }, + { + "auxiliary_loss_clip": 0.01161779, + "auxiliary_loss_mlp": 0.01026695, + "balance_loss_clip": 1.04616547, + "balance_loss_mlp": 1.01846337, + "epoch": 0.6087897553057176, + "flos": 20157054917760.0, + "grad_norm": 1.741111531102161, + "language_loss": 0.77180582, + "learning_rate": 1.4017053323834884e-06, + "loss": 0.79369056, + "num_input_tokens_seen": 109128825, + "step": 5063, + "time_per_iteration": 2.6816654205322266 + }, + { + "auxiliary_loss_clip": 0.01155918, + "auxiliary_loss_mlp": 0.01033443, + "balance_loss_clip": 1.0433538, + "balance_loss_mlp": 1.02558684, + "epoch": 0.6089099981963566, + "flos": 25482535194240.0, + "grad_norm": 2.428774830914427, + "language_loss": 0.76147979, + "learning_rate": 1.4009620776355333e-06, + "loss": 0.78337336, + "num_input_tokens_seen": 109150425, + "step": 5064, + "time_per_iteration": 2.818281650543213 + }, + { + "auxiliary_loss_clip": 0.01161675, + "auxiliary_loss_mlp": 0.0102767, + "balance_loss_clip": 1.04741585, + "balance_loss_mlp": 1.01918244, + "epoch": 0.6090302410869958, + "flos": 25332895134720.0, + "grad_norm": 2.1180259534298114, + "language_loss": 0.79280341, + "learning_rate": 1.4002189137611553e-06, + "loss": 0.81469691, + "num_input_tokens_seen": 109169765, + "step": 5065, + "time_per_iteration": 2.7446179389953613 + }, + { + "auxiliary_loss_clip": 0.01162244, + "auxiliary_loss_mlp": 0.01029368, + "balance_loss_clip": 1.04542589, + "balance_loss_mlp": 1.02150607, + "epoch": 0.6091504839776348, + "flos": 23987358639360.0, + "grad_norm": 1.9660401251053026, + "language_loss": 0.69817662, + "learning_rate": 1.3994758408730901e-06, + "loss": 0.72009277, + "num_input_tokens_seen": 109188950, + "step": 5066, + "time_per_iteration": 2.7557849884033203 + }, + { + "auxiliary_loss_clip": 0.01157033, + "auxiliary_loss_mlp": 0.01037283, + "balance_loss_clip": 1.04811764, + "balance_loss_mlp": 1.02902746, + "epoch": 0.6092707268682739, + "flos": 29643037666560.0, + "grad_norm": 2.1924030526397322, + "language_loss": 0.76712167, + "learning_rate": 1.3987328590840629e-06, + "loss": 0.78906476, + "num_input_tokens_seen": 109209895, + "step": 5067, + "time_per_iteration": 2.7882161140441895 + }, + { + "auxiliary_loss_clip": 0.01161386, + "auxiliary_loss_mlp": 0.01028729, + "balance_loss_clip": 1.04695082, + "balance_loss_mlp": 1.02020562, + "epoch": 0.609390969758913, + "flos": 24024957200640.0, + "grad_norm": 1.9829732615699718, + "language_loss": 0.86184335, + "learning_rate": 1.397989968506783e-06, + "loss": 0.88374448, + "num_input_tokens_seen": 109228905, + "step": 5068, + "time_per_iteration": 2.699918746948242 + }, + { + "auxiliary_loss_clip": 0.01178299, + "auxiliary_loss_mlp": 0.01028485, + "balance_loss_clip": 1.05015254, + "balance_loss_mlp": 1.01994944, + "epoch": 0.6095112126495521, + "flos": 11102143288320.0, + "grad_norm": 5.670631461140425, + "language_loss": 0.72165644, + "learning_rate": 1.3972471692539458e-06, + "loss": 0.74372429, + "num_input_tokens_seen": 109243620, + "step": 5069, + "time_per_iteration": 2.7413527965545654 + }, + { + "auxiliary_loss_clip": 0.01152739, + "auxiliary_loss_mlp": 0.01030115, + "balance_loss_clip": 1.04785717, + "balance_loss_mlp": 1.02190125, + "epoch": 0.6096314555401912, + "flos": 17265491187840.0, + "grad_norm": 3.067699247068969, + "language_loss": 0.75012422, + "learning_rate": 1.3965044614382348e-06, + "loss": 0.77195275, + "num_input_tokens_seen": 109259070, + "step": 5070, + "time_per_iteration": 2.6383161544799805 + }, + { + "auxiliary_loss_clip": 0.01178296, + "auxiliary_loss_mlp": 0.0102773, + "balance_loss_clip": 1.0503751, + "balance_loss_mlp": 1.01912904, + "epoch": 0.6097516984308303, + "flos": 21645910679040.0, + "grad_norm": 2.480175409709351, + "language_loss": 0.75248641, + "learning_rate": 1.3957618451723162e-06, + "loss": 0.77454668, + "num_input_tokens_seen": 109275100, + "step": 5071, + "time_per_iteration": 2.6559829711914062 + }, + { + "auxiliary_loss_clip": 0.01156435, + "auxiliary_loss_mlp": 0.01031239, + "balance_loss_clip": 1.0456574, + "balance_loss_mlp": 1.02291834, + "epoch": 0.6098719413214694, + "flos": 27199208966400.0, + "grad_norm": 2.0994472137325655, + "language_loss": 0.71606565, + "learning_rate": 1.3950193205688457e-06, + "loss": 0.7379424, + "num_input_tokens_seen": 109294825, + "step": 5072, + "time_per_iteration": 2.7472023963928223 + }, + { + "auxiliary_loss_clip": 0.01158427, + "auxiliary_loss_mlp": 0.01030887, + "balance_loss_clip": 1.0492332, + "balance_loss_mlp": 1.02314401, + "epoch": 0.6099921842121084, + "flos": 20412954385920.0, + "grad_norm": 1.8131508182061502, + "language_loss": 0.83623427, + "learning_rate": 1.3942768877404627e-06, + "loss": 0.85812742, + "num_input_tokens_seen": 109313790, + "step": 5073, + "time_per_iteration": 2.758727550506592 + }, + { + "auxiliary_loss_clip": 0.01172244, + "auxiliary_loss_mlp": 0.01029709, + "balance_loss_clip": 1.04684126, + "balance_loss_mlp": 1.02147198, + "epoch": 0.6101124271027476, + "flos": 23366139897600.0, + "grad_norm": 1.580354486960704, + "language_loss": 0.73512733, + "learning_rate": 1.393534546799795e-06, + "loss": 0.75714684, + "num_input_tokens_seen": 109333490, + "step": 5074, + "time_per_iteration": 2.6543421745300293 + }, + { + "auxiliary_loss_clip": 0.01146511, + "auxiliary_loss_mlp": 0.01036644, + "balance_loss_clip": 1.04579866, + "balance_loss_mlp": 1.02810836, + "epoch": 0.6102326699933867, + "flos": 26687840993280.0, + "grad_norm": 1.7741881692330297, + "language_loss": 0.67560816, + "learning_rate": 1.3927922978594536e-06, + "loss": 0.69743973, + "num_input_tokens_seen": 109354575, + "step": 5075, + "time_per_iteration": 2.855154514312744 + }, + { + "auxiliary_loss_clip": 0.01057036, + "auxiliary_loss_mlp": 0.0100198, + "balance_loss_clip": 1.0113858, + "balance_loss_mlp": 1.00050819, + "epoch": 0.6103529128840257, + "flos": 60644612551680.0, + "grad_norm": 0.7805675154336346, + "language_loss": 0.57456088, + "learning_rate": 1.3920501410320387e-06, + "loss": 0.59515095, + "num_input_tokens_seen": 109410690, + "step": 5076, + "time_per_iteration": 3.2582478523254395 + }, + { + "auxiliary_loss_clip": 0.01156093, + "auxiliary_loss_mlp": 0.01030389, + "balance_loss_clip": 1.04652143, + "balance_loss_mlp": 1.02200878, + "epoch": 0.6104731557746649, + "flos": 19021307806080.0, + "grad_norm": 2.0267782645157544, + "language_loss": 0.7590009, + "learning_rate": 1.3913080764301333e-06, + "loss": 0.78086567, + "num_input_tokens_seen": 109427650, + "step": 5077, + "time_per_iteration": 3.6842312812805176 + }, + { + "auxiliary_loss_clip": 0.011454, + "auxiliary_loss_mlp": 0.01036664, + "balance_loss_clip": 1.04266155, + "balance_loss_mlp": 1.02833176, + "epoch": 0.6105933986653039, + "flos": 23366894083200.0, + "grad_norm": 1.888874355775627, + "language_loss": 0.71487439, + "learning_rate": 1.3905661041663085e-06, + "loss": 0.73669505, + "num_input_tokens_seen": 109448835, + "step": 5078, + "time_per_iteration": 3.9057741165161133 + }, + { + "auxiliary_loss_clip": 0.01167201, + "auxiliary_loss_mlp": 0.01030745, + "balance_loss_clip": 1.05003631, + "balance_loss_mlp": 1.02203691, + "epoch": 0.610713641555943, + "flos": 34637565006720.0, + "grad_norm": 1.8831993902942996, + "language_loss": 0.65054071, + "learning_rate": 1.389824224353122e-06, + "loss": 0.67252016, + "num_input_tokens_seen": 109470425, + "step": 5079, + "time_per_iteration": 2.794180154800415 + }, + { + "auxiliary_loss_clip": 0.01165526, + "auxiliary_loss_mlp": 0.01028759, + "balance_loss_clip": 1.05006015, + "balance_loss_mlp": 1.02107561, + "epoch": 0.610833884446582, + "flos": 26646471504000.0, + "grad_norm": 1.5521958867705248, + "language_loss": 0.76788223, + "learning_rate": 1.389082437103115e-06, + "loss": 0.78982508, + "num_input_tokens_seen": 109489695, + "step": 5080, + "time_per_iteration": 2.6694438457489014 + }, + { + "auxiliary_loss_clip": 0.01145819, + "auxiliary_loss_mlp": 0.01026407, + "balance_loss_clip": 1.0432725, + "balance_loss_mlp": 1.01772285, + "epoch": 0.6109541273372212, + "flos": 21215126868480.0, + "grad_norm": 2.269859363694815, + "language_loss": 0.78365952, + "learning_rate": 1.3883407425288172e-06, + "loss": 0.80538183, + "num_input_tokens_seen": 109510030, + "step": 5081, + "time_per_iteration": 3.844752788543701 + }, + { + "auxiliary_loss_clip": 0.01153761, + "auxiliary_loss_mlp": 0.01026408, + "balance_loss_clip": 1.0441184, + "balance_loss_mlp": 1.01755118, + "epoch": 0.6110743702278603, + "flos": 20084084438400.0, + "grad_norm": 2.79270518174357, + "language_loss": 0.79772776, + "learning_rate": 1.3875991407427417e-06, + "loss": 0.81952953, + "num_input_tokens_seen": 109528255, + "step": 5082, + "time_per_iteration": 2.7117416858673096 + }, + { + "auxiliary_loss_clip": 0.01051361, + "auxiliary_loss_mlp": 0.01002978, + "balance_loss_clip": 1.01107073, + "balance_loss_mlp": 1.00134468, + "epoch": 0.6111946131184993, + "flos": 68302957438080.0, + "grad_norm": 0.8701884175413666, + "language_loss": 0.58180642, + "learning_rate": 1.38685763185739e-06, + "loss": 0.60234976, + "num_input_tokens_seen": 109581915, + "step": 5083, + "time_per_iteration": 3.396982192993164 + }, + { + "auxiliary_loss_clip": 0.01175608, + "auxiliary_loss_mlp": 0.01032663, + "balance_loss_clip": 1.04990602, + "balance_loss_mlp": 1.02350783, + "epoch": 0.6113148560091385, + "flos": 19937676602880.0, + "grad_norm": 4.175285738450409, + "language_loss": 0.6782409, + "learning_rate": 1.3861162159852476e-06, + "loss": 0.70032358, + "num_input_tokens_seen": 109600050, + "step": 5084, + "time_per_iteration": 2.693451166152954 + }, + { + "auxiliary_loss_clip": 0.01161832, + "auxiliary_loss_mlp": 0.0102581, + "balance_loss_clip": 1.04850698, + "balance_loss_mlp": 1.01707244, + "epoch": 0.6114350988997775, + "flos": 23731854220800.0, + "grad_norm": 4.176133080421647, + "language_loss": 0.79645646, + "learning_rate": 1.3853748932387875e-06, + "loss": 0.81833285, + "num_input_tokens_seen": 109620690, + "step": 5085, + "time_per_iteration": 2.769961357116699 + }, + { + "auxiliary_loss_clip": 0.01146021, + "auxiliary_loss_mlp": 0.01031074, + "balance_loss_clip": 1.04375207, + "balance_loss_mlp": 1.02224088, + "epoch": 0.6115553417904166, + "flos": 24023700224640.0, + "grad_norm": 2.337406813298057, + "language_loss": 0.75014269, + "learning_rate": 1.3846336637304671e-06, + "loss": 0.77191365, + "num_input_tokens_seen": 109638960, + "step": 5086, + "time_per_iteration": 2.7987306118011475 + }, + { + "auxiliary_loss_clip": 0.01147436, + "auxiliary_loss_mlp": 0.0103112, + "balance_loss_clip": 1.04548192, + "balance_loss_mlp": 1.0229423, + "epoch": 0.6116755846810558, + "flos": 23733542160000.0, + "grad_norm": 2.033371216048536, + "language_loss": 0.83354855, + "learning_rate": 1.3838925275727316e-06, + "loss": 0.8553341, + "num_input_tokens_seen": 109659700, + "step": 5087, + "time_per_iteration": 2.76335072517395 + }, + { + "auxiliary_loss_clip": 0.01175953, + "auxiliary_loss_mlp": 0.01027328, + "balance_loss_clip": 1.05005527, + "balance_loss_mlp": 1.01915586, + "epoch": 0.6117958275716948, + "flos": 18661626967680.0, + "grad_norm": 1.9772704069097615, + "language_loss": 0.79050457, + "learning_rate": 1.3831514848780089e-06, + "loss": 0.81253731, + "num_input_tokens_seen": 109679275, + "step": 5088, + "time_per_iteration": 2.7821402549743652 + }, + { + "auxiliary_loss_clip": 0.01157263, + "auxiliary_loss_mlp": 0.0102936, + "balance_loss_clip": 1.04570341, + "balance_loss_mlp": 1.02145016, + "epoch": 0.6119160704623339, + "flos": 16471183783680.0, + "grad_norm": 3.2883447173822717, + "language_loss": 0.91960371, + "learning_rate": 1.3824105357587152e-06, + "loss": 0.94146991, + "num_input_tokens_seen": 109696380, + "step": 5089, + "time_per_iteration": 2.706277847290039 + }, + { + "auxiliary_loss_clip": 0.01149391, + "auxiliary_loss_mlp": 0.0102987, + "balance_loss_clip": 1.04344678, + "balance_loss_mlp": 1.02120972, + "epoch": 0.612036313352973, + "flos": 23915465568000.0, + "grad_norm": 1.7505796196510683, + "language_loss": 0.82398415, + "learning_rate": 1.381669680327253e-06, + "loss": 0.8457768, + "num_input_tokens_seen": 109718060, + "step": 5090, + "time_per_iteration": 2.882601261138916 + }, + { + "auxiliary_loss_clip": 0.01150754, + "auxiliary_loss_mlp": 0.01026278, + "balance_loss_clip": 1.04781604, + "balance_loss_mlp": 1.01787388, + "epoch": 0.6121565562436121, + "flos": 26974766833920.0, + "grad_norm": 2.9141978902357697, + "language_loss": 0.70742583, + "learning_rate": 1.380928918696008e-06, + "loss": 0.72919619, + "num_input_tokens_seen": 109736830, + "step": 5091, + "time_per_iteration": 2.8475611209869385 + }, + { + "auxiliary_loss_clip": 0.01163597, + "auxiliary_loss_mlp": 0.01030545, + "balance_loss_clip": 1.04685521, + "balance_loss_mlp": 1.02270055, + "epoch": 0.6122767991342511, + "flos": 15668867646720.0, + "grad_norm": 3.2525220159436827, + "language_loss": 0.71065533, + "learning_rate": 1.3801882509773548e-06, + "loss": 0.73259676, + "num_input_tokens_seen": 109754690, + "step": 5092, + "time_per_iteration": 2.6702563762664795 + }, + { + "auxiliary_loss_clip": 0.01159215, + "auxiliary_loss_mlp": 0.01031259, + "balance_loss_clip": 1.04472136, + "balance_loss_mlp": 1.02224708, + "epoch": 0.6123970420248903, + "flos": 27964321591680.0, + "grad_norm": 1.8663267924059501, + "language_loss": 0.81745255, + "learning_rate": 1.3794476772836503e-06, + "loss": 0.83935732, + "num_input_tokens_seen": 109775790, + "step": 5093, + "time_per_iteration": 2.8052618503570557 + }, + { + "auxiliary_loss_clip": 0.01139524, + "auxiliary_loss_mlp": 0.01028827, + "balance_loss_clip": 1.04522097, + "balance_loss_mlp": 1.0195291, + "epoch": 0.6125172849155294, + "flos": 21468727866240.0, + "grad_norm": 1.7088943841801005, + "language_loss": 0.84220868, + "learning_rate": 1.3787071977272402e-06, + "loss": 0.8638922, + "num_input_tokens_seen": 109795050, + "step": 5094, + "time_per_iteration": 2.9641895294189453 + }, + { + "auxiliary_loss_clip": 0.0113331, + "auxiliary_loss_mlp": 0.01029596, + "balance_loss_clip": 1.04585505, + "balance_loss_mlp": 1.02134705, + "epoch": 0.6126375278061684, + "flos": 16248321849600.0, + "grad_norm": 2.6234000091967644, + "language_loss": 0.71001542, + "learning_rate": 1.3779668124204535e-06, + "loss": 0.73164451, + "num_input_tokens_seen": 109811465, + "step": 5095, + "time_per_iteration": 2.9406049251556396 + }, + { + "auxiliary_loss_clip": 0.01150818, + "auxiliary_loss_mlp": 0.01026224, + "balance_loss_clip": 1.05002308, + "balance_loss_mlp": 1.01773047, + "epoch": 0.6127577706968076, + "flos": 20448865008000.0, + "grad_norm": 2.067075973245237, + "language_loss": 0.80633789, + "learning_rate": 1.3772265214756074e-06, + "loss": 0.82810831, + "num_input_tokens_seen": 109831225, + "step": 5096, + "time_per_iteration": 2.685781717300415 + }, + { + "auxiliary_loss_clip": 0.01167212, + "auxiliary_loss_mlp": 0.01031375, + "balance_loss_clip": 1.04543972, + "balance_loss_mlp": 1.02280426, + "epoch": 0.6128780135874466, + "flos": 18260397072000.0, + "grad_norm": 1.9102929717001862, + "language_loss": 0.75512207, + "learning_rate": 1.3764863250050025e-06, + "loss": 0.77710795, + "num_input_tokens_seen": 109849465, + "step": 5097, + "time_per_iteration": 2.752454996109009 + }, + { + "auxiliary_loss_clip": 0.01146809, + "auxiliary_loss_mlp": 0.01027455, + "balance_loss_clip": 1.04482818, + "balance_loss_mlp": 1.01917589, + "epoch": 0.6129982564780857, + "flos": 24937088192640.0, + "grad_norm": 1.7496204903632413, + "language_loss": 0.80533838, + "learning_rate": 1.3757462231209272e-06, + "loss": 0.82708108, + "num_input_tokens_seen": 109869770, + "step": 5098, + "time_per_iteration": 2.785696506500244 + }, + { + "auxiliary_loss_clip": 0.01154295, + "auxiliary_loss_mlp": 0.01031166, + "balance_loss_clip": 1.04791427, + "balance_loss_mlp": 1.02273822, + "epoch": 0.6131184993687249, + "flos": 22492038430080.0, + "grad_norm": 2.888062083114474, + "language_loss": 0.88825995, + "learning_rate": 1.3750062159356525e-06, + "loss": 0.91011453, + "num_input_tokens_seen": 109889120, + "step": 5099, + "time_per_iteration": 2.8288276195526123 + }, + { + "auxiliary_loss_clip": 0.0113642, + "auxiliary_loss_mlp": 0.01029412, + "balance_loss_clip": 1.04330444, + "balance_loss_mlp": 1.02122879, + "epoch": 0.6132387422593639, + "flos": 15885839750400.0, + "grad_norm": 1.944950539374122, + "language_loss": 0.83377218, + "learning_rate": 1.3742663035614382e-06, + "loss": 0.85543048, + "num_input_tokens_seen": 109906490, + "step": 5100, + "time_per_iteration": 2.8115062713623047 + }, + { + "auxiliary_loss_clip": 0.01178531, + "auxiliary_loss_mlp": 0.01029553, + "balance_loss_clip": 1.05149746, + "balance_loss_mlp": 1.02170897, + "epoch": 0.613358985150003, + "flos": 25411539962880.0, + "grad_norm": 1.807723323041259, + "language_loss": 0.80198616, + "learning_rate": 1.3735264861105283e-06, + "loss": 0.824067, + "num_input_tokens_seen": 109927130, + "step": 5101, + "time_per_iteration": 2.6827750205993652 + }, + { + "auxiliary_loss_clip": 0.01146484, + "auxiliary_loss_mlp": 0.0102781, + "balance_loss_clip": 1.04325151, + "balance_loss_mlp": 1.01994836, + "epoch": 0.6134792280406421, + "flos": 21361283308800.0, + "grad_norm": 2.0199060467183307, + "language_loss": 0.78488225, + "learning_rate": 1.372786763695152e-06, + "loss": 0.80662519, + "num_input_tokens_seen": 109945890, + "step": 5102, + "time_per_iteration": 2.836439371109009 + }, + { + "auxiliary_loss_clip": 0.01165436, + "auxiliary_loss_mlp": 0.01031937, + "balance_loss_clip": 1.04579091, + "balance_loss_mlp": 1.02303183, + "epoch": 0.6135994709312812, + "flos": 21211248199680.0, + "grad_norm": 1.695563440497856, + "language_loss": 0.77604049, + "learning_rate": 1.3720471364275257e-06, + "loss": 0.79801416, + "num_input_tokens_seen": 109965535, + "step": 5103, + "time_per_iteration": 3.6229212284088135 + }, + { + "auxiliary_loss_clip": 0.01143358, + "auxiliary_loss_mlp": 0.00886784, + "balance_loss_clip": 1.0441761, + "balance_loss_mlp": 1.00069344, + "epoch": 0.6137197138219203, + "flos": 14794047907200.0, + "grad_norm": 1.8677270340976775, + "language_loss": 0.77758229, + "learning_rate": 1.3713076044198486e-06, + "loss": 0.79788375, + "num_input_tokens_seen": 109982345, + "step": 5104, + "time_per_iteration": 4.643614292144775 + }, + { + "auxiliary_loss_clip": 0.01149592, + "auxiliary_loss_mlp": 0.01031521, + "balance_loss_clip": 1.04452944, + "balance_loss_mlp": 1.02304506, + "epoch": 0.6138399567125594, + "flos": 20084515401600.0, + "grad_norm": 2.5969383127319436, + "language_loss": 0.80969793, + "learning_rate": 1.3705681677843086e-06, + "loss": 0.83150905, + "num_input_tokens_seen": 110000940, + "step": 5105, + "time_per_iteration": 2.7599997520446777 + }, + { + "auxiliary_loss_clip": 0.01064415, + "auxiliary_loss_mlp": 0.0100148, + "balance_loss_clip": 1.0096097, + "balance_loss_mlp": 0.99989462, + "epoch": 0.6139601996031985, + "flos": 60123838193280.0, + "grad_norm": 0.9576072887391479, + "language_loss": 0.60603881, + "learning_rate": 1.3698288266330768e-06, + "loss": 0.62669778, + "num_input_tokens_seen": 110061565, + "step": 5106, + "time_per_iteration": 3.2908527851104736 + }, + { + "auxiliary_loss_clip": 0.01155591, + "auxiliary_loss_mlp": 0.01024474, + "balance_loss_clip": 1.05066109, + "balance_loss_mlp": 1.01659727, + "epoch": 0.6140804424938375, + "flos": 23586703361280.0, + "grad_norm": 2.330713414278607, + "language_loss": 0.72836196, + "learning_rate": 1.3690895810783113e-06, + "loss": 0.7501626, + "num_input_tokens_seen": 110080360, + "step": 5107, + "time_per_iteration": 2.7687768936157227 + }, + { + "auxiliary_loss_clip": 0.01125435, + "auxiliary_loss_mlp": 0.00886663, + "balance_loss_clip": 1.04064274, + "balance_loss_mlp": 1.0006423, + "epoch": 0.6142006853844767, + "flos": 21398199511680.0, + "grad_norm": 2.7088724569480713, + "language_loss": 0.71852684, + "learning_rate": 1.3683504312321543e-06, + "loss": 0.73864782, + "num_input_tokens_seen": 110100695, + "step": 5108, + "time_per_iteration": 3.856910228729248 + }, + { + "auxiliary_loss_clip": 0.01169013, + "auxiliary_loss_mlp": 0.01030299, + "balance_loss_clip": 1.04699957, + "balance_loss_mlp": 1.02211547, + "epoch": 0.6143209282751158, + "flos": 12057367622400.0, + "grad_norm": 2.1017426804904966, + "language_loss": 0.80030644, + "learning_rate": 1.3676113772067355e-06, + "loss": 0.8222996, + "num_input_tokens_seen": 110117750, + "step": 5109, + "time_per_iteration": 2.6925833225250244 + }, + { + "auxiliary_loss_clip": 0.01143491, + "auxiliary_loss_mlp": 0.01028888, + "balance_loss_clip": 1.04474759, + "balance_loss_mlp": 1.02015626, + "epoch": 0.6144411711657548, + "flos": 25082274965760.0, + "grad_norm": 1.9192839523712946, + "language_loss": 0.72867101, + "learning_rate": 1.3668724191141671e-06, + "loss": 0.75039482, + "num_input_tokens_seen": 110137020, + "step": 5110, + "time_per_iteration": 2.8377578258514404 + }, + { + "auxiliary_loss_clip": 0.01139097, + "auxiliary_loss_mlp": 0.01029526, + "balance_loss_clip": 1.04874086, + "balance_loss_mlp": 1.02109218, + "epoch": 0.6145614140563939, + "flos": 20114069316480.0, + "grad_norm": 2.6001665074763545, + "language_loss": 0.66303301, + "learning_rate": 1.3661335570665493e-06, + "loss": 0.68471926, + "num_input_tokens_seen": 110154930, + "step": 5111, + "time_per_iteration": 2.788236618041992 + }, + { + "auxiliary_loss_clip": 0.01160473, + "auxiliary_loss_mlp": 0.01029696, + "balance_loss_clip": 1.04997015, + "balance_loss_mlp": 1.02140522, + "epoch": 0.614681656947033, + "flos": 16800376953600.0, + "grad_norm": 2.9204295347443816, + "language_loss": 0.69971949, + "learning_rate": 1.3653947911759676e-06, + "loss": 0.72162116, + "num_input_tokens_seen": 110172480, + "step": 5112, + "time_per_iteration": 2.689457654953003 + }, + { + "auxiliary_loss_clip": 0.01129215, + "auxiliary_loss_mlp": 0.01030286, + "balance_loss_clip": 1.04133511, + "balance_loss_mlp": 1.02198875, + "epoch": 0.6148018998376721, + "flos": 38801587011840.0, + "grad_norm": 1.6615529107684457, + "language_loss": 0.74405724, + "learning_rate": 1.3646561215544904e-06, + "loss": 0.76565224, + "num_input_tokens_seen": 110197120, + "step": 5113, + "time_per_iteration": 2.957594156265259 + }, + { + "auxiliary_loss_clip": 0.01166769, + "auxiliary_loss_mlp": 0.01028797, + "balance_loss_clip": 1.04864669, + "balance_loss_mlp": 1.02012467, + "epoch": 0.6149221427283111, + "flos": 23327032965120.0, + "grad_norm": 2.000352118939182, + "language_loss": 0.79546714, + "learning_rate": 1.363917548314176e-06, + "loss": 0.81742275, + "num_input_tokens_seen": 110216385, + "step": 5114, + "time_per_iteration": 2.699981689453125 + }, + { + "auxiliary_loss_clip": 0.01165506, + "auxiliary_loss_mlp": 0.01028458, + "balance_loss_clip": 1.04838538, + "balance_loss_mlp": 1.01964247, + "epoch": 0.6150423856189503, + "flos": 22379494141440.0, + "grad_norm": 2.1259858799353535, + "language_loss": 0.73169047, + "learning_rate": 1.3631790715670626e-06, + "loss": 0.75363016, + "num_input_tokens_seen": 110234790, + "step": 5115, + "time_per_iteration": 2.7740468978881836 + }, + { + "auxiliary_loss_clip": 0.01107451, + "auxiliary_loss_mlp": 0.01026387, + "balance_loss_clip": 1.0414145, + "balance_loss_mlp": 1.01898956, + "epoch": 0.6151626285095894, + "flos": 18692078722560.0, + "grad_norm": 2.018630517687444, + "language_loss": 0.85887015, + "learning_rate": 1.3624406914251783e-06, + "loss": 0.88020849, + "num_input_tokens_seen": 110251910, + "step": 5116, + "time_per_iteration": 2.937206506729126 + }, + { + "auxiliary_loss_clip": 0.01165772, + "auxiliary_loss_mlp": 0.0103111, + "balance_loss_clip": 1.04677653, + "balance_loss_mlp": 1.02240717, + "epoch": 0.6152828714002284, + "flos": 15851688894720.0, + "grad_norm": 1.9426592076096032, + "language_loss": 0.88230824, + "learning_rate": 1.3617024080005335e-06, + "loss": 0.90427709, + "num_input_tokens_seen": 110268810, + "step": 5117, + "time_per_iteration": 2.7513325214385986 + }, + { + "auxiliary_loss_clip": 0.01154725, + "auxiliary_loss_mlp": 0.00886985, + "balance_loss_clip": 1.04692101, + "balance_loss_mlp": 1.00075233, + "epoch": 0.6154031142908676, + "flos": 24869792062080.0, + "grad_norm": 1.7022114561640451, + "language_loss": 0.74536967, + "learning_rate": 1.3609642214051266e-06, + "loss": 0.76578677, + "num_input_tokens_seen": 110293035, + "step": 5118, + "time_per_iteration": 2.829956531524658 + }, + { + "auxiliary_loss_clip": 0.01148474, + "auxiliary_loss_mlp": 0.01027438, + "balance_loss_clip": 1.04770446, + "balance_loss_mlp": 1.01853871, + "epoch": 0.6155233571815066, + "flos": 19244744357760.0, + "grad_norm": 1.7683522670896605, + "language_loss": 0.66132897, + "learning_rate": 1.3602261317509385e-06, + "loss": 0.68308806, + "num_input_tokens_seen": 110309695, + "step": 5119, + "time_per_iteration": 2.736034631729126 + }, + { + "auxiliary_loss_clip": 0.0116913, + "auxiliary_loss_mlp": 0.01028333, + "balance_loss_clip": 1.04923558, + "balance_loss_mlp": 1.01957703, + "epoch": 0.6156436000721457, + "flos": 18770077105920.0, + "grad_norm": 2.596395485861064, + "language_loss": 0.82720113, + "learning_rate": 1.3594881391499387e-06, + "loss": 0.84917581, + "num_input_tokens_seen": 110328610, + "step": 5120, + "time_per_iteration": 2.6979408264160156 + }, + { + "auxiliary_loss_clip": 0.01157792, + "auxiliary_loss_mlp": 0.01030713, + "balance_loss_clip": 1.04717958, + "balance_loss_mlp": 1.02251172, + "epoch": 0.6157638429627849, + "flos": 18041198325120.0, + "grad_norm": 1.8095317064588432, + "language_loss": 0.79228026, + "learning_rate": 1.3587502437140778e-06, + "loss": 0.81416535, + "num_input_tokens_seen": 110346775, + "step": 5121, + "time_per_iteration": 2.874053955078125 + }, + { + "auxiliary_loss_clip": 0.01161129, + "auxiliary_loss_mlp": 0.01033486, + "balance_loss_clip": 1.04622829, + "balance_loss_mlp": 1.02483749, + "epoch": 0.6158840858534239, + "flos": 25556726736000.0, + "grad_norm": 2.157593843873021, + "language_loss": 0.84905088, + "learning_rate": 1.3580124455552952e-06, + "loss": 0.87099701, + "num_input_tokens_seen": 110366140, + "step": 5122, + "time_per_iteration": 2.8729894161224365 + }, + { + "auxiliary_loss_clip": 0.01165712, + "auxiliary_loss_mlp": 0.00886189, + "balance_loss_clip": 1.0490123, + "balance_loss_mlp": 1.00080109, + "epoch": 0.616004328744063, + "flos": 24640788902400.0, + "grad_norm": 1.6902790514950519, + "language_loss": 0.87744111, + "learning_rate": 1.3572747447855148e-06, + "loss": 0.89796013, + "num_input_tokens_seen": 110386550, + "step": 5123, + "time_per_iteration": 2.73157000541687 + }, + { + "auxiliary_loss_clip": 0.0117666, + "auxiliary_loss_mlp": 0.01034966, + "balance_loss_clip": 1.05007839, + "balance_loss_mlp": 1.02564979, + "epoch": 0.6161245716347021, + "flos": 21689686379520.0, + "grad_norm": 3.353730024164327, + "language_loss": 0.69364178, + "learning_rate": 1.356537141516644e-06, + "loss": 0.71575797, + "num_input_tokens_seen": 110403970, + "step": 5124, + "time_per_iteration": 2.7087185382843018 + }, + { + "auxiliary_loss_clip": 0.01166199, + "auxiliary_loss_mlp": 0.01028485, + "balance_loss_clip": 1.05040145, + "balance_loss_mlp": 1.02086151, + "epoch": 0.6162448145253412, + "flos": 35189225061120.0, + "grad_norm": 1.9185374587343904, + "language_loss": 0.61800414, + "learning_rate": 1.3557996358605775e-06, + "loss": 0.63995099, + "num_input_tokens_seen": 110423890, + "step": 5125, + "time_per_iteration": 2.8386917114257812 + }, + { + "auxiliary_loss_clip": 0.0116645, + "auxiliary_loss_mlp": 0.01031581, + "balance_loss_clip": 1.04864156, + "balance_loss_mlp": 1.02340293, + "epoch": 0.6163650574159802, + "flos": 21615279356160.0, + "grad_norm": 2.2412980146956754, + "language_loss": 0.70009631, + "learning_rate": 1.3550622279291941e-06, + "loss": 0.72207665, + "num_input_tokens_seen": 110442035, + "step": 5126, + "time_per_iteration": 2.7602767944335938 + }, + { + "auxiliary_loss_clip": 0.01129376, + "auxiliary_loss_mlp": 0.01031022, + "balance_loss_clip": 1.04181981, + "balance_loss_mlp": 1.02257597, + "epoch": 0.6164853003066194, + "flos": 24572163968640.0, + "grad_norm": 1.567446591735042, + "language_loss": 0.83377761, + "learning_rate": 1.354324917834358e-06, + "loss": 0.85538155, + "num_input_tokens_seen": 110463280, + "step": 5127, + "time_per_iteration": 3.000945568084717 + }, + { + "auxiliary_loss_clip": 0.01122763, + "auxiliary_loss_mlp": 0.00886408, + "balance_loss_clip": 1.04128933, + "balance_loss_mlp": 1.00073338, + "epoch": 0.6166055431972585, + "flos": 21835986474240.0, + "grad_norm": 1.6298318187551166, + "language_loss": 0.77031833, + "learning_rate": 1.353587705687918e-06, + "loss": 0.79041004, + "num_input_tokens_seen": 110481455, + "step": 5128, + "time_per_iteration": 2.848991632461548 + }, + { + "auxiliary_loss_clip": 0.01163506, + "auxiliary_loss_mlp": 0.01029288, + "balance_loss_clip": 1.05140758, + "balance_loss_mlp": 1.02068102, + "epoch": 0.6167257860878975, + "flos": 17785262943360.0, + "grad_norm": 2.8161193091189816, + "language_loss": 0.72160769, + "learning_rate": 1.3528505916017096e-06, + "loss": 0.74353564, + "num_input_tokens_seen": 110499155, + "step": 5129, + "time_per_iteration": 3.9302761554718018 + }, + { + "auxiliary_loss_clip": 0.01166834, + "auxiliary_loss_mlp": 0.01029973, + "balance_loss_clip": 1.04730701, + "balance_loss_mlp": 1.02177715, + "epoch": 0.6168460289785367, + "flos": 23214811898880.0, + "grad_norm": 2.3742578103332086, + "language_loss": 0.88613862, + "learning_rate": 1.3521135756875514e-06, + "loss": 0.90810668, + "num_input_tokens_seen": 110515470, + "step": 5130, + "time_per_iteration": 3.645143508911133 + }, + { + "auxiliary_loss_clip": 0.0112291, + "auxiliary_loss_mlp": 0.01030937, + "balance_loss_clip": 1.04338562, + "balance_loss_mlp": 1.02259266, + "epoch": 0.6169662718691757, + "flos": 26213281482240.0, + "grad_norm": 1.4227258300472634, + "language_loss": 0.86508209, + "learning_rate": 1.3513766580572496e-06, + "loss": 0.88662058, + "num_input_tokens_seen": 110538290, + "step": 5131, + "time_per_iteration": 3.7825098037719727 + }, + { + "auxiliary_loss_clip": 0.01165316, + "auxiliary_loss_mlp": 0.01029031, + "balance_loss_clip": 1.0486002, + "balance_loss_mlp": 1.02142596, + "epoch": 0.6170865147598148, + "flos": 19026120228480.0, + "grad_norm": 2.047692718679026, + "language_loss": 0.77416599, + "learning_rate": 1.3506398388225924e-06, + "loss": 0.79610944, + "num_input_tokens_seen": 110555610, + "step": 5132, + "time_per_iteration": 2.7054593563079834 + }, + { + "auxiliary_loss_clip": 0.01175679, + "auxiliary_loss_mlp": 0.01025865, + "balance_loss_clip": 1.05121708, + "balance_loss_mlp": 1.01786041, + "epoch": 0.617206757650454, + "flos": 18260361158400.0, + "grad_norm": 2.364685015588888, + "language_loss": 0.71791518, + "learning_rate": 1.349903118095355e-06, + "loss": 0.73993063, + "num_input_tokens_seen": 110574745, + "step": 5133, + "time_per_iteration": 3.5239219665527344 + }, + { + "auxiliary_loss_clip": 0.01168954, + "auxiliary_loss_mlp": 0.01027972, + "balance_loss_clip": 1.0484736, + "balance_loss_mlp": 1.01948392, + "epoch": 0.617327000541093, + "flos": 18186959715840.0, + "grad_norm": 2.2661413979650615, + "language_loss": 0.73572201, + "learning_rate": 1.349166495987298e-06, + "loss": 0.75769126, + "num_input_tokens_seen": 110593310, + "step": 5134, + "time_per_iteration": 2.655313014984131 + }, + { + "auxiliary_loss_clip": 0.01052844, + "auxiliary_loss_mlp": 0.01004413, + "balance_loss_clip": 1.01419699, + "balance_loss_mlp": 1.00270832, + "epoch": 0.6174472434317321, + "flos": 61833796122240.0, + "grad_norm": 0.8179924109130036, + "language_loss": 0.60800004, + "learning_rate": 1.348429972610166e-06, + "loss": 0.62857264, + "num_input_tokens_seen": 110657615, + "step": 5135, + "time_per_iteration": 3.334967851638794 + }, + { + "auxiliary_loss_clip": 0.01037384, + "auxiliary_loss_mlp": 0.01006459, + "balance_loss_clip": 1.01431775, + "balance_loss_mlp": 1.00469446, + "epoch": 0.6175674863223712, + "flos": 71230970494080.0, + "grad_norm": 0.859445812730298, + "language_loss": 0.57820666, + "learning_rate": 1.3476935480756897e-06, + "loss": 0.59864503, + "num_input_tokens_seen": 110714365, + "step": 5136, + "time_per_iteration": 3.301150321960449 + }, + { + "auxiliary_loss_clip": 0.01138178, + "auxiliary_loss_mlp": 0.01032354, + "balance_loss_clip": 1.04381227, + "balance_loss_mlp": 1.02356195, + "epoch": 0.6176877292130103, + "flos": 21835447770240.0, + "grad_norm": 2.16571110782779, + "language_loss": 0.7525475, + "learning_rate": 1.346957222495583e-06, + "loss": 0.77425277, + "num_input_tokens_seen": 110732160, + "step": 5137, + "time_per_iteration": 2.847012996673584 + }, + { + "auxiliary_loss_clip": 0.01155766, + "auxiliary_loss_mlp": 0.00885852, + "balance_loss_clip": 1.04905832, + "balance_loss_mlp": 1.00065851, + "epoch": 0.6178079721036493, + "flos": 17741738638080.0, + "grad_norm": 3.192579553059702, + "language_loss": 0.71075433, + "learning_rate": 1.3462209959815466e-06, + "loss": 0.73117054, + "num_input_tokens_seen": 110746900, + "step": 5138, + "time_per_iteration": 2.773975372314453 + }, + { + "auxiliary_loss_clip": 0.0116027, + "auxiliary_loss_mlp": 0.01026345, + "balance_loss_clip": 1.04906368, + "balance_loss_mlp": 1.01873958, + "epoch": 0.6179282149942885, + "flos": 22633131052800.0, + "grad_norm": 1.825159771554409, + "language_loss": 0.7411983, + "learning_rate": 1.345484868645265e-06, + "loss": 0.76306438, + "num_input_tokens_seen": 110765710, + "step": 5139, + "time_per_iteration": 2.7809736728668213 + }, + { + "auxiliary_loss_clip": 0.01146541, + "auxiliary_loss_mlp": 0.01026073, + "balance_loss_clip": 1.04544973, + "balance_loss_mlp": 1.01743627, + "epoch": 0.6180484578849276, + "flos": 22310330503680.0, + "grad_norm": 2.399642569194441, + "language_loss": 0.78569877, + "learning_rate": 1.3447488405984088e-06, + "loss": 0.8074249, + "num_input_tokens_seen": 110783970, + "step": 5140, + "time_per_iteration": 2.880796432495117 + }, + { + "auxiliary_loss_clip": 0.01156037, + "auxiliary_loss_mlp": 0.01030982, + "balance_loss_clip": 1.0476656, + "balance_loss_mlp": 1.02229738, + "epoch": 0.6181687007755666, + "flos": 35225458905600.0, + "grad_norm": 1.9839045986538995, + "language_loss": 0.69826007, + "learning_rate": 1.3440129119526322e-06, + "loss": 0.72013021, + "num_input_tokens_seen": 110806395, + "step": 5141, + "time_per_iteration": 2.9093592166900635 + }, + { + "auxiliary_loss_clip": 0.01064764, + "auxiliary_loss_mlp": 0.01000605, + "balance_loss_clip": 1.010216, + "balance_loss_mlp": 0.99907917, + "epoch": 0.6182889436662057, + "flos": 61547370094080.0, + "grad_norm": 0.8154285503159787, + "language_loss": 0.51211405, + "learning_rate": 1.3432770828195762e-06, + "loss": 0.53276777, + "num_input_tokens_seen": 110867380, + "step": 5142, + "time_per_iteration": 3.3768208026885986 + }, + { + "auxiliary_loss_clip": 0.01137423, + "auxiliary_loss_mlp": 0.0103262, + "balance_loss_clip": 1.04157281, + "balance_loss_mlp": 1.02421021, + "epoch": 0.6184091865568448, + "flos": 19609991804160.0, + "grad_norm": 2.2930893405821853, + "language_loss": 0.70968938, + "learning_rate": 1.3425413533108635e-06, + "loss": 0.73138982, + "num_input_tokens_seen": 110885980, + "step": 5143, + "time_per_iteration": 2.830171823501587 + }, + { + "auxiliary_loss_clip": 0.01138157, + "auxiliary_loss_mlp": 0.01031877, + "balance_loss_clip": 1.04829061, + "balance_loss_mlp": 1.02387857, + "epoch": 0.6185294294474839, + "flos": 23586882929280.0, + "grad_norm": 2.175060974432728, + "language_loss": 0.70971632, + "learning_rate": 1.341805723538105e-06, + "loss": 0.73141664, + "num_input_tokens_seen": 110906085, + "step": 5144, + "time_per_iteration": 2.8041248321533203 + }, + { + "auxiliary_loss_clip": 0.01158718, + "auxiliary_loss_mlp": 0.01030357, + "balance_loss_clip": 1.05010426, + "balance_loss_mlp": 1.02225709, + "epoch": 0.618649672338123, + "flos": 26762032535040.0, + "grad_norm": 1.5309722535079835, + "language_loss": 0.77519166, + "learning_rate": 1.3410701936128948e-06, + "loss": 0.79708242, + "num_input_tokens_seen": 110928865, + "step": 5145, + "time_per_iteration": 2.7944915294647217 + }, + { + "auxiliary_loss_clip": 0.01164172, + "auxiliary_loss_mlp": 0.01025318, + "balance_loss_clip": 1.05054355, + "balance_loss_mlp": 1.0172596, + "epoch": 0.6187699152287621, + "flos": 14456630522880.0, + "grad_norm": 2.752297566069669, + "language_loss": 0.85315239, + "learning_rate": 1.340334763646812e-06, + "loss": 0.87504727, + "num_input_tokens_seen": 110943000, + "step": 5146, + "time_per_iteration": 2.6455864906311035 + }, + { + "auxiliary_loss_clip": 0.01178017, + "auxiliary_loss_mlp": 0.01032437, + "balance_loss_clip": 1.05041099, + "balance_loss_mlp": 1.02416062, + "epoch": 0.6188901581194012, + "flos": 20084766796800.0, + "grad_norm": 1.8432513470953755, + "language_loss": 0.74298048, + "learning_rate": 1.3395994337514218e-06, + "loss": 0.76508498, + "num_input_tokens_seen": 110963170, + "step": 5147, + "time_per_iteration": 2.6556155681610107 + }, + { + "auxiliary_loss_clip": 0.01155646, + "auxiliary_loss_mlp": 0.01028104, + "balance_loss_clip": 1.04474556, + "balance_loss_mlp": 1.01974785, + "epoch": 0.6190104010100402, + "flos": 25700728360320.0, + "grad_norm": 2.7022670265236615, + "language_loss": 0.78514111, + "learning_rate": 1.3388642040382725e-06, + "loss": 0.80697858, + "num_input_tokens_seen": 110983595, + "step": 5148, + "time_per_iteration": 2.8143234252929688 + }, + { + "auxiliary_loss_clip": 0.01148644, + "auxiliary_loss_mlp": 0.01024165, + "balance_loss_clip": 1.04169941, + "balance_loss_mlp": 1.01664925, + "epoch": 0.6191306439006794, + "flos": 30442372974720.0, + "grad_norm": 1.675879449414679, + "language_loss": 0.83943939, + "learning_rate": 1.3381290746188975e-06, + "loss": 0.86116743, + "num_input_tokens_seen": 111002965, + "step": 5149, + "time_per_iteration": 2.9188897609710693 + }, + { + "auxiliary_loss_clip": 0.01167796, + "auxiliary_loss_mlp": 0.01033283, + "balance_loss_clip": 1.05081391, + "balance_loss_mlp": 1.02495074, + "epoch": 0.6192508867913185, + "flos": 26685793918080.0, + "grad_norm": 1.7230863555135612, + "language_loss": 0.6708461, + "learning_rate": 1.3373940456048152e-06, + "loss": 0.69285685, + "num_input_tokens_seen": 111022990, + "step": 5150, + "time_per_iteration": 2.7629501819610596 + }, + { + "auxiliary_loss_clip": 0.01175913, + "auxiliary_loss_mlp": 0.01027111, + "balance_loss_clip": 1.05207109, + "balance_loss_mlp": 1.01991701, + "epoch": 0.6193711296819575, + "flos": 36722036090880.0, + "grad_norm": 1.532127816921188, + "language_loss": 0.59058964, + "learning_rate": 1.3366591171075299e-06, + "loss": 0.61261988, + "num_input_tokens_seen": 111046495, + "step": 5151, + "time_per_iteration": 2.796577215194702 + }, + { + "auxiliary_loss_clip": 0.01154467, + "auxiliary_loss_mlp": 0.01028537, + "balance_loss_clip": 1.04778123, + "balance_loss_mlp": 1.02045417, + "epoch": 0.6194913725725967, + "flos": 25192556697600.0, + "grad_norm": 1.9589561164254152, + "language_loss": 0.91061229, + "learning_rate": 1.335924289238529e-06, + "loss": 0.93244231, + "num_input_tokens_seen": 111065705, + "step": 5152, + "time_per_iteration": 2.8257663249969482 + }, + { + "auxiliary_loss_clip": 0.01164276, + "auxiliary_loss_mlp": 0.008865, + "balance_loss_clip": 1.0504446, + "balance_loss_mlp": 1.00076365, + "epoch": 0.6196116154632357, + "flos": 21178821196800.0, + "grad_norm": 1.526776188852923, + "language_loss": 0.76565552, + "learning_rate": 1.3351895621092859e-06, + "loss": 0.78616327, + "num_input_tokens_seen": 111086050, + "step": 5153, + "time_per_iteration": 2.8031351566314697 + }, + { + "auxiliary_loss_clip": 0.01090132, + "auxiliary_loss_mlp": 0.01034597, + "balance_loss_clip": 1.03028572, + "balance_loss_mlp": 1.02643096, + "epoch": 0.6197318583538748, + "flos": 16253744803200.0, + "grad_norm": 1.9895476062728548, + "language_loss": 0.76414299, + "learning_rate": 1.3344549358312567e-06, + "loss": 0.78539026, + "num_input_tokens_seen": 111104450, + "step": 5154, + "time_per_iteration": 3.043254852294922 + }, + { + "auxiliary_loss_clip": 0.01168303, + "auxiliary_loss_mlp": 0.01031906, + "balance_loss_clip": 1.04947388, + "balance_loss_mlp": 1.02358532, + "epoch": 0.619852101244514, + "flos": 24425612478720.0, + "grad_norm": 2.370833811984985, + "language_loss": 0.78221416, + "learning_rate": 1.3337204105158852e-06, + "loss": 0.80421627, + "num_input_tokens_seen": 111123320, + "step": 5155, + "time_per_iteration": 3.6956987380981445 + }, + { + "auxiliary_loss_clip": 0.01130342, + "auxiliary_loss_mlp": 0.01028475, + "balance_loss_clip": 1.03575087, + "balance_loss_mlp": 1.02028561, + "epoch": 0.619972344135153, + "flos": 16727298733440.0, + "grad_norm": 1.8871747134415588, + "language_loss": 0.72610956, + "learning_rate": 1.332985986274597e-06, + "loss": 0.74769771, + "num_input_tokens_seen": 111140950, + "step": 5156, + "time_per_iteration": 3.751706838607788 + }, + { + "auxiliary_loss_clip": 0.01121559, + "auxiliary_loss_mlp": 0.00885414, + "balance_loss_clip": 1.04352021, + "balance_loss_mlp": 1.00095582, + "epoch": 0.6200925870257921, + "flos": 12495190498560.0, + "grad_norm": 2.1385267614543824, + "language_loss": 0.75156486, + "learning_rate": 1.3322516632188047e-06, + "loss": 0.77163464, + "num_input_tokens_seen": 111157845, + "step": 5157, + "time_per_iteration": 3.8615732192993164 + }, + { + "auxiliary_loss_clip": 0.01143661, + "auxiliary_loss_mlp": 0.01028716, + "balance_loss_clip": 1.04473412, + "balance_loss_mlp": 1.02076435, + "epoch": 0.6202128299164312, + "flos": 26539350168960.0, + "grad_norm": 1.7023171288755923, + "language_loss": 0.66938412, + "learning_rate": 1.3315174414599045e-06, + "loss": 0.69110787, + "num_input_tokens_seen": 111179165, + "step": 5158, + "time_per_iteration": 2.921680212020874 + }, + { + "auxiliary_loss_clip": 0.01158379, + "auxiliary_loss_mlp": 0.01028129, + "balance_loss_clip": 1.04520273, + "balance_loss_mlp": 1.01966524, + "epoch": 0.6203330728070703, + "flos": 18770508069120.0, + "grad_norm": 2.0758276507537055, + "language_loss": 0.74973571, + "learning_rate": 1.3307833211092768e-06, + "loss": 0.77160084, + "num_input_tokens_seen": 111197830, + "step": 5159, + "time_per_iteration": 3.6153759956359863 + }, + { + "auxiliary_loss_clip": 0.01179288, + "auxiliary_loss_mlp": 0.01033626, + "balance_loss_clip": 1.05315685, + "balance_loss_mlp": 1.02528751, + "epoch": 0.6204533156977093, + "flos": 20629782835200.0, + "grad_norm": 1.5474972914732275, + "language_loss": 0.750696, + "learning_rate": 1.3300493022782873e-06, + "loss": 0.77282512, + "num_input_tokens_seen": 111218400, + "step": 5160, + "time_per_iteration": 2.7369089126586914 + }, + { + "auxiliary_loss_clip": 0.0112528, + "auxiliary_loss_mlp": 0.00886432, + "balance_loss_clip": 1.04264224, + "balance_loss_mlp": 1.00080419, + "epoch": 0.6205735585883485, + "flos": 17348050598400.0, + "grad_norm": 1.7762962488838159, + "language_loss": 0.72217768, + "learning_rate": 1.3293153850782855e-06, + "loss": 0.74229485, + "num_input_tokens_seen": 111236720, + "step": 5161, + "time_per_iteration": 2.816683530807495 + }, + { + "auxiliary_loss_clip": 0.01134262, + "auxiliary_loss_mlp": 0.01028417, + "balance_loss_clip": 1.0427984, + "balance_loss_mlp": 1.01922035, + "epoch": 0.6206938014789876, + "flos": 22965017742720.0, + "grad_norm": 1.8335002502291646, + "language_loss": 0.71270818, + "learning_rate": 1.3285815696206069e-06, + "loss": 0.73433495, + "num_input_tokens_seen": 111258265, + "step": 5162, + "time_per_iteration": 2.839899778366089 + }, + { + "auxiliary_loss_clip": 0.01146318, + "auxiliary_loss_mlp": 0.01027916, + "balance_loss_clip": 1.04225326, + "balance_loss_mlp": 1.01907623, + "epoch": 0.6208140443696266, + "flos": 23983192661760.0, + "grad_norm": 2.6382501383222454, + "language_loss": 0.77066517, + "learning_rate": 1.32784785601657e-06, + "loss": 0.79240751, + "num_input_tokens_seen": 111277675, + "step": 5163, + "time_per_iteration": 2.8342671394348145 + }, + { + "auxiliary_loss_clip": 0.01157279, + "auxiliary_loss_mlp": 0.01024915, + "balance_loss_clip": 1.04499948, + "balance_loss_mlp": 1.01610541, + "epoch": 0.6209342872602658, + "flos": 35077291303680.0, + "grad_norm": 1.8082862881217705, + "language_loss": 0.73685712, + "learning_rate": 1.3271142443774798e-06, + "loss": 0.75867903, + "num_input_tokens_seen": 111299910, + "step": 5164, + "time_per_iteration": 2.8624227046966553 + }, + { + "auxiliary_loss_clip": 0.01152636, + "auxiliary_loss_mlp": 0.01022915, + "balance_loss_clip": 1.04713225, + "balance_loss_mlp": 1.01499343, + "epoch": 0.6210545301509048, + "flos": 26979327861120.0, + "grad_norm": 1.725634087116362, + "language_loss": 0.81658375, + "learning_rate": 1.3263807348146228e-06, + "loss": 0.83833921, + "num_input_tokens_seen": 111319765, + "step": 5165, + "time_per_iteration": 2.8289129734039307 + }, + { + "auxiliary_loss_clip": 0.01151791, + "auxiliary_loss_mlp": 0.01033414, + "balance_loss_clip": 1.04296052, + "balance_loss_mlp": 1.0246346, + "epoch": 0.6211747730415439, + "flos": 33618240852480.0, + "grad_norm": 2.377337124429185, + "language_loss": 0.73441112, + "learning_rate": 1.3256473274392733e-06, + "loss": 0.75626314, + "num_input_tokens_seen": 111341110, + "step": 5166, + "time_per_iteration": 2.8285269737243652 + }, + { + "auxiliary_loss_clip": 0.01175152, + "auxiliary_loss_mlp": 0.01031574, + "balance_loss_clip": 1.05140734, + "balance_loss_mlp": 1.02333403, + "epoch": 0.6212950159321831, + "flos": 34167099646080.0, + "grad_norm": 1.8965539320685154, + "language_loss": 0.70064539, + "learning_rate": 1.3249140223626873e-06, + "loss": 0.7227127, + "num_input_tokens_seen": 111362730, + "step": 5167, + "time_per_iteration": 2.7961769104003906 + }, + { + "auxiliary_loss_clip": 0.01164558, + "auxiliary_loss_mlp": 0.01030469, + "balance_loss_clip": 1.05012286, + "balance_loss_mlp": 1.02232671, + "epoch": 0.6214152588228221, + "flos": 27965758135680.0, + "grad_norm": 1.8131140018100027, + "language_loss": 0.75529456, + "learning_rate": 1.3241808196961077e-06, + "loss": 0.77724487, + "num_input_tokens_seen": 111383855, + "step": 5168, + "time_per_iteration": 2.7392892837524414 + }, + { + "auxiliary_loss_clip": 0.01143361, + "auxiliary_loss_mlp": 0.01025195, + "balance_loss_clip": 1.0453409, + "balance_loss_mlp": 1.01785815, + "epoch": 0.6215355017134612, + "flos": 20230204965120.0, + "grad_norm": 1.74428946902028, + "language_loss": 0.70651174, + "learning_rate": 1.3234477195507608e-06, + "loss": 0.72819734, + "num_input_tokens_seen": 111402685, + "step": 5169, + "time_per_iteration": 2.871171712875366 + }, + { + "auxiliary_loss_clip": 0.01147361, + "auxiliary_loss_mlp": 0.01027573, + "balance_loss_clip": 1.0484879, + "balance_loss_mlp": 1.02003884, + "epoch": 0.6216557446041003, + "flos": 41428129219200.0, + "grad_norm": 2.050010090898102, + "language_loss": 0.62206757, + "learning_rate": 1.322714722037857e-06, + "loss": 0.64381689, + "num_input_tokens_seen": 111424130, + "step": 5170, + "time_per_iteration": 3.022178888320923 + }, + { + "auxiliary_loss_clip": 0.0115541, + "auxiliary_loss_mlp": 0.01037134, + "balance_loss_clip": 1.04626417, + "balance_loss_mlp": 1.02853298, + "epoch": 0.6217759874947394, + "flos": 27928770105600.0, + "grad_norm": 2.266879042334355, + "language_loss": 0.77089763, + "learning_rate": 1.321981827268591e-06, + "loss": 0.79282302, + "num_input_tokens_seen": 111444785, + "step": 5171, + "time_per_iteration": 3.0182700157165527 + }, + { + "auxiliary_loss_clip": 0.01156475, + "auxiliary_loss_mlp": 0.01026818, + "balance_loss_clip": 1.0454421, + "balance_loss_mlp": 1.0190872, + "epoch": 0.6218962303853784, + "flos": 21765673601280.0, + "grad_norm": 2.156246508705377, + "language_loss": 0.81481707, + "learning_rate": 1.3212490353541426e-06, + "loss": 0.83665001, + "num_input_tokens_seen": 111467045, + "step": 5172, + "time_per_iteration": 3.047499418258667 + }, + { + "auxiliary_loss_clip": 0.01178663, + "auxiliary_loss_mlp": 0.01034249, + "balance_loss_clip": 1.0522927, + "balance_loss_mlp": 1.02542794, + "epoch": 0.6220164732760175, + "flos": 21246260981760.0, + "grad_norm": 1.9025302530685833, + "language_loss": 0.80588269, + "learning_rate": 1.3205163464056762e-06, + "loss": 0.82801187, + "num_input_tokens_seen": 111483650, + "step": 5173, + "time_per_iteration": 2.7403883934020996 + }, + { + "auxiliary_loss_clip": 0.01163661, + "auxiliary_loss_mlp": 0.01022553, + "balance_loss_clip": 1.04823184, + "balance_loss_mlp": 1.01428032, + "epoch": 0.6221367161666567, + "flos": 26136360506880.0, + "grad_norm": 1.8720389866605498, + "language_loss": 0.72872514, + "learning_rate": 1.319783760534339e-06, + "loss": 0.75058722, + "num_input_tokens_seen": 111502895, + "step": 5174, + "time_per_iteration": 2.822693347930908 + }, + { + "auxiliary_loss_clip": 0.01167105, + "auxiliary_loss_mlp": 0.01034146, + "balance_loss_clip": 1.05090499, + "balance_loss_mlp": 1.02599251, + "epoch": 0.6222569590572957, + "flos": 16284196558080.0, + "grad_norm": 2.1671163199618477, + "language_loss": 0.75527263, + "learning_rate": 1.319051277851266e-06, + "loss": 0.7772851, + "num_input_tokens_seen": 111519180, + "step": 5175, + "time_per_iteration": 2.7322838306427 + }, + { + "auxiliary_loss_clip": 0.01160616, + "auxiliary_loss_mlp": 0.01029871, + "balance_loss_clip": 1.04828179, + "balance_loss_mlp": 1.02193153, + "epoch": 0.6223772019479348, + "flos": 18223840005120.0, + "grad_norm": 2.0014908695919, + "language_loss": 0.84096545, + "learning_rate": 1.3183188984675716e-06, + "loss": 0.86287034, + "num_input_tokens_seen": 111537545, + "step": 5176, + "time_per_iteration": 2.7356514930725098 + }, + { + "auxiliary_loss_clip": 0.01155484, + "auxiliary_loss_mlp": 0.01026831, + "balance_loss_clip": 1.04874265, + "balance_loss_mlp": 1.01864123, + "epoch": 0.6224974448385739, + "flos": 27489797994240.0, + "grad_norm": 2.4077838440213144, + "language_loss": 0.7134583, + "learning_rate": 1.3175866224943586e-06, + "loss": 0.73528147, + "num_input_tokens_seen": 111556265, + "step": 5177, + "time_per_iteration": 2.7951598167419434 + }, + { + "auxiliary_loss_clip": 0.01161559, + "auxiliary_loss_mlp": 0.01027369, + "balance_loss_clip": 1.04951882, + "balance_loss_mlp": 1.01864839, + "epoch": 0.622617687729213, + "flos": 19791951125760.0, + "grad_norm": 3.548697402261869, + "language_loss": 0.73608857, + "learning_rate": 1.316854450042712e-06, + "loss": 0.75797784, + "num_input_tokens_seen": 111574205, + "step": 5178, + "time_per_iteration": 2.735725164413452 + }, + { + "auxiliary_loss_clip": 0.01168897, + "auxiliary_loss_mlp": 0.01027873, + "balance_loss_clip": 1.0495888, + "balance_loss_mlp": 1.01980829, + "epoch": 0.622737930619852, + "flos": 23038886062080.0, + "grad_norm": 2.521206538753654, + "language_loss": 0.74406278, + "learning_rate": 1.3161223812237024e-06, + "loss": 0.76603043, + "num_input_tokens_seen": 111593560, + "step": 5179, + "time_per_iteration": 2.6916046142578125 + }, + { + "auxiliary_loss_clip": 0.01173073, + "auxiliary_loss_mlp": 0.01029648, + "balance_loss_clip": 1.04766881, + "balance_loss_mlp": 1.02163696, + "epoch": 0.6228581735104912, + "flos": 12634271959680.0, + "grad_norm": 2.713944004281933, + "language_loss": 0.84998429, + "learning_rate": 1.3153904161483842e-06, + "loss": 0.87201154, + "num_input_tokens_seen": 111608860, + "step": 5180, + "time_per_iteration": 2.576404571533203 + }, + { + "auxiliary_loss_clip": 0.01146442, + "auxiliary_loss_mlp": 0.01025394, + "balance_loss_clip": 1.04586983, + "balance_loss_mlp": 1.01695943, + "epoch": 0.6229784164011303, + "flos": 23802813538560.0, + "grad_norm": 1.9225965486608592, + "language_loss": 0.85586476, + "learning_rate": 1.3146585549277953e-06, + "loss": 0.87758315, + "num_input_tokens_seen": 111627500, + "step": 5181, + "time_per_iteration": 4.667169570922852 + }, + { + "auxiliary_loss_clip": 0.01161069, + "auxiliary_loss_mlp": 0.01029379, + "balance_loss_clip": 1.05057287, + "balance_loss_mlp": 1.02118921, + "epoch": 0.6230986592917693, + "flos": 22414219614720.0, + "grad_norm": 2.827745579828072, + "language_loss": 0.78145963, + "learning_rate": 1.3139267976729591e-06, + "loss": 0.80336416, + "num_input_tokens_seen": 111647690, + "step": 5182, + "time_per_iteration": 2.781644344329834 + }, + { + "auxiliary_loss_clip": 0.01170331, + "auxiliary_loss_mlp": 0.01031356, + "balance_loss_clip": 1.05208945, + "balance_loss_mlp": 1.0231663, + "epoch": 0.6232189021824085, + "flos": 34528217028480.0, + "grad_norm": 1.7113000896526809, + "language_loss": 0.71666217, + "learning_rate": 1.3131951444948815e-06, + "loss": 0.73867905, + "num_input_tokens_seen": 111667090, + "step": 5183, + "time_per_iteration": 3.762845993041992 + }, + { + "auxiliary_loss_clip": 0.01162974, + "auxiliary_loss_mlp": 0.01027658, + "balance_loss_clip": 1.05265069, + "balance_loss_mlp": 1.01961708, + "epoch": 0.6233391450730476, + "flos": 22237000888320.0, + "grad_norm": 1.7384447621947716, + "language_loss": 0.76033145, + "learning_rate": 1.3124635955045546e-06, + "loss": 0.78223777, + "num_input_tokens_seen": 111686905, + "step": 5184, + "time_per_iteration": 2.8203392028808594 + }, + { + "auxiliary_loss_clip": 0.01119468, + "auxiliary_loss_mlp": 0.00886459, + "balance_loss_clip": 1.04066014, + "balance_loss_mlp": 1.00071704, + "epoch": 0.6234593879636866, + "flos": 20332693445760.0, + "grad_norm": 1.7126247771321568, + "language_loss": 0.84272957, + "learning_rate": 1.3117321508129537e-06, + "loss": 0.8627888, + "num_input_tokens_seen": 111704985, + "step": 5185, + "time_per_iteration": 2.899507522583008 + }, + { + "auxiliary_loss_clip": 0.01157955, + "auxiliary_loss_mlp": 0.01026368, + "balance_loss_clip": 1.04816306, + "balance_loss_mlp": 1.01852369, + "epoch": 0.6235796308543258, + "flos": 20664903358080.0, + "grad_norm": 1.6105003901673691, + "language_loss": 0.76438987, + "learning_rate": 1.3110008105310388e-06, + "loss": 0.78623313, + "num_input_tokens_seen": 111724805, + "step": 5186, + "time_per_iteration": 3.5913922786712646 + }, + { + "auxiliary_loss_clip": 0.01177151, + "auxiliary_loss_mlp": 0.01035521, + "balance_loss_clip": 1.0492425, + "balance_loss_mlp": 1.02659774, + "epoch": 0.6236998737449648, + "flos": 26618641441920.0, + "grad_norm": 1.9195058736013906, + "language_loss": 0.78174782, + "learning_rate": 1.3102695747697526e-06, + "loss": 0.80387449, + "num_input_tokens_seen": 111747675, + "step": 5187, + "time_per_iteration": 2.7562992572784424 + }, + { + "auxiliary_loss_clip": 0.01129403, + "auxiliary_loss_mlp": 0.01035003, + "balance_loss_clip": 1.04528189, + "balance_loss_mlp": 1.02652121, + "epoch": 0.6238201166356039, + "flos": 12674599954560.0, + "grad_norm": 2.081873713358476, + "language_loss": 0.90672588, + "learning_rate": 1.3095384436400237e-06, + "loss": 0.92836994, + "num_input_tokens_seen": 111759205, + "step": 5188, + "time_per_iteration": 2.798616647720337 + }, + { + "auxiliary_loss_clip": 0.0115689, + "auxiliary_loss_mlp": 0.01031001, + "balance_loss_clip": 1.04742527, + "balance_loss_mlp": 1.02259672, + "epoch": 0.623940359526243, + "flos": 10452160730880.0, + "grad_norm": 2.114038248226741, + "language_loss": 0.82482171, + "learning_rate": 1.3088074172527633e-06, + "loss": 0.84670067, + "num_input_tokens_seen": 111776335, + "step": 5189, + "time_per_iteration": 2.7169790267944336 + }, + { + "auxiliary_loss_clip": 0.01160814, + "auxiliary_loss_mlp": 0.01026644, + "balance_loss_clip": 1.04628181, + "balance_loss_mlp": 1.01822186, + "epoch": 0.6240606024168821, + "flos": 29059525226880.0, + "grad_norm": 1.827605555194224, + "language_loss": 0.71596754, + "learning_rate": 1.3080764957188684e-06, + "loss": 0.73784208, + "num_input_tokens_seen": 111796580, + "step": 5190, + "time_per_iteration": 2.831500291824341 + }, + { + "auxiliary_loss_clip": 0.01132643, + "auxiliary_loss_mlp": 0.01032856, + "balance_loss_clip": 1.04226756, + "balance_loss_mlp": 1.02489293, + "epoch": 0.6241808453075212, + "flos": 22018089450240.0, + "grad_norm": 1.8585155413698435, + "language_loss": 0.70833373, + "learning_rate": 1.3073456791492192e-06, + "loss": 0.72998869, + "num_input_tokens_seen": 111816290, + "step": 5191, + "time_per_iteration": 2.837655782699585 + }, + { + "auxiliary_loss_clip": 0.01157294, + "auxiliary_loss_mlp": 0.0102759, + "balance_loss_clip": 1.04522383, + "balance_loss_mlp": 1.01947761, + "epoch": 0.6243010881981603, + "flos": 21138708683520.0, + "grad_norm": 1.8767167989447202, + "language_loss": 0.78327411, + "learning_rate": 1.3066149676546801e-06, + "loss": 0.80512297, + "num_input_tokens_seen": 111834470, + "step": 5192, + "time_per_iteration": 2.790552854537964 + }, + { + "auxiliary_loss_clip": 0.01152567, + "auxiliary_loss_mlp": 0.01027255, + "balance_loss_clip": 1.04884982, + "balance_loss_mlp": 1.01944685, + "epoch": 0.6244213310887994, + "flos": 22344948236160.0, + "grad_norm": 1.6647356502344033, + "language_loss": 0.66181123, + "learning_rate": 1.3058843613460985e-06, + "loss": 0.68360949, + "num_input_tokens_seen": 111852410, + "step": 5193, + "time_per_iteration": 2.7837722301483154 + }, + { + "auxiliary_loss_clip": 0.01149334, + "auxiliary_loss_mlp": 0.01031095, + "balance_loss_clip": 1.0467211, + "balance_loss_mlp": 1.02265501, + "epoch": 0.6245415739794384, + "flos": 15231978524160.0, + "grad_norm": 2.9270331549488606, + "language_loss": 0.74320638, + "learning_rate": 1.3051538603343075e-06, + "loss": 0.76501071, + "num_input_tokens_seen": 111870340, + "step": 5194, + "time_per_iteration": 2.910975456237793 + }, + { + "auxiliary_loss_clip": 0.01167341, + "auxiliary_loss_mlp": 0.01027939, + "balance_loss_clip": 1.05168986, + "balance_loss_mlp": 1.02026784, + "epoch": 0.6246618168700776, + "flos": 18879891960960.0, + "grad_norm": 2.0142583497717963, + "language_loss": 0.67893851, + "learning_rate": 1.3044234647301235e-06, + "loss": 0.70089126, + "num_input_tokens_seen": 111888365, + "step": 5195, + "time_per_iteration": 2.9475789070129395 + }, + { + "auxiliary_loss_clip": 0.01160709, + "auxiliary_loss_mlp": 0.01028927, + "balance_loss_clip": 1.04878163, + "balance_loss_mlp": 1.02107704, + "epoch": 0.6247820597607167, + "flos": 14319201087360.0, + "grad_norm": 1.839022598955724, + "language_loss": 0.72694105, + "learning_rate": 1.303693174644347e-06, + "loss": 0.74883747, + "num_input_tokens_seen": 111905840, + "step": 5196, + "time_per_iteration": 2.9591500759124756 + }, + { + "auxiliary_loss_clip": 0.01149187, + "auxiliary_loss_mlp": 0.0103093, + "balance_loss_clip": 1.04371226, + "balance_loss_mlp": 1.022663, + "epoch": 0.6249023026513557, + "flos": 22637979388800.0, + "grad_norm": 1.8891454683519733, + "language_loss": 0.8039217, + "learning_rate": 1.3029629901877625e-06, + "loss": 0.82572293, + "num_input_tokens_seen": 111925215, + "step": 5197, + "time_per_iteration": 2.74119234085083 + }, + { + "auxiliary_loss_clip": 0.01168197, + "auxiliary_loss_mlp": 0.01028223, + "balance_loss_clip": 1.05287886, + "balance_loss_mlp": 1.01956797, + "epoch": 0.6250225455419949, + "flos": 20266690204800.0, + "grad_norm": 2.969883527721747, + "language_loss": 0.77325749, + "learning_rate": 1.3022329114711376e-06, + "loss": 0.79522169, + "num_input_tokens_seen": 111943925, + "step": 5198, + "time_per_iteration": 2.7808356285095215 + }, + { + "auxiliary_loss_clip": 0.01154064, + "auxiliary_loss_mlp": 0.01026279, + "balance_loss_clip": 1.0476414, + "balance_loss_mlp": 1.01788044, + "epoch": 0.6251427884326339, + "flos": 23437853400960.0, + "grad_norm": 1.9404567964705706, + "language_loss": 0.69591671, + "learning_rate": 1.3015029386052256e-06, + "loss": 0.71772015, + "num_input_tokens_seen": 111964095, + "step": 5199, + "time_per_iteration": 2.7450153827667236 + }, + { + "auxiliary_loss_clip": 0.01149794, + "auxiliary_loss_mlp": 0.01028846, + "balance_loss_clip": 1.04732454, + "balance_loss_mlp": 1.02060258, + "epoch": 0.625263031323273, + "flos": 31723055464320.0, + "grad_norm": 1.8819760419292002, + "language_loss": 0.7275604, + "learning_rate": 1.3007730717007622e-06, + "loss": 0.74934685, + "num_input_tokens_seen": 111984910, + "step": 5200, + "time_per_iteration": 2.862957239151001 + }, + { + "auxiliary_loss_clip": 0.01178977, + "auxiliary_loss_mlp": 0.01032201, + "balance_loss_clip": 1.05268085, + "balance_loss_mlp": 1.02356434, + "epoch": 0.6253832742139122, + "flos": 24134341092480.0, + "grad_norm": 1.9383559804628232, + "language_loss": 0.75451028, + "learning_rate": 1.3000433108684676e-06, + "loss": 0.77662206, + "num_input_tokens_seen": 112005410, + "step": 5201, + "time_per_iteration": 2.6601064205169678 + }, + { + "auxiliary_loss_clip": 0.01164436, + "auxiliary_loss_mlp": 0.01029312, + "balance_loss_clip": 1.05113983, + "balance_loss_mlp": 1.02124763, + "epoch": 0.6255035171045512, + "flos": 27668812400640.0, + "grad_norm": 2.543848803504099, + "language_loss": 0.80398083, + "learning_rate": 1.2993136562190467e-06, + "loss": 0.82591832, + "num_input_tokens_seen": 112024530, + "step": 5202, + "time_per_iteration": 2.6091086864471436 + }, + { + "auxiliary_loss_clip": 0.01160702, + "auxiliary_loss_mlp": 0.01029604, + "balance_loss_clip": 1.04794669, + "balance_loss_mlp": 1.02125394, + "epoch": 0.6256237599951903, + "flos": 20227798753920.0, + "grad_norm": 1.6854908898209984, + "language_loss": 0.70367181, + "learning_rate": 1.2985841078631871e-06, + "loss": 0.72557485, + "num_input_tokens_seen": 112043850, + "step": 5203, + "time_per_iteration": 2.6203908920288086 + }, + { + "auxiliary_loss_clip": 0.01126707, + "auxiliary_loss_mlp": 0.01028304, + "balance_loss_clip": 1.04061675, + "balance_loss_mlp": 1.01963115, + "epoch": 0.6257440028858293, + "flos": 24170574936960.0, + "grad_norm": 2.1619646040847655, + "language_loss": 0.78014934, + "learning_rate": 1.2978546659115608e-06, + "loss": 0.8016994, + "num_input_tokens_seen": 112061930, + "step": 5204, + "time_per_iteration": 2.724414110183716 + }, + { + "auxiliary_loss_clip": 0.01161915, + "auxiliary_loss_mlp": 0.01030184, + "balance_loss_clip": 1.04960632, + "balance_loss_mlp": 1.02189624, + "epoch": 0.6258642457764685, + "flos": 15851940289920.0, + "grad_norm": 1.8611575091777794, + "language_loss": 0.8552134, + "learning_rate": 1.2971253304748228e-06, + "loss": 0.87713444, + "num_input_tokens_seen": 112079645, + "step": 5205, + "time_per_iteration": 2.6617801189422607 + }, + { + "auxiliary_loss_clip": 0.01169256, + "auxiliary_loss_mlp": 0.0103295, + "balance_loss_clip": 1.05120754, + "balance_loss_mlp": 1.02462292, + "epoch": 0.6259844886671075, + "flos": 11911354836480.0, + "grad_norm": 2.0023599189399066, + "language_loss": 0.75054002, + "learning_rate": 1.296396101663614e-06, + "loss": 0.77256215, + "num_input_tokens_seen": 112096205, + "step": 5206, + "time_per_iteration": 2.56260085105896 + }, + { + "auxiliary_loss_clip": 0.01169821, + "auxiliary_loss_mlp": 0.01029181, + "balance_loss_clip": 1.04944634, + "balance_loss_mlp": 1.02102077, + "epoch": 0.6261047315577466, + "flos": 15887958652800.0, + "grad_norm": 2.110132436681832, + "language_loss": 0.84030533, + "learning_rate": 1.2956669795885565e-06, + "loss": 0.86229539, + "num_input_tokens_seen": 112112835, + "step": 5207, + "time_per_iteration": 4.559883117675781 + }, + { + "auxiliary_loss_clip": 0.01141808, + "auxiliary_loss_mlp": 0.01036249, + "balance_loss_clip": 1.04755712, + "balance_loss_mlp": 1.02734983, + "epoch": 0.6262249744483858, + "flos": 31248926916480.0, + "grad_norm": 2.0707030955019774, + "language_loss": 0.68170953, + "learning_rate": 1.294937964360259e-06, + "loss": 0.70349008, + "num_input_tokens_seen": 112133105, + "step": 5208, + "time_per_iteration": 2.9727015495300293 + }, + { + "auxiliary_loss_clip": 0.01155147, + "auxiliary_loss_mlp": 0.010283, + "balance_loss_clip": 1.04710865, + "balance_loss_mlp": 1.01968694, + "epoch": 0.6263452173390248, + "flos": 27198598435200.0, + "grad_norm": 2.336858785712902, + "language_loss": 0.71633136, + "learning_rate": 1.2942090560893108e-06, + "loss": 0.73816586, + "num_input_tokens_seen": 112152510, + "step": 5209, + "time_per_iteration": 4.005906343460083 + }, + { + "auxiliary_loss_clip": 0.01176175, + "auxiliary_loss_mlp": 0.01035544, + "balance_loss_clip": 1.05159998, + "balance_loss_mlp": 1.02743804, + "epoch": 0.6264654602296639, + "flos": 37342069683840.0, + "grad_norm": 1.7628226021641729, + "language_loss": 0.6046676, + "learning_rate": 1.2934802548862882e-06, + "loss": 0.6267848, + "num_input_tokens_seen": 112175295, + "step": 5210, + "time_per_iteration": 2.8321120738983154 + }, + { + "auxiliary_loss_clip": 0.01149648, + "auxiliary_loss_mlp": 0.01030154, + "balance_loss_clip": 1.04463506, + "balance_loss_mlp": 1.02205324, + "epoch": 0.626585703120303, + "flos": 14756952136320.0, + "grad_norm": 1.798994203434956, + "language_loss": 0.8270551, + "learning_rate": 1.292751560861749e-06, + "loss": 0.84885311, + "num_input_tokens_seen": 112190200, + "step": 5211, + "time_per_iteration": 2.6958606243133545 + }, + { + "auxiliary_loss_clip": 0.01179795, + "auxiliary_loss_mlp": 0.01031404, + "balance_loss_clip": 1.05195975, + "balance_loss_mlp": 1.02275515, + "epoch": 0.6267059460109421, + "flos": 22347318533760.0, + "grad_norm": 1.7990209998476487, + "language_loss": 0.79564679, + "learning_rate": 1.2920229741262354e-06, + "loss": 0.8177588, + "num_input_tokens_seen": 112208205, + "step": 5212, + "time_per_iteration": 3.596820116043091 + }, + { + "auxiliary_loss_clip": 0.01159432, + "auxiliary_loss_mlp": 0.01029072, + "balance_loss_clip": 1.04821265, + "balance_loss_mlp": 1.02064371, + "epoch": 0.6268261889015811, + "flos": 17748813617280.0, + "grad_norm": 2.155062110042451, + "language_loss": 0.75090116, + "learning_rate": 1.2912944947902739e-06, + "loss": 0.7727862, + "num_input_tokens_seen": 112224690, + "step": 5213, + "time_per_iteration": 2.7864389419555664 + }, + { + "auxiliary_loss_clip": 0.01162061, + "auxiliary_loss_mlp": 0.01029708, + "balance_loss_clip": 1.04734015, + "balance_loss_mlp": 1.02123797, + "epoch": 0.6269464317922203, + "flos": 32846484211200.0, + "grad_norm": 32.22317069754642, + "language_loss": 0.71947789, + "learning_rate": 1.2905661229643742e-06, + "loss": 0.74139559, + "num_input_tokens_seen": 112244450, + "step": 5214, + "time_per_iteration": 2.827869176864624 + }, + { + "auxiliary_loss_clip": 0.01176083, + "auxiliary_loss_mlp": 0.01033263, + "balance_loss_clip": 1.0494082, + "balance_loss_mlp": 1.02527571, + "epoch": 0.6270666746828594, + "flos": 17929192740480.0, + "grad_norm": 2.405339315299411, + "language_loss": 0.84168798, + "learning_rate": 1.2898378587590299e-06, + "loss": 0.86378145, + "num_input_tokens_seen": 112261050, + "step": 5215, + "time_per_iteration": 2.741128921508789 + }, + { + "auxiliary_loss_clip": 0.01162063, + "auxiliary_loss_mlp": 0.01028709, + "balance_loss_clip": 1.04936874, + "balance_loss_mlp": 1.02032852, + "epoch": 0.6271869175734984, + "flos": 17457326749440.0, + "grad_norm": 1.9737277260185213, + "language_loss": 0.87508225, + "learning_rate": 1.2891097022847173e-06, + "loss": 0.89698994, + "num_input_tokens_seen": 112278395, + "step": 5216, + "time_per_iteration": 2.7269458770751953 + }, + { + "auxiliary_loss_clip": 0.01160335, + "auxiliary_loss_mlp": 0.01033373, + "balance_loss_clip": 1.05007684, + "balance_loss_mlp": 1.02451563, + "epoch": 0.6273071604641376, + "flos": 26868615166080.0, + "grad_norm": 2.176702876791785, + "language_loss": 0.66428906, + "learning_rate": 1.2883816536518978e-06, + "loss": 0.68622613, + "num_input_tokens_seen": 112299535, + "step": 5217, + "time_per_iteration": 2.7376832962036133 + }, + { + "auxiliary_loss_clip": 0.0116437, + "auxiliary_loss_mlp": 0.01023696, + "balance_loss_clip": 1.04915869, + "balance_loss_mlp": 1.01607275, + "epoch": 0.6274274033547766, + "flos": 26062384446720.0, + "grad_norm": 1.983337986221293, + "language_loss": 0.81958437, + "learning_rate": 1.2876537129710155e-06, + "loss": 0.84146506, + "num_input_tokens_seen": 112317265, + "step": 5218, + "time_per_iteration": 2.7081687450408936 + }, + { + "auxiliary_loss_clip": 0.01154453, + "auxiliary_loss_mlp": 0.0102505, + "balance_loss_clip": 1.04972672, + "balance_loss_mlp": 1.01638913, + "epoch": 0.6275476462454157, + "flos": 20266259241600.0, + "grad_norm": 1.968470219816445, + "language_loss": 0.75582486, + "learning_rate": 1.286925880352499e-06, + "loss": 0.7776199, + "num_input_tokens_seen": 112336125, + "step": 5219, + "time_per_iteration": 2.7579071521759033 + }, + { + "auxiliary_loss_clip": 0.0115239, + "auxiliary_loss_mlp": 0.01029636, + "balance_loss_clip": 1.04708076, + "balance_loss_mlp": 1.02166104, + "epoch": 0.6276678891360549, + "flos": 26320402817280.0, + "grad_norm": 1.8663290005985957, + "language_loss": 0.7125333, + "learning_rate": 1.2861981559067592e-06, + "loss": 0.73435354, + "num_input_tokens_seen": 112356730, + "step": 5220, + "time_per_iteration": 2.9001643657684326 + }, + { + "auxiliary_loss_clip": 0.01128315, + "auxiliary_loss_mlp": 0.01027839, + "balance_loss_clip": 1.04207993, + "balance_loss_mlp": 1.01945221, + "epoch": 0.6277881320266939, + "flos": 13912512324480.0, + "grad_norm": 1.8572552227501227, + "language_loss": 0.80307961, + "learning_rate": 1.2854705397441917e-06, + "loss": 0.82464117, + "num_input_tokens_seen": 112372270, + "step": 5221, + "time_per_iteration": 2.8110744953155518 + }, + { + "auxiliary_loss_clip": 0.01140822, + "auxiliary_loss_mlp": 0.01028759, + "balance_loss_clip": 1.04269755, + "balance_loss_mlp": 1.02061081, + "epoch": 0.627908374917333, + "flos": 27048922462080.0, + "grad_norm": 2.116422764410176, + "language_loss": 0.77964211, + "learning_rate": 1.2847430319751747e-06, + "loss": 0.80133796, + "num_input_tokens_seen": 112390365, + "step": 5222, + "time_per_iteration": 2.854255199432373 + }, + { + "auxiliary_loss_clip": 0.01164976, + "auxiliary_loss_mlp": 0.01022485, + "balance_loss_clip": 1.05144429, + "balance_loss_mlp": 1.01411414, + "epoch": 0.6280286178079721, + "flos": 23769201386880.0, + "grad_norm": 2.2114157990882632, + "language_loss": 0.67664689, + "learning_rate": 1.2840156327100712e-06, + "loss": 0.69852155, + "num_input_tokens_seen": 112407490, + "step": 5223, + "time_per_iteration": 2.7565014362335205 + }, + { + "auxiliary_loss_clip": 0.01175149, + "auxiliary_loss_mlp": 0.01025275, + "balance_loss_clip": 1.05175531, + "balance_loss_mlp": 1.01763368, + "epoch": 0.6281488606986112, + "flos": 26359150613760.0, + "grad_norm": 1.7334145394477276, + "language_loss": 0.72603154, + "learning_rate": 1.2832883420592272e-06, + "loss": 0.74803585, + "num_input_tokens_seen": 112426385, + "step": 5224, + "time_per_iteration": 2.6713783740997314 + }, + { + "auxiliary_loss_clip": 0.01151821, + "auxiliary_loss_mlp": 0.01030262, + "balance_loss_clip": 1.04685783, + "balance_loss_mlp": 1.02170873, + "epoch": 0.6282691035892503, + "flos": 36137194848000.0, + "grad_norm": 1.9754829807873278, + "language_loss": 0.64057368, + "learning_rate": 1.282561160132972e-06, + "loss": 0.66239452, + "num_input_tokens_seen": 112446905, + "step": 5225, + "time_per_iteration": 2.843830108642578 + }, + { + "auxiliary_loss_clip": 0.01161545, + "auxiliary_loss_mlp": 0.01031292, + "balance_loss_clip": 1.04473853, + "balance_loss_mlp": 1.02250671, + "epoch": 0.6283893464798894, + "flos": 26537231266560.0, + "grad_norm": 1.6620417646277235, + "language_loss": 0.80819625, + "learning_rate": 1.2818340870416186e-06, + "loss": 0.83012462, + "num_input_tokens_seen": 112468040, + "step": 5226, + "time_per_iteration": 2.7910492420196533 + }, + { + "auxiliary_loss_clip": 0.01156949, + "auxiliary_loss_mlp": 0.01031758, + "balance_loss_clip": 1.04631114, + "balance_loss_mlp": 1.02354431, + "epoch": 0.6285095893705285, + "flos": 22237216369920.0, + "grad_norm": 1.8689429676140759, + "language_loss": 0.75574744, + "learning_rate": 1.2811071228954626e-06, + "loss": 0.77763444, + "num_input_tokens_seen": 112486675, + "step": 5227, + "time_per_iteration": 2.8490893840789795 + }, + { + "auxiliary_loss_clip": 0.01158755, + "auxiliary_loss_mlp": 0.01029689, + "balance_loss_clip": 1.05079651, + "balance_loss_mlp": 1.02155888, + "epoch": 0.6286298322611675, + "flos": 26542259170560.0, + "grad_norm": 2.7887883760766434, + "language_loss": 0.80848503, + "learning_rate": 1.2803802678047846e-06, + "loss": 0.83036947, + "num_input_tokens_seen": 112506825, + "step": 5228, + "time_per_iteration": 2.7673728466033936 + }, + { + "auxiliary_loss_clip": 0.01160965, + "auxiliary_loss_mlp": 0.01025382, + "balance_loss_clip": 1.05080175, + "balance_loss_mlp": 1.01701951, + "epoch": 0.6287500751518067, + "flos": 21795227516160.0, + "grad_norm": 7.704022894128315, + "language_loss": 0.7414366, + "learning_rate": 1.279653521879848e-06, + "loss": 0.76330006, + "num_input_tokens_seen": 112526890, + "step": 5229, + "time_per_iteration": 2.836944103240967 + }, + { + "auxiliary_loss_clip": 0.01108863, + "auxiliary_loss_mlp": 0.01024943, + "balance_loss_clip": 1.04133916, + "balance_loss_mlp": 1.01730216, + "epoch": 0.6288703180424458, + "flos": 20009605587840.0, + "grad_norm": 1.8570733881232124, + "language_loss": 0.83369535, + "learning_rate": 1.2789268852308997e-06, + "loss": 0.8550334, + "num_input_tokens_seen": 112542100, + "step": 5230, + "time_per_iteration": 2.878920555114746 + }, + { + "auxiliary_loss_clip": 0.01160296, + "auxiliary_loss_mlp": 0.01026424, + "balance_loss_clip": 1.04900944, + "balance_loss_mlp": 1.0178231, + "epoch": 0.6289905609330848, + "flos": 22124923476480.0, + "grad_norm": 1.8598094444770918, + "language_loss": 0.70669639, + "learning_rate": 1.2782003579681688e-06, + "loss": 0.72856355, + "num_input_tokens_seen": 112561630, + "step": 5231, + "time_per_iteration": 2.783762216567993 + }, + { + "auxiliary_loss_clip": 0.01181072, + "auxiliary_loss_mlp": 0.01026504, + "balance_loss_clip": 1.05382848, + "balance_loss_mlp": 1.01806402, + "epoch": 0.629110803823724, + "flos": 25518481729920.0, + "grad_norm": 1.6101943134436627, + "language_loss": 0.74380034, + "learning_rate": 1.2774739402018701e-06, + "loss": 0.76587611, + "num_input_tokens_seen": 112582465, + "step": 5232, + "time_per_iteration": 3.677225351333618 + }, + { + "auxiliary_loss_clip": 0.01166317, + "auxiliary_loss_mlp": 0.01028041, + "balance_loss_clip": 1.05141068, + "balance_loss_mlp": 1.01920211, + "epoch": 0.629231046714363, + "flos": 20886616056960.0, + "grad_norm": 1.744429811296908, + "language_loss": 0.73265982, + "learning_rate": 1.2767476320422002e-06, + "loss": 0.75460339, + "num_input_tokens_seen": 112602390, + "step": 5233, + "time_per_iteration": 3.687952995300293 + }, + { + "auxiliary_loss_clip": 0.01052853, + "auxiliary_loss_mlp": 0.01000706, + "balance_loss_clip": 1.01199603, + "balance_loss_mlp": 0.99907321, + "epoch": 0.6293512896050021, + "flos": 65050027908480.0, + "grad_norm": 0.695085446303002, + "language_loss": 0.57236004, + "learning_rate": 1.2760214335993392e-06, + "loss": 0.59289563, + "num_input_tokens_seen": 112669035, + "step": 5234, + "time_per_iteration": 3.4894609451293945 + }, + { + "auxiliary_loss_clip": 0.01158823, + "auxiliary_loss_mlp": 0.01034168, + "balance_loss_clip": 1.04694533, + "balance_loss_mlp": 1.02658594, + "epoch": 0.6294715324956413, + "flos": 34677857088000.0, + "grad_norm": 1.9807992815558042, + "language_loss": 0.582901, + "learning_rate": 1.2752953449834514e-06, + "loss": 0.60483086, + "num_input_tokens_seen": 112691485, + "step": 5235, + "time_per_iteration": 3.769813299179077 + }, + { + "auxiliary_loss_clip": 0.0117576, + "auxiliary_loss_mlp": 0.01029252, + "balance_loss_clip": 1.05088043, + "balance_loss_mlp": 1.0218854, + "epoch": 0.6295917753862803, + "flos": 22784207656320.0, + "grad_norm": 1.7496368732335694, + "language_loss": 0.80475771, + "learning_rate": 1.2745693663046836e-06, + "loss": 0.82680774, + "num_input_tokens_seen": 112710555, + "step": 5236, + "time_per_iteration": 2.672722578048706 + }, + { + "auxiliary_loss_clip": 0.01160341, + "auxiliary_loss_mlp": 0.01026797, + "balance_loss_clip": 1.04873633, + "balance_loss_mlp": 1.01886964, + "epoch": 0.6297120182769194, + "flos": 20850454039680.0, + "grad_norm": 1.8428106130922075, + "language_loss": 0.80681753, + "learning_rate": 1.2738434976731662e-06, + "loss": 0.82868886, + "num_input_tokens_seen": 112728740, + "step": 5237, + "time_per_iteration": 2.840090751647949 + }, + { + "auxiliary_loss_clip": 0.01156124, + "auxiliary_loss_mlp": 0.01030362, + "balance_loss_clip": 1.04794788, + "balance_loss_mlp": 1.02237844, + "epoch": 0.6298322611675584, + "flos": 19497662997120.0, + "grad_norm": 1.522694941865016, + "language_loss": 0.75002277, + "learning_rate": 1.2731177391990125e-06, + "loss": 0.77188766, + "num_input_tokens_seen": 112748665, + "step": 5238, + "time_per_iteration": 3.7450835704803467 + }, + { + "auxiliary_loss_clip": 0.01158118, + "auxiliary_loss_mlp": 0.01023822, + "balance_loss_clip": 1.04640293, + "balance_loss_mlp": 1.0158236, + "epoch": 0.6299525040581976, + "flos": 12604466649600.0, + "grad_norm": 2.0963341515043443, + "language_loss": 0.81766057, + "learning_rate": 1.2723920909923203e-06, + "loss": 0.83947998, + "num_input_tokens_seen": 112764410, + "step": 5239, + "time_per_iteration": 2.7426917552948 + }, + { + "auxiliary_loss_clip": 0.01066334, + "auxiliary_loss_mlp": 0.01000438, + "balance_loss_clip": 1.0117625, + "balance_loss_mlp": 0.99895996, + "epoch": 0.6300727469488366, + "flos": 57725685636480.0, + "grad_norm": 0.8750892967766596, + "language_loss": 0.60430479, + "learning_rate": 1.2716665531631688e-06, + "loss": 0.62497258, + "num_input_tokens_seen": 112818695, + "step": 5240, + "time_per_iteration": 3.2583115100860596 + }, + { + "auxiliary_loss_clip": 0.01163694, + "auxiliary_loss_mlp": 0.01026287, + "balance_loss_clip": 1.04779935, + "balance_loss_mlp": 1.01794279, + "epoch": 0.6301929898394757, + "flos": 22527302607360.0, + "grad_norm": 1.6953608938314657, + "language_loss": 0.77032912, + "learning_rate": 1.270941125821623e-06, + "loss": 0.79222888, + "num_input_tokens_seen": 112839120, + "step": 5241, + "time_per_iteration": 2.7985715866088867 + }, + { + "auxiliary_loss_clip": 0.01160147, + "auxiliary_loss_mlp": 0.01029168, + "balance_loss_clip": 1.04528856, + "balance_loss_mlp": 1.0210911, + "epoch": 0.6303132327301149, + "flos": 28293550675200.0, + "grad_norm": 1.6693125699566416, + "language_loss": 0.75516343, + "learning_rate": 1.2702158090777278e-06, + "loss": 0.77705657, + "num_input_tokens_seen": 112860210, + "step": 5242, + "time_per_iteration": 2.766014337539673 + }, + { + "auxiliary_loss_clip": 0.01140122, + "auxiliary_loss_mlp": 0.01032634, + "balance_loss_clip": 1.04290795, + "balance_loss_mlp": 1.024122, + "epoch": 0.6304334756207539, + "flos": 25264521596160.0, + "grad_norm": 2.0331267576332426, + "language_loss": 0.74980491, + "learning_rate": 1.2694906030415148e-06, + "loss": 0.77153254, + "num_input_tokens_seen": 112877955, + "step": 5243, + "time_per_iteration": 2.8862617015838623 + }, + { + "auxiliary_loss_clip": 0.01167816, + "auxiliary_loss_mlp": 0.0103011, + "balance_loss_clip": 1.04947793, + "balance_loss_mlp": 1.0214262, + "epoch": 0.630553718511393, + "flos": 18033548728320.0, + "grad_norm": 2.3949087112824774, + "language_loss": 0.821805, + "learning_rate": 1.2687655078229958e-06, + "loss": 0.84378427, + "num_input_tokens_seen": 112892285, + "step": 5244, + "time_per_iteration": 2.7816450595855713 + }, + { + "auxiliary_loss_clip": 0.01156036, + "auxiliary_loss_mlp": 0.01026407, + "balance_loss_clip": 1.04910862, + "balance_loss_mlp": 1.01809764, + "epoch": 0.6306739614020321, + "flos": 27304103658240.0, + "grad_norm": 8.143542152252556, + "language_loss": 0.692083, + "learning_rate": 1.2680405235321678e-06, + "loss": 0.71390748, + "num_input_tokens_seen": 112913620, + "step": 5245, + "time_per_iteration": 2.8502955436706543 + }, + { + "auxiliary_loss_clip": 0.01164041, + "auxiliary_loss_mlp": 0.00886568, + "balance_loss_clip": 1.05363107, + "balance_loss_mlp": 1.0004425, + "epoch": 0.6307942042926712, + "flos": 15341434243200.0, + "grad_norm": 2.01308118960165, + "language_loss": 0.78769815, + "learning_rate": 1.267315650279011e-06, + "loss": 0.80820429, + "num_input_tokens_seen": 112932090, + "step": 5246, + "time_per_iteration": 2.715618371963501 + }, + { + "auxiliary_loss_clip": 0.01141098, + "auxiliary_loss_mlp": 0.01026581, + "balance_loss_clip": 1.04718781, + "balance_loss_mlp": 1.01808167, + "epoch": 0.6309144471833102, + "flos": 19606400444160.0, + "grad_norm": 2.13226168089014, + "language_loss": 0.74146128, + "learning_rate": 1.2665908881734874e-06, + "loss": 0.76313806, + "num_input_tokens_seen": 112950925, + "step": 5247, + "time_per_iteration": 2.7913103103637695 + }, + { + "auxiliary_loss_clip": 0.01166118, + "auxiliary_loss_mlp": 0.01029073, + "balance_loss_clip": 1.04964614, + "balance_loss_mlp": 1.02114224, + "epoch": 0.6310346900739494, + "flos": 17493345112320.0, + "grad_norm": 2.156274376790593, + "language_loss": 0.85247928, + "learning_rate": 1.2658662373255432e-06, + "loss": 0.87443119, + "num_input_tokens_seen": 112969315, + "step": 5248, + "time_per_iteration": 2.713200330734253 + }, + { + "auxiliary_loss_clip": 0.01052743, + "auxiliary_loss_mlp": 0.01003787, + "balance_loss_clip": 1.01167107, + "balance_loss_mlp": 1.00221348, + "epoch": 0.6311549329645885, + "flos": 55070164131840.0, + "grad_norm": 0.9503736172289511, + "language_loss": 0.52306902, + "learning_rate": 1.2651416978451063e-06, + "loss": 0.5436343, + "num_input_tokens_seen": 113034700, + "step": 5249, + "time_per_iteration": 3.4408507347106934 + }, + { + "auxiliary_loss_clip": 0.01180687, + "auxiliary_loss_mlp": 0.01029844, + "balance_loss_clip": 1.05339932, + "balance_loss_mlp": 1.02160954, + "epoch": 0.6312751758552275, + "flos": 41902545075840.0, + "grad_norm": 1.7482147487145316, + "language_loss": 0.65465832, + "learning_rate": 1.2644172698420903e-06, + "loss": 0.67676365, + "num_input_tokens_seen": 113056805, + "step": 5250, + "time_per_iteration": 2.8808484077453613 + }, + { + "auxiliary_loss_clip": 0.01150632, + "auxiliary_loss_mlp": 0.01030021, + "balance_loss_clip": 1.04761803, + "balance_loss_mlp": 1.02112246, + "epoch": 0.6313954187458667, + "flos": 19646800266240.0, + "grad_norm": 3.0180611324946165, + "language_loss": 0.84774572, + "learning_rate": 1.2636929534263892e-06, + "loss": 0.86955225, + "num_input_tokens_seen": 113075790, + "step": 5251, + "time_per_iteration": 2.8805277347564697 + }, + { + "auxiliary_loss_clip": 0.01149035, + "auxiliary_loss_mlp": 0.01032455, + "balance_loss_clip": 1.04291296, + "balance_loss_mlp": 1.02394927, + "epoch": 0.6315156616365057, + "flos": 22894273906560.0, + "grad_norm": 1.7044199317562014, + "language_loss": 0.77479905, + "learning_rate": 1.2629687487078821e-06, + "loss": 0.79661393, + "num_input_tokens_seen": 113094600, + "step": 5252, + "time_per_iteration": 2.844247341156006 + }, + { + "auxiliary_loss_clip": 0.01167211, + "auxiliary_loss_mlp": 0.01035316, + "balance_loss_clip": 1.04632497, + "balance_loss_mlp": 1.02647638, + "epoch": 0.6316359045271448, + "flos": 23726251699200.0, + "grad_norm": 2.0692845330596996, + "language_loss": 0.76060015, + "learning_rate": 1.2622446557964293e-06, + "loss": 0.78262538, + "num_input_tokens_seen": 113112605, + "step": 5253, + "time_per_iteration": 2.704660177230835 + }, + { + "auxiliary_loss_clip": 0.01153212, + "auxiliary_loss_mlp": 0.01032941, + "balance_loss_clip": 1.04186666, + "balance_loss_mlp": 1.02498949, + "epoch": 0.631756147417784, + "flos": 33108417164160.0, + "grad_norm": 1.690151324114497, + "language_loss": 0.71217263, + "learning_rate": 1.261520674801876e-06, + "loss": 0.73403418, + "num_input_tokens_seen": 113133200, + "step": 5254, + "time_per_iteration": 2.8590707778930664 + }, + { + "auxiliary_loss_clip": 0.01156029, + "auxiliary_loss_mlp": 0.01032862, + "balance_loss_clip": 1.05079508, + "balance_loss_mlp": 1.02420187, + "epoch": 0.631876390308423, + "flos": 31248424126080.0, + "grad_norm": 1.917750081354955, + "language_loss": 0.72034544, + "learning_rate": 1.2607968058340488e-06, + "loss": 0.74223435, + "num_input_tokens_seen": 113152895, + "step": 5255, + "time_per_iteration": 2.7826948165893555 + }, + { + "auxiliary_loss_clip": 0.01153402, + "auxiliary_loss_mlp": 0.01027564, + "balance_loss_clip": 1.04643536, + "balance_loss_mlp": 1.01930904, + "epoch": 0.6319966331990621, + "flos": 24681152810880.0, + "grad_norm": 1.730479293307759, + "language_loss": 0.73520321, + "learning_rate": 1.2600730490027583e-06, + "loss": 0.7570129, + "num_input_tokens_seen": 113173135, + "step": 5256, + "time_per_iteration": 2.748415231704712 + }, + { + "auxiliary_loss_clip": 0.0114603, + "auxiliary_loss_mlp": 0.01026268, + "balance_loss_clip": 1.04601645, + "balance_loss_mlp": 1.01827824, + "epoch": 0.6321168760897012, + "flos": 17491764913920.0, + "grad_norm": 1.945706922696344, + "language_loss": 0.8026852, + "learning_rate": 1.2593494044177984e-06, + "loss": 0.82440817, + "num_input_tokens_seen": 113191440, + "step": 5257, + "time_per_iteration": 2.79069185256958 + }, + { + "auxiliary_loss_clip": 0.01180848, + "auxiliary_loss_mlp": 0.01026218, + "balance_loss_clip": 1.0501231, + "balance_loss_mlp": 1.01736712, + "epoch": 0.6322371189803403, + "flos": 18295373940480.0, + "grad_norm": 2.374154443515548, + "language_loss": 0.80580622, + "learning_rate": 1.2586258721889448e-06, + "loss": 0.82787687, + "num_input_tokens_seen": 113208790, + "step": 5258, + "time_per_iteration": 3.5279042720794678 + }, + { + "auxiliary_loss_clip": 0.0112906, + "auxiliary_loss_mlp": 0.01029677, + "balance_loss_clip": 1.04318666, + "balance_loss_mlp": 1.02143919, + "epoch": 0.6323573618709794, + "flos": 20157270399360.0, + "grad_norm": 1.8421306496776024, + "language_loss": 0.81747717, + "learning_rate": 1.2579024524259573e-06, + "loss": 0.83906448, + "num_input_tokens_seen": 113225050, + "step": 5259, + "time_per_iteration": 3.812258243560791 + }, + { + "auxiliary_loss_clip": 0.01150289, + "auxiliary_loss_mlp": 0.01031405, + "balance_loss_clip": 1.04294395, + "balance_loss_mlp": 1.02318597, + "epoch": 0.6324776047616185, + "flos": 20042391726720.0, + "grad_norm": 2.290512579729305, + "language_loss": 0.90924948, + "learning_rate": 1.2571791452385768e-06, + "loss": 0.93106639, + "num_input_tokens_seen": 113242315, + "step": 5260, + "time_per_iteration": 2.7613532543182373 + }, + { + "auxiliary_loss_clip": 0.01157246, + "auxiliary_loss_mlp": 0.0102517, + "balance_loss_clip": 1.04847181, + "balance_loss_mlp": 1.01734376, + "epoch": 0.6325978476522576, + "flos": 30848235724800.0, + "grad_norm": 1.664781480487319, + "language_loss": 0.76911116, + "learning_rate": 1.2564559507365301e-06, + "loss": 0.79093528, + "num_input_tokens_seen": 113264720, + "step": 5261, + "time_per_iteration": 3.8907828330993652 + }, + { + "auxiliary_loss_clip": 0.01163027, + "auxiliary_loss_mlp": 0.01031229, + "balance_loss_clip": 1.05049205, + "balance_loss_mlp": 1.02216291, + "epoch": 0.6327180905428966, + "flos": 24535104111360.0, + "grad_norm": 3.120707209390585, + "language_loss": 0.78739291, + "learning_rate": 1.2557328690295244e-06, + "loss": 0.80933547, + "num_input_tokens_seen": 113282910, + "step": 5262, + "time_per_iteration": 2.8495662212371826 + }, + { + "auxiliary_loss_clip": 0.0114591, + "auxiliary_loss_mlp": 0.01032593, + "balance_loss_clip": 1.0481391, + "balance_loss_mlp": 1.02455258, + "epoch": 0.6328383334335358, + "flos": 21575274583680.0, + "grad_norm": 1.718781535853956, + "language_loss": 0.7669791, + "learning_rate": 1.255009900227251e-06, + "loss": 0.78876412, + "num_input_tokens_seen": 113301935, + "step": 5263, + "time_per_iteration": 2.877336263656616 + }, + { + "auxiliary_loss_clip": 0.01174836, + "auxiliary_loss_mlp": 0.01033107, + "balance_loss_clip": 1.05144072, + "balance_loss_mlp": 1.02529311, + "epoch": 0.6329585763241748, + "flos": 22929861306240.0, + "grad_norm": 1.7315190644427403, + "language_loss": 0.79191959, + "learning_rate": 1.254287044439383e-06, + "loss": 0.813999, + "num_input_tokens_seen": 113321540, + "step": 5264, + "time_per_iteration": 3.6281135082244873 + }, + { + "auxiliary_loss_clip": 0.01064546, + "auxiliary_loss_mlp": 0.01002049, + "balance_loss_clip": 1.01033461, + "balance_loss_mlp": 1.00063634, + "epoch": 0.6330788192148139, + "flos": 70936897847040.0, + "grad_norm": 0.7722137222511444, + "language_loss": 0.54466021, + "learning_rate": 1.2535643017755776e-06, + "loss": 0.56532621, + "num_input_tokens_seen": 113383730, + "step": 5265, + "time_per_iteration": 3.270925760269165 + }, + { + "auxiliary_loss_clip": 0.01141125, + "auxiliary_loss_mlp": 0.01027503, + "balance_loss_clip": 1.0436976, + "balance_loss_mlp": 1.01959968, + "epoch": 0.6331990621054531, + "flos": 21244501215360.0, + "grad_norm": 2.341910217848512, + "language_loss": 0.71853447, + "learning_rate": 1.2528416723454737e-06, + "loss": 0.74022073, + "num_input_tokens_seen": 113400400, + "step": 5266, + "time_per_iteration": 2.769346237182617 + }, + { + "auxiliary_loss_clip": 0.01173541, + "auxiliary_loss_mlp": 0.01024567, + "balance_loss_clip": 1.05169594, + "balance_loss_mlp": 1.01705718, + "epoch": 0.6333193049960921, + "flos": 34459412526720.0, + "grad_norm": 1.4867889523987647, + "language_loss": 0.70999008, + "learning_rate": 1.2521191562586945e-06, + "loss": 0.73197114, + "num_input_tokens_seen": 113424050, + "step": 5267, + "time_per_iteration": 2.785490036010742 + }, + { + "auxiliary_loss_clip": 0.01174419, + "auxiliary_loss_mlp": 0.00885925, + "balance_loss_clip": 1.04963136, + "balance_loss_mlp": 1.00047445, + "epoch": 0.6334395478867312, + "flos": 18329883932160.0, + "grad_norm": 2.256805260858315, + "language_loss": 0.76923203, + "learning_rate": 1.2513967536248445e-06, + "loss": 0.78983545, + "num_input_tokens_seen": 113440370, + "step": 5268, + "time_per_iteration": 2.6865384578704834 + }, + { + "auxiliary_loss_clip": 0.01163547, + "auxiliary_loss_mlp": 0.01030918, + "balance_loss_clip": 1.04920721, + "balance_loss_mlp": 1.02245712, + "epoch": 0.6335597907773702, + "flos": 23623152687360.0, + "grad_norm": 2.1444787124142497, + "language_loss": 0.80999553, + "learning_rate": 1.2506744645535117e-06, + "loss": 0.83194017, + "num_input_tokens_seen": 113460800, + "step": 5269, + "time_per_iteration": 2.7110981941223145 + }, + { + "auxiliary_loss_clip": 0.0114717, + "auxiliary_loss_mlp": 0.01032105, + "balance_loss_clip": 1.04065132, + "balance_loss_mlp": 1.02345049, + "epoch": 0.6336800336680094, + "flos": 22710913954560.0, + "grad_norm": 1.8772903812593136, + "language_loss": 0.60086417, + "learning_rate": 1.249952289154267e-06, + "loss": 0.62265694, + "num_input_tokens_seen": 113480840, + "step": 5270, + "time_per_iteration": 2.7307474613189697 + }, + { + "auxiliary_loss_clip": 0.01112952, + "auxiliary_loss_mlp": 0.01025745, + "balance_loss_clip": 1.04005873, + "balance_loss_mlp": 1.01770782, + "epoch": 0.6338002765586485, + "flos": 23622757637760.0, + "grad_norm": 1.8479061267550903, + "language_loss": 0.76464748, + "learning_rate": 1.2492302275366635e-06, + "loss": 0.78603452, + "num_input_tokens_seen": 113500515, + "step": 5271, + "time_per_iteration": 2.8618288040161133 + }, + { + "auxiliary_loss_clip": 0.01161608, + "auxiliary_loss_mlp": 0.01031788, + "balance_loss_clip": 1.047526, + "balance_loss_mlp": 1.02194107, + "epoch": 0.6339205194492875, + "flos": 26505450708480.0, + "grad_norm": 2.5566196236276477, + "language_loss": 0.65781522, + "learning_rate": 1.2485082798102377e-06, + "loss": 0.67974913, + "num_input_tokens_seen": 113520930, + "step": 5272, + "time_per_iteration": 2.8113605976104736 + }, + { + "auxiliary_loss_clip": 0.01151712, + "auxiliary_loss_mlp": 0.01025722, + "balance_loss_clip": 1.04414988, + "balance_loss_mlp": 1.01706791, + "epoch": 0.6340407623399267, + "flos": 18544306170240.0, + "grad_norm": 2.2187215947325636, + "language_loss": 0.68527853, + "learning_rate": 1.2477864460845084e-06, + "loss": 0.70705283, + "num_input_tokens_seen": 113537330, + "step": 5273, + "time_per_iteration": 2.756566047668457 + }, + { + "auxiliary_loss_clip": 0.01155111, + "auxiliary_loss_mlp": 0.01033024, + "balance_loss_clip": 1.04600644, + "balance_loss_mlp": 1.0233624, + "epoch": 0.6341610052305657, + "flos": 17712579772800.0, + "grad_norm": 2.536149976473281, + "language_loss": 0.73601079, + "learning_rate": 1.2470647264689776e-06, + "loss": 0.75789213, + "num_input_tokens_seen": 113555810, + "step": 5274, + "time_per_iteration": 2.776313304901123 + }, + { + "auxiliary_loss_clip": 0.01130206, + "auxiliary_loss_mlp": 0.01027808, + "balance_loss_clip": 1.03982043, + "balance_loss_mlp": 1.01933861, + "epoch": 0.6342812481212048, + "flos": 23587026583680.0, + "grad_norm": 3.5386881305716105, + "language_loss": 0.71127963, + "learning_rate": 1.2463431210731282e-06, + "loss": 0.73285973, + "num_input_tokens_seen": 113575395, + "step": 5275, + "time_per_iteration": 2.939124822616577 + }, + { + "auxiliary_loss_clip": 0.01139691, + "auxiliary_loss_mlp": 0.0103103, + "balance_loss_clip": 1.042732, + "balance_loss_mlp": 1.02213132, + "epoch": 0.634401491011844, + "flos": 17821927751040.0, + "grad_norm": 6.616327073379655, + "language_loss": 0.75689095, + "learning_rate": 1.2456216300064289e-06, + "loss": 0.77859819, + "num_input_tokens_seen": 113592945, + "step": 5276, + "time_per_iteration": 2.841041326522827 + }, + { + "auxiliary_loss_clip": 0.01148228, + "auxiliary_loss_mlp": 0.01030628, + "balance_loss_clip": 1.04356229, + "balance_loss_mlp": 1.02219355, + "epoch": 0.634521733902483, + "flos": 21358158825600.0, + "grad_norm": 1.8534390684367954, + "language_loss": 0.78374904, + "learning_rate": 1.244900253378328e-06, + "loss": 0.80553758, + "num_input_tokens_seen": 113613000, + "step": 5277, + "time_per_iteration": 2.796757936477661 + }, + { + "auxiliary_loss_clip": 0.01104854, + "auxiliary_loss_mlp": 0.01025667, + "balance_loss_clip": 1.04435194, + "balance_loss_mlp": 1.01792717, + "epoch": 0.6346419767931221, + "flos": 16545052103040.0, + "grad_norm": 2.1055093382150796, + "language_loss": 0.69435471, + "learning_rate": 1.2441789912982583e-06, + "loss": 0.71565992, + "num_input_tokens_seen": 113630085, + "step": 5278, + "time_per_iteration": 2.901512384414673 + }, + { + "auxiliary_loss_clip": 0.01169396, + "auxiliary_loss_mlp": 0.01031616, + "balance_loss_clip": 1.05014813, + "balance_loss_mlp": 1.02295566, + "epoch": 0.6347622196837612, + "flos": 24350989973760.0, + "grad_norm": 2.5372521075512036, + "language_loss": 0.64828396, + "learning_rate": 1.2434578438756346e-06, + "loss": 0.67029411, + "num_input_tokens_seen": 113650515, + "step": 5279, + "time_per_iteration": 2.774995803833008 + }, + { + "auxiliary_loss_clip": 0.01167384, + "auxiliary_loss_mlp": 0.01026539, + "balance_loss_clip": 1.04720616, + "balance_loss_mlp": 1.018242, + "epoch": 0.6348824625744003, + "flos": 64523178195840.0, + "grad_norm": 1.7994398407834218, + "language_loss": 0.7803483, + "learning_rate": 1.242736811219855e-06, + "loss": 0.80228752, + "num_input_tokens_seen": 113676475, + "step": 5280, + "time_per_iteration": 3.0812292098999023 + }, + { + "auxiliary_loss_clip": 0.01159531, + "auxiliary_loss_mlp": 0.01027121, + "balance_loss_clip": 1.04766619, + "balance_loss_mlp": 1.01926184, + "epoch": 0.6350027054650393, + "flos": 28622133313920.0, + "grad_norm": 1.7630464641422097, + "language_loss": 0.81613392, + "learning_rate": 1.2420158934402988e-06, + "loss": 0.83800042, + "num_input_tokens_seen": 113697090, + "step": 5281, + "time_per_iteration": 2.79262113571167 + }, + { + "auxiliary_loss_clip": 0.01131859, + "auxiliary_loss_mlp": 0.01032001, + "balance_loss_clip": 1.04081917, + "balance_loss_mlp": 1.02323949, + "epoch": 0.6351229483556785, + "flos": 23002544476800.0, + "grad_norm": 2.067491125149172, + "language_loss": 0.84745538, + "learning_rate": 1.2412950906463286e-06, + "loss": 0.86909401, + "num_input_tokens_seen": 113714395, + "step": 5282, + "time_per_iteration": 2.8329994678497314 + }, + { + "auxiliary_loss_clip": 0.01134489, + "auxiliary_loss_mlp": 0.0102823, + "balance_loss_clip": 1.04297614, + "balance_loss_mlp": 1.02045155, + "epoch": 0.6352431912463176, + "flos": 21939300967680.0, + "grad_norm": 2.060239854476515, + "language_loss": 0.8992793, + "learning_rate": 1.2405744029472902e-06, + "loss": 0.92090648, + "num_input_tokens_seen": 113733880, + "step": 5283, + "time_per_iteration": 2.9240314960479736 + }, + { + "auxiliary_loss_clip": 0.01152887, + "auxiliary_loss_mlp": 0.01028647, + "balance_loss_clip": 1.04574418, + "balance_loss_mlp": 1.02086878, + "epoch": 0.6353634341369566, + "flos": 13735257684480.0, + "grad_norm": 2.1991079766071553, + "language_loss": 0.75858033, + "learning_rate": 1.2398538304525108e-06, + "loss": 0.78039563, + "num_input_tokens_seen": 113752505, + "step": 5284, + "time_per_iteration": 3.71647572517395 + }, + { + "auxiliary_loss_clip": 0.0114139, + "auxiliary_loss_mlp": 0.01031477, + "balance_loss_clip": 1.04582334, + "balance_loss_mlp": 1.02090907, + "epoch": 0.6354836770275958, + "flos": 19316170552320.0, + "grad_norm": 2.778742746644367, + "language_loss": 0.75975072, + "learning_rate": 1.2391333732713016e-06, + "loss": 0.78147942, + "num_input_tokens_seen": 113770310, + "step": 5285, + "time_per_iteration": 3.8082900047302246 + }, + { + "auxiliary_loss_clip": 0.01144171, + "auxiliary_loss_mlp": 0.01028587, + "balance_loss_clip": 1.04332364, + "balance_loss_mlp": 1.02032554, + "epoch": 0.6356039199182348, + "flos": 21613375935360.0, + "grad_norm": 1.8906404682160665, + "language_loss": 0.78379154, + "learning_rate": 1.2384130315129543e-06, + "loss": 0.80551904, + "num_input_tokens_seen": 113788635, + "step": 5286, + "time_per_iteration": 2.8150229454040527 + }, + { + "auxiliary_loss_clip": 0.01107156, + "auxiliary_loss_mlp": 0.01030781, + "balance_loss_clip": 1.03742552, + "balance_loss_mlp": 1.02228689, + "epoch": 0.6357241628088739, + "flos": 18111978074880.0, + "grad_norm": 3.4711076782829258, + "language_loss": 0.73349988, + "learning_rate": 1.2376928052867447e-06, + "loss": 0.7548793, + "num_input_tokens_seen": 113807755, + "step": 5287, + "time_per_iteration": 4.064208984375 + }, + { + "auxiliary_loss_clip": 0.0115733, + "auxiliary_loss_mlp": 0.01032938, + "balance_loss_clip": 1.04855716, + "balance_loss_mlp": 1.02532029, + "epoch": 0.6358444056995131, + "flos": 24935256599040.0, + "grad_norm": 2.0403323385390277, + "language_loss": 0.77790689, + "learning_rate": 1.2369726947019299e-06, + "loss": 0.79980958, + "num_input_tokens_seen": 113828230, + "step": 5288, + "time_per_iteration": 3.324993848800659 + }, + { + "auxiliary_loss_clip": 0.01164353, + "auxiliary_loss_mlp": 0.01033806, + "balance_loss_clip": 1.04661489, + "balance_loss_mlp": 1.02578282, + "epoch": 0.6359646485901521, + "flos": 23293348986240.0, + "grad_norm": 2.7915080820059406, + "language_loss": 0.66997343, + "learning_rate": 1.2362526998677511e-06, + "loss": 0.69195497, + "num_input_tokens_seen": 113844595, + "step": 5289, + "time_per_iteration": 3.381802558898926 + }, + { + "auxiliary_loss_clip": 0.01158982, + "auxiliary_loss_mlp": 0.01031031, + "balance_loss_clip": 1.04707587, + "balance_loss_mlp": 1.02371693, + "epoch": 0.6360848914807912, + "flos": 20887442069760.0, + "grad_norm": 1.98313561790643, + "language_loss": 0.84589577, + "learning_rate": 1.2355328208934301e-06, + "loss": 0.86779588, + "num_input_tokens_seen": 113863470, + "step": 5290, + "time_per_iteration": 3.822021007537842 + }, + { + "auxiliary_loss_clip": 0.01165769, + "auxiliary_loss_mlp": 0.00885798, + "balance_loss_clip": 1.04722273, + "balance_loss_mlp": 1.00052071, + "epoch": 0.6362051343714303, + "flos": 18479775386880.0, + "grad_norm": 1.694840442217359, + "language_loss": 0.72423995, + "learning_rate": 1.2348130578881728e-06, + "loss": 0.74475563, + "num_input_tokens_seen": 113881690, + "step": 5291, + "time_per_iteration": 2.791527509689331 + }, + { + "auxiliary_loss_clip": 0.01177482, + "auxiliary_loss_mlp": 0.01028178, + "balance_loss_clip": 1.04953253, + "balance_loss_mlp": 1.02009559, + "epoch": 0.6363253772620694, + "flos": 24389594115840.0, + "grad_norm": 2.2342171532380517, + "language_loss": 0.76031041, + "learning_rate": 1.2340934109611664e-06, + "loss": 0.78236705, + "num_input_tokens_seen": 113902450, + "step": 5292, + "time_per_iteration": 2.7334461212158203 + }, + { + "auxiliary_loss_clip": 0.01161571, + "auxiliary_loss_mlp": 0.01027112, + "balance_loss_clip": 1.0491035, + "balance_loss_mlp": 1.01789093, + "epoch": 0.6364456201527084, + "flos": 25958243940480.0, + "grad_norm": 3.2573995942179366, + "language_loss": 0.68914932, + "learning_rate": 1.2333738802215798e-06, + "loss": 0.71103609, + "num_input_tokens_seen": 113922670, + "step": 5293, + "time_per_iteration": 2.851097583770752 + }, + { + "auxiliary_loss_clip": 0.01131815, + "auxiliary_loss_mlp": 0.01026742, + "balance_loss_clip": 1.04108787, + "balance_loss_mlp": 1.01806378, + "epoch": 0.6365658630433476, + "flos": 20740711011840.0, + "grad_norm": 2.189753191558886, + "language_loss": 0.80710888, + "learning_rate": 1.2326544657785668e-06, + "loss": 0.82869446, + "num_input_tokens_seen": 113942360, + "step": 5294, + "time_per_iteration": 2.857290029525757 + }, + { + "auxiliary_loss_clip": 0.01136495, + "auxiliary_loss_mlp": 0.01032709, + "balance_loss_clip": 1.04257154, + "balance_loss_mlp": 1.02417922, + "epoch": 0.6366861059339867, + "flos": 21434146047360.0, + "grad_norm": 2.262454282745199, + "language_loss": 0.74637306, + "learning_rate": 1.2319351677412608e-06, + "loss": 0.76806509, + "num_input_tokens_seen": 113959405, + "step": 5295, + "time_per_iteration": 2.814190626144409 + }, + { + "auxiliary_loss_clip": 0.01156607, + "auxiliary_loss_mlp": 0.01025832, + "balance_loss_clip": 1.04752755, + "balance_loss_mlp": 1.01674843, + "epoch": 0.6368063488246257, + "flos": 22267093507200.0, + "grad_norm": 3.822289391385106, + "language_loss": 0.74381924, + "learning_rate": 1.2312159862187796e-06, + "loss": 0.7656436, + "num_input_tokens_seen": 113977815, + "step": 5296, + "time_per_iteration": 2.80273175239563 + }, + { + "auxiliary_loss_clip": 0.01179894, + "auxiliary_loss_mlp": 0.01026684, + "balance_loss_clip": 1.05219436, + "balance_loss_mlp": 1.01844633, + "epoch": 0.6369265917152649, + "flos": 22420719976320.0, + "grad_norm": 1.6422570900455813, + "language_loss": 0.76038873, + "learning_rate": 1.2304969213202217e-06, + "loss": 0.78245449, + "num_input_tokens_seen": 113999075, + "step": 5297, + "time_per_iteration": 2.794584035873413 + }, + { + "auxiliary_loss_clip": 0.01150429, + "auxiliary_loss_mlp": 0.01024583, + "balance_loss_clip": 1.04489434, + "balance_loss_mlp": 1.016608, + "epoch": 0.6370468346059039, + "flos": 24718176754560.0, + "grad_norm": 2.1586051009834937, + "language_loss": 0.79326636, + "learning_rate": 1.2297779731546692e-06, + "loss": 0.81501651, + "num_input_tokens_seen": 114018170, + "step": 5298, + "time_per_iteration": 2.833811044692993 + }, + { + "auxiliary_loss_clip": 0.0115538, + "auxiliary_loss_mlp": 0.01031655, + "balance_loss_clip": 1.04930615, + "balance_loss_mlp": 1.02270222, + "epoch": 0.637167077496543, + "flos": 25296589463040.0, + "grad_norm": 1.858837989409181, + "language_loss": 0.77763301, + "learning_rate": 1.2290591418311853e-06, + "loss": 0.79950333, + "num_input_tokens_seen": 114035565, + "step": 5299, + "time_per_iteration": 2.8164775371551514 + }, + { + "auxiliary_loss_clip": 0.0116239, + "auxiliary_loss_mlp": 0.01029664, + "balance_loss_clip": 1.04894233, + "balance_loss_mlp": 1.02139711, + "epoch": 0.637287320387182, + "flos": 27671110871040.0, + "grad_norm": 1.6808586739273228, + "language_loss": 0.71981984, + "learning_rate": 1.2283404274588172e-06, + "loss": 0.74174041, + "num_input_tokens_seen": 114054510, + "step": 5300, + "time_per_iteration": 2.8211135864257812 + }, + { + "auxiliary_loss_clip": 0.01017551, + "auxiliary_loss_mlp": 0.01008397, + "balance_loss_clip": 1.00891256, + "balance_loss_mlp": 1.00689542, + "epoch": 0.6374075632778212, + "flos": 63173406873600.0, + "grad_norm": 0.7419200613094503, + "language_loss": 0.52843785, + "learning_rate": 1.227621830146592e-06, + "loss": 0.54869735, + "num_input_tokens_seen": 114109875, + "step": 5301, + "time_per_iteration": 3.3341410160064697 + }, + { + "auxiliary_loss_clip": 0.01151106, + "auxiliary_loss_mlp": 0.01026952, + "balance_loss_clip": 1.04876518, + "balance_loss_mlp": 1.01887524, + "epoch": 0.6375278061684603, + "flos": 25558127366400.0, + "grad_norm": 2.0992088596244303, + "language_loss": 0.79114175, + "learning_rate": 1.2269033500035217e-06, + "loss": 0.81292224, + "num_input_tokens_seen": 114130010, + "step": 5302, + "time_per_iteration": 3.409787178039551 + }, + { + "auxiliary_loss_clip": 0.01140267, + "auxiliary_loss_mlp": 0.01031791, + "balance_loss_clip": 1.04712868, + "balance_loss_mlp": 1.02357769, + "epoch": 0.6376480490590993, + "flos": 25666362023040.0, + "grad_norm": 1.8914139343151528, + "language_loss": 0.73412019, + "learning_rate": 1.2261849871385988e-06, + "loss": 0.75584078, + "num_input_tokens_seen": 114151115, + "step": 5303, + "time_per_iteration": 2.940035581588745 + }, + { + "auxiliary_loss_clip": 0.01176676, + "auxiliary_loss_mlp": 0.0102315, + "balance_loss_clip": 1.04921687, + "balance_loss_mlp": 1.01465011, + "epoch": 0.6377682919497385, + "flos": 31537684350720.0, + "grad_norm": 1.9511589793959736, + "language_loss": 0.62379712, + "learning_rate": 1.2254667416607972e-06, + "loss": 0.6457954, + "num_input_tokens_seen": 114172715, + "step": 5304, + "time_per_iteration": 2.7710020542144775 + }, + { + "auxiliary_loss_clip": 0.01165065, + "auxiliary_loss_mlp": 0.01029249, + "balance_loss_clip": 1.04934669, + "balance_loss_mlp": 1.02135181, + "epoch": 0.6378885348403776, + "flos": 23039209284480.0, + "grad_norm": 2.098264804753152, + "language_loss": 0.82934147, + "learning_rate": 1.2247486136790756e-06, + "loss": 0.85128462, + "num_input_tokens_seen": 114192195, + "step": 5305, + "time_per_iteration": 2.7960033416748047 + }, + { + "auxiliary_loss_clip": 0.0117069, + "auxiliary_loss_mlp": 0.01029845, + "balance_loss_clip": 1.05178499, + "balance_loss_mlp": 1.02205503, + "epoch": 0.6380087777310166, + "flos": 18697070712960.0, + "grad_norm": 2.179191632039183, + "language_loss": 0.8029902, + "learning_rate": 1.2240306033023726e-06, + "loss": 0.82499564, + "num_input_tokens_seen": 114210020, + "step": 5306, + "time_per_iteration": 2.6868607997894287 + }, + { + "auxiliary_loss_clip": 0.01147185, + "auxiliary_loss_mlp": 0.01029513, + "balance_loss_clip": 1.04104304, + "balance_loss_mlp": 1.02124548, + "epoch": 0.6381290206216558, + "flos": 23331558078720.0, + "grad_norm": 1.652423893885559, + "language_loss": 0.71927071, + "learning_rate": 1.223312710639611e-06, + "loss": 0.74103767, + "num_input_tokens_seen": 114228740, + "step": 5307, + "time_per_iteration": 2.771071434020996 + }, + { + "auxiliary_loss_clip": 0.01154628, + "auxiliary_loss_mlp": 0.01028811, + "balance_loss_clip": 1.04671419, + "balance_loss_mlp": 1.02016246, + "epoch": 0.6382492635122948, + "flos": 18880466578560.0, + "grad_norm": 2.021962571682219, + "language_loss": 0.87008017, + "learning_rate": 1.2225949357996928e-06, + "loss": 0.89191455, + "num_input_tokens_seen": 114246865, + "step": 5308, + "time_per_iteration": 2.7919743061065674 + }, + { + "auxiliary_loss_clip": 0.01162146, + "auxiliary_loss_mlp": 0.01025258, + "balance_loss_clip": 1.0496912, + "balance_loss_mlp": 1.01666903, + "epoch": 0.6383695064029339, + "flos": 27819134818560.0, + "grad_norm": 1.5859832879133626, + "language_loss": 0.80722737, + "learning_rate": 1.221877278891505e-06, + "loss": 0.82910138, + "num_input_tokens_seen": 114266120, + "step": 5309, + "time_per_iteration": 2.9034318923950195 + }, + { + "auxiliary_loss_clip": 0.01165571, + "auxiliary_loss_mlp": 0.01034112, + "balance_loss_clip": 1.04966927, + "balance_loss_mlp": 1.02530253, + "epoch": 0.638489749293573, + "flos": 26395635853440.0, + "grad_norm": 3.305557522766332, + "language_loss": 0.71644145, + "learning_rate": 1.221159740023915e-06, + "loss": 0.73843831, + "num_input_tokens_seen": 114285950, + "step": 5310, + "time_per_iteration": 3.872593879699707 + }, + { + "auxiliary_loss_clip": 0.01155344, + "auxiliary_loss_mlp": 0.00886422, + "balance_loss_clip": 1.04737329, + "balance_loss_mlp": 1.00051332, + "epoch": 0.6386099921842121, + "flos": 23988328306560.0, + "grad_norm": 2.2689514245072537, + "language_loss": 0.72639918, + "learning_rate": 1.2204423193057735e-06, + "loss": 0.74681687, + "num_input_tokens_seen": 114304780, + "step": 5311, + "time_per_iteration": 4.0356056690216064 + }, + { + "auxiliary_loss_clip": 0.01052944, + "auxiliary_loss_mlp": 0.01000729, + "balance_loss_clip": 1.01082897, + "balance_loss_mlp": 0.99925691, + "epoch": 0.6387302350748512, + "flos": 71731169337600.0, + "grad_norm": 0.8516958800533712, + "language_loss": 0.63347769, + "learning_rate": 1.2197250168459122e-06, + "loss": 0.65401447, + "num_input_tokens_seen": 114361180, + "step": 5312, + "time_per_iteration": 3.41454815864563 + }, + { + "auxiliary_loss_clip": 0.01169231, + "auxiliary_loss_mlp": 0.01030309, + "balance_loss_clip": 1.04919815, + "balance_loss_mlp": 1.02192307, + "epoch": 0.6388504779654903, + "flos": 14535778141440.0, + "grad_norm": 1.7430592748898805, + "language_loss": 0.74749684, + "learning_rate": 1.2190078327531454e-06, + "loss": 0.76949227, + "num_input_tokens_seen": 114377425, + "step": 5313, + "time_per_iteration": 3.712120771408081 + }, + { + "auxiliary_loss_clip": 0.01168222, + "auxiliary_loss_mlp": 0.01028721, + "balance_loss_clip": 1.04893661, + "balance_loss_mlp": 1.02068615, + "epoch": 0.6389707208561294, + "flos": 22346133384960.0, + "grad_norm": 2.195037996974825, + "language_loss": 0.72984779, + "learning_rate": 1.2182907671362697e-06, + "loss": 0.75181723, + "num_input_tokens_seen": 114398120, + "step": 5314, + "time_per_iteration": 2.854943037033081 + }, + { + "auxiliary_loss_clip": 0.01166494, + "auxiliary_loss_mlp": 0.01031821, + "balance_loss_clip": 1.05032039, + "balance_loss_mlp": 1.02359533, + "epoch": 0.6390909637467684, + "flos": 19426883247360.0, + "grad_norm": 4.730408388464143, + "language_loss": 0.78841126, + "learning_rate": 1.2175738201040626e-06, + "loss": 0.81039435, + "num_input_tokens_seen": 114415160, + "step": 5315, + "time_per_iteration": 2.696777582168579 + }, + { + "auxiliary_loss_clip": 0.01163213, + "auxiliary_loss_mlp": 0.01031441, + "balance_loss_clip": 1.04606521, + "balance_loss_mlp": 1.02302504, + "epoch": 0.6392112066374076, + "flos": 24090852700800.0, + "grad_norm": 1.8115959959265189, + "language_loss": 0.77949274, + "learning_rate": 1.2168569917652855e-06, + "loss": 0.80143929, + "num_input_tokens_seen": 114435015, + "step": 5316, + "time_per_iteration": 3.664194345474243 + }, + { + "auxiliary_loss_clip": 0.01163719, + "auxiliary_loss_mlp": 0.01030408, + "balance_loss_clip": 1.04823971, + "balance_loss_mlp": 1.02242136, + "epoch": 0.6393314495280467, + "flos": 26795141896320.0, + "grad_norm": 1.5896282652955323, + "language_loss": 0.63800085, + "learning_rate": 1.2161402822286797e-06, + "loss": 0.65994215, + "num_input_tokens_seen": 114455700, + "step": 5317, + "time_per_iteration": 2.7099969387054443 + }, + { + "auxiliary_loss_clip": 0.01148567, + "auxiliary_loss_mlp": 0.01031828, + "balance_loss_clip": 1.04705608, + "balance_loss_mlp": 1.02365661, + "epoch": 0.6394516924186857, + "flos": 20260692633600.0, + "grad_norm": 1.7866079023389334, + "language_loss": 0.79196334, + "learning_rate": 1.2154236916029703e-06, + "loss": 0.81376731, + "num_input_tokens_seen": 114473675, + "step": 5318, + "time_per_iteration": 2.8350298404693604 + }, + { + "auxiliary_loss_clip": 0.01135173, + "auxiliary_loss_mlp": 0.01024789, + "balance_loss_clip": 1.03964138, + "balance_loss_mlp": 1.01682639, + "epoch": 0.6395719353093249, + "flos": 18368847210240.0, + "grad_norm": 2.428389674877577, + "language_loss": 0.72826546, + "learning_rate": 1.2147072199968627e-06, + "loss": 0.74986506, + "num_input_tokens_seen": 114492310, + "step": 5319, + "time_per_iteration": 2.8523828983306885 + }, + { + "auxiliary_loss_clip": 0.01161426, + "auxiliary_loss_mlp": 0.01030923, + "balance_loss_clip": 1.04737437, + "balance_loss_mlp": 1.02321863, + "epoch": 0.6396921781999639, + "flos": 17566315591680.0, + "grad_norm": 1.7764699357598284, + "language_loss": 0.71641886, + "learning_rate": 1.2139908675190454e-06, + "loss": 0.73834234, + "num_input_tokens_seen": 114511520, + "step": 5320, + "time_per_iteration": 2.7451653480529785 + }, + { + "auxiliary_loss_clip": 0.01120527, + "auxiliary_loss_mlp": 0.01025139, + "balance_loss_clip": 1.04010284, + "balance_loss_mlp": 1.0164603, + "epoch": 0.639812421090603, + "flos": 21251252972160.0, + "grad_norm": 2.0994208047082723, + "language_loss": 0.75202954, + "learning_rate": 1.2132746342781883e-06, + "loss": 0.7734862, + "num_input_tokens_seen": 114532680, + "step": 5321, + "time_per_iteration": 2.958329200744629 + }, + { + "auxiliary_loss_clip": 0.01177987, + "auxiliary_loss_mlp": 0.01033296, + "balance_loss_clip": 1.05185091, + "balance_loss_mlp": 1.02495766, + "epoch": 0.6399326639812422, + "flos": 11180967684480.0, + "grad_norm": 2.4783998276351173, + "language_loss": 0.80320704, + "learning_rate": 1.2125585203829442e-06, + "loss": 0.82531989, + "num_input_tokens_seen": 114548320, + "step": 5322, + "time_per_iteration": 2.7121193408966064 + }, + { + "auxiliary_loss_clip": 0.01136948, + "auxiliary_loss_mlp": 0.01026535, + "balance_loss_clip": 1.04584289, + "balance_loss_mlp": 1.01838112, + "epoch": 0.6400529068718812, + "flos": 23911048195200.0, + "grad_norm": 2.056267436668487, + "language_loss": 0.73865867, + "learning_rate": 1.211842525941946e-06, + "loss": 0.76029354, + "num_input_tokens_seen": 114568115, + "step": 5323, + "time_per_iteration": 2.946331739425659 + }, + { + "auxiliary_loss_clip": 0.01132777, + "auxiliary_loss_mlp": 0.01024312, + "balance_loss_clip": 1.0449152, + "balance_loss_mlp": 1.01633143, + "epoch": 0.6401731497625203, + "flos": 44018724890880.0, + "grad_norm": 3.1271661553417327, + "language_loss": 0.79248559, + "learning_rate": 1.2111266510638105e-06, + "loss": 0.81405652, + "num_input_tokens_seen": 114591040, + "step": 5324, + "time_per_iteration": 3.0553970336914062 + }, + { + "auxiliary_loss_clip": 0.01120508, + "auxiliary_loss_mlp": 0.01024317, + "balance_loss_clip": 1.04110217, + "balance_loss_mlp": 1.01641977, + "epoch": 0.6402933926531594, + "flos": 20662209838080.0, + "grad_norm": 1.7269488301596823, + "language_loss": 0.8002395, + "learning_rate": 1.2104108958571346e-06, + "loss": 0.82168782, + "num_input_tokens_seen": 114609310, + "step": 5325, + "time_per_iteration": 2.8954734802246094 + }, + { + "auxiliary_loss_clip": 0.01165323, + "auxiliary_loss_mlp": 0.01031676, + "balance_loss_clip": 1.05040693, + "balance_loss_mlp": 1.02367377, + "epoch": 0.6404136355437985, + "flos": 24863327614080.0, + "grad_norm": 1.9748023993581665, + "language_loss": 0.75795114, + "learning_rate": 1.2096952604304975e-06, + "loss": 0.77992117, + "num_input_tokens_seen": 114629740, + "step": 5326, + "time_per_iteration": 2.9411673545837402 + }, + { + "auxiliary_loss_clip": 0.01167526, + "auxiliary_loss_mlp": 0.01027221, + "balance_loss_clip": 1.04738545, + "balance_loss_mlp": 1.01911473, + "epoch": 0.6405338784344375, + "flos": 40479548901120.0, + "grad_norm": 2.0101740743675585, + "language_loss": 0.70507658, + "learning_rate": 1.2089797448924616e-06, + "loss": 0.72702402, + "num_input_tokens_seen": 114653615, + "step": 5327, + "time_per_iteration": 2.935830593109131 + }, + { + "auxiliary_loss_clip": 0.01134757, + "auxiliary_loss_mlp": 0.01026708, + "balance_loss_clip": 1.04200637, + "balance_loss_mlp": 1.01800609, + "epoch": 0.6406541213250767, + "flos": 20886041439360.0, + "grad_norm": 1.9946104458195897, + "language_loss": 0.65911222, + "learning_rate": 1.2082643493515692e-06, + "loss": 0.68072689, + "num_input_tokens_seen": 114671935, + "step": 5328, + "time_per_iteration": 2.8907310962677 + }, + { + "auxiliary_loss_clip": 0.01166126, + "auxiliary_loss_mlp": 0.01031632, + "balance_loss_clip": 1.04949725, + "balance_loss_mlp": 1.02409196, + "epoch": 0.6407743642157158, + "flos": 23295970679040.0, + "grad_norm": 1.7831450650899425, + "language_loss": 0.81803119, + "learning_rate": 1.207549073916346e-06, + "loss": 0.8400088, + "num_input_tokens_seen": 114692870, + "step": 5329, + "time_per_iteration": 2.815514326095581 + }, + { + "auxiliary_loss_clip": 0.01147202, + "auxiliary_loss_mlp": 0.01025443, + "balance_loss_clip": 1.04541039, + "balance_loss_mlp": 1.01727152, + "epoch": 0.6408946071063548, + "flos": 15012636122880.0, + "grad_norm": 2.1449393642772843, + "language_loss": 0.77757955, + "learning_rate": 1.2068339186952976e-06, + "loss": 0.79930604, + "num_input_tokens_seen": 114710410, + "step": 5330, + "time_per_iteration": 2.748927116394043 + }, + { + "auxiliary_loss_clip": 0.0116886, + "auxiliary_loss_mlp": 0.01025502, + "balance_loss_clip": 1.05006194, + "balance_loss_mlp": 1.01740789, + "epoch": 0.6410148499969939, + "flos": 22528595496960.0, + "grad_norm": 1.7400982704595234, + "language_loss": 0.73110974, + "learning_rate": 1.2061188837969136e-06, + "loss": 0.75305337, + "num_input_tokens_seen": 114730020, + "step": 5331, + "time_per_iteration": 2.6836671829223633 + }, + { + "auxiliary_loss_clip": 0.01140237, + "auxiliary_loss_mlp": 0.01024793, + "balance_loss_clip": 1.04164505, + "balance_loss_mlp": 1.0157454, + "epoch": 0.641135092887633, + "flos": 12422004537600.0, + "grad_norm": 2.92688327785134, + "language_loss": 0.838552, + "learning_rate": 1.2054039693296631e-06, + "loss": 0.86020231, + "num_input_tokens_seen": 114748015, + "step": 5332, + "time_per_iteration": 2.7145769596099854 + }, + { + "auxiliary_loss_clip": 0.01137801, + "auxiliary_loss_mlp": 0.01032556, + "balance_loss_clip": 1.04197145, + "balance_loss_mlp": 1.02477765, + "epoch": 0.6412553357782721, + "flos": 22127329687680.0, + "grad_norm": 1.8792645389341298, + "language_loss": 0.80858386, + "learning_rate": 1.2046891754019992e-06, + "loss": 0.8302874, + "num_input_tokens_seen": 114768625, + "step": 5333, + "time_per_iteration": 2.760030746459961 + }, + { + "auxiliary_loss_clip": 0.01167474, + "auxiliary_loss_mlp": 0.01031245, + "balance_loss_clip": 1.04909205, + "balance_loss_mlp": 1.0228467, + "epoch": 0.6413755786689112, + "flos": 15888605097600.0, + "grad_norm": 1.7766899213160505, + "language_loss": 0.82451409, + "learning_rate": 1.2039745021223548e-06, + "loss": 0.84650129, + "num_input_tokens_seen": 114786045, + "step": 5334, + "time_per_iteration": 2.6369740962982178 + }, + { + "auxiliary_loss_clip": 0.01042985, + "auxiliary_loss_mlp": 0.01003805, + "balance_loss_clip": 1.01448607, + "balance_loss_mlp": 1.00206435, + "epoch": 0.6414958215595503, + "flos": 68039159955840.0, + "grad_norm": 0.7956718867200269, + "language_loss": 0.57060021, + "learning_rate": 1.2032599495991456e-06, + "loss": 0.59106809, + "num_input_tokens_seen": 114850785, + "step": 5335, + "time_per_iteration": 3.465202808380127 + }, + { + "auxiliary_loss_clip": 0.01168161, + "auxiliary_loss_mlp": 0.01031019, + "balance_loss_clip": 1.05025172, + "balance_loss_mlp": 1.02281761, + "epoch": 0.6416160644501894, + "flos": 44091300320640.0, + "grad_norm": 1.8385270239516194, + "language_loss": 0.6922034, + "learning_rate": 1.2025455179407685e-06, + "loss": 0.71419513, + "num_input_tokens_seen": 114871945, + "step": 5336, + "time_per_iteration": 4.788590669631958 + }, + { + "auxiliary_loss_clip": 0.01161546, + "auxiliary_loss_mlp": 0.00886362, + "balance_loss_clip": 1.04906654, + "balance_loss_mlp": 1.00045896, + "epoch": 0.6417363073408284, + "flos": 20959837931520.0, + "grad_norm": 1.9291884105850619, + "language_loss": 0.73887861, + "learning_rate": 1.2018312072556022e-06, + "loss": 0.75935769, + "num_input_tokens_seen": 114890445, + "step": 5337, + "time_per_iteration": 2.7274298667907715 + }, + { + "auxiliary_loss_clip": 0.0117648, + "auxiliary_loss_mlp": 0.00885788, + "balance_loss_clip": 1.05140853, + "balance_loss_mlp": 1.00055385, + "epoch": 0.6418565502314676, + "flos": 22455122227200.0, + "grad_norm": 1.8635188589060883, + "language_loss": 0.74629378, + "learning_rate": 1.2011170176520077e-06, + "loss": 0.76691651, + "num_input_tokens_seen": 114911360, + "step": 5338, + "time_per_iteration": 2.760903835296631 + }, + { + "auxiliary_loss_clip": 0.0111136, + "auxiliary_loss_mlp": 0.01027265, + "balance_loss_clip": 1.03946781, + "balance_loss_mlp": 1.0191288, + "epoch": 0.6419767931221066, + "flos": 25045502417280.0, + "grad_norm": 1.9273721207751329, + "language_loss": 0.81102753, + "learning_rate": 1.2004029492383256e-06, + "loss": 0.83241379, + "num_input_tokens_seen": 114932700, + "step": 5339, + "time_per_iteration": 3.874443769454956 + }, + { + "auxiliary_loss_clip": 0.011627, + "auxiliary_loss_mlp": 0.0102924, + "balance_loss_clip": 1.0483675, + "balance_loss_mlp": 1.02144384, + "epoch": 0.6420970360127457, + "flos": 19463691709440.0, + "grad_norm": 2.753246008618856, + "language_loss": 0.73492551, + "learning_rate": 1.1996890021228814e-06, + "loss": 0.75684488, + "num_input_tokens_seen": 114949475, + "step": 5340, + "time_per_iteration": 2.6795995235443115 + }, + { + "auxiliary_loss_clip": 0.01152225, + "auxiliary_loss_mlp": 0.01022754, + "balance_loss_clip": 1.04481399, + "balance_loss_mlp": 1.01463008, + "epoch": 0.6422172789033849, + "flos": 40406147458560.0, + "grad_norm": 1.5037255113129295, + "language_loss": 0.69849008, + "learning_rate": 1.1989751764139785e-06, + "loss": 0.72023988, + "num_input_tokens_seen": 114973125, + "step": 5341, + "time_per_iteration": 2.9288175106048584 + }, + { + "auxiliary_loss_clip": 0.01136187, + "auxiliary_loss_mlp": 0.01025818, + "balance_loss_clip": 1.04116392, + "balance_loss_mlp": 1.01747954, + "epoch": 0.6423375217940239, + "flos": 27672870637440.0, + "grad_norm": 1.5589840104825412, + "language_loss": 0.83169508, + "learning_rate": 1.1982614722199044e-06, + "loss": 0.85331517, + "num_input_tokens_seen": 114994300, + "step": 5342, + "time_per_iteration": 3.816462516784668 + }, + { + "auxiliary_loss_clip": 0.01159832, + "auxiliary_loss_mlp": 0.01027153, + "balance_loss_clip": 1.04508662, + "balance_loss_mlp": 1.01927328, + "epoch": 0.642457764684663, + "flos": 18369242259840.0, + "grad_norm": 2.1130588285469365, + "language_loss": 0.77886581, + "learning_rate": 1.1975478896489276e-06, + "loss": 0.80073571, + "num_input_tokens_seen": 115012135, + "step": 5343, + "time_per_iteration": 2.767624616622925 + }, + { + "auxiliary_loss_clip": 0.01171734, + "auxiliary_loss_mlp": 0.01028461, + "balance_loss_clip": 1.04804087, + "balance_loss_mlp": 1.02093279, + "epoch": 0.6425780075753021, + "flos": 19750509809280.0, + "grad_norm": 2.0158185945961997, + "language_loss": 0.76420516, + "learning_rate": 1.1968344288092981e-06, + "loss": 0.78620708, + "num_input_tokens_seen": 115028715, + "step": 5344, + "time_per_iteration": 2.6994993686676025 + }, + { + "auxiliary_loss_clip": 0.01162896, + "auxiliary_loss_mlp": 0.00885737, + "balance_loss_clip": 1.04693234, + "balance_loss_mlp": 1.00046432, + "epoch": 0.6426982504659412, + "flos": 20558536208640.0, + "grad_norm": 1.7350984598171004, + "language_loss": 0.64726675, + "learning_rate": 1.1961210898092468e-06, + "loss": 0.6677531, + "num_input_tokens_seen": 115047665, + "step": 5345, + "time_per_iteration": 2.7793920040130615 + }, + { + "auxiliary_loss_clip": 0.01164173, + "auxiliary_loss_mlp": 0.01030556, + "balance_loss_clip": 1.04998028, + "balance_loss_mlp": 1.02173495, + "epoch": 0.6428184933565803, + "flos": 17851984456320.0, + "grad_norm": 2.0281569588488377, + "language_loss": 0.79281807, + "learning_rate": 1.1954078727569874e-06, + "loss": 0.81476533, + "num_input_tokens_seen": 115064965, + "step": 5346, + "time_per_iteration": 2.8361148834228516 + }, + { + "auxiliary_loss_clip": 0.01151735, + "auxiliary_loss_mlp": 0.00886149, + "balance_loss_clip": 1.04595196, + "balance_loss_mlp": 1.00056052, + "epoch": 0.6429387362472194, + "flos": 22456953820800.0, + "grad_norm": 1.5286927833203894, + "language_loss": 0.78090358, + "learning_rate": 1.1946947777607141e-06, + "loss": 0.80128241, + "num_input_tokens_seen": 115086100, + "step": 5347, + "time_per_iteration": 2.859393835067749 + }, + { + "auxiliary_loss_clip": 0.01130766, + "auxiliary_loss_mlp": 0.01027784, + "balance_loss_clip": 1.04373264, + "balance_loss_mlp": 1.01854551, + "epoch": 0.6430589791378585, + "flos": 24752579005440.0, + "grad_norm": 1.9216374350152392, + "language_loss": 0.80670279, + "learning_rate": 1.1939818049286024e-06, + "loss": 0.82828832, + "num_input_tokens_seen": 115104260, + "step": 5348, + "time_per_iteration": 2.918943166732788 + }, + { + "auxiliary_loss_clip": 0.01115391, + "auxiliary_loss_mlp": 0.01026523, + "balance_loss_clip": 1.03964138, + "balance_loss_mlp": 1.01795149, + "epoch": 0.6431792220284975, + "flos": 24901249397760.0, + "grad_norm": 1.7083737899799076, + "language_loss": 0.7548399, + "learning_rate": 1.1932689543688101e-06, + "loss": 0.77625901, + "num_input_tokens_seen": 115125365, + "step": 5349, + "time_per_iteration": 2.978463888168335 + }, + { + "auxiliary_loss_clip": 0.01155903, + "auxiliary_loss_mlp": 0.01029023, + "balance_loss_clip": 1.05001211, + "balance_loss_mlp": 1.02069616, + "epoch": 0.6432994649191367, + "flos": 21032305620480.0, + "grad_norm": 2.0485486430495934, + "language_loss": 0.72753704, + "learning_rate": 1.1925562261894756e-06, + "loss": 0.74938631, + "num_input_tokens_seen": 115144445, + "step": 5350, + "time_per_iteration": 2.9080605506896973 + }, + { + "auxiliary_loss_clip": 0.01149259, + "auxiliary_loss_mlp": 0.01025301, + "balance_loss_clip": 1.04402947, + "balance_loss_mlp": 1.01704001, + "epoch": 0.6434197078097758, + "flos": 30884433655680.0, + "grad_norm": 1.6783459649062205, + "language_loss": 0.77482826, + "learning_rate": 1.1918436204987207e-06, + "loss": 0.79657388, + "num_input_tokens_seen": 115166305, + "step": 5351, + "time_per_iteration": 2.8582286834716797 + }, + { + "auxiliary_loss_clip": 0.01162199, + "auxiliary_loss_mlp": 0.01026525, + "balance_loss_clip": 1.0498935, + "balance_loss_mlp": 1.01808476, + "epoch": 0.6435399507004148, + "flos": 15012492468480.0, + "grad_norm": 1.994199158734672, + "language_loss": 0.81654739, + "learning_rate": 1.191131137404645e-06, + "loss": 0.83843464, + "num_input_tokens_seen": 115183045, + "step": 5352, + "time_per_iteration": 2.6729371547698975 + }, + { + "auxiliary_loss_clip": 0.01134221, + "auxiliary_loss_mlp": 0.01032338, + "balance_loss_clip": 1.04306686, + "balance_loss_mlp": 1.02400541, + "epoch": 0.643660193591054, + "flos": 19901981462400.0, + "grad_norm": 2.0888145838178462, + "language_loss": 0.77429867, + "learning_rate": 1.190418777015333e-06, + "loss": 0.79596424, + "num_input_tokens_seen": 115201955, + "step": 5353, + "time_per_iteration": 2.807953357696533 + }, + { + "auxiliary_loss_clip": 0.0115301, + "auxiliary_loss_mlp": 0.0102463, + "balance_loss_clip": 1.04647374, + "balance_loss_mlp": 1.01674438, + "epoch": 0.643780436481693, + "flos": 24133622820480.0, + "grad_norm": 1.4393446791971782, + "language_loss": 0.73594201, + "learning_rate": 1.1897065394388487e-06, + "loss": 0.75771832, + "num_input_tokens_seen": 115222395, + "step": 5354, + "time_per_iteration": 2.7237226963043213 + }, + { + "auxiliary_loss_clip": 0.0115642, + "auxiliary_loss_mlp": 0.01027294, + "balance_loss_clip": 1.04999399, + "balance_loss_mlp": 1.01949763, + "epoch": 0.6439006793723321, + "flos": 23148808657920.0, + "grad_norm": 1.5702965400460238, + "language_loss": 0.76289713, + "learning_rate": 1.1889944247832385e-06, + "loss": 0.78473431, + "num_input_tokens_seen": 115242635, + "step": 5355, + "time_per_iteration": 2.732752799987793 + }, + { + "auxiliary_loss_clip": 0.01165743, + "auxiliary_loss_mlp": 0.01026137, + "balance_loss_clip": 1.0462898, + "balance_loss_mlp": 1.01788735, + "epoch": 0.6440209222629713, + "flos": 23617909301760.0, + "grad_norm": 2.1408911062907556, + "language_loss": 0.70744413, + "learning_rate": 1.1882824331565283e-06, + "loss": 0.72936296, + "num_input_tokens_seen": 115262095, + "step": 5356, + "time_per_iteration": 2.7099931240081787 + }, + { + "auxiliary_loss_clip": 0.01139258, + "auxiliary_loss_mlp": 0.01023757, + "balance_loss_clip": 1.04108202, + "balance_loss_mlp": 1.01564479, + "epoch": 0.6441411651536103, + "flos": 16544872535040.0, + "grad_norm": 2.0021141410982386, + "language_loss": 0.89165699, + "learning_rate": 1.1875705646667287e-06, + "loss": 0.91328716, + "num_input_tokens_seen": 115279985, + "step": 5357, + "time_per_iteration": 2.7982492446899414 + }, + { + "auxiliary_loss_clip": 0.01159492, + "auxiliary_loss_mlp": 0.01027656, + "balance_loss_clip": 1.04423881, + "balance_loss_mlp": 1.0189898, + "epoch": 0.6442614080442494, + "flos": 25410965345280.0, + "grad_norm": 2.6292778263497834, + "language_loss": 0.75219578, + "learning_rate": 1.1868588194218282e-06, + "loss": 0.77406728, + "num_input_tokens_seen": 115300365, + "step": 5358, + "time_per_iteration": 2.812204122543335 + }, + { + "auxiliary_loss_clip": 0.01159513, + "auxiliary_loss_mlp": 0.01033585, + "balance_loss_clip": 1.04535127, + "balance_loss_mlp": 1.02540684, + "epoch": 0.6443816509348885, + "flos": 28294017552000.0, + "grad_norm": 1.5411564278943763, + "language_loss": 0.73916376, + "learning_rate": 1.1861471975297979e-06, + "loss": 0.76109475, + "num_input_tokens_seen": 115322060, + "step": 5359, + "time_per_iteration": 2.8450920581817627 + }, + { + "auxiliary_loss_clip": 0.01144351, + "auxiliary_loss_mlp": 0.01030003, + "balance_loss_clip": 1.0464468, + "balance_loss_mlp": 1.02126479, + "epoch": 0.6445018938255276, + "flos": 36690075964800.0, + "grad_norm": 1.5509761645644682, + "language_loss": 0.70924556, + "learning_rate": 1.185435699098591e-06, + "loss": 0.7309891, + "num_input_tokens_seen": 115348255, + "step": 5360, + "time_per_iteration": 2.9530367851257324 + }, + { + "auxiliary_loss_clip": 0.01157472, + "auxiliary_loss_mlp": 0.01031569, + "balance_loss_clip": 1.0461657, + "balance_loss_mlp": 1.02342081, + "epoch": 0.6446221367161666, + "flos": 14501411804160.0, + "grad_norm": 2.550663066984056, + "language_loss": 0.77638125, + "learning_rate": 1.1847243242361403e-06, + "loss": 0.79827166, + "num_input_tokens_seen": 115366845, + "step": 5361, + "time_per_iteration": 2.7168939113616943 + }, + { + "auxiliary_loss_clip": 0.01154633, + "auxiliary_loss_mlp": 0.01031334, + "balance_loss_clip": 1.04645503, + "balance_loss_mlp": 1.02264333, + "epoch": 0.6447423796068057, + "flos": 24609367480320.0, + "grad_norm": 1.5824847335536163, + "language_loss": 0.77901024, + "learning_rate": 1.1840130730503624e-06, + "loss": 0.80086994, + "num_input_tokens_seen": 115388125, + "step": 5362, + "time_per_iteration": 4.65622353553772 + }, + { + "auxiliary_loss_clip": 0.01175004, + "auxiliary_loss_mlp": 0.0102965, + "balance_loss_clip": 1.05105948, + "balance_loss_mlp": 1.02203846, + "epoch": 0.6448626224974449, + "flos": 25047298097280.0, + "grad_norm": 1.636113673415346, + "language_loss": 0.7473104, + "learning_rate": 1.1833019456491518e-06, + "loss": 0.76935697, + "num_input_tokens_seen": 115409655, + "step": 5363, + "time_per_iteration": 2.7591090202331543 + }, + { + "auxiliary_loss_clip": 0.01166156, + "auxiliary_loss_mlp": 0.01026609, + "balance_loss_clip": 1.0486635, + "balance_loss_mlp": 1.01826131, + "epoch": 0.6449828653880839, + "flos": 22530355263360.0, + "grad_norm": 2.043077941574622, + "language_loss": 0.79469812, + "learning_rate": 1.1825909421403871e-06, + "loss": 0.81662583, + "num_input_tokens_seen": 115428750, + "step": 5364, + "time_per_iteration": 2.721766233444214 + }, + { + "auxiliary_loss_clip": 0.01166064, + "auxiliary_loss_mlp": 0.01028602, + "balance_loss_clip": 1.04843926, + "balance_loss_mlp": 1.02050805, + "epoch": 0.645103108278723, + "flos": 25695736369920.0, + "grad_norm": 2.1581779755787336, + "language_loss": 0.76117367, + "learning_rate": 1.181880062631926e-06, + "loss": 0.78312027, + "num_input_tokens_seen": 115448085, + "step": 5365, + "time_per_iteration": 3.6616344451904297 + }, + { + "auxiliary_loss_clip": 0.01148917, + "auxiliary_loss_mlp": 0.01027116, + "balance_loss_clip": 1.04736114, + "balance_loss_mlp": 1.01887298, + "epoch": 0.6452233511693621, + "flos": 27450331925760.0, + "grad_norm": 2.0717532392649938, + "language_loss": 0.85280305, + "learning_rate": 1.1811693072316093e-06, + "loss": 0.87456346, + "num_input_tokens_seen": 115465765, + "step": 5366, + "time_per_iteration": 2.7813215255737305 + }, + { + "auxiliary_loss_clip": 0.01173574, + "auxiliary_loss_mlp": 0.00886209, + "balance_loss_clip": 1.04871511, + "balance_loss_mlp": 1.00047612, + "epoch": 0.6453435940600012, + "flos": 19208618254080.0, + "grad_norm": 2.9287541289320065, + "language_loss": 0.8400718, + "learning_rate": 1.1804586760472574e-06, + "loss": 0.86066961, + "num_input_tokens_seen": 115482230, + "step": 5367, + "time_per_iteration": 2.729764223098755 + }, + { + "auxiliary_loss_clip": 0.01140368, + "auxiliary_loss_mlp": 0.01026541, + "balance_loss_clip": 1.04253745, + "balance_loss_mlp": 1.01846433, + "epoch": 0.6454638369506402, + "flos": 25737680476800.0, + "grad_norm": 2.5514866560991725, + "language_loss": 0.80075824, + "learning_rate": 1.1797481691866736e-06, + "loss": 0.82242727, + "num_input_tokens_seen": 115499455, + "step": 5368, + "time_per_iteration": 3.8702194690704346 + }, + { + "auxiliary_loss_clip": 0.01143166, + "auxiliary_loss_mlp": 0.01025483, + "balance_loss_clip": 1.04469657, + "balance_loss_mlp": 1.01779413, + "epoch": 0.6455840798412794, + "flos": 20989176364800.0, + "grad_norm": 1.9757240947429546, + "language_loss": 0.83195639, + "learning_rate": 1.1790377867576393e-06, + "loss": 0.85364294, + "num_input_tokens_seen": 115517205, + "step": 5369, + "time_per_iteration": 2.857855796813965 + }, + { + "auxiliary_loss_clip": 0.01155226, + "auxiliary_loss_mlp": 0.01025898, + "balance_loss_clip": 1.04376864, + "balance_loss_mlp": 1.01762533, + "epoch": 0.6457043227319185, + "flos": 26067556005120.0, + "grad_norm": 1.752078595849393, + "language_loss": 0.76791358, + "learning_rate": 1.1783275288679203e-06, + "loss": 0.78972483, + "num_input_tokens_seen": 115534370, + "step": 5370, + "time_per_iteration": 2.7340240478515625 + }, + { + "auxiliary_loss_clip": 0.01058483, + "auxiliary_loss_mlp": 0.01001911, + "balance_loss_clip": 1.00889969, + "balance_loss_mlp": 1.00036168, + "epoch": 0.6458245656225575, + "flos": 60370831088640.0, + "grad_norm": 0.845449731779572, + "language_loss": 0.57144487, + "learning_rate": 1.177617395625262e-06, + "loss": 0.59204888, + "num_input_tokens_seen": 115592345, + "step": 5371, + "time_per_iteration": 3.23085880279541 + }, + { + "auxiliary_loss_clip": 0.01163837, + "auxiliary_loss_mlp": 0.01024864, + "balance_loss_clip": 1.04839849, + "balance_loss_mlp": 1.01708007, + "epoch": 0.6459448085131967, + "flos": 23076771932160.0, + "grad_norm": 2.023999384093336, + "language_loss": 0.75344032, + "learning_rate": 1.1769073871373908e-06, + "loss": 0.77532732, + "num_input_tokens_seen": 115612550, + "step": 5372, + "time_per_iteration": 2.6592955589294434 + }, + { + "auxiliary_loss_clip": 0.01140599, + "auxiliary_loss_mlp": 0.01021918, + "balance_loss_clip": 1.04059768, + "balance_loss_mlp": 1.01334703, + "epoch": 0.6460650514038357, + "flos": 22598190097920.0, + "grad_norm": 2.062239059764274, + "language_loss": 0.83882785, + "learning_rate": 1.176197503512015e-06, + "loss": 0.86045301, + "num_input_tokens_seen": 115632265, + "step": 5373, + "time_per_iteration": 2.772866725921631 + }, + { + "auxiliary_loss_clip": 0.01148012, + "auxiliary_loss_mlp": 0.01026366, + "balance_loss_clip": 1.04385221, + "balance_loss_mlp": 1.01847386, + "epoch": 0.6461852942944748, + "flos": 20266726118400.0, + "grad_norm": 2.0462291155917245, + "language_loss": 0.8245346, + "learning_rate": 1.1754877448568223e-06, + "loss": 0.84627837, + "num_input_tokens_seen": 115651720, + "step": 5374, + "time_per_iteration": 2.7449679374694824 + }, + { + "auxiliary_loss_clip": 0.01152773, + "auxiliary_loss_mlp": 0.01025217, + "balance_loss_clip": 1.04461193, + "balance_loss_mlp": 1.01757562, + "epoch": 0.646305537185114, + "flos": 23367109564800.0, + "grad_norm": 2.048115827011344, + "language_loss": 0.9008621, + "learning_rate": 1.1747781112794837e-06, + "loss": 0.92264193, + "num_input_tokens_seen": 115668215, + "step": 5375, + "time_per_iteration": 2.8386895656585693 + }, + { + "auxiliary_loss_clip": 0.01143722, + "auxiliary_loss_mlp": 0.01032678, + "balance_loss_clip": 1.04579365, + "balance_loss_mlp": 1.02503109, + "epoch": 0.646425780075753, + "flos": 24277480790400.0, + "grad_norm": 1.7799067728818718, + "language_loss": 0.83087444, + "learning_rate": 1.1740686028876487e-06, + "loss": 0.85263848, + "num_input_tokens_seen": 115687080, + "step": 5376, + "time_per_iteration": 2.905052423477173 + }, + { + "auxiliary_loss_clip": 0.01160418, + "auxiliary_loss_mlp": 0.01023196, + "balance_loss_clip": 1.04815936, + "balance_loss_mlp": 1.01534009, + "epoch": 0.6465460229663921, + "flos": 20813968800000.0, + "grad_norm": 2.843149694420271, + "language_loss": 0.74662256, + "learning_rate": 1.1733592197889507e-06, + "loss": 0.76845872, + "num_input_tokens_seen": 115703990, + "step": 5377, + "time_per_iteration": 0.07628440856933594 + }, + { + "auxiliary_loss_clip": 0.01158605, + "auxiliary_loss_mlp": 0.01029497, + "balance_loss_clip": 1.04751956, + "balance_loss_mlp": 1.02210307, + "epoch": 0.6466662658570312, + "flos": 22853299466880.0, + "grad_norm": 1.7767725370523555, + "language_loss": 0.72370917, + "learning_rate": 1.1726499620910014e-06, + "loss": 0.74559021, + "num_input_tokens_seen": 115724270, + "step": 5378, + "time_per_iteration": 2.7102243900299072 + }, + { + "auxiliary_loss_clip": 0.01160421, + "auxiliary_loss_mlp": 0.01028294, + "balance_loss_clip": 1.04740715, + "balance_loss_mlp": 1.02028298, + "epoch": 0.6467865087476703, + "flos": 15304553953920.0, + "grad_norm": 2.121350462418043, + "language_loss": 0.78181684, + "learning_rate": 1.1719408299013955e-06, + "loss": 0.80370402, + "num_input_tokens_seen": 115742995, + "step": 5379, + "time_per_iteration": 2.80778169631958 + }, + { + "auxiliary_loss_clip": 0.01170388, + "auxiliary_loss_mlp": 0.01029565, + "balance_loss_clip": 1.04892659, + "balance_loss_mlp": 1.02225113, + "epoch": 0.6469067516383094, + "flos": 19573650218880.0, + "grad_norm": 2.320843120327907, + "language_loss": 0.75982195, + "learning_rate": 1.1712318233277067e-06, + "loss": 0.78182149, + "num_input_tokens_seen": 115762015, + "step": 5380, + "time_per_iteration": 2.673990249633789 + }, + { + "auxiliary_loss_clip": 0.01057764, + "auxiliary_loss_mlp": 0.01001884, + "balance_loss_clip": 1.00965786, + "balance_loss_mlp": 1.00034666, + "epoch": 0.6470269945289485, + "flos": 65098002522240.0, + "grad_norm": 0.8089275185122924, + "language_loss": 0.57868922, + "learning_rate": 1.1705229424774916e-06, + "loss": 0.59928566, + "num_input_tokens_seen": 115816285, + "step": 5381, + "time_per_iteration": 3.135690689086914 + }, + { + "auxiliary_loss_clip": 0.01151816, + "auxiliary_loss_mlp": 0.0102899, + "balance_loss_clip": 1.04556799, + "balance_loss_mlp": 1.0211761, + "epoch": 0.6471472374195876, + "flos": 30696943639680.0, + "grad_norm": 1.6112759647045707, + "language_loss": 0.64038026, + "learning_rate": 1.1698141874582867e-06, + "loss": 0.66218835, + "num_input_tokens_seen": 115837330, + "step": 5382, + "time_per_iteration": 2.812626600265503 + }, + { + "auxiliary_loss_clip": 0.01170552, + "auxiliary_loss_mlp": 0.01036873, + "balance_loss_clip": 1.04867435, + "balance_loss_mlp": 1.02906132, + "epoch": 0.6472674803102266, + "flos": 20521835487360.0, + "grad_norm": 1.7489956064545527, + "language_loss": 0.71990001, + "learning_rate": 1.169105558377609e-06, + "loss": 0.74197423, + "num_input_tokens_seen": 115857420, + "step": 5383, + "time_per_iteration": 2.6695945262908936 + }, + { + "auxiliary_loss_clip": 0.0112371, + "auxiliary_loss_mlp": 0.00885684, + "balance_loss_clip": 1.04551399, + "balance_loss_mlp": 1.00053549, + "epoch": 0.6473877232008658, + "flos": 24715447320960.0, + "grad_norm": 1.6948224571896073, + "language_loss": 0.78315282, + "learning_rate": 1.1683970553429587e-06, + "loss": 0.80324674, + "num_input_tokens_seen": 115878875, + "step": 5384, + "time_per_iteration": 2.8681252002716064 + }, + { + "auxiliary_loss_clip": 0.01146112, + "auxiliary_loss_mlp": 0.01030511, + "balance_loss_clip": 1.04413939, + "balance_loss_mlp": 1.02264321, + "epoch": 0.6475079660915048, + "flos": 15885552441600.0, + "grad_norm": 1.8127984297237485, + "language_loss": 0.8198148, + "learning_rate": 1.1676886784618128e-06, + "loss": 0.84158105, + "num_input_tokens_seen": 115895540, + "step": 5385, + "time_per_iteration": 2.7715089321136475 + }, + { + "auxiliary_loss_clip": 0.01162437, + "auxiliary_loss_mlp": 0.01032786, + "balance_loss_clip": 1.04756188, + "balance_loss_mlp": 1.02430999, + "epoch": 0.6476282089821439, + "flos": 17381590922880.0, + "grad_norm": 2.795311663693217, + "language_loss": 0.8413614, + "learning_rate": 1.1669804278416332e-06, + "loss": 0.86331367, + "num_input_tokens_seen": 115910265, + "step": 5386, + "time_per_iteration": 2.720228433609009 + }, + { + "auxiliary_loss_clip": 0.01160698, + "auxiliary_loss_mlp": 0.01025237, + "balance_loss_clip": 1.04798245, + "balance_loss_mlp": 1.01674318, + "epoch": 0.6477484518727831, + "flos": 20194078861440.0, + "grad_norm": 2.2024907058408987, + "language_loss": 0.71616256, + "learning_rate": 1.1662723035898602e-06, + "loss": 0.73802191, + "num_input_tokens_seen": 115930025, + "step": 5387, + "time_per_iteration": 3.7177932262420654 + }, + { + "auxiliary_loss_clip": 0.01161703, + "auxiliary_loss_mlp": 0.010298, + "balance_loss_clip": 1.04751909, + "balance_loss_mlp": 1.02149117, + "epoch": 0.6478686947634221, + "flos": 25410426641280.0, + "grad_norm": 1.8670713013255835, + "language_loss": 0.8216328, + "learning_rate": 1.165564305813915e-06, + "loss": 0.84354782, + "num_input_tokens_seen": 115949025, + "step": 5388, + "time_per_iteration": 3.7230043411254883 + }, + { + "auxiliary_loss_clip": 0.01163301, + "auxiliary_loss_mlp": 0.01026147, + "balance_loss_clip": 1.04744387, + "balance_loss_mlp": 1.01846349, + "epoch": 0.6479889376540612, + "flos": 20083581648000.0, + "grad_norm": 1.7865903957395681, + "language_loss": 0.81080425, + "learning_rate": 1.1648564346212019e-06, + "loss": 0.8326987, + "num_input_tokens_seen": 115968145, + "step": 5389, + "time_per_iteration": 2.7495107650756836 + }, + { + "auxiliary_loss_clip": 0.01156508, + "auxiliary_loss_mlp": 0.01026686, + "balance_loss_clip": 1.04790461, + "balance_loss_mlp": 1.01929498, + "epoch": 0.6481091805447003, + "flos": 26758082039040.0, + "grad_norm": 2.2479728772664043, + "language_loss": 0.76113033, + "learning_rate": 1.164148690119104e-06, + "loss": 0.78296226, + "num_input_tokens_seen": 115989425, + "step": 5390, + "time_per_iteration": 2.9393348693847656 + }, + { + "auxiliary_loss_clip": 0.01170887, + "auxiliary_loss_mlp": 0.01024618, + "balance_loss_clip": 1.048877, + "balance_loss_mlp": 1.01694107, + "epoch": 0.6482294234353394, + "flos": 23952094462080.0, + "grad_norm": 2.1636890244783133, + "language_loss": 0.74077678, + "learning_rate": 1.163441072414985e-06, + "loss": 0.76273179, + "num_input_tokens_seen": 116009630, + "step": 5391, + "time_per_iteration": 3.708743095397949 + }, + { + "auxiliary_loss_clip": 0.01161302, + "auxiliary_loss_mlp": 0.0102602, + "balance_loss_clip": 1.04874074, + "balance_loss_mlp": 1.01797354, + "epoch": 0.6483496663259785, + "flos": 26209833776640.0, + "grad_norm": 1.8657715985938197, + "language_loss": 0.69908714, + "learning_rate": 1.16273358161619e-06, + "loss": 0.72096038, + "num_input_tokens_seen": 116029965, + "step": 5392, + "time_per_iteration": 2.863753318786621 + }, + { + "auxiliary_loss_clip": 0.01156386, + "auxiliary_loss_mlp": 0.01030171, + "balance_loss_clip": 1.04906678, + "balance_loss_mlp": 1.02090263, + "epoch": 0.6484699092166175, + "flos": 20922239370240.0, + "grad_norm": 1.7749447395788285, + "language_loss": 0.83519506, + "learning_rate": 1.1620262178300446e-06, + "loss": 0.85706061, + "num_input_tokens_seen": 116048580, + "step": 5393, + "time_per_iteration": 2.897982120513916 + }, + { + "auxiliary_loss_clip": 0.01143533, + "auxiliary_loss_mlp": 0.01026053, + "balance_loss_clip": 1.04245663, + "balance_loss_mlp": 1.01796424, + "epoch": 0.6485901521072567, + "flos": 33072865678080.0, + "grad_norm": 1.884366123298288, + "language_loss": 0.75993931, + "learning_rate": 1.1613189811638563e-06, + "loss": 0.78163517, + "num_input_tokens_seen": 116070305, + "step": 5394, + "time_per_iteration": 3.7449092864990234 + }, + { + "auxiliary_loss_clip": 0.01164763, + "auxiliary_loss_mlp": 0.01030582, + "balance_loss_clip": 1.0486424, + "balance_loss_mlp": 1.02336991, + "epoch": 0.6487103949978957, + "flos": 22274060745600.0, + "grad_norm": 1.8531723632516295, + "language_loss": 0.78053784, + "learning_rate": 1.1606118717249117e-06, + "loss": 0.80249119, + "num_input_tokens_seen": 116090405, + "step": 5395, + "time_per_iteration": 2.720914840698242 + }, + { + "auxiliary_loss_clip": 0.01173746, + "auxiliary_loss_mlp": 0.01030612, + "balance_loss_clip": 1.04727829, + "balance_loss_mlp": 1.02165377, + "epoch": 0.6488306378885348, + "flos": 22930400010240.0, + "grad_norm": 1.8451281240061026, + "language_loss": 0.67833757, + "learning_rate": 1.1599048896204787e-06, + "loss": 0.70038116, + "num_input_tokens_seen": 116110285, + "step": 5396, + "time_per_iteration": 2.7091381549835205 + }, + { + "auxiliary_loss_clip": 0.0114614, + "auxiliary_loss_mlp": 0.01027023, + "balance_loss_clip": 1.04485834, + "balance_loss_mlp": 1.01908374, + "epoch": 0.648950880779174, + "flos": 20376110010240.0, + "grad_norm": 1.800918183366077, + "language_loss": 0.80810773, + "learning_rate": 1.1591980349578061e-06, + "loss": 0.82983941, + "num_input_tokens_seen": 116128955, + "step": 5397, + "time_per_iteration": 2.771622896194458 + }, + { + "auxiliary_loss_clip": 0.01048075, + "auxiliary_loss_mlp": 0.01001444, + "balance_loss_clip": 1.01084566, + "balance_loss_mlp": 0.99987072, + "epoch": 0.649071123669813, + "flos": 59930889310080.0, + "grad_norm": 0.7371385982750359, + "language_loss": 0.54272705, + "learning_rate": 1.158491307844123e-06, + "loss": 0.56322223, + "num_input_tokens_seen": 116188875, + "step": 5398, + "time_per_iteration": 3.2798733711242676 + }, + { + "auxiliary_loss_clip": 0.01152522, + "auxiliary_loss_mlp": 0.01033003, + "balance_loss_clip": 1.04737449, + "balance_loss_mlp": 1.02505815, + "epoch": 0.6491913665604521, + "flos": 20446566537600.0, + "grad_norm": 1.5509239447582857, + "language_loss": 0.83743024, + "learning_rate": 1.1577847083866387e-06, + "loss": 0.85928547, + "num_input_tokens_seen": 116207910, + "step": 5399, + "time_per_iteration": 2.739912986755371 + }, + { + "auxiliary_loss_clip": 0.01140681, + "auxiliary_loss_mlp": 0.01028073, + "balance_loss_clip": 1.04326034, + "balance_loss_mlp": 1.01975846, + "epoch": 0.6493116094510912, + "flos": 16946820702720.0, + "grad_norm": 2.337675920711497, + "language_loss": 0.72154844, + "learning_rate": 1.1570782366925453e-06, + "loss": 0.74323601, + "num_input_tokens_seen": 116226425, + "step": 5400, + "time_per_iteration": 2.8201112747192383 + }, + { + "auxiliary_loss_clip": 0.01153369, + "auxiliary_loss_mlp": 0.01031055, + "balance_loss_clip": 1.04249382, + "balance_loss_mlp": 1.02285957, + "epoch": 0.6494318523417303, + "flos": 18802935072000.0, + "grad_norm": 2.5865656981110035, + "language_loss": 0.75498331, + "learning_rate": 1.1563718928690132e-06, + "loss": 0.77682757, + "num_input_tokens_seen": 116243860, + "step": 5401, + "time_per_iteration": 2.7919139862060547 + }, + { + "auxiliary_loss_clip": 0.01145532, + "auxiliary_loss_mlp": 0.01031232, + "balance_loss_clip": 1.04725003, + "balance_loss_mlp": 1.02269101, + "epoch": 0.6495520952323693, + "flos": 18982847318400.0, + "grad_norm": 2.0450550353951806, + "language_loss": 0.71292764, + "learning_rate": 1.1556656770231942e-06, + "loss": 0.73469532, + "num_input_tokens_seen": 116260055, + "step": 5402, + "time_per_iteration": 2.799975633621216 + }, + { + "auxiliary_loss_clip": 0.01161356, + "auxiliary_loss_mlp": 0.01024362, + "balance_loss_clip": 1.04388738, + "balance_loss_mlp": 1.01685214, + "epoch": 0.6496723381230085, + "flos": 22745388032640.0, + "grad_norm": 1.8480975245248696, + "language_loss": 0.76003551, + "learning_rate": 1.1549595892622207e-06, + "loss": 0.78189266, + "num_input_tokens_seen": 116278825, + "step": 5403, + "time_per_iteration": 2.7377305030822754 + }, + { + "auxiliary_loss_clip": 0.01035447, + "auxiliary_loss_mlp": 0.01004778, + "balance_loss_clip": 1.01187921, + "balance_loss_mlp": 1.00328815, + "epoch": 0.6497925810136476, + "flos": 62145283887360.0, + "grad_norm": 0.8324624527239168, + "language_loss": 0.58968121, + "learning_rate": 1.1542536296932047e-06, + "loss": 0.61008346, + "num_input_tokens_seen": 116342360, + "step": 5404, + "time_per_iteration": 3.4773757457733154 + }, + { + "auxiliary_loss_clip": 0.01150515, + "auxiliary_loss_mlp": 0.01031953, + "balance_loss_clip": 1.04373252, + "balance_loss_mlp": 1.02339375, + "epoch": 0.6499128239042866, + "flos": 20156731695360.0, + "grad_norm": 1.8800695896302622, + "language_loss": 0.70463002, + "learning_rate": 1.1535477984232414e-06, + "loss": 0.72645462, + "num_input_tokens_seen": 116362235, + "step": 5405, + "time_per_iteration": 2.9435815811157227 + }, + { + "auxiliary_loss_clip": 0.0112678, + "auxiliary_loss_mlp": 0.01026002, + "balance_loss_clip": 1.03867304, + "balance_loss_mlp": 1.01858723, + "epoch": 0.6500330667949258, + "flos": 24462420940800.0, + "grad_norm": 1.863019466528934, + "language_loss": 0.76930201, + "learning_rate": 1.152842095559404e-06, + "loss": 0.79082984, + "num_input_tokens_seen": 116382895, + "step": 5406, + "time_per_iteration": 2.9022510051727295 + }, + { + "auxiliary_loss_clip": 0.01158232, + "auxiliary_loss_mlp": 0.01023535, + "balance_loss_clip": 1.04637027, + "balance_loss_mlp": 1.01618576, + "epoch": 0.6501533096855648, + "flos": 25477399549440.0, + "grad_norm": 1.6698787143411127, + "language_loss": 0.76656657, + "learning_rate": 1.1521365212087474e-06, + "loss": 0.7883842, + "num_input_tokens_seen": 116402880, + "step": 5407, + "time_per_iteration": 2.9218404293060303 + }, + { + "auxiliary_loss_clip": 0.01162231, + "auxiliary_loss_mlp": 0.01022621, + "balance_loss_clip": 1.04616404, + "balance_loss_mlp": 1.01464581, + "epoch": 0.6502735525762039, + "flos": 44819245347840.0, + "grad_norm": 1.625802588284715, + "language_loss": 0.70528454, + "learning_rate": 1.1514310754783062e-06, + "loss": 0.72713304, + "num_input_tokens_seen": 116425830, + "step": 5408, + "time_per_iteration": 2.988680601119995 + }, + { + "auxiliary_loss_clip": 0.01154953, + "auxiliary_loss_mlp": 0.010263, + "balance_loss_clip": 1.04599643, + "balance_loss_mlp": 1.01842308, + "epoch": 0.6503937954668431, + "flos": 28658546726400.0, + "grad_norm": 2.0363449319068834, + "language_loss": 0.7318055, + "learning_rate": 1.1507257584750964e-06, + "loss": 0.753618, + "num_input_tokens_seen": 116446010, + "step": 5409, + "time_per_iteration": 2.8196356296539307 + }, + { + "auxiliary_loss_clip": 0.01172252, + "auxiliary_loss_mlp": 0.01029477, + "balance_loss_clip": 1.0489459, + "balance_loss_mlp": 1.02156126, + "epoch": 0.6505140383574821, + "flos": 20922562592640.0, + "grad_norm": 2.546088096924371, + "language_loss": 0.77525002, + "learning_rate": 1.150020570306113e-06, + "loss": 0.79726732, + "num_input_tokens_seen": 116465150, + "step": 5410, + "time_per_iteration": 2.6738343238830566 + }, + { + "auxiliary_loss_clip": 0.01147899, + "auxiliary_loss_mlp": 0.01028349, + "balance_loss_clip": 1.04249001, + "balance_loss_mlp": 1.01980186, + "epoch": 0.6506342812481212, + "flos": 20595236929920.0, + "grad_norm": 1.6845731135129098, + "language_loss": 0.74764925, + "learning_rate": 1.1493155110783338e-06, + "loss": 0.76941168, + "num_input_tokens_seen": 116483675, + "step": 5411, + "time_per_iteration": 2.8046793937683105 + }, + { + "auxiliary_loss_clip": 0.01164453, + "auxiliary_loss_mlp": 0.01025735, + "balance_loss_clip": 1.04903257, + "balance_loss_mlp": 1.01792109, + "epoch": 0.6507545241387603, + "flos": 30226478279040.0, + "grad_norm": 2.047753610567213, + "language_loss": 0.70396012, + "learning_rate": 1.1486105808987155e-06, + "loss": 0.72586203, + "num_input_tokens_seen": 116505165, + "step": 5412, + "time_per_iteration": 2.8170101642608643 + }, + { + "auxiliary_loss_clip": 0.01164714, + "auxiliary_loss_mlp": 0.0102735, + "balance_loss_clip": 1.04975033, + "balance_loss_mlp": 1.01926184, + "epoch": 0.6508747670293994, + "flos": 17128241320320.0, + "grad_norm": 1.8073820363223212, + "language_loss": 0.81110418, + "learning_rate": 1.1479057798741947e-06, + "loss": 0.83302474, + "num_input_tokens_seen": 116523220, + "step": 5413, + "time_per_iteration": 2.7606418132781982 + }, + { + "auxiliary_loss_clip": 0.01050053, + "auxiliary_loss_mlp": 0.01002084, + "balance_loss_clip": 1.01247311, + "balance_loss_mlp": 1.00053453, + "epoch": 0.6509950099200384, + "flos": 68559826573440.0, + "grad_norm": 0.7853008888824019, + "language_loss": 0.53377438, + "learning_rate": 1.14720110811169e-06, + "loss": 0.55429578, + "num_input_tokens_seen": 116580450, + "step": 5414, + "time_per_iteration": 5.052164316177368 + }, + { + "auxiliary_loss_clip": 0.01166568, + "auxiliary_loss_mlp": 0.01031181, + "balance_loss_clip": 1.04931593, + "balance_loss_mlp": 1.02276504, + "epoch": 0.6511152528106776, + "flos": 22347462188160.0, + "grad_norm": 1.9871776209287957, + "language_loss": 0.77176422, + "learning_rate": 1.146496565718098e-06, + "loss": 0.7937417, + "num_input_tokens_seen": 116601020, + "step": 5415, + "time_per_iteration": 2.744611978530884 + }, + { + "auxiliary_loss_clip": 0.01150548, + "auxiliary_loss_mlp": 0.01027651, + "balance_loss_clip": 1.0466485, + "balance_loss_mlp": 1.0197593, + "epoch": 0.6512354957013167, + "flos": 20522158709760.0, + "grad_norm": 2.4755262159518017, + "language_loss": 0.75604886, + "learning_rate": 1.1457921528002996e-06, + "loss": 0.77783084, + "num_input_tokens_seen": 116619455, + "step": 5416, + "time_per_iteration": 2.7219254970550537 + }, + { + "auxiliary_loss_clip": 0.01172103, + "auxiliary_loss_mlp": 0.00885958, + "balance_loss_clip": 1.04848218, + "balance_loss_mlp": 1.00057101, + "epoch": 0.6513557385919557, + "flos": 32337342881280.0, + "grad_norm": 2.638311108394784, + "language_loss": 0.71925634, + "learning_rate": 1.1450878694651522e-06, + "loss": 0.73983693, + "num_input_tokens_seen": 116640020, + "step": 5417, + "time_per_iteration": 3.647627115249634 + }, + { + "auxiliary_loss_clip": 0.01133379, + "auxiliary_loss_mlp": 0.01023737, + "balance_loss_clip": 1.04153538, + "balance_loss_mlp": 1.0158217, + "epoch": 0.6514759814825949, + "flos": 12093206417280.0, + "grad_norm": 2.445208962905718, + "language_loss": 0.6288898, + "learning_rate": 1.1443837158194954e-06, + "loss": 0.65046096, + "num_input_tokens_seen": 116655165, + "step": 5418, + "time_per_iteration": 2.76523756980896 + }, + { + "auxiliary_loss_clip": 0.01143976, + "auxiliary_loss_mlp": 0.01029307, + "balance_loss_clip": 1.04900765, + "balance_loss_mlp": 1.02182055, + "epoch": 0.651596224373234, + "flos": 22526907557760.0, + "grad_norm": 1.6657310194471222, + "language_loss": 0.74354273, + "learning_rate": 1.1436796919701484e-06, + "loss": 0.76527554, + "num_input_tokens_seen": 116673880, + "step": 5419, + "time_per_iteration": 3.6356146335601807 + }, + { + "auxiliary_loss_clip": 0.01152231, + "auxiliary_loss_mlp": 0.01030166, + "balance_loss_clip": 1.04734659, + "balance_loss_mlp": 1.02225661, + "epoch": 0.651716467263873, + "flos": 27818955250560.0, + "grad_norm": 1.8875176196797285, + "language_loss": 0.61939162, + "learning_rate": 1.1429757980239115e-06, + "loss": 0.64121556, + "num_input_tokens_seen": 116694305, + "step": 5420, + "time_per_iteration": 2.8247828483581543 + }, + { + "auxiliary_loss_clip": 0.01177485, + "auxiliary_loss_mlp": 0.01030962, + "balance_loss_clip": 1.05056751, + "balance_loss_mlp": 1.02193153, + "epoch": 0.6518367101545122, + "flos": 24316300414080.0, + "grad_norm": 2.5571440490289614, + "language_loss": 0.81529391, + "learning_rate": 1.1422720340875636e-06, + "loss": 0.83737832, + "num_input_tokens_seen": 116713055, + "step": 5421, + "time_per_iteration": 2.6096267700195312 + }, + { + "auxiliary_loss_clip": 0.0117049, + "auxiliary_loss_mlp": 0.01026697, + "balance_loss_clip": 1.04856348, + "balance_loss_mlp": 1.01861453, + "epoch": 0.6519569530451512, + "flos": 20011939971840.0, + "grad_norm": 2.223583491511272, + "language_loss": 0.78953928, + "learning_rate": 1.1415684002678671e-06, + "loss": 0.81151116, + "num_input_tokens_seen": 116731815, + "step": 5422, + "time_per_iteration": 2.824720859527588 + }, + { + "auxiliary_loss_clip": 0.01153849, + "auxiliary_loss_mlp": 0.01030414, + "balance_loss_clip": 1.04371119, + "balance_loss_mlp": 1.02178335, + "epoch": 0.6520771959357903, + "flos": 21576064682880.0, + "grad_norm": 2.2943319001612505, + "language_loss": 0.77877921, + "learning_rate": 1.1408648966715617e-06, + "loss": 0.80062187, + "num_input_tokens_seen": 116749335, + "step": 5423, + "time_per_iteration": 2.7509915828704834 + }, + { + "auxiliary_loss_clip": 0.01154945, + "auxiliary_loss_mlp": 0.01029664, + "balance_loss_clip": 1.04360175, + "balance_loss_mlp": 1.02138805, + "epoch": 0.6521974388264293, + "flos": 22711021695360.0, + "grad_norm": 1.741907829387333, + "language_loss": 0.72448051, + "learning_rate": 1.1401615234053683e-06, + "loss": 0.74632663, + "num_input_tokens_seen": 116768155, + "step": 5424, + "time_per_iteration": 2.768956184387207 + }, + { + "auxiliary_loss_clip": 0.01152814, + "auxiliary_loss_mlp": 0.01026736, + "balance_loss_clip": 1.04391575, + "balance_loss_mlp": 1.01800358, + "epoch": 0.6523176817170685, + "flos": 23002939526400.0, + "grad_norm": 2.0317089094897183, + "language_loss": 0.75807017, + "learning_rate": 1.1394582805759885e-06, + "loss": 0.77986568, + "num_input_tokens_seen": 116787435, + "step": 5425, + "time_per_iteration": 2.7360048294067383 + }, + { + "auxiliary_loss_clip": 0.01164669, + "auxiliary_loss_mlp": 0.01024038, + "balance_loss_clip": 1.04960811, + "balance_loss_mlp": 1.01629555, + "epoch": 0.6524379246077076, + "flos": 21688249835520.0, + "grad_norm": 1.8760128133211733, + "language_loss": 0.75841939, + "learning_rate": 1.1387551682901022e-06, + "loss": 0.78030646, + "num_input_tokens_seen": 116808040, + "step": 5426, + "time_per_iteration": 2.7438113689422607 + }, + { + "auxiliary_loss_clip": 0.0114075, + "auxiliary_loss_mlp": 0.01026377, + "balance_loss_clip": 1.04430819, + "balance_loss_mlp": 1.01821733, + "epoch": 0.6525581674983466, + "flos": 19390936711680.0, + "grad_norm": 2.4491421940902427, + "language_loss": 0.71040571, + "learning_rate": 1.138052186654373e-06, + "loss": 0.73207688, + "num_input_tokens_seen": 116825510, + "step": 5427, + "time_per_iteration": 2.8076932430267334 + }, + { + "auxiliary_loss_clip": 0.01157851, + "auxiliary_loss_mlp": 0.01037113, + "balance_loss_clip": 1.04674172, + "balance_loss_mlp": 1.02852976, + "epoch": 0.6526784103889858, + "flos": 17165444832000.0, + "grad_norm": 1.9710626013471333, + "language_loss": 0.88058686, + "learning_rate": 1.1373493357754417e-06, + "loss": 0.90253651, + "num_input_tokens_seen": 116844415, + "step": 5428, + "time_per_iteration": 2.7775776386260986 + }, + { + "auxiliary_loss_clip": 0.0116989, + "auxiliary_loss_mlp": 0.0102126, + "balance_loss_clip": 1.04686236, + "balance_loss_mlp": 1.01387501, + "epoch": 0.6527986532796248, + "flos": 18989168112000.0, + "grad_norm": 1.7056254174762229, + "language_loss": 0.77328581, + "learning_rate": 1.1366466157599303e-06, + "loss": 0.79519731, + "num_input_tokens_seen": 116863690, + "step": 5429, + "time_per_iteration": 2.6956498622894287 + }, + { + "auxiliary_loss_clip": 0.01131576, + "auxiliary_loss_mlp": 0.00885535, + "balance_loss_clip": 1.04216969, + "balance_loss_mlp": 1.00053334, + "epoch": 0.6529188961702639, + "flos": 14238581011200.0, + "grad_norm": 2.1506374585829886, + "language_loss": 0.75902009, + "learning_rate": 1.1359440267144412e-06, + "loss": 0.77919114, + "num_input_tokens_seen": 116881145, + "step": 5430, + "time_per_iteration": 2.7854647636413574 + }, + { + "auxiliary_loss_clip": 0.01164149, + "auxiliary_loss_mlp": 0.01030733, + "balance_loss_clip": 1.04594517, + "balance_loss_mlp": 1.02274597, + "epoch": 0.653039139060903, + "flos": 36682929158400.0, + "grad_norm": 1.8705133418236015, + "language_loss": 0.739402, + "learning_rate": 1.1352415687455556e-06, + "loss": 0.76135081, + "num_input_tokens_seen": 116902405, + "step": 5431, + "time_per_iteration": 2.878143787384033 + }, + { + "auxiliary_loss_clip": 0.01161093, + "auxiliary_loss_mlp": 0.01030232, + "balance_loss_clip": 1.04697132, + "balance_loss_mlp": 1.02098775, + "epoch": 0.6531593819515421, + "flos": 25376275785600.0, + "grad_norm": 2.335258675567611, + "language_loss": 0.63275713, + "learning_rate": 1.1345392419598362e-06, + "loss": 0.65467036, + "num_input_tokens_seen": 116921285, + "step": 5432, + "time_per_iteration": 2.7506027221679688 + }, + { + "auxiliary_loss_clip": 0.0115484, + "auxiliary_loss_mlp": 0.01030697, + "balance_loss_clip": 1.04435468, + "balance_loss_mlp": 1.02179813, + "epoch": 0.6532796248421812, + "flos": 21178533888000.0, + "grad_norm": 1.7303237209153497, + "language_loss": 0.72007072, + "learning_rate": 1.1338370464638263e-06, + "loss": 0.74192607, + "num_input_tokens_seen": 116940685, + "step": 5433, + "time_per_iteration": 2.7216038703918457 + }, + { + "auxiliary_loss_clip": 0.01170191, + "auxiliary_loss_mlp": 0.01030422, + "balance_loss_clip": 1.04655981, + "balance_loss_mlp": 1.02230406, + "epoch": 0.6533998677328203, + "flos": 17675950878720.0, + "grad_norm": 1.961227487972948, + "language_loss": 0.64006025, + "learning_rate": 1.1331349823640474e-06, + "loss": 0.66206634, + "num_input_tokens_seen": 116958115, + "step": 5434, + "time_per_iteration": 2.669193983078003 + }, + { + "auxiliary_loss_clip": 0.01165379, + "auxiliary_loss_mlp": 0.00885329, + "balance_loss_clip": 1.04802275, + "balance_loss_mlp": 1.00054574, + "epoch": 0.6535201106234594, + "flos": 28400384701440.0, + "grad_norm": 4.806765221187544, + "language_loss": 0.78376013, + "learning_rate": 1.132433049767003e-06, + "loss": 0.80426717, + "num_input_tokens_seen": 116976030, + "step": 5435, + "time_per_iteration": 2.810676336288452 + }, + { + "auxiliary_loss_clip": 0.01152234, + "auxiliary_loss_mlp": 0.0102576, + "balance_loss_clip": 1.04600215, + "balance_loss_mlp": 1.01770735, + "epoch": 0.6536403535140984, + "flos": 23586667447680.0, + "grad_norm": 1.5784824860936515, + "language_loss": 0.81261826, + "learning_rate": 1.1317312487791748e-06, + "loss": 0.83439821, + "num_input_tokens_seen": 116997680, + "step": 5436, + "time_per_iteration": 2.803196430206299 + }, + { + "auxiliary_loss_clip": 0.01155779, + "auxiliary_loss_mlp": 0.01028596, + "balance_loss_clip": 1.04443359, + "balance_loss_mlp": 1.02036452, + "epoch": 0.6537605964047376, + "flos": 21579476474880.0, + "grad_norm": 2.101565436559297, + "language_loss": 0.72970974, + "learning_rate": 1.1310295795070253e-06, + "loss": 0.75155348, + "num_input_tokens_seen": 117017620, + "step": 5437, + "time_per_iteration": 2.7640209197998047 + }, + { + "auxiliary_loss_clip": 0.01132692, + "auxiliary_loss_mlp": 0.01029409, + "balance_loss_clip": 1.04316688, + "balance_loss_mlp": 1.02171993, + "epoch": 0.6538808392953767, + "flos": 26833997433600.0, + "grad_norm": 1.8035673001744346, + "language_loss": 0.80858696, + "learning_rate": 1.1303280420569982e-06, + "loss": 0.83020794, + "num_input_tokens_seen": 117039505, + "step": 5438, + "time_per_iteration": 2.966024875640869 + }, + { + "auxiliary_loss_clip": 0.01156978, + "auxiliary_loss_mlp": 0.01030091, + "balance_loss_clip": 1.04508543, + "balance_loss_mlp": 1.02219033, + "epoch": 0.6540010821860157, + "flos": 30738241301760.0, + "grad_norm": 1.5578268666458646, + "language_loss": 0.77223581, + "learning_rate": 1.1296266365355158e-06, + "loss": 0.79410648, + "num_input_tokens_seen": 117062890, + "step": 5439, + "time_per_iteration": 2.8033173084259033 + }, + { + "auxiliary_loss_clip": 0.01147157, + "auxiliary_loss_mlp": 0.01024983, + "balance_loss_clip": 1.04679, + "balance_loss_mlp": 1.0166024, + "epoch": 0.6541213250766549, + "flos": 26907147480960.0, + "grad_norm": 2.1595478133288317, + "language_loss": 0.73498774, + "learning_rate": 1.1289253630489806e-06, + "loss": 0.75670916, + "num_input_tokens_seen": 117083940, + "step": 5440, + "time_per_iteration": 4.768436670303345 + }, + { + "auxiliary_loss_clip": 0.01170424, + "auxiliary_loss_mlp": 0.01030722, + "balance_loss_clip": 1.04794574, + "balance_loss_mlp": 1.02216268, + "epoch": 0.6542415679672939, + "flos": 19172384409600.0, + "grad_norm": 2.6924566383393667, + "language_loss": 0.720043, + "learning_rate": 1.1282242217037753e-06, + "loss": 0.74205446, + "num_input_tokens_seen": 117101440, + "step": 5441, + "time_per_iteration": 2.7105517387390137 + }, + { + "auxiliary_loss_clip": 0.0112625, + "auxiliary_loss_mlp": 0.01026973, + "balance_loss_clip": 1.0375433, + "balance_loss_mlp": 1.01875341, + "epoch": 0.654361810857933, + "flos": 48173517100800.0, + "grad_norm": 1.8795671658619195, + "language_loss": 0.61823887, + "learning_rate": 1.127523212606262e-06, + "loss": 0.6397711, + "num_input_tokens_seen": 117124265, + "step": 5442, + "time_per_iteration": 3.10695481300354 + }, + { + "auxiliary_loss_clip": 0.01159795, + "auxiliary_loss_mlp": 0.0102799, + "balance_loss_clip": 1.04658318, + "balance_loss_mlp": 1.02034831, + "epoch": 0.6544820537485722, + "flos": 26943165843840.0, + "grad_norm": 1.710831594144942, + "language_loss": 0.73125094, + "learning_rate": 1.1268223358627835e-06, + "loss": 0.75312877, + "num_input_tokens_seen": 117146755, + "step": 5443, + "time_per_iteration": 3.7140438556671143 + }, + { + "auxiliary_loss_clip": 0.01172714, + "auxiliary_loss_mlp": 0.01025045, + "balance_loss_clip": 1.04796433, + "balance_loss_mlp": 1.01738882, + "epoch": 0.6546022966392112, + "flos": 20886328748160.0, + "grad_norm": 2.1038032066081906, + "language_loss": 0.71568894, + "learning_rate": 1.126121591579663e-06, + "loss": 0.73766655, + "num_input_tokens_seen": 117165960, + "step": 5444, + "time_per_iteration": 2.704573392868042 + }, + { + "auxiliary_loss_clip": 0.01159179, + "auxiliary_loss_mlp": 0.01024073, + "balance_loss_clip": 1.04835165, + "balance_loss_mlp": 1.01644397, + "epoch": 0.6547225395298503, + "flos": 24936693143040.0, + "grad_norm": 1.5872691134870878, + "language_loss": 0.69023526, + "learning_rate": 1.1254209798632018e-06, + "loss": 0.71206772, + "num_input_tokens_seen": 117186980, + "step": 5445, + "time_per_iteration": 3.6768639087677 + }, + { + "auxiliary_loss_clip": 0.01115413, + "auxiliary_loss_mlp": 0.01022919, + "balance_loss_clip": 1.03952003, + "balance_loss_mlp": 1.01519418, + "epoch": 0.6548427824204894, + "flos": 22565942663040.0, + "grad_norm": 1.5784018070873853, + "language_loss": 0.84628743, + "learning_rate": 1.124720500819683e-06, + "loss": 0.86767077, + "num_input_tokens_seen": 117205135, + "step": 5446, + "time_per_iteration": 2.9594318866729736 + }, + { + "auxiliary_loss_clip": 0.01171949, + "auxiliary_loss_mlp": 0.01030703, + "balance_loss_clip": 1.048967, + "balance_loss_mlp": 1.02275229, + "epoch": 0.6549630253111285, + "flos": 18442500048000.0, + "grad_norm": 1.8399574384835247, + "language_loss": 0.82346034, + "learning_rate": 1.1240201545553682e-06, + "loss": 0.84548688, + "num_input_tokens_seen": 117222935, + "step": 5447, + "time_per_iteration": 2.650996446609497 + }, + { + "auxiliary_loss_clip": 0.01142304, + "auxiliary_loss_mlp": 0.0103368, + "balance_loss_clip": 1.04452682, + "balance_loss_mlp": 1.02543032, + "epoch": 0.6550832682017675, + "flos": 25187313312000.0, + "grad_norm": 1.8065635456557374, + "language_loss": 0.72997206, + "learning_rate": 1.1233199411764987e-06, + "loss": 0.75173193, + "num_input_tokens_seen": 117242370, + "step": 5448, + "time_per_iteration": 2.8159403800964355 + }, + { + "auxiliary_loss_clip": 0.01130566, + "auxiliary_loss_mlp": 0.01027397, + "balance_loss_clip": 1.04110694, + "balance_loss_mlp": 1.01957083, + "epoch": 0.6552035110924067, + "flos": 22748153379840.0, + "grad_norm": 1.8477687346171225, + "language_loss": 0.68718243, + "learning_rate": 1.1226198607892978e-06, + "loss": 0.70876205, + "num_input_tokens_seen": 117262930, + "step": 5449, + "time_per_iteration": 2.9043734073638916 + }, + { + "auxiliary_loss_clip": 0.0113873, + "auxiliary_loss_mlp": 0.01026821, + "balance_loss_clip": 1.04685903, + "balance_loss_mlp": 1.0182972, + "epoch": 0.6553237539830458, + "flos": 21799178012160.0, + "grad_norm": 1.779987366873609, + "language_loss": 0.79996026, + "learning_rate": 1.1219199134999664e-06, + "loss": 0.8216157, + "num_input_tokens_seen": 117281430, + "step": 5450, + "time_per_iteration": 2.9162404537200928 + }, + { + "auxiliary_loss_clip": 0.01151359, + "auxiliary_loss_mlp": 0.0103073, + "balance_loss_clip": 1.04779458, + "balance_loss_mlp": 1.02253413, + "epoch": 0.6554439968736848, + "flos": 20887226588160.0, + "grad_norm": 2.4712627391711415, + "language_loss": 0.78711051, + "learning_rate": 1.1212200994146863e-06, + "loss": 0.80893141, + "num_input_tokens_seen": 117299185, + "step": 5451, + "time_per_iteration": 2.8828911781311035 + }, + { + "auxiliary_loss_clip": 0.01139536, + "auxiliary_loss_mlp": 0.01025159, + "balance_loss_clip": 1.03907537, + "balance_loss_mlp": 1.01705241, + "epoch": 0.655564239764324, + "flos": 16139045698560.0, + "grad_norm": 2.026585795791069, + "language_loss": 0.75559437, + "learning_rate": 1.120520418639618e-06, + "loss": 0.77724129, + "num_input_tokens_seen": 117317720, + "step": 5452, + "time_per_iteration": 2.8414595127105713 + }, + { + "auxiliary_loss_clip": 0.0116375, + "auxiliary_loss_mlp": 0.01027338, + "balance_loss_clip": 1.04996359, + "balance_loss_mlp": 1.01948798, + "epoch": 0.655684482654963, + "flos": 29570354496000.0, + "grad_norm": 2.070279598615259, + "language_loss": 0.83460474, + "learning_rate": 1.119820871280903e-06, + "loss": 0.85651559, + "num_input_tokens_seen": 117338795, + "step": 5453, + "time_per_iteration": 2.84799861907959 + }, + { + "auxiliary_loss_clip": 0.01161572, + "auxiliary_loss_mlp": 0.01037733, + "balance_loss_clip": 1.04633069, + "balance_loss_mlp": 1.02981126, + "epoch": 0.6558047255456021, + "flos": 29789409588480.0, + "grad_norm": 1.8756861958790083, + "language_loss": 0.7357794, + "learning_rate": 1.1191214574446614e-06, + "loss": 0.75777245, + "num_input_tokens_seen": 117359040, + "step": 5454, + "time_per_iteration": 2.8627853393554688 + }, + { + "auxiliary_loss_clip": 0.01148485, + "auxiliary_loss_mlp": 0.01029769, + "balance_loss_clip": 1.04468191, + "balance_loss_mlp": 1.02184725, + "epoch": 0.6559249684362413, + "flos": 29059166090880.0, + "grad_norm": 1.4045598183551664, + "language_loss": 0.80049098, + "learning_rate": 1.118422177236995e-06, + "loss": 0.82227355, + "num_input_tokens_seen": 117380865, + "step": 5455, + "time_per_iteration": 2.9176011085510254 + }, + { + "auxiliary_loss_clip": 0.01153647, + "auxiliary_loss_mlp": 0.01029947, + "balance_loss_clip": 1.04460621, + "balance_loss_mlp": 1.02120292, + "epoch": 0.6560452113268803, + "flos": 20225464369920.0, + "grad_norm": 2.1874458146753097, + "language_loss": 0.85904449, + "learning_rate": 1.1177230307639835e-06, + "loss": 0.88088048, + "num_input_tokens_seen": 117398405, + "step": 5456, + "time_per_iteration": 2.7917916774749756 + }, + { + "auxiliary_loss_clip": 0.01140785, + "auxiliary_loss_mlp": 0.01026815, + "balance_loss_clip": 1.04331708, + "balance_loss_mlp": 1.01890588, + "epoch": 0.6561654542175194, + "flos": 25045538330880.0, + "grad_norm": 2.059355650265737, + "language_loss": 0.78679973, + "learning_rate": 1.1170240181316865e-06, + "loss": 0.80847573, + "num_input_tokens_seen": 117419850, + "step": 5457, + "time_per_iteration": 2.8838884830474854 + }, + { + "auxiliary_loss_clip": 0.01140624, + "auxiliary_loss_mlp": 0.01025017, + "balance_loss_clip": 1.04180384, + "balance_loss_mlp": 1.01700044, + "epoch": 0.6562856971081584, + "flos": 22856711258880.0, + "grad_norm": 2.1219828229951054, + "language_loss": 0.79186475, + "learning_rate": 1.1163251394461442e-06, + "loss": 0.81352115, + "num_input_tokens_seen": 117438330, + "step": 5458, + "time_per_iteration": 2.808412551879883 + }, + { + "auxiliary_loss_clip": 0.01160003, + "auxiliary_loss_mlp": 0.01020252, + "balance_loss_clip": 1.0461967, + "balance_loss_mlp": 1.01238418, + "epoch": 0.6564059399987976, + "flos": 18872565586560.0, + "grad_norm": 2.072961548157021, + "language_loss": 0.82181925, + "learning_rate": 1.1156263948133746e-06, + "loss": 0.84362179, + "num_input_tokens_seen": 117454985, + "step": 5459, + "time_per_iteration": 2.74422287940979 + }, + { + "auxiliary_loss_clip": 0.01122983, + "auxiliary_loss_mlp": 0.00886461, + "balance_loss_clip": 1.04126906, + "balance_loss_mlp": 1.00059414, + "epoch": 0.6565261828894366, + "flos": 25484187219840.0, + "grad_norm": 1.7862711025309987, + "language_loss": 0.77777916, + "learning_rate": 1.1149277843393787e-06, + "loss": 0.79787362, + "num_input_tokens_seen": 117476145, + "step": 5460, + "time_per_iteration": 2.8402252197265625 + }, + { + "auxiliary_loss_clip": 0.01113262, + "auxiliary_loss_mlp": 0.00886147, + "balance_loss_clip": 1.03690529, + "balance_loss_mlp": 1.0005995, + "epoch": 0.6566464257800757, + "flos": 19683500987520.0, + "grad_norm": 2.128093359812819, + "language_loss": 0.63571137, + "learning_rate": 1.1142293081301342e-06, + "loss": 0.65570545, + "num_input_tokens_seen": 117494025, + "step": 5461, + "time_per_iteration": 2.9532530307769775 + }, + { + "auxiliary_loss_clip": 0.01149068, + "auxiliary_loss_mlp": 0.0102685, + "balance_loss_clip": 1.04558325, + "balance_loss_mlp": 1.01962268, + "epoch": 0.6567666686707149, + "flos": 23514127931520.0, + "grad_norm": 1.5341245420629166, + "language_loss": 0.67707682, + "learning_rate": 1.1135309662915995e-06, + "loss": 0.69883597, + "num_input_tokens_seen": 117514190, + "step": 5462, + "time_per_iteration": 2.7328860759735107 + }, + { + "auxiliary_loss_clip": 0.01134363, + "auxiliary_loss_mlp": 0.0101903, + "balance_loss_clip": 1.04194474, + "balance_loss_mlp": 1.01113188, + "epoch": 0.6568869115613539, + "flos": 32781342896640.0, + "grad_norm": 1.9738858413338647, + "language_loss": 0.60393405, + "learning_rate": 1.112832758929712e-06, + "loss": 0.6254679, + "num_input_tokens_seen": 117536800, + "step": 5463, + "time_per_iteration": 2.9186747074127197 + }, + { + "auxiliary_loss_clip": 0.01161164, + "auxiliary_loss_mlp": 0.01025151, + "balance_loss_clip": 1.04679036, + "balance_loss_mlp": 1.01676488, + "epoch": 0.657007154451993, + "flos": 18442428220800.0, + "grad_norm": 2.1045083231418484, + "language_loss": 0.74961269, + "learning_rate": 1.11213468615039e-06, + "loss": 0.77147579, + "num_input_tokens_seen": 117556230, + "step": 5464, + "time_per_iteration": 2.6536002159118652 + }, + { + "auxiliary_loss_clip": 0.01112274, + "auxiliary_loss_mlp": 0.01028951, + "balance_loss_clip": 1.03869605, + "balance_loss_mlp": 1.02166152, + "epoch": 0.6571273973426321, + "flos": 25156717902720.0, + "grad_norm": 1.62702624727639, + "language_loss": 0.75680417, + "learning_rate": 1.1114367480595292e-06, + "loss": 0.77821648, + "num_input_tokens_seen": 117577310, + "step": 5465, + "time_per_iteration": 3.863784074783325 + }, + { + "auxiliary_loss_clip": 0.01121203, + "auxiliary_loss_mlp": 0.01030503, + "balance_loss_clip": 1.03991675, + "balance_loss_mlp": 1.02211058, + "epoch": 0.6572476402332712, + "flos": 17529830352000.0, + "grad_norm": 2.0557260841464213, + "language_loss": 0.81285912, + "learning_rate": 1.1107389447630086e-06, + "loss": 0.83437616, + "num_input_tokens_seen": 117596010, + "step": 5466, + "time_per_iteration": 3.863231897354126 + }, + { + "auxiliary_loss_clip": 0.0114594, + "auxiliary_loss_mlp": 0.00885516, + "balance_loss_clip": 1.04141688, + "balance_loss_mlp": 1.00061107, + "epoch": 0.6573678831239103, + "flos": 17014260487680.0, + "grad_norm": 1.980786514144899, + "language_loss": 0.78389806, + "learning_rate": 1.1100412763666818e-06, + "loss": 0.80421263, + "num_input_tokens_seen": 117611270, + "step": 5467, + "time_per_iteration": 2.730914354324341 + }, + { + "auxiliary_loss_clip": 0.01155144, + "auxiliary_loss_mlp": 0.01026339, + "balance_loss_clip": 1.04628265, + "balance_loss_mlp": 1.01844716, + "epoch": 0.6574881260145494, + "flos": 23910078528000.0, + "grad_norm": 1.5089265544937949, + "language_loss": 0.79877448, + "learning_rate": 1.1093437429763865e-06, + "loss": 0.8205893, + "num_input_tokens_seen": 117631535, + "step": 5468, + "time_per_iteration": 2.8218014240264893 + }, + { + "auxiliary_loss_clip": 0.01163354, + "auxiliary_loss_mlp": 0.01024384, + "balance_loss_clip": 1.0491178, + "balance_loss_mlp": 1.01740158, + "epoch": 0.6576083689051885, + "flos": 11218458504960.0, + "grad_norm": 2.064256050071582, + "language_loss": 0.73222691, + "learning_rate": 1.1086463446979361e-06, + "loss": 0.75410432, + "num_input_tokens_seen": 117649885, + "step": 5469, + "time_per_iteration": 3.630356788635254 + }, + { + "auxiliary_loss_clip": 0.01166466, + "auxiliary_loss_mlp": 0.01035326, + "balance_loss_clip": 1.04993212, + "balance_loss_mlp": 1.02680254, + "epoch": 0.6577286117958275, + "flos": 22455553190400.0, + "grad_norm": 1.7710231058708263, + "language_loss": 0.77393818, + "learning_rate": 1.1079490816371277e-06, + "loss": 0.79595608, + "num_input_tokens_seen": 117669650, + "step": 5470, + "time_per_iteration": 2.706542491912842 + }, + { + "auxiliary_loss_clip": 0.01161639, + "auxiliary_loss_mlp": 0.00886278, + "balance_loss_clip": 1.04520154, + "balance_loss_mlp": 1.0006671, + "epoch": 0.6578488546864667, + "flos": 21872184405120.0, + "grad_norm": 2.500056240350879, + "language_loss": 0.7446028, + "learning_rate": 1.1072519538997352e-06, + "loss": 0.765082, + "num_input_tokens_seen": 117688790, + "step": 5471, + "time_per_iteration": 2.6961517333984375 + }, + { + "auxiliary_loss_clip": 0.01153098, + "auxiliary_loss_mlp": 0.01028648, + "balance_loss_clip": 1.04205692, + "balance_loss_mlp": 1.02042258, + "epoch": 0.6579690975771058, + "flos": 23543753673600.0, + "grad_norm": 2.0921395805825647, + "language_loss": 0.82092023, + "learning_rate": 1.1065549615915095e-06, + "loss": 0.84273773, + "num_input_tokens_seen": 117708620, + "step": 5472, + "time_per_iteration": 3.6685121059417725 + }, + { + "auxiliary_loss_clip": 0.01164466, + "auxiliary_loss_mlp": 0.01036371, + "balance_loss_clip": 1.05149794, + "balance_loss_mlp": 1.02778792, + "epoch": 0.6580893404677448, + "flos": 32743995730560.0, + "grad_norm": 2.691281223334406, + "language_loss": 0.78039169, + "learning_rate": 1.105858104818187e-06, + "loss": 0.80240011, + "num_input_tokens_seen": 117729775, + "step": 5473, + "time_per_iteration": 2.8296761512756348 + }, + { + "auxiliary_loss_clip": 0.01165228, + "auxiliary_loss_mlp": 0.01028402, + "balance_loss_clip": 1.04813862, + "balance_loss_mlp": 1.02032542, + "epoch": 0.658209583358384, + "flos": 15888138220800.0, + "grad_norm": 2.5824662342486255, + "language_loss": 0.75094879, + "learning_rate": 1.105161383685478e-06, + "loss": 0.77288514, + "num_input_tokens_seen": 117746160, + "step": 5474, + "time_per_iteration": 2.683216094970703 + }, + { + "auxiliary_loss_clip": 0.01041935, + "auxiliary_loss_mlp": 0.01002494, + "balance_loss_clip": 1.00987506, + "balance_loss_mlp": 1.00105119, + "epoch": 0.658329826249023, + "flos": 62695902447360.0, + "grad_norm": 0.7242687830653236, + "language_loss": 0.56279689, + "learning_rate": 1.1044647982990771e-06, + "loss": 0.58324122, + "num_input_tokens_seen": 117808045, + "step": 5475, + "time_per_iteration": 3.3163130283355713 + }, + { + "auxiliary_loss_clip": 0.01157814, + "auxiliary_loss_mlp": 0.01025559, + "balance_loss_clip": 1.04856515, + "balance_loss_mlp": 1.01735175, + "epoch": 0.6584500691396621, + "flos": 31722624501120.0, + "grad_norm": 2.152461377066342, + "language_loss": 0.64737725, + "learning_rate": 1.1037683487646536e-06, + "loss": 0.66921103, + "num_input_tokens_seen": 117828330, + "step": 5476, + "time_per_iteration": 2.9030773639678955 + }, + { + "auxiliary_loss_clip": 0.01152214, + "auxiliary_loss_mlp": 0.00885026, + "balance_loss_clip": 1.04876459, + "balance_loss_mlp": 1.00066519, + "epoch": 0.6585703120303013, + "flos": 18406086635520.0, + "grad_norm": 2.039272295751717, + "language_loss": 0.76939464, + "learning_rate": 1.1030720351878583e-06, + "loss": 0.78976715, + "num_input_tokens_seen": 117846450, + "step": 5477, + "time_per_iteration": 2.754789352416992 + }, + { + "auxiliary_loss_clip": 0.01053238, + "auxiliary_loss_mlp": 0.01002133, + "balance_loss_clip": 1.01013279, + "balance_loss_mlp": 1.00049937, + "epoch": 0.6586905549209403, + "flos": 58309880434560.0, + "grad_norm": 0.8110221935211747, + "language_loss": 0.5765447, + "learning_rate": 1.102375857674323e-06, + "loss": 0.59709841, + "num_input_tokens_seen": 117908365, + "step": 5478, + "time_per_iteration": 3.290508985519409 + }, + { + "auxiliary_loss_clip": 0.01152745, + "auxiliary_loss_mlp": 0.0102486, + "balance_loss_clip": 1.04351914, + "balance_loss_mlp": 1.01702809, + "epoch": 0.6588107978115794, + "flos": 22782627457920.0, + "grad_norm": 1.6902595454272338, + "language_loss": 0.90199095, + "learning_rate": 1.1016798163296561e-06, + "loss": 0.92376697, + "num_input_tokens_seen": 117927565, + "step": 5479, + "time_per_iteration": 2.764519214630127 + }, + { + "auxiliary_loss_clip": 0.01155511, + "auxiliary_loss_mlp": 0.01023281, + "balance_loss_clip": 1.04533815, + "balance_loss_mlp": 1.0152638, + "epoch": 0.6589310407022185, + "flos": 20667525050880.0, + "grad_norm": 1.9026053360844584, + "language_loss": 0.65801919, + "learning_rate": 1.1009839112594471e-06, + "loss": 0.67980707, + "num_input_tokens_seen": 117945590, + "step": 5480, + "time_per_iteration": 2.707873821258545 + }, + { + "auxiliary_loss_clip": 0.01165051, + "auxiliary_loss_mlp": 0.0102808, + "balance_loss_clip": 1.04725528, + "balance_loss_mlp": 1.02006948, + "epoch": 0.6590512835928576, + "flos": 25630595055360.0, + "grad_norm": 1.9876932287259739, + "language_loss": 0.71792942, + "learning_rate": 1.1002881425692638e-06, + "loss": 0.73986071, + "num_input_tokens_seen": 117966020, + "step": 5481, + "time_per_iteration": 2.875127077102661 + }, + { + "auxiliary_loss_clip": 0.01154338, + "auxiliary_loss_mlp": 0.01028412, + "balance_loss_clip": 1.04320621, + "balance_loss_mlp": 1.01978135, + "epoch": 0.6591715264834966, + "flos": 23726108044800.0, + "grad_norm": 2.4519291935020964, + "language_loss": 0.75093967, + "learning_rate": 1.0995925103646532e-06, + "loss": 0.77276719, + "num_input_tokens_seen": 117984620, + "step": 5482, + "time_per_iteration": 2.7106595039367676 + }, + { + "auxiliary_loss_clip": 0.01137675, + "auxiliary_loss_mlp": 0.01029196, + "balance_loss_clip": 1.04405797, + "balance_loss_mlp": 1.02144742, + "epoch": 0.6592917693741358, + "flos": 35773850822400.0, + "grad_norm": 1.4895873130737312, + "language_loss": 0.66599631, + "learning_rate": 1.0988970147511437e-06, + "loss": 0.68766505, + "num_input_tokens_seen": 118006500, + "step": 5483, + "time_per_iteration": 2.915236234664917 + }, + { + "auxiliary_loss_clip": 0.01152317, + "auxiliary_loss_mlp": 0.01027757, + "balance_loss_clip": 1.04692388, + "balance_loss_mlp": 1.01953745, + "epoch": 0.6594120122647749, + "flos": 21396834794880.0, + "grad_norm": 2.106084134702577, + "language_loss": 0.80942714, + "learning_rate": 1.0982016558342405e-06, + "loss": 0.8312279, + "num_input_tokens_seen": 118025470, + "step": 5484, + "time_per_iteration": 2.7296087741851807 + }, + { + "auxiliary_loss_clip": 0.01174023, + "auxiliary_loss_mlp": 0.01023506, + "balance_loss_clip": 1.05002761, + "balance_loss_mlp": 1.01566172, + "epoch": 0.6595322551554139, + "flos": 19351829779200.0, + "grad_norm": 2.5501531437932794, + "language_loss": 0.70942891, + "learning_rate": 1.0975064337194291e-06, + "loss": 0.73140419, + "num_input_tokens_seen": 118043515, + "step": 5485, + "time_per_iteration": 2.655057191848755 + }, + { + "auxiliary_loss_clip": 0.01136773, + "auxiliary_loss_mlp": 0.01029222, + "balance_loss_clip": 1.04226899, + "balance_loss_mlp": 1.02107406, + "epoch": 0.6596524980460531, + "flos": 16837113588480.0, + "grad_norm": 1.6391893857884166, + "language_loss": 0.70431721, + "learning_rate": 1.0968113485121743e-06, + "loss": 0.72597718, + "num_input_tokens_seen": 118063105, + "step": 5486, + "time_per_iteration": 2.769280195236206 + }, + { + "auxiliary_loss_clip": 0.01162465, + "auxiliary_loss_mlp": 0.00886413, + "balance_loss_clip": 1.0446595, + "balance_loss_mlp": 1.00055695, + "epoch": 0.6597727409366921, + "flos": 21798567480960.0, + "grad_norm": 1.9804615759063997, + "language_loss": 0.80367321, + "learning_rate": 1.0961164003179185e-06, + "loss": 0.82416201, + "num_input_tokens_seen": 118081615, + "step": 5487, + "time_per_iteration": 2.7868826389312744 + }, + { + "auxiliary_loss_clip": 0.0113943, + "auxiliary_loss_mlp": 0.01026026, + "balance_loss_clip": 1.04108143, + "balance_loss_mlp": 1.01790822, + "epoch": 0.6598929838273312, + "flos": 23730704985600.0, + "grad_norm": 1.8148189237435104, + "language_loss": 0.83978313, + "learning_rate": 1.0954215892420884e-06, + "loss": 0.86143768, + "num_input_tokens_seen": 118102315, + "step": 5488, + "time_per_iteration": 2.7463624477386475 + }, + { + "auxiliary_loss_clip": 0.01149593, + "auxiliary_loss_mlp": 0.01035096, + "balance_loss_clip": 1.04740882, + "balance_loss_mlp": 1.02640533, + "epoch": 0.6600132267179702, + "flos": 19974520978560.0, + "grad_norm": 1.7734823531635113, + "language_loss": 0.70372528, + "learning_rate": 1.094726915390082e-06, + "loss": 0.72557217, + "num_input_tokens_seen": 118120650, + "step": 5489, + "time_per_iteration": 2.773024320602417 + }, + { + "auxiliary_loss_clip": 0.01162407, + "auxiliary_loss_mlp": 0.01027583, + "balance_loss_clip": 1.04743993, + "balance_loss_mlp": 1.01980448, + "epoch": 0.6601334696086094, + "flos": 22342649765760.0, + "grad_norm": 1.9506518713833967, + "language_loss": 0.69653761, + "learning_rate": 1.0940323788672836e-06, + "loss": 0.71843755, + "num_input_tokens_seen": 118139825, + "step": 5490, + "time_per_iteration": 2.6841113567352295 + }, + { + "auxiliary_loss_clip": 0.01157107, + "auxiliary_loss_mlp": 0.01025027, + "balance_loss_clip": 1.04509676, + "balance_loss_mlp": 1.01694477, + "epoch": 0.6602537124992485, + "flos": 25703098657920.0, + "grad_norm": 1.5744341978605667, + "language_loss": 0.74032068, + "learning_rate": 1.0933379797790522e-06, + "loss": 0.76214206, + "num_input_tokens_seen": 118159240, + "step": 5491, + "time_per_iteration": 3.5963456630706787 + }, + { + "auxiliary_loss_clip": 0.01173377, + "auxiliary_loss_mlp": 0.01026243, + "balance_loss_clip": 1.04963446, + "balance_loss_mlp": 1.01786244, + "epoch": 0.6603739553898875, + "flos": 25848572739840.0, + "grad_norm": 2.643363491851159, + "language_loss": 0.71548128, + "learning_rate": 1.0926437182307293e-06, + "loss": 0.73747748, + "num_input_tokens_seen": 118178050, + "step": 5492, + "time_per_iteration": 3.704558849334717 + }, + { + "auxiliary_loss_clip": 0.01155891, + "auxiliary_loss_mlp": 0.0102991, + "balance_loss_clip": 1.04356372, + "balance_loss_mlp": 1.02167892, + "epoch": 0.6604941982805267, + "flos": 24570296461440.0, + "grad_norm": 2.0421315214713713, + "language_loss": 0.78218913, + "learning_rate": 1.0919495943276338e-06, + "loss": 0.80404711, + "num_input_tokens_seen": 118199070, + "step": 5493, + "time_per_iteration": 2.847428798675537 + }, + { + "auxiliary_loss_clip": 0.01146293, + "auxiliary_loss_mlp": 0.0102768, + "balance_loss_clip": 1.04202914, + "balance_loss_mlp": 1.01916802, + "epoch": 0.6606144411711657, + "flos": 13261775581440.0, + "grad_norm": 6.120446086481192, + "language_loss": 0.76196527, + "learning_rate": 1.0912556081750611e-06, + "loss": 0.783705, + "num_input_tokens_seen": 118217000, + "step": 5494, + "time_per_iteration": 3.7719037532806396 + }, + { + "auxiliary_loss_clip": 0.01149017, + "auxiliary_loss_mlp": 0.01030438, + "balance_loss_clip": 1.04591119, + "balance_loss_mlp": 1.02244544, + "epoch": 0.6607346840618048, + "flos": 25155281358720.0, + "grad_norm": 1.787838962261575, + "language_loss": 0.76339042, + "learning_rate": 1.0905617598782909e-06, + "loss": 0.78518498, + "num_input_tokens_seen": 118237205, + "step": 5495, + "time_per_iteration": 2.768693685531616 + }, + { + "auxiliary_loss_clip": 0.01127913, + "auxiliary_loss_mlp": 0.01026083, + "balance_loss_clip": 1.0425384, + "balance_loss_mlp": 1.0182209, + "epoch": 0.660854926952444, + "flos": 17638029095040.0, + "grad_norm": 2.092825978092067, + "language_loss": 0.81503987, + "learning_rate": 1.0898680495425775e-06, + "loss": 0.8365798, + "num_input_tokens_seen": 118255495, + "step": 5496, + "time_per_iteration": 2.7909038066864014 + }, + { + "auxiliary_loss_clip": 0.01157286, + "auxiliary_loss_mlp": 0.01024996, + "balance_loss_clip": 1.04704618, + "balance_loss_mlp": 1.0175271, + "epoch": 0.660975169843083, + "flos": 16836000266880.0, + "grad_norm": 1.965126704719264, + "language_loss": 0.80497628, + "learning_rate": 1.0891744772731594e-06, + "loss": 0.82679915, + "num_input_tokens_seen": 118273310, + "step": 5497, + "time_per_iteration": 2.7610840797424316 + }, + { + "auxiliary_loss_clip": 0.01165794, + "auxiliary_loss_mlp": 0.01030477, + "balance_loss_clip": 1.04772496, + "balance_loss_mlp": 1.02291334, + "epoch": 0.6610954127337221, + "flos": 26870410846080.0, + "grad_norm": 1.6676375219070771, + "language_loss": 0.65711164, + "learning_rate": 1.088481043175248e-06, + "loss": 0.67907441, + "num_input_tokens_seen": 118293880, + "step": 5498, + "time_per_iteration": 3.785383701324463 + }, + { + "auxiliary_loss_clip": 0.01143271, + "auxiliary_loss_mlp": 0.01029348, + "balance_loss_clip": 1.04219794, + "balance_loss_mlp": 1.02103329, + "epoch": 0.6612156556243612, + "flos": 26465697331200.0, + "grad_norm": 1.8950810925696624, + "language_loss": 0.75588155, + "learning_rate": 1.0877877473540368e-06, + "loss": 0.7776078, + "num_input_tokens_seen": 118314465, + "step": 5499, + "time_per_iteration": 2.7840147018432617 + }, + { + "auxiliary_loss_clip": 0.01172906, + "auxiliary_loss_mlp": 0.01024523, + "balance_loss_clip": 1.04895973, + "balance_loss_mlp": 1.01673913, + "epoch": 0.6613358985150003, + "flos": 19791915212160.0, + "grad_norm": 1.772964306760707, + "language_loss": 0.724572, + "learning_rate": 1.0870945899147002e-06, + "loss": 0.74654627, + "num_input_tokens_seen": 118331110, + "step": 5500, + "time_per_iteration": 2.698580741882324 + }, + { + "auxiliary_loss_clip": 0.01159107, + "auxiliary_loss_mlp": 0.01026592, + "balance_loss_clip": 1.04684043, + "balance_loss_mlp": 1.01927543, + "epoch": 0.6614561414056394, + "flos": 26831627136000.0, + "grad_norm": 1.9204575185069759, + "language_loss": 0.76047385, + "learning_rate": 1.0864015709623879e-06, + "loss": 0.78233087, + "num_input_tokens_seen": 118351980, + "step": 5501, + "time_per_iteration": 2.790158987045288 + }, + { + "auxiliary_loss_clip": 0.01164983, + "auxiliary_loss_mlp": 0.01024119, + "balance_loss_clip": 1.0459162, + "balance_loss_mlp": 1.01623333, + "epoch": 0.6615763842962785, + "flos": 22894597128960.0, + "grad_norm": 2.5625858361162748, + "language_loss": 0.80200875, + "learning_rate": 1.0857086906022313e-06, + "loss": 0.82389975, + "num_input_tokens_seen": 118370315, + "step": 5502, + "time_per_iteration": 2.788722276687622 + }, + { + "auxiliary_loss_clip": 0.01113749, + "auxiliary_loss_mlp": 0.01028184, + "balance_loss_clip": 1.04285264, + "balance_loss_mlp": 1.0199523, + "epoch": 0.6616966271869176, + "flos": 24790321221120.0, + "grad_norm": 1.8933728471438203, + "language_loss": 0.73354268, + "learning_rate": 1.0850159489393388e-06, + "loss": 0.75496197, + "num_input_tokens_seen": 118389575, + "step": 5503, + "time_per_iteration": 3.048262119293213 + }, + { + "auxiliary_loss_clip": 0.01135621, + "auxiliary_loss_mlp": 0.01026645, + "balance_loss_clip": 1.03909945, + "balance_loss_mlp": 1.01858044, + "epoch": 0.6618168700775566, + "flos": 17202109639680.0, + "grad_norm": 2.2857771179431357, + "language_loss": 0.82145166, + "learning_rate": 1.0843233460787992e-06, + "loss": 0.84307438, + "num_input_tokens_seen": 118406790, + "step": 5504, + "time_per_iteration": 2.8317208290100098 + }, + { + "auxiliary_loss_clip": 0.01130305, + "auxiliary_loss_mlp": 0.01027581, + "balance_loss_clip": 1.04411459, + "balance_loss_mlp": 1.01950407, + "epoch": 0.6619371129681958, + "flos": 25447091448960.0, + "grad_norm": 1.9063730634226383, + "language_loss": 0.78026652, + "learning_rate": 1.0836308821256805e-06, + "loss": 0.80184543, + "num_input_tokens_seen": 118427590, + "step": 5505, + "time_per_iteration": 2.8717188835144043 + }, + { + "auxiliary_loss_clip": 0.01159416, + "auxiliary_loss_mlp": 0.01025111, + "balance_loss_clip": 1.0461055, + "balance_loss_mlp": 1.01741862, + "epoch": 0.6620573558588349, + "flos": 18040444139520.0, + "grad_norm": 2.0544541595253936, + "language_loss": 0.78253651, + "learning_rate": 1.0829385571850282e-06, + "loss": 0.80438185, + "num_input_tokens_seen": 118444570, + "step": 5506, + "time_per_iteration": 2.724126100540161 + }, + { + "auxiliary_loss_clip": 0.01177501, + "auxiliary_loss_mlp": 0.01026822, + "balance_loss_clip": 1.0498091, + "balance_loss_mlp": 1.01814985, + "epoch": 0.6621775987494739, + "flos": 17785586165760.0, + "grad_norm": 2.2211259556525547, + "language_loss": 0.83799112, + "learning_rate": 1.0822463713618679e-06, + "loss": 0.86003435, + "num_input_tokens_seen": 118461425, + "step": 5507, + "time_per_iteration": 2.6358959674835205 + }, + { + "auxiliary_loss_clip": 0.01144628, + "auxiliary_loss_mlp": 0.0102583, + "balance_loss_clip": 1.04307234, + "balance_loss_mlp": 1.01813769, + "epoch": 0.6622978416401131, + "flos": 17492590926720.0, + "grad_norm": 2.08145009269814, + "language_loss": 0.8487404, + "learning_rate": 1.0815543247612034e-06, + "loss": 0.87044495, + "num_input_tokens_seen": 118478495, + "step": 5508, + "time_per_iteration": 2.7519333362579346 + }, + { + "auxiliary_loss_clip": 0.01149876, + "auxiliary_loss_mlp": 0.01027305, + "balance_loss_clip": 1.0407939, + "balance_loss_mlp": 1.01915741, + "epoch": 0.6624180845307521, + "flos": 21648352803840.0, + "grad_norm": 1.7568533277031724, + "language_loss": 0.82846618, + "learning_rate": 1.0808624174880168e-06, + "loss": 0.85023797, + "num_input_tokens_seen": 118499145, + "step": 5509, + "time_per_iteration": 2.7089884281158447 + }, + { + "auxiliary_loss_clip": 0.01166827, + "auxiliary_loss_mlp": 0.01029864, + "balance_loss_clip": 1.04511273, + "balance_loss_mlp": 1.02189493, + "epoch": 0.6625383274213912, + "flos": 23805902108160.0, + "grad_norm": 1.7030240227105722, + "language_loss": 0.79852355, + "learning_rate": 1.080170649647272e-06, + "loss": 0.82049036, + "num_input_tokens_seen": 118518950, + "step": 5510, + "time_per_iteration": 2.6569976806640625 + }, + { + "auxiliary_loss_clip": 0.01170381, + "auxiliary_loss_mlp": 0.01028146, + "balance_loss_clip": 1.04768157, + "balance_loss_mlp": 1.01975965, + "epoch": 0.6626585703120303, + "flos": 33262941473280.0, + "grad_norm": 1.6620032759096919, + "language_loss": 0.67215633, + "learning_rate": 1.0794790213439068e-06, + "loss": 0.69414151, + "num_input_tokens_seen": 118545850, + "step": 5511, + "time_per_iteration": 2.84328556060791 + }, + { + "auxiliary_loss_clip": 0.01136161, + "auxiliary_loss_mlp": 0.01028625, + "balance_loss_clip": 1.04349184, + "balance_loss_mlp": 1.02025342, + "epoch": 0.6627788132026694, + "flos": 22085780630400.0, + "grad_norm": 1.869181692781431, + "language_loss": 0.78286731, + "learning_rate": 1.078787532682843e-06, + "loss": 0.80451512, + "num_input_tokens_seen": 118563325, + "step": 5512, + "time_per_iteration": 2.9309048652648926 + }, + { + "auxiliary_loss_clip": 0.01160209, + "auxiliary_loss_mlp": 0.0102861, + "balance_loss_clip": 1.0469563, + "balance_loss_mlp": 1.02076578, + "epoch": 0.6628990560933085, + "flos": 36173608260480.0, + "grad_norm": 2.217606734981389, + "language_loss": 0.76085043, + "learning_rate": 1.0780961837689773e-06, + "loss": 0.78273857, + "num_input_tokens_seen": 118582835, + "step": 5513, + "time_per_iteration": 2.859100341796875 + }, + { + "auxiliary_loss_clip": 0.01146668, + "auxiliary_loss_mlp": 0.01025714, + "balance_loss_clip": 1.04638624, + "balance_loss_mlp": 1.01784623, + "epoch": 0.6630192989839476, + "flos": 18513567106560.0, + "grad_norm": 1.4902463129715107, + "language_loss": 0.6990338, + "learning_rate": 1.0774049747071883e-06, + "loss": 0.7207576, + "num_input_tokens_seen": 118600715, + "step": 5514, + "time_per_iteration": 2.7977101802825928 + }, + { + "auxiliary_loss_clip": 0.01133003, + "auxiliary_loss_mlp": 0.0102378, + "balance_loss_clip": 1.04566383, + "balance_loss_mlp": 1.01635289, + "epoch": 0.6631395418745867, + "flos": 35809510049280.0, + "grad_norm": 2.4682798555845924, + "language_loss": 0.6804145, + "learning_rate": 1.076713905602332e-06, + "loss": 0.70198232, + "num_input_tokens_seen": 118621290, + "step": 5515, + "time_per_iteration": 3.076998710632324 + }, + { + "auxiliary_loss_clip": 0.0116659, + "auxiliary_loss_mlp": 0.0102507, + "balance_loss_clip": 1.04929161, + "balance_loss_mlp": 1.01675534, + "epoch": 0.6632597847652257, + "flos": 20047742853120.0, + "grad_norm": 1.7966916226620566, + "language_loss": 0.81274045, + "learning_rate": 1.07602297655924e-06, + "loss": 0.83465707, + "num_input_tokens_seen": 118639610, + "step": 5516, + "time_per_iteration": 2.9083614349365234 + }, + { + "auxiliary_loss_clip": 0.01174022, + "auxiliary_loss_mlp": 0.01027049, + "balance_loss_clip": 1.05133128, + "balance_loss_mlp": 1.01934791, + "epoch": 0.6633800276558649, + "flos": 21214480423680.0, + "grad_norm": 1.7175883659731275, + "language_loss": 0.81208187, + "learning_rate": 1.0753321876827292e-06, + "loss": 0.83409262, + "num_input_tokens_seen": 118658895, + "step": 5517, + "time_per_iteration": 3.635342597961426 + }, + { + "auxiliary_loss_clip": 0.01172807, + "auxiliary_loss_mlp": 0.01028022, + "balance_loss_clip": 1.04834771, + "balance_loss_mlp": 1.01991546, + "epoch": 0.663500270546504, + "flos": 23987753688960.0, + "grad_norm": 2.1183380072477043, + "language_loss": 0.73989224, + "learning_rate": 1.0746415390775893e-06, + "loss": 0.76190054, + "num_input_tokens_seen": 118677025, + "step": 5518, + "time_per_iteration": 3.7209672927856445 + }, + { + "auxiliary_loss_clip": 0.01172693, + "auxiliary_loss_mlp": 0.01035775, + "balance_loss_clip": 1.0502584, + "balance_loss_mlp": 1.02750158, + "epoch": 0.663620513437143, + "flos": 17932389050880.0, + "grad_norm": 1.895061189045562, + "language_loss": 0.76372665, + "learning_rate": 1.0739510308485939e-06, + "loss": 0.78581136, + "num_input_tokens_seen": 118694240, + "step": 5519, + "time_per_iteration": 2.718780279159546 + }, + { + "auxiliary_loss_clip": 0.01042957, + "auxiliary_loss_mlp": 0.01002037, + "balance_loss_clip": 1.01108682, + "balance_loss_mlp": 1.00045133, + "epoch": 0.6637407563277821, + "flos": 57840241086720.0, + "grad_norm": 0.8072296236879223, + "language_loss": 0.6252088, + "learning_rate": 1.07326066310049e-06, + "loss": 0.64565873, + "num_input_tokens_seen": 118758365, + "step": 5520, + "time_per_iteration": 4.4007954597473145 + }, + { + "auxiliary_loss_clip": 0.01142587, + "auxiliary_loss_mlp": 0.01032151, + "balance_loss_clip": 1.04393566, + "balance_loss_mlp": 1.02393198, + "epoch": 0.6638609992184212, + "flos": 27306007079040.0, + "grad_norm": 1.8235544918739126, + "language_loss": 0.79565769, + "learning_rate": 1.0725704359380059e-06, + "loss": 0.8174051, + "num_input_tokens_seen": 118778220, + "step": 5521, + "time_per_iteration": 2.824104070663452 + }, + { + "auxiliary_loss_clip": 0.01172513, + "auxiliary_loss_mlp": 0.01027786, + "balance_loss_clip": 1.04907727, + "balance_loss_mlp": 1.01983452, + "epoch": 0.6639812421090603, + "flos": 18624854419200.0, + "grad_norm": 2.066142151324978, + "language_loss": 0.71688378, + "learning_rate": 1.0718803494658497e-06, + "loss": 0.73888683, + "num_input_tokens_seen": 118797110, + "step": 5522, + "time_per_iteration": 2.6760752201080322 + }, + { + "auxiliary_loss_clip": 0.01088541, + "auxiliary_loss_mlp": 0.01027675, + "balance_loss_clip": 1.03677893, + "balance_loss_mlp": 1.0198164, + "epoch": 0.6641014849996993, + "flos": 15924479806080.0, + "grad_norm": 2.0844771070286248, + "language_loss": 0.8365851, + "learning_rate": 1.071190403788707e-06, + "loss": 0.85774726, + "num_input_tokens_seen": 118812415, + "step": 5523, + "time_per_iteration": 3.9031918048858643 + }, + { + "auxiliary_loss_clip": 0.0115127, + "auxiliary_loss_mlp": 0.01029115, + "balance_loss_clip": 1.04738438, + "balance_loss_mlp": 1.02040648, + "epoch": 0.6642217278903385, + "flos": 26505486622080.0, + "grad_norm": 2.3064082392011205, + "language_loss": 0.75617838, + "learning_rate": 1.0705005990112415e-06, + "loss": 0.77798218, + "num_input_tokens_seen": 118832195, + "step": 5524, + "time_per_iteration": 2.8326308727264404 + }, + { + "auxiliary_loss_clip": 0.01124225, + "auxiliary_loss_mlp": 0.01031905, + "balance_loss_clip": 1.0431819, + "balance_loss_mlp": 1.02390575, + "epoch": 0.6643419707809776, + "flos": 15377308951680.0, + "grad_norm": 2.47876852979586, + "language_loss": 0.74541914, + "learning_rate": 1.0698109352380957e-06, + "loss": 0.76698041, + "num_input_tokens_seen": 118849795, + "step": 5525, + "time_per_iteration": 2.874575614929199 + }, + { + "auxiliary_loss_clip": 0.01170735, + "auxiliary_loss_mlp": 0.01030148, + "balance_loss_clip": 1.04675221, + "balance_loss_mlp": 1.02217317, + "epoch": 0.6644622136716166, + "flos": 25117610970240.0, + "grad_norm": 1.8663687591608997, + "language_loss": 0.77936208, + "learning_rate": 1.0691214125738909e-06, + "loss": 0.80137092, + "num_input_tokens_seen": 118870000, + "step": 5526, + "time_per_iteration": 2.756622791290283 + }, + { + "auxiliary_loss_clip": 0.01063847, + "auxiliary_loss_mlp": 0.01002215, + "balance_loss_clip": 1.0106082, + "balance_loss_mlp": 1.00079596, + "epoch": 0.6645824565622558, + "flos": 66201717680640.0, + "grad_norm": 0.7898655145371458, + "language_loss": 0.57508183, + "learning_rate": 1.0684320311232287e-06, + "loss": 0.59574246, + "num_input_tokens_seen": 118932905, + "step": 5527, + "time_per_iteration": 3.3300561904907227 + }, + { + "auxiliary_loss_clip": 0.01150744, + "auxiliary_loss_mlp": 0.01028677, + "balance_loss_clip": 1.04555726, + "balance_loss_mlp": 1.02000999, + "epoch": 0.6647026994528948, + "flos": 25082131311360.0, + "grad_norm": 2.06966089231964, + "language_loss": 0.81664753, + "learning_rate": 1.0677427909906865e-06, + "loss": 0.83844179, + "num_input_tokens_seen": 118953355, + "step": 5528, + "time_per_iteration": 2.833625078201294 + }, + { + "auxiliary_loss_clip": 0.01173733, + "auxiliary_loss_mlp": 0.0102963, + "balance_loss_clip": 1.04757857, + "balance_loss_mlp": 1.02147603, + "epoch": 0.6648229423435339, + "flos": 18222187979520.0, + "grad_norm": 1.7848986962160918, + "language_loss": 0.72344756, + "learning_rate": 1.0670536922808216e-06, + "loss": 0.74548125, + "num_input_tokens_seen": 118973480, + "step": 5529, + "time_per_iteration": 2.6724631786346436 + }, + { + "auxiliary_loss_clip": 0.01153033, + "auxiliary_loss_mlp": 0.01029778, + "balance_loss_clip": 1.04589105, + "balance_loss_mlp": 1.0222795, + "epoch": 0.6649431852341731, + "flos": 18296882311680.0, + "grad_norm": 2.528376990668147, + "language_loss": 0.72226131, + "learning_rate": 1.06636473509817e-06, + "loss": 0.74408942, + "num_input_tokens_seen": 118989860, + "step": 5530, + "time_per_iteration": 2.786367654800415 + }, + { + "auxiliary_loss_clip": 0.01147491, + "auxiliary_loss_mlp": 0.0088573, + "balance_loss_clip": 1.04353428, + "balance_loss_mlp": 1.00066864, + "epoch": 0.6650634281248121, + "flos": 17019575700480.0, + "grad_norm": 2.028346209607598, + "language_loss": 0.80635852, + "learning_rate": 1.0656759195472447e-06, + "loss": 0.82669067, + "num_input_tokens_seen": 119007150, + "step": 5531, + "time_per_iteration": 2.7034695148468018 + }, + { + "auxiliary_loss_clip": 0.01050792, + "auxiliary_loss_mlp": 0.01004277, + "balance_loss_clip": 1.01032782, + "balance_loss_mlp": 1.00284672, + "epoch": 0.6651836710154512, + "flos": 69294810666240.0, + "grad_norm": 0.9017059189528099, + "language_loss": 0.5976398, + "learning_rate": 1.0649872457325414e-06, + "loss": 0.61819053, + "num_input_tokens_seen": 119068435, + "step": 5532, + "time_per_iteration": 3.2659945487976074 + }, + { + "auxiliary_loss_clip": 0.01060216, + "auxiliary_loss_mlp": 0.01005733, + "balance_loss_clip": 1.0111146, + "balance_loss_mlp": 1.00424314, + "epoch": 0.6653039139060903, + "flos": 66883444882560.0, + "grad_norm": 0.8461090455213026, + "language_loss": 0.55127829, + "learning_rate": 1.0642987137585278e-06, + "loss": 0.57193774, + "num_input_tokens_seen": 119127960, + "step": 5533, + "time_per_iteration": 3.1960055828094482 + }, + { + "auxiliary_loss_clip": 0.01149711, + "auxiliary_loss_mlp": 0.01028, + "balance_loss_clip": 1.04378211, + "balance_loss_mlp": 1.01989985, + "epoch": 0.6654241567967294, + "flos": 21470056669440.0, + "grad_norm": 1.5985366573574935, + "language_loss": 0.8235954, + "learning_rate": 1.0636103237296561e-06, + "loss": 0.8453725, + "num_input_tokens_seen": 119146885, + "step": 5534, + "time_per_iteration": 2.8079943656921387 + }, + { + "auxiliary_loss_clip": 0.01158621, + "auxiliary_loss_mlp": 0.01029584, + "balance_loss_clip": 1.04781461, + "balance_loss_mlp": 1.0223484, + "epoch": 0.6655443996873684, + "flos": 25119514391040.0, + "grad_norm": 2.284162214934499, + "language_loss": 0.84416008, + "learning_rate": 1.062922075750353e-06, + "loss": 0.86604214, + "num_input_tokens_seen": 119166900, + "step": 5535, + "time_per_iteration": 2.8158798217773438 + }, + { + "auxiliary_loss_clip": 0.01145815, + "auxiliary_loss_mlp": 0.01027494, + "balance_loss_clip": 1.04470086, + "balance_loss_mlp": 1.0193162, + "epoch": 0.6656646425780076, + "flos": 17457326749440.0, + "grad_norm": 2.5693868319679307, + "language_loss": 0.72241998, + "learning_rate": 1.0622339699250267e-06, + "loss": 0.74415302, + "num_input_tokens_seen": 119184820, + "step": 5536, + "time_per_iteration": 2.834340810775757 + }, + { + "auxiliary_loss_clip": 0.01144448, + "auxiliary_loss_mlp": 0.0102809, + "balance_loss_clip": 1.04295743, + "balance_loss_mlp": 1.0212419, + "epoch": 0.6657848854686467, + "flos": 23434190213760.0, + "grad_norm": 1.764941575608543, + "language_loss": 0.79323256, + "learning_rate": 1.0615460063580624e-06, + "loss": 0.81495792, + "num_input_tokens_seen": 119203295, + "step": 5537, + "time_per_iteration": 2.891911506652832 + }, + { + "auxiliary_loss_clip": 0.01156765, + "auxiliary_loss_mlp": 0.01027784, + "balance_loss_clip": 1.04537368, + "balance_loss_mlp": 1.02036619, + "epoch": 0.6659051283592857, + "flos": 11509909459200.0, + "grad_norm": 1.79446864555465, + "language_loss": 0.72823882, + "learning_rate": 1.060858185153821e-06, + "loss": 0.75008434, + "num_input_tokens_seen": 119221395, + "step": 5538, + "time_per_iteration": 2.727825403213501 + }, + { + "auxiliary_loss_clip": 0.0115908, + "auxiliary_loss_mlp": 0.0103387, + "balance_loss_clip": 1.04672384, + "balance_loss_mlp": 1.02501273, + "epoch": 0.6660253712499249, + "flos": 20594554571520.0, + "grad_norm": 2.1836727996713328, + "language_loss": 0.76229119, + "learning_rate": 1.0601705064166474e-06, + "loss": 0.7842207, + "num_input_tokens_seen": 119239790, + "step": 5539, + "time_per_iteration": 2.787451982498169 + }, + { + "auxiliary_loss_clip": 0.01149166, + "auxiliary_loss_mlp": 0.01024638, + "balance_loss_clip": 1.04642534, + "balance_loss_mlp": 1.0172832, + "epoch": 0.666145614140564, + "flos": 21251504367360.0, + "grad_norm": 1.910015210438189, + "language_loss": 0.73335934, + "learning_rate": 1.0594829702508596e-06, + "loss": 0.75509739, + "num_input_tokens_seen": 119257505, + "step": 5540, + "time_per_iteration": 2.7904586791992188 + }, + { + "auxiliary_loss_clip": 0.01146714, + "auxiliary_loss_mlp": 0.01033688, + "balance_loss_clip": 1.0439564, + "balance_loss_mlp": 1.0259397, + "epoch": 0.666265857031203, + "flos": 33726188200320.0, + "grad_norm": 1.6129757542619902, + "language_loss": 0.54795772, + "learning_rate": 1.0587955767607592e-06, + "loss": 0.56976169, + "num_input_tokens_seen": 119279365, + "step": 5541, + "time_per_iteration": 2.927387237548828 + }, + { + "auxiliary_loss_clip": 0.01172144, + "auxiliary_loss_mlp": 0.01033106, + "balance_loss_clip": 1.04888678, + "balance_loss_mlp": 1.02502322, + "epoch": 0.6663860999218422, + "flos": 17456644391040.0, + "grad_norm": 2.39413716056954, + "language_loss": 0.77286726, + "learning_rate": 1.0581083260506206e-06, + "loss": 0.79491973, + "num_input_tokens_seen": 119296150, + "step": 5542, + "time_per_iteration": 2.6654765605926514 + }, + { + "auxiliary_loss_clip": 0.01151872, + "auxiliary_loss_mlp": 0.01029887, + "balance_loss_clip": 1.04476714, + "balance_loss_mlp": 1.02219832, + "epoch": 0.6665063428124812, + "flos": 17676740977920.0, + "grad_norm": 2.418996457622319, + "language_loss": 0.76299453, + "learning_rate": 1.0574212182246993e-06, + "loss": 0.78481215, + "num_input_tokens_seen": 119314845, + "step": 5543, + "time_per_iteration": 3.6774232387542725 + }, + { + "auxiliary_loss_clip": 0.01157464, + "auxiliary_loss_mlp": 0.01031917, + "balance_loss_clip": 1.04495668, + "balance_loss_mlp": 1.02308941, + "epoch": 0.6666265857031203, + "flos": 27673265687040.0, + "grad_norm": 2.4528158540250184, + "language_loss": 0.75411832, + "learning_rate": 1.0567342533872303e-06, + "loss": 0.77601218, + "num_input_tokens_seen": 119334875, + "step": 5544, + "time_per_iteration": 3.757969856262207 + }, + { + "auxiliary_loss_clip": 0.01155941, + "auxiliary_loss_mlp": 0.01027735, + "balance_loss_clip": 1.04762793, + "balance_loss_mlp": 1.02001619, + "epoch": 0.6667468285937594, + "flos": 25046831220480.0, + "grad_norm": 1.7219769222677688, + "language_loss": 0.81009305, + "learning_rate": 1.0560474316424255e-06, + "loss": 0.8319298, + "num_input_tokens_seen": 119354635, + "step": 5545, + "time_per_iteration": 2.8852474689483643 + }, + { + "auxiliary_loss_clip": 0.01152158, + "auxiliary_loss_mlp": 0.01024889, + "balance_loss_clip": 1.04356623, + "balance_loss_mlp": 1.01662815, + "epoch": 0.6668670714843985, + "flos": 22780472641920.0, + "grad_norm": 2.6918224947689353, + "language_loss": 0.7340315, + "learning_rate": 1.0553607530944746e-06, + "loss": 0.75580192, + "num_input_tokens_seen": 119372690, + "step": 5546, + "time_per_iteration": 2.7934610843658447 + }, + { + "auxiliary_loss_clip": 0.01143484, + "auxiliary_loss_mlp": 0.01030439, + "balance_loss_clip": 1.04218578, + "balance_loss_mlp": 1.02220798, + "epoch": 0.6669873143750376, + "flos": 22163886754560.0, + "grad_norm": 1.979618399534855, + "language_loss": 0.89478278, + "learning_rate": 1.0546742178475463e-06, + "loss": 0.91652191, + "num_input_tokens_seen": 119391685, + "step": 5547, + "time_per_iteration": 3.698092460632324 + }, + { + "auxiliary_loss_clip": 0.01132324, + "auxiliary_loss_mlp": 0.01027319, + "balance_loss_clip": 1.04445302, + "balance_loss_mlp": 1.01984477, + "epoch": 0.6671075572656767, + "flos": 20514832335360.0, + "grad_norm": 1.9734769914583625, + "language_loss": 0.86881483, + "learning_rate": 1.0539878260057868e-06, + "loss": 0.8904112, + "num_input_tokens_seen": 119410725, + "step": 5548, + "time_per_iteration": 3.040161609649658 + }, + { + "auxiliary_loss_clip": 0.01167093, + "auxiliary_loss_mlp": 0.01029056, + "balance_loss_clip": 1.05120349, + "balance_loss_mlp": 1.02128339, + "epoch": 0.6672278001563158, + "flos": 17931203902080.0, + "grad_norm": 2.826334205004473, + "language_loss": 0.68900895, + "learning_rate": 1.0533015776733226e-06, + "loss": 0.71097046, + "num_input_tokens_seen": 119426875, + "step": 5549, + "time_per_iteration": 3.637707471847534 + }, + { + "auxiliary_loss_clip": 0.01151346, + "auxiliary_loss_mlp": 0.01024946, + "balance_loss_clip": 1.04801226, + "balance_loss_mlp": 1.01691747, + "epoch": 0.6673480430469548, + "flos": 22342146975360.0, + "grad_norm": 2.24037286652019, + "language_loss": 0.78636551, + "learning_rate": 1.0526154729542566e-06, + "loss": 0.80812842, + "num_input_tokens_seen": 119446935, + "step": 5550, + "time_per_iteration": 2.9655516147613525 + }, + { + "auxiliary_loss_clip": 0.01144472, + "auxiliary_loss_mlp": 0.0102695, + "balance_loss_clip": 1.04520524, + "balance_loss_mlp": 1.0187484, + "epoch": 0.6674682859375939, + "flos": 20703830722560.0, + "grad_norm": 2.737191453055811, + "language_loss": 0.80019975, + "learning_rate": 1.0519295119526699e-06, + "loss": 0.82191396, + "num_input_tokens_seen": 119463240, + "step": 5551, + "time_per_iteration": 2.838160991668701 + }, + { + "auxiliary_loss_clip": 0.01154047, + "auxiliary_loss_mlp": 0.01025196, + "balance_loss_clip": 1.04325724, + "balance_loss_mlp": 1.01767433, + "epoch": 0.667588528828233, + "flos": 26206673379840.0, + "grad_norm": 1.6086620844815636, + "language_loss": 0.83071607, + "learning_rate": 1.0512436947726227e-06, + "loss": 0.85250849, + "num_input_tokens_seen": 119484655, + "step": 5552, + "time_per_iteration": 2.876326084136963 + }, + { + "auxiliary_loss_clip": 0.0114598, + "auxiliary_loss_mlp": 0.01029417, + "balance_loss_clip": 1.04441154, + "balance_loss_mlp": 1.02106071, + "epoch": 0.6677087717188721, + "flos": 23071025756160.0, + "grad_norm": 2.3503764452657565, + "language_loss": 0.65416908, + "learning_rate": 1.0505580215181517e-06, + "loss": 0.67592299, + "num_input_tokens_seen": 119502895, + "step": 5553, + "time_per_iteration": 2.843109130859375 + }, + { + "auxiliary_loss_clip": 0.01041931, + "auxiliary_loss_mlp": 0.01004855, + "balance_loss_clip": 1.01434779, + "balance_loss_mlp": 1.00325775, + "epoch": 0.6678290146095112, + "flos": 70941315219840.0, + "grad_norm": 0.7817280610664248, + "language_loss": 0.56657326, + "learning_rate": 1.0498724922932753e-06, + "loss": 0.58704114, + "num_input_tokens_seen": 119561010, + "step": 5554, + "time_per_iteration": 3.2803375720977783 + }, + { + "auxiliary_loss_clip": 0.01178014, + "auxiliary_loss_mlp": 0.01026613, + "balance_loss_clip": 1.05110395, + "balance_loss_mlp": 1.01813745, + "epoch": 0.6679492575001503, + "flos": 18661088263680.0, + "grad_norm": 2.330261793314548, + "language_loss": 0.86669397, + "learning_rate": 1.0491871072019851e-06, + "loss": 0.8887403, + "num_input_tokens_seen": 119578900, + "step": 5555, + "time_per_iteration": 2.7372384071350098 + }, + { + "auxiliary_loss_clip": 0.01146815, + "auxiliary_loss_mlp": 0.010294, + "balance_loss_clip": 1.04179239, + "balance_loss_mlp": 1.02155018, + "epoch": 0.6680695003907894, + "flos": 29711985822720.0, + "grad_norm": 1.9050472348259975, + "language_loss": 0.63870198, + "learning_rate": 1.0485018663482555e-06, + "loss": 0.66046411, + "num_input_tokens_seen": 119598920, + "step": 5556, + "time_per_iteration": 2.800560712814331 + }, + { + "auxiliary_loss_clip": 0.01156956, + "auxiliary_loss_mlp": 0.01029395, + "balance_loss_clip": 1.04577827, + "balance_loss_mlp": 1.02111018, + "epoch": 0.6681897432814284, + "flos": 28218964083840.0, + "grad_norm": 2.436979904682932, + "language_loss": 0.70954871, + "learning_rate": 1.0478167698360354e-06, + "loss": 0.73141217, + "num_input_tokens_seen": 119618220, + "step": 5557, + "time_per_iteration": 2.692403554916382 + }, + { + "auxiliary_loss_clip": 0.01155237, + "auxiliary_loss_mlp": 0.0102593, + "balance_loss_clip": 1.04471219, + "balance_loss_mlp": 1.01854479, + "epoch": 0.6683099861720676, + "flos": 25046543911680.0, + "grad_norm": 1.8588757720726656, + "language_loss": 0.69855273, + "learning_rate": 1.0471318177692556e-06, + "loss": 0.72036445, + "num_input_tokens_seen": 119638520, + "step": 5558, + "time_per_iteration": 2.6683335304260254 + }, + { + "auxiliary_loss_clip": 0.01139048, + "auxiliary_loss_mlp": 0.0102857, + "balance_loss_clip": 1.04373693, + "balance_loss_mlp": 1.02113676, + "epoch": 0.6684302290627067, + "flos": 22996977868800.0, + "grad_norm": 2.1989996929060873, + "language_loss": 0.76398832, + "learning_rate": 1.046447010251821e-06, + "loss": 0.78566444, + "num_input_tokens_seen": 119655850, + "step": 5559, + "time_per_iteration": 2.8537275791168213 + }, + { + "auxiliary_loss_clip": 0.01151188, + "auxiliary_loss_mlp": 0.01028628, + "balance_loss_clip": 1.04641557, + "balance_loss_mlp": 1.02077246, + "epoch": 0.6685504719533457, + "flos": 26573824247040.0, + "grad_norm": 1.8006968368710663, + "language_loss": 0.75780231, + "learning_rate": 1.0457623473876157e-06, + "loss": 0.7796005, + "num_input_tokens_seen": 119675355, + "step": 5560, + "time_per_iteration": 2.7690048217773438 + }, + { + "auxiliary_loss_clip": 0.01170004, + "auxiliary_loss_mlp": 0.01029832, + "balance_loss_clip": 1.04788971, + "balance_loss_mlp": 1.02145147, + "epoch": 0.6686707148439849, + "flos": 28986087870720.0, + "grad_norm": 2.246185111411815, + "language_loss": 0.71421248, + "learning_rate": 1.0450778292805046e-06, + "loss": 0.73621082, + "num_input_tokens_seen": 119695340, + "step": 5561, + "time_per_iteration": 2.7457072734832764 + }, + { + "auxiliary_loss_clip": 0.01163833, + "auxiliary_loss_mlp": 0.01028476, + "balance_loss_clip": 1.04428124, + "balance_loss_mlp": 1.02022123, + "epoch": 0.6687909577346239, + "flos": 23623152687360.0, + "grad_norm": 1.6719378431262515, + "language_loss": 0.78616929, + "learning_rate": 1.0443934560343267e-06, + "loss": 0.80809242, + "num_input_tokens_seen": 119716750, + "step": 5562, + "time_per_iteration": 2.7538955211639404 + }, + { + "auxiliary_loss_clip": 0.01130253, + "auxiliary_loss_mlp": 0.01026746, + "balance_loss_clip": 1.04071987, + "balance_loss_mlp": 1.01895618, + "epoch": 0.668911200625263, + "flos": 23148593176320.0, + "grad_norm": 2.13603628114815, + "language_loss": 0.78438824, + "learning_rate": 1.0437092277529034e-06, + "loss": 0.80595827, + "num_input_tokens_seen": 119736005, + "step": 5563, + "time_per_iteration": 2.8152458667755127 + }, + { + "auxiliary_loss_clip": 0.01148054, + "auxiliary_loss_mlp": 0.01024713, + "balance_loss_clip": 1.04366887, + "balance_loss_mlp": 1.01717877, + "epoch": 0.6690314435159022, + "flos": 18551919853440.0, + "grad_norm": 2.01483448169392, + "language_loss": 0.73847175, + "learning_rate": 1.0430251445400292e-06, + "loss": 0.76019943, + "num_input_tokens_seen": 119754050, + "step": 5564, + "time_per_iteration": 2.84590744972229 + }, + { + "auxiliary_loss_clip": 0.01105936, + "auxiliary_loss_mlp": 0.01031136, + "balance_loss_clip": 1.04074895, + "balance_loss_mlp": 1.02356029, + "epoch": 0.6691516864065412, + "flos": 31759540704000.0, + "grad_norm": 2.979134563160004, + "language_loss": 0.62641197, + "learning_rate": 1.0423412064994787e-06, + "loss": 0.64778268, + "num_input_tokens_seen": 119774820, + "step": 5565, + "time_per_iteration": 3.1053237915039062 + }, + { + "auxiliary_loss_clip": 0.01142058, + "auxiliary_loss_mlp": 0.01023376, + "balance_loss_clip": 1.04077947, + "balance_loss_mlp": 1.01569903, + "epoch": 0.6692719292971803, + "flos": 34933864296960.0, + "grad_norm": 1.8266670498359343, + "language_loss": 0.73947215, + "learning_rate": 1.0416574137350064e-06, + "loss": 0.76112652, + "num_input_tokens_seen": 119795525, + "step": 5566, + "time_per_iteration": 2.9602348804473877 + }, + { + "auxiliary_loss_clip": 0.01156452, + "auxiliary_loss_mlp": 0.01027694, + "balance_loss_clip": 1.04533482, + "balance_loss_mlp": 1.01967144, + "epoch": 0.6693921721878194, + "flos": 20449188230400.0, + "grad_norm": 2.3978243886225, + "language_loss": 0.80690229, + "learning_rate": 1.0409737663503428e-06, + "loss": 0.8287437, + "num_input_tokens_seen": 119813905, + "step": 5567, + "time_per_iteration": 2.7181010246276855 + }, + { + "auxiliary_loss_clip": 0.01157479, + "auxiliary_loss_mlp": 0.01026474, + "balance_loss_clip": 1.0423317, + "balance_loss_mlp": 1.01891577, + "epoch": 0.6695124150784585, + "flos": 16614538963200.0, + "grad_norm": 1.763936567982725, + "language_loss": 0.82866383, + "learning_rate": 1.040290264449196e-06, + "loss": 0.85050333, + "num_input_tokens_seen": 119832010, + "step": 5568, + "time_per_iteration": 2.7180099487304688 + }, + { + "auxiliary_loss_clip": 0.01158933, + "auxiliary_loss_mlp": 0.01026874, + "balance_loss_clip": 1.04631627, + "balance_loss_mlp": 1.0192802, + "epoch": 0.6696326579690975, + "flos": 26652145852800.0, + "grad_norm": 2.2934800143619105, + "language_loss": 0.6425308, + "learning_rate": 1.0396069081352532e-06, + "loss": 0.6643889, + "num_input_tokens_seen": 119851165, + "step": 5569, + "time_per_iteration": 3.657156229019165 + }, + { + "auxiliary_loss_clip": 0.0106322, + "auxiliary_loss_mlp": 0.0100073, + "balance_loss_clip": 1.0103575, + "balance_loss_mlp": 0.99928778, + "epoch": 0.6697529008597367, + "flos": 66964603662720.0, + "grad_norm": 0.7757231533191071, + "language_loss": 0.56030846, + "learning_rate": 1.0389236975121782e-06, + "loss": 0.58094794, + "num_input_tokens_seen": 119906015, + "step": 5570, + "time_per_iteration": 4.096133708953857 + }, + { + "auxiliary_loss_clip": 0.01173934, + "auxiliary_loss_mlp": 0.01028597, + "balance_loss_clip": 1.04782617, + "balance_loss_mlp": 1.02031839, + "epoch": 0.6698731437503758, + "flos": 20886939279360.0, + "grad_norm": 1.8567219931990833, + "language_loss": 0.71327263, + "learning_rate": 1.0382406326836147e-06, + "loss": 0.73529798, + "num_input_tokens_seen": 119925160, + "step": 5571, + "time_per_iteration": 2.637160062789917 + }, + { + "auxiliary_loss_clip": 0.01169048, + "auxiliary_loss_mlp": 0.01030308, + "balance_loss_clip": 1.04915333, + "balance_loss_mlp": 1.02212977, + "epoch": 0.6699933866410148, + "flos": 20409470766720.0, + "grad_norm": 2.1854747305821904, + "language_loss": 0.76207125, + "learning_rate": 1.0375577137531828e-06, + "loss": 0.78406477, + "num_input_tokens_seen": 119943720, + "step": 5572, + "time_per_iteration": 2.794271230697632 + }, + { + "auxiliary_loss_clip": 0.01156128, + "auxiliary_loss_mlp": 0.0103103, + "balance_loss_clip": 1.04659534, + "balance_loss_mlp": 1.02213693, + "epoch": 0.670113629531654, + "flos": 29023075900800.0, + "grad_norm": 1.5903844212976364, + "language_loss": 0.72135758, + "learning_rate": 1.0368749408244802e-06, + "loss": 0.74322909, + "num_input_tokens_seen": 119966640, + "step": 5573, + "time_per_iteration": 3.7175793647766113 + }, + { + "auxiliary_loss_clip": 0.01158029, + "auxiliary_loss_mlp": 0.01030019, + "balance_loss_clip": 1.04665637, + "balance_loss_mlp": 1.02183557, + "epoch": 0.670233872422293, + "flos": 19791699730560.0, + "grad_norm": 2.30423894451003, + "language_loss": 0.78646421, + "learning_rate": 1.0361923140010836e-06, + "loss": 0.80834472, + "num_input_tokens_seen": 119985125, + "step": 5574, + "time_per_iteration": 2.746506690979004 + }, + { + "auxiliary_loss_clip": 0.01168419, + "auxiliary_loss_mlp": 0.01025068, + "balance_loss_clip": 1.04789257, + "balance_loss_mlp": 1.01646698, + "epoch": 0.6703541153129321, + "flos": 24243689070720.0, + "grad_norm": 2.5353161201608856, + "language_loss": 0.63460875, + "learning_rate": 1.0355098333865455e-06, + "loss": 0.65654361, + "num_input_tokens_seen": 120004355, + "step": 5575, + "time_per_iteration": 3.7902791500091553 + }, + { + "auxiliary_loss_clip": 0.01155194, + "auxiliary_loss_mlp": 0.01022716, + "balance_loss_clip": 1.04753518, + "balance_loss_mlp": 1.0152775, + "epoch": 0.6704743582035713, + "flos": 26688523351680.0, + "grad_norm": 1.5735757421309533, + "language_loss": 0.69172275, + "learning_rate": 1.0348274990844006e-06, + "loss": 0.71350181, + "num_input_tokens_seen": 120027115, + "step": 5576, + "time_per_iteration": 2.718376636505127 + }, + { + "auxiliary_loss_clip": 0.01158116, + "auxiliary_loss_mlp": 0.01028206, + "balance_loss_clip": 1.04453254, + "balance_loss_mlp": 1.02025461, + "epoch": 0.6705946010942103, + "flos": 23514379326720.0, + "grad_norm": 1.7941016788020778, + "language_loss": 0.72857326, + "learning_rate": 1.034145311198155e-06, + "loss": 0.75043648, + "num_input_tokens_seen": 120047130, + "step": 5577, + "time_per_iteration": 2.7463300228118896 + }, + { + "auxiliary_loss_clip": 0.01169103, + "auxiliary_loss_mlp": 0.01025288, + "balance_loss_clip": 1.04780614, + "balance_loss_mlp": 1.01794195, + "epoch": 0.6707148439848494, + "flos": 24061011477120.0, + "grad_norm": 1.778378489162466, + "language_loss": 0.63893545, + "learning_rate": 1.0334632698312989e-06, + "loss": 0.66087937, + "num_input_tokens_seen": 120067925, + "step": 5578, + "time_per_iteration": 2.675557851791382 + }, + { + "auxiliary_loss_clip": 0.01145761, + "auxiliary_loss_mlp": 0.01030295, + "balance_loss_clip": 1.04466367, + "balance_loss_mlp": 1.02226043, + "epoch": 0.6708350868754885, + "flos": 22528667324160.0, + "grad_norm": 2.7562919497613123, + "language_loss": 0.75669301, + "learning_rate": 1.032781375087295e-06, + "loss": 0.77845359, + "num_input_tokens_seen": 120087825, + "step": 5579, + "time_per_iteration": 2.767033100128174 + }, + { + "auxiliary_loss_clip": 0.01155372, + "auxiliary_loss_mlp": 0.01030988, + "balance_loss_clip": 1.04819393, + "balance_loss_mlp": 1.02290273, + "epoch": 0.6709553297661276, + "flos": 25227749047680.0, + "grad_norm": 2.0341792582969376, + "language_loss": 0.6751585, + "learning_rate": 1.0320996270695891e-06, + "loss": 0.69702208, + "num_input_tokens_seen": 120108895, + "step": 5580, + "time_per_iteration": 2.7344167232513428 + }, + { + "auxiliary_loss_clip": 0.01143072, + "auxiliary_loss_mlp": 0.01031342, + "balance_loss_clip": 1.04230857, + "balance_loss_mlp": 1.02305079, + "epoch": 0.6710755726567667, + "flos": 20448757267200.0, + "grad_norm": 1.808905573465168, + "language_loss": 0.73233491, + "learning_rate": 1.0314180258815998e-06, + "loss": 0.75407904, + "num_input_tokens_seen": 120127535, + "step": 5581, + "time_per_iteration": 2.789503574371338 + }, + { + "auxiliary_loss_clip": 0.01134599, + "auxiliary_loss_mlp": 0.01026028, + "balance_loss_clip": 1.04168129, + "balance_loss_mlp": 1.01884913, + "epoch": 0.6711958155474057, + "flos": 25995411538560.0, + "grad_norm": 1.671473697631473, + "language_loss": 0.74115247, + "learning_rate": 1.0307365716267247e-06, + "loss": 0.76275873, + "num_input_tokens_seen": 120147980, + "step": 5582, + "time_per_iteration": 2.8601009845733643 + }, + { + "auxiliary_loss_clip": 0.01160676, + "auxiliary_loss_mlp": 0.01028853, + "balance_loss_clip": 1.04658973, + "balance_loss_mlp": 1.02096689, + "epoch": 0.6713160584380449, + "flos": 19937712516480.0, + "grad_norm": 1.8917343789246897, + "language_loss": 0.78106844, + "learning_rate": 1.0300552644083423e-06, + "loss": 0.80296373, + "num_input_tokens_seen": 120166905, + "step": 5583, + "time_per_iteration": 2.7336556911468506 + }, + { + "auxiliary_loss_clip": 0.01147895, + "auxiliary_loss_mlp": 0.01028082, + "balance_loss_clip": 1.04685724, + "balance_loss_mlp": 1.01969004, + "epoch": 0.6714363013286839, + "flos": 18223373128320.0, + "grad_norm": 2.482485307706155, + "language_loss": 0.7224313, + "learning_rate": 1.0293741043298036e-06, + "loss": 0.74419111, + "num_input_tokens_seen": 120185255, + "step": 5584, + "time_per_iteration": 2.810934543609619 + }, + { + "auxiliary_loss_clip": 0.01145062, + "auxiliary_loss_mlp": 0.01029953, + "balance_loss_clip": 1.04652286, + "balance_loss_mlp": 1.02184057, + "epoch": 0.671556544219323, + "flos": 25812374808960.0, + "grad_norm": 2.535654611220838, + "language_loss": 0.71428359, + "learning_rate": 1.0286930914944436e-06, + "loss": 0.73603368, + "num_input_tokens_seen": 120205070, + "step": 5585, + "time_per_iteration": 2.772637128829956 + }, + { + "auxiliary_loss_clip": 0.01171295, + "auxiliary_loss_mlp": 0.01030901, + "balance_loss_clip": 1.04622316, + "balance_loss_mlp": 1.02304471, + "epoch": 0.6716767871099621, + "flos": 15850431918720.0, + "grad_norm": 2.255996725730528, + "language_loss": 0.77234757, + "learning_rate": 1.0280122260055684e-06, + "loss": 0.79436958, + "num_input_tokens_seen": 120220780, + "step": 5586, + "time_per_iteration": 2.7254528999328613 + }, + { + "auxiliary_loss_clip": 0.01172848, + "auxiliary_loss_mlp": 0.01027735, + "balance_loss_clip": 1.04909909, + "balance_loss_mlp": 1.01945591, + "epoch": 0.6717970300006012, + "flos": 19756112330880.0, + "grad_norm": 1.9978311786171008, + "language_loss": 0.8213734, + "learning_rate": 1.0273315079664652e-06, + "loss": 0.8433792, + "num_input_tokens_seen": 120238735, + "step": 5587, + "time_per_iteration": 2.658466339111328 + }, + { + "auxiliary_loss_clip": 0.01166137, + "auxiliary_loss_mlp": 0.01025916, + "balance_loss_clip": 1.04869831, + "balance_loss_mlp": 1.01818228, + "epoch": 0.6719172728912403, + "flos": 25485049146240.0, + "grad_norm": 2.5952387590774237, + "language_loss": 0.7414785, + "learning_rate": 1.0266509374803992e-06, + "loss": 0.763399, + "num_input_tokens_seen": 120259895, + "step": 5588, + "time_per_iteration": 2.77945613861084 + }, + { + "auxiliary_loss_clip": 0.01174659, + "auxiliary_loss_mlp": 0.00886334, + "balance_loss_clip": 1.05013478, + "balance_loss_mlp": 1.00067496, + "epoch": 0.6720375157818794, + "flos": 15880344969600.0, + "grad_norm": 2.4679476173774693, + "language_loss": 0.84276497, + "learning_rate": 1.0259705146506123e-06, + "loss": 0.86337495, + "num_input_tokens_seen": 120274790, + "step": 5589, + "time_per_iteration": 2.6327197551727295 + }, + { + "auxiliary_loss_clip": 0.01167008, + "auxiliary_loss_mlp": 0.01027754, + "balance_loss_clip": 1.04914784, + "balance_loss_mlp": 1.01895618, + "epoch": 0.6721577586725185, + "flos": 32010843231360.0, + "grad_norm": 1.8693290038049437, + "language_loss": 0.77123225, + "learning_rate": 1.025290239580324e-06, + "loss": 0.79317987, + "num_input_tokens_seen": 120295460, + "step": 5590, + "time_per_iteration": 2.7783725261688232 + }, + { + "auxiliary_loss_clip": 0.01135539, + "auxiliary_loss_mlp": 0.01031935, + "balance_loss_clip": 1.04138172, + "balance_loss_mlp": 1.02381074, + "epoch": 0.6722780015631575, + "flos": 20737873837440.0, + "grad_norm": 1.6119628672793958, + "language_loss": 0.75420988, + "learning_rate": 1.0246101123727313e-06, + "loss": 0.77588463, + "num_input_tokens_seen": 120314440, + "step": 5591, + "time_per_iteration": 2.85444974899292 + }, + { + "auxiliary_loss_clip": 0.0116176, + "auxiliary_loss_mlp": 0.01025907, + "balance_loss_clip": 1.04567027, + "balance_loss_mlp": 1.01861167, + "epoch": 0.6723982444537967, + "flos": 16909617191040.0, + "grad_norm": 1.8863228795217133, + "language_loss": 0.78698742, + "learning_rate": 1.0239301331310085e-06, + "loss": 0.80886406, + "num_input_tokens_seen": 120332060, + "step": 5592, + "time_per_iteration": 2.6303224563598633 + }, + { + "auxiliary_loss_clip": 0.01159842, + "auxiliary_loss_mlp": 0.0103348, + "balance_loss_clip": 1.04631352, + "balance_loss_mlp": 1.02609515, + "epoch": 0.6725184873444358, + "flos": 20667812359680.0, + "grad_norm": 1.6747430110390507, + "language_loss": 0.88380593, + "learning_rate": 1.0232503019583088e-06, + "loss": 0.90573913, + "num_input_tokens_seen": 120351670, + "step": 5593, + "time_per_iteration": 2.6429615020751953 + }, + { + "auxiliary_loss_clip": 0.01159599, + "auxiliary_loss_mlp": 0.01027282, + "balance_loss_clip": 1.04856324, + "balance_loss_mlp": 1.01945257, + "epoch": 0.6726387302350748, + "flos": 23727616416000.0, + "grad_norm": 2.015920150921492, + "language_loss": 0.69803011, + "learning_rate": 1.0225706189577619e-06, + "loss": 0.71989888, + "num_input_tokens_seen": 120370195, + "step": 5594, + "time_per_iteration": 2.7298905849456787 + }, + { + "auxiliary_loss_clip": 0.0116408, + "auxiliary_loss_mlp": 0.01028058, + "balance_loss_clip": 1.04807365, + "balance_loss_mlp": 1.02007103, + "epoch": 0.672758973125714, + "flos": 15188274650880.0, + "grad_norm": 3.9850958232498495, + "language_loss": 0.74732429, + "learning_rate": 1.021891084232475e-06, + "loss": 0.76924568, + "num_input_tokens_seen": 120388130, + "step": 5595, + "time_per_iteration": 3.649395227432251 + }, + { + "auxiliary_loss_clip": 0.01161981, + "auxiliary_loss_mlp": 0.01031426, + "balance_loss_clip": 1.04533339, + "balance_loss_mlp": 1.02312887, + "epoch": 0.672879216016353, + "flos": 18077252601600.0, + "grad_norm": 2.951309554788455, + "language_loss": 0.80154443, + "learning_rate": 1.0212116978855325e-06, + "loss": 0.82347852, + "num_input_tokens_seen": 120406145, + "step": 5596, + "time_per_iteration": 3.689671039581299 + }, + { + "auxiliary_loss_clip": 0.01140849, + "auxiliary_loss_mlp": 0.01026625, + "balance_loss_clip": 1.04355645, + "balance_loss_mlp": 1.01893866, + "epoch": 0.6729994589069921, + "flos": 23476349802240.0, + "grad_norm": 1.7695297258186493, + "language_loss": 0.78767455, + "learning_rate": 1.020532460019997e-06, + "loss": 0.8093493, + "num_input_tokens_seen": 120425395, + "step": 5597, + "time_per_iteration": 2.794301986694336 + }, + { + "auxiliary_loss_clip": 0.01120472, + "auxiliary_loss_mlp": 0.01028324, + "balance_loss_clip": 1.04098237, + "balance_loss_mlp": 1.01988411, + "epoch": 0.6731197017976313, + "flos": 26322018929280.0, + "grad_norm": 2.3673259603529337, + "language_loss": 0.7083261, + "learning_rate": 1.0198533707389096e-06, + "loss": 0.72981411, + "num_input_tokens_seen": 120446270, + "step": 5598, + "time_per_iteration": 3.0260300636291504 + }, + { + "auxiliary_loss_clip": 0.01159286, + "auxiliary_loss_mlp": 0.00886413, + "balance_loss_clip": 1.04703617, + "balance_loss_mlp": 1.00070512, + "epoch": 0.6732399446882703, + "flos": 21616428591360.0, + "grad_norm": 1.7397024233719978, + "language_loss": 0.7309165, + "learning_rate": 1.0191744301452853e-06, + "loss": 0.75137347, + "num_input_tokens_seen": 120465570, + "step": 5599, + "time_per_iteration": 3.657668352127075 + }, + { + "auxiliary_loss_clip": 0.01168833, + "auxiliary_loss_mlp": 0.01030032, + "balance_loss_clip": 1.04563034, + "balance_loss_mlp": 1.02249777, + "epoch": 0.6733601875789094, + "flos": 25880173729920.0, + "grad_norm": 1.7277656377108357, + "language_loss": 0.70498508, + "learning_rate": 1.0184956383421208e-06, + "loss": 0.72697371, + "num_input_tokens_seen": 120484220, + "step": 5600, + "time_per_iteration": 3.6002213954925537 + }, + { + "auxiliary_loss_clip": 0.01165988, + "auxiliary_loss_mlp": 0.01028469, + "balance_loss_clip": 1.04787874, + "balance_loss_mlp": 1.02072644, + "epoch": 0.6734804304695485, + "flos": 22929573997440.0, + "grad_norm": 2.2014124045134884, + "language_loss": 0.65532917, + "learning_rate": 1.017816995432387e-06, + "loss": 0.67727369, + "num_input_tokens_seen": 120503320, + "step": 5601, + "time_per_iteration": 2.8129050731658936 + }, + { + "auxiliary_loss_clip": 0.011535, + "auxiliary_loss_mlp": 0.01027741, + "balance_loss_clip": 1.04750562, + "balance_loss_mlp": 1.01996899, + "epoch": 0.6736006733601876, + "flos": 18697968552960.0, + "grad_norm": 2.0284694823809954, + "language_loss": 0.74233818, + "learning_rate": 1.0171385015190353e-06, + "loss": 0.76415062, + "num_input_tokens_seen": 120523180, + "step": 5602, + "time_per_iteration": 2.77778959274292 + }, + { + "auxiliary_loss_clip": 0.01146116, + "auxiliary_loss_mlp": 0.00885536, + "balance_loss_clip": 1.04643607, + "balance_loss_mlp": 1.00065231, + "epoch": 0.6737209162508266, + "flos": 19427745173760.0, + "grad_norm": 2.065770212591738, + "language_loss": 0.73150337, + "learning_rate": 1.0164601567049908e-06, + "loss": 0.75181985, + "num_input_tokens_seen": 120541710, + "step": 5603, + "time_per_iteration": 2.85379958152771 + }, + { + "auxiliary_loss_clip": 0.01152952, + "auxiliary_loss_mlp": 0.01028804, + "balance_loss_clip": 1.04669547, + "balance_loss_mlp": 1.02124035, + "epoch": 0.6738411591414658, + "flos": 20158060498560.0, + "grad_norm": 1.8549389920461985, + "language_loss": 0.80369866, + "learning_rate": 1.015781961093158e-06, + "loss": 0.82551622, + "num_input_tokens_seen": 120561030, + "step": 5604, + "time_per_iteration": 2.792268753051758 + }, + { + "auxiliary_loss_clip": 0.01155718, + "auxiliary_loss_mlp": 0.01025191, + "balance_loss_clip": 1.04309142, + "balance_loss_mlp": 1.01719189, + "epoch": 0.6739614020321049, + "flos": 21653847584640.0, + "grad_norm": 1.5281226599814528, + "language_loss": 0.76889193, + "learning_rate": 1.0151039147864197e-06, + "loss": 0.79070103, + "num_input_tokens_seen": 120581005, + "step": 5605, + "time_per_iteration": 2.958329439163208 + }, + { + "auxiliary_loss_clip": 0.01113608, + "auxiliary_loss_mlp": 0.01029272, + "balance_loss_clip": 1.04651356, + "balance_loss_mlp": 1.0213443, + "epoch": 0.6740816449227439, + "flos": 19171702051200.0, + "grad_norm": 2.220116720737841, + "language_loss": 0.66145277, + "learning_rate": 1.0144260178876336e-06, + "loss": 0.68288159, + "num_input_tokens_seen": 120600350, + "step": 5606, + "time_per_iteration": 3.1158902645111084 + }, + { + "auxiliary_loss_clip": 0.01160424, + "auxiliary_loss_mlp": 0.01025458, + "balance_loss_clip": 1.04643524, + "balance_loss_mlp": 1.01747143, + "epoch": 0.6742018878133831, + "flos": 21097015971840.0, + "grad_norm": 2.556964035279693, + "language_loss": 0.67829543, + "learning_rate": 1.0137482704996388e-06, + "loss": 0.7001543, + "num_input_tokens_seen": 120614700, + "step": 5607, + "time_per_iteration": 2.9809975624084473 + }, + { + "auxiliary_loss_clip": 0.01149833, + "auxiliary_loss_mlp": 0.0102967, + "balance_loss_clip": 1.0471046, + "balance_loss_mlp": 1.02195764, + "epoch": 0.6743221307040221, + "flos": 23549966726400.0, + "grad_norm": 1.9590282261237528, + "language_loss": 0.78980279, + "learning_rate": 1.0130706727252461e-06, + "loss": 0.81159782, + "num_input_tokens_seen": 120631755, + "step": 5608, + "time_per_iteration": 2.783320188522339 + }, + { + "auxiliary_loss_clip": 0.01150106, + "auxiliary_loss_mlp": 0.01029158, + "balance_loss_clip": 1.04762554, + "balance_loss_mlp": 1.02121544, + "epoch": 0.6744423735946612, + "flos": 16249542912000.0, + "grad_norm": 2.613864841075268, + "language_loss": 0.67585057, + "learning_rate": 1.0123932246672468e-06, + "loss": 0.69764322, + "num_input_tokens_seen": 120645900, + "step": 5609, + "time_per_iteration": 2.8508856296539307 + }, + { + "auxiliary_loss_clip": 0.01039239, + "auxiliary_loss_mlp": 0.00877309, + "balance_loss_clip": 1.00995493, + "balance_loss_mlp": 1.00133228, + "epoch": 0.6745626164853004, + "flos": 57843257829120.0, + "grad_norm": 0.7462788585916417, + "language_loss": 0.55842471, + "learning_rate": 1.0117159264284114e-06, + "loss": 0.57759011, + "num_input_tokens_seen": 120709070, + "step": 5610, + "time_per_iteration": 3.339794158935547 + }, + { + "auxiliary_loss_clip": 0.0115273, + "auxiliary_loss_mlp": 0.01034214, + "balance_loss_clip": 1.04412913, + "balance_loss_mlp": 1.02590537, + "epoch": 0.6746828593759394, + "flos": 20485027025280.0, + "grad_norm": 1.7184750090123795, + "language_loss": 0.77321279, + "learning_rate": 1.0110387781114837e-06, + "loss": 0.79508221, + "num_input_tokens_seen": 120727685, + "step": 5611, + "time_per_iteration": 2.7839856147766113 + }, + { + "auxiliary_loss_clip": 0.01170908, + "auxiliary_loss_mlp": 0.01026756, + "balance_loss_clip": 1.04811358, + "balance_loss_mlp": 1.01887012, + "epoch": 0.6748031022665785, + "flos": 19208223204480.0, + "grad_norm": 2.105498058470423, + "language_loss": 0.77310944, + "learning_rate": 1.0103617798191872e-06, + "loss": 0.79508603, + "num_input_tokens_seen": 120747160, + "step": 5612, + "time_per_iteration": 2.6205902099609375 + }, + { + "auxiliary_loss_clip": 0.01148307, + "auxiliary_loss_mlp": 0.01027709, + "balance_loss_clip": 1.04530931, + "balance_loss_mlp": 1.01980507, + "epoch": 0.6749233451572175, + "flos": 15195026407680.0, + "grad_norm": 2.6963531373437286, + "language_loss": 0.82978344, + "learning_rate": 1.0096849316542217e-06, + "loss": 0.85154366, + "num_input_tokens_seen": 120763710, + "step": 5613, + "time_per_iteration": 2.7122652530670166 + }, + { + "auxiliary_loss_clip": 0.01100533, + "auxiliary_loss_mlp": 0.01027191, + "balance_loss_clip": 1.03615487, + "balance_loss_mlp": 1.01847112, + "epoch": 0.6750435880478567, + "flos": 26499489050880.0, + "grad_norm": 2.121723571136979, + "language_loss": 0.74557114, + "learning_rate": 1.0090082337192643e-06, + "loss": 0.76684844, + "num_input_tokens_seen": 120783355, + "step": 5614, + "time_per_iteration": 3.034308433532715 + }, + { + "auxiliary_loss_clip": 0.01116532, + "auxiliary_loss_mlp": 0.01028887, + "balance_loss_clip": 1.03606057, + "balance_loss_mlp": 1.02125716, + "epoch": 0.6751638309384957, + "flos": 23404313076480.0, + "grad_norm": 2.5407662462666636, + "language_loss": 0.78564721, + "learning_rate": 1.0083316861169705e-06, + "loss": 0.80710137, + "num_input_tokens_seen": 120802090, + "step": 5615, + "time_per_iteration": 2.982926607131958 + }, + { + "auxiliary_loss_clip": 0.01150759, + "auxiliary_loss_mlp": 0.01027638, + "balance_loss_clip": 1.0444243, + "balance_loss_mlp": 1.01857185, + "epoch": 0.6752840738291348, + "flos": 23441408847360.0, + "grad_norm": 2.052543697318421, + "language_loss": 0.71127433, + "learning_rate": 1.0076552889499713e-06, + "loss": 0.73305827, + "num_input_tokens_seen": 120822855, + "step": 5616, + "time_per_iteration": 2.848651170730591 + }, + { + "auxiliary_loss_clip": 0.01162805, + "auxiliary_loss_mlp": 0.01030037, + "balance_loss_clip": 1.04875731, + "balance_loss_mlp": 1.02187753, + "epoch": 0.675404316719774, + "flos": 30335826257280.0, + "grad_norm": 7.27471306246946, + "language_loss": 0.73764729, + "learning_rate": 1.006979042320876e-06, + "loss": 0.75957572, + "num_input_tokens_seen": 120843070, + "step": 5617, + "time_per_iteration": 2.7821903228759766 + }, + { + "auxiliary_loss_clip": 0.01148807, + "auxiliary_loss_mlp": 0.01020139, + "balance_loss_clip": 1.043167, + "balance_loss_mlp": 1.01222682, + "epoch": 0.675524559610413, + "flos": 23622613983360.0, + "grad_norm": 2.201443056412673, + "language_loss": 0.6287505, + "learning_rate": 1.0063029463322702e-06, + "loss": 0.65043998, + "num_input_tokens_seen": 120863345, + "step": 5618, + "time_per_iteration": 2.891073226928711 + }, + { + "auxiliary_loss_clip": 0.0113013, + "auxiliary_loss_mlp": 0.00885828, + "balance_loss_clip": 1.04141068, + "balance_loss_mlp": 1.00061703, + "epoch": 0.6756448025010521, + "flos": 21248631279360.0, + "grad_norm": 2.617085865709295, + "language_loss": 0.75570285, + "learning_rate": 1.0056270010867164e-06, + "loss": 0.7758624, + "num_input_tokens_seen": 120880915, + "step": 5619, + "time_per_iteration": 2.850674629211426 + }, + { + "auxiliary_loss_clip": 0.01154579, + "auxiliary_loss_mlp": 0.01026354, + "balance_loss_clip": 1.0424341, + "balance_loss_mlp": 1.01761591, + "epoch": 0.6757650453916912, + "flos": 21646521210240.0, + "grad_norm": 2.4048547391086976, + "language_loss": 0.77916801, + "learning_rate": 1.004951206686758e-06, + "loss": 0.80097735, + "num_input_tokens_seen": 120899190, + "step": 5620, + "time_per_iteration": 3.6955552101135254 + }, + { + "auxiliary_loss_clip": 0.01156998, + "auxiliary_loss_mlp": 0.01034402, + "balance_loss_clip": 1.04443061, + "balance_loss_mlp": 1.02629614, + "epoch": 0.6758852882823303, + "flos": 21795658479360.0, + "grad_norm": 1.989458519022935, + "language_loss": 0.71866316, + "learning_rate": 1.0042755632349087e-06, + "loss": 0.7405771, + "num_input_tokens_seen": 120916080, + "step": 5621, + "time_per_iteration": 2.727579355239868 + }, + { + "auxiliary_loss_clip": 0.01143787, + "auxiliary_loss_mlp": 0.0102591, + "balance_loss_clip": 1.04516137, + "balance_loss_mlp": 1.01737463, + "epoch": 0.6760055311729694, + "flos": 27088783580160.0, + "grad_norm": 2.061441784067372, + "language_loss": 0.62477213, + "learning_rate": 1.0036000708336653e-06, + "loss": 0.64646912, + "num_input_tokens_seen": 120935210, + "step": 5622, + "time_per_iteration": 3.8603405952453613 + }, + { + "auxiliary_loss_clip": 0.01158841, + "auxiliary_loss_mlp": 0.01027173, + "balance_loss_clip": 1.04736829, + "balance_loss_mlp": 1.01897144, + "epoch": 0.6761257740636085, + "flos": 17999792922240.0, + "grad_norm": 2.0554199066557786, + "language_loss": 0.79621398, + "learning_rate": 1.0029247295854984e-06, + "loss": 0.81807411, + "num_input_tokens_seen": 120951830, + "step": 5623, + "time_per_iteration": 2.7725207805633545 + }, + { + "auxiliary_loss_clip": 0.01151012, + "auxiliary_loss_mlp": 0.01030126, + "balance_loss_clip": 1.04654396, + "balance_loss_mlp": 1.02189445, + "epoch": 0.6762460169542476, + "flos": 15121912273920.0, + "grad_norm": 2.0430957924956417, + "language_loss": 0.72059494, + "learning_rate": 1.0022495395928588e-06, + "loss": 0.74240631, + "num_input_tokens_seen": 120970310, + "step": 5624, + "time_per_iteration": 2.720852851867676 + }, + { + "auxiliary_loss_clip": 0.01063341, + "auxiliary_loss_mlp": 0.01002089, + "balance_loss_clip": 1.01049066, + "balance_loss_mlp": 1.00052726, + "epoch": 0.6763662598448866, + "flos": 67886970030720.0, + "grad_norm": 0.7924429998937633, + "language_loss": 0.62359178, + "learning_rate": 1.0015745009581697e-06, + "loss": 0.6442461, + "num_input_tokens_seen": 121031915, + "step": 5625, + "time_per_iteration": 4.344590902328491 + }, + { + "auxiliary_loss_clip": 0.01163361, + "auxiliary_loss_mlp": 0.01025446, + "balance_loss_clip": 1.04877806, + "balance_loss_mlp": 1.01704764, + "epoch": 0.6764865027355258, + "flos": 20631829910400.0, + "grad_norm": 1.9957547004288043, + "language_loss": 0.66807693, + "learning_rate": 1.0008996137838343e-06, + "loss": 0.68996501, + "num_input_tokens_seen": 121050890, + "step": 5626, + "time_per_iteration": 3.6899075508117676 + }, + { + "auxiliary_loss_clip": 0.01177299, + "auxiliary_loss_mlp": 0.01026399, + "balance_loss_clip": 1.04914927, + "balance_loss_mlp": 1.01757169, + "epoch": 0.6766067456261649, + "flos": 21215809226880.0, + "grad_norm": 1.9584504395901758, + "language_loss": 0.80179322, + "learning_rate": 1.000224878172234e-06, + "loss": 0.82383025, + "num_input_tokens_seen": 121070015, + "step": 5627, + "time_per_iteration": 2.6366639137268066 + }, + { + "auxiliary_loss_clip": 0.01157065, + "auxiliary_loss_mlp": 0.01029947, + "balance_loss_clip": 1.04648948, + "balance_loss_mlp": 1.02185619, + "epoch": 0.6767269885168039, + "flos": 19938251220480.0, + "grad_norm": 2.1853311026965585, + "language_loss": 0.72801268, + "learning_rate": 9.99550294225724e-07, + "loss": 0.74988282, + "num_input_tokens_seen": 121089170, + "step": 5628, + "time_per_iteration": 2.7375500202178955 + }, + { + "auxiliary_loss_clip": 0.01134162, + "auxiliary_loss_mlp": 0.01026611, + "balance_loss_clip": 1.03892386, + "balance_loss_mlp": 1.01818848, + "epoch": 0.6768472314074431, + "flos": 20814076540800.0, + "grad_norm": 1.8762036391596286, + "language_loss": 0.72464895, + "learning_rate": 9.988758620466402e-07, + "loss": 0.74625665, + "num_input_tokens_seen": 121108040, + "step": 5629, + "time_per_iteration": 2.8002495765686035 + }, + { + "auxiliary_loss_clip": 0.01124388, + "auxiliary_loss_mlp": 0.01028132, + "balance_loss_clip": 1.04193759, + "balance_loss_mlp": 1.02095532, + "epoch": 0.6769674742980821, + "flos": 23186012169600.0, + "grad_norm": 1.7770605567750193, + "language_loss": 0.76075053, + "learning_rate": 9.982015817372917e-07, + "loss": 0.78227568, + "num_input_tokens_seen": 121128480, + "step": 5630, + "time_per_iteration": 2.9612152576446533 + }, + { + "auxiliary_loss_clip": 0.01132253, + "auxiliary_loss_mlp": 0.01030303, + "balance_loss_clip": 1.04088354, + "balance_loss_mlp": 1.02195263, + "epoch": 0.6770877171887212, + "flos": 24242934885120.0, + "grad_norm": 1.7577764795785078, + "language_loss": 0.82023072, + "learning_rate": 9.975274533999657e-07, + "loss": 0.84185636, + "num_input_tokens_seen": 121148010, + "step": 5631, + "time_per_iteration": 2.915558099746704 + }, + { + "auxiliary_loss_clip": 0.01173476, + "auxiliary_loss_mlp": 0.01029675, + "balance_loss_clip": 1.04828691, + "balance_loss_mlp": 1.0216527, + "epoch": 0.6772079600793603, + "flos": 18141567903360.0, + "grad_norm": 2.4329166790981613, + "language_loss": 0.84291208, + "learning_rate": 9.96853477136929e-07, + "loss": 0.86494362, + "num_input_tokens_seen": 121162755, + "step": 5632, + "time_per_iteration": 2.628520965576172 + }, + { + "auxiliary_loss_clip": 0.01135018, + "auxiliary_loss_mlp": 0.01031739, + "balance_loss_clip": 1.03912735, + "balance_loss_mlp": 1.02374017, + "epoch": 0.6773282029699994, + "flos": 22452069571200.0, + "grad_norm": 2.043387420055712, + "language_loss": 0.75251704, + "learning_rate": 9.96179653050422e-07, + "loss": 0.77418458, + "num_input_tokens_seen": 121182915, + "step": 5633, + "time_per_iteration": 2.8516228199005127 + }, + { + "auxiliary_loss_clip": 0.0113504, + "auxiliary_loss_mlp": 0.01030375, + "balance_loss_clip": 1.04384327, + "balance_loss_mlp": 1.02204251, + "epoch": 0.6774484458606385, + "flos": 18693730748160.0, + "grad_norm": 1.923072867476387, + "language_loss": 0.74403381, + "learning_rate": 9.955059812426635e-07, + "loss": 0.765688, + "num_input_tokens_seen": 121200445, + "step": 5634, + "time_per_iteration": 2.8048250675201416 + }, + { + "auxiliary_loss_clip": 0.01171511, + "auxiliary_loss_mlp": 0.01027285, + "balance_loss_clip": 1.04934049, + "balance_loss_mlp": 1.01896167, + "epoch": 0.6775686887512776, + "flos": 25994046821760.0, + "grad_norm": 2.2391106343802054, + "language_loss": 0.82765591, + "learning_rate": 9.948324618158493e-07, + "loss": 0.84964389, + "num_input_tokens_seen": 121220785, + "step": 5635, + "time_per_iteration": 2.72512149810791 + }, + { + "auxiliary_loss_clip": 0.01163231, + "auxiliary_loss_mlp": 0.01040717, + "balance_loss_clip": 1.04563618, + "balance_loss_mlp": 1.03213394, + "epoch": 0.6776889316419167, + "flos": 13587987922560.0, + "grad_norm": 2.285304310399449, + "language_loss": 0.77610463, + "learning_rate": 9.941590948721502e-07, + "loss": 0.7981441, + "num_input_tokens_seen": 121237985, + "step": 5636, + "time_per_iteration": 2.7441189289093018 + }, + { + "auxiliary_loss_clip": 0.01148278, + "auxiliary_loss_mlp": 0.01022214, + "balance_loss_clip": 1.04462278, + "balance_loss_mlp": 1.01423037, + "epoch": 0.6778091745325557, + "flos": 27601121220480.0, + "grad_norm": 1.7615118533922172, + "language_loss": 0.76647472, + "learning_rate": 9.934858805137188e-07, + "loss": 0.78817964, + "num_input_tokens_seen": 121258635, + "step": 5637, + "time_per_iteration": 2.839334726333618 + }, + { + "auxiliary_loss_clip": 0.01157711, + "auxiliary_loss_mlp": 0.01026912, + "balance_loss_clip": 1.04629803, + "balance_loss_mlp": 1.01894259, + "epoch": 0.6779294174231949, + "flos": 18734058743040.0, + "grad_norm": 1.5670723513319929, + "language_loss": 0.80677319, + "learning_rate": 9.92812818842677e-07, + "loss": 0.82861942, + "num_input_tokens_seen": 121277810, + "step": 5638, + "time_per_iteration": 2.7614858150482178 + }, + { + "auxiliary_loss_clip": 0.01160171, + "auxiliary_loss_mlp": 0.01026097, + "balance_loss_clip": 1.04808283, + "balance_loss_mlp": 1.0178299, + "epoch": 0.678049660313834, + "flos": 45873797765760.0, + "grad_norm": 1.7011261620059603, + "language_loss": 0.64389908, + "learning_rate": 9.921399099611306e-07, + "loss": 0.66576171, + "num_input_tokens_seen": 121298975, + "step": 5639, + "time_per_iteration": 2.8809638023376465 + }, + { + "auxiliary_loss_clip": 0.01153377, + "auxiliary_loss_mlp": 0.0102738, + "balance_loss_clip": 1.04548836, + "balance_loss_mlp": 1.01951778, + "epoch": 0.678169903204473, + "flos": 19974556892160.0, + "grad_norm": 1.541312807096302, + "language_loss": 0.69043398, + "learning_rate": 9.914671539711588e-07, + "loss": 0.71224159, + "num_input_tokens_seen": 121318495, + "step": 5640, + "time_per_iteration": 2.7473032474517822 + }, + { + "auxiliary_loss_clip": 0.01108811, + "auxiliary_loss_mlp": 0.00886575, + "balance_loss_clip": 1.03910744, + "balance_loss_mlp": 1.00058413, + "epoch": 0.6782901460951122, + "flos": 21395613732480.0, + "grad_norm": 1.9214571379915995, + "language_loss": 0.78200048, + "learning_rate": 9.90794550974817e-07, + "loss": 0.80195433, + "num_input_tokens_seen": 121338890, + "step": 5641, + "time_per_iteration": 3.015805959701538 + }, + { + "auxiliary_loss_clip": 0.01139865, + "auxiliary_loss_mlp": 0.01032649, + "balance_loss_clip": 1.04370427, + "balance_loss_mlp": 1.02450967, + "epoch": 0.6784103889857512, + "flos": 21434002392960.0, + "grad_norm": 2.8370512669718435, + "language_loss": 0.81322151, + "learning_rate": 9.901221010741407e-07, + "loss": 0.83494663, + "num_input_tokens_seen": 121358210, + "step": 5642, + "time_per_iteration": 2.9923603534698486 + }, + { + "auxiliary_loss_clip": 0.01167265, + "auxiliary_loss_mlp": 0.01033411, + "balance_loss_clip": 1.04775429, + "balance_loss_mlp": 1.02498853, + "epoch": 0.6785306318763903, + "flos": 32671923091200.0, + "grad_norm": 2.607955489031604, + "language_loss": 0.74443078, + "learning_rate": 9.894498043711375e-07, + "loss": 0.76643753, + "num_input_tokens_seen": 121379955, + "step": 5643, + "time_per_iteration": 2.809718132019043 + }, + { + "auxiliary_loss_clip": 0.01149856, + "auxiliary_loss_mlp": 0.01026072, + "balance_loss_clip": 1.04372025, + "balance_loss_mlp": 1.01815665, + "epoch": 0.6786508747670293, + "flos": 25632139340160.0, + "grad_norm": 1.9803994496277262, + "language_loss": 0.69126749, + "learning_rate": 9.887776609677962e-07, + "loss": 0.71302682, + "num_input_tokens_seen": 121401325, + "step": 5644, + "time_per_iteration": 2.8648905754089355 + }, + { + "auxiliary_loss_clip": 0.01132135, + "auxiliary_loss_mlp": 0.01028403, + "balance_loss_clip": 1.03884733, + "balance_loss_mlp": 1.02046382, + "epoch": 0.6787711176576685, + "flos": 19171881619200.0, + "grad_norm": 1.660000725787369, + "language_loss": 0.72180945, + "learning_rate": 9.88105670966079e-07, + "loss": 0.74341488, + "num_input_tokens_seen": 121419785, + "step": 5645, + "time_per_iteration": 2.834561824798584 + }, + { + "auxiliary_loss_clip": 0.0112176, + "auxiliary_loss_mlp": 0.01027605, + "balance_loss_clip": 1.04320383, + "balance_loss_mlp": 1.01963568, + "epoch": 0.6788913605483076, + "flos": 13985159581440.0, + "grad_norm": 1.7902665006184475, + "language_loss": 0.78918827, + "learning_rate": 9.874338344679283e-07, + "loss": 0.81068188, + "num_input_tokens_seen": 121435630, + "step": 5646, + "time_per_iteration": 3.766116142272949 + }, + { + "auxiliary_loss_clip": 0.01170687, + "auxiliary_loss_mlp": 0.01028926, + "balance_loss_clip": 1.04840064, + "balance_loss_mlp": 1.02126122, + "epoch": 0.6790116034389466, + "flos": 22017586659840.0, + "grad_norm": 1.6367507527203682, + "language_loss": 0.73835576, + "learning_rate": 9.86762151575259e-07, + "loss": 0.7603519, + "num_input_tokens_seen": 121455625, + "step": 5647, + "time_per_iteration": 2.679421901702881 + }, + { + "auxiliary_loss_clip": 0.01127328, + "auxiliary_loss_mlp": 0.00885106, + "balance_loss_clip": 1.04450965, + "balance_loss_mlp": 1.00069952, + "epoch": 0.6791318463295858, + "flos": 20922454851840.0, + "grad_norm": 1.4941299322148722, + "language_loss": 0.80052257, + "learning_rate": 9.860906223899651e-07, + "loss": 0.82064694, + "num_input_tokens_seen": 121475020, + "step": 5648, + "time_per_iteration": 3.857788562774658 + }, + { + "auxiliary_loss_clip": 0.0116066, + "auxiliary_loss_mlp": 0.01029445, + "balance_loss_clip": 1.04716361, + "balance_loss_mlp": 1.02150548, + "epoch": 0.6792520892202248, + "flos": 28512749422080.0, + "grad_norm": 1.6989832795651325, + "language_loss": 0.75722533, + "learning_rate": 9.854192470139184e-07, + "loss": 0.77912641, + "num_input_tokens_seen": 121496500, + "step": 5649, + "time_per_iteration": 2.8028783798217773 + }, + { + "auxiliary_loss_clip": 0.0115439, + "auxiliary_loss_mlp": 0.01033465, + "balance_loss_clip": 1.04945111, + "balance_loss_mlp": 1.02563906, + "epoch": 0.6793723321108639, + "flos": 20011904058240.0, + "grad_norm": 2.0296809164266456, + "language_loss": 0.71507305, + "learning_rate": 9.847480255489645e-07, + "loss": 0.73695159, + "num_input_tokens_seen": 121515525, + "step": 5650, + "time_per_iteration": 2.8050525188446045 + }, + { + "auxiliary_loss_clip": 0.01159074, + "auxiliary_loss_mlp": 0.0103366, + "balance_loss_clip": 1.04697847, + "balance_loss_mlp": 1.02590561, + "epoch": 0.6794925750015031, + "flos": 26649488246400.0, + "grad_norm": 1.8670701333811406, + "language_loss": 0.69224292, + "learning_rate": 9.840769580969295e-07, + "loss": 0.71417028, + "num_input_tokens_seen": 121535965, + "step": 5651, + "time_per_iteration": 3.654453992843628 + }, + { + "auxiliary_loss_clip": 0.01155248, + "auxiliary_loss_mlp": 0.01023428, + "balance_loss_clip": 1.04540801, + "balance_loss_mlp": 1.01525056, + "epoch": 0.6796128178921421, + "flos": 21580374314880.0, + "grad_norm": 1.830864315226745, + "language_loss": 0.80184603, + "learning_rate": 9.834060447596114e-07, + "loss": 0.82363278, + "num_input_tokens_seen": 121555235, + "step": 5652, + "time_per_iteration": 2.64216685295105 + }, + { + "auxiliary_loss_clip": 0.01161244, + "auxiliary_loss_mlp": 0.01033281, + "balance_loss_clip": 1.04449737, + "balance_loss_mlp": 1.02502608, + "epoch": 0.6797330607827812, + "flos": 22492002516480.0, + "grad_norm": 1.8756473526727446, + "language_loss": 0.77995205, + "learning_rate": 9.827352856387868e-07, + "loss": 0.80189729, + "num_input_tokens_seen": 121574945, + "step": 5653, + "time_per_iteration": 3.6631100177764893 + }, + { + "auxiliary_loss_clip": 0.01036247, + "auxiliary_loss_mlp": 0.01003222, + "balance_loss_clip": 1.01206684, + "balance_loss_mlp": 1.00138593, + "epoch": 0.6798533036734203, + "flos": 66306648286080.0, + "grad_norm": 0.7738960926138123, + "language_loss": 0.64248991, + "learning_rate": 9.820646808362118e-07, + "loss": 0.66288459, + "num_input_tokens_seen": 121641200, + "step": 5654, + "time_per_iteration": 3.431635856628418 + }, + { + "auxiliary_loss_clip": 0.01149851, + "auxiliary_loss_mlp": 0.01026888, + "balance_loss_clip": 1.04714417, + "balance_loss_mlp": 1.01889539, + "epoch": 0.6799735465640594, + "flos": 16180163792640.0, + "grad_norm": 2.0683880673349435, + "language_loss": 0.72370577, + "learning_rate": 9.813942304536154e-07, + "loss": 0.74547315, + "num_input_tokens_seen": 121659170, + "step": 5655, + "time_per_iteration": 2.7533063888549805 + }, + { + "auxiliary_loss_clip": 0.01152794, + "auxiliary_loss_mlp": 0.01028936, + "balance_loss_clip": 1.04523551, + "balance_loss_mlp": 1.02104473, + "epoch": 0.6800937894546984, + "flos": 22125749489280.0, + "grad_norm": 1.9299062527934878, + "language_loss": 0.63354504, + "learning_rate": 9.807239345927043e-07, + "loss": 0.65536231, + "num_input_tokens_seen": 121679180, + "step": 5656, + "time_per_iteration": 2.706716537475586 + }, + { + "auxiliary_loss_clip": 0.01147255, + "auxiliary_loss_mlp": 0.01025915, + "balance_loss_clip": 1.04266953, + "balance_loss_mlp": 1.01780295, + "epoch": 0.6802140323453376, + "flos": 31612953300480.0, + "grad_norm": 2.3817409248049866, + "language_loss": 0.71902561, + "learning_rate": 9.80053793355162e-07, + "loss": 0.74075729, + "num_input_tokens_seen": 121697875, + "step": 5657, + "time_per_iteration": 2.805431365966797 + }, + { + "auxiliary_loss_clip": 0.01129102, + "auxiliary_loss_mlp": 0.0102596, + "balance_loss_clip": 1.04180384, + "balance_loss_mlp": 1.01806808, + "epoch": 0.6803342752359767, + "flos": 17712938908800.0, + "grad_norm": 1.921217704428516, + "language_loss": 0.75151849, + "learning_rate": 9.793838068426472e-07, + "loss": 0.77306914, + "num_input_tokens_seen": 121715570, + "step": 5658, + "time_per_iteration": 2.923877716064453 + }, + { + "auxiliary_loss_clip": 0.01171267, + "auxiliary_loss_mlp": 0.01025541, + "balance_loss_clip": 1.04856408, + "balance_loss_mlp": 1.01752746, + "epoch": 0.6804545181266157, + "flos": 11326800902400.0, + "grad_norm": 2.2301791044557144, + "language_loss": 0.60940677, + "learning_rate": 9.78713975156799e-07, + "loss": 0.63137484, + "num_input_tokens_seen": 121731435, + "step": 5659, + "time_per_iteration": 2.5944137573242188 + }, + { + "auxiliary_loss_clip": 0.01143617, + "auxiliary_loss_mlp": 0.01031157, + "balance_loss_clip": 1.046947, + "balance_loss_mlp": 1.0230689, + "epoch": 0.6805747610172549, + "flos": 29350976181120.0, + "grad_norm": 1.7082987496822115, + "language_loss": 0.71605682, + "learning_rate": 9.780442983992273e-07, + "loss": 0.73780453, + "num_input_tokens_seen": 121749950, + "step": 5660, + "time_per_iteration": 2.866015911102295 + }, + { + "auxiliary_loss_clip": 0.01146691, + "auxiliary_loss_mlp": 0.01025396, + "balance_loss_clip": 1.04552841, + "balance_loss_mlp": 1.01739705, + "epoch": 0.680695003907894, + "flos": 37631868612480.0, + "grad_norm": 4.307993458829562, + "language_loss": 0.71864015, + "learning_rate": 9.773747766715238e-07, + "loss": 0.74036098, + "num_input_tokens_seen": 121770770, + "step": 5661, + "time_per_iteration": 2.8410966396331787 + }, + { + "auxiliary_loss_clip": 0.01153655, + "auxiliary_loss_mlp": 0.01023576, + "balance_loss_clip": 1.04400527, + "balance_loss_mlp": 1.01583958, + "epoch": 0.680815246798533, + "flos": 22127365601280.0, + "grad_norm": 2.02360029845121, + "language_loss": 0.80295682, + "learning_rate": 9.767054100752536e-07, + "loss": 0.82472914, + "num_input_tokens_seen": 121790720, + "step": 5662, + "time_per_iteration": 2.8637008666992188 + }, + { + "auxiliary_loss_clip": 0.01143151, + "auxiliary_loss_mlp": 0.01024806, + "balance_loss_clip": 1.04735327, + "balance_loss_mlp": 1.01669347, + "epoch": 0.6809354896891722, + "flos": 17201822330880.0, + "grad_norm": 1.9347182512936385, + "language_loss": 0.81578171, + "learning_rate": 9.760361987119584e-07, + "loss": 0.83746123, + "num_input_tokens_seen": 121808455, + "step": 5663, + "time_per_iteration": 3.010998487472534 + }, + { + "auxiliary_loss_clip": 0.01152011, + "auxiliary_loss_mlp": 0.01032387, + "balance_loss_clip": 1.0459404, + "balance_loss_mlp": 1.02429247, + "epoch": 0.6810557325798112, + "flos": 12458166554880.0, + "grad_norm": 1.9443752649561596, + "language_loss": 0.67794627, + "learning_rate": 9.753671426831592e-07, + "loss": 0.69979024, + "num_input_tokens_seen": 121824470, + "step": 5664, + "time_per_iteration": 2.760655641555786 + }, + { + "auxiliary_loss_clip": 0.01153437, + "auxiliary_loss_mlp": 0.01027165, + "balance_loss_clip": 1.04267001, + "balance_loss_mlp": 1.01869476, + "epoch": 0.6811759754704503, + "flos": 22156165330560.0, + "grad_norm": 2.844268257582314, + "language_loss": 0.79648656, + "learning_rate": 9.746982420903483e-07, + "loss": 0.81829256, + "num_input_tokens_seen": 121842665, + "step": 5665, + "time_per_iteration": 2.6894724369049072 + }, + { + "auxiliary_loss_clip": 0.01156224, + "auxiliary_loss_mlp": 0.01021432, + "balance_loss_clip": 1.04720151, + "balance_loss_mlp": 1.01389825, + "epoch": 0.6812962183610894, + "flos": 17525377065600.0, + "grad_norm": 1.872046733313311, + "language_loss": 0.74821162, + "learning_rate": 9.740294970349993e-07, + "loss": 0.76998818, + "num_input_tokens_seen": 121859080, + "step": 5666, + "time_per_iteration": 2.673424005508423 + }, + { + "auxiliary_loss_clip": 0.01053211, + "auxiliary_loss_mlp": 0.0100311, + "balance_loss_clip": 1.01128662, + "balance_loss_mlp": 1.00145328, + "epoch": 0.6814164612517285, + "flos": 60274480855680.0, + "grad_norm": 0.8918699506085159, + "language_loss": 0.60857427, + "learning_rate": 9.733609076185594e-07, + "loss": 0.62913752, + "num_input_tokens_seen": 121915485, + "step": 5667, + "time_per_iteration": 3.18315052986145 + }, + { + "auxiliary_loss_clip": 0.01162587, + "auxiliary_loss_mlp": 0.01026874, + "balance_loss_clip": 1.048527, + "balance_loss_mlp": 1.01864874, + "epoch": 0.6815367041423676, + "flos": 19317750750720.0, + "grad_norm": 1.8294111093447631, + "language_loss": 0.84003139, + "learning_rate": 9.72692473942455e-07, + "loss": 0.86192596, + "num_input_tokens_seen": 121932710, + "step": 5668, + "time_per_iteration": 2.6882050037384033 + }, + { + "auxiliary_loss_clip": 0.01138426, + "auxiliary_loss_mlp": 0.01031081, + "balance_loss_clip": 1.0463984, + "balance_loss_mlp": 1.02284956, + "epoch": 0.6816569470330067, + "flos": 22161696024960.0, + "grad_norm": 2.1797415809123315, + "language_loss": 0.77796483, + "learning_rate": 9.720241961080849e-07, + "loss": 0.79965997, + "num_input_tokens_seen": 121952025, + "step": 5669, + "time_per_iteration": 2.8275952339172363 + }, + { + "auxiliary_loss_clip": 0.01167996, + "auxiliary_loss_mlp": 0.01030378, + "balance_loss_clip": 1.04564345, + "balance_loss_mlp": 1.02208161, + "epoch": 0.6817771899236458, + "flos": 41463501137280.0, + "grad_norm": 1.8952792863543624, + "language_loss": 0.72974396, + "learning_rate": 9.713560742168259e-07, + "loss": 0.7517277, + "num_input_tokens_seen": 121974650, + "step": 5670, + "time_per_iteration": 2.8493900299072266 + }, + { + "auxiliary_loss_clip": 0.01132927, + "auxiliary_loss_mlp": 0.01035261, + "balance_loss_clip": 1.042045, + "balance_loss_mlp": 1.02790618, + "epoch": 0.6818974328142848, + "flos": 21106138026240.0, + "grad_norm": 2.7346455666797516, + "language_loss": 0.71421134, + "learning_rate": 9.706881083700333e-07, + "loss": 0.73589319, + "num_input_tokens_seen": 121994335, + "step": 5671, + "time_per_iteration": 2.7636280059814453 + }, + { + "auxiliary_loss_clip": 0.01123448, + "auxiliary_loss_mlp": 0.0103055, + "balance_loss_clip": 1.04435956, + "balance_loss_mlp": 1.02293229, + "epoch": 0.682017675704924, + "flos": 20441897769600.0, + "grad_norm": 2.1979597542886515, + "language_loss": 0.82474148, + "learning_rate": 9.700202986690357e-07, + "loss": 0.84628141, + "num_input_tokens_seen": 122012635, + "step": 5672, + "time_per_iteration": 3.7970545291900635 + }, + { + "auxiliary_loss_clip": 0.01159671, + "auxiliary_loss_mlp": 0.00886, + "balance_loss_clip": 1.04616618, + "balance_loss_mlp": 1.00078678, + "epoch": 0.682137918595563, + "flos": 20044438801920.0, + "grad_norm": 2.0796453894038014, + "language_loss": 0.66951722, + "learning_rate": 9.693526452151413e-07, + "loss": 0.68997389, + "num_input_tokens_seen": 122031685, + "step": 5673, + "time_per_iteration": 2.7225141525268555 + }, + { + "auxiliary_loss_clip": 0.0114905, + "auxiliary_loss_mlp": 0.01028404, + "balance_loss_clip": 1.04392838, + "balance_loss_mlp": 1.01939523, + "epoch": 0.6822581614862021, + "flos": 31684559063040.0, + "grad_norm": 2.2736268971358156, + "language_loss": 0.75339055, + "learning_rate": 9.686851481096305e-07, + "loss": 0.77516508, + "num_input_tokens_seen": 122052995, + "step": 5674, + "time_per_iteration": 3.8635754585266113 + }, + { + "auxiliary_loss_clip": 0.01112728, + "auxiliary_loss_mlp": 0.0102889, + "balance_loss_clip": 1.04020786, + "balance_loss_mlp": 1.0206883, + "epoch": 0.6823784043768413, + "flos": 23477570864640.0, + "grad_norm": 2.083521831472441, + "language_loss": 0.71868432, + "learning_rate": 9.68017807453762e-07, + "loss": 0.7401005, + "num_input_tokens_seen": 122071740, + "step": 5675, + "time_per_iteration": 2.8763391971588135 + }, + { + "auxiliary_loss_clip": 0.01157111, + "auxiliary_loss_mlp": 0.00885251, + "balance_loss_clip": 1.04791164, + "balance_loss_mlp": 1.00067186, + "epoch": 0.6824986472674803, + "flos": 14137134024960.0, + "grad_norm": 1.8352501408944697, + "language_loss": 0.7376585, + "learning_rate": 9.673506233487721e-07, + "loss": 0.75808215, + "num_input_tokens_seen": 122089705, + "step": 5676, + "time_per_iteration": 2.7237184047698975 + }, + { + "auxiliary_loss_clip": 0.01149705, + "auxiliary_loss_mlp": 0.00885064, + "balance_loss_clip": 1.0431639, + "balance_loss_mlp": 1.00073504, + "epoch": 0.6826188901581194, + "flos": 21504997624320.0, + "grad_norm": 1.7185757031144047, + "language_loss": 0.85972047, + "learning_rate": 9.666835958958717e-07, + "loss": 0.88006818, + "num_input_tokens_seen": 122109025, + "step": 5677, + "time_per_iteration": 3.6875479221343994 + }, + { + "auxiliary_loss_clip": 0.01169784, + "auxiliary_loss_mlp": 0.01026828, + "balance_loss_clip": 1.04814553, + "balance_loss_mlp": 1.01889503, + "epoch": 0.6827391330487584, + "flos": 20810126044800.0, + "grad_norm": 1.9151839522500977, + "language_loss": 0.80442536, + "learning_rate": 9.660167251962484e-07, + "loss": 0.82639146, + "num_input_tokens_seen": 122127385, + "step": 5678, + "time_per_iteration": 2.6828815937042236 + }, + { + "auxiliary_loss_clip": 0.01143913, + "auxiliary_loss_mlp": 0.01025562, + "balance_loss_clip": 1.04205549, + "balance_loss_mlp": 1.01806939, + "epoch": 0.6828593759393976, + "flos": 21688788539520.0, + "grad_norm": 1.6439123163594798, + "language_loss": 0.77960742, + "learning_rate": 9.653500113510654e-07, + "loss": 0.80130213, + "num_input_tokens_seen": 122146500, + "step": 5679, + "time_per_iteration": 3.6658477783203125 + }, + { + "auxiliary_loss_clip": 0.01145962, + "auxiliary_loss_mlp": 0.01028461, + "balance_loss_clip": 1.04265106, + "balance_loss_mlp": 1.02023518, + "epoch": 0.6829796188300367, + "flos": 25337707557120.0, + "grad_norm": 3.816334433257923, + "language_loss": 0.67043978, + "learning_rate": 9.646834544614627e-07, + "loss": 0.69218397, + "num_input_tokens_seen": 122167000, + "step": 5680, + "time_per_iteration": 2.888221025466919 + }, + { + "auxiliary_loss_clip": 0.01143864, + "auxiliary_loss_mlp": 0.01025156, + "balance_loss_clip": 1.04506588, + "balance_loss_mlp": 1.0177232, + "epoch": 0.6830998617206757, + "flos": 20704800389760.0, + "grad_norm": 2.5601369124746824, + "language_loss": 0.76178539, + "learning_rate": 9.64017054628558e-07, + "loss": 0.78347564, + "num_input_tokens_seen": 122185825, + "step": 5681, + "time_per_iteration": 2.7797811031341553 + }, + { + "auxiliary_loss_clip": 0.01133547, + "auxiliary_loss_mlp": 0.01029881, + "balance_loss_clip": 1.04092932, + "balance_loss_mlp": 1.02198339, + "epoch": 0.6832201046113149, + "flos": 21726638496000.0, + "grad_norm": 1.6309339210728266, + "language_loss": 0.78815192, + "learning_rate": 9.63350811953441e-07, + "loss": 0.8097862, + "num_input_tokens_seen": 122206200, + "step": 5682, + "time_per_iteration": 2.884786367416382 + }, + { + "auxiliary_loss_clip": 0.01141905, + "auxiliary_loss_mlp": 0.01025734, + "balance_loss_clip": 1.04224849, + "balance_loss_mlp": 1.01808643, + "epoch": 0.6833403475019539, + "flos": 19536554448000.0, + "grad_norm": 2.178879962657848, + "language_loss": 0.70271736, + "learning_rate": 9.626847265371826e-07, + "loss": 0.72439378, + "num_input_tokens_seen": 122225520, + "step": 5683, + "time_per_iteration": 2.8242948055267334 + }, + { + "auxiliary_loss_clip": 0.0114356, + "auxiliary_loss_mlp": 0.01027353, + "balance_loss_clip": 1.04257512, + "balance_loss_mlp": 1.01924109, + "epoch": 0.683460590392593, + "flos": 19352153001600.0, + "grad_norm": 1.880202062421156, + "language_loss": 0.78385413, + "learning_rate": 9.620187984808262e-07, + "loss": 0.80556315, + "num_input_tokens_seen": 122244320, + "step": 5684, + "time_per_iteration": 2.774660587310791 + }, + { + "auxiliary_loss_clip": 0.01153588, + "auxiliary_loss_mlp": 0.0088564, + "balance_loss_clip": 1.04729116, + "balance_loss_mlp": 1.00071514, + "epoch": 0.6835808332832322, + "flos": 23288500650240.0, + "grad_norm": 1.7660292709088854, + "language_loss": 0.85742331, + "learning_rate": 9.613530278853919e-07, + "loss": 0.8778156, + "num_input_tokens_seen": 122264295, + "step": 5685, + "time_per_iteration": 2.7768332958221436 + }, + { + "auxiliary_loss_clip": 0.01159634, + "auxiliary_loss_mlp": 0.01026841, + "balance_loss_clip": 1.04768658, + "balance_loss_mlp": 1.01903903, + "epoch": 0.6837010761738712, + "flos": 21653416621440.0, + "grad_norm": 1.9303705893261978, + "language_loss": 0.73900962, + "learning_rate": 9.60687414851879e-07, + "loss": 0.76087439, + "num_input_tokens_seen": 122285300, + "step": 5686, + "time_per_iteration": 2.7126047611236572 + }, + { + "auxiliary_loss_clip": 0.01147173, + "auxiliary_loss_mlp": 0.01032262, + "balance_loss_clip": 1.04587412, + "balance_loss_mlp": 1.02388167, + "epoch": 0.6838213190645103, + "flos": 17566387418880.0, + "grad_norm": 4.847021124821998, + "language_loss": 0.7731539, + "learning_rate": 9.600219594812575e-07, + "loss": 0.79494828, + "num_input_tokens_seen": 122303240, + "step": 5687, + "time_per_iteration": 2.746727705001831 + }, + { + "auxiliary_loss_clip": 0.01167997, + "auxiliary_loss_mlp": 0.01026617, + "balance_loss_clip": 1.0462569, + "balance_loss_mlp": 1.01955032, + "epoch": 0.6839415619551494, + "flos": 23112538899840.0, + "grad_norm": 1.7448202865657698, + "language_loss": 0.72941619, + "learning_rate": 9.593566618744786e-07, + "loss": 0.75136232, + "num_input_tokens_seen": 122323390, + "step": 5688, + "time_per_iteration": 2.6602022647857666 + }, + { + "auxiliary_loss_clip": 0.01168108, + "auxiliary_loss_mlp": 0.01026289, + "balance_loss_clip": 1.04608083, + "balance_loss_mlp": 1.01854897, + "epoch": 0.6840618048457885, + "flos": 22127868391680.0, + "grad_norm": 2.6105581150858534, + "language_loss": 0.74003935, + "learning_rate": 9.58691522132466e-07, + "loss": 0.76198334, + "num_input_tokens_seen": 122342200, + "step": 5689, + "time_per_iteration": 2.7235915660858154 + }, + { + "auxiliary_loss_clip": 0.01157041, + "auxiliary_loss_mlp": 0.01029312, + "balance_loss_clip": 1.04652524, + "balance_loss_mlp": 1.02155125, + "epoch": 0.6841820477364275, + "flos": 22015898720640.0, + "grad_norm": 2.4936896439121834, + "language_loss": 0.84591001, + "learning_rate": 9.58026540356123e-07, + "loss": 0.86777353, + "num_input_tokens_seen": 122360465, + "step": 5690, + "time_per_iteration": 2.8473119735717773 + }, + { + "auxiliary_loss_clip": 0.01164419, + "auxiliary_loss_mlp": 0.01023275, + "balance_loss_clip": 1.04702103, + "balance_loss_mlp": 1.01524043, + "epoch": 0.6843022906270667, + "flos": 24900531125760.0, + "grad_norm": 1.9199475746259491, + "language_loss": 0.86335051, + "learning_rate": 9.573617166463246e-07, + "loss": 0.88522738, + "num_input_tokens_seen": 122381680, + "step": 5691, + "time_per_iteration": 2.7294743061065674 + }, + { + "auxiliary_loss_clip": 0.01154421, + "auxiliary_loss_mlp": 0.01031465, + "balance_loss_clip": 1.04331779, + "balance_loss_mlp": 1.02321601, + "epoch": 0.6844225335177058, + "flos": 19969924037760.0, + "grad_norm": 5.440339360705573, + "language_loss": 0.59887457, + "learning_rate": 9.56697051103924e-07, + "loss": 0.62073338, + "num_input_tokens_seen": 122399120, + "step": 5692, + "time_per_iteration": 2.7808499336242676 + }, + { + "auxiliary_loss_clip": 0.01148191, + "auxiliary_loss_mlp": 0.01026712, + "balance_loss_clip": 1.04390764, + "balance_loss_mlp": 1.01963079, + "epoch": 0.6845427764083448, + "flos": 25883334126720.0, + "grad_norm": 2.246874608803116, + "language_loss": 0.81284213, + "learning_rate": 9.560325438297522e-07, + "loss": 0.83459121, + "num_input_tokens_seen": 122417430, + "step": 5693, + "time_per_iteration": 2.7379162311553955 + }, + { + "auxiliary_loss_clip": 0.01153802, + "auxiliary_loss_mlp": 0.01033547, + "balance_loss_clip": 1.04926002, + "balance_loss_mlp": 1.02550042, + "epoch": 0.684663019298984, + "flos": 18880143356160.0, + "grad_norm": 2.1631957303839013, + "language_loss": 0.87007689, + "learning_rate": 9.553681949246127e-07, + "loss": 0.89195037, + "num_input_tokens_seen": 122435055, + "step": 5694, + "time_per_iteration": 2.714916467666626 + }, + { + "auxiliary_loss_clip": 0.0115298, + "auxiliary_loss_mlp": 0.01029666, + "balance_loss_clip": 1.04742968, + "balance_loss_mlp": 1.02124405, + "epoch": 0.684783262189623, + "flos": 54193725302400.0, + "grad_norm": 1.929754273353501, + "language_loss": 0.75353426, + "learning_rate": 9.547040044892886e-07, + "loss": 0.7753607, + "num_input_tokens_seen": 122462570, + "step": 5695, + "time_per_iteration": 3.0794801712036133 + }, + { + "auxiliary_loss_clip": 0.01057051, + "auxiliary_loss_mlp": 0.01003987, + "balance_loss_clip": 1.00791025, + "balance_loss_mlp": 1.00242543, + "epoch": 0.6849035050802621, + "flos": 63970264143360.0, + "grad_norm": 0.8668354177668672, + "language_loss": 0.60131365, + "learning_rate": 9.540399726245354e-07, + "loss": 0.62192404, + "num_input_tokens_seen": 122519275, + "step": 5696, + "time_per_iteration": 3.209820508956909 + }, + { + "auxiliary_loss_clip": 0.0114738, + "auxiliary_loss_mlp": 0.01024981, + "balance_loss_clip": 1.04248977, + "balance_loss_mlp": 1.01643956, + "epoch": 0.6850237479709013, + "flos": 25224121774080.0, + "grad_norm": 1.9471876181885748, + "language_loss": 0.68971908, + "learning_rate": 9.533760994310859e-07, + "loss": 0.71144271, + "num_input_tokens_seen": 122539675, + "step": 5697, + "time_per_iteration": 2.7706692218780518 + }, + { + "auxiliary_loss_clip": 0.0117145, + "auxiliary_loss_mlp": 0.01024935, + "balance_loss_clip": 1.04770136, + "balance_loss_mlp": 1.01669216, + "epoch": 0.6851439908615403, + "flos": 19354128249600.0, + "grad_norm": 2.2335381671377275, + "language_loss": 0.75361687, + "learning_rate": 9.527123850096508e-07, + "loss": 0.7755807, + "num_input_tokens_seen": 122558035, + "step": 5698, + "time_per_iteration": 3.5648720264434814 + }, + { + "auxiliary_loss_clip": 0.01156913, + "auxiliary_loss_mlp": 0.01027181, + "balance_loss_clip": 1.04524231, + "balance_loss_mlp": 1.01937556, + "epoch": 0.6852642337521794, + "flos": 23182133500800.0, + "grad_norm": 2.220101568234814, + "language_loss": 0.71731007, + "learning_rate": 9.520488294609142e-07, + "loss": 0.739151, + "num_input_tokens_seen": 122576815, + "step": 5699, + "time_per_iteration": 2.7674694061279297 + }, + { + "auxiliary_loss_clip": 0.0103767, + "auxiliary_loss_mlp": 0.01000535, + "balance_loss_clip": 1.00954151, + "balance_loss_mlp": 0.9988541, + "epoch": 0.6853844766428185, + "flos": 62647206583680.0, + "grad_norm": 0.781070405949586, + "language_loss": 0.53809291, + "learning_rate": 9.513854328855368e-07, + "loss": 0.55847496, + "num_input_tokens_seen": 122634690, + "step": 5700, + "time_per_iteration": 4.3319292068481445 + }, + { + "auxiliary_loss_clip": 0.01166639, + "auxiliary_loss_mlp": 0.01021534, + "balance_loss_clip": 1.0456655, + "balance_loss_mlp": 1.01394677, + "epoch": 0.6855047195334576, + "flos": 23437242869760.0, + "grad_norm": 1.9546221513999704, + "language_loss": 0.81305492, + "learning_rate": 9.507221953841558e-07, + "loss": 0.83493674, + "num_input_tokens_seen": 122652320, + "step": 5701, + "time_per_iteration": 2.70487117767334 + }, + { + "auxiliary_loss_clip": 0.01162064, + "auxiliary_loss_mlp": 0.0102982, + "balance_loss_clip": 1.0487088, + "balance_loss_mlp": 1.02198517, + "epoch": 0.6856249624240967, + "flos": 20664831530880.0, + "grad_norm": 1.6154537447940642, + "language_loss": 0.77770424, + "learning_rate": 9.500591170573824e-07, + "loss": 0.79962307, + "num_input_tokens_seen": 122672340, + "step": 5702, + "time_per_iteration": 2.926936388015747 + }, + { + "auxiliary_loss_clip": 0.01129911, + "auxiliary_loss_mlp": 0.01030653, + "balance_loss_clip": 1.04328561, + "balance_loss_mlp": 1.02282667, + "epoch": 0.6857452053147358, + "flos": 17087302794240.0, + "grad_norm": 2.365889156069518, + "language_loss": 0.73850954, + "learning_rate": 9.493961980058078e-07, + "loss": 0.76011515, + "num_input_tokens_seen": 122689935, + "step": 5703, + "time_per_iteration": 3.71329665184021 + }, + { + "auxiliary_loss_clip": 0.01105962, + "auxiliary_loss_mlp": 0.01024937, + "balance_loss_clip": 1.03795397, + "balance_loss_mlp": 1.0170157, + "epoch": 0.6858654482053749, + "flos": 30847266057600.0, + "grad_norm": 1.8265682594267598, + "language_loss": 0.67608076, + "learning_rate": 9.48733438329993e-07, + "loss": 0.69738972, + "num_input_tokens_seen": 122710200, + "step": 5704, + "time_per_iteration": 2.9158244132995605 + }, + { + "auxiliary_loss_clip": 0.01168061, + "auxiliary_loss_mlp": 0.00885154, + "balance_loss_clip": 1.0473026, + "balance_loss_mlp": 1.00064993, + "epoch": 0.6859856910960139, + "flos": 28877314510080.0, + "grad_norm": 1.5637245283902887, + "language_loss": 0.74295437, + "learning_rate": 9.480708381304807e-07, + "loss": 0.7634865, + "num_input_tokens_seen": 122731495, + "step": 5705, + "time_per_iteration": 3.653747797012329 + }, + { + "auxiliary_loss_clip": 0.01131361, + "auxiliary_loss_mlp": 0.01025764, + "balance_loss_clip": 1.04380012, + "balance_loss_mlp": 1.01771188, + "epoch": 0.6861059339866531, + "flos": 19354523299200.0, + "grad_norm": 2.34785019718303, + "language_loss": 0.83372456, + "learning_rate": 9.474083975077858e-07, + "loss": 0.85529584, + "num_input_tokens_seen": 122748620, + "step": 5706, + "time_per_iteration": 2.8075485229492188 + }, + { + "auxiliary_loss_clip": 0.01155945, + "auxiliary_loss_mlp": 0.01029543, + "balance_loss_clip": 1.04543686, + "balance_loss_mlp": 1.02107358, + "epoch": 0.6862261768772921, + "flos": 22199976944640.0, + "grad_norm": 2.7124962147221026, + "language_loss": 0.79874945, + "learning_rate": 9.467461165623994e-07, + "loss": 0.82060444, + "num_input_tokens_seen": 122767670, + "step": 5707, + "time_per_iteration": 2.7192864418029785 + }, + { + "auxiliary_loss_clip": 0.01163566, + "auxiliary_loss_mlp": 0.01026362, + "balance_loss_clip": 1.04489255, + "balance_loss_mlp": 1.01872945, + "epoch": 0.6863464197679312, + "flos": 26285677344000.0, + "grad_norm": 1.8807194973449628, + "language_loss": 0.7955606, + "learning_rate": 9.46083995394791e-07, + "loss": 0.81745988, + "num_input_tokens_seen": 122785480, + "step": 5708, + "time_per_iteration": 2.6926121711730957 + }, + { + "auxiliary_loss_clip": 0.01162525, + "auxiliary_loss_mlp": 0.00885264, + "balance_loss_clip": 1.04772627, + "balance_loss_mlp": 1.000705, + "epoch": 0.6864666626585703, + "flos": 37815228564480.0, + "grad_norm": 2.5180990428603236, + "language_loss": 0.63331056, + "learning_rate": 9.454220341054012e-07, + "loss": 0.65378845, + "num_input_tokens_seen": 122810265, + "step": 5709, + "time_per_iteration": 2.837308168411255 + }, + { + "auxiliary_loss_clip": 0.01139344, + "auxiliary_loss_mlp": 0.01027521, + "balance_loss_clip": 1.04239535, + "balance_loss_mlp": 1.0198859, + "epoch": 0.6865869055492094, + "flos": 19391152193280.0, + "grad_norm": 2.171501606069549, + "language_loss": 0.80728084, + "learning_rate": 9.447602327946512e-07, + "loss": 0.82894951, + "num_input_tokens_seen": 122828905, + "step": 5710, + "time_per_iteration": 2.7294421195983887 + }, + { + "auxiliary_loss_clip": 0.01147932, + "auxiliary_loss_mlp": 0.0102877, + "balance_loss_clip": 1.0419699, + "balance_loss_mlp": 1.02077997, + "epoch": 0.6867071484398485, + "flos": 20375966355840.0, + "grad_norm": 1.777736779022514, + "language_loss": 0.76521081, + "learning_rate": 9.440985915629338e-07, + "loss": 0.78697777, + "num_input_tokens_seen": 122846235, + "step": 5711, + "time_per_iteration": 2.742029905319214 + }, + { + "auxiliary_loss_clip": 0.01168987, + "auxiliary_loss_mlp": 0.01025145, + "balance_loss_clip": 1.04910767, + "balance_loss_mlp": 1.01795101, + "epoch": 0.6868273913304875, + "flos": 15889143801600.0, + "grad_norm": 1.964071364433267, + "language_loss": 0.7324506, + "learning_rate": 9.434371105106223e-07, + "loss": 0.75439191, + "num_input_tokens_seen": 122863835, + "step": 5712, + "time_per_iteration": 2.6557936668395996 + }, + { + "auxiliary_loss_clip": 0.01136035, + "auxiliary_loss_mlp": 0.01020861, + "balance_loss_clip": 1.04151893, + "balance_loss_mlp": 1.01256716, + "epoch": 0.6869476342211267, + "flos": 24462492768000.0, + "grad_norm": 1.8289550948669053, + "language_loss": 0.70235437, + "learning_rate": 9.427757897380602e-07, + "loss": 0.72392333, + "num_input_tokens_seen": 122883235, + "step": 5713, + "time_per_iteration": 2.87449049949646 + }, + { + "auxiliary_loss_clip": 0.01141176, + "auxiliary_loss_mlp": 0.01036853, + "balance_loss_clip": 1.04605281, + "balance_loss_mlp": 1.02861524, + "epoch": 0.6870678771117658, + "flos": 18442571875200.0, + "grad_norm": 1.8995396701755443, + "language_loss": 0.8455025, + "learning_rate": 9.421146293455695e-07, + "loss": 0.86728275, + "num_input_tokens_seen": 122898975, + "step": 5714, + "time_per_iteration": 2.8482422828674316 + }, + { + "auxiliary_loss_clip": 0.01147561, + "auxiliary_loss_mlp": 0.01032647, + "balance_loss_clip": 1.04263306, + "balance_loss_mlp": 1.02501178, + "epoch": 0.6871881200024048, + "flos": 22200371994240.0, + "grad_norm": 2.079860796947039, + "language_loss": 0.68328357, + "learning_rate": 9.414536294334489e-07, + "loss": 0.70508564, + "num_input_tokens_seen": 122918995, + "step": 5715, + "time_per_iteration": 2.797546625137329 + }, + { + "auxiliary_loss_clip": 0.0115277, + "auxiliary_loss_mlp": 0.01026457, + "balance_loss_clip": 1.0424304, + "balance_loss_mlp": 1.0186553, + "epoch": 0.687308362893044, + "flos": 22127724737280.0, + "grad_norm": 2.018283355764644, + "language_loss": 0.69749796, + "learning_rate": 9.407927901019708e-07, + "loss": 0.71929026, + "num_input_tokens_seen": 122938125, + "step": 5716, + "time_per_iteration": 2.800673484802246 + }, + { + "auxiliary_loss_clip": 0.01158623, + "auxiliary_loss_mlp": 0.01030679, + "balance_loss_clip": 1.04479945, + "balance_loss_mlp": 1.02297187, + "epoch": 0.687428605783683, + "flos": 25040546340480.0, + "grad_norm": 2.2451074123286214, + "language_loss": 0.76726544, + "learning_rate": 9.401321114513854e-07, + "loss": 0.78915846, + "num_input_tokens_seen": 122957020, + "step": 5717, + "time_per_iteration": 2.929511547088623 + }, + { + "auxiliary_loss_clip": 0.01171507, + "auxiliary_loss_mlp": 0.01024841, + "balance_loss_clip": 1.04828084, + "balance_loss_mlp": 1.01688361, + "epoch": 0.6875488486743221, + "flos": 23770063313280.0, + "grad_norm": 3.3386205367305815, + "language_loss": 0.7565034, + "learning_rate": 9.394715935819155e-07, + "loss": 0.77846688, + "num_input_tokens_seen": 122977410, + "step": 5718, + "time_per_iteration": 2.655686378479004 + }, + { + "auxiliary_loss_clip": 0.01164592, + "auxiliary_loss_mlp": 0.01030365, + "balance_loss_clip": 1.04591537, + "balance_loss_mlp": 1.02234185, + "epoch": 0.6876690915649613, + "flos": 25516937445120.0, + "grad_norm": 2.1055105276987085, + "language_loss": 0.6241585, + "learning_rate": 9.388112365937608e-07, + "loss": 0.64610809, + "num_input_tokens_seen": 122996875, + "step": 5719, + "time_per_iteration": 2.710456371307373 + }, + { + "auxiliary_loss_clip": 0.01141796, + "auxiliary_loss_mlp": 0.01030091, + "balance_loss_clip": 1.04411709, + "balance_loss_mlp": 1.02200222, + "epoch": 0.6877893344556003, + "flos": 19427996568960.0, + "grad_norm": 1.8950846568958546, + "language_loss": 0.82496011, + "learning_rate": 9.381510405870985e-07, + "loss": 0.84667897, + "num_input_tokens_seen": 123015890, + "step": 5720, + "time_per_iteration": 2.803819179534912 + }, + { + "auxiliary_loss_clip": 0.011576, + "auxiliary_loss_mlp": 0.01039811, + "balance_loss_clip": 1.04426479, + "balance_loss_mlp": 1.03168726, + "epoch": 0.6879095773462394, + "flos": 18661303745280.0, + "grad_norm": 2.3758579502012602, + "language_loss": 0.77374631, + "learning_rate": 9.374910056620791e-07, + "loss": 0.7957204, + "num_input_tokens_seen": 123034955, + "step": 5721, + "time_per_iteration": 2.6429190635681152 + }, + { + "auxiliary_loss_clip": 0.01164854, + "auxiliary_loss_mlp": 0.01031478, + "balance_loss_clip": 1.04931986, + "balance_loss_mlp": 1.02341914, + "epoch": 0.6880298202368785, + "flos": 20883132437760.0, + "grad_norm": 1.760403249902407, + "language_loss": 0.80947566, + "learning_rate": 9.368311319188293e-07, + "loss": 0.83143902, + "num_input_tokens_seen": 123052770, + "step": 5722, + "time_per_iteration": 2.75022554397583 + }, + { + "auxiliary_loss_clip": 0.01139695, + "auxiliary_loss_mlp": 0.0102675, + "balance_loss_clip": 1.04164886, + "balance_loss_mlp": 1.01929319, + "epoch": 0.6881500631275176, + "flos": 30153292318080.0, + "grad_norm": 1.8081547969561762, + "language_loss": 0.79158092, + "learning_rate": 9.361714194574515e-07, + "loss": 0.8132453, + "num_input_tokens_seen": 123075105, + "step": 5723, + "time_per_iteration": 2.955435276031494 + }, + { + "auxiliary_loss_clip": 0.01061295, + "auxiliary_loss_mlp": 0.01003222, + "balance_loss_clip": 1.00841498, + "balance_loss_mlp": 1.00166059, + "epoch": 0.6882703060181566, + "flos": 66181537215360.0, + "grad_norm": 0.7375979664034454, + "language_loss": 0.58277774, + "learning_rate": 9.355118683780228e-07, + "loss": 0.603423, + "num_input_tokens_seen": 123145175, + "step": 5724, + "time_per_iteration": 4.286202430725098 + }, + { + "auxiliary_loss_clip": 0.01169359, + "auxiliary_loss_mlp": 0.0102752, + "balance_loss_clip": 1.04666877, + "balance_loss_mlp": 1.01972997, + "epoch": 0.6883905489087958, + "flos": 18214646123520.0, + "grad_norm": 2.0347422852127908, + "language_loss": 0.78943437, + "learning_rate": 9.348524787805987e-07, + "loss": 0.81140316, + "num_input_tokens_seen": 123160365, + "step": 5725, + "time_per_iteration": 2.54514479637146 + }, + { + "auxiliary_loss_clip": 0.0114464, + "auxiliary_loss_mlp": 0.01025311, + "balance_loss_clip": 1.0411551, + "balance_loss_mlp": 1.01710892, + "epoch": 0.6885107917994349, + "flos": 14056262553600.0, + "grad_norm": 2.641384607574854, + "language_loss": 0.85165906, + "learning_rate": 9.341932507652053e-07, + "loss": 0.87335861, + "num_input_tokens_seen": 123174855, + "step": 5726, + "time_per_iteration": 3.6355526447296143 + }, + { + "auxiliary_loss_clip": 0.01148301, + "auxiliary_loss_mlp": 0.01024253, + "balance_loss_clip": 1.0412482, + "balance_loss_mlp": 1.01569355, + "epoch": 0.6886310346900739, + "flos": 28690722334080.0, + "grad_norm": 2.0293983764734795, + "language_loss": 0.78610694, + "learning_rate": 9.335341844318489e-07, + "loss": 0.80783248, + "num_input_tokens_seen": 123194995, + "step": 5727, + "time_per_iteration": 2.6155686378479004 + }, + { + "auxiliary_loss_clip": 0.01150558, + "auxiliary_loss_mlp": 0.0102676, + "balance_loss_clip": 1.04511905, + "balance_loss_mlp": 1.01850486, + "epoch": 0.6887512775807131, + "flos": 24535319592960.0, + "grad_norm": 1.9624315370180174, + "language_loss": 0.73592216, + "learning_rate": 9.328752798805091e-07, + "loss": 0.75769526, + "num_input_tokens_seen": 123213465, + "step": 5728, + "time_per_iteration": 2.734328031539917 + }, + { + "auxiliary_loss_clip": 0.01161936, + "auxiliary_loss_mlp": 0.01028544, + "balance_loss_clip": 1.04754543, + "balance_loss_mlp": 1.02016997, + "epoch": 0.6888715204713521, + "flos": 22414363269120.0, + "grad_norm": 2.1491252595344434, + "language_loss": 0.75892198, + "learning_rate": 9.322165372111399e-07, + "loss": 0.78082681, + "num_input_tokens_seen": 123231610, + "step": 5729, + "time_per_iteration": 3.3628156185150146 + }, + { + "auxiliary_loss_clip": 0.01139237, + "auxiliary_loss_mlp": 0.01027301, + "balance_loss_clip": 1.0453496, + "balance_loss_mlp": 1.01985598, + "epoch": 0.6889917633619912, + "flos": 22054323294720.0, + "grad_norm": 1.8410048964376469, + "language_loss": 0.75773597, + "learning_rate": 9.315579565236747e-07, + "loss": 0.7794013, + "num_input_tokens_seen": 123250715, + "step": 5730, + "time_per_iteration": 2.6710314750671387 + }, + { + "auxiliary_loss_clip": 0.01149725, + "auxiliary_loss_mlp": 0.01031875, + "balance_loss_clip": 1.04811919, + "balance_loss_mlp": 1.02353036, + "epoch": 0.6891120062526304, + "flos": 23949724164480.0, + "grad_norm": 1.7252345828945765, + "language_loss": 0.74158311, + "learning_rate": 9.308995379180162e-07, + "loss": 0.76339912, + "num_input_tokens_seen": 123270270, + "step": 5731, + "time_per_iteration": 3.574864387512207 + }, + { + "auxiliary_loss_clip": 0.01057005, + "auxiliary_loss_mlp": 0.0100203, + "balance_loss_clip": 1.0084393, + "balance_loss_mlp": 1.00044489, + "epoch": 0.6892322491432694, + "flos": 64117354337280.0, + "grad_norm": 0.7384895442856348, + "language_loss": 0.59513009, + "learning_rate": 9.302412814940488e-07, + "loss": 0.61572051, + "num_input_tokens_seen": 123333045, + "step": 5732, + "time_per_iteration": 3.3175599575042725 + }, + { + "auxiliary_loss_clip": 0.01150578, + "auxiliary_loss_mlp": 0.01027769, + "balance_loss_clip": 1.04240656, + "balance_loss_mlp": 1.01970708, + "epoch": 0.6893524920339085, + "flos": 23002436736000.0, + "grad_norm": 3.020672858075604, + "language_loss": 0.71283692, + "learning_rate": 9.295831873516276e-07, + "loss": 0.73462039, + "num_input_tokens_seen": 123352320, + "step": 5733, + "time_per_iteration": 2.895110607147217 + }, + { + "auxiliary_loss_clip": 0.01169057, + "auxiliary_loss_mlp": 0.01030918, + "balance_loss_clip": 1.04718041, + "balance_loss_mlp": 1.02253759, + "epoch": 0.6894727349245476, + "flos": 21396260177280.0, + "grad_norm": 1.5844679599198108, + "language_loss": 0.76203787, + "learning_rate": 9.289252555905873e-07, + "loss": 0.78403759, + "num_input_tokens_seen": 123372400, + "step": 5734, + "time_per_iteration": 2.8439345359802246 + }, + { + "auxiliary_loss_clip": 0.01163943, + "auxiliary_loss_mlp": 0.01032229, + "balance_loss_clip": 1.04986966, + "balance_loss_mlp": 1.02445054, + "epoch": 0.6895929778151867, + "flos": 19865316654720.0, + "grad_norm": 2.31093008465398, + "language_loss": 0.76153129, + "learning_rate": 9.282674863107334e-07, + "loss": 0.78349298, + "num_input_tokens_seen": 123390215, + "step": 5735, + "time_per_iteration": 2.758314371109009 + }, + { + "auxiliary_loss_clip": 0.01160664, + "auxiliary_loss_mlp": 0.01030824, + "balance_loss_clip": 1.04810023, + "balance_loss_mlp": 1.02326655, + "epoch": 0.6897132207058257, + "flos": 18179166464640.0, + "grad_norm": 2.1819333045391427, + "language_loss": 0.75442374, + "learning_rate": 9.276098796118488e-07, + "loss": 0.77633858, + "num_input_tokens_seen": 123406700, + "step": 5736, + "time_per_iteration": 2.673274517059326 + }, + { + "auxiliary_loss_clip": 0.01151596, + "auxiliary_loss_mlp": 0.01023744, + "balance_loss_clip": 1.04629588, + "balance_loss_mlp": 1.0159061, + "epoch": 0.6898334635964649, + "flos": 32561641359360.0, + "grad_norm": 1.82680602157723, + "language_loss": 0.661282, + "learning_rate": 9.269524355936938e-07, + "loss": 0.68303537, + "num_input_tokens_seen": 123429880, + "step": 5737, + "time_per_iteration": 2.7675585746765137 + }, + { + "auxiliary_loss_clip": 0.01146077, + "auxiliary_loss_mlp": 0.01019883, + "balance_loss_clip": 1.04296803, + "balance_loss_mlp": 1.01222086, + "epoch": 0.689953706487104, + "flos": 22819004956800.0, + "grad_norm": 1.6961099159424906, + "language_loss": 0.84927762, + "learning_rate": 9.262951543560002e-07, + "loss": 0.87093723, + "num_input_tokens_seen": 123449105, + "step": 5738, + "time_per_iteration": 2.6843740940093994 + }, + { + "auxiliary_loss_clip": 0.01151447, + "auxiliary_loss_mlp": 0.01031014, + "balance_loss_clip": 1.04794025, + "balance_loss_mlp": 1.02281237, + "epoch": 0.690073949377743, + "flos": 18515362786560.0, + "grad_norm": 2.220511821941698, + "language_loss": 0.86142921, + "learning_rate": 9.256380359984795e-07, + "loss": 0.88325381, + "num_input_tokens_seen": 123466215, + "step": 5739, + "time_per_iteration": 2.7324934005737305 + }, + { + "auxiliary_loss_clip": 0.01136521, + "auxiliary_loss_mlp": 0.01026505, + "balance_loss_clip": 1.03795886, + "balance_loss_mlp": 1.01860762, + "epoch": 0.6901941922683821, + "flos": 34857194716800.0, + "grad_norm": 1.8639722601611277, + "language_loss": 0.74547344, + "learning_rate": 9.249810806208139e-07, + "loss": 0.76710367, + "num_input_tokens_seen": 123485480, + "step": 5740, + "time_per_iteration": 2.9694864749908447 + }, + { + "auxiliary_loss_clip": 0.01128062, + "auxiliary_loss_mlp": 0.008857, + "balance_loss_clip": 1.03741384, + "balance_loss_mlp": 1.00072563, + "epoch": 0.6903144351590212, + "flos": 16253672976000.0, + "grad_norm": 1.8576860325326334, + "language_loss": 0.80407286, + "learning_rate": 9.243242883226627e-07, + "loss": 0.82421041, + "num_input_tokens_seen": 123504575, + "step": 5741, + "time_per_iteration": 2.818251848220825 + }, + { + "auxiliary_loss_clip": 0.01163996, + "auxiliary_loss_mlp": 0.01023155, + "balance_loss_clip": 1.04232073, + "balance_loss_mlp": 1.01503694, + "epoch": 0.6904346780496603, + "flos": 28035137255040.0, + "grad_norm": 3.2444784274207867, + "language_loss": 0.69662535, + "learning_rate": 9.236676592036628e-07, + "loss": 0.7184968, + "num_input_tokens_seen": 123524250, + "step": 5742, + "time_per_iteration": 2.7346510887145996 + }, + { + "auxiliary_loss_clip": 0.01151105, + "auxiliary_loss_mlp": 0.01030795, + "balance_loss_clip": 1.05031347, + "balance_loss_mlp": 1.02278447, + "epoch": 0.6905549209402994, + "flos": 23624266008960.0, + "grad_norm": 1.6796200278369338, + "language_loss": 0.73584783, + "learning_rate": 9.230111933634228e-07, + "loss": 0.75766689, + "num_input_tokens_seen": 123545845, + "step": 5743, + "time_per_iteration": 2.8441267013549805 + }, + { + "auxiliary_loss_clip": 0.01165621, + "auxiliary_loss_mlp": 0.01031058, + "balance_loss_clip": 1.04883075, + "balance_loss_mlp": 1.02292442, + "epoch": 0.6906751638309385, + "flos": 23114945111040.0, + "grad_norm": 1.7745680047787513, + "language_loss": 0.80858278, + "learning_rate": 9.223548909015288e-07, + "loss": 0.8305496, + "num_input_tokens_seen": 123567535, + "step": 5744, + "time_per_iteration": 2.7357571125030518 + }, + { + "auxiliary_loss_clip": 0.01126514, + "auxiliary_loss_mlp": 0.01023728, + "balance_loss_clip": 1.04146099, + "balance_loss_mlp": 1.01617277, + "epoch": 0.6907954067215776, + "flos": 27305468375040.0, + "grad_norm": 2.2095824596106683, + "language_loss": 0.71994603, + "learning_rate": 9.216987519175407e-07, + "loss": 0.74144846, + "num_input_tokens_seen": 123587710, + "step": 5745, + "time_per_iteration": 2.992863655090332 + }, + { + "auxiliary_loss_clip": 0.01154677, + "auxiliary_loss_mlp": 0.01023973, + "balance_loss_clip": 1.04455233, + "balance_loss_mlp": 1.01645708, + "epoch": 0.6909156496122166, + "flos": 21689399070720.0, + "grad_norm": 1.723542153990597, + "language_loss": 0.68974936, + "learning_rate": 9.210427765109942e-07, + "loss": 0.71153587, + "num_input_tokens_seen": 123607385, + "step": 5746, + "time_per_iteration": 2.7165322303771973 + }, + { + "auxiliary_loss_clip": 0.01155179, + "auxiliary_loss_mlp": 0.01034246, + "balance_loss_clip": 1.04479456, + "balance_loss_mlp": 1.02593148, + "epoch": 0.6910358925028558, + "flos": 22561453463040.0, + "grad_norm": 1.940782336799389, + "language_loss": 0.81292319, + "learning_rate": 9.20386964781402e-07, + "loss": 0.83481747, + "num_input_tokens_seen": 123625405, + "step": 5747, + "time_per_iteration": 2.9129698276519775 + }, + { + "auxiliary_loss_clip": 0.01149782, + "auxiliary_loss_mlp": 0.01027025, + "balance_loss_clip": 1.04369521, + "balance_loss_mlp": 1.01903749, + "epoch": 0.6911561353934949, + "flos": 22054107813120.0, + "grad_norm": 2.0244983801975596, + "language_loss": 0.84246421, + "learning_rate": 9.197313168282472e-07, + "loss": 0.86423224, + "num_input_tokens_seen": 123642850, + "step": 5748, + "time_per_iteration": 2.780673027038574 + }, + { + "auxiliary_loss_clip": 0.01154318, + "auxiliary_loss_mlp": 0.01030342, + "balance_loss_clip": 1.04227114, + "balance_loss_mlp": 1.02203286, + "epoch": 0.6912763782841339, + "flos": 24206557386240.0, + "grad_norm": 2.0238770577604743, + "language_loss": 0.72155452, + "learning_rate": 9.190758327509935e-07, + "loss": 0.74340111, + "num_input_tokens_seen": 123661595, + "step": 5749, + "time_per_iteration": 2.8233633041381836 + }, + { + "auxiliary_loss_clip": 0.01033593, + "auxiliary_loss_mlp": 0.00876757, + "balance_loss_clip": 1.0086596, + "balance_loss_mlp": 1.00116479, + "epoch": 0.6913966211747731, + "flos": 52329641091840.0, + "grad_norm": 0.9443456610692813, + "language_loss": 0.6445775, + "learning_rate": 9.184205126490767e-07, + "loss": 0.66368103, + "num_input_tokens_seen": 123710490, + "step": 5750, + "time_per_iteration": 4.2185986042022705 + }, + { + "auxiliary_loss_clip": 0.01047447, + "auxiliary_loss_mlp": 0.00877236, + "balance_loss_clip": 1.01069522, + "balance_loss_mlp": 1.00142181, + "epoch": 0.6915168640654121, + "flos": 66741274851840.0, + "grad_norm": 1.20117717286339, + "language_loss": 0.59664768, + "learning_rate": 9.177653566219075e-07, + "loss": 0.6158945, + "num_input_tokens_seen": 123765215, + "step": 5751, + "time_per_iteration": 4.2461159229278564 + }, + { + "auxiliary_loss_clip": 0.01145609, + "auxiliary_loss_mlp": 0.01025359, + "balance_loss_clip": 1.04186261, + "balance_loss_mlp": 1.01745534, + "epoch": 0.6916371069560512, + "flos": 18296523175680.0, + "grad_norm": 2.135988090048372, + "language_loss": 0.76202703, + "learning_rate": 9.171103647688744e-07, + "loss": 0.78373677, + "num_input_tokens_seen": 123783955, + "step": 5752, + "time_per_iteration": 2.8424668312072754 + }, + { + "auxiliary_loss_clip": 0.01105516, + "auxiliary_loss_mlp": 0.01025049, + "balance_loss_clip": 1.03763485, + "balance_loss_mlp": 1.01743126, + "epoch": 0.6917573498466904, + "flos": 19645794685440.0, + "grad_norm": 1.8118784523017002, + "language_loss": 0.68814135, + "learning_rate": 9.164555371893367e-07, + "loss": 0.70944703, + "num_input_tokens_seen": 123803885, + "step": 5753, + "time_per_iteration": 2.874793529510498 + }, + { + "auxiliary_loss_clip": 0.01162315, + "auxiliary_loss_mlp": 0.00885745, + "balance_loss_clip": 1.049178, + "balance_loss_mlp": 1.00074291, + "epoch": 0.6918775927373294, + "flos": 14210319985920.0, + "grad_norm": 3.5700598253221827, + "language_loss": 0.75475335, + "learning_rate": 9.158008739826333e-07, + "loss": 0.77523398, + "num_input_tokens_seen": 123821485, + "step": 5754, + "time_per_iteration": 2.730786085128784 + }, + { + "auxiliary_loss_clip": 0.01148155, + "auxiliary_loss_mlp": 0.01031566, + "balance_loss_clip": 1.04609203, + "balance_loss_mlp": 1.02334058, + "epoch": 0.6919978356279685, + "flos": 23985455218560.0, + "grad_norm": 1.6720386437903363, + "language_loss": 0.86557424, + "learning_rate": 9.151463752480744e-07, + "loss": 0.88737142, + "num_input_tokens_seen": 123840215, + "step": 5755, + "time_per_iteration": 3.6971628665924072 + }, + { + "auxiliary_loss_clip": 0.01133984, + "auxiliary_loss_mlp": 0.01029347, + "balance_loss_clip": 1.04235339, + "balance_loss_mlp": 1.02098489, + "epoch": 0.6921180785186076, + "flos": 23622937205760.0, + "grad_norm": 1.3870165959921408, + "language_loss": 0.79966205, + "learning_rate": 9.144920410849493e-07, + "loss": 0.82129538, + "num_input_tokens_seen": 123861450, + "step": 5756, + "time_per_iteration": 2.7724578380584717 + }, + { + "auxiliary_loss_clip": 0.01156187, + "auxiliary_loss_mlp": 0.01027259, + "balance_loss_clip": 1.04541826, + "balance_loss_mlp": 1.01940942, + "epoch": 0.6922383214092467, + "flos": 21142623265920.0, + "grad_norm": 1.6931551925606272, + "language_loss": 0.80521464, + "learning_rate": 9.138378715925176e-07, + "loss": 0.82704914, + "num_input_tokens_seen": 123880545, + "step": 5757, + "time_per_iteration": 3.6755452156066895 + }, + { + "auxiliary_loss_clip": 0.01146087, + "auxiliary_loss_mlp": 0.0102767, + "balance_loss_clip": 1.04271579, + "balance_loss_mlp": 1.01946211, + "epoch": 0.6923585642998857, + "flos": 21470667200640.0, + "grad_norm": 1.6713441871834236, + "language_loss": 0.80787933, + "learning_rate": 9.131838668700167e-07, + "loss": 0.82961684, + "num_input_tokens_seen": 123900615, + "step": 5758, + "time_per_iteration": 2.7610960006713867 + }, + { + "auxiliary_loss_clip": 0.01142325, + "auxiliary_loss_mlp": 0.01026595, + "balance_loss_clip": 1.04237747, + "balance_loss_mlp": 1.01802063, + "epoch": 0.6924788071905249, + "flos": 21105204272640.0, + "grad_norm": 1.71354896114843, + "language_loss": 0.86443138, + "learning_rate": 9.125300270166598e-07, + "loss": 0.88612056, + "num_input_tokens_seen": 123921220, + "step": 5759, + "time_per_iteration": 2.8091280460357666 + }, + { + "auxiliary_loss_clip": 0.01143366, + "auxiliary_loss_mlp": 0.01029155, + "balance_loss_clip": 1.04337788, + "balance_loss_mlp": 1.02174616, + "epoch": 0.692599050081164, + "flos": 26250018117120.0, + "grad_norm": 2.3312987451269493, + "language_loss": 0.8583529, + "learning_rate": 9.118763521316324e-07, + "loss": 0.88007808, + "num_input_tokens_seen": 123941795, + "step": 5760, + "time_per_iteration": 2.8186676502227783 + }, + { + "auxiliary_loss_clip": 0.01172017, + "auxiliary_loss_mlp": 0.0088608, + "balance_loss_clip": 1.04732931, + "balance_loss_mlp": 1.00069761, + "epoch": 0.692719292971803, + "flos": 20885215426560.0, + "grad_norm": 1.7359105863435962, + "language_loss": 0.75917882, + "learning_rate": 9.112228423140987e-07, + "loss": 0.77975976, + "num_input_tokens_seen": 123960715, + "step": 5761, + "time_per_iteration": 2.702672243118286 + }, + { + "auxiliary_loss_clip": 0.01157197, + "auxiliary_loss_mlp": 0.01024637, + "balance_loss_clip": 1.04732883, + "balance_loss_mlp": 1.01667356, + "epoch": 0.6928395358624422, + "flos": 25921938268800.0, + "grad_norm": 2.4163357528130054, + "language_loss": 0.85958284, + "learning_rate": 9.105694976631932e-07, + "loss": 0.88140118, + "num_input_tokens_seen": 123978625, + "step": 5762, + "time_per_iteration": 2.714547634124756 + }, + { + "auxiliary_loss_clip": 0.01165786, + "auxiliary_loss_mlp": 0.0102542, + "balance_loss_clip": 1.0499599, + "balance_loss_mlp": 1.01774895, + "epoch": 0.6929597787530812, + "flos": 23586559706880.0, + "grad_norm": 2.1755198595664167, + "language_loss": 0.72264367, + "learning_rate": 9.099163182780283e-07, + "loss": 0.74455571, + "num_input_tokens_seen": 123996780, + "step": 5763, + "time_per_iteration": 2.7392780780792236 + }, + { + "auxiliary_loss_clip": 0.01151621, + "auxiliary_loss_mlp": 0.01031139, + "balance_loss_clip": 1.04612482, + "balance_loss_mlp": 1.02321136, + "epoch": 0.6930800216437203, + "flos": 18255656476800.0, + "grad_norm": 6.152782673759096, + "language_loss": 0.49058479, + "learning_rate": 9.092633042576916e-07, + "loss": 0.51241243, + "num_input_tokens_seen": 124014045, + "step": 5764, + "time_per_iteration": 2.719339609146118 + }, + { + "auxiliary_loss_clip": 0.0114949, + "auxiliary_loss_mlp": 0.01031618, + "balance_loss_clip": 1.0465486, + "balance_loss_mlp": 1.02358913, + "epoch": 0.6932002645343595, + "flos": 29168621809920.0, + "grad_norm": 2.722443035925104, + "language_loss": 0.56655157, + "learning_rate": 9.086104557012446e-07, + "loss": 0.58836257, + "num_input_tokens_seen": 124034615, + "step": 5765, + "time_per_iteration": 2.81429123878479 + }, + { + "auxiliary_loss_clip": 0.0115357, + "auxiliary_loss_mlp": 0.01026565, + "balance_loss_clip": 1.04490507, + "balance_loss_mlp": 1.01863217, + "epoch": 0.6933205074249985, + "flos": 23842746483840.0, + "grad_norm": 1.9815202381786468, + "language_loss": 0.65850621, + "learning_rate": 9.079577727077239e-07, + "loss": 0.68030763, + "num_input_tokens_seen": 124053445, + "step": 5766, + "time_per_iteration": 2.7008607387542725 + }, + { + "auxiliary_loss_clip": 0.01163466, + "auxiliary_loss_mlp": 0.01026701, + "balance_loss_clip": 1.0484699, + "balance_loss_mlp": 1.01815319, + "epoch": 0.6934407503156376, + "flos": 24166696268160.0, + "grad_norm": 7.4555569655974905, + "language_loss": 0.72077686, + "learning_rate": 9.073052553761404e-07, + "loss": 0.74267852, + "num_input_tokens_seen": 124072810, + "step": 5767, + "time_per_iteration": 2.7147560119628906 + }, + { + "auxiliary_loss_clip": 0.01137525, + "auxiliary_loss_mlp": 0.01033114, + "balance_loss_clip": 1.04341054, + "balance_loss_mlp": 1.02401781, + "epoch": 0.6935609932062767, + "flos": 20631327120000.0, + "grad_norm": 1.7686761310333632, + "language_loss": 0.78057528, + "learning_rate": 9.066529038054805e-07, + "loss": 0.80228168, + "num_input_tokens_seen": 124092875, + "step": 5768, + "time_per_iteration": 2.837407350540161 + }, + { + "auxiliary_loss_clip": 0.01151478, + "auxiliary_loss_mlp": 0.01027022, + "balance_loss_clip": 1.04546762, + "balance_loss_mlp": 1.01945245, + "epoch": 0.6936812360969158, + "flos": 18254184019200.0, + "grad_norm": 2.9950415727729234, + "language_loss": 0.74223745, + "learning_rate": 9.060007180947071e-07, + "loss": 0.76402247, + "num_input_tokens_seen": 124110930, + "step": 5769, + "time_per_iteration": 2.810720205307007 + }, + { + "auxiliary_loss_clip": 0.0112856, + "auxiliary_loss_mlp": 0.01033922, + "balance_loss_clip": 1.04154086, + "balance_loss_mlp": 1.02545178, + "epoch": 0.6938014789875548, + "flos": 31317336368640.0, + "grad_norm": 2.091092042762548, + "language_loss": 0.73170269, + "learning_rate": 9.053486983427534e-07, + "loss": 0.75332749, + "num_input_tokens_seen": 124132180, + "step": 5770, + "time_per_iteration": 2.9078853130340576 + }, + { + "auxiliary_loss_clip": 0.01154787, + "auxiliary_loss_mlp": 0.01026534, + "balance_loss_clip": 1.04407513, + "balance_loss_mlp": 1.01840365, + "epoch": 0.6939217218781939, + "flos": 17528429721600.0, + "grad_norm": 1.8471851437792286, + "language_loss": 0.70339906, + "learning_rate": 9.046968446485326e-07, + "loss": 0.72521222, + "num_input_tokens_seen": 124150585, + "step": 5771, + "time_per_iteration": 2.694159984588623 + }, + { + "auxiliary_loss_clip": 0.01166164, + "auxiliary_loss_mlp": 0.01029623, + "balance_loss_clip": 1.04864001, + "balance_loss_mlp": 1.02075946, + "epoch": 0.6940419647688331, + "flos": 18551776199040.0, + "grad_norm": 3.479729824915304, + "language_loss": 0.70752215, + "learning_rate": 9.040451571109295e-07, + "loss": 0.72948003, + "num_input_tokens_seen": 124166205, + "step": 5772, + "time_per_iteration": 2.668020248413086 + }, + { + "auxiliary_loss_clip": 0.01039957, + "auxiliary_loss_mlp": 0.01002084, + "balance_loss_clip": 1.00838327, + "balance_loss_mlp": 1.00035524, + "epoch": 0.6941622076594721, + "flos": 66926286829440.0, + "grad_norm": 0.8293230569901648, + "language_loss": 0.6037792, + "learning_rate": 9.033936358288042e-07, + "loss": 0.62419963, + "num_input_tokens_seen": 124219940, + "step": 5773, + "time_per_iteration": 3.1677403450012207 + }, + { + "auxiliary_loss_clip": 0.01173549, + "auxiliary_loss_mlp": 0.0103091, + "balance_loss_clip": 1.04856002, + "balance_loss_mlp": 1.02248216, + "epoch": 0.6942824505501112, + "flos": 26578062051840.0, + "grad_norm": 1.8430105809496478, + "language_loss": 0.82374442, + "learning_rate": 9.027422809009937e-07, + "loss": 0.84578907, + "num_input_tokens_seen": 124239885, + "step": 5774, + "time_per_iteration": 2.6813900470733643 + }, + { + "auxiliary_loss_clip": 0.01163542, + "auxiliary_loss_mlp": 0.01026421, + "balance_loss_clip": 1.04498446, + "balance_loss_mlp": 1.01851106, + "epoch": 0.6944026934407503, + "flos": 21248308056960.0, + "grad_norm": 1.6719187279083871, + "language_loss": 0.83442879, + "learning_rate": 9.020910924263054e-07, + "loss": 0.85632843, + "num_input_tokens_seen": 124258410, + "step": 5775, + "time_per_iteration": 2.687211036682129 + }, + { + "auxiliary_loss_clip": 0.01040105, + "auxiliary_loss_mlp": 0.01006545, + "balance_loss_clip": 1.01051712, + "balance_loss_mlp": 1.00491166, + "epoch": 0.6945229363313894, + "flos": 70677191537280.0, + "grad_norm": 1.6245408440599, + "language_loss": 0.58160657, + "learning_rate": 9.014400705035261e-07, + "loss": 0.60207301, + "num_input_tokens_seen": 124315315, + "step": 5776, + "time_per_iteration": 4.259497880935669 + }, + { + "auxiliary_loss_clip": 0.01171772, + "auxiliary_loss_mlp": 0.01024173, + "balance_loss_clip": 1.05096018, + "balance_loss_mlp": 1.01620436, + "epoch": 0.6946431792220285, + "flos": 18952934267520.0, + "grad_norm": 1.9039478283763696, + "language_loss": 0.76438653, + "learning_rate": 9.00789215231414e-07, + "loss": 0.78634608, + "num_input_tokens_seen": 124333710, + "step": 5777, + "time_per_iteration": 3.600471019744873 + }, + { + "auxiliary_loss_clip": 0.01142353, + "auxiliary_loss_mlp": 0.00887181, + "balance_loss_clip": 1.04165721, + "balance_loss_mlp": 1.00073469, + "epoch": 0.6947634221126676, + "flos": 20338834671360.0, + "grad_norm": 1.679543104969234, + "language_loss": 0.82084286, + "learning_rate": 9.001385267087056e-07, + "loss": 0.84113812, + "num_input_tokens_seen": 124352855, + "step": 5778, + "time_per_iteration": 2.8828823566436768 + }, + { + "auxiliary_loss_clip": 0.01166524, + "auxiliary_loss_mlp": 0.01025856, + "balance_loss_clip": 1.04901111, + "balance_loss_mlp": 1.01766026, + "epoch": 0.6948836650033067, + "flos": 21833723917440.0, + "grad_norm": 1.510289370047414, + "language_loss": 0.70406407, + "learning_rate": 8.994880050341072e-07, + "loss": 0.72598785, + "num_input_tokens_seen": 124372960, + "step": 5779, + "time_per_iteration": 2.8169450759887695 + }, + { + "auxiliary_loss_clip": 0.01145545, + "auxiliary_loss_mlp": 0.01035069, + "balance_loss_clip": 1.04444611, + "balance_loss_mlp": 1.02637887, + "epoch": 0.6950039078939457, + "flos": 23657519024640.0, + "grad_norm": 1.6809721621220357, + "language_loss": 0.77388567, + "learning_rate": 8.988376503063026e-07, + "loss": 0.79569179, + "num_input_tokens_seen": 124394220, + "step": 5780, + "time_per_iteration": 3.700092077255249 + }, + { + "auxiliary_loss_clip": 0.011411, + "auxiliary_loss_mlp": 0.01033122, + "balance_loss_clip": 1.04353333, + "balance_loss_mlp": 1.02475333, + "epoch": 0.6951241507845849, + "flos": 21792462168960.0, + "grad_norm": 1.93748188880113, + "language_loss": 0.8143962, + "learning_rate": 8.981874626239521e-07, + "loss": 0.83613843, + "num_input_tokens_seen": 124412795, + "step": 5781, + "time_per_iteration": 2.801206588745117 + }, + { + "auxiliary_loss_clip": 0.01164956, + "auxiliary_loss_mlp": 0.01029836, + "balance_loss_clip": 1.05041504, + "balance_loss_mlp": 1.0218308, + "epoch": 0.695244393675224, + "flos": 14647568244480.0, + "grad_norm": 2.0015850557748824, + "language_loss": 0.88453424, + "learning_rate": 8.975374420856872e-07, + "loss": 0.90648222, + "num_input_tokens_seen": 124429690, + "step": 5782, + "time_per_iteration": 2.772038221359253 + }, + { + "auxiliary_loss_clip": 0.01133926, + "auxiliary_loss_mlp": 0.01028785, + "balance_loss_clip": 1.0414412, + "balance_loss_mlp": 1.02114916, + "epoch": 0.695364636565863, + "flos": 16873203778560.0, + "grad_norm": 2.0266541522471826, + "language_loss": 0.72602975, + "learning_rate": 8.968875887901157e-07, + "loss": 0.74765688, + "num_input_tokens_seen": 124447070, + "step": 5783, + "time_per_iteration": 3.8048532009124756 + }, + { + "auxiliary_loss_clip": 0.01153081, + "auxiliary_loss_mlp": 0.01029416, + "balance_loss_clip": 1.04432774, + "balance_loss_mlp": 1.02112508, + "epoch": 0.6954848794565022, + "flos": 19354523299200.0, + "grad_norm": 2.097820031270314, + "language_loss": 0.62719351, + "learning_rate": 8.9623790283582e-07, + "loss": 0.64901853, + "num_input_tokens_seen": 124464950, + "step": 5784, + "time_per_iteration": 2.7533342838287354 + }, + { + "auxiliary_loss_clip": 0.01144556, + "auxiliary_loss_mlp": 0.01031511, + "balance_loss_clip": 1.04477501, + "balance_loss_mlp": 1.02321362, + "epoch": 0.6956051223471412, + "flos": 18990209606400.0, + "grad_norm": 2.240745599921708, + "language_loss": 0.76077747, + "learning_rate": 8.955883843213561e-07, + "loss": 0.78253818, + "num_input_tokens_seen": 124483965, + "step": 5785, + "time_per_iteration": 2.7736194133758545 + }, + { + "auxiliary_loss_clip": 0.01169861, + "auxiliary_loss_mlp": 0.01033694, + "balance_loss_clip": 1.04843521, + "balance_loss_mlp": 1.02508664, + "epoch": 0.6957253652377803, + "flos": 16107229226880.0, + "grad_norm": 2.1066330017828014, + "language_loss": 0.86383307, + "learning_rate": 8.949390333452569e-07, + "loss": 0.88586855, + "num_input_tokens_seen": 124501910, + "step": 5786, + "time_per_iteration": 2.8677284717559814 + }, + { + "auxiliary_loss_clip": 0.01172121, + "auxiliary_loss_mlp": 0.01026639, + "balance_loss_clip": 1.04961026, + "balance_loss_mlp": 1.01876545, + "epoch": 0.6958456081284194, + "flos": 29388646569600.0, + "grad_norm": 1.9313632033866046, + "language_loss": 0.67616171, + "learning_rate": 8.942898500060279e-07, + "loss": 0.69814932, + "num_input_tokens_seen": 124521625, + "step": 5787, + "time_per_iteration": 2.7624964714050293 + }, + { + "auxiliary_loss_clip": 0.01136845, + "auxiliary_loss_mlp": 0.01030217, + "balance_loss_clip": 1.04621851, + "balance_loss_mlp": 1.021348, + "epoch": 0.6959658510190585, + "flos": 25154850395520.0, + "grad_norm": 5.596599156430686, + "language_loss": 0.71143138, + "learning_rate": 8.936408344021493e-07, + "loss": 0.73310196, + "num_input_tokens_seen": 124538540, + "step": 5788, + "time_per_iteration": 2.829779863357544 + }, + { + "auxiliary_loss_clip": 0.01155375, + "auxiliary_loss_mlp": 0.01036686, + "balance_loss_clip": 1.04841328, + "balance_loss_mlp": 1.02826357, + "epoch": 0.6960860939096976, + "flos": 42814388759040.0, + "grad_norm": 2.501823054621285, + "language_loss": 0.71292925, + "learning_rate": 8.929919866320765e-07, + "loss": 0.73484981, + "num_input_tokens_seen": 124559355, + "step": 5789, + "time_per_iteration": 2.9532885551452637 + }, + { + "auxiliary_loss_clip": 0.01147286, + "auxiliary_loss_mlp": 0.0088659, + "balance_loss_clip": 1.04315329, + "balance_loss_mlp": 1.00068927, + "epoch": 0.6962063368003367, + "flos": 17566566986880.0, + "grad_norm": 2.368198379815106, + "language_loss": 0.80984521, + "learning_rate": 8.923433067942385e-07, + "loss": 0.83018398, + "num_input_tokens_seen": 124577920, + "step": 5790, + "time_per_iteration": 2.8464415073394775 + }, + { + "auxiliary_loss_clip": 0.01145799, + "auxiliary_loss_mlp": 0.01032397, + "balance_loss_clip": 1.04694104, + "balance_loss_mlp": 1.02441001, + "epoch": 0.6963265796909758, + "flos": 21251648021760.0, + "grad_norm": 2.6174536658356073, + "language_loss": 0.68584692, + "learning_rate": 8.916947949870417e-07, + "loss": 0.70762891, + "num_input_tokens_seen": 124597585, + "step": 5791, + "time_per_iteration": 3.0080766677856445 + }, + { + "auxiliary_loss_clip": 0.01056546, + "auxiliary_loss_mlp": 0.01002048, + "balance_loss_clip": 1.007967, + "balance_loss_mlp": 1.00039124, + "epoch": 0.6964468225816148, + "flos": 68828295801600.0, + "grad_norm": 0.7392796778483126, + "language_loss": 0.58110023, + "learning_rate": 8.910464513088615e-07, + "loss": 0.60168612, + "num_input_tokens_seen": 124661625, + "step": 5792, + "time_per_iteration": 3.367567777633667 + }, + { + "auxiliary_loss_clip": 0.01145242, + "auxiliary_loss_mlp": 0.0102803, + "balance_loss_clip": 1.04366422, + "balance_loss_mlp": 1.01947057, + "epoch": 0.696567065472254, + "flos": 18950887192320.0, + "grad_norm": 1.8314293825246273, + "language_loss": 0.78373981, + "learning_rate": 8.903982758580542e-07, + "loss": 0.80547249, + "num_input_tokens_seen": 124680565, + "step": 5793, + "time_per_iteration": 2.7682323455810547 + }, + { + "auxiliary_loss_clip": 0.0115078, + "auxiliary_loss_mlp": 0.01026777, + "balance_loss_clip": 1.04694974, + "balance_loss_mlp": 1.01841486, + "epoch": 0.696687308362893, + "flos": 22856675345280.0, + "grad_norm": 2.0010248944908686, + "language_loss": 0.80587912, + "learning_rate": 8.897502687329457e-07, + "loss": 0.82765472, + "num_input_tokens_seen": 124700365, + "step": 5794, + "time_per_iteration": 2.787348985671997 + }, + { + "auxiliary_loss_clip": 0.01140572, + "auxiliary_loss_mlp": 0.01025833, + "balance_loss_clip": 1.04451895, + "balance_loss_mlp": 1.01734579, + "epoch": 0.6968075512535321, + "flos": 24972926987520.0, + "grad_norm": 1.9175363966352974, + "language_loss": 0.80120975, + "learning_rate": 8.891024300318382e-07, + "loss": 0.82287383, + "num_input_tokens_seen": 124718935, + "step": 5795, + "time_per_iteration": 2.894641637802124 + }, + { + "auxiliary_loss_clip": 0.01133199, + "auxiliary_loss_mlp": 0.01029123, + "balance_loss_clip": 1.04136252, + "balance_loss_mlp": 1.02160096, + "epoch": 0.6969277941441713, + "flos": 21030438113280.0, + "grad_norm": 1.459244074392914, + "language_loss": 0.75738716, + "learning_rate": 8.884547598530103e-07, + "loss": 0.77901042, + "num_input_tokens_seen": 124739505, + "step": 5796, + "time_per_iteration": 2.847702741622925 + }, + { + "auxiliary_loss_clip": 0.01106443, + "auxiliary_loss_mlp": 0.01030491, + "balance_loss_clip": 1.03933859, + "balance_loss_mlp": 1.02266526, + "epoch": 0.6970480370348103, + "flos": 21579404647680.0, + "grad_norm": 2.2908991264825005, + "language_loss": 0.75290549, + "learning_rate": 8.8780725829471e-07, + "loss": 0.77427483, + "num_input_tokens_seen": 124757410, + "step": 5797, + "time_per_iteration": 2.9821619987487793 + }, + { + "auxiliary_loss_clip": 0.01172768, + "auxiliary_loss_mlp": 0.01031211, + "balance_loss_clip": 1.0487175, + "balance_loss_mlp": 1.02244878, + "epoch": 0.6971682799254494, + "flos": 22419175691520.0, + "grad_norm": 3.5992574755016804, + "language_loss": 0.77866399, + "learning_rate": 8.87159925455165e-07, + "loss": 0.80070382, + "num_input_tokens_seen": 124777240, + "step": 5798, + "time_per_iteration": 2.6499319076538086 + }, + { + "auxiliary_loss_clip": 0.01140639, + "auxiliary_loss_mlp": 0.01026152, + "balance_loss_clip": 1.04415369, + "balance_loss_mlp": 1.01777124, + "epoch": 0.6972885228160886, + "flos": 20005834659840.0, + "grad_norm": 1.8952021226108975, + "language_loss": 0.73237872, + "learning_rate": 8.865127614325738e-07, + "loss": 0.75404656, + "num_input_tokens_seen": 124795670, + "step": 5799, + "time_per_iteration": 2.8068807125091553 + }, + { + "auxiliary_loss_clip": 0.01147548, + "auxiliary_loss_mlp": 0.01039427, + "balance_loss_clip": 1.0451045, + "balance_loss_mlp": 1.03080261, + "epoch": 0.6974087657067276, + "flos": 37853437656960.0, + "grad_norm": 3.20785582767562, + "language_loss": 0.66845798, + "learning_rate": 8.85865766325113e-07, + "loss": 0.69032782, + "num_input_tokens_seen": 124819600, + "step": 5800, + "time_per_iteration": 2.8456664085388184 + }, + { + "auxiliary_loss_clip": 0.01150516, + "auxiliary_loss_mlp": 0.0102571, + "balance_loss_clip": 1.04528689, + "balance_loss_mlp": 1.01727045, + "epoch": 0.6975290085973667, + "flos": 29489267543040.0, + "grad_norm": 2.3859541135878386, + "language_loss": 0.72263587, + "learning_rate": 8.852189402309287e-07, + "loss": 0.74439818, + "num_input_tokens_seen": 124838785, + "step": 5801, + "time_per_iteration": 2.7616612911224365 + }, + { + "auxiliary_loss_clip": 0.01163951, + "auxiliary_loss_mlp": 0.01031878, + "balance_loss_clip": 1.04802239, + "balance_loss_mlp": 1.0229311, + "epoch": 0.6976492514880057, + "flos": 12895630295040.0, + "grad_norm": 3.030470096754316, + "language_loss": 0.74523234, + "learning_rate": 8.845722832481441e-07, + "loss": 0.76719064, + "num_input_tokens_seen": 124854215, + "step": 5802, + "time_per_iteration": 3.5119130611419678 + }, + { + "auxiliary_loss_clip": 0.01161523, + "auxiliary_loss_mlp": 0.01029759, + "balance_loss_clip": 1.04608738, + "balance_loss_mlp": 1.02164662, + "epoch": 0.6977694943786449, + "flos": 24352929308160.0, + "grad_norm": 1.860592424707388, + "language_loss": 0.77582955, + "learning_rate": 8.83925795474858e-07, + "loss": 0.79774237, + "num_input_tokens_seen": 124874340, + "step": 5803, + "time_per_iteration": 3.6471304893493652 + }, + { + "auxiliary_loss_clip": 0.01144438, + "auxiliary_loss_mlp": 0.01030298, + "balance_loss_clip": 1.04754555, + "balance_loss_mlp": 1.02175021, + "epoch": 0.6978897372692839, + "flos": 29898470257920.0, + "grad_norm": 3.2427557481315814, + "language_loss": 0.58864236, + "learning_rate": 8.832794770091414e-07, + "loss": 0.61038971, + "num_input_tokens_seen": 124895175, + "step": 5804, + "time_per_iteration": 2.831669330596924 + }, + { + "auxiliary_loss_clip": 0.01156868, + "auxiliary_loss_mlp": 0.01028931, + "balance_loss_clip": 1.04438806, + "balance_loss_mlp": 1.02081907, + "epoch": 0.698009980159923, + "flos": 21761579450880.0, + "grad_norm": 1.948610562919791, + "language_loss": 0.82453299, + "learning_rate": 8.826333279490401e-07, + "loss": 0.84639096, + "num_input_tokens_seen": 124915810, + "step": 5805, + "time_per_iteration": 2.7294435501098633 + }, + { + "auxiliary_loss_clip": 0.01156381, + "auxiliary_loss_mlp": 0.01025649, + "balance_loss_clip": 1.04761243, + "balance_loss_mlp": 1.01793647, + "epoch": 0.6981302230505622, + "flos": 19857164267520.0, + "grad_norm": 2.0271413994221796, + "language_loss": 0.68242693, + "learning_rate": 8.819873483925748e-07, + "loss": 0.70424724, + "num_input_tokens_seen": 124932930, + "step": 5806, + "time_per_iteration": 3.6291935443878174 + }, + { + "auxiliary_loss_clip": 0.01150485, + "auxiliary_loss_mlp": 0.00886255, + "balance_loss_clip": 1.04750907, + "balance_loss_mlp": 1.00064671, + "epoch": 0.6982504659412012, + "flos": 22198648141440.0, + "grad_norm": 2.2012345266811826, + "language_loss": 0.74448764, + "learning_rate": 8.81341538437739e-07, + "loss": 0.76485503, + "num_input_tokens_seen": 124951220, + "step": 5807, + "time_per_iteration": 2.8215279579162598 + }, + { + "auxiliary_loss_clip": 0.01159183, + "auxiliary_loss_mlp": 0.01027615, + "balance_loss_clip": 1.04549503, + "balance_loss_mlp": 1.01962185, + "epoch": 0.6983707088318403, + "flos": 35588479708800.0, + "grad_norm": 1.700499640021266, + "language_loss": 0.68210542, + "learning_rate": 8.80695898182503e-07, + "loss": 0.70397347, + "num_input_tokens_seen": 124972200, + "step": 5808, + "time_per_iteration": 2.8691768646240234 + }, + { + "auxiliary_loss_clip": 0.01049805, + "auxiliary_loss_mlp": 0.01005043, + "balance_loss_clip": 1.00963473, + "balance_loss_mlp": 1.00344551, + "epoch": 0.6984909517224794, + "flos": 65440052760960.0, + "grad_norm": 0.8207046143643415, + "language_loss": 0.65040052, + "learning_rate": 8.800504277248093e-07, + "loss": 0.67094898, + "num_input_tokens_seen": 125036950, + "step": 5809, + "time_per_iteration": 4.216948986053467 + }, + { + "auxiliary_loss_clip": 0.01139669, + "auxiliary_loss_mlp": 0.00885429, + "balance_loss_clip": 1.05013657, + "balance_loss_mlp": 1.00070405, + "epoch": 0.6986111946131185, + "flos": 18546927863040.0, + "grad_norm": 1.7971987208094056, + "language_loss": 0.75161755, + "learning_rate": 8.794051271625753e-07, + "loss": 0.77186853, + "num_input_tokens_seen": 125054585, + "step": 5810, + "time_per_iteration": 2.8474230766296387 + }, + { + "auxiliary_loss_clip": 0.01152647, + "auxiliary_loss_mlp": 0.0102583, + "balance_loss_clip": 1.04625428, + "balance_loss_mlp": 1.01845694, + "epoch": 0.6987314375037575, + "flos": 23039173370880.0, + "grad_norm": 1.5307636831277593, + "language_loss": 0.8319875, + "learning_rate": 8.787599965936925e-07, + "loss": 0.85377228, + "num_input_tokens_seen": 125075515, + "step": 5811, + "time_per_iteration": 2.7161598205566406 + }, + { + "auxiliary_loss_clip": 0.01140602, + "auxiliary_loss_mlp": 0.01030437, + "balance_loss_clip": 1.04696417, + "balance_loss_mlp": 1.02283704, + "epoch": 0.6988516803943967, + "flos": 38400393029760.0, + "grad_norm": 1.7809072648020476, + "language_loss": 0.71931064, + "learning_rate": 8.781150361160261e-07, + "loss": 0.74102104, + "num_input_tokens_seen": 125097425, + "step": 5812, + "time_per_iteration": 2.902928590774536 + }, + { + "auxiliary_loss_clip": 0.01140305, + "auxiliary_loss_mlp": 0.01025599, + "balance_loss_clip": 1.04615927, + "balance_loss_mlp": 1.01754642, + "epoch": 0.6989719232850358, + "flos": 24096993926400.0, + "grad_norm": 1.8791887989155904, + "language_loss": 0.7355358, + "learning_rate": 8.774702458274181e-07, + "loss": 0.75719488, + "num_input_tokens_seen": 125117830, + "step": 5813, + "time_per_iteration": 2.7859623432159424 + }, + { + "auxiliary_loss_clip": 0.01161243, + "auxiliary_loss_mlp": 0.01025461, + "balance_loss_clip": 1.04710722, + "balance_loss_mlp": 1.01705062, + "epoch": 0.6990921661756748, + "flos": 14866838818560.0, + "grad_norm": 3.711048712051334, + "language_loss": 0.70677131, + "learning_rate": 8.768256258256799e-07, + "loss": 0.72863835, + "num_input_tokens_seen": 125134455, + "step": 5814, + "time_per_iteration": 2.6200385093688965 + }, + { + "auxiliary_loss_clip": 0.01165691, + "auxiliary_loss_mlp": 0.01029063, + "balance_loss_clip": 1.04801965, + "balance_loss_mlp": 1.02109957, + "epoch": 0.699212409066314, + "flos": 20193719725440.0, + "grad_norm": 1.8632785679953174, + "language_loss": 0.73437929, + "learning_rate": 8.76181176208602e-07, + "loss": 0.75632679, + "num_input_tokens_seen": 125152555, + "step": 5815, + "time_per_iteration": 2.769343614578247 + }, + { + "auxiliary_loss_clip": 0.01120391, + "auxiliary_loss_mlp": 0.01029594, + "balance_loss_clip": 1.03878784, + "balance_loss_mlp": 1.02092123, + "epoch": 0.699332651956953, + "flos": 19427888828160.0, + "grad_norm": 1.733994249149162, + "language_loss": 0.73387372, + "learning_rate": 8.755368970739461e-07, + "loss": 0.7553736, + "num_input_tokens_seen": 125171915, + "step": 5816, + "time_per_iteration": 2.858660936355591 + }, + { + "auxiliary_loss_clip": 0.01149756, + "auxiliary_loss_mlp": 0.01030313, + "balance_loss_clip": 1.04338288, + "balance_loss_mlp": 1.02172947, + "epoch": 0.6994528948475921, + "flos": 16143714466560.0, + "grad_norm": 7.4518557417408315, + "language_loss": 0.61647105, + "learning_rate": 8.748927885194479e-07, + "loss": 0.63827175, + "num_input_tokens_seen": 125190220, + "step": 5817, + "time_per_iteration": 2.963751792907715 + }, + { + "auxiliary_loss_clip": 0.0104265, + "auxiliary_loss_mlp": 0.01001303, + "balance_loss_clip": 1.0122509, + "balance_loss_mlp": 0.99956274, + "epoch": 0.6995731377382313, + "flos": 64952420699520.0, + "grad_norm": 0.7913188465000518, + "language_loss": 0.57377946, + "learning_rate": 8.742488506428209e-07, + "loss": 0.59421903, + "num_input_tokens_seen": 125249310, + "step": 5818, + "time_per_iteration": 3.3112218379974365 + }, + { + "auxiliary_loss_clip": 0.01156019, + "auxiliary_loss_mlp": 0.0088563, + "balance_loss_clip": 1.04545915, + "balance_loss_mlp": 1.00074232, + "epoch": 0.6996933806288703, + "flos": 24900136076160.0, + "grad_norm": 1.8803172432400481, + "language_loss": 0.78091282, + "learning_rate": 8.736050835417466e-07, + "loss": 0.80132926, + "num_input_tokens_seen": 125269350, + "step": 5819, + "time_per_iteration": 2.796117067337036 + }, + { + "auxiliary_loss_clip": 0.01166396, + "auxiliary_loss_mlp": 0.01035976, + "balance_loss_clip": 1.04852283, + "balance_loss_mlp": 1.02798891, + "epoch": 0.6998136235195094, + "flos": 20777806782720.0, + "grad_norm": 1.9552005038711593, + "language_loss": 0.61126512, + "learning_rate": 8.729614873138862e-07, + "loss": 0.63328886, + "num_input_tokens_seen": 125286985, + "step": 5820, + "time_per_iteration": 2.944317579269409 + }, + { + "auxiliary_loss_clip": 0.01142795, + "auxiliary_loss_mlp": 0.01031568, + "balance_loss_clip": 1.04854953, + "balance_loss_mlp": 1.02249622, + "epoch": 0.6999338664101485, + "flos": 23733470332800.0, + "grad_norm": 2.319349035626825, + "language_loss": 0.77406478, + "learning_rate": 8.723180620568716e-07, + "loss": 0.79580843, + "num_input_tokens_seen": 125306240, + "step": 5821, + "time_per_iteration": 2.916242837905884 + }, + { + "auxiliary_loss_clip": 0.0114559, + "auxiliary_loss_mlp": 0.01028719, + "balance_loss_clip": 1.0433135, + "balance_loss_mlp": 1.02151835, + "epoch": 0.7000541093007876, + "flos": 19864598382720.0, + "grad_norm": 2.271322564848437, + "language_loss": 0.85216016, + "learning_rate": 8.716748078683116e-07, + "loss": 0.87390327, + "num_input_tokens_seen": 125323015, + "step": 5822, + "time_per_iteration": 2.771364212036133 + }, + { + "auxiliary_loss_clip": 0.01105418, + "auxiliary_loss_mlp": 0.01028031, + "balance_loss_clip": 1.03930056, + "balance_loss_mlp": 1.01885176, + "epoch": 0.7001743521914267, + "flos": 29679056029440.0, + "grad_norm": 2.1285063710964067, + "language_loss": 0.68506122, + "learning_rate": 8.710317248457855e-07, + "loss": 0.70639575, + "num_input_tokens_seen": 125342630, + "step": 5823, + "time_per_iteration": 2.950935125350952 + }, + { + "auxiliary_loss_clip": 0.01150765, + "auxiliary_loss_mlp": 0.01034798, + "balance_loss_clip": 1.04758561, + "balance_loss_mlp": 1.02746403, + "epoch": 0.7002945950820658, + "flos": 27489762080640.0, + "grad_norm": 1.772972403466729, + "language_loss": 0.71981215, + "learning_rate": 8.703888130868482e-07, + "loss": 0.74166787, + "num_input_tokens_seen": 125364480, + "step": 5824, + "time_per_iteration": 2.8356199264526367 + }, + { + "auxiliary_loss_clip": 0.01141055, + "auxiliary_loss_mlp": 0.01022582, + "balance_loss_clip": 1.04404354, + "balance_loss_mlp": 1.01436257, + "epoch": 0.7004148379727049, + "flos": 22158463800960.0, + "grad_norm": 2.0828887350365277, + "language_loss": 0.81885666, + "learning_rate": 8.697460726890307e-07, + "loss": 0.84049296, + "num_input_tokens_seen": 125381625, + "step": 5825, + "time_per_iteration": 2.7663843631744385 + }, + { + "auxiliary_loss_clip": 0.01141481, + "auxiliary_loss_mlp": 0.00886531, + "balance_loss_clip": 1.04150295, + "balance_loss_mlp": 1.00070953, + "epoch": 0.7005350808633439, + "flos": 19423758764160.0, + "grad_norm": 2.994481694178624, + "language_loss": 0.90393907, + "learning_rate": 8.691035037498354e-07, + "loss": 0.92421919, + "num_input_tokens_seen": 125397615, + "step": 5826, + "time_per_iteration": 2.9200074672698975 + }, + { + "auxiliary_loss_clip": 0.01148439, + "auxiliary_loss_mlp": 0.01027993, + "balance_loss_clip": 1.04282463, + "balance_loss_mlp": 1.02014923, + "epoch": 0.7006553237539831, + "flos": 23476708938240.0, + "grad_norm": 1.6416389593662306, + "language_loss": 0.72396314, + "learning_rate": 8.684611063667391e-07, + "loss": 0.74572742, + "num_input_tokens_seen": 125418080, + "step": 5827, + "time_per_iteration": 2.86930251121521 + }, + { + "auxiliary_loss_clip": 0.01159905, + "auxiliary_loss_mlp": 0.01026742, + "balance_loss_clip": 1.04514277, + "balance_loss_mlp": 1.01875544, + "epoch": 0.7007755666446221, + "flos": 31212872640000.0, + "grad_norm": 1.9052188172061555, + "language_loss": 0.76948547, + "learning_rate": 8.678188806371935e-07, + "loss": 0.79135191, + "num_input_tokens_seen": 125440115, + "step": 5828, + "time_per_iteration": 3.7274363040924072 + }, + { + "auxiliary_loss_clip": 0.01160727, + "auxiliary_loss_mlp": 0.01028467, + "balance_loss_clip": 1.04536581, + "balance_loss_mlp": 1.02090907, + "epoch": 0.7008958095352612, + "flos": 18149899858560.0, + "grad_norm": 1.8267910773492049, + "language_loss": 0.85337806, + "learning_rate": 8.671768266586228e-07, + "loss": 0.87526995, + "num_input_tokens_seen": 125458240, + "step": 5829, + "time_per_iteration": 3.6535494327545166 + }, + { + "auxiliary_loss_clip": 0.01142554, + "auxiliary_loss_mlp": 0.0102649, + "balance_loss_clip": 1.04522169, + "balance_loss_mlp": 1.01831794, + "epoch": 0.7010160524259004, + "flos": 27452307173760.0, + "grad_norm": 1.7486420175916375, + "language_loss": 0.7813251, + "learning_rate": 8.665349445284275e-07, + "loss": 0.80301553, + "num_input_tokens_seen": 125477980, + "step": 5830, + "time_per_iteration": 2.849496841430664 + }, + { + "auxiliary_loss_clip": 0.0114177, + "auxiliary_loss_mlp": 0.01026262, + "balance_loss_clip": 1.04476821, + "balance_loss_mlp": 1.01811433, + "epoch": 0.7011362953165394, + "flos": 23842064125440.0, + "grad_norm": 1.5192923636490738, + "language_loss": 0.80927569, + "learning_rate": 8.658932343439799e-07, + "loss": 0.83095604, + "num_input_tokens_seen": 125497765, + "step": 5831, + "time_per_iteration": 2.7808334827423096 + }, + { + "auxiliary_loss_clip": 0.01174485, + "auxiliary_loss_mlp": 0.01028163, + "balance_loss_clip": 1.0504843, + "balance_loss_mlp": 1.01928759, + "epoch": 0.7012565382071785, + "flos": 24823430582400.0, + "grad_norm": 3.611432380879751, + "language_loss": 0.77653313, + "learning_rate": 8.65251696202627e-07, + "loss": 0.79855961, + "num_input_tokens_seen": 125514145, + "step": 5832, + "time_per_iteration": 3.664884090423584 + }, + { + "auxiliary_loss_clip": 0.01138927, + "auxiliary_loss_mlp": 0.01029811, + "balance_loss_clip": 1.04758739, + "balance_loss_mlp": 1.02142763, + "epoch": 0.7013767810978175, + "flos": 21397445326080.0, + "grad_norm": 2.0456468375455996, + "language_loss": 0.87456203, + "learning_rate": 8.646103302016896e-07, + "loss": 0.89624941, + "num_input_tokens_seen": 125533115, + "step": 5833, + "time_per_iteration": 2.733854055404663 + }, + { + "auxiliary_loss_clip": 0.01146591, + "auxiliary_loss_mlp": 0.01031155, + "balance_loss_clip": 1.04380369, + "balance_loss_mlp": 1.02332306, + "epoch": 0.7014970239884567, + "flos": 16687150306560.0, + "grad_norm": 2.0049815083423423, + "language_loss": 0.88505781, + "learning_rate": 8.639691364384614e-07, + "loss": 0.9068352, + "num_input_tokens_seen": 125550740, + "step": 5834, + "time_per_iteration": 2.7911343574523926 + }, + { + "auxiliary_loss_clip": 0.01155442, + "auxiliary_loss_mlp": 0.01030263, + "balance_loss_clip": 1.04788184, + "balance_loss_mlp": 1.02186465, + "epoch": 0.7016172668790958, + "flos": 12568268718720.0, + "grad_norm": 1.8474115357962104, + "language_loss": 0.72672832, + "learning_rate": 8.633281150102136e-07, + "loss": 0.7485854, + "num_input_tokens_seen": 125567590, + "step": 5835, + "time_per_iteration": 2.778407335281372 + }, + { + "auxiliary_loss_clip": 0.01153945, + "auxiliary_loss_mlp": 0.01023988, + "balance_loss_clip": 1.04771018, + "balance_loss_mlp": 1.01656747, + "epoch": 0.7017375097697348, + "flos": 17452729808640.0, + "grad_norm": 2.6006044127443784, + "language_loss": 0.67922854, + "learning_rate": 8.626872660141855e-07, + "loss": 0.70100784, + "num_input_tokens_seen": 125585500, + "step": 5836, + "time_per_iteration": 3.6179521083831787 + }, + { + "auxiliary_loss_clip": 0.01134914, + "auxiliary_loss_mlp": 0.01023426, + "balance_loss_clip": 1.04430175, + "balance_loss_mlp": 1.01508188, + "epoch": 0.701857752660374, + "flos": 18513028402560.0, + "grad_norm": 1.6930537006517994, + "language_loss": 0.74859941, + "learning_rate": 8.620465895475957e-07, + "loss": 0.77018285, + "num_input_tokens_seen": 125603720, + "step": 5837, + "time_per_iteration": 2.874955654144287 + }, + { + "auxiliary_loss_clip": 0.01129637, + "auxiliary_loss_mlp": 0.0103417, + "balance_loss_clip": 1.04408026, + "balance_loss_mlp": 1.02605188, + "epoch": 0.701977995551013, + "flos": 24425971614720.0, + "grad_norm": 2.136466297657435, + "language_loss": 0.74784768, + "learning_rate": 8.614060857076333e-07, + "loss": 0.76948571, + "num_input_tokens_seen": 125624390, + "step": 5838, + "time_per_iteration": 2.799900770187378 + }, + { + "auxiliary_loss_clip": 0.01147622, + "auxiliary_loss_mlp": 0.01028664, + "balance_loss_clip": 1.04371762, + "balance_loss_mlp": 1.02042031, + "epoch": 0.7020982384416521, + "flos": 23002759958400.0, + "grad_norm": 2.036887586671783, + "language_loss": 0.7501263, + "learning_rate": 8.60765754591462e-07, + "loss": 0.77188915, + "num_input_tokens_seen": 125644085, + "step": 5839, + "time_per_iteration": 2.780158281326294 + }, + { + "auxiliary_loss_clip": 0.0117023, + "auxiliary_loss_mlp": 0.01029451, + "balance_loss_clip": 1.04779971, + "balance_loss_mlp": 1.0212729, + "epoch": 0.7022184813322913, + "flos": 20449080489600.0, + "grad_norm": 2.419291829381878, + "language_loss": 0.72747421, + "learning_rate": 8.601255962962211e-07, + "loss": 0.74947101, + "num_input_tokens_seen": 125663095, + "step": 5840, + "time_per_iteration": 2.6837940216064453 + }, + { + "auxiliary_loss_clip": 0.01174098, + "auxiliary_loss_mlp": 0.0103233, + "balance_loss_clip": 1.05161715, + "balance_loss_mlp": 1.02361584, + "epoch": 0.7023387242229303, + "flos": 19790514581760.0, + "grad_norm": 2.5552964988660696, + "language_loss": 0.7224192, + "learning_rate": 8.594856109190194e-07, + "loss": 0.74448347, + "num_input_tokens_seen": 125680125, + "step": 5841, + "time_per_iteration": 2.7146682739257812 + }, + { + "auxiliary_loss_clip": 0.01173442, + "auxiliary_loss_mlp": 0.01027041, + "balance_loss_clip": 1.04937041, + "balance_loss_mlp": 1.01851797, + "epoch": 0.7024589671135694, + "flos": 33259278286080.0, + "grad_norm": 1.5643983094785354, + "language_loss": 0.6898219, + "learning_rate": 8.588457985569446e-07, + "loss": 0.7118268, + "num_input_tokens_seen": 125703035, + "step": 5842, + "time_per_iteration": 2.8569180965423584 + }, + { + "auxiliary_loss_clip": 0.01176591, + "auxiliary_loss_mlp": 0.01029302, + "balance_loss_clip": 1.05074549, + "balance_loss_mlp": 1.02073669, + "epoch": 0.7025792100042085, + "flos": 19098982967040.0, + "grad_norm": 2.113433737765792, + "language_loss": 0.71722949, + "learning_rate": 8.582061593070542e-07, + "loss": 0.73928845, + "num_input_tokens_seen": 125723765, + "step": 5843, + "time_per_iteration": 2.701918363571167 + }, + { + "auxiliary_loss_clip": 0.0117194, + "auxiliary_loss_mlp": 0.00886136, + "balance_loss_clip": 1.04789972, + "balance_loss_mlp": 1.00082183, + "epoch": 0.7026994528948476, + "flos": 18952611045120.0, + "grad_norm": 2.5964639481785783, + "language_loss": 0.76975882, + "learning_rate": 8.57566693266383e-07, + "loss": 0.79033959, + "num_input_tokens_seen": 125741455, + "step": 5844, + "time_per_iteration": 2.679427146911621 + }, + { + "auxiliary_loss_clip": 0.01158585, + "auxiliary_loss_mlp": 0.00886282, + "balance_loss_clip": 1.0458771, + "balance_loss_mlp": 1.00082612, + "epoch": 0.7028196957854866, + "flos": 19536662188800.0, + "grad_norm": 2.1371089688519707, + "language_loss": 0.69202077, + "learning_rate": 8.569274005319354e-07, + "loss": 0.71246946, + "num_input_tokens_seen": 125759855, + "step": 5845, + "time_per_iteration": 3.0789544582366943 + }, + { + "auxiliary_loss_clip": 0.01158957, + "auxiliary_loss_mlp": 0.0102714, + "balance_loss_clip": 1.04678106, + "balance_loss_mlp": 1.01956439, + "epoch": 0.7029399386761258, + "flos": 20845318394880.0, + "grad_norm": 1.7559880508503485, + "language_loss": 0.79658175, + "learning_rate": 8.562882812006913e-07, + "loss": 0.81844276, + "num_input_tokens_seen": 125777345, + "step": 5846, + "time_per_iteration": 2.715287923812866 + }, + { + "auxiliary_loss_clip": 0.0117103, + "auxiliary_loss_mlp": 0.01029353, + "balance_loss_clip": 1.0475471, + "balance_loss_mlp": 1.02146697, + "epoch": 0.7030601815667649, + "flos": 22055005653120.0, + "grad_norm": 2.1289151015669536, + "language_loss": 0.7746011, + "learning_rate": 8.556493353696066e-07, + "loss": 0.79660493, + "num_input_tokens_seen": 125796345, + "step": 5847, + "time_per_iteration": 2.646925449371338 + }, + { + "auxiliary_loss_clip": 0.01164134, + "auxiliary_loss_mlp": 0.00886133, + "balance_loss_clip": 1.04743934, + "balance_loss_mlp": 1.0007093, + "epoch": 0.7031804244574039, + "flos": 27198742089600.0, + "grad_norm": 2.4642588432898265, + "language_loss": 0.67718589, + "learning_rate": 8.550105631356077e-07, + "loss": 0.69768858, + "num_input_tokens_seen": 125816070, + "step": 5848, + "time_per_iteration": 2.7638700008392334 + }, + { + "auxiliary_loss_clip": 0.01138521, + "auxiliary_loss_mlp": 0.01030107, + "balance_loss_clip": 1.04083991, + "balance_loss_mlp": 1.02165508, + "epoch": 0.7033006673480431, + "flos": 22379853277440.0, + "grad_norm": 2.2787214644415092, + "language_loss": 0.77406687, + "learning_rate": 8.543719645955961e-07, + "loss": 0.79575312, + "num_input_tokens_seen": 125834400, + "step": 5849, + "time_per_iteration": 2.7371606826782227 + }, + { + "auxiliary_loss_clip": 0.0115461, + "auxiliary_loss_mlp": 0.01026929, + "balance_loss_clip": 1.04683089, + "balance_loss_mlp": 1.01848888, + "epoch": 0.7034209102386821, + "flos": 24715986024960.0, + "grad_norm": 1.6140006064897392, + "language_loss": 0.74545658, + "learning_rate": 8.537335398464467e-07, + "loss": 0.76727194, + "num_input_tokens_seen": 125854720, + "step": 5850, + "time_per_iteration": 2.774641513824463 + }, + { + "auxiliary_loss_clip": 0.01153806, + "auxiliary_loss_mlp": 0.01028574, + "balance_loss_clip": 1.04383552, + "balance_loss_mlp": 1.01957989, + "epoch": 0.7035411531293212, + "flos": 22556174163840.0, + "grad_norm": 7.929391150852338, + "language_loss": 0.84938705, + "learning_rate": 8.53095288985007e-07, + "loss": 0.87121081, + "num_input_tokens_seen": 125868455, + "step": 5851, + "time_per_iteration": 2.7008891105651855 + }, + { + "auxiliary_loss_clip": 0.01169764, + "auxiliary_loss_mlp": 0.01026738, + "balance_loss_clip": 1.0480504, + "balance_loss_mlp": 1.0186851, + "epoch": 0.7036613960199604, + "flos": 22674967418880.0, + "grad_norm": 1.6077413446394213, + "language_loss": 0.82450515, + "learning_rate": 8.524572121081009e-07, + "loss": 0.84647012, + "num_input_tokens_seen": 125888555, + "step": 5852, + "time_per_iteration": 2.6486737728118896 + }, + { + "auxiliary_loss_clip": 0.01166255, + "auxiliary_loss_mlp": 0.01030162, + "balance_loss_clip": 1.0465889, + "balance_loss_mlp": 1.0217998, + "epoch": 0.7037816389105994, + "flos": 22492146170880.0, + "grad_norm": 2.178790764946486, + "language_loss": 0.62021565, + "learning_rate": 8.518193093125232e-07, + "loss": 0.64217985, + "num_input_tokens_seen": 125907610, + "step": 5853, + "time_per_iteration": 2.7348504066467285 + }, + { + "auxiliary_loss_clip": 0.01158521, + "auxiliary_loss_mlp": 0.0102776, + "balance_loss_clip": 1.04630566, + "balance_loss_mlp": 1.01993942, + "epoch": 0.7039018818012385, + "flos": 27087490690560.0, + "grad_norm": 1.7134182950221821, + "language_loss": 0.81033492, + "learning_rate": 8.511815806950436e-07, + "loss": 0.83219773, + "num_input_tokens_seen": 125928640, + "step": 5854, + "time_per_iteration": 3.654940128326416 + }, + { + "auxiliary_loss_clip": 0.01158177, + "auxiliary_loss_mlp": 0.01024188, + "balance_loss_clip": 1.04337835, + "balance_loss_mlp": 1.01599813, + "epoch": 0.7040221246918776, + "flos": 17749819198080.0, + "grad_norm": 1.7919893112357446, + "language_loss": 0.78080964, + "learning_rate": 8.505440263524044e-07, + "loss": 0.80263329, + "num_input_tokens_seen": 125947485, + "step": 5855, + "time_per_iteration": 3.592759609222412 + }, + { + "auxiliary_loss_clip": 0.01165338, + "auxiliary_loss_mlp": 0.01031891, + "balance_loss_clip": 1.04659343, + "balance_loss_mlp": 1.02278352, + "epoch": 0.7041423675825167, + "flos": 16279851012480.0, + "grad_norm": 2.5236812397933015, + "language_loss": 0.88258433, + "learning_rate": 8.49906646381322e-07, + "loss": 0.90455663, + "num_input_tokens_seen": 125960320, + "step": 5856, + "time_per_iteration": 2.6832334995269775 + }, + { + "auxiliary_loss_clip": 0.011391, + "auxiliary_loss_mlp": 0.01028186, + "balance_loss_clip": 1.04752278, + "balance_loss_mlp": 1.0207113, + "epoch": 0.7042626104731557, + "flos": 25483181639040.0, + "grad_norm": 1.8143140858977331, + "language_loss": 0.72267807, + "learning_rate": 8.492694408784884e-07, + "loss": 0.74435091, + "num_input_tokens_seen": 125980575, + "step": 5857, + "time_per_iteration": 2.802838087081909 + }, + { + "auxiliary_loss_clip": 0.01164652, + "auxiliary_loss_mlp": 0.01028452, + "balance_loss_clip": 1.04619098, + "balance_loss_mlp": 1.02084088, + "epoch": 0.7043828533637949, + "flos": 17857622891520.0, + "grad_norm": 2.490687545867199, + "language_loss": 0.62148988, + "learning_rate": 8.486324099405642e-07, + "loss": 0.64342088, + "num_input_tokens_seen": 125997420, + "step": 5858, + "time_per_iteration": 3.5728204250335693 + }, + { + "auxiliary_loss_clip": 0.0115925, + "auxiliary_loss_mlp": 0.01027418, + "balance_loss_clip": 1.0444454, + "balance_loss_mlp": 1.01920438, + "epoch": 0.704503096254434, + "flos": 29494259533440.0, + "grad_norm": 1.5949255433702976, + "language_loss": 0.75022554, + "learning_rate": 8.479955536641887e-07, + "loss": 0.77209228, + "num_input_tokens_seen": 126018915, + "step": 5859, + "time_per_iteration": 2.7304980754852295 + }, + { + "auxiliary_loss_clip": 0.01144188, + "auxiliary_loss_mlp": 0.01023956, + "balance_loss_clip": 1.04032469, + "balance_loss_mlp": 1.01593947, + "epoch": 0.704623339145073, + "flos": 30920739327360.0, + "grad_norm": 1.8613983062226538, + "language_loss": 0.66129744, + "learning_rate": 8.473588721459716e-07, + "loss": 0.68297887, + "num_input_tokens_seen": 126038825, + "step": 5860, + "time_per_iteration": 2.8243160247802734 + }, + { + "auxiliary_loss_clip": 0.01168517, + "auxiliary_loss_mlp": 0.01028886, + "balance_loss_clip": 1.05263436, + "balance_loss_mlp": 1.02010584, + "epoch": 0.7047435820357122, + "flos": 23914747296000.0, + "grad_norm": 2.9472064555309094, + "language_loss": 0.70358157, + "learning_rate": 8.467223654824967e-07, + "loss": 0.7255556, + "num_input_tokens_seen": 126058280, + "step": 5861, + "time_per_iteration": 2.783275604248047 + }, + { + "auxiliary_loss_clip": 0.01156528, + "auxiliary_loss_mlp": 0.01030336, + "balance_loss_clip": 1.04645538, + "balance_loss_mlp": 1.02242637, + "epoch": 0.7048638249263512, + "flos": 46494010926720.0, + "grad_norm": 2.0389115132552362, + "language_loss": 0.62505531, + "learning_rate": 8.460860337703233e-07, + "loss": 0.64692396, + "num_input_tokens_seen": 126078885, + "step": 5862, + "time_per_iteration": 3.8620073795318604 + }, + { + "auxiliary_loss_clip": 0.01129365, + "auxiliary_loss_mlp": 0.01028661, + "balance_loss_clip": 1.04051948, + "balance_loss_mlp": 1.01982141, + "epoch": 0.7049840678169903, + "flos": 21689219502720.0, + "grad_norm": 1.8052203450266893, + "language_loss": 0.70491409, + "learning_rate": 8.454498771059797e-07, + "loss": 0.72649431, + "num_input_tokens_seen": 126098260, + "step": 5863, + "time_per_iteration": 2.811204195022583 + }, + { + "auxiliary_loss_clip": 0.01131419, + "auxiliary_loss_mlp": 0.01030013, + "balance_loss_clip": 1.04232132, + "balance_loss_mlp": 1.02211571, + "epoch": 0.7051043107076294, + "flos": 18405081054720.0, + "grad_norm": 2.6456265322614003, + "language_loss": 0.83240581, + "learning_rate": 8.448138955859725e-07, + "loss": 0.85402012, + "num_input_tokens_seen": 126114845, + "step": 5864, + "time_per_iteration": 2.8687875270843506 + }, + { + "auxiliary_loss_clip": 0.01152248, + "auxiliary_loss_mlp": 0.01030467, + "balance_loss_clip": 1.04515338, + "balance_loss_mlp": 1.02247405, + "epoch": 0.7052245535982685, + "flos": 19319043640320.0, + "grad_norm": 1.8309602270620196, + "language_loss": 0.90197247, + "learning_rate": 8.44178089306778e-07, + "loss": 0.92379963, + "num_input_tokens_seen": 126132780, + "step": 5865, + "time_per_iteration": 2.8206064701080322 + }, + { + "auxiliary_loss_clip": 0.01172175, + "auxiliary_loss_mlp": 0.01025807, + "balance_loss_clip": 1.04982686, + "balance_loss_mlp": 1.01779056, + "epoch": 0.7053447964889076, + "flos": 19062138591360.0, + "grad_norm": 1.9382759313049167, + "language_loss": 0.77280593, + "learning_rate": 8.4354245836485e-07, + "loss": 0.79478586, + "num_input_tokens_seen": 126151225, + "step": 5866, + "time_per_iteration": 2.680295705795288 + }, + { + "auxiliary_loss_clip": 0.01146211, + "auxiliary_loss_mlp": 0.01034476, + "balance_loss_clip": 1.04445028, + "balance_loss_mlp": 1.02595806, + "epoch": 0.7054650393795466, + "flos": 27379228953600.0, + "grad_norm": 1.5373403711697151, + "language_loss": 0.7291826, + "learning_rate": 8.429070028566108e-07, + "loss": 0.75098944, + "num_input_tokens_seen": 126172535, + "step": 5867, + "time_per_iteration": 2.752457618713379 + }, + { + "auxiliary_loss_clip": 0.0116128, + "auxiliary_loss_mlp": 0.01025455, + "balance_loss_clip": 1.0468868, + "balance_loss_mlp": 1.01775146, + "epoch": 0.7055852822701858, + "flos": 16102201322880.0, + "grad_norm": 2.1408192776969073, + "language_loss": 0.74784791, + "learning_rate": 8.422717228784586e-07, + "loss": 0.76971531, + "num_input_tokens_seen": 126189410, + "step": 5868, + "time_per_iteration": 2.7033023834228516 + }, + { + "auxiliary_loss_clip": 0.01135576, + "auxiliary_loss_mlp": 0.01025067, + "balance_loss_clip": 1.04811788, + "balance_loss_mlp": 1.0167644, + "epoch": 0.7057055251608249, + "flos": 11692299744000.0, + "grad_norm": 1.9091608986306114, + "language_loss": 0.69663262, + "learning_rate": 8.416366185267663e-07, + "loss": 0.71823907, + "num_input_tokens_seen": 126206910, + "step": 5869, + "time_per_iteration": 2.974409818649292 + }, + { + "auxiliary_loss_clip": 0.01161802, + "auxiliary_loss_mlp": 0.01029877, + "balance_loss_clip": 1.04504716, + "balance_loss_mlp": 1.02147925, + "epoch": 0.7058257680514639, + "flos": 22711560399360.0, + "grad_norm": 1.6257834322427938, + "language_loss": 0.77873421, + "learning_rate": 8.410016898978778e-07, + "loss": 0.80065101, + "num_input_tokens_seen": 126224385, + "step": 5870, + "time_per_iteration": 2.8027830123901367 + }, + { + "auxiliary_loss_clip": 0.01123057, + "auxiliary_loss_mlp": 0.01028688, + "balance_loss_clip": 1.04255033, + "balance_loss_mlp": 1.02099347, + "epoch": 0.7059460109421031, + "flos": 17529543043200.0, + "grad_norm": 1.9919775515670917, + "language_loss": 0.79082549, + "learning_rate": 8.403669370881115e-07, + "loss": 0.81234294, + "num_input_tokens_seen": 126243120, + "step": 5871, + "time_per_iteration": 2.797869920730591 + }, + { + "auxiliary_loss_clip": 0.01171819, + "auxiliary_loss_mlp": 0.0102788, + "balance_loss_clip": 1.04806745, + "balance_loss_mlp": 1.02020335, + "epoch": 0.7060662538327421, + "flos": 23544687427200.0, + "grad_norm": 1.6575235000291435, + "language_loss": 0.78598952, + "learning_rate": 8.397323601937587e-07, + "loss": 0.8079865, + "num_input_tokens_seen": 126263020, + "step": 5872, + "time_per_iteration": 2.6879501342773438 + }, + { + "auxiliary_loss_clip": 0.01138607, + "auxiliary_loss_mlp": 0.01027293, + "balance_loss_clip": 1.04384351, + "balance_loss_mlp": 1.01928771, + "epoch": 0.7061864967233812, + "flos": 30260736875520.0, + "grad_norm": 1.8892588157297998, + "language_loss": 0.76766205, + "learning_rate": 8.390979593110838e-07, + "loss": 0.78932106, + "num_input_tokens_seen": 126285150, + "step": 5873, + "time_per_iteration": 2.8491804599761963 + }, + { + "auxiliary_loss_clip": 0.0116069, + "auxiliary_loss_mlp": 0.01027591, + "balance_loss_clip": 1.05063033, + "balance_loss_mlp": 1.01911497, + "epoch": 0.7063067396140204, + "flos": 20701460424960.0, + "grad_norm": 1.5141185023763815, + "language_loss": 0.81616348, + "learning_rate": 8.384637345363262e-07, + "loss": 0.83804631, + "num_input_tokens_seen": 126304340, + "step": 5874, + "time_per_iteration": 2.6894097328186035 + }, + { + "auxiliary_loss_clip": 0.01143835, + "auxiliary_loss_mlp": 0.01027571, + "balance_loss_clip": 1.04100132, + "balance_loss_mlp": 1.01898217, + "epoch": 0.7064269825046594, + "flos": 32266168081920.0, + "grad_norm": 1.8913443084748311, + "language_loss": 0.76799715, + "learning_rate": 8.378296859656964e-07, + "loss": 0.78971124, + "num_input_tokens_seen": 126325495, + "step": 5875, + "time_per_iteration": 2.834286689758301 + }, + { + "auxiliary_loss_clip": 0.01155732, + "auxiliary_loss_mlp": 0.01027485, + "balance_loss_clip": 1.04687762, + "balance_loss_mlp": 1.01901555, + "epoch": 0.7065472253952985, + "flos": 30227124723840.0, + "grad_norm": 2.055725907078453, + "language_loss": 0.67857695, + "learning_rate": 8.371958136953792e-07, + "loss": 0.70040911, + "num_input_tokens_seen": 126345525, + "step": 5876, + "time_per_iteration": 2.811786651611328 + }, + { + "auxiliary_loss_clip": 0.01145725, + "auxiliary_loss_mlp": 0.01030108, + "balance_loss_clip": 1.04151738, + "balance_loss_mlp": 1.02131605, + "epoch": 0.7066674682859376, + "flos": 16216720859520.0, + "grad_norm": 2.681408034420322, + "language_loss": 0.66373527, + "learning_rate": 8.365621178215326e-07, + "loss": 0.68549359, + "num_input_tokens_seen": 126361995, + "step": 5877, + "time_per_iteration": 2.7695930004119873 + }, + { + "auxiliary_loss_clip": 0.0115781, + "auxiliary_loss_mlp": 0.01023089, + "balance_loss_clip": 1.04535544, + "balance_loss_mlp": 1.01480114, + "epoch": 0.7067877111765767, + "flos": 14830461319680.0, + "grad_norm": 2.01530307313066, + "language_loss": 0.75293076, + "learning_rate": 8.359285984402871e-07, + "loss": 0.77473968, + "num_input_tokens_seen": 126379260, + "step": 5878, + "time_per_iteration": 2.759819269180298 + }, + { + "auxiliary_loss_clip": 0.01150118, + "auxiliary_loss_mlp": 0.01028968, + "balance_loss_clip": 1.04514098, + "balance_loss_mlp": 1.02167833, + "epoch": 0.7069079540672157, + "flos": 25440196037760.0, + "grad_norm": 1.9020597683387248, + "language_loss": 0.74198896, + "learning_rate": 8.352952556477489e-07, + "loss": 0.76377982, + "num_input_tokens_seen": 126397170, + "step": 5879, + "time_per_iteration": 2.7296230792999268 + }, + { + "auxiliary_loss_clip": 0.01162341, + "auxiliary_loss_mlp": 0.01027008, + "balance_loss_clip": 1.0482378, + "balance_loss_mlp": 1.01882398, + "epoch": 0.7070281969578549, + "flos": 24607751368320.0, + "grad_norm": 1.9877474031616655, + "language_loss": 0.76668078, + "learning_rate": 8.34662089539993e-07, + "loss": 0.78857422, + "num_input_tokens_seen": 126416680, + "step": 5880, + "time_per_iteration": 3.6657607555389404 + }, + { + "auxiliary_loss_clip": 0.01168556, + "auxiliary_loss_mlp": 0.01027126, + "balance_loss_clip": 1.04828036, + "balance_loss_mlp": 1.0194366, + "epoch": 0.707148439848494, + "flos": 26724469887360.0, + "grad_norm": 2.125475632503464, + "language_loss": 0.78841269, + "learning_rate": 8.340291002130722e-07, + "loss": 0.81036949, + "num_input_tokens_seen": 126435870, + "step": 5881, + "time_per_iteration": 3.5742223262786865 + }, + { + "auxiliary_loss_clip": 0.01175401, + "auxiliary_loss_mlp": 0.01029384, + "balance_loss_clip": 1.05040729, + "balance_loss_mlp": 1.02066982, + "epoch": 0.707268682739133, + "flos": 15085750256640.0, + "grad_norm": 2.598221488134716, + "language_loss": 0.79516047, + "learning_rate": 8.3339628776301e-07, + "loss": 0.81720829, + "num_input_tokens_seen": 126454010, + "step": 5882, + "time_per_iteration": 2.7693967819213867 + }, + { + "auxiliary_loss_clip": 0.01169973, + "auxiliary_loss_mlp": 0.01027907, + "balance_loss_clip": 1.0476923, + "balance_loss_mlp": 1.02026284, + "epoch": 0.7073889256297722, + "flos": 34313148345600.0, + "grad_norm": 2.0461618848079723, + "language_loss": 0.57362026, + "learning_rate": 8.327636522858033e-07, + "loss": 0.59559906, + "num_input_tokens_seen": 126473615, + "step": 5883, + "time_per_iteration": 2.70422625541687 + }, + { + "auxiliary_loss_clip": 0.0112788, + "auxiliary_loss_mlp": 0.01026862, + "balance_loss_clip": 1.04616559, + "balance_loss_mlp": 1.01883888, + "epoch": 0.7075091685204112, + "flos": 20083940784000.0, + "grad_norm": 4.200574151511155, + "language_loss": 0.77326077, + "learning_rate": 8.321311938774225e-07, + "loss": 0.79480815, + "num_input_tokens_seen": 126492705, + "step": 5884, + "time_per_iteration": 3.7046148777008057 + }, + { + "auxiliary_loss_clip": 0.01176505, + "auxiliary_loss_mlp": 0.01032265, + "balance_loss_clip": 1.05127966, + "balance_loss_mlp": 1.02405703, + "epoch": 0.7076294114110503, + "flos": 20777124424320.0, + "grad_norm": 2.031176003120811, + "language_loss": 0.79427308, + "learning_rate": 8.314989126338104e-07, + "loss": 0.81636077, + "num_input_tokens_seen": 126512715, + "step": 5885, + "time_per_iteration": 2.665639877319336 + }, + { + "auxiliary_loss_clip": 0.01162119, + "auxiliary_loss_mlp": 0.01027505, + "balance_loss_clip": 1.04479003, + "balance_loss_mlp": 1.01920772, + "epoch": 0.7077496543016895, + "flos": 17967689141760.0, + "grad_norm": 1.7587095893884293, + "language_loss": 0.84105092, + "learning_rate": 8.308668086508847e-07, + "loss": 0.86294717, + "num_input_tokens_seen": 126530795, + "step": 5886, + "time_per_iteration": 2.6747326850891113 + }, + { + "auxiliary_loss_clip": 0.01145948, + "auxiliary_loss_mlp": 0.01031906, + "balance_loss_clip": 1.04129398, + "balance_loss_mlp": 1.02309632, + "epoch": 0.7078698971923285, + "flos": 45478098564480.0, + "grad_norm": 1.8599265310629132, + "language_loss": 0.73748326, + "learning_rate": 8.302348820245342e-07, + "loss": 0.75926185, + "num_input_tokens_seen": 126553360, + "step": 5887, + "time_per_iteration": 3.011125326156616 + }, + { + "auxiliary_loss_clip": 0.01136097, + "auxiliary_loss_mlp": 0.01024877, + "balance_loss_clip": 1.04379749, + "balance_loss_mlp": 1.01625836, + "epoch": 0.7079901400829676, + "flos": 26943704547840.0, + "grad_norm": 2.45767663175848, + "language_loss": 0.7027182, + "learning_rate": 8.296031328506232e-07, + "loss": 0.72432792, + "num_input_tokens_seen": 126573110, + "step": 5888, + "time_per_iteration": 3.74731707572937 + }, + { + "auxiliary_loss_clip": 0.01154655, + "auxiliary_loss_mlp": 0.01030278, + "balance_loss_clip": 1.04549503, + "balance_loss_mlp": 1.02227855, + "epoch": 0.7081103829736067, + "flos": 24423206267520.0, + "grad_norm": 1.918669100339438, + "language_loss": 0.75420988, + "learning_rate": 8.289715612249857e-07, + "loss": 0.77605915, + "num_input_tokens_seen": 126593725, + "step": 5889, + "time_per_iteration": 2.741863250732422 + }, + { + "auxiliary_loss_clip": 0.01147661, + "auxiliary_loss_mlp": 0.01029476, + "balance_loss_clip": 1.04448915, + "balance_loss_mlp": 1.02102995, + "epoch": 0.7082306258642458, + "flos": 18543300589440.0, + "grad_norm": 2.7709776745552817, + "language_loss": 0.77869117, + "learning_rate": 8.283401672434305e-07, + "loss": 0.80046248, + "num_input_tokens_seen": 126608950, + "step": 5890, + "time_per_iteration": 2.7222073078155518 + }, + { + "auxiliary_loss_clip": 0.01152042, + "auxiliary_loss_mlp": 0.01025245, + "balance_loss_clip": 1.04838395, + "balance_loss_mlp": 1.01752889, + "epoch": 0.7083508687548848, + "flos": 23477534951040.0, + "grad_norm": 1.9162073886792832, + "language_loss": 0.70123947, + "learning_rate": 8.277089510017412e-07, + "loss": 0.72301233, + "num_input_tokens_seen": 126629755, + "step": 5891, + "time_per_iteration": 2.7505650520324707 + }, + { + "auxiliary_loss_clip": 0.01149252, + "auxiliary_loss_mlp": 0.01026241, + "balance_loss_clip": 1.04620957, + "balance_loss_mlp": 1.01897573, + "epoch": 0.708471111645524, + "flos": 22419463000320.0, + "grad_norm": 1.7015085276223936, + "language_loss": 0.81916219, + "learning_rate": 8.270779125956719e-07, + "loss": 0.84091711, + "num_input_tokens_seen": 126650135, + "step": 5892, + "time_per_iteration": 2.834312915802002 + }, + { + "auxiliary_loss_clip": 0.01127972, + "auxiliary_loss_mlp": 0.01029537, + "balance_loss_clip": 1.04367113, + "balance_loss_mlp": 1.02165687, + "epoch": 0.7085913545361631, + "flos": 20922885815040.0, + "grad_norm": 2.261220936779811, + "language_loss": 0.79688573, + "learning_rate": 8.264470521209505e-07, + "loss": 0.81846082, + "num_input_tokens_seen": 126668500, + "step": 5893, + "time_per_iteration": 3.050421714782715 + }, + { + "auxiliary_loss_clip": 0.01153942, + "auxiliary_loss_mlp": 0.01030866, + "balance_loss_clip": 1.0449295, + "balance_loss_mlp": 1.02219057, + "epoch": 0.7087115974268021, + "flos": 15012384727680.0, + "grad_norm": 2.630024502269483, + "language_loss": 0.76379293, + "learning_rate": 8.258163696732785e-07, + "loss": 0.78564095, + "num_input_tokens_seen": 126686090, + "step": 5894, + "time_per_iteration": 2.9002878665924072 + }, + { + "auxiliary_loss_clip": 0.01158035, + "auxiliary_loss_mlp": 0.01027146, + "balance_loss_clip": 1.04676783, + "balance_loss_mlp": 1.01941574, + "epoch": 0.7088318403174413, + "flos": 21539040739200.0, + "grad_norm": 2.090757241134485, + "language_loss": 0.77245748, + "learning_rate": 8.251858653483288e-07, + "loss": 0.79430926, + "num_input_tokens_seen": 126704255, + "step": 5895, + "time_per_iteration": 2.7789058685302734 + }, + { + "auxiliary_loss_clip": 0.01163863, + "auxiliary_loss_mlp": 0.01029436, + "balance_loss_clip": 1.04877579, + "balance_loss_mlp": 1.02134216, + "epoch": 0.7089520832080803, + "flos": 15516785462400.0, + "grad_norm": 2.769965306046548, + "language_loss": 0.85846913, + "learning_rate": 8.245555392417501e-07, + "loss": 0.88040209, + "num_input_tokens_seen": 126718910, + "step": 5896, + "time_per_iteration": 2.738290786743164 + }, + { + "auxiliary_loss_clip": 0.01118892, + "auxiliary_loss_mlp": 0.01032752, + "balance_loss_clip": 1.03939557, + "balance_loss_mlp": 1.02479506, + "epoch": 0.7090723260987194, + "flos": 20412667077120.0, + "grad_norm": 1.7762273331538392, + "language_loss": 0.7884407, + "learning_rate": 8.239253914491613e-07, + "loss": 0.80995715, + "num_input_tokens_seen": 126737235, + "step": 5897, + "time_per_iteration": 3.1037449836730957 + }, + { + "auxiliary_loss_clip": 0.01140084, + "auxiliary_loss_mlp": 0.01028155, + "balance_loss_clip": 1.04518676, + "balance_loss_mlp": 1.02040052, + "epoch": 0.7091925689893585, + "flos": 25668337271040.0, + "grad_norm": 2.1033868148342565, + "language_loss": 0.75271261, + "learning_rate": 8.232954220661556e-07, + "loss": 0.77439499, + "num_input_tokens_seen": 126759970, + "step": 5898, + "time_per_iteration": 2.8672094345092773 + }, + { + "auxiliary_loss_clip": 0.0117141, + "auxiliary_loss_mlp": 0.0103233, + "balance_loss_clip": 1.05078864, + "balance_loss_mlp": 1.02429795, + "epoch": 0.7093128118799976, + "flos": 24206629213440.0, + "grad_norm": 2.3024235803013045, + "language_loss": 0.70087016, + "learning_rate": 8.226656311882989e-07, + "loss": 0.72290754, + "num_input_tokens_seen": 126779280, + "step": 5899, + "time_per_iteration": 2.845200538635254 + }, + { + "auxiliary_loss_clip": 0.01160227, + "auxiliary_loss_mlp": 0.01028096, + "balance_loss_clip": 1.04892099, + "balance_loss_mlp": 1.02043033, + "epoch": 0.7094330547706367, + "flos": 16646786398080.0, + "grad_norm": 2.0405588882247048, + "language_loss": 0.76567918, + "learning_rate": 8.22036018911129e-07, + "loss": 0.78756237, + "num_input_tokens_seen": 126797310, + "step": 5900, + "time_per_iteration": 2.6822357177734375 + }, + { + "auxiliary_loss_clip": 0.01175377, + "auxiliary_loss_mlp": 0.01031219, + "balance_loss_clip": 1.04902637, + "balance_loss_mlp": 1.02231431, + "epoch": 0.7095532976612757, + "flos": 16283370545280.0, + "grad_norm": 2.0375055882686826, + "language_loss": 0.80601013, + "learning_rate": 8.214065853301599e-07, + "loss": 0.82807606, + "num_input_tokens_seen": 126812840, + "step": 5901, + "time_per_iteration": 2.688098430633545 + }, + { + "auxiliary_loss_clip": 0.01057779, + "auxiliary_loss_mlp": 0.01001604, + "balance_loss_clip": 1.0085752, + "balance_loss_mlp": 0.99993497, + "epoch": 0.7096735405519149, + "flos": 70722080559360.0, + "grad_norm": 0.8022975046757506, + "language_loss": 0.58154428, + "learning_rate": 8.207773305408734e-07, + "loss": 0.60213816, + "num_input_tokens_seen": 126880060, + "step": 5902, + "time_per_iteration": 3.4818930625915527 + }, + { + "auxiliary_loss_clip": 0.01141939, + "auxiliary_loss_mlp": 0.01034955, + "balance_loss_clip": 1.04253197, + "balance_loss_mlp": 1.02685511, + "epoch": 0.709793783442554, + "flos": 23621500661760.0, + "grad_norm": 4.718880325812836, + "language_loss": 0.80087447, + "learning_rate": 8.201482546387288e-07, + "loss": 0.8226434, + "num_input_tokens_seen": 126899535, + "step": 5903, + "time_per_iteration": 2.9121522903442383 + }, + { + "auxiliary_loss_clip": 0.01162569, + "auxiliary_loss_mlp": 0.01028544, + "balance_loss_clip": 1.04771876, + "balance_loss_mlp": 1.02120662, + "epoch": 0.709914026333193, + "flos": 25993472204160.0, + "grad_norm": 1.9073765946842172, + "language_loss": 0.91998422, + "learning_rate": 8.195193577191553e-07, + "loss": 0.94189543, + "num_input_tokens_seen": 126921365, + "step": 5904, + "time_per_iteration": 2.8720874786376953 + }, + { + "auxiliary_loss_clip": 0.01152888, + "auxiliary_loss_mlp": 0.0088646, + "balance_loss_clip": 1.04635286, + "balance_loss_mlp": 1.00074792, + "epoch": 0.7100342692238322, + "flos": 24861531934080.0, + "grad_norm": 3.0182859445007715, + "language_loss": 0.84608841, + "learning_rate": 8.188906398775579e-07, + "loss": 0.86648184, + "num_input_tokens_seen": 126941910, + "step": 5905, + "time_per_iteration": 2.8218905925750732 + }, + { + "auxiliary_loss_clip": 0.01173967, + "auxiliary_loss_mlp": 0.00886188, + "balance_loss_clip": 1.04856682, + "balance_loss_mlp": 1.00076234, + "epoch": 0.7101545121144712, + "flos": 24932203943040.0, + "grad_norm": 1.832915279899737, + "language_loss": 0.68616354, + "learning_rate": 8.18262101209311e-07, + "loss": 0.70676506, + "num_input_tokens_seen": 126961120, + "step": 5906, + "time_per_iteration": 4.638651609420776 + }, + { + "auxiliary_loss_clip": 0.011571, + "auxiliary_loss_mlp": 0.01026668, + "balance_loss_clip": 1.04626095, + "balance_loss_mlp": 1.01905608, + "epoch": 0.7102747550051103, + "flos": 23768842250880.0, + "grad_norm": 1.9287389660497058, + "language_loss": 0.70168567, + "learning_rate": 8.176337418097626e-07, + "loss": 0.72352338, + "num_input_tokens_seen": 126981590, + "step": 5907, + "time_per_iteration": 2.829833507537842 + }, + { + "auxiliary_loss_clip": 0.01160418, + "auxiliary_loss_mlp": 0.0088601, + "balance_loss_clip": 1.04767275, + "balance_loss_mlp": 1.00076687, + "epoch": 0.7103949978957494, + "flos": 15303907509120.0, + "grad_norm": 1.9338167480191666, + "language_loss": 0.79300773, + "learning_rate": 8.170055617742364e-07, + "loss": 0.81347203, + "num_input_tokens_seen": 126998870, + "step": 5908, + "time_per_iteration": 2.776137113571167 + }, + { + "auxiliary_loss_clip": 0.01148344, + "auxiliary_loss_mlp": 0.01033848, + "balance_loss_clip": 1.04537177, + "balance_loss_mlp": 1.02512217, + "epoch": 0.7105152407863885, + "flos": 22638805401600.0, + "grad_norm": 1.7567489279193176, + "language_loss": 0.7067337, + "learning_rate": 8.163775611980252e-07, + "loss": 0.72855556, + "num_input_tokens_seen": 127017980, + "step": 5909, + "time_per_iteration": 3.6890857219696045 + }, + { + "auxiliary_loss_clip": 0.01152484, + "auxiliary_loss_mlp": 0.01024307, + "balance_loss_clip": 1.04620695, + "balance_loss_mlp": 1.01611745, + "epoch": 0.7106354836770276, + "flos": 17238594879360.0, + "grad_norm": 2.062861506298314, + "language_loss": 0.7880013, + "learning_rate": 8.157497401763982e-07, + "loss": 0.80976921, + "num_input_tokens_seen": 127035645, + "step": 5910, + "time_per_iteration": 2.8263330459594727 + }, + { + "auxiliary_loss_clip": 0.01160026, + "auxiliary_loss_mlp": 0.01025435, + "balance_loss_clip": 1.04665124, + "balance_loss_mlp": 1.01793027, + "epoch": 0.7107557265676667, + "flos": 20193647898240.0, + "grad_norm": 2.0612365666003774, + "language_loss": 0.77747047, + "learning_rate": 8.151220988045935e-07, + "loss": 0.79932499, + "num_input_tokens_seen": 127054900, + "step": 5911, + "time_per_iteration": 2.724705219268799 + }, + { + "auxiliary_loss_clip": 0.01161623, + "auxiliary_loss_mlp": 0.0102846, + "balance_loss_clip": 1.04712892, + "balance_loss_mlp": 1.02075267, + "epoch": 0.7108759694583058, + "flos": 21507080613120.0, + "grad_norm": 1.838304744415265, + "language_loss": 0.82936895, + "learning_rate": 8.144946371778234e-07, + "loss": 0.85126978, + "num_input_tokens_seen": 127075010, + "step": 5912, + "time_per_iteration": 2.7567262649536133 + }, + { + "auxiliary_loss_clip": 0.01151641, + "auxiliary_loss_mlp": 0.00886381, + "balance_loss_clip": 1.04743493, + "balance_loss_mlp": 1.00072527, + "epoch": 0.7109962123489448, + "flos": 24061909317120.0, + "grad_norm": 2.321929500236099, + "language_loss": 0.77788889, + "learning_rate": 8.138673553912751e-07, + "loss": 0.79826915, + "num_input_tokens_seen": 127095570, + "step": 5913, + "time_per_iteration": 3.8496439456939697 + }, + { + "auxiliary_loss_clip": 0.0112767, + "auxiliary_loss_mlp": 0.0103174, + "balance_loss_clip": 1.04384887, + "balance_loss_mlp": 1.02395582, + "epoch": 0.711116455239584, + "flos": 30480474326400.0, + "grad_norm": 2.918393875946937, + "language_loss": 0.56509316, + "learning_rate": 8.132402535401059e-07, + "loss": 0.58668727, + "num_input_tokens_seen": 127116825, + "step": 5914, + "time_per_iteration": 2.909902572631836 + }, + { + "auxiliary_loss_clip": 0.01157711, + "auxiliary_loss_mlp": 0.01024505, + "balance_loss_clip": 1.04826808, + "balance_loss_mlp": 1.01670849, + "epoch": 0.711236698130223, + "flos": 25045610158080.0, + "grad_norm": 1.7725018223316171, + "language_loss": 0.74151647, + "learning_rate": 8.126133317194465e-07, + "loss": 0.76333869, + "num_input_tokens_seen": 127137015, + "step": 5915, + "time_per_iteration": 2.791980028152466 + }, + { + "auxiliary_loss_clip": 0.01124555, + "auxiliary_loss_mlp": 0.01032838, + "balance_loss_clip": 1.04053724, + "balance_loss_mlp": 1.02386141, + "epoch": 0.7113569410208621, + "flos": 24206701040640.0, + "grad_norm": 1.72033901491944, + "language_loss": 0.74208248, + "learning_rate": 8.11986590024401e-07, + "loss": 0.76365644, + "num_input_tokens_seen": 127156755, + "step": 5916, + "time_per_iteration": 2.85154390335083 + }, + { + "auxiliary_loss_clip": 0.0115787, + "auxiliary_loss_mlp": 0.01024761, + "balance_loss_clip": 1.04896462, + "balance_loss_mlp": 1.01587427, + "epoch": 0.7114771839115013, + "flos": 35439306526080.0, + "grad_norm": 1.7546360809608996, + "language_loss": 0.68923348, + "learning_rate": 8.113600285500442e-07, + "loss": 0.71105981, + "num_input_tokens_seen": 127176965, + "step": 5917, + "time_per_iteration": 2.8499484062194824 + }, + { + "auxiliary_loss_clip": 0.01172218, + "auxiliary_loss_mlp": 0.01025, + "balance_loss_clip": 1.048208, + "balance_loss_mlp": 1.0168283, + "epoch": 0.7115974268021403, + "flos": 21099458096640.0, + "grad_norm": 1.967872751203376, + "language_loss": 0.74337208, + "learning_rate": 8.107336473914268e-07, + "loss": 0.76534426, + "num_input_tokens_seen": 127195595, + "step": 5918, + "time_per_iteration": 2.640516757965088 + }, + { + "auxiliary_loss_clip": 0.01048712, + "auxiliary_loss_mlp": 0.0100493, + "balance_loss_clip": 1.00918937, + "balance_loss_mlp": 1.00318992, + "epoch": 0.7117176696927794, + "flos": 56752866616320.0, + "grad_norm": 0.8235420968485608, + "language_loss": 0.55735672, + "learning_rate": 8.101074466435694e-07, + "loss": 0.57789314, + "num_input_tokens_seen": 127255070, + "step": 5919, + "time_per_iteration": 3.3045871257781982 + }, + { + "auxiliary_loss_clip": 0.01158296, + "auxiliary_loss_mlp": 0.01032139, + "balance_loss_clip": 1.04742932, + "balance_loss_mlp": 1.02358532, + "epoch": 0.7118379125834186, + "flos": 15925269905280.0, + "grad_norm": 1.9749920565000247, + "language_loss": 0.68171751, + "learning_rate": 8.094814264014662e-07, + "loss": 0.70362186, + "num_input_tokens_seen": 127273825, + "step": 5920, + "time_per_iteration": 2.737701892852783 + }, + { + "auxiliary_loss_clip": 0.01174991, + "auxiliary_loss_mlp": 0.01030966, + "balance_loss_clip": 1.04852104, + "balance_loss_mlp": 1.02187634, + "epoch": 0.7119581554740576, + "flos": 20193360589440.0, + "grad_norm": 2.1233386546151363, + "language_loss": 0.81454253, + "learning_rate": 8.088555867600844e-07, + "loss": 0.83660203, + "num_input_tokens_seen": 127289990, + "step": 5921, + "time_per_iteration": 2.6429409980773926 + }, + { + "auxiliary_loss_clip": 0.01138423, + "auxiliary_loss_mlp": 0.01027442, + "balance_loss_clip": 1.04198158, + "balance_loss_mlp": 1.01948512, + "epoch": 0.7120783983646967, + "flos": 34715383822080.0, + "grad_norm": 2.2392601866742248, + "language_loss": 0.60459048, + "learning_rate": 8.08229927814362e-07, + "loss": 0.62624913, + "num_input_tokens_seen": 127312880, + "step": 5922, + "time_per_iteration": 2.9467568397521973 + }, + { + "auxiliary_loss_clip": 0.01140555, + "auxiliary_loss_mlp": 0.01027655, + "balance_loss_clip": 1.04220355, + "balance_loss_mlp": 1.02020478, + "epoch": 0.7121986412553358, + "flos": 26359114700160.0, + "grad_norm": 1.82918976068363, + "language_loss": 0.64949465, + "learning_rate": 8.076044496592134e-07, + "loss": 0.67117679, + "num_input_tokens_seen": 127334730, + "step": 5923, + "time_per_iteration": 2.886601448059082 + }, + { + "auxiliary_loss_clip": 0.01152975, + "auxiliary_loss_mlp": 0.01028593, + "balance_loss_clip": 1.04661584, + "balance_loss_mlp": 1.0202837, + "epoch": 0.7123188841459749, + "flos": 11145344371200.0, + "grad_norm": 2.2352224520358677, + "language_loss": 0.78209424, + "learning_rate": 8.069791523895204e-07, + "loss": 0.80390996, + "num_input_tokens_seen": 127351180, + "step": 5924, + "time_per_iteration": 2.7638392448425293 + }, + { + "auxiliary_loss_clip": 0.01131409, + "auxiliary_loss_mlp": 0.01029159, + "balance_loss_clip": 1.03958511, + "balance_loss_mlp": 1.02070427, + "epoch": 0.7124391270366139, + "flos": 20811670329600.0, + "grad_norm": 2.2896879996467723, + "language_loss": 0.77622676, + "learning_rate": 8.063540361001422e-07, + "loss": 0.79783249, + "num_input_tokens_seen": 127369750, + "step": 5925, + "time_per_iteration": 2.7811453342437744 + }, + { + "auxiliary_loss_clip": 0.01136542, + "auxiliary_loss_mlp": 0.01032735, + "balance_loss_clip": 1.04327452, + "balance_loss_mlp": 1.02495933, + "epoch": 0.7125593699272531, + "flos": 17603734584960.0, + "grad_norm": 2.5363633422477005, + "language_loss": 0.79658902, + "learning_rate": 8.057291008859069e-07, + "loss": 0.81828177, + "num_input_tokens_seen": 127387910, + "step": 5926, + "time_per_iteration": 2.9381887912750244 + }, + { + "auxiliary_loss_clip": 0.01157625, + "auxiliary_loss_mlp": 0.01029212, + "balance_loss_clip": 1.04462099, + "balance_loss_mlp": 1.02176738, + "epoch": 0.7126796128178922, + "flos": 28654057526400.0, + "grad_norm": 2.0667730983770016, + "language_loss": 0.68106079, + "learning_rate": 8.051043468416187e-07, + "loss": 0.70292914, + "num_input_tokens_seen": 127409160, + "step": 5927, + "time_per_iteration": 2.853879928588867 + }, + { + "auxiliary_loss_clip": 0.01170065, + "auxiliary_loss_mlp": 0.01033695, + "balance_loss_clip": 1.04878187, + "balance_loss_mlp": 1.02626538, + "epoch": 0.7127998557085312, + "flos": 16034438315520.0, + "grad_norm": 2.542685736037904, + "language_loss": 0.8207891, + "learning_rate": 8.044797740620506e-07, + "loss": 0.84282672, + "num_input_tokens_seen": 127427765, + "step": 5928, + "time_per_iteration": 2.8184638023376465 + }, + { + "auxiliary_loss_clip": 0.01126656, + "auxiliary_loss_mlp": 0.01030345, + "balance_loss_clip": 1.04258955, + "balance_loss_mlp": 1.02232265, + "epoch": 0.7129200985991703, + "flos": 23403271582080.0, + "grad_norm": 2.1708011968731284, + "language_loss": 0.78742027, + "learning_rate": 8.038553826419494e-07, + "loss": 0.80899024, + "num_input_tokens_seen": 127446475, + "step": 5929, + "time_per_iteration": 2.991718292236328 + }, + { + "auxiliary_loss_clip": 0.01167657, + "auxiliary_loss_mlp": 0.01027056, + "balance_loss_clip": 1.04547489, + "balance_loss_mlp": 1.01928949, + "epoch": 0.7130403414898094, + "flos": 21397445326080.0, + "grad_norm": 1.8078360420549455, + "language_loss": 0.81269658, + "learning_rate": 8.032311726760364e-07, + "loss": 0.83464372, + "num_input_tokens_seen": 127467695, + "step": 5930, + "time_per_iteration": 2.8211801052093506 + }, + { + "auxiliary_loss_clip": 0.01133793, + "auxiliary_loss_mlp": 0.01026961, + "balance_loss_clip": 1.04353166, + "balance_loss_mlp": 1.01833034, + "epoch": 0.7131605843804485, + "flos": 74739045306240.0, + "grad_norm": 1.765633523899542, + "language_loss": 0.68960351, + "learning_rate": 8.026071442590022e-07, + "loss": 0.71121103, + "num_input_tokens_seen": 127494590, + "step": 5931, + "time_per_iteration": 3.167409896850586 + }, + { + "auxiliary_loss_clip": 0.0116135, + "auxiliary_loss_mlp": 0.010302, + "balance_loss_clip": 1.04943407, + "balance_loss_mlp": 1.02218866, + "epoch": 0.7132808272710875, + "flos": 18368739469440.0, + "grad_norm": 1.7670822352174294, + "language_loss": 0.8087706, + "learning_rate": 8.019832974855134e-07, + "loss": 0.83068609, + "num_input_tokens_seen": 127512550, + "step": 5932, + "time_per_iteration": 3.588709831237793 + }, + { + "auxiliary_loss_clip": 0.01140566, + "auxiliary_loss_mlp": 0.01031172, + "balance_loss_clip": 1.04513848, + "balance_loss_mlp": 1.0235163, + "epoch": 0.7134010701617267, + "flos": 23253380127360.0, + "grad_norm": 2.226227595320668, + "language_loss": 0.82600862, + "learning_rate": 8.013596324502052e-07, + "loss": 0.84772611, + "num_input_tokens_seen": 127531015, + "step": 5933, + "time_per_iteration": 2.866131544113159 + }, + { + "auxiliary_loss_clip": 0.01155125, + "auxiliary_loss_mlp": 0.01029735, + "balance_loss_clip": 1.04713809, + "balance_loss_mlp": 1.02205801, + "epoch": 0.7135213130523658, + "flos": 23653137565440.0, + "grad_norm": 1.8289770427207588, + "language_loss": 0.78374177, + "learning_rate": 8.007361492476872e-07, + "loss": 0.80559033, + "num_input_tokens_seen": 127550340, + "step": 5934, + "time_per_iteration": 2.7555577754974365 + }, + { + "auxiliary_loss_clip": 0.01150614, + "auxiliary_loss_mlp": 0.01035159, + "balance_loss_clip": 1.04371333, + "balance_loss_mlp": 1.02718127, + "epoch": 0.7136415559430048, + "flos": 24790644443520.0, + "grad_norm": 1.5271875756920292, + "language_loss": 0.7887857, + "learning_rate": 8.001128479725426e-07, + "loss": 0.81064343, + "num_input_tokens_seen": 127572245, + "step": 5935, + "time_per_iteration": 3.811380624771118 + }, + { + "auxiliary_loss_clip": 0.01124567, + "auxiliary_loss_mlp": 0.01032483, + "balance_loss_clip": 1.03850412, + "balance_loss_mlp": 1.02482116, + "epoch": 0.713761798833644, + "flos": 18296954138880.0, + "grad_norm": 1.846014217825516, + "language_loss": 0.80872327, + "learning_rate": 7.994897287193248e-07, + "loss": 0.83029377, + "num_input_tokens_seen": 127591625, + "step": 5936, + "time_per_iteration": 2.8348000049591064 + }, + { + "auxiliary_loss_clip": 0.01166569, + "auxiliary_loss_mlp": 0.01027079, + "balance_loss_clip": 1.04840279, + "balance_loss_mlp": 1.01903796, + "epoch": 0.713882041724283, + "flos": 15558262692480.0, + "grad_norm": 2.377111619691094, + "language_loss": 0.83564228, + "learning_rate": 7.988667915825605e-07, + "loss": 0.85757875, + "num_input_tokens_seen": 127608690, + "step": 5937, + "time_per_iteration": 2.706411600112915 + }, + { + "auxiliary_loss_clip": 0.01153656, + "auxiliary_loss_mlp": 0.01028395, + "balance_loss_clip": 1.04713082, + "balance_loss_mlp": 1.02028906, + "epoch": 0.7140022846149221, + "flos": 24061011477120.0, + "grad_norm": 2.571474907365754, + "language_loss": 0.75274515, + "learning_rate": 7.982440366567491e-07, + "loss": 0.77456564, + "num_input_tokens_seen": 127627180, + "step": 5938, + "time_per_iteration": 2.7345938682556152 + }, + { + "auxiliary_loss_clip": 0.01154943, + "auxiliary_loss_mlp": 0.01023471, + "balance_loss_clip": 1.04436624, + "balance_loss_mlp": 1.01584136, + "epoch": 0.7141225275055613, + "flos": 27891710248320.0, + "grad_norm": 1.6059074875476687, + "language_loss": 0.75185287, + "learning_rate": 7.97621464036361e-07, + "loss": 0.773637, + "num_input_tokens_seen": 127648940, + "step": 5939, + "time_per_iteration": 3.6490108966827393 + }, + { + "auxiliary_loss_clip": 0.01165807, + "auxiliary_loss_mlp": 0.01025132, + "balance_loss_clip": 1.04799414, + "balance_loss_mlp": 1.0171752, + "epoch": 0.7142427703962003, + "flos": 19682603147520.0, + "grad_norm": 1.6528414460784702, + "language_loss": 0.68231285, + "learning_rate": 7.969990738158417e-07, + "loss": 0.70422226, + "num_input_tokens_seen": 127667350, + "step": 5940, + "time_per_iteration": 2.7546443939208984 + }, + { + "auxiliary_loss_clip": 0.01161615, + "auxiliary_loss_mlp": 0.0103488, + "balance_loss_clip": 1.04738688, + "balance_loss_mlp": 1.026649, + "epoch": 0.7143630132868394, + "flos": 21032377447680.0, + "grad_norm": 2.1111171974674057, + "language_loss": 0.85235173, + "learning_rate": 7.963768660896062e-07, + "loss": 0.87431669, + "num_input_tokens_seen": 127685760, + "step": 5941, + "time_per_iteration": 2.7591915130615234 + }, + { + "auxiliary_loss_clip": 0.01164255, + "auxiliary_loss_mlp": 0.01025903, + "balance_loss_clip": 1.04669917, + "balance_loss_mlp": 1.01812434, + "epoch": 0.7144832561774785, + "flos": 24129923719680.0, + "grad_norm": 4.7284407160544095, + "language_loss": 0.82431889, + "learning_rate": 7.957548409520432e-07, + "loss": 0.84622049, + "num_input_tokens_seen": 127704985, + "step": 5942, + "time_per_iteration": 2.872976779937744 + }, + { + "auxiliary_loss_clip": 0.01142236, + "auxiliary_loss_mlp": 0.01030081, + "balance_loss_clip": 1.04229999, + "balance_loss_mlp": 1.02178431, + "epoch": 0.7146034990681176, + "flos": 16325817442560.0, + "grad_norm": 1.8987264557129888, + "language_loss": 0.83950448, + "learning_rate": 7.951329984975135e-07, + "loss": 0.86122763, + "num_input_tokens_seen": 127721925, + "step": 5943, + "time_per_iteration": 2.7125465869903564 + }, + { + "auxiliary_loss_clip": 0.01040749, + "auxiliary_loss_mlp": 0.01003623, + "balance_loss_clip": 1.008986, + "balance_loss_mlp": 1.00192988, + "epoch": 0.7147237419587567, + "flos": 69627164232960.0, + "grad_norm": 0.715593114440303, + "language_loss": 0.54273832, + "learning_rate": 7.94511338820349e-07, + "loss": 0.56318212, + "num_input_tokens_seen": 127784230, + "step": 5944, + "time_per_iteration": 3.3184683322906494 + }, + { + "auxiliary_loss_clip": 0.01153568, + "auxiliary_loss_mlp": 0.00885908, + "balance_loss_clip": 1.04680181, + "balance_loss_mlp": 1.00074303, + "epoch": 0.7148439848493958, + "flos": 22266806198400.0, + "grad_norm": 2.465740504261866, + "language_loss": 0.78402627, + "learning_rate": 7.938898620148575e-07, + "loss": 0.80442101, + "num_input_tokens_seen": 127801990, + "step": 5945, + "time_per_iteration": 2.7417070865631104 + }, + { + "auxiliary_loss_clip": 0.01152985, + "auxiliary_loss_mlp": 0.01026209, + "balance_loss_clip": 1.04784584, + "balance_loss_mlp": 1.01857972, + "epoch": 0.7149642277400349, + "flos": 17931383470080.0, + "grad_norm": 1.9024927860396563, + "language_loss": 0.70881283, + "learning_rate": 7.932685681753135e-07, + "loss": 0.73060477, + "num_input_tokens_seen": 127819270, + "step": 5946, + "time_per_iteration": 2.7553296089172363 + }, + { + "auxiliary_loss_clip": 0.01167282, + "auxiliary_loss_mlp": 0.01032877, + "balance_loss_clip": 1.04780948, + "balance_loss_mlp": 1.02476525, + "epoch": 0.7150844706306739, + "flos": 31681937370240.0, + "grad_norm": 1.9088670272933543, + "language_loss": 0.62851495, + "learning_rate": 7.92647457395969e-07, + "loss": 0.65051651, + "num_input_tokens_seen": 127841095, + "step": 5947, + "time_per_iteration": 2.715003728866577 + }, + { + "auxiliary_loss_clip": 0.01127862, + "auxiliary_loss_mlp": 0.01029706, + "balance_loss_clip": 1.03719175, + "balance_loss_mlp": 1.02054489, + "epoch": 0.7152047135213131, + "flos": 10926217451520.0, + "grad_norm": 2.621236800714716, + "language_loss": 0.74275964, + "learning_rate": 7.920265297710444e-07, + "loss": 0.76433533, + "num_input_tokens_seen": 127858485, + "step": 5948, + "time_per_iteration": 2.8462345600128174 + }, + { + "auxiliary_loss_clip": 0.01164647, + "auxiliary_loss_mlp": 0.01025533, + "balance_loss_clip": 1.04771638, + "balance_loss_mlp": 1.01863647, + "epoch": 0.7153249564119522, + "flos": 20995640812800.0, + "grad_norm": 2.5256927289953772, + "language_loss": 0.73461276, + "learning_rate": 7.914057853947363e-07, + "loss": 0.75651455, + "num_input_tokens_seen": 127877665, + "step": 5949, + "time_per_iteration": 2.806689977645874 + }, + { + "auxiliary_loss_clip": 0.01138395, + "auxiliary_loss_mlp": 0.010314, + "balance_loss_clip": 1.04261374, + "balance_loss_mlp": 1.02285242, + "epoch": 0.7154451993025912, + "flos": 24243114453120.0, + "grad_norm": 1.8686347321264976, + "language_loss": 0.62672591, + "learning_rate": 7.907852243612089e-07, + "loss": 0.64842379, + "num_input_tokens_seen": 127898070, + "step": 5950, + "time_per_iteration": 2.9029979705810547 + }, + { + "auxiliary_loss_clip": 0.01148387, + "auxiliary_loss_mlp": 0.01026629, + "balance_loss_clip": 1.04386497, + "balance_loss_mlp": 1.01854086, + "epoch": 0.7155654421932304, + "flos": 23330947547520.0, + "grad_norm": 1.881275625427491, + "language_loss": 0.72597343, + "learning_rate": 7.901648467646009e-07, + "loss": 0.74772358, + "num_input_tokens_seen": 127917010, + "step": 5951, + "time_per_iteration": 2.8006949424743652 + }, + { + "auxiliary_loss_clip": 0.01172511, + "auxiliary_loss_mlp": 0.01030014, + "balance_loss_clip": 1.04883242, + "balance_loss_mlp": 1.02177989, + "epoch": 0.7156856850838694, + "flos": 22711883621760.0, + "grad_norm": 2.12200210707876, + "language_loss": 0.72396874, + "learning_rate": 7.895446526990244e-07, + "loss": 0.74599409, + "num_input_tokens_seen": 127937025, + "step": 5952, + "time_per_iteration": 2.6483547687530518 + }, + { + "auxiliary_loss_clip": 0.0114265, + "auxiliary_loss_mlp": 0.01031842, + "balance_loss_clip": 1.04414594, + "balance_loss_mlp": 1.023664, + "epoch": 0.7158059279745085, + "flos": 19865424395520.0, + "grad_norm": 1.6446904384066592, + "language_loss": 0.7565183, + "learning_rate": 7.889246422585609e-07, + "loss": 0.77826321, + "num_input_tokens_seen": 127956410, + "step": 5953, + "time_per_iteration": 2.854036331176758 + }, + { + "auxiliary_loss_clip": 0.01171587, + "auxiliary_loss_mlp": 0.01026008, + "balance_loss_clip": 1.04863608, + "balance_loss_mlp": 1.01807427, + "epoch": 0.7159261708651476, + "flos": 24134772055680.0, + "grad_norm": 1.898797910552329, + "language_loss": 0.73561263, + "learning_rate": 7.883048155372675e-07, + "loss": 0.75758851, + "num_input_tokens_seen": 127974925, + "step": 5954, + "time_per_iteration": 2.6741976737976074 + }, + { + "auxiliary_loss_clip": 0.01155906, + "auxiliary_loss_mlp": 0.01027369, + "balance_loss_clip": 1.04526472, + "balance_loss_mlp": 1.01948309, + "epoch": 0.7160464137557867, + "flos": 16983198201600.0, + "grad_norm": 2.3559957250232513, + "language_loss": 0.71204841, + "learning_rate": 7.876851726291698e-07, + "loss": 0.73388112, + "num_input_tokens_seen": 127993225, + "step": 5955, + "time_per_iteration": 2.8168272972106934 + }, + { + "auxiliary_loss_clip": 0.01145665, + "auxiliary_loss_mlp": 0.01029897, + "balance_loss_clip": 1.04253542, + "balance_loss_mlp": 1.02236319, + "epoch": 0.7161666566464258, + "flos": 25228251838080.0, + "grad_norm": 3.6739974030926676, + "language_loss": 0.78589517, + "learning_rate": 7.870657136282666e-07, + "loss": 0.8076508, + "num_input_tokens_seen": 128012085, + "step": 5956, + "time_per_iteration": 2.8500914573669434 + }, + { + "auxiliary_loss_clip": 0.01156964, + "auxiliary_loss_mlp": 0.01029278, + "balance_loss_clip": 1.04542994, + "balance_loss_mlp": 1.02135658, + "epoch": 0.7162868995370649, + "flos": 26468390851200.0, + "grad_norm": 1.5506090325647457, + "language_loss": 0.82217526, + "learning_rate": 7.86446438628531e-07, + "loss": 0.84403765, + "num_input_tokens_seen": 128033155, + "step": 5957, + "time_per_iteration": 2.7505617141723633 + }, + { + "auxiliary_loss_clip": 0.01061959, + "auxiliary_loss_mlp": 0.01003522, + "balance_loss_clip": 1.00929713, + "balance_loss_mlp": 1.00206172, + "epoch": 0.716407142427704, + "flos": 69998912040960.0, + "grad_norm": 0.767059985791482, + "language_loss": 0.56904244, + "learning_rate": 7.858273477239059e-07, + "loss": 0.58969724, + "num_input_tokens_seen": 128101575, + "step": 5958, + "time_per_iteration": 4.230075836181641 + }, + { + "auxiliary_loss_clip": 0.01121231, + "auxiliary_loss_mlp": 0.01024303, + "balance_loss_clip": 1.04082072, + "balance_loss_mlp": 1.01604819, + "epoch": 0.716527385318343, + "flos": 20740459616640.0, + "grad_norm": 1.5827603257836405, + "language_loss": 0.71122277, + "learning_rate": 7.852084410083067e-07, + "loss": 0.73267806, + "num_input_tokens_seen": 128120395, + "step": 5959, + "time_per_iteration": 3.8220794200897217 + }, + { + "auxiliary_loss_clip": 0.01147084, + "auxiliary_loss_mlp": 0.01026502, + "balance_loss_clip": 1.04530966, + "balance_loss_mlp": 1.01964164, + "epoch": 0.7166476282089821, + "flos": 25371966153600.0, + "grad_norm": 1.8386337274995859, + "language_loss": 0.63733351, + "learning_rate": 7.84589718575621e-07, + "loss": 0.65906942, + "num_input_tokens_seen": 128140840, + "step": 5960, + "time_per_iteration": 2.7362406253814697 + }, + { + "auxiliary_loss_clip": 0.01151336, + "auxiliary_loss_mlp": 0.01028826, + "balance_loss_clip": 1.04137158, + "balance_loss_mlp": 1.02065969, + "epoch": 0.7167678710996213, + "flos": 24133730561280.0, + "grad_norm": 1.914016617053417, + "language_loss": 0.68801457, + "learning_rate": 7.83971180519708e-07, + "loss": 0.70981616, + "num_input_tokens_seen": 128159695, + "step": 5961, + "time_per_iteration": 3.610443353652954 + }, + { + "auxiliary_loss_clip": 0.01173689, + "auxiliary_loss_mlp": 0.01027161, + "balance_loss_clip": 1.04902077, + "balance_loss_mlp": 1.01881075, + "epoch": 0.7168881139902603, + "flos": 30226586019840.0, + "grad_norm": 2.2266764771421372, + "language_loss": 0.75654536, + "learning_rate": 7.833528269344008e-07, + "loss": 0.7785539, + "num_input_tokens_seen": 128179600, + "step": 5962, + "time_per_iteration": 2.7890915870666504 + }, + { + "auxiliary_loss_clip": 0.01143032, + "auxiliary_loss_mlp": 0.01032882, + "balance_loss_clip": 1.04668283, + "balance_loss_mlp": 1.02487135, + "epoch": 0.7170083568808994, + "flos": 14606414236800.0, + "grad_norm": 2.0370457523392633, + "language_loss": 0.77782422, + "learning_rate": 7.827346579135023e-07, + "loss": 0.79958338, + "num_input_tokens_seen": 128196940, + "step": 5963, + "time_per_iteration": 2.8691468238830566 + }, + { + "auxiliary_loss_clip": 0.01150805, + "auxiliary_loss_mlp": 0.01031633, + "balance_loss_clip": 1.04603171, + "balance_loss_mlp": 1.0234313, + "epoch": 0.7171285997715385, + "flos": 23331091201920.0, + "grad_norm": 1.9128104093817528, + "language_loss": 0.83116078, + "learning_rate": 7.821166735507885e-07, + "loss": 0.85298508, + "num_input_tokens_seen": 128215970, + "step": 5964, + "time_per_iteration": 2.900036096572876 + }, + { + "auxiliary_loss_clip": 0.01169157, + "auxiliary_loss_mlp": 0.01025493, + "balance_loss_clip": 1.04813874, + "balance_loss_mlp": 1.01772618, + "epoch": 0.7172488426621776, + "flos": 16543543731840.0, + "grad_norm": 1.7270731613976167, + "language_loss": 0.68244636, + "learning_rate": 7.81498873940007e-07, + "loss": 0.70439285, + "num_input_tokens_seen": 128233185, + "step": 5965, + "time_per_iteration": 3.6160192489624023 + }, + { + "auxiliary_loss_clip": 0.01166683, + "auxiliary_loss_mlp": 0.01026559, + "balance_loss_clip": 1.04549825, + "balance_loss_mlp": 1.01820803, + "epoch": 0.7173690855528166, + "flos": 26541612725760.0, + "grad_norm": 2.0177380510139518, + "language_loss": 0.77050936, + "learning_rate": 7.808812591748768e-07, + "loss": 0.79244173, + "num_input_tokens_seen": 128253565, + "step": 5966, + "time_per_iteration": 2.882654905319214 + }, + { + "auxiliary_loss_clip": 0.01140835, + "auxiliary_loss_mlp": 0.01032897, + "balance_loss_clip": 1.04463589, + "balance_loss_mlp": 1.02400386, + "epoch": 0.7174893284434558, + "flos": 22784099915520.0, + "grad_norm": 1.9558944552641293, + "language_loss": 0.65063304, + "learning_rate": 7.802638293490915e-07, + "loss": 0.67237037, + "num_input_tokens_seen": 128273210, + "step": 5967, + "time_per_iteration": 2.8393445014953613 + }, + { + "auxiliary_loss_clip": 0.01155845, + "auxiliary_loss_mlp": 0.01036168, + "balance_loss_clip": 1.0454824, + "balance_loss_mlp": 1.02790117, + "epoch": 0.7176095713340949, + "flos": 23293564467840.0, + "grad_norm": 1.7291356151825883, + "language_loss": 0.76964724, + "learning_rate": 7.796465845563123e-07, + "loss": 0.79156733, + "num_input_tokens_seen": 128292085, + "step": 5968, + "time_per_iteration": 2.8148319721221924 + }, + { + "auxiliary_loss_clip": 0.01147388, + "auxiliary_loss_mlp": 0.00885889, + "balance_loss_clip": 1.04365098, + "balance_loss_mlp": 1.00089884, + "epoch": 0.7177298142247339, + "flos": 25591631777280.0, + "grad_norm": 1.9149791922130421, + "language_loss": 0.79581416, + "learning_rate": 7.790295248901766e-07, + "loss": 0.81614697, + "num_input_tokens_seen": 128313215, + "step": 5969, + "time_per_iteration": 2.779865264892578 + }, + { + "auxiliary_loss_clip": 0.01160191, + "auxiliary_loss_mlp": 0.01035881, + "balance_loss_clip": 1.04619098, + "balance_loss_mlp": 1.02820969, + "epoch": 0.7178500571153731, + "flos": 31652778504960.0, + "grad_norm": 1.8899996890651651, + "language_loss": 0.62336028, + "learning_rate": 7.784126504442902e-07, + "loss": 0.64532101, + "num_input_tokens_seen": 128336445, + "step": 5970, + "time_per_iteration": 2.838789463043213 + }, + { + "auxiliary_loss_clip": 0.01132574, + "auxiliary_loss_mlp": 0.01024703, + "balance_loss_clip": 1.04205287, + "balance_loss_mlp": 1.01696026, + "epoch": 0.7179703000060121, + "flos": 19427242383360.0, + "grad_norm": 1.3346662328358854, + "language_loss": 0.67886519, + "learning_rate": 7.777959613122351e-07, + "loss": 0.70043802, + "num_input_tokens_seen": 128356270, + "step": 5971, + "time_per_iteration": 2.8613269329071045 + }, + { + "auxiliary_loss_clip": 0.01147655, + "auxiliary_loss_mlp": 0.01028561, + "balance_loss_clip": 1.04688573, + "balance_loss_mlp": 1.02061582, + "epoch": 0.7180905428966512, + "flos": 28839249072000.0, + "grad_norm": 1.82779362221984, + "language_loss": 0.77959812, + "learning_rate": 7.771794575875604e-07, + "loss": 0.80136025, + "num_input_tokens_seen": 128378140, + "step": 5972, + "time_per_iteration": 2.8608076572418213 + }, + { + "auxiliary_loss_clip": 0.01163195, + "auxiliary_loss_mlp": 0.01033852, + "balance_loss_clip": 1.05007148, + "balance_loss_mlp": 1.02540016, + "epoch": 0.7182107857872904, + "flos": 20047563285120.0, + "grad_norm": 2.570556157784129, + "language_loss": 0.77752113, + "learning_rate": 7.765631393637888e-07, + "loss": 0.79949158, + "num_input_tokens_seen": 128396335, + "step": 5973, + "time_per_iteration": 2.8466336727142334 + }, + { + "auxiliary_loss_clip": 0.01156693, + "auxiliary_loss_mlp": 0.01034983, + "balance_loss_clip": 1.04462135, + "balance_loss_mlp": 1.02622676, + "epoch": 0.7183310286779294, + "flos": 22747686503040.0, + "grad_norm": 5.473885993757213, + "language_loss": 0.48677847, + "learning_rate": 7.75947006734417e-07, + "loss": 0.50869524, + "num_input_tokens_seen": 128414115, + "step": 5974, + "time_per_iteration": 2.759573221206665 + }, + { + "auxiliary_loss_clip": 0.01172253, + "auxiliary_loss_mlp": 0.01035008, + "balance_loss_clip": 1.0478512, + "balance_loss_mlp": 1.02618039, + "epoch": 0.7184512715685685, + "flos": 17158262112000.0, + "grad_norm": 2.1872077471686038, + "language_loss": 0.82566106, + "learning_rate": 7.753310597929101e-07, + "loss": 0.84773368, + "num_input_tokens_seen": 128430755, + "step": 5975, + "time_per_iteration": 2.889431953430176 + }, + { + "auxiliary_loss_clip": 0.01062356, + "auxiliary_loss_mlp": 0.01001653, + "balance_loss_clip": 1.00975478, + "balance_loss_mlp": 1.00007939, + "epoch": 0.7185715144592076, + "flos": 65509611448320.0, + "grad_norm": 0.7516652007059564, + "language_loss": 0.55088437, + "learning_rate": 7.747152986327095e-07, + "loss": 0.57152444, + "num_input_tokens_seen": 128491300, + "step": 5976, + "time_per_iteration": 3.2016820907592773 + }, + { + "auxiliary_loss_clip": 0.0113822, + "auxiliary_loss_mlp": 0.01027594, + "balance_loss_clip": 1.04314637, + "balance_loss_mlp": 1.01980376, + "epoch": 0.7186917573498467, + "flos": 16180522928640.0, + "grad_norm": 2.2295534694780312, + "language_loss": 0.68303394, + "learning_rate": 7.740997233472228e-07, + "loss": 0.70469201, + "num_input_tokens_seen": 128508920, + "step": 5977, + "time_per_iteration": 3.034677028656006 + }, + { + "auxiliary_loss_clip": 0.01148202, + "auxiliary_loss_mlp": 0.01025218, + "balance_loss_clip": 1.0422945, + "balance_loss_mlp": 1.01807427, + "epoch": 0.7188120002404857, + "flos": 29242274647680.0, + "grad_norm": 2.3172456898232237, + "language_loss": 0.70349395, + "learning_rate": 7.734843340298329e-07, + "loss": 0.72522819, + "num_input_tokens_seen": 128528745, + "step": 5978, + "time_per_iteration": 2.933091878890991 + }, + { + "auxiliary_loss_clip": 0.01158498, + "auxiliary_loss_mlp": 0.01034568, + "balance_loss_clip": 1.04455566, + "balance_loss_mlp": 1.02635193, + "epoch": 0.7189322431311249, + "flos": 33401161008000.0, + "grad_norm": 2.3301974798971417, + "language_loss": 0.74824798, + "learning_rate": 7.72869130773895e-07, + "loss": 0.77017862, + "num_input_tokens_seen": 128549345, + "step": 5979, + "time_per_iteration": 2.972029447555542 + }, + { + "auxiliary_loss_clip": 0.01057662, + "auxiliary_loss_mlp": 0.01001529, + "balance_loss_clip": 1.00910926, + "balance_loss_mlp": 1.0000155, + "epoch": 0.719052486021764, + "flos": 61351263792000.0, + "grad_norm": 0.8007689088311614, + "language_loss": 0.59406805, + "learning_rate": 7.722541136727343e-07, + "loss": 0.61465997, + "num_input_tokens_seen": 128605360, + "step": 5980, + "time_per_iteration": 3.128570318222046 + }, + { + "auxiliary_loss_clip": 0.01162984, + "auxiliary_loss_mlp": 0.01031917, + "balance_loss_clip": 1.04903781, + "balance_loss_mlp": 1.02414417, + "epoch": 0.719172728912403, + "flos": 15596795007360.0, + "grad_norm": 1.9095861100142517, + "language_loss": 0.80446875, + "learning_rate": 7.716392828196483e-07, + "loss": 0.82641774, + "num_input_tokens_seen": 128623160, + "step": 5981, + "time_per_iteration": 2.7240707874298096 + }, + { + "auxiliary_loss_clip": 0.01161148, + "auxiliary_loss_mlp": 0.01028497, + "balance_loss_clip": 1.04663837, + "balance_loss_mlp": 1.02036655, + "epoch": 0.7192929718030422, + "flos": 15553162961280.0, + "grad_norm": 2.4238339189310047, + "language_loss": 0.77381247, + "learning_rate": 7.710246383079064e-07, + "loss": 0.79570895, + "num_input_tokens_seen": 128638545, + "step": 5982, + "time_per_iteration": 2.703415870666504 + }, + { + "auxiliary_loss_clip": 0.0115572, + "auxiliary_loss_mlp": 0.01028197, + "balance_loss_clip": 1.04526305, + "balance_loss_mlp": 1.02026403, + "epoch": 0.7194132146936812, + "flos": 21862487733120.0, + "grad_norm": 3.121446881430223, + "language_loss": 0.91933453, + "learning_rate": 7.704101802307492e-07, + "loss": 0.94117373, + "num_input_tokens_seen": 128650845, + "step": 5983, + "time_per_iteration": 2.7121875286102295 + }, + { + "auxiliary_loss_clip": 0.01134162, + "auxiliary_loss_mlp": 0.01025082, + "balance_loss_clip": 1.04201484, + "balance_loss_mlp": 1.01686299, + "epoch": 0.7195334575843203, + "flos": 27338900958720.0, + "grad_norm": 2.3351296575912066, + "language_loss": 0.87198508, + "learning_rate": 7.697959086813912e-07, + "loss": 0.89357752, + "num_input_tokens_seen": 128667010, + "step": 5984, + "time_per_iteration": 4.681714296340942 + }, + { + "auxiliary_loss_clip": 0.01133117, + "auxiliary_loss_mlp": 0.01023951, + "balance_loss_clip": 1.04042721, + "balance_loss_mlp": 1.01639342, + "epoch": 0.7196537004749595, + "flos": 18770615809920.0, + "grad_norm": 1.5844933296996853, + "language_loss": 0.80235487, + "learning_rate": 7.691818237530145e-07, + "loss": 0.82392561, + "num_input_tokens_seen": 128685870, + "step": 5985, + "time_per_iteration": 2.85807204246521 + }, + { + "auxiliary_loss_clip": 0.01143775, + "auxiliary_loss_mlp": 0.0102863, + "balance_loss_clip": 1.04242611, + "balance_loss_mlp": 1.02094698, + "epoch": 0.7197739433655985, + "flos": 24531009960960.0, + "grad_norm": 5.183436133740082, + "language_loss": 0.77732396, + "learning_rate": 7.685679255387774e-07, + "loss": 0.79904801, + "num_input_tokens_seen": 128704185, + "step": 5986, + "time_per_iteration": 2.8916306495666504 + }, + { + "auxiliary_loss_clip": 0.01148569, + "auxiliary_loss_mlp": 0.01029282, + "balance_loss_clip": 1.04636884, + "balance_loss_mlp": 1.02150989, + "epoch": 0.7198941862562376, + "flos": 18040587793920.0, + "grad_norm": 1.9637847489577775, + "language_loss": 0.76663733, + "learning_rate": 7.679542141318065e-07, + "loss": 0.78841579, + "num_input_tokens_seen": 128721290, + "step": 5987, + "time_per_iteration": 3.66526460647583 + }, + { + "auxiliary_loss_clip": 0.01140431, + "auxiliary_loss_mlp": 0.01025934, + "balance_loss_clip": 1.0427165, + "balance_loss_mlp": 1.01810801, + "epoch": 0.7200144291468767, + "flos": 29022393542400.0, + "grad_norm": 1.7015216170846406, + "language_loss": 0.75515157, + "learning_rate": 7.673406896252013e-07, + "loss": 0.77681524, + "num_input_tokens_seen": 128742665, + "step": 5988, + "time_per_iteration": 2.908595323562622 + }, + { + "auxiliary_loss_clip": 0.01140527, + "auxiliary_loss_mlp": 0.01031586, + "balance_loss_clip": 1.04206014, + "balance_loss_mlp": 1.02269888, + "epoch": 0.7201346720375158, + "flos": 25374264624000.0, + "grad_norm": 1.5352802004033799, + "language_loss": 0.77998167, + "learning_rate": 7.667273521120347e-07, + "loss": 0.80170274, + "num_input_tokens_seen": 128762225, + "step": 5989, + "time_per_iteration": 3.024871587753296 + }, + { + "auxiliary_loss_clip": 0.01145143, + "auxiliary_loss_mlp": 0.0102973, + "balance_loss_clip": 1.04412365, + "balance_loss_mlp": 1.02146292, + "epoch": 0.7202549149281549, + "flos": 14355614499840.0, + "grad_norm": 1.9241345985661629, + "language_loss": 0.79570711, + "learning_rate": 7.661142016853468e-07, + "loss": 0.81745577, + "num_input_tokens_seen": 128779585, + "step": 5990, + "time_per_iteration": 2.816570997238159 + }, + { + "auxiliary_loss_clip": 0.01130792, + "auxiliary_loss_mlp": 0.01029264, + "balance_loss_clip": 1.04314649, + "balance_loss_mlp": 1.02102637, + "epoch": 0.7203751578187939, + "flos": 23001682550400.0, + "grad_norm": 1.8536344828900464, + "language_loss": 0.75115037, + "learning_rate": 7.655012384381543e-07, + "loss": 0.77275097, + "num_input_tokens_seen": 128799070, + "step": 5991, + "time_per_iteration": 3.7714390754699707 + }, + { + "auxiliary_loss_clip": 0.01149934, + "auxiliary_loss_mlp": 0.0102945, + "balance_loss_clip": 1.0490005, + "balance_loss_mlp": 1.021451, + "epoch": 0.7204954007094331, + "flos": 23692424065920.0, + "grad_norm": 2.077988288569174, + "language_loss": 0.81761754, + "learning_rate": 7.648884624634415e-07, + "loss": 0.83941138, + "num_input_tokens_seen": 128817620, + "step": 5992, + "time_per_iteration": 2.7651584148406982 + }, + { + "auxiliary_loss_clip": 0.0116095, + "auxiliary_loss_mlp": 0.01033162, + "balance_loss_clip": 1.04785287, + "balance_loss_mlp": 1.02489448, + "epoch": 0.7206156436000721, + "flos": 16253026531200.0, + "grad_norm": 1.9212024333131708, + "language_loss": 0.88731831, + "learning_rate": 7.642758738541683e-07, + "loss": 0.90925944, + "num_input_tokens_seen": 128834200, + "step": 5993, + "time_per_iteration": 2.6692755222320557 + }, + { + "auxiliary_loss_clip": 0.01056458, + "auxiliary_loss_mlp": 0.01001919, + "balance_loss_clip": 1.00957322, + "balance_loss_mlp": 1.00039911, + "epoch": 0.7207358864907112, + "flos": 54377806504320.0, + "grad_norm": 0.7621171796323821, + "language_loss": 0.60746098, + "learning_rate": 7.636634727032621e-07, + "loss": 0.62804472, + "num_input_tokens_seen": 128891305, + "step": 5994, + "time_per_iteration": 3.166229009628296 + }, + { + "auxiliary_loss_clip": 0.0114233, + "auxiliary_loss_mlp": 0.01027448, + "balance_loss_clip": 1.03935337, + "balance_loss_mlp": 1.01887059, + "epoch": 0.7208561293813504, + "flos": 19135540033920.0, + "grad_norm": 1.9922379247248745, + "language_loss": 0.79005623, + "learning_rate": 7.630512591036231e-07, + "loss": 0.81175405, + "num_input_tokens_seen": 128910615, + "step": 5995, + "time_per_iteration": 2.755199670791626 + }, + { + "auxiliary_loss_clip": 0.01162245, + "auxiliary_loss_mlp": 0.01026468, + "balance_loss_clip": 1.04652405, + "balance_loss_mlp": 1.01766419, + "epoch": 0.7209763722719894, + "flos": 17748526308480.0, + "grad_norm": 2.2978717691113935, + "language_loss": 0.64661658, + "learning_rate": 7.624392331481255e-07, + "loss": 0.6685037, + "num_input_tokens_seen": 128928270, + "step": 5996, + "time_per_iteration": 2.704942464828491 + }, + { + "auxiliary_loss_clip": 0.01053859, + "auxiliary_loss_mlp": 0.01001707, + "balance_loss_clip": 1.00888991, + "balance_loss_mlp": 1.00013375, + "epoch": 0.7210966151626285, + "flos": 66819488716800.0, + "grad_norm": 0.7427124981565238, + "language_loss": 0.51813781, + "learning_rate": 7.618273949296115e-07, + "loss": 0.53869343, + "num_input_tokens_seen": 128987780, + "step": 5997, + "time_per_iteration": 3.1373324394226074 + }, + { + "auxiliary_loss_clip": 0.01146924, + "auxiliary_loss_mlp": 0.01027195, + "balance_loss_clip": 1.04284787, + "balance_loss_mlp": 1.0187732, + "epoch": 0.7212168580532676, + "flos": 21141869080320.0, + "grad_norm": 2.026500314203717, + "language_loss": 0.68474728, + "learning_rate": 7.612157445408987e-07, + "loss": 0.70648849, + "num_input_tokens_seen": 129005590, + "step": 5998, + "time_per_iteration": 2.819648265838623 + }, + { + "auxiliary_loss_clip": 0.01158522, + "auxiliary_loss_mlp": 0.01025731, + "balance_loss_clip": 1.05031419, + "balance_loss_mlp": 1.01811397, + "epoch": 0.7213371009439067, + "flos": 22345738335360.0, + "grad_norm": 4.966864578740766, + "language_loss": 0.74305898, + "learning_rate": 7.606042820747716e-07, + "loss": 0.76490146, + "num_input_tokens_seen": 129021995, + "step": 5999, + "time_per_iteration": 2.751469850540161 + }, + { + "auxiliary_loss_clip": 0.0116166, + "auxiliary_loss_mlp": 0.01036884, + "balance_loss_clip": 1.05060971, + "balance_loss_mlp": 1.02864087, + "epoch": 0.7214573438345457, + "flos": 18515901490560.0, + "grad_norm": 1.7584615811369257, + "language_loss": 0.85797083, + "learning_rate": 7.599930076239889e-07, + "loss": 0.87995625, + "num_input_tokens_seen": 129039280, + "step": 6000, + "time_per_iteration": 2.791428327560425 + }, + { + "auxiliary_loss_clip": 0.01138113, + "auxiliary_loss_mlp": 0.00886362, + "balance_loss_clip": 1.04574466, + "balance_loss_mlp": 1.00075912, + "epoch": 0.7215775867251849, + "flos": 35736108606720.0, + "grad_norm": 2.163011864330132, + "language_loss": 0.70687568, + "learning_rate": 7.593819212812818e-07, + "loss": 0.7271204, + "num_input_tokens_seen": 129060860, + "step": 6001, + "time_per_iteration": 3.033396005630493 + }, + { + "auxiliary_loss_clip": 0.01160266, + "auxiliary_loss_mlp": 0.01036515, + "balance_loss_clip": 1.04744923, + "balance_loss_mlp": 1.0283668, + "epoch": 0.721697829615824, + "flos": 20372410909440.0, + "grad_norm": 2.3161451689184074, + "language_loss": 0.71623904, + "learning_rate": 7.587710231393508e-07, + "loss": 0.73820686, + "num_input_tokens_seen": 129079215, + "step": 6002, + "time_per_iteration": 2.6945266723632812 + }, + { + "auxiliary_loss_clip": 0.01101194, + "auxiliary_loss_mlp": 0.01025634, + "balance_loss_clip": 1.03697681, + "balance_loss_mlp": 1.01778364, + "epoch": 0.721818072506463, + "flos": 20229809915520.0, + "grad_norm": 1.9416574437667418, + "language_loss": 0.83456206, + "learning_rate": 7.581603132908685e-07, + "loss": 0.85583037, + "num_input_tokens_seen": 129097185, + "step": 6003, + "time_per_iteration": 2.9710042476654053 + }, + { + "auxiliary_loss_clip": 0.01135825, + "auxiliary_loss_mlp": 0.01025233, + "balance_loss_clip": 1.04250646, + "balance_loss_mlp": 1.01684666, + "epoch": 0.7219383153971022, + "flos": 18186887888640.0, + "grad_norm": 2.359119360303587, + "language_loss": 0.78337628, + "learning_rate": 7.575497918284795e-07, + "loss": 0.80498683, + "num_input_tokens_seen": 129114730, + "step": 6004, + "time_per_iteration": 2.7785892486572266 + }, + { + "auxiliary_loss_clip": 0.01174304, + "auxiliary_loss_mlp": 0.01029195, + "balance_loss_clip": 1.04705071, + "balance_loss_mlp": 1.02073145, + "epoch": 0.7220585582877412, + "flos": 17342124854400.0, + "grad_norm": 5.220020270795282, + "language_loss": 0.74690944, + "learning_rate": 7.569394588447984e-07, + "loss": 0.76894444, + "num_input_tokens_seen": 129131745, + "step": 6005, + "time_per_iteration": 2.690945863723755 + }, + { + "auxiliary_loss_clip": 0.01153116, + "auxiliary_loss_mlp": 0.01024789, + "balance_loss_clip": 1.04390132, + "balance_loss_mlp": 1.01701617, + "epoch": 0.7221788011783803, + "flos": 16976338704000.0, + "grad_norm": 2.3750323256160963, + "language_loss": 0.77930647, + "learning_rate": 7.563293144324146e-07, + "loss": 0.80108547, + "num_input_tokens_seen": 129147295, + "step": 6006, + "time_per_iteration": 2.667130708694458 + }, + { + "auxiliary_loss_clip": 0.01170016, + "auxiliary_loss_mlp": 0.01032203, + "balance_loss_clip": 1.04896426, + "balance_loss_mlp": 1.02434683, + "epoch": 0.7222990440690195, + "flos": 26286359702400.0, + "grad_norm": 1.8608489820737988, + "language_loss": 0.79925823, + "learning_rate": 7.557193586838834e-07, + "loss": 0.82128042, + "num_input_tokens_seen": 129162660, + "step": 6007, + "time_per_iteration": 2.7544143199920654 + }, + { + "auxiliary_loss_clip": 0.01145649, + "auxiliary_loss_mlp": 0.01032789, + "balance_loss_clip": 1.04257226, + "balance_loss_mlp": 1.02482557, + "epoch": 0.7224192869596585, + "flos": 17601687509760.0, + "grad_norm": 2.5442992918845944, + "language_loss": 0.70473242, + "learning_rate": 7.551095916917371e-07, + "loss": 0.72651684, + "num_input_tokens_seen": 129179990, + "step": 6008, + "time_per_iteration": 2.7860445976257324 + }, + { + "auxiliary_loss_clip": 0.01144044, + "auxiliary_loss_mlp": 0.01032212, + "balance_loss_clip": 1.04369032, + "balance_loss_mlp": 1.02280664, + "epoch": 0.7225395298502976, + "flos": 12932331016320.0, + "grad_norm": 2.7342403671890243, + "language_loss": 0.66173625, + "learning_rate": 7.545000135484758e-07, + "loss": 0.68349886, + "num_input_tokens_seen": 129197425, + "step": 6009, + "time_per_iteration": 2.7331759929656982 + }, + { + "auxiliary_loss_clip": 0.01170738, + "auxiliary_loss_mlp": 0.00886091, + "balance_loss_clip": 1.04826343, + "balance_loss_mlp": 1.00085175, + "epoch": 0.7226597727409367, + "flos": 29643899592960.0, + "grad_norm": 2.175596899147445, + "language_loss": 0.63005531, + "learning_rate": 7.538906243465714e-07, + "loss": 0.65062362, + "num_input_tokens_seen": 129217560, + "step": 6010, + "time_per_iteration": 3.83545184135437 + }, + { + "auxiliary_loss_clip": 0.01172941, + "auxiliary_loss_mlp": 0.01031035, + "balance_loss_clip": 1.04919171, + "balance_loss_mlp": 1.02313709, + "epoch": 0.7227800156315758, + "flos": 13771635183360.0, + "grad_norm": 2.022341629360125, + "language_loss": 0.78771955, + "learning_rate": 7.5328142417847e-07, + "loss": 0.80975938, + "num_input_tokens_seen": 129234325, + "step": 6011, + "time_per_iteration": 3.961350679397583 + }, + { + "auxiliary_loss_clip": 0.01157009, + "auxiliary_loss_mlp": 0.01033962, + "balance_loss_clip": 1.04434097, + "balance_loss_mlp": 1.0269258, + "epoch": 0.7229002585222148, + "flos": 20301882554880.0, + "grad_norm": 1.5974501811568185, + "language_loss": 0.69215286, + "learning_rate": 7.526724131365838e-07, + "loss": 0.71406257, + "num_input_tokens_seen": 129255280, + "step": 6012, + "time_per_iteration": 2.8314411640167236 + }, + { + "auxiliary_loss_clip": 0.01155202, + "auxiliary_loss_mlp": 0.01026354, + "balance_loss_clip": 1.04977918, + "balance_loss_mlp": 1.01801491, + "epoch": 0.723020501412854, + "flos": 16581250033920.0, + "grad_norm": 1.925923565188096, + "language_loss": 0.70403194, + "learning_rate": 7.520635913133017e-07, + "loss": 0.72584754, + "num_input_tokens_seen": 129273910, + "step": 6013, + "time_per_iteration": 2.795550584793091 + }, + { + "auxiliary_loss_clip": 0.01166928, + "auxiliary_loss_mlp": 0.01027756, + "balance_loss_clip": 1.04672337, + "balance_loss_mlp": 1.01815414, + "epoch": 0.7231407443034931, + "flos": 28548300908160.0, + "grad_norm": 2.502980695635196, + "language_loss": 0.82412541, + "learning_rate": 7.514549588009798e-07, + "loss": 0.84607232, + "num_input_tokens_seen": 129294785, + "step": 6014, + "time_per_iteration": 3.682129144668579 + }, + { + "auxiliary_loss_clip": 0.01154919, + "auxiliary_loss_mlp": 0.01034475, + "balance_loss_clip": 1.04572499, + "balance_loss_mlp": 1.02613628, + "epoch": 0.7232609871941321, + "flos": 30008536508160.0, + "grad_norm": 2.298671040669195, + "language_loss": 0.70616448, + "learning_rate": 7.508465156919492e-07, + "loss": 0.72805846, + "num_input_tokens_seen": 129318295, + "step": 6015, + "time_per_iteration": 2.7783026695251465 + }, + { + "auxiliary_loss_clip": 0.01151534, + "auxiliary_loss_mlp": 0.0102799, + "balance_loss_clip": 1.0443517, + "balance_loss_mlp": 1.01949036, + "epoch": 0.7233812300847713, + "flos": 16654005031680.0, + "grad_norm": 2.385949071193917, + "language_loss": 0.61047214, + "learning_rate": 7.502382620785083e-07, + "loss": 0.63226736, + "num_input_tokens_seen": 129334845, + "step": 6016, + "time_per_iteration": 2.7890994548797607 + }, + { + "auxiliary_loss_clip": 0.01041956, + "auxiliary_loss_mlp": 0.01005481, + "balance_loss_clip": 1.01148832, + "balance_loss_mlp": 1.00388336, + "epoch": 0.7235014729754103, + "flos": 67258784050560.0, + "grad_norm": 0.8066683565143561, + "language_loss": 0.62528795, + "learning_rate": 7.496301980529289e-07, + "loss": 0.64576232, + "num_input_tokens_seen": 129398055, + "step": 6017, + "time_per_iteration": 4.317505598068237 + }, + { + "auxiliary_loss_clip": 0.01170604, + "auxiliary_loss_mlp": 0.01029899, + "balance_loss_clip": 1.04774094, + "balance_loss_mlp": 1.02232349, + "epoch": 0.7236217158660494, + "flos": 26943237671040.0, + "grad_norm": 2.310877777497456, + "language_loss": 0.74338293, + "learning_rate": 7.490223237074547e-07, + "loss": 0.76538789, + "num_input_tokens_seen": 129417765, + "step": 6018, + "time_per_iteration": 2.725480079650879 + }, + { + "auxiliary_loss_clip": 0.0114013, + "auxiliary_loss_mlp": 0.01026738, + "balance_loss_clip": 1.04230928, + "balance_loss_mlp": 1.01887584, + "epoch": 0.7237419587566886, + "flos": 29423372042880.0, + "grad_norm": 1.8323813815488303, + "language_loss": 0.6594348, + "learning_rate": 7.484146391342989e-07, + "loss": 0.68110353, + "num_input_tokens_seen": 129437560, + "step": 6019, + "time_per_iteration": 2.921788215637207 + }, + { + "auxiliary_loss_clip": 0.01143974, + "auxiliary_loss_mlp": 0.01033819, + "balance_loss_clip": 1.04299498, + "balance_loss_mlp": 1.02586746, + "epoch": 0.7238622016473276, + "flos": 17821496787840.0, + "grad_norm": 3.6733548972739247, + "language_loss": 0.5701353, + "learning_rate": 7.478071444256484e-07, + "loss": 0.59191322, + "num_input_tokens_seen": 129455320, + "step": 6020, + "time_per_iteration": 2.730806350708008 + }, + { + "auxiliary_loss_clip": 0.01144191, + "auxiliary_loss_mlp": 0.01028323, + "balance_loss_clip": 1.04448628, + "balance_loss_mlp": 1.02065158, + "epoch": 0.7239824445379667, + "flos": 25739117020800.0, + "grad_norm": 1.7131160706032633, + "language_loss": 0.79381108, + "learning_rate": 7.471998396736579e-07, + "loss": 0.81553626, + "num_input_tokens_seen": 129475700, + "step": 6021, + "time_per_iteration": 2.8936197757720947 + }, + { + "auxiliary_loss_clip": 0.0114429, + "auxiliary_loss_mlp": 0.0102971, + "balance_loss_clip": 1.0449487, + "balance_loss_mlp": 1.02189016, + "epoch": 0.7241026874286057, + "flos": 23148916398720.0, + "grad_norm": 1.9223363642326785, + "language_loss": 0.76231325, + "learning_rate": 7.465927249704549e-07, + "loss": 0.78405321, + "num_input_tokens_seen": 129493585, + "step": 6022, + "time_per_iteration": 2.8259456157684326 + }, + { + "auxiliary_loss_clip": 0.01160513, + "auxiliary_loss_mlp": 0.01027378, + "balance_loss_clip": 1.04598606, + "balance_loss_mlp": 1.01969457, + "epoch": 0.7242229303192449, + "flos": 20266905686400.0, + "grad_norm": 2.560873302975877, + "language_loss": 0.77677226, + "learning_rate": 7.459858004081398e-07, + "loss": 0.79865122, + "num_input_tokens_seen": 129511555, + "step": 6023, + "time_per_iteration": 2.699686050415039 + }, + { + "auxiliary_loss_clip": 0.01036201, + "auxiliary_loss_mlp": 0.01005836, + "balance_loss_clip": 1.00895905, + "balance_loss_mlp": 1.0043875, + "epoch": 0.724343173209884, + "flos": 62311659684480.0, + "grad_norm": 0.6565675561622671, + "language_loss": 0.58047676, + "learning_rate": 7.453790660787815e-07, + "loss": 0.60089707, + "num_input_tokens_seen": 129579650, + "step": 6024, + "time_per_iteration": 3.456890344619751 + }, + { + "auxiliary_loss_clip": 0.01156296, + "auxiliary_loss_mlp": 0.01029944, + "balance_loss_clip": 1.04776168, + "balance_loss_mlp": 1.02140832, + "epoch": 0.724463416100523, + "flos": 35006403813120.0, + "grad_norm": 3.3567825031405714, + "language_loss": 0.63707554, + "learning_rate": 7.447725220744214e-07, + "loss": 0.65893793, + "num_input_tokens_seen": 129601895, + "step": 6025, + "time_per_iteration": 2.8205699920654297 + }, + { + "auxiliary_loss_clip": 0.01173257, + "auxiliary_loss_mlp": 0.0103522, + "balance_loss_clip": 1.04858434, + "balance_loss_mlp": 1.02674413, + "epoch": 0.7245836589911622, + "flos": 21871968923520.0, + "grad_norm": 2.6821487432282898, + "language_loss": 0.7721349, + "learning_rate": 7.441661684870717e-07, + "loss": 0.79421961, + "num_input_tokens_seen": 129622150, + "step": 6026, + "time_per_iteration": 2.828665256500244 + }, + { + "auxiliary_loss_clip": 0.01172924, + "auxiliary_loss_mlp": 0.01029596, + "balance_loss_clip": 1.04915738, + "balance_loss_mlp": 1.02199042, + "epoch": 0.7247039018818012, + "flos": 23006494972800.0, + "grad_norm": 1.831825972448472, + "language_loss": 0.81585503, + "learning_rate": 7.435600054087152e-07, + "loss": 0.83788025, + "num_input_tokens_seen": 129644315, + "step": 6027, + "time_per_iteration": 2.6881308555603027 + }, + { + "auxiliary_loss_clip": 0.01171367, + "auxiliary_loss_mlp": 0.01030152, + "balance_loss_clip": 1.04805732, + "balance_loss_mlp": 1.02167058, + "epoch": 0.7248241447724403, + "flos": 31722588587520.0, + "grad_norm": 2.089590553014591, + "language_loss": 0.74527699, + "learning_rate": 7.42954032931308e-07, + "loss": 0.76729214, + "num_input_tokens_seen": 129665355, + "step": 6028, + "time_per_iteration": 2.767408609390259 + }, + { + "auxiliary_loss_clip": 0.01156025, + "auxiliary_loss_mlp": 0.0103343, + "balance_loss_clip": 1.04713988, + "balance_loss_mlp": 1.02581263, + "epoch": 0.7249443876630794, + "flos": 34896984007680.0, + "grad_norm": 1.9383876619309217, + "language_loss": 0.74996901, + "learning_rate": 7.423482511467733e-07, + "loss": 0.77186358, + "num_input_tokens_seen": 129686125, + "step": 6029, + "time_per_iteration": 2.8471317291259766 + }, + { + "auxiliary_loss_clip": 0.01112684, + "auxiliary_loss_mlp": 0.01022933, + "balance_loss_clip": 1.03927445, + "balance_loss_mlp": 1.01495206, + "epoch": 0.7250646305537185, + "flos": 26359294268160.0, + "grad_norm": 2.2019673300571445, + "language_loss": 0.64950746, + "learning_rate": 7.417426601470099e-07, + "loss": 0.67086363, + "num_input_tokens_seen": 129706485, + "step": 6030, + "time_per_iteration": 2.9215290546417236 + }, + { + "auxiliary_loss_clip": 0.01163243, + "auxiliary_loss_mlp": 0.01028099, + "balance_loss_clip": 1.04765415, + "balance_loss_mlp": 1.01910472, + "epoch": 0.7251848734443576, + "flos": 30081614728320.0, + "grad_norm": 2.8749804417434257, + "language_loss": 0.78508979, + "learning_rate": 7.411372600238841e-07, + "loss": 0.8070032, + "num_input_tokens_seen": 129727100, + "step": 6031, + "time_per_iteration": 2.7560665607452393 + }, + { + "auxiliary_loss_clip": 0.01172797, + "auxiliary_loss_mlp": 0.01025711, + "balance_loss_clip": 1.04963529, + "balance_loss_mlp": 1.01827252, + "epoch": 0.7253051163349967, + "flos": 17785262943360.0, + "grad_norm": 1.8855999805332426, + "language_loss": 0.73918706, + "learning_rate": 7.405320508692346e-07, + "loss": 0.76117218, + "num_input_tokens_seen": 129745840, + "step": 6032, + "time_per_iteration": 2.6459224224090576 + }, + { + "auxiliary_loss_clip": 0.01167668, + "auxiliary_loss_mlp": 0.0102417, + "balance_loss_clip": 1.04773498, + "balance_loss_mlp": 1.01677608, + "epoch": 0.7254253592256358, + "flos": 12641346938880.0, + "grad_norm": 2.038236865785806, + "language_loss": 0.75089347, + "learning_rate": 7.399270327748727e-07, + "loss": 0.77281189, + "num_input_tokens_seen": 129763500, + "step": 6033, + "time_per_iteration": 2.641756772994995 + }, + { + "auxiliary_loss_clip": 0.01145158, + "auxiliary_loss_mlp": 0.00885695, + "balance_loss_clip": 1.04407871, + "balance_loss_mlp": 1.00079465, + "epoch": 0.7255456021162748, + "flos": 27199208966400.0, + "grad_norm": 1.8525166006112956, + "language_loss": 0.74482441, + "learning_rate": 7.39322205832577e-07, + "loss": 0.76513296, + "num_input_tokens_seen": 129784390, + "step": 6034, + "time_per_iteration": 2.834139585494995 + }, + { + "auxiliary_loss_clip": 0.01148308, + "auxiliary_loss_mlp": 0.01030775, + "balance_loss_clip": 1.04495215, + "balance_loss_mlp": 1.02261543, + "epoch": 0.725665845006914, + "flos": 21288205088640.0, + "grad_norm": 2.1497983179463755, + "language_loss": 0.80946708, + "learning_rate": 7.387175701341009e-07, + "loss": 0.83125794, + "num_input_tokens_seen": 129803060, + "step": 6035, + "time_per_iteration": 2.800917387008667 + }, + { + "auxiliary_loss_clip": 0.0116223, + "auxiliary_loss_mlp": 0.01026473, + "balance_loss_clip": 1.04564977, + "balance_loss_mlp": 1.01827693, + "epoch": 0.7257860878975531, + "flos": 16033684129920.0, + "grad_norm": 2.3386051288590273, + "language_loss": 0.72275424, + "learning_rate": 7.381131257711659e-07, + "loss": 0.7446413, + "num_input_tokens_seen": 129820165, + "step": 6036, + "time_per_iteration": 3.599729537963867 + }, + { + "auxiliary_loss_clip": 0.0114851, + "auxiliary_loss_mlp": 0.01026788, + "balance_loss_clip": 1.04809058, + "balance_loss_mlp": 1.01936722, + "epoch": 0.7259063307881921, + "flos": 12129943052160.0, + "grad_norm": 1.8064747101203995, + "language_loss": 0.83418852, + "learning_rate": 7.375088728354677e-07, + "loss": 0.85594153, + "num_input_tokens_seen": 129835195, + "step": 6037, + "time_per_iteration": 3.641573429107666 + }, + { + "auxiliary_loss_clip": 0.01143638, + "auxiliary_loss_mlp": 0.01033234, + "balance_loss_clip": 1.04232717, + "balance_loss_mlp": 1.02521741, + "epoch": 0.7260265736788313, + "flos": 30443845432320.0, + "grad_norm": 2.291956851365193, + "language_loss": 0.67719549, + "learning_rate": 7.369048114186691e-07, + "loss": 0.69896424, + "num_input_tokens_seen": 129856240, + "step": 6038, + "time_per_iteration": 2.944990396499634 + }, + { + "auxiliary_loss_clip": 0.01142058, + "auxiliary_loss_mlp": 0.00885553, + "balance_loss_clip": 1.04327703, + "balance_loss_mlp": 1.00070715, + "epoch": 0.7261468165694703, + "flos": 21142264129920.0, + "grad_norm": 1.7761164138776064, + "language_loss": 0.82721573, + "learning_rate": 7.363009416124055e-07, + "loss": 0.84749186, + "num_input_tokens_seen": 129875565, + "step": 6039, + "time_per_iteration": 3.7765870094299316 + }, + { + "auxiliary_loss_clip": 0.01145156, + "auxiliary_loss_mlp": 0.0102698, + "balance_loss_clip": 1.04404509, + "balance_loss_mlp": 1.01898754, + "epoch": 0.7262670594601094, + "flos": 22306308180480.0, + "grad_norm": 2.1708956207995795, + "language_loss": 0.62211865, + "learning_rate": 7.356972635082852e-07, + "loss": 0.64384001, + "num_input_tokens_seen": 129894420, + "step": 6040, + "time_per_iteration": 2.8297462463378906 + }, + { + "auxiliary_loss_clip": 0.01130142, + "auxiliary_loss_mlp": 0.01026188, + "balance_loss_clip": 1.0454185, + "balance_loss_mlp": 1.01823699, + "epoch": 0.7263873023507486, + "flos": 25335049950720.0, + "grad_norm": 2.187771471932375, + "language_loss": 0.75366116, + "learning_rate": 7.35093777197884e-07, + "loss": 0.77522445, + "num_input_tokens_seen": 129914490, + "step": 6041, + "time_per_iteration": 2.847252130508423 + }, + { + "auxiliary_loss_clip": 0.01149044, + "auxiliary_loss_mlp": 0.01025876, + "balance_loss_clip": 1.04637814, + "balance_loss_mlp": 1.01800847, + "epoch": 0.7265075452413876, + "flos": 23878621192320.0, + "grad_norm": 2.5232869229036408, + "language_loss": 0.85812497, + "learning_rate": 7.344904827727525e-07, + "loss": 0.87987411, + "num_input_tokens_seen": 129931670, + "step": 6042, + "time_per_iteration": 2.7143452167510986 + }, + { + "auxiliary_loss_clip": 0.01143277, + "auxiliary_loss_mlp": 0.01027539, + "balance_loss_clip": 1.04124308, + "balance_loss_mlp": 1.01930141, + "epoch": 0.7266277881320267, + "flos": 28724549967360.0, + "grad_norm": 2.102334877808614, + "language_loss": 0.74288505, + "learning_rate": 7.338873803244076e-07, + "loss": 0.76459324, + "num_input_tokens_seen": 129946905, + "step": 6043, + "time_per_iteration": 3.7058002948760986 + }, + { + "auxiliary_loss_clip": 0.01146032, + "auxiliary_loss_mlp": 0.0102585, + "balance_loss_clip": 1.04481494, + "balance_loss_mlp": 1.01770782, + "epoch": 0.7267480310226658, + "flos": 24863507182080.0, + "grad_norm": 1.6813004818747295, + "language_loss": 0.80795467, + "learning_rate": 7.332844699443401e-07, + "loss": 0.82967347, + "num_input_tokens_seen": 129965505, + "step": 6044, + "time_per_iteration": 2.7458322048187256 + }, + { + "auxiliary_loss_clip": 0.01124343, + "auxiliary_loss_mlp": 0.0102863, + "balance_loss_clip": 1.0403372, + "balance_loss_mlp": 1.02140009, + "epoch": 0.7268682739133049, + "flos": 27198490694400.0, + "grad_norm": 1.8190240258678945, + "language_loss": 0.7520228, + "learning_rate": 7.326817517240121e-07, + "loss": 0.77355254, + "num_input_tokens_seen": 129987210, + "step": 6045, + "time_per_iteration": 2.8947441577911377 + }, + { + "auxiliary_loss_clip": 0.0116058, + "auxiliary_loss_mlp": 0.00885058, + "balance_loss_clip": 1.04470634, + "balance_loss_mlp": 1.00079274, + "epoch": 0.7269885168039439, + "flos": 33508138688640.0, + "grad_norm": 1.9151974344616756, + "language_loss": 0.83087409, + "learning_rate": 7.320792257548545e-07, + "loss": 0.85133052, + "num_input_tokens_seen": 130008385, + "step": 6046, + "time_per_iteration": 2.859049081802368 + }, + { + "auxiliary_loss_clip": 0.01157366, + "auxiliary_loss_mlp": 0.01027653, + "balance_loss_clip": 1.04645193, + "balance_loss_mlp": 1.01979685, + "epoch": 0.7271087596945831, + "flos": 24313750548480.0, + "grad_norm": 1.9378926253990345, + "language_loss": 0.76304436, + "learning_rate": 7.314768921282704e-07, + "loss": 0.78489453, + "num_input_tokens_seen": 130029040, + "step": 6047, + "time_per_iteration": 2.7594516277313232 + }, + { + "auxiliary_loss_clip": 0.01165215, + "auxiliary_loss_mlp": 0.01026533, + "balance_loss_clip": 1.04833221, + "balance_loss_mlp": 1.01908243, + "epoch": 0.7272290025852222, + "flos": 23805147922560.0, + "grad_norm": 3.0423749959464237, + "language_loss": 0.71941864, + "learning_rate": 7.30874750935633e-07, + "loss": 0.74133617, + "num_input_tokens_seen": 130048725, + "step": 6048, + "time_per_iteration": 2.8812451362609863 + }, + { + "auxiliary_loss_clip": 0.01142032, + "auxiliary_loss_mlp": 0.01030375, + "balance_loss_clip": 1.0459578, + "balance_loss_mlp": 1.02285314, + "epoch": 0.7273492454758612, + "flos": 16720367408640.0, + "grad_norm": 1.9432160950897384, + "language_loss": 0.79119146, + "learning_rate": 7.30272802268286e-07, + "loss": 0.81291556, + "num_input_tokens_seen": 130065720, + "step": 6049, + "time_per_iteration": 2.826399803161621 + }, + { + "auxiliary_loss_clip": 0.0109974, + "auxiliary_loss_mlp": 0.01028567, + "balance_loss_clip": 1.0384115, + "balance_loss_mlp": 1.02088416, + "epoch": 0.7274694883665004, + "flos": 28031330413440.0, + "grad_norm": 1.7365687853765353, + "language_loss": 0.75931144, + "learning_rate": 7.29671046217547e-07, + "loss": 0.78059453, + "num_input_tokens_seen": 130084830, + "step": 6050, + "time_per_iteration": 2.985410690307617 + }, + { + "auxiliary_loss_clip": 0.01145212, + "auxiliary_loss_mlp": 0.01029783, + "balance_loss_clip": 1.04605699, + "balance_loss_mlp": 1.02253532, + "epoch": 0.7275897312571394, + "flos": 30372706546560.0, + "grad_norm": 2.167372899314996, + "language_loss": 0.81519485, + "learning_rate": 7.290694828746988e-07, + "loss": 0.83694488, + "num_input_tokens_seen": 130104495, + "step": 6051, + "time_per_iteration": 2.847409248352051 + }, + { + "auxiliary_loss_clip": 0.01146422, + "auxiliary_loss_mlp": 0.01025625, + "balance_loss_clip": 1.04294419, + "balance_loss_mlp": 1.01735771, + "epoch": 0.7277099741477785, + "flos": 19204775498880.0, + "grad_norm": 1.7949438254498287, + "language_loss": 0.85877961, + "learning_rate": 7.284681123310004e-07, + "loss": 0.88050008, + "num_input_tokens_seen": 130123210, + "step": 6052, + "time_per_iteration": 2.810504674911499 + }, + { + "auxiliary_loss_clip": 0.01162021, + "auxiliary_loss_mlp": 0.01033365, + "balance_loss_clip": 1.04654336, + "balance_loss_mlp": 1.0250026, + "epoch": 0.7278302170384175, + "flos": 20667884186880.0, + "grad_norm": 1.9169563003766303, + "language_loss": 0.79495174, + "learning_rate": 7.27866934677678e-07, + "loss": 0.81690562, + "num_input_tokens_seen": 130142880, + "step": 6053, + "time_per_iteration": 2.745201587677002 + }, + { + "auxiliary_loss_clip": 0.01130807, + "auxiliary_loss_mlp": 0.01027874, + "balance_loss_clip": 1.04230225, + "balance_loss_mlp": 1.02006638, + "epoch": 0.7279504599290567, + "flos": 19093200877440.0, + "grad_norm": 1.9206118898568427, + "language_loss": 0.78275335, + "learning_rate": 7.272659500059297e-07, + "loss": 0.80434024, + "num_input_tokens_seen": 130160220, + "step": 6054, + "time_per_iteration": 2.899362802505493 + }, + { + "auxiliary_loss_clip": 0.01157429, + "auxiliary_loss_mlp": 0.01032171, + "balance_loss_clip": 1.04686356, + "balance_loss_mlp": 1.02382684, + "epoch": 0.7280707028196958, + "flos": 19062174504960.0, + "grad_norm": 2.120801415036663, + "language_loss": 0.80431294, + "learning_rate": 7.266651584069264e-07, + "loss": 0.82620895, + "num_input_tokens_seen": 130177885, + "step": 6055, + "time_per_iteration": 2.7634830474853516 + }, + { + "auxiliary_loss_clip": 0.01165732, + "auxiliary_loss_mlp": 0.01031869, + "balance_loss_clip": 1.04910636, + "balance_loss_mlp": 1.02326798, + "epoch": 0.7281909457103348, + "flos": 37196308293120.0, + "grad_norm": 3.5689191219920193, + "language_loss": 0.56871969, + "learning_rate": 7.260645599718045e-07, + "loss": 0.59069574, + "num_input_tokens_seen": 130204240, + "step": 6056, + "time_per_iteration": 2.875328540802002 + }, + { + "auxiliary_loss_clip": 0.01150191, + "auxiliary_loss_mlp": 0.0102695, + "balance_loss_clip": 1.0439775, + "balance_loss_mlp": 1.01822972, + "epoch": 0.728311188600974, + "flos": 20667094087680.0, + "grad_norm": 2.198337309582186, + "language_loss": 0.67155373, + "learning_rate": 7.254641547916767e-07, + "loss": 0.6933251, + "num_input_tokens_seen": 130221735, + "step": 6057, + "time_per_iteration": 2.7740468978881836 + }, + { + "auxiliary_loss_clip": 0.01173943, + "auxiliary_loss_mlp": 0.01028823, + "balance_loss_clip": 1.05179548, + "balance_loss_mlp": 1.02112508, + "epoch": 0.728431431491613, + "flos": 28840685616000.0, + "grad_norm": 1.7958638657871802, + "language_loss": 0.69479203, + "learning_rate": 7.248639429576226e-07, + "loss": 0.7168197, + "num_input_tokens_seen": 130241190, + "step": 6058, + "time_per_iteration": 2.6544909477233887 + }, + { + "auxiliary_loss_clip": 0.0116277, + "auxiliary_loss_mlp": 0.01025192, + "balance_loss_clip": 1.04699111, + "balance_loss_mlp": 1.01735687, + "epoch": 0.7285516743822521, + "flos": 25991856092160.0, + "grad_norm": 1.6443368456498508, + "language_loss": 0.72126061, + "learning_rate": 7.242639245606959e-07, + "loss": 0.74314022, + "num_input_tokens_seen": 130260980, + "step": 6059, + "time_per_iteration": 2.786759853363037 + }, + { + "auxiliary_loss_clip": 0.01159343, + "auxiliary_loss_mlp": 0.01030456, + "balance_loss_clip": 1.04685962, + "balance_loss_mlp": 1.02204585, + "epoch": 0.7286719172728913, + "flos": 16399721675520.0, + "grad_norm": 1.7945903123355385, + "language_loss": 0.82315981, + "learning_rate": 7.236640996919168e-07, + "loss": 0.84505785, + "num_input_tokens_seen": 130280025, + "step": 6060, + "time_per_iteration": 2.7455153465270996 + }, + { + "auxiliary_loss_clip": 0.01165058, + "auxiliary_loss_mlp": 0.01025797, + "balance_loss_clip": 1.04867375, + "balance_loss_mlp": 1.01823592, + "epoch": 0.7287921601635303, + "flos": 22018161277440.0, + "grad_norm": 2.932480525865668, + "language_loss": 0.70635402, + "learning_rate": 7.230644684422782e-07, + "loss": 0.72826254, + "num_input_tokens_seen": 130300255, + "step": 6061, + "time_per_iteration": 2.6939237117767334 + }, + { + "auxiliary_loss_clip": 0.01142414, + "auxiliary_loss_mlp": 0.01029244, + "balance_loss_clip": 1.04431629, + "balance_loss_mlp": 1.02095604, + "epoch": 0.7289124030541694, + "flos": 24600927784320.0, + "grad_norm": 1.667241891181801, + "language_loss": 0.81377631, + "learning_rate": 7.224650309027451e-07, + "loss": 0.83549285, + "num_input_tokens_seen": 130320005, + "step": 6062, + "time_per_iteration": 3.829216241836548 + }, + { + "auxiliary_loss_clip": 0.01168074, + "auxiliary_loss_mlp": 0.01032324, + "balance_loss_clip": 1.050547, + "balance_loss_mlp": 1.02462876, + "epoch": 0.7290326459448085, + "flos": 21393638484480.0, + "grad_norm": 4.524152858867397, + "language_loss": 0.68606645, + "learning_rate": 7.218657871642506e-07, + "loss": 0.70807046, + "num_input_tokens_seen": 130338810, + "step": 6063, + "time_per_iteration": 3.6117632389068604 + }, + { + "auxiliary_loss_clip": 0.01173217, + "auxiliary_loss_mlp": 0.01025118, + "balance_loss_clip": 1.0496192, + "balance_loss_mlp": 1.01665401, + "epoch": 0.7291528888354476, + "flos": 18587686821120.0, + "grad_norm": 2.0961451289244866, + "language_loss": 0.61950576, + "learning_rate": 7.212667373177012e-07, + "loss": 0.64148915, + "num_input_tokens_seen": 130353805, + "step": 6064, + "time_per_iteration": 2.6789069175720215 + }, + { + "auxiliary_loss_clip": 0.01143248, + "auxiliary_loss_mlp": 0.01027971, + "balance_loss_clip": 1.04345083, + "balance_loss_mlp": 1.02002525, + "epoch": 0.7292731317260867, + "flos": 18951066760320.0, + "grad_norm": 1.6907893203354054, + "language_loss": 0.75204325, + "learning_rate": 7.206678814539704e-07, + "loss": 0.77375537, + "num_input_tokens_seen": 130372105, + "step": 6065, + "time_per_iteration": 3.7680790424346924 + }, + { + "auxiliary_loss_clip": 0.01141522, + "auxiliary_loss_mlp": 0.01026769, + "balance_loss_clip": 1.04263496, + "balance_loss_mlp": 1.01942015, + "epoch": 0.7293933746167258, + "flos": 21067569797760.0, + "grad_norm": 1.4541858134159158, + "language_loss": 0.72884744, + "learning_rate": 7.20069219663904e-07, + "loss": 0.75053036, + "num_input_tokens_seen": 130391990, + "step": 6066, + "time_per_iteration": 2.875319480895996 + }, + { + "auxiliary_loss_clip": 0.01164368, + "auxiliary_loss_mlp": 0.01023664, + "balance_loss_clip": 1.04521286, + "balance_loss_mlp": 1.01553392, + "epoch": 0.7295136175073649, + "flos": 22453326547200.0, + "grad_norm": 1.6287537339564577, + "language_loss": 0.7952922, + "learning_rate": 7.1947075203832e-07, + "loss": 0.81717259, + "num_input_tokens_seen": 130411970, + "step": 6067, + "time_per_iteration": 2.798784017562866 + }, + { + "auxiliary_loss_clip": 0.0106001, + "auxiliary_loss_mlp": 0.01002403, + "balance_loss_clip": 1.0077281, + "balance_loss_mlp": 1.00088263, + "epoch": 0.7296338603980039, + "flos": 56125506648960.0, + "grad_norm": 0.9401147416294576, + "language_loss": 0.60129654, + "learning_rate": 7.188724786680049e-07, + "loss": 0.62192065, + "num_input_tokens_seen": 130472440, + "step": 6068, + "time_per_iteration": 3.218158721923828 + }, + { + "auxiliary_loss_clip": 0.01150821, + "auxiliary_loss_mlp": 0.01029768, + "balance_loss_clip": 1.04353189, + "balance_loss_mlp": 1.02203178, + "epoch": 0.7297541032886431, + "flos": 25228287751680.0, + "grad_norm": 1.5110340375189342, + "language_loss": 0.75779724, + "learning_rate": 7.182743996437162e-07, + "loss": 0.77960318, + "num_input_tokens_seen": 130491975, + "step": 6069, + "time_per_iteration": 3.7021267414093018 + }, + { + "auxiliary_loss_clip": 0.01153277, + "auxiliary_loss_mlp": 0.01026297, + "balance_loss_clip": 1.04494691, + "balance_loss_mlp": 1.01785731, + "epoch": 0.7298743461792822, + "flos": 26467600752000.0, + "grad_norm": 2.026995121447764, + "language_loss": 0.68632853, + "learning_rate": 7.176765150561819e-07, + "loss": 0.70812428, + "num_input_tokens_seen": 130510580, + "step": 6070, + "time_per_iteration": 2.853680372238159 + }, + { + "auxiliary_loss_clip": 0.0117199, + "auxiliary_loss_mlp": 0.01029817, + "balance_loss_clip": 1.04687142, + "balance_loss_mlp": 1.02143097, + "epoch": 0.7299945890699212, + "flos": 19569053278080.0, + "grad_norm": 2.340313547196506, + "language_loss": 0.79761767, + "learning_rate": 7.170788249961002e-07, + "loss": 0.81963575, + "num_input_tokens_seen": 130529090, + "step": 6071, + "time_per_iteration": 2.609806776046753 + }, + { + "auxiliary_loss_clip": 0.01168996, + "auxiliary_loss_mlp": 0.01028038, + "balance_loss_clip": 1.04648554, + "balance_loss_mlp": 1.01968181, + "epoch": 0.7301148319605604, + "flos": 22928963466240.0, + "grad_norm": 2.1787007071799476, + "language_loss": 0.88147509, + "learning_rate": 7.164813295541418e-07, + "loss": 0.90344548, + "num_input_tokens_seen": 130548655, + "step": 6072, + "time_per_iteration": 2.648879051208496 + }, + { + "auxiliary_loss_clip": 0.01155338, + "auxiliary_loss_mlp": 0.01028316, + "balance_loss_clip": 1.04556394, + "balance_loss_mlp": 1.02026904, + "epoch": 0.7302350748511994, + "flos": 25369703596800.0, + "grad_norm": 1.7534984652541608, + "language_loss": 0.70293403, + "learning_rate": 7.15884028820944e-07, + "loss": 0.72477055, + "num_input_tokens_seen": 130567710, + "step": 6073, + "time_per_iteration": 2.7076029777526855 + }, + { + "auxiliary_loss_clip": 0.01138404, + "auxiliary_loss_mlp": 0.01031174, + "balance_loss_clip": 1.04206967, + "balance_loss_mlp": 1.02267432, + "epoch": 0.7303553177418385, + "flos": 27819170732160.0, + "grad_norm": 2.125658735811979, + "language_loss": 0.59760857, + "learning_rate": 7.152869228871185e-07, + "loss": 0.6193043, + "num_input_tokens_seen": 130590195, + "step": 6074, + "time_per_iteration": 2.827880620956421 + }, + { + "auxiliary_loss_clip": 0.01150527, + "auxiliary_loss_mlp": 0.01027474, + "balance_loss_clip": 1.04703045, + "balance_loss_mlp": 1.01930213, + "epoch": 0.7304755606324776, + "flos": 24426510318720.0, + "grad_norm": 2.252274585218699, + "language_loss": 0.72165364, + "learning_rate": 7.146900118432457e-07, + "loss": 0.74343365, + "num_input_tokens_seen": 130609940, + "step": 6075, + "time_per_iteration": 2.7821125984191895 + }, + { + "auxiliary_loss_clip": 0.01108407, + "auxiliary_loss_mlp": 0.01025931, + "balance_loss_clip": 1.03363824, + "balance_loss_mlp": 1.01800358, + "epoch": 0.7305958035231167, + "flos": 23840483927040.0, + "grad_norm": 1.6751553757419115, + "language_loss": 0.85978466, + "learning_rate": 7.140932957798753e-07, + "loss": 0.88112807, + "num_input_tokens_seen": 130628380, + "step": 6076, + "time_per_iteration": 2.9375176429748535 + }, + { + "auxiliary_loss_clip": 0.01156292, + "auxiliary_loss_mlp": 0.01029125, + "balance_loss_clip": 1.04365396, + "balance_loss_mlp": 1.0209589, + "epoch": 0.7307160464137558, + "flos": 16726939597440.0, + "grad_norm": 2.195945547191571, + "language_loss": 0.71544778, + "learning_rate": 7.134967747875309e-07, + "loss": 0.73730195, + "num_input_tokens_seen": 130646590, + "step": 6077, + "time_per_iteration": 2.796818733215332 + }, + { + "auxiliary_loss_clip": 0.01157158, + "auxiliary_loss_mlp": 0.01033484, + "balance_loss_clip": 1.04408157, + "balance_loss_mlp": 1.02494287, + "epoch": 0.7308362893043949, + "flos": 21798280172160.0, + "grad_norm": 1.7922486412851095, + "language_loss": 0.81793666, + "learning_rate": 7.129004489567014e-07, + "loss": 0.83984303, + "num_input_tokens_seen": 130664070, + "step": 6078, + "time_per_iteration": 2.6625919342041016 + }, + { + "auxiliary_loss_clip": 0.0114627, + "auxiliary_loss_mlp": 0.01026295, + "balance_loss_clip": 1.04366112, + "balance_loss_mlp": 1.01869571, + "epoch": 0.730956532195034, + "flos": 10707377840640.0, + "grad_norm": 2.52064063535587, + "language_loss": 0.77892363, + "learning_rate": 7.123043183778512e-07, + "loss": 0.80064929, + "num_input_tokens_seen": 130681400, + "step": 6079, + "time_per_iteration": 2.75089955329895 + }, + { + "auxiliary_loss_clip": 0.01149493, + "auxiliary_loss_mlp": 0.0103118, + "balance_loss_clip": 1.04654324, + "balance_loss_mlp": 1.02311587, + "epoch": 0.731076775085673, + "flos": 19791987039360.0, + "grad_norm": 1.5512891609452013, + "language_loss": 0.65249217, + "learning_rate": 7.117083831414114e-07, + "loss": 0.67429894, + "num_input_tokens_seen": 130700675, + "step": 6080, + "time_per_iteration": 2.7414302825927734 + }, + { + "auxiliary_loss_clip": 0.01168639, + "auxiliary_loss_mlp": 0.01027648, + "balance_loss_clip": 1.04724813, + "balance_loss_mlp": 1.01963723, + "epoch": 0.7311970179763122, + "flos": 20447033414400.0, + "grad_norm": 2.4257761913540716, + "language_loss": 0.69673145, + "learning_rate": 7.11112643337787e-07, + "loss": 0.71869427, + "num_input_tokens_seen": 130719720, + "step": 6081, + "time_per_iteration": 2.771679639816284 + }, + { + "auxiliary_loss_clip": 0.01154455, + "auxiliary_loss_mlp": 0.01025114, + "balance_loss_clip": 1.04861426, + "balance_loss_mlp": 1.01685858, + "epoch": 0.7313172608669513, + "flos": 18513818501760.0, + "grad_norm": 2.5162473094404305, + "language_loss": 0.76263189, + "learning_rate": 7.10517099057349e-07, + "loss": 0.78442752, + "num_input_tokens_seen": 130736670, + "step": 6082, + "time_per_iteration": 2.694031000137329 + }, + { + "auxiliary_loss_clip": 0.01158954, + "auxiliary_loss_mlp": 0.01030594, + "balance_loss_clip": 1.04864872, + "balance_loss_mlp": 1.02184391, + "epoch": 0.7314375037575903, + "flos": 16180738410240.0, + "grad_norm": 2.3215659542248983, + "language_loss": 0.60976607, + "learning_rate": 7.099217503904411e-07, + "loss": 0.63166153, + "num_input_tokens_seen": 130754525, + "step": 6083, + "time_per_iteration": 2.7289419174194336 + }, + { + "auxiliary_loss_clip": 0.01159807, + "auxiliary_loss_mlp": 0.01027199, + "balance_loss_clip": 1.04856622, + "balance_loss_mlp": 1.01992738, + "epoch": 0.7315577466482295, + "flos": 17967940536960.0, + "grad_norm": 1.7700322486524198, + "language_loss": 0.90352184, + "learning_rate": 7.093265974273788e-07, + "loss": 0.92539185, + "num_input_tokens_seen": 130772420, + "step": 6084, + "time_per_iteration": 2.730264902114868 + }, + { + "auxiliary_loss_clip": 0.01164131, + "auxiliary_loss_mlp": 0.01027071, + "balance_loss_clip": 1.04641962, + "balance_loss_mlp": 1.01986194, + "epoch": 0.7316779895388685, + "flos": 18405440190720.0, + "grad_norm": 2.1255977715110563, + "language_loss": 0.71659917, + "learning_rate": 7.087316402584447e-07, + "loss": 0.73851115, + "num_input_tokens_seen": 130791245, + "step": 6085, + "time_per_iteration": 2.6819674968719482 + }, + { + "auxiliary_loss_clip": 0.01173403, + "auxiliary_loss_mlp": 0.01027104, + "balance_loss_clip": 1.04912353, + "balance_loss_mlp": 1.01867032, + "epoch": 0.7317982324295076, + "flos": 17928294900480.0, + "grad_norm": 2.542692729153414, + "language_loss": 0.86161757, + "learning_rate": 7.081368789738953e-07, + "loss": 0.88362265, + "num_input_tokens_seen": 130808445, + "step": 6086, + "time_per_iteration": 2.6880831718444824 + }, + { + "auxiliary_loss_clip": 0.01150665, + "auxiliary_loss_mlp": 0.01031345, + "balance_loss_clip": 1.04390168, + "balance_loss_mlp": 1.02356005, + "epoch": 0.7319184753201466, + "flos": 27229840289280.0, + "grad_norm": 1.9593156728862817, + "language_loss": 0.77912652, + "learning_rate": 7.075423136639537e-07, + "loss": 0.80094659, + "num_input_tokens_seen": 130827700, + "step": 6087, + "time_per_iteration": 2.8186733722686768 + }, + { + "auxiliary_loss_clip": 0.01140208, + "auxiliary_loss_mlp": 0.010305, + "balance_loss_clip": 1.04289544, + "balance_loss_mlp": 1.02287626, + "epoch": 0.7320387182107858, + "flos": 37448544574080.0, + "grad_norm": 1.8834239325458444, + "language_loss": 0.74978727, + "learning_rate": 7.069479444188149e-07, + "loss": 0.77149427, + "num_input_tokens_seen": 130848290, + "step": 6088, + "time_per_iteration": 3.871060609817505 + }, + { + "auxiliary_loss_clip": 0.01144482, + "auxiliary_loss_mlp": 0.01030658, + "balance_loss_clip": 1.04433608, + "balance_loss_mlp": 1.02238524, + "epoch": 0.7321589611014249, + "flos": 17859023521920.0, + "grad_norm": 1.6435086085977235, + "language_loss": 0.82224846, + "learning_rate": 7.063537713286453e-07, + "loss": 0.84399992, + "num_input_tokens_seen": 130865970, + "step": 6089, + "time_per_iteration": 3.65580153465271 + }, + { + "auxiliary_loss_clip": 0.01160106, + "auxiliary_loss_mlp": 0.01028494, + "balance_loss_clip": 1.04577994, + "balance_loss_mlp": 1.02036083, + "epoch": 0.7322792039920639, + "flos": 26100593539200.0, + "grad_norm": 1.8946934747588944, + "language_loss": 0.80726743, + "learning_rate": 7.057597944835803e-07, + "loss": 0.82915342, + "num_input_tokens_seen": 130885245, + "step": 6090, + "time_per_iteration": 2.7860236167907715 + }, + { + "auxiliary_loss_clip": 0.01148945, + "auxiliary_loss_mlp": 0.01025179, + "balance_loss_clip": 1.0434618, + "balance_loss_mlp": 1.01732945, + "epoch": 0.7323994468827031, + "flos": 25369093065600.0, + "grad_norm": 1.6917373703757153, + "language_loss": 0.74687958, + "learning_rate": 7.051660139737253e-07, + "loss": 0.76862085, + "num_input_tokens_seen": 130903465, + "step": 6091, + "time_per_iteration": 3.9053642749786377 + }, + { + "auxiliary_loss_clip": 0.01160592, + "auxiliary_loss_mlp": 0.00885979, + "balance_loss_clip": 1.04723382, + "balance_loss_mlp": 1.00065541, + "epoch": 0.7325196897733421, + "flos": 26907075653760.0, + "grad_norm": 3.2841887209980354, + "language_loss": 0.76771879, + "learning_rate": 7.045724298891565e-07, + "loss": 0.78818452, + "num_input_tokens_seen": 130922935, + "step": 6092, + "time_per_iteration": 2.7914345264434814 + }, + { + "auxiliary_loss_clip": 0.01161663, + "auxiliary_loss_mlp": 0.0102898, + "balance_loss_clip": 1.0479722, + "balance_loss_mlp": 1.02066541, + "epoch": 0.7326399326639812, + "flos": 25775781828480.0, + "grad_norm": 1.9782168449880864, + "language_loss": 0.69153571, + "learning_rate": 7.039790423199192e-07, + "loss": 0.71344209, + "num_input_tokens_seen": 130942575, + "step": 6093, + "time_per_iteration": 2.763784885406494 + }, + { + "auxiliary_loss_clip": 0.01158736, + "auxiliary_loss_mlp": 0.01028726, + "balance_loss_clip": 1.0481962, + "balance_loss_mlp": 1.01996374, + "epoch": 0.7327601755546204, + "flos": 21032269706880.0, + "grad_norm": 2.050412764131393, + "language_loss": 0.77713674, + "learning_rate": 7.033858513560322e-07, + "loss": 0.79901135, + "num_input_tokens_seen": 130958870, + "step": 6094, + "time_per_iteration": 2.7820091247558594 + }, + { + "auxiliary_loss_clip": 0.01165869, + "auxiliary_loss_mlp": 0.01028234, + "balance_loss_clip": 1.04985738, + "balance_loss_mlp": 1.02024102, + "epoch": 0.7328804184452594, + "flos": 16289224462080.0, + "grad_norm": 2.1372952053633294, + "language_loss": 0.7582776, + "learning_rate": 7.027928570874794e-07, + "loss": 0.7802186, + "num_input_tokens_seen": 130977060, + "step": 6095, + "time_per_iteration": 3.664645195007324 + }, + { + "auxiliary_loss_clip": 0.01170328, + "auxiliary_loss_mlp": 0.01027058, + "balance_loss_clip": 1.04804206, + "balance_loss_mlp": 1.01929164, + "epoch": 0.7330006613358985, + "flos": 17858233422720.0, + "grad_norm": 1.8122625438486244, + "language_loss": 0.8528223, + "learning_rate": 7.022000596042194e-07, + "loss": 0.87479615, + "num_input_tokens_seen": 130994160, + "step": 6096, + "time_per_iteration": 2.840904712677002 + }, + { + "auxiliary_loss_clip": 0.0114544, + "auxiliary_loss_mlp": 0.01023835, + "balance_loss_clip": 1.041996, + "balance_loss_mlp": 1.01615167, + "epoch": 0.7331209042265376, + "flos": 22492074343680.0, + "grad_norm": 1.9817130750662162, + "language_loss": 0.81960565, + "learning_rate": 7.016074589961784e-07, + "loss": 0.84129834, + "num_input_tokens_seen": 131012725, + "step": 6097, + "time_per_iteration": 2.989971160888672 + }, + { + "auxiliary_loss_clip": 0.01148717, + "auxiliary_loss_mlp": 0.01030349, + "balance_loss_clip": 1.04553843, + "balance_loss_mlp": 1.02200449, + "epoch": 0.7332411471171767, + "flos": 33072757937280.0, + "grad_norm": 1.8359789766148704, + "language_loss": 0.66964972, + "learning_rate": 7.01015055353253e-07, + "loss": 0.69144046, + "num_input_tokens_seen": 131035150, + "step": 6098, + "time_per_iteration": 2.8789455890655518 + }, + { + "auxiliary_loss_clip": 0.01122579, + "auxiliary_loss_mlp": 0.0102702, + "balance_loss_clip": 1.04093385, + "balance_loss_mlp": 1.01907504, + "epoch": 0.7333613900078157, + "flos": 22743017735040.0, + "grad_norm": 1.8623371877691388, + "language_loss": 0.77871805, + "learning_rate": 7.004228487653123e-07, + "loss": 0.80021411, + "num_input_tokens_seen": 131055955, + "step": 6099, + "time_per_iteration": 2.8821017742156982 + }, + { + "auxiliary_loss_clip": 0.0113748, + "auxiliary_loss_mlp": 0.01026061, + "balance_loss_clip": 1.03824425, + "balance_loss_mlp": 1.0182023, + "epoch": 0.7334816328984549, + "flos": 22346133384960.0, + "grad_norm": 1.6934253571700244, + "language_loss": 0.78203058, + "learning_rate": 6.998308393221906e-07, + "loss": 0.803666, + "num_input_tokens_seen": 131074360, + "step": 6100, + "time_per_iteration": 2.864049196243286 + }, + { + "auxiliary_loss_clip": 0.01148529, + "auxiliary_loss_mlp": 0.01028053, + "balance_loss_clip": 1.04763961, + "balance_loss_mlp": 1.02084029, + "epoch": 0.733601875789094, + "flos": 20736149984640.0, + "grad_norm": 2.3626462866574305, + "language_loss": 0.71322906, + "learning_rate": 6.992390271136977e-07, + "loss": 0.73499489, + "num_input_tokens_seen": 131090070, + "step": 6101, + "time_per_iteration": 2.9237637519836426 + }, + { + "auxiliary_loss_clip": 0.01154908, + "auxiliary_loss_mlp": 0.0102277, + "balance_loss_clip": 1.04490733, + "balance_loss_mlp": 1.0142349, + "epoch": 0.733722118679733, + "flos": 22564362464640.0, + "grad_norm": 1.8075207817616912, + "language_loss": 0.8568368, + "learning_rate": 6.986474122296094e-07, + "loss": 0.87861353, + "num_input_tokens_seen": 131109185, + "step": 6102, + "time_per_iteration": 2.7305877208709717 + }, + { + "auxiliary_loss_clip": 0.01178348, + "auxiliary_loss_mlp": 0.01026402, + "balance_loss_clip": 1.05216372, + "balance_loss_mlp": 1.01792622, + "epoch": 0.7338423615703722, + "flos": 20084192179200.0, + "grad_norm": 2.1722924291388814, + "language_loss": 0.72458351, + "learning_rate": 6.980559947596751e-07, + "loss": 0.74663097, + "num_input_tokens_seen": 131127725, + "step": 6103, + "time_per_iteration": 2.6936285495758057 + }, + { + "auxiliary_loss_clip": 0.01132766, + "auxiliary_loss_mlp": 0.01030932, + "balance_loss_clip": 1.04191458, + "balance_loss_mlp": 1.02347505, + "epoch": 0.7339626044610112, + "flos": 21687675217920.0, + "grad_norm": 1.8443729175368468, + "language_loss": 0.75952595, + "learning_rate": 6.974647747936109e-07, + "loss": 0.78116286, + "num_input_tokens_seen": 131146110, + "step": 6104, + "time_per_iteration": 2.855794906616211 + }, + { + "auxiliary_loss_clip": 0.01171544, + "auxiliary_loss_mlp": 0.00886199, + "balance_loss_clip": 1.04799378, + "balance_loss_mlp": 1.00074446, + "epoch": 0.7340828473516503, + "flos": 15268248282240.0, + "grad_norm": 2.0763150335233767, + "language_loss": 0.82383764, + "learning_rate": 6.968737524211039e-07, + "loss": 0.84441507, + "num_input_tokens_seen": 131162920, + "step": 6105, + "time_per_iteration": 2.612809181213379 + }, + { + "auxiliary_loss_clip": 0.01161448, + "auxiliary_loss_mlp": 0.01025919, + "balance_loss_clip": 1.04861593, + "balance_loss_mlp": 1.01743758, + "epoch": 0.7342030902422895, + "flos": 22930112701440.0, + "grad_norm": 1.8909769563796088, + "language_loss": 0.80388391, + "learning_rate": 6.962829277318132e-07, + "loss": 0.82575756, + "num_input_tokens_seen": 131182515, + "step": 6106, + "time_per_iteration": 2.7718451023101807 + }, + { + "auxiliary_loss_clip": 0.01164755, + "auxiliary_loss_mlp": 0.01024095, + "balance_loss_clip": 1.04895794, + "balance_loss_mlp": 1.01593471, + "epoch": 0.7343233331329285, + "flos": 25847890381440.0, + "grad_norm": 1.8137061825663383, + "language_loss": 0.83457893, + "learning_rate": 6.956923008153652e-07, + "loss": 0.85646743, + "num_input_tokens_seen": 131202280, + "step": 6107, + "time_per_iteration": 2.7167534828186035 + }, + { + "auxiliary_loss_clip": 0.01162915, + "auxiliary_loss_mlp": 0.01025657, + "balance_loss_clip": 1.04630458, + "balance_loss_mlp": 1.01789308, + "epoch": 0.7344435760235676, + "flos": 18478985287680.0, + "grad_norm": 3.1258008585229966, + "language_loss": 0.84159839, + "learning_rate": 6.951018717613593e-07, + "loss": 0.86348403, + "num_input_tokens_seen": 131221295, + "step": 6108, + "time_per_iteration": 2.7170073986053467 + }, + { + "auxiliary_loss_clip": 0.01160764, + "auxiliary_loss_mlp": 0.01031169, + "balance_loss_clip": 1.04639482, + "balance_loss_mlp": 1.02347159, + "epoch": 0.7345638189142067, + "flos": 17640040256640.0, + "grad_norm": 1.8162305318301066, + "language_loss": 0.7818898, + "learning_rate": 6.945116406593614e-07, + "loss": 0.80380911, + "num_input_tokens_seen": 131240150, + "step": 6109, + "time_per_iteration": 2.7001898288726807 + }, + { + "auxiliary_loss_clip": 0.01138864, + "auxiliary_loss_mlp": 0.01032066, + "balance_loss_clip": 1.04511285, + "balance_loss_mlp": 1.02350712, + "epoch": 0.7346840618048458, + "flos": 20260225756800.0, + "grad_norm": 4.189343962595735, + "language_loss": 0.74099302, + "learning_rate": 6.939216075989089e-07, + "loss": 0.76270235, + "num_input_tokens_seen": 131258080, + "step": 6110, + "time_per_iteration": 2.847627878189087 + }, + { + "auxiliary_loss_clip": 0.01150522, + "auxiliary_loss_mlp": 0.01034157, + "balance_loss_clip": 1.04355955, + "balance_loss_mlp": 1.02553201, + "epoch": 0.7348043046954849, + "flos": 29023183641600.0, + "grad_norm": 2.116586599434844, + "language_loss": 0.66081977, + "learning_rate": 6.933317726695109e-07, + "loss": 0.68266654, + "num_input_tokens_seen": 131279310, + "step": 6111, + "time_per_iteration": 2.7671427726745605 + }, + { + "auxiliary_loss_clip": 0.01140125, + "auxiliary_loss_mlp": 0.01029543, + "balance_loss_clip": 1.04603004, + "balance_loss_mlp": 1.02157986, + "epoch": 0.734924547586124, + "flos": 17931203902080.0, + "grad_norm": 2.3295712362324377, + "language_loss": 0.80382097, + "learning_rate": 6.92742135960644e-07, + "loss": 0.82551771, + "num_input_tokens_seen": 131297010, + "step": 6112, + "time_per_iteration": 2.800184488296509 + }, + { + "auxiliary_loss_clip": 0.0105431, + "auxiliary_loss_mlp": 0.01001903, + "balance_loss_clip": 1.00747681, + "balance_loss_mlp": 1.00034165, + "epoch": 0.7350447904767631, + "flos": 63588319850880.0, + "grad_norm": 0.8324033238708543, + "language_loss": 0.55717903, + "learning_rate": 6.921526975617556e-07, + "loss": 0.57774115, + "num_input_tokens_seen": 131356470, + "step": 6113, + "time_per_iteration": 3.3522300720214844 + }, + { + "auxiliary_loss_clip": 0.01147629, + "auxiliary_loss_mlp": 0.01029647, + "balance_loss_clip": 1.04374075, + "balance_loss_mlp": 1.02078426, + "epoch": 0.7351650333674021, + "flos": 21580015178880.0, + "grad_norm": 1.7370722718784315, + "language_loss": 0.75275666, + "learning_rate": 6.915634575622631e-07, + "loss": 0.77452934, + "num_input_tokens_seen": 131374985, + "step": 6114, + "time_per_iteration": 3.7267162799835205 + }, + { + "auxiliary_loss_clip": 0.01169853, + "auxiliary_loss_mlp": 0.01031926, + "balance_loss_clip": 1.04690874, + "balance_loss_mlp": 1.02398705, + "epoch": 0.7352852762580413, + "flos": 18186349184640.0, + "grad_norm": 2.0668166400011105, + "language_loss": 0.70862103, + "learning_rate": 6.909744160515532e-07, + "loss": 0.73063886, + "num_input_tokens_seen": 131393125, + "step": 6115, + "time_per_iteration": 2.7241320610046387 + }, + { + "auxiliary_loss_clip": 0.01154133, + "auxiliary_loss_mlp": 0.01027082, + "balance_loss_clip": 1.04746783, + "balance_loss_mlp": 1.01890445, + "epoch": 0.7354055191486804, + "flos": 38910073063680.0, + "grad_norm": 1.993921737766611, + "language_loss": 0.69632781, + "learning_rate": 6.903855731189849e-07, + "loss": 0.71814001, + "num_input_tokens_seen": 131415760, + "step": 6116, + "time_per_iteration": 3.8498799800872803 + }, + { + "auxiliary_loss_clip": 0.01163956, + "auxiliary_loss_mlp": 0.01025862, + "balance_loss_clip": 1.04903018, + "balance_loss_mlp": 1.01720762, + "epoch": 0.7355257620393194, + "flos": 16289978647680.0, + "grad_norm": 2.4247564185404067, + "language_loss": 0.81740111, + "learning_rate": 6.897969288538825e-07, + "loss": 0.83929932, + "num_input_tokens_seen": 131433705, + "step": 6117, + "time_per_iteration": 3.6464598178863525 + }, + { + "auxiliary_loss_clip": 0.01149117, + "auxiliary_loss_mlp": 0.01031186, + "balance_loss_clip": 1.0444876, + "balance_loss_mlp": 1.02372301, + "epoch": 0.7356460049299585, + "flos": 18114240631680.0, + "grad_norm": 2.8024962619127285, + "language_loss": 0.81718773, + "learning_rate": 6.892084833455452e-07, + "loss": 0.83899081, + "num_input_tokens_seen": 131453275, + "step": 6118, + "time_per_iteration": 2.8376073837280273 + }, + { + "auxiliary_loss_clip": 0.01162161, + "auxiliary_loss_mlp": 0.01026954, + "balance_loss_clip": 1.04888773, + "balance_loss_mlp": 1.01906514, + "epoch": 0.7357662478205976, + "flos": 21325193118720.0, + "grad_norm": 1.6111473542046952, + "language_loss": 0.83692455, + "learning_rate": 6.886202366832384e-07, + "loss": 0.85881567, + "num_input_tokens_seen": 131474960, + "step": 6119, + "time_per_iteration": 2.712639808654785 + }, + { + "auxiliary_loss_clip": 0.0113143, + "auxiliary_loss_mlp": 0.01026257, + "balance_loss_clip": 1.04505861, + "balance_loss_mlp": 1.01852012, + "epoch": 0.7358864907112367, + "flos": 14246841139200.0, + "grad_norm": 1.7413558351182954, + "language_loss": 0.73283845, + "learning_rate": 6.880321889561987e-07, + "loss": 0.75441533, + "num_input_tokens_seen": 131492935, + "step": 6120, + "time_per_iteration": 2.8253931999206543 + }, + { + "auxiliary_loss_clip": 0.01138544, + "auxiliary_loss_mlp": 0.01029516, + "balance_loss_clip": 1.04416752, + "balance_loss_mlp": 1.02062261, + "epoch": 0.7360067336018757, + "flos": 22309684058880.0, + "grad_norm": 2.1633062248793884, + "language_loss": 0.65448731, + "learning_rate": 6.874443402536338e-07, + "loss": 0.67616791, + "num_input_tokens_seen": 131512025, + "step": 6121, + "time_per_iteration": 3.7302770614624023 + }, + { + "auxiliary_loss_clip": 0.01158632, + "auxiliary_loss_mlp": 0.01032535, + "balance_loss_clip": 1.04834414, + "balance_loss_mlp": 1.0241127, + "epoch": 0.7361269764925149, + "flos": 25554607833600.0, + "grad_norm": 1.6695754748944789, + "language_loss": 0.80312079, + "learning_rate": 6.868566906647177e-07, + "loss": 0.82503247, + "num_input_tokens_seen": 131532975, + "step": 6122, + "time_per_iteration": 2.8968751430511475 + }, + { + "auxiliary_loss_clip": 0.0116085, + "auxiliary_loss_mlp": 0.01034507, + "balance_loss_clip": 1.04472613, + "balance_loss_mlp": 1.02619243, + "epoch": 0.736247219383154, + "flos": 20376505059840.0, + "grad_norm": 1.8978793784982042, + "language_loss": 0.83576524, + "learning_rate": 6.862692402785984e-07, + "loss": 0.85771883, + "num_input_tokens_seen": 131553225, + "step": 6123, + "time_per_iteration": 2.921787977218628 + }, + { + "auxiliary_loss_clip": 0.01036158, + "auxiliary_loss_mlp": 0.01003762, + "balance_loss_clip": 1.00718844, + "balance_loss_mlp": 1.00208116, + "epoch": 0.736367462273793, + "flos": 70339525735680.0, + "grad_norm": 0.6819532797550607, + "language_loss": 0.4961856, + "learning_rate": 6.856819891843899e-07, + "loss": 0.51658481, + "num_input_tokens_seen": 131617930, + "step": 6124, + "time_per_iteration": 3.519895076751709 + }, + { + "auxiliary_loss_clip": 0.01128795, + "auxiliary_loss_mlp": 0.01028005, + "balance_loss_clip": 1.04393864, + "balance_loss_mlp": 1.0199821, + "epoch": 0.7364877051644322, + "flos": 22412711243520.0, + "grad_norm": 2.0441840262867625, + "language_loss": 0.71988291, + "learning_rate": 6.8509493747118e-07, + "loss": 0.74145091, + "num_input_tokens_seen": 131636740, + "step": 6125, + "time_per_iteration": 3.104764461517334 + }, + { + "auxiliary_loss_clip": 0.01172224, + "auxiliary_loss_mlp": 0.01031204, + "balance_loss_clip": 1.04971385, + "balance_loss_mlp": 1.022964, + "epoch": 0.7366079480550712, + "flos": 12130266274560.0, + "grad_norm": 2.1785787204090075, + "language_loss": 0.88188463, + "learning_rate": 6.845080852280221e-07, + "loss": 0.90391898, + "num_input_tokens_seen": 131653810, + "step": 6126, + "time_per_iteration": 2.776871919631958 + }, + { + "auxiliary_loss_clip": 0.01142766, + "auxiliary_loss_mlp": 0.01024482, + "balance_loss_clip": 1.04282618, + "balance_loss_mlp": 1.01707363, + "epoch": 0.7367281909457103, + "flos": 15049336844160.0, + "grad_norm": 1.5408095531041777, + "language_loss": 0.74219507, + "learning_rate": 6.839214325439409e-07, + "loss": 0.7638675, + "num_input_tokens_seen": 131671505, + "step": 6127, + "time_per_iteration": 2.725992441177368 + }, + { + "auxiliary_loss_clip": 0.01150321, + "auxiliary_loss_mlp": 0.01026016, + "balance_loss_clip": 1.04768348, + "balance_loss_mlp": 1.01841676, + "epoch": 0.7368484338363495, + "flos": 23510752053120.0, + "grad_norm": 1.6628948435619446, + "language_loss": 0.71903229, + "learning_rate": 6.833349795079327e-07, + "loss": 0.74079567, + "num_input_tokens_seen": 131690615, + "step": 6128, + "time_per_iteration": 2.74600887298584 + }, + { + "auxiliary_loss_clip": 0.01141968, + "auxiliary_loss_mlp": 0.01025191, + "balance_loss_clip": 1.04618847, + "balance_loss_mlp": 1.01710892, + "epoch": 0.7369686767269885, + "flos": 27417833095680.0, + "grad_norm": 1.801830741086987, + "language_loss": 0.68685454, + "learning_rate": 6.827487262089613e-07, + "loss": 0.70852619, + "num_input_tokens_seen": 131711120, + "step": 6129, + "time_per_iteration": 2.8155598640441895 + }, + { + "auxiliary_loss_clip": 0.01044949, + "auxiliary_loss_mlp": 0.0100268, + "balance_loss_clip": 1.00770879, + "balance_loss_mlp": 1.00122583, + "epoch": 0.7370889196176276, + "flos": 70293343824000.0, + "grad_norm": 0.881066203219365, + "language_loss": 0.5680564, + "learning_rate": 6.821626727359606e-07, + "loss": 0.58853269, + "num_input_tokens_seen": 131776680, + "step": 6130, + "time_per_iteration": 3.388101816177368 + }, + { + "auxiliary_loss_clip": 0.0115203, + "auxiliary_loss_mlp": 0.01029493, + "balance_loss_clip": 1.04813874, + "balance_loss_mlp": 1.02121985, + "epoch": 0.7372091625082667, + "flos": 18040839189120.0, + "grad_norm": 2.0596815200308543, + "language_loss": 0.76705301, + "learning_rate": 6.815768191778348e-07, + "loss": 0.78886825, + "num_input_tokens_seen": 131794760, + "step": 6131, + "time_per_iteration": 2.8083863258361816 + }, + { + "auxiliary_loss_clip": 0.0115937, + "auxiliary_loss_mlp": 0.01030764, + "balance_loss_clip": 1.04584241, + "balance_loss_mlp": 1.02264023, + "epoch": 0.7373294053989058, + "flos": 33726331854720.0, + "grad_norm": 1.6747808896910787, + "language_loss": 0.72991288, + "learning_rate": 6.809911656234569e-07, + "loss": 0.75181425, + "num_input_tokens_seen": 131816735, + "step": 6132, + "time_per_iteration": 2.8575069904327393 + }, + { + "auxiliary_loss_clip": 0.01145079, + "auxiliary_loss_mlp": 0.01029885, + "balance_loss_clip": 1.04142654, + "balance_loss_mlp": 1.02208865, + "epoch": 0.7374496482895448, + "flos": 21506326427520.0, + "grad_norm": 2.1993246271864844, + "language_loss": 0.78489578, + "learning_rate": 6.804057121616707e-07, + "loss": 0.80664539, + "num_input_tokens_seen": 131834940, + "step": 6133, + "time_per_iteration": 2.8474605083465576 + }, + { + "auxiliary_loss_clip": 0.01163511, + "auxiliary_loss_mlp": 0.01022193, + "balance_loss_clip": 1.04691911, + "balance_loss_mlp": 1.01439404, + "epoch": 0.737569891180184, + "flos": 24936908624640.0, + "grad_norm": 2.274087788984432, + "language_loss": 0.72372097, + "learning_rate": 6.798204588812888e-07, + "loss": 0.74557805, + "num_input_tokens_seen": 131854355, + "step": 6134, + "time_per_iteration": 2.7099502086639404 + }, + { + "auxiliary_loss_clip": 0.01113908, + "auxiliary_loss_mlp": 0.00885852, + "balance_loss_clip": 1.03879905, + "balance_loss_mlp": 1.00076151, + "epoch": 0.7376901340708231, + "flos": 20664544222080.0, + "grad_norm": 1.6142112999473759, + "language_loss": 0.7509104, + "learning_rate": 6.792354058710937e-07, + "loss": 0.770908, + "num_input_tokens_seen": 131871825, + "step": 6135, + "time_per_iteration": 2.9509315490722656 + }, + { + "auxiliary_loss_clip": 0.01165182, + "auxiliary_loss_mlp": 0.0102759, + "balance_loss_clip": 1.04574704, + "balance_loss_mlp": 1.02021122, + "epoch": 0.7378103769614621, + "flos": 23805794367360.0, + "grad_norm": 1.9134004528456023, + "language_loss": 0.65117466, + "learning_rate": 6.786505532198374e-07, + "loss": 0.67310238, + "num_input_tokens_seen": 131890770, + "step": 6136, + "time_per_iteration": 2.746309518814087 + }, + { + "auxiliary_loss_clip": 0.01173339, + "auxiliary_loss_mlp": 0.01026459, + "balance_loss_clip": 1.04971182, + "balance_loss_mlp": 1.01806676, + "epoch": 0.7379306198521013, + "flos": 22237216369920.0, + "grad_norm": 1.704608598294317, + "language_loss": 0.85302603, + "learning_rate": 6.780659010162411e-07, + "loss": 0.87502396, + "num_input_tokens_seen": 131909720, + "step": 6137, + "time_per_iteration": 2.647900104522705 + }, + { + "auxiliary_loss_clip": 0.01148839, + "auxiliary_loss_mlp": 0.01028776, + "balance_loss_clip": 1.04628599, + "balance_loss_mlp": 1.02158499, + "epoch": 0.7380508627427403, + "flos": 14903108576640.0, + "grad_norm": 1.7469178059452333, + "language_loss": 0.83599812, + "learning_rate": 6.774814493489975e-07, + "loss": 0.85777432, + "num_input_tokens_seen": 131927395, + "step": 6138, + "time_per_iteration": 2.7912404537200928 + }, + { + "auxiliary_loss_clip": 0.0115802, + "auxiliary_loss_mlp": 0.01026678, + "balance_loss_clip": 1.04684114, + "balance_loss_mlp": 1.01906037, + "epoch": 0.7381711056333794, + "flos": 21685843624320.0, + "grad_norm": 2.1675857242724628, + "language_loss": 0.66014576, + "learning_rate": 6.768971983067655e-07, + "loss": 0.68199265, + "num_input_tokens_seen": 131947725, + "step": 6139, + "time_per_iteration": 2.6599159240722656 + }, + { + "auxiliary_loss_clip": 0.01059682, + "auxiliary_loss_mlp": 0.01002514, + "balance_loss_clip": 1.00741553, + "balance_loss_mlp": 1.0009526, + "epoch": 0.7382913485240186, + "flos": 52404263596800.0, + "grad_norm": 1.0026674388096037, + "language_loss": 0.67822433, + "learning_rate": 6.763131479781772e-07, + "loss": 0.69884628, + "num_input_tokens_seen": 131997485, + "step": 6140, + "time_per_iteration": 3.9378552436828613 + }, + { + "auxiliary_loss_clip": 0.01144613, + "auxiliary_loss_mlp": 0.01031911, + "balance_loss_clip": 1.04664135, + "balance_loss_mlp": 1.02404928, + "epoch": 0.7384115914146576, + "flos": 21798818876160.0, + "grad_norm": 1.9617664831913533, + "language_loss": 0.75894547, + "learning_rate": 6.757292984518316e-07, + "loss": 0.7807107, + "num_input_tokens_seen": 132016885, + "step": 6141, + "time_per_iteration": 3.60658597946167 + }, + { + "auxiliary_loss_clip": 0.01055086, + "auxiliary_loss_mlp": 0.01004286, + "balance_loss_clip": 1.00785685, + "balance_loss_mlp": 1.00260544, + "epoch": 0.7385318343052967, + "flos": 61494331662720.0, + "grad_norm": 0.7509370436754255, + "language_loss": 0.56454277, + "learning_rate": 6.751456498162981e-07, + "loss": 0.58513653, + "num_input_tokens_seen": 132075920, + "step": 6142, + "time_per_iteration": 3.1212618350982666 + }, + { + "auxiliary_loss_clip": 0.01158291, + "auxiliary_loss_mlp": 0.01029959, + "balance_loss_clip": 1.04279006, + "balance_loss_mlp": 1.02213907, + "epoch": 0.7386520771959358, + "flos": 17013757697280.0, + "grad_norm": 1.8812378161702903, + "language_loss": 0.85526729, + "learning_rate": 6.745622021601174e-07, + "loss": 0.87714976, + "num_input_tokens_seen": 132092945, + "step": 6143, + "time_per_iteration": 3.579390525817871 + }, + { + "auxiliary_loss_clip": 0.01145885, + "auxiliary_loss_mlp": 0.01024774, + "balance_loss_clip": 1.04418874, + "balance_loss_mlp": 1.01670313, + "epoch": 0.7387723200865749, + "flos": 18770759464320.0, + "grad_norm": 2.335722636359681, + "language_loss": 0.69771868, + "learning_rate": 6.739789555717954e-07, + "loss": 0.71942532, + "num_input_tokens_seen": 132109920, + "step": 6144, + "time_per_iteration": 2.8906126022338867 + }, + { + "auxiliary_loss_clip": 0.01168626, + "auxiliary_loss_mlp": 0.01031492, + "balance_loss_clip": 1.04639006, + "balance_loss_mlp": 1.02389288, + "epoch": 0.738892562977214, + "flos": 22525542840960.0, + "grad_norm": 2.3443455777882143, + "language_loss": 0.773561, + "learning_rate": 6.733959101398124e-07, + "loss": 0.79556221, + "num_input_tokens_seen": 132128050, + "step": 6145, + "time_per_iteration": 2.6441702842712402 + }, + { + "auxiliary_loss_clip": 0.01146906, + "auxiliary_loss_mlp": 0.01026846, + "balance_loss_clip": 1.04139709, + "balance_loss_mlp": 1.01860857, + "epoch": 0.7390128058678531, + "flos": 21501478091520.0, + "grad_norm": 1.9601089276039436, + "language_loss": 0.81552881, + "learning_rate": 6.728130659526143e-07, + "loss": 0.83726633, + "num_input_tokens_seen": 132145860, + "step": 6146, + "time_per_iteration": 2.8708839416503906 + }, + { + "auxiliary_loss_clip": 0.01154979, + "auxiliary_loss_mlp": 0.01026637, + "balance_loss_clip": 1.0472765, + "balance_loss_mlp": 1.01886475, + "epoch": 0.7391330487584922, + "flos": 25776176878080.0, + "grad_norm": 2.511593010363277, + "language_loss": 0.70856935, + "learning_rate": 6.7223042309862e-07, + "loss": 0.73038554, + "num_input_tokens_seen": 132166060, + "step": 6147, + "time_per_iteration": 3.7004075050354004 + }, + { + "auxiliary_loss_clip": 0.01157347, + "auxiliary_loss_mlp": 0.01024595, + "balance_loss_clip": 1.04401624, + "balance_loss_mlp": 1.01687002, + "epoch": 0.7392532916491312, + "flos": 28366736636160.0, + "grad_norm": 2.2547892159321945, + "language_loss": 0.7373364, + "learning_rate": 6.716479816662144e-07, + "loss": 0.75915587, + "num_input_tokens_seen": 132187790, + "step": 6148, + "time_per_iteration": 2.7369444370269775 + }, + { + "auxiliary_loss_clip": 0.01154767, + "auxiliary_loss_mlp": 0.01027427, + "balance_loss_clip": 1.04450536, + "balance_loss_mlp": 1.01951122, + "epoch": 0.7393735345397703, + "flos": 23585877348480.0, + "grad_norm": 1.9691562291000781, + "language_loss": 0.73308837, + "learning_rate": 6.710657417437531e-07, + "loss": 0.75491035, + "num_input_tokens_seen": 132207495, + "step": 6149, + "time_per_iteration": 2.8187344074249268 + }, + { + "auxiliary_loss_clip": 0.0115069, + "auxiliary_loss_mlp": 0.01025292, + "balance_loss_clip": 1.04349232, + "balance_loss_mlp": 1.01778781, + "epoch": 0.7394937774304094, + "flos": 19974772373760.0, + "grad_norm": 2.0549551579196157, + "language_loss": 0.79896849, + "learning_rate": 6.704837034195628e-07, + "loss": 0.8207283, + "num_input_tokens_seen": 132225960, + "step": 6150, + "time_per_iteration": 2.736180067062378 + }, + { + "auxiliary_loss_clip": 0.01154564, + "auxiliary_loss_mlp": 0.0103234, + "balance_loss_clip": 1.04594207, + "balance_loss_mlp": 1.02418602, + "epoch": 0.7396140203210485, + "flos": 23478037741440.0, + "grad_norm": 2.0417280290465563, + "language_loss": 0.8457762, + "learning_rate": 6.699018667819376e-07, + "loss": 0.8676452, + "num_input_tokens_seen": 132245360, + "step": 6151, + "time_per_iteration": 2.7097842693328857 + }, + { + "auxiliary_loss_clip": 0.01157112, + "auxiliary_loss_mlp": 0.01027881, + "balance_loss_clip": 1.04473019, + "balance_loss_mlp": 1.01987028, + "epoch": 0.7397342632116876, + "flos": 25555433846400.0, + "grad_norm": 1.8060262277468826, + "language_loss": 0.72773725, + "learning_rate": 6.693202319191415e-07, + "loss": 0.74958718, + "num_input_tokens_seen": 132267095, + "step": 6152, + "time_per_iteration": 2.693312406539917 + }, + { + "auxiliary_loss_clip": 0.01171083, + "auxiliary_loss_mlp": 0.01035506, + "balance_loss_clip": 1.05055821, + "balance_loss_mlp": 1.0274291, + "epoch": 0.7398545061023267, + "flos": 24755021130240.0, + "grad_norm": 2.1420591448689024, + "language_loss": 0.752648, + "learning_rate": 6.687387989194084e-07, + "loss": 0.77471387, + "num_input_tokens_seen": 132286610, + "step": 6153, + "time_per_iteration": 2.6887049674987793 + }, + { + "auxiliary_loss_clip": 0.01144958, + "auxiliary_loss_mlp": 0.0102897, + "balance_loss_clip": 1.04449058, + "balance_loss_mlp": 1.0211494, + "epoch": 0.7399747489929658, + "flos": 16508602776960.0, + "grad_norm": 1.773467956444431, + "language_loss": 0.79394293, + "learning_rate": 6.681575678709404e-07, + "loss": 0.81568217, + "num_input_tokens_seen": 132305300, + "step": 6154, + "time_per_iteration": 2.6908793449401855 + }, + { + "auxiliary_loss_clip": 0.0115864, + "auxiliary_loss_mlp": 0.01027193, + "balance_loss_clip": 1.0469501, + "balance_loss_mlp": 1.01945615, + "epoch": 0.7400949918836048, + "flos": 24097065753600.0, + "grad_norm": 1.7917620129095941, + "language_loss": 0.70838785, + "learning_rate": 6.67576538861911e-07, + "loss": 0.73024613, + "num_input_tokens_seen": 132323875, + "step": 6155, + "time_per_iteration": 2.741016387939453 + }, + { + "auxiliary_loss_clip": 0.01148219, + "auxiliary_loss_mlp": 0.01027788, + "balance_loss_clip": 1.04419899, + "balance_loss_mlp": 1.01965165, + "epoch": 0.740215234774244, + "flos": 21802517976960.0, + "grad_norm": 1.6312294927619344, + "language_loss": 0.81968236, + "learning_rate": 6.669957119804612e-07, + "loss": 0.84144241, + "num_input_tokens_seen": 132345510, + "step": 6156, + "time_per_iteration": 2.7358336448669434 + }, + { + "auxiliary_loss_clip": 0.01157959, + "auxiliary_loss_mlp": 0.01027439, + "balance_loss_clip": 1.04563761, + "balance_loss_mlp": 1.01931512, + "epoch": 0.7403354776648831, + "flos": 18733196816640.0, + "grad_norm": 3.747407655178064, + "language_loss": 0.72308838, + "learning_rate": 6.66415087314702e-07, + "loss": 0.74494243, + "num_input_tokens_seen": 132360465, + "step": 6157, + "time_per_iteration": 2.7896621227264404 + }, + { + "auxiliary_loss_clip": 0.0115016, + "auxiliary_loss_mlp": 0.01025349, + "balance_loss_clip": 1.04230225, + "balance_loss_mlp": 1.01758206, + "epoch": 0.7404557205555221, + "flos": 16909581277440.0, + "grad_norm": 2.2144929159958284, + "language_loss": 0.73081535, + "learning_rate": 6.65834664952714e-07, + "loss": 0.75257045, + "num_input_tokens_seen": 132377915, + "step": 6158, + "time_per_iteration": 2.7761762142181396 + }, + { + "auxiliary_loss_clip": 0.01142595, + "auxiliary_loss_mlp": 0.01023296, + "balance_loss_clip": 1.04323077, + "balance_loss_mlp": 1.01563966, + "epoch": 0.7405759634461613, + "flos": 21214408596480.0, + "grad_norm": 4.342156645791408, + "language_loss": 0.760378, + "learning_rate": 6.652544449825457e-07, + "loss": 0.7820369, + "num_input_tokens_seen": 132398170, + "step": 6159, + "time_per_iteration": 2.916201114654541 + }, + { + "auxiliary_loss_clip": 0.01152985, + "auxiliary_loss_mlp": 0.01026202, + "balance_loss_clip": 1.04574597, + "balance_loss_mlp": 1.01744056, + "epoch": 0.7406962063368003, + "flos": 20480106862080.0, + "grad_norm": 1.7787438688774664, + "language_loss": 0.76395458, + "learning_rate": 6.646744274922182e-07, + "loss": 0.78574646, + "num_input_tokens_seen": 132416615, + "step": 6160, + "time_per_iteration": 2.8280205726623535 + }, + { + "auxiliary_loss_clip": 0.01150092, + "auxiliary_loss_mlp": 0.01025224, + "balance_loss_clip": 1.04232812, + "balance_loss_mlp": 1.01774907, + "epoch": 0.7408164492274394, + "flos": 19791915212160.0, + "grad_norm": 2.5392633556588606, + "language_loss": 0.75738907, + "learning_rate": 6.640946125697171e-07, + "loss": 0.7791422, + "num_input_tokens_seen": 132434145, + "step": 6161, + "time_per_iteration": 2.7206358909606934 + }, + { + "auxiliary_loss_clip": 0.01162796, + "auxiliary_loss_mlp": 0.01028874, + "balance_loss_clip": 1.04603577, + "balance_loss_mlp": 1.0203203, + "epoch": 0.7409366921180786, + "flos": 29204855654400.0, + "grad_norm": 1.9754049831520928, + "language_loss": 0.75369668, + "learning_rate": 6.635150003030017e-07, + "loss": 0.77561337, + "num_input_tokens_seen": 132452670, + "step": 6162, + "time_per_iteration": 2.8657126426696777 + }, + { + "auxiliary_loss_clip": 0.01122629, + "auxiliary_loss_mlp": 0.01024485, + "balance_loss_clip": 1.03756332, + "balance_loss_mlp": 1.01633108, + "epoch": 0.7410569350087176, + "flos": 22930004960640.0, + "grad_norm": 2.0719820158388442, + "language_loss": 0.86109453, + "learning_rate": 6.629355907799981e-07, + "loss": 0.88256562, + "num_input_tokens_seen": 132472475, + "step": 6163, + "time_per_iteration": 2.8910841941833496 + }, + { + "auxiliary_loss_clip": 0.01161776, + "auxiliary_loss_mlp": 0.01029017, + "balance_loss_clip": 1.04497159, + "balance_loss_mlp": 1.02011847, + "epoch": 0.7411771778993567, + "flos": 30440397726720.0, + "grad_norm": 1.7392216121227453, + "language_loss": 0.69324756, + "learning_rate": 6.623563840886015e-07, + "loss": 0.71515548, + "num_input_tokens_seen": 132493400, + "step": 6164, + "time_per_iteration": 2.8220698833465576 + }, + { + "auxiliary_loss_clip": 0.01154014, + "auxiliary_loss_mlp": 0.01028468, + "balance_loss_clip": 1.04293752, + "balance_loss_mlp": 1.02090967, + "epoch": 0.7412974207899958, + "flos": 20522050968960.0, + "grad_norm": 4.058637090820378, + "language_loss": 0.6965754, + "learning_rate": 6.617773803166795e-07, + "loss": 0.71840018, + "num_input_tokens_seen": 132511725, + "step": 6165, + "time_per_iteration": 2.6711041927337646 + }, + { + "auxiliary_loss_clip": 0.01155149, + "auxiliary_loss_mlp": 0.00887049, + "balance_loss_clip": 1.04596686, + "balance_loss_mlp": 1.0006727, + "epoch": 0.7414176636806349, + "flos": 22090700793600.0, + "grad_norm": 4.664798693416947, + "language_loss": 0.82395852, + "learning_rate": 6.611985795520634e-07, + "loss": 0.8443805, + "num_input_tokens_seen": 132530270, + "step": 6166, + "time_per_iteration": 3.692157030105591 + }, + { + "auxiliary_loss_clip": 0.01152983, + "auxiliary_loss_mlp": 0.01025208, + "balance_loss_clip": 1.04733455, + "balance_loss_mlp": 1.01713181, + "epoch": 0.7415379065712739, + "flos": 25155245445120.0, + "grad_norm": 1.9415052798074857, + "language_loss": 0.77149582, + "learning_rate": 6.606199818825588e-07, + "loss": 0.79327774, + "num_input_tokens_seen": 132550725, + "step": 6167, + "time_per_iteration": 2.793710470199585 + }, + { + "auxiliary_loss_clip": 0.01153536, + "auxiliary_loss_mlp": 0.01026373, + "balance_loss_clip": 1.04201698, + "balance_loss_mlp": 1.01851678, + "epoch": 0.7416581494619131, + "flos": 16871731320960.0, + "grad_norm": 2.0843279369260923, + "language_loss": 0.81822419, + "learning_rate": 6.600415873959377e-07, + "loss": 0.84002328, + "num_input_tokens_seen": 132568600, + "step": 6168, + "time_per_iteration": 4.669190168380737 + }, + { + "auxiliary_loss_clip": 0.01119163, + "auxiliary_loss_mlp": 0.00884807, + "balance_loss_clip": 1.03767657, + "balance_loss_mlp": 1.00067353, + "epoch": 0.7417783923525522, + "flos": 28438881102720.0, + "grad_norm": 1.8459838537590412, + "language_loss": 0.64536887, + "learning_rate": 6.594633961799437e-07, + "loss": 0.66540855, + "num_input_tokens_seen": 132587640, + "step": 6169, + "time_per_iteration": 2.9336414337158203 + }, + { + "auxiliary_loss_clip": 0.01151379, + "auxiliary_loss_mlp": 0.01027092, + "balance_loss_clip": 1.04575634, + "balance_loss_mlp": 1.0191704, + "epoch": 0.7418986352431912, + "flos": 20084299920000.0, + "grad_norm": 1.5565645619370276, + "language_loss": 0.81797761, + "learning_rate": 6.588854083222857e-07, + "loss": 0.83976227, + "num_input_tokens_seen": 132607075, + "step": 6170, + "time_per_iteration": 2.860833168029785 + }, + { + "auxiliary_loss_clip": 0.01158198, + "auxiliary_loss_mlp": 0.01031688, + "balance_loss_clip": 1.04710793, + "balance_loss_mlp": 1.02351046, + "epoch": 0.7420188781338304, + "flos": 18259571059200.0, + "grad_norm": 2.038959878944515, + "language_loss": 0.80928445, + "learning_rate": 6.583076239106444e-07, + "loss": 0.83118331, + "num_input_tokens_seen": 132625580, + "step": 6171, + "time_per_iteration": 2.715021848678589 + }, + { + "auxiliary_loss_clip": 0.01157008, + "auxiliary_loss_mlp": 0.01024881, + "balance_loss_clip": 1.04525006, + "balance_loss_mlp": 1.01711464, + "epoch": 0.7421391210244694, + "flos": 13771994319360.0, + "grad_norm": 2.0979205474261775, + "language_loss": 0.75417209, + "learning_rate": 6.577300430326707e-07, + "loss": 0.77599096, + "num_input_tokens_seen": 132640525, + "step": 6172, + "time_per_iteration": 2.820817232131958 + }, + { + "auxiliary_loss_clip": 0.01139012, + "auxiliary_loss_mlp": 0.01029732, + "balance_loss_clip": 1.04510927, + "balance_loss_mlp": 1.02179849, + "epoch": 0.7422593639151085, + "flos": 15961683317760.0, + "grad_norm": 4.70658075615799, + "language_loss": 0.71645355, + "learning_rate": 6.571526657759821e-07, + "loss": 0.73814094, + "num_input_tokens_seen": 132656265, + "step": 6173, + "time_per_iteration": 3.753612518310547 + }, + { + "auxiliary_loss_clip": 0.01156508, + "auxiliary_loss_mlp": 0.01020389, + "balance_loss_clip": 1.04530883, + "balance_loss_mlp": 1.01203847, + "epoch": 0.7423796068057477, + "flos": 30114400867200.0, + "grad_norm": 1.6267704273155053, + "language_loss": 0.70923162, + "learning_rate": 6.565754922281663e-07, + "loss": 0.7310006, + "num_input_tokens_seen": 132678510, + "step": 6174, + "time_per_iteration": 2.894134759902954 + }, + { + "auxiliary_loss_clip": 0.0114867, + "auxiliary_loss_mlp": 0.01021268, + "balance_loss_clip": 1.04304051, + "balance_loss_mlp": 1.01319158, + "epoch": 0.7424998496963867, + "flos": 20521907314560.0, + "grad_norm": 1.7496950627626926, + "language_loss": 0.78702891, + "learning_rate": 6.559985224767801e-07, + "loss": 0.80872822, + "num_input_tokens_seen": 132696385, + "step": 6175, + "time_per_iteration": 2.7775044441223145 + }, + { + "auxiliary_loss_clip": 0.01140265, + "auxiliary_loss_mlp": 0.01026148, + "balance_loss_clip": 1.04364085, + "balance_loss_mlp": 1.01813722, + "epoch": 0.7426200925870258, + "flos": 21871573873920.0, + "grad_norm": 2.365012966916098, + "language_loss": 0.75722992, + "learning_rate": 6.55421756609349e-07, + "loss": 0.77889407, + "num_input_tokens_seen": 132714640, + "step": 6176, + "time_per_iteration": 2.7911713123321533 + }, + { + "auxiliary_loss_clip": 0.01158042, + "auxiliary_loss_mlp": 0.01028903, + "balance_loss_clip": 1.04814172, + "balance_loss_mlp": 1.01958096, + "epoch": 0.7427403354776649, + "flos": 26432049265920.0, + "grad_norm": 1.7918792851272827, + "language_loss": 0.78552461, + "learning_rate": 6.54845194713369e-07, + "loss": 0.80739403, + "num_input_tokens_seen": 132735590, + "step": 6177, + "time_per_iteration": 2.763882875442505 + }, + { + "auxiliary_loss_clip": 0.01159238, + "auxiliary_loss_mlp": 0.01034563, + "balance_loss_clip": 1.04762614, + "balance_loss_mlp": 1.02683806, + "epoch": 0.742860578368304, + "flos": 19898390102400.0, + "grad_norm": 1.9916873825593193, + "language_loss": 0.79717857, + "learning_rate": 6.542688368763034e-07, + "loss": 0.81911659, + "num_input_tokens_seen": 132753995, + "step": 6178, + "time_per_iteration": 2.6949076652526855 + }, + { + "auxiliary_loss_clip": 0.01160183, + "auxiliary_loss_mlp": 0.01022186, + "balance_loss_clip": 1.04821742, + "balance_loss_mlp": 1.01472926, + "epoch": 0.742980821258943, + "flos": 24827201510400.0, + "grad_norm": 2.073522880429396, + "language_loss": 0.77159542, + "learning_rate": 6.536926831855854e-07, + "loss": 0.79341912, + "num_input_tokens_seen": 132773160, + "step": 6179, + "time_per_iteration": 2.781747341156006 + }, + { + "auxiliary_loss_clip": 0.01152289, + "auxiliary_loss_mlp": 0.01021701, + "balance_loss_clip": 1.04794908, + "balance_loss_mlp": 1.013767, + "epoch": 0.7431010641495821, + "flos": 25228646887680.0, + "grad_norm": 2.8674588769398777, + "language_loss": 0.73239273, + "learning_rate": 6.531167337286165e-07, + "loss": 0.75413263, + "num_input_tokens_seen": 132793180, + "step": 6180, + "time_per_iteration": 2.7448904514312744 + }, + { + "auxiliary_loss_clip": 0.01149257, + "auxiliary_loss_mlp": 0.01027806, + "balance_loss_clip": 1.04690111, + "balance_loss_mlp": 1.01989079, + "epoch": 0.7432213070402213, + "flos": 21762369550080.0, + "grad_norm": 1.7765773223120571, + "language_loss": 0.79797113, + "learning_rate": 6.52540988592768e-07, + "loss": 0.81974173, + "num_input_tokens_seen": 132814200, + "step": 6181, + "time_per_iteration": 2.711434841156006 + }, + { + "auxiliary_loss_clip": 0.01154125, + "auxiliary_loss_mlp": 0.01024023, + "balance_loss_clip": 1.04460549, + "balance_loss_mlp": 1.0161072, + "epoch": 0.7433415499308603, + "flos": 14793832425600.0, + "grad_norm": 2.2950771042686493, + "language_loss": 0.83370495, + "learning_rate": 6.519654478653814e-07, + "loss": 0.85548639, + "num_input_tokens_seen": 132832565, + "step": 6182, + "time_per_iteration": 2.854720115661621 + }, + { + "auxiliary_loss_clip": 0.0104971, + "auxiliary_loss_mlp": 0.0100388, + "balance_loss_clip": 1.00593328, + "balance_loss_mlp": 1.0023545, + "epoch": 0.7434617928214994, + "flos": 67155577297920.0, + "grad_norm": 0.7421869109850868, + "language_loss": 0.5607003, + "learning_rate": 6.51390111633763e-07, + "loss": 0.58123618, + "num_input_tokens_seen": 132897840, + "step": 6183, + "time_per_iteration": 3.321390151977539 + }, + { + "auxiliary_loss_clip": 0.0112165, + "auxiliary_loss_mlp": 0.01027798, + "balance_loss_clip": 1.03766918, + "balance_loss_mlp": 1.02006149, + "epoch": 0.7435820357121385, + "flos": 27377576928000.0, + "grad_norm": 1.5416525640742256, + "language_loss": 0.76184469, + "learning_rate": 6.508149799851932e-07, + "loss": 0.78333914, + "num_input_tokens_seen": 132919505, + "step": 6184, + "time_per_iteration": 2.915224313735962 + }, + { + "auxiliary_loss_clip": 0.01147934, + "auxiliary_loss_mlp": 0.01027969, + "balance_loss_clip": 1.04613066, + "balance_loss_mlp": 1.02059019, + "epoch": 0.7437022786027776, + "flos": 23987645948160.0, + "grad_norm": 3.0004780799350805, + "language_loss": 0.61561096, + "learning_rate": 6.502400530069183e-07, + "loss": 0.63736999, + "num_input_tokens_seen": 132939390, + "step": 6185, + "time_per_iteration": 2.7405693531036377 + }, + { + "auxiliary_loss_clip": 0.01143665, + "auxiliary_loss_mlp": 0.01033689, + "balance_loss_clip": 1.04438436, + "balance_loss_mlp": 1.02529931, + "epoch": 0.7438225214934167, + "flos": 21866761451520.0, + "grad_norm": 1.5266936930451855, + "language_loss": 0.68705487, + "learning_rate": 6.496653307861535e-07, + "loss": 0.70882839, + "num_input_tokens_seen": 132960060, + "step": 6186, + "time_per_iteration": 2.8318018913269043 + }, + { + "auxiliary_loss_clip": 0.01164521, + "auxiliary_loss_mlp": 0.0102804, + "balance_loss_clip": 1.04520941, + "balance_loss_mlp": 1.01949906, + "epoch": 0.7439427643840558, + "flos": 20230097224320.0, + "grad_norm": 1.8924612923228827, + "language_loss": 0.66018581, + "learning_rate": 6.490908134100857e-07, + "loss": 0.68211144, + "num_input_tokens_seen": 132978525, + "step": 6187, + "time_per_iteration": 2.794602155685425 + }, + { + "auxiliary_loss_clip": 0.01165291, + "auxiliary_loss_mlp": 0.01028637, + "balance_loss_clip": 1.04658675, + "balance_loss_mlp": 1.02041125, + "epoch": 0.7440630072746949, + "flos": 20849915335680.0, + "grad_norm": 2.108816062085017, + "language_loss": 0.69295591, + "learning_rate": 6.48516500965866e-07, + "loss": 0.71489519, + "num_input_tokens_seen": 132998460, + "step": 6188, + "time_per_iteration": 2.7111704349517822 + }, + { + "auxiliary_loss_clip": 0.01161406, + "auxiliary_loss_mlp": 0.0102833, + "balance_loss_clip": 1.04174554, + "balance_loss_mlp": 1.02048564, + "epoch": 0.7441832501653339, + "flos": 26503762769280.0, + "grad_norm": 1.6035034331189113, + "language_loss": 0.81341308, + "learning_rate": 6.479423935406192e-07, + "loss": 0.8353104, + "num_input_tokens_seen": 133018445, + "step": 6189, + "time_per_iteration": 2.7881510257720947 + }, + { + "auxiliary_loss_clip": 0.01048077, + "auxiliary_loss_mlp": 0.01003157, + "balance_loss_clip": 1.01408935, + "balance_loss_mlp": 1.00178015, + "epoch": 0.7443034930559731, + "flos": 68602848088320.0, + "grad_norm": 0.8329465550642631, + "language_loss": 0.62055409, + "learning_rate": 6.473684912214357e-07, + "loss": 0.64106643, + "num_input_tokens_seen": 133082005, + "step": 6190, + "time_per_iteration": 3.476379156112671 + }, + { + "auxiliary_loss_clip": 0.01164016, + "auxiliary_loss_mlp": 0.01027192, + "balance_loss_clip": 1.04977906, + "balance_loss_mlp": 1.0192858, + "epoch": 0.7444237359466122, + "flos": 18654982951680.0, + "grad_norm": 1.8844173793318493, + "language_loss": 0.69400471, + "learning_rate": 6.467947940953778e-07, + "loss": 0.71591675, + "num_input_tokens_seen": 133100530, + "step": 6191, + "time_per_iteration": 2.776597499847412 + }, + { + "auxiliary_loss_clip": 0.01149044, + "auxiliary_loss_mlp": 0.01025327, + "balance_loss_clip": 1.04357123, + "balance_loss_mlp": 1.01816273, + "epoch": 0.7445439788372512, + "flos": 22817604326400.0, + "grad_norm": 1.762934475520245, + "language_loss": 0.72386014, + "learning_rate": 6.462213022494732e-07, + "loss": 0.74560386, + "num_input_tokens_seen": 133119775, + "step": 6192, + "time_per_iteration": 3.767618179321289 + }, + { + "auxiliary_loss_clip": 0.01054628, + "auxiliary_loss_mlp": 0.010029, + "balance_loss_clip": 1.00672889, + "balance_loss_mlp": 1.00138044, + "epoch": 0.7446642217278904, + "flos": 67045690615680.0, + "grad_norm": 0.77087614311373, + "language_loss": 0.61035717, + "learning_rate": 6.456480157707201e-07, + "loss": 0.63093245, + "num_input_tokens_seen": 133184550, + "step": 6193, + "time_per_iteration": 3.2342207431793213 + }, + { + "auxiliary_loss_clip": 0.01134659, + "auxiliary_loss_mlp": 0.0102599, + "balance_loss_clip": 1.04178023, + "balance_loss_mlp": 1.01817584, + "epoch": 0.7447844646185294, + "flos": 17417465631360.0, + "grad_norm": 1.844262528934812, + "language_loss": 0.85030937, + "learning_rate": 6.450749347460866e-07, + "loss": 0.87191582, + "num_input_tokens_seen": 133201525, + "step": 6194, + "time_per_iteration": 4.622639179229736 + }, + { + "auxiliary_loss_clip": 0.01171219, + "auxiliary_loss_mlp": 0.01030833, + "balance_loss_clip": 1.04766536, + "balance_loss_mlp": 1.0227685, + "epoch": 0.7449047075091685, + "flos": 26615876094720.0, + "grad_norm": 2.102626221400523, + "language_loss": 0.7887758, + "learning_rate": 6.445020592625083e-07, + "loss": 0.81079638, + "num_input_tokens_seen": 133222175, + "step": 6195, + "time_per_iteration": 2.7356889247894287 + }, + { + "auxiliary_loss_clip": 0.01168113, + "auxiliary_loss_mlp": 0.01026976, + "balance_loss_clip": 1.04561138, + "balance_loss_mlp": 1.01927841, + "epoch": 0.7450249503998077, + "flos": 14170458867840.0, + "grad_norm": 3.560165551884857, + "language_loss": 0.80285823, + "learning_rate": 6.4392938940689e-07, + "loss": 0.82480913, + "num_input_tokens_seen": 133237590, + "step": 6196, + "time_per_iteration": 2.6456589698791504 + }, + { + "auxiliary_loss_clip": 0.01127201, + "auxiliary_loss_mlp": 0.00885903, + "balance_loss_clip": 1.0414989, + "balance_loss_mlp": 1.00059736, + "epoch": 0.7451451932904467, + "flos": 19606687752960.0, + "grad_norm": 2.3932088009631114, + "language_loss": 0.71251959, + "learning_rate": 6.433569252661049e-07, + "loss": 0.73265064, + "num_input_tokens_seen": 133255590, + "step": 6197, + "time_per_iteration": 3.0090525150299072 + }, + { + "auxiliary_loss_clip": 0.0113805, + "auxiliary_loss_mlp": 0.01028492, + "balance_loss_clip": 1.04353666, + "balance_loss_mlp": 1.02139664, + "epoch": 0.7452654361810858, + "flos": 12495405980160.0, + "grad_norm": 1.7980500043785743, + "language_loss": 0.71220672, + "learning_rate": 6.427846669269952e-07, + "loss": 0.73387218, + "num_input_tokens_seen": 133273210, + "step": 6198, + "time_per_iteration": 2.7774837017059326 + }, + { + "auxiliary_loss_clip": 0.01173034, + "auxiliary_loss_mlp": 0.01027798, + "balance_loss_clip": 1.05065489, + "balance_loss_mlp": 1.02001321, + "epoch": 0.7453856790717249, + "flos": 22127329687680.0, + "grad_norm": 2.111953831784675, + "language_loss": 0.82295614, + "learning_rate": 6.422126144763729e-07, + "loss": 0.84496444, + "num_input_tokens_seen": 133292600, + "step": 6199, + "time_per_iteration": 3.61259126663208 + }, + { + "auxiliary_loss_clip": 0.01141814, + "auxiliary_loss_mlp": 0.00886464, + "balance_loss_clip": 1.04107046, + "balance_loss_mlp": 1.00063324, + "epoch": 0.745505921962364, + "flos": 20010682995840.0, + "grad_norm": 2.1679467335502767, + "language_loss": 0.7686106, + "learning_rate": 6.416407680010174e-07, + "loss": 0.78889334, + "num_input_tokens_seen": 133306960, + "step": 6200, + "time_per_iteration": 2.887047529220581 + }, + { + "auxiliary_loss_clip": 0.01138853, + "auxiliary_loss_mlp": 0.01026081, + "balance_loss_clip": 1.04603302, + "balance_loss_mlp": 1.01827884, + "epoch": 0.745626164853003, + "flos": 24677884673280.0, + "grad_norm": 2.0609600749913533, + "language_loss": 0.812814, + "learning_rate": 6.410691275876774e-07, + "loss": 0.83446336, + "num_input_tokens_seen": 133326380, + "step": 6201, + "time_per_iteration": 2.8394362926483154 + }, + { + "auxiliary_loss_clip": 0.01152933, + "auxiliary_loss_mlp": 0.01033401, + "balance_loss_clip": 1.04758716, + "balance_loss_mlp": 1.02522612, + "epoch": 0.7457464077436422, + "flos": 14538830797440.0, + "grad_norm": 2.194808766740119, + "language_loss": 0.76516449, + "learning_rate": 6.404976933230704e-07, + "loss": 0.78702784, + "num_input_tokens_seen": 133342900, + "step": 6202, + "time_per_iteration": 2.7381765842437744 + }, + { + "auxiliary_loss_clip": 0.0115663, + "auxiliary_loss_mlp": 0.01030931, + "balance_loss_clip": 1.04421568, + "balance_loss_mlp": 1.02268195, + "epoch": 0.7458666506342813, + "flos": 34021194600960.0, + "grad_norm": 2.0025842264642444, + "language_loss": 0.7278986, + "learning_rate": 6.399264652938813e-07, + "loss": 0.74977428, + "num_input_tokens_seen": 133363805, + "step": 6203, + "time_per_iteration": 2.8288400173187256 + }, + { + "auxiliary_loss_clip": 0.01151863, + "auxiliary_loss_mlp": 0.01025346, + "balance_loss_clip": 1.04523206, + "balance_loss_mlp": 1.01769924, + "epoch": 0.7459868935249203, + "flos": 24279025075200.0, + "grad_norm": 2.0634932740446015, + "language_loss": 0.7443105, + "learning_rate": 6.393554435867679e-07, + "loss": 0.76608253, + "num_input_tokens_seen": 133384655, + "step": 6204, + "time_per_iteration": 2.7915492057800293 + }, + { + "auxiliary_loss_clip": 0.01136221, + "auxiliary_loss_mlp": 0.01031324, + "balance_loss_clip": 1.03997946, + "balance_loss_mlp": 1.02282441, + "epoch": 0.7461071364155595, + "flos": 21908777385600.0, + "grad_norm": 1.974585569117103, + "language_loss": 0.83870852, + "learning_rate": 6.387846282883502e-07, + "loss": 0.86038393, + "num_input_tokens_seen": 133401185, + "step": 6205, + "time_per_iteration": 2.867932081222534 + }, + { + "auxiliary_loss_clip": 0.01169521, + "auxiliary_loss_mlp": 0.01022527, + "balance_loss_clip": 1.04829836, + "balance_loss_mlp": 1.01531768, + "epoch": 0.7462273793061985, + "flos": 22889712879360.0, + "grad_norm": 2.1088398373195467, + "language_loss": 0.76890349, + "learning_rate": 6.38214019485223e-07, + "loss": 0.79082394, + "num_input_tokens_seen": 133420010, + "step": 6206, + "time_per_iteration": 2.6530885696411133 + }, + { + "auxiliary_loss_clip": 0.01117974, + "auxiliary_loss_mlp": 0.01023855, + "balance_loss_clip": 1.04038882, + "balance_loss_mlp": 1.01601744, + "epoch": 0.7463476221968376, + "flos": 19968451580160.0, + "grad_norm": 1.7653574620691619, + "language_loss": 0.71565431, + "learning_rate": 6.376436172639461e-07, + "loss": 0.73707259, + "num_input_tokens_seen": 133437855, + "step": 6207, + "time_per_iteration": 3.049100160598755 + }, + { + "auxiliary_loss_clip": 0.01108489, + "auxiliary_loss_mlp": 0.0102751, + "balance_loss_clip": 1.04140508, + "balance_loss_mlp": 1.01939178, + "epoch": 0.7464678650874768, + "flos": 16836610798080.0, + "grad_norm": 2.6065536235767572, + "language_loss": 0.65467775, + "learning_rate": 6.370734217110487e-07, + "loss": 0.67603773, + "num_input_tokens_seen": 133456600, + "step": 6208, + "time_per_iteration": 2.8847293853759766 + }, + { + "auxiliary_loss_clip": 0.0115714, + "auxiliary_loss_mlp": 0.0103278, + "balance_loss_clip": 1.04883075, + "balance_loss_mlp": 1.0244112, + "epoch": 0.7465881079781158, + "flos": 48100869843840.0, + "grad_norm": 1.4473815803699097, + "language_loss": 0.64193314, + "learning_rate": 6.36503432913031e-07, + "loss": 0.66383231, + "num_input_tokens_seen": 133479745, + "step": 6209, + "time_per_iteration": 2.9752066135406494 + }, + { + "auxiliary_loss_clip": 0.01161806, + "auxiliary_loss_mlp": 0.01030061, + "balance_loss_clip": 1.0468061, + "balance_loss_mlp": 1.02219915, + "epoch": 0.7467083508687549, + "flos": 19677359761920.0, + "grad_norm": 2.331301239867833, + "language_loss": 0.68830389, + "learning_rate": 6.359336509563569e-07, + "loss": 0.71022254, + "num_input_tokens_seen": 133495765, + "step": 6210, + "time_per_iteration": 2.6550090312957764 + }, + { + "auxiliary_loss_clip": 0.01131292, + "auxiliary_loss_mlp": 0.01027348, + "balance_loss_clip": 1.0426898, + "balance_loss_mlp": 1.01940286, + "epoch": 0.7468285937593939, + "flos": 17895436934400.0, + "grad_norm": 1.7877244579366742, + "language_loss": 0.80809784, + "learning_rate": 6.353640759274641e-07, + "loss": 0.8296842, + "num_input_tokens_seen": 133514655, + "step": 6211, + "time_per_iteration": 2.8385186195373535 + }, + { + "auxiliary_loss_clip": 0.01158012, + "auxiliary_loss_mlp": 0.01025433, + "balance_loss_clip": 1.04367423, + "balance_loss_mlp": 1.01769042, + "epoch": 0.7469488366500331, + "flos": 23141446369920.0, + "grad_norm": 2.2572043660712113, + "language_loss": 0.75365031, + "learning_rate": 6.347947079127556e-07, + "loss": 0.77548474, + "num_input_tokens_seen": 133532555, + "step": 6212, + "time_per_iteration": 2.9730892181396484 + }, + { + "auxiliary_loss_clip": 0.01143778, + "auxiliary_loss_mlp": 0.01029632, + "balance_loss_clip": 1.0425117, + "balance_loss_mlp": 1.02153754, + "epoch": 0.7470690795406721, + "flos": 16690849407360.0, + "grad_norm": 2.289596371378527, + "language_loss": 0.77024794, + "learning_rate": 6.342255469986053e-07, + "loss": 0.79198205, + "num_input_tokens_seen": 133551300, + "step": 6213, + "time_per_iteration": 2.7032065391540527 + }, + { + "auxiliary_loss_clip": 0.01171475, + "auxiliary_loss_mlp": 0.010312, + "balance_loss_clip": 1.04940963, + "balance_loss_mlp": 1.02371335, + "epoch": 0.7471893224313112, + "flos": 25192700352000.0, + "grad_norm": 1.940291139213515, + "language_loss": 0.76229155, + "learning_rate": 6.336565932713533e-07, + "loss": 0.78431827, + "num_input_tokens_seen": 133570725, + "step": 6214, + "time_per_iteration": 2.740368366241455 + }, + { + "auxiliary_loss_clip": 0.01148603, + "auxiliary_loss_mlp": 0.01037987, + "balance_loss_clip": 1.047405, + "balance_loss_mlp": 1.02969599, + "epoch": 0.7473095653219504, + "flos": 22526225199360.0, + "grad_norm": 3.944851153298799, + "language_loss": 0.77652973, + "learning_rate": 6.330878468173088e-07, + "loss": 0.79839563, + "num_input_tokens_seen": 133590790, + "step": 6215, + "time_per_iteration": 2.7387890815734863 + }, + { + "auxiliary_loss_clip": 0.01151184, + "auxiliary_loss_mlp": 0.01022793, + "balance_loss_clip": 1.04286385, + "balance_loss_mlp": 1.01527643, + "epoch": 0.7474298082125894, + "flos": 18113989236480.0, + "grad_norm": 1.8565068569984946, + "language_loss": 0.73061156, + "learning_rate": 6.32519307722752e-07, + "loss": 0.75235134, + "num_input_tokens_seen": 133608685, + "step": 6216, + "time_per_iteration": 2.698035955429077 + }, + { + "auxiliary_loss_clip": 0.01036072, + "auxiliary_loss_mlp": 0.01002039, + "balance_loss_clip": 1.00818133, + "balance_loss_mlp": 1.00022745, + "epoch": 0.7475500511032285, + "flos": 62086535193600.0, + "grad_norm": 0.826466882464957, + "language_loss": 0.55007923, + "learning_rate": 6.31950976073929e-07, + "loss": 0.57046032, + "num_input_tokens_seen": 133662775, + "step": 6217, + "time_per_iteration": 3.252528429031372 + }, + { + "auxiliary_loss_clip": 0.0112788, + "auxiliary_loss_mlp": 0.01032876, + "balance_loss_clip": 1.04381251, + "balance_loss_mlp": 1.02473414, + "epoch": 0.7476702939938676, + "flos": 17785586165760.0, + "grad_norm": 2.4885288457236405, + "language_loss": 0.80562842, + "learning_rate": 6.31382851957055e-07, + "loss": 0.82723594, + "num_input_tokens_seen": 133679595, + "step": 6218, + "time_per_iteration": 3.7601852416992188 + }, + { + "auxiliary_loss_clip": 0.0113652, + "auxiliary_loss_mlp": 0.00885642, + "balance_loss_clip": 1.04359722, + "balance_loss_mlp": 1.00065029, + "epoch": 0.7477905368845067, + "flos": 27927944092800.0, + "grad_norm": 2.1722107637225907, + "language_loss": 0.7193982, + "learning_rate": 6.308149354583143e-07, + "loss": 0.73961985, + "num_input_tokens_seen": 133699000, + "step": 6219, + "time_per_iteration": 3.738563060760498 + }, + { + "auxiliary_loss_clip": 0.01165257, + "auxiliary_loss_mlp": 0.01035927, + "balance_loss_clip": 1.04749417, + "balance_loss_mlp": 1.02710581, + "epoch": 0.7479107797751458, + "flos": 26870374932480.0, + "grad_norm": 1.9438508967059405, + "language_loss": 0.81624103, + "learning_rate": 6.302472266638586e-07, + "loss": 0.8382529, + "num_input_tokens_seen": 133719540, + "step": 6220, + "time_per_iteration": 3.594531297683716 + }, + { + "auxiliary_loss_clip": 0.01177401, + "auxiliary_loss_mlp": 0.0103683, + "balance_loss_clip": 1.04993105, + "balance_loss_mlp": 1.02783561, + "epoch": 0.7480310226657849, + "flos": 33943375785600.0, + "grad_norm": 2.8012453824781507, + "language_loss": 0.70164394, + "learning_rate": 6.296797256598101e-07, + "loss": 0.72378623, + "num_input_tokens_seen": 133741020, + "step": 6221, + "time_per_iteration": 2.7026407718658447 + }, + { + "auxiliary_loss_clip": 0.01131355, + "auxiliary_loss_mlp": 0.01030153, + "balance_loss_clip": 1.04121256, + "balance_loss_mlp": 1.02280331, + "epoch": 0.748151265556424, + "flos": 24826555065600.0, + "grad_norm": 1.7613041292184886, + "language_loss": 0.81109363, + "learning_rate": 6.291124325322576e-07, + "loss": 0.83270872, + "num_input_tokens_seen": 133761145, + "step": 6222, + "time_per_iteration": 2.8824875354766846 + }, + { + "auxiliary_loss_clip": 0.01156822, + "auxiliary_loss_mlp": 0.0102692, + "balance_loss_clip": 1.0459547, + "balance_loss_mlp": 1.01916504, + "epoch": 0.748271508447063, + "flos": 38399351535360.0, + "grad_norm": 1.5454323504873109, + "language_loss": 0.62350476, + "learning_rate": 6.285453473672595e-07, + "loss": 0.64534211, + "num_input_tokens_seen": 133783715, + "step": 6223, + "time_per_iteration": 2.9326069355010986 + }, + { + "auxiliary_loss_clip": 0.01168224, + "auxiliary_loss_mlp": 0.0102148, + "balance_loss_clip": 1.0470953, + "balance_loss_mlp": 1.01403522, + "epoch": 0.7483917513377022, + "flos": 21541842000000.0, + "grad_norm": 1.9385150208858748, + "language_loss": 0.75203061, + "learning_rate": 6.279784702508415e-07, + "loss": 0.77392763, + "num_input_tokens_seen": 133804465, + "step": 6224, + "time_per_iteration": 3.8112666606903076 + }, + { + "auxiliary_loss_clip": 0.01037843, + "auxiliary_loss_mlp": 0.0100387, + "balance_loss_clip": 1.00653934, + "balance_loss_mlp": 1.00241554, + "epoch": 0.7485119942283412, + "flos": 62314532772480.0, + "grad_norm": 0.7887494226684182, + "language_loss": 0.58537984, + "learning_rate": 6.274118012689979e-07, + "loss": 0.60579693, + "num_input_tokens_seen": 133866365, + "step": 6225, + "time_per_iteration": 3.438730001449585 + }, + { + "auxiliary_loss_clip": 0.01142138, + "auxiliary_loss_mlp": 0.01028393, + "balance_loss_clip": 1.04297423, + "balance_loss_mlp": 1.02024555, + "epoch": 0.7486322371189803, + "flos": 29937613104000.0, + "grad_norm": 1.5036804671469615, + "language_loss": 0.68547845, + "learning_rate": 6.268453405076943e-07, + "loss": 0.70718372, + "num_input_tokens_seen": 133888760, + "step": 6226, + "time_per_iteration": 2.814692497253418 + }, + { + "auxiliary_loss_clip": 0.01149766, + "auxiliary_loss_mlp": 0.01029182, + "balance_loss_clip": 1.04349589, + "balance_loss_mlp": 1.02175498, + "epoch": 0.7487524800096195, + "flos": 18949414734720.0, + "grad_norm": 2.3000124199365537, + "language_loss": 0.81650698, + "learning_rate": 6.262790880528592e-07, + "loss": 0.83829641, + "num_input_tokens_seen": 133906380, + "step": 6227, + "time_per_iteration": 2.758910655975342 + }, + { + "auxiliary_loss_clip": 0.01152013, + "auxiliary_loss_mlp": 0.01030994, + "balance_loss_clip": 1.04203153, + "balance_loss_mlp": 1.02269101, + "epoch": 0.7488727229002585, + "flos": 18697393935360.0, + "grad_norm": 2.351463774140287, + "language_loss": 0.79631448, + "learning_rate": 6.257130439903951e-07, + "loss": 0.81814456, + "num_input_tokens_seen": 133922875, + "step": 6228, + "time_per_iteration": 2.745296001434326 + }, + { + "auxiliary_loss_clip": 0.01171974, + "auxiliary_loss_mlp": 0.01027905, + "balance_loss_clip": 1.04910421, + "balance_loss_mlp": 1.02037966, + "epoch": 0.7489929657908976, + "flos": 23623368168960.0, + "grad_norm": 1.7250333180540538, + "language_loss": 0.81227666, + "learning_rate": 6.251472084061695e-07, + "loss": 0.83427548, + "num_input_tokens_seen": 133941795, + "step": 6229, + "time_per_iteration": 2.7049946784973145 + }, + { + "auxiliary_loss_clip": 0.01157513, + "auxiliary_loss_mlp": 0.0102819, + "balance_loss_clip": 1.04593825, + "balance_loss_mlp": 1.02068281, + "epoch": 0.7491132086815367, + "flos": 20551533056640.0, + "grad_norm": 3.678463295996263, + "language_loss": 0.88824993, + "learning_rate": 6.245815813860191e-07, + "loss": 0.91010702, + "num_input_tokens_seen": 133957305, + "step": 6230, + "time_per_iteration": 2.733779191970825 + }, + { + "auxiliary_loss_clip": 0.01173882, + "auxiliary_loss_mlp": 0.01026738, + "balance_loss_clip": 1.04843807, + "balance_loss_mlp": 1.01853001, + "epoch": 0.7492334515721758, + "flos": 23003011353600.0, + "grad_norm": 2.939542145367924, + "language_loss": 0.70296037, + "learning_rate": 6.240161630157495e-07, + "loss": 0.72496659, + "num_input_tokens_seen": 133976660, + "step": 6231, + "time_per_iteration": 2.738945722579956 + }, + { + "auxiliary_loss_clip": 0.01173309, + "auxiliary_loss_mlp": 0.01028882, + "balance_loss_clip": 1.04887128, + "balance_loss_mlp": 1.02022123, + "epoch": 0.7493536944628149, + "flos": 16398823835520.0, + "grad_norm": 2.4420659448233244, + "language_loss": 0.69618487, + "learning_rate": 6.23450953381133e-07, + "loss": 0.71820682, + "num_input_tokens_seen": 133994750, + "step": 6232, + "time_per_iteration": 2.722985029220581 + }, + { + "auxiliary_loss_clip": 0.01142086, + "auxiliary_loss_mlp": 0.0102483, + "balance_loss_clip": 1.04240096, + "balance_loss_mlp": 1.01718879, + "epoch": 0.749473937353454, + "flos": 15338561155200.0, + "grad_norm": 3.3916450106498637, + "language_loss": 0.68302286, + "learning_rate": 6.228859525679131e-07, + "loss": 0.70469201, + "num_input_tokens_seen": 134009165, + "step": 6233, + "time_per_iteration": 2.781827449798584 + }, + { + "auxiliary_loss_clip": 0.01163046, + "auxiliary_loss_mlp": 0.01028584, + "balance_loss_clip": 1.04805756, + "balance_loss_mlp": 1.02124023, + "epoch": 0.7495941802440931, + "flos": 18951138587520.0, + "grad_norm": 3.690729182714101, + "language_loss": 0.80041552, + "learning_rate": 6.223211606617986e-07, + "loss": 0.82233185, + "num_input_tokens_seen": 134027585, + "step": 6234, + "time_per_iteration": 2.780012369155884 + }, + { + "auxiliary_loss_clip": 0.01160274, + "auxiliary_loss_mlp": 0.01023957, + "balance_loss_clip": 1.04984593, + "balance_loss_mlp": 1.01732278, + "epoch": 0.7497144231347321, + "flos": 22492469393280.0, + "grad_norm": 1.7704286931177113, + "language_loss": 0.83743399, + "learning_rate": 6.217565777484701e-07, + "loss": 0.85927629, + "num_input_tokens_seen": 134046680, + "step": 6235, + "time_per_iteration": 2.7556493282318115 + }, + { + "auxiliary_loss_clip": 0.01148686, + "auxiliary_loss_mlp": 0.00885316, + "balance_loss_clip": 1.04536211, + "balance_loss_mlp": 1.00061202, + "epoch": 0.7498346660253713, + "flos": 24243509502720.0, + "grad_norm": 2.020886320497836, + "language_loss": 0.80209243, + "learning_rate": 6.211922039135722e-07, + "loss": 0.8224324, + "num_input_tokens_seen": 134066825, + "step": 6236, + "time_per_iteration": 2.7692971229553223 + }, + { + "auxiliary_loss_clip": 0.01172537, + "auxiliary_loss_mlp": 0.01027184, + "balance_loss_clip": 1.05029225, + "balance_loss_mlp": 1.01981103, + "epoch": 0.7499549089160104, + "flos": 24387080163840.0, + "grad_norm": 2.253037610223063, + "language_loss": 0.8094424, + "learning_rate": 6.206280392427201e-07, + "loss": 0.83143961, + "num_input_tokens_seen": 134086410, + "step": 6237, + "time_per_iteration": 2.6578927040100098 + }, + { + "auxiliary_loss_clip": 0.01153979, + "auxiliary_loss_mlp": 0.0102888, + "balance_loss_clip": 1.0442059, + "balance_loss_mlp": 1.02094674, + "epoch": 0.7500751518066494, + "flos": 34057320704640.0, + "grad_norm": 1.788010320705475, + "language_loss": 0.73804891, + "learning_rate": 6.200640838214983e-07, + "loss": 0.75987756, + "num_input_tokens_seen": 134109185, + "step": 6238, + "time_per_iteration": 2.8445146083831787 + }, + { + "auxiliary_loss_clip": 0.01168574, + "auxiliary_loss_mlp": 0.01022231, + "balance_loss_clip": 1.04596949, + "balance_loss_mlp": 1.01460123, + "epoch": 0.7501953946972886, + "flos": 18843586289280.0, + "grad_norm": 1.8080506062051802, + "language_loss": 0.66620111, + "learning_rate": 6.195003377354578e-07, + "loss": 0.68810916, + "num_input_tokens_seen": 134128455, + "step": 6239, + "time_per_iteration": 2.8171515464782715 + }, + { + "auxiliary_loss_clip": 0.01157761, + "auxiliary_loss_mlp": 0.01030304, + "balance_loss_clip": 1.0454762, + "balance_loss_mlp": 1.02189362, + "epoch": 0.7503156375879276, + "flos": 20257675891200.0, + "grad_norm": 2.4426714165656254, + "language_loss": 0.73495746, + "learning_rate": 6.189368010701183e-07, + "loss": 0.75683808, + "num_input_tokens_seen": 134145515, + "step": 6240, + "time_per_iteration": 2.6962850093841553 + }, + { + "auxiliary_loss_clip": 0.01165225, + "auxiliary_loss_mlp": 0.010263, + "balance_loss_clip": 1.04411674, + "balance_loss_mlp": 1.01858711, + "epoch": 0.7504358804785667, + "flos": 13480040574720.0, + "grad_norm": 1.8878464298365598, + "language_loss": 0.76311266, + "learning_rate": 6.183734739109683e-07, + "loss": 0.78502792, + "num_input_tokens_seen": 134163335, + "step": 6241, + "time_per_iteration": 2.7826926708221436 + }, + { + "auxiliary_loss_clip": 0.01161499, + "auxiliary_loss_mlp": 0.01031808, + "balance_loss_clip": 1.04913199, + "balance_loss_mlp": 1.02312934, + "epoch": 0.7505561233692057, + "flos": 29461042431360.0, + "grad_norm": 2.6260231778573626, + "language_loss": 0.68685609, + "learning_rate": 6.178103563434629e-07, + "loss": 0.70878911, + "num_input_tokens_seen": 134182335, + "step": 6242, + "time_per_iteration": 2.776747941970825 + }, + { + "auxiliary_loss_clip": 0.01170872, + "auxiliary_loss_mlp": 0.01028008, + "balance_loss_clip": 1.04750979, + "balance_loss_mlp": 1.02053916, + "epoch": 0.7506763662598449, + "flos": 20302457172480.0, + "grad_norm": 1.750229841966453, + "language_loss": 0.83979428, + "learning_rate": 6.172474484530283e-07, + "loss": 0.86178309, + "num_input_tokens_seen": 134201070, + "step": 6243, + "time_per_iteration": 2.7568411827087402 + }, + { + "auxiliary_loss_clip": 0.01140648, + "auxiliary_loss_mlp": 0.01032277, + "balance_loss_clip": 1.04122162, + "balance_loss_mlp": 1.02431393, + "epoch": 0.750796609150484, + "flos": 37230961939200.0, + "grad_norm": 2.060056481242751, + "language_loss": 0.76095808, + "learning_rate": 6.166847503250563e-07, + "loss": 0.78268737, + "num_input_tokens_seen": 134223310, + "step": 6244, + "time_per_iteration": 3.8363826274871826 + }, + { + "auxiliary_loss_clip": 0.01152581, + "auxiliary_loss_mlp": 0.01026544, + "balance_loss_clip": 1.04540324, + "balance_loss_mlp": 1.01884925, + "epoch": 0.750916852041123, + "flos": 19609417186560.0, + "grad_norm": 2.9682981953744116, + "language_loss": 0.79352033, + "learning_rate": 6.161222620449078e-07, + "loss": 0.81531155, + "num_input_tokens_seen": 134242085, + "step": 6245, + "time_per_iteration": 2.793433427810669 + }, + { + "auxiliary_loss_clip": 0.01144498, + "auxiliary_loss_mlp": 0.01031714, + "balance_loss_clip": 1.04459, + "balance_loss_mlp": 1.02389431, + "epoch": 0.7510370949317622, + "flos": 25112690807040.0, + "grad_norm": 9.049766138804367, + "language_loss": 0.80074859, + "learning_rate": 6.155599836979117e-07, + "loss": 0.82251072, + "num_input_tokens_seen": 134260770, + "step": 6246, + "time_per_iteration": 3.75304913520813 + }, + { + "auxiliary_loss_clip": 0.01134617, + "auxiliary_loss_mlp": 0.01027679, + "balance_loss_clip": 1.04208684, + "balance_loss_mlp": 1.0194627, + "epoch": 0.7511573378224012, + "flos": 19062282245760.0, + "grad_norm": 2.038018564432958, + "language_loss": 0.82159638, + "learning_rate": 6.149979153693649e-07, + "loss": 0.84321928, + "num_input_tokens_seen": 134278025, + "step": 6247, + "time_per_iteration": 2.7858023643493652 + }, + { + "auxiliary_loss_clip": 0.0115831, + "auxiliary_loss_mlp": 0.01029381, + "balance_loss_clip": 1.04630756, + "balance_loss_mlp": 1.02108359, + "epoch": 0.7512775807130403, + "flos": 19937676602880.0, + "grad_norm": 2.1749897479119227, + "language_loss": 0.7608515, + "learning_rate": 6.144360571445343e-07, + "loss": 0.78272837, + "num_input_tokens_seen": 134297170, + "step": 6248, + "time_per_iteration": 2.6648781299591064 + }, + { + "auxiliary_loss_clip": 0.01156728, + "auxiliary_loss_mlp": 0.01024396, + "balance_loss_clip": 1.046772, + "balance_loss_mlp": 1.01631951, + "epoch": 0.7513978236036795, + "flos": 20739920912640.0, + "grad_norm": 1.720325827728107, + "language_loss": 0.80228251, + "learning_rate": 6.138744091086509e-07, + "loss": 0.82409376, + "num_input_tokens_seen": 134316755, + "step": 6249, + "time_per_iteration": 2.7136077880859375 + }, + { + "auxiliary_loss_clip": 0.01145702, + "auxiliary_loss_mlp": 0.01032759, + "balance_loss_clip": 1.04631925, + "balance_loss_mlp": 1.02503419, + "epoch": 0.7515180664943185, + "flos": 27563163523200.0, + "grad_norm": 2.8383093021289003, + "language_loss": 0.72735631, + "learning_rate": 6.133129713469183e-07, + "loss": 0.74914086, + "num_input_tokens_seen": 134335960, + "step": 6250, + "time_per_iteration": 3.7655575275421143 + }, + { + "auxiliary_loss_clip": 0.0115037, + "auxiliary_loss_mlp": 0.01027096, + "balance_loss_clip": 1.04195356, + "balance_loss_mlp": 1.01904917, + "epoch": 0.7516383093849576, + "flos": 33803181002880.0, + "grad_norm": 1.7382780505830697, + "language_loss": 0.63776708, + "learning_rate": 6.127517439445053e-07, + "loss": 0.65954173, + "num_input_tokens_seen": 134356805, + "step": 6251, + "time_per_iteration": 2.9177494049072266 + }, + { + "auxiliary_loss_clip": 0.01117371, + "auxiliary_loss_mlp": 0.01031515, + "balance_loss_clip": 1.04155052, + "balance_loss_mlp": 1.02438331, + "epoch": 0.7517585522755967, + "flos": 29746172592000.0, + "grad_norm": 2.2099646569908424, + "language_loss": 0.81555033, + "learning_rate": 6.121907269865498e-07, + "loss": 0.83703923, + "num_input_tokens_seen": 134376295, + "step": 6252, + "time_per_iteration": 2.83120059967041 + }, + { + "auxiliary_loss_clip": 0.01037673, + "auxiliary_loss_mlp": 0.01001905, + "balance_loss_clip": 1.00775313, + "balance_loss_mlp": 1.0003556, + "epoch": 0.7518787951662358, + "flos": 69807974319360.0, + "grad_norm": 1.202858001511816, + "language_loss": 0.67265666, + "learning_rate": 6.116299205581577e-07, + "loss": 0.69305241, + "num_input_tokens_seen": 134431125, + "step": 6253, + "time_per_iteration": 3.278616189956665 + }, + { + "auxiliary_loss_clip": 0.01175916, + "auxiliary_loss_mlp": 0.01031156, + "balance_loss_clip": 1.04891622, + "balance_loss_mlp": 1.02179205, + "epoch": 0.7519990380568748, + "flos": 34203225749760.0, + "grad_norm": 15.519978703917996, + "language_loss": 0.68785775, + "learning_rate": 6.110693247444018e-07, + "loss": 0.70992851, + "num_input_tokens_seen": 134452960, + "step": 6254, + "time_per_iteration": 2.7181758880615234 + }, + { + "auxiliary_loss_clip": 0.01130943, + "auxiliary_loss_mlp": 0.0102084, + "balance_loss_clip": 1.04121304, + "balance_loss_mlp": 1.01324069, + "epoch": 0.752119280947514, + "flos": 21725704742400.0, + "grad_norm": 1.6998519277239708, + "language_loss": 0.82295227, + "learning_rate": 6.105089396303258e-07, + "loss": 0.84447014, + "num_input_tokens_seen": 134471350, + "step": 6255, + "time_per_iteration": 2.7446532249450684 + }, + { + "auxiliary_loss_clip": 0.0115213, + "auxiliary_loss_mlp": 0.01029243, + "balance_loss_clip": 1.04449129, + "balance_loss_mlp": 1.02107131, + "epoch": 0.7522395238381531, + "flos": 32742774668160.0, + "grad_norm": 1.947823359732076, + "language_loss": 0.75465655, + "learning_rate": 6.099487653009383e-07, + "loss": 0.7764703, + "num_input_tokens_seen": 134490695, + "step": 6256, + "time_per_iteration": 2.7181472778320312 + }, + { + "auxiliary_loss_clip": 0.0115762, + "auxiliary_loss_mlp": 0.01024648, + "balance_loss_clip": 1.04413581, + "balance_loss_mlp": 1.01762688, + "epoch": 0.7523597667287921, + "flos": 23476026579840.0, + "grad_norm": 2.0915094258403535, + "language_loss": 0.82916892, + "learning_rate": 6.093888018412192e-07, + "loss": 0.85099161, + "num_input_tokens_seen": 134506885, + "step": 6257, + "time_per_iteration": 2.5548315048217773 + }, + { + "auxiliary_loss_clip": 0.01055392, + "auxiliary_loss_mlp": 0.01001388, + "balance_loss_clip": 1.00706673, + "balance_loss_mlp": 0.99983805, + "epoch": 0.7524800096194313, + "flos": 67346730501120.0, + "grad_norm": 0.7055680698458747, + "language_loss": 0.54685599, + "learning_rate": 6.088290493361125e-07, + "loss": 0.56742376, + "num_input_tokens_seen": 134571770, + "step": 6258, + "time_per_iteration": 3.2873969078063965 + }, + { + "auxiliary_loss_clip": 0.01122944, + "auxiliary_loss_mlp": 0.01029704, + "balance_loss_clip": 1.04036736, + "balance_loss_mlp": 1.02149606, + "epoch": 0.7526002525100703, + "flos": 13006055681280.0, + "grad_norm": 2.252693397557538, + "language_loss": 0.71452343, + "learning_rate": 6.082695078705322e-07, + "loss": 0.73604995, + "num_input_tokens_seen": 134589250, + "step": 6259, + "time_per_iteration": 2.629344940185547 + }, + { + "auxiliary_loss_clip": 0.01153767, + "auxiliary_loss_mlp": 0.01029068, + "balance_loss_clip": 1.04434633, + "balance_loss_mlp": 1.0212388, + "epoch": 0.7527204954007094, + "flos": 21397229844480.0, + "grad_norm": 2.2388127412995993, + "language_loss": 0.68959171, + "learning_rate": 6.077101775293618e-07, + "loss": 0.71142006, + "num_input_tokens_seen": 134608075, + "step": 6260, + "time_per_iteration": 2.58267879486084 + }, + { + "auxiliary_loss_clip": 0.01162478, + "auxiliary_loss_mlp": 0.01028788, + "balance_loss_clip": 1.04595506, + "balance_loss_mlp": 1.02053857, + "epoch": 0.7528407382913486, + "flos": 18947188091520.0, + "grad_norm": 2.692858018775739, + "language_loss": 0.82521963, + "learning_rate": 6.071510583974504e-07, + "loss": 0.84713233, + "num_input_tokens_seen": 134623260, + "step": 6261, + "time_per_iteration": 2.6861090660095215 + }, + { + "auxiliary_loss_clip": 0.0117062, + "auxiliary_loss_mlp": 0.01035881, + "balance_loss_clip": 1.04779768, + "balance_loss_mlp": 1.02753019, + "epoch": 0.7529609811819876, + "flos": 15231798956160.0, + "grad_norm": 1.848604147423112, + "language_loss": 0.72087586, + "learning_rate": 6.065921505596161e-07, + "loss": 0.74294084, + "num_input_tokens_seen": 134641540, + "step": 6262, + "time_per_iteration": 2.703667163848877 + }, + { + "auxiliary_loss_clip": 0.0114141, + "auxiliary_loss_mlp": 0.01026405, + "balance_loss_clip": 1.04543257, + "balance_loss_mlp": 1.01861477, + "epoch": 0.7530812240726267, + "flos": 19354487385600.0, + "grad_norm": 1.5741289128202716, + "language_loss": 0.76704907, + "learning_rate": 6.060334541006445e-07, + "loss": 0.78872716, + "num_input_tokens_seen": 134660035, + "step": 6263, + "time_per_iteration": 2.745358943939209 + }, + { + "auxiliary_loss_clip": 0.01140256, + "auxiliary_loss_mlp": 0.0102537, + "balance_loss_clip": 1.03747046, + "balance_loss_mlp": 1.01725817, + "epoch": 0.7532014669632658, + "flos": 27748247328000.0, + "grad_norm": 1.479311850616888, + "language_loss": 0.69139016, + "learning_rate": 6.05474969105289e-07, + "loss": 0.71304643, + "num_input_tokens_seen": 134683025, + "step": 6264, + "time_per_iteration": 2.8461670875549316 + }, + { + "auxiliary_loss_clip": 0.01163091, + "auxiliary_loss_mlp": 0.01029502, + "balance_loss_clip": 1.04752374, + "balance_loss_mlp": 1.02134252, + "epoch": 0.7533217098539049, + "flos": 14137421333760.0, + "grad_norm": 2.1410842516617103, + "language_loss": 0.73564804, + "learning_rate": 6.049166956582725e-07, + "loss": 0.75757396, + "num_input_tokens_seen": 134701290, + "step": 6265, + "time_per_iteration": 2.7104363441467285 + }, + { + "auxiliary_loss_clip": 0.01155456, + "auxiliary_loss_mlp": 0.01028898, + "balance_loss_clip": 1.04538596, + "balance_loss_mlp": 1.02124524, + "epoch": 0.753441952744544, + "flos": 26429068437120.0, + "grad_norm": 1.936534316294834, + "language_loss": 0.87401378, + "learning_rate": 6.043586338442841e-07, + "loss": 0.89585733, + "num_input_tokens_seen": 134720345, + "step": 6266, + "time_per_iteration": 2.9737985134124756 + }, + { + "auxiliary_loss_clip": 0.01166546, + "auxiliary_loss_mlp": 0.01026463, + "balance_loss_clip": 1.0477854, + "balance_loss_mlp": 1.01917315, + "epoch": 0.7535621956351831, + "flos": 23878621192320.0, + "grad_norm": 1.4819325108822605, + "language_loss": 0.73329347, + "learning_rate": 6.038007837479815e-07, + "loss": 0.75522351, + "num_input_tokens_seen": 134741450, + "step": 6267, + "time_per_iteration": 2.744643449783325 + }, + { + "auxiliary_loss_clip": 0.01157675, + "auxiliary_loss_mlp": 0.01024554, + "balance_loss_clip": 1.04601336, + "balance_loss_mlp": 1.01744366, + "epoch": 0.7536824385258222, + "flos": 21795873960960.0, + "grad_norm": 2.5301640079808014, + "language_loss": 0.63935852, + "learning_rate": 6.032431454539897e-07, + "loss": 0.66118079, + "num_input_tokens_seen": 134760295, + "step": 6268, + "time_per_iteration": 2.686066150665283 + }, + { + "auxiliary_loss_clip": 0.01143569, + "auxiliary_loss_mlp": 0.01026596, + "balance_loss_clip": 1.04339325, + "balance_loss_mlp": 1.0187993, + "epoch": 0.7538026814164612, + "flos": 28911644933760.0, + "grad_norm": 2.148674272603314, + "language_loss": 0.8127566, + "learning_rate": 6.026857190469014e-07, + "loss": 0.83445829, + "num_input_tokens_seen": 134782050, + "step": 6269, + "time_per_iteration": 3.739572286605835 + }, + { + "auxiliary_loss_clip": 0.01152298, + "auxiliary_loss_mlp": 0.01028062, + "balance_loss_clip": 1.04497552, + "balance_loss_mlp": 1.02048326, + "epoch": 0.7539229243071004, + "flos": 21104701482240.0, + "grad_norm": 1.787140459768276, + "language_loss": 0.74348116, + "learning_rate": 6.0212850461128e-07, + "loss": 0.76528478, + "num_input_tokens_seen": 134801170, + "step": 6270, + "time_per_iteration": 2.7084624767303467 + }, + { + "auxiliary_loss_clip": 0.0115354, + "auxiliary_loss_mlp": 0.01029166, + "balance_loss_clip": 1.04348397, + "balance_loss_mlp": 1.02021909, + "epoch": 0.7540431671977395, + "flos": 15158469340800.0, + "grad_norm": 2.022929332714753, + "language_loss": 0.74970257, + "learning_rate": 6.015715022316516e-07, + "loss": 0.77152956, + "num_input_tokens_seen": 134819150, + "step": 6271, + "time_per_iteration": 2.703336715698242 + }, + { + "auxiliary_loss_clip": 0.01134873, + "auxiliary_loss_mlp": 0.01026599, + "balance_loss_clip": 1.04121995, + "balance_loss_mlp": 1.01851106, + "epoch": 0.7541634100883785, + "flos": 18770579896320.0, + "grad_norm": 3.935104300516719, + "language_loss": 0.77800977, + "learning_rate": 6.010147119925154e-07, + "loss": 0.79962444, + "num_input_tokens_seen": 134836905, + "step": 6272, + "time_per_iteration": 3.8258252143859863 + }, + { + "auxiliary_loss_clip": 0.01132411, + "auxiliary_loss_mlp": 0.01023888, + "balance_loss_clip": 1.04187238, + "balance_loss_mlp": 1.01644969, + "epoch": 0.7542836529790176, + "flos": 20594770053120.0, + "grad_norm": 2.755932626203063, + "language_loss": 0.66930419, + "learning_rate": 6.004581339783348e-07, + "loss": 0.69086719, + "num_input_tokens_seen": 134855225, + "step": 6273, + "time_per_iteration": 3.7488961219787598 + }, + { + "auxiliary_loss_clip": 0.01166879, + "auxiliary_loss_mlp": 0.01028968, + "balance_loss_clip": 1.04803205, + "balance_loss_mlp": 1.02053356, + "epoch": 0.7544038958696567, + "flos": 19095104298240.0, + "grad_norm": 2.4609176268280715, + "language_loss": 0.68668604, + "learning_rate": 5.999017682735425e-07, + "loss": 0.70864451, + "num_input_tokens_seen": 134871615, + "step": 6274, + "time_per_iteration": 2.7182986736297607 + }, + { + "auxiliary_loss_clip": 0.01127764, + "auxiliary_loss_mlp": 0.01029776, + "balance_loss_clip": 1.04016733, + "balance_loss_mlp": 1.02172959, + "epoch": 0.7545241387602958, + "flos": 31723306859520.0, + "grad_norm": 1.9807204364373683, + "language_loss": 0.66937512, + "learning_rate": 5.993456149625387e-07, + "loss": 0.69095051, + "num_input_tokens_seen": 134892765, + "step": 6275, + "time_per_iteration": 2.9463651180267334 + }, + { + "auxiliary_loss_clip": 0.01130696, + "auxiliary_loss_mlp": 0.01026121, + "balance_loss_clip": 1.04147685, + "balance_loss_mlp": 1.01843202, + "epoch": 0.7546443816509348, + "flos": 20296495514880.0, + "grad_norm": 1.6553374924194066, + "language_loss": 0.82203448, + "learning_rate": 5.987896741296909e-07, + "loss": 0.84360266, + "num_input_tokens_seen": 134910505, + "step": 6276, + "time_per_iteration": 3.8852853775024414 + }, + { + "auxiliary_loss_clip": 0.01151885, + "auxiliary_loss_mlp": 0.01026611, + "balance_loss_clip": 1.04761744, + "balance_loss_mlp": 1.01908016, + "epoch": 0.754764624541574, + "flos": 23696159080320.0, + "grad_norm": 2.002606690419233, + "language_loss": 0.78291523, + "learning_rate": 5.982339458593361e-07, + "loss": 0.80470014, + "num_input_tokens_seen": 134930445, + "step": 6277, + "time_per_iteration": 2.7718701362609863 + }, + { + "auxiliary_loss_clip": 0.01155597, + "auxiliary_loss_mlp": 0.00884994, + "balance_loss_clip": 1.0462997, + "balance_loss_mlp": 1.00059175, + "epoch": 0.7548848674322131, + "flos": 25337204766720.0, + "grad_norm": 1.5933037171047173, + "language_loss": 0.83839452, + "learning_rate": 5.976784302357767e-07, + "loss": 0.85880041, + "num_input_tokens_seen": 134951010, + "step": 6278, + "time_per_iteration": 2.861232280731201 + }, + { + "auxiliary_loss_clip": 0.01164034, + "auxiliary_loss_mlp": 0.01029635, + "balance_loss_clip": 1.04668128, + "balance_loss_mlp": 1.02136171, + "epoch": 0.7550051103228521, + "flos": 19573147428480.0, + "grad_norm": 1.9804667229464243, + "language_loss": 0.73404622, + "learning_rate": 5.971231273432855e-07, + "loss": 0.75598294, + "num_input_tokens_seen": 134970495, + "step": 6279, + "time_per_iteration": 2.6991610527038574 + }, + { + "auxiliary_loss_clip": 0.01056511, + "auxiliary_loss_mlp": 0.01001363, + "balance_loss_clip": 1.00949073, + "balance_loss_mlp": 0.99992055, + "epoch": 0.7551253532134913, + "flos": 64150068648960.0, + "grad_norm": 0.8124360123205435, + "language_loss": 0.5457716, + "learning_rate": 5.965680372661e-07, + "loss": 0.56635034, + "num_input_tokens_seen": 135028060, + "step": 6280, + "time_per_iteration": 3.288626194000244 + }, + { + "auxiliary_loss_clip": 0.01151835, + "auxiliary_loss_mlp": 0.01025449, + "balance_loss_clip": 1.04504919, + "balance_loss_mlp": 1.01852262, + "epoch": 0.7552455961041303, + "flos": 26067986968320.0, + "grad_norm": 2.0696032075082007, + "language_loss": 0.55969882, + "learning_rate": 5.960131600884266e-07, + "loss": 0.58147168, + "num_input_tokens_seen": 135047330, + "step": 6281, + "time_per_iteration": 2.766636610031128 + }, + { + "auxiliary_loss_clip": 0.01140642, + "auxiliary_loss_mlp": 0.01028898, + "balance_loss_clip": 1.04253232, + "balance_loss_mlp": 1.02178121, + "epoch": 0.7553658389947694, + "flos": 24498223822080.0, + "grad_norm": 1.5994894622353222, + "language_loss": 0.75940734, + "learning_rate": 5.954584958944413e-07, + "loss": 0.78110272, + "num_input_tokens_seen": 135065995, + "step": 6282, + "time_per_iteration": 2.795907974243164 + }, + { + "auxiliary_loss_clip": 0.01141074, + "auxiliary_loss_mlp": 0.00885394, + "balance_loss_clip": 1.04063249, + "balance_loss_mlp": 1.00063396, + "epoch": 0.7554860818854086, + "flos": 21799465320960.0, + "grad_norm": 3.926709862712693, + "language_loss": 0.8145501, + "learning_rate": 5.949040447682854e-07, + "loss": 0.83481485, + "num_input_tokens_seen": 135085820, + "step": 6283, + "time_per_iteration": 2.835355281829834 + }, + { + "auxiliary_loss_clip": 0.01155882, + "auxiliary_loss_mlp": 0.01034577, + "balance_loss_clip": 1.04420972, + "balance_loss_mlp": 1.02684057, + "epoch": 0.7556063247760476, + "flos": 16362123114240.0, + "grad_norm": 2.1660681250227, + "language_loss": 0.68264663, + "learning_rate": 5.943498067940686e-07, + "loss": 0.70455122, + "num_input_tokens_seen": 135102845, + "step": 6284, + "time_per_iteration": 2.9184446334838867 + }, + { + "auxiliary_loss_clip": 0.01145111, + "auxiliary_loss_mlp": 0.01027163, + "balance_loss_clip": 1.05025935, + "balance_loss_mlp": 1.0201354, + "epoch": 0.7557265676666867, + "flos": 27235155502080.0, + "grad_norm": 1.8303203785784636, + "language_loss": 0.81207907, + "learning_rate": 5.937957820558686e-07, + "loss": 0.83380175, + "num_input_tokens_seen": 135122190, + "step": 6285, + "time_per_iteration": 2.872591257095337 + }, + { + "auxiliary_loss_clip": 0.01042316, + "auxiliary_loss_mlp": 0.01001362, + "balance_loss_clip": 1.00689268, + "balance_loss_mlp": 0.99987775, + "epoch": 0.7558468105573258, + "flos": 62189131415040.0, + "grad_norm": 0.85601798759763, + "language_loss": 0.6540606, + "learning_rate": 5.932419706377296e-07, + "loss": 0.67449743, + "num_input_tokens_seen": 135180495, + "step": 6286, + "time_per_iteration": 3.253126382827759 + }, + { + "auxiliary_loss_clip": 0.0113806, + "auxiliary_loss_mlp": 0.01025258, + "balance_loss_clip": 1.04669023, + "balance_loss_mlp": 1.01743209, + "epoch": 0.7559670534479649, + "flos": 33249078823680.0, + "grad_norm": 2.21031721892312, + "language_loss": 0.73961884, + "learning_rate": 5.92688372623666e-07, + "loss": 0.76125205, + "num_input_tokens_seen": 135199200, + "step": 6287, + "time_per_iteration": 2.97714900970459 + }, + { + "auxiliary_loss_clip": 0.01159571, + "auxiliary_loss_mlp": 0.01025042, + "balance_loss_clip": 1.04386044, + "balance_loss_mlp": 1.01686788, + "epoch": 0.7560872963386039, + "flos": 14064379027200.0, + "grad_norm": 3.6202651206832526, + "language_loss": 0.74177742, + "learning_rate": 5.921349880976574e-07, + "loss": 0.76362348, + "num_input_tokens_seen": 135217035, + "step": 6288, + "time_per_iteration": 2.7627413272857666 + }, + { + "auxiliary_loss_clip": 0.01153856, + "auxiliary_loss_mlp": 0.00885896, + "balance_loss_clip": 1.043437, + "balance_loss_mlp": 1.0006572, + "epoch": 0.7562075392292431, + "flos": 20412307941120.0, + "grad_norm": 2.561517511492292, + "language_loss": 0.81522512, + "learning_rate": 5.915818171436515e-07, + "loss": 0.83562267, + "num_input_tokens_seen": 135236370, + "step": 6289, + "time_per_iteration": 2.826967239379883 + }, + { + "auxiliary_loss_clip": 0.01148934, + "auxiliary_loss_mlp": 0.01028903, + "balance_loss_clip": 1.04046822, + "balance_loss_mlp": 1.0206362, + "epoch": 0.7563277821198822, + "flos": 20376792368640.0, + "grad_norm": 1.7278722425729536, + "language_loss": 0.74403989, + "learning_rate": 5.910288598455642e-07, + "loss": 0.76581824, + "num_input_tokens_seen": 135255720, + "step": 6290, + "time_per_iteration": 2.7734241485595703 + }, + { + "auxiliary_loss_clip": 0.01166045, + "auxiliary_loss_mlp": 0.01033771, + "balance_loss_clip": 1.04626393, + "balance_loss_mlp": 1.02471077, + "epoch": 0.7564480250105212, + "flos": 18588261438720.0, + "grad_norm": 8.922159390567545, + "language_loss": 0.74313033, + "learning_rate": 5.90476116287278e-07, + "loss": 0.76512849, + "num_input_tokens_seen": 135273320, + "step": 6291, + "time_per_iteration": 2.6389036178588867 + }, + { + "auxiliary_loss_clip": 0.01152176, + "auxiliary_loss_mlp": 0.0103223, + "balance_loss_clip": 1.04801619, + "balance_loss_mlp": 1.02481508, + "epoch": 0.7565682679011604, + "flos": 21215521918080.0, + "grad_norm": 1.795319097762087, + "language_loss": 0.67843723, + "learning_rate": 5.899235865526456e-07, + "loss": 0.70028132, + "num_input_tokens_seen": 135292615, + "step": 6292, + "time_per_iteration": 2.8342721462249756 + }, + { + "auxiliary_loss_clip": 0.01133898, + "auxiliary_loss_mlp": 0.01024737, + "balance_loss_clip": 1.04299688, + "balance_loss_mlp": 1.01757836, + "epoch": 0.7566885107917994, + "flos": 20449008662400.0, + "grad_norm": 2.102180771902799, + "language_loss": 0.81983125, + "learning_rate": 5.893712707254825e-07, + "loss": 0.84141761, + "num_input_tokens_seen": 135310075, + "step": 6293, + "time_per_iteration": 2.7622382640838623 + }, + { + "auxiliary_loss_clip": 0.0112992, + "auxiliary_loss_mlp": 0.01023724, + "balance_loss_clip": 1.04027915, + "balance_loss_mlp": 1.0156213, + "epoch": 0.7568087536824385, + "flos": 19025832919680.0, + "grad_norm": 2.7478855856512165, + "language_loss": 0.65617537, + "learning_rate": 5.888191688895769e-07, + "loss": 0.67771184, + "num_input_tokens_seen": 135327335, + "step": 6294, + "time_per_iteration": 2.8722496032714844 + }, + { + "auxiliary_loss_clip": 0.01169439, + "auxiliary_loss_mlp": 0.01025724, + "balance_loss_clip": 1.0453887, + "balance_loss_mlp": 1.01646161, + "epoch": 0.7569289965730777, + "flos": 15225442248960.0, + "grad_norm": 2.149147822830442, + "language_loss": 0.61964953, + "learning_rate": 5.882672811286813e-07, + "loss": 0.6416012, + "num_input_tokens_seen": 135343615, + "step": 6295, + "time_per_iteration": 3.5575127601623535 + }, + { + "auxiliary_loss_clip": 0.01172169, + "auxiliary_loss_mlp": 0.01024819, + "balance_loss_clip": 1.04716134, + "balance_loss_mlp": 1.01674867, + "epoch": 0.7570492394637167, + "flos": 20769367086720.0, + "grad_norm": 2.1714749806097404, + "language_loss": 0.69934106, + "learning_rate": 5.877156075265166e-07, + "loss": 0.72131091, + "num_input_tokens_seen": 135359880, + "step": 6296, + "time_per_iteration": 2.5758883953094482 + }, + { + "auxiliary_loss_clip": 0.01146363, + "auxiliary_loss_mlp": 0.01026157, + "balance_loss_clip": 1.04140449, + "balance_loss_mlp": 1.0185219, + "epoch": 0.7571694823543558, + "flos": 15664091137920.0, + "grad_norm": 2.710388302865423, + "language_loss": 0.70384628, + "learning_rate": 5.871641481667715e-07, + "loss": 0.72557145, + "num_input_tokens_seen": 135374325, + "step": 6297, + "time_per_iteration": 2.7515993118286133 + }, + { + "auxiliary_loss_clip": 0.0113371, + "auxiliary_loss_mlp": 0.01028161, + "balance_loss_clip": 1.04191148, + "balance_loss_mlp": 1.02001929, + "epoch": 0.7572897252449949, + "flos": 25409241492480.0, + "grad_norm": 1.6371496427612307, + "language_loss": 0.84383619, + "learning_rate": 5.866129031331011e-07, + "loss": 0.86545491, + "num_input_tokens_seen": 135393980, + "step": 6298, + "time_per_iteration": 3.7437453269958496 + }, + { + "auxiliary_loss_clip": 0.01152431, + "auxiliary_loss_mlp": 0.01023295, + "balance_loss_clip": 1.0434283, + "balance_loss_mlp": 1.01517963, + "epoch": 0.757409968135634, + "flos": 24279348297600.0, + "grad_norm": 2.4609521449635725, + "language_loss": 0.83405757, + "learning_rate": 5.8606187250913e-07, + "loss": 0.85581475, + "num_input_tokens_seen": 135412030, + "step": 6299, + "time_per_iteration": 3.6663951873779297 + }, + { + "auxiliary_loss_clip": 0.01160805, + "auxiliary_loss_mlp": 0.0088548, + "balance_loss_clip": 1.04926157, + "balance_loss_mlp": 1.00073647, + "epoch": 0.757530211026273, + "flos": 24133766474880.0, + "grad_norm": 2.53658266806569, + "language_loss": 0.84047306, + "learning_rate": 5.855110563784482e-07, + "loss": 0.86093593, + "num_input_tokens_seen": 135430565, + "step": 6300, + "time_per_iteration": 2.8257529735565186 + }, + { + "auxiliary_loss_clip": 0.01153641, + "auxiliary_loss_mlp": 0.0088533, + "balance_loss_clip": 1.04320264, + "balance_loss_mlp": 1.00075483, + "epoch": 0.7576504539169122, + "flos": 23951807153280.0, + "grad_norm": 1.8045060414197664, + "language_loss": 0.64179438, + "learning_rate": 5.849604548246156e-07, + "loss": 0.66218412, + "num_input_tokens_seen": 135451675, + "step": 6301, + "time_per_iteration": 2.785991668701172 + }, + { + "auxiliary_loss_clip": 0.01156005, + "auxiliary_loss_mlp": 0.00885972, + "balance_loss_clip": 1.0459789, + "balance_loss_mlp": 1.00072896, + "epoch": 0.7577706968075513, + "flos": 21251360712960.0, + "grad_norm": 2.3635091366862575, + "language_loss": 0.79856545, + "learning_rate": 5.844100679311565e-07, + "loss": 0.81898522, + "num_input_tokens_seen": 135470635, + "step": 6302, + "time_per_iteration": 3.7951643466949463 + }, + { + "auxiliary_loss_clip": 0.01152239, + "auxiliary_loss_mlp": 0.01032894, + "balance_loss_clip": 1.04681468, + "balance_loss_mlp": 1.02432895, + "epoch": 0.7578909396981903, + "flos": 18296595002880.0, + "grad_norm": 2.304972532570155, + "language_loss": 0.76405996, + "learning_rate": 5.838598957815637e-07, + "loss": 0.78591132, + "num_input_tokens_seen": 135487865, + "step": 6303, + "time_per_iteration": 2.8032171726226807 + }, + { + "auxiliary_loss_clip": 0.01145659, + "auxiliary_loss_mlp": 0.01022736, + "balance_loss_clip": 1.04386806, + "balance_loss_mlp": 1.01502919, + "epoch": 0.7580111825888295, + "flos": 25373869574400.0, + "grad_norm": 1.5420987504130428, + "language_loss": 0.85143054, + "learning_rate": 5.833099384592996e-07, + "loss": 0.87311447, + "num_input_tokens_seen": 135508440, + "step": 6304, + "time_per_iteration": 2.8347280025482178 + }, + { + "auxiliary_loss_clip": 0.01147587, + "auxiliary_loss_mlp": 0.01028345, + "balance_loss_clip": 1.04481387, + "balance_loss_mlp": 1.02013755, + "epoch": 0.7581314254794685, + "flos": 23768662682880.0, + "grad_norm": 2.2629209342292453, + "language_loss": 0.71595669, + "learning_rate": 5.827601960477913e-07, + "loss": 0.73771608, + "num_input_tokens_seen": 135526365, + "step": 6305, + "time_per_iteration": 2.7571473121643066 + }, + { + "auxiliary_loss_clip": 0.01156255, + "auxiliary_loss_mlp": 0.01025739, + "balance_loss_clip": 1.0436213, + "balance_loss_mlp": 1.01873279, + "epoch": 0.7582516683701076, + "flos": 22054610603520.0, + "grad_norm": 2.5583462049449226, + "language_loss": 0.70062482, + "learning_rate": 5.822106686304344e-07, + "loss": 0.72244471, + "num_input_tokens_seen": 135545655, + "step": 6306, + "time_per_iteration": 2.64176344871521 + }, + { + "auxiliary_loss_clip": 0.01147807, + "auxiliary_loss_mlp": 0.01027709, + "balance_loss_clip": 1.04374552, + "balance_loss_mlp": 1.01944757, + "epoch": 0.7583719112607467, + "flos": 31649725848960.0, + "grad_norm": 1.7687251013179268, + "language_loss": 0.57629764, + "learning_rate": 5.816613562905919e-07, + "loss": 0.5980528, + "num_input_tokens_seen": 135566840, + "step": 6307, + "time_per_iteration": 2.8659322261810303 + }, + { + "auxiliary_loss_clip": 0.01135212, + "auxiliary_loss_mlp": 0.0102247, + "balance_loss_clip": 1.04600024, + "balance_loss_mlp": 1.01484048, + "epoch": 0.7584921541513858, + "flos": 33068376478080.0, + "grad_norm": 1.601539700363497, + "language_loss": 0.69850338, + "learning_rate": 5.811122591115933e-07, + "loss": 0.72008014, + "num_input_tokens_seen": 135587825, + "step": 6308, + "time_per_iteration": 2.8306691646575928 + }, + { + "auxiliary_loss_clip": 0.01140088, + "auxiliary_loss_mlp": 0.01026842, + "balance_loss_clip": 1.04758036, + "balance_loss_mlp": 1.01926661, + "epoch": 0.7586123970420249, + "flos": 23326350606720.0, + "grad_norm": 2.4502554946791424, + "language_loss": 0.71970701, + "learning_rate": 5.805633771767376e-07, + "loss": 0.74137628, + "num_input_tokens_seen": 135605220, + "step": 6309, + "time_per_iteration": 2.8112175464630127 + }, + { + "auxiliary_loss_clip": 0.01141032, + "auxiliary_loss_mlp": 0.01028712, + "balance_loss_clip": 1.04545593, + "balance_loss_mlp": 1.02112389, + "epoch": 0.7587326399326639, + "flos": 18334229477760.0, + "grad_norm": 1.7359800063494464, + "language_loss": 0.77585053, + "learning_rate": 5.800147105692888e-07, + "loss": 0.79754794, + "num_input_tokens_seen": 135624795, + "step": 6310, + "time_per_iteration": 2.872763156890869 + }, + { + "auxiliary_loss_clip": 0.01160834, + "auxiliary_loss_mlp": 0.01029888, + "balance_loss_clip": 1.04421294, + "balance_loss_mlp": 1.02192497, + "epoch": 0.7588528828233031, + "flos": 17275080119040.0, + "grad_norm": 1.7672367031625245, + "language_loss": 0.79160738, + "learning_rate": 5.794662593724795e-07, + "loss": 0.81351459, + "num_input_tokens_seen": 135643800, + "step": 6311, + "time_per_iteration": 2.8105685710906982 + }, + { + "auxiliary_loss_clip": 0.01169935, + "auxiliary_loss_mlp": 0.01027087, + "balance_loss_clip": 1.04848087, + "balance_loss_mlp": 1.01831329, + "epoch": 0.7589731257139422, + "flos": 17713621267200.0, + "grad_norm": 2.2552812356923035, + "language_loss": 0.75132066, + "learning_rate": 5.789180236695091e-07, + "loss": 0.77329087, + "num_input_tokens_seen": 135660655, + "step": 6312, + "time_per_iteration": 2.6516356468200684 + }, + { + "auxiliary_loss_clip": 0.01155305, + "auxiliary_loss_mlp": 0.01021187, + "balance_loss_clip": 1.04531717, + "balance_loss_mlp": 1.01400161, + "epoch": 0.7590933686045812, + "flos": 15961072786560.0, + "grad_norm": 2.012736075235672, + "language_loss": 0.84900057, + "learning_rate": 5.78370003543544e-07, + "loss": 0.87076545, + "num_input_tokens_seen": 135679410, + "step": 6313, + "time_per_iteration": 2.689924716949463 + }, + { + "auxiliary_loss_clip": 0.01162194, + "auxiliary_loss_mlp": 0.00885945, + "balance_loss_clip": 1.04674268, + "balance_loss_mlp": 1.00066876, + "epoch": 0.7592136114952204, + "flos": 21068072588160.0, + "grad_norm": 1.9143956791644858, + "language_loss": 0.84063613, + "learning_rate": 5.778221990777203e-07, + "loss": 0.86111748, + "num_input_tokens_seen": 135697150, + "step": 6314, + "time_per_iteration": 2.7040255069732666 + }, + { + "auxiliary_loss_clip": 0.01155158, + "auxiliary_loss_mlp": 0.01031873, + "balance_loss_clip": 1.04756594, + "balance_loss_mlp": 1.02421927, + "epoch": 0.7593338543858594, + "flos": 25297666871040.0, + "grad_norm": 2.1715777037941244, + "language_loss": 0.82827652, + "learning_rate": 5.772746103551372e-07, + "loss": 0.85014683, + "num_input_tokens_seen": 135712545, + "step": 6315, + "time_per_iteration": 2.7358644008636475 + }, + { + "auxiliary_loss_clip": 0.01148913, + "auxiliary_loss_mlp": 0.01030017, + "balance_loss_clip": 1.04688454, + "balance_loss_mlp": 1.02191699, + "epoch": 0.7594540972764985, + "flos": 31832367528960.0, + "grad_norm": 1.852494696865466, + "language_loss": 0.71768928, + "learning_rate": 5.767272374588648e-07, + "loss": 0.73947853, + "num_input_tokens_seen": 135733950, + "step": 6316, + "time_per_iteration": 2.8269155025482178 + }, + { + "auxiliary_loss_clip": 0.01159917, + "auxiliary_loss_mlp": 0.01028537, + "balance_loss_clip": 1.04810441, + "balance_loss_mlp": 1.02078867, + "epoch": 0.7595743401671377, + "flos": 37597250880000.0, + "grad_norm": 2.506661539700177, + "language_loss": 0.78175998, + "learning_rate": 5.76180080471939e-07, + "loss": 0.80364442, + "num_input_tokens_seen": 135757120, + "step": 6317, + "time_per_iteration": 2.9918832778930664 + }, + { + "auxiliary_loss_clip": 0.0117331, + "auxiliary_loss_mlp": 0.01029967, + "balance_loss_clip": 1.04755926, + "balance_loss_mlp": 1.02147937, + "epoch": 0.7596945830577767, + "flos": 18287724343680.0, + "grad_norm": 2.877406505993945, + "language_loss": 0.72081733, + "learning_rate": 5.756331394773631e-07, + "loss": 0.74285007, + "num_input_tokens_seen": 135773335, + "step": 6318, + "time_per_iteration": 2.6456785202026367 + }, + { + "auxiliary_loss_clip": 0.01120773, + "auxiliary_loss_mlp": 0.00886204, + "balance_loss_clip": 1.04005432, + "balance_loss_mlp": 1.00066626, + "epoch": 0.7598148259484158, + "flos": 22233122219520.0, + "grad_norm": 1.945936127274189, + "language_loss": 0.76094919, + "learning_rate": 5.750864145581071e-07, + "loss": 0.78101897, + "num_input_tokens_seen": 135792555, + "step": 6319, + "time_per_iteration": 2.8638508319854736 + }, + { + "auxiliary_loss_clip": 0.01169338, + "auxiliary_loss_mlp": 0.01021117, + "balance_loss_clip": 1.048908, + "balance_loss_mlp": 1.01392853, + "epoch": 0.7599350688390549, + "flos": 27161718145920.0, + "grad_norm": 2.1358552519559293, + "language_loss": 0.86081266, + "learning_rate": 5.745399057971085e-07, + "loss": 0.88271725, + "num_input_tokens_seen": 135813690, + "step": 6320, + "time_per_iteration": 2.6922905445098877 + }, + { + "auxiliary_loss_clip": 0.01163832, + "auxiliary_loss_mlp": 0.01026081, + "balance_loss_clip": 1.0458281, + "balance_loss_mlp": 1.0176115, + "epoch": 0.760055311729694, + "flos": 15560704817280.0, + "grad_norm": 1.952664469495035, + "language_loss": 0.75566924, + "learning_rate": 5.739936132772738e-07, + "loss": 0.7775684, + "num_input_tokens_seen": 135832255, + "step": 6321, + "time_per_iteration": 3.767751455307007 + }, + { + "auxiliary_loss_clip": 0.01170711, + "auxiliary_loss_mlp": 0.01026498, + "balance_loss_clip": 1.04812527, + "balance_loss_mlp": 1.01851666, + "epoch": 0.760175554620333, + "flos": 25155496840320.0, + "grad_norm": 2.351823615085245, + "language_loss": 0.74721515, + "learning_rate": 5.734475370814733e-07, + "loss": 0.76918727, + "num_input_tokens_seen": 135851935, + "step": 6322, + "time_per_iteration": 2.6378703117370605 + }, + { + "auxiliary_loss_clip": 0.01162174, + "auxiliary_loss_mlp": 0.01028816, + "balance_loss_clip": 1.04516244, + "balance_loss_mlp": 1.02134156, + "epoch": 0.7602957975109722, + "flos": 24353791234560.0, + "grad_norm": 1.5959931780157146, + "language_loss": 0.78552598, + "learning_rate": 5.729016772925483e-07, + "loss": 0.80743587, + "num_input_tokens_seen": 135873510, + "step": 6323, + "time_per_iteration": 2.7853915691375732 + }, + { + "auxiliary_loss_clip": 0.01129733, + "auxiliary_loss_mlp": 0.01022214, + "balance_loss_clip": 1.04480982, + "balance_loss_mlp": 1.01326156, + "epoch": 0.7604160404016113, + "flos": 25192664438400.0, + "grad_norm": 1.7044812491959993, + "language_loss": 0.70666575, + "learning_rate": 5.723560339933038e-07, + "loss": 0.72818524, + "num_input_tokens_seen": 135893845, + "step": 6324, + "time_per_iteration": 2.8607425689697266 + }, + { + "auxiliary_loss_clip": 0.01157641, + "auxiliary_loss_mlp": 0.00885532, + "balance_loss_clip": 1.04330087, + "balance_loss_mlp": 1.00060177, + "epoch": 0.7605362832922503, + "flos": 29861841363840.0, + "grad_norm": 1.9463907503597537, + "language_loss": 0.65801722, + "learning_rate": 5.71810607266513e-07, + "loss": 0.67844892, + "num_input_tokens_seen": 135912430, + "step": 6325, + "time_per_iteration": 4.544022560119629 + }, + { + "auxiliary_loss_clip": 0.01164133, + "auxiliary_loss_mlp": 0.01027254, + "balance_loss_clip": 1.04746914, + "balance_loss_mlp": 1.01906395, + "epoch": 0.7606565261828895, + "flos": 13917935278080.0, + "grad_norm": 2.4136894453214173, + "language_loss": 0.60321486, + "learning_rate": 5.712653971949184e-07, + "loss": 0.62512875, + "num_input_tokens_seen": 135930550, + "step": 6326, + "time_per_iteration": 2.67527174949646 + }, + { + "auxiliary_loss_clip": 0.01153861, + "auxiliary_loss_mlp": 0.0103058, + "balance_loss_clip": 1.04349613, + "balance_loss_mlp": 1.0223186, + "epoch": 0.7607767690735285, + "flos": 18551273408640.0, + "grad_norm": 2.4886707734159716, + "language_loss": 0.75511354, + "learning_rate": 5.707204038612268e-07, + "loss": 0.77695793, + "num_input_tokens_seen": 135947980, + "step": 6327, + "time_per_iteration": 2.6808881759643555 + }, + { + "auxiliary_loss_clip": 0.01159402, + "auxiliary_loss_mlp": 0.01036673, + "balance_loss_clip": 1.05121756, + "balance_loss_mlp": 1.02769065, + "epoch": 0.7608970119641676, + "flos": 20922993555840.0, + "grad_norm": 3.778058089652268, + "language_loss": 0.74161577, + "learning_rate": 5.701756273481138e-07, + "loss": 0.76357651, + "num_input_tokens_seen": 135965400, + "step": 6328, + "time_per_iteration": 3.6212236881256104 + }, + { + "auxiliary_loss_clip": 0.01146291, + "auxiliary_loss_mlp": 0.0102619, + "balance_loss_clip": 1.04408526, + "balance_loss_mlp": 1.01838183, + "epoch": 0.7610172548548068, + "flos": 23807302738560.0, + "grad_norm": 1.5826847221892446, + "language_loss": 0.73859763, + "learning_rate": 5.696310677382212e-07, + "loss": 0.76032245, + "num_input_tokens_seen": 135986795, + "step": 6329, + "time_per_iteration": 2.8345327377319336 + }, + { + "auxiliary_loss_clip": 0.0104454, + "auxiliary_loss_mlp": 0.01001088, + "balance_loss_clip": 1.01008236, + "balance_loss_mlp": 0.99949056, + "epoch": 0.7611374977454458, + "flos": 66496580426880.0, + "grad_norm": 0.918042704560164, + "language_loss": 0.61693209, + "learning_rate": 5.690867251141576e-07, + "loss": 0.63738841, + "num_input_tokens_seen": 136053450, + "step": 6330, + "time_per_iteration": 3.4258105754852295 + }, + { + "auxiliary_loss_clip": 0.01160347, + "auxiliary_loss_mlp": 0.01029996, + "balance_loss_clip": 1.04665637, + "balance_loss_mlp": 1.02156758, + "epoch": 0.7612577406360849, + "flos": 15633136592640.0, + "grad_norm": 2.4687270564484067, + "language_loss": 0.91563725, + "learning_rate": 5.685425995585013e-07, + "loss": 0.93754077, + "num_input_tokens_seen": 136071375, + "step": 6331, + "time_per_iteration": 2.813215970993042 + }, + { + "auxiliary_loss_clip": 0.01048935, + "auxiliary_loss_mlp": 0.01001196, + "balance_loss_clip": 1.00660825, + "balance_loss_mlp": 0.99963427, + "epoch": 0.761377983526724, + "flos": 60526253237760.0, + "grad_norm": 0.7795658482365034, + "language_loss": 0.59036267, + "learning_rate": 5.679986911537935e-07, + "loss": 0.61086404, + "num_input_tokens_seen": 136138905, + "step": 6332, + "time_per_iteration": 3.445390224456787 + }, + { + "auxiliary_loss_clip": 0.01117054, + "auxiliary_loss_mlp": 0.01023738, + "balance_loss_clip": 1.03913116, + "balance_loss_mlp": 1.01556015, + "epoch": 0.7614982264173631, + "flos": 35772522019200.0, + "grad_norm": 1.9566821660211438, + "language_loss": 0.67569774, + "learning_rate": 5.674549999825462e-07, + "loss": 0.69710565, + "num_input_tokens_seen": 136161720, + "step": 6333, + "time_per_iteration": 2.9442811012268066 + }, + { + "auxiliary_loss_clip": 0.01053939, + "auxiliary_loss_mlp": 0.01001131, + "balance_loss_clip": 1.00648487, + "balance_loss_mlp": 0.99962348, + "epoch": 0.7616184693080021, + "flos": 67925502345600.0, + "grad_norm": 0.9156297609259416, + "language_loss": 0.71401989, + "learning_rate": 5.669115261272363e-07, + "loss": 0.73457062, + "num_input_tokens_seen": 136222040, + "step": 6334, + "time_per_iteration": 3.309530258178711 + }, + { + "auxiliary_loss_clip": 0.01160729, + "auxiliary_loss_mlp": 0.01029708, + "balance_loss_clip": 1.04613674, + "balance_loss_mlp": 1.02129221, + "epoch": 0.7617387121986413, + "flos": 20521979141760.0, + "grad_norm": 2.3397931384309043, + "language_loss": 0.72569394, + "learning_rate": 5.663682696703081e-07, + "loss": 0.74759829, + "num_input_tokens_seen": 136240305, + "step": 6335, + "time_per_iteration": 2.7470028400421143 + }, + { + "auxiliary_loss_clip": 0.0116778, + "auxiliary_loss_mlp": 0.01024288, + "balance_loss_clip": 1.04689586, + "balance_loss_mlp": 1.01676595, + "epoch": 0.7618589550892804, + "flos": 18624495283200.0, + "grad_norm": 1.8186719873102346, + "language_loss": 0.82273388, + "learning_rate": 5.658252306941746e-07, + "loss": 0.84465462, + "num_input_tokens_seen": 136259625, + "step": 6336, + "time_per_iteration": 2.6862690448760986 + }, + { + "auxiliary_loss_clip": 0.01133251, + "auxiliary_loss_mlp": 0.01031615, + "balance_loss_clip": 1.04318845, + "balance_loss_mlp": 1.02299595, + "epoch": 0.7619791979799194, + "flos": 17453735389440.0, + "grad_norm": 2.030622574822976, + "language_loss": 0.75327599, + "learning_rate": 5.65282409281212e-07, + "loss": 0.77492464, + "num_input_tokens_seen": 136277090, + "step": 6337, + "time_per_iteration": 2.779212474822998 + }, + { + "auxiliary_loss_clip": 0.01148678, + "auxiliary_loss_mlp": 0.01026581, + "balance_loss_clip": 1.04515743, + "balance_loss_mlp": 1.01774192, + "epoch": 0.7620994408705585, + "flos": 14137421333760.0, + "grad_norm": 2.444700553374446, + "language_loss": 0.69989121, + "learning_rate": 5.64739805513768e-07, + "loss": 0.72164381, + "num_input_tokens_seen": 136294635, + "step": 6338, + "time_per_iteration": 2.7393062114715576 + }, + { + "auxiliary_loss_clip": 0.01050852, + "auxiliary_loss_mlp": 0.00877016, + "balance_loss_clip": 1.0075767, + "balance_loss_mlp": 1.00139081, + "epoch": 0.7622196837611976, + "flos": 70708792527360.0, + "grad_norm": 0.7867996511524488, + "language_loss": 0.55741656, + "learning_rate": 5.641974194741541e-07, + "loss": 0.57669526, + "num_input_tokens_seen": 136350320, + "step": 6339, + "time_per_iteration": 3.2064402103424072 + }, + { + "auxiliary_loss_clip": 0.01040858, + "auxiliary_loss_mlp": 0.01003363, + "balance_loss_clip": 1.01054811, + "balance_loss_mlp": 1.00159919, + "epoch": 0.7623399266518367, + "flos": 60684150447360.0, + "grad_norm": 0.7881580135684653, + "language_loss": 0.63676322, + "learning_rate": 5.636552512446502e-07, + "loss": 0.65720546, + "num_input_tokens_seen": 136411375, + "step": 6340, + "time_per_iteration": 3.222289800643921 + }, + { + "auxiliary_loss_clip": 0.01159065, + "auxiliary_loss_mlp": 0.01028291, + "balance_loss_clip": 1.04618073, + "balance_loss_mlp": 1.02029753, + "epoch": 0.7624601695424758, + "flos": 26468893641600.0, + "grad_norm": 1.7396559739314672, + "language_loss": 0.78213215, + "learning_rate": 5.631133009075027e-07, + "loss": 0.80400574, + "num_input_tokens_seen": 136430560, + "step": 6341, + "time_per_iteration": 2.8233752250671387 + }, + { + "auxiliary_loss_clip": 0.01161502, + "auxiliary_loss_mlp": 0.00885314, + "balance_loss_clip": 1.04673839, + "balance_loss_mlp": 1.00069976, + "epoch": 0.7625804124331149, + "flos": 19135755515520.0, + "grad_norm": 1.8198062638956958, + "language_loss": 0.68252176, + "learning_rate": 5.625715685449242e-07, + "loss": 0.70298994, + "num_input_tokens_seen": 136448665, + "step": 6342, + "time_per_iteration": 2.721400022506714 + }, + { + "auxiliary_loss_clip": 0.0113146, + "auxiliary_loss_mlp": 0.01023608, + "balance_loss_clip": 1.04774857, + "balance_loss_mlp": 1.0157733, + "epoch": 0.762700655323754, + "flos": 26213101914240.0, + "grad_norm": 1.6226407945885488, + "language_loss": 0.71605432, + "learning_rate": 5.620300542390966e-07, + "loss": 0.73760504, + "num_input_tokens_seen": 136469710, + "step": 6343, + "time_per_iteration": 2.8574485778808594 + }, + { + "auxiliary_loss_clip": 0.01137574, + "auxiliary_loss_mlp": 0.01026504, + "balance_loss_clip": 1.04159093, + "balance_loss_mlp": 1.01918507, + "epoch": 0.762820898214393, + "flos": 22382582711040.0, + "grad_norm": 1.8509698991363284, + "language_loss": 0.85161209, + "learning_rate": 5.614887580721659e-07, + "loss": 0.87325287, + "num_input_tokens_seen": 136489855, + "step": 6344, + "time_per_iteration": 2.762927293777466 + }, + { + "auxiliary_loss_clip": 0.01139343, + "auxiliary_loss_mlp": 0.01031013, + "balance_loss_clip": 1.0472486, + "balance_loss_mlp": 1.02288246, + "epoch": 0.7629411411050322, + "flos": 15700504550400.0, + "grad_norm": 2.9013583976895148, + "language_loss": 0.74094534, + "learning_rate": 5.609476801262481e-07, + "loss": 0.76264894, + "num_input_tokens_seen": 136504715, + "step": 6345, + "time_per_iteration": 2.7202343940734863 + }, + { + "auxiliary_loss_clip": 0.01137242, + "auxiliary_loss_mlp": 0.01025198, + "balance_loss_clip": 1.0430963, + "balance_loss_mlp": 1.01752687, + "epoch": 0.7630613839956712, + "flos": 13770342293760.0, + "grad_norm": 2.68447365421414, + "language_loss": 0.64260113, + "learning_rate": 5.604068204834223e-07, + "loss": 0.66422558, + "num_input_tokens_seen": 136521610, + "step": 6346, + "time_per_iteration": 2.802245616912842 + }, + { + "auxiliary_loss_clip": 0.01128607, + "auxiliary_loss_mlp": 0.00885504, + "balance_loss_clip": 1.0418222, + "balance_loss_mlp": 1.00056028, + "epoch": 0.7631816268863103, + "flos": 14569569861120.0, + "grad_norm": 2.996752122314375, + "language_loss": 0.76164532, + "learning_rate": 5.598661792257367e-07, + "loss": 0.78178644, + "num_input_tokens_seen": 136538655, + "step": 6347, + "time_per_iteration": 2.831312894821167 + }, + { + "auxiliary_loss_clip": 0.01156713, + "auxiliary_loss_mlp": 0.01025348, + "balance_loss_clip": 1.04352903, + "balance_loss_mlp": 1.0180074, + "epoch": 0.7633018697769495, + "flos": 19062210418560.0, + "grad_norm": 2.2563716802830727, + "language_loss": 0.75652349, + "learning_rate": 5.593257564352071e-07, + "loss": 0.77834409, + "num_input_tokens_seen": 136557095, + "step": 6348, + "time_per_iteration": 3.697831153869629 + }, + { + "auxiliary_loss_clip": 0.01156075, + "auxiliary_loss_mlp": 0.01028244, + "balance_loss_clip": 1.0448395, + "balance_loss_mlp": 1.02054024, + "epoch": 0.7634221126675885, + "flos": 22052958577920.0, + "grad_norm": 2.0933639089425107, + "language_loss": 0.75692087, + "learning_rate": 5.58785552193815e-07, + "loss": 0.77876407, + "num_input_tokens_seen": 136577340, + "step": 6349, + "time_per_iteration": 2.7284934520721436 + }, + { + "auxiliary_loss_clip": 0.01168389, + "auxiliary_loss_mlp": 0.01022864, + "balance_loss_clip": 1.04656291, + "balance_loss_mlp": 1.01537192, + "epoch": 0.7635423555582276, + "flos": 29382720825600.0, + "grad_norm": 1.8568865624747666, + "language_loss": 0.75851601, + "learning_rate": 5.582455665835086e-07, + "loss": 0.78042859, + "num_input_tokens_seen": 136597635, + "step": 6350, + "time_per_iteration": 2.7235991954803467 + }, + { + "auxiliary_loss_clip": 0.01158886, + "auxiliary_loss_mlp": 0.01033935, + "balance_loss_clip": 1.04310715, + "balance_loss_mlp": 1.02458286, + "epoch": 0.7636625984488667, + "flos": 17784903807360.0, + "grad_norm": 3.1131383342631462, + "language_loss": 0.72550684, + "learning_rate": 5.577057996862036e-07, + "loss": 0.74743503, + "num_input_tokens_seen": 136615260, + "step": 6351, + "time_per_iteration": 4.511592864990234 + }, + { + "auxiliary_loss_clip": 0.0116424, + "auxiliary_loss_mlp": 0.0102335, + "balance_loss_clip": 1.04580927, + "balance_loss_mlp": 1.01599169, + "epoch": 0.7637828413395058, + "flos": 23734583654400.0, + "grad_norm": 1.6509009675345403, + "language_loss": 0.75904948, + "learning_rate": 5.571662515837814e-07, + "loss": 0.78092533, + "num_input_tokens_seen": 136637220, + "step": 6352, + "time_per_iteration": 2.7128639221191406 + }, + { + "auxiliary_loss_clip": 0.01152856, + "auxiliary_loss_mlp": 0.01025302, + "balance_loss_clip": 1.04547775, + "balance_loss_mlp": 1.01757169, + "epoch": 0.7639030842301449, + "flos": 36283279461120.0, + "grad_norm": 1.7832104475574377, + "language_loss": 0.8374238, + "learning_rate": 5.566269223580926e-07, + "loss": 0.85920542, + "num_input_tokens_seen": 136658930, + "step": 6353, + "time_per_iteration": 2.979222059249878 + }, + { + "auxiliary_loss_clip": 0.01160484, + "auxiliary_loss_mlp": 0.01027433, + "balance_loss_clip": 1.044402, + "balance_loss_mlp": 1.02011919, + "epoch": 0.764023327120784, + "flos": 28878104609280.0, + "grad_norm": 1.73053320314229, + "language_loss": 0.75402248, + "learning_rate": 5.560878120909511e-07, + "loss": 0.77590168, + "num_input_tokens_seen": 136681530, + "step": 6354, + "time_per_iteration": 3.776824712753296 + }, + { + "auxiliary_loss_clip": 0.01054579, + "auxiliary_loss_mlp": 0.01001921, + "balance_loss_clip": 1.00676572, + "balance_loss_mlp": 1.0004667, + "epoch": 0.7641435700114231, + "flos": 64789711067520.0, + "grad_norm": 0.8452991102582298, + "language_loss": 0.58582127, + "learning_rate": 5.55548920864141e-07, + "loss": 0.6063863, + "num_input_tokens_seen": 136742185, + "step": 6355, + "time_per_iteration": 3.272632122039795 + }, + { + "auxiliary_loss_clip": 0.01159154, + "auxiliary_loss_mlp": 0.01029095, + "balance_loss_clip": 1.04813588, + "balance_loss_mlp": 1.02087569, + "epoch": 0.7642638129020621, + "flos": 16835784785280.0, + "grad_norm": 1.6965157245954712, + "language_loss": 0.77957994, + "learning_rate": 5.550102487594113e-07, + "loss": 0.80146241, + "num_input_tokens_seen": 136760855, + "step": 6356, + "time_per_iteration": 2.6748178005218506 + }, + { + "auxiliary_loss_clip": 0.01135377, + "auxiliary_loss_mlp": 0.00885802, + "balance_loss_clip": 1.03996956, + "balance_loss_mlp": 1.00063491, + "epoch": 0.7643840557927013, + "flos": 30408940391040.0, + "grad_norm": 1.4716368695812676, + "language_loss": 0.71289957, + "learning_rate": 5.54471795858477e-07, + "loss": 0.73311138, + "num_input_tokens_seen": 136780925, + "step": 6357, + "time_per_iteration": 2.9443516731262207 + }, + { + "auxiliary_loss_clip": 0.01139187, + "auxiliary_loss_mlp": 0.01027761, + "balance_loss_clip": 1.03949213, + "balance_loss_mlp": 1.01947629, + "epoch": 0.7645042986833404, + "flos": 16983234115200.0, + "grad_norm": 2.264994348086826, + "language_loss": 0.82661057, + "learning_rate": 5.539335622430235e-07, + "loss": 0.84828007, + "num_input_tokens_seen": 136799545, + "step": 6358, + "time_per_iteration": 2.8918070793151855 + }, + { + "auxiliary_loss_clip": 0.01153042, + "auxiliary_loss_mlp": 0.01028407, + "balance_loss_clip": 1.04314661, + "balance_loss_mlp": 1.0200206, + "epoch": 0.7646245415739794, + "flos": 17311493531520.0, + "grad_norm": 2.150617662666669, + "language_loss": 0.74877048, + "learning_rate": 5.533955479946975e-07, + "loss": 0.770585, + "num_input_tokens_seen": 136818325, + "step": 6359, + "time_per_iteration": 2.765894651412964 + }, + { + "auxiliary_loss_clip": 0.01026917, + "auxiliary_loss_mlp": 0.00877159, + "balance_loss_clip": 1.00961971, + "balance_loss_mlp": 1.00128579, + "epoch": 0.7647447844646186, + "flos": 70402332666240.0, + "grad_norm": 0.8674309111709727, + "language_loss": 0.65800631, + "learning_rate": 5.528577531951173e-07, + "loss": 0.67704701, + "num_input_tokens_seen": 136878730, + "step": 6360, + "time_per_iteration": 3.3250174522399902 + }, + { + "auxiliary_loss_clip": 0.01156714, + "auxiliary_loss_mlp": 0.01034091, + "balance_loss_clip": 1.04561257, + "balance_loss_mlp": 1.02689075, + "epoch": 0.7648650273552576, + "flos": 17675914965120.0, + "grad_norm": 1.933754458022507, + "language_loss": 0.735623, + "learning_rate": 5.523201779258653e-07, + "loss": 0.75753111, + "num_input_tokens_seen": 136897705, + "step": 6361, + "time_per_iteration": 2.819336414337158 + }, + { + "auxiliary_loss_clip": 0.01169278, + "auxiliary_loss_mlp": 0.01036364, + "balance_loss_clip": 1.04629123, + "balance_loss_mlp": 1.02834725, + "epoch": 0.7649852702458967, + "flos": 22162019247360.0, + "grad_norm": 1.7387233746624893, + "language_loss": 0.84108204, + "learning_rate": 5.517828222684912e-07, + "loss": 0.86313844, + "num_input_tokens_seen": 136918360, + "step": 6362, + "time_per_iteration": 2.9412953853607178 + }, + { + "auxiliary_loss_clip": 0.0104673, + "auxiliary_loss_mlp": 0.01002136, + "balance_loss_clip": 1.00850606, + "balance_loss_mlp": 1.00062811, + "epoch": 0.7651055131365359, + "flos": 69848338227840.0, + "grad_norm": 0.7757188995796765, + "language_loss": 0.59109527, + "learning_rate": 5.512456863045117e-07, + "loss": 0.61158395, + "num_input_tokens_seen": 136979050, + "step": 6363, + "time_per_iteration": 3.3832178115844727 + }, + { + "auxiliary_loss_clip": 0.01170664, + "auxiliary_loss_mlp": 0.01027165, + "balance_loss_clip": 1.04659271, + "balance_loss_mlp": 1.01904702, + "epoch": 0.7652257560271749, + "flos": 19464014931840.0, + "grad_norm": 2.0706475131901687, + "language_loss": 0.74151194, + "learning_rate": 5.507087701154089e-07, + "loss": 0.76349032, + "num_input_tokens_seen": 136998970, + "step": 6364, + "time_per_iteration": 2.792962074279785 + }, + { + "auxiliary_loss_clip": 0.01134223, + "auxiliary_loss_mlp": 0.01023344, + "balance_loss_clip": 1.04227996, + "balance_loss_mlp": 1.01607537, + "epoch": 0.765345998917814, + "flos": 15961108700160.0, + "grad_norm": 2.0123997682997214, + "language_loss": 0.75582254, + "learning_rate": 5.50172073782634e-07, + "loss": 0.77739823, + "num_input_tokens_seen": 137016950, + "step": 6365, + "time_per_iteration": 3.074793577194214 + }, + { + "auxiliary_loss_clip": 0.01141389, + "auxiliary_loss_mlp": 0.01028663, + "balance_loss_clip": 1.04594862, + "balance_loss_mlp": 1.02171886, + "epoch": 0.7654662418084531, + "flos": 23659853408640.0, + "grad_norm": 1.7051414436588554, + "language_loss": 0.87746108, + "learning_rate": 5.496355973876023e-07, + "loss": 0.89916158, + "num_input_tokens_seen": 137036205, + "step": 6366, + "time_per_iteration": 2.950012683868408 + }, + { + "auxiliary_loss_clip": 0.01137715, + "auxiliary_loss_mlp": 0.00886667, + "balance_loss_clip": 1.04225349, + "balance_loss_mlp": 1.00057983, + "epoch": 0.7655864846990922, + "flos": 41463608878080.0, + "grad_norm": 1.6671649906079762, + "language_loss": 0.71010876, + "learning_rate": 5.490993410116984e-07, + "loss": 0.73035252, + "num_input_tokens_seen": 137059195, + "step": 6367, + "time_per_iteration": 2.933966875076294 + }, + { + "auxiliary_loss_clip": 0.01134661, + "auxiliary_loss_mlp": 0.01023008, + "balance_loss_clip": 1.0430243, + "balance_loss_mlp": 1.01538515, + "epoch": 0.7657067275897312, + "flos": 43142684088960.0, + "grad_norm": 1.5793420248706955, + "language_loss": 0.69244075, + "learning_rate": 5.485633047362704e-07, + "loss": 0.71401739, + "num_input_tokens_seen": 137081200, + "step": 6368, + "time_per_iteration": 2.9620556831359863 + }, + { + "auxiliary_loss_clip": 0.011751, + "auxiliary_loss_mlp": 0.01032085, + "balance_loss_clip": 1.05015099, + "balance_loss_mlp": 1.02369261, + "epoch": 0.7658269704803703, + "flos": 17311780840320.0, + "grad_norm": 1.7536531433813909, + "language_loss": 0.78923345, + "learning_rate": 5.480274886426341e-07, + "loss": 0.81130528, + "num_input_tokens_seen": 137097840, + "step": 6369, + "time_per_iteration": 2.700002908706665 + }, + { + "auxiliary_loss_clip": 0.01153901, + "auxiliary_loss_mlp": 0.01024572, + "balance_loss_clip": 1.04712963, + "balance_loss_mlp": 1.01716352, + "epoch": 0.7659472133710095, + "flos": 12568160977920.0, + "grad_norm": 3.869631015855314, + "language_loss": 0.77948123, + "learning_rate": 5.474918928120744e-07, + "loss": 0.80126595, + "num_input_tokens_seen": 137114335, + "step": 6370, + "time_per_iteration": 2.6353931427001953 + }, + { + "auxiliary_loss_clip": 0.01157299, + "auxiliary_loss_mlp": 0.0102609, + "balance_loss_clip": 1.04531288, + "balance_loss_mlp": 1.01860332, + "epoch": 0.7660674562616485, + "flos": 22707430335360.0, + "grad_norm": 1.8505388667536051, + "language_loss": 0.87532508, + "learning_rate": 5.469565173258392e-07, + "loss": 0.89715898, + "num_input_tokens_seen": 137132850, + "step": 6371, + "time_per_iteration": 2.7058796882629395 + }, + { + "auxiliary_loss_clip": 0.01172904, + "auxiliary_loss_mlp": 0.01028376, + "balance_loss_clip": 1.04736185, + "balance_loss_mlp": 1.01941729, + "epoch": 0.7661876991522876, + "flos": 17056455989760.0, + "grad_norm": 3.702643499179303, + "language_loss": 0.63629949, + "learning_rate": 5.464213622651454e-07, + "loss": 0.65831232, + "num_input_tokens_seen": 137150665, + "step": 6372, + "time_per_iteration": 2.6403727531433105 + }, + { + "auxiliary_loss_clip": 0.01147251, + "auxiliary_loss_mlp": 0.01025791, + "balance_loss_clip": 1.04408908, + "balance_loss_mlp": 1.01753569, + "epoch": 0.7663079420429267, + "flos": 20084228092800.0, + "grad_norm": 1.6119123713372236, + "language_loss": 0.84069121, + "learning_rate": 5.458864277111753e-07, + "loss": 0.86242163, + "num_input_tokens_seen": 137168500, + "step": 6373, + "time_per_iteration": 3.769151210784912 + }, + { + "auxiliary_loss_clip": 0.0114337, + "auxiliary_loss_mlp": 0.00885507, + "balance_loss_clip": 1.04173398, + "balance_loss_mlp": 1.00057244, + "epoch": 0.7664281849335658, + "flos": 12677473042560.0, + "grad_norm": 2.448512630120872, + "language_loss": 0.69077307, + "learning_rate": 5.453517137450769e-07, + "loss": 0.71106184, + "num_input_tokens_seen": 137185075, + "step": 6374, + "time_per_iteration": 2.784628391265869 + }, + { + "auxiliary_loss_clip": 0.01163042, + "auxiliary_loss_mlp": 0.01023593, + "balance_loss_clip": 1.04869294, + "balance_loss_mlp": 1.01578498, + "epoch": 0.7665484278242048, + "flos": 22345271458560.0, + "grad_norm": 1.950990766883273, + "language_loss": 0.75953442, + "learning_rate": 5.448172204479684e-07, + "loss": 0.7814008, + "num_input_tokens_seen": 137204355, + "step": 6375, + "time_per_iteration": 2.791478157043457 + }, + { + "auxiliary_loss_clip": 0.01169324, + "auxiliary_loss_mlp": 0.01025744, + "balance_loss_clip": 1.04625773, + "balance_loss_mlp": 1.01766777, + "epoch": 0.766668670714844, + "flos": 23617909301760.0, + "grad_norm": 2.8002794977124923, + "language_loss": 0.74956524, + "learning_rate": 5.442829479009294e-07, + "loss": 0.77151591, + "num_input_tokens_seen": 137223135, + "step": 6376, + "time_per_iteration": 3.881225109100342 + }, + { + "auxiliary_loss_clip": 0.01159143, + "auxiliary_loss_mlp": 0.01026663, + "balance_loss_clip": 1.04593444, + "balance_loss_mlp": 1.01837778, + "epoch": 0.7667889136054831, + "flos": 19427134642560.0, + "grad_norm": 1.6950781814132871, + "language_loss": 0.71847951, + "learning_rate": 5.437488961850103e-07, + "loss": 0.74033761, + "num_input_tokens_seen": 137242935, + "step": 6377, + "time_per_iteration": 3.670452356338501 + }, + { + "auxiliary_loss_clip": 0.01132462, + "auxiliary_loss_mlp": 0.01033508, + "balance_loss_clip": 1.04319048, + "balance_loss_mlp": 1.02621293, + "epoch": 0.7669091564961221, + "flos": 26866352609280.0, + "grad_norm": 1.8892995040608478, + "language_loss": 0.75467235, + "learning_rate": 5.432150653812258e-07, + "loss": 0.77633202, + "num_input_tokens_seen": 137262970, + "step": 6378, + "time_per_iteration": 2.8141064643859863 + }, + { + "auxiliary_loss_clip": 0.01159024, + "auxiliary_loss_mlp": 0.01022305, + "balance_loss_clip": 1.04936445, + "balance_loss_mlp": 1.01435328, + "epoch": 0.7670293993867613, + "flos": 12385303816320.0, + "grad_norm": 3.439708726280419, + "language_loss": 0.8224113, + "learning_rate": 5.42681455570557e-07, + "loss": 0.84422457, + "num_input_tokens_seen": 137279500, + "step": 6379, + "time_per_iteration": 2.727769136428833 + }, + { + "auxiliary_loss_clip": 0.01167128, + "auxiliary_loss_mlp": 0.01029678, + "balance_loss_clip": 1.04651856, + "balance_loss_mlp": 1.0221436, + "epoch": 0.7671496422774003, + "flos": 21762944167680.0, + "grad_norm": 1.7640997502535511, + "language_loss": 0.64949286, + "learning_rate": 5.42148066833954e-07, + "loss": 0.67146087, + "num_input_tokens_seen": 137298745, + "step": 6380, + "time_per_iteration": 3.645250082015991 + }, + { + "auxiliary_loss_clip": 0.01167716, + "auxiliary_loss_mlp": 0.01030503, + "balance_loss_clip": 1.04703879, + "balance_loss_mlp": 1.02302599, + "epoch": 0.7672698851680394, + "flos": 21069221823360.0, + "grad_norm": 3.3476378611836606, + "language_loss": 0.75275129, + "learning_rate": 5.416148992523289e-07, + "loss": 0.77473342, + "num_input_tokens_seen": 137317320, + "step": 6381, + "time_per_iteration": 2.722773790359497 + }, + { + "auxiliary_loss_clip": 0.01109829, + "auxiliary_loss_mlp": 0.01031578, + "balance_loss_clip": 1.03828716, + "balance_loss_mlp": 1.02354884, + "epoch": 0.7673901280586786, + "flos": 16976697840000.0, + "grad_norm": 1.9713671856558914, + "language_loss": 0.78570712, + "learning_rate": 5.410819529065644e-07, + "loss": 0.80712116, + "num_input_tokens_seen": 137335275, + "step": 6382, + "time_per_iteration": 2.9492430686950684 + }, + { + "auxiliary_loss_clip": 0.01132326, + "auxiliary_loss_mlp": 0.01022887, + "balance_loss_clip": 1.04148686, + "balance_loss_mlp": 1.01515341, + "epoch": 0.7675103709493176, + "flos": 29242669697280.0, + "grad_norm": 2.7144986467678764, + "language_loss": 0.65635777, + "learning_rate": 5.405492278775079e-07, + "loss": 0.67790985, + "num_input_tokens_seen": 137355055, + "step": 6383, + "time_per_iteration": 2.8270814418792725 + }, + { + "auxiliary_loss_clip": 0.01150014, + "auxiliary_loss_mlp": 0.01036186, + "balance_loss_clip": 1.04196489, + "balance_loss_mlp": 1.0275259, + "epoch": 0.7676306138399567, + "flos": 29023004073600.0, + "grad_norm": 5.590538881095942, + "language_loss": 0.79985452, + "learning_rate": 5.400167242459732e-07, + "loss": 0.82171649, + "num_input_tokens_seen": 137374015, + "step": 6384, + "time_per_iteration": 2.8380870819091797 + }, + { + "auxiliary_loss_clip": 0.011569, + "auxiliary_loss_mlp": 0.01025786, + "balance_loss_clip": 1.04465187, + "balance_loss_mlp": 1.01795387, + "epoch": 0.7677508567305958, + "flos": 22565116650240.0, + "grad_norm": 1.7116225159727059, + "language_loss": 0.80752194, + "learning_rate": 5.394844420927405e-07, + "loss": 0.8293488, + "num_input_tokens_seen": 137393625, + "step": 6385, + "time_per_iteration": 2.7430615425109863 + }, + { + "auxiliary_loss_clip": 0.01168665, + "auxiliary_loss_mlp": 0.01025909, + "balance_loss_clip": 1.04760194, + "balance_loss_mlp": 1.0176959, + "epoch": 0.7678710996212349, + "flos": 25411432222080.0, + "grad_norm": 2.085593294629148, + "language_loss": 0.73687083, + "learning_rate": 5.389523814985562e-07, + "loss": 0.7588166, + "num_input_tokens_seen": 137413045, + "step": 6386, + "time_per_iteration": 2.7544257640838623 + }, + { + "auxiliary_loss_clip": 0.0112956, + "auxiliary_loss_mlp": 0.01032174, + "balance_loss_clip": 1.0406363, + "balance_loss_mlp": 1.02325106, + "epoch": 0.767991342511874, + "flos": 26756825063040.0, + "grad_norm": 2.0396169593607127, + "language_loss": 0.76152778, + "learning_rate": 5.384205425441344e-07, + "loss": 0.78314507, + "num_input_tokens_seen": 137433955, + "step": 6387, + "time_per_iteration": 2.89713191986084 + }, + { + "auxiliary_loss_clip": 0.01154706, + "auxiliary_loss_mlp": 0.01023001, + "balance_loss_clip": 1.0455699, + "balance_loss_mlp": 1.01593792, + "epoch": 0.7681115854025131, + "flos": 26359509749760.0, + "grad_norm": 1.9829398679600059, + "language_loss": 0.84124953, + "learning_rate": 5.378889253101537e-07, + "loss": 0.86302662, + "num_input_tokens_seen": 137454510, + "step": 6388, + "time_per_iteration": 2.781179666519165 + }, + { + "auxiliary_loss_clip": 0.01157475, + "auxiliary_loss_mlp": 0.01033544, + "balance_loss_clip": 1.0419147, + "balance_loss_mlp": 1.02618289, + "epoch": 0.7682318282931522, + "flos": 23257043314560.0, + "grad_norm": 1.6659735492456045, + "language_loss": 0.80849922, + "learning_rate": 5.373575298772617e-07, + "loss": 0.83040941, + "num_input_tokens_seen": 137473630, + "step": 6389, + "time_per_iteration": 2.707995653152466 + }, + { + "auxiliary_loss_clip": 0.01054006, + "auxiliary_loss_mlp": 0.01002154, + "balance_loss_clip": 1.00623751, + "balance_loss_mlp": 1.00054514, + "epoch": 0.7683520711837912, + "flos": 70072457137920.0, + "grad_norm": 0.7716667260457836, + "language_loss": 0.61371505, + "learning_rate": 5.368263563260689e-07, + "loss": 0.63427675, + "num_input_tokens_seen": 137538765, + "step": 6390, + "time_per_iteration": 3.3282623291015625 + }, + { + "auxiliary_loss_clip": 0.01158585, + "auxiliary_loss_mlp": 0.0102707, + "balance_loss_clip": 1.04354405, + "balance_loss_mlp": 1.01930928, + "epoch": 0.7684723140744304, + "flos": 18624890332800.0, + "grad_norm": 1.5676514259395757, + "language_loss": 0.64359248, + "learning_rate": 5.362954047371537e-07, + "loss": 0.66544902, + "num_input_tokens_seen": 137557875, + "step": 6391, + "time_per_iteration": 2.6352035999298096 + }, + { + "auxiliary_loss_clip": 0.01141327, + "auxiliary_loss_mlp": 0.01027129, + "balance_loss_clip": 1.04730761, + "balance_loss_mlp": 1.01905227, + "epoch": 0.7685925569650695, + "flos": 27452989532160.0, + "grad_norm": 1.937900765270325, + "language_loss": 0.71978897, + "learning_rate": 5.357646751910627e-07, + "loss": 0.74147356, + "num_input_tokens_seen": 137579055, + "step": 6392, + "time_per_iteration": 2.8828654289245605 + }, + { + "auxiliary_loss_clip": 0.01149277, + "auxiliary_loss_mlp": 0.01028331, + "balance_loss_clip": 1.04483485, + "balance_loss_mlp": 1.01995635, + "epoch": 0.7687127998557085, + "flos": 24535714642560.0, + "grad_norm": 2.015733734111528, + "language_loss": 0.79967213, + "learning_rate": 5.352341677683061e-07, + "loss": 0.82144821, + "num_input_tokens_seen": 137600355, + "step": 6393, + "time_per_iteration": 2.852456569671631 + }, + { + "auxiliary_loss_clip": 0.011469, + "auxiliary_loss_mlp": 0.01031786, + "balance_loss_clip": 1.04247189, + "balance_loss_mlp": 1.02447224, + "epoch": 0.7688330427463477, + "flos": 25155963717120.0, + "grad_norm": 1.8568655665848337, + "language_loss": 0.79115677, + "learning_rate": 5.347038825493617e-07, + "loss": 0.8129437, + "num_input_tokens_seen": 137621885, + "step": 6394, + "time_per_iteration": 2.974518299102783 + }, + { + "auxiliary_loss_clip": 0.01146031, + "auxiliary_loss_mlp": 0.01027049, + "balance_loss_clip": 1.04492068, + "balance_loss_mlp": 1.02000988, + "epoch": 0.7689532856369867, + "flos": 21211284113280.0, + "grad_norm": 1.9391086505255284, + "language_loss": 0.68518198, + "learning_rate": 5.341738196146732e-07, + "loss": 0.70691288, + "num_input_tokens_seen": 137640230, + "step": 6395, + "time_per_iteration": 2.767728567123413 + }, + { + "auxiliary_loss_clip": 0.01156088, + "auxiliary_loss_mlp": 0.01025046, + "balance_loss_clip": 1.04315162, + "balance_loss_mlp": 1.01698709, + "epoch": 0.7690735285276258, + "flos": 25119083427840.0, + "grad_norm": 2.2527831281248036, + "language_loss": 0.74111497, + "learning_rate": 5.336439790446503e-07, + "loss": 0.76292622, + "num_input_tokens_seen": 137659330, + "step": 6396, + "time_per_iteration": 2.7938590049743652 + }, + { + "auxiliary_loss_clip": 0.0113696, + "auxiliary_loss_mlp": 0.0102959, + "balance_loss_clip": 1.04042292, + "balance_loss_mlp": 1.0210309, + "epoch": 0.769193771418265, + "flos": 54744020640000.0, + "grad_norm": 1.727775642767777, + "language_loss": 0.62743169, + "learning_rate": 5.331143609196711e-07, + "loss": 0.64909714, + "num_input_tokens_seen": 137683145, + "step": 6397, + "time_per_iteration": 3.2088842391967773 + }, + { + "auxiliary_loss_clip": 0.01157963, + "auxiliary_loss_mlp": 0.01024093, + "balance_loss_clip": 1.0470643, + "balance_loss_mlp": 1.01577246, + "epoch": 0.769314014308904, + "flos": 37341890115840.0, + "grad_norm": 1.8757019017589975, + "language_loss": 0.77443129, + "learning_rate": 5.325849653200758e-07, + "loss": 0.79625189, + "num_input_tokens_seen": 137707095, + "step": 6398, + "time_per_iteration": 2.893498420715332 + }, + { + "auxiliary_loss_clip": 0.01167974, + "auxiliary_loss_mlp": 0.01031585, + "balance_loss_clip": 1.04662347, + "balance_loss_mlp": 1.02343142, + "epoch": 0.7694342571995431, + "flos": 20631686256000.0, + "grad_norm": 1.904909302724302, + "language_loss": 0.76431227, + "learning_rate": 5.32055792326175e-07, + "loss": 0.78630793, + "num_input_tokens_seen": 137725520, + "step": 6399, + "time_per_iteration": 2.6533384323120117 + }, + { + "auxiliary_loss_clip": 0.01155005, + "auxiliary_loss_mlp": 0.01030014, + "balance_loss_clip": 1.04845643, + "balance_loss_mlp": 1.02216971, + "epoch": 0.7695545000901821, + "flos": 24207706621440.0, + "grad_norm": 2.1655942662902494, + "language_loss": 0.72302449, + "learning_rate": 5.315268420182437e-07, + "loss": 0.74487472, + "num_input_tokens_seen": 137744195, + "step": 6400, + "time_per_iteration": 3.7383601665496826 + }, + { + "auxiliary_loss_clip": 0.01146278, + "auxiliary_loss_mlp": 0.00886083, + "balance_loss_clip": 1.04380679, + "balance_loss_mlp": 1.00060081, + "epoch": 0.7696747429808213, + "flos": 28001273708160.0, + "grad_norm": 1.6461877436839216, + "language_loss": 0.76541781, + "learning_rate": 5.309981144765221e-07, + "loss": 0.78574139, + "num_input_tokens_seen": 137764340, + "step": 6401, + "time_per_iteration": 2.8215575218200684 + }, + { + "auxiliary_loss_clip": 0.01138363, + "auxiliary_loss_mlp": 0.0102261, + "balance_loss_clip": 1.04224038, + "balance_loss_mlp": 1.01518679, + "epoch": 0.7697949858714603, + "flos": 11509550323200.0, + "grad_norm": 2.141403858694902, + "language_loss": 0.75492686, + "learning_rate": 5.304696097812196e-07, + "loss": 0.77653658, + "num_input_tokens_seen": 137780940, + "step": 6402, + "time_per_iteration": 3.846653461456299 + }, + { + "auxiliary_loss_clip": 0.01148837, + "auxiliary_loss_mlp": 0.01025641, + "balance_loss_clip": 1.04517567, + "balance_loss_mlp": 1.01644397, + "epoch": 0.7699152287620994, + "flos": 26688271956480.0, + "grad_norm": 2.971727198094374, + "language_loss": 0.60752976, + "learning_rate": 5.299413280125078e-07, + "loss": 0.62927449, + "num_input_tokens_seen": 137799250, + "step": 6403, + "time_per_iteration": 3.661050319671631 + }, + { + "auxiliary_loss_clip": 0.01151948, + "auxiliary_loss_mlp": 0.01026897, + "balance_loss_clip": 1.04447758, + "balance_loss_mlp": 1.01902342, + "epoch": 0.7700354716527386, + "flos": 16544944362240.0, + "grad_norm": 2.091732806373797, + "language_loss": 0.72892547, + "learning_rate": 5.294132692505284e-07, + "loss": 0.75071388, + "num_input_tokens_seen": 137817660, + "step": 6404, + "time_per_iteration": 2.816574811935425 + }, + { + "auxiliary_loss_clip": 0.01121826, + "auxiliary_loss_mlp": 0.0103389, + "balance_loss_clip": 1.04124308, + "balance_loss_mlp": 1.02619505, + "epoch": 0.7701557145433776, + "flos": 19242733196160.0, + "grad_norm": 3.522352986948875, + "language_loss": 0.79094458, + "learning_rate": 5.288854335753861e-07, + "loss": 0.81250179, + "num_input_tokens_seen": 137835920, + "step": 6405, + "time_per_iteration": 2.8881890773773193 + }, + { + "auxiliary_loss_clip": 0.01161959, + "auxiliary_loss_mlp": 0.01029318, + "balance_loss_clip": 1.04524505, + "balance_loss_mlp": 1.02139688, + "epoch": 0.7702759574340167, + "flos": 31685744211840.0, + "grad_norm": 1.6075387726319426, + "language_loss": 0.75374484, + "learning_rate": 5.283578210671551e-07, + "loss": 0.77565759, + "num_input_tokens_seen": 137858160, + "step": 6406, + "time_per_iteration": 2.778517723083496 + }, + { + "auxiliary_loss_clip": 0.01154343, + "auxiliary_loss_mlp": 0.01029137, + "balance_loss_clip": 1.04529321, + "balance_loss_mlp": 1.0212096, + "epoch": 0.7703962003246558, + "flos": 16800089644800.0, + "grad_norm": 1.961616697328101, + "language_loss": 0.76121932, + "learning_rate": 5.278304318058719e-07, + "loss": 0.78305411, + "num_input_tokens_seen": 137876015, + "step": 6407, + "time_per_iteration": 3.757410764694214 + }, + { + "auxiliary_loss_clip": 0.01118542, + "auxiliary_loss_mlp": 0.0103065, + "balance_loss_clip": 1.03819454, + "balance_loss_mlp": 1.0224607, + "epoch": 0.7705164432152949, + "flos": 35736072693120.0, + "grad_norm": 1.7363573020779053, + "language_loss": 0.78941977, + "learning_rate": 5.273032658715411e-07, + "loss": 0.81091172, + "num_input_tokens_seen": 137898825, + "step": 6408, + "time_per_iteration": 3.004408836364746 + }, + { + "auxiliary_loss_clip": 0.01127974, + "auxiliary_loss_mlp": 0.0103175, + "balance_loss_clip": 1.04207885, + "balance_loss_mlp": 1.02356601, + "epoch": 0.7706366861059339, + "flos": 23365960329600.0, + "grad_norm": 1.7036149944038375, + "language_loss": 0.76463509, + "learning_rate": 5.267763233441347e-07, + "loss": 0.78623235, + "num_input_tokens_seen": 137919455, + "step": 6409, + "time_per_iteration": 2.8907828330993652 + }, + { + "auxiliary_loss_clip": 0.01163853, + "auxiliary_loss_mlp": 0.01026595, + "balance_loss_clip": 1.04659462, + "balance_loss_mlp": 1.01805341, + "epoch": 0.7707569289965731, + "flos": 22929897219840.0, + "grad_norm": 2.3187040084894455, + "language_loss": 0.69554889, + "learning_rate": 5.26249604303588e-07, + "loss": 0.71745336, + "num_input_tokens_seen": 137937960, + "step": 6410, + "time_per_iteration": 2.747110366821289 + }, + { + "auxiliary_loss_clip": 0.01171748, + "auxiliary_loss_mlp": 0.01027126, + "balance_loss_clip": 1.04798198, + "balance_loss_mlp": 1.01925206, + "epoch": 0.7708771718872122, + "flos": 17420661941760.0, + "grad_norm": 2.332294143638694, + "language_loss": 0.78410393, + "learning_rate": 5.257231088298057e-07, + "loss": 0.80609262, + "num_input_tokens_seen": 137956370, + "step": 6411, + "time_per_iteration": 2.7135300636291504 + }, + { + "auxiliary_loss_clip": 0.01038165, + "auxiliary_loss_mlp": 0.01002669, + "balance_loss_clip": 1.00661886, + "balance_loss_mlp": 1.00115478, + "epoch": 0.7709974147778512, + "flos": 72241316248320.0, + "grad_norm": 0.8038546312767241, + "language_loss": 0.53934813, + "learning_rate": 5.25196837002655e-07, + "loss": 0.55975646, + "num_input_tokens_seen": 138016080, + "step": 6412, + "time_per_iteration": 3.362565040588379 + }, + { + "auxiliary_loss_clip": 0.01147444, + "auxiliary_loss_mlp": 0.0102743, + "balance_loss_clip": 1.04457521, + "balance_loss_mlp": 1.01891279, + "epoch": 0.7711176576684904, + "flos": 39859694876160.0, + "grad_norm": 2.109980238006077, + "language_loss": 0.68224698, + "learning_rate": 5.24670788901971e-07, + "loss": 0.7039957, + "num_input_tokens_seen": 138039170, + "step": 6413, + "time_per_iteration": 2.8644461631774902 + }, + { + "auxiliary_loss_clip": 0.01150024, + "auxiliary_loss_mlp": 0.01026284, + "balance_loss_clip": 1.04612064, + "balance_loss_mlp": 1.01781452, + "epoch": 0.7712379005591294, + "flos": 36976391274240.0, + "grad_norm": 2.2467144200898503, + "language_loss": 0.6876229, + "learning_rate": 5.241449646075557e-07, + "loss": 0.70938599, + "num_input_tokens_seen": 138062395, + "step": 6414, + "time_per_iteration": 2.866446018218994 + }, + { + "auxiliary_loss_clip": 0.01167393, + "auxiliary_loss_mlp": 0.01027395, + "balance_loss_clip": 1.046929, + "balance_loss_mlp": 1.01955664, + "epoch": 0.7713581434497685, + "flos": 22776773541120.0, + "grad_norm": 1.875649967149126, + "language_loss": 0.72350919, + "learning_rate": 5.236193641991762e-07, + "loss": 0.74545705, + "num_input_tokens_seen": 138080325, + "step": 6415, + "time_per_iteration": 2.701063632965088 + }, + { + "auxiliary_loss_clip": 0.0114872, + "auxiliary_loss_mlp": 0.01029328, + "balance_loss_clip": 1.04427993, + "balance_loss_mlp": 1.02171636, + "epoch": 0.7714783863404077, + "flos": 24097460803200.0, + "grad_norm": 2.482029523676293, + "language_loss": 0.69907069, + "learning_rate": 5.23093987756565e-07, + "loss": 0.72085118, + "num_input_tokens_seen": 138099020, + "step": 6416, + "time_per_iteration": 2.8608131408691406 + }, + { + "auxiliary_loss_clip": 0.01147096, + "auxiliary_loss_mlp": 0.01029846, + "balance_loss_clip": 1.04098296, + "balance_loss_mlp": 1.02119112, + "epoch": 0.7715986292310467, + "flos": 21063655215360.0, + "grad_norm": 1.7959559025104912, + "language_loss": 0.75453418, + "learning_rate": 5.225688353594217e-07, + "loss": 0.77630365, + "num_input_tokens_seen": 138118650, + "step": 6417, + "time_per_iteration": 2.901333808898926 + }, + { + "auxiliary_loss_clip": 0.01158019, + "auxiliary_loss_mlp": 0.00886156, + "balance_loss_clip": 1.04815137, + "balance_loss_mlp": 1.00064218, + "epoch": 0.7717188721216858, + "flos": 20594877793920.0, + "grad_norm": 2.133632457310335, + "language_loss": 0.77655464, + "learning_rate": 5.220439070874108e-07, + "loss": 0.79699641, + "num_input_tokens_seen": 138137890, + "step": 6418, + "time_per_iteration": 2.8254799842834473 + }, + { + "auxiliary_loss_clip": 0.0116145, + "auxiliary_loss_mlp": 0.01029094, + "balance_loss_clip": 1.04883814, + "balance_loss_mlp": 1.02135158, + "epoch": 0.7718391150123249, + "flos": 26250951870720.0, + "grad_norm": 1.7607599344262588, + "language_loss": 0.70881891, + "learning_rate": 5.215192030201652e-07, + "loss": 0.73072433, + "num_input_tokens_seen": 138158880, + "step": 6419, + "time_per_iteration": 2.792027473449707 + }, + { + "auxiliary_loss_clip": 0.01130266, + "auxiliary_loss_mlp": 0.01025339, + "balance_loss_clip": 1.03931057, + "balance_loss_mlp": 1.01745892, + "epoch": 0.771959357902964, + "flos": 22049762267520.0, + "grad_norm": 1.7561882577546784, + "language_loss": 0.86230874, + "learning_rate": 5.209947232372798e-07, + "loss": 0.88386476, + "num_input_tokens_seen": 138176370, + "step": 6420, + "time_per_iteration": 2.7329492568969727 + }, + { + "auxiliary_loss_clip": 0.01162941, + "auxiliary_loss_mlp": 0.00886424, + "balance_loss_clip": 1.04473031, + "balance_loss_mlp": 1.00068855, + "epoch": 0.772079600793603, + "flos": 30446000248320.0, + "grad_norm": 1.8401133609836744, + "language_loss": 0.81092525, + "learning_rate": 5.204704678183196e-07, + "loss": 0.83141887, + "num_input_tokens_seen": 138195105, + "step": 6421, + "time_per_iteration": 2.8133437633514404 + }, + { + "auxiliary_loss_clip": 0.01171875, + "auxiliary_loss_mlp": 0.01025067, + "balance_loss_clip": 1.04936075, + "balance_loss_mlp": 1.01654303, + "epoch": 0.7721998436842422, + "flos": 12969857750400.0, + "grad_norm": 1.7892277215620005, + "language_loss": 0.85457152, + "learning_rate": 5.19946436842813e-07, + "loss": 0.8765409, + "num_input_tokens_seen": 138212235, + "step": 6422, + "time_per_iteration": 2.5986921787261963 + }, + { + "auxiliary_loss_clip": 0.01140107, + "auxiliary_loss_mlp": 0.01024069, + "balance_loss_clip": 1.04449427, + "balance_loss_mlp": 1.01677608, + "epoch": 0.7723200865748813, + "flos": 32635509678720.0, + "grad_norm": 1.5287052657402904, + "language_loss": 0.68408126, + "learning_rate": 5.194226303902546e-07, + "loss": 0.70572305, + "num_input_tokens_seen": 138231970, + "step": 6423, + "time_per_iteration": 2.8687708377838135 + }, + { + "auxiliary_loss_clip": 0.01149561, + "auxiliary_loss_mlp": 0.01023823, + "balance_loss_clip": 1.04298782, + "balance_loss_mlp": 1.0158, + "epoch": 0.7724403294655203, + "flos": 21105707063040.0, + "grad_norm": 1.8376121213576846, + "language_loss": 0.70606637, + "learning_rate": 5.188990485401072e-07, + "loss": 0.72780025, + "num_input_tokens_seen": 138251175, + "step": 6424, + "time_per_iteration": 2.6982555389404297 + }, + { + "auxiliary_loss_clip": 0.01160742, + "auxiliary_loss_mlp": 0.01025394, + "balance_loss_clip": 1.046543, + "balance_loss_mlp": 1.01741314, + "epoch": 0.7725605723561595, + "flos": 22090736707200.0, + "grad_norm": 2.009534708874606, + "language_loss": 0.86136854, + "learning_rate": 5.183756913717954e-07, + "loss": 0.88322985, + "num_input_tokens_seen": 138270950, + "step": 6425, + "time_per_iteration": 2.707984685897827 + }, + { + "auxiliary_loss_clip": 0.01148034, + "auxiliary_loss_mlp": 0.01027507, + "balance_loss_clip": 1.04670739, + "balance_loss_mlp": 1.01964211, + "epoch": 0.7726808152467985, + "flos": 34495610457600.0, + "grad_norm": 1.7581034940398828, + "language_loss": 0.73351324, + "learning_rate": 5.178525589647136e-07, + "loss": 0.75526869, + "num_input_tokens_seen": 138292590, + "step": 6426, + "time_per_iteration": 3.810053586959839 + }, + { + "auxiliary_loss_clip": 0.01155791, + "auxiliary_loss_mlp": 0.0102545, + "balance_loss_clip": 1.04452229, + "balance_loss_mlp": 1.01771975, + "epoch": 0.7728010581374376, + "flos": 22306344094080.0, + "grad_norm": 6.284178530563206, + "language_loss": 0.78890246, + "learning_rate": 5.173296513982197e-07, + "loss": 0.8107149, + "num_input_tokens_seen": 138311115, + "step": 6427, + "time_per_iteration": 2.7096588611602783 + }, + { + "auxiliary_loss_clip": 0.01154675, + "auxiliary_loss_mlp": 0.01031429, + "balance_loss_clip": 1.04747558, + "balance_loss_mlp": 1.0231055, + "epoch": 0.7729213010280768, + "flos": 27126453968640.0, + "grad_norm": 2.1512322410602454, + "language_loss": 0.64916039, + "learning_rate": 5.168069687516398e-07, + "loss": 0.67102146, + "num_input_tokens_seen": 138330885, + "step": 6428, + "time_per_iteration": 4.699500322341919 + }, + { + "auxiliary_loss_clip": 0.01153983, + "auxiliary_loss_mlp": 0.01028103, + "balance_loss_clip": 1.04873538, + "balance_loss_mlp": 1.02014029, + "epoch": 0.7730415439187158, + "flos": 18150223080960.0, + "grad_norm": 2.052507232043789, + "language_loss": 0.71742177, + "learning_rate": 5.16284511104263e-07, + "loss": 0.73924267, + "num_input_tokens_seen": 138350020, + "step": 6429, + "time_per_iteration": 2.831617832183838 + }, + { + "auxiliary_loss_clip": 0.01155173, + "auxiliary_loss_mlp": 0.01030617, + "balance_loss_clip": 1.04766023, + "balance_loss_mlp": 1.02227855, + "epoch": 0.7731617868093549, + "flos": 11947480940160.0, + "grad_norm": 2.9156784123721224, + "language_loss": 0.81016123, + "learning_rate": 5.157622785353457e-07, + "loss": 0.83201915, + "num_input_tokens_seen": 138368135, + "step": 6430, + "time_per_iteration": 2.658630132675171 + }, + { + "auxiliary_loss_clip": 0.010543, + "auxiliary_loss_mlp": 0.01001746, + "balance_loss_clip": 1.00718939, + "balance_loss_mlp": 1.00025558, + "epoch": 0.7732820296999939, + "flos": 64201027069440.0, + "grad_norm": 0.6490980938488422, + "language_loss": 0.60367084, + "learning_rate": 5.152402711241113e-07, + "loss": 0.62423128, + "num_input_tokens_seen": 138436040, + "step": 6431, + "time_per_iteration": 3.300231456756592 + }, + { + "auxiliary_loss_clip": 0.01138004, + "auxiliary_loss_mlp": 0.01024117, + "balance_loss_clip": 1.04142141, + "balance_loss_mlp": 1.01669896, + "epoch": 0.7734022725906331, + "flos": 25302191984640.0, + "grad_norm": 1.6445762788948552, + "language_loss": 0.82844424, + "learning_rate": 5.147184889497465e-07, + "loss": 0.85006541, + "num_input_tokens_seen": 138455510, + "step": 6432, + "time_per_iteration": 3.7640910148620605 + }, + { + "auxiliary_loss_clip": 0.01135576, + "auxiliary_loss_mlp": 0.01026925, + "balance_loss_clip": 1.043244, + "balance_loss_mlp": 1.01865768, + "epoch": 0.7735225154812722, + "flos": 17347440067200.0, + "grad_norm": 2.0826646547723384, + "language_loss": 0.79869819, + "learning_rate": 5.141969320914072e-07, + "loss": 0.82032323, + "num_input_tokens_seen": 138473015, + "step": 6433, + "time_per_iteration": 2.760573148727417 + }, + { + "auxiliary_loss_clip": 0.01173117, + "auxiliary_loss_mlp": 0.01029275, + "balance_loss_clip": 1.04794073, + "balance_loss_mlp": 1.02123404, + "epoch": 0.7736427583719112, + "flos": 32630086725120.0, + "grad_norm": 2.650440088798427, + "language_loss": 0.62604141, + "learning_rate": 5.136756006282113e-07, + "loss": 0.64806533, + "num_input_tokens_seen": 138491680, + "step": 6434, + "time_per_iteration": 2.792210817337036 + }, + { + "auxiliary_loss_clip": 0.0117122, + "auxiliary_loss_mlp": 0.01024763, + "balance_loss_clip": 1.04793119, + "balance_loss_mlp": 1.01687169, + "epoch": 0.7737630012625504, + "flos": 19860073269120.0, + "grad_norm": 2.1740655529621336, + "language_loss": 0.85418844, + "learning_rate": 5.131544946392446e-07, + "loss": 0.87614822, + "num_input_tokens_seen": 138506960, + "step": 6435, + "time_per_iteration": 2.6208839416503906 + }, + { + "auxiliary_loss_clip": 0.01151616, + "auxiliary_loss_mlp": 0.01031945, + "balance_loss_clip": 1.04816937, + "balance_loss_mlp": 1.02392828, + "epoch": 0.7738832441531894, + "flos": 36022639397760.0, + "grad_norm": 2.766550913628017, + "language_loss": 0.63984275, + "learning_rate": 5.126336142035592e-07, + "loss": 0.66167831, + "num_input_tokens_seen": 138526995, + "step": 6436, + "time_per_iteration": 2.8954248428344727 + }, + { + "auxiliary_loss_clip": 0.01149763, + "auxiliary_loss_mlp": 0.01025062, + "balance_loss_clip": 1.04342556, + "balance_loss_mlp": 1.01711655, + "epoch": 0.7740034870438285, + "flos": 13405274415360.0, + "grad_norm": 2.251614921828636, + "language_loss": 0.72606391, + "learning_rate": 5.121129594001721e-07, + "loss": 0.74781215, + "num_input_tokens_seen": 138541260, + "step": 6437, + "time_per_iteration": 2.702967643737793 + }, + { + "auxiliary_loss_clip": 0.01162445, + "auxiliary_loss_mlp": 0.01026148, + "balance_loss_clip": 1.04835904, + "balance_loss_mlp": 1.01794028, + "epoch": 0.7741237299344677, + "flos": 22086714384000.0, + "grad_norm": 1.664329065680147, + "language_loss": 0.81070495, + "learning_rate": 5.115925303080661e-07, + "loss": 0.83259094, + "num_input_tokens_seen": 138560970, + "step": 6438, + "time_per_iteration": 2.695066213607788 + }, + { + "auxiliary_loss_clip": 0.0114339, + "auxiliary_loss_mlp": 0.01025812, + "balance_loss_clip": 1.04502153, + "balance_loss_mlp": 1.01771724, + "epoch": 0.7742439728251067, + "flos": 19864777950720.0, + "grad_norm": 2.0398894262907326, + "language_loss": 0.78961957, + "learning_rate": 5.110723270061899e-07, + "loss": 0.8113116, + "num_input_tokens_seen": 138577460, + "step": 6439, + "time_per_iteration": 2.7366857528686523 + }, + { + "auxiliary_loss_clip": 0.01167665, + "auxiliary_loss_mlp": 0.01025838, + "balance_loss_clip": 1.04711318, + "balance_loss_mlp": 1.01815248, + "epoch": 0.7743642157157458, + "flos": 16690167048960.0, + "grad_norm": 1.7896416580151306, + "language_loss": 0.79441303, + "learning_rate": 5.105523495734572e-07, + "loss": 0.81634808, + "num_input_tokens_seen": 138594860, + "step": 6440, + "time_per_iteration": 2.643336772918701 + }, + { + "auxiliary_loss_clip": 0.01169379, + "auxiliary_loss_mlp": 0.01027784, + "balance_loss_clip": 1.04628193, + "balance_loss_mlp": 1.02019048, + "epoch": 0.7744844586063849, + "flos": 20304360593280.0, + "grad_norm": 1.492318388746716, + "language_loss": 0.75264555, + "learning_rate": 5.100325980887499e-07, + "loss": 0.7746172, + "num_input_tokens_seen": 138614785, + "step": 6441, + "time_per_iteration": 2.6368138790130615 + }, + { + "auxiliary_loss_clip": 0.01153896, + "auxiliary_loss_mlp": 0.0102886, + "balance_loss_clip": 1.04716921, + "balance_loss_mlp": 1.0211947, + "epoch": 0.774604701497024, + "flos": 22966705681920.0, + "grad_norm": 2.1008369129944766, + "language_loss": 0.83661652, + "learning_rate": 5.095130726309116e-07, + "loss": 0.85844409, + "num_input_tokens_seen": 138634960, + "step": 6442, + "time_per_iteration": 2.7466933727264404 + }, + { + "auxiliary_loss_clip": 0.01058434, + "auxiliary_loss_mlp": 0.01001644, + "balance_loss_clip": 1.00680161, + "balance_loss_mlp": 1.0002079, + "epoch": 0.774724944387663, + "flos": 60288523073280.0, + "grad_norm": 0.7984002585515194, + "language_loss": 0.58994806, + "learning_rate": 5.089937732787559e-07, + "loss": 0.61054885, + "num_input_tokens_seen": 138699520, + "step": 6443, + "time_per_iteration": 3.2566468715667725 + }, + { + "auxiliary_loss_clip": 0.0114202, + "auxiliary_loss_mlp": 0.01029371, + "balance_loss_clip": 1.04290843, + "balance_loss_mlp": 1.02128291, + "epoch": 0.7748451872783022, + "flos": 26761026954240.0, + "grad_norm": 2.472252374821544, + "language_loss": 0.66072154, + "learning_rate": 5.084747001110592e-07, + "loss": 0.68243551, + "num_input_tokens_seen": 138719145, + "step": 6444, + "time_per_iteration": 2.7927968502044678 + }, + { + "auxiliary_loss_clip": 0.01154563, + "auxiliary_loss_mlp": 0.00885984, + "balance_loss_clip": 1.04745233, + "balance_loss_mlp": 1.00067687, + "epoch": 0.7749654301689413, + "flos": 30338627518080.0, + "grad_norm": 1.7486126756711293, + "language_loss": 0.70216787, + "learning_rate": 5.07955853206564e-07, + "loss": 0.7225734, + "num_input_tokens_seen": 138743850, + "step": 6445, + "time_per_iteration": 2.7727837562561035 + }, + { + "auxiliary_loss_clip": 0.01165899, + "auxiliary_loss_mlp": 0.01024384, + "balance_loss_clip": 1.04819417, + "balance_loss_mlp": 1.01659954, + "epoch": 0.7750856730595803, + "flos": 43179851687040.0, + "grad_norm": 2.395883403485001, + "language_loss": 0.70944417, + "learning_rate": 5.074372326439807e-07, + "loss": 0.73134696, + "num_input_tokens_seen": 138766860, + "step": 6446, + "time_per_iteration": 2.849189519882202 + }, + { + "auxiliary_loss_clip": 0.01146218, + "auxiliary_loss_mlp": 0.01028715, + "balance_loss_clip": 1.04424381, + "balance_loss_mlp": 1.02088928, + "epoch": 0.7752059159502195, + "flos": 17640040256640.0, + "grad_norm": 2.3717437849013656, + "language_loss": 0.73808289, + "learning_rate": 5.069188385019814e-07, + "loss": 0.7598322, + "num_input_tokens_seen": 138784560, + "step": 6447, + "time_per_iteration": 2.7471230030059814 + }, + { + "auxiliary_loss_clip": 0.01138859, + "auxiliary_loss_mlp": 0.0102733, + "balance_loss_clip": 1.04153287, + "balance_loss_mlp": 1.01944423, + "epoch": 0.7753261588408585, + "flos": 12677688524160.0, + "grad_norm": 3.3066213783186362, + "language_loss": 0.61228228, + "learning_rate": 5.064006708592077e-07, + "loss": 0.63394415, + "num_input_tokens_seen": 138800805, + "step": 6448, + "time_per_iteration": 2.752129554748535 + }, + { + "auxiliary_loss_clip": 0.01144612, + "auxiliary_loss_mlp": 0.01026575, + "balance_loss_clip": 1.04481506, + "balance_loss_mlp": 1.01899004, + "epoch": 0.7754464017314976, + "flos": 16690741666560.0, + "grad_norm": 2.2789435220738454, + "language_loss": 0.75649601, + "learning_rate": 5.058827297942641e-07, + "loss": 0.7782079, + "num_input_tokens_seen": 138815910, + "step": 6449, + "time_per_iteration": 2.7409112453460693 + }, + { + "auxiliary_loss_clip": 0.01149571, + "auxiliary_loss_mlp": 0.01027337, + "balance_loss_clip": 1.04494572, + "balance_loss_mlp": 1.01938629, + "epoch": 0.7755666446221368, + "flos": 19718944732800.0, + "grad_norm": 1.8398734356818327, + "language_loss": 0.75210494, + "learning_rate": 5.053650153857237e-07, + "loss": 0.77387404, + "num_input_tokens_seen": 138834920, + "step": 6450, + "time_per_iteration": 2.8145105838775635 + }, + { + "auxiliary_loss_clip": 0.0116125, + "auxiliary_loss_mlp": 0.01028572, + "balance_loss_clip": 1.04834843, + "balance_loss_mlp": 1.02021503, + "epoch": 0.7756868875127758, + "flos": 18693623007360.0, + "grad_norm": 1.678740811302926, + "language_loss": 0.6991601, + "learning_rate": 5.048475277121214e-07, + "loss": 0.72105831, + "num_input_tokens_seen": 138852135, + "step": 6451, + "time_per_iteration": 2.659606456756592 + }, + { + "auxiliary_loss_clip": 0.01160786, + "auxiliary_loss_mlp": 0.01026911, + "balance_loss_clip": 1.04672813, + "balance_loss_mlp": 1.01875687, + "epoch": 0.7758071304034149, + "flos": 28404191543040.0, + "grad_norm": 3.4999127266065115, + "language_loss": 0.76979256, + "learning_rate": 5.043302668519598e-07, + "loss": 0.79166949, + "num_input_tokens_seen": 138871470, + "step": 6452, + "time_per_iteration": 3.720755100250244 + }, + { + "auxiliary_loss_clip": 0.01163644, + "auxiliary_loss_mlp": 0.01030259, + "balance_loss_clip": 1.04551721, + "balance_loss_mlp": 1.02206361, + "epoch": 0.775927373294054, + "flos": 20595344670720.0, + "grad_norm": 1.9330744511373805, + "language_loss": 0.71861196, + "learning_rate": 5.038132328837079e-07, + "loss": 0.74055099, + "num_input_tokens_seen": 138889860, + "step": 6453, + "time_per_iteration": 2.7444748878479004 + }, + { + "auxiliary_loss_clip": 0.01162193, + "auxiliary_loss_mlp": 0.01025526, + "balance_loss_clip": 1.04655218, + "balance_loss_mlp": 1.01775324, + "epoch": 0.7760476161846931, + "flos": 22526368853760.0, + "grad_norm": 1.9909851996938563, + "language_loss": 0.7367478, + "learning_rate": 5.032964258857993e-07, + "loss": 0.75862497, + "num_input_tokens_seen": 138909955, + "step": 6454, + "time_per_iteration": 4.6292266845703125 + }, + { + "auxiliary_loss_clip": 0.01157973, + "auxiliary_loss_mlp": 0.01028841, + "balance_loss_clip": 1.04249358, + "balance_loss_mlp": 1.02062774, + "epoch": 0.7761678590753321, + "flos": 48651488403840.0, + "grad_norm": 1.5375858262846924, + "language_loss": 0.68609017, + "learning_rate": 5.027798459366329e-07, + "loss": 0.70795834, + "num_input_tokens_seen": 138935320, + "step": 6455, + "time_per_iteration": 2.9076056480407715 + }, + { + "auxiliary_loss_clip": 0.01163478, + "auxiliary_loss_mlp": 0.01024444, + "balance_loss_clip": 1.04509687, + "balance_loss_mlp": 1.01619494, + "epoch": 0.7762881019659713, + "flos": 26177047637760.0, + "grad_norm": 1.618081430363688, + "language_loss": 0.63347048, + "learning_rate": 5.02263493114573e-07, + "loss": 0.65534973, + "num_input_tokens_seen": 138957115, + "step": 6456, + "time_per_iteration": 2.781003475189209 + }, + { + "auxiliary_loss_clip": 0.01167882, + "auxiliary_loss_mlp": 0.01027537, + "balance_loss_clip": 1.04625833, + "balance_loss_mlp": 1.01923728, + "epoch": 0.7764083448566104, + "flos": 20588341518720.0, + "grad_norm": 2.9452849642913557, + "language_loss": 0.76877868, + "learning_rate": 5.017473674979502e-07, + "loss": 0.79073286, + "num_input_tokens_seen": 138973140, + "step": 6457, + "time_per_iteration": 2.7962422370910645 + }, + { + "auxiliary_loss_clip": 0.01037858, + "auxiliary_loss_mlp": 0.01001483, + "balance_loss_clip": 1.01115048, + "balance_loss_mlp": 1.00002861, + "epoch": 0.7765285877472494, + "flos": 67293078560640.0, + "grad_norm": 0.7391165750597587, + "language_loss": 0.5838232, + "learning_rate": 5.01231469165061e-07, + "loss": 0.60421669, + "num_input_tokens_seen": 139028965, + "step": 6458, + "time_per_iteration": 3.2247114181518555 + }, + { + "auxiliary_loss_clip": 0.01053039, + "auxiliary_loss_mlp": 0.01001387, + "balance_loss_clip": 1.00629747, + "balance_loss_mlp": 0.99977791, + "epoch": 0.7766488306378886, + "flos": 61344476121600.0, + "grad_norm": 0.8266805964150472, + "language_loss": 0.5687139, + "learning_rate": 5.007157981941663e-07, + "loss": 0.58925819, + "num_input_tokens_seen": 139094325, + "step": 6459, + "time_per_iteration": 4.361464262008667 + }, + { + "auxiliary_loss_clip": 0.01049428, + "auxiliary_loss_mlp": 0.01002931, + "balance_loss_clip": 1.00782549, + "balance_loss_mlp": 1.00142908, + "epoch": 0.7767690735285276, + "flos": 62946199393920.0, + "grad_norm": 0.8733179340914952, + "language_loss": 0.67515874, + "learning_rate": 5.002003546634928e-07, + "loss": 0.69568241, + "num_input_tokens_seen": 139150425, + "step": 6460, + "time_per_iteration": 3.2426183223724365 + }, + { + "auxiliary_loss_clip": 0.01127767, + "auxiliary_loss_mlp": 0.01026424, + "balance_loss_clip": 1.0415349, + "balance_loss_mlp": 1.01890206, + "epoch": 0.7768893164191667, + "flos": 20886400575360.0, + "grad_norm": 1.6402956381596951, + "language_loss": 0.75865722, + "learning_rate": 4.996851386512331e-07, + "loss": 0.78019911, + "num_input_tokens_seen": 139169130, + "step": 6461, + "time_per_iteration": 2.81331205368042 + }, + { + "auxiliary_loss_clip": 0.01150856, + "auxiliary_loss_mlp": 0.01027327, + "balance_loss_clip": 1.04466176, + "balance_loss_mlp": 1.01938128, + "epoch": 0.7770095593098058, + "flos": 20704584908160.0, + "grad_norm": 2.1057995208344886, + "language_loss": 0.83084464, + "learning_rate": 4.991701502355444e-07, + "loss": 0.85262644, + "num_input_tokens_seen": 139189595, + "step": 6462, + "time_per_iteration": 2.719038248062134 + }, + { + "auxiliary_loss_clip": 0.01164845, + "auxiliary_loss_mlp": 0.01029407, + "balance_loss_clip": 1.0465281, + "balance_loss_mlp": 1.02094877, + "epoch": 0.7771298022004449, + "flos": 24717709877760.0, + "grad_norm": 1.5226176103716187, + "language_loss": 0.76113981, + "learning_rate": 4.986553894945518e-07, + "loss": 0.78308225, + "num_input_tokens_seen": 139210805, + "step": 6463, + "time_per_iteration": 2.8163058757781982 + }, + { + "auxiliary_loss_clip": 0.01122824, + "auxiliary_loss_mlp": 0.01027595, + "balance_loss_clip": 1.03809094, + "balance_loss_mlp": 1.02028763, + "epoch": 0.777250045091084, + "flos": 25009232659200.0, + "grad_norm": 1.9060028477953286, + "language_loss": 0.86251527, + "learning_rate": 4.981408565063416e-07, + "loss": 0.88401943, + "num_input_tokens_seen": 139230750, + "step": 6464, + "time_per_iteration": 2.792097568511963 + }, + { + "auxiliary_loss_clip": 0.01170151, + "auxiliary_loss_mlp": 0.01025385, + "balance_loss_clip": 1.04678917, + "balance_loss_mlp": 1.01734459, + "epoch": 0.777370287981723, + "flos": 20119887319680.0, + "grad_norm": 1.8657194709844105, + "language_loss": 0.76511526, + "learning_rate": 4.976265513489701e-07, + "loss": 0.78707063, + "num_input_tokens_seen": 139250720, + "step": 6465, + "time_per_iteration": 2.644108295440674 + }, + { + "auxiliary_loss_clip": 0.01157531, + "auxiliary_loss_mlp": 0.01026398, + "balance_loss_clip": 1.04304087, + "balance_loss_mlp": 1.01844645, + "epoch": 0.7774905308723622, + "flos": 21718809331200.0, + "grad_norm": 1.7379898480283291, + "language_loss": 0.80326235, + "learning_rate": 4.971124741004562e-07, + "loss": 0.82510161, + "num_input_tokens_seen": 139269720, + "step": 6466, + "time_per_iteration": 2.6681010723114014 + }, + { + "auxiliary_loss_clip": 0.01160284, + "auxiliary_loss_mlp": 0.01031815, + "balance_loss_clip": 1.0449121, + "balance_loss_mlp": 1.02385807, + "epoch": 0.7776107737630013, + "flos": 16034115093120.0, + "grad_norm": 1.7175869054940383, + "language_loss": 0.76142311, + "learning_rate": 4.965986248387846e-07, + "loss": 0.78334415, + "num_input_tokens_seen": 139288035, + "step": 6467, + "time_per_iteration": 2.8859241008758545 + }, + { + "auxiliary_loss_clip": 0.01152668, + "auxiliary_loss_mlp": 0.01025016, + "balance_loss_clip": 1.04391623, + "balance_loss_mlp": 1.01688576, + "epoch": 0.7777310166536403, + "flos": 24790895838720.0, + "grad_norm": 6.60317914338319, + "language_loss": 0.77220076, + "learning_rate": 4.960850036419073e-07, + "loss": 0.79397762, + "num_input_tokens_seen": 139307135, + "step": 6468, + "time_per_iteration": 2.747373104095459 + }, + { + "auxiliary_loss_clip": 0.01147891, + "auxiliary_loss_mlp": 0.01029149, + "balance_loss_clip": 1.04395223, + "balance_loss_mlp": 1.02085781, + "epoch": 0.7778512595442795, + "flos": 17272530253440.0, + "grad_norm": 1.684877848624507, + "language_loss": 0.78631079, + "learning_rate": 4.955716105877378e-07, + "loss": 0.80808121, + "num_input_tokens_seen": 139325905, + "step": 6469, + "time_per_iteration": 2.6317806243896484 + }, + { + "auxiliary_loss_clip": 0.01163316, + "auxiliary_loss_mlp": 0.00886396, + "balance_loss_clip": 1.04554093, + "balance_loss_mlp": 1.00076914, + "epoch": 0.7779715024349185, + "flos": 17748418567680.0, + "grad_norm": 2.088430453689837, + "language_loss": 0.83087087, + "learning_rate": 4.950584457541598e-07, + "loss": 0.85136807, + "num_input_tokens_seen": 139344370, + "step": 6470, + "time_per_iteration": 2.711958885192871 + }, + { + "auxiliary_loss_clip": 0.0116269, + "auxiliary_loss_mlp": 0.01026737, + "balance_loss_clip": 1.04564369, + "balance_loss_mlp": 1.01844311, + "epoch": 0.7780917453255576, + "flos": 24316875031680.0, + "grad_norm": 1.3767846030784003, + "language_loss": 0.81741416, + "learning_rate": 4.945455092190183e-07, + "loss": 0.83930838, + "num_input_tokens_seen": 139365625, + "step": 6471, + "time_per_iteration": 2.901092290878296 + }, + { + "auxiliary_loss_clip": 0.01057993, + "auxiliary_loss_mlp": 0.01002239, + "balance_loss_clip": 1.00629401, + "balance_loss_mlp": 1.0008502, + "epoch": 0.7782119882161967, + "flos": 56364601530240.0, + "grad_norm": 0.6845885910441908, + "language_loss": 0.56041378, + "learning_rate": 4.940328010601271e-07, + "loss": 0.58101612, + "num_input_tokens_seen": 139430540, + "step": 6472, + "time_per_iteration": 3.2583844661712646 + }, + { + "auxiliary_loss_clip": 0.01155154, + "auxiliary_loss_mlp": 0.01038221, + "balance_loss_clip": 1.05152428, + "balance_loss_mlp": 1.02935231, + "epoch": 0.7783322311068358, + "flos": 46789986994560.0, + "grad_norm": 2.5884799406831283, + "language_loss": 0.76694053, + "learning_rate": 4.935203213552621e-07, + "loss": 0.78887427, + "num_input_tokens_seen": 139454280, + "step": 6473, + "time_per_iteration": 2.967893362045288 + }, + { + "auxiliary_loss_clip": 0.0115128, + "auxiliary_loss_mlp": 0.0102852, + "balance_loss_clip": 1.04481554, + "balance_loss_mlp": 1.0205096, + "epoch": 0.7784524739974749, + "flos": 19057864872960.0, + "grad_norm": 2.118957828862329, + "language_loss": 0.66887265, + "learning_rate": 4.930080701821662e-07, + "loss": 0.69067061, + "num_input_tokens_seen": 139471745, + "step": 6474, + "time_per_iteration": 2.702944755554199 + }, + { + "auxiliary_loss_clip": 0.01149635, + "auxiliary_loss_mlp": 0.01029987, + "balance_loss_clip": 1.04365349, + "balance_loss_mlp": 1.02157688, + "epoch": 0.778572716888114, + "flos": 24791111320320.0, + "grad_norm": 2.0780319184240685, + "language_loss": 0.76965916, + "learning_rate": 4.92496047618548e-07, + "loss": 0.79145539, + "num_input_tokens_seen": 139491505, + "step": 6475, + "time_per_iteration": 2.7919774055480957 + }, + { + "auxiliary_loss_clip": 0.01161813, + "auxiliary_loss_mlp": 0.01025157, + "balance_loss_clip": 1.04816866, + "balance_loss_mlp": 1.01711679, + "epoch": 0.7786929597787531, + "flos": 20078086867200.0, + "grad_norm": 1.9464157685695829, + "language_loss": 0.77531052, + "learning_rate": 4.919842537420811e-07, + "loss": 0.79718018, + "num_input_tokens_seen": 139508620, + "step": 6476, + "time_per_iteration": 2.6912331581115723 + }, + { + "auxiliary_loss_clip": 0.011526, + "auxiliary_loss_mlp": 0.01030014, + "balance_loss_clip": 1.047418, + "balance_loss_mlp": 1.02164531, + "epoch": 0.7788132026693921, + "flos": 21872220318720.0, + "grad_norm": 1.6170112776248402, + "language_loss": 0.79052037, + "learning_rate": 4.91472688630404e-07, + "loss": 0.81234646, + "num_input_tokens_seen": 139529360, + "step": 6477, + "time_per_iteration": 3.79380202293396 + }, + { + "auxiliary_loss_clip": 0.01168047, + "auxiliary_loss_mlp": 0.0103373, + "balance_loss_clip": 1.04663825, + "balance_loss_mlp": 1.02623796, + "epoch": 0.7789334455600313, + "flos": 11181937351680.0, + "grad_norm": 1.6317659271691016, + "language_loss": 0.73836422, + "learning_rate": 4.909613523611202e-07, + "loss": 0.76038194, + "num_input_tokens_seen": 139546240, + "step": 6478, + "time_per_iteration": 2.6302852630615234 + }, + { + "auxiliary_loss_clip": 0.0111978, + "auxiliary_loss_mlp": 0.00886091, + "balance_loss_clip": 1.03933954, + "balance_loss_mlp": 1.00057983, + "epoch": 0.7790536884506704, + "flos": 28695427015680.0, + "grad_norm": 1.6886991943388725, + "language_loss": 0.74567199, + "learning_rate": 4.904502450117991e-07, + "loss": 0.76573062, + "num_input_tokens_seen": 139567200, + "step": 6479, + "time_per_iteration": 3.8151326179504395 + }, + { + "auxiliary_loss_clip": 0.01145116, + "auxiliary_loss_mlp": 0.010334, + "balance_loss_clip": 1.04482687, + "balance_loss_mlp": 1.02506733, + "epoch": 0.7791739313413094, + "flos": 11072302064640.0, + "grad_norm": 2.070269833376294, + "language_loss": 0.72338313, + "learning_rate": 4.899393666599762e-07, + "loss": 0.74516833, + "num_input_tokens_seen": 139583775, + "step": 6480, + "time_per_iteration": 3.5186240673065186 + }, + { + "auxiliary_loss_clip": 0.01168686, + "auxiliary_loss_mlp": 0.01025876, + "balance_loss_clip": 1.04594457, + "balance_loss_mlp": 1.01765013, + "epoch": 0.7792941742319486, + "flos": 14679276975360.0, + "grad_norm": 2.2724110218779083, + "language_loss": 0.72236812, + "learning_rate": 4.894287173831506e-07, + "loss": 0.74431372, + "num_input_tokens_seen": 139599735, + "step": 6481, + "time_per_iteration": 2.6329939365386963 + }, + { + "auxiliary_loss_clip": 0.01150998, + "auxiliary_loss_mlp": 0.01024718, + "balance_loss_clip": 1.04278147, + "balance_loss_mlp": 1.01682651, + "epoch": 0.7794144171225876, + "flos": 23258874908160.0, + "grad_norm": 2.0334338901465547, + "language_loss": 0.8413251, + "learning_rate": 4.889182972587877e-07, + "loss": 0.86308229, + "num_input_tokens_seen": 139619030, + "step": 6482, + "time_per_iteration": 2.756432056427002 + }, + { + "auxiliary_loss_clip": 0.01151809, + "auxiliary_loss_mlp": 0.01030228, + "balance_loss_clip": 1.04662716, + "balance_loss_mlp": 1.02257454, + "epoch": 0.7795346600132267, + "flos": 21507080613120.0, + "grad_norm": 1.8989579755511474, + "language_loss": 0.66543359, + "learning_rate": 4.884081063643177e-07, + "loss": 0.68725395, + "num_input_tokens_seen": 139637690, + "step": 6483, + "time_per_iteration": 2.7521755695343018 + }, + { + "auxiliary_loss_clip": 0.01041806, + "auxiliary_loss_mlp": 0.01000238, + "balance_loss_clip": 1.01064169, + "balance_loss_mlp": 0.99881929, + "epoch": 0.7796549029038659, + "flos": 70052273694720.0, + "grad_norm": 0.8725554631078318, + "language_loss": 0.52512074, + "learning_rate": 4.878981447771353e-07, + "loss": 0.54554117, + "num_input_tokens_seen": 139692070, + "step": 6484, + "time_per_iteration": 3.2995660305023193 + }, + { + "auxiliary_loss_clip": 0.01137684, + "auxiliary_loss_mlp": 0.01025317, + "balance_loss_clip": 1.04456973, + "balance_loss_mlp": 1.01697826, + "epoch": 0.7797751457945049, + "flos": 23989405714560.0, + "grad_norm": 1.5570419353972071, + "language_loss": 0.72865021, + "learning_rate": 4.873884125746035e-07, + "loss": 0.75028026, + "num_input_tokens_seen": 139713745, + "step": 6485, + "time_per_iteration": 3.6949124336242676 + }, + { + "auxiliary_loss_clip": 0.01144788, + "auxiliary_loss_mlp": 0.01026807, + "balance_loss_clip": 1.04304099, + "balance_loss_mlp": 1.0190649, + "epoch": 0.779895388685144, + "flos": 22674751937280.0, + "grad_norm": 2.330050023520517, + "language_loss": 0.72167379, + "learning_rate": 4.868789098340456e-07, + "loss": 0.74338973, + "num_input_tokens_seen": 139731650, + "step": 6486, + "time_per_iteration": 2.6859097480773926 + }, + { + "auxiliary_loss_clip": 0.011406, + "auxiliary_loss_mlp": 0.01032245, + "balance_loss_clip": 1.04295599, + "balance_loss_mlp": 1.02432323, + "epoch": 0.7800156315757831, + "flos": 23768698596480.0, + "grad_norm": 3.913673174920504, + "language_loss": 0.73099113, + "learning_rate": 4.863696366327543e-07, + "loss": 0.75271952, + "num_input_tokens_seen": 139750820, + "step": 6487, + "time_per_iteration": 2.803351402282715 + }, + { + "auxiliary_loss_clip": 0.01162026, + "auxiliary_loss_mlp": 0.01027618, + "balance_loss_clip": 1.04516792, + "balance_loss_mlp": 1.01955962, + "epoch": 0.7801358744664222, + "flos": 26429714881920.0, + "grad_norm": 2.052725433746748, + "language_loss": 0.77906317, + "learning_rate": 4.85860593047986e-07, + "loss": 0.80095959, + "num_input_tokens_seen": 139770885, + "step": 6488, + "time_per_iteration": 2.76714825630188 + }, + { + "auxiliary_loss_clip": 0.01133924, + "auxiliary_loss_mlp": 0.01031353, + "balance_loss_clip": 1.03840303, + "balance_loss_mlp": 1.02347898, + "epoch": 0.7802561173570612, + "flos": 26322162583680.0, + "grad_norm": 1.519776049239545, + "language_loss": 0.7467494, + "learning_rate": 4.853517791569613e-07, + "loss": 0.76840222, + "num_input_tokens_seen": 139793065, + "step": 6489, + "time_per_iteration": 2.8047585487365723 + }, + { + "auxiliary_loss_clip": 0.01154104, + "auxiliary_loss_mlp": 0.00886648, + "balance_loss_clip": 1.04313779, + "balance_loss_mlp": 1.00061011, + "epoch": 0.7803763602477004, + "flos": 40333751596800.0, + "grad_norm": 1.7585469575886203, + "language_loss": 0.66093194, + "learning_rate": 4.848431950368684e-07, + "loss": 0.68133944, + "num_input_tokens_seen": 139815625, + "step": 6490, + "time_per_iteration": 2.913753032684326 + }, + { + "auxiliary_loss_clip": 0.01058037, + "auxiliary_loss_mlp": 0.00876977, + "balance_loss_clip": 1.00636411, + "balance_loss_mlp": 1.00141573, + "epoch": 0.7804966031383395, + "flos": 67001448038400.0, + "grad_norm": 0.7089998966599189, + "language_loss": 0.55717212, + "learning_rate": 4.843348407648569e-07, + "loss": 0.57652223, + "num_input_tokens_seen": 139876905, + "step": 6491, + "time_per_iteration": 3.121347188949585 + }, + { + "auxiliary_loss_clip": 0.01160703, + "auxiliary_loss_mlp": 0.01025382, + "balance_loss_clip": 1.04204774, + "balance_loss_mlp": 1.01694798, + "epoch": 0.7806168460289785, + "flos": 17740733057280.0, + "grad_norm": 3.9611279707656157, + "language_loss": 0.83441806, + "learning_rate": 4.838267164180457e-07, + "loss": 0.8562789, + "num_input_tokens_seen": 139892575, + "step": 6492, + "time_per_iteration": 2.7504749298095703 + }, + { + "auxiliary_loss_clip": 0.0117128, + "auxiliary_loss_mlp": 0.01035511, + "balance_loss_clip": 1.04712188, + "balance_loss_mlp": 1.02732754, + "epoch": 0.7807370889196176, + "flos": 23946240545280.0, + "grad_norm": 3.300111926566273, + "language_loss": 0.8410725, + "learning_rate": 4.833188220735156e-07, + "loss": 0.86314046, + "num_input_tokens_seen": 139912245, + "step": 6493, + "time_per_iteration": 2.616575002670288 + }, + { + "auxiliary_loss_clip": 0.01159219, + "auxiliary_loss_mlp": 0.01035246, + "balance_loss_clip": 1.04605579, + "balance_loss_mlp": 1.02744365, + "epoch": 0.7808573318102567, + "flos": 18989024457600.0, + "grad_norm": 2.0075336499621996, + "language_loss": 0.74427313, + "learning_rate": 4.828111578083152e-07, + "loss": 0.76621771, + "num_input_tokens_seen": 139929150, + "step": 6494, + "time_per_iteration": 2.6680033206939697 + }, + { + "auxiliary_loss_clip": 0.0114959, + "auxiliary_loss_mlp": 0.01025467, + "balance_loss_clip": 1.04623008, + "balance_loss_mlp": 1.01768851, + "epoch": 0.7809775747008958, + "flos": 23980750536960.0, + "grad_norm": 1.9768888494553385, + "language_loss": 0.81173301, + "learning_rate": 4.823037236994556e-07, + "loss": 0.83348358, + "num_input_tokens_seen": 139947315, + "step": 6495, + "time_per_iteration": 2.7368898391723633 + }, + { + "auxiliary_loss_clip": 0.01053658, + "auxiliary_loss_mlp": 0.01001734, + "balance_loss_clip": 1.00653887, + "balance_loss_mlp": 1.0002383, + "epoch": 0.7810978175915348, + "flos": 68535875180160.0, + "grad_norm": 0.71980858768118, + "language_loss": 0.56369364, + "learning_rate": 4.817965198239136e-07, + "loss": 0.58424759, + "num_input_tokens_seen": 140013775, + "step": 6496, + "time_per_iteration": 3.244133949279785 + }, + { + "auxiliary_loss_clip": 0.0113923, + "auxiliary_loss_mlp": 0.01027667, + "balance_loss_clip": 1.04113972, + "balance_loss_mlp": 1.01984668, + "epoch": 0.781218060482174, + "flos": 19642131498240.0, + "grad_norm": 1.9465006089615444, + "language_loss": 0.74278533, + "learning_rate": 4.812895462586331e-07, + "loss": 0.76445425, + "num_input_tokens_seen": 140031600, + "step": 6497, + "time_per_iteration": 2.741476535797119 + }, + { + "auxiliary_loss_clip": 0.01144864, + "auxiliary_loss_mlp": 0.01027355, + "balance_loss_clip": 1.04454684, + "balance_loss_mlp": 1.01996362, + "epoch": 0.7813383033728131, + "flos": 25627865621760.0, + "grad_norm": 1.7009745854791176, + "language_loss": 0.81594348, + "learning_rate": 4.807828030805207e-07, + "loss": 0.83766568, + "num_input_tokens_seen": 140050590, + "step": 6498, + "time_per_iteration": 2.7854177951812744 + }, + { + "auxiliary_loss_clip": 0.01156241, + "auxiliary_loss_mlp": 0.0102359, + "balance_loss_clip": 1.04523015, + "balance_loss_mlp": 1.0155437, + "epoch": 0.7814585462634521, + "flos": 20485924865280.0, + "grad_norm": 2.1303915666116353, + "language_loss": 0.67915142, + "learning_rate": 4.802762903664495e-07, + "loss": 0.70094979, + "num_input_tokens_seen": 140069770, + "step": 6499, + "time_per_iteration": 2.7148728370666504 + }, + { + "auxiliary_loss_clip": 0.01156732, + "auxiliary_loss_mlp": 0.01030336, + "balance_loss_clip": 1.04681826, + "balance_loss_mlp": 1.02196753, + "epoch": 0.7815787891540913, + "flos": 22304297018880.0, + "grad_norm": 2.053016660719519, + "language_loss": 0.73928177, + "learning_rate": 4.797700081932565e-07, + "loss": 0.76115251, + "num_input_tokens_seen": 140087635, + "step": 6500, + "time_per_iteration": 2.683384418487549 + }, + { + "auxiliary_loss_clip": 0.01116835, + "auxiliary_loss_mlp": 0.01032583, + "balance_loss_clip": 1.03803682, + "balance_loss_mlp": 1.02464986, + "epoch": 0.7816990320447303, + "flos": 22600668136320.0, + "grad_norm": 2.4828961366425286, + "language_loss": 0.81833494, + "learning_rate": 4.792639566377442e-07, + "loss": 0.83982921, + "num_input_tokens_seen": 140105045, + "step": 6501, + "time_per_iteration": 2.8739840984344482 + }, + { + "auxiliary_loss_clip": 0.01152832, + "auxiliary_loss_mlp": 0.01028293, + "balance_loss_clip": 1.04309702, + "balance_loss_mlp": 1.02001357, + "epoch": 0.7818192749353694, + "flos": 24935974871040.0, + "grad_norm": 1.827169696856918, + "language_loss": 0.77452016, + "learning_rate": 4.78758135776681e-07, + "loss": 0.79633141, + "num_input_tokens_seen": 140124900, + "step": 6502, + "time_per_iteration": 2.66874361038208 + }, + { + "auxiliary_loss_clip": 0.01151284, + "auxiliary_loss_mlp": 0.01024468, + "balance_loss_clip": 1.0457263, + "balance_loss_mlp": 1.01685095, + "epoch": 0.7819395178260086, + "flos": 23733039369600.0, + "grad_norm": 1.9675440493991794, + "language_loss": 0.78358054, + "learning_rate": 4.782525456867989e-07, + "loss": 0.80533803, + "num_input_tokens_seen": 140143755, + "step": 6503, + "time_per_iteration": 2.7275874614715576 + }, + { + "auxiliary_loss_clip": 0.01137849, + "auxiliary_loss_mlp": 0.01035116, + "balance_loss_clip": 1.04458618, + "balance_loss_mlp": 1.02712309, + "epoch": 0.7820597607166476, + "flos": 23221671396480.0, + "grad_norm": 1.5303442515946424, + "language_loss": 0.83164513, + "learning_rate": 4.777471864447959e-07, + "loss": 0.85337472, + "num_input_tokens_seen": 140164495, + "step": 6504, + "time_per_iteration": 3.7669291496276855 + }, + { + "auxiliary_loss_clip": 0.01151196, + "auxiliary_loss_mlp": 0.0102887, + "balance_loss_clip": 1.04189718, + "balance_loss_mlp": 1.02056086, + "epoch": 0.7821800036072867, + "flos": 22309540404480.0, + "grad_norm": 1.8627995390724206, + "language_loss": 0.80770779, + "learning_rate": 4.772420581273344e-07, + "loss": 0.82950842, + "num_input_tokens_seen": 140181980, + "step": 6505, + "time_per_iteration": 2.7461698055267334 + }, + { + "auxiliary_loss_clip": 0.01156971, + "auxiliary_loss_mlp": 0.01024383, + "balance_loss_clip": 1.04522395, + "balance_loss_mlp": 1.01708698, + "epoch": 0.7823002464979258, + "flos": 21544176384000.0, + "grad_norm": 2.1718539616049632, + "language_loss": 0.76112759, + "learning_rate": 4.7673716081104134e-07, + "loss": 0.7829411, + "num_input_tokens_seen": 140202155, + "step": 6506, + "time_per_iteration": 4.465746879577637 + }, + { + "auxiliary_loss_clip": 0.01159462, + "auxiliary_loss_mlp": 0.01034894, + "balance_loss_clip": 1.0486778, + "balance_loss_mlp": 1.02743745, + "epoch": 0.7824204893885649, + "flos": 24535642815360.0, + "grad_norm": 1.792382892492564, + "language_loss": 0.8468411, + "learning_rate": 4.762324945725109e-07, + "loss": 0.86878467, + "num_input_tokens_seen": 140221600, + "step": 6507, + "time_per_iteration": 2.87489652633667 + }, + { + "auxiliary_loss_clip": 0.01150192, + "auxiliary_loss_mlp": 0.01027104, + "balance_loss_clip": 1.04723895, + "balance_loss_mlp": 1.02021933, + "epoch": 0.782540732279204, + "flos": 27415211402880.0, + "grad_norm": 1.7116016837140813, + "language_loss": 0.75969738, + "learning_rate": 4.7572805948829844e-07, + "loss": 0.78147042, + "num_input_tokens_seen": 140241860, + "step": 6508, + "time_per_iteration": 2.803004264831543 + }, + { + "auxiliary_loss_clip": 0.01135653, + "auxiliary_loss_mlp": 0.01031514, + "balance_loss_clip": 1.04238582, + "balance_loss_mlp": 1.02396202, + "epoch": 0.7826609751698431, + "flos": 24353216616960.0, + "grad_norm": 2.3149512199325, + "language_loss": 0.7088055, + "learning_rate": 4.7522385563492795e-07, + "loss": 0.73047721, + "num_input_tokens_seen": 140262160, + "step": 6509, + "time_per_iteration": 2.8644535541534424 + }, + { + "auxiliary_loss_clip": 0.01146919, + "auxiliary_loss_mlp": 0.01029313, + "balance_loss_clip": 1.046978, + "balance_loss_mlp": 1.02177894, + "epoch": 0.7827812180604822, + "flos": 23988543788160.0, + "grad_norm": 1.7402778083278394, + "language_loss": 0.70361614, + "learning_rate": 4.747198830888863e-07, + "loss": 0.72537851, + "num_input_tokens_seen": 140282030, + "step": 6510, + "time_per_iteration": 2.7618069648742676 + }, + { + "auxiliary_loss_clip": 0.01145481, + "auxiliary_loss_mlp": 0.0102744, + "balance_loss_clip": 1.04350936, + "balance_loss_mlp": 1.01984, + "epoch": 0.7829014609511212, + "flos": 27454318335360.0, + "grad_norm": 9.266087758038475, + "language_loss": 0.68584573, + "learning_rate": 4.742161419266251e-07, + "loss": 0.70757496, + "num_input_tokens_seen": 140301190, + "step": 6511, + "time_per_iteration": 3.7411410808563232 + }, + { + "auxiliary_loss_clip": 0.01164549, + "auxiliary_loss_mlp": 0.01030143, + "balance_loss_clip": 1.0468998, + "balance_loss_mlp": 1.021119, + "epoch": 0.7830217038417604, + "flos": 29204532432000.0, + "grad_norm": 2.957691129602829, + "language_loss": 0.65037036, + "learning_rate": 4.7371263222456304e-07, + "loss": 0.67231727, + "num_input_tokens_seen": 140318510, + "step": 6512, + "time_per_iteration": 2.7280287742614746 + }, + { + "auxiliary_loss_clip": 0.0104997, + "auxiliary_loss_mlp": 0.01001443, + "balance_loss_clip": 1.0068773, + "balance_loss_mlp": 1.0000782, + "epoch": 0.7831419467323995, + "flos": 60950895822720.0, + "grad_norm": 0.8145525198671836, + "language_loss": 0.61350834, + "learning_rate": 4.7320935405908004e-07, + "loss": 0.63402247, + "num_input_tokens_seen": 140379380, + "step": 6513, + "time_per_iteration": 3.1840932369232178 + }, + { + "auxiliary_loss_clip": 0.01172137, + "auxiliary_loss_mlp": 0.01027148, + "balance_loss_clip": 1.04656863, + "balance_loss_mlp": 1.01861286, + "epoch": 0.7832621896230385, + "flos": 19682531320320.0, + "grad_norm": 2.3337487377361796, + "language_loss": 0.84008545, + "learning_rate": 4.7270630750652475e-07, + "loss": 0.86207831, + "num_input_tokens_seen": 140395335, + "step": 6514, + "time_per_iteration": 2.6630783081054688 + }, + { + "auxiliary_loss_clip": 0.01156963, + "auxiliary_loss_mlp": 0.0103374, + "balance_loss_clip": 1.04464138, + "balance_loss_mlp": 1.02577972, + "epoch": 0.7833824325136777, + "flos": 25009232659200.0, + "grad_norm": 1.680813148506717, + "language_loss": 0.80220723, + "learning_rate": 4.7220349264320746e-07, + "loss": 0.8241142, + "num_input_tokens_seen": 140414420, + "step": 6515, + "time_per_iteration": 2.724604606628418 + }, + { + "auxiliary_loss_clip": 0.0105258, + "auxiliary_loss_mlp": 0.01001427, + "balance_loss_clip": 1.00692666, + "balance_loss_mlp": 0.99999678, + "epoch": 0.7835026754043167, + "flos": 68800142517120.0, + "grad_norm": 0.7374232680680444, + "language_loss": 0.54948539, + "learning_rate": 4.71700909545407e-07, + "loss": 0.5700255, + "num_input_tokens_seen": 140477365, + "step": 6516, + "time_per_iteration": 3.2637805938720703 + }, + { + "auxiliary_loss_clip": 0.01162809, + "auxiliary_loss_mlp": 0.01028824, + "balance_loss_clip": 1.04558635, + "balance_loss_mlp": 1.02101874, + "epoch": 0.7836229182949558, + "flos": 19864598382720.0, + "grad_norm": 1.9710230113050136, + "language_loss": 0.77040613, + "learning_rate": 4.711985582893627e-07, + "loss": 0.7923224, + "num_input_tokens_seen": 140495885, + "step": 6517, + "time_per_iteration": 2.687347650527954 + }, + { + "auxiliary_loss_clip": 0.01131534, + "auxiliary_loss_mlp": 0.01030692, + "balance_loss_clip": 1.03983808, + "balance_loss_mlp": 1.02229357, + "epoch": 0.783743161185595, + "flos": 22965843755520.0, + "grad_norm": 1.994751817999139, + "language_loss": 0.71738118, + "learning_rate": 4.706964389512811e-07, + "loss": 0.73900342, + "num_input_tokens_seen": 140515920, + "step": 6518, + "time_per_iteration": 2.836113452911377 + }, + { + "auxiliary_loss_clip": 0.01166617, + "auxiliary_loss_mlp": 0.01030852, + "balance_loss_clip": 1.04595566, + "balance_loss_mlp": 1.02344298, + "epoch": 0.783863404076234, + "flos": 12458489777280.0, + "grad_norm": 1.9397136158028738, + "language_loss": 0.87540483, + "learning_rate": 4.701945516073345e-07, + "loss": 0.89737958, + "num_input_tokens_seen": 140533395, + "step": 6519, + "time_per_iteration": 2.7481369972229004 + }, + { + "auxiliary_loss_clip": 0.01139308, + "auxiliary_loss_mlp": 0.01027282, + "balance_loss_clip": 1.04470873, + "balance_loss_mlp": 1.01897955, + "epoch": 0.7839836469668731, + "flos": 24243940465920.0, + "grad_norm": 1.779253250025266, + "language_loss": 0.75036716, + "learning_rate": 4.696928963336577e-07, + "loss": 0.77203304, + "num_input_tokens_seen": 140552825, + "step": 6520, + "time_per_iteration": 2.816253662109375 + }, + { + "auxiliary_loss_clip": 0.01049435, + "auxiliary_loss_mlp": 0.01001847, + "balance_loss_clip": 1.00658512, + "balance_loss_mlp": 1.00037479, + "epoch": 0.7841038898575122, + "flos": 62121978938880.0, + "grad_norm": 0.8507488092344131, + "language_loss": 0.61024415, + "learning_rate": 4.6919147320635224e-07, + "loss": 0.63075697, + "num_input_tokens_seen": 140615535, + "step": 6521, + "time_per_iteration": 3.2010538578033447 + }, + { + "auxiliary_loss_clip": 0.01162585, + "auxiliary_loss_mlp": 0.01029438, + "balance_loss_clip": 1.04587412, + "balance_loss_mlp": 1.02136755, + "epoch": 0.7842241327481513, + "flos": 20193899293440.0, + "grad_norm": 2.266461243612319, + "language_loss": 0.73154795, + "learning_rate": 4.6869028230148286e-07, + "loss": 0.75346822, + "num_input_tokens_seen": 140633330, + "step": 6522, + "time_per_iteration": 2.636672019958496 + }, + { + "auxiliary_loss_clip": 0.01133113, + "auxiliary_loss_mlp": 0.01023785, + "balance_loss_clip": 1.03918028, + "balance_loss_mlp": 1.01585746, + "epoch": 0.7843443756387903, + "flos": 28074531496320.0, + "grad_norm": 2.1206210032443717, + "language_loss": 0.59402442, + "learning_rate": 4.6818932369507957e-07, + "loss": 0.61559343, + "num_input_tokens_seen": 140652830, + "step": 6523, + "time_per_iteration": 2.815065622329712 + }, + { + "auxiliary_loss_clip": 0.01161032, + "auxiliary_loss_mlp": 0.01028124, + "balance_loss_clip": 1.04786718, + "balance_loss_mlp": 1.02039373, + "epoch": 0.7844646185294295, + "flos": 21323397438720.0, + "grad_norm": 3.309701097133162, + "language_loss": 0.89210057, + "learning_rate": 4.676885974631386e-07, + "loss": 0.91399217, + "num_input_tokens_seen": 140671190, + "step": 6524, + "time_per_iteration": 2.739379644393921 + }, + { + "auxiliary_loss_clip": 0.01162279, + "auxiliary_loss_mlp": 0.0102816, + "balance_loss_clip": 1.0480181, + "balance_loss_mlp": 1.019768, + "epoch": 0.7845848614200686, + "flos": 23656585271040.0, + "grad_norm": 2.2526672602376077, + "language_loss": 0.81206894, + "learning_rate": 4.67188103681619e-07, + "loss": 0.83397335, + "num_input_tokens_seen": 140690975, + "step": 6525, + "time_per_iteration": 2.7316343784332275 + }, + { + "auxiliary_loss_clip": 0.01155074, + "auxiliary_loss_mlp": 0.00885694, + "balance_loss_clip": 1.04620755, + "balance_loss_mlp": 1.0006485, + "epoch": 0.7847051043107076, + "flos": 23402194174080.0, + "grad_norm": 2.401929834956108, + "language_loss": 0.69690144, + "learning_rate": 4.666878424264453e-07, + "loss": 0.71730918, + "num_input_tokens_seen": 140710930, + "step": 6526, + "time_per_iteration": 2.7225821018218994 + }, + { + "auxiliary_loss_clip": 0.01140685, + "auxiliary_loss_mlp": 0.01023115, + "balance_loss_clip": 1.04237604, + "balance_loss_mlp": 1.01604629, + "epoch": 0.7848253472013467, + "flos": 19022277473280.0, + "grad_norm": 1.5440427521548183, + "language_loss": 0.73848629, + "learning_rate": 4.661878137735069e-07, + "loss": 0.76012433, + "num_input_tokens_seen": 140729120, + "step": 6527, + "time_per_iteration": 2.6701152324676514 + }, + { + "auxiliary_loss_clip": 0.01149007, + "auxiliary_loss_mlp": 0.01035561, + "balance_loss_clip": 1.04480958, + "balance_loss_mlp": 1.02799153, + "epoch": 0.7849455900919858, + "flos": 21179180332800.0, + "grad_norm": 1.7004266871509002, + "language_loss": 0.7438848, + "learning_rate": 4.656880177986571e-07, + "loss": 0.7657305, + "num_input_tokens_seen": 140747665, + "step": 6528, + "time_per_iteration": 2.7003111839294434 + }, + { + "auxiliary_loss_clip": 0.01154005, + "auxiliary_loss_mlp": 0.01028595, + "balance_loss_clip": 1.04401159, + "balance_loss_mlp": 1.02029216, + "epoch": 0.7850658329826249, + "flos": 19536482620800.0, + "grad_norm": 1.8773066771546876, + "language_loss": 0.81496048, + "learning_rate": 4.6518845457771607e-07, + "loss": 0.83678651, + "num_input_tokens_seen": 140766525, + "step": 6529, + "time_per_iteration": 3.7299752235412598 + }, + { + "auxiliary_loss_clip": 0.01154997, + "auxiliary_loss_mlp": 0.00886155, + "balance_loss_clip": 1.04584205, + "balance_loss_mlp": 1.00056934, + "epoch": 0.7851860758732639, + "flos": 12495334152960.0, + "grad_norm": 1.837618786385187, + "language_loss": 0.78732991, + "learning_rate": 4.646891241864652e-07, + "loss": 0.8077414, + "num_input_tokens_seen": 140785090, + "step": 6530, + "time_per_iteration": 2.6628363132476807 + }, + { + "auxiliary_loss_clip": 0.01157765, + "auxiliary_loss_mlp": 0.01029354, + "balance_loss_clip": 1.04399884, + "balance_loss_mlp": 1.02027035, + "epoch": 0.7853063187639031, + "flos": 22960959505920.0, + "grad_norm": 2.201999195213768, + "language_loss": 0.73217905, + "learning_rate": 4.6419002670065397e-07, + "loss": 0.75405025, + "num_input_tokens_seen": 140804670, + "step": 6531, + "time_per_iteration": 3.60634446144104 + }, + { + "auxiliary_loss_clip": 0.01147268, + "auxiliary_loss_mlp": 0.01029449, + "balance_loss_clip": 1.04638577, + "balance_loss_mlp": 1.02093744, + "epoch": 0.7854265616545422, + "flos": 17347260499200.0, + "grad_norm": 2.0577875533211705, + "language_loss": 0.86406487, + "learning_rate": 4.6369116219599445e-07, + "loss": 0.88583201, + "num_input_tokens_seen": 140820655, + "step": 6532, + "time_per_iteration": 3.641718864440918 + }, + { + "auxiliary_loss_clip": 0.01139274, + "auxiliary_loss_mlp": 0.01024252, + "balance_loss_clip": 1.04224563, + "balance_loss_mlp": 1.01666999, + "epoch": 0.7855468045451812, + "flos": 23838293197440.0, + "grad_norm": 1.599209835726505, + "language_loss": 0.79057288, + "learning_rate": 4.631925307481637e-07, + "loss": 0.81220818, + "num_input_tokens_seen": 140840470, + "step": 6533, + "time_per_iteration": 2.7782044410705566 + }, + { + "auxiliary_loss_clip": 0.01151492, + "auxiliary_loss_mlp": 0.01026782, + "balance_loss_clip": 1.04709983, + "balance_loss_mlp": 1.0193975, + "epoch": 0.7856670474358204, + "flos": 25666792986240.0, + "grad_norm": 2.5298465597383144, + "language_loss": 0.75771457, + "learning_rate": 4.6269413243280533e-07, + "loss": 0.77949727, + "num_input_tokens_seen": 140859890, + "step": 6534, + "time_per_iteration": 2.7754967212677 + }, + { + "auxiliary_loss_clip": 0.01155115, + "auxiliary_loss_mlp": 0.01026378, + "balance_loss_clip": 1.04653585, + "balance_loss_mlp": 1.01793766, + "epoch": 0.7857872903264594, + "flos": 18144656472960.0, + "grad_norm": 2.749683677540575, + "language_loss": 0.73716176, + "learning_rate": 4.621959673255236e-07, + "loss": 0.75897664, + "num_input_tokens_seen": 140876190, + "step": 6535, + "time_per_iteration": 2.767388343811035 + }, + { + "auxiliary_loss_clip": 0.01131563, + "auxiliary_loss_mlp": 0.010258, + "balance_loss_clip": 1.04288435, + "balance_loss_mlp": 1.01883495, + "epoch": 0.7859075332170985, + "flos": 14386138081920.0, + "grad_norm": 2.1329534272363353, + "language_loss": 0.90294278, + "learning_rate": 4.6169803550189135e-07, + "loss": 0.92451632, + "num_input_tokens_seen": 140891885, + "step": 6536, + "time_per_iteration": 3.0562121868133545 + }, + { + "auxiliary_loss_clip": 0.01127961, + "auxiliary_loss_mlp": 0.01031168, + "balance_loss_clip": 1.04373789, + "balance_loss_mlp": 1.02295411, + "epoch": 0.7860277761077377, + "flos": 19864059678720.0, + "grad_norm": 2.1089080397114937, + "language_loss": 0.77228987, + "learning_rate": 4.6120033703744355e-07, + "loss": 0.79388106, + "num_input_tokens_seen": 140910780, + "step": 6537, + "time_per_iteration": 3.8124351501464844 + }, + { + "auxiliary_loss_clip": 0.01141739, + "auxiliary_loss_mlp": 0.01031155, + "balance_loss_clip": 1.04229212, + "balance_loss_mlp": 1.02309084, + "epoch": 0.7861480189983767, + "flos": 26396174557440.0, + "grad_norm": 1.7811837686263807, + "language_loss": 0.78430933, + "learning_rate": 4.607028720076822e-07, + "loss": 0.80603826, + "num_input_tokens_seen": 140927460, + "step": 6538, + "time_per_iteration": 2.714449405670166 + }, + { + "auxiliary_loss_clip": 0.01158389, + "auxiliary_loss_mlp": 0.01027892, + "balance_loss_clip": 1.04533911, + "balance_loss_mlp": 1.01975572, + "epoch": 0.7862682618890158, + "flos": 24236578177920.0, + "grad_norm": 2.7978199492523252, + "language_loss": 0.73227692, + "learning_rate": 4.6020564048807074e-07, + "loss": 0.75413966, + "num_input_tokens_seen": 140945135, + "step": 6539, + "time_per_iteration": 2.7175545692443848 + }, + { + "auxiliary_loss_clip": 0.0116039, + "auxiliary_loss_mlp": 0.01031536, + "balance_loss_clip": 1.04633951, + "balance_loss_mlp": 1.02342427, + "epoch": 0.7863885047796549, + "flos": 47551508259840.0, + "grad_norm": 2.129559511759809, + "language_loss": 0.71797329, + "learning_rate": 4.5970864255403883e-07, + "loss": 0.73989254, + "num_input_tokens_seen": 140966660, + "step": 6540, + "time_per_iteration": 2.875267744064331 + }, + { + "auxiliary_loss_clip": 0.01149567, + "auxiliary_loss_mlp": 0.01028996, + "balance_loss_clip": 1.04361916, + "balance_loss_mlp": 1.02193904, + "epoch": 0.786508747670294, + "flos": 24389234979840.0, + "grad_norm": 2.016720093358715, + "language_loss": 0.82476491, + "learning_rate": 4.59211878280982e-07, + "loss": 0.84655058, + "num_input_tokens_seen": 140986175, + "step": 6541, + "time_per_iteration": 2.6767475605010986 + }, + { + "auxiliary_loss_clip": 0.01150807, + "auxiliary_loss_mlp": 0.01026056, + "balance_loss_clip": 1.04434979, + "balance_loss_mlp": 1.01793182, + "epoch": 0.786628990560933, + "flos": 18041234238720.0, + "grad_norm": 4.492612516239852, + "language_loss": 0.69949353, + "learning_rate": 4.587153477442578e-07, + "loss": 0.72126216, + "num_input_tokens_seen": 141002490, + "step": 6542, + "time_per_iteration": 2.8438868522644043 + }, + { + "auxiliary_loss_clip": 0.0117426, + "auxiliary_loss_mlp": 0.01027022, + "balance_loss_clip": 1.04868793, + "balance_loss_mlp": 1.01876092, + "epoch": 0.7867492334515722, + "flos": 25848860048640.0, + "grad_norm": 3.728374419756463, + "language_loss": 0.81591249, + "learning_rate": 4.582190510191899e-07, + "loss": 0.83792531, + "num_input_tokens_seen": 141021150, + "step": 6543, + "time_per_iteration": 2.7489242553710938 + }, + { + "auxiliary_loss_clip": 0.01138382, + "auxiliary_loss_mlp": 0.01024158, + "balance_loss_clip": 1.04443181, + "balance_loss_mlp": 1.0159086, + "epoch": 0.7868694763422113, + "flos": 16580819070720.0, + "grad_norm": 2.281306109522994, + "language_loss": 0.87208652, + "learning_rate": 4.5772298818106625e-07, + "loss": 0.89371192, + "num_input_tokens_seen": 141036940, + "step": 6544, + "time_per_iteration": 2.730653762817383 + }, + { + "auxiliary_loss_clip": 0.01141139, + "auxiliary_loss_mlp": 0.01025827, + "balance_loss_clip": 1.04686213, + "balance_loss_mlp": 1.01751769, + "epoch": 0.7869897192328503, + "flos": 29386276272000.0, + "grad_norm": 2.7623272079635424, + "language_loss": 0.71673501, + "learning_rate": 4.572271593051384e-07, + "loss": 0.73840463, + "num_input_tokens_seen": 141054295, + "step": 6545, + "time_per_iteration": 2.8048624992370605 + }, + { + "auxiliary_loss_clip": 0.0112644, + "auxiliary_loss_mlp": 0.01034071, + "balance_loss_clip": 1.04440165, + "balance_loss_mlp": 1.0256846, + "epoch": 0.7871099621234895, + "flos": 17128923678720.0, + "grad_norm": 1.785143684622032, + "language_loss": 0.78293735, + "learning_rate": 4.567315644666245e-07, + "loss": 0.80454248, + "num_input_tokens_seen": 141073090, + "step": 6546, + "time_per_iteration": 2.7607619762420654 + }, + { + "auxiliary_loss_clip": 0.01134601, + "auxiliary_loss_mlp": 0.01021036, + "balance_loss_clip": 1.04428959, + "balance_loss_mlp": 1.01297712, + "epoch": 0.7872302050141285, + "flos": 23440187784960.0, + "grad_norm": 2.0526558846745506, + "language_loss": 0.84750563, + "learning_rate": 4.5623620374070507e-07, + "loss": 0.86906207, + "num_input_tokens_seen": 141092405, + "step": 6547, + "time_per_iteration": 2.817378282546997 + }, + { + "auxiliary_loss_clip": 0.01040369, + "auxiliary_loss_mlp": 0.01007419, + "balance_loss_clip": 1.00700307, + "balance_loss_mlp": 1.00590503, + "epoch": 0.7873504479047676, + "flos": 65959752689280.0, + "grad_norm": 0.7649063876368674, + "language_loss": 0.58357453, + "learning_rate": 4.557410772025263e-07, + "loss": 0.60405242, + "num_input_tokens_seen": 141154355, + "step": 6548, + "time_per_iteration": 3.424680709838867 + }, + { + "auxiliary_loss_clip": 0.01147293, + "auxiliary_loss_mlp": 0.01031664, + "balance_loss_clip": 1.04278326, + "balance_loss_mlp": 1.0240705, + "epoch": 0.7874706907954068, + "flos": 23258336204160.0, + "grad_norm": 1.8530157341338793, + "language_loss": 0.66254389, + "learning_rate": 4.5524618492719803e-07, + "loss": 0.68433344, + "num_input_tokens_seen": 141173575, + "step": 6549, + "time_per_iteration": 2.7567532062530518 + }, + { + "auxiliary_loss_clip": 0.01161891, + "auxiliary_loss_mlp": 0.01020489, + "balance_loss_clip": 1.0460217, + "balance_loss_mlp": 1.01256764, + "epoch": 0.7875909336860458, + "flos": 28767786963840.0, + "grad_norm": 1.4910469760272431, + "language_loss": 0.78871191, + "learning_rate": 4.54751526989795e-07, + "loss": 0.81053573, + "num_input_tokens_seen": 141195415, + "step": 6550, + "time_per_iteration": 2.77826189994812 + }, + { + "auxiliary_loss_clip": 0.01161937, + "auxiliary_loss_mlp": 0.01027372, + "balance_loss_clip": 1.04552126, + "balance_loss_mlp": 1.01969242, + "epoch": 0.7877111765766849, + "flos": 18697286194560.0, + "grad_norm": 2.033673374836695, + "language_loss": 0.79207206, + "learning_rate": 4.5425710346535775e-07, + "loss": 0.8139652, + "num_input_tokens_seen": 141213360, + "step": 6551, + "time_per_iteration": 2.641226291656494 + }, + { + "auxiliary_loss_clip": 0.0116183, + "auxiliary_loss_mlp": 0.01021394, + "balance_loss_clip": 1.04592538, + "balance_loss_mlp": 1.01349616, + "epoch": 0.787831419467324, + "flos": 27592968833280.0, + "grad_norm": 2.6222107025999395, + "language_loss": 0.81921995, + "learning_rate": 4.537629144288877e-07, + "loss": 0.84105217, + "num_input_tokens_seen": 141230815, + "step": 6552, + "time_per_iteration": 2.733976125717163 + }, + { + "auxiliary_loss_clip": 0.01130262, + "auxiliary_loss_mlp": 0.01023645, + "balance_loss_clip": 1.04030728, + "balance_loss_mlp": 1.01578951, + "epoch": 0.7879516623579631, + "flos": 18150187167360.0, + "grad_norm": 1.9285687098137474, + "language_loss": 0.74941307, + "learning_rate": 4.5326895995535477e-07, + "loss": 0.77095217, + "num_input_tokens_seen": 141249715, + "step": 6553, + "time_per_iteration": 2.778001308441162 + }, + { + "auxiliary_loss_clip": 0.01157335, + "auxiliary_loss_mlp": 0.01028496, + "balance_loss_clip": 1.04512453, + "balance_loss_mlp": 1.02055633, + "epoch": 0.7880719052486022, + "flos": 20339193807360.0, + "grad_norm": 2.1246095213104224, + "language_loss": 0.8410539, + "learning_rate": 4.527752401196907e-07, + "loss": 0.86291218, + "num_input_tokens_seen": 141267730, + "step": 6554, + "time_per_iteration": 2.681755781173706 + }, + { + "auxiliary_loss_clip": 0.01144355, + "auxiliary_loss_mlp": 0.01028229, + "balance_loss_clip": 1.04331994, + "balance_loss_mlp": 1.02033103, + "epoch": 0.7881921481392413, + "flos": 21653237053440.0, + "grad_norm": 12.562485329206964, + "language_loss": 0.66645151, + "learning_rate": 4.5228175499679254e-07, + "loss": 0.68817735, + "num_input_tokens_seen": 141287315, + "step": 6555, + "time_per_iteration": 2.8402767181396484 + }, + { + "auxiliary_loss_clip": 0.01051618, + "auxiliary_loss_mlp": 0.01001397, + "balance_loss_clip": 1.00580966, + "balance_loss_mlp": 1.00001454, + "epoch": 0.7883123910298804, + "flos": 68565860058240.0, + "grad_norm": 0.8290790381959467, + "language_loss": 0.54550982, + "learning_rate": 4.5178850466152174e-07, + "loss": 0.56603992, + "num_input_tokens_seen": 141346145, + "step": 6556, + "time_per_iteration": 4.214515209197998 + }, + { + "auxiliary_loss_clip": 0.01145202, + "auxiliary_loss_mlp": 0.01027896, + "balance_loss_clip": 1.04094875, + "balance_loss_mlp": 1.02060604, + "epoch": 0.7884326339205194, + "flos": 19318217627520.0, + "grad_norm": 1.9972845158455852, + "language_loss": 0.81950104, + "learning_rate": 4.512954891887031e-07, + "loss": 0.841232, + "num_input_tokens_seen": 141364445, + "step": 6557, + "time_per_iteration": 3.7920897006988525 + }, + { + "auxiliary_loss_clip": 0.01142863, + "auxiliary_loss_mlp": 0.0103023, + "balance_loss_clip": 1.04469371, + "balance_loss_mlp": 1.02151608, + "epoch": 0.7885528768111585, + "flos": 17784903807360.0, + "grad_norm": 2.5143274692120214, + "language_loss": 0.8338263, + "learning_rate": 4.5080270865312806e-07, + "loss": 0.8555572, + "num_input_tokens_seen": 141381640, + "step": 6558, + "time_per_iteration": 3.6617894172668457 + }, + { + "auxiliary_loss_clip": 0.01160694, + "auxiliary_loss_mlp": 0.01025563, + "balance_loss_clip": 1.04552841, + "balance_loss_mlp": 1.01802874, + "epoch": 0.7886731197017977, + "flos": 18807639753600.0, + "grad_norm": 3.217666818671749, + "language_loss": 0.71246874, + "learning_rate": 4.5031016312954985e-07, + "loss": 0.73433131, + "num_input_tokens_seen": 141399955, + "step": 6559, + "time_per_iteration": 2.736480712890625 + }, + { + "auxiliary_loss_clip": 0.01170571, + "auxiliary_loss_mlp": 0.0102739, + "balance_loss_clip": 1.05045831, + "balance_loss_mlp": 1.01890182, + "epoch": 0.7887933625924367, + "flos": 33365358126720.0, + "grad_norm": 1.8389480767195068, + "language_loss": 0.74402809, + "learning_rate": 4.498178526926886e-07, + "loss": 0.76600772, + "num_input_tokens_seen": 141420820, + "step": 6560, + "time_per_iteration": 2.7760860919952393 + }, + { + "auxiliary_loss_clip": 0.01168665, + "auxiliary_loss_mlp": 0.0103398, + "balance_loss_clip": 1.0479939, + "balance_loss_mlp": 1.02609396, + "epoch": 0.7889136054830758, + "flos": 17019360218880.0, + "grad_norm": 2.185553551505263, + "language_loss": 0.72508883, + "learning_rate": 4.4932577741722635e-07, + "loss": 0.74711525, + "num_input_tokens_seen": 141439350, + "step": 6561, + "time_per_iteration": 2.6322476863861084 + }, + { + "auxiliary_loss_clip": 0.01147677, + "auxiliary_loss_mlp": 0.01032984, + "balance_loss_clip": 1.04411352, + "balance_loss_mlp": 1.02458, + "epoch": 0.7890338483737149, + "flos": 29424629018880.0, + "grad_norm": 1.6106577352288514, + "language_loss": 0.74382126, + "learning_rate": 4.4883393737780985e-07, + "loss": 0.76562786, + "num_input_tokens_seen": 141460300, + "step": 6562, + "time_per_iteration": 2.7987067699432373 + }, + { + "auxiliary_loss_clip": 0.01154227, + "auxiliary_loss_mlp": 0.01027215, + "balance_loss_clip": 1.04430056, + "balance_loss_mlp": 1.02017546, + "epoch": 0.789154091264354, + "flos": 19971576063360.0, + "grad_norm": 1.8520496789946201, + "language_loss": 0.78334904, + "learning_rate": 4.4834233264905254e-07, + "loss": 0.8051635, + "num_input_tokens_seen": 141477315, + "step": 6563, + "time_per_iteration": 3.575915813446045 + }, + { + "auxiliary_loss_clip": 0.01133195, + "auxiliary_loss_mlp": 0.01028396, + "balance_loss_clip": 1.04022896, + "balance_loss_mlp": 1.01994371, + "epoch": 0.789274334154993, + "flos": 14537825216640.0, + "grad_norm": 2.7659338787568344, + "language_loss": 0.71757424, + "learning_rate": 4.478509633055294e-07, + "loss": 0.73919016, + "num_input_tokens_seen": 141495025, + "step": 6564, + "time_per_iteration": 2.824885845184326 + }, + { + "auxiliary_loss_clip": 0.01151917, + "auxiliary_loss_mlp": 0.01028507, + "balance_loss_clip": 1.04367507, + "balance_loss_mlp": 1.0200907, + "epoch": 0.7893945770456322, + "flos": 21827403123840.0, + "grad_norm": 3.7319013241447303, + "language_loss": 0.80562896, + "learning_rate": 4.473598294217813e-07, + "loss": 0.82743323, + "num_input_tokens_seen": 141510450, + "step": 6565, + "time_per_iteration": 2.773522138595581 + }, + { + "auxiliary_loss_clip": 0.01161491, + "auxiliary_loss_mlp": 0.01033319, + "balance_loss_clip": 1.04866195, + "balance_loss_mlp": 1.02583814, + "epoch": 0.7895148199362713, + "flos": 20740639184640.0, + "grad_norm": 2.5267140734421196, + "language_loss": 0.7128756, + "learning_rate": 4.468689310723124e-07, + "loss": 0.7348237, + "num_input_tokens_seen": 141528265, + "step": 6566, + "time_per_iteration": 2.768284320831299 + }, + { + "auxiliary_loss_clip": 0.0114152, + "auxiliary_loss_mlp": 0.01029265, + "balance_loss_clip": 1.04037976, + "balance_loss_mlp": 1.02113807, + "epoch": 0.7896350628269103, + "flos": 16690669839360.0, + "grad_norm": 2.090551655641089, + "language_loss": 0.78427935, + "learning_rate": 4.463782683315913e-07, + "loss": 0.80598724, + "num_input_tokens_seen": 141547270, + "step": 6567, + "time_per_iteration": 2.770108938217163 + }, + { + "auxiliary_loss_clip": 0.01168602, + "auxiliary_loss_mlp": 0.01026987, + "balance_loss_clip": 1.04712343, + "balance_loss_mlp": 1.01932764, + "epoch": 0.7897553057175495, + "flos": 22638374438400.0, + "grad_norm": 1.7562594293010991, + "language_loss": 0.73384476, + "learning_rate": 4.458878412740523e-07, + "loss": 0.75580066, + "num_input_tokens_seen": 141566050, + "step": 6568, + "time_per_iteration": 2.7388813495635986 + }, + { + "auxiliary_loss_clip": 0.01160319, + "auxiliary_loss_mlp": 0.01022486, + "balance_loss_clip": 1.04761815, + "balance_loss_mlp": 1.01504779, + "epoch": 0.7898755486081885, + "flos": 14537573821440.0, + "grad_norm": 2.754226082651683, + "language_loss": 0.77576828, + "learning_rate": 4.453976499740919e-07, + "loss": 0.79759634, + "num_input_tokens_seen": 141583695, + "step": 6569, + "time_per_iteration": 2.601247787475586 + }, + { + "auxiliary_loss_clip": 0.01160593, + "auxiliary_loss_mlp": 0.01026197, + "balance_loss_clip": 1.04831958, + "balance_loss_mlp": 1.01762581, + "epoch": 0.7899957914988276, + "flos": 17238487138560.0, + "grad_norm": 2.6469152018126274, + "language_loss": 0.78161776, + "learning_rate": 4.4490769450607215e-07, + "loss": 0.80348563, + "num_input_tokens_seen": 141601320, + "step": 6570, + "time_per_iteration": 2.682683229446411 + }, + { + "auxiliary_loss_clip": 0.01136617, + "auxiliary_loss_mlp": 0.01030548, + "balance_loss_clip": 1.03950298, + "balance_loss_mlp": 1.02243328, + "epoch": 0.7901160343894668, + "flos": 41279351086080.0, + "grad_norm": 2.078045912634499, + "language_loss": 0.72578502, + "learning_rate": 4.4441797494431845e-07, + "loss": 0.74745667, + "num_input_tokens_seen": 141623125, + "step": 6571, + "time_per_iteration": 2.9088594913482666 + }, + { + "auxiliary_loss_clip": 0.01158616, + "auxiliary_loss_mlp": 0.01022639, + "balance_loss_clip": 1.04713368, + "balance_loss_mlp": 1.01508749, + "epoch": 0.7902362772801058, + "flos": 16837005847680.0, + "grad_norm": 2.4669506758180937, + "language_loss": 0.77940285, + "learning_rate": 4.439284913631207e-07, + "loss": 0.80121541, + "num_input_tokens_seen": 141640335, + "step": 6572, + "time_per_iteration": 2.7329797744750977 + }, + { + "auxiliary_loss_clip": 0.01136421, + "auxiliary_loss_mlp": 0.0102807, + "balance_loss_clip": 1.04544806, + "balance_loss_mlp": 1.01982653, + "epoch": 0.7903565201707449, + "flos": 27125987091840.0, + "grad_norm": 2.1583404510610658, + "language_loss": 0.83878177, + "learning_rate": 4.434392438367347e-07, + "loss": 0.86042666, + "num_input_tokens_seen": 141659760, + "step": 6573, + "time_per_iteration": 2.8264129161834717 + }, + { + "auxiliary_loss_clip": 0.01156678, + "auxiliary_loss_mlp": 0.01026684, + "balance_loss_clip": 1.04594731, + "balance_loss_mlp": 1.01875651, + "epoch": 0.790476763061384, + "flos": 31025167142400.0, + "grad_norm": 1.728454559188017, + "language_loss": 0.73908627, + "learning_rate": 4.4295023243937677e-07, + "loss": 0.76091993, + "num_input_tokens_seen": 141679965, + "step": 6574, + "time_per_iteration": 2.7356302738189697 + }, + { + "auxiliary_loss_clip": 0.01164717, + "auxiliary_loss_mlp": 0.01030074, + "balance_loss_clip": 1.05037057, + "balance_loss_mlp": 1.02191997, + "epoch": 0.7905970059520231, + "flos": 22089084681600.0, + "grad_norm": 1.8503601651704362, + "language_loss": 0.80168319, + "learning_rate": 4.4246145724523123e-07, + "loss": 0.82363105, + "num_input_tokens_seen": 141697710, + "step": 6575, + "time_per_iteration": 2.659797191619873 + }, + { + "auxiliary_loss_clip": 0.01143163, + "auxiliary_loss_mlp": 0.01028173, + "balance_loss_clip": 1.04513991, + "balance_loss_mlp": 1.02013803, + "epoch": 0.7907172488426621, + "flos": 20558141159040.0, + "grad_norm": 2.3235149583359487, + "language_loss": 0.77265352, + "learning_rate": 4.41972918328444e-07, + "loss": 0.7943669, + "num_input_tokens_seen": 141715145, + "step": 6576, + "time_per_iteration": 2.773097515106201 + }, + { + "auxiliary_loss_clip": 0.01157745, + "auxiliary_loss_mlp": 0.01029854, + "balance_loss_clip": 1.04740357, + "balance_loss_mlp": 1.02131259, + "epoch": 0.7908374917333013, + "flos": 30081542901120.0, + "grad_norm": 3.109149331705783, + "language_loss": 0.77552366, + "learning_rate": 4.4148461576312646e-07, + "loss": 0.79739958, + "num_input_tokens_seen": 141734810, + "step": 6577, + "time_per_iteration": 2.768889904022217 + }, + { + "auxiliary_loss_clip": 0.01160688, + "auxiliary_loss_mlp": 0.0102629, + "balance_loss_clip": 1.04827869, + "balance_loss_mlp": 1.0193994, + "epoch": 0.7909577346239404, + "flos": 20996359084800.0, + "grad_norm": 1.4636605226039612, + "language_loss": 0.74738395, + "learning_rate": 4.4099654962335343e-07, + "loss": 0.76925373, + "num_input_tokens_seen": 141755260, + "step": 6578, + "time_per_iteration": 2.6783342361450195 + }, + { + "auxiliary_loss_clip": 0.01157922, + "auxiliary_loss_mlp": 0.01028828, + "balance_loss_clip": 1.04646301, + "balance_loss_mlp": 1.02096057, + "epoch": 0.7910779775145794, + "flos": 26247935128320.0, + "grad_norm": 1.6862473790412067, + "language_loss": 0.7477712, + "learning_rate": 4.405087199831636e-07, + "loss": 0.76963872, + "num_input_tokens_seen": 141775500, + "step": 6579, + "time_per_iteration": 2.741645574569702 + }, + { + "auxiliary_loss_clip": 0.01152762, + "auxiliary_loss_mlp": 0.00885452, + "balance_loss_clip": 1.04363418, + "balance_loss_mlp": 1.00053144, + "epoch": 0.7911982204052186, + "flos": 22564434291840.0, + "grad_norm": 2.1122067346445506, + "language_loss": 0.67299539, + "learning_rate": 4.400211269165619e-07, + "loss": 0.69337749, + "num_input_tokens_seen": 141791955, + "step": 6580, + "time_per_iteration": 2.785081148147583 + }, + { + "auxiliary_loss_clip": 0.01170698, + "auxiliary_loss_mlp": 0.01024936, + "balance_loss_clip": 1.05033231, + "balance_loss_mlp": 1.01705074, + "epoch": 0.7913184632958576, + "flos": 23112538899840.0, + "grad_norm": 1.5143031621747025, + "language_loss": 0.76590914, + "learning_rate": 4.3953377049751416e-07, + "loss": 0.78786552, + "num_input_tokens_seen": 141812380, + "step": 6581, + "time_per_iteration": 3.5948235988616943 + }, + { + "auxiliary_loss_clip": 0.01155937, + "auxiliary_loss_mlp": 0.01029088, + "balance_loss_clip": 1.04550028, + "balance_loss_mlp": 1.0216608, + "epoch": 0.7914387061864967, + "flos": 12311758719360.0, + "grad_norm": 2.532520882896888, + "language_loss": 0.77997023, + "learning_rate": 4.390466507999537e-07, + "loss": 0.80182052, + "num_input_tokens_seen": 141828130, + "step": 6582, + "time_per_iteration": 2.6658289432525635 + }, + { + "auxiliary_loss_clip": 0.01139194, + "auxiliary_loss_mlp": 0.01028975, + "balance_loss_clip": 1.04202795, + "balance_loss_mlp": 1.0211308, + "epoch": 0.7915589490771359, + "flos": 17603267708160.0, + "grad_norm": 2.4142748971403423, + "language_loss": 0.75807852, + "learning_rate": 4.385597678977748e-07, + "loss": 0.77976024, + "num_input_tokens_seen": 141846965, + "step": 6583, + "time_per_iteration": 3.683518409729004 + }, + { + "auxiliary_loss_clip": 0.01145841, + "auxiliary_loss_mlp": 0.01030954, + "balance_loss_clip": 1.04265702, + "balance_loss_mlp": 1.02292514, + "epoch": 0.7916791919677749, + "flos": 25591272641280.0, + "grad_norm": 1.6060671450815724, + "language_loss": 0.75435376, + "learning_rate": 4.3807312186483726e-07, + "loss": 0.77612174, + "num_input_tokens_seen": 141867685, + "step": 6584, + "time_per_iteration": 3.614445447921753 + }, + { + "auxiliary_loss_clip": 0.0115918, + "auxiliary_loss_mlp": 0.01025754, + "balance_loss_clip": 1.04920077, + "balance_loss_mlp": 1.0178926, + "epoch": 0.791799434858414, + "flos": 18844340474880.0, + "grad_norm": 2.6587289340240146, + "language_loss": 0.78333992, + "learning_rate": 4.375867127749655e-07, + "loss": 0.80518931, + "num_input_tokens_seen": 141885960, + "step": 6585, + "time_per_iteration": 2.659254789352417 + }, + { + "auxiliary_loss_clip": 0.01142869, + "auxiliary_loss_mlp": 0.01037607, + "balance_loss_clip": 1.04658842, + "balance_loss_mlp": 1.02971852, + "epoch": 0.7919196777490531, + "flos": 25812015672960.0, + "grad_norm": 2.1652085682920204, + "language_loss": 0.67343307, + "learning_rate": 4.3710054070194744e-07, + "loss": 0.69523776, + "num_input_tokens_seen": 141905655, + "step": 6586, + "time_per_iteration": 2.8362979888916016 + }, + { + "auxiliary_loss_clip": 0.01170116, + "auxiliary_loss_mlp": 0.00885883, + "balance_loss_clip": 1.0467149, + "balance_loss_mlp": 1.00054693, + "epoch": 0.7920399206396922, + "flos": 11947624594560.0, + "grad_norm": 2.654641671051255, + "language_loss": 0.66112012, + "learning_rate": 4.3661460571953455e-07, + "loss": 0.68168008, + "num_input_tokens_seen": 141922390, + "step": 6587, + "time_per_iteration": 2.588172674179077 + }, + { + "auxiliary_loss_clip": 0.01157722, + "auxiliary_loss_mlp": 0.01026226, + "balance_loss_clip": 1.0418644, + "balance_loss_mlp": 1.01916289, + "epoch": 0.7921601635303313, + "flos": 21579907438080.0, + "grad_norm": 1.6587995062282128, + "language_loss": 0.6833269, + "learning_rate": 4.36128907901443e-07, + "loss": 0.70516634, + "num_input_tokens_seen": 141941985, + "step": 6588, + "time_per_iteration": 2.7841057777404785 + }, + { + "auxiliary_loss_clip": 0.0114309, + "auxiliary_loss_mlp": 0.01024915, + "balance_loss_clip": 1.04271066, + "balance_loss_mlp": 1.01747024, + "epoch": 0.7922804064209703, + "flos": 18113989236480.0, + "grad_norm": 2.0513970512667212, + "language_loss": 0.72996664, + "learning_rate": 4.356434473213519e-07, + "loss": 0.75164664, + "num_input_tokens_seen": 141959435, + "step": 6589, + "time_per_iteration": 2.762722969055176 + }, + { + "auxiliary_loss_clip": 0.01151995, + "auxiliary_loss_mlp": 0.01028617, + "balance_loss_clip": 1.0484066, + "balance_loss_mlp": 1.02103496, + "epoch": 0.7924006493116095, + "flos": 21652806090240.0, + "grad_norm": 1.5716623517392887, + "language_loss": 0.79625231, + "learning_rate": 4.351582240529068e-07, + "loss": 0.81805843, + "num_input_tokens_seen": 141980265, + "step": 6590, + "time_per_iteration": 3.605776786804199 + }, + { + "auxiliary_loss_clip": 0.01049293, + "auxiliary_loss_mlp": 0.01002307, + "balance_loss_clip": 1.0059731, + "balance_loss_mlp": 1.00096023, + "epoch": 0.7925208922022485, + "flos": 64242755694720.0, + "grad_norm": 0.6794406429926948, + "language_loss": 0.58238328, + "learning_rate": 4.346732381697149e-07, + "loss": 0.60289925, + "num_input_tokens_seen": 142044395, + "step": 6591, + "time_per_iteration": 3.308084726333618 + }, + { + "auxiliary_loss_clip": 0.01143297, + "auxiliary_loss_mlp": 0.01028111, + "balance_loss_clip": 1.04519057, + "balance_loss_mlp": 1.0204463, + "epoch": 0.7926411350928876, + "flos": 16941541403520.0, + "grad_norm": 1.9338256241295753, + "language_loss": 0.80851364, + "learning_rate": 4.3418848974534825e-07, + "loss": 0.83022779, + "num_input_tokens_seen": 142061335, + "step": 6592, + "time_per_iteration": 2.732692003250122 + }, + { + "auxiliary_loss_clip": 0.01144669, + "auxiliary_loss_mlp": 0.01032529, + "balance_loss_clip": 1.04394305, + "balance_loss_mlp": 1.02479839, + "epoch": 0.7927613779835267, + "flos": 34459987144320.0, + "grad_norm": 1.916334694775034, + "language_loss": 0.69122684, + "learning_rate": 4.3370397885334276e-07, + "loss": 0.71299887, + "num_input_tokens_seen": 142081965, + "step": 6593, + "time_per_iteration": 2.836304187774658 + }, + { + "auxiliary_loss_clip": 0.01153974, + "auxiliary_loss_mlp": 0.01030044, + "balance_loss_clip": 1.04601908, + "balance_loss_mlp": 1.02217603, + "epoch": 0.7928816208741658, + "flos": 18951174501120.0, + "grad_norm": 1.7667340372445473, + "language_loss": 0.75681466, + "learning_rate": 4.3321970556719777e-07, + "loss": 0.77865481, + "num_input_tokens_seen": 142100260, + "step": 6594, + "time_per_iteration": 2.6971023082733154 + }, + { + "auxiliary_loss_clip": 0.01169873, + "auxiliary_loss_mlp": 0.01034459, + "balance_loss_clip": 1.0480839, + "balance_loss_mlp": 1.02613783, + "epoch": 0.7930018637648049, + "flos": 18623022825600.0, + "grad_norm": 2.559790676953224, + "language_loss": 0.72278917, + "learning_rate": 4.3273566996037856e-07, + "loss": 0.74483246, + "num_input_tokens_seen": 142116955, + "step": 6595, + "time_per_iteration": 2.6179897785186768 + }, + { + "auxiliary_loss_clip": 0.01150399, + "auxiliary_loss_mlp": 0.01024381, + "balance_loss_clip": 1.04491949, + "balance_loss_mlp": 1.01718414, + "epoch": 0.793122106655444, + "flos": 24530650824960.0, + "grad_norm": 3.812541348957293, + "language_loss": 0.80415487, + "learning_rate": 4.322518721063113e-07, + "loss": 0.8259027, + "num_input_tokens_seen": 142135505, + "step": 6596, + "time_per_iteration": 2.8098948001861572 + }, + { + "auxiliary_loss_clip": 0.01162713, + "auxiliary_loss_mlp": 0.01025211, + "balance_loss_clip": 1.04949903, + "balance_loss_mlp": 1.01712251, + "epoch": 0.7932423495460831, + "flos": 34421203434240.0, + "grad_norm": 1.9751710120123696, + "language_loss": 0.70407355, + "learning_rate": 4.3176831207838906e-07, + "loss": 0.7259528, + "num_input_tokens_seen": 142158915, + "step": 6597, + "time_per_iteration": 2.805807113647461 + }, + { + "auxiliary_loss_clip": 0.01161084, + "auxiliary_loss_mlp": 0.01034353, + "balance_loss_clip": 1.04902864, + "balance_loss_mlp": 1.02611613, + "epoch": 0.7933625924367221, + "flos": 26980333441920.0, + "grad_norm": 1.8163251439538013, + "language_loss": 0.74709165, + "learning_rate": 4.3128498994996685e-07, + "loss": 0.76904607, + "num_input_tokens_seen": 142178390, + "step": 6598, + "time_per_iteration": 2.7690622806549072 + }, + { + "auxiliary_loss_clip": 0.01165025, + "auxiliary_loss_mlp": 0.01025946, + "balance_loss_clip": 1.04685116, + "balance_loss_mlp": 1.01749969, + "epoch": 0.7934828353273613, + "flos": 29568630643200.0, + "grad_norm": 2.3815221429952835, + "language_loss": 0.71792686, + "learning_rate": 4.308019057943646e-07, + "loss": 0.73983657, + "num_input_tokens_seen": 142200115, + "step": 6599, + "time_per_iteration": 2.7157952785491943 + }, + { + "auxiliary_loss_clip": 0.01131925, + "auxiliary_loss_mlp": 0.01027743, + "balance_loss_clip": 1.04371881, + "balance_loss_mlp": 1.01987565, + "epoch": 0.7936030782180004, + "flos": 28615381557120.0, + "grad_norm": 1.605886368757025, + "language_loss": 0.74269009, + "learning_rate": 4.3031905968486535e-07, + "loss": 0.76428682, + "num_input_tokens_seen": 142220945, + "step": 6600, + "time_per_iteration": 2.8824353218078613 + }, + { + "auxiliary_loss_clip": 0.01130919, + "auxiliary_loss_mlp": 0.0102913, + "balance_loss_clip": 1.04527092, + "balance_loss_mlp": 1.02142918, + "epoch": 0.7937233211086394, + "flos": 16392574869120.0, + "grad_norm": 1.9955368951365973, + "language_loss": 0.68065888, + "learning_rate": 4.298364516947162e-07, + "loss": 0.70225942, + "num_input_tokens_seen": 142238175, + "step": 6601, + "time_per_iteration": 2.788081645965576 + }, + { + "auxiliary_loss_clip": 0.01127553, + "auxiliary_loss_mlp": 0.01028399, + "balance_loss_clip": 1.04030395, + "balance_loss_mlp": 1.02128243, + "epoch": 0.7938435639992786, + "flos": 22013420682240.0, + "grad_norm": 1.7464996025243058, + "language_loss": 0.65868425, + "learning_rate": 4.293540818971295e-07, + "loss": 0.68024379, + "num_input_tokens_seen": 142255980, + "step": 6602, + "time_per_iteration": 2.8058178424835205 + }, + { + "auxiliary_loss_clip": 0.01164175, + "auxiliary_loss_mlp": 0.01024532, + "balance_loss_clip": 1.04580355, + "balance_loss_mlp": 1.0165869, + "epoch": 0.7939638068899176, + "flos": 22197032029440.0, + "grad_norm": 1.9604094872350033, + "language_loss": 0.76442003, + "learning_rate": 4.2887195036527934e-07, + "loss": 0.7863071, + "num_input_tokens_seen": 142274785, + "step": 6603, + "time_per_iteration": 2.707226514816284 + }, + { + "auxiliary_loss_clip": 0.01152294, + "auxiliary_loss_mlp": 0.01029268, + "balance_loss_clip": 1.04306126, + "balance_loss_mlp": 1.02072024, + "epoch": 0.7940840497805567, + "flos": 17745186343680.0, + "grad_norm": 2.8503313823272722, + "language_loss": 0.73616064, + "learning_rate": 4.28390057172306e-07, + "loss": 0.75797629, + "num_input_tokens_seen": 142291290, + "step": 6604, + "time_per_iteration": 2.615807056427002 + }, + { + "auxiliary_loss_clip": 0.01136472, + "auxiliary_loss_mlp": 0.01034256, + "balance_loss_clip": 1.04116988, + "balance_loss_mlp": 1.02608967, + "epoch": 0.7942042926711959, + "flos": 23805435231360.0, + "grad_norm": 3.3591903318688683, + "language_loss": 0.72328728, + "learning_rate": 4.279084023913111e-07, + "loss": 0.74499452, + "num_input_tokens_seen": 142309165, + "step": 6605, + "time_per_iteration": 2.782057762145996 + }, + { + "auxiliary_loss_clip": 0.01157825, + "auxiliary_loss_mlp": 0.01029117, + "balance_loss_clip": 1.04616261, + "balance_loss_mlp": 1.02152348, + "epoch": 0.7943245355618349, + "flos": 19244959839360.0, + "grad_norm": 1.882176136356108, + "language_loss": 0.69570005, + "learning_rate": 4.2742698609536096e-07, + "loss": 0.71756947, + "num_input_tokens_seen": 142327475, + "step": 6606, + "time_per_iteration": 2.6143879890441895 + }, + { + "auxiliary_loss_clip": 0.01153367, + "auxiliary_loss_mlp": 0.01025364, + "balance_loss_clip": 1.04567671, + "balance_loss_mlp": 1.01747227, + "epoch": 0.794444778452474, + "flos": 25007616547200.0, + "grad_norm": 2.6226309451916276, + "language_loss": 0.7837162, + "learning_rate": 4.2694580835748706e-07, + "loss": 0.80550349, + "num_input_tokens_seen": 142347335, + "step": 6607, + "time_per_iteration": 3.6679017543792725 + }, + { + "auxiliary_loss_clip": 0.01143858, + "auxiliary_loss_mlp": 0.01026208, + "balance_loss_clip": 1.04485154, + "balance_loss_mlp": 1.01859951, + "epoch": 0.7945650213431131, + "flos": 23221491828480.0, + "grad_norm": 1.9633457000400032, + "language_loss": 0.74017262, + "learning_rate": 4.264648692506836e-07, + "loss": 0.7618733, + "num_input_tokens_seen": 142366125, + "step": 6608, + "time_per_iteration": 2.728832244873047 + }, + { + "auxiliary_loss_clip": 0.01148167, + "auxiliary_loss_mlp": 0.01032209, + "balance_loss_clip": 1.04580534, + "balance_loss_mlp": 1.02424598, + "epoch": 0.7946852642337522, + "flos": 26062887237120.0, + "grad_norm": 2.130182236874565, + "language_loss": 0.72101313, + "learning_rate": 4.2598416884790824e-07, + "loss": 0.74281693, + "num_input_tokens_seen": 142385175, + "step": 6609, + "time_per_iteration": 3.690972089767456 + }, + { + "auxiliary_loss_clip": 0.01160487, + "auxiliary_loss_mlp": 0.0102626, + "balance_loss_clip": 1.04387045, + "balance_loss_mlp": 1.01770043, + "epoch": 0.7948055071243912, + "flos": 23769704177280.0, + "grad_norm": 1.9834933397827603, + "language_loss": 0.80774862, + "learning_rate": 4.255037072220828e-07, + "loss": 0.82961607, + "num_input_tokens_seen": 142406545, + "step": 6610, + "time_per_iteration": 3.6575019359588623 + }, + { + "auxiliary_loss_clip": 0.01166553, + "auxiliary_loss_mlp": 0.01027421, + "balance_loss_clip": 1.04720223, + "balance_loss_mlp": 1.01985109, + "epoch": 0.7949257500150304, + "flos": 21980814111360.0, + "grad_norm": 1.7689287320897782, + "language_loss": 0.71713775, + "learning_rate": 4.2502348444609293e-07, + "loss": 0.73907751, + "num_input_tokens_seen": 142426165, + "step": 6611, + "time_per_iteration": 2.6586039066314697 + }, + { + "auxiliary_loss_clip": 0.01127125, + "auxiliary_loss_mlp": 0.01030674, + "balance_loss_clip": 1.03807223, + "balance_loss_mlp": 1.0231936, + "epoch": 0.7950459929056695, + "flos": 25774129802880.0, + "grad_norm": 1.7833141122863574, + "language_loss": 0.69465017, + "learning_rate": 4.2454350059278844e-07, + "loss": 0.71622819, + "num_input_tokens_seen": 142447225, + "step": 6612, + "time_per_iteration": 2.8764100074768066 + }, + { + "auxiliary_loss_clip": 0.01143524, + "auxiliary_loss_mlp": 0.01028635, + "balance_loss_clip": 1.03979063, + "balance_loss_mlp": 1.02062964, + "epoch": 0.7951662357963085, + "flos": 22158068751360.0, + "grad_norm": 1.7344751193981767, + "language_loss": 0.84468269, + "learning_rate": 4.240637557349824e-07, + "loss": 0.86640429, + "num_input_tokens_seen": 142464440, + "step": 6613, + "time_per_iteration": 2.760003089904785 + }, + { + "auxiliary_loss_clip": 0.01140385, + "auxiliary_loss_mlp": 0.01024922, + "balance_loss_clip": 1.04420054, + "balance_loss_mlp": 1.0168519, + "epoch": 0.7952864786869477, + "flos": 24641938137600.0, + "grad_norm": 1.9715283673343527, + "language_loss": 0.66509068, + "learning_rate": 4.235842499454516e-07, + "loss": 0.68674374, + "num_input_tokens_seen": 142484355, + "step": 6614, + "time_per_iteration": 2.7212603092193604 + }, + { + "auxiliary_loss_clip": 0.01152869, + "auxiliary_loss_mlp": 0.01028723, + "balance_loss_clip": 1.0474056, + "balance_loss_mlp": 1.02057457, + "epoch": 0.7954067215775867, + "flos": 21830922656640.0, + "grad_norm": 2.1351519890178783, + "language_loss": 0.82510066, + "learning_rate": 4.2310498329693687e-07, + "loss": 0.84691656, + "num_input_tokens_seen": 142505255, + "step": 6615, + "time_per_iteration": 3.648629665374756 + }, + { + "auxiliary_loss_clip": 0.01161969, + "auxiliary_loss_mlp": 0.01026241, + "balance_loss_clip": 1.04735267, + "balance_loss_mlp": 1.01796198, + "epoch": 0.7955269644682258, + "flos": 24060652341120.0, + "grad_norm": 2.5873087460895934, + "language_loss": 0.80973649, + "learning_rate": 4.2262595586214164e-07, + "loss": 0.83161861, + "num_input_tokens_seen": 142526350, + "step": 6616, + "time_per_iteration": 2.6572296619415283 + }, + { + "auxiliary_loss_clip": 0.01167469, + "auxiliary_loss_mlp": 0.0102555, + "balance_loss_clip": 1.04921734, + "balance_loss_mlp": 1.01741982, + "epoch": 0.795647207358865, + "flos": 25010741030400.0, + "grad_norm": 1.6167158950864782, + "language_loss": 0.76805454, + "learning_rate": 4.221471677137358e-07, + "loss": 0.7899847, + "num_input_tokens_seen": 142547165, + "step": 6617, + "time_per_iteration": 2.6713192462921143 + }, + { + "auxiliary_loss_clip": 0.01139052, + "auxiliary_loss_mlp": 0.01030422, + "balance_loss_clip": 1.04344845, + "balance_loss_mlp": 1.02264929, + "epoch": 0.795767450249504, + "flos": 14648358343680.0, + "grad_norm": 1.7021113911049237, + "language_loss": 0.70523512, + "learning_rate": 4.216686189243492e-07, + "loss": 0.7269299, + "num_input_tokens_seen": 142565955, + "step": 6618, + "time_per_iteration": 2.7049057483673096 + }, + { + "auxiliary_loss_clip": 0.01137106, + "auxiliary_loss_mlp": 0.01029216, + "balance_loss_clip": 1.04483366, + "balance_loss_mlp": 1.02176547, + "epoch": 0.7958876931401431, + "flos": 18547897530240.0, + "grad_norm": 1.7119717767234708, + "language_loss": 0.7304945, + "learning_rate": 4.211903095665785e-07, + "loss": 0.75215775, + "num_input_tokens_seen": 142585340, + "step": 6619, + "time_per_iteration": 2.7509846687316895 + }, + { + "auxiliary_loss_clip": 0.01154, + "auxiliary_loss_mlp": 0.0102678, + "balance_loss_clip": 1.0457356, + "balance_loss_mlp": 1.01876903, + "epoch": 0.7960079360307821, + "flos": 21543960902400.0, + "grad_norm": 1.8223481688192853, + "language_loss": 0.75782156, + "learning_rate": 4.2071223971298277e-07, + "loss": 0.77962935, + "num_input_tokens_seen": 142602525, + "step": 6620, + "time_per_iteration": 2.721139907836914 + }, + { + "auxiliary_loss_clip": 0.01161089, + "auxiliary_loss_mlp": 0.01030833, + "balance_loss_clip": 1.04557204, + "balance_loss_mlp": 1.02279854, + "epoch": 0.7961281789214213, + "flos": 25481745095040.0, + "grad_norm": 2.444195380492295, + "language_loss": 0.61199713, + "learning_rate": 4.2023440943608433e-07, + "loss": 0.63391638, + "num_input_tokens_seen": 142622490, + "step": 6621, + "time_per_iteration": 2.779311180114746 + }, + { + "auxiliary_loss_clip": 0.01157784, + "auxiliary_loss_mlp": 0.01029793, + "balance_loss_clip": 1.0435307, + "balance_loss_mlp": 1.02236629, + "epoch": 0.7962484218120603, + "flos": 21944436612480.0, + "grad_norm": 1.632350119225419, + "language_loss": 0.78395057, + "learning_rate": 4.1975681880837023e-07, + "loss": 0.80582631, + "num_input_tokens_seen": 142642495, + "step": 6622, + "time_per_iteration": 2.7203431129455566 + }, + { + "auxiliary_loss_clip": 0.0113578, + "auxiliary_loss_mlp": 0.01026544, + "balance_loss_clip": 1.04000413, + "balance_loss_mlp": 1.01868796, + "epoch": 0.7963686647026994, + "flos": 18876264687360.0, + "grad_norm": 1.7213855819975672, + "language_loss": 0.82406795, + "learning_rate": 4.192794679022895e-07, + "loss": 0.8456912, + "num_input_tokens_seen": 142660820, + "step": 6623, + "time_per_iteration": 2.7005958557128906 + }, + { + "auxiliary_loss_clip": 0.0116194, + "auxiliary_loss_mlp": 0.01028711, + "balance_loss_clip": 1.04585838, + "balance_loss_mlp": 1.0206461, + "epoch": 0.7964889075933386, + "flos": 29716582763520.0, + "grad_norm": 1.8835933734827475, + "language_loss": 0.7195909, + "learning_rate": 4.1880235679025743e-07, + "loss": 0.7414974, + "num_input_tokens_seen": 142680915, + "step": 6624, + "time_per_iteration": 2.7893009185791016 + }, + { + "auxiliary_loss_clip": 0.01125523, + "auxiliary_loss_mlp": 0.01026545, + "balance_loss_clip": 1.04178357, + "balance_loss_mlp": 1.01839662, + "epoch": 0.7966091504839776, + "flos": 29491458272640.0, + "grad_norm": 1.8480159021733282, + "language_loss": 0.63471389, + "learning_rate": 4.1832548554464986e-07, + "loss": 0.65623462, + "num_input_tokens_seen": 142699210, + "step": 6625, + "time_per_iteration": 2.876232385635376 + }, + { + "auxiliary_loss_clip": 0.01048499, + "auxiliary_loss_mlp": 0.01003256, + "balance_loss_clip": 1.00613749, + "balance_loss_mlp": 1.00193918, + "epoch": 0.7967293933746167, + "flos": 67288697101440.0, + "grad_norm": 0.7418583251929431, + "language_loss": 0.5865041, + "learning_rate": 4.178488542378098e-07, + "loss": 0.60702169, + "num_input_tokens_seen": 142756790, + "step": 6626, + "time_per_iteration": 3.128384590148926 + }, + { + "auxiliary_loss_clip": 0.01175846, + "auxiliary_loss_mlp": 0.01026602, + "balance_loss_clip": 1.04984152, + "balance_loss_mlp": 1.0182451, + "epoch": 0.7968496362652558, + "flos": 25554679660800.0, + "grad_norm": 1.6401794101507408, + "language_loss": 0.89047164, + "learning_rate": 4.173724629420401e-07, + "loss": 0.91249609, + "num_input_tokens_seen": 142778150, + "step": 6627, + "time_per_iteration": 2.754058599472046 + }, + { + "auxiliary_loss_clip": 0.01157712, + "auxiliary_loss_mlp": 0.01027101, + "balance_loss_clip": 1.04658806, + "balance_loss_mlp": 1.01863742, + "epoch": 0.7969698791558949, + "flos": 14501088581760.0, + "grad_norm": 3.071388938449107, + "language_loss": 0.6835531, + "learning_rate": 4.168963117296087e-07, + "loss": 0.70540118, + "num_input_tokens_seen": 142795485, + "step": 6628, + "time_per_iteration": 2.7218735218048096 + }, + { + "auxiliary_loss_clip": 0.01172379, + "auxiliary_loss_mlp": 0.0103071, + "balance_loss_clip": 1.05008757, + "balance_loss_mlp": 1.02274656, + "epoch": 0.797090122046534, + "flos": 22127545169280.0, + "grad_norm": 5.424093475025655, + "language_loss": 0.75885266, + "learning_rate": 4.1642040067274876e-07, + "loss": 0.78088361, + "num_input_tokens_seen": 142815155, + "step": 6629, + "time_per_iteration": 2.6397547721862793 + }, + { + "auxiliary_loss_clip": 0.01155766, + "auxiliary_loss_mlp": 0.01025778, + "balance_loss_clip": 1.04525995, + "balance_loss_mlp": 1.01819038, + "epoch": 0.7972103649371731, + "flos": 19897671830400.0, + "grad_norm": 1.68123757316556, + "language_loss": 0.72536778, + "learning_rate": 4.1594472984365493e-07, + "loss": 0.7471832, + "num_input_tokens_seen": 142833840, + "step": 6630, + "time_per_iteration": 2.750676155090332 + }, + { + "auxiliary_loss_clip": 0.01156069, + "auxiliary_loss_mlp": 0.01026028, + "balance_loss_clip": 1.04579115, + "balance_loss_mlp": 1.01789212, + "epoch": 0.7973306078278122, + "flos": 36058621847040.0, + "grad_norm": 1.7200394828457501, + "language_loss": 0.78044212, + "learning_rate": 4.154692993144862e-07, + "loss": 0.80226308, + "num_input_tokens_seen": 142853610, + "step": 6631, + "time_per_iteration": 2.8207528591156006 + }, + { + "auxiliary_loss_clip": 0.01169338, + "auxiliary_loss_mlp": 0.00886015, + "balance_loss_clip": 1.04689467, + "balance_loss_mlp": 1.00061107, + "epoch": 0.7974508507184512, + "flos": 21360600950400.0, + "grad_norm": 1.9345891162786522, + "language_loss": 0.71604252, + "learning_rate": 4.1499410915736476e-07, + "loss": 0.73659599, + "num_input_tokens_seen": 142872540, + "step": 6632, + "time_per_iteration": 2.662078380584717 + }, + { + "auxiliary_loss_clip": 0.01054281, + "auxiliary_loss_mlp": 0.01001867, + "balance_loss_clip": 1.00792718, + "balance_loss_mlp": 1.00048971, + "epoch": 0.7975710936090904, + "flos": 68253115317120.0, + "grad_norm": 0.7675754625316706, + "language_loss": 0.64239407, + "learning_rate": 4.145191594443762e-07, + "loss": 0.66295552, + "num_input_tokens_seen": 142936895, + "step": 6633, + "time_per_iteration": 4.312676429748535 + }, + { + "auxiliary_loss_clip": 0.01134422, + "auxiliary_loss_mlp": 0.01029052, + "balance_loss_clip": 1.04218626, + "balance_loss_mlp": 1.02141428, + "epoch": 0.7976913364997295, + "flos": 22492433479680.0, + "grad_norm": 1.6220514624125488, + "language_loss": 0.70654333, + "learning_rate": 4.140444502475713e-07, + "loss": 0.72817808, + "num_input_tokens_seen": 142956445, + "step": 6634, + "time_per_iteration": 3.7508723735809326 + }, + { + "auxiliary_loss_clip": 0.01155237, + "auxiliary_loss_mlp": 0.01027569, + "balance_loss_clip": 1.04403019, + "balance_loss_mlp": 1.01952243, + "epoch": 0.7978115793903685, + "flos": 15263220378240.0, + "grad_norm": 1.752687071401569, + "language_loss": 0.69830465, + "learning_rate": 4.1356998163896216e-07, + "loss": 0.72013271, + "num_input_tokens_seen": 142973495, + "step": 6635, + "time_per_iteration": 2.6025760173797607 + }, + { + "auxiliary_loss_clip": 0.01146625, + "auxiliary_loss_mlp": 0.01025116, + "balance_loss_clip": 1.04503965, + "balance_loss_mlp": 1.01717699, + "epoch": 0.7979318222810077, + "flos": 19719232041600.0, + "grad_norm": 2.162255108036857, + "language_loss": 0.74780089, + "learning_rate": 4.130957536905255e-07, + "loss": 0.76951832, + "num_input_tokens_seen": 142991510, + "step": 6636, + "time_per_iteration": 3.5975182056427 + }, + { + "auxiliary_loss_clip": 0.01152571, + "auxiliary_loss_mlp": 0.01024061, + "balance_loss_clip": 1.04649711, + "balance_loss_mlp": 1.01562142, + "epoch": 0.7980520651716467, + "flos": 15560273854080.0, + "grad_norm": 2.321207585066494, + "language_loss": 0.71161366, + "learning_rate": 4.1262176647420134e-07, + "loss": 0.73337996, + "num_input_tokens_seen": 143009675, + "step": 6637, + "time_per_iteration": 2.6746556758880615 + }, + { + "auxiliary_loss_clip": 0.01156892, + "auxiliary_loss_mlp": 0.01029063, + "balance_loss_clip": 1.04569876, + "balance_loss_mlp": 1.02125156, + "epoch": 0.7981723080622858, + "flos": 22309432663680.0, + "grad_norm": 1.6765183520391642, + "language_loss": 0.79926294, + "learning_rate": 4.121480200618923e-07, + "loss": 0.82112247, + "num_input_tokens_seen": 143029330, + "step": 6638, + "time_per_iteration": 2.7319228649139404 + }, + { + "auxiliary_loss_clip": 0.01143994, + "auxiliary_loss_mlp": 0.01029202, + "balance_loss_clip": 1.0431788, + "balance_loss_mlp": 1.02103925, + "epoch": 0.798292550952925, + "flos": 22929573997440.0, + "grad_norm": 1.7306037345862124, + "language_loss": 0.80297506, + "learning_rate": 4.116745145254674e-07, + "loss": 0.82470703, + "num_input_tokens_seen": 143048865, + "step": 6639, + "time_per_iteration": 2.745957851409912 + }, + { + "auxiliary_loss_clip": 0.01040282, + "auxiliary_loss_mlp": 0.01000794, + "balance_loss_clip": 1.00645888, + "balance_loss_mlp": 0.99945307, + "epoch": 0.798412793843564, + "flos": 64497936890880.0, + "grad_norm": 0.7617389390856535, + "language_loss": 0.57990092, + "learning_rate": 4.1120124993675476e-07, + "loss": 0.60031164, + "num_input_tokens_seen": 143113295, + "step": 6640, + "time_per_iteration": 3.286728858947754 + }, + { + "auxiliary_loss_clip": 0.01160008, + "auxiliary_loss_mlp": 0.01022369, + "balance_loss_clip": 1.04637945, + "balance_loss_mlp": 1.01367855, + "epoch": 0.7985330367342031, + "flos": 13586910514560.0, + "grad_norm": 1.9113765711437558, + "language_loss": 0.61633652, + "learning_rate": 4.107282263675498e-07, + "loss": 0.63816035, + "num_input_tokens_seen": 143130965, + "step": 6641, + "time_per_iteration": 3.6709280014038086 + }, + { + "auxiliary_loss_clip": 0.01048082, + "auxiliary_loss_mlp": 0.00877038, + "balance_loss_clip": 1.00969875, + "balance_loss_mlp": 1.00109696, + "epoch": 0.7986532796248422, + "flos": 67698797656320.0, + "grad_norm": 0.7852249122724416, + "language_loss": 0.52567655, + "learning_rate": 4.1025544388960907e-07, + "loss": 0.54492772, + "num_input_tokens_seen": 143192005, + "step": 6642, + "time_per_iteration": 3.3197968006134033 + }, + { + "auxiliary_loss_clip": 0.01161164, + "auxiliary_loss_mlp": 0.01025117, + "balance_loss_clip": 1.04876328, + "balance_loss_mlp": 1.01768398, + "epoch": 0.7987735225154813, + "flos": 22455373622400.0, + "grad_norm": 1.7592495597663558, + "language_loss": 0.71523178, + "learning_rate": 4.097829025746538e-07, + "loss": 0.73709464, + "num_input_tokens_seen": 143213550, + "step": 6643, + "time_per_iteration": 2.7101926803588867 + }, + { + "auxiliary_loss_clip": 0.0105095, + "auxiliary_loss_mlp": 0.01001776, + "balance_loss_clip": 1.00672865, + "balance_loss_mlp": 1.000471, + "epoch": 0.7988937654061203, + "flos": 68864098682880.0, + "grad_norm": 0.6641559885693112, + "language_loss": 0.61010748, + "learning_rate": 4.0931060249436757e-07, + "loss": 0.63063473, + "num_input_tokens_seen": 143277390, + "step": 6644, + "time_per_iteration": 3.2950265407562256 + }, + { + "auxiliary_loss_clip": 0.01158181, + "auxiliary_loss_mlp": 0.01023633, + "balance_loss_clip": 1.04751313, + "balance_loss_mlp": 1.01623654, + "epoch": 0.7990140082967595, + "flos": 20806893820800.0, + "grad_norm": 1.8976591501966382, + "language_loss": 0.69973576, + "learning_rate": 4.088385437203978e-07, + "loss": 0.72155392, + "num_input_tokens_seen": 143294400, + "step": 6645, + "time_per_iteration": 2.860687017440796 + }, + { + "auxiliary_loss_clip": 0.01167118, + "auxiliary_loss_mlp": 0.01026646, + "balance_loss_clip": 1.04389215, + "balance_loss_mlp": 1.01844406, + "epoch": 0.7991342511873986, + "flos": 18985289443200.0, + "grad_norm": 1.9854546428289763, + "language_loss": 0.7719667, + "learning_rate": 4.083667263243564e-07, + "loss": 0.79390436, + "num_input_tokens_seen": 143312745, + "step": 6646, + "time_per_iteration": 2.6111152172088623 + }, + { + "auxiliary_loss_clip": 0.01154805, + "auxiliary_loss_mlp": 0.01026686, + "balance_loss_clip": 1.04644275, + "balance_loss_mlp": 1.0190568, + "epoch": 0.7992544940780376, + "flos": 20816805974400.0, + "grad_norm": 1.6988091908520637, + "language_loss": 0.71468806, + "learning_rate": 4.0789515037781653e-07, + "loss": 0.73650289, + "num_input_tokens_seen": 143333470, + "step": 6647, + "time_per_iteration": 2.6720621585845947 + }, + { + "auxiliary_loss_clip": 0.01162795, + "auxiliary_loss_mlp": 0.01033109, + "balance_loss_clip": 1.04572654, + "balance_loss_mlp": 1.02527702, + "epoch": 0.7993747369686768, + "flos": 12640772321280.0, + "grad_norm": 1.7575613806340806, + "language_loss": 0.82429969, + "learning_rate": 4.0742381595231755e-07, + "loss": 0.84625876, + "num_input_tokens_seen": 143350195, + "step": 6648, + "time_per_iteration": 2.6198432445526123 + }, + { + "auxiliary_loss_clip": 0.01137839, + "auxiliary_loss_mlp": 0.01032204, + "balance_loss_clip": 1.04453945, + "balance_loss_mlp": 1.02437806, + "epoch": 0.7994949798593158, + "flos": 20078769225600.0, + "grad_norm": 1.5505008286467932, + "language_loss": 0.78230214, + "learning_rate": 4.06952723119359e-07, + "loss": 0.80400252, + "num_input_tokens_seen": 143370070, + "step": 6649, + "time_per_iteration": 2.783993721008301 + }, + { + "auxiliary_loss_clip": 0.01141747, + "auxiliary_loss_mlp": 0.01032875, + "balance_loss_clip": 1.04361749, + "balance_loss_mlp": 1.02494144, + "epoch": 0.7996152227499549, + "flos": 38654209509120.0, + "grad_norm": 1.9313019067615664, + "language_loss": 0.67356759, + "learning_rate": 4.0648187195040504e-07, + "loss": 0.69531381, + "num_input_tokens_seen": 143392275, + "step": 6650, + "time_per_iteration": 2.815295934677124 + }, + { + "auxiliary_loss_clip": 0.01047892, + "auxiliary_loss_mlp": 0.01001587, + "balance_loss_clip": 1.0057199, + "balance_loss_mlp": 1.00031745, + "epoch": 0.799735465640594, + "flos": 70243821947520.0, + "grad_norm": 0.9367238312902167, + "language_loss": 0.67563933, + "learning_rate": 4.060112625168848e-07, + "loss": 0.69613409, + "num_input_tokens_seen": 143457385, + "step": 6651, + "time_per_iteration": 3.296517848968506 + }, + { + "auxiliary_loss_clip": 0.01170074, + "auxiliary_loss_mlp": 0.01029577, + "balance_loss_clip": 1.04920363, + "balance_loss_mlp": 1.02126849, + "epoch": 0.7998557085312331, + "flos": 24240995550720.0, + "grad_norm": 3.589219262465667, + "language_loss": 0.73804975, + "learning_rate": 4.055408948901886e-07, + "loss": 0.76004624, + "num_input_tokens_seen": 143478785, + "step": 6652, + "time_per_iteration": 2.647115468978882 + }, + { + "auxiliary_loss_clip": 0.01163714, + "auxiliary_loss_mlp": 0.01028582, + "balance_loss_clip": 1.04640865, + "balance_loss_mlp": 1.02085125, + "epoch": 0.7999759514218722, + "flos": 27564025449600.0, + "grad_norm": 1.8296812722002973, + "language_loss": 0.71336091, + "learning_rate": 4.050707691416708e-07, + "loss": 0.73528385, + "num_input_tokens_seen": 143500095, + "step": 6653, + "time_per_iteration": 2.7188260555267334 + }, + { + "auxiliary_loss_clip": 0.01048012, + "auxiliary_loss_mlp": 0.01002586, + "balance_loss_clip": 1.0059123, + "balance_loss_mlp": 1.00137043, + "epoch": 0.8000961943125112, + "flos": 67337428878720.0, + "grad_norm": 0.6714731250345416, + "language_loss": 0.59776521, + "learning_rate": 4.046008853426495e-07, + "loss": 0.61827111, + "num_input_tokens_seen": 143563410, + "step": 6654, + "time_per_iteration": 3.293491840362549 + }, + { + "auxiliary_loss_clip": 0.01136029, + "auxiliary_loss_mlp": 0.01026301, + "balance_loss_clip": 1.04257751, + "balance_loss_mlp": 1.01854658, + "epoch": 0.8002164372031504, + "flos": 28733815676160.0, + "grad_norm": 3.953444163814838, + "language_loss": 0.62911749, + "learning_rate": 4.0413124356440464e-07, + "loss": 0.6507408, + "num_input_tokens_seen": 143587455, + "step": 6655, + "time_per_iteration": 2.8272058963775635 + }, + { + "auxiliary_loss_clip": 0.01134661, + "auxiliary_loss_mlp": 0.01031087, + "balance_loss_clip": 1.04184198, + "balance_loss_mlp": 1.02315927, + "epoch": 0.8003366800937894, + "flos": 17639429725440.0, + "grad_norm": 1.7614231036796342, + "language_loss": 0.81950009, + "learning_rate": 4.0366184387818223e-07, + "loss": 0.84115756, + "num_input_tokens_seen": 143605915, + "step": 6656, + "time_per_iteration": 2.8267409801483154 + }, + { + "auxiliary_loss_clip": 0.01176912, + "auxiliary_loss_mlp": 0.01028069, + "balance_loss_clip": 1.04999745, + "balance_loss_mlp": 1.01954579, + "epoch": 0.8004569229844285, + "flos": 25995303797760.0, + "grad_norm": 1.79472131463868, + "language_loss": 0.85285556, + "learning_rate": 4.0319268635518797e-07, + "loss": 0.87490541, + "num_input_tokens_seen": 143626490, + "step": 6657, + "time_per_iteration": 304.3151707649231 + }, + { + "auxiliary_loss_clip": 0.01159825, + "auxiliary_loss_mlp": 0.01026039, + "balance_loss_clip": 1.04519081, + "balance_loss_mlp": 1.0182811, + "epoch": 0.8005771658750677, + "flos": 20812352688000.0, + "grad_norm": 1.522851898022585, + "language_loss": 0.75119889, + "learning_rate": 4.027237710665943e-07, + "loss": 0.77305758, + "num_input_tokens_seen": 143644955, + "step": 6658, + "time_per_iteration": 2.762040615081787 + }, + { + "auxiliary_loss_clip": 0.01146194, + "auxiliary_loss_mlp": 0.01029014, + "balance_loss_clip": 1.04221964, + "balance_loss_mlp": 1.02056241, + "epoch": 0.8006974087657067, + "flos": 25812626204160.0, + "grad_norm": 1.7850397893839938, + "language_loss": 0.69284374, + "learning_rate": 4.022550980835344e-07, + "loss": 0.71459579, + "num_input_tokens_seen": 143667200, + "step": 6659, + "time_per_iteration": 3.8690357208251953 + }, + { + "auxiliary_loss_clip": 0.01139614, + "auxiliary_loss_mlp": 0.01026161, + "balance_loss_clip": 1.04085112, + "balance_loss_mlp": 1.01844859, + "epoch": 0.8008176516563458, + "flos": 17164690646400.0, + "grad_norm": 2.6437665567512436, + "language_loss": 0.79519928, + "learning_rate": 4.017866674771051e-07, + "loss": 0.81685698, + "num_input_tokens_seen": 143684685, + "step": 6660, + "time_per_iteration": 3.8156065940856934 + }, + { + "auxiliary_loss_clip": 0.0112628, + "auxiliary_loss_mlp": 0.01028325, + "balance_loss_clip": 1.04060626, + "balance_loss_mlp": 1.02013576, + "epoch": 0.8009378945469849, + "flos": 24207311571840.0, + "grad_norm": 1.833838778369798, + "language_loss": 0.74603546, + "learning_rate": 4.013184793183688e-07, + "loss": 0.76758146, + "num_input_tokens_seen": 143706780, + "step": 6661, + "time_per_iteration": 3.1417250633239746 + }, + { + "auxiliary_loss_clip": 0.01158616, + "auxiliary_loss_mlp": 0.010316, + "balance_loss_clip": 1.04532325, + "balance_loss_mlp": 1.02405977, + "epoch": 0.801058137437624, + "flos": 19787318271360.0, + "grad_norm": 1.8108257848569191, + "language_loss": 0.7272976, + "learning_rate": 4.008505336783472e-07, + "loss": 0.74919981, + "num_input_tokens_seen": 143724505, + "step": 6662, + "time_per_iteration": 3.7761950492858887 + }, + { + "auxiliary_loss_clip": 0.0115208, + "auxiliary_loss_mlp": 0.01027303, + "balance_loss_clip": 1.04551733, + "balance_loss_mlp": 1.01998377, + "epoch": 0.801178380328263, + "flos": 18659400324480.0, + "grad_norm": 2.224841057095926, + "language_loss": 0.80934966, + "learning_rate": 4.003828306280284e-07, + "loss": 0.8311435, + "num_input_tokens_seen": 143742180, + "step": 6663, + "time_per_iteration": 2.884514808654785 + }, + { + "auxiliary_loss_clip": 0.01151971, + "auxiliary_loss_mlp": 0.0102688, + "balance_loss_clip": 1.04588258, + "balance_loss_mlp": 1.01965284, + "epoch": 0.8012986232189022, + "flos": 15706573948800.0, + "grad_norm": 1.9920897350583249, + "language_loss": 0.77684057, + "learning_rate": 3.999153702383626e-07, + "loss": 0.79862911, + "num_input_tokens_seen": 143760070, + "step": 6664, + "time_per_iteration": 2.6241772174835205 + }, + { + "auxiliary_loss_clip": 0.0116441, + "auxiliary_loss_mlp": 0.01031023, + "balance_loss_clip": 1.04664683, + "balance_loss_mlp": 1.02286267, + "epoch": 0.8014188661095413, + "flos": 28584139703040.0, + "grad_norm": 1.7155495861553498, + "language_loss": 0.73879337, + "learning_rate": 3.9944815258026263e-07, + "loss": 0.76074767, + "num_input_tokens_seen": 143781890, + "step": 6665, + "time_per_iteration": 2.779268503189087 + }, + { + "auxiliary_loss_clip": 0.01164778, + "auxiliary_loss_mlp": 0.0103064, + "balance_loss_clip": 1.04750395, + "balance_loss_mlp": 1.02234864, + "epoch": 0.8015391090001803, + "flos": 29310360877440.0, + "grad_norm": 1.9952081720796848, + "language_loss": 0.8305341, + "learning_rate": 3.989811777246057e-07, + "loss": 0.85248822, + "num_input_tokens_seen": 143802060, + "step": 6666, + "time_per_iteration": 2.825155019760132 + }, + { + "auxiliary_loss_clip": 0.01056298, + "auxiliary_loss_mlp": 0.0100258, + "balance_loss_clip": 1.00522614, + "balance_loss_mlp": 1.00122666, + "epoch": 0.8016593518908195, + "flos": 70397340675840.0, + "grad_norm": 0.8574394209791819, + "language_loss": 0.66255218, + "learning_rate": 3.985144457422305e-07, + "loss": 0.68314093, + "num_input_tokens_seen": 143856345, + "step": 6667, + "time_per_iteration": 4.09240198135376 + }, + { + "auxiliary_loss_clip": 0.01168298, + "auxiliary_loss_mlp": 0.01025782, + "balance_loss_clip": 1.04653966, + "balance_loss_mlp": 1.01824784, + "epoch": 0.8017795947814585, + "flos": 26026114688640.0, + "grad_norm": 1.8954080734089305, + "language_loss": 0.76961756, + "learning_rate": 3.9804795670394096e-07, + "loss": 0.79155838, + "num_input_tokens_seen": 143876470, + "step": 6668, + "time_per_iteration": 2.665942907333374 + }, + { + "auxiliary_loss_clip": 0.01141564, + "auxiliary_loss_mlp": 0.01026676, + "balance_loss_clip": 1.04278398, + "balance_loss_mlp": 1.01883233, + "epoch": 0.8018998376720976, + "flos": 22087181260800.0, + "grad_norm": 1.650769992844497, + "language_loss": 0.70510173, + "learning_rate": 3.975817106805022e-07, + "loss": 0.72678417, + "num_input_tokens_seen": 143895170, + "step": 6669, + "time_per_iteration": 2.7722105979919434 + }, + { + "auxiliary_loss_clip": 0.01141663, + "auxiliary_loss_mlp": 0.01027615, + "balance_loss_clip": 1.04330826, + "balance_loss_mlp": 1.01954412, + "epoch": 0.8020200805627368, + "flos": 34568545023360.0, + "grad_norm": 1.716809820085747, + "language_loss": 0.65091097, + "learning_rate": 3.97115707742645e-07, + "loss": 0.67260379, + "num_input_tokens_seen": 143915845, + "step": 6670, + "time_per_iteration": 2.973520040512085 + }, + { + "auxiliary_loss_clip": 0.01153688, + "auxiliary_loss_mlp": 0.01023598, + "balance_loss_clip": 1.04660225, + "balance_loss_mlp": 1.01602781, + "epoch": 0.8021403234533758, + "flos": 20120354196480.0, + "grad_norm": 2.041527985443413, + "language_loss": 0.64755589, + "learning_rate": 3.966499479610599e-07, + "loss": 0.66932875, + "num_input_tokens_seen": 143933940, + "step": 6671, + "time_per_iteration": 2.724839687347412 + }, + { + "auxiliary_loss_clip": 0.01143316, + "auxiliary_loss_mlp": 0.01030954, + "balance_loss_clip": 1.04691339, + "balance_loss_mlp": 1.02382493, + "epoch": 0.8022605663440149, + "flos": 27746200252800.0, + "grad_norm": 1.7661421392920464, + "language_loss": 0.65302682, + "learning_rate": 3.9618443140640225e-07, + "loss": 0.67476952, + "num_input_tokens_seen": 143952850, + "step": 6672, + "time_per_iteration": 2.8401620388031006 + }, + { + "auxiliary_loss_clip": 0.01032679, + "auxiliary_loss_mlp": 0.01002934, + "balance_loss_clip": 1.00669384, + "balance_loss_mlp": 1.0014317, + "epoch": 0.802380809234654, + "flos": 60244998768000.0, + "grad_norm": 3.050407249099782, + "language_loss": 0.51351976, + "learning_rate": 3.957191581492918e-07, + "loss": 0.53387588, + "num_input_tokens_seen": 144013610, + "step": 6673, + "time_per_iteration": 3.382108211517334 + }, + { + "auxiliary_loss_clip": 0.01149007, + "auxiliary_loss_mlp": 0.01032384, + "balance_loss_clip": 1.04442453, + "balance_loss_mlp": 1.02450442, + "epoch": 0.8025010521252931, + "flos": 15080722352640.0, + "grad_norm": 2.7869880055213017, + "language_loss": 0.70855772, + "learning_rate": 3.952541282603097e-07, + "loss": 0.73037159, + "num_input_tokens_seen": 144028715, + "step": 6674, + "time_per_iteration": 2.7307546138763428 + }, + { + "auxiliary_loss_clip": 0.01156911, + "auxiliary_loss_mlp": 0.01024502, + "balance_loss_clip": 1.04619992, + "balance_loss_mlp": 1.0165031, + "epoch": 0.8026212950159322, + "flos": 22163527618560.0, + "grad_norm": 1.8375458971624707, + "language_loss": 0.83265239, + "learning_rate": 3.9478934181000013e-07, + "loss": 0.8544665, + "num_input_tokens_seen": 144048740, + "step": 6675, + "time_per_iteration": 2.7621169090270996 + }, + { + "auxiliary_loss_clip": 0.01170431, + "auxiliary_loss_mlp": 0.0102854, + "balance_loss_clip": 1.045964, + "balance_loss_mlp": 1.02053475, + "epoch": 0.8027415379065713, + "flos": 17675986792320.0, + "grad_norm": 2.2670418991553767, + "language_loss": 0.84331554, + "learning_rate": 3.943247988688714e-07, + "loss": 0.86530524, + "num_input_tokens_seen": 144067435, + "step": 6676, + "time_per_iteration": 2.733884334564209 + }, + { + "auxiliary_loss_clip": 0.01160182, + "auxiliary_loss_mlp": 0.01027891, + "balance_loss_clip": 1.04546046, + "balance_loss_mlp": 1.02033317, + "epoch": 0.8028617807972104, + "flos": 21979593048960.0, + "grad_norm": 2.526354857767045, + "language_loss": 0.72021556, + "learning_rate": 3.938604995073933e-07, + "loss": 0.7420963, + "num_input_tokens_seen": 144085905, + "step": 6677, + "time_per_iteration": 2.7767210006713867 + }, + { + "auxiliary_loss_clip": 0.01153102, + "auxiliary_loss_mlp": 0.01033091, + "balance_loss_clip": 1.0447371, + "balance_loss_mlp": 1.02529442, + "epoch": 0.8029820236878494, + "flos": 26428457905920.0, + "grad_norm": 1.800110330431966, + "language_loss": 0.65404534, + "learning_rate": 3.9339644379600157e-07, + "loss": 0.67590725, + "num_input_tokens_seen": 144105735, + "step": 6678, + "time_per_iteration": 2.7856192588806152 + }, + { + "auxiliary_loss_clip": 0.01156302, + "auxiliary_loss_mlp": 0.0102772, + "balance_loss_clip": 1.04831839, + "balance_loss_mlp": 1.01961398, + "epoch": 0.8031022665784886, + "flos": 17676489582720.0, + "grad_norm": 2.035219303612223, + "language_loss": 0.71006596, + "learning_rate": 3.929326318050907e-07, + "loss": 0.73190624, + "num_input_tokens_seen": 144123405, + "step": 6679, + "time_per_iteration": 2.7189509868621826 + }, + { + "auxiliary_loss_clip": 0.01164573, + "auxiliary_loss_mlp": 0.0102777, + "balance_loss_clip": 1.04416811, + "balance_loss_mlp": 1.01984239, + "epoch": 0.8032225094691277, + "flos": 15450279431040.0, + "grad_norm": 1.8292539264200423, + "language_loss": 0.7859267, + "learning_rate": 3.924690636050225e-07, + "loss": 0.80785012, + "num_input_tokens_seen": 144140815, + "step": 6680, + "time_per_iteration": 2.5684397220611572 + }, + { + "auxiliary_loss_clip": 0.01163462, + "auxiliary_loss_mlp": 0.01027867, + "balance_loss_clip": 1.04826784, + "balance_loss_mlp": 1.01972473, + "epoch": 0.8033427523597667, + "flos": 26179202453760.0, + "grad_norm": 2.4689848946166095, + "language_loss": 0.72880888, + "learning_rate": 3.9200573926611915e-07, + "loss": 0.75072217, + "num_input_tokens_seen": 144162230, + "step": 6681, + "time_per_iteration": 2.7249724864959717 + }, + { + "auxiliary_loss_clip": 0.01158502, + "auxiliary_loss_mlp": 0.0102477, + "balance_loss_clip": 1.04759955, + "balance_loss_mlp": 1.01633644, + "epoch": 0.8034629952504058, + "flos": 21324905809920.0, + "grad_norm": 2.2409029850954583, + "language_loss": 0.73229802, + "learning_rate": 3.9154265885866613e-07, + "loss": 0.75413072, + "num_input_tokens_seen": 144181540, + "step": 6682, + "time_per_iteration": 2.7514028549194336 + }, + { + "auxiliary_loss_clip": 0.01160314, + "auxiliary_loss_mlp": 0.01028064, + "balance_loss_clip": 1.04720664, + "balance_loss_mlp": 1.01942754, + "epoch": 0.8035832381410449, + "flos": 21651585027840.0, + "grad_norm": 4.13005790087241, + "language_loss": 0.75014406, + "learning_rate": 3.9107982245291394e-07, + "loss": 0.77202785, + "num_input_tokens_seen": 144199665, + "step": 6683, + "time_per_iteration": 2.795833110809326 + }, + { + "auxiliary_loss_clip": 0.01143103, + "auxiliary_loss_mlp": 0.01028425, + "balance_loss_clip": 1.04471517, + "balance_loss_mlp": 1.02027678, + "epoch": 0.803703481031684, + "flos": 20518818744960.0, + "grad_norm": 2.1247923034344063, + "language_loss": 0.77526021, + "learning_rate": 3.9061723011907245e-07, + "loss": 0.79697549, + "num_input_tokens_seen": 144219020, + "step": 6684, + "time_per_iteration": 3.652433156967163 + }, + { + "auxiliary_loss_clip": 0.01151995, + "auxiliary_loss_mlp": 0.01027326, + "balance_loss_clip": 1.04600096, + "balance_loss_mlp": 1.01914847, + "epoch": 0.803823723922323, + "flos": 22854807838080.0, + "grad_norm": 1.6746908847652981, + "language_loss": 0.79225147, + "learning_rate": 3.901548819273179e-07, + "loss": 0.81404471, + "num_input_tokens_seen": 144239035, + "step": 6685, + "time_per_iteration": 2.7205255031585693 + }, + { + "auxiliary_loss_clip": 0.0116213, + "auxiliary_loss_mlp": 0.01026843, + "balance_loss_clip": 1.04855895, + "balance_loss_mlp": 1.01858783, + "epoch": 0.8039439668129622, + "flos": 21362145235200.0, + "grad_norm": 3.5814061070096876, + "language_loss": 0.69130987, + "learning_rate": 3.896927779477881e-07, + "loss": 0.71319962, + "num_input_tokens_seen": 144258295, + "step": 6686, + "time_per_iteration": 3.9062142372131348 + }, + { + "auxiliary_loss_clip": 0.01139121, + "auxiliary_loss_mlp": 0.01028974, + "balance_loss_clip": 1.04173791, + "balance_loss_mlp": 1.02159214, + "epoch": 0.8040642097036013, + "flos": 23802382575360.0, + "grad_norm": 3.458877363085688, + "language_loss": 0.67093182, + "learning_rate": 3.892309182505833e-07, + "loss": 0.69261283, + "num_input_tokens_seen": 144276110, + "step": 6687, + "time_per_iteration": 2.9471986293792725 + }, + { + "auxiliary_loss_clip": 0.0116782, + "auxiliary_loss_mlp": 0.01026913, + "balance_loss_clip": 1.04649842, + "balance_loss_mlp": 1.0191648, + "epoch": 0.8041844525942403, + "flos": 25922046009600.0, + "grad_norm": 2.2928272172712485, + "language_loss": 0.860699, + "learning_rate": 3.887693029057675e-07, + "loss": 0.88264632, + "num_input_tokens_seen": 144295620, + "step": 6688, + "time_per_iteration": 3.954676866531372 + }, + { + "auxiliary_loss_clip": 0.01149756, + "auxiliary_loss_mlp": 0.0102662, + "balance_loss_clip": 1.04371393, + "balance_loss_mlp": 1.01984334, + "epoch": 0.8043046954848795, + "flos": 25191120153600.0, + "grad_norm": 1.6098299215662686, + "language_loss": 0.81286609, + "learning_rate": 3.8830793198336684e-07, + "loss": 0.83462983, + "num_input_tokens_seen": 144315210, + "step": 6689, + "time_per_iteration": 2.8849611282348633 + }, + { + "auxiliary_loss_clip": 0.0115561, + "auxiliary_loss_mlp": 0.0102784, + "balance_loss_clip": 1.04559326, + "balance_loss_mlp": 1.01984692, + "epoch": 0.8044249383755185, + "flos": 41719185123840.0, + "grad_norm": 1.8320954634886137, + "language_loss": 0.70435047, + "learning_rate": 3.878468055533721e-07, + "loss": 0.72618496, + "num_input_tokens_seen": 144337750, + "step": 6690, + "time_per_iteration": 2.9640274047851562 + }, + { + "auxiliary_loss_clip": 0.01149864, + "auxiliary_loss_mlp": 0.01031084, + "balance_loss_clip": 1.04675174, + "balance_loss_mlp": 1.02264392, + "epoch": 0.8045451812661576, + "flos": 20631434860800.0, + "grad_norm": 2.412108341235735, + "language_loss": 0.85134339, + "learning_rate": 3.8738592368573464e-07, + "loss": 0.87315285, + "num_input_tokens_seen": 144355305, + "step": 6691, + "time_per_iteration": 2.771845817565918 + }, + { + "auxiliary_loss_clip": 0.0113493, + "auxiliary_loss_mlp": 0.01029983, + "balance_loss_clip": 1.0441314, + "balance_loss_mlp": 1.02110195, + "epoch": 0.8046654241567968, + "flos": 29711806254720.0, + "grad_norm": 1.8604832044779076, + "language_loss": 0.88325328, + "learning_rate": 3.8692528645037137e-07, + "loss": 0.9049024, + "num_input_tokens_seen": 144374485, + "step": 6692, + "time_per_iteration": 2.8544886112213135 + }, + { + "auxiliary_loss_clip": 0.01170162, + "auxiliary_loss_mlp": 0.01028761, + "balance_loss_clip": 1.04868317, + "balance_loss_mlp": 1.0206306, + "epoch": 0.8047856670474358, + "flos": 17671389851520.0, + "grad_norm": 2.183219145236387, + "language_loss": 0.77413684, + "learning_rate": 3.8646489391715907e-07, + "loss": 0.79612607, + "num_input_tokens_seen": 144388780, + "step": 6693, + "time_per_iteration": 3.5532593727111816 + }, + { + "auxiliary_loss_clip": 0.01154822, + "auxiliary_loss_mlp": 0.01034202, + "balance_loss_clip": 1.04673457, + "balance_loss_mlp": 1.0255239, + "epoch": 0.8049059099380749, + "flos": 17120699464320.0, + "grad_norm": 2.57797529594764, + "language_loss": 0.88193887, + "learning_rate": 3.8600474615593903e-07, + "loss": 0.9038291, + "num_input_tokens_seen": 144403395, + "step": 6694, + "time_per_iteration": 2.852569341659546 + }, + { + "auxiliary_loss_clip": 0.01042973, + "auxiliary_loss_mlp": 0.01001665, + "balance_loss_clip": 1.00710702, + "balance_loss_mlp": 1.00017512, + "epoch": 0.805026152828714, + "flos": 62212903240320.0, + "grad_norm": 0.7772218236790794, + "language_loss": 0.5967288, + "learning_rate": 3.8554484323651605e-07, + "loss": 0.61717522, + "num_input_tokens_seen": 144465265, + "step": 6695, + "time_per_iteration": 3.4329400062561035 + }, + { + "auxiliary_loss_clip": 0.01158064, + "auxiliary_loss_mlp": 0.00885379, + "balance_loss_clip": 1.04628646, + "balance_loss_mlp": 1.00056803, + "epoch": 0.8051463957193531, + "flos": 21688608971520.0, + "grad_norm": 1.5907351918461379, + "language_loss": 0.78708649, + "learning_rate": 3.85085185228657e-07, + "loss": 0.80752093, + "num_input_tokens_seen": 144484235, + "step": 6696, + "time_per_iteration": 2.707014799118042 + }, + { + "auxiliary_loss_clip": 0.0114468, + "auxiliary_loss_mlp": 0.01029205, + "balance_loss_clip": 1.04440093, + "balance_loss_mlp": 1.02157545, + "epoch": 0.8052666386099921, + "flos": 32051458535040.0, + "grad_norm": 3.150292237350344, + "language_loss": 0.73119807, + "learning_rate": 3.8462577220209114e-07, + "loss": 0.75293696, + "num_input_tokens_seen": 144504610, + "step": 6697, + "time_per_iteration": 2.8426969051361084 + }, + { + "auxiliary_loss_clip": 0.01056603, + "auxiliary_loss_mlp": 0.01002252, + "balance_loss_clip": 1.00531626, + "balance_loss_mlp": 1.00086915, + "epoch": 0.8053868815006313, + "flos": 67157875768320.0, + "grad_norm": 0.7140902613729962, + "language_loss": 0.59038985, + "learning_rate": 3.8416660422651127e-07, + "loss": 0.61097836, + "num_input_tokens_seen": 144574260, + "step": 6698, + "time_per_iteration": 3.3250536918640137 + }, + { + "auxiliary_loss_clip": 0.01144214, + "auxiliary_loss_mlp": 0.01024998, + "balance_loss_clip": 1.04235184, + "balance_loss_mlp": 1.01730299, + "epoch": 0.8055071243912704, + "flos": 23837000307840.0, + "grad_norm": 3.9589600343674256, + "language_loss": 0.67882574, + "learning_rate": 3.837076813715723e-07, + "loss": 0.70051789, + "num_input_tokens_seen": 144594145, + "step": 6699, + "time_per_iteration": 2.8890256881713867 + }, + { + "auxiliary_loss_clip": 0.01138283, + "auxiliary_loss_mlp": 0.01030063, + "balance_loss_clip": 1.04099894, + "balance_loss_mlp": 1.02172971, + "epoch": 0.8056273672819094, + "flos": 21324510760320.0, + "grad_norm": 1.8578146738231136, + "language_loss": 0.75378597, + "learning_rate": 3.832490037068941e-07, + "loss": 0.77546948, + "num_input_tokens_seen": 144612935, + "step": 6700, + "time_per_iteration": 2.787797689437866 + }, + { + "auxiliary_loss_clip": 0.01115628, + "auxiliary_loss_mlp": 0.01025091, + "balance_loss_clip": 1.03918719, + "balance_loss_mlp": 1.01727104, + "epoch": 0.8057476101725486, + "flos": 25768383626880.0, + "grad_norm": 2.2865652183393523, + "language_loss": 0.76597387, + "learning_rate": 3.827905713020554e-07, + "loss": 0.78738105, + "num_input_tokens_seen": 144630580, + "step": 6701, + "time_per_iteration": 2.8811936378479004 + }, + { + "auxiliary_loss_clip": 0.01144543, + "auxiliary_loss_mlp": 0.01037292, + "balance_loss_clip": 1.04191899, + "balance_loss_mlp": 1.02810693, + "epoch": 0.8058678530631876, + "flos": 24535283679360.0, + "grad_norm": 2.161749790352874, + "language_loss": 0.68993819, + "learning_rate": 3.823323842266017e-07, + "loss": 0.71175659, + "num_input_tokens_seen": 144649975, + "step": 6702, + "time_per_iteration": 2.9092061519622803 + }, + { + "auxiliary_loss_clip": 0.01162594, + "auxiliary_loss_mlp": 0.0102345, + "balance_loss_clip": 1.04535484, + "balance_loss_mlp": 1.01514161, + "epoch": 0.8059880959538267, + "flos": 24753728240640.0, + "grad_norm": 2.2969077988369397, + "language_loss": 0.73028958, + "learning_rate": 3.818744425500393e-07, + "loss": 0.75214994, + "num_input_tokens_seen": 144667990, + "step": 6703, + "time_per_iteration": 2.769368886947632 + }, + { + "auxiliary_loss_clip": 0.01133955, + "auxiliary_loss_mlp": 0.0102955, + "balance_loss_clip": 1.040012, + "balance_loss_mlp": 1.02111006, + "epoch": 0.8061083388444659, + "flos": 22196349671040.0, + "grad_norm": 2.1469854892921005, + "language_loss": 0.8058393, + "learning_rate": 3.8141674634183675e-07, + "loss": 0.82747436, + "num_input_tokens_seen": 144687020, + "step": 6704, + "time_per_iteration": 2.8539998531341553 + }, + { + "auxiliary_loss_clip": 0.01133221, + "auxiliary_loss_mlp": 0.01033227, + "balance_loss_clip": 1.04578114, + "balance_loss_mlp": 1.02543664, + "epoch": 0.8062285817351049, + "flos": 30044195735040.0, + "grad_norm": 1.7365302195569732, + "language_loss": 0.66060901, + "learning_rate": 3.809592956714278e-07, + "loss": 0.68227351, + "num_input_tokens_seen": 144710255, + "step": 6705, + "time_per_iteration": 2.9921910762786865 + }, + { + "auxiliary_loss_clip": 0.0116962, + "auxiliary_loss_mlp": 0.01034827, + "balance_loss_clip": 1.05080652, + "balance_loss_mlp": 1.02664363, + "epoch": 0.806348824625744, + "flos": 22782591544320.0, + "grad_norm": 2.047766092997203, + "language_loss": 0.74480504, + "learning_rate": 3.805020906082057e-07, + "loss": 0.76684952, + "num_input_tokens_seen": 144728830, + "step": 6706, + "time_per_iteration": 2.8021695613861084 + }, + { + "auxiliary_loss_clip": 0.01155067, + "auxiliary_loss_mlp": 0.01034749, + "balance_loss_clip": 1.04635453, + "balance_loss_mlp": 1.02647614, + "epoch": 0.8064690675163831, + "flos": 23404600385280.0, + "grad_norm": 2.0638777325714877, + "language_loss": 0.81066394, + "learning_rate": 3.8004513122152917e-07, + "loss": 0.83256209, + "num_input_tokens_seen": 144747140, + "step": 6707, + "time_per_iteration": 2.857422351837158 + }, + { + "auxiliary_loss_clip": 0.01140406, + "auxiliary_loss_mlp": 0.01028446, + "balance_loss_clip": 1.04579854, + "balance_loss_mlp": 1.02095056, + "epoch": 0.8065893104070222, + "flos": 24060903736320.0, + "grad_norm": 1.7444302825318807, + "language_loss": 0.67418134, + "learning_rate": 3.79588417580718e-07, + "loss": 0.6958698, + "num_input_tokens_seen": 144765250, + "step": 6708, + "time_per_iteration": 2.887669801712036 + }, + { + "auxiliary_loss_clip": 0.01164388, + "auxiliary_loss_mlp": 0.01031222, + "balance_loss_clip": 1.04815578, + "balance_loss_mlp": 1.02369428, + "epoch": 0.8067095532976613, + "flos": 22305410340480.0, + "grad_norm": 1.9044400672115704, + "language_loss": 0.76562244, + "learning_rate": 3.791319497550558e-07, + "loss": 0.78757858, + "num_input_tokens_seen": 144783080, + "step": 6709, + "time_per_iteration": 2.8303768634796143 + }, + { + "auxiliary_loss_clip": 0.01139947, + "auxiliary_loss_mlp": 0.0088569, + "balance_loss_clip": 1.04577327, + "balance_loss_mlp": 1.00058246, + "epoch": 0.8068297961883004, + "flos": 17129498296320.0, + "grad_norm": 1.8160882142189154, + "language_loss": 0.70627296, + "learning_rate": 3.78675727813788e-07, + "loss": 0.72652936, + "num_input_tokens_seen": 144800645, + "step": 6710, + "time_per_iteration": 3.872958183288574 + }, + { + "auxiliary_loss_clip": 0.01149878, + "auxiliary_loss_mlp": 0.01029397, + "balance_loss_clip": 1.04598534, + "balance_loss_mlp": 1.02157736, + "epoch": 0.8069500390789395, + "flos": 22018843635840.0, + "grad_norm": 1.930647942857088, + "language_loss": 0.73440701, + "learning_rate": 3.782197518261225e-07, + "loss": 0.75619978, + "num_input_tokens_seen": 144820085, + "step": 6711, + "time_per_iteration": 2.869077444076538 + }, + { + "auxiliary_loss_clip": 0.01154019, + "auxiliary_loss_mlp": 0.01026472, + "balance_loss_clip": 1.04491258, + "balance_loss_mlp": 1.01887894, + "epoch": 0.8070702819695785, + "flos": 19244241567360.0, + "grad_norm": 2.0933167386567026, + "language_loss": 0.95704406, + "learning_rate": 3.777640218612319e-07, + "loss": 0.97884899, + "num_input_tokens_seen": 144838070, + "step": 6712, + "time_per_iteration": 3.738757371902466 + }, + { + "auxiliary_loss_clip": 0.0115872, + "auxiliary_loss_mlp": 0.01029783, + "balance_loss_clip": 1.04755974, + "balance_loss_mlp": 1.02173638, + "epoch": 0.8071905248602176, + "flos": 21544320038400.0, + "grad_norm": 2.3321505915192478, + "language_loss": 0.71883559, + "learning_rate": 3.773085379882488e-07, + "loss": 0.74072063, + "num_input_tokens_seen": 144857125, + "step": 6713, + "time_per_iteration": 3.6494176387786865 + }, + { + "auxiliary_loss_clip": 0.0116058, + "auxiliary_loss_mlp": 0.00886026, + "balance_loss_clip": 1.04362154, + "balance_loss_mlp": 1.00063896, + "epoch": 0.8073107677508568, + "flos": 37268309105280.0, + "grad_norm": 2.029015650907539, + "language_loss": 0.75955683, + "learning_rate": 3.768533002762715e-07, + "loss": 0.78002286, + "num_input_tokens_seen": 144880660, + "step": 6714, + "time_per_iteration": 2.7329800128936768 + }, + { + "auxiliary_loss_clip": 0.01146706, + "auxiliary_loss_mlp": 0.01026374, + "balance_loss_clip": 1.04068148, + "balance_loss_mlp": 1.01926911, + "epoch": 0.8074310106414958, + "flos": 28366269759360.0, + "grad_norm": 1.8833469331923764, + "language_loss": 0.77204525, + "learning_rate": 3.763983087943572e-07, + "loss": 0.79377604, + "num_input_tokens_seen": 144900050, + "step": 6715, + "time_per_iteration": 2.7253715991973877 + }, + { + "auxiliary_loss_clip": 0.01152102, + "auxiliary_loss_mlp": 0.00885368, + "balance_loss_clip": 1.04458547, + "balance_loss_mlp": 1.00059032, + "epoch": 0.8075512535321349, + "flos": 24281646768000.0, + "grad_norm": 1.8870997679860677, + "language_loss": 0.80804753, + "learning_rate": 3.759435636115282e-07, + "loss": 0.82842225, + "num_input_tokens_seen": 144920835, + "step": 6716, + "time_per_iteration": 2.5803871154785156 + }, + { + "auxiliary_loss_clip": 0.01114173, + "auxiliary_loss_mlp": 0.00885603, + "balance_loss_clip": 1.04117084, + "balance_loss_mlp": 1.00055432, + "epoch": 0.807671496422774, + "flos": 26030855283840.0, + "grad_norm": 1.886473644032921, + "language_loss": 0.73070991, + "learning_rate": 3.7548906479676967e-07, + "loss": 0.75070775, + "num_input_tokens_seen": 144940430, + "step": 6717, + "time_per_iteration": 2.7617475986480713 + }, + { + "auxiliary_loss_clip": 0.01164216, + "auxiliary_loss_mlp": 0.01030464, + "balance_loss_clip": 1.04587126, + "balance_loss_mlp": 1.02228582, + "epoch": 0.8077917393134131, + "flos": 23730740899200.0, + "grad_norm": 1.808620123634107, + "language_loss": 0.71426702, + "learning_rate": 3.7503481241902855e-07, + "loss": 0.7362138, + "num_input_tokens_seen": 144960405, + "step": 6718, + "time_per_iteration": 2.5438222885131836 + }, + { + "auxiliary_loss_clip": 0.01149284, + "auxiliary_loss_mlp": 0.00885091, + "balance_loss_clip": 1.04449141, + "balance_loss_mlp": 1.00048518, + "epoch": 0.8079119822040521, + "flos": 18402028398720.0, + "grad_norm": 1.71083510095696, + "language_loss": 0.7995649, + "learning_rate": 3.745808065472145e-07, + "loss": 0.81990862, + "num_input_tokens_seen": 144977700, + "step": 6719, + "time_per_iteration": 3.3914904594421387 + }, + { + "auxiliary_loss_clip": 0.01153729, + "auxiliary_loss_mlp": 0.01028203, + "balance_loss_clip": 1.04775381, + "balance_loss_mlp": 1.02037072, + "epoch": 0.8080322250946913, + "flos": 23621787970560.0, + "grad_norm": 1.6955609818632427, + "language_loss": 0.76484001, + "learning_rate": 3.741270472501994e-07, + "loss": 0.78665924, + "num_input_tokens_seen": 144998340, + "step": 6720, + "time_per_iteration": 2.6022720336914062 + }, + { + "auxiliary_loss_clip": 0.01150925, + "auxiliary_loss_mlp": 0.01025497, + "balance_loss_clip": 1.04766655, + "balance_loss_mlp": 1.01863325, + "epoch": 0.8081524679853304, + "flos": 22820692896000.0, + "grad_norm": 2.682860203692889, + "language_loss": 0.72618079, + "learning_rate": 3.736735345968183e-07, + "loss": 0.74794507, + "num_input_tokens_seen": 145017950, + "step": 6721, + "time_per_iteration": 2.8482019901275635 + }, + { + "auxiliary_loss_clip": 0.01160807, + "auxiliary_loss_mlp": 0.01031217, + "balance_loss_clip": 1.04666829, + "balance_loss_mlp": 1.0235157, + "epoch": 0.8082727108759694, + "flos": 17640004343040.0, + "grad_norm": 1.5333399217020476, + "language_loss": 0.78843904, + "learning_rate": 3.7322026865586986e-07, + "loss": 0.81035924, + "num_input_tokens_seen": 145036985, + "step": 6722, + "time_per_iteration": 2.834564685821533 + }, + { + "auxiliary_loss_clip": 0.01167594, + "auxiliary_loss_mlp": 0.01028379, + "balance_loss_clip": 1.04906881, + "balance_loss_mlp": 1.02017188, + "epoch": 0.8083929537666086, + "flos": 25958172113280.0, + "grad_norm": 2.2597225874776257, + "language_loss": 0.73474967, + "learning_rate": 3.7276724949611206e-07, + "loss": 0.75670946, + "num_input_tokens_seen": 145057095, + "step": 6723, + "time_per_iteration": 2.859388828277588 + }, + { + "auxiliary_loss_clip": 0.01155152, + "auxiliary_loss_mlp": 0.01024445, + "balance_loss_clip": 1.04677606, + "balance_loss_mlp": 1.01621342, + "epoch": 0.8085131966572476, + "flos": 27089178629760.0, + "grad_norm": 2.7289101199651205, + "language_loss": 0.75175416, + "learning_rate": 3.723144771862694e-07, + "loss": 0.77355015, + "num_input_tokens_seen": 145077735, + "step": 6724, + "time_per_iteration": 3.023319721221924 + }, + { + "auxiliary_loss_clip": 0.01145871, + "auxiliary_loss_mlp": 0.01027182, + "balance_loss_clip": 1.04393363, + "balance_loss_mlp": 1.01947474, + "epoch": 0.8086334395478867, + "flos": 23988543788160.0, + "grad_norm": 1.4851517481144598, + "language_loss": 0.76942199, + "learning_rate": 3.718619517950263e-07, + "loss": 0.79115248, + "num_input_tokens_seen": 145098330, + "step": 6725, + "time_per_iteration": 2.953160524368286 + }, + { + "auxiliary_loss_clip": 0.01169925, + "auxiliary_loss_mlp": 0.01032458, + "balance_loss_clip": 1.04869962, + "balance_loss_mlp": 1.02427435, + "epoch": 0.8087536824385259, + "flos": 20405879406720.0, + "grad_norm": 2.0701436879400625, + "language_loss": 0.76982206, + "learning_rate": 3.714096733910301e-07, + "loss": 0.79184592, + "num_input_tokens_seen": 145115855, + "step": 6726, + "time_per_iteration": 2.6603307723999023 + }, + { + "auxiliary_loss_clip": 0.01162468, + "auxiliary_loss_mlp": 0.01033515, + "balance_loss_clip": 1.04940212, + "balance_loss_mlp": 1.0245204, + "epoch": 0.8088739253291649, + "flos": 25919639798400.0, + "grad_norm": 1.9799538899810771, + "language_loss": 0.69732749, + "learning_rate": 3.709576420428926e-07, + "loss": 0.71928734, + "num_input_tokens_seen": 145136655, + "step": 6727, + "time_per_iteration": 2.6638224124908447 + }, + { + "auxiliary_loss_clip": 0.01154534, + "auxiliary_loss_mlp": 0.01029898, + "balance_loss_clip": 1.04427826, + "balance_loss_mlp": 1.0215416, + "epoch": 0.808994168219804, + "flos": 28402072640640.0, + "grad_norm": 2.1978750094147435, + "language_loss": 0.73329735, + "learning_rate": 3.7050585781918463e-07, + "loss": 0.75514162, + "num_input_tokens_seen": 145156955, + "step": 6728, + "time_per_iteration": 2.6580891609191895 + }, + { + "auxiliary_loss_clip": 0.01166598, + "auxiliary_loss_mlp": 0.01028796, + "balance_loss_clip": 1.04717159, + "balance_loss_mlp": 1.02060628, + "epoch": 0.8091144111104431, + "flos": 17421056991360.0, + "grad_norm": 1.8721334340075046, + "language_loss": 0.68903387, + "learning_rate": 3.700543207884428e-07, + "loss": 0.71098775, + "num_input_tokens_seen": 145173865, + "step": 6729, + "time_per_iteration": 2.6298694610595703 + }, + { + "auxiliary_loss_clip": 0.01159298, + "auxiliary_loss_mlp": 0.01025211, + "balance_loss_clip": 1.04730177, + "balance_loss_mlp": 1.01725364, + "epoch": 0.8092346540010822, + "flos": 32153803361280.0, + "grad_norm": 1.855899057609299, + "language_loss": 0.70899689, + "learning_rate": 3.6960303101916466e-07, + "loss": 0.73084199, + "num_input_tokens_seen": 145193780, + "step": 6730, + "time_per_iteration": 2.746795177459717 + }, + { + "auxiliary_loss_clip": 0.01056871, + "auxiliary_loss_mlp": 0.0087676, + "balance_loss_clip": 1.00554502, + "balance_loss_mlp": 1.00121164, + "epoch": 0.8093548968917212, + "flos": 58035093390720.0, + "grad_norm": 0.7441085572397257, + "language_loss": 0.55562997, + "learning_rate": 3.6915198857981047e-07, + "loss": 0.57496631, + "num_input_tokens_seen": 145258980, + "step": 6731, + "time_per_iteration": 3.2693939208984375 + }, + { + "auxiliary_loss_clip": 0.01137434, + "auxiliary_loss_mlp": 0.01023238, + "balance_loss_clip": 1.04300392, + "balance_loss_mlp": 1.01545358, + "epoch": 0.8094751397823604, + "flos": 27381599251200.0, + "grad_norm": 3.770184209843505, + "language_loss": 0.68271518, + "learning_rate": 3.687011935388027e-07, + "loss": 0.70432186, + "num_input_tokens_seen": 145281875, + "step": 6732, + "time_per_iteration": 2.7767207622528076 + }, + { + "auxiliary_loss_clip": 0.01160092, + "auxiliary_loss_mlp": 0.01029014, + "balance_loss_clip": 1.0464325, + "balance_loss_mlp": 1.0207113, + "epoch": 0.8095953826729995, + "flos": 24061083304320.0, + "grad_norm": 1.8938798184961365, + "language_loss": 0.7277565, + "learning_rate": 3.6825064596452646e-07, + "loss": 0.7496475, + "num_input_tokens_seen": 145302220, + "step": 6733, + "time_per_iteration": 2.777932643890381 + }, + { + "auxiliary_loss_clip": 0.0115967, + "auxiliary_loss_mlp": 0.01027151, + "balance_loss_clip": 1.04467666, + "balance_loss_mlp": 1.01983798, + "epoch": 0.8097156255636385, + "flos": 23951412103680.0, + "grad_norm": 1.72889179095167, + "language_loss": 0.70629889, + "learning_rate": 3.678003459253305e-07, + "loss": 0.72816706, + "num_input_tokens_seen": 145323070, + "step": 6734, + "time_per_iteration": 2.8130335807800293 + }, + { + "auxiliary_loss_clip": 0.01139827, + "auxiliary_loss_mlp": 0.01027276, + "balance_loss_clip": 1.04331052, + "balance_loss_mlp": 1.01967609, + "epoch": 0.8098358684542777, + "flos": 21799142098560.0, + "grad_norm": 2.0706109685886323, + "language_loss": 0.74745381, + "learning_rate": 3.673502934895236e-07, + "loss": 0.76912481, + "num_input_tokens_seen": 145342575, + "step": 6735, + "time_per_iteration": 2.8424320220947266 + }, + { + "auxiliary_loss_clip": 0.01056295, + "auxiliary_loss_mlp": 0.01002365, + "balance_loss_clip": 1.00524056, + "balance_loss_mlp": 1.00105977, + "epoch": 0.8099561113449167, + "flos": 68809515966720.0, + "grad_norm": 0.6963804570463379, + "language_loss": 0.57982832, + "learning_rate": 3.669004887253802e-07, + "loss": 0.60041493, + "num_input_tokens_seen": 145408865, + "step": 6736, + "time_per_iteration": 4.348278999328613 + }, + { + "auxiliary_loss_clip": 0.01157303, + "auxiliary_loss_mlp": 0.01030999, + "balance_loss_clip": 1.04788554, + "balance_loss_mlp": 1.02345347, + "epoch": 0.8100763542355558, + "flos": 23586056916480.0, + "grad_norm": 1.6416209929126047, + "language_loss": 0.78806257, + "learning_rate": 3.664509317011335e-07, + "loss": 0.80994558, + "num_input_tokens_seen": 145429200, + "step": 6737, + "time_per_iteration": 2.818411350250244 + }, + { + "auxiliary_loss_clip": 0.01162425, + "auxiliary_loss_mlp": 0.01027201, + "balance_loss_clip": 1.05037165, + "balance_loss_mlp": 1.01883233, + "epoch": 0.810196597126195, + "flos": 31650408207360.0, + "grad_norm": 1.9951335493551103, + "language_loss": 0.73658484, + "learning_rate": 3.6600162248498134e-07, + "loss": 0.75848103, + "num_input_tokens_seen": 145452830, + "step": 6738, + "time_per_iteration": 3.854339599609375 + }, + { + "auxiliary_loss_clip": 0.01107574, + "auxiliary_loss_mlp": 0.01022564, + "balance_loss_clip": 1.03925276, + "balance_loss_mlp": 1.01501834, + "epoch": 0.810316840016834, + "flos": 24900459298560.0, + "grad_norm": 2.277617416754832, + "language_loss": 0.76281917, + "learning_rate": 3.6555256114508426e-07, + "loss": 0.78412062, + "num_input_tokens_seen": 145472625, + "step": 6739, + "time_per_iteration": 3.82944917678833 + }, + { + "auxiliary_loss_clip": 0.01148481, + "auxiliary_loss_mlp": 0.01029564, + "balance_loss_clip": 1.0404563, + "balance_loss_mlp": 1.02166033, + "epoch": 0.8104370829074731, + "flos": 27965003950080.0, + "grad_norm": 1.8030651007100809, + "language_loss": 0.72568333, + "learning_rate": 3.651037477495642e-07, + "loss": 0.74746382, + "num_input_tokens_seen": 145494075, + "step": 6740, + "time_per_iteration": 2.840151309967041 + }, + { + "auxiliary_loss_clip": 0.01169988, + "auxiliary_loss_mlp": 0.01025598, + "balance_loss_clip": 1.04659581, + "balance_loss_mlp": 1.01731312, + "epoch": 0.8105573257981122, + "flos": 24640752988800.0, + "grad_norm": 2.118693328668492, + "language_loss": 0.6809755, + "learning_rate": 3.6465518236650584e-07, + "loss": 0.7029314, + "num_input_tokens_seen": 145514220, + "step": 6741, + "time_per_iteration": 2.676632881164551 + }, + { + "auxiliary_loss_clip": 0.01140779, + "auxiliary_loss_mlp": 0.01025605, + "balance_loss_clip": 1.04209399, + "balance_loss_mlp": 1.01835418, + "epoch": 0.8106775686887513, + "flos": 26358935132160.0, + "grad_norm": 1.6372867915696454, + "language_loss": 0.78297734, + "learning_rate": 3.642068650639558e-07, + "loss": 0.80464119, + "num_input_tokens_seen": 145533965, + "step": 6742, + "time_per_iteration": 2.8643829822540283 + }, + { + "auxiliary_loss_clip": 0.01143068, + "auxiliary_loss_mlp": 0.01031379, + "balance_loss_clip": 1.03995609, + "balance_loss_mlp": 1.02350783, + "epoch": 0.8107978115793903, + "flos": 27271892136960.0, + "grad_norm": 1.9283205597819002, + "language_loss": 0.64437151, + "learning_rate": 3.6375879590992334e-07, + "loss": 0.66611594, + "num_input_tokens_seen": 145554310, + "step": 6743, + "time_per_iteration": 2.796562671661377 + }, + { + "auxiliary_loss_clip": 0.01145071, + "auxiliary_loss_mlp": 0.01021331, + "balance_loss_clip": 1.04260719, + "balance_loss_mlp": 1.01379085, + "epoch": 0.8109180544700295, + "flos": 24934322845440.0, + "grad_norm": 1.8835124830588652, + "language_loss": 0.80977881, + "learning_rate": 3.6331097497238173e-07, + "loss": 0.83144283, + "num_input_tokens_seen": 145573755, + "step": 6744, + "time_per_iteration": 2.7963314056396484 + }, + { + "auxiliary_loss_clip": 0.01140115, + "auxiliary_loss_mlp": 0.01025901, + "balance_loss_clip": 1.04419446, + "balance_loss_mlp": 1.01874828, + "epoch": 0.8110382973606686, + "flos": 21105383840640.0, + "grad_norm": 2.816870513344445, + "language_loss": 0.80378973, + "learning_rate": 3.628634023192627e-07, + "loss": 0.82544982, + "num_input_tokens_seen": 145594000, + "step": 6745, + "time_per_iteration": 3.689183235168457 + }, + { + "auxiliary_loss_clip": 0.01163204, + "auxiliary_loss_mlp": 0.01024841, + "balance_loss_clip": 1.04705131, + "balance_loss_mlp": 1.01677084, + "epoch": 0.8111585402513076, + "flos": 15414081500160.0, + "grad_norm": 2.2007937729555644, + "language_loss": 0.75268334, + "learning_rate": 3.624160780184644e-07, + "loss": 0.77456379, + "num_input_tokens_seen": 145611215, + "step": 6746, + "time_per_iteration": 2.6741466522216797 + }, + { + "auxiliary_loss_clip": 0.01145796, + "auxiliary_loss_mlp": 0.01027403, + "balance_loss_clip": 1.04442048, + "balance_loss_mlp": 1.01958263, + "epoch": 0.8112787831419467, + "flos": 24095736950400.0, + "grad_norm": 1.7368817335066515, + "language_loss": 0.74626696, + "learning_rate": 3.6196900213784496e-07, + "loss": 0.76799893, + "num_input_tokens_seen": 145630530, + "step": 6747, + "time_per_iteration": 2.826136827468872 + }, + { + "auxiliary_loss_clip": 0.0116066, + "auxiliary_loss_mlp": 0.01027403, + "balance_loss_clip": 1.04518199, + "balance_loss_mlp": 1.01991665, + "epoch": 0.8113990260325858, + "flos": 20483374999680.0, + "grad_norm": 2.1693029409193265, + "language_loss": 0.86744648, + "learning_rate": 3.6152217474522527e-07, + "loss": 0.88932711, + "num_input_tokens_seen": 145647345, + "step": 6748, + "time_per_iteration": 2.671523332595825 + }, + { + "auxiliary_loss_clip": 0.01161221, + "auxiliary_loss_mlp": 0.01026656, + "balance_loss_clip": 1.04830551, + "balance_loss_mlp": 1.01887155, + "epoch": 0.8115192689232249, + "flos": 24901141656960.0, + "grad_norm": 1.5580671315445178, + "language_loss": 0.72758591, + "learning_rate": 3.6107559590838975e-07, + "loss": 0.74946463, + "num_input_tokens_seen": 145666330, + "step": 6749, + "time_per_iteration": 2.7824673652648926 + }, + { + "auxiliary_loss_clip": 0.01115029, + "auxiliary_loss_mlp": 0.01027193, + "balance_loss_clip": 1.03972018, + "balance_loss_mlp": 1.01887178, + "epoch": 0.811639511813864, + "flos": 24057204635520.0, + "grad_norm": 2.356589872052153, + "language_loss": 0.66272151, + "learning_rate": 3.606292656950822e-07, + "loss": 0.68414366, + "num_input_tokens_seen": 145684740, + "step": 6750, + "time_per_iteration": 2.837040901184082 + }, + { + "auxiliary_loss_clip": 0.01145393, + "auxiliary_loss_mlp": 0.01026646, + "balance_loss_clip": 1.04253602, + "balance_loss_mlp": 1.01852822, + "epoch": 0.8117597547045031, + "flos": 23185150243200.0, + "grad_norm": 2.4795418521999912, + "language_loss": 0.86816657, + "learning_rate": 3.601831841730121e-07, + "loss": 0.88988698, + "num_input_tokens_seen": 145702660, + "step": 6751, + "time_per_iteration": 2.833159923553467 + }, + { + "auxiliary_loss_clip": 0.01158098, + "auxiliary_loss_mlp": 0.01027416, + "balance_loss_clip": 1.04598069, + "balance_loss_mlp": 1.02019525, + "epoch": 0.8118799975951422, + "flos": 23040250778880.0, + "grad_norm": 1.5861055329702578, + "language_loss": 0.72426784, + "learning_rate": 3.5973735140984916e-07, + "loss": 0.74612296, + "num_input_tokens_seen": 145722830, + "step": 6752, + "time_per_iteration": 2.742847204208374 + }, + { + "auxiliary_loss_clip": 0.01120838, + "auxiliary_loss_mlp": 0.00885398, + "balance_loss_clip": 1.04049373, + "balance_loss_mlp": 1.00059533, + "epoch": 0.8120002404857812, + "flos": 24639962889600.0, + "grad_norm": 1.9361229985870527, + "language_loss": 0.79342473, + "learning_rate": 3.5929176747322607e-07, + "loss": 0.81348705, + "num_input_tokens_seen": 145741935, + "step": 6753, + "time_per_iteration": 2.937614917755127 + }, + { + "auxiliary_loss_clip": 0.01048418, + "auxiliary_loss_mlp": 0.01000965, + "balance_loss_clip": 1.00580335, + "balance_loss_mlp": 0.99957579, + "epoch": 0.8121204833764204, + "flos": 57415742156160.0, + "grad_norm": 0.8220318007447835, + "language_loss": 0.56241041, + "learning_rate": 3.588464324307372e-07, + "loss": 0.58290422, + "num_input_tokens_seen": 145805560, + "step": 6754, + "time_per_iteration": 3.4612154960632324 + }, + { + "auxiliary_loss_clip": 0.01160289, + "auxiliary_loss_mlp": 0.01021801, + "balance_loss_clip": 1.04269624, + "balance_loss_mlp": 1.01390922, + "epoch": 0.8122407262670595, + "flos": 19464589549440.0, + "grad_norm": 2.0206584884825145, + "language_loss": 0.75396037, + "learning_rate": 3.584013463499391e-07, + "loss": 0.77578133, + "num_input_tokens_seen": 145824180, + "step": 6755, + "time_per_iteration": 2.702244520187378 + }, + { + "auxiliary_loss_clip": 0.01048184, + "auxiliary_loss_mlp": 0.01001709, + "balance_loss_clip": 1.00889254, + "balance_loss_mlp": 1.00037372, + "epoch": 0.8123609691576985, + "flos": 56425325472000.0, + "grad_norm": 0.7309117270903431, + "language_loss": 0.64478743, + "learning_rate": 3.579565092983521e-07, + "loss": 0.66528642, + "num_input_tokens_seen": 145885300, + "step": 6756, + "time_per_iteration": 3.1933648586273193 + }, + { + "auxiliary_loss_clip": 0.01171897, + "auxiliary_loss_mlp": 0.01026019, + "balance_loss_clip": 1.04914427, + "balance_loss_mlp": 1.01807928, + "epoch": 0.8124812120483377, + "flos": 20631973564800.0, + "grad_norm": 15.822344388144424, + "language_loss": 0.83881867, + "learning_rate": 3.575119213434565e-07, + "loss": 0.86079782, + "num_input_tokens_seen": 145903815, + "step": 6757, + "time_per_iteration": 2.6100666522979736 + }, + { + "auxiliary_loss_clip": 0.0115849, + "auxiliary_loss_mlp": 0.01028127, + "balance_loss_clip": 1.04617476, + "balance_loss_mlp": 1.01998472, + "epoch": 0.8126014549389767, + "flos": 22492397566080.0, + "grad_norm": 1.772415639683999, + "language_loss": 0.81885284, + "learning_rate": 3.5706758255269765e-07, + "loss": 0.84071898, + "num_input_tokens_seen": 145922270, + "step": 6758, + "time_per_iteration": 2.8287103176116943 + }, + { + "auxiliary_loss_clip": 0.01155752, + "auxiliary_loss_mlp": 0.01030952, + "balance_loss_clip": 1.04550111, + "balance_loss_mlp": 1.02261949, + "epoch": 0.8127216978296158, + "flos": 23287961946240.0, + "grad_norm": 1.4706832535649408, + "language_loss": 0.69622183, + "learning_rate": 3.566234929934795e-07, + "loss": 0.71808892, + "num_input_tokens_seen": 145941470, + "step": 6759, + "time_per_iteration": 2.753504991531372 + }, + { + "auxiliary_loss_clip": 0.01157149, + "auxiliary_loss_mlp": 0.01026981, + "balance_loss_clip": 1.0474422, + "balance_loss_mlp": 1.01917887, + "epoch": 0.812841940720255, + "flos": 25154994049920.0, + "grad_norm": 1.5353279908660236, + "language_loss": 0.71841347, + "learning_rate": 3.561796527331706e-07, + "loss": 0.74025476, + "num_input_tokens_seen": 145963145, + "step": 6760, + "time_per_iteration": 2.7232182025909424 + }, + { + "auxiliary_loss_clip": 0.01142796, + "auxiliary_loss_mlp": 0.01030784, + "balance_loss_clip": 1.04303765, + "balance_loss_mlp": 1.02285981, + "epoch": 0.812962183610894, + "flos": 26648446752000.0, + "grad_norm": 1.8534788016908117, + "language_loss": 0.77418411, + "learning_rate": 3.5573606183910163e-07, + "loss": 0.7959199, + "num_input_tokens_seen": 145983150, + "step": 6761, + "time_per_iteration": 2.8370161056518555 + }, + { + "auxiliary_loss_clip": 0.01165129, + "auxiliary_loss_mlp": 0.01026668, + "balance_loss_clip": 1.04453242, + "balance_loss_mlp": 1.01849651, + "epoch": 0.8130824265015331, + "flos": 24966965329920.0, + "grad_norm": 2.0113908220926366, + "language_loss": 0.78612542, + "learning_rate": 3.5529272037856493e-07, + "loss": 0.80804336, + "num_input_tokens_seen": 146001365, + "step": 6762, + "time_per_iteration": 3.6428720951080322 + }, + { + "auxiliary_loss_clip": 0.01034403, + "auxiliary_loss_mlp": 0.01001552, + "balance_loss_clip": 1.00642657, + "balance_loss_mlp": 0.99989504, + "epoch": 0.8132026693921722, + "flos": 67622918175360.0, + "grad_norm": 0.7084162235910093, + "language_loss": 0.53843594, + "learning_rate": 3.548496284188149e-07, + "loss": 0.55879551, + "num_input_tokens_seen": 146061570, + "step": 6763, + "time_per_iteration": 3.4584238529205322 + }, + { + "auxiliary_loss_clip": 0.01125807, + "auxiliary_loss_mlp": 0.01024298, + "balance_loss_clip": 1.04241335, + "balance_loss_mlp": 1.01735961, + "epoch": 0.8133229122828113, + "flos": 19495149045120.0, + "grad_norm": 1.8157756761621378, + "language_loss": 0.79499477, + "learning_rate": 3.544067860270681e-07, + "loss": 0.81649584, + "num_input_tokens_seen": 146079145, + "step": 6764, + "time_per_iteration": 3.7423765659332275 + }, + { + "auxiliary_loss_clip": 0.01145834, + "auxiliary_loss_mlp": 0.01029331, + "balance_loss_clip": 1.04336798, + "balance_loss_mlp": 1.02120066, + "epoch": 0.8134431551734503, + "flos": 20668135582080.0, + "grad_norm": 1.8299043954519958, + "language_loss": 0.7082687, + "learning_rate": 3.539641932705029e-07, + "loss": 0.7300204, + "num_input_tokens_seen": 146097625, + "step": 6765, + "time_per_iteration": 3.7237401008605957 + }, + { + "auxiliary_loss_clip": 0.01174241, + "auxiliary_loss_mlp": 0.01028522, + "balance_loss_clip": 1.04877102, + "balance_loss_mlp": 1.01974225, + "epoch": 0.8135633980640895, + "flos": 21507332008320.0, + "grad_norm": 4.232057580027849, + "language_loss": 0.77709955, + "learning_rate": 3.53521850216262e-07, + "loss": 0.79912722, + "num_input_tokens_seen": 146117195, + "step": 6766, + "time_per_iteration": 2.665140151977539 + }, + { + "auxiliary_loss_clip": 0.01172418, + "auxiliary_loss_mlp": 0.01033384, + "balance_loss_clip": 1.04951692, + "balance_loss_mlp": 1.02528942, + "epoch": 0.8136836409547286, + "flos": 20554442058240.0, + "grad_norm": 1.838751539686784, + "language_loss": 0.76928306, + "learning_rate": 3.530797569314461e-07, + "loss": 0.79134107, + "num_input_tokens_seen": 146136220, + "step": 6767, + "time_per_iteration": 2.6981992721557617 + }, + { + "auxiliary_loss_clip": 0.01170619, + "auxiliary_loss_mlp": 0.01029111, + "balance_loss_clip": 1.04865742, + "balance_loss_mlp": 1.02079582, + "epoch": 0.8138038838453676, + "flos": 20299045380480.0, + "grad_norm": 1.9587322180532338, + "language_loss": 0.77581751, + "learning_rate": 3.5263791348312235e-07, + "loss": 0.79781485, + "num_input_tokens_seen": 146155415, + "step": 6768, + "time_per_iteration": 2.68794846534729 + }, + { + "auxiliary_loss_clip": 0.01149997, + "auxiliary_loss_mlp": 0.01023855, + "balance_loss_clip": 1.04347157, + "balance_loss_mlp": 1.01564109, + "epoch": 0.8139241267360068, + "flos": 29789840551680.0, + "grad_norm": 1.8466646639270605, + "language_loss": 0.70500392, + "learning_rate": 3.521963199383171e-07, + "loss": 0.72674251, + "num_input_tokens_seen": 146178370, + "step": 6769, + "time_per_iteration": 2.869183301925659 + }, + { + "auxiliary_loss_clip": 0.01131946, + "auxiliary_loss_mlp": 0.01025221, + "balance_loss_clip": 1.0424962, + "balance_loss_mlp": 1.01639974, + "epoch": 0.8140443696266458, + "flos": 19713270384000.0, + "grad_norm": 2.0836502838662745, + "language_loss": 0.76997709, + "learning_rate": 3.517549763640197e-07, + "loss": 0.79154873, + "num_input_tokens_seen": 146196010, + "step": 6770, + "time_per_iteration": 2.8430230617523193 + }, + { + "auxiliary_loss_clip": 0.01159151, + "auxiliary_loss_mlp": 0.00885834, + "balance_loss_clip": 1.04937387, + "balance_loss_mlp": 1.00050521, + "epoch": 0.8141646125172849, + "flos": 27160568910720.0, + "grad_norm": 1.767355118021529, + "language_loss": 0.71557826, + "learning_rate": 3.513138828271829e-07, + "loss": 0.73602819, + "num_input_tokens_seen": 146215880, + "step": 6771, + "time_per_iteration": 3.751099109649658 + }, + { + "auxiliary_loss_clip": 0.01133723, + "auxiliary_loss_mlp": 0.01027017, + "balance_loss_clip": 1.04187512, + "balance_loss_mlp": 1.01927471, + "epoch": 0.8142848554079241, + "flos": 39673102700160.0, + "grad_norm": 2.4772660256280727, + "language_loss": 0.70240629, + "learning_rate": 3.508730393947179e-07, + "loss": 0.72401369, + "num_input_tokens_seen": 146239135, + "step": 6772, + "time_per_iteration": 2.961970567703247 + }, + { + "auxiliary_loss_clip": 0.01138579, + "auxiliary_loss_mlp": 0.01027975, + "balance_loss_clip": 1.04250669, + "balance_loss_mlp": 1.01984525, + "epoch": 0.8144050982985631, + "flos": 22237288197120.0, + "grad_norm": 1.9973875340355085, + "language_loss": 0.71949536, + "learning_rate": 3.504324461335024e-07, + "loss": 0.74116093, + "num_input_tokens_seen": 146259245, + "step": 6773, + "time_per_iteration": 2.844827890396118 + }, + { + "auxiliary_loss_clip": 0.0112278, + "auxiliary_loss_mlp": 0.01028539, + "balance_loss_clip": 1.03966975, + "balance_loss_mlp": 1.02017045, + "epoch": 0.8145253411892022, + "flos": 23038239617280.0, + "grad_norm": 1.540726244039686, + "language_loss": 0.88358593, + "learning_rate": 3.499921031103732e-07, + "loss": 0.90509915, + "num_input_tokens_seen": 146280015, + "step": 6774, + "time_per_iteration": 2.8831403255462646 + }, + { + "auxiliary_loss_clip": 0.01145291, + "auxiliary_loss_mlp": 0.01027812, + "balance_loss_clip": 1.04376304, + "balance_loss_mlp": 1.020087, + "epoch": 0.8146455840798413, + "flos": 24827668387200.0, + "grad_norm": 4.480742322136781, + "language_loss": 0.78355497, + "learning_rate": 3.4955201039212987e-07, + "loss": 0.80528593, + "num_input_tokens_seen": 146300935, + "step": 6775, + "time_per_iteration": 2.779841184616089 + }, + { + "auxiliary_loss_clip": 0.01158039, + "auxiliary_loss_mlp": 0.0102973, + "balance_loss_clip": 1.04762006, + "balance_loss_mlp": 1.02161813, + "epoch": 0.8147658269704804, + "flos": 19974520978560.0, + "grad_norm": 2.401647008792956, + "language_loss": 0.6539886, + "learning_rate": 3.4911216804553465e-07, + "loss": 0.67586625, + "num_input_tokens_seen": 146319835, + "step": 6776, + "time_per_iteration": 2.7529356479644775 + }, + { + "auxiliary_loss_clip": 0.01150314, + "auxiliary_loss_mlp": 0.010306, + "balance_loss_clip": 1.04460859, + "balance_loss_mlp": 1.02203441, + "epoch": 0.8148860698611194, + "flos": 21178031097600.0, + "grad_norm": 1.9393137079107927, + "language_loss": 0.7011869, + "learning_rate": 3.4867257613731017e-07, + "loss": 0.72299606, + "num_input_tokens_seen": 146339030, + "step": 6777, + "time_per_iteration": 2.7406983375549316 + }, + { + "auxiliary_loss_clip": 0.01153401, + "auxiliary_loss_mlp": 0.01031082, + "balance_loss_clip": 1.04515839, + "balance_loss_mlp": 1.02339351, + "epoch": 0.8150063127517585, + "flos": 19606903234560.0, + "grad_norm": 1.6187257755736983, + "language_loss": 0.8571769, + "learning_rate": 3.4823323473414343e-07, + "loss": 0.87902164, + "num_input_tokens_seen": 146358550, + "step": 6778, + "time_per_iteration": 2.676478624343872 + }, + { + "auxiliary_loss_clip": 0.01142031, + "auxiliary_loss_mlp": 0.01029756, + "balance_loss_clip": 1.04430461, + "balance_loss_mlp": 1.02176285, + "epoch": 0.8151265556423977, + "flos": 22638374438400.0, + "grad_norm": 2.774150351484917, + "language_loss": 0.75896561, + "learning_rate": 3.477941439026812e-07, + "loss": 0.78068352, + "num_input_tokens_seen": 146376770, + "step": 6779, + "time_per_iteration": 2.8499741554260254 + }, + { + "auxiliary_loss_clip": 0.01142467, + "auxiliary_loss_mlp": 0.01029692, + "balance_loss_clip": 1.04570234, + "balance_loss_mlp": 1.02237308, + "epoch": 0.8152467985330367, + "flos": 17968048277760.0, + "grad_norm": 1.7275258042141592, + "language_loss": 0.7313956, + "learning_rate": 3.473553037095349e-07, + "loss": 0.7531172, + "num_input_tokens_seen": 146395795, + "step": 6780, + "time_per_iteration": 2.762838125228882 + }, + { + "auxiliary_loss_clip": 0.01142824, + "auxiliary_loss_mlp": 0.01024701, + "balance_loss_clip": 1.04206049, + "balance_loss_mlp": 1.0170716, + "epoch": 0.8153670414236758, + "flos": 24969012405120.0, + "grad_norm": 1.7462310139682116, + "language_loss": 0.82947063, + "learning_rate": 3.469167142212743e-07, + "loss": 0.85114592, + "num_input_tokens_seen": 146417640, + "step": 6781, + "time_per_iteration": 2.865711212158203 + }, + { + "auxiliary_loss_clip": 0.01161396, + "auxiliary_loss_mlp": 0.01030103, + "balance_loss_clip": 1.04657221, + "balance_loss_mlp": 1.02163887, + "epoch": 0.8154872843143149, + "flos": 31066069754880.0, + "grad_norm": 2.2876296734671535, + "language_loss": 0.62769866, + "learning_rate": 3.4647837550443337e-07, + "loss": 0.64961362, + "num_input_tokens_seen": 146436205, + "step": 6782, + "time_per_iteration": 2.776331901550293 + }, + { + "auxiliary_loss_clip": 0.01142497, + "auxiliary_loss_mlp": 0.01027064, + "balance_loss_clip": 1.04280066, + "balance_loss_mlp": 1.01922584, + "epoch": 0.815607527204954, + "flos": 19391654983680.0, + "grad_norm": 2.046491277445308, + "language_loss": 0.74523199, + "learning_rate": 3.460402876255086e-07, + "loss": 0.7669276, + "num_input_tokens_seen": 146453595, + "step": 6783, + "time_per_iteration": 2.835191011428833 + }, + { + "auxiliary_loss_clip": 0.01162366, + "auxiliary_loss_mlp": 0.01028755, + "balance_loss_clip": 1.04694366, + "balance_loss_mlp": 1.0213933, + "epoch": 0.815727770095593, + "flos": 26140418743680.0, + "grad_norm": 1.9690555885366485, + "language_loss": 0.72121775, + "learning_rate": 3.456024506509574e-07, + "loss": 0.74312901, + "num_input_tokens_seen": 146474515, + "step": 6784, + "time_per_iteration": 2.7340211868286133 + }, + { + "auxiliary_loss_clip": 0.01162336, + "auxiliary_loss_mlp": 0.00885864, + "balance_loss_clip": 1.04937363, + "balance_loss_mlp": 1.00057423, + "epoch": 0.8158480129862322, + "flos": 25337527989120.0, + "grad_norm": 4.59155107730472, + "language_loss": 0.73821247, + "learning_rate": 3.4516486464719873e-07, + "loss": 0.75869453, + "num_input_tokens_seen": 146493905, + "step": 6785, + "time_per_iteration": 2.7800254821777344 + }, + { + "auxiliary_loss_clip": 0.01126997, + "auxiliary_loss_mlp": 0.01027281, + "balance_loss_clip": 1.04196978, + "balance_loss_mlp": 1.01940703, + "epoch": 0.8159682558768713, + "flos": 34423645559040.0, + "grad_norm": 2.084706541187286, + "language_loss": 0.62214285, + "learning_rate": 3.4472752968061445e-07, + "loss": 0.64368564, + "num_input_tokens_seen": 146518335, + "step": 6786, + "time_per_iteration": 2.9126133918762207 + }, + { + "auxiliary_loss_clip": 0.01160595, + "auxiliary_loss_mlp": 0.01028678, + "balance_loss_clip": 1.04491043, + "balance_loss_mlp": 1.02089977, + "epoch": 0.8160884987675103, + "flos": 18653223185280.0, + "grad_norm": 1.775627763461335, + "language_loss": 0.73487765, + "learning_rate": 3.442904458175475e-07, + "loss": 0.75677037, + "num_input_tokens_seen": 146535655, + "step": 6787, + "time_per_iteration": 2.6935720443725586 + }, + { + "auxiliary_loss_clip": 0.01156981, + "auxiliary_loss_mlp": 0.01028949, + "balance_loss_clip": 1.04341936, + "balance_loss_mlp": 1.02149224, + "epoch": 0.8162087416581495, + "flos": 31430527102080.0, + "grad_norm": 1.520052563202923, + "language_loss": 0.75936908, + "learning_rate": 3.438536131243044e-07, + "loss": 0.78122836, + "num_input_tokens_seen": 146556815, + "step": 6788, + "time_per_iteration": 3.7794322967529297 + }, + { + "auxiliary_loss_clip": 0.01155596, + "auxiliary_loss_mlp": 0.01032736, + "balance_loss_clip": 1.04528427, + "balance_loss_mlp": 1.02387261, + "epoch": 0.8163289845487885, + "flos": 37593910915200.0, + "grad_norm": 3.9713559865538812, + "language_loss": 0.62509161, + "learning_rate": 3.434170316671503e-07, + "loss": 0.64697492, + "num_input_tokens_seen": 146581845, + "step": 6789, + "time_per_iteration": 2.840437173843384 + }, + { + "auxiliary_loss_clip": 0.01135193, + "auxiliary_loss_mlp": 0.01023649, + "balance_loss_clip": 1.04610443, + "balance_loss_mlp": 1.01633263, + "epoch": 0.8164492274394276, + "flos": 13953989554560.0, + "grad_norm": 4.751691521577746, + "language_loss": 0.8946631, + "learning_rate": 3.4298070151231583e-07, + "loss": 0.91625148, + "num_input_tokens_seen": 146597245, + "step": 6790, + "time_per_iteration": 3.783560276031494 + }, + { + "auxiliary_loss_clip": 0.01155539, + "auxiliary_loss_mlp": 0.01026364, + "balance_loss_clip": 1.04528546, + "balance_loss_mlp": 1.01900923, + "epoch": 0.8165694703300668, + "flos": 28986554747520.0, + "grad_norm": 1.8107013189264514, + "language_loss": 0.59852004, + "learning_rate": 3.425446227259916e-07, + "loss": 0.62033904, + "num_input_tokens_seen": 146618210, + "step": 6791, + "time_per_iteration": 3.8559153079986572 + }, + { + "auxiliary_loss_clip": 0.0115264, + "auxiliary_loss_mlp": 0.01022623, + "balance_loss_clip": 1.04489183, + "balance_loss_mlp": 1.01483858, + "epoch": 0.8166897132207058, + "flos": 25118365155840.0, + "grad_norm": 2.0810956633128295, + "language_loss": 0.82695234, + "learning_rate": 3.421087953743296e-07, + "loss": 0.84870505, + "num_input_tokens_seen": 146637975, + "step": 6792, + "time_per_iteration": 2.7378337383270264 + }, + { + "auxiliary_loss_clip": 0.01159176, + "auxiliary_loss_mlp": 0.01029072, + "balance_loss_clip": 1.04320848, + "balance_loss_mlp": 1.02099562, + "epoch": 0.8168099561113449, + "flos": 23148593176320.0, + "grad_norm": 2.1325461577169107, + "language_loss": 0.79719377, + "learning_rate": 3.416732195234464e-07, + "loss": 0.8190763, + "num_input_tokens_seen": 146658030, + "step": 6793, + "time_per_iteration": 2.7852258682250977 + }, + { + "auxiliary_loss_clip": 0.01164076, + "auxiliary_loss_mlp": 0.01023001, + "balance_loss_clip": 1.04600763, + "balance_loss_mlp": 1.01512742, + "epoch": 0.816930199001984, + "flos": 18407666833920.0, + "grad_norm": 1.4354325772881857, + "language_loss": 0.79572624, + "learning_rate": 3.4123789523941613e-07, + "loss": 0.81759697, + "num_input_tokens_seen": 146677855, + "step": 6794, + "time_per_iteration": 2.7138359546661377 + }, + { + "auxiliary_loss_clip": 0.01154482, + "auxiliary_loss_mlp": 0.01028633, + "balance_loss_clip": 1.0436573, + "balance_loss_mlp": 1.02061021, + "epoch": 0.8170504418926231, + "flos": 21251324799360.0, + "grad_norm": 1.5709980707839877, + "language_loss": 0.63557684, + "learning_rate": 3.4080282258827884e-07, + "loss": 0.657408, + "num_input_tokens_seen": 146696230, + "step": 6795, + "time_per_iteration": 2.7390003204345703 + }, + { + "auxiliary_loss_clip": 0.01162165, + "auxiliary_loss_mlp": 0.01024471, + "balance_loss_clip": 1.04583192, + "balance_loss_mlp": 1.01700234, + "epoch": 0.8171706847832622, + "flos": 19099234362240.0, + "grad_norm": 2.09968785790556, + "language_loss": 0.72335625, + "learning_rate": 3.403680016360342e-07, + "loss": 0.74522257, + "num_input_tokens_seen": 146714835, + "step": 6796, + "time_per_iteration": 2.7875685691833496 + }, + { + "auxiliary_loss_clip": 0.01154501, + "auxiliary_loss_mlp": 0.01023238, + "balance_loss_clip": 1.04644561, + "balance_loss_mlp": 1.01503062, + "epoch": 0.8172909276739013, + "flos": 21470128496640.0, + "grad_norm": 5.249983620426196, + "language_loss": 0.67569673, + "learning_rate": 3.3993343244864403e-07, + "loss": 0.69747412, + "num_input_tokens_seen": 146734425, + "step": 6797, + "time_per_iteration": 2.657111406326294 + }, + { + "auxiliary_loss_clip": 0.01157028, + "auxiliary_loss_mlp": 0.01023556, + "balance_loss_clip": 1.04484296, + "balance_loss_mlp": 1.01585782, + "epoch": 0.8174111705645404, + "flos": 27599792417280.0, + "grad_norm": 1.5721169501932875, + "language_loss": 0.72529471, + "learning_rate": 3.394991150920323e-07, + "loss": 0.74710059, + "num_input_tokens_seen": 146757545, + "step": 6798, + "time_per_iteration": 3.744391679763794 + }, + { + "auxiliary_loss_clip": 0.01134635, + "auxiliary_loss_mlp": 0.00886431, + "balance_loss_clip": 1.04263854, + "balance_loss_mlp": 1.00061226, + "epoch": 0.8175314134551794, + "flos": 14064594508800.0, + "grad_norm": 1.9621228813280438, + "language_loss": 0.74645704, + "learning_rate": 3.3906504963208396e-07, + "loss": 0.76666772, + "num_input_tokens_seen": 146774240, + "step": 6799, + "time_per_iteration": 2.8059184551239014 + }, + { + "auxiliary_loss_clip": 0.01127865, + "auxiliary_loss_mlp": 0.01030522, + "balance_loss_clip": 1.04443061, + "balance_loss_mlp": 1.02172399, + "epoch": 0.8176516563458186, + "flos": 22708076780160.0, + "grad_norm": 1.8707461659701348, + "language_loss": 0.66464418, + "learning_rate": 3.3863123613464774e-07, + "loss": 0.68622804, + "num_input_tokens_seen": 146793140, + "step": 6800, + "time_per_iteration": 2.904254674911499 + }, + { + "auxiliary_loss_clip": 0.01152149, + "auxiliary_loss_mlp": 0.01028784, + "balance_loss_clip": 1.0411818, + "balance_loss_mlp": 1.02095795, + "epoch": 0.8177718992364577, + "flos": 21945406279680.0, + "grad_norm": 1.7422116274279678, + "language_loss": 0.75024748, + "learning_rate": 3.381976746655317e-07, + "loss": 0.77205676, + "num_input_tokens_seen": 146812895, + "step": 6801, + "time_per_iteration": 2.8486056327819824 + }, + { + "auxiliary_loss_clip": 0.01124596, + "auxiliary_loss_mlp": 0.01026006, + "balance_loss_clip": 1.04386401, + "balance_loss_mlp": 1.01819229, + "epoch": 0.8178921421270967, + "flos": 22017443005440.0, + "grad_norm": 15.660433716998664, + "language_loss": 0.67494649, + "learning_rate": 3.3776436529050756e-07, + "loss": 0.6964525, + "num_input_tokens_seen": 146832445, + "step": 6802, + "time_per_iteration": 2.7733347415924072 + }, + { + "auxiliary_loss_clip": 0.0116717, + "auxiliary_loss_mlp": 0.01030844, + "balance_loss_clip": 1.04554176, + "balance_loss_mlp": 1.023018, + "epoch": 0.8180123850177359, + "flos": 33183111496320.0, + "grad_norm": 1.6961701912695812, + "language_loss": 0.72497743, + "learning_rate": 3.373313080753073e-07, + "loss": 0.74695754, + "num_input_tokens_seen": 146856505, + "step": 6803, + "time_per_iteration": 2.882972240447998 + }, + { + "auxiliary_loss_clip": 0.01157158, + "auxiliary_loss_mlp": 0.01025676, + "balance_loss_clip": 1.044945, + "balance_loss_mlp": 1.01731896, + "epoch": 0.8181326279083749, + "flos": 22091167670400.0, + "grad_norm": 1.5327161157290012, + "language_loss": 0.77645004, + "learning_rate": 3.3689850308562527e-07, + "loss": 0.79827833, + "num_input_tokens_seen": 146876950, + "step": 6804, + "time_per_iteration": 2.6475424766540527 + }, + { + "auxiliary_loss_clip": 0.01124937, + "auxiliary_loss_mlp": 0.01025533, + "balance_loss_clip": 1.0448482, + "balance_loss_mlp": 1.01805246, + "epoch": 0.818252870799014, + "flos": 15705747936000.0, + "grad_norm": 5.898133704119684, + "language_loss": 0.7786805, + "learning_rate": 3.364659503871183e-07, + "loss": 0.8001852, + "num_input_tokens_seen": 146894885, + "step": 6805, + "time_per_iteration": 2.844888210296631 + }, + { + "auxiliary_loss_clip": 0.01137818, + "auxiliary_loss_mlp": 0.01024067, + "balance_loss_clip": 1.0411787, + "balance_loss_mlp": 1.01729298, + "epoch": 0.8183731136896532, + "flos": 18770687637120.0, + "grad_norm": 2.0481282226651687, + "language_loss": 0.83951473, + "learning_rate": 3.3603365004540417e-07, + "loss": 0.86113358, + "num_input_tokens_seen": 146913180, + "step": 6806, + "time_per_iteration": 2.7449657917022705 + }, + { + "auxiliary_loss_clip": 0.01169946, + "auxiliary_loss_mlp": 0.01025039, + "balance_loss_clip": 1.04892349, + "balance_loss_mlp": 1.01758218, + "epoch": 0.8184933565802922, + "flos": 26541792293760.0, + "grad_norm": 1.8954883356583079, + "language_loss": 0.7693463, + "learning_rate": 3.356016021260624e-07, + "loss": 0.79129612, + "num_input_tokens_seen": 146933510, + "step": 6807, + "time_per_iteration": 2.681753158569336 + }, + { + "auxiliary_loss_clip": 0.01165418, + "auxiliary_loss_mlp": 0.01030305, + "balance_loss_clip": 1.04952061, + "balance_loss_mlp": 1.02202618, + "epoch": 0.8186135994709313, + "flos": 17530117660800.0, + "grad_norm": 5.772407138637987, + "language_loss": 0.65105242, + "learning_rate": 3.35169806694634e-07, + "loss": 0.67300963, + "num_input_tokens_seen": 146951760, + "step": 6808, + "time_per_iteration": 2.702373504638672 + }, + { + "auxiliary_loss_clip": 0.01042249, + "auxiliary_loss_mlp": 0.01002064, + "balance_loss_clip": 1.0082283, + "balance_loss_mlp": 1.00047803, + "epoch": 0.8187338423615703, + "flos": 63480300675840.0, + "grad_norm": 0.7487009858172237, + "language_loss": 0.60680753, + "learning_rate": 3.3473826381662186e-07, + "loss": 0.62725061, + "num_input_tokens_seen": 147022900, + "step": 6809, + "time_per_iteration": 3.436190366744995 + }, + { + "auxiliary_loss_clip": 0.01156342, + "auxiliary_loss_mlp": 0.01027695, + "balance_loss_clip": 1.04659522, + "balance_loss_mlp": 1.02002144, + "epoch": 0.8188540852522095, + "flos": 17529974006400.0, + "grad_norm": 2.0947183074164033, + "language_loss": 0.81487387, + "learning_rate": 3.3430697355749216e-07, + "loss": 0.83671427, + "num_input_tokens_seen": 147040590, + "step": 6810, + "time_per_iteration": 2.6912527084350586 + }, + { + "auxiliary_loss_clip": 0.01127988, + "auxiliary_loss_mlp": 0.010279, + "balance_loss_clip": 1.04156327, + "balance_loss_mlp": 1.0197345, + "epoch": 0.8189743281428485, + "flos": 14392530702720.0, + "grad_norm": 1.892849674726773, + "language_loss": 0.75016296, + "learning_rate": 3.3387593598266907e-07, + "loss": 0.77172184, + "num_input_tokens_seen": 147057200, + "step": 6811, + "time_per_iteration": 2.806729316711426 + }, + { + "auxiliary_loss_clip": 0.01131259, + "auxiliary_loss_mlp": 0.01029082, + "balance_loss_clip": 1.03842139, + "balance_loss_mlp": 1.02149415, + "epoch": 0.8190945710334876, + "flos": 25080479285760.0, + "grad_norm": 1.6344847090668573, + "language_loss": 0.78271806, + "learning_rate": 3.3344515115754225e-07, + "loss": 0.80432141, + "num_input_tokens_seen": 147076180, + "step": 6812, + "time_per_iteration": 2.850825786590576 + }, + { + "auxiliary_loss_clip": 0.01135891, + "auxiliary_loss_mlp": 0.01023689, + "balance_loss_clip": 1.04206991, + "balance_loss_mlp": 1.01627445, + "epoch": 0.8192148139241268, + "flos": 21507152440320.0, + "grad_norm": 2.227016914561257, + "language_loss": 0.80256605, + "learning_rate": 3.33014619147461e-07, + "loss": 0.82416189, + "num_input_tokens_seen": 147094205, + "step": 6813, + "time_per_iteration": 2.756610870361328 + }, + { + "auxiliary_loss_clip": 0.01151742, + "auxiliary_loss_mlp": 0.0103074, + "balance_loss_clip": 1.0471307, + "balance_loss_mlp": 1.02240682, + "epoch": 0.8193350568147658, + "flos": 23952166289280.0, + "grad_norm": 2.084028699714716, + "language_loss": 0.71817183, + "learning_rate": 3.325843400177362e-07, + "loss": 0.73999667, + "num_input_tokens_seen": 147115545, + "step": 6814, + "time_per_iteration": 3.661276340484619 + }, + { + "auxiliary_loss_clip": 0.01163573, + "auxiliary_loss_mlp": 0.00886068, + "balance_loss_clip": 1.04546428, + "balance_loss_mlp": 1.00049305, + "epoch": 0.8194552997054049, + "flos": 20559469962240.0, + "grad_norm": 1.7274980401547924, + "language_loss": 0.73626113, + "learning_rate": 3.32154313833642e-07, + "loss": 0.75675756, + "num_input_tokens_seen": 147135700, + "step": 6815, + "time_per_iteration": 2.80987286567688 + }, + { + "auxiliary_loss_clip": 0.01172403, + "auxiliary_loss_mlp": 0.01034284, + "balance_loss_clip": 1.04878831, + "balance_loss_mlp": 1.02603769, + "epoch": 0.819575542596044, + "flos": 26031753123840.0, + "grad_norm": 2.438047943479732, + "language_loss": 0.59335876, + "learning_rate": 3.3172454066041164e-07, + "loss": 0.61542559, + "num_input_tokens_seen": 147155205, + "step": 6816, + "time_per_iteration": 3.6799864768981934 + }, + { + "auxiliary_loss_clip": 0.0111932, + "auxiliary_loss_mlp": 0.00884644, + "balance_loss_clip": 1.04209757, + "balance_loss_mlp": 1.00052583, + "epoch": 0.8196957854866831, + "flos": 29096944220160.0, + "grad_norm": 1.9842214824858637, + "language_loss": 0.7597121, + "learning_rate": 3.3129502056324234e-07, + "loss": 0.77975172, + "num_input_tokens_seen": 147176570, + "step": 6817, + "time_per_iteration": 3.913295269012451 + }, + { + "auxiliary_loss_clip": 0.01028337, + "auxiliary_loss_mlp": 0.0100035, + "balance_loss_clip": 1.01297951, + "balance_loss_mlp": 0.99872857, + "epoch": 0.8198160283773221, + "flos": 69033631898880.0, + "grad_norm": 1.0500122331617094, + "language_loss": 0.59762621, + "learning_rate": 3.3086575360729165e-07, + "loss": 0.61791313, + "num_input_tokens_seen": 147234105, + "step": 6818, + "time_per_iteration": 3.4332196712493896 + }, + { + "auxiliary_loss_clip": 0.01147217, + "auxiliary_loss_mlp": 0.01032105, + "balance_loss_clip": 1.04405665, + "balance_loss_mlp": 1.02380252, + "epoch": 0.8199362712679613, + "flos": 16618058496000.0, + "grad_norm": 2.0408598540228566, + "language_loss": 0.71529448, + "learning_rate": 3.3043673985767906e-07, + "loss": 0.73708767, + "num_input_tokens_seen": 147253170, + "step": 6819, + "time_per_iteration": 3.005835771560669 + }, + { + "auxiliary_loss_clip": 0.01130208, + "auxiliary_loss_mlp": 0.01026451, + "balance_loss_clip": 1.03983402, + "balance_loss_mlp": 1.01819611, + "epoch": 0.8200565141586004, + "flos": 21757664868480.0, + "grad_norm": 1.722713013527094, + "language_loss": 0.77601546, + "learning_rate": 3.3000797937948564e-07, + "loss": 0.79758203, + "num_input_tokens_seen": 147271465, + "step": 6820, + "time_per_iteration": 2.848019599914551 + }, + { + "auxiliary_loss_clip": 0.01032254, + "auxiliary_loss_mlp": 0.01005001, + "balance_loss_clip": 1.00579476, + "balance_loss_mlp": 1.00346339, + "epoch": 0.8201767570492394, + "flos": 69807112392960.0, + "grad_norm": 0.936249340516657, + "language_loss": 0.65044045, + "learning_rate": 3.295794722377534e-07, + "loss": 0.67081308, + "num_input_tokens_seen": 147335070, + "step": 6821, + "time_per_iteration": 3.359116792678833 + }, + { + "auxiliary_loss_clip": 0.01165852, + "auxiliary_loss_mlp": 0.01023605, + "balance_loss_clip": 1.04454207, + "balance_loss_mlp": 1.01613641, + "epoch": 0.8202969999398786, + "flos": 23111892455040.0, + "grad_norm": 1.5474560986967776, + "language_loss": 0.79991049, + "learning_rate": 3.291512184974876e-07, + "loss": 0.82180506, + "num_input_tokens_seen": 147355460, + "step": 6822, + "time_per_iteration": 2.6694540977478027 + }, + { + "auxiliary_loss_clip": 0.01149064, + "auxiliary_loss_mlp": 0.01031371, + "balance_loss_clip": 1.04253018, + "balance_loss_mlp": 1.02345598, + "epoch": 0.8204172428305176, + "flos": 28220616109440.0, + "grad_norm": 1.6568841166446242, + "language_loss": 0.66366917, + "learning_rate": 3.2872321822365346e-07, + "loss": 0.6854735, + "num_input_tokens_seen": 147375675, + "step": 6823, + "time_per_iteration": 3.858959913253784 + }, + { + "auxiliary_loss_clip": 0.0115796, + "auxiliary_loss_mlp": 0.01032333, + "balance_loss_clip": 1.04691362, + "balance_loss_mlp": 1.02447069, + "epoch": 0.8205374857211567, + "flos": 20887011106560.0, + "grad_norm": 1.749634217231541, + "language_loss": 0.73356581, + "learning_rate": 3.282954714811783e-07, + "loss": 0.75546873, + "num_input_tokens_seen": 147394580, + "step": 6824, + "time_per_iteration": 2.7430574893951416 + }, + { + "auxiliary_loss_clip": 0.01138376, + "auxiliary_loss_mlp": 0.01025436, + "balance_loss_clip": 1.04019535, + "balance_loss_mlp": 1.01750219, + "epoch": 0.8206577286117959, + "flos": 13152140294400.0, + "grad_norm": 2.315736263332174, + "language_loss": 0.71017098, + "learning_rate": 3.2786797833495093e-07, + "loss": 0.73180902, + "num_input_tokens_seen": 147409935, + "step": 6825, + "time_per_iteration": 2.7155842781066895 + }, + { + "auxiliary_loss_clip": 0.01167725, + "auxiliary_loss_mlp": 0.01025097, + "balance_loss_clip": 1.0464617, + "balance_loss_mlp": 1.01743746, + "epoch": 0.8207779715024349, + "flos": 25265634917760.0, + "grad_norm": 1.8514642310001597, + "language_loss": 0.73051131, + "learning_rate": 3.274407388498213e-07, + "loss": 0.7524395, + "num_input_tokens_seen": 147428065, + "step": 6826, + "time_per_iteration": 2.740401029586792 + }, + { + "auxiliary_loss_clip": 0.01136694, + "auxiliary_loss_mlp": 0.01025576, + "balance_loss_clip": 1.04203963, + "balance_loss_mlp": 1.01759458, + "epoch": 0.820898214393074, + "flos": 19610243199360.0, + "grad_norm": 1.803771853138431, + "language_loss": 0.74119282, + "learning_rate": 3.270137530906021e-07, + "loss": 0.76281559, + "num_input_tokens_seen": 147447300, + "step": 6827, + "time_per_iteration": 2.766890287399292 + }, + { + "auxiliary_loss_clip": 0.01122974, + "auxiliary_loss_mlp": 0.01024366, + "balance_loss_clip": 1.04419994, + "balance_loss_mlp": 1.01688302, + "epoch": 0.8210184572837131, + "flos": 15596615439360.0, + "grad_norm": 1.7512493877709434, + "language_loss": 0.83713114, + "learning_rate": 3.265870211220665e-07, + "loss": 0.85860455, + "num_input_tokens_seen": 147465135, + "step": 6828, + "time_per_iteration": 2.8680906295776367 + }, + { + "auxiliary_loss_clip": 0.01136562, + "auxiliary_loss_mlp": 0.01027297, + "balance_loss_clip": 1.04448867, + "balance_loss_mlp": 1.01902401, + "epoch": 0.8211387001743522, + "flos": 20813932886400.0, + "grad_norm": 2.502921895951799, + "language_loss": 0.81810111, + "learning_rate": 3.2616054300894934e-07, + "loss": 0.83973968, + "num_input_tokens_seen": 147484585, + "step": 6829, + "time_per_iteration": 2.797886610031128 + }, + { + "auxiliary_loss_clip": 0.01138741, + "auxiliary_loss_mlp": 0.01027243, + "balance_loss_clip": 1.04268754, + "balance_loss_mlp": 1.019068, + "epoch": 0.8212589430649913, + "flos": 27704579368320.0, + "grad_norm": 2.5362424086597932, + "language_loss": 0.84019339, + "learning_rate": 3.2573431881594693e-07, + "loss": 0.86185318, + "num_input_tokens_seen": 147504130, + "step": 6830, + "time_per_iteration": 2.746410846710205 + }, + { + "auxiliary_loss_clip": 0.0112148, + "auxiliary_loss_mlp": 0.01031351, + "balance_loss_clip": 1.0391314, + "balance_loss_mlp": 1.02314901, + "epoch": 0.8213791859556304, + "flos": 22455625017600.0, + "grad_norm": 2.3342969586256705, + "language_loss": 0.66024584, + "learning_rate": 3.2530834860771663e-07, + "loss": 0.68177414, + "num_input_tokens_seen": 147523510, + "step": 6831, + "time_per_iteration": 2.886148691177368 + }, + { + "auxiliary_loss_clip": 0.01160762, + "auxiliary_loss_mlp": 0.01026169, + "balance_loss_clip": 1.04581654, + "balance_loss_mlp": 1.01790166, + "epoch": 0.8214994288462695, + "flos": 16654471908480.0, + "grad_norm": 1.9069800665748027, + "language_loss": 0.74115133, + "learning_rate": 3.248826324488794e-07, + "loss": 0.76302063, + "num_input_tokens_seen": 147540805, + "step": 6832, + "time_per_iteration": 2.745124101638794 + }, + { + "auxiliary_loss_clip": 0.01169579, + "auxiliary_loss_mlp": 0.01023355, + "balance_loss_clip": 1.05006242, + "balance_loss_mlp": 1.01582444, + "epoch": 0.8216196717369085, + "flos": 25221787390080.0, + "grad_norm": 1.7252389164044224, + "language_loss": 0.87752283, + "learning_rate": 3.244571704040138e-07, + "loss": 0.89945221, + "num_input_tokens_seen": 147560965, + "step": 6833, + "time_per_iteration": 2.7322659492492676 + }, + { + "auxiliary_loss_clip": 0.01157219, + "auxiliary_loss_mlp": 0.01027993, + "balance_loss_clip": 1.04302168, + "balance_loss_mlp": 1.02001214, + "epoch": 0.8217399146275477, + "flos": 25371930240000.0, + "grad_norm": 1.8078071923460388, + "language_loss": 0.7345587, + "learning_rate": 3.2403196253766374e-07, + "loss": 0.75641084, + "num_input_tokens_seen": 147580045, + "step": 6834, + "time_per_iteration": 2.648326873779297 + }, + { + "auxiliary_loss_clip": 0.01157202, + "auxiliary_loss_mlp": 0.01029897, + "balance_loss_clip": 1.04538822, + "balance_loss_mlp": 1.02158201, + "epoch": 0.8218601575181868, + "flos": 25629625388160.0, + "grad_norm": 2.8108942739930454, + "language_loss": 0.79038286, + "learning_rate": 3.2360700891433254e-07, + "loss": 0.81225383, + "num_input_tokens_seen": 147599070, + "step": 6835, + "time_per_iteration": 2.8146777153015137 + }, + { + "auxiliary_loss_clip": 0.01034648, + "auxiliary_loss_mlp": 0.01004078, + "balance_loss_clip": 1.00682211, + "balance_loss_mlp": 1.0024929, + "epoch": 0.8219804004088258, + "flos": 67660229427840.0, + "grad_norm": 0.7934660905475361, + "language_loss": 0.57294452, + "learning_rate": 3.231823095984847e-07, + "loss": 0.59333175, + "num_input_tokens_seen": 147653710, + "step": 6836, + "time_per_iteration": 3.2441213130950928 + }, + { + "auxiliary_loss_clip": 0.01148825, + "auxiliary_loss_mlp": 0.01028398, + "balance_loss_clip": 1.0449264, + "balance_loss_mlp": 1.020486, + "epoch": 0.822100643299465, + "flos": 19464266327040.0, + "grad_norm": 2.0068428409238943, + "language_loss": 0.76017296, + "learning_rate": 3.2275786465454814e-07, + "loss": 0.78194523, + "num_input_tokens_seen": 147670360, + "step": 6837, + "time_per_iteration": 2.678302526473999 + }, + { + "auxiliary_loss_clip": 0.01140495, + "auxiliary_loss_mlp": 0.01024539, + "balance_loss_clip": 1.04268336, + "balance_loss_mlp": 1.01656437, + "epoch": 0.822220886190104, + "flos": 24681368292480.0, + "grad_norm": 1.7748432056569454, + "language_loss": 0.75883049, + "learning_rate": 3.2233367414690917e-07, + "loss": 0.78048086, + "num_input_tokens_seen": 147692550, + "step": 6838, + "time_per_iteration": 2.7960901260375977 + }, + { + "auxiliary_loss_clip": 0.01137144, + "auxiliary_loss_mlp": 0.01024034, + "balance_loss_clip": 1.04015779, + "balance_loss_mlp": 1.01658654, + "epoch": 0.8223411290807431, + "flos": 27819062991360.0, + "grad_norm": 2.70887283570885, + "language_loss": 0.8499831, + "learning_rate": 3.219097381399183e-07, + "loss": 0.87159491, + "num_input_tokens_seen": 147709725, + "step": 6839, + "time_per_iteration": 2.8097164630889893 + }, + { + "auxiliary_loss_clip": 0.01157527, + "auxiliary_loss_mlp": 0.01031428, + "balance_loss_clip": 1.04541421, + "balance_loss_mlp": 1.02374458, + "epoch": 0.8224613719713821, + "flos": 23218546913280.0, + "grad_norm": 1.818126222459677, + "language_loss": 0.81098652, + "learning_rate": 3.2148605669788584e-07, + "loss": 0.83287609, + "num_input_tokens_seen": 147729615, + "step": 6840, + "time_per_iteration": 3.7553012371063232 + }, + { + "auxiliary_loss_clip": 0.01153012, + "auxiliary_loss_mlp": 0.0102377, + "balance_loss_clip": 1.04746127, + "balance_loss_mlp": 1.01652861, + "epoch": 0.8225816148620213, + "flos": 15706250726400.0, + "grad_norm": 2.1854844506385156, + "language_loss": 0.77893591, + "learning_rate": 3.2106262988508405e-07, + "loss": 0.8007037, + "num_input_tokens_seen": 147747665, + "step": 6841, + "time_per_iteration": 3.6573781967163086 + }, + { + "auxiliary_loss_clip": 0.01149801, + "auxiliary_loss_mlp": 0.01028203, + "balance_loss_clip": 1.04507768, + "balance_loss_mlp": 1.02005517, + "epoch": 0.8227018577526604, + "flos": 18515111391360.0, + "grad_norm": 1.9937850977599463, + "language_loss": 0.74120605, + "learning_rate": 3.206394577657465e-07, + "loss": 0.76298606, + "num_input_tokens_seen": 147765445, + "step": 6842, + "time_per_iteration": 3.7100706100463867 + }, + { + "auxiliary_loss_clip": 0.01164272, + "auxiliary_loss_mlp": 0.01028324, + "balance_loss_clip": 1.04619694, + "balance_loss_mlp": 1.01976514, + "epoch": 0.8228221006432994, + "flos": 22236785406720.0, + "grad_norm": 2.693610196616268, + "language_loss": 0.72479081, + "learning_rate": 3.202165404040675e-07, + "loss": 0.74671674, + "num_input_tokens_seen": 147783365, + "step": 6843, + "time_per_iteration": 2.8473548889160156 + }, + { + "auxiliary_loss_clip": 0.01119962, + "auxiliary_loss_mlp": 0.0102631, + "balance_loss_clip": 1.04107153, + "balance_loss_mlp": 1.01846623, + "epoch": 0.8229423435339386, + "flos": 24097532630400.0, + "grad_norm": 2.10567970433406, + "language_loss": 0.74790943, + "learning_rate": 3.1979387786420396e-07, + "loss": 0.76937217, + "num_input_tokens_seen": 147803605, + "step": 6844, + "time_per_iteration": 2.853895902633667 + }, + { + "auxiliary_loss_clip": 0.0114814, + "auxiliary_loss_mlp": 0.01023339, + "balance_loss_clip": 1.0400486, + "balance_loss_mlp": 1.01564956, + "epoch": 0.8230625864245776, + "flos": 23878549365120.0, + "grad_norm": 1.8876716163980187, + "language_loss": 0.82067168, + "learning_rate": 3.1937147021027346e-07, + "loss": 0.84238642, + "num_input_tokens_seen": 147822060, + "step": 6845, + "time_per_iteration": 2.824204683303833 + }, + { + "auxiliary_loss_clip": 0.01157522, + "auxiliary_loss_mlp": 0.01021315, + "balance_loss_clip": 1.04541135, + "balance_loss_mlp": 1.0138557, + "epoch": 0.8231828293152167, + "flos": 16581106379520.0, + "grad_norm": 2.3169445285695542, + "language_loss": 0.76630652, + "learning_rate": 3.189493175063547e-07, + "loss": 0.78809488, + "num_input_tokens_seen": 147839295, + "step": 6846, + "time_per_iteration": 2.692462205886841 + }, + { + "auxiliary_loss_clip": 0.01156079, + "auxiliary_loss_mlp": 0.01030247, + "balance_loss_clip": 1.0489881, + "balance_loss_mlp": 1.02205777, + "epoch": 0.8233030722058559, + "flos": 18880071528960.0, + "grad_norm": 2.2405635400733463, + "language_loss": 0.67481327, + "learning_rate": 3.1852741981648776e-07, + "loss": 0.69667649, + "num_input_tokens_seen": 147857945, + "step": 6847, + "time_per_iteration": 2.7368502616882324 + }, + { + "auxiliary_loss_clip": 0.01129318, + "auxiliary_loss_mlp": 0.01026847, + "balance_loss_clip": 1.04308152, + "balance_loss_mlp": 1.01853812, + "epoch": 0.8234233150964949, + "flos": 28439024757120.0, + "grad_norm": 1.9524660262657825, + "language_loss": 0.69785345, + "learning_rate": 3.1810577720467404e-07, + "loss": 0.71941519, + "num_input_tokens_seen": 147879675, + "step": 6848, + "time_per_iteration": 2.8299388885498047 + }, + { + "auxiliary_loss_clip": 0.0115456, + "auxiliary_loss_mlp": 0.01027641, + "balance_loss_clip": 1.04526854, + "balance_loss_mlp": 1.01937342, + "epoch": 0.823543557987134, + "flos": 33765941577600.0, + "grad_norm": 1.521675922851043, + "language_loss": 0.56602144, + "learning_rate": 3.176843897348769e-07, + "loss": 0.58784342, + "num_input_tokens_seen": 147902870, + "step": 6849, + "time_per_iteration": 3.901047468185425 + }, + { + "auxiliary_loss_clip": 0.01147894, + "auxiliary_loss_mlp": 0.0103511, + "balance_loss_clip": 1.04566503, + "balance_loss_mlp": 1.02647948, + "epoch": 0.8236638008777731, + "flos": 17092366611840.0, + "grad_norm": 3.2497591788443247, + "language_loss": 0.75461292, + "learning_rate": 3.1726325747102034e-07, + "loss": 0.776443, + "num_input_tokens_seen": 147921245, + "step": 6850, + "time_per_iteration": 2.7744438648223877 + }, + { + "auxiliary_loss_clip": 0.01124506, + "auxiliary_loss_mlp": 0.0102182, + "balance_loss_clip": 1.03789699, + "balance_loss_mlp": 1.0139699, + "epoch": 0.8237840437684122, + "flos": 61639982334720.0, + "grad_norm": 1.6182810748786511, + "language_loss": 0.64148462, + "learning_rate": 3.1684238047698974e-07, + "loss": 0.66294789, + "num_input_tokens_seen": 147949515, + "step": 6851, + "time_per_iteration": 3.2928884029388428 + }, + { + "auxiliary_loss_clip": 0.01152657, + "auxiliary_loss_mlp": 0.01027079, + "balance_loss_clip": 1.04464555, + "balance_loss_mlp": 1.01856112, + "epoch": 0.8239042866590512, + "flos": 27309023821440.0, + "grad_norm": 1.8553370817208732, + "language_loss": 0.5265584, + "learning_rate": 3.1642175881663155e-07, + "loss": 0.5483557, + "num_input_tokens_seen": 147969245, + "step": 6852, + "time_per_iteration": 2.8584189414978027 + }, + { + "auxiliary_loss_clip": 0.0116746, + "auxiliary_loss_mlp": 0.01028264, + "balance_loss_clip": 1.04628026, + "balance_loss_mlp": 1.02075374, + "epoch": 0.8240245295496904, + "flos": 21726351187200.0, + "grad_norm": 1.9322659057756255, + "language_loss": 0.83597428, + "learning_rate": 3.160013925537537e-07, + "loss": 0.85793155, + "num_input_tokens_seen": 147990080, + "step": 6853, + "time_per_iteration": 2.7845051288604736 + }, + { + "auxiliary_loss_clip": 0.01136294, + "auxiliary_loss_mlp": 0.01022318, + "balance_loss_clip": 1.0437727, + "balance_loss_mlp": 1.01461744, + "epoch": 0.8241447724403295, + "flos": 20009318279040.0, + "grad_norm": 1.8907005441805897, + "language_loss": 0.75542682, + "learning_rate": 3.155812817521266e-07, + "loss": 0.77701294, + "num_input_tokens_seen": 148010455, + "step": 6854, + "time_per_iteration": 2.9381377696990967 + }, + { + "auxiliary_loss_clip": 0.01154738, + "auxiliary_loss_mlp": 0.01024177, + "balance_loss_clip": 1.04867816, + "balance_loss_mlp": 1.01577902, + "epoch": 0.8242650153309685, + "flos": 22272983337600.0, + "grad_norm": 2.0996803421409544, + "language_loss": 0.77928895, + "learning_rate": 3.151614264754787e-07, + "loss": 0.80107808, + "num_input_tokens_seen": 148028400, + "step": 6855, + "time_per_iteration": 2.7878410816192627 + }, + { + "auxiliary_loss_clip": 0.01171246, + "auxiliary_loss_mlp": 0.0102911, + "balance_loss_clip": 1.04701281, + "balance_loss_mlp": 1.02086639, + "epoch": 0.8243852582216077, + "flos": 22309971367680.0, + "grad_norm": 2.0857159860387133, + "language_loss": 0.79250658, + "learning_rate": 3.147418267875035e-07, + "loss": 0.81451011, + "num_input_tokens_seen": 148046530, + "step": 6856, + "time_per_iteration": 2.6389846801757812 + }, + { + "auxiliary_loss_clip": 0.01117137, + "auxiliary_loss_mlp": 0.00885726, + "balance_loss_clip": 1.03677201, + "balance_loss_mlp": 1.00051403, + "epoch": 0.8245055011122467, + "flos": 24645421756800.0, + "grad_norm": 2.2285230702136003, + "language_loss": 0.65481853, + "learning_rate": 3.1432248275185315e-07, + "loss": 0.67484713, + "num_input_tokens_seen": 148067040, + "step": 6857, + "time_per_iteration": 2.8867409229278564 + }, + { + "auxiliary_loss_clip": 0.01157658, + "auxiliary_loss_mlp": 0.01026589, + "balance_loss_clip": 1.04655004, + "balance_loss_mlp": 1.01854289, + "epoch": 0.8246257440028858, + "flos": 17487275713920.0, + "grad_norm": 2.093567844262956, + "language_loss": 0.77061164, + "learning_rate": 3.139033944321412e-07, + "loss": 0.79245406, + "num_input_tokens_seen": 148084400, + "step": 6858, + "time_per_iteration": 2.636235475540161 + }, + { + "auxiliary_loss_clip": 0.01160606, + "auxiliary_loss_mlp": 0.01029475, + "balance_loss_clip": 1.04409432, + "balance_loss_mlp": 1.02147603, + "epoch": 0.824745986893525, + "flos": 25010130499200.0, + "grad_norm": 1.7104172559087485, + "language_loss": 0.78922307, + "learning_rate": 3.1348456189194507e-07, + "loss": 0.81112385, + "num_input_tokens_seen": 148104860, + "step": 6859, + "time_per_iteration": 2.7122035026550293 + }, + { + "auxiliary_loss_clip": 0.0113066, + "auxiliary_loss_mlp": 0.01025679, + "balance_loss_clip": 1.04101276, + "balance_loss_mlp": 1.01727509, + "epoch": 0.824866229784164, + "flos": 18772698798720.0, + "grad_norm": 1.9382820163101862, + "language_loss": 0.82897639, + "learning_rate": 3.1306598519479876e-07, + "loss": 0.8505398, + "num_input_tokens_seen": 148124680, + "step": 6860, + "time_per_iteration": 2.77766489982605 + }, + { + "auxiliary_loss_clip": 0.01148507, + "auxiliary_loss_mlp": 0.01030826, + "balance_loss_clip": 1.04519141, + "balance_loss_mlp": 1.02323282, + "epoch": 0.8249864726748031, + "flos": 23842171866240.0, + "grad_norm": 1.717586758668225, + "language_loss": 0.78605193, + "learning_rate": 3.1264766440420177e-07, + "loss": 0.80784529, + "num_input_tokens_seen": 148147150, + "step": 6861, + "time_per_iteration": 2.8872947692871094 + }, + { + "auxiliary_loss_clip": 0.01154649, + "auxiliary_loss_mlp": 0.01025612, + "balance_loss_clip": 1.04576778, + "balance_loss_mlp": 1.01811719, + "epoch": 0.8251067155654422, + "flos": 20303103617280.0, + "grad_norm": 2.235830045321377, + "language_loss": 0.69160134, + "learning_rate": 3.122295995836124e-07, + "loss": 0.713404, + "num_input_tokens_seen": 148167020, + "step": 6862, + "time_per_iteration": 2.7113490104675293 + }, + { + "auxiliary_loss_clip": 0.01163398, + "auxiliary_loss_mlp": 0.0102637, + "balance_loss_clip": 1.04417515, + "balance_loss_mlp": 1.01803112, + "epoch": 0.8252269584560813, + "flos": 25009699536000.0, + "grad_norm": 1.6535669081997106, + "language_loss": 0.77341616, + "learning_rate": 3.118117907964508e-07, + "loss": 0.79531384, + "num_input_tokens_seen": 148188965, + "step": 6863, + "time_per_iteration": 2.770292282104492 + }, + { + "auxiliary_loss_clip": 0.01148114, + "auxiliary_loss_mlp": 0.01023723, + "balance_loss_clip": 1.04362416, + "balance_loss_mlp": 1.01653135, + "epoch": 0.8253472013467203, + "flos": 17128564542720.0, + "grad_norm": 1.8288876225542525, + "language_loss": 0.80250716, + "learning_rate": 3.1139423810609856e-07, + "loss": 0.82422549, + "num_input_tokens_seen": 148205660, + "step": 6864, + "time_per_iteration": 2.834240436553955 + }, + { + "auxiliary_loss_clip": 0.01166649, + "auxiliary_loss_mlp": 0.01024979, + "balance_loss_clip": 1.04456687, + "balance_loss_mlp": 1.01669955, + "epoch": 0.8254674442373595, + "flos": 22414794232320.0, + "grad_norm": 2.6809029570657783, + "language_loss": 0.75112492, + "learning_rate": 3.1097694157589714e-07, + "loss": 0.77304119, + "num_input_tokens_seen": 148225545, + "step": 6865, + "time_per_iteration": 2.84049129486084 + }, + { + "auxiliary_loss_clip": 0.01157355, + "auxiliary_loss_mlp": 0.0103071, + "balance_loss_clip": 1.04748487, + "balance_loss_mlp": 1.02322996, + "epoch": 0.8255876871279986, + "flos": 24786765774720.0, + "grad_norm": 4.392429020623105, + "language_loss": 0.76030576, + "learning_rate": 3.105599012691511e-07, + "loss": 0.78218645, + "num_input_tokens_seen": 148243975, + "step": 6866, + "time_per_iteration": 3.6642706394195557 + }, + { + "auxiliary_loss_clip": 0.01156481, + "auxiliary_loss_mlp": 0.01021239, + "balance_loss_clip": 1.04521036, + "balance_loss_mlp": 1.01365709, + "epoch": 0.8257079300186376, + "flos": 27455431656960.0, + "grad_norm": 1.4607109020833422, + "language_loss": 0.82363921, + "learning_rate": 3.101431172491249e-07, + "loss": 0.84541637, + "num_input_tokens_seen": 148265520, + "step": 6867, + "time_per_iteration": 2.791576385498047 + }, + { + "auxiliary_loss_clip": 0.01144184, + "auxiliary_loss_mlp": 0.00886405, + "balance_loss_clip": 1.04126298, + "balance_loss_mlp": 1.00060344, + "epoch": 0.8258281729092768, + "flos": 16471866142080.0, + "grad_norm": 2.2466732157170717, + "language_loss": 0.71952671, + "learning_rate": 3.097265895790444e-07, + "loss": 0.73983258, + "num_input_tokens_seen": 148283730, + "step": 6868, + "time_per_iteration": 4.6736109256744385 + }, + { + "auxiliary_loss_clip": 0.01142028, + "auxiliary_loss_mlp": 0.01029393, + "balance_loss_clip": 1.0431087, + "balance_loss_mlp": 1.02163815, + "epoch": 0.8259484157999158, + "flos": 21433822824960.0, + "grad_norm": 2.6475205184607913, + "language_loss": 0.82955056, + "learning_rate": 3.093103183220962e-07, + "loss": 0.85126483, + "num_input_tokens_seen": 148303775, + "step": 6869, + "time_per_iteration": 2.8103294372558594 + }, + { + "auxiliary_loss_clip": 0.01050946, + "auxiliary_loss_mlp": 0.01002048, + "balance_loss_clip": 1.0055604, + "balance_loss_mlp": 1.00057554, + "epoch": 0.8260686586905549, + "flos": 58322342453760.0, + "grad_norm": 0.8174781638989546, + "language_loss": 0.59382856, + "learning_rate": 3.0889430354142796e-07, + "loss": 0.61435843, + "num_input_tokens_seen": 148365285, + "step": 6870, + "time_per_iteration": 3.268340826034546 + }, + { + "auxiliary_loss_clip": 0.01144262, + "auxiliary_loss_mlp": 0.01025465, + "balance_loss_clip": 1.04238605, + "balance_loss_mlp": 1.01776385, + "epoch": 0.826188901581194, + "flos": 27527288814720.0, + "grad_norm": 1.9156281196714178, + "language_loss": 0.69718128, + "learning_rate": 3.084785453001497e-07, + "loss": 0.71887851, + "num_input_tokens_seen": 148386200, + "step": 6871, + "time_per_iteration": 2.8120479583740234 + }, + { + "auxiliary_loss_clip": 0.01149778, + "auxiliary_loss_mlp": 0.0088573, + "balance_loss_clip": 1.04653287, + "balance_loss_mlp": 1.00060427, + "epoch": 0.8263091444718331, + "flos": 23696051339520.0, + "grad_norm": 2.6196131570342174, + "language_loss": 0.81779677, + "learning_rate": 3.080630436613314e-07, + "loss": 0.83815187, + "num_input_tokens_seen": 148403970, + "step": 6872, + "time_per_iteration": 2.784489393234253 + }, + { + "auxiliary_loss_clip": 0.01150781, + "auxiliary_loss_mlp": 0.01024645, + "balance_loss_clip": 1.04353595, + "balance_loss_mlp": 1.01682472, + "epoch": 0.8264293873624722, + "flos": 17165157523200.0, + "grad_norm": 2.056943242505639, + "language_loss": 0.85830927, + "learning_rate": 3.076477986880039e-07, + "loss": 0.88006353, + "num_input_tokens_seen": 148421765, + "step": 6873, + "time_per_iteration": 2.7429256439208984 + }, + { + "auxiliary_loss_clip": 0.01150797, + "auxiliary_loss_mlp": 0.01023546, + "balance_loss_clip": 1.04600239, + "balance_loss_mlp": 1.01568079, + "epoch": 0.8265496302531112, + "flos": 24098645952000.0, + "grad_norm": 2.2225893999557362, + "language_loss": 0.69373953, + "learning_rate": 3.0723281044315986e-07, + "loss": 0.71548295, + "num_input_tokens_seen": 148443720, + "step": 6874, + "time_per_iteration": 2.9365668296813965 + }, + { + "auxiliary_loss_clip": 0.01163603, + "auxiliary_loss_mlp": 0.01025649, + "balance_loss_clip": 1.04345989, + "balance_loss_mlp": 1.01824594, + "epoch": 0.8266698731437504, + "flos": 14099894599680.0, + "grad_norm": 2.0732372036000597, + "language_loss": 0.76829231, + "learning_rate": 3.068180789897521e-07, + "loss": 0.7901848, + "num_input_tokens_seen": 148462130, + "step": 6875, + "time_per_iteration": 3.6671054363250732 + }, + { + "auxiliary_loss_clip": 0.01162292, + "auxiliary_loss_mlp": 0.01027826, + "balance_loss_clip": 1.04546881, + "balance_loss_mlp": 1.01994038, + "epoch": 0.8267901160343895, + "flos": 30777563715840.0, + "grad_norm": 1.4046958583674267, + "language_loss": 0.81769472, + "learning_rate": 3.064036043906966e-07, + "loss": 0.83959591, + "num_input_tokens_seen": 148485570, + "step": 6876, + "time_per_iteration": 2.7723517417907715 + }, + { + "auxiliary_loss_clip": 0.01148335, + "auxiliary_loss_mlp": 0.01028253, + "balance_loss_clip": 1.04350352, + "balance_loss_mlp": 1.01971769, + "epoch": 0.8269103589250285, + "flos": 40624915242240.0, + "grad_norm": 2.149660827452391, + "language_loss": 0.68115669, + "learning_rate": 3.059893867088668e-07, + "loss": 0.70292258, + "num_input_tokens_seen": 148509715, + "step": 6877, + "time_per_iteration": 2.9512288570404053 + }, + { + "auxiliary_loss_clip": 0.01159921, + "auxiliary_loss_mlp": 0.01022578, + "balance_loss_clip": 1.04833055, + "balance_loss_mlp": 1.0151459, + "epoch": 0.8270306018156677, + "flos": 30263645877120.0, + "grad_norm": 1.7359152550344865, + "language_loss": 0.66844368, + "learning_rate": 3.055754260071004e-07, + "loss": 0.6902687, + "num_input_tokens_seen": 148532010, + "step": 6878, + "time_per_iteration": 2.7634599208831787 + }, + { + "auxiliary_loss_clip": 0.0115949, + "auxiliary_loss_mlp": 0.01023798, + "balance_loss_clip": 1.04455733, + "balance_loss_mlp": 1.01686025, + "epoch": 0.8271508447063067, + "flos": 25226599812480.0, + "grad_norm": 2.1022656229669088, + "language_loss": 0.73695582, + "learning_rate": 3.051617223481948e-07, + "loss": 0.7587887, + "num_input_tokens_seen": 148553330, + "step": 6879, + "time_per_iteration": 2.8563787937164307 + }, + { + "auxiliary_loss_clip": 0.01144001, + "auxiliary_loss_mlp": 0.0103031, + "balance_loss_clip": 1.043455, + "balance_loss_mlp": 1.02236176, + "epoch": 0.8272710875969458, + "flos": 17566602900480.0, + "grad_norm": 2.412766546298222, + "language_loss": 0.75214696, + "learning_rate": 3.047482757949078e-07, + "loss": 0.77389002, + "num_input_tokens_seen": 148570960, + "step": 6880, + "time_per_iteration": 3.084068775177002 + }, + { + "auxiliary_loss_clip": 0.01134139, + "auxiliary_loss_mlp": 0.00885547, + "balance_loss_clip": 1.03906012, + "balance_loss_mlp": 1.00057507, + "epoch": 0.827391330487585, + "flos": 19755465886080.0, + "grad_norm": 2.593678172481295, + "language_loss": 0.85573834, + "learning_rate": 3.043350864099605e-07, + "loss": 0.8759352, + "num_input_tokens_seen": 148589520, + "step": 6881, + "time_per_iteration": 2.9191501140594482 + }, + { + "auxiliary_loss_clip": 0.01160526, + "auxiliary_loss_mlp": 0.0102694, + "balance_loss_clip": 1.0437324, + "balance_loss_mlp": 1.01905441, + "epoch": 0.827511573378224, + "flos": 16835174254080.0, + "grad_norm": 3.3238186620495105, + "language_loss": 0.80884635, + "learning_rate": 3.039221542560315e-07, + "loss": 0.83072102, + "num_input_tokens_seen": 148606085, + "step": 6882, + "time_per_iteration": 2.681912660598755 + }, + { + "auxiliary_loss_clip": 0.01158329, + "auxiliary_loss_mlp": 0.01020551, + "balance_loss_clip": 1.04563165, + "balance_loss_mlp": 1.01270676, + "epoch": 0.8276318162688631, + "flos": 18369242259840.0, + "grad_norm": 1.9303168650337506, + "language_loss": 0.73430741, + "learning_rate": 3.0350947939576356e-07, + "loss": 0.75609624, + "num_input_tokens_seen": 148625240, + "step": 6883, + "time_per_iteration": 2.6598947048187256 + }, + { + "auxiliary_loss_clip": 0.011646, + "auxiliary_loss_mlp": 0.01027396, + "balance_loss_clip": 1.04666448, + "balance_loss_mlp": 1.01923883, + "epoch": 0.8277520591595022, + "flos": 19352691705600.0, + "grad_norm": 1.8385283406026864, + "language_loss": 0.72440374, + "learning_rate": 3.0309706189175876e-07, + "loss": 0.7463237, + "num_input_tokens_seen": 148645075, + "step": 6884, + "time_per_iteration": 2.7471213340759277 + }, + { + "auxiliary_loss_clip": 0.01047462, + "auxiliary_loss_mlp": 0.01002603, + "balance_loss_clip": 1.00592518, + "balance_loss_mlp": 1.00121391, + "epoch": 0.8278723020501413, + "flos": 67918858329600.0, + "grad_norm": 0.7789256604977269, + "language_loss": 0.57364088, + "learning_rate": 3.0268490180658045e-07, + "loss": 0.59414154, + "num_input_tokens_seen": 148707855, + "step": 6885, + "time_per_iteration": 3.3082406520843506 + }, + { + "auxiliary_loss_clip": 0.01174119, + "auxiliary_loss_mlp": 0.01026053, + "balance_loss_clip": 1.05054426, + "balance_loss_mlp": 1.01811111, + "epoch": 0.8279925449407803, + "flos": 18185738653440.0, + "grad_norm": 2.2961031212731653, + "language_loss": 0.79321343, + "learning_rate": 3.0227299920275305e-07, + "loss": 0.81521511, + "num_input_tokens_seen": 148724170, + "step": 6886, + "time_per_iteration": 2.7174856662750244 + }, + { + "auxiliary_loss_clip": 0.01145147, + "auxiliary_loss_mlp": 0.01027639, + "balance_loss_clip": 1.04577804, + "balance_loss_mlp": 1.01860857, + "epoch": 0.8281127878314195, + "flos": 20631434860800.0, + "grad_norm": 2.7350009356443814, + "language_loss": 0.85866696, + "learning_rate": 3.018613541427613e-07, + "loss": 0.88039482, + "num_input_tokens_seen": 148743690, + "step": 6887, + "time_per_iteration": 2.836181163787842 + }, + { + "auxiliary_loss_clip": 0.01167077, + "auxiliary_loss_mlp": 0.01025668, + "balance_loss_clip": 1.04530168, + "balance_loss_mlp": 1.01792192, + "epoch": 0.8282330307220586, + "flos": 18004282122240.0, + "grad_norm": 1.6259000158290742, + "language_loss": 0.73511004, + "learning_rate": 3.0144996668905243e-07, + "loss": 0.75703752, + "num_input_tokens_seen": 148761070, + "step": 6888, + "time_per_iteration": 2.742135524749756 + }, + { + "auxiliary_loss_clip": 0.01123204, + "auxiliary_loss_mlp": 0.0088591, + "balance_loss_clip": 1.03731394, + "balance_loss_mlp": 1.00064862, + "epoch": 0.8283532736126976, + "flos": 20084120352000.0, + "grad_norm": 2.0922050137734898, + "language_loss": 0.82083309, + "learning_rate": 3.010388369040331e-07, + "loss": 0.8409242, + "num_input_tokens_seen": 148779730, + "step": 6889, + "time_per_iteration": 3.0480732917785645 + }, + { + "auxiliary_loss_clip": 0.01159774, + "auxiliary_loss_mlp": 0.01027529, + "balance_loss_clip": 1.04632807, + "balance_loss_mlp": 1.02049577, + "epoch": 0.8284735165033368, + "flos": 31868421805440.0, + "grad_norm": 1.465369679272681, + "language_loss": 0.82853174, + "learning_rate": 3.0062796485007156e-07, + "loss": 0.85040474, + "num_input_tokens_seen": 148800670, + "step": 6890, + "time_per_iteration": 2.7945311069488525 + }, + { + "auxiliary_loss_clip": 0.01168924, + "auxiliary_loss_mlp": 0.00885782, + "balance_loss_clip": 1.04678047, + "balance_loss_mlp": 1.00045812, + "epoch": 0.8285937593939758, + "flos": 26651319840000.0, + "grad_norm": 2.376671520637755, + "language_loss": 0.65198183, + "learning_rate": 3.002173505894965e-07, + "loss": 0.67252886, + "num_input_tokens_seen": 148819820, + "step": 6891, + "time_per_iteration": 2.680140495300293 + }, + { + "auxiliary_loss_clip": 0.01162476, + "auxiliary_loss_mlp": 0.0102875, + "balance_loss_clip": 1.04414618, + "balance_loss_mlp": 1.02041173, + "epoch": 0.8287140022846149, + "flos": 20193683811840.0, + "grad_norm": 2.6745153276450173, + "language_loss": 0.62048018, + "learning_rate": 2.998069941845973e-07, + "loss": 0.6423924, + "num_input_tokens_seen": 148838890, + "step": 6892, + "time_per_iteration": 3.7870588302612305 + }, + { + "auxiliary_loss_clip": 0.01056551, + "auxiliary_loss_mlp": 0.01002435, + "balance_loss_clip": 1.00540924, + "balance_loss_mlp": 1.00102866, + "epoch": 0.8288342451752541, + "flos": 70755980019840.0, + "grad_norm": 0.7500419593206463, + "language_loss": 0.57409227, + "learning_rate": 2.993968956976258e-07, + "loss": 0.59468216, + "num_input_tokens_seen": 148906635, + "step": 6893, + "time_per_iteration": 3.3331923484802246 + }, + { + "auxiliary_loss_clip": 0.01176225, + "auxiliary_loss_mlp": 0.01025273, + "balance_loss_clip": 1.04853964, + "balance_loss_mlp": 1.01693463, + "epoch": 0.8289544880658931, + "flos": 24572235795840.0, + "grad_norm": 2.1572569888530966, + "language_loss": 0.70093584, + "learning_rate": 2.9898705519079313e-07, + "loss": 0.72295082, + "num_input_tokens_seen": 148925740, + "step": 6894, + "time_per_iteration": 4.6035380363464355 + }, + { + "auxiliary_loss_clip": 0.01141257, + "auxiliary_loss_mlp": 0.01032237, + "balance_loss_clip": 1.04170609, + "balance_loss_mlp": 1.02459598, + "epoch": 0.8290747309565322, + "flos": 22273378387200.0, + "grad_norm": 1.753663568503286, + "language_loss": 0.74955982, + "learning_rate": 2.985774727262715e-07, + "loss": 0.77129477, + "num_input_tokens_seen": 148944585, + "step": 6895, + "time_per_iteration": 2.760498523712158 + }, + { + "auxiliary_loss_clip": 0.01165319, + "auxiliary_loss_mlp": 0.01027507, + "balance_loss_clip": 1.04524279, + "balance_loss_mlp": 1.02010703, + "epoch": 0.8291949738471713, + "flos": 23255570856960.0, + "grad_norm": 1.7802339031583303, + "language_loss": 0.8142339, + "learning_rate": 2.981681483661949e-07, + "loss": 0.83616221, + "num_input_tokens_seen": 148964170, + "step": 6896, + "time_per_iteration": 2.8291592597961426 + }, + { + "auxiliary_loss_clip": 0.01162609, + "auxiliary_loss_mlp": 0.0102587, + "balance_loss_clip": 1.04916501, + "balance_loss_mlp": 1.01832366, + "epoch": 0.8293152167378104, + "flos": 52555768185600.0, + "grad_norm": 1.5978831946914371, + "language_loss": 0.71153545, + "learning_rate": 2.9775908217265633e-07, + "loss": 0.73342025, + "num_input_tokens_seen": 148989405, + "step": 6897, + "time_per_iteration": 3.0553855895996094 + }, + { + "auxiliary_loss_clip": 0.01026347, + "auxiliary_loss_mlp": 0.01003657, + "balance_loss_clip": 1.00685561, + "balance_loss_mlp": 1.00205982, + "epoch": 0.8294354596284494, + "flos": 63356156294400.0, + "grad_norm": 0.828563117382008, + "language_loss": 0.50351465, + "learning_rate": 2.9735027420771253e-07, + "loss": 0.52381468, + "num_input_tokens_seen": 149049740, + "step": 6898, + "time_per_iteration": 3.418835163116455 + }, + { + "auxiliary_loss_clip": 0.01144583, + "auxiliary_loss_mlp": 0.01031717, + "balance_loss_clip": 1.04679918, + "balance_loss_mlp": 1.02495241, + "epoch": 0.8295557025190886, + "flos": 24827021942400.0, + "grad_norm": 1.6720155530898893, + "language_loss": 0.71476042, + "learning_rate": 2.969417245333774e-07, + "loss": 0.73652345, + "num_input_tokens_seen": 149069120, + "step": 6899, + "time_per_iteration": 2.953841209411621 + }, + { + "auxiliary_loss_clip": 0.01136542, + "auxiliary_loss_mlp": 0.01030237, + "balance_loss_clip": 1.04502559, + "balance_loss_mlp": 1.02207756, + "epoch": 0.8296759454097277, + "flos": 25118580637440.0, + "grad_norm": 2.4330192862271565, + "language_loss": 0.78183198, + "learning_rate": 2.9653343321162915e-07, + "loss": 0.8034997, + "num_input_tokens_seen": 149088630, + "step": 6900, + "time_per_iteration": 2.806950092315674 + }, + { + "auxiliary_loss_clip": 0.01143243, + "auxiliary_loss_mlp": 0.01031932, + "balance_loss_clip": 1.0475359, + "balance_loss_mlp": 1.02377808, + "epoch": 0.8297961883003667, + "flos": 24132581326080.0, + "grad_norm": 2.351194017721152, + "language_loss": 0.64983392, + "learning_rate": 2.9612540030440446e-07, + "loss": 0.67158568, + "num_input_tokens_seen": 149109175, + "step": 6901, + "time_per_iteration": 3.669731616973877 + }, + { + "auxiliary_loss_clip": 0.01044015, + "auxiliary_loss_mlp": 0.01001698, + "balance_loss_clip": 1.00540519, + "balance_loss_mlp": 1.00020838, + "epoch": 0.8299164311910058, + "flos": 67446561375360.0, + "grad_norm": 0.840491595033259, + "language_loss": 0.64108062, + "learning_rate": 2.9571762587360206e-07, + "loss": 0.66153777, + "num_input_tokens_seen": 149165560, + "step": 6902, + "time_per_iteration": 3.284834146499634 + }, + { + "auxiliary_loss_clip": 0.0112604, + "auxiliary_loss_mlp": 0.01026788, + "balance_loss_clip": 1.03505063, + "balance_loss_mlp": 1.01900387, + "epoch": 0.8300366740816449, + "flos": 25228682801280.0, + "grad_norm": 2.80402058083465, + "language_loss": 0.73595738, + "learning_rate": 2.953101099810806e-07, + "loss": 0.75748575, + "num_input_tokens_seen": 149185165, + "step": 6903, + "time_per_iteration": 2.910625457763672 + }, + { + "auxiliary_loss_clip": 0.01156046, + "auxiliary_loss_mlp": 0.01022718, + "balance_loss_clip": 1.04696941, + "balance_loss_mlp": 1.01531482, + "epoch": 0.830156916972284, + "flos": 18041018757120.0, + "grad_norm": 2.7930487416274765, + "language_loss": 0.82446742, + "learning_rate": 2.9490285268865965e-07, + "loss": 0.84625512, + "num_input_tokens_seen": 149202655, + "step": 6904, + "time_per_iteration": 2.715390205383301 + }, + { + "auxiliary_loss_clip": 0.01165633, + "auxiliary_loss_mlp": 0.01036241, + "balance_loss_clip": 1.04856145, + "balance_loss_mlp": 1.02753568, + "epoch": 0.830277159862923, + "flos": 26322485806080.0, + "grad_norm": 2.142319574415293, + "language_loss": 0.795663, + "learning_rate": 2.9449585405812085e-07, + "loss": 0.81768179, + "num_input_tokens_seen": 149220035, + "step": 6905, + "time_per_iteration": 2.7894554138183594 + }, + { + "auxiliary_loss_clip": 0.01137999, + "auxiliary_loss_mlp": 0.01028597, + "balance_loss_clip": 1.04470837, + "balance_loss_mlp": 1.0209794, + "epoch": 0.8303974027535622, + "flos": 19938861751680.0, + "grad_norm": 1.832590638411908, + "language_loss": 0.73613131, + "learning_rate": 2.940891141512043e-07, + "loss": 0.75779724, + "num_input_tokens_seen": 149238055, + "step": 6906, + "time_per_iteration": 2.8588640689849854 + }, + { + "auxiliary_loss_clip": 0.01149718, + "auxiliary_loss_mlp": 0.01028371, + "balance_loss_clip": 1.04523695, + "balance_loss_mlp": 1.01973391, + "epoch": 0.8305176456442013, + "flos": 17165552572800.0, + "grad_norm": 2.5642003564442373, + "language_loss": 0.71966654, + "learning_rate": 2.9368263302961385e-07, + "loss": 0.74144745, + "num_input_tokens_seen": 149256755, + "step": 6907, + "time_per_iteration": 2.8748111724853516 + }, + { + "auxiliary_loss_clip": 0.01118461, + "auxiliary_loss_mlp": 0.0102259, + "balance_loss_clip": 1.0359199, + "balance_loss_mlp": 1.01406395, + "epoch": 0.8306378885348403, + "flos": 25627614226560.0, + "grad_norm": 1.8397677112638795, + "language_loss": 0.7968877, + "learning_rate": 2.9327641075501075e-07, + "loss": 0.81829822, + "num_input_tokens_seen": 149275745, + "step": 6908, + "time_per_iteration": 2.9772229194641113 + }, + { + "auxiliary_loss_clip": 0.01144059, + "auxiliary_loss_mlp": 0.01029883, + "balance_loss_clip": 1.0407896, + "balance_loss_mlp": 1.02134192, + "epoch": 0.8307581314254795, + "flos": 33947864985600.0, + "grad_norm": 2.4464506164655995, + "language_loss": 0.6652258, + "learning_rate": 2.9287044738901866e-07, + "loss": 0.68696523, + "num_input_tokens_seen": 149293730, + "step": 6909, + "time_per_iteration": 2.885080099105835 + }, + { + "auxiliary_loss_clip": 0.01160708, + "auxiliary_loss_mlp": 0.00885357, + "balance_loss_clip": 1.04504657, + "balance_loss_mlp": 1.00048983, + "epoch": 0.8308783743161186, + "flos": 17562724231680.0, + "grad_norm": 3.24731565913331, + "language_loss": 0.90784848, + "learning_rate": 2.9246474299322274e-07, + "loss": 0.92830908, + "num_input_tokens_seen": 149309290, + "step": 6910, + "time_per_iteration": 2.736135721206665 + }, + { + "auxiliary_loss_clip": 0.01028369, + "auxiliary_loss_mlp": 0.01002354, + "balance_loss_clip": 1.00509644, + "balance_loss_mlp": 1.00093496, + "epoch": 0.8309986172067576, + "flos": 69412885649280.0, + "grad_norm": 1.0109013066392953, + "language_loss": 0.63181996, + "learning_rate": 2.920592976291678e-07, + "loss": 0.65212721, + "num_input_tokens_seen": 149366620, + "step": 6911, + "time_per_iteration": 3.203713893890381 + }, + { + "auxiliary_loss_clip": 0.01157911, + "auxiliary_loss_mlp": 0.0102639, + "balance_loss_clip": 1.0446744, + "balance_loss_mlp": 1.01863503, + "epoch": 0.8311188600973968, + "flos": 22309755886080.0, + "grad_norm": 1.7983918786322532, + "language_loss": 0.8073234, + "learning_rate": 2.916541113583595e-07, + "loss": 0.82916635, + "num_input_tokens_seen": 149385120, + "step": 6912, + "time_per_iteration": 2.743807077407837 + }, + { + "auxiliary_loss_clip": 0.01137903, + "auxiliary_loss_mlp": 0.01029037, + "balance_loss_clip": 1.04683769, + "balance_loss_mlp": 1.02111578, + "epoch": 0.8312391029880358, + "flos": 18770077105920.0, + "grad_norm": 2.1897448554332657, + "language_loss": 0.66467738, + "learning_rate": 2.912491842422642e-07, + "loss": 0.68634677, + "num_input_tokens_seen": 149402825, + "step": 6913, + "time_per_iteration": 2.8367714881896973 + }, + { + "auxiliary_loss_clip": 0.01160654, + "auxiliary_loss_mlp": 0.0102973, + "balance_loss_clip": 1.04676652, + "balance_loss_mlp": 1.02196336, + "epoch": 0.8313593458786749, + "flos": 20376648714240.0, + "grad_norm": 1.9759717401430088, + "language_loss": 0.71247959, + "learning_rate": 2.9084451634230857e-07, + "loss": 0.7343834, + "num_input_tokens_seen": 149422125, + "step": 6914, + "time_per_iteration": 2.86010479927063 + }, + { + "auxiliary_loss_clip": 0.01140489, + "auxiliary_loss_mlp": 0.01027897, + "balance_loss_clip": 1.04260719, + "balance_loss_mlp": 1.01951623, + "epoch": 0.831479588769314, + "flos": 32124069878400.0, + "grad_norm": 2.263224024021926, + "language_loss": 0.72069687, + "learning_rate": 2.9044010771988125e-07, + "loss": 0.74238074, + "num_input_tokens_seen": 149441940, + "step": 6915, + "time_per_iteration": 2.974116563796997 + }, + { + "auxiliary_loss_clip": 0.01144094, + "auxiliary_loss_mlp": 0.0102989, + "balance_loss_clip": 1.04345489, + "balance_loss_mlp": 1.02204573, + "epoch": 0.8315998316599531, + "flos": 45185929338240.0, + "grad_norm": 1.846795157904739, + "language_loss": 0.72343814, + "learning_rate": 2.900359584363303e-07, + "loss": 0.74517798, + "num_input_tokens_seen": 149465045, + "step": 6916, + "time_per_iteration": 3.0219662189483643 + }, + { + "auxiliary_loss_clip": 0.01129968, + "auxiliary_loss_mlp": 0.01030053, + "balance_loss_clip": 1.04527485, + "balance_loss_mlp": 1.02149057, + "epoch": 0.8317200745505922, + "flos": 18363747479040.0, + "grad_norm": 2.1651677350337017, + "language_loss": 0.84287697, + "learning_rate": 2.8963206855296494e-07, + "loss": 0.86447716, + "num_input_tokens_seen": 149481285, + "step": 6917, + "time_per_iteration": 2.9135618209838867 + }, + { + "auxiliary_loss_clip": 0.01159915, + "auxiliary_loss_mlp": 0.01027609, + "balance_loss_clip": 1.04453552, + "balance_loss_mlp": 1.02003932, + "epoch": 0.8318403174412313, + "flos": 24206557386240.0, + "grad_norm": 1.6513038364806172, + "language_loss": 0.77237517, + "learning_rate": 2.892284381310548e-07, + "loss": 0.79425043, + "num_input_tokens_seen": 149502700, + "step": 6918, + "time_per_iteration": 3.919813394546509 + }, + { + "auxiliary_loss_clip": 0.01147958, + "auxiliary_loss_mlp": 0.01027411, + "balance_loss_clip": 1.04448223, + "balance_loss_mlp": 1.01955247, + "epoch": 0.8319605603318704, + "flos": 22418780641920.0, + "grad_norm": 2.7983095582531115, + "language_loss": 0.72279006, + "learning_rate": 2.888250672318302e-07, + "loss": 0.74454379, + "num_input_tokens_seen": 149520100, + "step": 6919, + "time_per_iteration": 2.7740988731384277 + }, + { + "auxiliary_loss_clip": 0.01172199, + "auxiliary_loss_mlp": 0.0102854, + "balance_loss_clip": 1.04934382, + "balance_loss_mlp": 1.0208751, + "epoch": 0.8320808032225094, + "flos": 37414501459200.0, + "grad_norm": 2.1074225665890904, + "language_loss": 0.68802774, + "learning_rate": 2.884219559164831e-07, + "loss": 0.71003509, + "num_input_tokens_seen": 149543245, + "step": 6920, + "time_per_iteration": 4.60952615737915 + }, + { + "auxiliary_loss_clip": 0.01158893, + "auxiliary_loss_mlp": 0.01025144, + "balance_loss_clip": 1.04647756, + "balance_loss_mlp": 1.01685286, + "epoch": 0.8322010461131486, + "flos": 12787395638400.0, + "grad_norm": 4.162003762501876, + "language_loss": 0.81259131, + "learning_rate": 2.880191042461635e-07, + "loss": 0.83443165, + "num_input_tokens_seen": 149559185, + "step": 6921, + "time_per_iteration": 2.710078239440918 + }, + { + "auxiliary_loss_clip": 0.01137152, + "auxiliary_loss_mlp": 0.01025649, + "balance_loss_clip": 1.04133463, + "balance_loss_mlp": 1.01833534, + "epoch": 0.8323212890037877, + "flos": 15815455050240.0, + "grad_norm": 1.808878354819191, + "language_loss": 0.80361658, + "learning_rate": 2.876165122819849e-07, + "loss": 0.82524461, + "num_input_tokens_seen": 149577165, + "step": 6922, + "time_per_iteration": 2.818105697631836 + }, + { + "auxiliary_loss_clip": 0.01166857, + "auxiliary_loss_mlp": 0.01025037, + "balance_loss_clip": 1.04627824, + "balance_loss_mlp": 1.01725245, + "epoch": 0.8324415318944267, + "flos": 21719276208000.0, + "grad_norm": 1.6284236055906436, + "language_loss": 0.79329711, + "learning_rate": 2.872141800850201e-07, + "loss": 0.81521606, + "num_input_tokens_seen": 149594340, + "step": 6923, + "time_per_iteration": 2.722827196121216 + }, + { + "auxiliary_loss_clip": 0.01169056, + "auxiliary_loss_mlp": 0.01029841, + "balance_loss_clip": 1.04741895, + "balance_loss_mlp": 1.02200317, + "epoch": 0.8325617747850659, + "flos": 34198700636160.0, + "grad_norm": 2.443291096188202, + "language_loss": 0.73369932, + "learning_rate": 2.868121077163024e-07, + "loss": 0.75568831, + "num_input_tokens_seen": 149613895, + "step": 6924, + "time_per_iteration": 2.7348086833953857 + }, + { + "auxiliary_loss_clip": 0.01161633, + "auxiliary_loss_mlp": 0.01025455, + "balance_loss_clip": 1.04462063, + "balance_loss_mlp": 1.01765227, + "epoch": 0.8326820176757049, + "flos": 18369457741440.0, + "grad_norm": 1.786198390451614, + "language_loss": 0.72395861, + "learning_rate": 2.864102952368257e-07, + "loss": 0.74582952, + "num_input_tokens_seen": 149631820, + "step": 6925, + "time_per_iteration": 2.757730007171631 + }, + { + "auxiliary_loss_clip": 0.0112036, + "auxiliary_loss_mlp": 0.01029084, + "balance_loss_clip": 1.03785729, + "balance_loss_mlp": 1.02022111, + "epoch": 0.832802260566344, + "flos": 35991325716480.0, + "grad_norm": 1.3029616501080015, + "language_loss": 0.59359378, + "learning_rate": 2.860087427075444e-07, + "loss": 0.61508822, + "num_input_tokens_seen": 149656070, + "step": 6926, + "time_per_iteration": 3.0261073112487793 + }, + { + "auxiliary_loss_clip": 0.01145287, + "auxiliary_loss_mlp": 0.01028451, + "balance_loss_clip": 1.04344881, + "balance_loss_mlp": 1.0207684, + "epoch": 0.8329225034569832, + "flos": 14244434928000.0, + "grad_norm": 2.475928003253703, + "language_loss": 0.86110735, + "learning_rate": 2.856074501893744e-07, + "loss": 0.88284469, + "num_input_tokens_seen": 149671270, + "step": 6927, + "time_per_iteration": 3.763124465942383 + }, + { + "auxiliary_loss_clip": 0.01164092, + "auxiliary_loss_mlp": 0.01035064, + "balance_loss_clip": 1.05008972, + "balance_loss_mlp": 1.0270294, + "epoch": 0.8330427463476222, + "flos": 18077468083200.0, + "grad_norm": 1.7143442972001364, + "language_loss": 0.81378126, + "learning_rate": 2.8520641774319054e-07, + "loss": 0.83577287, + "num_input_tokens_seen": 149689360, + "step": 6928, + "time_per_iteration": 2.9055335521698 + }, + { + "auxiliary_loss_clip": 0.01149759, + "auxiliary_loss_mlp": 0.01026524, + "balance_loss_clip": 1.04086745, + "balance_loss_mlp": 1.01820946, + "epoch": 0.8331629892382613, + "flos": 18040839189120.0, + "grad_norm": 2.089262077466029, + "language_loss": 0.75777602, + "learning_rate": 2.848056454298309e-07, + "loss": 0.77953893, + "num_input_tokens_seen": 149706685, + "step": 6929, + "time_per_iteration": 2.817572593688965 + }, + { + "auxiliary_loss_clip": 0.01148777, + "auxiliary_loss_mlp": 0.01028158, + "balance_loss_clip": 1.04517221, + "balance_loss_mlp": 1.02012897, + "epoch": 0.8332832321289004, + "flos": 17457398576640.0, + "grad_norm": 2.660077881883285, + "language_loss": 0.65287852, + "learning_rate": 2.844051333100905e-07, + "loss": 0.67464787, + "num_input_tokens_seen": 149724230, + "step": 6930, + "time_per_iteration": 2.8694651126861572 + }, + { + "auxiliary_loss_clip": 0.0115038, + "auxiliary_loss_mlp": 0.01024769, + "balance_loss_clip": 1.04630816, + "balance_loss_mlp": 1.0179832, + "epoch": 0.8334034750195395, + "flos": 15084852416640.0, + "grad_norm": 2.063016794810725, + "language_loss": 0.8322252, + "learning_rate": 2.840048814447269e-07, + "loss": 0.85397673, + "num_input_tokens_seen": 149742395, + "step": 6931, + "time_per_iteration": 2.874403953552246 + }, + { + "auxiliary_loss_clip": 0.01141636, + "auxiliary_loss_mlp": 0.01027855, + "balance_loss_clip": 1.04221082, + "balance_loss_mlp": 1.01975441, + "epoch": 0.8335237179101785, + "flos": 19427170556160.0, + "grad_norm": 4.899265501727095, + "language_loss": 0.73848808, + "learning_rate": 2.836048898944587e-07, + "loss": 0.76018304, + "num_input_tokens_seen": 149760820, + "step": 6932, + "time_per_iteration": 2.865506887435913 + }, + { + "auxiliary_loss_clip": 0.01146071, + "auxiliary_loss_mlp": 0.01027734, + "balance_loss_clip": 1.04255795, + "balance_loss_mlp": 1.02047443, + "epoch": 0.8336439608008177, + "flos": 21762046327680.0, + "grad_norm": 2.382820918577689, + "language_loss": 0.7232579, + "learning_rate": 2.832051587199642e-07, + "loss": 0.74499595, + "num_input_tokens_seen": 149778075, + "step": 6933, + "time_per_iteration": 2.8345069885253906 + }, + { + "auxiliary_loss_clip": 0.01052824, + "auxiliary_loss_mlp": 0.01002669, + "balance_loss_clip": 1.00572705, + "balance_loss_mlp": 1.00131023, + "epoch": 0.8337642036914568, + "flos": 59702783990400.0, + "grad_norm": 0.811322589667753, + "language_loss": 0.577613, + "learning_rate": 2.828056879818821e-07, + "loss": 0.5981679, + "num_input_tokens_seen": 149837150, + "step": 6934, + "time_per_iteration": 3.2033607959747314 + }, + { + "auxiliary_loss_clip": 0.01137021, + "auxiliary_loss_mlp": 0.010268, + "balance_loss_clip": 1.03835166, + "balance_loss_mlp": 1.01925159, + "epoch": 0.8338844465820958, + "flos": 27162185022720.0, + "grad_norm": 2.565706425296944, + "language_loss": 0.83282793, + "learning_rate": 2.824064777408117e-07, + "loss": 0.85446608, + "num_input_tokens_seen": 149856940, + "step": 6935, + "time_per_iteration": 2.8632898330688477 + }, + { + "auxiliary_loss_clip": 0.01159965, + "auxiliary_loss_mlp": 0.01026439, + "balance_loss_clip": 1.04840899, + "balance_loss_mlp": 1.01830304, + "epoch": 0.8340046894727349, + "flos": 30481264425600.0, + "grad_norm": 1.6478963947835463, + "language_loss": 0.75752544, + "learning_rate": 2.8200752805731263e-07, + "loss": 0.7793895, + "num_input_tokens_seen": 149879930, + "step": 6936, + "time_per_iteration": 2.8625824451446533 + }, + { + "auxiliary_loss_clip": 0.01161771, + "auxiliary_loss_mlp": 0.01028466, + "balance_loss_clip": 1.04850078, + "balance_loss_mlp": 1.02085781, + "epoch": 0.834124932363374, + "flos": 27126166659840.0, + "grad_norm": 4.644196476962356, + "language_loss": 0.8096931, + "learning_rate": 2.8160883899190625e-07, + "loss": 0.83159542, + "num_input_tokens_seen": 149903200, + "step": 6937, + "time_per_iteration": 2.790393352508545 + }, + { + "auxiliary_loss_clip": 0.01132689, + "auxiliary_loss_mlp": 0.01028457, + "balance_loss_clip": 1.04482722, + "balance_loss_mlp": 1.0211432, + "epoch": 0.8342451752540131, + "flos": 24569865498240.0, + "grad_norm": 2.1833994679072837, + "language_loss": 0.73082459, + "learning_rate": 2.8121041060507234e-07, + "loss": 0.75243604, + "num_input_tokens_seen": 149922230, + "step": 6938, + "time_per_iteration": 2.9343838691711426 + }, + { + "auxiliary_loss_clip": 0.01164457, + "auxiliary_loss_mlp": 0.01028096, + "balance_loss_clip": 1.04676628, + "balance_loss_mlp": 1.02009726, + "epoch": 0.8343654181446521, + "flos": 26615085995520.0, + "grad_norm": 1.7491223644055627, + "language_loss": 0.71626282, + "learning_rate": 2.808122429572528e-07, + "loss": 0.73818833, + "num_input_tokens_seen": 149942435, + "step": 6939, + "time_per_iteration": 2.834690809249878 + }, + { + "auxiliary_loss_clip": 0.01146245, + "auxiliary_loss_mlp": 0.01026146, + "balance_loss_clip": 1.04260206, + "balance_loss_mlp": 1.01792037, + "epoch": 0.8344856610352913, + "flos": 20777268078720.0, + "grad_norm": 2.697988655170884, + "language_loss": 0.75905263, + "learning_rate": 2.804143361088489e-07, + "loss": 0.7807765, + "num_input_tokens_seen": 149961615, + "step": 6940, + "time_per_iteration": 2.915060043334961 + }, + { + "auxiliary_loss_clip": 0.01143556, + "auxiliary_loss_mlp": 0.01027149, + "balance_loss_clip": 1.04398644, + "balance_loss_mlp": 1.01894152, + "epoch": 0.8346059039259304, + "flos": 26095960684800.0, + "grad_norm": 2.6846547737236266, + "language_loss": 0.77914131, + "learning_rate": 2.8001669012022277e-07, + "loss": 0.80084836, + "num_input_tokens_seen": 149979585, + "step": 6941, + "time_per_iteration": 2.810288667678833 + }, + { + "auxiliary_loss_clip": 0.01161609, + "auxiliary_loss_mlp": 0.01025995, + "balance_loss_clip": 1.05005693, + "balance_loss_mlp": 1.01850235, + "epoch": 0.8347261468165694, + "flos": 29027708755200.0, + "grad_norm": 1.7336574577505104, + "language_loss": 0.69329399, + "learning_rate": 2.7961930505169795e-07, + "loss": 0.71517003, + "num_input_tokens_seen": 150003830, + "step": 6942, + "time_per_iteration": 2.8108363151550293 + }, + { + "auxiliary_loss_clip": 0.01162818, + "auxiliary_loss_mlp": 0.00886049, + "balance_loss_clip": 1.04741168, + "balance_loss_mlp": 1.0005362, + "epoch": 0.8348463897072086, + "flos": 26396461866240.0, + "grad_norm": 2.1420103805669553, + "language_loss": 0.76350892, + "learning_rate": 2.792221809635558e-07, + "loss": 0.78399754, + "num_input_tokens_seen": 150024460, + "step": 6943, + "time_per_iteration": 2.77736759185791 + }, + { + "auxiliary_loss_clip": 0.01114831, + "auxiliary_loss_mlp": 0.01025569, + "balance_loss_clip": 1.04269588, + "balance_loss_mlp": 1.01748633, + "epoch": 0.8349666325978476, + "flos": 23367720096000.0, + "grad_norm": 3.4121386494449113, + "language_loss": 0.74889672, + "learning_rate": 2.788253179160411e-07, + "loss": 0.77030075, + "num_input_tokens_seen": 150045620, + "step": 6944, + "time_per_iteration": 4.044082403182983 + }, + { + "auxiliary_loss_clip": 0.01150073, + "auxiliary_loss_mlp": 0.01029309, + "balance_loss_clip": 1.04484427, + "balance_loss_mlp": 1.02210319, + "epoch": 0.8350868754884867, + "flos": 12896528135040.0, + "grad_norm": 1.8371923623238862, + "language_loss": 0.64963812, + "learning_rate": 2.7842871596935725e-07, + "loss": 0.67143202, + "num_input_tokens_seen": 150064135, + "step": 6945, + "time_per_iteration": 2.8842201232910156 + }, + { + "auxiliary_loss_clip": 0.01157294, + "auxiliary_loss_mlp": 0.01023295, + "balance_loss_clip": 1.04649591, + "balance_loss_mlp": 1.01564741, + "epoch": 0.8352071183791259, + "flos": 26505522535680.0, + "grad_norm": 1.647921844896116, + "language_loss": 0.68875206, + "learning_rate": 2.780323751836682e-07, + "loss": 0.71055794, + "num_input_tokens_seen": 150085350, + "step": 6946, + "time_per_iteration": 3.822718381881714 + }, + { + "auxiliary_loss_clip": 0.01147754, + "auxiliary_loss_mlp": 0.00885578, + "balance_loss_clip": 1.04120195, + "balance_loss_mlp": 1.00056934, + "epoch": 0.8353273612697649, + "flos": 20668063754880.0, + "grad_norm": 1.8472444784176347, + "language_loss": 0.78971303, + "learning_rate": 2.7763629561909876e-07, + "loss": 0.81004632, + "num_input_tokens_seen": 150106180, + "step": 6947, + "time_per_iteration": 3.7258641719818115 + }, + { + "auxiliary_loss_clip": 0.01166301, + "auxiliary_loss_mlp": 0.01027842, + "balance_loss_clip": 1.0456506, + "balance_loss_mlp": 1.02005804, + "epoch": 0.835447604160404, + "flos": 19754137082880.0, + "grad_norm": 1.9744901013008125, + "language_loss": 0.77147317, + "learning_rate": 2.772404773357335e-07, + "loss": 0.79341459, + "num_input_tokens_seen": 150125585, + "step": 6948, + "time_per_iteration": 2.7096762657165527 + }, + { + "auxiliary_loss_clip": 0.01135132, + "auxiliary_loss_mlp": 0.01028153, + "balance_loss_clip": 1.04365027, + "balance_loss_mlp": 1.0197016, + "epoch": 0.8355678470510431, + "flos": 23435842239360.0, + "grad_norm": 1.9432776922152453, + "language_loss": 0.78497684, + "learning_rate": 2.7684492039361853e-07, + "loss": 0.80660969, + "num_input_tokens_seen": 150144810, + "step": 6949, + "time_per_iteration": 2.7806894779205322 + }, + { + "auxiliary_loss_clip": 0.0117323, + "auxiliary_loss_mlp": 0.01025905, + "balance_loss_clip": 1.04974055, + "balance_loss_mlp": 1.01816869, + "epoch": 0.8356880899416822, + "flos": 21214588164480.0, + "grad_norm": 1.6644610838471459, + "language_loss": 0.83599842, + "learning_rate": 2.764496248527586e-07, + "loss": 0.85798979, + "num_input_tokens_seen": 150163785, + "step": 6950, + "time_per_iteration": 2.722078323364258 + }, + { + "auxiliary_loss_clip": 0.01142005, + "auxiliary_loss_mlp": 0.0102564, + "balance_loss_clip": 1.04312897, + "balance_loss_mlp": 1.01723003, + "epoch": 0.8358083328323213, + "flos": 28037543466240.0, + "grad_norm": 1.841490458207708, + "language_loss": 0.78487647, + "learning_rate": 2.760545907731211e-07, + "loss": 0.80655295, + "num_input_tokens_seen": 150184360, + "step": 6951, + "time_per_iteration": 2.8021678924560547 + }, + { + "auxiliary_loss_clip": 0.01162017, + "auxiliary_loss_mlp": 0.0102707, + "balance_loss_clip": 1.04526877, + "balance_loss_mlp": 1.01852894, + "epoch": 0.8359285757229604, + "flos": 27783655159680.0, + "grad_norm": 1.6779967888144633, + "language_loss": 0.67845023, + "learning_rate": 2.75659818214631e-07, + "loss": 0.70034105, + "num_input_tokens_seen": 150205465, + "step": 6952, + "time_per_iteration": 2.801417827606201 + }, + { + "auxiliary_loss_clip": 0.01151776, + "auxiliary_loss_mlp": 0.01029592, + "balance_loss_clip": 1.04363608, + "balance_loss_mlp": 1.02149189, + "epoch": 0.8360488186135995, + "flos": 21435115714560.0, + "grad_norm": 1.8732785321883871, + "language_loss": 0.78096575, + "learning_rate": 2.752653072371749e-07, + "loss": 0.80277938, + "num_input_tokens_seen": 150224900, + "step": 6953, + "time_per_iteration": 3.6686108112335205 + }, + { + "auxiliary_loss_clip": 0.01135194, + "auxiliary_loss_mlp": 0.01031077, + "balance_loss_clip": 1.04382777, + "balance_loss_mlp": 1.02306008, + "epoch": 0.8361690615042385, + "flos": 27632327160960.0, + "grad_norm": 1.940607246677412, + "language_loss": 0.74552268, + "learning_rate": 2.7487105790060105e-07, + "loss": 0.76718545, + "num_input_tokens_seen": 150244310, + "step": 6954, + "time_per_iteration": 2.8913161754608154 + }, + { + "auxiliary_loss_clip": 0.01162009, + "auxiliary_loss_mlp": 0.01025276, + "balance_loss_clip": 1.04635453, + "balance_loss_mlp": 1.0177058, + "epoch": 0.8362893043948777, + "flos": 39202529598720.0, + "grad_norm": 1.9123543005923893, + "language_loss": 0.69836956, + "learning_rate": 2.7447707026471587e-07, + "loss": 0.72024238, + "num_input_tokens_seen": 150267285, + "step": 6955, + "time_per_iteration": 2.975628137588501 + }, + { + "auxiliary_loss_clip": 0.01142531, + "auxiliary_loss_mlp": 0.0103208, + "balance_loss_clip": 1.04254818, + "balance_loss_mlp": 1.02465272, + "epoch": 0.8364095472855168, + "flos": 24785329230720.0, + "grad_norm": 2.0078923805821316, + "language_loss": 0.7952286, + "learning_rate": 2.740833443892874e-07, + "loss": 0.81697476, + "num_input_tokens_seen": 150285455, + "step": 6956, + "time_per_iteration": 2.890105724334717 + }, + { + "auxiliary_loss_clip": 0.01149439, + "auxiliary_loss_mlp": 0.01025297, + "balance_loss_clip": 1.04403329, + "balance_loss_mlp": 1.01769102, + "epoch": 0.8365297901761558, + "flos": 22743412784640.0, + "grad_norm": 1.8622202860930317, + "language_loss": 0.79765093, + "learning_rate": 2.7368988033404327e-07, + "loss": 0.81939828, + "num_input_tokens_seen": 150302970, + "step": 6957, + "time_per_iteration": 2.7692129611968994 + }, + { + "auxiliary_loss_clip": 0.0114497, + "auxiliary_loss_mlp": 0.01024832, + "balance_loss_clip": 1.04463065, + "balance_loss_mlp": 1.01743174, + "epoch": 0.836650033066795, + "flos": 28396003242240.0, + "grad_norm": 1.5120721431845563, + "language_loss": 0.84436464, + "learning_rate": 2.732966781586712e-07, + "loss": 0.86606264, + "num_input_tokens_seen": 150322715, + "step": 6958, + "time_per_iteration": 2.8658571243286133 + }, + { + "auxiliary_loss_clip": 0.01153241, + "auxiliary_loss_mlp": 0.01027157, + "balance_loss_clip": 1.0420754, + "balance_loss_mlp": 1.01955128, + "epoch": 0.836770275957434, + "flos": 22236857233920.0, + "grad_norm": 1.6471486308496566, + "language_loss": 0.66582859, + "learning_rate": 2.729037379228205e-07, + "loss": 0.68763262, + "num_input_tokens_seen": 150342900, + "step": 6959, + "time_per_iteration": 2.7307169437408447 + }, + { + "auxiliary_loss_clip": 0.01151296, + "auxiliary_loss_mlp": 0.01023248, + "balance_loss_clip": 1.04839826, + "balance_loss_mlp": 1.01582754, + "epoch": 0.8368905188480731, + "flos": 22491930689280.0, + "grad_norm": 1.4431765520989726, + "language_loss": 0.80747855, + "learning_rate": 2.725110596860998e-07, + "loss": 0.82922399, + "num_input_tokens_seen": 150363580, + "step": 6960, + "time_per_iteration": 2.851405620574951 + }, + { + "auxiliary_loss_clip": 0.01129772, + "auxiliary_loss_mlp": 0.0102685, + "balance_loss_clip": 1.04410207, + "balance_loss_mlp": 1.01854753, + "epoch": 0.8370107617387123, + "flos": 13370405287680.0, + "grad_norm": 2.609353714255138, + "language_loss": 0.69764328, + "learning_rate": 2.7211864350807776e-07, + "loss": 0.71920955, + "num_input_tokens_seen": 150381780, + "step": 6961, + "time_per_iteration": 2.9665515422821045 + }, + { + "auxiliary_loss_clip": 0.01170459, + "auxiliary_loss_mlp": 0.01029105, + "balance_loss_clip": 1.04840088, + "balance_loss_mlp": 1.02035499, + "epoch": 0.8371310046293513, + "flos": 25261289372160.0, + "grad_norm": 1.6171823299866233, + "language_loss": 0.73466986, + "learning_rate": 2.717264894482836e-07, + "loss": 0.75666547, + "num_input_tokens_seen": 150402120, + "step": 6962, + "time_per_iteration": 2.748783826828003 + }, + { + "auxiliary_loss_clip": 0.01165151, + "auxiliary_loss_mlp": 0.01028299, + "balance_loss_clip": 1.04816449, + "balance_loss_mlp": 1.02026427, + "epoch": 0.8372512475199904, + "flos": 19792705311360.0, + "grad_norm": 3.6124007166002645, + "language_loss": 0.81352711, + "learning_rate": 2.7133459756620646e-07, + "loss": 0.83546162, + "num_input_tokens_seen": 150419315, + "step": 6963, + "time_per_iteration": 2.716430187225342 + }, + { + "auxiliary_loss_clip": 0.01155036, + "auxiliary_loss_mlp": 0.01031693, + "balance_loss_clip": 1.04566503, + "balance_loss_mlp": 1.02328897, + "epoch": 0.8373714904106295, + "flos": 19391224020480.0, + "grad_norm": 1.6338969263081222, + "language_loss": 0.73919606, + "learning_rate": 2.7094296792129733e-07, + "loss": 0.76106334, + "num_input_tokens_seen": 150438915, + "step": 6964, + "time_per_iteration": 2.732682704925537 + }, + { + "auxiliary_loss_clip": 0.01160494, + "auxiliary_loss_mlp": 0.01029485, + "balance_loss_clip": 1.04689598, + "balance_loss_mlp": 1.02184963, + "epoch": 0.8374917333012686, + "flos": 14975935401600.0, + "grad_norm": 1.764156450414326, + "language_loss": 0.75584322, + "learning_rate": 2.7055160057296424e-07, + "loss": 0.77774304, + "num_input_tokens_seen": 150456155, + "step": 6965, + "time_per_iteration": 2.7116799354553223 + }, + { + "auxiliary_loss_clip": 0.01137071, + "auxiliary_loss_mlp": 0.01030413, + "balance_loss_clip": 1.04077435, + "balance_loss_mlp": 1.02182424, + "epoch": 0.8376119761919076, + "flos": 30331839847680.0, + "grad_norm": 1.598856566703946, + "language_loss": 0.72309184, + "learning_rate": 2.7016049558057896e-07, + "loss": 0.74476671, + "num_input_tokens_seen": 150478115, + "step": 6966, + "time_per_iteration": 2.884434461593628 + }, + { + "auxiliary_loss_clip": 0.01157883, + "auxiliary_loss_mlp": 0.01023777, + "balance_loss_clip": 1.04657984, + "balance_loss_mlp": 1.01589715, + "epoch": 0.8377322190825467, + "flos": 29423336129280.0, + "grad_norm": 1.726831636857982, + "language_loss": 0.70738643, + "learning_rate": 2.6976965300347074e-07, + "loss": 0.72920305, + "num_input_tokens_seen": 150500725, + "step": 6967, + "time_per_iteration": 2.8787667751312256 + }, + { + "auxiliary_loss_clip": 0.011441, + "auxiliary_loss_mlp": 0.01026789, + "balance_loss_clip": 1.04209423, + "balance_loss_mlp": 1.01877213, + "epoch": 0.8378524619731859, + "flos": 26687086807680.0, + "grad_norm": 2.461666379085168, + "language_loss": 0.69385099, + "learning_rate": 2.693790729009309e-07, + "loss": 0.7155599, + "num_input_tokens_seen": 150522335, + "step": 6968, + "time_per_iteration": 2.8726155757904053 + }, + { + "auxiliary_loss_clip": 0.01149602, + "auxiliary_loss_mlp": 0.01029015, + "balance_loss_clip": 1.04333305, + "balance_loss_mlp": 1.02129006, + "epoch": 0.8379727048638249, + "flos": 20703866636160.0, + "grad_norm": 1.7741532562752775, + "language_loss": 0.88206583, + "learning_rate": 2.6898875533220946e-07, + "loss": 0.90385205, + "num_input_tokens_seen": 150541640, + "step": 6969, + "time_per_iteration": 4.008011102676392 + }, + { + "auxiliary_loss_clip": 0.01164971, + "auxiliary_loss_mlp": 0.01022993, + "balance_loss_clip": 1.04756474, + "balance_loss_mlp": 1.01593304, + "epoch": 0.838092947754464, + "flos": 20084084438400.0, + "grad_norm": 1.7871753158297141, + "language_loss": 0.81909335, + "learning_rate": 2.685987003565171e-07, + "loss": 0.84097296, + "num_input_tokens_seen": 150559680, + "step": 6970, + "time_per_iteration": 2.7531869411468506 + }, + { + "auxiliary_loss_clip": 0.01134079, + "auxiliary_loss_mlp": 0.01028143, + "balance_loss_clip": 1.04607606, + "balance_loss_mlp": 1.02047765, + "epoch": 0.8382131906451031, + "flos": 18113270964480.0, + "grad_norm": 3.062050913641307, + "language_loss": 0.75064218, + "learning_rate": 2.6820890803302566e-07, + "loss": 0.77226436, + "num_input_tokens_seen": 150575205, + "step": 6971, + "time_per_iteration": 2.874586582183838 + }, + { + "auxiliary_loss_clip": 0.01142082, + "auxiliary_loss_mlp": 0.01026638, + "balance_loss_clip": 1.04703605, + "balance_loss_mlp": 1.01865971, + "epoch": 0.8383334335357422, + "flos": 17092653920640.0, + "grad_norm": 2.742371734941802, + "language_loss": 0.81916559, + "learning_rate": 2.6781937842086557e-07, + "loss": 0.84085274, + "num_input_tokens_seen": 150593995, + "step": 6972, + "time_per_iteration": 3.6689162254333496 + }, + { + "auxiliary_loss_clip": 0.01161155, + "auxiliary_loss_mlp": 0.01033266, + "balance_loss_clip": 1.04475868, + "balance_loss_mlp": 1.0249753, + "epoch": 0.8384536764263812, + "flos": 20704728562560.0, + "grad_norm": 1.9097477181911215, + "language_loss": 0.67254138, + "learning_rate": 2.6743011157912933e-07, + "loss": 0.69448555, + "num_input_tokens_seen": 150613715, + "step": 6973, + "time_per_iteration": 3.6681530475616455 + }, + { + "auxiliary_loss_clip": 0.01129433, + "auxiliary_loss_mlp": 0.01026229, + "balance_loss_clip": 1.03918338, + "balance_loss_mlp": 1.01877546, + "epoch": 0.8385739193170204, + "flos": 28986842056320.0, + "grad_norm": 2.2447402954908093, + "language_loss": 0.65060592, + "learning_rate": 2.6704110756686725e-07, + "loss": 0.67216253, + "num_input_tokens_seen": 150634540, + "step": 6974, + "time_per_iteration": 2.8360137939453125 + }, + { + "auxiliary_loss_clip": 0.01144908, + "auxiliary_loss_mlp": 0.00886103, + "balance_loss_clip": 1.04119766, + "balance_loss_mlp": 1.0004493, + "epoch": 0.8386941622076595, + "flos": 23438068882560.0, + "grad_norm": 1.7853320226983078, + "language_loss": 0.83886713, + "learning_rate": 2.6665236644309085e-07, + "loss": 0.85917723, + "num_input_tokens_seen": 150654850, + "step": 6975, + "time_per_iteration": 2.8743505477905273 + }, + { + "auxiliary_loss_clip": 0.01157891, + "auxiliary_loss_mlp": 0.01031563, + "balance_loss_clip": 1.04439616, + "balance_loss_mlp": 1.02395749, + "epoch": 0.8388144050982985, + "flos": 23002724044800.0, + "grad_norm": 1.8853819785841432, + "language_loss": 0.79586053, + "learning_rate": 2.662638882667727e-07, + "loss": 0.81775504, + "num_input_tokens_seen": 150673790, + "step": 6976, + "time_per_iteration": 2.6941635608673096 + }, + { + "auxiliary_loss_clip": 0.01173614, + "auxiliary_loss_mlp": 0.01029309, + "balance_loss_clip": 1.04767728, + "balance_loss_mlp": 1.02107775, + "epoch": 0.8389346479889377, + "flos": 24280353878400.0, + "grad_norm": 1.791170310690439, + "language_loss": 0.72977716, + "learning_rate": 2.658756730968443e-07, + "loss": 0.75180638, + "num_input_tokens_seen": 150692255, + "step": 6977, + "time_per_iteration": 2.6925816535949707 + }, + { + "auxiliary_loss_clip": 0.0115527, + "auxiliary_loss_mlp": 0.01028796, + "balance_loss_clip": 1.04784179, + "balance_loss_mlp": 1.02113056, + "epoch": 0.8390548908795767, + "flos": 21215019127680.0, + "grad_norm": 3.1215503678170413, + "language_loss": 0.88346744, + "learning_rate": 2.654877209921975e-07, + "loss": 0.90530807, + "num_input_tokens_seen": 150709790, + "step": 6978, + "time_per_iteration": 2.721257209777832 + }, + { + "auxiliary_loss_clip": 0.01135506, + "auxiliary_loss_mlp": 0.0103099, + "balance_loss_clip": 1.03957057, + "balance_loss_mlp": 1.02212691, + "epoch": 0.8391751337702158, + "flos": 35627299332480.0, + "grad_norm": 2.334328399989941, + "language_loss": 0.63479948, + "learning_rate": 2.651000320116843e-07, + "loss": 0.65646446, + "num_input_tokens_seen": 150730675, + "step": 6979, + "time_per_iteration": 3.8726091384887695 + }, + { + "auxiliary_loss_clip": 0.01134972, + "auxiliary_loss_mlp": 0.0088634, + "balance_loss_clip": 1.04142487, + "balance_loss_mlp": 1.00054026, + "epoch": 0.839295376660855, + "flos": 21325229032320.0, + "grad_norm": 3.1038042336880003, + "language_loss": 0.75993323, + "learning_rate": 2.647126062141163e-07, + "loss": 0.78014636, + "num_input_tokens_seen": 150749750, + "step": 6980, + "time_per_iteration": 2.7816197872161865 + }, + { + "auxiliary_loss_clip": 0.01153046, + "auxiliary_loss_mlp": 0.01027126, + "balance_loss_clip": 1.0428704, + "balance_loss_mlp": 1.01952648, + "epoch": 0.839415619551494, + "flos": 18442535961600.0, + "grad_norm": 1.8004133854377673, + "language_loss": 0.83822513, + "learning_rate": 2.643254436582669e-07, + "loss": 0.86002684, + "num_input_tokens_seen": 150769240, + "step": 6981, + "time_per_iteration": 2.7244622707366943 + }, + { + "auxiliary_loss_clip": 0.01133045, + "auxiliary_loss_mlp": 0.01030268, + "balance_loss_clip": 1.04126954, + "balance_loss_mlp": 1.02188134, + "epoch": 0.8395358624421331, + "flos": 23221958705280.0, + "grad_norm": 1.7380816073919911, + "language_loss": 0.82196581, + "learning_rate": 2.6393854440286743e-07, + "loss": 0.84359896, + "num_input_tokens_seen": 150788410, + "step": 6982, + "time_per_iteration": 3.005112648010254 + }, + { + "auxiliary_loss_clip": 0.01169345, + "auxiliary_loss_mlp": 0.01028431, + "balance_loss_clip": 1.04845238, + "balance_loss_mlp": 1.02111125, + "epoch": 0.8396561053327722, + "flos": 24381657210240.0, + "grad_norm": 1.8702136596935723, + "language_loss": 0.70360959, + "learning_rate": 2.6355190850661045e-07, + "loss": 0.72558737, + "num_input_tokens_seen": 150805245, + "step": 6983, + "time_per_iteration": 2.7243752479553223 + }, + { + "auxiliary_loss_clip": 0.01150007, + "auxiliary_loss_mlp": 0.01026436, + "balance_loss_clip": 1.04642558, + "balance_loss_mlp": 1.01887178, + "epoch": 0.8397763482234113, + "flos": 22237755073920.0, + "grad_norm": 1.702483038622019, + "language_loss": 0.8667649, + "learning_rate": 2.631655360281486e-07, + "loss": 0.8885293, + "num_input_tokens_seen": 150824920, + "step": 6984, + "time_per_iteration": 2.7750966548919678 + }, + { + "auxiliary_loss_clip": 0.01156378, + "auxiliary_loss_mlp": 0.00886597, + "balance_loss_clip": 1.04651487, + "balance_loss_mlp": 1.00061774, + "epoch": 0.8398965911140504, + "flos": 22163743100160.0, + "grad_norm": 1.8014101581919761, + "language_loss": 0.65345442, + "learning_rate": 2.6277942702609323e-07, + "loss": 0.67388415, + "num_input_tokens_seen": 150844400, + "step": 6985, + "time_per_iteration": 2.861689805984497 + }, + { + "auxiliary_loss_clip": 0.01142344, + "auxiliary_loss_mlp": 0.01031941, + "balance_loss_clip": 1.04352343, + "balance_loss_mlp": 1.02394772, + "epoch": 0.8400168340046895, + "flos": 21542775753600.0, + "grad_norm": 1.8551156633466683, + "language_loss": 0.87116891, + "learning_rate": 2.623935815590186e-07, + "loss": 0.89291179, + "num_input_tokens_seen": 150862780, + "step": 6986, + "time_per_iteration": 2.7540767192840576 + }, + { + "auxiliary_loss_clip": 0.0115164, + "auxiliary_loss_mlp": 0.01033673, + "balance_loss_clip": 1.04792738, + "balance_loss_mlp": 1.02622271, + "epoch": 0.8401370768953286, + "flos": 22491966602880.0, + "grad_norm": 1.9422246779324837, + "language_loss": 0.80927789, + "learning_rate": 2.6200799968545516e-07, + "loss": 0.83113104, + "num_input_tokens_seen": 150883075, + "step": 6987, + "time_per_iteration": 2.780195713043213 + }, + { + "auxiliary_loss_clip": 0.0104646, + "auxiliary_loss_mlp": 0.01002527, + "balance_loss_clip": 1.00935793, + "balance_loss_mlp": 1.0010314, + "epoch": 0.8402573197859676, + "flos": 59238890818560.0, + "grad_norm": 0.7875530719534418, + "language_loss": 0.56441516, + "learning_rate": 2.616226814638969e-07, + "loss": 0.58490497, + "num_input_tokens_seen": 150948180, + "step": 6988, + "time_per_iteration": 3.3382108211517334 + }, + { + "auxiliary_loss_clip": 0.01144229, + "auxiliary_loss_mlp": 0.01026345, + "balance_loss_clip": 1.04582512, + "balance_loss_mlp": 1.01857281, + "epoch": 0.8403775626766068, + "flos": 22674608282880.0, + "grad_norm": 1.8526586762825965, + "language_loss": 0.77504337, + "learning_rate": 2.612376269527954e-07, + "loss": 0.79674911, + "num_input_tokens_seen": 150967885, + "step": 6989, + "time_per_iteration": 2.8409202098846436 + }, + { + "auxiliary_loss_clip": 0.01145958, + "auxiliary_loss_mlp": 0.01028538, + "balance_loss_clip": 1.04477644, + "balance_loss_mlp": 1.02018785, + "epoch": 0.8404978055672458, + "flos": 19609704495360.0, + "grad_norm": 1.838534652933294, + "language_loss": 0.67441308, + "learning_rate": 2.608528362105635e-07, + "loss": 0.69615811, + "num_input_tokens_seen": 150987255, + "step": 6990, + "time_per_iteration": 2.7862017154693604 + }, + { + "auxiliary_loss_clip": 0.01140791, + "auxiliary_loss_mlp": 0.0103514, + "balance_loss_clip": 1.03996456, + "balance_loss_mlp": 1.02704549, + "epoch": 0.8406180484578849, + "flos": 27526929678720.0, + "grad_norm": 1.7508266463179905, + "language_loss": 0.73620993, + "learning_rate": 2.6046830929557374e-07, + "loss": 0.75796926, + "num_input_tokens_seen": 151006905, + "step": 6991, + "time_per_iteration": 2.860635995864868 + }, + { + "auxiliary_loss_clip": 0.0113674, + "auxiliary_loss_mlp": 0.01026795, + "balance_loss_clip": 1.04275799, + "balance_loss_mlp": 1.0185039, + "epoch": 0.8407382913485241, + "flos": 22127473342080.0, + "grad_norm": 5.57983351791814, + "language_loss": 0.84872389, + "learning_rate": 2.6008404626615776e-07, + "loss": 0.87035924, + "num_input_tokens_seen": 151025405, + "step": 6992, + "time_per_iteration": 2.9238522052764893 + }, + { + "auxiliary_loss_clip": 0.01162778, + "auxiliary_loss_mlp": 0.01030921, + "balance_loss_clip": 1.04613554, + "balance_loss_mlp": 1.02336955, + "epoch": 0.8408585342391631, + "flos": 13918473982080.0, + "grad_norm": 2.320748003081078, + "language_loss": 0.73816013, + "learning_rate": 2.597000471806092e-07, + "loss": 0.76009715, + "num_input_tokens_seen": 151041970, + "step": 6993, + "time_per_iteration": 2.813467025756836 + }, + { + "auxiliary_loss_clip": 0.01148308, + "auxiliary_loss_mlp": 0.0102929, + "balance_loss_clip": 1.04794788, + "balance_loss_mlp": 1.02111804, + "epoch": 0.8409787771298022, + "flos": 20187865808640.0, + "grad_norm": 2.1119621362273797, + "language_loss": 0.72951722, + "learning_rate": 2.593163120971793e-07, + "loss": 0.75129318, + "num_input_tokens_seen": 151060835, + "step": 6994, + "time_per_iteration": 2.82710337638855 + }, + { + "auxiliary_loss_clip": 0.01121259, + "auxiliary_loss_mlp": 0.01026855, + "balance_loss_clip": 1.03766668, + "balance_loss_mlp": 1.01915431, + "epoch": 0.8410990200204413, + "flos": 23142523777920.0, + "grad_norm": 2.0124551296863573, + "language_loss": 0.68934667, + "learning_rate": 2.5893284107408165e-07, + "loss": 0.71082777, + "num_input_tokens_seen": 151078205, + "step": 6995, + "time_per_iteration": 3.8087923526763916 + }, + { + "auxiliary_loss_clip": 0.01130076, + "auxiliary_loss_mlp": 0.01027864, + "balance_loss_clip": 1.043203, + "balance_loss_mlp": 1.02022874, + "epoch": 0.8412192629110804, + "flos": 24027219757440.0, + "grad_norm": 1.7021501088355175, + "language_loss": 0.77832186, + "learning_rate": 2.5854963416948726e-07, + "loss": 0.79990125, + "num_input_tokens_seen": 151100470, + "step": 6996, + "time_per_iteration": 3.0048739910125732 + }, + { + "auxiliary_loss_clip": 0.01120981, + "auxiliary_loss_mlp": 0.01030105, + "balance_loss_clip": 1.03646791, + "balance_loss_mlp": 1.02150977, + "epoch": 0.8413395058017195, + "flos": 25591703604480.0, + "grad_norm": 1.7407296774451309, + "language_loss": 0.69322538, + "learning_rate": 2.5816669144152816e-07, + "loss": 0.71473622, + "num_input_tokens_seen": 151121650, + "step": 6997, + "time_per_iteration": 2.9141111373901367 + }, + { + "auxiliary_loss_clip": 0.01056267, + "auxiliary_loss_mlp": 0.01000981, + "balance_loss_clip": 1.00549555, + "balance_loss_mlp": 0.99962771, + "epoch": 0.8414597486923585, + "flos": 63635396624640.0, + "grad_norm": 0.859679143125677, + "language_loss": 0.66315842, + "learning_rate": 2.5778401294829777e-07, + "loss": 0.68373096, + "num_input_tokens_seen": 151180390, + "step": 6998, + "time_per_iteration": 4.207540512084961 + }, + { + "auxiliary_loss_clip": 0.01157742, + "auxiliary_loss_mlp": 0.00885712, + "balance_loss_clip": 1.04698837, + "balance_loss_mlp": 1.00052476, + "epoch": 0.8415799915829977, + "flos": 19098731571840.0, + "grad_norm": 1.6283028603489496, + "language_loss": 0.64825547, + "learning_rate": 2.574015987478473e-07, + "loss": 0.66868997, + "num_input_tokens_seen": 151198520, + "step": 6999, + "time_per_iteration": 3.669487237930298 + }, + { + "auxiliary_loss_clip": 0.01156873, + "auxiliary_loss_mlp": 0.01034489, + "balance_loss_clip": 1.04600608, + "balance_loss_mlp": 1.02609658, + "epoch": 0.8417002344736367, + "flos": 19821612781440.0, + "grad_norm": 3.229278561116962, + "language_loss": 0.86718357, + "learning_rate": 2.570194488981887e-07, + "loss": 0.88909721, + "num_input_tokens_seen": 151215065, + "step": 7000, + "time_per_iteration": 2.7945785522460938 + }, + { + "auxiliary_loss_clip": 0.0105617, + "auxiliary_loss_mlp": 0.01001201, + "balance_loss_clip": 1.00532889, + "balance_loss_mlp": 0.99984759, + "epoch": 0.8418204773642758, + "flos": 62161516834560.0, + "grad_norm": 0.8399104147631801, + "language_loss": 0.60389185, + "learning_rate": 2.566375634572939e-07, + "loss": 0.62446558, + "num_input_tokens_seen": 151275705, + "step": 7001, + "time_per_iteration": 3.1919562816619873 + }, + { + "auxiliary_loss_clip": 0.01146437, + "auxiliary_loss_mlp": 0.01034241, + "balance_loss_clip": 1.04296517, + "balance_loss_mlp": 1.02628422, + "epoch": 0.841940720254915, + "flos": 17092905315840.0, + "grad_norm": 1.7114759537397424, + "language_loss": 0.7626245, + "learning_rate": 2.562559424830943e-07, + "loss": 0.78443134, + "num_input_tokens_seen": 151293665, + "step": 7002, + "time_per_iteration": 2.9448232650756836 + }, + { + "auxiliary_loss_clip": 0.01146596, + "auxiliary_loss_mlp": 0.01027997, + "balance_loss_clip": 1.0440104, + "balance_loss_mlp": 1.01989663, + "epoch": 0.842060963145554, + "flos": 16283586026880.0, + "grad_norm": 2.5003229364770325, + "language_loss": 0.70107818, + "learning_rate": 2.5587458603348256e-07, + "loss": 0.7228241, + "num_input_tokens_seen": 151310955, + "step": 7003, + "time_per_iteration": 2.8247382640838623 + }, + { + "auxiliary_loss_clip": 0.01135523, + "auxiliary_loss_mlp": 0.01023149, + "balance_loss_clip": 1.04167473, + "balance_loss_mlp": 1.01520991, + "epoch": 0.8421812060361931, + "flos": 21908238681600.0, + "grad_norm": 1.785086855162189, + "language_loss": 0.8396073, + "learning_rate": 2.554934941663085e-07, + "loss": 0.86119401, + "num_input_tokens_seen": 151328490, + "step": 7004, + "time_per_iteration": 2.8638916015625 + }, + { + "auxiliary_loss_clip": 0.01139211, + "auxiliary_loss_mlp": 0.01024774, + "balance_loss_clip": 1.04126358, + "balance_loss_mlp": 1.01648259, + "epoch": 0.8423014489268322, + "flos": 27777693502080.0, + "grad_norm": 5.6724307777443, + "language_loss": 0.73490691, + "learning_rate": 2.5511266693938484e-07, + "loss": 0.75654674, + "num_input_tokens_seen": 151346950, + "step": 7005, + "time_per_iteration": 3.8915610313415527 + }, + { + "auxiliary_loss_clip": 0.01150445, + "auxiliary_loss_mlp": 0.01027442, + "balance_loss_clip": 1.04598892, + "balance_loss_mlp": 1.01931763, + "epoch": 0.8424216918174713, + "flos": 25117610970240.0, + "grad_norm": 1.5848351171415864, + "language_loss": 0.77756202, + "learning_rate": 2.547321044104822e-07, + "loss": 0.7993409, + "num_input_tokens_seen": 151368445, + "step": 7006, + "time_per_iteration": 2.7958765029907227 + }, + { + "auxiliary_loss_clip": 0.01174719, + "auxiliary_loss_mlp": 0.01030083, + "balance_loss_clip": 1.05059743, + "balance_loss_mlp": 1.02150559, + "epoch": 0.8425419347081103, + "flos": 24748448941440.0, + "grad_norm": 1.830568455427657, + "language_loss": 0.76892054, + "learning_rate": 2.5435180663733113e-07, + "loss": 0.79096854, + "num_input_tokens_seen": 151388745, + "step": 7007, + "time_per_iteration": 2.720982313156128 + }, + { + "auxiliary_loss_clip": 0.01131076, + "auxiliary_loss_mlp": 0.01026008, + "balance_loss_clip": 1.04032445, + "balance_loss_mlp": 1.01819992, + "epoch": 0.8426621775987495, + "flos": 24820916630400.0, + "grad_norm": 3.086906647412224, + "language_loss": 0.71868622, + "learning_rate": 2.539717736776241e-07, + "loss": 0.74025702, + "num_input_tokens_seen": 151404970, + "step": 7008, + "time_per_iteration": 2.9191484451293945 + }, + { + "auxiliary_loss_clip": 0.01157991, + "auxiliary_loss_mlp": 0.01022677, + "balance_loss_clip": 1.04675555, + "balance_loss_mlp": 1.01510096, + "epoch": 0.8427824204893886, + "flos": 23550074467200.0, + "grad_norm": 1.961144789977891, + "language_loss": 0.76509023, + "learning_rate": 2.535920055890097e-07, + "loss": 0.78689694, + "num_input_tokens_seen": 151426265, + "step": 7009, + "time_per_iteration": 2.7906112670898438 + }, + { + "auxiliary_loss_clip": 0.0112829, + "auxiliary_loss_mlp": 0.01029572, + "balance_loss_clip": 1.04252696, + "balance_loss_mlp": 1.02111411, + "epoch": 0.8429026633800276, + "flos": 16143858120960.0, + "grad_norm": 2.1050222459191748, + "language_loss": 0.64680719, + "learning_rate": 2.5321250242910006e-07, + "loss": 0.66838586, + "num_input_tokens_seen": 151444180, + "step": 7010, + "time_per_iteration": 2.8002865314483643 + }, + { + "auxiliary_loss_clip": 0.01172654, + "auxiliary_loss_mlp": 0.01025498, + "balance_loss_clip": 1.05084097, + "balance_loss_mlp": 1.01816654, + "epoch": 0.8430229062706668, + "flos": 22198540400640.0, + "grad_norm": 1.6520442420537926, + "language_loss": 0.86288911, + "learning_rate": 2.5283326425546493e-07, + "loss": 0.88487065, + "num_input_tokens_seen": 151463290, + "step": 7011, + "time_per_iteration": 2.6022419929504395 + }, + { + "auxiliary_loss_clip": 0.0113266, + "auxiliary_loss_mlp": 0.01027158, + "balance_loss_clip": 1.04496765, + "balance_loss_mlp": 1.0201695, + "epoch": 0.8431431491613058, + "flos": 35330317683840.0, + "grad_norm": 1.9226496538749032, + "language_loss": 0.70073009, + "learning_rate": 2.5245429112563443e-07, + "loss": 0.72232831, + "num_input_tokens_seen": 151483965, + "step": 7012, + "time_per_iteration": 2.9430723190307617 + }, + { + "auxiliary_loss_clip": 0.01157698, + "auxiliary_loss_mlp": 0.01028502, + "balance_loss_clip": 1.04625261, + "balance_loss_mlp": 1.02111709, + "epoch": 0.8432633920519449, + "flos": 25812374808960.0, + "grad_norm": 4.372215781023085, + "language_loss": 0.81951284, + "learning_rate": 2.5207558309709865e-07, + "loss": 0.84137487, + "num_input_tokens_seen": 151503700, + "step": 7013, + "time_per_iteration": 2.6984610557556152 + }, + { + "auxiliary_loss_clip": 0.01044352, + "auxiliary_loss_mlp": 0.00876698, + "balance_loss_clip": 1.00576854, + "balance_loss_mlp": 1.00118053, + "epoch": 0.8433836349425841, + "flos": 64959531592320.0, + "grad_norm": 0.6594422123676215, + "language_loss": 0.56324339, + "learning_rate": 2.516971402273065e-07, + "loss": 0.58245385, + "num_input_tokens_seen": 151569765, + "step": 7014, + "time_per_iteration": 3.454909563064575 + }, + { + "auxiliary_loss_clip": 0.01151648, + "auxiliary_loss_mlp": 0.01025917, + "balance_loss_clip": 1.04463434, + "balance_loss_mlp": 1.01762593, + "epoch": 0.8435038778332231, + "flos": 20229989483520.0, + "grad_norm": 1.8690538897252262, + "language_loss": 0.67679483, + "learning_rate": 2.513189625736687e-07, + "loss": 0.69857049, + "num_input_tokens_seen": 151586660, + "step": 7015, + "time_per_iteration": 2.7836191654205322 + }, + { + "auxiliary_loss_clip": 0.01145916, + "auxiliary_loss_mlp": 0.01032242, + "balance_loss_clip": 1.04417849, + "balance_loss_mlp": 1.02391505, + "epoch": 0.8436241207238622, + "flos": 20992229020800.0, + "grad_norm": 2.3839066779222757, + "language_loss": 0.71955776, + "learning_rate": 2.509410501935534e-07, + "loss": 0.74133933, + "num_input_tokens_seen": 151602295, + "step": 7016, + "time_per_iteration": 2.8375954627990723 + }, + { + "auxiliary_loss_clip": 0.01158278, + "auxiliary_loss_mlp": 0.01033201, + "balance_loss_clip": 1.04921222, + "balance_loss_mlp": 1.02499068, + "epoch": 0.8437443636145013, + "flos": 14682257804160.0, + "grad_norm": 2.683525770998386, + "language_loss": 0.7541405, + "learning_rate": 2.5056340314429116e-07, + "loss": 0.77605534, + "num_input_tokens_seen": 151619760, + "step": 7017, + "time_per_iteration": 2.8506882190704346 + }, + { + "auxiliary_loss_clip": 0.01135752, + "auxiliary_loss_mlp": 0.01030672, + "balance_loss_clip": 1.0418849, + "balance_loss_mlp": 1.0222137, + "epoch": 0.8438646065051404, + "flos": 21608814908160.0, + "grad_norm": 2.1314935474012957, + "language_loss": 0.80329674, + "learning_rate": 2.5018602148316904e-07, + "loss": 0.82496095, + "num_input_tokens_seen": 151635795, + "step": 7018, + "time_per_iteration": 2.8523473739624023 + }, + { + "auxiliary_loss_clip": 0.01168133, + "auxiliary_loss_mlp": 0.01030614, + "balance_loss_clip": 1.04932618, + "balance_loss_mlp": 1.02231097, + "epoch": 0.8439848493957794, + "flos": 23289937194240.0, + "grad_norm": 2.0241318044503798, + "language_loss": 0.80575752, + "learning_rate": 2.498089052674359e-07, + "loss": 0.82774496, + "num_input_tokens_seen": 151653770, + "step": 7019, + "time_per_iteration": 2.7295804023742676 + }, + { + "auxiliary_loss_clip": 0.01162328, + "auxiliary_loss_mlp": 0.01029652, + "balance_loss_clip": 1.0492084, + "balance_loss_mlp": 1.02183199, + "epoch": 0.8441050922864186, + "flos": 19719339782400.0, + "grad_norm": 1.8693969789383724, + "language_loss": 0.75059223, + "learning_rate": 2.494320545543007e-07, + "loss": 0.77251208, + "num_input_tokens_seen": 151673340, + "step": 7020, + "time_per_iteration": 2.7469980716705322 + }, + { + "auxiliary_loss_clip": 0.01173558, + "auxiliary_loss_mlp": 0.01028791, + "balance_loss_clip": 1.04896402, + "balance_loss_mlp": 1.02089357, + "epoch": 0.8442253351770577, + "flos": 21835268202240.0, + "grad_norm": 1.6785491490985676, + "language_loss": 0.66339099, + "learning_rate": 2.490554694009308e-07, + "loss": 0.68541449, + "num_input_tokens_seen": 151694205, + "step": 7021, + "time_per_iteration": 3.648951768875122 + }, + { + "auxiliary_loss_clip": 0.01164233, + "auxiliary_loss_mlp": 0.01022413, + "balance_loss_clip": 1.04580498, + "balance_loss_mlp": 1.01529646, + "epoch": 0.8443455780676967, + "flos": 34346365447680.0, + "grad_norm": 1.9958882190520062, + "language_loss": 0.78411567, + "learning_rate": 2.4867914986445426e-07, + "loss": 0.80598211, + "num_input_tokens_seen": 151716595, + "step": 7022, + "time_per_iteration": 2.8478450775146484 + }, + { + "auxiliary_loss_clip": 0.01153771, + "auxiliary_loss_mlp": 0.01025858, + "balance_loss_clip": 1.04281521, + "balance_loss_mlp": 1.01830029, + "epoch": 0.8444658209583359, + "flos": 48214599281280.0, + "grad_norm": 1.9838153509023448, + "language_loss": 0.70978922, + "learning_rate": 2.483030960019581e-07, + "loss": 0.7315855, + "num_input_tokens_seen": 151740525, + "step": 7023, + "time_per_iteration": 2.976363182067871 + }, + { + "auxiliary_loss_clip": 0.01033003, + "auxiliary_loss_mlp": 0.01005354, + "balance_loss_clip": 1.00597429, + "balance_loss_mlp": 1.00388169, + "epoch": 0.8445860638489749, + "flos": 68484773105280.0, + "grad_norm": 0.733651102256556, + "language_loss": 0.55458987, + "learning_rate": 2.479273078704891e-07, + "loss": 0.57497334, + "num_input_tokens_seen": 151793890, + "step": 7024, + "time_per_iteration": 4.154437065124512 + }, + { + "auxiliary_loss_clip": 0.01032303, + "auxiliary_loss_mlp": 0.01001041, + "balance_loss_clip": 1.01121569, + "balance_loss_mlp": 0.99961692, + "epoch": 0.844706306739614, + "flos": 62833331882880.0, + "grad_norm": 0.8496182843087503, + "language_loss": 0.64741683, + "learning_rate": 2.475517855270552e-07, + "loss": 0.6677503, + "num_input_tokens_seen": 151853970, + "step": 7025, + "time_per_iteration": 4.205377578735352 + }, + { + "auxiliary_loss_clip": 0.01168459, + "auxiliary_loss_mlp": 0.01026403, + "balance_loss_clip": 1.0484916, + "balance_loss_mlp": 1.0191164, + "epoch": 0.8448265496302532, + "flos": 14976114969600.0, + "grad_norm": 1.9740774093736455, + "language_loss": 0.72423482, + "learning_rate": 2.4717652902862143e-07, + "loss": 0.74618345, + "num_input_tokens_seen": 151872945, + "step": 7026, + "time_per_iteration": 2.6701161861419678 + }, + { + "auxiliary_loss_clip": 0.01145849, + "auxiliary_loss_mlp": 0.0102585, + "balance_loss_clip": 1.04414392, + "balance_loss_mlp": 1.01767516, + "epoch": 0.8449467925208922, + "flos": 23441265192960.0, + "grad_norm": 1.7789534320942737, + "language_loss": 0.8154906, + "learning_rate": 2.4680153843211495e-07, + "loss": 0.83720756, + "num_input_tokens_seen": 151892875, + "step": 7027, + "time_per_iteration": 2.7659456729888916 + }, + { + "auxiliary_loss_clip": 0.01150106, + "auxiliary_loss_mlp": 0.01023716, + "balance_loss_clip": 1.04686546, + "balance_loss_mlp": 1.01621735, + "epoch": 0.8450670354115313, + "flos": 22748045639040.0, + "grad_norm": 1.7131409636744488, + "language_loss": 0.72588336, + "learning_rate": 2.464268137944212e-07, + "loss": 0.7476216, + "num_input_tokens_seen": 151914170, + "step": 7028, + "time_per_iteration": 2.852752685546875 + }, + { + "auxiliary_loss_clip": 0.01118789, + "auxiliary_loss_mlp": 0.01026656, + "balance_loss_clip": 1.03977573, + "balance_loss_mlp": 1.01812708, + "epoch": 0.8451872783021703, + "flos": 29825571605760.0, + "grad_norm": 1.8590484364011732, + "language_loss": 0.78228009, + "learning_rate": 2.46052355172385e-07, + "loss": 0.80373454, + "num_input_tokens_seen": 151932210, + "step": 7029, + "time_per_iteration": 2.8038830757141113 + }, + { + "auxiliary_loss_clip": 0.01170576, + "auxiliary_loss_mlp": 0.01025401, + "balance_loss_clip": 1.04687047, + "balance_loss_mlp": 1.01747394, + "epoch": 0.8453075211928095, + "flos": 21870029589120.0, + "grad_norm": 1.8994842748122591, + "language_loss": 0.74675858, + "learning_rate": 2.456781626228128e-07, + "loss": 0.76871836, + "num_input_tokens_seen": 151951715, + "step": 7030, + "time_per_iteration": 3.5603084564208984 + }, + { + "auxiliary_loss_clip": 0.01031483, + "auxiliary_loss_mlp": 0.00876772, + "balance_loss_clip": 1.00550258, + "balance_loss_mlp": 1.00114334, + "epoch": 0.8454277640834486, + "flos": 58751869288320.0, + "grad_norm": 0.9168125530518325, + "language_loss": 0.6633088, + "learning_rate": 2.453042362024675e-07, + "loss": 0.68239135, + "num_input_tokens_seen": 152004960, + "step": 7031, + "time_per_iteration": 3.3579421043395996 + }, + { + "auxiliary_loss_clip": 0.01167854, + "auxiliary_loss_mlp": 0.01024851, + "balance_loss_clip": 1.04597199, + "balance_loss_mlp": 1.01756191, + "epoch": 0.8455480069740876, + "flos": 27090076469760.0, + "grad_norm": 1.4634390967304525, + "language_loss": 0.73311174, + "learning_rate": 2.449305759680751e-07, + "loss": 0.7550388, + "num_input_tokens_seen": 152026285, + "step": 7032, + "time_per_iteration": 2.674605369567871 + }, + { + "auxiliary_loss_clip": 0.011384, + "auxiliary_loss_mlp": 0.01027598, + "balance_loss_clip": 1.04382825, + "balance_loss_mlp": 1.01898515, + "epoch": 0.8456682498647268, + "flos": 27198670262400.0, + "grad_norm": 1.4927261163998253, + "language_loss": 0.75224662, + "learning_rate": 2.445571819763188e-07, + "loss": 0.77390659, + "num_input_tokens_seen": 152048585, + "step": 7033, + "time_per_iteration": 2.888862133026123 + }, + { + "auxiliary_loss_clip": 0.01168533, + "auxiliary_loss_mlp": 0.01030779, + "balance_loss_clip": 1.04790306, + "balance_loss_mlp": 1.02272677, + "epoch": 0.8457884927553658, + "flos": 20631901737600.0, + "grad_norm": 1.617860567747778, + "language_loss": 0.58646542, + "learning_rate": 2.4418405428384227e-07, + "loss": 0.60845852, + "num_input_tokens_seen": 152068795, + "step": 7034, + "time_per_iteration": 2.629869222640991 + }, + { + "auxiliary_loss_clip": 0.01166073, + "auxiliary_loss_mlp": 0.00886093, + "balance_loss_clip": 1.04545414, + "balance_loss_mlp": 1.00054729, + "epoch": 0.8459087356460049, + "flos": 15299023259520.0, + "grad_norm": 1.7887172952848591, + "language_loss": 0.71641815, + "learning_rate": 2.4381119294724864e-07, + "loss": 0.73693991, + "num_input_tokens_seen": 152086240, + "step": 7035, + "time_per_iteration": 2.6716701984405518 + }, + { + "auxiliary_loss_clip": 0.01170507, + "auxiliary_loss_mlp": 0.01026382, + "balance_loss_clip": 1.04904127, + "balance_loss_mlp": 1.01875901, + "epoch": 0.846028978536644, + "flos": 18843155326080.0, + "grad_norm": 1.8169531196067357, + "language_loss": 0.53603518, + "learning_rate": 2.434385980231004e-07, + "loss": 0.55800402, + "num_input_tokens_seen": 152105080, + "step": 7036, + "time_per_iteration": 2.6683318614959717 + }, + { + "auxiliary_loss_clip": 0.0115777, + "auxiliary_loss_mlp": 0.01023582, + "balance_loss_clip": 1.04538774, + "balance_loss_mlp": 1.01639974, + "epoch": 0.8461492214272831, + "flos": 52661740285440.0, + "grad_norm": 1.591448620536241, + "language_loss": 0.65691245, + "learning_rate": 2.4306626956792043e-07, + "loss": 0.67872596, + "num_input_tokens_seen": 152130025, + "step": 7037, + "time_per_iteration": 3.0393548011779785 + }, + { + "auxiliary_loss_clip": 0.01159565, + "auxiliary_loss_mlp": 0.0102392, + "balance_loss_clip": 1.04512417, + "balance_loss_mlp": 1.01576018, + "epoch": 0.8462694643179222, + "flos": 18588405093120.0, + "grad_norm": 1.7196851858040425, + "language_loss": 0.75785059, + "learning_rate": 2.4269420763819017e-07, + "loss": 0.77968544, + "num_input_tokens_seen": 152148070, + "step": 7038, + "time_per_iteration": 2.760693073272705 + }, + { + "auxiliary_loss_clip": 0.01154898, + "auxiliary_loss_mlp": 0.01023052, + "balance_loss_clip": 1.04342508, + "balance_loss_mlp": 1.01554203, + "epoch": 0.8463897072085613, + "flos": 24387080163840.0, + "grad_norm": 3.9213124694730683, + "language_loss": 0.83484602, + "learning_rate": 2.4232241229035223e-07, + "loss": 0.85662556, + "num_input_tokens_seen": 152165825, + "step": 7039, + "time_per_iteration": 2.8471808433532715 + }, + { + "auxiliary_loss_clip": 0.01051833, + "auxiliary_loss_mlp": 0.01002891, + "balance_loss_clip": 1.00516129, + "balance_loss_mlp": 1.00154352, + "epoch": 0.8465099500992004, + "flos": 68702140258560.0, + "grad_norm": 0.7463722191001343, + "language_loss": 0.56728971, + "learning_rate": 2.419508835808064e-07, + "loss": 0.58783698, + "num_input_tokens_seen": 152222380, + "step": 7040, + "time_per_iteration": 3.33720064163208 + }, + { + "auxiliary_loss_clip": 0.01148825, + "auxiliary_loss_mlp": 0.01024603, + "balance_loss_clip": 1.04537761, + "balance_loss_mlp": 1.01673508, + "epoch": 0.8466301929898394, + "flos": 13735724561280.0, + "grad_norm": 2.1189593383041485, + "language_loss": 0.62431371, + "learning_rate": 2.415796215659134e-07, + "loss": 0.64604807, + "num_input_tokens_seen": 152239085, + "step": 7041, + "time_per_iteration": 2.7469756603240967 + }, + { + "auxiliary_loss_clip": 0.01141115, + "auxiliary_loss_mlp": 0.01033189, + "balance_loss_clip": 1.03831851, + "balance_loss_mlp": 1.02539551, + "epoch": 0.8467504358804786, + "flos": 19241260738560.0, + "grad_norm": 1.835685653627334, + "language_loss": 0.77128106, + "learning_rate": 2.412086263019939e-07, + "loss": 0.79302412, + "num_input_tokens_seen": 152257110, + "step": 7042, + "time_per_iteration": 2.8416781425476074 + }, + { + "auxiliary_loss_clip": 0.01163946, + "auxiliary_loss_mlp": 0.01026234, + "balance_loss_clip": 1.04730785, + "balance_loss_mlp": 1.01937902, + "epoch": 0.8468706787711177, + "flos": 21324115710720.0, + "grad_norm": 1.96065300445104, + "language_loss": 0.8013202, + "learning_rate": 2.408378978453276e-07, + "loss": 0.82322204, + "num_input_tokens_seen": 152277230, + "step": 7043, + "time_per_iteration": 2.6700432300567627 + }, + { + "auxiliary_loss_clip": 0.01052131, + "auxiliary_loss_mlp": 0.01003096, + "balance_loss_clip": 1.00513601, + "balance_loss_mlp": 1.00173128, + "epoch": 0.8469909216617567, + "flos": 64877439058560.0, + "grad_norm": 0.8174997673682569, + "language_loss": 0.63925558, + "learning_rate": 2.404674362521533e-07, + "loss": 0.65980786, + "num_input_tokens_seen": 152335725, + "step": 7044, + "time_per_iteration": 3.1776952743530273 + }, + { + "auxiliary_loss_clip": 0.01158919, + "auxiliary_loss_mlp": 0.01028637, + "balance_loss_clip": 1.04712987, + "balance_loss_mlp": 1.02064967, + "epoch": 0.8471111645523959, + "flos": 19280583152640.0, + "grad_norm": 2.617761920768185, + "language_loss": 0.74540651, + "learning_rate": 2.4009724157866997e-07, + "loss": 0.76728213, + "num_input_tokens_seen": 152352785, + "step": 7045, + "time_per_iteration": 2.7251980304718018 + }, + { + "auxiliary_loss_clip": 0.01167165, + "auxiliary_loss_mlp": 0.01024539, + "balance_loss_clip": 1.04635155, + "balance_loss_mlp": 1.01750553, + "epoch": 0.8472314074430349, + "flos": 22015826893440.0, + "grad_norm": 2.1156679715843785, + "language_loss": 0.76439607, + "learning_rate": 2.3972731388103564e-07, + "loss": 0.78631312, + "num_input_tokens_seen": 152371265, + "step": 7046, + "time_per_iteration": 2.685084342956543 + }, + { + "auxiliary_loss_clip": 0.01019769, + "auxiliary_loss_mlp": 0.0100365, + "balance_loss_clip": 1.00584173, + "balance_loss_mlp": 1.00227284, + "epoch": 0.847351650333674, + "flos": 57882580243200.0, + "grad_norm": 0.8071061006633152, + "language_loss": 0.62358797, + "learning_rate": 2.393576532153687e-07, + "loss": 0.64382207, + "num_input_tokens_seen": 152435050, + "step": 7047, + "time_per_iteration": 4.444588899612427 + }, + { + "auxiliary_loss_clip": 0.01049591, + "auxiliary_loss_mlp": 0.01003392, + "balance_loss_clip": 1.00552118, + "balance_loss_mlp": 1.00207448, + "epoch": 0.8474718932243132, + "flos": 41284238313600.0, + "grad_norm": 0.937722494931121, + "language_loss": 0.57789516, + "learning_rate": 2.389882596377453e-07, + "loss": 0.59842503, + "num_input_tokens_seen": 152489315, + "step": 7048, + "time_per_iteration": 3.2302043437957764 + }, + { + "auxiliary_loss_clip": 0.01166898, + "auxiliary_loss_mlp": 0.01027563, + "balance_loss_clip": 1.04500091, + "balance_loss_mlp": 1.01927781, + "epoch": 0.8475921361149522, + "flos": 38180906974080.0, + "grad_norm": 1.8630628241437048, + "language_loss": 0.76486087, + "learning_rate": 2.386191332042031e-07, + "loss": 0.78680545, + "num_input_tokens_seen": 152511210, + "step": 7049, + "time_per_iteration": 2.8195102214813232 + }, + { + "auxiliary_loss_clip": 0.0117425, + "auxiliary_loss_mlp": 0.01026795, + "balance_loss_clip": 1.04968715, + "balance_loss_mlp": 1.01860857, + "epoch": 0.8477123790055913, + "flos": 25375054723200.0, + "grad_norm": 1.7287682181239765, + "language_loss": 0.7248075, + "learning_rate": 2.3825027397073794e-07, + "loss": 0.74681789, + "num_input_tokens_seen": 152531685, + "step": 7050, + "time_per_iteration": 3.7637085914611816 + }, + { + "auxiliary_loss_clip": 0.01155232, + "auxiliary_loss_mlp": 0.01024893, + "balance_loss_clip": 1.04733682, + "balance_loss_mlp": 1.01750755, + "epoch": 0.8478326218962304, + "flos": 30225185389440.0, + "grad_norm": 2.1050531130861776, + "language_loss": 0.66528547, + "learning_rate": 2.3788168199330515e-07, + "loss": 0.68708664, + "num_input_tokens_seen": 152553245, + "step": 7051, + "time_per_iteration": 3.6854124069213867 + }, + { + "auxiliary_loss_clip": 0.01134727, + "auxiliary_loss_mlp": 0.01024023, + "balance_loss_clip": 1.03829455, + "balance_loss_mlp": 1.01645875, + "epoch": 0.8479528647868695, + "flos": 38213800853760.0, + "grad_norm": 1.6372269084628448, + "language_loss": 0.72478569, + "learning_rate": 2.3751335732782074e-07, + "loss": 0.74637318, + "num_input_tokens_seen": 152574505, + "step": 7052, + "time_per_iteration": 2.919926166534424 + }, + { + "auxiliary_loss_clip": 0.01158734, + "auxiliary_loss_mlp": 0.01027255, + "balance_loss_clip": 1.04849696, + "balance_loss_mlp": 1.02014422, + "epoch": 0.8480731076775085, + "flos": 20957790856320.0, + "grad_norm": 1.825002101709725, + "language_loss": 0.79365385, + "learning_rate": 2.371453000301582e-07, + "loss": 0.81551373, + "num_input_tokens_seen": 152593190, + "step": 7053, + "time_per_iteration": 2.785433769226074 + }, + { + "auxiliary_loss_clip": 0.01135709, + "auxiliary_loss_mlp": 0.01027579, + "balance_loss_clip": 1.04249346, + "balance_loss_mlp": 1.02018845, + "epoch": 0.8481933505681477, + "flos": 32596510487040.0, + "grad_norm": 1.7523797545571196, + "language_loss": 0.74408448, + "learning_rate": 2.3677751015615222e-07, + "loss": 0.76571739, + "num_input_tokens_seen": 152615265, + "step": 7054, + "time_per_iteration": 2.787872552871704 + }, + { + "auxiliary_loss_clip": 0.01138814, + "auxiliary_loss_mlp": 0.01030704, + "balance_loss_clip": 1.04232454, + "balance_loss_mlp": 1.02281845, + "epoch": 0.8483135934587868, + "flos": 20741177888640.0, + "grad_norm": 1.7019814677093288, + "language_loss": 0.85251921, + "learning_rate": 2.3640998776159593e-07, + "loss": 0.87421441, + "num_input_tokens_seen": 152632770, + "step": 7055, + "time_per_iteration": 2.9329073429107666 + }, + { + "auxiliary_loss_clip": 0.01151482, + "auxiliary_loss_mlp": 0.01023738, + "balance_loss_clip": 1.04577208, + "balance_loss_mlp": 1.01685417, + "epoch": 0.8484338363494258, + "flos": 21653057485440.0, + "grad_norm": 1.6650437370398496, + "language_loss": 0.81360662, + "learning_rate": 2.3604273290224253e-07, + "loss": 0.83535886, + "num_input_tokens_seen": 152653485, + "step": 7056, + "time_per_iteration": 3.7492499351501465 + }, + { + "auxiliary_loss_clip": 0.01153262, + "auxiliary_loss_mlp": 0.01030525, + "balance_loss_clip": 1.04683685, + "balance_loss_mlp": 1.0225265, + "epoch": 0.848554079240065, + "flos": 15013964926080.0, + "grad_norm": 1.9264940751609072, + "language_loss": 0.74506396, + "learning_rate": 2.356757456338039e-07, + "loss": 0.76690185, + "num_input_tokens_seen": 152670970, + "step": 7057, + "time_per_iteration": 2.861180543899536 + }, + { + "auxiliary_loss_clip": 0.0104355, + "auxiliary_loss_mlp": 0.01002144, + "balance_loss_clip": 1.00919199, + "balance_loss_mlp": 1.00078475, + "epoch": 0.848674322130704, + "flos": 68060453742720.0, + "grad_norm": 0.7494427417366016, + "language_loss": 0.59010732, + "learning_rate": 2.3530902601195147e-07, + "loss": 0.61056435, + "num_input_tokens_seen": 152739460, + "step": 7058, + "time_per_iteration": 3.4971680641174316 + }, + { + "auxiliary_loss_clip": 0.01162187, + "auxiliary_loss_mlp": 0.01032636, + "balance_loss_clip": 1.04797411, + "balance_loss_mlp": 1.02421999, + "epoch": 0.8487945650213431, + "flos": 18475788977280.0, + "grad_norm": 2.568561203455433, + "language_loss": 0.78848886, + "learning_rate": 2.34942574092317e-07, + "loss": 0.81043708, + "num_input_tokens_seen": 152754710, + "step": 7059, + "time_per_iteration": 2.6530654430389404 + }, + { + "auxiliary_loss_clip": 0.01167412, + "auxiliary_loss_mlp": 0.01023939, + "balance_loss_clip": 1.04780483, + "balance_loss_mlp": 1.01619649, + "epoch": 0.8489148079119821, + "flos": 23473189405440.0, + "grad_norm": 1.9614043362937945, + "language_loss": 0.76708698, + "learning_rate": 2.3457638993049045e-07, + "loss": 0.78900051, + "num_input_tokens_seen": 152772700, + "step": 7060, + "time_per_iteration": 2.737461805343628 + }, + { + "auxiliary_loss_clip": 0.01122359, + "auxiliary_loss_mlp": 0.01028038, + "balance_loss_clip": 1.04324496, + "balance_loss_mlp": 1.01980698, + "epoch": 0.8490350508026213, + "flos": 19937604775680.0, + "grad_norm": 2.0184979161732577, + "language_loss": 0.64176726, + "learning_rate": 2.3421047358202252e-07, + "loss": 0.66327119, + "num_input_tokens_seen": 152791550, + "step": 7061, + "time_per_iteration": 2.957981586456299 + }, + { + "auxiliary_loss_clip": 0.0116144, + "auxiliary_loss_mlp": 0.01034615, + "balance_loss_clip": 1.04706573, + "balance_loss_mlp": 1.02683032, + "epoch": 0.8491552936932604, + "flos": 24279958828800.0, + "grad_norm": 3.0696868692761425, + "language_loss": 0.82973158, + "learning_rate": 2.3384482510242144e-07, + "loss": 0.85169208, + "num_input_tokens_seen": 152809410, + "step": 7062, + "time_per_iteration": 2.701094627380371 + }, + { + "auxiliary_loss_clip": 0.0116875, + "auxiliary_loss_mlp": 0.01027936, + "balance_loss_clip": 1.04483247, + "balance_loss_mlp": 1.01981807, + "epoch": 0.8492755365838994, + "flos": 22522526098560.0, + "grad_norm": 2.0942344137542466, + "language_loss": 0.77283609, + "learning_rate": 2.3347944454715575e-07, + "loss": 0.79480296, + "num_input_tokens_seen": 152825800, + "step": 7063, + "time_per_iteration": 2.699568510055542 + }, + { + "auxiliary_loss_clip": 0.01173354, + "auxiliary_loss_mlp": 0.01028884, + "balance_loss_clip": 1.04804087, + "balance_loss_mlp": 1.0207237, + "epoch": 0.8493957794745386, + "flos": 26980441182720.0, + "grad_norm": 1.6553350227558918, + "language_loss": 0.67274183, + "learning_rate": 2.331143319716542e-07, + "loss": 0.69476426, + "num_input_tokens_seen": 152845330, + "step": 7064, + "time_per_iteration": 2.651792287826538 + }, + { + "auxiliary_loss_clip": 0.01139315, + "auxiliary_loss_mlp": 0.01023675, + "balance_loss_clip": 1.0440948, + "balance_loss_mlp": 1.01571202, + "epoch": 0.8495160223651776, + "flos": 29861985018240.0, + "grad_norm": 1.8928074346814365, + "language_loss": 0.65715253, + "learning_rate": 2.3274948743130363e-07, + "loss": 0.67878246, + "num_input_tokens_seen": 152865165, + "step": 7065, + "time_per_iteration": 2.8650336265563965 + }, + { + "auxiliary_loss_clip": 0.01167676, + "auxiliary_loss_mlp": 0.0102345, + "balance_loss_clip": 1.04490376, + "balance_loss_mlp": 1.01537621, + "epoch": 0.8496362652558167, + "flos": 23075443128960.0, + "grad_norm": 1.6605912153176547, + "language_loss": 0.79303527, + "learning_rate": 2.3238491098145085e-07, + "loss": 0.81494647, + "num_input_tokens_seen": 152884695, + "step": 7066, + "time_per_iteration": 2.9770634174346924 + }, + { + "auxiliary_loss_clip": 0.01157903, + "auxiliary_loss_mlp": 0.01031058, + "balance_loss_clip": 1.04538465, + "balance_loss_mlp": 1.02334571, + "epoch": 0.8497565081464559, + "flos": 14609107756800.0, + "grad_norm": 2.336021684976828, + "language_loss": 0.72857523, + "learning_rate": 2.3202060267740141e-07, + "loss": 0.7504648, + "num_input_tokens_seen": 152902220, + "step": 7067, + "time_per_iteration": 2.7216641902923584 + }, + { + "auxiliary_loss_clip": 0.01128809, + "auxiliary_loss_mlp": 0.01026557, + "balance_loss_clip": 1.03946579, + "balance_loss_mlp": 1.01892722, + "epoch": 0.8498767510370949, + "flos": 21136446126720.0, + "grad_norm": 2.074562264327627, + "language_loss": 0.76742083, + "learning_rate": 2.3165656257442044e-07, + "loss": 0.78897452, + "num_input_tokens_seen": 152920740, + "step": 7068, + "time_per_iteration": 2.9056713581085205 + }, + { + "auxiliary_loss_clip": 0.01155043, + "auxiliary_loss_mlp": 0.01024958, + "balance_loss_clip": 1.04548895, + "balance_loss_mlp": 1.01743913, + "epoch": 0.849996993927734, + "flos": 23654538195840.0, + "grad_norm": 1.7771842194732446, + "language_loss": 0.89851403, + "learning_rate": 2.31292790727734e-07, + "loss": 0.92031401, + "num_input_tokens_seen": 152938305, + "step": 7069, + "time_per_iteration": 2.7155487537384033 + }, + { + "auxiliary_loss_clip": 0.0116566, + "auxiliary_loss_mlp": 0.01029899, + "balance_loss_clip": 1.04481292, + "balance_loss_mlp": 1.02243662, + "epoch": 0.8501172368183731, + "flos": 20558069331840.0, + "grad_norm": 2.1437587795235973, + "language_loss": 0.80037522, + "learning_rate": 2.3092928719252392e-07, + "loss": 0.82233083, + "num_input_tokens_seen": 152956705, + "step": 7070, + "time_per_iteration": 2.691263437271118 + }, + { + "auxiliary_loss_clip": 0.01159449, + "auxiliary_loss_mlp": 0.01023135, + "balance_loss_clip": 1.04585266, + "balance_loss_mlp": 1.01570272, + "epoch": 0.8502374797090122, + "flos": 22272624201600.0, + "grad_norm": 1.9876319068204555, + "language_loss": 0.78517854, + "learning_rate": 2.3056605202393475e-07, + "loss": 0.80700433, + "num_input_tokens_seen": 152974265, + "step": 7071, + "time_per_iteration": 2.6962873935699463 + }, + { + "auxiliary_loss_clip": 0.01153824, + "auxiliary_loss_mlp": 0.00886221, + "balance_loss_clip": 1.04192364, + "balance_loss_mlp": 1.00056291, + "epoch": 0.8503577225996513, + "flos": 23659817495040.0, + "grad_norm": 1.7448132161911853, + "language_loss": 0.66819942, + "learning_rate": 2.3020308527706888e-07, + "loss": 0.68859982, + "num_input_tokens_seen": 152993680, + "step": 7072, + "time_per_iteration": 2.7464792728424072 + }, + { + "auxiliary_loss_clip": 0.01154447, + "auxiliary_loss_mlp": 0.01024352, + "balance_loss_clip": 1.04303169, + "balance_loss_mlp": 1.01687777, + "epoch": 0.8504779654902904, + "flos": 26758513002240.0, + "grad_norm": 2.868446386523635, + "language_loss": 0.89082193, + "learning_rate": 2.2984038700698715e-07, + "loss": 0.91260993, + "num_input_tokens_seen": 153012990, + "step": 7073, + "time_per_iteration": 3.7950730323791504 + }, + { + "auxiliary_loss_clip": 0.01154315, + "auxiliary_loss_mlp": 0.01029841, + "balance_loss_clip": 1.04558945, + "balance_loss_mlp": 1.02243507, + "epoch": 0.8505982083809295, + "flos": 26468247196800.0, + "grad_norm": 1.843493994218175, + "language_loss": 0.79048514, + "learning_rate": 2.2947795726871222e-07, + "loss": 0.81232667, + "num_input_tokens_seen": 153034015, + "step": 7074, + "time_per_iteration": 2.7084145545959473 + }, + { + "auxiliary_loss_clip": 0.01158876, + "auxiliary_loss_mlp": 0.00885069, + "balance_loss_clip": 1.05058599, + "balance_loss_mlp": 1.00058556, + "epoch": 0.8507184512715685, + "flos": 20303390926080.0, + "grad_norm": 2.050234619101193, + "language_loss": 0.85677516, + "learning_rate": 2.2911579611722253e-07, + "loss": 0.87721455, + "num_input_tokens_seen": 153053160, + "step": 7075, + "time_per_iteration": 2.7737603187561035 + }, + { + "auxiliary_loss_clip": 0.0114026, + "auxiliary_loss_mlp": 0.01025916, + "balance_loss_clip": 1.04367447, + "balance_loss_mlp": 1.01869214, + "epoch": 0.8508386941622077, + "flos": 19025186474880.0, + "grad_norm": 1.8920344671508829, + "language_loss": 0.87443715, + "learning_rate": 2.2875390360745905e-07, + "loss": 0.89609885, + "num_input_tokens_seen": 153072565, + "step": 7076, + "time_per_iteration": 3.7016453742980957 + }, + { + "auxiliary_loss_clip": 0.0114628, + "auxiliary_loss_mlp": 0.01034102, + "balance_loss_clip": 1.0424118, + "balance_loss_mlp": 1.02522063, + "epoch": 0.8509589370528468, + "flos": 16433405654400.0, + "grad_norm": 2.7858750117612874, + "language_loss": 0.77726007, + "learning_rate": 2.2839227979432008e-07, + "loss": 0.79906392, + "num_input_tokens_seen": 153090215, + "step": 7077, + "time_per_iteration": 2.854445457458496 + }, + { + "auxiliary_loss_clip": 0.0115253, + "auxiliary_loss_mlp": 0.01023445, + "balance_loss_clip": 1.04449427, + "balance_loss_mlp": 1.01529145, + "epoch": 0.8510791799434858, + "flos": 18259714713600.0, + "grad_norm": 1.8277924797756182, + "language_loss": 0.85303539, + "learning_rate": 2.2803092473266373e-07, + "loss": 0.87479514, + "num_input_tokens_seen": 153107740, + "step": 7078, + "time_per_iteration": 3.771756887435913 + }, + { + "auxiliary_loss_clip": 0.01172005, + "auxiliary_loss_mlp": 0.01028354, + "balance_loss_clip": 1.0500344, + "balance_loss_mlp": 1.02043545, + "epoch": 0.851199422834125, + "flos": 23441372933760.0, + "grad_norm": 2.71228810997616, + "language_loss": 0.86603272, + "learning_rate": 2.2766983847730724e-07, + "loss": 0.88803625, + "num_input_tokens_seen": 153127410, + "step": 7079, + "time_per_iteration": 2.670128583908081 + }, + { + "auxiliary_loss_clip": 0.01152135, + "auxiliary_loss_mlp": 0.01032117, + "balance_loss_clip": 1.04510975, + "balance_loss_mlp": 1.02390933, + "epoch": 0.851319665724764, + "flos": 16289404030080.0, + "grad_norm": 1.9036424677292736, + "language_loss": 0.66425085, + "learning_rate": 2.2730902108302663e-07, + "loss": 0.68609333, + "num_input_tokens_seen": 153144325, + "step": 7080, + "time_per_iteration": 2.966655969619751 + }, + { + "auxiliary_loss_clip": 0.01143725, + "auxiliary_loss_mlp": 0.01035187, + "balance_loss_clip": 1.04187512, + "balance_loss_mlp": 1.02716398, + "epoch": 0.8514399086154031, + "flos": 18989347680000.0, + "grad_norm": 1.8609784463855783, + "language_loss": 0.68328601, + "learning_rate": 2.269484726045583e-07, + "loss": 0.70507514, + "num_input_tokens_seen": 153163240, + "step": 7081, + "time_per_iteration": 2.937155246734619 + }, + { + "auxiliary_loss_clip": 0.01144214, + "auxiliary_loss_mlp": 0.01034566, + "balance_loss_clip": 1.0436703, + "balance_loss_mlp": 1.02696645, + "epoch": 0.8515601515060423, + "flos": 24571194301440.0, + "grad_norm": 1.6158557842172356, + "language_loss": 0.79009151, + "learning_rate": 2.2658819309659672e-07, + "loss": 0.81187928, + "num_input_tokens_seen": 153183440, + "step": 7082, + "time_per_iteration": 3.845686197280884 + }, + { + "auxiliary_loss_clip": 0.01144464, + "auxiliary_loss_mlp": 0.01021974, + "balance_loss_clip": 1.04495192, + "balance_loss_mlp": 1.015239, + "epoch": 0.8516803943966813, + "flos": 19529443555200.0, + "grad_norm": 1.8539185717108115, + "language_loss": 0.84630126, + "learning_rate": 2.2622818261379706e-07, + "loss": 0.86796564, + "num_input_tokens_seen": 153200460, + "step": 7083, + "time_per_iteration": 2.851834535598755 + }, + { + "auxiliary_loss_clip": 0.01148534, + "auxiliary_loss_mlp": 0.01031737, + "balance_loss_clip": 1.04399562, + "balance_loss_mlp": 1.02330256, + "epoch": 0.8518006372873204, + "flos": 20265792364800.0, + "grad_norm": 1.6345186928807582, + "language_loss": 0.7497381, + "learning_rate": 2.2586844121077142e-07, + "loss": 0.77154088, + "num_input_tokens_seen": 153218970, + "step": 7084, + "time_per_iteration": 2.8177788257598877 + }, + { + "auxiliary_loss_clip": 0.01124833, + "auxiliary_loss_mlp": 0.01027475, + "balance_loss_clip": 1.04069495, + "balance_loss_mlp": 1.01995909, + "epoch": 0.8519208801779595, + "flos": 24133227770880.0, + "grad_norm": 1.872832847110266, + "language_loss": 0.71714664, + "learning_rate": 2.2550896894209215e-07, + "loss": 0.73866975, + "num_input_tokens_seen": 153238485, + "step": 7085, + "time_per_iteration": 2.8472280502319336 + }, + { + "auxiliary_loss_clip": 0.01033152, + "auxiliary_loss_mlp": 0.01002146, + "balance_loss_clip": 1.01042843, + "balance_loss_mlp": 1.00080454, + "epoch": 0.8520411230685986, + "flos": 63035223252480.0, + "grad_norm": 0.6879533981489377, + "language_loss": 0.5667541, + "learning_rate": 2.2514976586229184e-07, + "loss": 0.58710706, + "num_input_tokens_seen": 153306430, + "step": 7086, + "time_per_iteration": 3.4509246349334717 + }, + { + "auxiliary_loss_clip": 0.01051271, + "auxiliary_loss_mlp": 0.01001033, + "balance_loss_clip": 1.00538445, + "balance_loss_mlp": 0.9995724, + "epoch": 0.8521613659592376, + "flos": 65836865283840.0, + "grad_norm": 0.7738727935287759, + "language_loss": 0.54715514, + "learning_rate": 2.247908320258609e-07, + "loss": 0.56767821, + "num_input_tokens_seen": 153366520, + "step": 7087, + "time_per_iteration": 3.1968111991882324 + }, + { + "auxiliary_loss_clip": 0.01127679, + "auxiliary_loss_mlp": 0.01030299, + "balance_loss_clip": 1.04328489, + "balance_loss_mlp": 1.02212119, + "epoch": 0.8522816088498768, + "flos": 23112323418240.0, + "grad_norm": 2.1331834444990823, + "language_loss": 0.79103076, + "learning_rate": 2.2443216748724914e-07, + "loss": 0.81261051, + "num_input_tokens_seen": 153387230, + "step": 7088, + "time_per_iteration": 2.8828225135803223 + }, + { + "auxiliary_loss_clip": 0.01164962, + "auxiliary_loss_mlp": 0.0088551, + "balance_loss_clip": 1.04861236, + "balance_loss_mlp": 1.00061023, + "epoch": 0.8524018517405159, + "flos": 31758140073600.0, + "grad_norm": 2.0633513053616768, + "language_loss": 0.74196529, + "learning_rate": 2.2407377230086588e-07, + "loss": 0.76246995, + "num_input_tokens_seen": 153409585, + "step": 7089, + "time_per_iteration": 2.9248287677764893 + }, + { + "auxiliary_loss_clip": 0.01137664, + "auxiliary_loss_mlp": 0.01027272, + "balance_loss_clip": 1.04410553, + "balance_loss_mlp": 1.01954746, + "epoch": 0.8525220946311549, + "flos": 18690318956160.0, + "grad_norm": 2.212755494418706, + "language_loss": 0.83195579, + "learning_rate": 2.23715646521079e-07, + "loss": 0.85360515, + "num_input_tokens_seen": 153427105, + "step": 7090, + "time_per_iteration": 2.678600549697876 + }, + { + "auxiliary_loss_clip": 0.01166199, + "auxiliary_loss_mlp": 0.00886333, + "balance_loss_clip": 1.04740763, + "balance_loss_mlp": 1.00051928, + "epoch": 0.852642337521794, + "flos": 21793216354560.0, + "grad_norm": 1.8930940498965438, + "language_loss": 0.83860636, + "learning_rate": 2.2335779020221724e-07, + "loss": 0.85913169, + "num_input_tokens_seen": 153443725, + "step": 7091, + "time_per_iteration": 2.6680235862731934 + }, + { + "auxiliary_loss_clip": 0.01045354, + "auxiliary_loss_mlp": 0.01002347, + "balance_loss_clip": 1.00655317, + "balance_loss_mlp": 1.00083947, + "epoch": 0.8527625804124331, + "flos": 69040132260480.0, + "grad_norm": 0.8048178486619252, + "language_loss": 0.564376, + "learning_rate": 2.2300020339856497e-07, + "loss": 0.58485293, + "num_input_tokens_seen": 153506410, + "step": 7092, + "time_per_iteration": 3.3063392639160156 + }, + { + "auxiliary_loss_clip": 0.01146425, + "auxiliary_loss_mlp": 0.01028637, + "balance_loss_clip": 1.04494452, + "balance_loss_mlp": 1.021106, + "epoch": 0.8528828233030722, + "flos": 26979399688320.0, + "grad_norm": 2.18744533086472, + "language_loss": 0.78211236, + "learning_rate": 2.2264288616436966e-07, + "loss": 0.80386299, + "num_input_tokens_seen": 153526665, + "step": 7093, + "time_per_iteration": 2.8077924251556396 + }, + { + "auxiliary_loss_clip": 0.0114924, + "auxiliary_loss_mlp": 0.01026751, + "balance_loss_clip": 1.04635847, + "balance_loss_mlp": 1.01908565, + "epoch": 0.8530030661937112, + "flos": 17487598936320.0, + "grad_norm": 2.325486071812548, + "language_loss": 0.7242316, + "learning_rate": 2.222858385538351e-07, + "loss": 0.74599147, + "num_input_tokens_seen": 153543465, + "step": 7094, + "time_per_iteration": 2.6948609352111816 + }, + { + "auxiliary_loss_clip": 0.0115626, + "auxiliary_loss_mlp": 0.01025895, + "balance_loss_clip": 1.04436624, + "balance_loss_mlp": 1.0184232, + "epoch": 0.8531233090843504, + "flos": 22160798184960.0, + "grad_norm": 2.0035294879822043, + "language_loss": 0.67827249, + "learning_rate": 2.2192906062112527e-07, + "loss": 0.7000941, + "num_input_tokens_seen": 153563340, + "step": 7095, + "time_per_iteration": 2.82934308052063 + }, + { + "auxiliary_loss_clip": 0.01167498, + "auxiliary_loss_mlp": 0.01023027, + "balance_loss_clip": 1.04534304, + "balance_loss_mlp": 1.01503396, + "epoch": 0.8532435519749895, + "flos": 37635388145280.0, + "grad_norm": 1.599328588377138, + "language_loss": 0.70581585, + "learning_rate": 2.2157255242036377e-07, + "loss": 0.7277211, + "num_input_tokens_seen": 153587005, + "step": 7096, + "time_per_iteration": 2.7422571182250977 + }, + { + "auxiliary_loss_clip": 0.01136888, + "auxiliary_loss_mlp": 0.01030411, + "balance_loss_clip": 1.04327106, + "balance_loss_mlp": 1.02267146, + "epoch": 0.8533637948656285, + "flos": 21398163598080.0, + "grad_norm": 1.6373866028484718, + "language_loss": 0.74767041, + "learning_rate": 2.2121631400563135e-07, + "loss": 0.76934338, + "num_input_tokens_seen": 153606835, + "step": 7097, + "time_per_iteration": 2.86261248588562 + }, + { + "auxiliary_loss_clip": 0.01048661, + "auxiliary_loss_mlp": 0.01000701, + "balance_loss_clip": 1.00912738, + "balance_loss_mlp": 0.99937195, + "epoch": 0.8534840377562677, + "flos": 53345122490880.0, + "grad_norm": 0.7773232892829423, + "language_loss": 0.52937418, + "learning_rate": 2.208603454309701e-07, + "loss": 0.54986781, + "num_input_tokens_seen": 153664925, + "step": 7098, + "time_per_iteration": 3.2504429817199707 + }, + { + "auxiliary_loss_clip": 0.01127644, + "auxiliary_loss_mlp": 0.01026146, + "balance_loss_clip": 1.04105592, + "balance_loss_mlp": 1.01843369, + "epoch": 0.8536042806469067, + "flos": 20814148368000.0, + "grad_norm": 2.242026738522391, + "language_loss": 0.70926034, + "learning_rate": 2.2050464675037994e-07, + "loss": 0.73079824, + "num_input_tokens_seen": 153683550, + "step": 7099, + "time_per_iteration": 3.937634229660034 + }, + { + "auxiliary_loss_clip": 0.01151686, + "auxiliary_loss_mlp": 0.01027059, + "balance_loss_clip": 1.04579437, + "balance_loss_mlp": 1.01867843, + "epoch": 0.8537245235375458, + "flos": 24681368292480.0, + "grad_norm": 2.0722964997216535, + "language_loss": 0.73209637, + "learning_rate": 2.2014921801782016e-07, + "loss": 0.75388384, + "num_input_tokens_seen": 153703040, + "step": 7100, + "time_per_iteration": 2.8641908168792725 + }, + { + "auxiliary_loss_clip": 0.01147614, + "auxiliary_loss_mlp": 0.01030777, + "balance_loss_clip": 1.03830123, + "balance_loss_mlp": 1.02296317, + "epoch": 0.853844766428185, + "flos": 24384817607040.0, + "grad_norm": 1.806424927919527, + "language_loss": 0.73794967, + "learning_rate": 2.1979405928720872e-07, + "loss": 0.75973356, + "num_input_tokens_seen": 153722695, + "step": 7101, + "time_per_iteration": 2.9415500164031982 + }, + { + "auxiliary_loss_clip": 0.01151756, + "auxiliary_loss_mlp": 0.01024103, + "balance_loss_clip": 1.04383445, + "balance_loss_mlp": 1.01588953, + "epoch": 0.853965009318824, + "flos": 20955707867520.0, + "grad_norm": 1.4544248565738924, + "language_loss": 0.79212648, + "learning_rate": 2.1943917061242257e-07, + "loss": 0.81388503, + "num_input_tokens_seen": 153742550, + "step": 7102, + "time_per_iteration": 3.700575590133667 + }, + { + "auxiliary_loss_clip": 0.01164347, + "auxiliary_loss_mlp": 0.00886779, + "balance_loss_clip": 1.0463903, + "balance_loss_mlp": 1.00056469, + "epoch": 0.8540852522094631, + "flos": 24201816791040.0, + "grad_norm": 1.6645603691335276, + "language_loss": 0.66175038, + "learning_rate": 2.1908455204729903e-07, + "loss": 0.68226159, + "num_input_tokens_seen": 153761700, + "step": 7103, + "time_per_iteration": 2.7628068923950195 + }, + { + "auxiliary_loss_clip": 0.01148674, + "auxiliary_loss_mlp": 0.01031125, + "balance_loss_clip": 1.04271197, + "balance_loss_mlp": 1.02297711, + "epoch": 0.8542054951001022, + "flos": 25082921410560.0, + "grad_norm": 2.0371794700936428, + "language_loss": 0.78505933, + "learning_rate": 2.1873020364563265e-07, + "loss": 0.80685735, + "num_input_tokens_seen": 153780765, + "step": 7104, + "time_per_iteration": 3.741948366165161 + }, + { + "auxiliary_loss_clip": 0.01156191, + "auxiliary_loss_mlp": 0.01026861, + "balance_loss_clip": 1.04406333, + "balance_loss_mlp": 1.01926112, + "epoch": 0.8543257379907413, + "flos": 24316551809280.0, + "grad_norm": 2.332178175587671, + "language_loss": 0.76485157, + "learning_rate": 2.183761254611789e-07, + "loss": 0.78668207, + "num_input_tokens_seen": 153801090, + "step": 7105, + "time_per_iteration": 2.665377378463745 + }, + { + "auxiliary_loss_clip": 0.0115909, + "auxiliary_loss_mlp": 0.01026646, + "balance_loss_clip": 1.04824567, + "balance_loss_mlp": 1.01916564, + "epoch": 0.8544459808813804, + "flos": 55286630467200.0, + "grad_norm": 1.9353796175998486, + "language_loss": 0.70332205, + "learning_rate": 2.1802231754764987e-07, + "loss": 0.72517943, + "num_input_tokens_seen": 153826530, + "step": 7106, + "time_per_iteration": 3.021930694580078 + }, + { + "auxiliary_loss_clip": 0.01152435, + "auxiliary_loss_mlp": 0.01026864, + "balance_loss_clip": 1.04307306, + "balance_loss_mlp": 1.01884723, + "epoch": 0.8545662237720195, + "flos": 25776248705280.0, + "grad_norm": 1.8067476179830266, + "language_loss": 0.76559001, + "learning_rate": 2.17668779958718e-07, + "loss": 0.78738302, + "num_input_tokens_seen": 153849110, + "step": 7107, + "time_per_iteration": 2.747908353805542 + }, + { + "auxiliary_loss_clip": 0.01166343, + "auxiliary_loss_mlp": 0.01031381, + "balance_loss_clip": 1.04497337, + "balance_loss_mlp": 1.02337027, + "epoch": 0.8546864666626586, + "flos": 11108320427520.0, + "grad_norm": 2.3798381763565084, + "language_loss": 0.8021993, + "learning_rate": 2.1731551274801553e-07, + "loss": 0.82417655, + "num_input_tokens_seen": 153865550, + "step": 7108, + "time_per_iteration": 2.6312384605407715 + }, + { + "auxiliary_loss_clip": 0.01149277, + "auxiliary_loss_mlp": 0.01028357, + "balance_loss_clip": 1.04440546, + "balance_loss_mlp": 1.02033448, + "epoch": 0.8548067095532976, + "flos": 25520169669120.0, + "grad_norm": 2.498837959858074, + "language_loss": 0.62162876, + "learning_rate": 2.169625159691324e-07, + "loss": 0.64340514, + "num_input_tokens_seen": 153885425, + "step": 7109, + "time_per_iteration": 3.798003673553467 + }, + { + "auxiliary_loss_clip": 0.0113882, + "auxiliary_loss_mlp": 0.01033277, + "balance_loss_clip": 1.0403223, + "balance_loss_mlp": 1.0246644, + "epoch": 0.8549269524439368, + "flos": 24717853532160.0, + "grad_norm": 8.081983504605843, + "language_loss": 0.74471551, + "learning_rate": 2.1660978967561784e-07, + "loss": 0.76643646, + "num_input_tokens_seen": 153904760, + "step": 7110, + "time_per_iteration": 2.91518497467041 + }, + { + "auxiliary_loss_clip": 0.01167381, + "auxiliary_loss_mlp": 0.01023519, + "balance_loss_clip": 1.04571176, + "balance_loss_mlp": 1.01565683, + "epoch": 0.8550471953345758, + "flos": 19825599191040.0, + "grad_norm": 2.292191003490961, + "language_loss": 0.79318011, + "learning_rate": 2.1625733392098035e-07, + "loss": 0.81508911, + "num_input_tokens_seen": 153920370, + "step": 7111, + "time_per_iteration": 2.6810755729675293 + }, + { + "auxiliary_loss_clip": 0.01166008, + "auxiliary_loss_mlp": 0.01026002, + "balance_loss_clip": 1.04510546, + "balance_loss_mlp": 1.01883459, + "epoch": 0.8551674382252149, + "flos": 22820441500800.0, + "grad_norm": 1.7741179107923584, + "language_loss": 0.79553658, + "learning_rate": 2.159051487586867e-07, + "loss": 0.81745672, + "num_input_tokens_seen": 153940500, + "step": 7112, + "time_per_iteration": 2.7095015048980713 + }, + { + "auxiliary_loss_clip": 0.01155941, + "auxiliary_loss_mlp": 0.01027954, + "balance_loss_clip": 1.04879272, + "balance_loss_mlp": 1.01919818, + "epoch": 0.8552876811158541, + "flos": 20631255292800.0, + "grad_norm": 2.148784666501629, + "language_loss": 0.72330648, + "learning_rate": 2.155532342421642e-07, + "loss": 0.74514544, + "num_input_tokens_seen": 153958500, + "step": 7113, + "time_per_iteration": 2.8204753398895264 + }, + { + "auxiliary_loss_clip": 0.01162052, + "auxiliary_loss_mlp": 0.01030577, + "balance_loss_clip": 1.04711843, + "balance_loss_mlp": 1.02270317, + "epoch": 0.8554079240064931, + "flos": 23112359331840.0, + "grad_norm": 1.8379475671047862, + "language_loss": 0.78422147, + "learning_rate": 2.1520159042479636e-07, + "loss": 0.80614775, + "num_input_tokens_seen": 153976790, + "step": 7114, + "time_per_iteration": 2.750577449798584 + }, + { + "auxiliary_loss_clip": 0.01159478, + "auxiliary_loss_mlp": 0.01030948, + "balance_loss_clip": 1.0473448, + "balance_loss_mlp": 1.02339613, + "epoch": 0.8555281668971322, + "flos": 22128047959680.0, + "grad_norm": 4.810632578389086, + "language_loss": 0.7113418, + "learning_rate": 2.148502173599287e-07, + "loss": 0.73324609, + "num_input_tokens_seen": 153994930, + "step": 7115, + "time_per_iteration": 2.8674564361572266 + }, + { + "auxiliary_loss_clip": 0.01147591, + "auxiliary_loss_mlp": 0.01025121, + "balance_loss_clip": 1.04412556, + "balance_loss_mlp": 1.0172925, + "epoch": 0.8556484097877713, + "flos": 31139040234240.0, + "grad_norm": 1.7203428395684928, + "language_loss": 0.6559279, + "learning_rate": 2.1449911510086372e-07, + "loss": 0.67765498, + "num_input_tokens_seen": 154014400, + "step": 7116, + "time_per_iteration": 2.8413755893707275 + }, + { + "auxiliary_loss_clip": 0.01158566, + "auxiliary_loss_mlp": 0.0103049, + "balance_loss_clip": 1.04707778, + "balance_loss_mlp": 1.02278948, + "epoch": 0.8557686526784104, + "flos": 24316551809280.0, + "grad_norm": 1.955631579298787, + "language_loss": 0.76839107, + "learning_rate": 2.141482837008628e-07, + "loss": 0.79028165, + "num_input_tokens_seen": 154034940, + "step": 7117, + "time_per_iteration": 2.7926619052886963 + }, + { + "auxiliary_loss_clip": 0.01151388, + "auxiliary_loss_mlp": 0.01028467, + "balance_loss_clip": 1.04443717, + "balance_loss_mlp": 1.02030671, + "epoch": 0.8558888955690495, + "flos": 17712723427200.0, + "grad_norm": 2.107234228963361, + "language_loss": 0.72239792, + "learning_rate": 2.1379772321314826e-07, + "loss": 0.74419647, + "num_input_tokens_seen": 154052985, + "step": 7118, + "time_per_iteration": 2.704115867614746 + }, + { + "auxiliary_loss_clip": 0.01111525, + "auxiliary_loss_mlp": 0.01025766, + "balance_loss_clip": 1.03979552, + "balance_loss_mlp": 1.01805902, + "epoch": 0.8560091384596886, + "flos": 19171702051200.0, + "grad_norm": 2.2750452278899367, + "language_loss": 0.81649292, + "learning_rate": 2.1344743369089802e-07, + "loss": 0.83786577, + "num_input_tokens_seen": 154068765, + "step": 7119, + "time_per_iteration": 2.8461883068084717 + }, + { + "auxiliary_loss_clip": 0.0114932, + "auxiliary_loss_mlp": 0.01027141, + "balance_loss_clip": 1.04530466, + "balance_loss_mlp": 1.01920164, + "epoch": 0.8561293813503277, + "flos": 23914855036800.0, + "grad_norm": 1.8175313263125754, + "language_loss": 0.82002854, + "learning_rate": 2.130974151872522e-07, + "loss": 0.84179312, + "num_input_tokens_seen": 154089100, + "step": 7120, + "time_per_iteration": 2.747328996658325 + }, + { + "auxiliary_loss_clip": 0.01132209, + "auxiliary_loss_mlp": 0.0103191, + "balance_loss_clip": 1.04403329, + "balance_loss_mlp": 1.02360702, + "epoch": 0.8562496242409667, + "flos": 22529206028160.0, + "grad_norm": 2.0074824861103377, + "language_loss": 0.78555131, + "learning_rate": 2.1274766775530773e-07, + "loss": 0.80719244, + "num_input_tokens_seen": 154108965, + "step": 7121, + "time_per_iteration": 2.8964247703552246 + }, + { + "auxiliary_loss_clip": 0.01172133, + "auxiliary_loss_mlp": 0.01025236, + "balance_loss_clip": 1.0469923, + "balance_loss_mlp": 1.01730275, + "epoch": 0.8563698671316058, + "flos": 14712745472640.0, + "grad_norm": 2.188669592599364, + "language_loss": 0.79531604, + "learning_rate": 2.1239819144812077e-07, + "loss": 0.81728971, + "num_input_tokens_seen": 154123425, + "step": 7122, + "time_per_iteration": 2.6399149894714355 + }, + { + "auxiliary_loss_clip": 0.01131497, + "auxiliary_loss_mlp": 0.01029875, + "balance_loss_clip": 1.03821385, + "balance_loss_mlp": 1.0222218, + "epoch": 0.856490110022245, + "flos": 39167768211840.0, + "grad_norm": 2.5797423209674464, + "language_loss": 0.69941103, + "learning_rate": 2.1204898631870716e-07, + "loss": 0.72102481, + "num_input_tokens_seen": 154148315, + "step": 7123, + "time_per_iteration": 2.930727481842041 + }, + { + "auxiliary_loss_clip": 0.01152269, + "auxiliary_loss_mlp": 0.01023643, + "balance_loss_clip": 1.04719996, + "balance_loss_mlp": 1.01587665, + "epoch": 0.856610352912884, + "flos": 29059345658880.0, + "grad_norm": 1.8741364401261857, + "language_loss": 0.76205015, + "learning_rate": 2.1170005242004006e-07, + "loss": 0.7838093, + "num_input_tokens_seen": 154169665, + "step": 7124, + "time_per_iteration": 2.86287784576416 + }, + { + "auxiliary_loss_clip": 0.01143799, + "auxiliary_loss_mlp": 0.01025215, + "balance_loss_clip": 1.04228401, + "balance_loss_mlp": 1.0184108, + "epoch": 0.8567305958035231, + "flos": 23878333883520.0, + "grad_norm": 1.8193242992742837, + "language_loss": 0.77956498, + "learning_rate": 2.1135138980505384e-07, + "loss": 0.80125511, + "num_input_tokens_seen": 154190335, + "step": 7125, + "time_per_iteration": 3.745800495147705 + }, + { + "auxiliary_loss_clip": 0.01146086, + "auxiliary_loss_mlp": 0.01028799, + "balance_loss_clip": 1.04639244, + "balance_loss_mlp": 1.02086532, + "epoch": 0.8568508386941622, + "flos": 22200120599040.0, + "grad_norm": 1.6026160511095735, + "language_loss": 0.7240352, + "learning_rate": 2.110029985266395e-07, + "loss": 0.74578404, + "num_input_tokens_seen": 154210040, + "step": 7126, + "time_per_iteration": 2.7402381896972656 + }, + { + "auxiliary_loss_clip": 0.01146079, + "auxiliary_loss_mlp": 0.01027269, + "balance_loss_clip": 1.04400933, + "balance_loss_mlp": 1.019449, + "epoch": 0.8569710815848013, + "flos": 17307507121920.0, + "grad_norm": 1.6729185177079875, + "language_loss": 0.73679221, + "learning_rate": 2.1065487863764787e-07, + "loss": 0.75852573, + "num_input_tokens_seen": 154228385, + "step": 7127, + "time_per_iteration": 2.7506418228149414 + }, + { + "auxiliary_loss_clip": 0.01121804, + "auxiliary_loss_mlp": 0.0102581, + "balance_loss_clip": 1.03588891, + "balance_loss_mlp": 1.01779282, + "epoch": 0.8570913244754403, + "flos": 23732285184000.0, + "grad_norm": 4.484338702488571, + "language_loss": 0.85708964, + "learning_rate": 2.1030703019088846e-07, + "loss": 0.87856579, + "num_input_tokens_seen": 154249015, + "step": 7128, + "time_per_iteration": 3.7417449951171875 + }, + { + "auxiliary_loss_clip": 0.01152796, + "auxiliary_loss_mlp": 0.0102807, + "balance_loss_clip": 1.04477334, + "balance_loss_mlp": 1.02061367, + "epoch": 0.8572115673660795, + "flos": 20048748433920.0, + "grad_norm": 2.433815836438044, + "language_loss": 0.7067495, + "learning_rate": 2.099594532391291e-07, + "loss": 0.72855818, + "num_input_tokens_seen": 154267700, + "step": 7129, + "time_per_iteration": 2.7767932415008545 + }, + { + "auxiliary_loss_clip": 0.01148601, + "auxiliary_loss_mlp": 0.0103095, + "balance_loss_clip": 1.04169643, + "balance_loss_mlp": 1.02296615, + "epoch": 0.8573318102567186, + "flos": 27160389342720.0, + "grad_norm": 2.0274158486288547, + "language_loss": 0.79044998, + "learning_rate": 2.0961214783509806e-07, + "loss": 0.81224549, + "num_input_tokens_seen": 154290580, + "step": 7130, + "time_per_iteration": 3.7009825706481934 + }, + { + "auxiliary_loss_clip": 0.01152178, + "auxiliary_loss_mlp": 0.01026927, + "balance_loss_clip": 1.04280639, + "balance_loss_mlp": 1.019274, + "epoch": 0.8574520531473576, + "flos": 24936585402240.0, + "grad_norm": 1.653694626078175, + "language_loss": 0.7464205, + "learning_rate": 2.0926511403148051e-07, + "loss": 0.76821148, + "num_input_tokens_seen": 154309545, + "step": 7131, + "time_per_iteration": 2.8320417404174805 + }, + { + "auxiliary_loss_clip": 0.01137218, + "auxiliary_loss_mlp": 0.01025736, + "balance_loss_clip": 1.04311192, + "balance_loss_mlp": 1.01829171, + "epoch": 0.8575722960379968, + "flos": 18771154513920.0, + "grad_norm": 1.8096545323488014, + "language_loss": 0.75974983, + "learning_rate": 2.0891835188092143e-07, + "loss": 0.7813794, + "num_input_tokens_seen": 154326545, + "step": 7132, + "time_per_iteration": 2.7996973991394043 + }, + { + "auxiliary_loss_clip": 0.01137215, + "auxiliary_loss_mlp": 0.01023529, + "balance_loss_clip": 1.04289901, + "balance_loss_mlp": 1.0158875, + "epoch": 0.8576925389286358, + "flos": 22200300167040.0, + "grad_norm": 2.148382639771783, + "language_loss": 0.81210172, + "learning_rate": 2.0857186143602434e-07, + "loss": 0.83370912, + "num_input_tokens_seen": 154345190, + "step": 7133, + "time_per_iteration": 2.7534546852111816 + }, + { + "auxiliary_loss_clip": 0.01129802, + "auxiliary_loss_mlp": 0.01023795, + "balance_loss_clip": 1.04005682, + "balance_loss_mlp": 1.01653504, + "epoch": 0.8578127818192749, + "flos": 22894345733760.0, + "grad_norm": 1.7736882989707796, + "language_loss": 0.67704678, + "learning_rate": 2.0822564274935094e-07, + "loss": 0.69858277, + "num_input_tokens_seen": 154364615, + "step": 7134, + "time_per_iteration": 3.649690866470337 + }, + { + "auxiliary_loss_clip": 0.01149444, + "auxiliary_loss_mlp": 0.01029194, + "balance_loss_clip": 1.04648077, + "balance_loss_mlp": 1.02043235, + "epoch": 0.8579330247099141, + "flos": 34824839541120.0, + "grad_norm": 1.7047414903827756, + "language_loss": 0.66485327, + "learning_rate": 2.078796958734239e-07, + "loss": 0.68663961, + "num_input_tokens_seen": 154387335, + "step": 7135, + "time_per_iteration": 2.860173463821411 + }, + { + "auxiliary_loss_clip": 0.01159632, + "auxiliary_loss_mlp": 0.01030937, + "balance_loss_clip": 1.04692042, + "balance_loss_mlp": 1.02311683, + "epoch": 0.8580532676005531, + "flos": 19755681367680.0, + "grad_norm": 1.9174131912376504, + "language_loss": 0.75021565, + "learning_rate": 2.0753402086072124e-07, + "loss": 0.77212131, + "num_input_tokens_seen": 154405965, + "step": 7136, + "time_per_iteration": 2.6694159507751465 + }, + { + "auxiliary_loss_clip": 0.01116952, + "auxiliary_loss_mlp": 0.01027253, + "balance_loss_clip": 1.04068816, + "balance_loss_mlp": 1.01983178, + "epoch": 0.8581735104911922, + "flos": 22739318634240.0, + "grad_norm": 20.697328251875938, + "language_loss": 0.75257152, + "learning_rate": 2.071886177636828e-07, + "loss": 0.77401352, + "num_input_tokens_seen": 154422750, + "step": 7137, + "time_per_iteration": 2.9765803813934326 + }, + { + "auxiliary_loss_clip": 0.01156365, + "auxiliary_loss_mlp": 0.01026897, + "balance_loss_clip": 1.0456059, + "balance_loss_mlp": 1.01919591, + "epoch": 0.8582937533818313, + "flos": 23149131880320.0, + "grad_norm": 1.8490611886429562, + "language_loss": 0.83211023, + "learning_rate": 2.0684348663470575e-07, + "loss": 0.85394287, + "num_input_tokens_seen": 154442930, + "step": 7138, + "time_per_iteration": 2.7790846824645996 + }, + { + "auxiliary_loss_clip": 0.01146496, + "auxiliary_loss_mlp": 0.01030909, + "balance_loss_clip": 1.03913605, + "balance_loss_mlp": 1.02274299, + "epoch": 0.8584139962724704, + "flos": 19498668577920.0, + "grad_norm": 1.7220875266053786, + "language_loss": 0.61436313, + "learning_rate": 2.0649862752614555e-07, + "loss": 0.63613719, + "num_input_tokens_seen": 154461640, + "step": 7139, + "time_per_iteration": 2.7598297595977783 + }, + { + "auxiliary_loss_clip": 0.01046984, + "auxiliary_loss_mlp": 0.01002163, + "balance_loss_clip": 1.00578451, + "balance_loss_mlp": 1.00072694, + "epoch": 0.8585342391631094, + "flos": 71276577788160.0, + "grad_norm": 0.7522336308977733, + "language_loss": 0.5707407, + "learning_rate": 2.0615404049031838e-07, + "loss": 0.59123212, + "num_input_tokens_seen": 154518610, + "step": 7140, + "time_per_iteration": 3.3394432067871094 + }, + { + "auxiliary_loss_clip": 0.01163948, + "auxiliary_loss_mlp": 0.01026648, + "balance_loss_clip": 1.04890299, + "balance_loss_mlp": 1.0183804, + "epoch": 0.8586544820537486, + "flos": 10815432929280.0, + "grad_norm": 2.2538667771804857, + "language_loss": 0.77873623, + "learning_rate": 2.0580972557949616e-07, + "loss": 0.80064213, + "num_input_tokens_seen": 154533700, + "step": 7141, + "time_per_iteration": 2.669612169265747 + }, + { + "auxiliary_loss_clip": 0.0105131, + "auxiliary_loss_mlp": 0.01001971, + "balance_loss_clip": 1.00453782, + "balance_loss_mlp": 1.00049913, + "epoch": 0.8587747249443877, + "flos": 64811184422400.0, + "grad_norm": 0.7959734980538199, + "language_loss": 0.54276478, + "learning_rate": 2.054656828459125e-07, + "loss": 0.56329763, + "num_input_tokens_seen": 154597810, + "step": 7142, + "time_per_iteration": 3.3991525173187256 + }, + { + "auxiliary_loss_clip": 0.01127294, + "auxiliary_loss_mlp": 0.01027841, + "balance_loss_clip": 1.04212999, + "balance_loss_mlp": 1.02007425, + "epoch": 0.8588949678350267, + "flos": 26834607964800.0, + "grad_norm": 3.016094384276907, + "language_loss": 0.77129292, + "learning_rate": 2.051219123417578e-07, + "loss": 0.7928443, + "num_input_tokens_seen": 154617870, + "step": 7143, + "time_per_iteration": 2.9335620403289795 + }, + { + "auxiliary_loss_clip": 0.0117088, + "auxiliary_loss_mlp": 0.01028318, + "balance_loss_clip": 1.04675198, + "balance_loss_mlp": 1.02044392, + "epoch": 0.8590152107256659, + "flos": 26104256726400.0, + "grad_norm": 1.8588314368576009, + "language_loss": 0.60817903, + "learning_rate": 2.0477841411918196e-07, + "loss": 0.63017094, + "num_input_tokens_seen": 154637395, + "step": 7144, + "time_per_iteration": 2.7712745666503906 + }, + { + "auxiliary_loss_clip": 0.01154011, + "auxiliary_loss_mlp": 0.0102702, + "balance_loss_clip": 1.04356813, + "balance_loss_mlp": 1.01950932, + "epoch": 0.859135453616305, + "flos": 26140885620480.0, + "grad_norm": 2.005410068540173, + "language_loss": 0.73929262, + "learning_rate": 2.0443518823029326e-07, + "loss": 0.76110291, + "num_input_tokens_seen": 154657935, + "step": 7145, + "time_per_iteration": 2.838608503341675 + }, + { + "auxiliary_loss_clip": 0.01136634, + "auxiliary_loss_mlp": 0.01028275, + "balance_loss_clip": 1.04305637, + "balance_loss_mlp": 1.02041602, + "epoch": 0.859255696506944, + "flos": 12969319046400.0, + "grad_norm": 2.3936296445118015, + "language_loss": 0.7644791, + "learning_rate": 2.0409223472715854e-07, + "loss": 0.78612816, + "num_input_tokens_seen": 154675080, + "step": 7146, + "time_per_iteration": 2.981334686279297 + }, + { + "auxiliary_loss_clip": 0.01141283, + "auxiliary_loss_mlp": 0.00885527, + "balance_loss_clip": 1.04425752, + "balance_loss_mlp": 1.0005827, + "epoch": 0.8593759393975832, + "flos": 18475753063680.0, + "grad_norm": 1.7231335553238363, + "language_loss": 0.75126606, + "learning_rate": 2.0374955366180434e-07, + "loss": 0.77153414, + "num_input_tokens_seen": 154692720, + "step": 7147, + "time_per_iteration": 2.807452917098999 + }, + { + "auxiliary_loss_clip": 0.0114214, + "auxiliary_loss_mlp": 0.01027067, + "balance_loss_clip": 1.0411315, + "balance_loss_mlp": 1.01884735, + "epoch": 0.8594961822882222, + "flos": 22200156512640.0, + "grad_norm": 1.6948188078664206, + "language_loss": 0.72577727, + "learning_rate": 2.034071450862147e-07, + "loss": 0.74746931, + "num_input_tokens_seen": 154710190, + "step": 7148, + "time_per_iteration": 2.8919949531555176 + }, + { + "auxiliary_loss_clip": 0.0115238, + "auxiliary_loss_mlp": 0.0103656, + "balance_loss_clip": 1.04270196, + "balance_loss_mlp": 1.02823329, + "epoch": 0.8596164251788613, + "flos": 23294749616640.0, + "grad_norm": 1.7646452116848184, + "language_loss": 0.76854074, + "learning_rate": 2.030650090523327e-07, + "loss": 0.79043007, + "num_input_tokens_seen": 154729380, + "step": 7149, + "time_per_iteration": 2.7606701850891113 + }, + { + "auxiliary_loss_clip": 0.01136313, + "auxiliary_loss_mlp": 0.01025071, + "balance_loss_clip": 1.04078722, + "balance_loss_mlp": 1.01720905, + "epoch": 0.8597366680695004, + "flos": 31649905416960.0, + "grad_norm": 1.8474916532423018, + "language_loss": 0.59542423, + "learning_rate": 2.0272314561205995e-07, + "loss": 0.61703807, + "num_input_tokens_seen": 154749775, + "step": 7150, + "time_per_iteration": 2.83316707611084 + }, + { + "auxiliary_loss_clip": 0.01132135, + "auxiliary_loss_mlp": 0.01023894, + "balance_loss_clip": 1.03853774, + "balance_loss_mlp": 1.01641965, + "epoch": 0.8598569109601395, + "flos": 21287738211840.0, + "grad_norm": 1.9968519779676046, + "language_loss": 0.72667396, + "learning_rate": 2.023815548172567e-07, + "loss": 0.74823427, + "num_input_tokens_seen": 154769845, + "step": 7151, + "time_per_iteration": 3.8095436096191406 + }, + { + "auxiliary_loss_clip": 0.01158671, + "auxiliary_loss_mlp": 0.0102785, + "balance_loss_clip": 1.04474115, + "balance_loss_mlp": 1.01981521, + "epoch": 0.8599771538507786, + "flos": 25447809720960.0, + "grad_norm": 1.5317697114836009, + "language_loss": 0.66089869, + "learning_rate": 2.0204023671974267e-07, + "loss": 0.68276387, + "num_input_tokens_seen": 154789230, + "step": 7152, + "time_per_iteration": 2.8376569747924805 + }, + { + "auxiliary_loss_clip": 0.01155894, + "auxiliary_loss_mlp": 0.01025103, + "balance_loss_clip": 1.04420519, + "balance_loss_mlp": 1.01706243, + "epoch": 0.8600973967414177, + "flos": 16723958768640.0, + "grad_norm": 4.794954841225615, + "language_loss": 0.81060696, + "learning_rate": 2.0169919137129532e-07, + "loss": 0.83241689, + "num_input_tokens_seen": 154807670, + "step": 7153, + "time_per_iteration": 2.702805995941162 + }, + { + "auxiliary_loss_clip": 0.0116249, + "auxiliary_loss_mlp": 0.0103262, + "balance_loss_clip": 1.04890227, + "balance_loss_mlp": 1.02476954, + "epoch": 0.8602176396320568, + "flos": 25227928615680.0, + "grad_norm": 2.3159418862298593, + "language_loss": 0.70912957, + "learning_rate": 2.013584188236508e-07, + "loss": 0.73108065, + "num_input_tokens_seen": 154825575, + "step": 7154, + "time_per_iteration": 3.624342918395996 + }, + { + "auxiliary_loss_clip": 0.01170462, + "auxiliary_loss_mlp": 0.0102932, + "balance_loss_clip": 1.04714406, + "balance_loss_mlp": 1.02133298, + "epoch": 0.8603378825226958, + "flos": 20412236113920.0, + "grad_norm": 1.653391200550046, + "language_loss": 0.79238659, + "learning_rate": 2.0101791912850396e-07, + "loss": 0.81438446, + "num_input_tokens_seen": 154845115, + "step": 7155, + "time_per_iteration": 2.659022092819214 + }, + { + "auxiliary_loss_clip": 0.01152458, + "auxiliary_loss_mlp": 0.01032551, + "balance_loss_clip": 1.04776406, + "balance_loss_mlp": 1.02480292, + "epoch": 0.8604581254133349, + "flos": 34930201109760.0, + "grad_norm": 1.8051901889861968, + "language_loss": 0.6406216, + "learning_rate": 2.006776923375082e-07, + "loss": 0.66247165, + "num_input_tokens_seen": 154866770, + "step": 7156, + "time_per_iteration": 3.686223030090332 + }, + { + "auxiliary_loss_clip": 0.01166668, + "auxiliary_loss_mlp": 0.01025837, + "balance_loss_clip": 1.04616547, + "balance_loss_mlp": 1.01754594, + "epoch": 0.860578368303974, + "flos": 22596538072320.0, + "grad_norm": 1.5740896177345935, + "language_loss": 0.71152323, + "learning_rate": 2.003377385022764e-07, + "loss": 0.73344827, + "num_input_tokens_seen": 154885595, + "step": 7157, + "time_per_iteration": 2.6892287731170654 + }, + { + "auxiliary_loss_clip": 0.01151309, + "auxiliary_loss_mlp": 0.01023842, + "balance_loss_clip": 1.04448295, + "balance_loss_mlp": 1.01579559, + "epoch": 0.8606986111946131, + "flos": 21324331192320.0, + "grad_norm": 2.286957506210953, + "language_loss": 0.77299714, + "learning_rate": 1.9999805767437826e-07, + "loss": 0.79474866, + "num_input_tokens_seen": 154904485, + "step": 7158, + "time_per_iteration": 2.8325741291046143 + }, + { + "auxiliary_loss_clip": 0.0114021, + "auxiliary_loss_mlp": 0.01027695, + "balance_loss_clip": 1.04122925, + "balance_loss_mlp": 1.01940751, + "epoch": 0.8608188540852522, + "flos": 28877206769280.0, + "grad_norm": 1.8212810517749307, + "language_loss": 0.71652597, + "learning_rate": 1.9965864990534386e-07, + "loss": 0.73820502, + "num_input_tokens_seen": 154925010, + "step": 7159, + "time_per_iteration": 2.8250858783721924 + }, + { + "auxiliary_loss_clip": 0.01130637, + "auxiliary_loss_mlp": 0.01024287, + "balance_loss_clip": 1.03850567, + "balance_loss_mlp": 1.01693797, + "epoch": 0.8609390969758913, + "flos": 29716187713920.0, + "grad_norm": 2.9466406402595293, + "language_loss": 0.77821565, + "learning_rate": 1.9931951524666092e-07, + "loss": 0.79976487, + "num_input_tokens_seen": 154946100, + "step": 7160, + "time_per_iteration": 2.9873859882354736 + }, + { + "auxiliary_loss_clip": 0.0116368, + "auxiliary_loss_mlp": 0.00885181, + "balance_loss_clip": 1.04688287, + "balance_loss_mlp": 1.00050855, + "epoch": 0.8610593398665304, + "flos": 21249349551360.0, + "grad_norm": 2.126110057924139, + "language_loss": 0.80971229, + "learning_rate": 1.9898065374977534e-07, + "loss": 0.83020091, + "num_input_tokens_seen": 154966305, + "step": 7161, + "time_per_iteration": 3.7512426376342773 + }, + { + "auxiliary_loss_clip": 0.01131259, + "auxiliary_loss_mlp": 0.01021448, + "balance_loss_clip": 1.04252326, + "balance_loss_mlp": 1.01449835, + "epoch": 0.8611795827571694, + "flos": 14830102183680.0, + "grad_norm": 1.865161893859953, + "language_loss": 0.72840846, + "learning_rate": 1.9864206546609342e-07, + "loss": 0.74993557, + "num_input_tokens_seen": 154985145, + "step": 7162, + "time_per_iteration": 2.8535258769989014 + }, + { + "auxiliary_loss_clip": 0.01166914, + "auxiliary_loss_mlp": 0.01028125, + "balance_loss_clip": 1.0447073, + "balance_loss_mlp": 1.02044201, + "epoch": 0.8612998256478086, + "flos": 24243258107520.0, + "grad_norm": 1.8236662244522162, + "language_loss": 0.84206355, + "learning_rate": 1.983037504469771e-07, + "loss": 0.86401391, + "num_input_tokens_seen": 155003855, + "step": 7163, + "time_per_iteration": 2.811229705810547 + }, + { + "auxiliary_loss_clip": 0.01160339, + "auxiliary_loss_mlp": 0.01033765, + "balance_loss_clip": 1.04671526, + "balance_loss_mlp": 1.02646375, + "epoch": 0.8614200685384477, + "flos": 21252653602560.0, + "grad_norm": 1.7737608701241607, + "language_loss": 0.66657698, + "learning_rate": 1.9796570874374984e-07, + "loss": 0.68851805, + "num_input_tokens_seen": 155023960, + "step": 7164, + "time_per_iteration": 2.814000129699707 + }, + { + "auxiliary_loss_clip": 0.01150772, + "auxiliary_loss_mlp": 0.01032153, + "balance_loss_clip": 1.04273748, + "balance_loss_mlp": 1.02420819, + "epoch": 0.8615403114290867, + "flos": 20007738080640.0, + "grad_norm": 1.766610836582026, + "language_loss": 0.77627206, + "learning_rate": 1.976279404076917e-07, + "loss": 0.79810137, + "num_input_tokens_seen": 155043360, + "step": 7165, + "time_per_iteration": 2.894883155822754 + }, + { + "auxiliary_loss_clip": 0.01143101, + "auxiliary_loss_mlp": 0.0102862, + "balance_loss_clip": 1.04724956, + "balance_loss_mlp": 1.02071023, + "epoch": 0.8616605543197259, + "flos": 29789373674880.0, + "grad_norm": 7.834091266128361, + "language_loss": 0.76132751, + "learning_rate": 1.9729044549004193e-07, + "loss": 0.7830447, + "num_input_tokens_seen": 155064745, + "step": 7166, + "time_per_iteration": 2.89385986328125 + }, + { + "auxiliary_loss_clip": 0.01158614, + "auxiliary_loss_mlp": 0.01024124, + "balance_loss_clip": 1.04680789, + "balance_loss_mlp": 1.01655126, + "epoch": 0.8617807972103649, + "flos": 28911609020160.0, + "grad_norm": 1.4616516993384692, + "language_loss": 0.70188165, + "learning_rate": 1.9695322404199822e-07, + "loss": 0.72370905, + "num_input_tokens_seen": 155086790, + "step": 7167, + "time_per_iteration": 2.8119266033172607 + }, + { + "auxiliary_loss_clip": 0.01152507, + "auxiliary_loss_mlp": 0.01032319, + "balance_loss_clip": 1.04739857, + "balance_loss_mlp": 1.02365267, + "epoch": 0.861901040101004, + "flos": 27673804391040.0, + "grad_norm": 1.855669689455432, + "language_loss": 0.82263172, + "learning_rate": 1.9661627611471654e-07, + "loss": 0.84447992, + "num_input_tokens_seen": 155106585, + "step": 7168, + "time_per_iteration": 2.8169338703155518 + }, + { + "auxiliary_loss_clip": 0.01147872, + "auxiliary_loss_mlp": 0.01025878, + "balance_loss_clip": 1.04388714, + "balance_loss_mlp": 1.01793218, + "epoch": 0.8620212829916432, + "flos": 49748056755840.0, + "grad_norm": 1.838587212146956, + "language_loss": 0.70417082, + "learning_rate": 1.9627960175931246e-07, + "loss": 0.72590828, + "num_input_tokens_seen": 155131285, + "step": 7169, + "time_per_iteration": 2.9482901096343994 + }, + { + "auxiliary_loss_clip": 0.011585, + "auxiliary_loss_mlp": 0.01034294, + "balance_loss_clip": 1.045856, + "balance_loss_mlp": 1.02658677, + "epoch": 0.8621415258822822, + "flos": 21138672769920.0, + "grad_norm": 1.8522066442778486, + "language_loss": 0.7419157, + "learning_rate": 1.9594320102685847e-07, + "loss": 0.76384366, + "num_input_tokens_seen": 155150555, + "step": 7170, + "time_per_iteration": 2.8071961402893066 + }, + { + "auxiliary_loss_clip": 0.01138687, + "auxiliary_loss_mlp": 0.00885536, + "balance_loss_clip": 1.04186392, + "balance_loss_mlp": 1.0005579, + "epoch": 0.8622617687729213, + "flos": 21689039934720.0, + "grad_norm": 2.0736392217762036, + "language_loss": 0.639727, + "learning_rate": 1.956070739683864e-07, + "loss": 0.65996921, + "num_input_tokens_seen": 155169890, + "step": 7171, + "time_per_iteration": 2.8805699348449707 + }, + { + "auxiliary_loss_clip": 0.01123648, + "auxiliary_loss_mlp": 0.01028141, + "balance_loss_clip": 1.03863716, + "balance_loss_mlp": 1.02064848, + "epoch": 0.8623820116635604, + "flos": 26250592734720.0, + "grad_norm": 1.5622082820773666, + "language_loss": 0.74313569, + "learning_rate": 1.9527122063488678e-07, + "loss": 0.76465356, + "num_input_tokens_seen": 155191005, + "step": 7172, + "time_per_iteration": 2.965463876724243 + }, + { + "auxiliary_loss_clip": 0.01146101, + "auxiliary_loss_mlp": 0.01022246, + "balance_loss_clip": 1.04052985, + "balance_loss_mlp": 1.01443231, + "epoch": 0.8625022545541995, + "flos": 19647554451840.0, + "grad_norm": 2.694817479939643, + "language_loss": 0.80258203, + "learning_rate": 1.9493564107730755e-07, + "loss": 0.82426554, + "num_input_tokens_seen": 155211005, + "step": 7173, + "time_per_iteration": 2.853290319442749 + }, + { + "auxiliary_loss_clip": 0.01142599, + "auxiliary_loss_mlp": 0.01027364, + "balance_loss_clip": 1.04111624, + "balance_loss_mlp": 1.01942492, + "epoch": 0.8626224974448385, + "flos": 21908382336000.0, + "grad_norm": 1.751147316109011, + "language_loss": 0.61003768, + "learning_rate": 1.9460033534655684e-07, + "loss": 0.63173735, + "num_input_tokens_seen": 155230365, + "step": 7174, + "time_per_iteration": 2.9675750732421875 + }, + { + "auxiliary_loss_clip": 0.01142307, + "auxiliary_loss_mlp": 0.01025651, + "balance_loss_clip": 1.03821564, + "balance_loss_mlp": 1.01831388, + "epoch": 0.8627427403354777, + "flos": 23331198942720.0, + "grad_norm": 1.6854682742008624, + "language_loss": 0.84216911, + "learning_rate": 1.9426530349349978e-07, + "loss": 0.86384869, + "num_input_tokens_seen": 155250815, + "step": 7175, + "time_per_iteration": 2.9711925983428955 + }, + { + "auxiliary_loss_clip": 0.01156172, + "auxiliary_loss_mlp": 0.00885457, + "balance_loss_clip": 1.04312801, + "balance_loss_mlp": 1.00056124, + "epoch": 0.8628629832261168, + "flos": 16362877299840.0, + "grad_norm": 1.7641865083237736, + "language_loss": 0.64742512, + "learning_rate": 1.9393054556896038e-07, + "loss": 0.66784143, + "num_input_tokens_seen": 155268515, + "step": 7176, + "time_per_iteration": 2.756861686706543 + }, + { + "auxiliary_loss_clip": 0.01131112, + "auxiliary_loss_mlp": 0.01029139, + "balance_loss_clip": 1.04270077, + "balance_loss_mlp": 1.02133727, + "epoch": 0.8629832261167558, + "flos": 28103941756800.0, + "grad_norm": 2.3257659359671914, + "language_loss": 0.69118869, + "learning_rate": 1.9359606162372133e-07, + "loss": 0.7127912, + "num_input_tokens_seen": 155290120, + "step": 7177, + "time_per_iteration": 3.8597447872161865 + }, + { + "auxiliary_loss_clip": 0.01167817, + "auxiliary_loss_mlp": 0.01028899, + "balance_loss_clip": 1.0478822, + "balance_loss_mlp": 1.02124238, + "epoch": 0.863103469007395, + "flos": 20230061310720.0, + "grad_norm": 1.6933255102853526, + "language_loss": 0.70628834, + "learning_rate": 1.9326185170852293e-07, + "loss": 0.72825551, + "num_input_tokens_seen": 155309085, + "step": 7178, + "time_per_iteration": 2.7109487056732178 + }, + { + "auxiliary_loss_clip": 0.01157076, + "auxiliary_loss_mlp": 0.01023777, + "balance_loss_clip": 1.04490542, + "balance_loss_mlp": 1.0157181, + "epoch": 0.863223711898034, + "flos": 24498547044480.0, + "grad_norm": 1.977806234964694, + "language_loss": 0.72135162, + "learning_rate": 1.9292791587406598e-07, + "loss": 0.74316013, + "num_input_tokens_seen": 155327945, + "step": 7179, + "time_per_iteration": 2.77545428276062 + }, + { + "auxiliary_loss_clip": 0.01156369, + "auxiliary_loss_mlp": 0.00885465, + "balance_loss_clip": 1.04278135, + "balance_loss_mlp": 1.00054324, + "epoch": 0.8633439547886731, + "flos": 17675376261120.0, + "grad_norm": 2.3651360626581615, + "language_loss": 0.86411947, + "learning_rate": 1.9259425417100661e-07, + "loss": 0.88453782, + "num_input_tokens_seen": 155344060, + "step": 7180, + "time_per_iteration": 2.734562397003174 + }, + { + "auxiliary_loss_clip": 0.01114812, + "auxiliary_loss_mlp": 0.01028051, + "balance_loss_clip": 1.03263617, + "balance_loss_mlp": 1.01988554, + "epoch": 0.8634641976793123, + "flos": 12895055677440.0, + "grad_norm": 2.254198055255694, + "language_loss": 0.74766779, + "learning_rate": 1.9226086664996234e-07, + "loss": 0.76909637, + "num_input_tokens_seen": 155362305, + "step": 7181, + "time_per_iteration": 4.163851022720337 + }, + { + "auxiliary_loss_clip": 0.01143464, + "auxiliary_loss_mlp": 0.01031754, + "balance_loss_clip": 1.04475856, + "balance_loss_mlp": 1.02348137, + "epoch": 0.8635844405699513, + "flos": 23878980328320.0, + "grad_norm": 2.1327119887686483, + "language_loss": 0.74263173, + "learning_rate": 1.9192775336150712e-07, + "loss": 0.76438391, + "num_input_tokens_seen": 155382605, + "step": 7182, + "time_per_iteration": 3.796112537384033 + }, + { + "auxiliary_loss_clip": 0.01047671, + "auxiliary_loss_mlp": 0.01001692, + "balance_loss_clip": 1.00542593, + "balance_loss_mlp": 1.00026119, + "epoch": 0.8637046834605904, + "flos": 60453387521280.0, + "grad_norm": 0.7654036035503433, + "language_loss": 0.56242007, + "learning_rate": 1.915949143561739e-07, + "loss": 0.5829137, + "num_input_tokens_seen": 155437280, + "step": 7183, + "time_per_iteration": 3.2781825065612793 + }, + { + "auxiliary_loss_clip": 0.0116259, + "auxiliary_loss_mlp": 0.01033039, + "balance_loss_clip": 1.04895008, + "balance_loss_mlp": 1.02500391, + "epoch": 0.8638249263512295, + "flos": 20558751690240.0, + "grad_norm": 2.7376472650847354, + "language_loss": 0.78176272, + "learning_rate": 1.9126234968445498e-07, + "loss": 0.80371904, + "num_input_tokens_seen": 155456970, + "step": 7184, + "time_per_iteration": 2.872619152069092 + }, + { + "auxiliary_loss_clip": 0.01168014, + "auxiliary_loss_mlp": 0.01027187, + "balance_loss_clip": 1.04657316, + "balance_loss_mlp": 1.01938474, + "epoch": 0.8639451692418686, + "flos": 26615768353920.0, + "grad_norm": 7.573820092282439, + "language_loss": 0.67932218, + "learning_rate": 1.9093005939679884e-07, + "loss": 0.70127416, + "num_input_tokens_seen": 155478925, + "step": 7185, + "time_per_iteration": 2.818570852279663 + }, + { + "auxiliary_loss_clip": 0.01161226, + "auxiliary_loss_mlp": 0.01028763, + "balance_loss_clip": 1.04883361, + "balance_loss_mlp": 1.02081168, + "epoch": 0.8640654121325076, + "flos": 15122450977920.0, + "grad_norm": 1.9061818273283182, + "language_loss": 0.76727074, + "learning_rate": 1.9059804354361452e-07, + "loss": 0.78917062, + "num_input_tokens_seen": 155496700, + "step": 7186, + "time_per_iteration": 2.6966116428375244 + }, + { + "auxiliary_loss_clip": 0.01141905, + "auxiliary_loss_mlp": 0.01031055, + "balance_loss_clip": 1.04007256, + "balance_loss_mlp": 1.02300847, + "epoch": 0.8641856550231467, + "flos": 31869068250240.0, + "grad_norm": 2.0380570372091493, + "language_loss": 0.70361632, + "learning_rate": 1.902663021752684e-07, + "loss": 0.72534585, + "num_input_tokens_seen": 155518130, + "step": 7187, + "time_per_iteration": 3.8834640979766846 + }, + { + "auxiliary_loss_clip": 0.01171754, + "auxiliary_loss_mlp": 0.01030894, + "balance_loss_clip": 1.04852879, + "balance_loss_mlp": 1.02337217, + "epoch": 0.8643058979137859, + "flos": 14976545932800.0, + "grad_norm": 2.1993502214592264, + "language_loss": 0.82699394, + "learning_rate": 1.8993483534208556e-07, + "loss": 0.84902042, + "num_input_tokens_seen": 155537040, + "step": 7188, + "time_per_iteration": 2.794426918029785 + }, + { + "auxiliary_loss_clip": 0.01142293, + "auxiliary_loss_mlp": 0.01026473, + "balance_loss_clip": 1.0444957, + "balance_loss_mlp": 1.0189507, + "epoch": 0.8644261408044249, + "flos": 13115726881920.0, + "grad_norm": 3.152277338687502, + "language_loss": 0.75137734, + "learning_rate": 1.8960364309434884e-07, + "loss": 0.77306503, + "num_input_tokens_seen": 155554535, + "step": 7189, + "time_per_iteration": 2.8333048820495605 + }, + { + "auxiliary_loss_clip": 0.01118035, + "auxiliary_loss_mlp": 0.00884488, + "balance_loss_clip": 1.04031682, + "balance_loss_mlp": 1.00047398, + "epoch": 0.864546383695064, + "flos": 20850920916480.0, + "grad_norm": 1.711839927372636, + "language_loss": 0.78231889, + "learning_rate": 1.8927272548229967e-07, + "loss": 0.80234414, + "num_input_tokens_seen": 155574225, + "step": 7190, + "time_per_iteration": 2.916761636734009 + }, + { + "auxiliary_loss_clip": 0.01128868, + "auxiliary_loss_mlp": 0.01026499, + "balance_loss_clip": 1.03996372, + "balance_loss_mlp": 1.0193584, + "epoch": 0.8646666265857031, + "flos": 21324582587520.0, + "grad_norm": 1.4805457479911122, + "language_loss": 0.83158123, + "learning_rate": 1.8894208255613876e-07, + "loss": 0.85313499, + "num_input_tokens_seen": 155593540, + "step": 7191, + "time_per_iteration": 2.83931303024292 + }, + { + "auxiliary_loss_clip": 0.01168531, + "auxiliary_loss_mlp": 0.01028019, + "balance_loss_clip": 1.04687655, + "balance_loss_mlp": 1.02028871, + "epoch": 0.8647868694763422, + "flos": 19750833031680.0, + "grad_norm": 1.8756800095635877, + "language_loss": 0.77814984, + "learning_rate": 1.8861171436602397e-07, + "loss": 0.80011535, + "num_input_tokens_seen": 155610655, + "step": 7192, + "time_per_iteration": 2.672027826309204 + }, + { + "auxiliary_loss_clip": 0.01162304, + "auxiliary_loss_mlp": 0.01025832, + "balance_loss_clip": 1.04736435, + "balance_loss_mlp": 1.01782131, + "epoch": 0.8649071123669813, + "flos": 26176760328960.0, + "grad_norm": 3.390521988133777, + "language_loss": 0.8026042, + "learning_rate": 1.882816209620719e-07, + "loss": 0.82448554, + "num_input_tokens_seen": 155627365, + "step": 7193, + "time_per_iteration": 2.711489200592041 + }, + { + "auxiliary_loss_clip": 0.01151898, + "auxiliary_loss_mlp": 0.01031679, + "balance_loss_clip": 1.04621077, + "balance_loss_mlp": 1.02378178, + "epoch": 0.8650273552576204, + "flos": 20302888135680.0, + "grad_norm": 1.913164406821284, + "language_loss": 0.77000135, + "learning_rate": 1.8795180239435738e-07, + "loss": 0.7918371, + "num_input_tokens_seen": 155646220, + "step": 7194, + "time_per_iteration": 2.759842872619629 + }, + { + "auxiliary_loss_clip": 0.01151985, + "auxiliary_loss_mlp": 0.01026662, + "balance_loss_clip": 1.04230332, + "balance_loss_mlp": 1.01833582, + "epoch": 0.8651475981482595, + "flos": 23951088881280.0, + "grad_norm": 2.4747330040525113, + "language_loss": 0.75954467, + "learning_rate": 1.8762225871291348e-07, + "loss": 0.78133112, + "num_input_tokens_seen": 155662095, + "step": 7195, + "time_per_iteration": 2.808541774749756 + }, + { + "auxiliary_loss_clip": 0.01169159, + "auxiliary_loss_mlp": 0.0088559, + "balance_loss_clip": 1.04715991, + "balance_loss_mlp": 1.0005312, + "epoch": 0.8652678410388985, + "flos": 21684622561920.0, + "grad_norm": 1.669615666600001, + "language_loss": 0.80685532, + "learning_rate": 1.8729298996773201e-07, + "loss": 0.82740277, + "num_input_tokens_seen": 155680845, + "step": 7196, + "time_per_iteration": 2.6682286262512207 + }, + { + "auxiliary_loss_clip": 0.01046783, + "auxiliary_loss_mlp": 0.01001494, + "balance_loss_clip": 1.00517309, + "balance_loss_mlp": 1.00018311, + "epoch": 0.8653880839295377, + "flos": 65224660855680.0, + "grad_norm": 0.8341222916992735, + "language_loss": 0.60929728, + "learning_rate": 1.8696399620876301e-07, + "loss": 0.62978005, + "num_input_tokens_seen": 155737875, + "step": 7197, + "time_per_iteration": 3.2252705097198486 + }, + { + "auxiliary_loss_clip": 0.01133255, + "auxiliary_loss_mlp": 0.01029978, + "balance_loss_clip": 1.03656542, + "balance_loss_mlp": 1.02209532, + "epoch": 0.8655083268201768, + "flos": 17749172753280.0, + "grad_norm": 2.3116350937898944, + "language_loss": 0.79213357, + "learning_rate": 1.866352774859141e-07, + "loss": 0.81376582, + "num_input_tokens_seen": 155753100, + "step": 7198, + "time_per_iteration": 2.7634942531585693 + }, + { + "auxiliary_loss_clip": 0.01141695, + "auxiliary_loss_mlp": 0.01027981, + "balance_loss_clip": 1.03981209, + "balance_loss_mlp": 1.02039385, + "epoch": 0.8656285697108158, + "flos": 20703974376960.0, + "grad_norm": 2.1486615472425963, + "language_loss": 0.6943053, + "learning_rate": 1.8630683384905188e-07, + "loss": 0.71600205, + "num_input_tokens_seen": 155772430, + "step": 7199, + "time_per_iteration": 2.833479404449463 + }, + { + "auxiliary_loss_clip": 0.01169437, + "auxiliary_loss_mlp": 0.00885843, + "balance_loss_clip": 1.04821873, + "balance_loss_mlp": 1.0005753, + "epoch": 0.865748812601455, + "flos": 18653833716480.0, + "grad_norm": 2.4192181917977553, + "language_loss": 0.88205087, + "learning_rate": 1.8597866534800045e-07, + "loss": 0.90260369, + "num_input_tokens_seen": 155787545, + "step": 7200, + "time_per_iteration": 2.6514322757720947 + }, + { + "auxiliary_loss_clip": 0.0116382, + "auxiliary_loss_mlp": 0.00885478, + "balance_loss_clip": 1.04730535, + "balance_loss_mlp": 1.00065267, + "epoch": 0.865869055492094, + "flos": 70652554807680.0, + "grad_norm": 1.8730030736754812, + "language_loss": 0.74771059, + "learning_rate": 1.8565077203254398e-07, + "loss": 0.76820356, + "num_input_tokens_seen": 155813005, + "step": 7201, + "time_per_iteration": 3.110816717147827 + }, + { + "auxiliary_loss_clip": 0.01143551, + "auxiliary_loss_mlp": 0.01027372, + "balance_loss_clip": 1.04861104, + "balance_loss_mlp": 1.01922441, + "epoch": 0.8659892983827331, + "flos": 17383961220480.0, + "grad_norm": 3.4457134280964317, + "language_loss": 0.72729093, + "learning_rate": 1.8532315395242203e-07, + "loss": 0.74900013, + "num_input_tokens_seen": 155829455, + "step": 7202, + "time_per_iteration": 2.853248357772827 + }, + { + "auxiliary_loss_clip": 0.01141719, + "auxiliary_loss_mlp": 0.01023258, + "balance_loss_clip": 1.04247105, + "balance_loss_mlp": 1.01612294, + "epoch": 0.8661095412733723, + "flos": 17895221452800.0, + "grad_norm": 2.3366486545535783, + "language_loss": 0.72001517, + "learning_rate": 1.849958111573353e-07, + "loss": 0.74166501, + "num_input_tokens_seen": 155848060, + "step": 7203, + "time_per_iteration": 3.694256067276001 + }, + { + "auxiliary_loss_clip": 0.01166959, + "auxiliary_loss_mlp": 0.01026168, + "balance_loss_clip": 1.04600334, + "balance_loss_mlp": 1.01847363, + "epoch": 0.8662297841640113, + "flos": 18224163227520.0, + "grad_norm": 1.8137062068928056, + "language_loss": 0.64223301, + "learning_rate": 1.8466874369694074e-07, + "loss": 0.6641643, + "num_input_tokens_seen": 155865755, + "step": 7204, + "time_per_iteration": 2.6994950771331787 + }, + { + "auxiliary_loss_clip": 0.01139878, + "auxiliary_loss_mlp": 0.0102172, + "balance_loss_clip": 1.0392077, + "balance_loss_mlp": 1.01446629, + "epoch": 0.8663500270546504, + "flos": 16362159027840.0, + "grad_norm": 2.2509867654576254, + "language_loss": 0.69972968, + "learning_rate": 1.843419516208542e-07, + "loss": 0.7213456, + "num_input_tokens_seen": 155882680, + "step": 7205, + "time_per_iteration": 2.7728655338287354 + }, + { + "auxiliary_loss_clip": 0.01163729, + "auxiliary_loss_mlp": 0.01029394, + "balance_loss_clip": 1.05052686, + "balance_loss_mlp": 1.02110028, + "epoch": 0.8664702699452895, + "flos": 17894431353600.0, + "grad_norm": 2.2787006443070736, + "language_loss": 0.80031049, + "learning_rate": 1.8401543497865047e-07, + "loss": 0.82224166, + "num_input_tokens_seen": 155900680, + "step": 7206, + "time_per_iteration": 3.687674045562744 + }, + { + "auxiliary_loss_clip": 0.01161503, + "auxiliary_loss_mlp": 0.00885232, + "balance_loss_clip": 1.04553795, + "balance_loss_mlp": 1.00051415, + "epoch": 0.8665905128359286, + "flos": 30736373794560.0, + "grad_norm": 2.1252337812526516, + "language_loss": 0.64247173, + "learning_rate": 1.836891938198608e-07, + "loss": 0.66293913, + "num_input_tokens_seen": 155921105, + "step": 7207, + "time_per_iteration": 2.8635659217834473 + }, + { + "auxiliary_loss_clip": 0.01148018, + "auxiliary_loss_mlp": 0.01026144, + "balance_loss_clip": 1.04442346, + "balance_loss_mlp": 1.01841056, + "epoch": 0.8667107557265676, + "flos": 18656419495680.0, + "grad_norm": 2.149630409620414, + "language_loss": 0.71330136, + "learning_rate": 1.8336322819397677e-07, + "loss": 0.73504299, + "num_input_tokens_seen": 155938640, + "step": 7208, + "time_per_iteration": 3.6358132362365723 + }, + { + "auxiliary_loss_clip": 0.01143244, + "auxiliary_loss_mlp": 0.01033617, + "balance_loss_clip": 1.04017425, + "balance_loss_mlp": 1.02596951, + "epoch": 0.8668309986172068, + "flos": 20083725302400.0, + "grad_norm": 1.799505166953558, + "language_loss": 0.62806022, + "learning_rate": 1.8303753815044654e-07, + "loss": 0.64982879, + "num_input_tokens_seen": 155957945, + "step": 7209, + "time_per_iteration": 2.88615345954895 + }, + { + "auxiliary_loss_clip": 0.01160226, + "auxiliary_loss_mlp": 0.01031459, + "balance_loss_clip": 1.04535127, + "balance_loss_mlp": 1.02252984, + "epoch": 0.8669512415078459, + "flos": 21615099788160.0, + "grad_norm": 2.5990407449640553, + "language_loss": 0.69576299, + "learning_rate": 1.827121237386773e-07, + "loss": 0.71767986, + "num_input_tokens_seen": 155975390, + "step": 7210, + "time_per_iteration": 2.7956247329711914 + }, + { + "auxiliary_loss_clip": 0.01152191, + "auxiliary_loss_mlp": 0.0102658, + "balance_loss_clip": 1.04452527, + "balance_loss_mlp": 1.01874816, + "epoch": 0.8670714843984849, + "flos": 17703601372800.0, + "grad_norm": 2.68459055278505, + "language_loss": 0.75151473, + "learning_rate": 1.8238698500803374e-07, + "loss": 0.7733025, + "num_input_tokens_seen": 155988155, + "step": 7211, + "time_per_iteration": 2.744170904159546 + }, + { + "auxiliary_loss_clip": 0.01052221, + "auxiliary_loss_mlp": 0.01002495, + "balance_loss_clip": 1.0052228, + "balance_loss_mlp": 1.00110006, + "epoch": 0.8671917272891241, + "flos": 60705483125760.0, + "grad_norm": 0.7168306032956818, + "language_loss": 0.56272233, + "learning_rate": 1.820621220078391e-07, + "loss": 0.58326948, + "num_input_tokens_seen": 156052065, + "step": 7212, + "time_per_iteration": 4.350037574768066 + }, + { + "auxiliary_loss_clip": 0.01168028, + "auxiliary_loss_mlp": 0.01031055, + "balance_loss_clip": 1.04646373, + "balance_loss_mlp": 1.02332425, + "epoch": 0.8673119701797631, + "flos": 20451881750400.0, + "grad_norm": 1.5848443410556126, + "language_loss": 0.67786521, + "learning_rate": 1.8173753478737553e-07, + "loss": 0.69985604, + "num_input_tokens_seen": 156072500, + "step": 7213, + "time_per_iteration": 2.8004653453826904 + }, + { + "auxiliary_loss_clip": 0.01170227, + "auxiliary_loss_mlp": 0.01027442, + "balance_loss_clip": 1.04756188, + "balance_loss_mlp": 1.01925194, + "epoch": 0.8674322130704022, + "flos": 19647410797440.0, + "grad_norm": 2.111824511873485, + "language_loss": 0.7944417, + "learning_rate": 1.8141322339588205e-07, + "loss": 0.81641835, + "num_input_tokens_seen": 156089840, + "step": 7214, + "time_per_iteration": 2.6943345069885254 + }, + { + "auxiliary_loss_clip": 0.01168779, + "auxiliary_loss_mlp": 0.01026048, + "balance_loss_clip": 1.04774213, + "balance_loss_mlp": 1.01813221, + "epoch": 0.8675524559610414, + "flos": 26025001367040.0, + "grad_norm": 2.0181892578434826, + "language_loss": 0.70349085, + "learning_rate": 1.810891878825569e-07, + "loss": 0.72543919, + "num_input_tokens_seen": 156109815, + "step": 7215, + "time_per_iteration": 2.843027114868164 + }, + { + "auxiliary_loss_clip": 0.01146881, + "auxiliary_loss_mlp": 0.01027279, + "balance_loss_clip": 1.04192495, + "balance_loss_mlp": 1.01941085, + "epoch": 0.8676726988516804, + "flos": 15049444584960.0, + "grad_norm": 2.637471066824862, + "language_loss": 0.71609628, + "learning_rate": 1.8076542829655561e-07, + "loss": 0.73783791, + "num_input_tokens_seen": 156128620, + "step": 7216, + "time_per_iteration": 2.696910858154297 + }, + { + "auxiliary_loss_clip": 0.01149696, + "auxiliary_loss_mlp": 0.01026537, + "balance_loss_clip": 1.04560649, + "balance_loss_mlp": 1.01831722, + "epoch": 0.8677929417423195, + "flos": 16288111140480.0, + "grad_norm": 2.6841923121766538, + "language_loss": 0.79388553, + "learning_rate": 1.8044194468699203e-07, + "loss": 0.81564784, + "num_input_tokens_seen": 156145930, + "step": 7217, + "time_per_iteration": 2.778640031814575 + }, + { + "auxiliary_loss_clip": 0.01148607, + "auxiliary_loss_mlp": 0.01024326, + "balance_loss_clip": 1.0474577, + "balance_loss_mlp": 1.01679206, + "epoch": 0.8679131846329585, + "flos": 18844160906880.0, + "grad_norm": 4.816951617958628, + "language_loss": 0.75559199, + "learning_rate": 1.8011873710293912e-07, + "loss": 0.77732122, + "num_input_tokens_seen": 156164435, + "step": 7218, + "time_per_iteration": 2.665119171142578 + }, + { + "auxiliary_loss_clip": 0.01158393, + "auxiliary_loss_mlp": 0.01027012, + "balance_loss_clip": 1.04613376, + "balance_loss_mlp": 1.01917124, + "epoch": 0.8680334275235977, + "flos": 33620718890880.0, + "grad_norm": 2.054480623948904, + "language_loss": 0.69026792, + "learning_rate": 1.7979580559342677e-07, + "loss": 0.71212196, + "num_input_tokens_seen": 156185165, + "step": 7219, + "time_per_iteration": 2.8445205688476562 + }, + { + "auxiliary_loss_clip": 0.01147465, + "auxiliary_loss_mlp": 0.01025021, + "balance_loss_clip": 1.04420912, + "balance_loss_mlp": 1.01673651, + "epoch": 0.8681536704142367, + "flos": 24681152810880.0, + "grad_norm": 1.6357621019150088, + "language_loss": 0.66874743, + "learning_rate": 1.7947315020744358e-07, + "loss": 0.69047225, + "num_input_tokens_seen": 156206260, + "step": 7220, + "time_per_iteration": 2.857907772064209 + }, + { + "auxiliary_loss_clip": 0.01147919, + "auxiliary_loss_mlp": 0.01026369, + "balance_loss_clip": 1.04234207, + "balance_loss_mlp": 1.01880562, + "epoch": 0.8682739133048758, + "flos": 20011042131840.0, + "grad_norm": 1.8233898558197796, + "language_loss": 0.80569911, + "learning_rate": 1.7915077099393594e-07, + "loss": 0.82744205, + "num_input_tokens_seen": 156222860, + "step": 7221, + "time_per_iteration": 2.8049333095550537 + }, + { + "auxiliary_loss_clip": 0.0116091, + "auxiliary_loss_mlp": 0.0102631, + "balance_loss_clip": 1.04438853, + "balance_loss_mlp": 1.01828766, + "epoch": 0.868394156195515, + "flos": 16654759217280.0, + "grad_norm": 2.133695407392764, + "language_loss": 0.73183036, + "learning_rate": 1.788286680018083e-07, + "loss": 0.75370252, + "num_input_tokens_seen": 156241570, + "step": 7222, + "time_per_iteration": 2.678046226501465 + }, + { + "auxiliary_loss_clip": 0.01152055, + "auxiliary_loss_mlp": 0.01026295, + "balance_loss_clip": 1.04449594, + "balance_loss_mlp": 1.01859105, + "epoch": 0.868514399086154, + "flos": 28001381448960.0, + "grad_norm": 9.252161342119054, + "language_loss": 0.72214127, + "learning_rate": 1.7850684127992443e-07, + "loss": 0.74392474, + "num_input_tokens_seen": 156261315, + "step": 7223, + "time_per_iteration": 2.767103672027588 + }, + { + "auxiliary_loss_clip": 0.01141755, + "auxiliary_loss_mlp": 0.01026019, + "balance_loss_clip": 1.04618478, + "balance_loss_mlp": 1.01872933, + "epoch": 0.8686346419767931, + "flos": 20084587228800.0, + "grad_norm": 1.89534171591495, + "language_loss": 0.70398277, + "learning_rate": 1.7818529087710378e-07, + "loss": 0.7256605, + "num_input_tokens_seen": 156281670, + "step": 7224, + "time_per_iteration": 2.887176036834717 + }, + { + "auxiliary_loss_clip": 0.01157538, + "auxiliary_loss_mlp": 0.00885886, + "balance_loss_clip": 1.0452981, + "balance_loss_mlp": 1.00057387, + "epoch": 0.8687548848674322, + "flos": 18223516782720.0, + "grad_norm": 1.746327807812503, + "language_loss": 0.83985841, + "learning_rate": 1.7786401684212637e-07, + "loss": 0.86029267, + "num_input_tokens_seen": 156300500, + "step": 7225, + "time_per_iteration": 2.7343366146087646 + }, + { + "auxiliary_loss_clip": 0.01032037, + "auxiliary_loss_mlp": 0.01005051, + "balance_loss_clip": 1.00704813, + "balance_loss_mlp": 1.00355458, + "epoch": 0.8688751277580713, + "flos": 70457885049600.0, + "grad_norm": 0.7386400225930355, + "language_loss": 0.55953908, + "learning_rate": 1.7754301922372883e-07, + "loss": 0.57990998, + "num_input_tokens_seen": 156350145, + "step": 7226, + "time_per_iteration": 3.33971905708313 + }, + { + "auxiliary_loss_clip": 0.01126695, + "auxiliary_loss_mlp": 0.01028007, + "balance_loss_clip": 1.04035974, + "balance_loss_mlp": 1.02005005, + "epoch": 0.8689953706487104, + "flos": 26906788344960.0, + "grad_norm": 1.7850803978859302, + "language_loss": 0.80791426, + "learning_rate": 1.7722229807060617e-07, + "loss": 0.82946134, + "num_input_tokens_seen": 156368725, + "step": 7227, + "time_per_iteration": 3.042764902114868 + }, + { + "auxiliary_loss_clip": 0.01127202, + "auxiliary_loss_mlp": 0.01030393, + "balance_loss_clip": 1.03784108, + "balance_loss_mlp": 1.02256668, + "epoch": 0.8691156135393495, + "flos": 34637385438720.0, + "grad_norm": 2.2348371226750627, + "language_loss": 0.81791747, + "learning_rate": 1.7690185343141172e-07, + "loss": 0.83949345, + "num_input_tokens_seen": 156388640, + "step": 7228, + "time_per_iteration": 3.0131492614746094 + }, + { + "auxiliary_loss_clip": 0.01147273, + "auxiliary_loss_mlp": 0.01023368, + "balance_loss_clip": 1.0409379, + "balance_loss_mlp": 1.0159471, + "epoch": 0.8692358564299886, + "flos": 18989814556800.0, + "grad_norm": 2.085976776508945, + "language_loss": 0.69898558, + "learning_rate": 1.7658168535475615e-07, + "loss": 0.72069192, + "num_input_tokens_seen": 156406425, + "step": 7229, + "time_per_iteration": 3.7974178791046143 + }, + { + "auxiliary_loss_clip": 0.01156057, + "auxiliary_loss_mlp": 0.01023775, + "balance_loss_clip": 1.04724288, + "balance_loss_mlp": 1.01576447, + "epoch": 0.8693560993206276, + "flos": 30370839039360.0, + "grad_norm": 1.4922736653546425, + "language_loss": 0.64280897, + "learning_rate": 1.7626179388920948e-07, + "loss": 0.66460729, + "num_input_tokens_seen": 156427705, + "step": 7230, + "time_per_iteration": 2.8456902503967285 + }, + { + "auxiliary_loss_clip": 0.01149527, + "auxiliary_loss_mlp": 0.00885515, + "balance_loss_clip": 1.04397082, + "balance_loss_mlp": 1.00068295, + "epoch": 0.8694763422112668, + "flos": 27200430028800.0, + "grad_norm": 1.6426719479051077, + "language_loss": 0.80467176, + "learning_rate": 1.7594217908329866e-07, + "loss": 0.82502216, + "num_input_tokens_seen": 156449890, + "step": 7231, + "time_per_iteration": 3.060258388519287 + }, + { + "auxiliary_loss_clip": 0.01139569, + "auxiliary_loss_mlp": 0.01025319, + "balance_loss_clip": 1.04216552, + "balance_loss_mlp": 1.017946, + "epoch": 0.8695965851019059, + "flos": 26139161767680.0, + "grad_norm": 1.8080553753462771, + "language_loss": 0.74056697, + "learning_rate": 1.7562284098550895e-07, + "loss": 0.76221591, + "num_input_tokens_seen": 156469600, + "step": 7232, + "time_per_iteration": 3.8092517852783203 + }, + { + "auxiliary_loss_clip": 0.0104302, + "auxiliary_loss_mlp": 0.01002947, + "balance_loss_clip": 1.01019764, + "balance_loss_mlp": 1.00157583, + "epoch": 0.8697168279925449, + "flos": 67332616456320.0, + "grad_norm": 0.8560130945503779, + "language_loss": 0.62215281, + "learning_rate": 1.753037796442838e-07, + "loss": 0.64261258, + "num_input_tokens_seen": 156529040, + "step": 7233, + "time_per_iteration": 4.172805070877075 + }, + { + "auxiliary_loss_clip": 0.01167159, + "auxiliary_loss_mlp": 0.01027564, + "balance_loss_clip": 1.04502404, + "balance_loss_mlp": 1.01994634, + "epoch": 0.8698370708831841, + "flos": 19718693337600.0, + "grad_norm": 2.2289413762783488, + "language_loss": 0.75153816, + "learning_rate": 1.74984995108024e-07, + "loss": 0.77348536, + "num_input_tokens_seen": 156546970, + "step": 7234, + "time_per_iteration": 2.714588165283203 + }, + { + "auxiliary_loss_clip": 0.01161237, + "auxiliary_loss_mlp": 0.01028399, + "balance_loss_clip": 1.0468595, + "balance_loss_mlp": 1.02020931, + "epoch": 0.8699573137738231, + "flos": 12859971068160.0, + "grad_norm": 2.0667972873719513, + "language_loss": 0.83287513, + "learning_rate": 1.7466648742508981e-07, + "loss": 0.85477144, + "num_input_tokens_seen": 156563155, + "step": 7235, + "time_per_iteration": 2.767467498779297 + }, + { + "auxiliary_loss_clip": 0.01151651, + "auxiliary_loss_mlp": 0.01030642, + "balance_loss_clip": 1.04703224, + "balance_loss_mlp": 1.0230186, + "epoch": 0.8700775566644622, + "flos": 17420733768960.0, + "grad_norm": 1.8968966781973904, + "language_loss": 0.84493959, + "learning_rate": 1.7434825664379837e-07, + "loss": 0.86676252, + "num_input_tokens_seen": 156581660, + "step": 7236, + "time_per_iteration": 2.726156234741211 + }, + { + "auxiliary_loss_clip": 0.01160307, + "auxiliary_loss_mlp": 0.01029228, + "balance_loss_clip": 1.04579616, + "balance_loss_mlp": 1.02143133, + "epoch": 0.8701977995551013, + "flos": 13735221770880.0, + "grad_norm": 2.4661655882100577, + "language_loss": 0.85866058, + "learning_rate": 1.740303028124246e-07, + "loss": 0.88055593, + "num_input_tokens_seen": 156597720, + "step": 7237, + "time_per_iteration": 2.7358412742614746 + }, + { + "auxiliary_loss_clip": 0.01110112, + "auxiliary_loss_mlp": 0.0102724, + "balance_loss_clip": 1.03623986, + "balance_loss_mlp": 1.01964009, + "epoch": 0.8703180424457404, + "flos": 30555707362560.0, + "grad_norm": 1.7420479967910691, + "language_loss": 0.75207007, + "learning_rate": 1.7371262597920212e-07, + "loss": 0.77344358, + "num_input_tokens_seen": 156619780, + "step": 7238, + "time_per_iteration": 3.965378522872925 + }, + { + "auxiliary_loss_clip": 0.01129674, + "auxiliary_loss_mlp": 0.01032243, + "balance_loss_clip": 1.04398966, + "balance_loss_mlp": 1.02476287, + "epoch": 0.8704382853363795, + "flos": 19608986223360.0, + "grad_norm": 1.7071015974369501, + "language_loss": 0.76153404, + "learning_rate": 1.7339522619232195e-07, + "loss": 0.78315318, + "num_input_tokens_seen": 156638160, + "step": 7239, + "time_per_iteration": 2.895993232727051 + }, + { + "auxiliary_loss_clip": 0.01158096, + "auxiliary_loss_mlp": 0.01028004, + "balance_loss_clip": 1.04444909, + "balance_loss_mlp": 1.02030301, + "epoch": 0.8705585282270186, + "flos": 26613900846720.0, + "grad_norm": 1.9870787426716303, + "language_loss": 0.75416481, + "learning_rate": 1.730781034999338e-07, + "loss": 0.77602589, + "num_input_tokens_seen": 156659740, + "step": 7240, + "time_per_iteration": 2.8679757118225098 + }, + { + "auxiliary_loss_clip": 0.01165451, + "auxiliary_loss_mlp": 0.01030494, + "balance_loss_clip": 1.0483973, + "balance_loss_mlp": 1.02255166, + "epoch": 0.8706787711176577, + "flos": 34090465979520.0, + "grad_norm": 1.8409869100060585, + "language_loss": 0.73232889, + "learning_rate": 1.7276125795014497e-07, + "loss": 0.75428838, + "num_input_tokens_seen": 156678190, + "step": 7241, + "time_per_iteration": 2.735013008117676 + }, + { + "auxiliary_loss_clip": 0.01152284, + "auxiliary_loss_mlp": 0.01028287, + "balance_loss_clip": 1.04175806, + "balance_loss_mlp": 1.01959109, + "epoch": 0.8707990140082967, + "flos": 14611513968000.0, + "grad_norm": 1.7905214664664852, + "language_loss": 0.67349768, + "learning_rate": 1.7244468959102054e-07, + "loss": 0.69530338, + "num_input_tokens_seen": 156695245, + "step": 7242, + "time_per_iteration": 2.8848493099212646 + }, + { + "auxiliary_loss_clip": 0.01159547, + "auxiliary_loss_mlp": 0.01030343, + "balance_loss_clip": 1.04596162, + "balance_loss_mlp": 1.02205777, + "epoch": 0.8709192568989359, + "flos": 20084156265600.0, + "grad_norm": 2.0836461763556704, + "language_loss": 0.85165566, + "learning_rate": 1.7212839847058348e-07, + "loss": 0.87355459, + "num_input_tokens_seen": 156710375, + "step": 7243, + "time_per_iteration": 2.871346950531006 + }, + { + "auxiliary_loss_clip": 0.01121967, + "auxiliary_loss_mlp": 0.01025647, + "balance_loss_clip": 1.03918862, + "balance_loss_mlp": 1.01800013, + "epoch": 0.871039499789575, + "flos": 16727083251840.0, + "grad_norm": 1.867195742405476, + "language_loss": 0.73687619, + "learning_rate": 1.718123846368147e-07, + "loss": 0.75835234, + "num_input_tokens_seen": 156729420, + "step": 7244, + "time_per_iteration": 2.8030834197998047 + }, + { + "auxiliary_loss_clip": 0.01148493, + "auxiliary_loss_mlp": 0.00885089, + "balance_loss_clip": 1.04553151, + "balance_loss_mlp": 1.00056124, + "epoch": 0.871159742680214, + "flos": 21068790860160.0, + "grad_norm": 1.7644280620915154, + "language_loss": 0.71354228, + "learning_rate": 1.714966481376543e-07, + "loss": 0.73387814, + "num_input_tokens_seen": 156746100, + "step": 7245, + "time_per_iteration": 2.668722629547119 + }, + { + "auxiliary_loss_clip": 0.01157988, + "auxiliary_loss_mlp": 0.01024267, + "balance_loss_clip": 1.04485369, + "balance_loss_mlp": 1.01638103, + "epoch": 0.8712799855708532, + "flos": 28256526731520.0, + "grad_norm": 1.9083458292810433, + "language_loss": 0.83039969, + "learning_rate": 1.7118118902099797e-07, + "loss": 0.8522222, + "num_input_tokens_seen": 156764185, + "step": 7246, + "time_per_iteration": 2.619445323944092 + }, + { + "auxiliary_loss_clip": 0.01159754, + "auxiliary_loss_mlp": 0.01030922, + "balance_loss_clip": 1.04535198, + "balance_loss_mlp": 1.02327478, + "epoch": 0.8714002284614922, + "flos": 22236677665920.0, + "grad_norm": 1.7436645423437798, + "language_loss": 0.80659449, + "learning_rate": 1.7086600733470146e-07, + "loss": 0.82850122, + "num_input_tokens_seen": 156784855, + "step": 7247, + "time_per_iteration": 2.7117843627929688 + }, + { + "auxiliary_loss_clip": 0.0115487, + "auxiliary_loss_mlp": 0.01024009, + "balance_loss_clip": 1.04528069, + "balance_loss_mlp": 1.01655269, + "epoch": 0.8715204713521313, + "flos": 21431919404160.0, + "grad_norm": 1.8012716481105953, + "language_loss": 0.7702108, + "learning_rate": 1.7055110312657738e-07, + "loss": 0.79199958, + "num_input_tokens_seen": 156804350, + "step": 7248, + "time_per_iteration": 2.5608227252960205 + }, + { + "auxiliary_loss_clip": 0.01143887, + "auxiliary_loss_mlp": 0.01028404, + "balance_loss_clip": 1.04319715, + "balance_loss_mlp": 1.02086437, + "epoch": 0.8716407142427703, + "flos": 23440439180160.0, + "grad_norm": 2.241630860618792, + "language_loss": 0.73760879, + "learning_rate": 1.702364764443962e-07, + "loss": 0.75933176, + "num_input_tokens_seen": 156823425, + "step": 7249, + "time_per_iteration": 2.64359450340271 + }, + { + "auxiliary_loss_clip": 0.01117283, + "auxiliary_loss_mlp": 0.01031926, + "balance_loss_clip": 1.03709269, + "balance_loss_mlp": 1.02387333, + "epoch": 0.8717609571334095, + "flos": 27958683156480.0, + "grad_norm": 1.965646752566735, + "language_loss": 0.72516441, + "learning_rate": 1.6992212733588685e-07, + "loss": 0.74665654, + "num_input_tokens_seen": 156843090, + "step": 7250, + "time_per_iteration": 2.7948243618011475 + }, + { + "auxiliary_loss_clip": 0.01145528, + "auxiliary_loss_mlp": 0.01031939, + "balance_loss_clip": 1.04258251, + "balance_loss_mlp": 1.02422893, + "epoch": 0.8718812000240486, + "flos": 25479482538240.0, + "grad_norm": 1.7626973410165623, + "language_loss": 0.74879241, + "learning_rate": 1.6960805584873538e-07, + "loss": 0.77056706, + "num_input_tokens_seen": 156861090, + "step": 7251, + "time_per_iteration": 2.860063314437866 + }, + { + "auxiliary_loss_clip": 0.01133783, + "auxiliary_loss_mlp": 0.01026029, + "balance_loss_clip": 1.04131079, + "balance_loss_mlp": 1.0183636, + "epoch": 0.8720014429146876, + "flos": 23403056100480.0, + "grad_norm": 1.5346837818753662, + "language_loss": 0.77959156, + "learning_rate": 1.6929426203058684e-07, + "loss": 0.80118966, + "num_input_tokens_seen": 156881515, + "step": 7252, + "time_per_iteration": 2.889223575592041 + }, + { + "auxiliary_loss_clip": 0.01169811, + "auxiliary_loss_mlp": 0.00886192, + "balance_loss_clip": 1.044626, + "balance_loss_mlp": 1.00058162, + "epoch": 0.8721216858053268, + "flos": 24352821567360.0, + "grad_norm": 2.135827345698193, + "language_loss": 0.79898143, + "learning_rate": 1.689807459290431e-07, + "loss": 0.81954145, + "num_input_tokens_seen": 156900170, + "step": 7253, + "time_per_iteration": 2.8055574893951416 + }, + { + "auxiliary_loss_clip": 0.0115186, + "auxiliary_loss_mlp": 0.01028396, + "balance_loss_clip": 1.04628849, + "balance_loss_mlp": 1.0208416, + "epoch": 0.8722419286959658, + "flos": 33869687034240.0, + "grad_norm": 2.1252811594044507, + "language_loss": 0.70626003, + "learning_rate": 1.6866750759166437e-07, + "loss": 0.72806257, + "num_input_tokens_seen": 156920150, + "step": 7254, + "time_per_iteration": 2.995288133621216 + }, + { + "auxiliary_loss_clip": 0.01136319, + "auxiliary_loss_mlp": 0.01023463, + "balance_loss_clip": 1.03972685, + "balance_loss_mlp": 1.0153749, + "epoch": 0.8723621715866049, + "flos": 18369385914240.0, + "grad_norm": 8.496481602594109, + "language_loss": 0.77175593, + "learning_rate": 1.6835454706596865e-07, + "loss": 0.7933538, + "num_input_tokens_seen": 156937980, + "step": 7255, + "time_per_iteration": 3.775668144226074 + }, + { + "auxiliary_loss_clip": 0.01170554, + "auxiliary_loss_mlp": 0.01031011, + "balance_loss_clip": 1.04900599, + "balance_loss_mlp": 1.02273214, + "epoch": 0.8724824144772441, + "flos": 22013348855040.0, + "grad_norm": 1.8740479353425656, + "language_loss": 0.73646307, + "learning_rate": 1.680418643994317e-07, + "loss": 0.7584787, + "num_input_tokens_seen": 156956550, + "step": 7256, + "time_per_iteration": 2.6476197242736816 + }, + { + "auxiliary_loss_clip": 0.0105542, + "auxiliary_loss_mlp": 0.01002156, + "balance_loss_clip": 1.00467539, + "balance_loss_mlp": 1.00079727, + "epoch": 0.8726026573678831, + "flos": 66698720213760.0, + "grad_norm": 0.8858133345024193, + "language_loss": 0.64473408, + "learning_rate": 1.6772945963948738e-07, + "loss": 0.66530985, + "num_input_tokens_seen": 157014715, + "step": 7257, + "time_per_iteration": 3.170048236846924 + }, + { + "auxiliary_loss_clip": 0.01147859, + "auxiliary_loss_mlp": 0.01028204, + "balance_loss_clip": 1.04395974, + "balance_loss_mlp": 1.0203656, + "epoch": 0.8727229002585222, + "flos": 13370908078080.0, + "grad_norm": 2.152798800696649, + "language_loss": 0.77820122, + "learning_rate": 1.6741733283352733e-07, + "loss": 0.79996181, + "num_input_tokens_seen": 157032320, + "step": 7258, + "time_per_iteration": 3.586106300354004 + }, + { + "auxiliary_loss_clip": 0.01127775, + "auxiliary_loss_mlp": 0.01027535, + "balance_loss_clip": 1.04217815, + "balance_loss_mlp": 1.01921427, + "epoch": 0.8728431431491613, + "flos": 21796987282560.0, + "grad_norm": 1.5881402442427721, + "language_loss": 0.83719587, + "learning_rate": 1.6710548402890102e-07, + "loss": 0.85874897, + "num_input_tokens_seen": 157052845, + "step": 7259, + "time_per_iteration": 3.789445161819458 + }, + { + "auxiliary_loss_clip": 0.01174745, + "auxiliary_loss_mlp": 0.01025812, + "balance_loss_clip": 1.04889441, + "balance_loss_mlp": 1.01746702, + "epoch": 0.8729633860398004, + "flos": 36173823742080.0, + "grad_norm": 3.14026938767594, + "language_loss": 0.66732752, + "learning_rate": 1.6679391327291527e-07, + "loss": 0.68933308, + "num_input_tokens_seen": 157074050, + "step": 7260, + "time_per_iteration": 2.845470428466797 + }, + { + "auxiliary_loss_clip": 0.01147438, + "auxiliary_loss_mlp": 0.01024263, + "balance_loss_clip": 1.03995275, + "balance_loss_mlp": 1.01674414, + "epoch": 0.8730836289304394, + "flos": 16359680989440.0, + "grad_norm": 2.752998999790271, + "language_loss": 0.6790446, + "learning_rate": 1.6648262061283492e-07, + "loss": 0.70076168, + "num_input_tokens_seen": 157089350, + "step": 7261, + "time_per_iteration": 2.757227659225464 + }, + { + "auxiliary_loss_clip": 0.0113849, + "auxiliary_loss_mlp": 0.01027257, + "balance_loss_clip": 1.0387733, + "balance_loss_mlp": 1.01905537, + "epoch": 0.8732038718210786, + "flos": 21215126868480.0, + "grad_norm": 1.9520015799063017, + "language_loss": 0.73255301, + "learning_rate": 1.6617160609588353e-07, + "loss": 0.75421047, + "num_input_tokens_seen": 157108525, + "step": 7262, + "time_per_iteration": 2.825744390487671 + }, + { + "auxiliary_loss_clip": 0.01145699, + "auxiliary_loss_mlp": 0.01027679, + "balance_loss_clip": 1.04483712, + "balance_loss_mlp": 1.01981676, + "epoch": 0.8733241147117177, + "flos": 16610696208000.0, + "grad_norm": 2.2679517596649656, + "language_loss": 0.72304547, + "learning_rate": 1.6586086976924163e-07, + "loss": 0.74477923, + "num_input_tokens_seen": 157124025, + "step": 7263, + "time_per_iteration": 2.832374334335327 + }, + { + "auxiliary_loss_clip": 0.01159249, + "auxiliary_loss_mlp": 0.01021901, + "balance_loss_clip": 1.04484057, + "balance_loss_mlp": 1.01490676, + "epoch": 0.8734443576023567, + "flos": 20193935207040.0, + "grad_norm": 1.9819574453108992, + "language_loss": 0.78554368, + "learning_rate": 1.6555041168004747e-07, + "loss": 0.80735523, + "num_input_tokens_seen": 157143345, + "step": 7264, + "time_per_iteration": 3.717222213745117 + }, + { + "auxiliary_loss_clip": 0.01143476, + "auxiliary_loss_mlp": 0.01024887, + "balance_loss_clip": 1.04229975, + "balance_loss_mlp": 1.01751995, + "epoch": 0.8735646004929959, + "flos": 18041162411520.0, + "grad_norm": 1.9304724096389816, + "language_loss": 0.69138718, + "learning_rate": 1.6524023187539715e-07, + "loss": 0.71307081, + "num_input_tokens_seen": 157161630, + "step": 7265, + "time_per_iteration": 2.6551201343536377 + }, + { + "auxiliary_loss_clip": 0.01147513, + "auxiliary_loss_mlp": 0.01026572, + "balance_loss_clip": 1.04251814, + "balance_loss_mlp": 1.0187223, + "epoch": 0.873684843383635, + "flos": 20262344659200.0, + "grad_norm": 2.0856569380165837, + "language_loss": 0.74744934, + "learning_rate": 1.649303304023446e-07, + "loss": 0.76919019, + "num_input_tokens_seen": 157181385, + "step": 7266, + "time_per_iteration": 2.696655511856079 + }, + { + "auxiliary_loss_clip": 0.0113562, + "auxiliary_loss_mlp": 0.01029646, + "balance_loss_clip": 1.04387391, + "balance_loss_mlp": 1.0219512, + "epoch": 0.873805086274274, + "flos": 16947287579520.0, + "grad_norm": 1.634212403109975, + "language_loss": 0.78754711, + "learning_rate": 1.6462070730790246e-07, + "loss": 0.80919975, + "num_input_tokens_seen": 157200545, + "step": 7267, + "time_per_iteration": 2.775653123855591 + }, + { + "auxiliary_loss_clip": 0.01144146, + "auxiliary_loss_mlp": 0.01022886, + "balance_loss_clip": 1.04020882, + "balance_loss_mlp": 1.01538825, + "epoch": 0.8739253291649132, + "flos": 18041270152320.0, + "grad_norm": 2.7201992446657752, + "language_loss": 0.78524572, + "learning_rate": 1.6431136263903912e-07, + "loss": 0.806916, + "num_input_tokens_seen": 157219545, + "step": 7268, + "time_per_iteration": 2.6577234268188477 + }, + { + "auxiliary_loss_clip": 0.01162292, + "auxiliary_loss_mlp": 0.00885637, + "balance_loss_clip": 1.04396248, + "balance_loss_mlp": 1.00061607, + "epoch": 0.8740455720555522, + "flos": 21325085377920.0, + "grad_norm": 1.8231839522060116, + "language_loss": 0.73498726, + "learning_rate": 1.6400229644268282e-07, + "loss": 0.75546658, + "num_input_tokens_seen": 157237900, + "step": 7269, + "time_per_iteration": 2.675971031188965 + }, + { + "auxiliary_loss_clip": 0.01135658, + "auxiliary_loss_mlp": 0.01028208, + "balance_loss_clip": 1.04602492, + "balance_loss_mlp": 1.02007818, + "epoch": 0.8741658149461913, + "flos": 15158684822400.0, + "grad_norm": 2.054584958481651, + "language_loss": 0.8120181, + "learning_rate": 1.6369350876571852e-07, + "loss": 0.83365679, + "num_input_tokens_seen": 157256055, + "step": 7270, + "time_per_iteration": 2.7677223682403564 + }, + { + "auxiliary_loss_clip": 0.01125962, + "auxiliary_loss_mlp": 0.01023403, + "balance_loss_clip": 1.03933012, + "balance_loss_mlp": 1.01581514, + "epoch": 0.8742860578368304, + "flos": 23039855729280.0, + "grad_norm": 2.4550326399815385, + "language_loss": 0.81571132, + "learning_rate": 1.6338499965498874e-07, + "loss": 0.83720499, + "num_input_tokens_seen": 157274785, + "step": 7271, + "time_per_iteration": 2.832347869873047 + }, + { + "auxiliary_loss_clip": 0.01134204, + "auxiliary_loss_mlp": 0.01026932, + "balance_loss_clip": 1.04083633, + "balance_loss_mlp": 1.0191772, + "epoch": 0.8744063007274695, + "flos": 28145347159680.0, + "grad_norm": 1.4665945715078266, + "language_loss": 0.77178788, + "learning_rate": 1.630767691572943e-07, + "loss": 0.79339921, + "num_input_tokens_seen": 157294805, + "step": 7272, + "time_per_iteration": 2.810999870300293 + }, + { + "auxiliary_loss_clip": 0.01046655, + "auxiliary_loss_mlp": 0.01002175, + "balance_loss_clip": 1.00500774, + "balance_loss_mlp": 1.00062561, + "epoch": 0.8745265436181086, + "flos": 64034076654720.0, + "grad_norm": 0.7365585634311795, + "language_loss": 0.53560054, + "learning_rate": 1.6276881731939306e-07, + "loss": 0.55608881, + "num_input_tokens_seen": 157356695, + "step": 7273, + "time_per_iteration": 3.3687400817871094 + }, + { + "auxiliary_loss_clip": 0.01154093, + "auxiliary_loss_mlp": 0.01026847, + "balance_loss_clip": 1.04476607, + "balance_loss_mlp": 1.01905715, + "epoch": 0.8746467865087477, + "flos": 28658618553600.0, + "grad_norm": 1.7198007630564562, + "language_loss": 0.75096434, + "learning_rate": 1.6246114418800193e-07, + "loss": 0.77277374, + "num_input_tokens_seen": 157376975, + "step": 7274, + "time_per_iteration": 2.8250010013580322 + }, + { + "auxiliary_loss_clip": 0.01153644, + "auxiliary_loss_mlp": 0.01032125, + "balance_loss_clip": 1.04350007, + "balance_loss_mlp": 1.02440047, + "epoch": 0.8747670293993868, + "flos": 23985850268160.0, + "grad_norm": 1.8412398080690384, + "language_loss": 0.76726592, + "learning_rate": 1.6215374980979423e-07, + "loss": 0.78912365, + "num_input_tokens_seen": 157397385, + "step": 7275, + "time_per_iteration": 2.8514914512634277 + }, + { + "auxiliary_loss_clip": 0.01157199, + "auxiliary_loss_mlp": 0.01028336, + "balance_loss_clip": 1.04687226, + "balance_loss_mlp": 1.02057588, + "epoch": 0.8748872722900258, + "flos": 45221624478720.0, + "grad_norm": 2.094538961834441, + "language_loss": 0.6855951, + "learning_rate": 1.6184663423140133e-07, + "loss": 0.70745051, + "num_input_tokens_seen": 157417685, + "step": 7276, + "time_per_iteration": 3.030693292617798 + }, + { + "auxiliary_loss_clip": 0.01133507, + "auxiliary_loss_mlp": 0.01030981, + "balance_loss_clip": 1.04293323, + "balance_loss_mlp": 1.02376902, + "epoch": 0.875007515180665, + "flos": 19754280737280.0, + "grad_norm": 2.498940731812475, + "language_loss": 0.639772, + "learning_rate": 1.615397974994126e-07, + "loss": 0.66141689, + "num_input_tokens_seen": 157435490, + "step": 7277, + "time_per_iteration": 2.8729312419891357 + }, + { + "auxiliary_loss_clip": 0.01164338, + "auxiliary_loss_mlp": 0.01028626, + "balance_loss_clip": 1.04542696, + "balance_loss_mlp": 1.02181053, + "epoch": 0.875127758071304, + "flos": 22710734386560.0, + "grad_norm": 1.4894028036842502, + "language_loss": 0.80745494, + "learning_rate": 1.6123323966037438e-07, + "loss": 0.82938457, + "num_input_tokens_seen": 157454010, + "step": 7278, + "time_per_iteration": 2.7835452556610107 + }, + { + "auxiliary_loss_clip": 0.01168649, + "auxiliary_loss_mlp": 0.01029148, + "balance_loss_clip": 1.04848373, + "balance_loss_mlp": 1.02144706, + "epoch": 0.8752480009619431, + "flos": 23403846199680.0, + "grad_norm": 2.020256570691683, + "language_loss": 0.79114795, + "learning_rate": 1.6092696076079216e-07, + "loss": 0.81312591, + "num_input_tokens_seen": 157472385, + "step": 7279, + "time_per_iteration": 2.6860361099243164 + }, + { + "auxiliary_loss_clip": 0.01127785, + "auxiliary_loss_mlp": 0.01022454, + "balance_loss_clip": 1.04133296, + "balance_loss_mlp": 1.01456249, + "epoch": 0.8753682438525822, + "flos": 26213101914240.0, + "grad_norm": 1.8452170860302772, + "language_loss": 0.73895657, + "learning_rate": 1.6062096084712785e-07, + "loss": 0.76045895, + "num_input_tokens_seen": 157493735, + "step": 7280, + "time_per_iteration": 2.851945638656616 + }, + { + "auxiliary_loss_clip": 0.01139754, + "auxiliary_loss_mlp": 0.00886219, + "balance_loss_clip": 1.04063058, + "balance_loss_mlp": 1.00050247, + "epoch": 0.8754884867432213, + "flos": 23326745656320.0, + "grad_norm": 1.751967932885645, + "language_loss": 0.70644748, + "learning_rate": 1.6031523996580098e-07, + "loss": 0.72670722, + "num_input_tokens_seen": 157511295, + "step": 7281, + "time_per_iteration": 3.7781224250793457 + }, + { + "auxiliary_loss_clip": 0.01148375, + "auxiliary_loss_mlp": 0.01032892, + "balance_loss_clip": 1.04292881, + "balance_loss_mlp": 1.02415395, + "epoch": 0.8756087296338604, + "flos": 12495226412160.0, + "grad_norm": 2.3884236240788344, + "language_loss": 0.66240859, + "learning_rate": 1.6000979816318981e-07, + "loss": 0.68422127, + "num_input_tokens_seen": 157529760, + "step": 7282, + "time_per_iteration": 2.7399673461914062 + }, + { + "auxiliary_loss_clip": 0.01156754, + "auxiliary_loss_mlp": 0.01028171, + "balance_loss_clip": 1.04610872, + "balance_loss_mlp": 1.0204227, + "epoch": 0.8757289725244994, + "flos": 18952898353920.0, + "grad_norm": 2.516205821000991, + "language_loss": 0.75198698, + "learning_rate": 1.5970463548562886e-07, + "loss": 0.77383626, + "num_input_tokens_seen": 157548915, + "step": 7283, + "time_per_iteration": 2.7205471992492676 + }, + { + "auxiliary_loss_clip": 0.01147793, + "auxiliary_loss_mlp": 0.01021681, + "balance_loss_clip": 1.04539132, + "balance_loss_mlp": 1.01429284, + "epoch": 0.8758492154151386, + "flos": 25265958140160.0, + "grad_norm": 1.7478710026585105, + "language_loss": 0.71541977, + "learning_rate": 1.5939975197941192e-07, + "loss": 0.73711455, + "num_input_tokens_seen": 157570570, + "step": 7284, + "time_per_iteration": 3.7422454357147217 + }, + { + "auxiliary_loss_clip": 0.01046554, + "auxiliary_loss_mlp": 0.01003581, + "balance_loss_clip": 1.00538433, + "balance_loss_mlp": 1.00211477, + "epoch": 0.8759694583057777, + "flos": 65571664193280.0, + "grad_norm": 0.8627137332631463, + "language_loss": 0.53392988, + "learning_rate": 1.5909514769078892e-07, + "loss": 0.5544312, + "num_input_tokens_seen": 157635675, + "step": 7285, + "time_per_iteration": 4.241755962371826 + }, + { + "auxiliary_loss_clip": 0.01129461, + "auxiliary_loss_mlp": 0.01025928, + "balance_loss_clip": 1.04338336, + "balance_loss_mlp": 1.0182811, + "epoch": 0.8760897011964167, + "flos": 25446193608960.0, + "grad_norm": 1.6179051785823335, + "language_loss": 0.77653426, + "learning_rate": 1.5879082266596867e-07, + "loss": 0.79808819, + "num_input_tokens_seen": 157657015, + "step": 7286, + "time_per_iteration": 2.905348539352417 + }, + { + "auxiliary_loss_clip": 0.01140031, + "auxiliary_loss_mlp": 0.01026881, + "balance_loss_clip": 1.03780198, + "balance_loss_mlp": 1.01906991, + "epoch": 0.8762099440870559, + "flos": 28984830894720.0, + "grad_norm": 1.7651746191349702, + "language_loss": 0.71820378, + "learning_rate": 1.5848677695111645e-07, + "loss": 0.73987287, + "num_input_tokens_seen": 157678615, + "step": 7287, + "time_per_iteration": 2.8121862411499023 + }, + { + "auxiliary_loss_clip": 0.01148594, + "auxiliary_loss_mlp": 0.01039852, + "balance_loss_clip": 1.045573, + "balance_loss_mlp": 1.03114998, + "epoch": 0.8763301869776949, + "flos": 21609461352960.0, + "grad_norm": 2.2583084674077156, + "language_loss": 0.69427383, + "learning_rate": 1.5818301059235562e-07, + "loss": 0.71615827, + "num_input_tokens_seen": 157693790, + "step": 7288, + "time_per_iteration": 2.7687735557556152 + }, + { + "auxiliary_loss_clip": 0.01150331, + "auxiliary_loss_mlp": 0.01026566, + "balance_loss_clip": 1.04581988, + "balance_loss_mlp": 1.01838243, + "epoch": 0.876450429868334, + "flos": 24644416176000.0, + "grad_norm": 2.671011226360607, + "language_loss": 0.81288022, + "learning_rate": 1.578795236357684e-07, + "loss": 0.83464915, + "num_input_tokens_seen": 157715255, + "step": 7289, + "time_per_iteration": 2.7909958362579346 + }, + { + "auxiliary_loss_clip": 0.01149702, + "auxiliary_loss_mlp": 0.01025289, + "balance_loss_clip": 1.0451138, + "balance_loss_mlp": 1.01802289, + "epoch": 0.8765706727589732, + "flos": 20260046188800.0, + "grad_norm": 2.110785109260213, + "language_loss": 0.85758781, + "learning_rate": 1.5757631612739218e-07, + "loss": 0.87933779, + "num_input_tokens_seen": 157728800, + "step": 7290, + "time_per_iteration": 3.6425750255584717 + }, + { + "auxiliary_loss_clip": 0.01055664, + "auxiliary_loss_mlp": 0.01001303, + "balance_loss_clip": 1.00494421, + "balance_loss_mlp": 0.99996758, + "epoch": 0.8766909156496122, + "flos": 71371165276800.0, + "grad_norm": 0.7774644996266439, + "language_loss": 0.61473, + "learning_rate": 1.572733881132242e-07, + "loss": 0.63529968, + "num_input_tokens_seen": 157789445, + "step": 7291, + "time_per_iteration": 3.2871623039245605 + }, + { + "auxiliary_loss_clip": 0.01040282, + "auxiliary_loss_mlp": 0.0100321, + "balance_loss_clip": 1.01241434, + "balance_loss_mlp": 1.00192893, + "epoch": 0.8768111585402513, + "flos": 69523490603520.0, + "grad_norm": 0.7859577445995742, + "language_loss": 0.58489966, + "learning_rate": 1.5697073963921814e-07, + "loss": 0.60533458, + "num_input_tokens_seen": 157848685, + "step": 7292, + "time_per_iteration": 3.246462106704712 + }, + { + "auxiliary_loss_clip": 0.01162957, + "auxiliary_loss_mlp": 0.01029158, + "balance_loss_clip": 1.0479728, + "balance_loss_mlp": 1.02065897, + "epoch": 0.8769314014308904, + "flos": 18838558385280.0, + "grad_norm": 2.7459507193920962, + "language_loss": 0.84904873, + "learning_rate": 1.566683707512857e-07, + "loss": 0.87096989, + "num_input_tokens_seen": 157866360, + "step": 7293, + "time_per_iteration": 2.7218222618103027 + }, + { + "auxiliary_loss_clip": 0.01147376, + "auxiliary_loss_mlp": 0.01028207, + "balance_loss_clip": 1.04351592, + "balance_loss_mlp": 1.02004087, + "epoch": 0.8770516443215295, + "flos": 14976402278400.0, + "grad_norm": 5.090223225594815, + "language_loss": 0.79426849, + "learning_rate": 1.5636628149529553e-07, + "loss": 0.8160243, + "num_input_tokens_seen": 157884150, + "step": 7294, + "time_per_iteration": 2.7982935905456543 + }, + { + "auxiliary_loss_clip": 0.01151372, + "auxiliary_loss_mlp": 0.01027245, + "balance_loss_clip": 1.04443991, + "balance_loss_mlp": 1.01932991, + "epoch": 0.8771718872121685, + "flos": 31649654021760.0, + "grad_norm": 3.217198441031732, + "language_loss": 0.79647768, + "learning_rate": 1.560644719170743e-07, + "loss": 0.81826389, + "num_input_tokens_seen": 157905020, + "step": 7295, + "time_per_iteration": 2.8206419944763184 + }, + { + "auxiliary_loss_clip": 0.01137332, + "auxiliary_loss_mlp": 0.01023156, + "balance_loss_clip": 1.04023457, + "balance_loss_mlp": 1.01514518, + "epoch": 0.8772921301028077, + "flos": 36095466222720.0, + "grad_norm": 1.9846595360766135, + "language_loss": 0.72134173, + "learning_rate": 1.5576294206240692e-07, + "loss": 0.74294657, + "num_input_tokens_seen": 157924545, + "step": 7296, + "time_per_iteration": 2.9025979042053223 + }, + { + "auxiliary_loss_clip": 0.01144427, + "auxiliary_loss_mlp": 0.01025871, + "balance_loss_clip": 1.04469919, + "balance_loss_mlp": 1.01771736, + "epoch": 0.8774123729934468, + "flos": 57116961849600.0, + "grad_norm": 1.5696678804736588, + "language_loss": 0.67555422, + "learning_rate": 1.5546169197703507e-07, + "loss": 0.69725722, + "num_input_tokens_seen": 157950820, + "step": 7297, + "time_per_iteration": 3.0956625938415527 + }, + { + "auxiliary_loss_clip": 0.01152729, + "auxiliary_loss_mlp": 0.01023494, + "balance_loss_clip": 1.04210663, + "balance_loss_mlp": 1.01623702, + "epoch": 0.8775326158840858, + "flos": 23914495900800.0, + "grad_norm": 3.9510262524405095, + "language_loss": 0.77572775, + "learning_rate": 1.5516072170665774e-07, + "loss": 0.79749, + "num_input_tokens_seen": 157968790, + "step": 7298, + "time_per_iteration": 2.7687294483184814 + }, + { + "auxiliary_loss_clip": 0.01160897, + "auxiliary_loss_mlp": 0.01029014, + "balance_loss_clip": 1.04631126, + "balance_loss_mlp": 1.02164364, + "epoch": 0.877652858774725, + "flos": 17123285243520.0, + "grad_norm": 1.8206306020459058, + "language_loss": 0.86659849, + "learning_rate": 1.5486003129693214e-07, + "loss": 0.88849759, + "num_input_tokens_seen": 157986155, + "step": 7299, + "time_per_iteration": 2.7281837463378906 + }, + { + "auxiliary_loss_clip": 0.01160916, + "auxiliary_loss_mlp": 0.01027091, + "balance_loss_clip": 1.04608774, + "balance_loss_mlp": 1.01934564, + "epoch": 0.877773101665364, + "flos": 16508961912960.0, + "grad_norm": 1.959337106492257, + "language_loss": 0.77950746, + "learning_rate": 1.545596207934725e-07, + "loss": 0.80138755, + "num_input_tokens_seen": 158004640, + "step": 7300, + "time_per_iteration": 2.669358730316162 + }, + { + "auxiliary_loss_clip": 0.01144446, + "auxiliary_loss_mlp": 0.01029478, + "balance_loss_clip": 1.04329073, + "balance_loss_mlp": 1.02175951, + "epoch": 0.8778933445560031, + "flos": 22053209973120.0, + "grad_norm": 1.902738703724181, + "language_loss": 0.77637112, + "learning_rate": 1.5425949024185147e-07, + "loss": 0.79811037, + "num_input_tokens_seen": 158024665, + "step": 7301, + "time_per_iteration": 2.758373260498047 + }, + { + "auxiliary_loss_clip": 0.0114851, + "auxiliary_loss_mlp": 0.01026085, + "balance_loss_clip": 1.04136062, + "balance_loss_mlp": 1.01836956, + "epoch": 0.8780135874466423, + "flos": 22564757514240.0, + "grad_norm": 1.7500054262301, + "language_loss": 0.68050021, + "learning_rate": 1.5395963968759818e-07, + "loss": 0.70224613, + "num_input_tokens_seen": 158044940, + "step": 7302, + "time_per_iteration": 2.7278330326080322 + }, + { + "auxiliary_loss_clip": 0.01147691, + "auxiliary_loss_mlp": 0.01030462, + "balance_loss_clip": 1.04150629, + "balance_loss_mlp": 1.02223694, + "epoch": 0.8781338303372813, + "flos": 61531999073280.0, + "grad_norm": 2.146781860331606, + "language_loss": 0.64457929, + "learning_rate": 1.536600691761998e-07, + "loss": 0.6663608, + "num_input_tokens_seen": 158070770, + "step": 7303, + "time_per_iteration": 3.073286771774292 + }, + { + "auxiliary_loss_clip": 0.01144238, + "auxiliary_loss_mlp": 0.01033443, + "balance_loss_clip": 1.04603696, + "balance_loss_mlp": 1.02596283, + "epoch": 0.8782540732279204, + "flos": 22674751937280.0, + "grad_norm": 1.6637499920636407, + "language_loss": 0.71394336, + "learning_rate": 1.5336077875310084e-07, + "loss": 0.73572016, + "num_input_tokens_seen": 158089995, + "step": 7304, + "time_per_iteration": 2.7939419746398926 + }, + { + "auxiliary_loss_clip": 0.0113589, + "auxiliary_loss_mlp": 0.01026162, + "balance_loss_clip": 1.04232991, + "balance_loss_mlp": 1.01862514, + "epoch": 0.8783743161185595, + "flos": 16070348937600.0, + "grad_norm": 2.4919260221046104, + "language_loss": 0.73726118, + "learning_rate": 1.5306176846370321e-07, + "loss": 0.75888169, + "num_input_tokens_seen": 158108140, + "step": 7305, + "time_per_iteration": 2.8527650833129883 + }, + { + "auxiliary_loss_clip": 0.01155939, + "auxiliary_loss_mlp": 0.01026344, + "balance_loss_clip": 1.04269552, + "balance_loss_mlp": 1.01866734, + "epoch": 0.8784945590091986, + "flos": 26067879227520.0, + "grad_norm": 5.8371877006983635, + "language_loss": 0.73971266, + "learning_rate": 1.5276303835336712e-07, + "loss": 0.76153541, + "num_input_tokens_seen": 158128680, + "step": 7306, + "time_per_iteration": 2.844982624053955 + }, + { + "auxiliary_loss_clip": 0.0105182, + "auxiliary_loss_mlp": 0.01001795, + "balance_loss_clip": 1.00507271, + "balance_loss_mlp": 1.0003407, + "epoch": 0.8786148018998376, + "flos": 62720643939840.0, + "grad_norm": 0.7616487494050262, + "language_loss": 0.53524667, + "learning_rate": 1.524645884674094e-07, + "loss": 0.55578279, + "num_input_tokens_seen": 158185610, + "step": 7307, + "time_per_iteration": 3.224883556365967 + }, + { + "auxiliary_loss_clip": 0.01170167, + "auxiliary_loss_mlp": 0.00886637, + "balance_loss_clip": 1.04758012, + "balance_loss_mlp": 1.00047779, + "epoch": 0.8787350447904768, + "flos": 21652734263040.0, + "grad_norm": 2.093199938883683, + "language_loss": 0.78939939, + "learning_rate": 1.521664188511047e-07, + "loss": 0.8099674, + "num_input_tokens_seen": 158205635, + "step": 7308, + "time_per_iteration": 3.663632392883301 + }, + { + "auxiliary_loss_clip": 0.01146258, + "auxiliary_loss_mlp": 0.00886092, + "balance_loss_clip": 1.04476881, + "balance_loss_mlp": 1.00066435, + "epoch": 0.8788552876811159, + "flos": 25478476957440.0, + "grad_norm": 1.9458776170700074, + "language_loss": 0.80215991, + "learning_rate": 1.518685295496851e-07, + "loss": 0.82248342, + "num_input_tokens_seen": 158223495, + "step": 7309, + "time_per_iteration": 2.8756649494171143 + }, + { + "auxiliary_loss_clip": 0.01156827, + "auxiliary_loss_mlp": 0.01027147, + "balance_loss_clip": 1.04233885, + "balance_loss_mlp": 1.01974988, + "epoch": 0.8789755305717549, + "flos": 22310222762880.0, + "grad_norm": 4.0480361407076515, + "language_loss": 0.85339105, + "learning_rate": 1.5157092060833975e-07, + "loss": 0.87523079, + "num_input_tokens_seen": 158243145, + "step": 7310, + "time_per_iteration": 4.742863655090332 + }, + { + "auxiliary_loss_clip": 0.01137699, + "auxiliary_loss_mlp": 0.0102691, + "balance_loss_clip": 1.04161966, + "balance_loss_mlp": 1.01874447, + "epoch": 0.879095773462394, + "flos": 29310971408640.0, + "grad_norm": 1.739683343534417, + "language_loss": 0.65748519, + "learning_rate": 1.5127359207221658e-07, + "loss": 0.67913127, + "num_input_tokens_seen": 158262625, + "step": 7311, + "time_per_iteration": 2.9025909900665283 + }, + { + "auxiliary_loss_clip": 0.01107294, + "auxiliary_loss_mlp": 0.01030927, + "balance_loss_clip": 1.03411961, + "balance_loss_mlp": 1.02233243, + "epoch": 0.8792160163530331, + "flos": 16690023394560.0, + "grad_norm": 1.8228090677504292, + "language_loss": 0.73292553, + "learning_rate": 1.5097654398641923e-07, + "loss": 0.75430781, + "num_input_tokens_seen": 158280530, + "step": 7312, + "time_per_iteration": 2.9101061820983887 + }, + { + "auxiliary_loss_clip": 0.01157471, + "auxiliary_loss_mlp": 0.01031141, + "balance_loss_clip": 1.0479548, + "balance_loss_mlp": 1.02385736, + "epoch": 0.8793362592436722, + "flos": 24499301230080.0, + "grad_norm": 1.2975017868838175, + "language_loss": 0.72971642, + "learning_rate": 1.5067977639601014e-07, + "loss": 0.75160253, + "num_input_tokens_seen": 158303290, + "step": 7313, + "time_per_iteration": 2.8089873790740967 + }, + { + "auxiliary_loss_clip": 0.0114867, + "auxiliary_loss_mlp": 0.01031093, + "balance_loss_clip": 1.0451901, + "balance_loss_mlp": 1.02407789, + "epoch": 0.8794565021343113, + "flos": 14538399834240.0, + "grad_norm": 2.2439529344333224, + "language_loss": 0.70954299, + "learning_rate": 1.5038328934600864e-07, + "loss": 0.73134065, + "num_input_tokens_seen": 158319925, + "step": 7314, + "time_per_iteration": 2.7584965229034424 + }, + { + "auxiliary_loss_clip": 0.01150152, + "auxiliary_loss_mlp": 0.01026576, + "balance_loss_clip": 1.04619992, + "balance_loss_mlp": 1.01936674, + "epoch": 0.8795767450249504, + "flos": 39530286224640.0, + "grad_norm": 2.01904646221237, + "language_loss": 0.69666076, + "learning_rate": 1.5008708288139161e-07, + "loss": 0.71842802, + "num_input_tokens_seen": 158342285, + "step": 7315, + "time_per_iteration": 2.8892836570739746 + }, + { + "auxiliary_loss_clip": 0.01157381, + "auxiliary_loss_mlp": 0.01026651, + "balance_loss_clip": 1.04656494, + "balance_loss_mlp": 1.01813984, + "epoch": 0.8796969879155895, + "flos": 22960672197120.0, + "grad_norm": 1.9387628752107033, + "language_loss": 0.73509014, + "learning_rate": 1.497911570470931e-07, + "loss": 0.75693047, + "num_input_tokens_seen": 158362290, + "step": 7316, + "time_per_iteration": 3.6842851638793945 + }, + { + "auxiliary_loss_clip": 0.01127996, + "auxiliary_loss_mlp": 0.01025576, + "balance_loss_clip": 1.04070413, + "balance_loss_mlp": 1.01775599, + "epoch": 0.8798172308062285, + "flos": 28362427004160.0, + "grad_norm": 1.7458336836480852, + "language_loss": 0.85656714, + "learning_rate": 1.494955118880048e-07, + "loss": 0.8781029, + "num_input_tokens_seen": 158383275, + "step": 7317, + "time_per_iteration": 2.8489296436309814 + }, + { + "auxiliary_loss_clip": 0.01158234, + "auxiliary_loss_mlp": 0.01024214, + "balance_loss_clip": 1.04420233, + "balance_loss_mlp": 1.01676035, + "epoch": 0.8799374736968677, + "flos": 23988974751360.0, + "grad_norm": 1.9679223151648617, + "language_loss": 0.73049808, + "learning_rate": 1.4920014744897634e-07, + "loss": 0.75232255, + "num_input_tokens_seen": 158402690, + "step": 7318, + "time_per_iteration": 2.7177343368530273 + }, + { + "auxiliary_loss_clip": 0.01142987, + "auxiliary_loss_mlp": 0.01029225, + "balance_loss_clip": 1.04376769, + "balance_loss_mlp": 1.02126741, + "epoch": 0.8800577165875068, + "flos": 25630271832960.0, + "grad_norm": 2.179338913455042, + "language_loss": 0.86211032, + "learning_rate": 1.4890506377481392e-07, + "loss": 0.88383245, + "num_input_tokens_seen": 158421780, + "step": 7319, + "time_per_iteration": 2.8579163551330566 + }, + { + "auxiliary_loss_clip": 0.01108727, + "auxiliary_loss_mlp": 0.01026792, + "balance_loss_clip": 1.03951252, + "balance_loss_mlp": 1.0194788, + "epoch": 0.8801779594781458, + "flos": 23440331439360.0, + "grad_norm": 1.5523788648917147, + "language_loss": 0.64029062, + "learning_rate": 1.486102609102815e-07, + "loss": 0.66164577, + "num_input_tokens_seen": 158442330, + "step": 7320, + "time_per_iteration": 2.9012959003448486 + }, + { + "auxiliary_loss_clip": 0.01146534, + "auxiliary_loss_mlp": 0.01024189, + "balance_loss_clip": 1.04396009, + "balance_loss_mlp": 1.01686358, + "epoch": 0.880298202368785, + "flos": 11508580656000.0, + "grad_norm": 4.7768779823742085, + "language_loss": 0.8568086, + "learning_rate": 1.483157389001004e-07, + "loss": 0.87851584, + "num_input_tokens_seen": 158459890, + "step": 7321, + "time_per_iteration": 2.9075469970703125 + }, + { + "auxiliary_loss_clip": 0.01148256, + "auxiliary_loss_mlp": 0.01026515, + "balance_loss_clip": 1.04237807, + "balance_loss_mlp": 1.018296, + "epoch": 0.880418445259424, + "flos": 22671447886080.0, + "grad_norm": 2.317381237086776, + "language_loss": 0.78647888, + "learning_rate": 1.4802149778894933e-07, + "loss": 0.80822653, + "num_input_tokens_seen": 158478680, + "step": 7322, + "time_per_iteration": 2.72540283203125 + }, + { + "auxiliary_loss_clip": 0.01148738, + "auxiliary_loss_mlp": 0.01023118, + "balance_loss_clip": 1.04032421, + "balance_loss_mlp": 1.01532769, + "epoch": 0.8805386881500631, + "flos": 20522158709760.0, + "grad_norm": 1.8633946435143673, + "language_loss": 0.87468255, + "learning_rate": 1.4772753762146484e-07, + "loss": 0.89640111, + "num_input_tokens_seen": 158497935, + "step": 7323, + "time_per_iteration": 2.796736240386963 + }, + { + "auxiliary_loss_clip": 0.01151585, + "auxiliary_loss_mlp": 0.01027412, + "balance_loss_clip": 1.04230988, + "balance_loss_mlp": 1.01933289, + "epoch": 0.8806589310407023, + "flos": 36538891620480.0, + "grad_norm": 1.7011891791980283, + "language_loss": 0.70540726, + "learning_rate": 1.474338584422401e-07, + "loss": 0.72719729, + "num_input_tokens_seen": 158523145, + "step": 7324, + "time_per_iteration": 2.9275546073913574 + }, + { + "auxiliary_loss_clip": 0.01155014, + "auxiliary_loss_mlp": 0.01026654, + "balance_loss_clip": 1.04587483, + "balance_loss_mlp": 1.01904261, + "epoch": 0.8807791739313413, + "flos": 23440187784960.0, + "grad_norm": 1.558573407422698, + "language_loss": 0.75597692, + "learning_rate": 1.4714046029582595e-07, + "loss": 0.77779365, + "num_input_tokens_seen": 158542210, + "step": 7325, + "time_per_iteration": 2.774027109146118 + }, + { + "auxiliary_loss_clip": 0.01143154, + "auxiliary_loss_mlp": 0.01026989, + "balance_loss_clip": 1.04358971, + "balance_loss_mlp": 1.01934767, + "epoch": 0.8808994168219804, + "flos": 25956843310080.0, + "grad_norm": 1.7917755169908909, + "language_loss": 0.75947207, + "learning_rate": 1.46847343226731e-07, + "loss": 0.78117347, + "num_input_tokens_seen": 158563250, + "step": 7326, + "time_per_iteration": 2.9291956424713135 + }, + { + "auxiliary_loss_clip": 0.0116343, + "auxiliary_loss_mlp": 0.01030345, + "balance_loss_clip": 1.04545391, + "balance_loss_mlp": 1.02282286, + "epoch": 0.8810196597126195, + "flos": 17092079303040.0, + "grad_norm": 2.419595464312706, + "language_loss": 0.69488764, + "learning_rate": 1.465545072794203e-07, + "loss": 0.71682549, + "num_input_tokens_seen": 158581125, + "step": 7327, + "time_per_iteration": 2.6949567794799805 + }, + { + "auxiliary_loss_clip": 0.01113507, + "auxiliary_loss_mlp": 0.01021902, + "balance_loss_clip": 1.04061699, + "balance_loss_mlp": 1.01497269, + "epoch": 0.8811399026032586, + "flos": 23002831785600.0, + "grad_norm": 1.5699774391504266, + "language_loss": 0.75610614, + "learning_rate": 1.4626195249831774e-07, + "loss": 0.77746022, + "num_input_tokens_seen": 158602025, + "step": 7328, + "time_per_iteration": 2.9461276531219482 + }, + { + "auxiliary_loss_clip": 0.01155717, + "auxiliary_loss_mlp": 0.01027397, + "balance_loss_clip": 1.0436945, + "balance_loss_mlp": 1.02028072, + "epoch": 0.8812601454938976, + "flos": 14463813242880.0, + "grad_norm": 1.7757106204642832, + "language_loss": 0.71500313, + "learning_rate": 1.4596967892780244e-07, + "loss": 0.73683429, + "num_input_tokens_seen": 158618355, + "step": 7329, + "time_per_iteration": 2.7190959453582764 + }, + { + "auxiliary_loss_clip": 0.01167948, + "auxiliary_loss_mlp": 0.01029909, + "balance_loss_clip": 1.04829311, + "balance_loss_mlp": 1.02236307, + "epoch": 0.8813803883845368, + "flos": 22493223578880.0, + "grad_norm": 1.6266816327960147, + "language_loss": 0.74694121, + "learning_rate": 1.4567768661221314e-07, + "loss": 0.76891977, + "num_input_tokens_seen": 158638925, + "step": 7330, + "time_per_iteration": 2.7386951446533203 + }, + { + "auxiliary_loss_clip": 0.01163601, + "auxiliary_loss_mlp": 0.00886179, + "balance_loss_clip": 1.0466249, + "balance_loss_mlp": 1.00060201, + "epoch": 0.8815006312751759, + "flos": 21506901045120.0, + "grad_norm": 2.186375902741381, + "language_loss": 0.74410903, + "learning_rate": 1.4538597559584442e-07, + "loss": 0.76460695, + "num_input_tokens_seen": 158656715, + "step": 7331, + "time_per_iteration": 2.731515884399414 + }, + { + "auxiliary_loss_clip": 0.01144727, + "auxiliary_loss_mlp": 0.01033036, + "balance_loss_clip": 1.04204464, + "balance_loss_mlp": 1.02512074, + "epoch": 0.8816208741658149, + "flos": 22784566792320.0, + "grad_norm": 1.7698039865463648, + "language_loss": 0.78717983, + "learning_rate": 1.4509454592294823e-07, + "loss": 0.80895746, + "num_input_tokens_seen": 158677200, + "step": 7332, + "time_per_iteration": 2.7621796131134033 + }, + { + "auxiliary_loss_clip": 0.01138679, + "auxiliary_loss_mlp": 0.00885666, + "balance_loss_clip": 1.04361272, + "balance_loss_mlp": 1.0006268, + "epoch": 0.8817411170564541, + "flos": 17779409026560.0, + "grad_norm": 1.9454973338436201, + "language_loss": 0.78836179, + "learning_rate": 1.448033976377354e-07, + "loss": 0.80860519, + "num_input_tokens_seen": 158692185, + "step": 7333, + "time_per_iteration": 3.7998268604278564 + }, + { + "auxiliary_loss_clip": 0.01160692, + "auxiliary_loss_mlp": 0.0102241, + "balance_loss_clip": 1.04396129, + "balance_loss_mlp": 1.01508451, + "epoch": 0.8818613599470931, + "flos": 18551812112640.0, + "grad_norm": 1.7407800150915582, + "language_loss": 0.74008083, + "learning_rate": 1.445125307843713e-07, + "loss": 0.76191187, + "num_input_tokens_seen": 158710410, + "step": 7334, + "time_per_iteration": 2.7748377323150635 + }, + { + "auxiliary_loss_clip": 0.01156187, + "auxiliary_loss_mlp": 0.01026558, + "balance_loss_clip": 1.04584813, + "balance_loss_mlp": 1.01940894, + "epoch": 0.8819816028377322, + "flos": 27599792417280.0, + "grad_norm": 1.6114949209113794, + "language_loss": 0.75728106, + "learning_rate": 1.442219454069813e-07, + "loss": 0.77910846, + "num_input_tokens_seen": 158731435, + "step": 7335, + "time_per_iteration": 2.8097341060638428 + }, + { + "auxiliary_loss_clip": 0.01133357, + "auxiliary_loss_mlp": 0.01030091, + "balance_loss_clip": 1.04058814, + "balance_loss_mlp": 1.02242589, + "epoch": 0.8821018457283714, + "flos": 23404600385280.0, + "grad_norm": 2.0802699834201706, + "language_loss": 0.66430485, + "learning_rate": 1.4393164154964676e-07, + "loss": 0.68593937, + "num_input_tokens_seen": 158750965, + "step": 7336, + "time_per_iteration": 3.900573253631592 + }, + { + "auxiliary_loss_clip": 0.01155837, + "auxiliary_loss_mlp": 0.0102741, + "balance_loss_clip": 1.04670763, + "balance_loss_mlp": 1.0193162, + "epoch": 0.8822220886190104, + "flos": 29132459792640.0, + "grad_norm": 1.7378103236595066, + "language_loss": 0.94189143, + "learning_rate": 1.4364161925640649e-07, + "loss": 0.96372396, + "num_input_tokens_seen": 158772365, + "step": 7337, + "time_per_iteration": 3.7647247314453125 + }, + { + "auxiliary_loss_clip": 0.01166163, + "auxiliary_loss_mlp": 0.01026154, + "balance_loss_clip": 1.04580998, + "balance_loss_mlp": 1.01912665, + "epoch": 0.8823423315096495, + "flos": 20485422074880.0, + "grad_norm": 1.851359869395952, + "language_loss": 0.84939849, + "learning_rate": 1.4335187857125663e-07, + "loss": 0.87132168, + "num_input_tokens_seen": 158791065, + "step": 7338, + "time_per_iteration": 2.8426694869995117 + }, + { + "auxiliary_loss_clip": 0.01161283, + "auxiliary_loss_mlp": 0.01031479, + "balance_loss_clip": 1.04560947, + "balance_loss_mlp": 1.02386725, + "epoch": 0.8824625744002886, + "flos": 24206377818240.0, + "grad_norm": 1.6305811794784923, + "language_loss": 0.7501148, + "learning_rate": 1.4306241953815023e-07, + "loss": 0.77204239, + "num_input_tokens_seen": 158812125, + "step": 7339, + "time_per_iteration": 2.8599650859832764 + }, + { + "auxiliary_loss_clip": 0.01160643, + "auxiliary_loss_mlp": 0.01026013, + "balance_loss_clip": 1.04607415, + "balance_loss_mlp": 1.01803184, + "epoch": 0.8825828172909277, + "flos": 24679500785280.0, + "grad_norm": 2.808390870065563, + "language_loss": 0.71249056, + "learning_rate": 1.4277324220099862e-07, + "loss": 0.73435712, + "num_input_tokens_seen": 158834035, + "step": 7340, + "time_per_iteration": 3.129459857940674 + }, + { + "auxiliary_loss_clip": 0.01136407, + "auxiliary_loss_mlp": 0.01025594, + "balance_loss_clip": 1.04015458, + "balance_loss_mlp": 1.01808381, + "epoch": 0.8827030601815667, + "flos": 22456163721600.0, + "grad_norm": 2.0355343181125534, + "language_loss": 0.74051917, + "learning_rate": 1.4248434660366938e-07, + "loss": 0.76213914, + "num_input_tokens_seen": 158853510, + "step": 7341, + "time_per_iteration": 2.951280117034912 + }, + { + "auxiliary_loss_clip": 0.01147488, + "auxiliary_loss_mlp": 0.01026208, + "balance_loss_clip": 1.04402518, + "balance_loss_mlp": 1.01866472, + "epoch": 0.8828233030722058, + "flos": 19865639877120.0, + "grad_norm": 3.7731584156650584, + "language_loss": 0.70469797, + "learning_rate": 1.4219573278998808e-07, + "loss": 0.72643495, + "num_input_tokens_seen": 158871970, + "step": 7342, + "time_per_iteration": 3.7721664905548096 + }, + { + "auxiliary_loss_clip": 0.0114581, + "auxiliary_loss_mlp": 0.01028613, + "balance_loss_clip": 1.04077816, + "balance_loss_mlp": 1.02020919, + "epoch": 0.882943545962845, + "flos": 39347213581440.0, + "grad_norm": 2.367706951158691, + "language_loss": 0.65005565, + "learning_rate": 1.4190740080373685e-07, + "loss": 0.6717999, + "num_input_tokens_seen": 158892250, + "step": 7343, + "time_per_iteration": 2.858978509902954 + }, + { + "auxiliary_loss_clip": 0.01129508, + "auxiliary_loss_mlp": 0.01027689, + "balance_loss_clip": 1.04333329, + "balance_loss_mlp": 1.01952338, + "epoch": 0.883063788853484, + "flos": 19054524908160.0, + "grad_norm": 2.6102344089943985, + "language_loss": 0.83913529, + "learning_rate": 1.4161935068865538e-07, + "loss": 0.86070728, + "num_input_tokens_seen": 158907395, + "step": 7344, + "time_per_iteration": 2.89105224609375 + }, + { + "auxiliary_loss_clip": 0.01166639, + "auxiliary_loss_mlp": 0.0102801, + "balance_loss_clip": 1.04541135, + "balance_loss_mlp": 1.02032137, + "epoch": 0.8831840317441231, + "flos": 18733196816640.0, + "grad_norm": 2.162838215465626, + "language_loss": 0.75805229, + "learning_rate": 1.4133158248844113e-07, + "loss": 0.77999878, + "num_input_tokens_seen": 158926300, + "step": 7345, + "time_per_iteration": 2.591273307800293 + }, + { + "auxiliary_loss_clip": 0.01133864, + "auxiliary_loss_mlp": 0.01025857, + "balance_loss_clip": 1.0425446, + "balance_loss_mlp": 1.01795077, + "epoch": 0.8833042746347622, + "flos": 26827712553600.0, + "grad_norm": 1.9035687994020516, + "language_loss": 0.73312569, + "learning_rate": 1.4104409624674785e-07, + "loss": 0.75472295, + "num_input_tokens_seen": 158946085, + "step": 7346, + "time_per_iteration": 2.874427080154419 + }, + { + "auxiliary_loss_clip": 0.01159866, + "auxiliary_loss_mlp": 0.01028262, + "balance_loss_clip": 1.04778504, + "balance_loss_mlp": 1.0210557, + "epoch": 0.8834245175254013, + "flos": 26104077158400.0, + "grad_norm": 1.9125395290760496, + "language_loss": 0.78353083, + "learning_rate": 1.407568920071873e-07, + "loss": 0.80541205, + "num_input_tokens_seen": 158964950, + "step": 7347, + "time_per_iteration": 2.7197518348693848 + }, + { + "auxiliary_loss_clip": 0.01175111, + "auxiliary_loss_mlp": 0.01029559, + "balance_loss_clip": 1.04910755, + "balance_loss_mlp": 1.02132177, + "epoch": 0.8835447604160404, + "flos": 30629036977920.0, + "grad_norm": 1.910657648218946, + "language_loss": 0.68418217, + "learning_rate": 1.4046996981332782e-07, + "loss": 0.70622891, + "num_input_tokens_seen": 158984835, + "step": 7348, + "time_per_iteration": 2.7225468158721924 + }, + { + "auxiliary_loss_clip": 0.01132731, + "auxiliary_loss_mlp": 0.01027993, + "balance_loss_clip": 1.04275537, + "balance_loss_mlp": 1.02015519, + "epoch": 0.8836650033066795, + "flos": 24718356322560.0, + "grad_norm": 1.9499552468479917, + "language_loss": 0.78250921, + "learning_rate": 1.4018332970869516e-07, + "loss": 0.80411643, + "num_input_tokens_seen": 159002775, + "step": 7349, + "time_per_iteration": 2.8074655532836914 + }, + { + "auxiliary_loss_clip": 0.01142181, + "auxiliary_loss_mlp": 0.01030318, + "balance_loss_clip": 1.04426455, + "balance_loss_mlp": 1.02280164, + "epoch": 0.8837852461973186, + "flos": 25413371556480.0, + "grad_norm": 1.799933482906562, + "language_loss": 0.85144413, + "learning_rate": 1.398969717367733e-07, + "loss": 0.87316912, + "num_input_tokens_seen": 159024100, + "step": 7350, + "time_per_iteration": 2.7757813930511475 + }, + { + "auxiliary_loss_clip": 0.01120973, + "auxiliary_loss_mlp": 0.01022894, + "balance_loss_clip": 1.04474163, + "balance_loss_mlp": 1.01577771, + "epoch": 0.8839054890879576, + "flos": 17822574195840.0, + "grad_norm": 1.6884009094409125, + "language_loss": 0.76499712, + "learning_rate": 1.396108959410014e-07, + "loss": 0.78643578, + "num_input_tokens_seen": 159043315, + "step": 7351, + "time_per_iteration": 2.8668618202209473 + }, + { + "auxiliary_loss_clip": 0.01158496, + "auxiliary_loss_mlp": 0.00885834, + "balance_loss_clip": 1.04768062, + "balance_loss_mlp": 1.00060856, + "epoch": 0.8840257319785968, + "flos": 23769021818880.0, + "grad_norm": 1.844113770001648, + "language_loss": 0.81510532, + "learning_rate": 1.3932510236477745e-07, + "loss": 0.8355487, + "num_input_tokens_seen": 159063985, + "step": 7352, + "time_per_iteration": 2.7411692142486572 + }, + { + "auxiliary_loss_clip": 0.01155909, + "auxiliary_loss_mlp": 0.01024151, + "balance_loss_clip": 1.04233718, + "balance_loss_mlp": 1.01554418, + "epoch": 0.8841459748692359, + "flos": 29059776622080.0, + "grad_norm": 1.720115875092806, + "language_loss": 0.5578804, + "learning_rate": 1.3903959105145636e-07, + "loss": 0.57968092, + "num_input_tokens_seen": 159084475, + "step": 7353, + "time_per_iteration": 2.7427353858947754 + }, + { + "auxiliary_loss_clip": 0.01168647, + "auxiliary_loss_mlp": 0.01027399, + "balance_loss_clip": 1.04780078, + "balance_loss_mlp": 1.0196979, + "epoch": 0.8842662177598749, + "flos": 24311523905280.0, + "grad_norm": 2.201342112768662, + "language_loss": 0.83274651, + "learning_rate": 1.387543620443492e-07, + "loss": 0.854707, + "num_input_tokens_seen": 159101320, + "step": 7354, + "time_per_iteration": 2.6495285034179688 + }, + { + "auxiliary_loss_clip": 0.01167373, + "auxiliary_loss_mlp": 0.01029608, + "balance_loss_clip": 1.04738712, + "balance_loss_mlp": 1.02247906, + "epoch": 0.8843864606505141, + "flos": 25007867942400.0, + "grad_norm": 2.119302815210363, + "language_loss": 0.84266049, + "learning_rate": 1.3846941538672606e-07, + "loss": 0.86463034, + "num_input_tokens_seen": 159120025, + "step": 7355, + "time_per_iteration": 2.8096418380737305 + }, + { + "auxiliary_loss_clip": 0.01134258, + "auxiliary_loss_mlp": 0.01025271, + "balance_loss_clip": 1.04359651, + "balance_loss_mlp": 1.0176357, + "epoch": 0.8845067035411531, + "flos": 28183915388160.0, + "grad_norm": 2.3902345852105342, + "language_loss": 0.80692613, + "learning_rate": 1.3818475112181193e-07, + "loss": 0.82852137, + "num_input_tokens_seen": 159138820, + "step": 7356, + "time_per_iteration": 3.007692575454712 + }, + { + "auxiliary_loss_clip": 0.01139398, + "auxiliary_loss_mlp": 0.01025592, + "balance_loss_clip": 1.04378617, + "balance_loss_mlp": 1.01840711, + "epoch": 0.8846269464317922, + "flos": 12853219311360.0, + "grad_norm": 2.109168374867887, + "language_loss": 0.79616284, + "learning_rate": 1.3790036929279091e-07, + "loss": 0.8178128, + "num_input_tokens_seen": 159155975, + "step": 7357, + "time_per_iteration": 2.843398332595825 + }, + { + "auxiliary_loss_clip": 0.01160347, + "auxiliary_loss_mlp": 0.0088506, + "balance_loss_clip": 1.04687405, + "balance_loss_mlp": 1.00067282, + "epoch": 0.8847471893224313, + "flos": 18624351628800.0, + "grad_norm": 2.8218321658326517, + "language_loss": 0.58799171, + "learning_rate": 1.3761626994280363e-07, + "loss": 0.60844582, + "num_input_tokens_seen": 159173445, + "step": 7358, + "time_per_iteration": 2.7919206619262695 + }, + { + "auxiliary_loss_clip": 0.01141949, + "auxiliary_loss_mlp": 0.01025622, + "balance_loss_clip": 1.04300046, + "balance_loss_mlp": 1.01801062, + "epoch": 0.8848674322130704, + "flos": 35769433449600.0, + "grad_norm": 1.6711987767409116, + "language_loss": 0.73780966, + "learning_rate": 1.3733245311494735e-07, + "loss": 0.75948536, + "num_input_tokens_seen": 159196100, + "step": 7359, + "time_per_iteration": 3.795475482940674 + }, + { + "auxiliary_loss_clip": 0.01160788, + "auxiliary_loss_mlp": 0.01024207, + "balance_loss_clip": 1.0471797, + "balance_loss_mlp": 1.01628518, + "epoch": 0.8849876751037095, + "flos": 24243760897920.0, + "grad_norm": 2.2960731452652445, + "language_loss": 0.70531404, + "learning_rate": 1.3704891885227676e-07, + "loss": 0.72716397, + "num_input_tokens_seen": 159216145, + "step": 7360, + "time_per_iteration": 2.768127679824829 + }, + { + "auxiliary_loss_clip": 0.01139012, + "auxiliary_loss_mlp": 0.01026337, + "balance_loss_clip": 1.03954649, + "balance_loss_mlp": 1.01821876, + "epoch": 0.8851079179943486, + "flos": 21500580251520.0, + "grad_norm": 2.0155549342307895, + "language_loss": 0.77600861, + "learning_rate": 1.367656671978037e-07, + "loss": 0.79766214, + "num_input_tokens_seen": 159233610, + "step": 7361, + "time_per_iteration": 2.7578580379486084 + }, + { + "auxiliary_loss_clip": 0.01155466, + "auxiliary_loss_mlp": 0.01020303, + "balance_loss_clip": 1.04373932, + "balance_loss_mlp": 1.01326966, + "epoch": 0.8852281608849877, + "flos": 15300711198720.0, + "grad_norm": 1.8134301461999087, + "language_loss": 0.73442733, + "learning_rate": 1.36482698194498e-07, + "loss": 0.75618505, + "num_input_tokens_seen": 159250155, + "step": 7362, + "time_per_iteration": 3.661642074584961 + }, + { + "auxiliary_loss_clip": 0.01149015, + "auxiliary_loss_mlp": 0.01025277, + "balance_loss_clip": 1.04243731, + "balance_loss_mlp": 1.01671743, + "epoch": 0.8853484037756267, + "flos": 23295719283840.0, + "grad_norm": 1.895019485786073, + "language_loss": 0.72028154, + "learning_rate": 1.3620001188528506e-07, + "loss": 0.74202442, + "num_input_tokens_seen": 159270875, + "step": 7363, + "time_per_iteration": 3.714170455932617 + }, + { + "auxiliary_loss_clip": 0.01161661, + "auxiliary_loss_mlp": 0.01029025, + "balance_loss_clip": 1.04459345, + "balance_loss_mlp": 1.02081132, + "epoch": 0.8854686466662659, + "flos": 25114773795840.0, + "grad_norm": 2.414727769187321, + "language_loss": 0.73564619, + "learning_rate": 1.3591760831304865e-07, + "loss": 0.75755304, + "num_input_tokens_seen": 159288565, + "step": 7364, + "time_per_iteration": 2.9223921298980713 + }, + { + "auxiliary_loss_clip": 0.01168017, + "auxiliary_loss_mlp": 0.0103159, + "balance_loss_clip": 1.04715419, + "balance_loss_mlp": 1.02387726, + "epoch": 0.885588889556905, + "flos": 21390873137280.0, + "grad_norm": 1.695754333873571, + "language_loss": 0.79446876, + "learning_rate": 1.356354875206287e-07, + "loss": 0.81646484, + "num_input_tokens_seen": 159306400, + "step": 7365, + "time_per_iteration": 2.710103750228882 + }, + { + "auxiliary_loss_clip": 0.01137987, + "auxiliary_loss_mlp": 0.01025554, + "balance_loss_clip": 1.0436821, + "balance_loss_mlp": 1.01854193, + "epoch": 0.885709132447544, + "flos": 26906752431360.0, + "grad_norm": 2.069353469288387, + "language_loss": 0.69574964, + "learning_rate": 1.3535364955082296e-07, + "loss": 0.71738505, + "num_input_tokens_seen": 159326250, + "step": 7366, + "time_per_iteration": 2.9073398113250732 + }, + { + "auxiliary_loss_clip": 0.01165497, + "auxiliary_loss_mlp": 0.01027875, + "balance_loss_clip": 1.04602647, + "balance_loss_mlp": 1.02048731, + "epoch": 0.8858293753381832, + "flos": 26103394800000.0, + "grad_norm": 1.7031583919197213, + "language_loss": 0.64601016, + "learning_rate": 1.3507209444638613e-07, + "loss": 0.66794384, + "num_input_tokens_seen": 159348250, + "step": 7367, + "time_per_iteration": 3.9990808963775635 + }, + { + "auxiliary_loss_clip": 0.01161298, + "auxiliary_loss_mlp": 0.01022091, + "balance_loss_clip": 1.04726505, + "balance_loss_mlp": 1.01455379, + "epoch": 0.8859496182288222, + "flos": 23292810282240.0, + "grad_norm": 2.714158043831515, + "language_loss": 0.74173939, + "learning_rate": 1.347908222500298e-07, + "loss": 0.76357329, + "num_input_tokens_seen": 159368325, + "step": 7368, + "time_per_iteration": 2.7559046745300293 + }, + { + "auxiliary_loss_clip": 0.01124787, + "auxiliary_loss_mlp": 0.01028241, + "balance_loss_clip": 1.04184818, + "balance_loss_mlp": 1.0205698, + "epoch": 0.8860698611194613, + "flos": 16872916469760.0, + "grad_norm": 1.7939125714664312, + "language_loss": 0.69728303, + "learning_rate": 1.3450983300442276e-07, + "loss": 0.71881324, + "num_input_tokens_seen": 159387555, + "step": 7369, + "time_per_iteration": 2.8744921684265137 + }, + { + "auxiliary_loss_clip": 0.0116007, + "auxiliary_loss_mlp": 0.0102575, + "balance_loss_clip": 1.04564679, + "balance_loss_mlp": 1.01759636, + "epoch": 0.8861901040101005, + "flos": 24681404206080.0, + "grad_norm": 1.9734714588207667, + "language_loss": 0.73309374, + "learning_rate": 1.3422912675219068e-07, + "loss": 0.75495195, + "num_input_tokens_seen": 159407310, + "step": 7370, + "time_per_iteration": 2.9875681400299072 + }, + { + "auxiliary_loss_clip": 0.01164145, + "auxiliary_loss_mlp": 0.01027584, + "balance_loss_clip": 1.04588723, + "balance_loss_mlp": 1.0201571, + "epoch": 0.8863103469007395, + "flos": 24423026699520.0, + "grad_norm": 2.0639562889071756, + "language_loss": 0.78917956, + "learning_rate": 1.339487035359166e-07, + "loss": 0.81109691, + "num_input_tokens_seen": 159427680, + "step": 7371, + "time_per_iteration": 2.8221535682678223 + }, + { + "auxiliary_loss_clip": 0.01151279, + "auxiliary_loss_mlp": 0.00885178, + "balance_loss_clip": 1.0461967, + "balance_loss_mlp": 1.00059462, + "epoch": 0.8864305897913786, + "flos": 22053964158720.0, + "grad_norm": 1.5079253310038998, + "language_loss": 0.84927535, + "learning_rate": 1.336685633981409e-07, + "loss": 0.86963993, + "num_input_tokens_seen": 159448765, + "step": 7372, + "time_per_iteration": 2.821773052215576 + }, + { + "auxiliary_loss_clip": 0.01159632, + "auxiliary_loss_mlp": 0.01031337, + "balance_loss_clip": 1.04547417, + "balance_loss_mlp": 1.02307594, + "epoch": 0.8865508326820177, + "flos": 19099449843840.0, + "grad_norm": 1.718855370053027, + "language_loss": 0.75127184, + "learning_rate": 1.333887063813597e-07, + "loss": 0.7731815, + "num_input_tokens_seen": 159466870, + "step": 7373, + "time_per_iteration": 2.8180744647979736 + }, + { + "auxiliary_loss_clip": 0.01149966, + "auxiliary_loss_mlp": 0.01030115, + "balance_loss_clip": 1.04213047, + "balance_loss_mlp": 1.02277732, + "epoch": 0.8866710755726568, + "flos": 15414189240960.0, + "grad_norm": 1.7193002962378006, + "language_loss": 0.65991825, + "learning_rate": 1.331091325280278e-07, + "loss": 0.68171906, + "num_input_tokens_seen": 159485840, + "step": 7374, + "time_per_iteration": 2.7948436737060547 + }, + { + "auxiliary_loss_clip": 0.01121259, + "auxiliary_loss_mlp": 0.01029293, + "balance_loss_clip": 1.03920889, + "balance_loss_mlp": 1.02126396, + "epoch": 0.8867913184632958, + "flos": 20083689388800.0, + "grad_norm": 2.0654051962144706, + "language_loss": 0.78422689, + "learning_rate": 1.3282984188055625e-07, + "loss": 0.80573243, + "num_input_tokens_seen": 159505630, + "step": 7375, + "time_per_iteration": 2.9039955139160156 + }, + { + "auxiliary_loss_clip": 0.01167397, + "auxiliary_loss_mlp": 0.0102805, + "balance_loss_clip": 1.04622006, + "balance_loss_mlp": 1.02048373, + "epoch": 0.8869115613539349, + "flos": 23365852588800.0, + "grad_norm": 2.246395935124186, + "language_loss": 0.79758114, + "learning_rate": 1.3255083448131288e-07, + "loss": 0.81953561, + "num_input_tokens_seen": 159524675, + "step": 7376, + "time_per_iteration": 2.6398704051971436 + }, + { + "auxiliary_loss_clip": 0.0116163, + "auxiliary_loss_mlp": 0.01027573, + "balance_loss_clip": 1.04426706, + "balance_loss_mlp": 1.01940739, + "epoch": 0.8870318042445741, + "flos": 21286840371840.0, + "grad_norm": 2.076942024484867, + "language_loss": 0.78687668, + "learning_rate": 1.3227211037262365e-07, + "loss": 0.80876869, + "num_input_tokens_seen": 159541915, + "step": 7377, + "time_per_iteration": 2.6912474632263184 + }, + { + "auxiliary_loss_clip": 0.01125253, + "auxiliary_loss_mlp": 0.01032296, + "balance_loss_clip": 1.04044151, + "balance_loss_mlp": 1.02379382, + "epoch": 0.8871520471352131, + "flos": 20010862563840.0, + "grad_norm": 2.21921570948943, + "language_loss": 0.85393322, + "learning_rate": 1.319936695967696e-07, + "loss": 0.87550873, + "num_input_tokens_seen": 159559740, + "step": 7378, + "time_per_iteration": 2.980977773666382 + }, + { + "auxiliary_loss_clip": 0.01176546, + "auxiliary_loss_mlp": 0.01026843, + "balance_loss_clip": 1.04967093, + "balance_loss_mlp": 1.01856422, + "epoch": 0.8872722900258522, + "flos": 22601422321920.0, + "grad_norm": 2.5279517600271, + "language_loss": 0.82270616, + "learning_rate": 1.3171551219599097e-07, + "loss": 0.84474009, + "num_input_tokens_seen": 159578265, + "step": 7379, + "time_per_iteration": 2.709876298904419 + }, + { + "auxiliary_loss_clip": 0.0116932, + "auxiliary_loss_mlp": 0.0103151, + "balance_loss_clip": 1.04841292, + "balance_loss_mlp": 1.02323663, + "epoch": 0.8873925329164913, + "flos": 22163276223360.0, + "grad_norm": 3.035552704058582, + "language_loss": 0.78223526, + "learning_rate": 1.3143763821248377e-07, + "loss": 0.80424356, + "num_input_tokens_seen": 159595350, + "step": 7380, + "time_per_iteration": 2.6889209747314453 + }, + { + "auxiliary_loss_clip": 0.01164703, + "auxiliary_loss_mlp": 0.01029027, + "balance_loss_clip": 1.04482675, + "balance_loss_mlp": 1.02198768, + "epoch": 0.8875127758071304, + "flos": 19208223204480.0, + "grad_norm": 1.7625315910546995, + "language_loss": 0.72136265, + "learning_rate": 1.3116004768840118e-07, + "loss": 0.7432999, + "num_input_tokens_seen": 159613725, + "step": 7381, + "time_per_iteration": 2.628701686859131 + }, + { + "auxiliary_loss_clip": 0.01171683, + "auxiliary_loss_mlp": 0.01029352, + "balance_loss_clip": 1.04891694, + "balance_loss_mlp": 1.02180624, + "epoch": 0.8876330186977694, + "flos": 18110900666880.0, + "grad_norm": 1.6425278928071365, + "language_loss": 0.74086082, + "learning_rate": 1.3088274066585348e-07, + "loss": 0.76287121, + "num_input_tokens_seen": 159631335, + "step": 7382, + "time_per_iteration": 2.743115186691284 + }, + { + "auxiliary_loss_clip": 0.01141962, + "auxiliary_loss_mlp": 0.01024738, + "balance_loss_clip": 1.04054773, + "balance_loss_mlp": 1.01695347, + "epoch": 0.8877532615884086, + "flos": 22009434272640.0, + "grad_norm": 2.1008957578603056, + "language_loss": 0.90359747, + "learning_rate": 1.3060571718690749e-07, + "loss": 0.92526454, + "num_input_tokens_seen": 159648830, + "step": 7383, + "time_per_iteration": 2.764253854751587 + }, + { + "auxiliary_loss_clip": 0.01041332, + "auxiliary_loss_mlp": 0.00876702, + "balance_loss_clip": 1.00487137, + "balance_loss_mlp": 1.00119734, + "epoch": 0.8878735044790477, + "flos": 72136924346880.0, + "grad_norm": 0.7455687761787294, + "language_loss": 0.56911945, + "learning_rate": 1.3032897729358805e-07, + "loss": 0.58829975, + "num_input_tokens_seen": 159709785, + "step": 7384, + "time_per_iteration": 3.3608932495117188 + }, + { + "auxiliary_loss_clip": 0.0111224, + "auxiliary_loss_mlp": 0.00886279, + "balance_loss_clip": 1.03765488, + "balance_loss_mlp": 1.00055146, + "epoch": 0.8879937473696867, + "flos": 27526355061120.0, + "grad_norm": 2.3315888488825935, + "language_loss": 0.80106336, + "learning_rate": 1.3005252102787645e-07, + "loss": 0.82104856, + "num_input_tokens_seen": 159728725, + "step": 7385, + "time_per_iteration": 3.868485927581787 + }, + { + "auxiliary_loss_clip": 0.01163267, + "auxiliary_loss_mlp": 0.0102788, + "balance_loss_clip": 1.04760134, + "balance_loss_mlp": 1.02031589, + "epoch": 0.8881139902603259, + "flos": 22234091886720.0, + "grad_norm": 1.5153910859316395, + "language_loss": 0.73221666, + "learning_rate": 1.297763484317105e-07, + "loss": 0.75412816, + "num_input_tokens_seen": 159747020, + "step": 7386, + "time_per_iteration": 2.7326574325561523 + }, + { + "auxiliary_loss_clip": 0.01119799, + "auxiliary_loss_mlp": 0.00886024, + "balance_loss_clip": 1.03955698, + "balance_loss_mlp": 1.00062895, + "epoch": 0.888234233150965, + "flos": 20299548170880.0, + "grad_norm": 2.1420049297299295, + "language_loss": 0.702618, + "learning_rate": 1.2950045954698551e-07, + "loss": 0.72267628, + "num_input_tokens_seen": 159764855, + "step": 7387, + "time_per_iteration": 2.837002754211426 + }, + { + "auxiliary_loss_clip": 0.01129677, + "auxiliary_loss_mlp": 0.01025489, + "balance_loss_clip": 1.04097772, + "balance_loss_mlp": 1.01798439, + "epoch": 0.888354476041604, + "flos": 18147996437760.0, + "grad_norm": 2.0851686502676987, + "language_loss": 0.75440419, + "learning_rate": 1.2922485441555343e-07, + "loss": 0.77595592, + "num_input_tokens_seen": 159783935, + "step": 7388, + "time_per_iteration": 3.759586811065674 + }, + { + "auxiliary_loss_clip": 0.01167197, + "auxiliary_loss_mlp": 0.0102728, + "balance_loss_clip": 1.04619718, + "balance_loss_mlp": 1.01971006, + "epoch": 0.8884747189322432, + "flos": 22014282608640.0, + "grad_norm": 1.8444835078868036, + "language_loss": 0.81762147, + "learning_rate": 1.2894953307922363e-07, + "loss": 0.83956623, + "num_input_tokens_seen": 159802895, + "step": 7389, + "time_per_iteration": 2.699246406555176 + }, + { + "auxiliary_loss_clip": 0.01133493, + "auxiliary_loss_mlp": 0.01029039, + "balance_loss_clip": 1.04231143, + "balance_loss_mlp": 1.02107549, + "epoch": 0.8885949618228822, + "flos": 19786779567360.0, + "grad_norm": 1.855190024960081, + "language_loss": 0.84040004, + "learning_rate": 1.2867449557976208e-07, + "loss": 0.86202538, + "num_input_tokens_seen": 159820995, + "step": 7390, + "time_per_iteration": 3.7441277503967285 + }, + { + "auxiliary_loss_clip": 0.01157057, + "auxiliary_loss_mlp": 0.01030412, + "balance_loss_clip": 1.04741883, + "balance_loss_mlp": 1.02331924, + "epoch": 0.8887152047135213, + "flos": 20047599198720.0, + "grad_norm": 1.8623845532823966, + "language_loss": 0.75231659, + "learning_rate": 1.283997419588916e-07, + "loss": 0.77419138, + "num_input_tokens_seen": 159840465, + "step": 7391, + "time_per_iteration": 2.6428070068359375 + }, + { + "auxiliary_loss_clip": 0.01159778, + "auxiliary_loss_mlp": 0.01026387, + "balance_loss_clip": 1.04319525, + "balance_loss_mlp": 1.01893306, + "epoch": 0.8888354476041604, + "flos": 18588117784320.0, + "grad_norm": 2.5001105160646078, + "language_loss": 0.61961997, + "learning_rate": 1.2812527225829216e-07, + "loss": 0.64148164, + "num_input_tokens_seen": 159858690, + "step": 7392, + "time_per_iteration": 2.6477317810058594 + }, + { + "auxiliary_loss_clip": 0.01164956, + "auxiliary_loss_mlp": 0.0102784, + "balance_loss_clip": 1.04728711, + "balance_loss_mlp": 1.01970375, + "epoch": 0.8889556904947995, + "flos": 21689794120320.0, + "grad_norm": 2.7156440281946566, + "language_loss": 0.76351726, + "learning_rate": 1.2785108651960052e-07, + "loss": 0.78544521, + "num_input_tokens_seen": 159880325, + "step": 7393, + "time_per_iteration": 3.7028727531433105 + }, + { + "auxiliary_loss_clip": 0.01161576, + "auxiliary_loss_mlp": 0.01029909, + "balance_loss_clip": 1.04444194, + "balance_loss_mlp": 1.02183819, + "epoch": 0.8890759333854386, + "flos": 27381204201600.0, + "grad_norm": 1.7469860777611776, + "language_loss": 0.80413413, + "learning_rate": 1.2757718478441094e-07, + "loss": 0.82604897, + "num_input_tokens_seen": 159901070, + "step": 7394, + "time_per_iteration": 2.700819253921509 + }, + { + "auxiliary_loss_clip": 0.01147425, + "auxiliary_loss_mlp": 0.01025363, + "balance_loss_clip": 1.04304051, + "balance_loss_mlp": 1.0180496, + "epoch": 0.8891961762760777, + "flos": 24498834353280.0, + "grad_norm": 3.7235400492625548, + "language_loss": 0.77285659, + "learning_rate": 1.2730356709427302e-07, + "loss": 0.79458451, + "num_input_tokens_seen": 159919750, + "step": 7395, + "time_per_iteration": 2.8061604499816895 + }, + { + "auxiliary_loss_clip": 0.01156033, + "auxiliary_loss_mlp": 0.01027151, + "balance_loss_clip": 1.04738927, + "balance_loss_mlp": 1.01986098, + "epoch": 0.8893164191667168, + "flos": 41499770895360.0, + "grad_norm": 1.7608157288956439, + "language_loss": 0.59580809, + "learning_rate": 1.2703023349069542e-07, + "loss": 0.6176399, + "num_input_tokens_seen": 159944600, + "step": 7396, + "time_per_iteration": 2.8119592666625977 + }, + { + "auxiliary_loss_clip": 0.01156608, + "auxiliary_loss_mlp": 0.01022161, + "balance_loss_clip": 1.04704773, + "balance_loss_mlp": 1.01449561, + "epoch": 0.8894366620573558, + "flos": 33583623120000.0, + "grad_norm": 1.9819526256985391, + "language_loss": 0.61918342, + "learning_rate": 1.2675718401514223e-07, + "loss": 0.64097112, + "num_input_tokens_seen": 159968780, + "step": 7397, + "time_per_iteration": 2.8194420337677 + }, + { + "auxiliary_loss_clip": 0.01148917, + "auxiliary_loss_mlp": 0.01030292, + "balance_loss_clip": 1.04504657, + "balance_loss_mlp": 1.02270436, + "epoch": 0.889556904947995, + "flos": 16909832672640.0, + "grad_norm": 2.5006933606750916, + "language_loss": 0.74357784, + "learning_rate": 1.264844187090346e-07, + "loss": 0.76536989, + "num_input_tokens_seen": 159985905, + "step": 7398, + "time_per_iteration": 2.734858989715576 + }, + { + "auxiliary_loss_clip": 0.01145049, + "auxiliary_loss_mlp": 0.01035022, + "balance_loss_clip": 1.04250824, + "balance_loss_mlp": 1.02761006, + "epoch": 0.889677147838634, + "flos": 26030855283840.0, + "grad_norm": 1.7730245131477638, + "language_loss": 0.75052398, + "learning_rate": 1.262119376137516e-07, + "loss": 0.77232468, + "num_input_tokens_seen": 160006965, + "step": 7399, + "time_per_iteration": 2.841204881668091 + }, + { + "auxiliary_loss_clip": 0.01149718, + "auxiliary_loss_mlp": 0.01025639, + "balance_loss_clip": 1.04328132, + "balance_loss_mlp": 1.01803362, + "epoch": 0.8897973907292731, + "flos": 26468283110400.0, + "grad_norm": 1.517312102206471, + "language_loss": 0.85148263, + "learning_rate": 1.2593974077062707e-07, + "loss": 0.87323618, + "num_input_tokens_seen": 160028585, + "step": 7400, + "time_per_iteration": 2.804121971130371 + }, + { + "auxiliary_loss_clip": 0.01135465, + "auxiliary_loss_mlp": 0.01028226, + "balance_loss_clip": 1.04341269, + "balance_loss_mlp": 1.02061486, + "epoch": 0.8899176336199123, + "flos": 26249694894720.0, + "grad_norm": 1.494295714540381, + "language_loss": 0.6358878, + "learning_rate": 1.2566782822095423e-07, + "loss": 0.6575247, + "num_input_tokens_seen": 160048840, + "step": 7401, + "time_per_iteration": 2.86118745803833 + }, + { + "auxiliary_loss_clip": 0.01147684, + "auxiliary_loss_mlp": 0.01023038, + "balance_loss_clip": 1.04628301, + "balance_loss_mlp": 1.01565289, + "epoch": 0.8900378765105513, + "flos": 20811742156800.0, + "grad_norm": 1.793176486178132, + "language_loss": 0.71170807, + "learning_rate": 1.2539620000598162e-07, + "loss": 0.73341531, + "num_input_tokens_seen": 160068175, + "step": 7402, + "time_per_iteration": 2.8219850063323975 + }, + { + "auxiliary_loss_clip": 0.01168079, + "auxiliary_loss_mlp": 0.01026731, + "balance_loss_clip": 1.04662299, + "balance_loss_mlp": 1.01900649, + "epoch": 0.8901581194011904, + "flos": 16472333018880.0, + "grad_norm": 2.1999410186725585, + "language_loss": 0.79915738, + "learning_rate": 1.2512485616691492e-07, + "loss": 0.82110548, + "num_input_tokens_seen": 160085230, + "step": 7403, + "time_per_iteration": 2.722052812576294 + }, + { + "auxiliary_loss_clip": 0.01141252, + "auxiliary_loss_mlp": 0.01032851, + "balance_loss_clip": 1.04212594, + "balance_loss_mlp": 1.02490008, + "epoch": 0.8902783622918296, + "flos": 35155253773440.0, + "grad_norm": 1.9104286426899897, + "language_loss": 0.80820942, + "learning_rate": 1.2485379674491681e-07, + "loss": 0.82995045, + "num_input_tokens_seen": 160111425, + "step": 7404, + "time_per_iteration": 3.019947052001953 + }, + { + "auxiliary_loss_clip": 0.01148888, + "auxiliary_loss_mlp": 0.01027735, + "balance_loss_clip": 1.04618478, + "balance_loss_mlp": 1.01997471, + "epoch": 0.8903986051824686, + "flos": 17201068145280.0, + "grad_norm": 2.1900863313009773, + "language_loss": 0.78751785, + "learning_rate": 1.2458302178110657e-07, + "loss": 0.80928409, + "num_input_tokens_seen": 160129790, + "step": 7405, + "time_per_iteration": 2.850806474685669 + }, + { + "auxiliary_loss_clip": 0.01128517, + "auxiliary_loss_mlp": 0.01024493, + "balance_loss_clip": 1.03917241, + "balance_loss_mlp": 1.01697719, + "epoch": 0.8905188480731077, + "flos": 25483863997440.0, + "grad_norm": 2.0006492944843868, + "language_loss": 0.82263112, + "learning_rate": 1.2431253131656118e-07, + "loss": 0.84416127, + "num_input_tokens_seen": 160149265, + "step": 7406, + "time_per_iteration": 2.9290826320648193 + }, + { + "auxiliary_loss_clip": 0.01142653, + "auxiliary_loss_mlp": 0.01024039, + "balance_loss_clip": 1.04316175, + "balance_loss_mlp": 1.01617682, + "epoch": 0.8906390909637467, + "flos": 23365888502400.0, + "grad_norm": 8.722850097796147, + "language_loss": 0.76512945, + "learning_rate": 1.240423253923133e-07, + "loss": 0.78679633, + "num_input_tokens_seen": 160168870, + "step": 7407, + "time_per_iteration": 2.819685935974121 + }, + { + "auxiliary_loss_clip": 0.01160144, + "auxiliary_loss_mlp": 0.01032845, + "balance_loss_clip": 1.04509997, + "balance_loss_mlp": 1.02521586, + "epoch": 0.8907593338543859, + "flos": 21068790860160.0, + "grad_norm": 1.9078877070989215, + "language_loss": 0.69400513, + "learning_rate": 1.237724040493533e-07, + "loss": 0.71593499, + "num_input_tokens_seen": 160187495, + "step": 7408, + "time_per_iteration": 3.1222264766693115 + }, + { + "auxiliary_loss_clip": 0.01173999, + "auxiliary_loss_mlp": 0.01030411, + "balance_loss_clip": 1.04985285, + "balance_loss_mlp": 1.02252555, + "epoch": 0.8908795767450249, + "flos": 21869562712320.0, + "grad_norm": 2.366145909288511, + "language_loss": 0.72522497, + "learning_rate": 1.2350276732862773e-07, + "loss": 0.74726909, + "num_input_tokens_seen": 160208520, + "step": 7409, + "time_per_iteration": 2.965451955795288 + }, + { + "auxiliary_loss_clip": 0.01050596, + "auxiliary_loss_mlp": 0.01000441, + "balance_loss_clip": 1.00462413, + "balance_loss_mlp": 0.99902838, + "epoch": 0.890999819635664, + "flos": 66307869348480.0, + "grad_norm": 0.8334793117868664, + "language_loss": 0.56687129, + "learning_rate": 1.2323341527103993e-07, + "loss": 0.58738166, + "num_input_tokens_seen": 160263720, + "step": 7410, + "time_per_iteration": 3.1537725925445557 + }, + { + "auxiliary_loss_clip": 0.01165442, + "auxiliary_loss_mlp": 0.01024876, + "balance_loss_clip": 1.04521036, + "balance_loss_mlp": 1.01724076, + "epoch": 0.8911200625263032, + "flos": 26869908055680.0, + "grad_norm": 2.1557696872787364, + "language_loss": 0.84851712, + "learning_rate": 1.2296434791745135e-07, + "loss": 0.87042034, + "num_input_tokens_seen": 160282170, + "step": 7411, + "time_per_iteration": 2.7582364082336426 + }, + { + "auxiliary_loss_clip": 0.01161052, + "auxiliary_loss_mlp": 0.01026519, + "balance_loss_clip": 1.04595447, + "balance_loss_mlp": 1.0187403, + "epoch": 0.8912403054169422, + "flos": 20885825957760.0, + "grad_norm": 1.5902076846741544, + "language_loss": 0.766563, + "learning_rate": 1.2269556530867875e-07, + "loss": 0.78843868, + "num_input_tokens_seen": 160300725, + "step": 7412, + "time_per_iteration": 3.7310495376586914 + }, + { + "auxiliary_loss_clip": 0.0117445, + "auxiliary_loss_mlp": 0.01034737, + "balance_loss_clip": 1.04874086, + "balance_loss_mlp": 1.02650571, + "epoch": 0.8913605483075813, + "flos": 27016567286400.0, + "grad_norm": 1.789355609706221, + "language_loss": 0.81746691, + "learning_rate": 1.2242706748549614e-07, + "loss": 0.83955884, + "num_input_tokens_seen": 160318720, + "step": 7413, + "time_per_iteration": 3.23826003074646 + }, + { + "auxiliary_loss_clip": 0.01149488, + "auxiliary_loss_mlp": 0.01027817, + "balance_loss_clip": 1.0405935, + "balance_loss_mlp": 1.01996088, + "epoch": 0.8914807911982204, + "flos": 23621500661760.0, + "grad_norm": 1.8841753552155138, + "language_loss": 0.81986296, + "learning_rate": 1.2215885448863473e-07, + "loss": 0.841636, + "num_input_tokens_seen": 160339595, + "step": 7414, + "time_per_iteration": 3.7772421836853027 + }, + { + "auxiliary_loss_clip": 0.01147317, + "auxiliary_loss_mlp": 0.01031491, + "balance_loss_clip": 1.04466891, + "balance_loss_mlp": 1.02359664, + "epoch": 0.8916010340888595, + "flos": 24462277286400.0, + "grad_norm": 3.66545956819492, + "language_loss": 0.80315542, + "learning_rate": 1.2189092635878152e-07, + "loss": 0.82494354, + "num_input_tokens_seen": 160361045, + "step": 7415, + "time_per_iteration": 3.7096073627471924 + }, + { + "auxiliary_loss_clip": 0.01130366, + "auxiliary_loss_mlp": 0.01027362, + "balance_loss_clip": 1.04055929, + "balance_loss_mlp": 1.01958942, + "epoch": 0.8917212769794985, + "flos": 21215773313280.0, + "grad_norm": 1.960382949627764, + "language_loss": 0.77159286, + "learning_rate": 1.216232831365822e-07, + "loss": 0.79317021, + "num_input_tokens_seen": 160379990, + "step": 7416, + "time_per_iteration": 3.0026774406433105 + }, + { + "auxiliary_loss_clip": 0.01157497, + "auxiliary_loss_mlp": 0.01029294, + "balance_loss_clip": 1.04616332, + "balance_loss_mlp": 1.02116394, + "epoch": 0.8918415198701377, + "flos": 25513992529920.0, + "grad_norm": 1.6855223025923924, + "language_loss": 0.80836236, + "learning_rate": 1.2135592486263678e-07, + "loss": 0.83023018, + "num_input_tokens_seen": 160399240, + "step": 7417, + "time_per_iteration": 2.874154567718506 + }, + { + "auxiliary_loss_clip": 0.01149706, + "auxiliary_loss_mlp": 0.01024772, + "balance_loss_clip": 1.04286432, + "balance_loss_mlp": 1.01758337, + "epoch": 0.8919617627607768, + "flos": 37853006693760.0, + "grad_norm": 1.6303646883668865, + "language_loss": 0.61393964, + "learning_rate": 1.2108885157750415e-07, + "loss": 0.63568437, + "num_input_tokens_seen": 160421600, + "step": 7418, + "time_per_iteration": 2.923516273498535 + }, + { + "auxiliary_loss_clip": 0.01136767, + "auxiliary_loss_mlp": 0.008856, + "balance_loss_clip": 1.04585814, + "balance_loss_mlp": 1.00053346, + "epoch": 0.8920820056514158, + "flos": 26213676531840.0, + "grad_norm": 1.6159307576466475, + "language_loss": 0.79968786, + "learning_rate": 1.2082206332169897e-07, + "loss": 0.81991154, + "num_input_tokens_seen": 160441695, + "step": 7419, + "time_per_iteration": 2.803525924682617 + }, + { + "auxiliary_loss_clip": 0.01142202, + "auxiliary_loss_mlp": 0.01027578, + "balance_loss_clip": 1.04417753, + "balance_loss_mlp": 1.02039564, + "epoch": 0.892202248542055, + "flos": 17383135207680.0, + "grad_norm": 3.457581256519279, + "language_loss": 0.73130161, + "learning_rate": 1.2055556013569225e-07, + "loss": 0.75299942, + "num_input_tokens_seen": 160457205, + "step": 7420, + "time_per_iteration": 3.6776950359344482 + }, + { + "auxiliary_loss_clip": 0.01142781, + "auxiliary_loss_mlp": 0.01026599, + "balance_loss_clip": 1.04325068, + "balance_loss_mlp": 1.01838827, + "epoch": 0.892322491432694, + "flos": 21324223451520.0, + "grad_norm": 2.55918102580378, + "language_loss": 0.82273757, + "learning_rate": 1.2028934205991315e-07, + "loss": 0.8444314, + "num_input_tokens_seen": 160476525, + "step": 7421, + "time_per_iteration": 2.9097230434417725 + }, + { + "auxiliary_loss_clip": 0.01156726, + "auxiliary_loss_mlp": 0.01027261, + "balance_loss_clip": 1.04270923, + "balance_loss_mlp": 1.01903892, + "epoch": 0.8924427343233331, + "flos": 24029374573440.0, + "grad_norm": 1.4283100107588877, + "language_loss": 0.76619989, + "learning_rate": 1.2002340913474607e-07, + "loss": 0.7880398, + "num_input_tokens_seen": 160500160, + "step": 7422, + "time_per_iteration": 2.796576738357544 + }, + { + "auxiliary_loss_clip": 0.01170025, + "auxiliary_loss_mlp": 0.0102893, + "balance_loss_clip": 1.04637671, + "balance_loss_mlp": 1.02043629, + "epoch": 0.8925629772139723, + "flos": 30008069631360.0, + "grad_norm": 2.1938212221440994, + "language_loss": 0.73646331, + "learning_rate": 1.1975776140053317e-07, + "loss": 0.75845289, + "num_input_tokens_seen": 160520130, + "step": 7423, + "time_per_iteration": 2.7334518432617188 + }, + { + "auxiliary_loss_clip": 0.0113236, + "auxiliary_loss_mlp": 0.01032965, + "balance_loss_clip": 1.04277551, + "balance_loss_mlp": 1.02492452, + "epoch": 0.8926832201046113, + "flos": 22601709630720.0, + "grad_norm": 1.9856092695957746, + "language_loss": 0.73208416, + "learning_rate": 1.194923988975729e-07, + "loss": 0.75373739, + "num_input_tokens_seen": 160539730, + "step": 7424, + "time_per_iteration": 2.7966504096984863 + }, + { + "auxiliary_loss_clip": 0.01138025, + "auxiliary_loss_mlp": 0.01024003, + "balance_loss_clip": 1.04442775, + "balance_loss_mlp": 1.0161593, + "epoch": 0.8928034629952504, + "flos": 13297722117120.0, + "grad_norm": 2.305882424277766, + "language_loss": 0.737023, + "learning_rate": 1.192273216661206e-07, + "loss": 0.75864327, + "num_input_tokens_seen": 160557820, + "step": 7425, + "time_per_iteration": 2.7484827041625977 + }, + { + "auxiliary_loss_clip": 0.01027543, + "auxiliary_loss_mlp": 0.01005238, + "balance_loss_clip": 1.00502539, + "balance_loss_mlp": 1.00355721, + "epoch": 0.8929237058858895, + "flos": 54854556744960.0, + "grad_norm": 0.7664012852185124, + "language_loss": 0.57470524, + "learning_rate": 1.189625297463881e-07, + "loss": 0.59503305, + "num_input_tokens_seen": 160619510, + "step": 7426, + "time_per_iteration": 3.3621537685394287 + }, + { + "auxiliary_loss_clip": 0.01119722, + "auxiliary_loss_mlp": 0.01026614, + "balance_loss_clip": 1.039258, + "balance_loss_mlp": 1.01932752, + "epoch": 0.8930439487765286, + "flos": 28883850785280.0, + "grad_norm": 1.5969906589393845, + "language_loss": 0.79610157, + "learning_rate": 1.1869802317854394e-07, + "loss": 0.81756496, + "num_input_tokens_seen": 160643295, + "step": 7427, + "time_per_iteration": 2.9803712368011475 + }, + { + "auxiliary_loss_clip": 0.01133446, + "auxiliary_loss_mlp": 0.01023372, + "balance_loss_clip": 1.04135585, + "balance_loss_mlp": 1.01450253, + "epoch": 0.8931641916671677, + "flos": 22419283432320.0, + "grad_norm": 1.8791048491671989, + "language_loss": 0.72682607, + "learning_rate": 1.1843380200271425e-07, + "loss": 0.74839431, + "num_input_tokens_seen": 160662495, + "step": 7428, + "time_per_iteration": 2.8613507747650146 + }, + { + "auxiliary_loss_clip": 0.01126281, + "auxiliary_loss_mlp": 0.01026436, + "balance_loss_clip": 1.04278469, + "balance_loss_mlp": 1.01895607, + "epoch": 0.8932844345578068, + "flos": 25843149786240.0, + "grad_norm": 1.7012468066044566, + "language_loss": 0.80792797, + "learning_rate": 1.181698662589805e-07, + "loss": 0.82945514, + "num_input_tokens_seen": 160682080, + "step": 7429, + "time_per_iteration": 2.822420835494995 + }, + { + "auxiliary_loss_clip": 0.01156632, + "auxiliary_loss_mlp": 0.01026722, + "balance_loss_clip": 1.04393327, + "balance_loss_mlp": 1.01911056, + "epoch": 0.8934046774484459, + "flos": 22925803069440.0, + "grad_norm": 1.7476102043880715, + "language_loss": 0.76178485, + "learning_rate": 1.1790621598738249e-07, + "loss": 0.78361839, + "num_input_tokens_seen": 160700395, + "step": 7430, + "time_per_iteration": 2.7444162368774414 + }, + { + "auxiliary_loss_clip": 0.01164577, + "auxiliary_loss_mlp": 0.01027837, + "balance_loss_clip": 1.04592752, + "balance_loss_mlp": 1.02075863, + "epoch": 0.8935249203390849, + "flos": 24462097718400.0, + "grad_norm": 1.9259762092450543, + "language_loss": 0.74695933, + "learning_rate": 1.1764285122791461e-07, + "loss": 0.76888347, + "num_input_tokens_seen": 160721115, + "step": 7431, + "time_per_iteration": 2.7434427738189697 + }, + { + "auxiliary_loss_clip": 0.01159321, + "auxiliary_loss_mlp": 0.01030931, + "balance_loss_clip": 1.04454672, + "balance_loss_mlp": 1.02382886, + "epoch": 0.8936451632297241, + "flos": 15742735966080.0, + "grad_norm": 1.9617614931313128, + "language_loss": 0.77133125, + "learning_rate": 1.173797720205294e-07, + "loss": 0.79323375, + "num_input_tokens_seen": 160739150, + "step": 7432, + "time_per_iteration": 2.9344475269317627 + }, + { + "auxiliary_loss_clip": 0.01162457, + "auxiliary_loss_mlp": 0.01022925, + "balance_loss_clip": 1.0484432, + "balance_loss_mlp": 1.01467013, + "epoch": 0.8937654061203631, + "flos": 35115500396160.0, + "grad_norm": 2.822701705903028, + "language_loss": 0.71534204, + "learning_rate": 1.1711697840513602e-07, + "loss": 0.73719585, + "num_input_tokens_seen": 160758585, + "step": 7433, + "time_per_iteration": 2.8501315116882324 + }, + { + "auxiliary_loss_clip": 0.0115176, + "auxiliary_loss_mlp": 0.01024833, + "balance_loss_clip": 1.04239917, + "balance_loss_mlp": 1.01709652, + "epoch": 0.8938856490110022, + "flos": 16107444708480.0, + "grad_norm": 2.123968706862265, + "language_loss": 0.71336579, + "learning_rate": 1.1685447042160012e-07, + "loss": 0.73513174, + "num_input_tokens_seen": 160776620, + "step": 7434, + "time_per_iteration": 2.717083692550659 + }, + { + "auxiliary_loss_clip": 0.01170648, + "auxiliary_loss_mlp": 0.01029556, + "balance_loss_clip": 1.04708242, + "balance_loss_mlp": 1.02106214, + "epoch": 0.8940058919016414, + "flos": 20704189858560.0, + "grad_norm": 1.5771401638372198, + "language_loss": 0.7172972, + "learning_rate": 1.1659224810974367e-07, + "loss": 0.73929918, + "num_input_tokens_seen": 160796580, + "step": 7435, + "time_per_iteration": 2.7117176055908203 + }, + { + "auxiliary_loss_clip": 0.01149295, + "auxiliary_loss_mlp": 0.01029075, + "balance_loss_clip": 1.04713595, + "balance_loss_mlp": 1.02179384, + "epoch": 0.8941261347922804, + "flos": 25229041937280.0, + "grad_norm": 1.3950559828489955, + "language_loss": 0.68328047, + "learning_rate": 1.1633031150934591e-07, + "loss": 0.70506418, + "num_input_tokens_seen": 160819610, + "step": 7436, + "time_per_iteration": 2.876708984375 + }, + { + "auxiliary_loss_clip": 0.01161072, + "auxiliary_loss_mlp": 0.01027889, + "balance_loss_clip": 1.04706585, + "balance_loss_mlp": 1.02025962, + "epoch": 0.8942463776829195, + "flos": 19537236806400.0, + "grad_norm": 1.9136359347886298, + "language_loss": 0.7967205, + "learning_rate": 1.1606866066014176e-07, + "loss": 0.81861007, + "num_input_tokens_seen": 160838660, + "step": 7437, + "time_per_iteration": 2.692415714263916 + }, + { + "auxiliary_loss_clip": 0.01137886, + "auxiliary_loss_mlp": 0.01024883, + "balance_loss_clip": 1.04370046, + "balance_loss_mlp": 1.01715231, + "epoch": 0.8943666205735585, + "flos": 22301567585280.0, + "grad_norm": 7.983210870652214, + "language_loss": 0.75572169, + "learning_rate": 1.1580729560182434e-07, + "loss": 0.77734935, + "num_input_tokens_seen": 160854515, + "step": 7438, + "time_per_iteration": 3.7730329036712646 + }, + { + "auxiliary_loss_clip": 0.01166386, + "auxiliary_loss_mlp": 0.00885439, + "balance_loss_clip": 1.04557657, + "balance_loss_mlp": 1.0005995, + "epoch": 0.8944868634641977, + "flos": 18912893581440.0, + "grad_norm": 1.724103631340188, + "language_loss": 0.71029806, + "learning_rate": 1.1554621637404171e-07, + "loss": 0.73081625, + "num_input_tokens_seen": 160872605, + "step": 7439, + "time_per_iteration": 2.711482286453247 + }, + { + "auxiliary_loss_clip": 0.01159663, + "auxiliary_loss_mlp": 0.01025011, + "balance_loss_clip": 1.04546404, + "balance_loss_mlp": 1.01756632, + "epoch": 0.8946071063548368, + "flos": 14460904241280.0, + "grad_norm": 2.5923264327051627, + "language_loss": 0.61368692, + "learning_rate": 1.1528542301639999e-07, + "loss": 0.63553363, + "num_input_tokens_seen": 160889395, + "step": 7440, + "time_per_iteration": 3.6521058082580566 + }, + { + "auxiliary_loss_clip": 0.01139285, + "auxiliary_loss_mlp": 0.01029938, + "balance_loss_clip": 1.03995335, + "balance_loss_mlp": 1.02159309, + "epoch": 0.8947273492454758, + "flos": 20084084438400.0, + "grad_norm": 2.4389524858227825, + "language_loss": 0.82363099, + "learning_rate": 1.1502491556846105e-07, + "loss": 0.8453232, + "num_input_tokens_seen": 160907890, + "step": 7441, + "time_per_iteration": 3.640993356704712 + }, + { + "auxiliary_loss_clip": 0.01149617, + "auxiliary_loss_mlp": 0.01028902, + "balance_loss_clip": 1.04557419, + "balance_loss_mlp": 1.02098703, + "epoch": 0.894847592136115, + "flos": 18550555136640.0, + "grad_norm": 2.2248108838209144, + "language_loss": 0.81540096, + "learning_rate": 1.1476469406974331e-07, + "loss": 0.8371861, + "num_input_tokens_seen": 160923490, + "step": 7442, + "time_per_iteration": 2.7459654808044434 + }, + { + "auxiliary_loss_clip": 0.01164151, + "auxiliary_loss_mlp": 0.01026455, + "balance_loss_clip": 1.04485631, + "balance_loss_mlp": 1.01847386, + "epoch": 0.894967835026754, + "flos": 23478468704640.0, + "grad_norm": 1.6388578095518058, + "language_loss": 0.76944977, + "learning_rate": 1.1450475855972341e-07, + "loss": 0.79135579, + "num_input_tokens_seen": 160944280, + "step": 7443, + "time_per_iteration": 2.740523338317871 + }, + { + "auxiliary_loss_clip": 0.01146195, + "auxiliary_loss_mlp": 0.00885634, + "balance_loss_clip": 1.04144382, + "balance_loss_mlp": 1.00054908, + "epoch": 0.8950880779173931, + "flos": 15188310564480.0, + "grad_norm": 2.0329913003587654, + "language_loss": 0.70592153, + "learning_rate": 1.1424510907783158e-07, + "loss": 0.72623992, + "num_input_tokens_seen": 160961560, + "step": 7444, + "time_per_iteration": 2.7621593475341797 + }, + { + "auxiliary_loss_clip": 0.01154906, + "auxiliary_loss_mlp": 0.01026767, + "balance_loss_clip": 1.04272413, + "balance_loss_mlp": 1.01995122, + "epoch": 0.8952083208080323, + "flos": 22091957769600.0, + "grad_norm": 1.5528022623868531, + "language_loss": 0.82536829, + "learning_rate": 1.1398574566345787e-07, + "loss": 0.84718502, + "num_input_tokens_seen": 160982195, + "step": 7445, + "time_per_iteration": 3.6839449405670166 + }, + { + "auxiliary_loss_clip": 0.01153514, + "auxiliary_loss_mlp": 0.01026913, + "balance_loss_clip": 1.04277849, + "balance_loss_mlp": 1.01914024, + "epoch": 0.8953285636986713, + "flos": 23254026572160.0, + "grad_norm": 2.1332101914439723, + "language_loss": 0.82155937, + "learning_rate": 1.1372666835594702e-07, + "loss": 0.84336364, + "num_input_tokens_seen": 161000520, + "step": 7446, + "time_per_iteration": 2.751713514328003 + }, + { + "auxiliary_loss_clip": 0.01147418, + "auxiliary_loss_mlp": 0.01027574, + "balance_loss_clip": 1.04414308, + "balance_loss_mlp": 1.0198791, + "epoch": 0.8954488065893104, + "flos": 16362661818240.0, + "grad_norm": 2.1950612467879673, + "language_loss": 0.72037423, + "learning_rate": 1.1346787719460071e-07, + "loss": 0.74212414, + "num_input_tokens_seen": 161019405, + "step": 7447, + "time_per_iteration": 2.7682137489318848 + }, + { + "auxiliary_loss_clip": 0.01147445, + "auxiliary_loss_mlp": 0.01025147, + "balance_loss_clip": 1.04329824, + "balance_loss_mlp": 1.01781905, + "epoch": 0.8955690494799495, + "flos": 18257883120000.0, + "grad_norm": 1.9493047794553944, + "language_loss": 0.72167033, + "learning_rate": 1.1320937221867732e-07, + "loss": 0.74339622, + "num_input_tokens_seen": 161036985, + "step": 7448, + "time_per_iteration": 2.782609701156616 + }, + { + "auxiliary_loss_clip": 0.01147761, + "auxiliary_loss_mlp": 0.01028214, + "balance_loss_clip": 1.04264426, + "balance_loss_mlp": 1.0212698, + "epoch": 0.8956892923705886, + "flos": 25447486498560.0, + "grad_norm": 1.7346927751470156, + "language_loss": 0.79412115, + "learning_rate": 1.1295115346739192e-07, + "loss": 0.81588089, + "num_input_tokens_seen": 161056985, + "step": 7449, + "time_per_iteration": 2.866950035095215 + }, + { + "auxiliary_loss_clip": 0.01154944, + "auxiliary_loss_mlp": 0.01024493, + "balance_loss_clip": 1.04626846, + "balance_loss_mlp": 1.01629782, + "epoch": 0.8958095352612276, + "flos": 52661883939840.0, + "grad_norm": 2.4301832076570413, + "language_loss": 0.73246086, + "learning_rate": 1.1269322097991629e-07, + "loss": 0.75425524, + "num_input_tokens_seen": 161080270, + "step": 7450, + "time_per_iteration": 3.088146924972534 + }, + { + "auxiliary_loss_clip": 0.01165276, + "auxiliary_loss_mlp": 0.01027663, + "balance_loss_clip": 1.04913235, + "balance_loss_mlp": 1.01888943, + "epoch": 0.8959297781518668, + "flos": 23186335392000.0, + "grad_norm": 2.040110077653717, + "language_loss": 0.67998743, + "learning_rate": 1.1243557479537846e-07, + "loss": 0.70191681, + "num_input_tokens_seen": 161100160, + "step": 7451, + "time_per_iteration": 2.7590837478637695 + }, + { + "auxiliary_loss_clip": 0.01168714, + "auxiliary_loss_mlp": 0.01022671, + "balance_loss_clip": 1.04514611, + "balance_loss_mlp": 1.01482105, + "epoch": 0.8960500210425059, + "flos": 20334309557760.0, + "grad_norm": 2.2644344322970142, + "language_loss": 0.68805718, + "learning_rate": 1.121782149528634e-07, + "loss": 0.70997107, + "num_input_tokens_seen": 161117260, + "step": 7452, + "time_per_iteration": 2.69095516204834 + }, + { + "auxiliary_loss_clip": 0.01145168, + "auxiliary_loss_mlp": 0.01028376, + "balance_loss_clip": 1.04639721, + "balance_loss_mlp": 1.02067494, + "epoch": 0.8961702639331449, + "flos": 19901694153600.0, + "grad_norm": 2.2301883760974577, + "language_loss": 0.78413981, + "learning_rate": 1.1192114149141208e-07, + "loss": 0.80587518, + "num_input_tokens_seen": 161136895, + "step": 7453, + "time_per_iteration": 2.726846933364868 + }, + { + "auxiliary_loss_clip": 0.01155036, + "auxiliary_loss_mlp": 0.01028136, + "balance_loss_clip": 1.04483271, + "balance_loss_mlp": 1.01992249, + "epoch": 0.8962905068237841, + "flos": 12896348567040.0, + "grad_norm": 2.1771680848224646, + "language_loss": 0.65353179, + "learning_rate": 1.1166435445002197e-07, + "loss": 0.67536354, + "num_input_tokens_seen": 161154565, + "step": 7454, + "time_per_iteration": 2.9009642601013184 + }, + { + "auxiliary_loss_clip": 0.0116119, + "auxiliary_loss_mlp": 0.01032656, + "balance_loss_clip": 1.04588223, + "balance_loss_mlp": 1.02466869, + "epoch": 0.8964107497144231, + "flos": 23440331439360.0, + "grad_norm": 2.1728972265509587, + "language_loss": 0.68315983, + "learning_rate": 1.1140785386764818e-07, + "loss": 0.70509833, + "num_input_tokens_seen": 161173265, + "step": 7455, + "time_per_iteration": 2.632031202316284 + }, + { + "auxiliary_loss_clip": 0.01153954, + "auxiliary_loss_mlp": 0.01027958, + "balance_loss_clip": 1.04433751, + "balance_loss_mlp": 1.01977146, + "epoch": 0.8965309926050622, + "flos": 19500176949120.0, + "grad_norm": 1.9890659372108368, + "language_loss": 0.69523549, + "learning_rate": 1.1115163978320153e-07, + "loss": 0.71705461, + "num_input_tokens_seen": 161191995, + "step": 7456, + "time_per_iteration": 2.666085958480835 + }, + { + "auxiliary_loss_clip": 0.01164601, + "auxiliary_loss_mlp": 0.00886009, + "balance_loss_clip": 1.04782009, + "balance_loss_mlp": 1.00049543, + "epoch": 0.8966512354957014, + "flos": 28658008022400.0, + "grad_norm": 1.9736810869142314, + "language_loss": 0.82365239, + "learning_rate": 1.1089571223554917e-07, + "loss": 0.84415847, + "num_input_tokens_seen": 161212880, + "step": 7457, + "time_per_iteration": 2.7390964031219482 + }, + { + "auxiliary_loss_clip": 0.01160406, + "auxiliary_loss_mlp": 0.01028823, + "balance_loss_clip": 1.04390466, + "balance_loss_mlp": 1.0209074, + "epoch": 0.8967714783863404, + "flos": 23370916406400.0, + "grad_norm": 2.4877934376664808, + "language_loss": 0.85432744, + "learning_rate": 1.1064007126351537e-07, + "loss": 0.87621975, + "num_input_tokens_seen": 161233595, + "step": 7458, + "time_per_iteration": 2.6656410694122314 + }, + { + "auxiliary_loss_clip": 0.01145262, + "auxiliary_loss_mlp": 0.01027869, + "balance_loss_clip": 1.04490805, + "balance_loss_mlp": 1.02025723, + "epoch": 0.8968917212769795, + "flos": 24535175938560.0, + "grad_norm": 2.1537452577607583, + "language_loss": 0.7649371, + "learning_rate": 1.1038471690588003e-07, + "loss": 0.78666842, + "num_input_tokens_seen": 161252740, + "step": 7459, + "time_per_iteration": 2.734471559524536 + }, + { + "auxiliary_loss_clip": 0.01128008, + "auxiliary_loss_mlp": 0.01025068, + "balance_loss_clip": 1.04275978, + "balance_loss_mlp": 1.01769495, + "epoch": 0.8970119641676186, + "flos": 23475416048640.0, + "grad_norm": 1.8955636764587422, + "language_loss": 0.79892504, + "learning_rate": 1.1012964920138145e-07, + "loss": 0.82045579, + "num_input_tokens_seen": 161272325, + "step": 7460, + "time_per_iteration": 2.8140664100646973 + }, + { + "auxiliary_loss_clip": 0.01143847, + "auxiliary_loss_mlp": 0.01026643, + "balance_loss_clip": 1.04095852, + "balance_loss_mlp": 1.01918674, + "epoch": 0.8971322070582577, + "flos": 24538192680960.0, + "grad_norm": 1.5274829483136014, + "language_loss": 0.75949621, + "learning_rate": 1.0987486818871205e-07, + "loss": 0.78120112, + "num_input_tokens_seen": 161295915, + "step": 7461, + "time_per_iteration": 2.7826743125915527 + }, + { + "auxiliary_loss_clip": 0.0115786, + "auxiliary_loss_mlp": 0.00886255, + "balance_loss_clip": 1.04556084, + "balance_loss_mlp": 1.00062346, + "epoch": 0.8972524499488967, + "flos": 21797454159360.0, + "grad_norm": 2.2643155408403834, + "language_loss": 0.72961581, + "learning_rate": 1.0962037390652245e-07, + "loss": 0.75005698, + "num_input_tokens_seen": 161314935, + "step": 7462, + "time_per_iteration": 2.6585328578948975 + }, + { + "auxiliary_loss_clip": 0.01147873, + "auxiliary_loss_mlp": 0.0102294, + "balance_loss_clip": 1.0440439, + "balance_loss_mlp": 1.01464295, + "epoch": 0.8973726928395359, + "flos": 21726243446400.0, + "grad_norm": 1.7092835660424037, + "language_loss": 0.72228104, + "learning_rate": 1.0936616639341911e-07, + "loss": 0.74398911, + "num_input_tokens_seen": 161335225, + "step": 7463, + "time_per_iteration": 2.73610520362854 + }, + { + "auxiliary_loss_clip": 0.01049552, + "auxiliary_loss_mlp": 0.01001663, + "balance_loss_clip": 1.00788021, + "balance_loss_mlp": 1.00036991, + "epoch": 0.897492935730175, + "flos": 53837100097920.0, + "grad_norm": 0.735119531860407, + "language_loss": 0.54776782, + "learning_rate": 1.0911224568796473e-07, + "loss": 0.56827998, + "num_input_tokens_seen": 161393420, + "step": 7464, + "time_per_iteration": 3.2772319316864014 + }, + { + "auxiliary_loss_clip": 0.01158014, + "auxiliary_loss_mlp": 0.01026611, + "balance_loss_clip": 1.04741096, + "balance_loss_mlp": 1.0195272, + "epoch": 0.897613178620814, + "flos": 18290346036480.0, + "grad_norm": 2.0548729691795296, + "language_loss": 0.7088623, + "learning_rate": 1.0885861182867984e-07, + "loss": 0.7307086, + "num_input_tokens_seen": 161411525, + "step": 7465, + "time_per_iteration": 3.638655424118042 + }, + { + "auxiliary_loss_clip": 0.01150898, + "auxiliary_loss_mlp": 0.01027152, + "balance_loss_clip": 1.04332638, + "balance_loss_mlp": 1.01982629, + "epoch": 0.8977334215114532, + "flos": 32993718059520.0, + "grad_norm": 1.8175160612484356, + "language_loss": 0.70879889, + "learning_rate": 1.0860526485403942e-07, + "loss": 0.73057944, + "num_input_tokens_seen": 161432800, + "step": 7466, + "time_per_iteration": 3.771218776702881 + }, + { + "auxiliary_loss_clip": 0.01167483, + "auxiliary_loss_mlp": 0.01026936, + "balance_loss_clip": 1.0462501, + "balance_loss_mlp": 1.01970875, + "epoch": 0.8978536644020922, + "flos": 15195636938880.0, + "grad_norm": 1.6912602371540895, + "language_loss": 0.77140826, + "learning_rate": 1.0835220480247675e-07, + "loss": 0.79335248, + "num_input_tokens_seen": 161451295, + "step": 7467, + "time_per_iteration": 3.565678834915161 + }, + { + "auxiliary_loss_clip": 0.01149589, + "auxiliary_loss_mlp": 0.01025364, + "balance_loss_clip": 1.04547763, + "balance_loss_mlp": 1.01793146, + "epoch": 0.8979739072927313, + "flos": 18004389863040.0, + "grad_norm": 2.4083717606358377, + "language_loss": 0.83220452, + "learning_rate": 1.0809943171238067e-07, + "loss": 0.85395408, + "num_input_tokens_seen": 161469220, + "step": 7468, + "time_per_iteration": 2.9424071311950684 + }, + { + "auxiliary_loss_clip": 0.0115791, + "auxiliary_loss_mlp": 0.01030948, + "balance_loss_clip": 1.04633439, + "balance_loss_mlp": 1.02262139, + "epoch": 0.8980941501833704, + "flos": 22271546793600.0, + "grad_norm": 2.2608542013352486, + "language_loss": 0.63024175, + "learning_rate": 1.078469456220965e-07, + "loss": 0.65213037, + "num_input_tokens_seen": 161489375, + "step": 7469, + "time_per_iteration": 2.8312227725982666 + }, + { + "auxiliary_loss_clip": 0.01150409, + "auxiliary_loss_mlp": 0.01026804, + "balance_loss_clip": 1.04351008, + "balance_loss_mlp": 1.01893067, + "epoch": 0.8982143930740095, + "flos": 37560729726720.0, + "grad_norm": 1.9608041899309867, + "language_loss": 0.69703424, + "learning_rate": 1.0759474656992606e-07, + "loss": 0.71880639, + "num_input_tokens_seen": 161512145, + "step": 7470, + "time_per_iteration": 2.7859597206115723 + }, + { + "auxiliary_loss_clip": 0.01154108, + "auxiliary_loss_mlp": 0.01029204, + "balance_loss_clip": 1.04237008, + "balance_loss_mlp": 1.0209074, + "epoch": 0.8983346359646486, + "flos": 18076893465600.0, + "grad_norm": 2.7598256356194852, + "language_loss": 0.77672136, + "learning_rate": 1.0734283459412785e-07, + "loss": 0.79855448, + "num_input_tokens_seen": 161528995, + "step": 7471, + "time_per_iteration": 3.660151958465576 + }, + { + "auxiliary_loss_clip": 0.01133712, + "auxiliary_loss_mlp": 0.01035886, + "balance_loss_clip": 1.0406034, + "balance_loss_mlp": 1.02738047, + "epoch": 0.8984548788552876, + "flos": 20558895344640.0, + "grad_norm": 1.7123536122867584, + "language_loss": 0.80665243, + "learning_rate": 1.0709120973291707e-07, + "loss": 0.8283484, + "num_input_tokens_seen": 161548775, + "step": 7472, + "time_per_iteration": 2.929550886154175 + }, + { + "auxiliary_loss_clip": 0.01173082, + "auxiliary_loss_mlp": 0.0103023, + "balance_loss_clip": 1.04943514, + "balance_loss_mlp": 1.02180159, + "epoch": 0.8985751217459268, + "flos": 17785442511360.0, + "grad_norm": 2.3600444392632784, + "language_loss": 0.77813989, + "learning_rate": 1.0683987202446475e-07, + "loss": 0.80017298, + "num_input_tokens_seen": 161566960, + "step": 7473, + "time_per_iteration": 2.6145429611206055 + }, + { + "auxiliary_loss_clip": 0.01161981, + "auxiliary_loss_mlp": 0.01022913, + "balance_loss_clip": 1.04482996, + "balance_loss_mlp": 1.01516449, + "epoch": 0.8986953646365659, + "flos": 21617003208960.0, + "grad_norm": 1.8608256774545753, + "language_loss": 0.69941092, + "learning_rate": 1.0658882150689862e-07, + "loss": 0.72125983, + "num_input_tokens_seen": 161585820, + "step": 7474, + "time_per_iteration": 2.7513482570648193 + }, + { + "auxiliary_loss_clip": 0.01145254, + "auxiliary_loss_mlp": 0.01024744, + "balance_loss_clip": 1.04283679, + "balance_loss_mlp": 1.01629198, + "epoch": 0.8988156075272049, + "flos": 14027355083520.0, + "grad_norm": 2.4554834159644754, + "language_loss": 0.78846818, + "learning_rate": 1.0633805821830288e-07, + "loss": 0.81016815, + "num_input_tokens_seen": 161602505, + "step": 7475, + "time_per_iteration": 2.7072060108184814 + }, + { + "auxiliary_loss_clip": 0.01151459, + "auxiliary_loss_mlp": 0.01028815, + "balance_loss_clip": 1.04615545, + "balance_loss_mlp": 1.02096474, + "epoch": 0.8989358504178441, + "flos": 29059202004480.0, + "grad_norm": 3.332510638575725, + "language_loss": 0.82753533, + "learning_rate": 1.0608758219671753e-07, + "loss": 0.84933805, + "num_input_tokens_seen": 161621545, + "step": 7476, + "time_per_iteration": 2.812938928604126 + }, + { + "auxiliary_loss_clip": 0.0115623, + "auxiliary_loss_mlp": 0.01027405, + "balance_loss_clip": 1.04649818, + "balance_loss_mlp": 1.01995397, + "epoch": 0.8990560933084831, + "flos": 20230420446720.0, + "grad_norm": 1.9269853743839154, + "language_loss": 0.70561171, + "learning_rate": 1.0583739348014065e-07, + "loss": 0.72744811, + "num_input_tokens_seen": 161642630, + "step": 7477, + "time_per_iteration": 2.708930492401123 + }, + { + "auxiliary_loss_clip": 0.01169325, + "auxiliary_loss_mlp": 0.01022983, + "balance_loss_clip": 1.04786158, + "balance_loss_mlp": 1.01531148, + "epoch": 0.8991763361991222, + "flos": 25520672459520.0, + "grad_norm": 2.01491085709289, + "language_loss": 0.84630412, + "learning_rate": 1.0558749210652518e-07, + "loss": 0.86822718, + "num_input_tokens_seen": 161662560, + "step": 7478, + "time_per_iteration": 2.6595230102539062 + }, + { + "auxiliary_loss_clip": 0.01145913, + "auxiliary_loss_mlp": 0.0102832, + "balance_loss_clip": 1.04385734, + "balance_loss_mlp": 1.02022529, + "epoch": 0.8992965790897613, + "flos": 25119191168640.0, + "grad_norm": 1.714664343266138, + "language_loss": 0.85225987, + "learning_rate": 1.053378781137808e-07, + "loss": 0.87400222, + "num_input_tokens_seen": 161683480, + "step": 7479, + "time_per_iteration": 2.7968764305114746 + }, + { + "auxiliary_loss_clip": 0.01144806, + "auxiliary_loss_mlp": 0.01028274, + "balance_loss_clip": 1.04502237, + "balance_loss_mlp": 1.02004278, + "epoch": 0.8994168219804004, + "flos": 16070815814400.0, + "grad_norm": 1.892800676860529, + "language_loss": 0.7773056, + "learning_rate": 1.0508855153977392e-07, + "loss": 0.79903638, + "num_input_tokens_seen": 161699945, + "step": 7480, + "time_per_iteration": 2.7646865844726562 + }, + { + "auxiliary_loss_clip": 0.01157261, + "auxiliary_loss_mlp": 0.01027672, + "balance_loss_clip": 1.04194784, + "balance_loss_mlp": 1.02013814, + "epoch": 0.8995370648710395, + "flos": 24825764966400.0, + "grad_norm": 2.9966404873510606, + "language_loss": 0.6713317, + "learning_rate": 1.0483951242232669e-07, + "loss": 0.69318104, + "num_input_tokens_seen": 161720420, + "step": 7481, + "time_per_iteration": 2.7210991382598877 + }, + { + "auxiliary_loss_clip": 0.0105512, + "auxiliary_loss_mlp": 0.01001601, + "balance_loss_clip": 1.00454855, + "balance_loss_mlp": 1.00024831, + "epoch": 0.8996573077616786, + "flos": 63116238378240.0, + "grad_norm": 0.9718518730612731, + "language_loss": 0.57721126, + "learning_rate": 1.0459076079921936e-07, + "loss": 0.59777856, + "num_input_tokens_seen": 161773080, + "step": 7482, + "time_per_iteration": 3.283956289291382 + }, + { + "auxiliary_loss_clip": 0.01146772, + "auxiliary_loss_mlp": 0.01030137, + "balance_loss_clip": 1.04648817, + "balance_loss_mlp": 1.02224529, + "epoch": 0.8997775506523177, + "flos": 18219674027520.0, + "grad_norm": 2.2217709495075626, + "language_loss": 0.84889108, + "learning_rate": 1.0434229670818618e-07, + "loss": 0.87066019, + "num_input_tokens_seen": 161789755, + "step": 7483, + "time_per_iteration": 2.9407670497894287 + }, + { + "auxiliary_loss_clip": 0.01141265, + "auxiliary_loss_mlp": 0.01025979, + "balance_loss_clip": 1.04287434, + "balance_loss_mlp": 1.0177182, + "epoch": 0.8998977935429567, + "flos": 24166768095360.0, + "grad_norm": 1.4798946695535364, + "language_loss": 0.7992878, + "learning_rate": 1.0409412018691944e-07, + "loss": 0.82096028, + "num_input_tokens_seen": 161810220, + "step": 7484, + "time_per_iteration": 2.850292444229126 + }, + { + "auxiliary_loss_clip": 0.0114567, + "auxiliary_loss_mlp": 0.01036288, + "balance_loss_clip": 1.04429412, + "balance_loss_mlp": 1.02758551, + "epoch": 0.9000180364335959, + "flos": 20773030273920.0, + "grad_norm": 1.9046787168897017, + "language_loss": 0.75206614, + "learning_rate": 1.0384623127306724e-07, + "loss": 0.77388573, + "num_input_tokens_seen": 161827565, + "step": 7485, + "time_per_iteration": 2.8567373752593994 + }, + { + "auxiliary_loss_clip": 0.01137566, + "auxiliary_loss_mlp": 0.01024893, + "balance_loss_clip": 1.04108357, + "balance_loss_mlp": 1.01743007, + "epoch": 0.900138279324235, + "flos": 19205745166080.0, + "grad_norm": 1.7299264601401405, + "language_loss": 0.7957961, + "learning_rate": 1.0359863000423397e-07, + "loss": 0.81742066, + "num_input_tokens_seen": 161845700, + "step": 7486, + "time_per_iteration": 3.0635218620300293 + }, + { + "auxiliary_loss_clip": 0.01167638, + "auxiliary_loss_mlp": 0.01027594, + "balance_loss_clip": 1.04510641, + "balance_loss_mlp": 1.02004242, + "epoch": 0.900258522214874, + "flos": 28731158069760.0, + "grad_norm": 1.6879248752782592, + "language_loss": 0.72079903, + "learning_rate": 1.0335131641798112e-07, + "loss": 0.74275136, + "num_input_tokens_seen": 161867660, + "step": 7487, + "time_per_iteration": 2.800881862640381 + }, + { + "auxiliary_loss_clip": 0.01042533, + "auxiliary_loss_mlp": 0.01001393, + "balance_loss_clip": 1.00493145, + "balance_loss_mlp": 1.00001037, + "epoch": 0.9003787651055132, + "flos": 58280685655680.0, + "grad_norm": 0.812104730338037, + "language_loss": 0.55643964, + "learning_rate": 1.0310429055182512e-07, + "loss": 0.57687891, + "num_input_tokens_seen": 161921980, + "step": 7488, + "time_per_iteration": 3.1307835578918457 + }, + { + "auxiliary_loss_clip": 0.0114148, + "auxiliary_loss_mlp": 0.01030468, + "balance_loss_clip": 1.04274309, + "balance_loss_mlp": 1.02221239, + "epoch": 0.9004990079961522, + "flos": 25556475340800.0, + "grad_norm": 2.060642550426612, + "language_loss": 0.73990625, + "learning_rate": 1.0285755244324024e-07, + "loss": 0.76162571, + "num_input_tokens_seen": 161942725, + "step": 7489, + "time_per_iteration": 2.81839656829834 + }, + { + "auxiliary_loss_clip": 0.01151479, + "auxiliary_loss_mlp": 0.00885138, + "balance_loss_clip": 1.04206157, + "balance_loss_mlp": 1.00062537, + "epoch": 0.9006192508867913, + "flos": 23335185352320.0, + "grad_norm": 1.4807313755336298, + "language_loss": 0.68742228, + "learning_rate": 1.0261110212965629e-07, + "loss": 0.70778841, + "num_input_tokens_seen": 161964520, + "step": 7490, + "time_per_iteration": 2.8502144813537598 + }, + { + "auxiliary_loss_clip": 0.01149268, + "auxiliary_loss_mlp": 0.01026845, + "balance_loss_clip": 1.04391313, + "balance_loss_mlp": 1.01963878, + "epoch": 0.9007394937774305, + "flos": 18040300485120.0, + "grad_norm": 2.1668255448119176, + "language_loss": 0.79233778, + "learning_rate": 1.023649396484596e-07, + "loss": 0.81409889, + "num_input_tokens_seen": 161983575, + "step": 7491, + "time_per_iteration": 3.952707290649414 + }, + { + "auxiliary_loss_clip": 0.01169261, + "auxiliary_loss_mlp": 0.01035764, + "balance_loss_clip": 1.04746675, + "balance_loss_mlp": 1.02809262, + "epoch": 0.9008597366680695, + "flos": 43068456633600.0, + "grad_norm": 1.8494698334502144, + "language_loss": 0.67211711, + "learning_rate": 1.0211906503699275e-07, + "loss": 0.69416738, + "num_input_tokens_seen": 162006550, + "step": 7492, + "time_per_iteration": 3.823385238647461 + }, + { + "auxiliary_loss_clip": 0.01165099, + "auxiliary_loss_mlp": 0.01026207, + "balance_loss_clip": 1.04959095, + "balance_loss_mlp": 1.01792741, + "epoch": 0.9009799795587086, + "flos": 14939055112320.0, + "grad_norm": 7.32405127735126, + "language_loss": 0.81990206, + "learning_rate": 1.0187347833255455e-07, + "loss": 0.84181511, + "num_input_tokens_seen": 162022455, + "step": 7493, + "time_per_iteration": 3.648843288421631 + }, + { + "auxiliary_loss_clip": 0.0116751, + "auxiliary_loss_mlp": 0.01024625, + "balance_loss_clip": 1.04786646, + "balance_loss_mlp": 1.01703715, + "epoch": 0.9011002224493477, + "flos": 21579584215680.0, + "grad_norm": 1.8060977211969347, + "language_loss": 0.78968364, + "learning_rate": 1.0162817957240056e-07, + "loss": 0.81160498, + "num_input_tokens_seen": 162042350, + "step": 7494, + "time_per_iteration": 2.7010550498962402 + }, + { + "auxiliary_loss_clip": 0.01050945, + "auxiliary_loss_mlp": 0.01001658, + "balance_loss_clip": 1.00551653, + "balance_loss_mlp": 1.00030541, + "epoch": 0.9012204653399868, + "flos": 71166367883520.0, + "grad_norm": 0.8997689248053978, + "language_loss": 0.63075411, + "learning_rate": 1.0138316879374253e-07, + "loss": 0.65128016, + "num_input_tokens_seen": 162111640, + "step": 7495, + "time_per_iteration": 3.41225004196167 + }, + { + "auxiliary_loss_clip": 0.01152305, + "auxiliary_loss_mlp": 0.01028223, + "balance_loss_clip": 1.046031, + "balance_loss_mlp": 1.02089167, + "epoch": 0.9013407082306258, + "flos": 15594963413760.0, + "grad_norm": 3.1093542880387695, + "language_loss": 0.74320936, + "learning_rate": 1.0113844603374833e-07, + "loss": 0.76501459, + "num_input_tokens_seen": 162128165, + "step": 7496, + "time_per_iteration": 3.498748302459717 + }, + { + "auxiliary_loss_clip": 0.01147902, + "auxiliary_loss_mlp": 0.01030368, + "balance_loss_clip": 1.042238, + "balance_loss_mlp": 1.02192175, + "epoch": 0.901460951121265, + "flos": 15049157276160.0, + "grad_norm": 2.7805473604963225, + "language_loss": 0.72395623, + "learning_rate": 1.0089401132954178e-07, + "loss": 0.74573886, + "num_input_tokens_seen": 162146145, + "step": 7497, + "time_per_iteration": 3.7569103240966797 + }, + { + "auxiliary_loss_clip": 0.01150898, + "auxiliary_loss_mlp": 0.01029713, + "balance_loss_clip": 1.04702592, + "balance_loss_mlp": 1.02187502, + "epoch": 0.9015811940119041, + "flos": 22236857233920.0, + "grad_norm": 1.6295116013358129, + "language_loss": 0.72376192, + "learning_rate": 1.006498647182037e-07, + "loss": 0.74556798, + "num_input_tokens_seen": 162164800, + "step": 7498, + "time_per_iteration": 2.7518818378448486 + }, + { + "auxiliary_loss_clip": 0.01116534, + "auxiliary_loss_mlp": 0.01027098, + "balance_loss_clip": 1.03780603, + "balance_loss_mlp": 1.01926005, + "epoch": 0.9017014369025431, + "flos": 24973824827520.0, + "grad_norm": 2.08842225854856, + "language_loss": 0.71497232, + "learning_rate": 1.004060062367713e-07, + "loss": 0.73640859, + "num_input_tokens_seen": 162185895, + "step": 7499, + "time_per_iteration": 2.972752094268799 + }, + { + "auxiliary_loss_clip": 0.01162864, + "auxiliary_loss_mlp": 0.01028832, + "balance_loss_clip": 1.04738951, + "balance_loss_mlp": 1.02060699, + "epoch": 0.9018216797931822, + "flos": 18114168804480.0, + "grad_norm": 1.6613322667445722, + "language_loss": 0.69752431, + "learning_rate": 1.0016243592223728e-07, + "loss": 0.71944135, + "num_input_tokens_seen": 162206295, + "step": 7500, + "time_per_iteration": 2.675417184829712 + }, + { + "auxiliary_loss_clip": 0.01118489, + "auxiliary_loss_mlp": 0.01031752, + "balance_loss_clip": 1.03953671, + "balance_loss_mlp": 1.02437639, + "epoch": 0.9019419226838213, + "flos": 37268452759680.0, + "grad_norm": 1.7773417967839518, + "language_loss": 0.653467, + "learning_rate": 9.991915381155114e-08, + "loss": 0.67496943, + "num_input_tokens_seen": 162229275, + "step": 7501, + "time_per_iteration": 2.9877710342407227 + }, + { + "auxiliary_loss_clip": 0.01161956, + "auxiliary_loss_mlp": 0.01030541, + "balance_loss_clip": 1.04472923, + "balance_loss_mlp": 1.02226758, + "epoch": 0.9020621655744604, + "flos": 23441121538560.0, + "grad_norm": 5.453084634707804, + "language_loss": 0.74629354, + "learning_rate": 9.967615994161871e-08, + "loss": 0.76821852, + "num_input_tokens_seen": 162248935, + "step": 7502, + "time_per_iteration": 2.727959632873535 + }, + { + "auxiliary_loss_clip": 0.01168002, + "auxiliary_loss_mlp": 0.01028938, + "balance_loss_clip": 1.04751086, + "balance_loss_mlp": 1.02124023, + "epoch": 0.9021824084650995, + "flos": 22857465444480.0, + "grad_norm": 1.6403535604530908, + "language_loss": 0.78220612, + "learning_rate": 9.943345434930161e-08, + "loss": 0.8041755, + "num_input_tokens_seen": 162269185, + "step": 7503, + "time_per_iteration": 2.6525306701660156 + }, + { + "auxiliary_loss_clip": 0.0113914, + "auxiliary_loss_mlp": 0.0102989, + "balance_loss_clip": 1.04481816, + "balance_loss_mlp": 1.02233863, + "epoch": 0.9023026513557386, + "flos": 22127581082880.0, + "grad_norm": 1.9941655233508524, + "language_loss": 0.68999672, + "learning_rate": 9.919103707141885e-08, + "loss": 0.71168697, + "num_input_tokens_seen": 162288065, + "step": 7504, + "time_per_iteration": 2.817530870437622 + }, + { + "auxiliary_loss_clip": 0.01161637, + "auxiliary_loss_mlp": 0.01026618, + "balance_loss_clip": 1.04684246, + "balance_loss_mlp": 1.01918793, + "epoch": 0.9024228942463777, + "flos": 24199087357440.0, + "grad_norm": 8.043649091545715, + "language_loss": 0.76419592, + "learning_rate": 9.89489081447441e-08, + "loss": 0.78607845, + "num_input_tokens_seen": 162305265, + "step": 7505, + "time_per_iteration": 2.7032344341278076 + }, + { + "auxiliary_loss_clip": 0.01147493, + "auxiliary_loss_mlp": 0.01023114, + "balance_loss_clip": 1.04286492, + "balance_loss_mlp": 1.01475191, + "epoch": 0.9025431371370167, + "flos": 25008262992000.0, + "grad_norm": 1.8093644999216185, + "language_loss": 0.82938021, + "learning_rate": 9.870706760600844e-08, + "loss": 0.85108626, + "num_input_tokens_seen": 162325215, + "step": 7506, + "time_per_iteration": 2.7473323345184326 + }, + { + "auxiliary_loss_clip": 0.01132027, + "auxiliary_loss_mlp": 0.01025653, + "balance_loss_clip": 1.04814935, + "balance_loss_mlp": 1.01756465, + "epoch": 0.9026633800276559, + "flos": 18952862440320.0, + "grad_norm": 2.1566784732066577, + "language_loss": 0.73035848, + "learning_rate": 9.846551549189918e-08, + "loss": 0.7519353, + "num_input_tokens_seen": 162344820, + "step": 7507, + "time_per_iteration": 2.7327213287353516 + }, + { + "auxiliary_loss_clip": 0.01148592, + "auxiliary_loss_mlp": 0.01027701, + "balance_loss_clip": 1.04525304, + "balance_loss_mlp": 1.01978886, + "epoch": 0.902783622918295, + "flos": 32416059536640.0, + "grad_norm": 2.5834287649441534, + "language_loss": 0.685516, + "learning_rate": 9.822425183905902e-08, + "loss": 0.70727891, + "num_input_tokens_seen": 162365345, + "step": 7508, + "time_per_iteration": 2.877720355987549 + }, + { + "auxiliary_loss_clip": 0.01039021, + "auxiliary_loss_mlp": 0.0100477, + "balance_loss_clip": 1.00626111, + "balance_loss_mlp": 1.00333345, + "epoch": 0.902903865808934, + "flos": 63717453244800.0, + "grad_norm": 0.9218300605246217, + "language_loss": 0.75128543, + "learning_rate": 9.798327668408823e-08, + "loss": 0.77172339, + "num_input_tokens_seen": 162426980, + "step": 7509, + "time_per_iteration": 3.421144485473633 + }, + { + "auxiliary_loss_clip": 0.01173608, + "auxiliary_loss_mlp": 0.01029806, + "balance_loss_clip": 1.04830909, + "balance_loss_mlp": 1.02099621, + "epoch": 0.9030241086995732, + "flos": 23804034600960.0, + "grad_norm": 2.147819222195321, + "language_loss": 0.68897307, + "learning_rate": 9.774259006354158e-08, + "loss": 0.71100724, + "num_input_tokens_seen": 162447050, + "step": 7510, + "time_per_iteration": 2.7187156677246094 + }, + { + "auxiliary_loss_clip": 0.01157396, + "auxiliary_loss_mlp": 0.0102955, + "balance_loss_clip": 1.0455066, + "balance_loss_mlp": 1.02170634, + "epoch": 0.9031443515902122, + "flos": 26395887248640.0, + "grad_norm": 2.511119311598583, + "language_loss": 0.75962579, + "learning_rate": 9.750219201393184e-08, + "loss": 0.78149521, + "num_input_tokens_seen": 162467015, + "step": 7511, + "time_per_iteration": 2.8666317462921143 + }, + { + "auxiliary_loss_clip": 0.01158965, + "auxiliary_loss_mlp": 0.01026373, + "balance_loss_clip": 1.04519355, + "balance_loss_mlp": 1.0185467, + "epoch": 0.9032645944808513, + "flos": 24939350749440.0, + "grad_norm": 1.8724123665366126, + "language_loss": 0.77720195, + "learning_rate": 9.726208257172697e-08, + "loss": 0.79905534, + "num_input_tokens_seen": 162488710, + "step": 7512, + "time_per_iteration": 2.742260217666626 + }, + { + "auxiliary_loss_clip": 0.01168985, + "auxiliary_loss_mlp": 0.01022989, + "balance_loss_clip": 1.04780054, + "balance_loss_mlp": 1.01522219, + "epoch": 0.9033848373714904, + "flos": 21178821196800.0, + "grad_norm": 1.9888923526565072, + "language_loss": 0.74866098, + "learning_rate": 9.702226177335115e-08, + "loss": 0.77058077, + "num_input_tokens_seen": 162507205, + "step": 7513, + "time_per_iteration": 2.687685489654541 + }, + { + "auxiliary_loss_clip": 0.01152526, + "auxiliary_loss_mlp": 0.01025695, + "balance_loss_clip": 1.04637456, + "balance_loss_mlp": 1.01768744, + "epoch": 0.9035050802621295, + "flos": 26286359702400.0, + "grad_norm": 1.5728317522458763, + "language_loss": 0.72259891, + "learning_rate": 9.67827296551853e-08, + "loss": 0.74438113, + "num_input_tokens_seen": 162528490, + "step": 7514, + "time_per_iteration": 2.726285696029663 + }, + { + "auxiliary_loss_clip": 0.01142167, + "auxiliary_loss_mlp": 0.00885327, + "balance_loss_clip": 1.04135144, + "balance_loss_mlp": 1.00060534, + "epoch": 0.9036253231527686, + "flos": 24204546224640.0, + "grad_norm": 2.04244448686236, + "language_loss": 0.68684936, + "learning_rate": 9.65434862535659e-08, + "loss": 0.70712423, + "num_input_tokens_seen": 162547860, + "step": 7515, + "time_per_iteration": 2.7731733322143555 + }, + { + "auxiliary_loss_clip": 0.01153844, + "auxiliary_loss_mlp": 0.01030069, + "balance_loss_clip": 1.04474926, + "balance_loss_mlp": 1.02205789, + "epoch": 0.9037455660434077, + "flos": 18072655660800.0, + "grad_norm": 2.4859361078417646, + "language_loss": 0.64953876, + "learning_rate": 9.630453160478635e-08, + "loss": 0.6713779, + "num_input_tokens_seen": 162563215, + "step": 7516, + "time_per_iteration": 3.6360421180725098 + }, + { + "auxiliary_loss_clip": 0.01131276, + "auxiliary_loss_mlp": 0.01024607, + "balance_loss_clip": 1.04064465, + "balance_loss_mlp": 1.01733816, + "epoch": 0.9038658089340468, + "flos": 24060795995520.0, + "grad_norm": 1.5806354922671284, + "language_loss": 0.82498682, + "learning_rate": 9.60658657450959e-08, + "loss": 0.84654558, + "num_input_tokens_seen": 162583515, + "step": 7517, + "time_per_iteration": 2.839529514312744 + }, + { + "auxiliary_loss_clip": 0.01137781, + "auxiliary_loss_mlp": 0.01029058, + "balance_loss_clip": 1.04016984, + "balance_loss_mlp": 1.02152419, + "epoch": 0.9039860518246858, + "flos": 21834298535040.0, + "grad_norm": 1.63779996631907, + "language_loss": 0.79484367, + "learning_rate": 9.582748871069979e-08, + "loss": 0.81651211, + "num_input_tokens_seen": 162602955, + "step": 7518, + "time_per_iteration": 3.6993725299835205 + }, + { + "auxiliary_loss_clip": 0.01149624, + "auxiliary_loss_mlp": 0.00885407, + "balance_loss_clip": 1.04169941, + "balance_loss_mlp": 1.00061798, + "epoch": 0.904106294715325, + "flos": 26614870513920.0, + "grad_norm": 1.9236908908483252, + "language_loss": 0.83203548, + "learning_rate": 9.558940053775954e-08, + "loss": 0.85238582, + "num_input_tokens_seen": 162621595, + "step": 7519, + "time_per_iteration": 3.7377543449401855 + }, + { + "auxiliary_loss_clip": 0.01158592, + "auxiliary_loss_mlp": 0.01026433, + "balance_loss_clip": 1.04546463, + "balance_loss_mlp": 1.01855922, + "epoch": 0.904226537605964, + "flos": 17785693906560.0, + "grad_norm": 1.9055289181635964, + "language_loss": 0.67977047, + "learning_rate": 9.535160126239294e-08, + "loss": 0.70162064, + "num_input_tokens_seen": 162638220, + "step": 7520, + "time_per_iteration": 2.6470720767974854 + }, + { + "auxiliary_loss_clip": 0.01159323, + "auxiliary_loss_mlp": 0.01024219, + "balance_loss_clip": 1.0472126, + "balance_loss_mlp": 1.0168457, + "epoch": 0.9043467804966031, + "flos": 24790428961920.0, + "grad_norm": 1.4841166129033654, + "language_loss": 0.70687199, + "learning_rate": 9.511409092067424e-08, + "loss": 0.72870743, + "num_input_tokens_seen": 162658575, + "step": 7521, + "time_per_iteration": 2.747518301010132 + }, + { + "auxiliary_loss_clip": 0.01152989, + "auxiliary_loss_mlp": 0.01027805, + "balance_loss_clip": 1.04754281, + "balance_loss_mlp": 1.01929331, + "epoch": 0.9044670233872423, + "flos": 22632125472000.0, + "grad_norm": 1.646753393783433, + "language_loss": 0.67247474, + "learning_rate": 9.487686954863327e-08, + "loss": 0.69428265, + "num_input_tokens_seen": 162678295, + "step": 7522, + "time_per_iteration": 2.777167320251465 + }, + { + "auxiliary_loss_clip": 0.01160076, + "auxiliary_loss_mlp": 0.01027035, + "balance_loss_clip": 1.0461278, + "balance_loss_mlp": 1.01933956, + "epoch": 0.9045872662778813, + "flos": 23771320289280.0, + "grad_norm": 2.020009304898649, + "language_loss": 0.77182412, + "learning_rate": 9.46399371822566e-08, + "loss": 0.79369527, + "num_input_tokens_seen": 162698070, + "step": 7523, + "time_per_iteration": 3.615610361099243 + }, + { + "auxiliary_loss_clip": 0.01169957, + "auxiliary_loss_mlp": 0.01025021, + "balance_loss_clip": 1.04641509, + "balance_loss_mlp": 1.01727295, + "epoch": 0.9047075091685204, + "flos": 15191039998080.0, + "grad_norm": 2.119440119815896, + "language_loss": 0.72318482, + "learning_rate": 9.440329385748657e-08, + "loss": 0.74513459, + "num_input_tokens_seen": 162715140, + "step": 7524, + "time_per_iteration": 2.659951686859131 + }, + { + "auxiliary_loss_clip": 0.0113959, + "auxiliary_loss_mlp": 0.01025018, + "balance_loss_clip": 1.04375219, + "balance_loss_mlp": 1.01810992, + "epoch": 0.9048277520591596, + "flos": 18003707504640.0, + "grad_norm": 1.68612511157763, + "language_loss": 0.70335054, + "learning_rate": 9.416693961022137e-08, + "loss": 0.72499663, + "num_input_tokens_seen": 162733390, + "step": 7525, + "time_per_iteration": 2.751079559326172 + }, + { + "auxiliary_loss_clip": 0.01106392, + "auxiliary_loss_mlp": 0.01027443, + "balance_loss_clip": 1.03671265, + "balance_loss_mlp": 1.01987934, + "epoch": 0.9049479949497986, + "flos": 21872471713920.0, + "grad_norm": 2.1491186412762886, + "language_loss": 0.77113974, + "learning_rate": 9.393087447631654e-08, + "loss": 0.79247808, + "num_input_tokens_seen": 162751670, + "step": 7526, + "time_per_iteration": 2.87345552444458 + }, + { + "auxiliary_loss_clip": 0.01143436, + "auxiliary_loss_mlp": 0.01023674, + "balance_loss_clip": 1.04370975, + "balance_loss_mlp": 1.01628006, + "epoch": 0.9050682378404377, + "flos": 20773928113920.0, + "grad_norm": 1.786672769905981, + "language_loss": 0.72982168, + "learning_rate": 9.36950984915823e-08, + "loss": 0.75149274, + "num_input_tokens_seen": 162770025, + "step": 7527, + "time_per_iteration": 2.6812379360198975 + }, + { + "auxiliary_loss_clip": 0.01172061, + "auxiliary_loss_mlp": 0.01026874, + "balance_loss_clip": 1.04949927, + "balance_loss_mlp": 1.01892281, + "epoch": 0.9051884807310768, + "flos": 21580015178880.0, + "grad_norm": 2.444938729009153, + "language_loss": 0.69088101, + "learning_rate": 9.345961169178607e-08, + "loss": 0.71287036, + "num_input_tokens_seen": 162789710, + "step": 7528, + "time_per_iteration": 2.638920307159424 + }, + { + "auxiliary_loss_clip": 0.01129821, + "auxiliary_loss_mlp": 0.01029126, + "balance_loss_clip": 1.04533768, + "balance_loss_mlp": 1.02169645, + "epoch": 0.9053087236217159, + "flos": 21908059113600.0, + "grad_norm": 1.4701044719272622, + "language_loss": 0.72870553, + "learning_rate": 9.322441411265081e-08, + "loss": 0.75029504, + "num_input_tokens_seen": 162810695, + "step": 7529, + "time_per_iteration": 2.7386748790740967 + }, + { + "auxiliary_loss_clip": 0.01144851, + "auxiliary_loss_mlp": 0.01031866, + "balance_loss_clip": 1.04487634, + "balance_loss_mlp": 1.02396798, + "epoch": 0.9054289665123549, + "flos": 17055809544960.0, + "grad_norm": 2.067442927038434, + "language_loss": 0.73187149, + "learning_rate": 9.298950578985554e-08, + "loss": 0.75363863, + "num_input_tokens_seen": 162827770, + "step": 7530, + "time_per_iteration": 2.8750574588775635 + }, + { + "auxiliary_loss_clip": 0.01159729, + "auxiliary_loss_mlp": 0.00886044, + "balance_loss_clip": 1.04888654, + "balance_loss_mlp": 1.00054324, + "epoch": 0.905549209402994, + "flos": 20777268078720.0, + "grad_norm": 2.7305441311698835, + "language_loss": 0.70852762, + "learning_rate": 9.275488675903665e-08, + "loss": 0.72898543, + "num_input_tokens_seen": 162846715, + "step": 7531, + "time_per_iteration": 2.6818878650665283 + }, + { + "auxiliary_loss_clip": 0.01126389, + "auxiliary_loss_mlp": 0.01025989, + "balance_loss_clip": 1.04069006, + "balance_loss_mlp": 1.01778746, + "epoch": 0.9056694522936332, + "flos": 21686813291520.0, + "grad_norm": 2.750663100565317, + "language_loss": 0.74202776, + "learning_rate": 9.252055705578454e-08, + "loss": 0.76355153, + "num_input_tokens_seen": 162866215, + "step": 7532, + "time_per_iteration": 2.825450897216797 + }, + { + "auxiliary_loss_clip": 0.01158334, + "auxiliary_loss_mlp": 0.01025878, + "balance_loss_clip": 1.04456568, + "balance_loss_mlp": 1.01896048, + "epoch": 0.9057896951842722, + "flos": 29569133433600.0, + "grad_norm": 2.088177747329651, + "language_loss": 0.71857285, + "learning_rate": 9.228651671564747e-08, + "loss": 0.74041498, + "num_input_tokens_seen": 162888245, + "step": 7533, + "time_per_iteration": 2.7940621376037598 + }, + { + "auxiliary_loss_clip": 0.01127883, + "auxiliary_loss_mlp": 0.01033773, + "balance_loss_clip": 1.04296565, + "balance_loss_mlp": 1.02610183, + "epoch": 0.9059099380749113, + "flos": 27892248952320.0, + "grad_norm": 1.4674082134717936, + "language_loss": 0.77804422, + "learning_rate": 9.205276577412901e-08, + "loss": 0.7996608, + "num_input_tokens_seen": 162911025, + "step": 7534, + "time_per_iteration": 2.836517810821533 + }, + { + "auxiliary_loss_clip": 0.01158375, + "auxiliary_loss_mlp": 0.00886731, + "balance_loss_clip": 1.04513383, + "balance_loss_mlp": 1.00054622, + "epoch": 0.9060301809655504, + "flos": 17748993185280.0, + "grad_norm": 3.0574785207901924, + "language_loss": 0.77112615, + "learning_rate": 9.181930426668905e-08, + "loss": 0.79157722, + "num_input_tokens_seen": 162927820, + "step": 7535, + "time_per_iteration": 2.962538480758667 + }, + { + "auxiliary_loss_clip": 0.01115504, + "auxiliary_loss_mlp": 0.01030625, + "balance_loss_clip": 1.03993142, + "balance_loss_mlp": 1.02273679, + "epoch": 0.9061504238561895, + "flos": 31759432963200.0, + "grad_norm": 1.5305830591136538, + "language_loss": 0.67889094, + "learning_rate": 9.158613222874346e-08, + "loss": 0.70035219, + "num_input_tokens_seen": 162949445, + "step": 7536, + "time_per_iteration": 2.960496425628662 + }, + { + "auxiliary_loss_clip": 0.01148542, + "auxiliary_loss_mlp": 0.01026625, + "balance_loss_clip": 1.04343534, + "balance_loss_mlp": 1.01866734, + "epoch": 0.9062706667468285, + "flos": 20048066075520.0, + "grad_norm": 1.7247998614854472, + "language_loss": 0.8215549, + "learning_rate": 9.135324969566394e-08, + "loss": 0.84330654, + "num_input_tokens_seen": 162968945, + "step": 7537, + "time_per_iteration": 2.836319923400879 + }, + { + "auxiliary_loss_clip": 0.01165093, + "auxiliary_loss_mlp": 0.01030131, + "balance_loss_clip": 1.04709506, + "balance_loss_mlp": 1.0224452, + "epoch": 0.9063909096374677, + "flos": 18437292576000.0, + "grad_norm": 1.9818950974010168, + "language_loss": 0.75566053, + "learning_rate": 9.112065670277913e-08, + "loss": 0.77761281, + "num_input_tokens_seen": 162985310, + "step": 7538, + "time_per_iteration": 2.6593337059020996 + }, + { + "auxiliary_loss_clip": 0.01142428, + "auxiliary_loss_mlp": 0.01023147, + "balance_loss_clip": 1.03976142, + "balance_loss_mlp": 1.01575565, + "epoch": 0.9065111525281068, + "flos": 33547353361920.0, + "grad_norm": 1.7078313140317223, + "language_loss": 0.7293011, + "learning_rate": 9.088835328537303e-08, + "loss": 0.75095689, + "num_input_tokens_seen": 163006900, + "step": 7539, + "time_per_iteration": 2.8551876544952393 + }, + { + "auxiliary_loss_clip": 0.01153021, + "auxiliary_loss_mlp": 0.01027132, + "balance_loss_clip": 1.04531372, + "balance_loss_mlp": 1.01965737, + "epoch": 0.9066313954187458, + "flos": 23367863750400.0, + "grad_norm": 2.920843409178114, + "language_loss": 0.71698529, + "learning_rate": 9.065633947868568e-08, + "loss": 0.73878682, + "num_input_tokens_seen": 163026505, + "step": 7540, + "time_per_iteration": 2.8026504516601562 + }, + { + "auxiliary_loss_clip": 0.01140537, + "auxiliary_loss_mlp": 0.00885406, + "balance_loss_clip": 1.04640388, + "balance_loss_mlp": 1.00056887, + "epoch": 0.906751638309385, + "flos": 26249623067520.0, + "grad_norm": 2.8837156421588714, + "language_loss": 0.80043447, + "learning_rate": 9.042461531791379e-08, + "loss": 0.82069391, + "num_input_tokens_seen": 163044925, + "step": 7541, + "time_per_iteration": 2.96232008934021 + }, + { + "auxiliary_loss_clip": 0.01164687, + "auxiliary_loss_mlp": 0.01028332, + "balance_loss_clip": 1.0450865, + "balance_loss_mlp": 1.02074444, + "epoch": 0.906871881200024, + "flos": 16544477485440.0, + "grad_norm": 1.743170585772125, + "language_loss": 0.78174222, + "learning_rate": 9.019318083820903e-08, + "loss": 0.80367237, + "num_input_tokens_seen": 163063505, + "step": 7542, + "time_per_iteration": 3.8776087760925293 + }, + { + "auxiliary_loss_clip": 0.01156859, + "auxiliary_loss_mlp": 0.01028874, + "balance_loss_clip": 1.04507029, + "balance_loss_mlp": 1.02097607, + "epoch": 0.9069921240906631, + "flos": 24605129675520.0, + "grad_norm": 1.713770256497396, + "language_loss": 0.85503006, + "learning_rate": 8.996203607468045e-08, + "loss": 0.87688738, + "num_input_tokens_seen": 163082505, + "step": 7543, + "time_per_iteration": 2.737527370452881 + }, + { + "auxiliary_loss_clip": 0.01153772, + "auxiliary_loss_mlp": 0.01025704, + "balance_loss_clip": 1.04292512, + "balance_loss_mlp": 1.01771688, + "epoch": 0.9071123669813023, + "flos": 25374731500800.0, + "grad_norm": 2.5532714953013618, + "language_loss": 0.75331843, + "learning_rate": 8.973118106239241e-08, + "loss": 0.77511322, + "num_input_tokens_seen": 163105110, + "step": 7544, + "time_per_iteration": 3.6878821849823 + }, + { + "auxiliary_loss_clip": 0.01119753, + "auxiliary_loss_mlp": 0.01027814, + "balance_loss_clip": 1.03681946, + "balance_loss_mlp": 1.02039289, + "epoch": 0.9072326098719413, + "flos": 26725798690560.0, + "grad_norm": 2.393808120565819, + "language_loss": 0.94581574, + "learning_rate": 8.95006158363656e-08, + "loss": 0.96729141, + "num_input_tokens_seen": 163125295, + "step": 7545, + "time_per_iteration": 3.8781516551971436 + }, + { + "auxiliary_loss_clip": 0.01159943, + "auxiliary_loss_mlp": 0.01024026, + "balance_loss_clip": 1.04826427, + "balance_loss_mlp": 1.01613998, + "epoch": 0.9073528527625804, + "flos": 23878800760320.0, + "grad_norm": 1.9478034707147536, + "language_loss": 0.77614748, + "learning_rate": 8.9270340431576e-08, + "loss": 0.79798716, + "num_input_tokens_seen": 163144385, + "step": 7546, + "time_per_iteration": 2.80607008934021 + }, + { + "auxiliary_loss_clip": 0.01159514, + "auxiliary_loss_mlp": 0.01025246, + "balance_loss_clip": 1.04436326, + "balance_loss_mlp": 1.01733649, + "epoch": 0.9074730956532195, + "flos": 37852144767360.0, + "grad_norm": 2.3562054578920555, + "language_loss": 0.73723179, + "learning_rate": 8.904035488295658e-08, + "loss": 0.7590794, + "num_input_tokens_seen": 163163885, + "step": 7547, + "time_per_iteration": 3.0574357509613037 + }, + { + "auxiliary_loss_clip": 0.01049229, + "auxiliary_loss_mlp": 0.00876834, + "balance_loss_clip": 1.00417137, + "balance_loss_mlp": 1.00121832, + "epoch": 0.9075933385438586, + "flos": 65173307385600.0, + "grad_norm": 0.6635790110353315, + "language_loss": 0.53254402, + "learning_rate": 8.881065922539632e-08, + "loss": 0.55180466, + "num_input_tokens_seen": 163224325, + "step": 7548, + "time_per_iteration": 3.2547686100006104 + }, + { + "auxiliary_loss_clip": 0.01135303, + "auxiliary_loss_mlp": 0.0102945, + "balance_loss_clip": 1.04337096, + "balance_loss_mlp": 1.0215044, + "epoch": 0.9077135814344977, + "flos": 19931571290880.0, + "grad_norm": 1.7474796648226327, + "language_loss": 0.73624694, + "learning_rate": 8.85812534937389e-08, + "loss": 0.75789446, + "num_input_tokens_seen": 163242425, + "step": 7549, + "time_per_iteration": 3.8441240787506104 + }, + { + "auxiliary_loss_clip": 0.01157992, + "auxiliary_loss_mlp": 0.01029614, + "balance_loss_clip": 1.04785132, + "balance_loss_mlp": 1.02162099, + "epoch": 0.9078338243251368, + "flos": 17529650784000.0, + "grad_norm": 2.7393730837350296, + "language_loss": 0.67368215, + "learning_rate": 8.835213772278583e-08, + "loss": 0.69555819, + "num_input_tokens_seen": 163259280, + "step": 7550, + "time_per_iteration": 2.6822147369384766 + }, + { + "auxiliary_loss_clip": 0.01128831, + "auxiliary_loss_mlp": 0.01020726, + "balance_loss_clip": 1.04275548, + "balance_loss_mlp": 1.01314449, + "epoch": 0.9079540672157759, + "flos": 28803410277120.0, + "grad_norm": 1.6691641975569438, + "language_loss": 0.78814006, + "learning_rate": 8.812331194729373e-08, + "loss": 0.80963564, + "num_input_tokens_seen": 163278925, + "step": 7551, + "time_per_iteration": 2.8605475425720215 + }, + { + "auxiliary_loss_clip": 0.0117558, + "auxiliary_loss_mlp": 0.01030127, + "balance_loss_clip": 1.05192304, + "balance_loss_mlp": 1.02180028, + "epoch": 0.9080743101064149, + "flos": 23513840622720.0, + "grad_norm": 1.9204821879795453, + "language_loss": 0.71796811, + "learning_rate": 8.789477620197461e-08, + "loss": 0.74002516, + "num_input_tokens_seen": 163298450, + "step": 7552, + "time_per_iteration": 2.59798002243042 + }, + { + "auxiliary_loss_clip": 0.01149454, + "auxiliary_loss_mlp": 0.01029853, + "balance_loss_clip": 1.0448637, + "balance_loss_mlp": 1.02137744, + "epoch": 0.9081945529970541, + "flos": 22778102344320.0, + "grad_norm": 2.064270220831766, + "language_loss": 0.79278892, + "learning_rate": 8.766653052149831e-08, + "loss": 0.81458199, + "num_input_tokens_seen": 163313635, + "step": 7553, + "time_per_iteration": 2.742933750152588 + }, + { + "auxiliary_loss_clip": 0.0114789, + "auxiliary_loss_mlp": 0.01027898, + "balance_loss_clip": 1.04455817, + "balance_loss_mlp": 1.02019072, + "epoch": 0.9083147958876931, + "flos": 18873714821760.0, + "grad_norm": 2.3019417003030687, + "language_loss": 0.74614006, + "learning_rate": 8.743857494048823e-08, + "loss": 0.76789796, + "num_input_tokens_seen": 163330450, + "step": 7554, + "time_per_iteration": 2.7899789810180664 + }, + { + "auxiliary_loss_clip": 0.01140618, + "auxiliary_loss_mlp": 0.01038285, + "balance_loss_clip": 1.04459882, + "balance_loss_mlp": 1.03043818, + "epoch": 0.9084350387783322, + "flos": 18909374048640.0, + "grad_norm": 2.493526439653506, + "language_loss": 0.63100725, + "learning_rate": 8.721090949352605e-08, + "loss": 0.65279639, + "num_input_tokens_seen": 163346690, + "step": 7555, + "time_per_iteration": 2.833448886871338 + }, + { + "auxiliary_loss_clip": 0.0116313, + "auxiliary_loss_mlp": 0.0102703, + "balance_loss_clip": 1.05011201, + "balance_loss_mlp": 1.01910305, + "epoch": 0.9085552816689714, + "flos": 20595488325120.0, + "grad_norm": 1.8376484892330098, + "language_loss": 0.7301662, + "learning_rate": 8.698353421514793e-08, + "loss": 0.7520678, + "num_input_tokens_seen": 163365065, + "step": 7556, + "time_per_iteration": 2.6535840034484863 + }, + { + "auxiliary_loss_clip": 0.011599, + "auxiliary_loss_mlp": 0.01028738, + "balance_loss_clip": 1.0465765, + "balance_loss_mlp": 1.02103114, + "epoch": 0.9086755245596104, + "flos": 18113163223680.0, + "grad_norm": 2.328021862172087, + "language_loss": 0.80255163, + "learning_rate": 8.67564491398467e-08, + "loss": 0.82443798, + "num_input_tokens_seen": 163382070, + "step": 7557, + "time_per_iteration": 2.685107707977295 + }, + { + "auxiliary_loss_clip": 0.01161509, + "auxiliary_loss_mlp": 0.01031358, + "balance_loss_clip": 1.04555273, + "balance_loss_mlp": 1.02360368, + "epoch": 0.9087957674502495, + "flos": 19129793857920.0, + "grad_norm": 1.7156058753559482, + "language_loss": 0.7364167, + "learning_rate": 8.652965430207104e-08, + "loss": 0.75834537, + "num_input_tokens_seen": 163399975, + "step": 7558, + "time_per_iteration": 2.793827772140503 + }, + { + "auxiliary_loss_clip": 0.01162777, + "auxiliary_loss_mlp": 0.01032466, + "balance_loss_clip": 1.04492557, + "balance_loss_mlp": 1.02446151, + "epoch": 0.9089160103408886, + "flos": 18109930999680.0, + "grad_norm": 2.4046591185853754, + "language_loss": 0.65597749, + "learning_rate": 8.630314973622521e-08, + "loss": 0.67792988, + "num_input_tokens_seen": 163417520, + "step": 7559, + "time_per_iteration": 2.8920481204986572 + }, + { + "auxiliary_loss_clip": 0.01155757, + "auxiliary_loss_mlp": 0.01032344, + "balance_loss_clip": 1.0463953, + "balance_loss_mlp": 1.02535534, + "epoch": 0.9090362532315277, + "flos": 33364855336320.0, + "grad_norm": 2.0280960244952673, + "language_loss": 0.71182215, + "learning_rate": 8.607693547666995e-08, + "loss": 0.73370314, + "num_input_tokens_seen": 163440060, + "step": 7560, + "time_per_iteration": 2.859787702560425 + }, + { + "auxiliary_loss_clip": 0.01042658, + "auxiliary_loss_mlp": 0.01001808, + "balance_loss_clip": 1.00691056, + "balance_loss_mlp": 1.00032985, + "epoch": 0.9091564961221668, + "flos": 71480585082240.0, + "grad_norm": 0.8876626764782175, + "language_loss": 0.57949471, + "learning_rate": 8.585101155772201e-08, + "loss": 0.59993935, + "num_input_tokens_seen": 163502180, + "step": 7561, + "time_per_iteration": 3.390439748764038 + }, + { + "auxiliary_loss_clip": 0.0114268, + "auxiliary_loss_mlp": 0.0102495, + "balance_loss_clip": 1.04051447, + "balance_loss_mlp": 1.01646245, + "epoch": 0.9092767390128058, + "flos": 24712574232960.0, + "grad_norm": 1.802217021111896, + "language_loss": 0.68349671, + "learning_rate": 8.562537801365377e-08, + "loss": 0.70517302, + "num_input_tokens_seen": 163521915, + "step": 7562, + "time_per_iteration": 2.815396308898926 + }, + { + "auxiliary_loss_clip": 0.01170857, + "auxiliary_loss_mlp": 0.01030056, + "balance_loss_clip": 1.04829931, + "balance_loss_mlp": 1.02158606, + "epoch": 0.909396981903445, + "flos": 23586487879680.0, + "grad_norm": 1.9012495179008424, + "language_loss": 0.69620788, + "learning_rate": 8.540003487869362e-08, + "loss": 0.71821702, + "num_input_tokens_seen": 163543585, + "step": 7563, + "time_per_iteration": 2.8141374588012695 + }, + { + "auxiliary_loss_clip": 0.01132417, + "auxiliary_loss_mlp": 0.01023916, + "balance_loss_clip": 1.04174376, + "balance_loss_mlp": 1.01638818, + "epoch": 0.909517224794084, + "flos": 23404169422080.0, + "grad_norm": 2.1803507096836, + "language_loss": 0.79642695, + "learning_rate": 8.517498218702557e-08, + "loss": 0.8179903, + "num_input_tokens_seen": 163561515, + "step": 7564, + "time_per_iteration": 2.8216209411621094 + }, + { + "auxiliary_loss_clip": 0.01133713, + "auxiliary_loss_mlp": 0.01023465, + "balance_loss_clip": 1.04039252, + "balance_loss_mlp": 1.01540089, + "epoch": 0.9096374676847231, + "flos": 19208618254080.0, + "grad_norm": 1.761900111128631, + "language_loss": 0.69383389, + "learning_rate": 8.49502199727905e-08, + "loss": 0.7154057, + "num_input_tokens_seen": 163579540, + "step": 7565, + "time_per_iteration": 2.746356964111328 + }, + { + "auxiliary_loss_clip": 0.01155815, + "auxiliary_loss_mlp": 0.01027457, + "balance_loss_clip": 1.04310894, + "balance_loss_mlp": 1.01955914, + "epoch": 0.9097577105753623, + "flos": 33292495388160.0, + "grad_norm": 2.0610487844055614, + "language_loss": 0.66487312, + "learning_rate": 8.472574827008428e-08, + "loss": 0.68670583, + "num_input_tokens_seen": 163600425, + "step": 7566, + "time_per_iteration": 2.7681055068969727 + }, + { + "auxiliary_loss_clip": 0.01158038, + "auxiliary_loss_mlp": 0.01023679, + "balance_loss_clip": 1.04380679, + "balance_loss_mlp": 1.0157876, + "epoch": 0.9098779534660013, + "flos": 21906443001600.0, + "grad_norm": 1.7291507128490946, + "language_loss": 0.83994627, + "learning_rate": 8.450156711295942e-08, + "loss": 0.86176348, + "num_input_tokens_seen": 163620595, + "step": 7567, + "time_per_iteration": 2.720520496368408 + }, + { + "auxiliary_loss_clip": 0.01147368, + "auxiliary_loss_mlp": 0.01027842, + "balance_loss_clip": 1.04711592, + "balance_loss_mlp": 1.02041507, + "epoch": 0.9099981963566404, + "flos": 25730354102400.0, + "grad_norm": 2.160514764676082, + "language_loss": 0.86716294, + "learning_rate": 8.427767653542383e-08, + "loss": 0.888915, + "num_input_tokens_seen": 163635765, + "step": 7568, + "time_per_iteration": 3.626923084259033 + }, + { + "auxiliary_loss_clip": 0.01126019, + "auxiliary_loss_mlp": 0.0102303, + "balance_loss_clip": 1.03980374, + "balance_loss_mlp": 1.01546311, + "epoch": 0.9101184392472795, + "flos": 21069437304960.0, + "grad_norm": 1.8438404394946872, + "language_loss": 0.70094705, + "learning_rate": 8.405407657144125e-08, + "loss": 0.7224375, + "num_input_tokens_seen": 163654925, + "step": 7569, + "time_per_iteration": 2.7909719944000244 + }, + { + "auxiliary_loss_clip": 0.01141761, + "auxiliary_loss_mlp": 0.01025927, + "balance_loss_clip": 1.04182363, + "balance_loss_mlp": 1.01884842, + "epoch": 0.9102386821379186, + "flos": 24752614919040.0, + "grad_norm": 1.868173786709701, + "language_loss": 0.7278415, + "learning_rate": 8.383076725493232e-08, + "loss": 0.74951839, + "num_input_tokens_seen": 163672245, + "step": 7570, + "time_per_iteration": 3.6518983840942383 + }, + { + "auxiliary_loss_clip": 0.011598, + "auxiliary_loss_mlp": 0.01026697, + "balance_loss_clip": 1.04609323, + "balance_loss_mlp": 1.01906729, + "epoch": 0.9103589250285576, + "flos": 22562818179840.0, + "grad_norm": 2.035050035810512, + "language_loss": 0.67821169, + "learning_rate": 8.360774861977216e-08, + "loss": 0.7000767, + "num_input_tokens_seen": 163691365, + "step": 7571, + "time_per_iteration": 3.671008586883545 + }, + { + "auxiliary_loss_clip": 0.01148656, + "auxiliary_loss_mlp": 0.01026607, + "balance_loss_clip": 1.04097414, + "balance_loss_mlp": 1.01893938, + "epoch": 0.9104791679191968, + "flos": 25373474524800.0, + "grad_norm": 1.7707466939249055, + "language_loss": 0.7463392, + "learning_rate": 8.338502069979281e-08, + "loss": 0.7680918, + "num_input_tokens_seen": 163711675, + "step": 7572, + "time_per_iteration": 2.9504780769348145 + }, + { + "auxiliary_loss_clip": 0.01161237, + "auxiliary_loss_mlp": 0.01029684, + "balance_loss_clip": 1.0440805, + "balance_loss_mlp": 1.0221262, + "epoch": 0.9105994108098359, + "flos": 14426681558400.0, + "grad_norm": 2.7154924262919264, + "language_loss": 0.79826397, + "learning_rate": 8.316258352878214e-08, + "loss": 0.82017314, + "num_input_tokens_seen": 163728095, + "step": 7573, + "time_per_iteration": 2.736647129058838 + }, + { + "auxiliary_loss_clip": 0.01163855, + "auxiliary_loss_mlp": 0.01027814, + "balance_loss_clip": 1.04563749, + "balance_loss_mlp": 1.01952875, + "epoch": 0.9107196537004749, + "flos": 26718292748160.0, + "grad_norm": 6.555952851001472, + "language_loss": 0.70929766, + "learning_rate": 8.294043714048338e-08, + "loss": 0.73121434, + "num_input_tokens_seen": 163747175, + "step": 7574, + "time_per_iteration": 2.7025394439697266 + }, + { + "auxiliary_loss_clip": 0.01044603, + "auxiliary_loss_mlp": 0.01002044, + "balance_loss_clip": 1.00535583, + "balance_loss_mlp": 1.00059593, + "epoch": 0.9108398965911141, + "flos": 66532634703360.0, + "grad_norm": 0.7478814327772523, + "language_loss": 0.60483366, + "learning_rate": 8.271858156859624e-08, + "loss": 0.62530011, + "num_input_tokens_seen": 163812545, + "step": 7575, + "time_per_iteration": 4.162235736846924 + }, + { + "auxiliary_loss_clip": 0.01166104, + "auxiliary_loss_mlp": 0.01025891, + "balance_loss_clip": 1.04578424, + "balance_loss_mlp": 1.01825535, + "epoch": 0.9109601394817531, + "flos": 25411073086080.0, + "grad_norm": 1.659329175359307, + "language_loss": 0.73792911, + "learning_rate": 8.249701684677557e-08, + "loss": 0.75984907, + "num_input_tokens_seen": 163833870, + "step": 7576, + "time_per_iteration": 2.683530807495117 + }, + { + "auxiliary_loss_clip": 0.0116005, + "auxiliary_loss_mlp": 0.01027823, + "balance_loss_clip": 1.04861975, + "balance_loss_mlp": 1.02058077, + "epoch": 0.9110803823723922, + "flos": 22747794243840.0, + "grad_norm": 1.8303355946877655, + "language_loss": 0.80804861, + "learning_rate": 8.227574300863294e-08, + "loss": 0.82992733, + "num_input_tokens_seen": 163854040, + "step": 7577, + "time_per_iteration": 2.698458194732666 + }, + { + "auxiliary_loss_clip": 0.01156286, + "auxiliary_loss_mlp": 0.0102972, + "balance_loss_clip": 1.04785454, + "balance_loss_mlp": 1.02154779, + "epoch": 0.9112006252630314, + "flos": 48469924131840.0, + "grad_norm": 1.89512697989017, + "language_loss": 0.69834942, + "learning_rate": 8.205476008773548e-08, + "loss": 0.72020948, + "num_input_tokens_seen": 163878040, + "step": 7578, + "time_per_iteration": 3.0282223224639893 + }, + { + "auxiliary_loss_clip": 0.01134625, + "auxiliary_loss_mlp": 0.01027058, + "balance_loss_clip": 1.04586244, + "balance_loss_mlp": 1.0185051, + "epoch": 0.9113208681536704, + "flos": 30009649829760.0, + "grad_norm": 2.057322389083467, + "language_loss": 0.82253087, + "learning_rate": 8.183406811760596e-08, + "loss": 0.84414774, + "num_input_tokens_seen": 163897770, + "step": 7579, + "time_per_iteration": 2.856651782989502 + }, + { + "auxiliary_loss_clip": 0.01126041, + "auxiliary_loss_mlp": 0.010276, + "balance_loss_clip": 1.04023015, + "balance_loss_mlp": 1.01973796, + "epoch": 0.9114411110443095, + "flos": 25594971742080.0, + "grad_norm": 1.4543993578190235, + "language_loss": 0.74052179, + "learning_rate": 8.161366713172313e-08, + "loss": 0.76205814, + "num_input_tokens_seen": 163920160, + "step": 7580, + "time_per_iteration": 2.880746364593506 + }, + { + "auxiliary_loss_clip": 0.01144319, + "auxiliary_loss_mlp": 0.01028762, + "balance_loss_clip": 1.04196286, + "balance_loss_mlp": 1.02025044, + "epoch": 0.9115613539349486, + "flos": 18399729928320.0, + "grad_norm": 2.5091988558145415, + "language_loss": 0.83963764, + "learning_rate": 8.139355716352137e-08, + "loss": 0.86136842, + "num_input_tokens_seen": 163935000, + "step": 7581, + "time_per_iteration": 2.778510570526123 + }, + { + "auxiliary_loss_clip": 0.01144616, + "auxiliary_loss_mlp": 0.01024626, + "balance_loss_clip": 1.04349959, + "balance_loss_mlp": 1.01660299, + "epoch": 0.9116815968255877, + "flos": 21726171619200.0, + "grad_norm": 1.7886343260980087, + "language_loss": 0.69880861, + "learning_rate": 8.117373824639196e-08, + "loss": 0.72050107, + "num_input_tokens_seen": 163955265, + "step": 7582, + "time_per_iteration": 2.80157208442688 + }, + { + "auxiliary_loss_clip": 0.01055029, + "auxiliary_loss_mlp": 0.01000819, + "balance_loss_clip": 1.00452828, + "balance_loss_mlp": 0.99940604, + "epoch": 0.9118018397162267, + "flos": 65363526835200.0, + "grad_norm": 0.7243789428512247, + "language_loss": 0.5926522, + "learning_rate": 8.095421041368067e-08, + "loss": 0.61321068, + "num_input_tokens_seen": 164014680, + "step": 7583, + "time_per_iteration": 3.1299455165863037 + }, + { + "auxiliary_loss_clip": 0.01147301, + "auxiliary_loss_mlp": 0.0088565, + "balance_loss_clip": 1.04562926, + "balance_loss_mlp": 1.00058627, + "epoch": 0.9119220826068659, + "flos": 20922885815040.0, + "grad_norm": 1.9711179977896176, + "language_loss": 0.70755219, + "learning_rate": 8.073497369868999e-08, + "loss": 0.72788167, + "num_input_tokens_seen": 164033140, + "step": 7584, + "time_per_iteration": 2.832791805267334 + }, + { + "auxiliary_loss_clip": 0.01156876, + "auxiliary_loss_mlp": 0.01028655, + "balance_loss_clip": 1.04472387, + "balance_loss_mlp": 1.0203228, + "epoch": 0.912042325497505, + "flos": 28366449327360.0, + "grad_norm": 1.811711312027086, + "language_loss": 0.75393891, + "learning_rate": 8.051602813467772e-08, + "loss": 0.77579415, + "num_input_tokens_seen": 164054995, + "step": 7585, + "time_per_iteration": 2.874178409576416 + }, + { + "auxiliary_loss_clip": 0.01162035, + "auxiliary_loss_mlp": 0.0102949, + "balance_loss_clip": 1.04618657, + "balance_loss_mlp": 1.02183676, + "epoch": 0.912162568388144, + "flos": 17566782468480.0, + "grad_norm": 1.7751457012340135, + "language_loss": 0.70863062, + "learning_rate": 8.029737375485756e-08, + "loss": 0.73054588, + "num_input_tokens_seen": 164074225, + "step": 7586, + "time_per_iteration": 2.731900930404663 + }, + { + "auxiliary_loss_clip": 0.01168007, + "auxiliary_loss_mlp": 0.01026199, + "balance_loss_clip": 1.04577637, + "balance_loss_mlp": 1.0184921, + "epoch": 0.9122828112787832, + "flos": 19827897661440.0, + "grad_norm": 1.689468137742601, + "language_loss": 0.72164714, + "learning_rate": 8.007901059239986e-08, + "loss": 0.74358916, + "num_input_tokens_seen": 164093505, + "step": 7587, + "time_per_iteration": 2.635099411010742 + }, + { + "auxiliary_loss_clip": 0.01149546, + "auxiliary_loss_mlp": 0.01023725, + "balance_loss_clip": 1.04297996, + "balance_loss_mlp": 1.01639962, + "epoch": 0.9124030541694222, + "flos": 20813789232000.0, + "grad_norm": 1.5888994181757061, + "language_loss": 0.80129611, + "learning_rate": 7.986093868042964e-08, + "loss": 0.8230288, + "num_input_tokens_seen": 164113750, + "step": 7588, + "time_per_iteration": 2.781120777130127 + }, + { + "auxiliary_loss_clip": 0.01154502, + "auxiliary_loss_mlp": 0.01028502, + "balance_loss_clip": 1.04381657, + "balance_loss_mlp": 1.02112865, + "epoch": 0.9125232970600613, + "flos": 25192305302400.0, + "grad_norm": 1.6924259167220614, + "language_loss": 0.67706418, + "learning_rate": 7.964315805202826e-08, + "loss": 0.69889426, + "num_input_tokens_seen": 164134330, + "step": 7589, + "time_per_iteration": 2.7535836696624756 + }, + { + "auxiliary_loss_clip": 0.01149705, + "auxiliary_loss_mlp": 0.01027098, + "balance_loss_clip": 1.04545903, + "balance_loss_mlp": 1.01943874, + "epoch": 0.9126435399507005, + "flos": 19719591177600.0, + "grad_norm": 1.7275312846705237, + "language_loss": 0.73417974, + "learning_rate": 7.942566874023304e-08, + "loss": 0.75594783, + "num_input_tokens_seen": 164153515, + "step": 7590, + "time_per_iteration": 2.8775665760040283 + }, + { + "auxiliary_loss_clip": 0.01138825, + "auxiliary_loss_mlp": 0.01027467, + "balance_loss_clip": 1.04263818, + "balance_loss_mlp": 1.01974249, + "epoch": 0.9127637828413395, + "flos": 19573614305280.0, + "grad_norm": 3.044279290657896, + "language_loss": 0.69857752, + "learning_rate": 7.920847077803649e-08, + "loss": 0.72024047, + "num_input_tokens_seen": 164171305, + "step": 7591, + "time_per_iteration": 2.7190990447998047 + }, + { + "auxiliary_loss_clip": 0.0111779, + "auxiliary_loss_mlp": 0.01027382, + "balance_loss_clip": 1.0361222, + "balance_loss_mlp": 1.01929414, + "epoch": 0.9128840257319786, + "flos": 20230635928320.0, + "grad_norm": 1.7952124820750552, + "language_loss": 0.82387173, + "learning_rate": 7.899156419838826e-08, + "loss": 0.84532344, + "num_input_tokens_seen": 164190275, + "step": 7592, + "time_per_iteration": 2.8210487365722656 + }, + { + "auxiliary_loss_clip": 0.01139937, + "auxiliary_loss_mlp": 0.01031785, + "balance_loss_clip": 1.04429996, + "balance_loss_mlp": 1.02380955, + "epoch": 0.9130042686226177, + "flos": 24858658846080.0, + "grad_norm": 2.1900045173382643, + "language_loss": 0.65493232, + "learning_rate": 7.87749490341918e-08, + "loss": 0.67664957, + "num_input_tokens_seen": 164210550, + "step": 7593, + "time_per_iteration": 2.8279316425323486 + }, + { + "auxiliary_loss_clip": 0.01172048, + "auxiliary_loss_mlp": 0.01026131, + "balance_loss_clip": 1.0490644, + "balance_loss_mlp": 1.0177809, + "epoch": 0.9131245115132568, + "flos": 23581747284480.0, + "grad_norm": 1.9415711585054864, + "language_loss": 0.83286846, + "learning_rate": 7.855862531830836e-08, + "loss": 0.85485023, + "num_input_tokens_seen": 164226660, + "step": 7594, + "time_per_iteration": 3.8647162914276123 + }, + { + "auxiliary_loss_clip": 0.01158032, + "auxiliary_loss_mlp": 0.01024236, + "balance_loss_clip": 1.04499125, + "balance_loss_mlp": 1.01620173, + "epoch": 0.9132447544038959, + "flos": 19931607204480.0, + "grad_norm": 1.624228396448796, + "language_loss": 0.72443366, + "learning_rate": 7.834259308355373e-08, + "loss": 0.74625635, + "num_input_tokens_seen": 164245425, + "step": 7595, + "time_per_iteration": 3.693951368331909 + }, + { + "auxiliary_loss_clip": 0.0110539, + "auxiliary_loss_mlp": 0.01024926, + "balance_loss_clip": 1.03805506, + "balance_loss_mlp": 1.01714122, + "epoch": 0.9133649972945349, + "flos": 21981747864960.0, + "grad_norm": 2.1815743320797742, + "language_loss": 0.74957752, + "learning_rate": 7.812685236269989e-08, + "loss": 0.7708807, + "num_input_tokens_seen": 164264085, + "step": 7596, + "time_per_iteration": 2.909467935562134 + }, + { + "auxiliary_loss_clip": 0.01040024, + "auxiliary_loss_mlp": 0.01002105, + "balance_loss_clip": 1.00954843, + "balance_loss_mlp": 1.0006566, + "epoch": 0.9134852401851741, + "flos": 71240523511680.0, + "grad_norm": 0.7908359374633754, + "language_loss": 0.58588457, + "learning_rate": 7.791140318847445e-08, + "loss": 0.60630584, + "num_input_tokens_seen": 164322220, + "step": 7597, + "time_per_iteration": 4.230167865753174 + }, + { + "auxiliary_loss_clip": 0.01142878, + "auxiliary_loss_mlp": 0.01024515, + "balance_loss_clip": 1.04505968, + "balance_loss_mlp": 1.01744294, + "epoch": 0.9136054830758131, + "flos": 23626923615360.0, + "grad_norm": 1.4301886776868633, + "language_loss": 0.80541909, + "learning_rate": 7.769624559356081e-08, + "loss": 0.82709301, + "num_input_tokens_seen": 164345615, + "step": 7598, + "time_per_iteration": 2.9804863929748535 + }, + { + "auxiliary_loss_clip": 0.01158458, + "auxiliary_loss_mlp": 0.01031639, + "balance_loss_clip": 1.04591823, + "balance_loss_mlp": 1.0234431, + "epoch": 0.9137257259664522, + "flos": 23438858981760.0, + "grad_norm": 2.571983936419719, + "language_loss": 0.75116122, + "learning_rate": 7.748137961059842e-08, + "loss": 0.77306223, + "num_input_tokens_seen": 164359595, + "step": 7599, + "time_per_iteration": 2.813589096069336 + }, + { + "auxiliary_loss_clip": 0.01164589, + "auxiliary_loss_mlp": 0.01025691, + "balance_loss_clip": 1.0463165, + "balance_loss_mlp": 1.01787663, + "epoch": 0.9138459688570914, + "flos": 19127854523520.0, + "grad_norm": 2.581286240108329, + "language_loss": 0.6573413, + "learning_rate": 7.726680527218211e-08, + "loss": 0.67924404, + "num_input_tokens_seen": 164376635, + "step": 7600, + "time_per_iteration": 2.7330679893493652 + }, + { + "auxiliary_loss_clip": 0.01168079, + "auxiliary_loss_mlp": 0.01028962, + "balance_loss_clip": 1.04534566, + "balance_loss_mlp": 1.02102542, + "epoch": 0.9139662117477304, + "flos": 46281240714240.0, + "grad_norm": 2.3012293393172816, + "language_loss": 0.75617725, + "learning_rate": 7.70525226108627e-08, + "loss": 0.77814764, + "num_input_tokens_seen": 164400305, + "step": 7601, + "time_per_iteration": 3.731661796569824 + }, + { + "auxiliary_loss_clip": 0.01163021, + "auxiliary_loss_mlp": 0.01026763, + "balance_loss_clip": 1.04943991, + "balance_loss_mlp": 1.01967561, + "epoch": 0.9140864546383695, + "flos": 22273198819200.0, + "grad_norm": 1.9948575066820378, + "language_loss": 0.79674113, + "learning_rate": 7.683853165914666e-08, + "loss": 0.81863904, + "num_input_tokens_seen": 164418075, + "step": 7602, + "time_per_iteration": 2.713113307952881 + }, + { + "auxiliary_loss_clip": 0.011302, + "auxiliary_loss_mlp": 0.01026463, + "balance_loss_clip": 1.04191434, + "balance_loss_mlp": 1.0188458, + "epoch": 0.9142066975290086, + "flos": 17530009920000.0, + "grad_norm": 1.7709451042771058, + "language_loss": 0.77106994, + "learning_rate": 7.662483244949602e-08, + "loss": 0.79263663, + "num_input_tokens_seen": 164435335, + "step": 7603, + "time_per_iteration": 3.348043441772461 + }, + { + "auxiliary_loss_clip": 0.01134349, + "auxiliary_loss_mlp": 0.01027653, + "balance_loss_clip": 1.04300606, + "balance_loss_mlp": 1.02057552, + "epoch": 0.9143269404196477, + "flos": 17712148809600.0, + "grad_norm": 2.1491842136837347, + "language_loss": 0.8076067, + "learning_rate": 7.641142501432951e-08, + "loss": 0.82922673, + "num_input_tokens_seen": 164451530, + "step": 7604, + "time_per_iteration": 2.842724561691284 + }, + { + "auxiliary_loss_clip": 0.01144652, + "auxiliary_loss_mlp": 0.01024523, + "balance_loss_clip": 1.04232693, + "balance_loss_mlp": 1.01740897, + "epoch": 0.9144471833102867, + "flos": 33323414019840.0, + "grad_norm": 1.5916043180822652, + "language_loss": 0.73581332, + "learning_rate": 7.619830938602013e-08, + "loss": 0.75750506, + "num_input_tokens_seen": 164472755, + "step": 7605, + "time_per_iteration": 2.8834521770477295 + }, + { + "auxiliary_loss_clip": 0.01155233, + "auxiliary_loss_mlp": 0.01032741, + "balance_loss_clip": 1.04594314, + "balance_loss_mlp": 1.02513885, + "epoch": 0.9145674262009259, + "flos": 21068970428160.0, + "grad_norm": 2.421606689365849, + "language_loss": 0.82488376, + "learning_rate": 7.598548559689777e-08, + "loss": 0.84676349, + "num_input_tokens_seen": 164491155, + "step": 7606, + "time_per_iteration": 2.8442940711975098 + }, + { + "auxiliary_loss_clip": 0.01137254, + "auxiliary_loss_mlp": 0.01028509, + "balance_loss_clip": 1.04273701, + "balance_loss_mlp": 1.02045608, + "epoch": 0.914687669091565, + "flos": 16800269212800.0, + "grad_norm": 2.091513902351616, + "language_loss": 0.81161141, + "learning_rate": 7.577295367924751e-08, + "loss": 0.83326906, + "num_input_tokens_seen": 164507555, + "step": 7607, + "time_per_iteration": 2.9754440784454346 + }, + { + "auxiliary_loss_clip": 0.01156794, + "auxiliary_loss_mlp": 0.01026975, + "balance_loss_clip": 1.04729581, + "balance_loss_mlp": 1.01923227, + "epoch": 0.914807911982204, + "flos": 25773627012480.0, + "grad_norm": 1.6608575197017512, + "language_loss": 0.82131374, + "learning_rate": 7.556071366531002e-08, + "loss": 0.84315145, + "num_input_tokens_seen": 164528525, + "step": 7608, + "time_per_iteration": 2.852332353591919 + }, + { + "auxiliary_loss_clip": 0.01161635, + "auxiliary_loss_mlp": 0.01031511, + "balance_loss_clip": 1.04841733, + "balance_loss_mlp": 1.02229619, + "epoch": 0.9149281548728432, + "flos": 19208043636480.0, + "grad_norm": 2.1660756740639306, + "language_loss": 0.78776264, + "learning_rate": 7.53487655872822e-08, + "loss": 0.80969405, + "num_input_tokens_seen": 164547695, + "step": 7609, + "time_per_iteration": 2.7231152057647705 + }, + { + "auxiliary_loss_clip": 0.01133348, + "auxiliary_loss_mlp": 0.01029021, + "balance_loss_clip": 1.03975391, + "balance_loss_mlp": 1.02119517, + "epoch": 0.9150483977634822, + "flos": 26870554500480.0, + "grad_norm": 1.7608753542172648, + "language_loss": 0.74360508, + "learning_rate": 7.513710947731656e-08, + "loss": 0.76522875, + "num_input_tokens_seen": 164568905, + "step": 7610, + "time_per_iteration": 2.907799482345581 + }, + { + "auxiliary_loss_clip": 0.01144039, + "auxiliary_loss_mlp": 0.01029849, + "balance_loss_clip": 1.04442596, + "balance_loss_mlp": 1.02212441, + "epoch": 0.9151686406541213, + "flos": 21908956953600.0, + "grad_norm": 1.7404298526095132, + "language_loss": 0.85026145, + "learning_rate": 7.492574536752095e-08, + "loss": 0.87200034, + "num_input_tokens_seen": 164588895, + "step": 7611, + "time_per_iteration": 2.8149051666259766 + }, + { + "auxiliary_loss_clip": 0.01154135, + "auxiliary_loss_mlp": 0.01027932, + "balance_loss_clip": 1.0448401, + "balance_loss_mlp": 1.02099752, + "epoch": 0.9152888835447605, + "flos": 27308556944640.0, + "grad_norm": 1.9938027573762063, + "language_loss": 0.78232104, + "learning_rate": 7.471467328995907e-08, + "loss": 0.8041417, + "num_input_tokens_seen": 164607705, + "step": 7612, + "time_per_iteration": 2.7538034915924072 + }, + { + "auxiliary_loss_clip": 0.01102018, + "auxiliary_loss_mlp": 0.01025635, + "balance_loss_clip": 1.03743768, + "balance_loss_mlp": 1.01852083, + "epoch": 0.9154091264353995, + "flos": 13370728510080.0, + "grad_norm": 2.3506002797082317, + "language_loss": 0.60657918, + "learning_rate": 7.450389327665018e-08, + "loss": 0.62785566, + "num_input_tokens_seen": 164625540, + "step": 7613, + "time_per_iteration": 3.21856689453125 + }, + { + "auxiliary_loss_clip": 0.01141972, + "auxiliary_loss_mlp": 0.01024726, + "balance_loss_clip": 1.04734206, + "balance_loss_mlp": 1.01641679, + "epoch": 0.9155293693260386, + "flos": 20193037367040.0, + "grad_norm": 4.499486467826496, + "language_loss": 0.67706013, + "learning_rate": 7.429340535957029e-08, + "loss": 0.69872701, + "num_input_tokens_seen": 164640735, + "step": 7614, + "time_per_iteration": 3.164492130279541 + }, + { + "auxiliary_loss_clip": 0.01150973, + "auxiliary_loss_mlp": 0.01031166, + "balance_loss_clip": 1.04353535, + "balance_loss_mlp": 1.02362585, + "epoch": 0.9156496122166777, + "flos": 19354990176000.0, + "grad_norm": 2.3313515265407956, + "language_loss": 0.70341992, + "learning_rate": 7.40832095706494e-08, + "loss": 0.7252413, + "num_input_tokens_seen": 164657430, + "step": 7615, + "time_per_iteration": 2.849231481552124 + }, + { + "auxiliary_loss_clip": 0.01147639, + "auxiliary_loss_mlp": 0.01025329, + "balance_loss_clip": 1.04521346, + "balance_loss_mlp": 1.01818883, + "epoch": 0.9157698551073168, + "flos": 21107287261440.0, + "grad_norm": 2.5280456450624005, + "language_loss": 0.80156052, + "learning_rate": 7.387330594177443e-08, + "loss": 0.82329023, + "num_input_tokens_seen": 164679505, + "step": 7616, + "time_per_iteration": 2.8409903049468994 + }, + { + "auxiliary_loss_clip": 0.01131502, + "auxiliary_loss_mlp": 0.01029625, + "balance_loss_clip": 1.04150295, + "balance_loss_mlp": 1.02165532, + "epoch": 0.9158900979979558, + "flos": 25193167228800.0, + "grad_norm": 1.8646367284669734, + "language_loss": 0.79590887, + "learning_rate": 7.366369450478749e-08, + "loss": 0.81752014, + "num_input_tokens_seen": 164700615, + "step": 7617, + "time_per_iteration": 2.844071626663208 + }, + { + "auxiliary_loss_clip": 0.01131412, + "auxiliary_loss_mlp": 0.01025368, + "balance_loss_clip": 1.03928447, + "balance_loss_mlp": 1.01809335, + "epoch": 0.916010340888595, + "flos": 30146648302080.0, + "grad_norm": 2.837834250308912, + "language_loss": 0.66682768, + "learning_rate": 7.345437529148646e-08, + "loss": 0.6883955, + "num_input_tokens_seen": 164719625, + "step": 7618, + "time_per_iteration": 2.868882417678833 + }, + { + "auxiliary_loss_clip": 0.01142664, + "auxiliary_loss_mlp": 0.01032407, + "balance_loss_clip": 1.04273367, + "balance_loss_mlp": 1.02472973, + "epoch": 0.9161305837792341, + "flos": 17091827907840.0, + "grad_norm": 2.088998647650497, + "language_loss": 0.72501719, + "learning_rate": 7.324534833362483e-08, + "loss": 0.74676794, + "num_input_tokens_seen": 164737200, + "step": 7619, + "time_per_iteration": 2.7666025161743164 + }, + { + "auxiliary_loss_clip": 0.01150718, + "auxiliary_loss_mlp": 0.01028378, + "balance_loss_clip": 1.04538083, + "balance_loss_mlp": 1.02063811, + "epoch": 0.9162508266698731, + "flos": 22893699288960.0, + "grad_norm": 1.704477352020199, + "language_loss": 0.68514431, + "learning_rate": 7.303661366291192e-08, + "loss": 0.70693529, + "num_input_tokens_seen": 164757870, + "step": 7620, + "time_per_iteration": 3.838618755340576 + }, + { + "auxiliary_loss_clip": 0.01127441, + "auxiliary_loss_mlp": 0.01031062, + "balance_loss_clip": 1.03901362, + "balance_loss_mlp": 1.02376676, + "epoch": 0.9163710695605123, + "flos": 19974808287360.0, + "grad_norm": 1.7646502225744007, + "language_loss": 0.81531322, + "learning_rate": 7.28281713110126e-08, + "loss": 0.83689821, + "num_input_tokens_seen": 164775945, + "step": 7621, + "time_per_iteration": 3.8544116020202637 + }, + { + "auxiliary_loss_clip": 0.01143074, + "auxiliary_loss_mlp": 0.01026127, + "balance_loss_clip": 1.04435349, + "balance_loss_mlp": 1.01905441, + "epoch": 0.9164913124511513, + "flos": 22783812606720.0, + "grad_norm": 2.509316535640966, + "language_loss": 0.77042878, + "learning_rate": 7.262002130954759e-08, + "loss": 0.79212075, + "num_input_tokens_seen": 164794400, + "step": 7622, + "time_per_iteration": 2.7809765338897705 + }, + { + "auxiliary_loss_clip": 0.01133361, + "auxiliary_loss_mlp": 0.01030896, + "balance_loss_clip": 1.04088819, + "balance_loss_mlp": 1.02314138, + "epoch": 0.9166115553417904, + "flos": 24900854348160.0, + "grad_norm": 1.7300636180884827, + "language_loss": 0.78868389, + "learning_rate": 7.241216369009296e-08, + "loss": 0.81032646, + "num_input_tokens_seen": 164814585, + "step": 7623, + "time_per_iteration": 3.805710792541504 + }, + { + "auxiliary_loss_clip": 0.01168564, + "auxiliary_loss_mlp": 0.0102627, + "balance_loss_clip": 1.0456481, + "balance_loss_mlp": 1.01867032, + "epoch": 0.9167317982324296, + "flos": 25702919089920.0, + "grad_norm": 1.877412225920319, + "language_loss": 0.66125298, + "learning_rate": 7.220459848418037e-08, + "loss": 0.68320125, + "num_input_tokens_seen": 164834660, + "step": 7624, + "time_per_iteration": 2.7062928676605225 + }, + { + "auxiliary_loss_clip": 0.01168638, + "auxiliary_loss_mlp": 0.01025893, + "balance_loss_clip": 1.04873061, + "balance_loss_mlp": 1.01877952, + "epoch": 0.9168520411230686, + "flos": 15632813370240.0, + "grad_norm": 1.7215317108553057, + "language_loss": 0.79769254, + "learning_rate": 7.199732572329708e-08, + "loss": 0.81963778, + "num_input_tokens_seen": 164852560, + "step": 7625, + "time_per_iteration": 2.6225228309631348 + }, + { + "auxiliary_loss_clip": 0.0113555, + "auxiliary_loss_mlp": 0.0102548, + "balance_loss_clip": 1.04291475, + "balance_loss_mlp": 1.01802325, + "epoch": 0.9169722840137077, + "flos": 30258151096320.0, + "grad_norm": 2.4445404835718896, + "language_loss": 0.76158828, + "learning_rate": 7.179034543888684e-08, + "loss": 0.7831986, + "num_input_tokens_seen": 164872065, + "step": 7626, + "time_per_iteration": 2.8795504570007324 + }, + { + "auxiliary_loss_clip": 0.01158684, + "auxiliary_loss_mlp": 0.01023917, + "balance_loss_clip": 1.04343402, + "balance_loss_mlp": 1.01595426, + "epoch": 0.9170925269043467, + "flos": 22491643380480.0, + "grad_norm": 1.9722785397798552, + "language_loss": 0.77441013, + "learning_rate": 7.158365766234808e-08, + "loss": 0.79623616, + "num_input_tokens_seen": 164890915, + "step": 7627, + "time_per_iteration": 3.8086650371551514 + }, + { + "auxiliary_loss_clip": 0.01132301, + "auxiliary_loss_mlp": 0.01027329, + "balance_loss_clip": 1.03967702, + "balance_loss_mlp": 1.01962852, + "epoch": 0.9172127697949859, + "flos": 22893914770560.0, + "grad_norm": 1.7416911425589903, + "language_loss": 0.72458607, + "learning_rate": 7.137726242503527e-08, + "loss": 0.74618238, + "num_input_tokens_seen": 164909835, + "step": 7628, + "time_per_iteration": 2.938415288925171 + }, + { + "auxiliary_loss_clip": 0.01160584, + "auxiliary_loss_mlp": 0.0088575, + "balance_loss_clip": 1.04809761, + "balance_loss_mlp": 1.00058627, + "epoch": 0.917333012685625, + "flos": 17451867882240.0, + "grad_norm": 2.2386357597706583, + "language_loss": 0.78085077, + "learning_rate": 7.11711597582585e-08, + "loss": 0.80131412, + "num_input_tokens_seen": 164927195, + "step": 7629, + "time_per_iteration": 2.6897661685943604 + }, + { + "auxiliary_loss_clip": 0.01139503, + "auxiliary_loss_mlp": 0.01030462, + "balance_loss_clip": 1.03857338, + "balance_loss_mlp": 1.02301693, + "epoch": 0.917453255576264, + "flos": 14318949692160.0, + "grad_norm": 1.6616628638460595, + "language_loss": 0.80020493, + "learning_rate": 7.096534969328271e-08, + "loss": 0.82190454, + "num_input_tokens_seen": 164944640, + "step": 7630, + "time_per_iteration": 2.884434938430786 + }, + { + "auxiliary_loss_clip": 0.01150593, + "auxiliary_loss_mlp": 0.01028562, + "balance_loss_clip": 1.04099417, + "balance_loss_mlp": 1.02079582, + "epoch": 0.9175734984669032, + "flos": 20741177888640.0, + "grad_norm": 2.8057891185841237, + "language_loss": 0.83826363, + "learning_rate": 7.075983226132987e-08, + "loss": 0.86005521, + "num_input_tokens_seen": 164963570, + "step": 7631, + "time_per_iteration": 2.8317766189575195 + }, + { + "auxiliary_loss_clip": 0.01156604, + "auxiliary_loss_mlp": 0.00886853, + "balance_loss_clip": 1.04556584, + "balance_loss_mlp": 1.0006392, + "epoch": 0.9176937413575422, + "flos": 14830497233280.0, + "grad_norm": 2.766128457262865, + "language_loss": 0.79280579, + "learning_rate": 7.055460749357656e-08, + "loss": 0.81324041, + "num_input_tokens_seen": 164979850, + "step": 7632, + "time_per_iteration": 2.878624677658081 + }, + { + "auxiliary_loss_clip": 0.01149677, + "auxiliary_loss_mlp": 0.01029245, + "balance_loss_clip": 1.0454464, + "balance_loss_mlp": 1.02185404, + "epoch": 0.9178139842481813, + "flos": 18474603828480.0, + "grad_norm": 1.7029800194275078, + "language_loss": 0.70556486, + "learning_rate": 7.034967542115521e-08, + "loss": 0.72735399, + "num_input_tokens_seen": 164998115, + "step": 7633, + "time_per_iteration": 2.767618179321289 + }, + { + "auxiliary_loss_clip": 0.01148311, + "auxiliary_loss_mlp": 0.00885988, + "balance_loss_clip": 1.04227817, + "balance_loss_mlp": 1.00055599, + "epoch": 0.9179342271388204, + "flos": 20047455544320.0, + "grad_norm": 2.23494408388092, + "language_loss": 0.75649738, + "learning_rate": 7.014503607515388e-08, + "loss": 0.77684039, + "num_input_tokens_seen": 165017420, + "step": 7634, + "time_per_iteration": 2.6553237438201904 + }, + { + "auxiliary_loss_clip": 0.01150388, + "auxiliary_loss_mlp": 0.01028241, + "balance_loss_clip": 1.04800522, + "balance_loss_mlp": 1.02058196, + "epoch": 0.9180544700294595, + "flos": 24676232647680.0, + "grad_norm": 2.102179183470335, + "language_loss": 0.67991614, + "learning_rate": 6.994068948661592e-08, + "loss": 0.70170248, + "num_input_tokens_seen": 165035575, + "step": 7635, + "time_per_iteration": 2.776508331298828 + }, + { + "auxiliary_loss_clip": 0.01159189, + "auxiliary_loss_mlp": 0.0102489, + "balance_loss_clip": 1.04525971, + "balance_loss_mlp": 1.01662254, + "epoch": 0.9181747129200986, + "flos": 16727478301440.0, + "grad_norm": 1.9910212339276354, + "language_loss": 0.76839364, + "learning_rate": 6.973663568654142e-08, + "loss": 0.79023445, + "num_input_tokens_seen": 165053280, + "step": 7636, + "time_per_iteration": 2.708275318145752 + }, + { + "auxiliary_loss_clip": 0.01166083, + "auxiliary_loss_mlp": 0.01030209, + "balance_loss_clip": 1.0461899, + "balance_loss_mlp": 1.02232027, + "epoch": 0.9182949558107377, + "flos": 24271626873600.0, + "grad_norm": 2.1974955996566874, + "language_loss": 0.64995909, + "learning_rate": 6.953287470588386e-08, + "loss": 0.67192197, + "num_input_tokens_seen": 165071235, + "step": 7637, + "time_per_iteration": 2.7033910751342773 + }, + { + "auxiliary_loss_clip": 0.01162876, + "auxiliary_loss_mlp": 0.01024133, + "balance_loss_clip": 1.04451048, + "balance_loss_mlp": 1.01589561, + "epoch": 0.9184151987013768, + "flos": 22082117443200.0, + "grad_norm": 2.330816095065835, + "language_loss": 0.85932589, + "learning_rate": 6.932940657555452e-08, + "loss": 0.88119596, + "num_input_tokens_seen": 165087365, + "step": 7638, + "time_per_iteration": 2.7168872356414795 + }, + { + "auxiliary_loss_clip": 0.01164538, + "auxiliary_loss_mlp": 0.01021796, + "balance_loss_clip": 1.04645896, + "balance_loss_mlp": 1.01478088, + "epoch": 0.9185354415920158, + "flos": 32166732257280.0, + "grad_norm": 1.4158018047629068, + "language_loss": 0.76276219, + "learning_rate": 6.912623132641938e-08, + "loss": 0.78462553, + "num_input_tokens_seen": 165112455, + "step": 7639, + "time_per_iteration": 2.7823805809020996 + }, + { + "auxiliary_loss_clip": 0.01153566, + "auxiliary_loss_mlp": 0.01029647, + "balance_loss_clip": 1.04547203, + "balance_loss_mlp": 1.02135003, + "epoch": 0.918655684482655, + "flos": 20997831542400.0, + "grad_norm": 1.8411006695194754, + "language_loss": 0.76399875, + "learning_rate": 6.892334898929952e-08, + "loss": 0.78583086, + "num_input_tokens_seen": 165132700, + "step": 7640, + "time_per_iteration": 2.7679803371429443 + }, + { + "auxiliary_loss_clip": 0.01154459, + "auxiliary_loss_mlp": 0.01026428, + "balance_loss_clip": 1.04472315, + "balance_loss_mlp": 1.0189538, + "epoch": 0.918775927373294, + "flos": 15560704817280.0, + "grad_norm": 1.8612105482711787, + "language_loss": 0.84814286, + "learning_rate": 6.872075959497236e-08, + "loss": 0.86995173, + "num_input_tokens_seen": 165151475, + "step": 7641, + "time_per_iteration": 2.7127325534820557 + }, + { + "auxiliary_loss_clip": 0.01154696, + "auxiliary_loss_mlp": 0.01030677, + "balance_loss_clip": 1.04474497, + "balance_loss_mlp": 1.02332187, + "epoch": 0.9188961702639331, + "flos": 29934057657600.0, + "grad_norm": 1.8392691347349164, + "language_loss": 0.82956231, + "learning_rate": 6.85184631741702e-08, + "loss": 0.85141599, + "num_input_tokens_seen": 165172040, + "step": 7642, + "time_per_iteration": 2.7275538444519043 + }, + { + "auxiliary_loss_clip": 0.01158098, + "auxiliary_loss_mlp": 0.01024486, + "balance_loss_clip": 1.04515982, + "balance_loss_mlp": 1.01633239, + "epoch": 0.9190164131545723, + "flos": 20701244943360.0, + "grad_norm": 1.814774945784006, + "language_loss": 0.77236944, + "learning_rate": 6.831645975758161e-08, + "loss": 0.79419529, + "num_input_tokens_seen": 165189980, + "step": 7643, + "time_per_iteration": 2.660507917404175 + }, + { + "auxiliary_loss_clip": 0.01143432, + "auxiliary_loss_mlp": 0.01030645, + "balance_loss_clip": 1.04596448, + "balance_loss_mlp": 1.02283716, + "epoch": 0.9191366560452113, + "flos": 25629912696960.0, + "grad_norm": 2.105452100464814, + "language_loss": 0.67669016, + "learning_rate": 6.811474937585026e-08, + "loss": 0.6984309, + "num_input_tokens_seen": 165209770, + "step": 7644, + "time_per_iteration": 2.801048517227173 + }, + { + "auxiliary_loss_clip": 0.01135468, + "auxiliary_loss_mlp": 0.01026738, + "balance_loss_clip": 1.04166198, + "balance_loss_mlp": 1.01960325, + "epoch": 0.9192568989358504, + "flos": 21434325615360.0, + "grad_norm": 1.5628361702082851, + "language_loss": 0.79217905, + "learning_rate": 6.79133320595755e-08, + "loss": 0.81380111, + "num_input_tokens_seen": 165229690, + "step": 7645, + "time_per_iteration": 2.800764560699463 + }, + { + "auxiliary_loss_clip": 0.01152492, + "auxiliary_loss_mlp": 0.01029762, + "balance_loss_clip": 1.04653108, + "balance_loss_mlp": 1.02197194, + "epoch": 0.9193771418264896, + "flos": 23185078416000.0, + "grad_norm": 1.854424393692408, + "language_loss": 0.75424337, + "learning_rate": 6.771220783931198e-08, + "loss": 0.77606589, + "num_input_tokens_seen": 165249850, + "step": 7646, + "time_per_iteration": 3.696138620376587 + }, + { + "auxiliary_loss_clip": 0.01018048, + "auxiliary_loss_mlp": 0.00877359, + "balance_loss_clip": 1.01134109, + "balance_loss_mlp": 1.00124645, + "epoch": 0.9194973847171286, + "flos": 70582963184640.0, + "grad_norm": 0.8538132195293969, + "language_loss": 0.64605379, + "learning_rate": 6.751137674556994e-08, + "loss": 0.66500783, + "num_input_tokens_seen": 165310235, + "step": 7647, + "time_per_iteration": 4.65202260017395 + }, + { + "auxiliary_loss_clip": 0.01161037, + "auxiliary_loss_mlp": 0.01026429, + "balance_loss_clip": 1.04404497, + "balance_loss_mlp": 1.01849008, + "epoch": 0.9196176276077677, + "flos": 14720682378240.0, + "grad_norm": 2.0104180992965666, + "language_loss": 0.77613628, + "learning_rate": 6.731083880881572e-08, + "loss": 0.79801095, + "num_input_tokens_seen": 165326455, + "step": 7648, + "time_per_iteration": 3.0776336193084717 + }, + { + "auxiliary_loss_clip": 0.01151853, + "auxiliary_loss_mlp": 0.01024897, + "balance_loss_clip": 1.0453968, + "balance_loss_mlp": 1.01731563, + "epoch": 0.9197378704984068, + "flos": 23294893271040.0, + "grad_norm": 2.583394658225541, + "language_loss": 0.81242377, + "learning_rate": 6.711059405947072e-08, + "loss": 0.83419126, + "num_input_tokens_seen": 165344645, + "step": 7649, + "time_per_iteration": 3.963249921798706 + }, + { + "auxiliary_loss_clip": 0.01135951, + "auxiliary_loss_mlp": 0.01025967, + "balance_loss_clip": 1.04373908, + "balance_loss_mlp": 1.01836193, + "epoch": 0.9198581133890459, + "flos": 20302564913280.0, + "grad_norm": 9.893744429992767, + "language_loss": 0.76823068, + "learning_rate": 6.691064252791156e-08, + "loss": 0.78984988, + "num_input_tokens_seen": 165364120, + "step": 7650, + "time_per_iteration": 2.871243715286255 + }, + { + "auxiliary_loss_clip": 0.0112266, + "auxiliary_loss_mlp": 0.01029406, + "balance_loss_clip": 1.0409013, + "balance_loss_mlp": 1.02177691, + "epoch": 0.9199783562796849, + "flos": 17675663569920.0, + "grad_norm": 2.2613670769160814, + "language_loss": 0.78044784, + "learning_rate": 6.67109842444713e-08, + "loss": 0.80196857, + "num_input_tokens_seen": 165383050, + "step": 7651, + "time_per_iteration": 2.879486083984375 + }, + { + "auxiliary_loss_clip": 0.01154608, + "auxiliary_loss_mlp": 0.00885963, + "balance_loss_clip": 1.04550123, + "balance_loss_mlp": 1.00063539, + "epoch": 0.9200985991703241, + "flos": 17676022705920.0, + "grad_norm": 1.865552114724451, + "language_loss": 0.767941, + "learning_rate": 6.651161923943704e-08, + "loss": 0.78834677, + "num_input_tokens_seen": 165400955, + "step": 7652, + "time_per_iteration": 2.7435874938964844 + }, + { + "auxiliary_loss_clip": 0.01155575, + "auxiliary_loss_mlp": 0.01023475, + "balance_loss_clip": 1.04446137, + "balance_loss_mlp": 1.0156188, + "epoch": 0.9202188420609632, + "flos": 20996574566400.0, + "grad_norm": 1.865102429631826, + "language_loss": 0.77001774, + "learning_rate": 6.631254754305326e-08, + "loss": 0.79180819, + "num_input_tokens_seen": 165420415, + "step": 7653, + "time_per_iteration": 3.7033820152282715 + }, + { + "auxiliary_loss_clip": 0.01170192, + "auxiliary_loss_mlp": 0.0102365, + "balance_loss_clip": 1.04644394, + "balance_loss_mlp": 1.01596677, + "epoch": 0.9203390849516022, + "flos": 13918222586880.0, + "grad_norm": 2.1313660213109626, + "language_loss": 0.78556585, + "learning_rate": 6.611376918551848e-08, + "loss": 0.8075043, + "num_input_tokens_seen": 165439200, + "step": 7654, + "time_per_iteration": 2.632223606109619 + }, + { + "auxiliary_loss_clip": 0.01127723, + "auxiliary_loss_mlp": 0.00885909, + "balance_loss_clip": 1.04070711, + "balance_loss_mlp": 1.00061107, + "epoch": 0.9204593278422414, + "flos": 21175912195200.0, + "grad_norm": 1.9657885856244346, + "language_loss": 0.79879308, + "learning_rate": 6.591528419698744e-08, + "loss": 0.81892937, + "num_input_tokens_seen": 165458985, + "step": 7655, + "time_per_iteration": 2.8199377059936523 + }, + { + "auxiliary_loss_clip": 0.01151254, + "auxiliary_loss_mlp": 0.01030253, + "balance_loss_clip": 1.04270935, + "balance_loss_mlp": 1.02253163, + "epoch": 0.9205795707328804, + "flos": 14501375890560.0, + "grad_norm": 2.6318539553158464, + "language_loss": 0.83472848, + "learning_rate": 6.571709260756986e-08, + "loss": 0.85654354, + "num_input_tokens_seen": 165475630, + "step": 7656, + "time_per_iteration": 2.735548257827759 + }, + { + "auxiliary_loss_clip": 0.01160939, + "auxiliary_loss_mlp": 0.01039928, + "balance_loss_clip": 1.04864502, + "balance_loss_mlp": 1.03169084, + "epoch": 0.9206998136235195, + "flos": 22417559579520.0, + "grad_norm": 2.3099495997264854, + "language_loss": 0.76478326, + "learning_rate": 6.551919444733122e-08, + "loss": 0.78679192, + "num_input_tokens_seen": 165493445, + "step": 7657, + "time_per_iteration": 2.711003541946411 + }, + { + "auxiliary_loss_clip": 0.01147236, + "auxiliary_loss_mlp": 0.01026515, + "balance_loss_clip": 1.04469705, + "balance_loss_mlp": 1.01880479, + "epoch": 0.9208200565141585, + "flos": 53358407544960.0, + "grad_norm": 2.0483972430108968, + "language_loss": 0.66159296, + "learning_rate": 6.53215897462931e-08, + "loss": 0.68333048, + "num_input_tokens_seen": 165517200, + "step": 7658, + "time_per_iteration": 3.0357162952423096 + }, + { + "auxiliary_loss_clip": 0.01157279, + "auxiliary_loss_mlp": 0.01038414, + "balance_loss_clip": 1.04407239, + "balance_loss_mlp": 1.03046298, + "epoch": 0.9209402994047977, + "flos": 30589139946240.0, + "grad_norm": 1.9222339935273491, + "language_loss": 0.75051337, + "learning_rate": 6.512427853443103e-08, + "loss": 0.77247036, + "num_input_tokens_seen": 165539280, + "step": 7659, + "time_per_iteration": 2.854076862335205 + }, + { + "auxiliary_loss_clip": 0.01161628, + "auxiliary_loss_mlp": 0.01030895, + "balance_loss_clip": 1.04528308, + "balance_loss_mlp": 1.02357006, + "epoch": 0.9210605422954368, + "flos": 29132711187840.0, + "grad_norm": 1.7824155436858957, + "language_loss": 0.75795501, + "learning_rate": 6.492726084167799e-08, + "loss": 0.77988023, + "num_input_tokens_seen": 165561395, + "step": 7660, + "time_per_iteration": 2.7824220657348633 + }, + { + "auxiliary_loss_clip": 0.01055178, + "auxiliary_loss_mlp": 0.01001328, + "balance_loss_clip": 1.00437927, + "balance_loss_mlp": 0.99998724, + "epoch": 0.9211807851860758, + "flos": 54853838472960.0, + "grad_norm": 0.7777576236145395, + "language_loss": 0.57505667, + "learning_rate": 6.473053669792072e-08, + "loss": 0.59562171, + "num_input_tokens_seen": 165616085, + "step": 7661, + "time_per_iteration": 3.1220145225524902 + }, + { + "auxiliary_loss_clip": 0.01158937, + "auxiliary_loss_mlp": 0.01021661, + "balance_loss_clip": 1.04518235, + "balance_loss_mlp": 1.0140909, + "epoch": 0.921301028076715, + "flos": 19201974238080.0, + "grad_norm": 2.254871420279956, + "language_loss": 0.72954392, + "learning_rate": 6.453410613300248e-08, + "loss": 0.75134993, + "num_input_tokens_seen": 165634015, + "step": 7662, + "time_per_iteration": 2.7597217559814453 + }, + { + "auxiliary_loss_clip": 0.01119649, + "auxiliary_loss_mlp": 0.01025488, + "balance_loss_clip": 1.04036808, + "balance_loss_mlp": 1.01691651, + "epoch": 0.921421270967354, + "flos": 27526893765120.0, + "grad_norm": 3.017194234921833, + "language_loss": 0.58403462, + "learning_rate": 6.43379691767214e-08, + "loss": 0.60548592, + "num_input_tokens_seen": 165653220, + "step": 7663, + "time_per_iteration": 2.92685604095459 + }, + { + "auxiliary_loss_clip": 0.01035361, + "auxiliary_loss_mlp": 0.01002398, + "balance_loss_clip": 1.00539684, + "balance_loss_mlp": 1.00087249, + "epoch": 0.9215415138579931, + "flos": 70209311955840.0, + "grad_norm": 0.7202170121060011, + "language_loss": 0.55130494, + "learning_rate": 6.414212585883105e-08, + "loss": 0.57168257, + "num_input_tokens_seen": 165715850, + "step": 7664, + "time_per_iteration": 3.424015522003174 + }, + { + "auxiliary_loss_clip": 0.01153184, + "auxiliary_loss_mlp": 0.01027999, + "balance_loss_clip": 1.0449369, + "balance_loss_mlp": 1.0199523, + "epoch": 0.9216617567486323, + "flos": 35553107790720.0, + "grad_norm": 1.619831680777605, + "language_loss": 0.69749898, + "learning_rate": 6.394657620904143e-08, + "loss": 0.71931076, + "num_input_tokens_seen": 165738960, + "step": 7665, + "time_per_iteration": 2.9488155841827393 + }, + { + "auxiliary_loss_clip": 0.01172833, + "auxiliary_loss_mlp": 0.01029737, + "balance_loss_clip": 1.04805851, + "balance_loss_mlp": 1.02139235, + "epoch": 0.9217819996392713, + "flos": 29533330552320.0, + "grad_norm": 1.6217674663761386, + "language_loss": 0.71761572, + "learning_rate": 6.375132025701657e-08, + "loss": 0.73964149, + "num_input_tokens_seen": 165761260, + "step": 7666, + "time_per_iteration": 2.8038926124572754 + }, + { + "auxiliary_loss_clip": 0.01171233, + "auxiliary_loss_mlp": 0.01030714, + "balance_loss_clip": 1.04890227, + "balance_loss_mlp": 1.02306402, + "epoch": 0.9219022425299104, + "flos": 14574669592320.0, + "grad_norm": 2.5060286293035983, + "language_loss": 0.69314301, + "learning_rate": 6.355635803237724e-08, + "loss": 0.71516252, + "num_input_tokens_seen": 165776960, + "step": 7667, + "time_per_iteration": 2.7398886680603027 + }, + { + "auxiliary_loss_clip": 0.01156752, + "auxiliary_loss_mlp": 0.01023361, + "balance_loss_clip": 1.0441432, + "balance_loss_mlp": 1.01523042, + "epoch": 0.9220224854205495, + "flos": 18077503996800.0, + "grad_norm": 2.3248046912783336, + "language_loss": 0.79530078, + "learning_rate": 6.336168956469867e-08, + "loss": 0.8171019, + "num_input_tokens_seen": 165795435, + "step": 7668, + "time_per_iteration": 2.769578456878662 + }, + { + "auxiliary_loss_clip": 0.0114236, + "auxiliary_loss_mlp": 0.01029061, + "balance_loss_clip": 1.04392016, + "balance_loss_mlp": 1.02125859, + "epoch": 0.9221427283111886, + "flos": 24790464875520.0, + "grad_norm": 1.6321675892446192, + "language_loss": 0.72221464, + "learning_rate": 6.316731488351168e-08, + "loss": 0.74392885, + "num_input_tokens_seen": 165816625, + "step": 7669, + "time_per_iteration": 2.810581684112549 + }, + { + "auxiliary_loss_clip": 0.01156326, + "auxiliary_loss_mlp": 0.01027554, + "balance_loss_clip": 1.04471374, + "balance_loss_mlp": 1.01990032, + "epoch": 0.9222629712018277, + "flos": 13845036625920.0, + "grad_norm": 1.8083897807645435, + "language_loss": 0.63572305, + "learning_rate": 6.297323401830334e-08, + "loss": 0.65756178, + "num_input_tokens_seen": 165835410, + "step": 7670, + "time_per_iteration": 2.732715368270874 + }, + { + "auxiliary_loss_clip": 0.01162502, + "auxiliary_loss_mlp": 0.01027763, + "balance_loss_clip": 1.04579258, + "balance_loss_mlp": 1.0199306, + "epoch": 0.9223832140924668, + "flos": 21616177196160.0, + "grad_norm": 1.9296451026027805, + "language_loss": 0.69043237, + "learning_rate": 6.277944699851523e-08, + "loss": 0.71233499, + "num_input_tokens_seen": 165854930, + "step": 7671, + "time_per_iteration": 2.751389265060425 + }, + { + "auxiliary_loss_clip": 0.0116816, + "auxiliary_loss_mlp": 0.01032078, + "balance_loss_clip": 1.04590738, + "balance_loss_mlp": 1.02386439, + "epoch": 0.9225034569831059, + "flos": 21142084561920.0, + "grad_norm": 1.8751162286920915, + "language_loss": 0.73196018, + "learning_rate": 6.25859538535447e-08, + "loss": 0.75396258, + "num_input_tokens_seen": 165875725, + "step": 7672, + "time_per_iteration": 3.6367828845977783 + }, + { + "auxiliary_loss_clip": 0.01149807, + "auxiliary_loss_mlp": 0.01031307, + "balance_loss_clip": 1.04584897, + "balance_loss_mlp": 1.02283764, + "epoch": 0.9226236998737449, + "flos": 12495046844160.0, + "grad_norm": 2.6265046484882246, + "language_loss": 0.77999848, + "learning_rate": 6.239275461274474e-08, + "loss": 0.80180967, + "num_input_tokens_seen": 165892100, + "step": 7673, + "time_per_iteration": 3.7187087535858154 + }, + { + "auxiliary_loss_clip": 0.01158297, + "auxiliary_loss_mlp": 0.01022465, + "balance_loss_clip": 1.0454917, + "balance_loss_mlp": 1.0151155, + "epoch": 0.9227439427643841, + "flos": 26214071581440.0, + "grad_norm": 1.730341705372336, + "language_loss": 0.85770869, + "learning_rate": 6.219984930542299e-08, + "loss": 0.87951636, + "num_input_tokens_seen": 165912840, + "step": 7674, + "time_per_iteration": 2.7854814529418945 + }, + { + "auxiliary_loss_clip": 0.01160617, + "auxiliary_loss_mlp": 0.01025203, + "balance_loss_clip": 1.0442307, + "balance_loss_mlp": 1.01734424, + "epoch": 0.9228641856550232, + "flos": 17967581400960.0, + "grad_norm": 2.617818823128622, + "language_loss": 0.75901282, + "learning_rate": 6.200723796084383e-08, + "loss": 0.78087103, + "num_input_tokens_seen": 165930935, + "step": 7675, + "time_per_iteration": 3.6042611598968506 + }, + { + "auxiliary_loss_clip": 0.0104307, + "auxiliary_loss_mlp": 0.01001589, + "balance_loss_clip": 1.00597572, + "balance_loss_mlp": 0.99993247, + "epoch": 0.9229844285456622, + "flos": 70420609710720.0, + "grad_norm": 0.761940151721824, + "language_loss": 0.62968564, + "learning_rate": 6.181492060822546e-08, + "loss": 0.65013218, + "num_input_tokens_seen": 165991110, + "step": 7676, + "time_per_iteration": 3.240063428878784 + }, + { + "auxiliary_loss_clip": 0.01126069, + "auxiliary_loss_mlp": 0.01028118, + "balance_loss_clip": 1.0394733, + "balance_loss_mlp": 1.02081013, + "epoch": 0.9231046714363014, + "flos": 17967832796160.0, + "grad_norm": 2.0274649977855366, + "language_loss": 0.81894845, + "learning_rate": 6.162289727674274e-08, + "loss": 0.84049028, + "num_input_tokens_seen": 166008790, + "step": 7677, + "time_per_iteration": 2.837216377258301 + }, + { + "auxiliary_loss_clip": 0.0114, + "auxiliary_loss_mlp": 0.01028434, + "balance_loss_clip": 1.04199934, + "balance_loss_mlp": 1.02132964, + "epoch": 0.9232249143269404, + "flos": 17858233422720.0, + "grad_norm": 2.2281966285749806, + "language_loss": 0.87745661, + "learning_rate": 6.143116799552527e-08, + "loss": 0.89914095, + "num_input_tokens_seen": 166025035, + "step": 7678, + "time_per_iteration": 2.733185052871704 + }, + { + "auxiliary_loss_clip": 0.01162534, + "auxiliary_loss_mlp": 0.01029619, + "balance_loss_clip": 1.04573131, + "balance_loss_mlp": 1.02184629, + "epoch": 0.9233451572175795, + "flos": 23404384903680.0, + "grad_norm": 4.3239145710422315, + "language_loss": 0.56163383, + "learning_rate": 6.123973279365802e-08, + "loss": 0.5835554, + "num_input_tokens_seen": 166044010, + "step": 7679, + "time_per_iteration": 3.6855623722076416 + }, + { + "auxiliary_loss_clip": 0.01163534, + "auxiliary_loss_mlp": 0.01034273, + "balance_loss_clip": 1.04785323, + "balance_loss_mlp": 1.02632153, + "epoch": 0.9234654001082186, + "flos": 17999326045440.0, + "grad_norm": 1.8475680262838956, + "language_loss": 0.77501571, + "learning_rate": 6.10485917001824e-08, + "loss": 0.79699379, + "num_input_tokens_seen": 166061865, + "step": 7680, + "time_per_iteration": 2.7707676887512207 + }, + { + "auxiliary_loss_clip": 0.01146658, + "auxiliary_loss_mlp": 0.01023939, + "balance_loss_clip": 1.04505527, + "balance_loss_mlp": 1.0168227, + "epoch": 0.9235856429988577, + "flos": 24750747411840.0, + "grad_norm": 1.9442042267408997, + "language_loss": 0.81109977, + "learning_rate": 6.085774474409322e-08, + "loss": 0.83280575, + "num_input_tokens_seen": 166082425, + "step": 7681, + "time_per_iteration": 2.7496156692504883 + }, + { + "auxiliary_loss_clip": 0.01152735, + "auxiliary_loss_mlp": 0.0103244, + "balance_loss_clip": 1.04851413, + "balance_loss_mlp": 1.02480805, + "epoch": 0.9237058858894968, + "flos": 14099894599680.0, + "grad_norm": 1.7898730195040589, + "language_loss": 0.70015788, + "learning_rate": 6.066719195434267e-08, + "loss": 0.72200966, + "num_input_tokens_seen": 166100225, + "step": 7682, + "time_per_iteration": 2.660834789276123 + }, + { + "auxiliary_loss_clip": 0.01163107, + "auxiliary_loss_mlp": 0.01029727, + "balance_loss_clip": 1.04748249, + "balance_loss_mlp": 1.02196038, + "epoch": 0.9238261287801359, + "flos": 28694529175680.0, + "grad_norm": 2.0639878633946567, + "language_loss": 0.66392881, + "learning_rate": 6.047693335983717e-08, + "loss": 0.68585718, + "num_input_tokens_seen": 166122570, + "step": 7683, + "time_per_iteration": 2.695065498352051 + }, + { + "auxiliary_loss_clip": 0.01161771, + "auxiliary_loss_mlp": 0.01025514, + "balance_loss_clip": 1.04620099, + "balance_loss_mlp": 1.01723766, + "epoch": 0.923946371670775, + "flos": 23111856541440.0, + "grad_norm": 2.3501665134095804, + "language_loss": 0.82177222, + "learning_rate": 6.028696898943853e-08, + "loss": 0.8436451, + "num_input_tokens_seen": 166141630, + "step": 7684, + "time_per_iteration": 2.66335391998291 + }, + { + "auxiliary_loss_clip": 0.01147806, + "auxiliary_loss_mlp": 0.00886295, + "balance_loss_clip": 1.04168177, + "balance_loss_mlp": 1.00054479, + "epoch": 0.924066614561414, + "flos": 21867120587520.0, + "grad_norm": 3.2070831229520165, + "language_loss": 0.70988095, + "learning_rate": 6.00972988719648e-08, + "loss": 0.73022193, + "num_input_tokens_seen": 166159865, + "step": 7685, + "time_per_iteration": 2.6754534244537354 + }, + { + "auxiliary_loss_clip": 0.01141348, + "auxiliary_loss_mlp": 0.00886055, + "balance_loss_clip": 1.04353499, + "balance_loss_mlp": 1.0005939, + "epoch": 0.9241868574520532, + "flos": 28511887495680.0, + "grad_norm": 2.370710986502765, + "language_loss": 0.70616066, + "learning_rate": 5.990792303618807e-08, + "loss": 0.72643471, + "num_input_tokens_seen": 166179445, + "step": 7686, + "time_per_iteration": 2.766779661178589 + }, + { + "auxiliary_loss_clip": 0.01138283, + "auxiliary_loss_mlp": 0.01027878, + "balance_loss_clip": 1.04429698, + "balance_loss_mlp": 1.02037954, + "epoch": 0.9243071003426923, + "flos": 30518324282880.0, + "grad_norm": 1.5347655260371789, + "language_loss": 0.6939224, + "learning_rate": 5.971884151083695e-08, + "loss": 0.71558398, + "num_input_tokens_seen": 166201855, + "step": 7687, + "time_per_iteration": 2.7782795429229736 + }, + { + "auxiliary_loss_clip": 0.01151911, + "auxiliary_loss_mlp": 0.01030519, + "balance_loss_clip": 1.04465461, + "balance_loss_mlp": 1.02321792, + "epoch": 0.9244273432333313, + "flos": 28658331244800.0, + "grad_norm": 1.6909048385628034, + "language_loss": 0.74346715, + "learning_rate": 5.9530054324595124e-08, + "loss": 0.76529145, + "num_input_tokens_seen": 166221970, + "step": 7688, + "time_per_iteration": 3.031475782394409 + }, + { + "auxiliary_loss_clip": 0.0104651, + "auxiliary_loss_mlp": 0.00876771, + "balance_loss_clip": 1.00504732, + "balance_loss_mlp": 1.00131309, + "epoch": 0.9245475861239704, + "flos": 66230589237120.0, + "grad_norm": 0.7166735232609349, + "language_loss": 0.575288, + "learning_rate": 5.934156150610103e-08, + "loss": 0.59452081, + "num_input_tokens_seen": 166279335, + "step": 7689, + "time_per_iteration": 3.305464029312134 + }, + { + "auxiliary_loss_clip": 0.01147738, + "auxiliary_loss_mlp": 0.01033629, + "balance_loss_clip": 1.04392529, + "balance_loss_mlp": 1.02558267, + "epoch": 0.9246678290146095, + "flos": 24239918142720.0, + "grad_norm": 2.276902361120007, + "language_loss": 0.78844523, + "learning_rate": 5.915336308394914e-08, + "loss": 0.81025887, + "num_input_tokens_seen": 166298170, + "step": 7690, + "time_per_iteration": 2.8311257362365723 + }, + { + "auxiliary_loss_clip": 0.01152941, + "auxiliary_loss_mlp": 0.01031455, + "balance_loss_clip": 1.04330385, + "balance_loss_mlp": 1.02399564, + "epoch": 0.9247880719052486, + "flos": 18988808976000.0, + "grad_norm": 3.6523072121986466, + "language_loss": 0.76846665, + "learning_rate": 5.89654590866886e-08, + "loss": 0.79031062, + "num_input_tokens_seen": 166317670, + "step": 7691, + "time_per_iteration": 2.8199188709259033 + }, + { + "auxiliary_loss_clip": 0.01124482, + "auxiliary_loss_mlp": 0.01026301, + "balance_loss_clip": 1.04353547, + "balance_loss_mlp": 1.0183053, + "epoch": 0.9249083147958876, + "flos": 24024095274240.0, + "grad_norm": 2.187346597701121, + "language_loss": 0.88494039, + "learning_rate": 5.877784954282483e-08, + "loss": 0.90644825, + "num_input_tokens_seen": 166337010, + "step": 7692, + "time_per_iteration": 2.90084171295166 + }, + { + "auxiliary_loss_clip": 0.01163656, + "auxiliary_loss_mlp": 0.01033174, + "balance_loss_clip": 1.04679346, + "balance_loss_mlp": 1.02496099, + "epoch": 0.9250285576865268, + "flos": 30773972355840.0, + "grad_norm": 1.8570208261346928, + "language_loss": 0.72302246, + "learning_rate": 5.8590534480817963e-08, + "loss": 0.74499077, + "num_input_tokens_seen": 166358735, + "step": 7693, + "time_per_iteration": 2.7961883544921875 + }, + { + "auxiliary_loss_clip": 0.01169612, + "auxiliary_loss_mlp": 0.0102943, + "balance_loss_clip": 1.04836226, + "balance_loss_mlp": 1.02161026, + "epoch": 0.9251488005771659, + "flos": 10633581348480.0, + "grad_norm": 2.25640626599157, + "language_loss": 0.72681928, + "learning_rate": 5.840351392908349e-08, + "loss": 0.7488097, + "num_input_tokens_seen": 166374455, + "step": 7694, + "time_per_iteration": 2.6966006755828857 + }, + { + "auxiliary_loss_clip": 0.01155821, + "auxiliary_loss_mlp": 0.00886209, + "balance_loss_clip": 1.04377401, + "balance_loss_mlp": 1.00057697, + "epoch": 0.9252690434678049, + "flos": 23586416052480.0, + "grad_norm": 2.087776005251002, + "language_loss": 0.70450497, + "learning_rate": 5.821678791599205e-08, + "loss": 0.72492528, + "num_input_tokens_seen": 166393900, + "step": 7695, + "time_per_iteration": 2.862175226211548 + }, + { + "auxiliary_loss_clip": 0.01149208, + "auxiliary_loss_mlp": 0.01027842, + "balance_loss_clip": 1.046314, + "balance_loss_mlp": 1.02042389, + "epoch": 0.9253892863584441, + "flos": 21469158829440.0, + "grad_norm": 1.8116319094997624, + "language_loss": 0.8080743, + "learning_rate": 5.803035646986965e-08, + "loss": 0.82984477, + "num_input_tokens_seen": 166413235, + "step": 7696, + "time_per_iteration": 2.8027994632720947 + }, + { + "auxiliary_loss_clip": 0.0117086, + "auxiliary_loss_mlp": 0.01031592, + "balance_loss_clip": 1.04706621, + "balance_loss_mlp": 1.02312779, + "epoch": 0.9255095292490831, + "flos": 17456680304640.0, + "grad_norm": 3.4289619023358724, + "language_loss": 0.6783312, + "learning_rate": 5.7844219618998766e-08, + "loss": 0.70035577, + "num_input_tokens_seen": 166427560, + "step": 7697, + "time_per_iteration": 2.6427459716796875 + }, + { + "auxiliary_loss_clip": 0.01124922, + "auxiliary_loss_mlp": 0.01027896, + "balance_loss_clip": 1.03739214, + "balance_loss_mlp": 1.01975965, + "epoch": 0.9256297721397222, + "flos": 24750675584640.0, + "grad_norm": 1.7756961644809781, + "language_loss": 0.71832198, + "learning_rate": 5.765837739161505e-08, + "loss": 0.73985016, + "num_input_tokens_seen": 166446680, + "step": 7698, + "time_per_iteration": 3.8976738452911377 + }, + { + "auxiliary_loss_clip": 0.01140924, + "auxiliary_loss_mlp": 0.01025943, + "balance_loss_clip": 1.04286194, + "balance_loss_mlp": 1.01829529, + "epoch": 0.9257500150303614, + "flos": 23112215677440.0, + "grad_norm": 2.0817440614631773, + "language_loss": 0.74449211, + "learning_rate": 5.7472829815911504e-08, + "loss": 0.76616073, + "num_input_tokens_seen": 166465505, + "step": 7699, + "time_per_iteration": 3.7380058765411377 + }, + { + "auxiliary_loss_clip": 0.01142451, + "auxiliary_loss_mlp": 0.01025364, + "balance_loss_clip": 1.04190946, + "balance_loss_mlp": 1.01725173, + "epoch": 0.9258702579210004, + "flos": 22564685687040.0, + "grad_norm": 1.6443873645463014, + "language_loss": 0.81372827, + "learning_rate": 5.7287576920035164e-08, + "loss": 0.83540642, + "num_input_tokens_seen": 166484520, + "step": 7700, + "time_per_iteration": 2.766221761703491 + }, + { + "auxiliary_loss_clip": 0.0113729, + "auxiliary_loss_mlp": 0.01028709, + "balance_loss_clip": 1.04230857, + "balance_loss_mlp": 1.02104402, + "epoch": 0.9259905008116395, + "flos": 30004298703360.0, + "grad_norm": 1.872716526097365, + "language_loss": 0.76472741, + "learning_rate": 5.7102618732088435e-08, + "loss": 0.78638732, + "num_input_tokens_seen": 166503850, + "step": 7701, + "time_per_iteration": 3.7598953247070312 + }, + { + "auxiliary_loss_clip": 0.0114686, + "auxiliary_loss_mlp": 0.01028127, + "balance_loss_clip": 1.0455097, + "balance_loss_mlp": 1.02056289, + "epoch": 0.9261107437022786, + "flos": 24572128055040.0, + "grad_norm": 1.6401595063554317, + "language_loss": 0.74520278, + "learning_rate": 5.6917955280130216e-08, + "loss": 0.76695263, + "num_input_tokens_seen": 166525330, + "step": 7702, + "time_per_iteration": 2.8183112144470215 + }, + { + "auxiliary_loss_clip": 0.01157344, + "auxiliary_loss_mlp": 0.01030583, + "balance_loss_clip": 1.04660606, + "balance_loss_mlp": 1.02247107, + "epoch": 0.9262309865929177, + "flos": 22018448586240.0, + "grad_norm": 2.393152177496653, + "language_loss": 0.71757138, + "learning_rate": 5.6733586592172755e-08, + "loss": 0.73945063, + "num_input_tokens_seen": 166544825, + "step": 7703, + "time_per_iteration": 2.695805549621582 + }, + { + "auxiliary_loss_clip": 0.0114263, + "auxiliary_loss_mlp": 0.00884806, + "balance_loss_clip": 1.04115045, + "balance_loss_mlp": 1.00058472, + "epoch": 0.9263512294835567, + "flos": 20339481116160.0, + "grad_norm": 1.9412208139641223, + "language_loss": 0.80257869, + "learning_rate": 5.6549512696185244e-08, + "loss": 0.82285303, + "num_input_tokens_seen": 166563325, + "step": 7704, + "time_per_iteration": 3.741800546646118 + }, + { + "auxiliary_loss_clip": 0.01166181, + "auxiliary_loss_mlp": 0.01028613, + "balance_loss_clip": 1.04620862, + "balance_loss_mlp": 1.02039635, + "epoch": 0.9264714723741959, + "flos": 21215378263680.0, + "grad_norm": 1.578019294174429, + "language_loss": 0.68217599, + "learning_rate": 5.636573362009156e-08, + "loss": 0.70412397, + "num_input_tokens_seen": 166583385, + "step": 7705, + "time_per_iteration": 2.6168062686920166 + }, + { + "auxiliary_loss_clip": 0.0117164, + "auxiliary_loss_mlp": 0.01025433, + "balance_loss_clip": 1.04785585, + "balance_loss_mlp": 1.01717758, + "epoch": 0.926591715264835, + "flos": 18004964480640.0, + "grad_norm": 2.7976545578762635, + "language_loss": 0.7682184, + "learning_rate": 5.618224939177074e-08, + "loss": 0.79018915, + "num_input_tokens_seen": 166601290, + "step": 7706, + "time_per_iteration": 2.6660616397857666 + }, + { + "auxiliary_loss_clip": 0.01141144, + "auxiliary_loss_mlp": 0.01029032, + "balance_loss_clip": 1.04457974, + "balance_loss_mlp": 1.02126527, + "epoch": 0.926711958155474, + "flos": 36167969825280.0, + "grad_norm": 1.7026449598332831, + "language_loss": 0.70060402, + "learning_rate": 5.599906003905719e-08, + "loss": 0.72230577, + "num_input_tokens_seen": 166623835, + "step": 7707, + "time_per_iteration": 2.933335304260254 + }, + { + "auxiliary_loss_clip": 0.01155248, + "auxiliary_loss_mlp": 0.01025081, + "balance_loss_clip": 1.04796958, + "balance_loss_mlp": 1.0177201, + "epoch": 0.9268322010461132, + "flos": 21032736583680.0, + "grad_norm": 2.11096393181088, + "language_loss": 0.81744009, + "learning_rate": 5.581616558974023e-08, + "loss": 0.83924341, + "num_input_tokens_seen": 166642400, + "step": 7708, + "time_per_iteration": 2.731635570526123 + }, + { + "auxiliary_loss_clip": 0.01167296, + "auxiliary_loss_mlp": 0.00885752, + "balance_loss_clip": 1.04766726, + "balance_loss_mlp": 1.00055289, + "epoch": 0.9269524439367522, + "flos": 22964838174720.0, + "grad_norm": 2.4212287708649467, + "language_loss": 0.7916193, + "learning_rate": 5.5633566071565444e-08, + "loss": 0.81214982, + "num_input_tokens_seen": 166661640, + "step": 7709, + "time_per_iteration": 2.7575595378875732 + }, + { + "auxiliary_loss_clip": 0.01124393, + "auxiliary_loss_mlp": 0.0102568, + "balance_loss_clip": 1.03922057, + "balance_loss_mlp": 1.0185957, + "epoch": 0.9270726868273913, + "flos": 41975551468800.0, + "grad_norm": 2.198935173901466, + "language_loss": 0.70981908, + "learning_rate": 5.5451261512232896e-08, + "loss": 0.73131979, + "num_input_tokens_seen": 166684320, + "step": 7710, + "time_per_iteration": 3.0141561031341553 + }, + { + "auxiliary_loss_clip": 0.01162863, + "auxiliary_loss_mlp": 0.01029082, + "balance_loss_clip": 1.04375339, + "balance_loss_mlp": 1.02130413, + "epoch": 0.9271929297180305, + "flos": 19791771557760.0, + "grad_norm": 1.8964958603755764, + "language_loss": 0.62567055, + "learning_rate": 5.5269251939397576e-08, + "loss": 0.64759004, + "num_input_tokens_seen": 166703835, + "step": 7711, + "time_per_iteration": 2.6682498455047607 + }, + { + "auxiliary_loss_clip": 0.011416, + "auxiliary_loss_mlp": 0.01022541, + "balance_loss_clip": 1.04021144, + "balance_loss_mlp": 1.01467359, + "epoch": 0.9273131726086695, + "flos": 19968343839360.0, + "grad_norm": 1.9600396951951287, + "language_loss": 0.76687562, + "learning_rate": 5.508753738067073e-08, + "loss": 0.788517, + "num_input_tokens_seen": 166723375, + "step": 7712, + "time_per_iteration": 2.868300199508667 + }, + { + "auxiliary_loss_clip": 0.01160251, + "auxiliary_loss_mlp": 0.01030327, + "balance_loss_clip": 1.04362202, + "balance_loss_mlp": 1.02244139, + "epoch": 0.9274334154993086, + "flos": 23258587599360.0, + "grad_norm": 1.828643581990433, + "language_loss": 0.78992629, + "learning_rate": 5.4906117863617875e-08, + "loss": 0.81183201, + "num_input_tokens_seen": 166742760, + "step": 7713, + "time_per_iteration": 2.749389171600342 + }, + { + "auxiliary_loss_clip": 0.01133385, + "auxiliary_loss_mlp": 0.0102757, + "balance_loss_clip": 1.03919244, + "balance_loss_mlp": 1.02036989, + "epoch": 0.9275536583899477, + "flos": 31795343585280.0, + "grad_norm": 1.8670207585510334, + "language_loss": 0.78149569, + "learning_rate": 5.4724993415760533e-08, + "loss": 0.80310524, + "num_input_tokens_seen": 166761115, + "step": 7714, + "time_per_iteration": 2.842787027359009 + }, + { + "auxiliary_loss_clip": 0.01148661, + "auxiliary_loss_mlp": 0.00885931, + "balance_loss_clip": 1.04213476, + "balance_loss_mlp": 1.00056505, + "epoch": 0.9276739012805868, + "flos": 18696998885760.0, + "grad_norm": 2.006574632001041, + "language_loss": 0.74640572, + "learning_rate": 5.454416406457496e-08, + "loss": 0.76675165, + "num_input_tokens_seen": 166780210, + "step": 7715, + "time_per_iteration": 2.79884672164917 + }, + { + "auxiliary_loss_clip": 0.01157841, + "auxiliary_loss_mlp": 0.01025828, + "balance_loss_clip": 1.04381013, + "balance_loss_mlp": 1.01821327, + "epoch": 0.9277941441712259, + "flos": 13879079740800.0, + "grad_norm": 3.5815842590787743, + "language_loss": 0.74309754, + "learning_rate": 5.436362983749299e-08, + "loss": 0.7649343, + "num_input_tokens_seen": 166795380, + "step": 7716, + "time_per_iteration": 2.738054037094116 + }, + { + "auxiliary_loss_clip": 0.01130397, + "auxiliary_loss_mlp": 0.01024691, + "balance_loss_clip": 1.04326987, + "balance_loss_mlp": 1.01737201, + "epoch": 0.927914387061865, + "flos": 23258659426560.0, + "grad_norm": 3.759687098431908, + "language_loss": 0.64212197, + "learning_rate": 5.418339076190137e-08, + "loss": 0.6636728, + "num_input_tokens_seen": 166814890, + "step": 7717, + "time_per_iteration": 2.9046053886413574 + }, + { + "auxiliary_loss_clip": 0.01142159, + "auxiliary_loss_mlp": 0.01028937, + "balance_loss_clip": 1.0435679, + "balance_loss_mlp": 1.02136683, + "epoch": 0.9280346299525041, + "flos": 18073733068800.0, + "grad_norm": 1.7577587964001222, + "language_loss": 0.88656735, + "learning_rate": 5.400344686514202e-08, + "loss": 0.90827829, + "num_input_tokens_seen": 166832475, + "step": 7718, + "time_per_iteration": 2.76644229888916 + }, + { + "auxiliary_loss_clip": 0.01159707, + "auxiliary_loss_mlp": 0.01028712, + "balance_loss_clip": 1.04747963, + "balance_loss_mlp": 1.02138114, + "epoch": 0.9281548728431431, + "flos": 22342901160960.0, + "grad_norm": 1.8705695907829665, + "language_loss": 0.66433823, + "learning_rate": 5.38237981745131e-08, + "loss": 0.68622243, + "num_input_tokens_seen": 166850590, + "step": 7719, + "time_per_iteration": 2.6542599201202393 + }, + { + "auxiliary_loss_clip": 0.01160116, + "auxiliary_loss_mlp": 0.00885443, + "balance_loss_clip": 1.04451382, + "balance_loss_mlp": 1.00053966, + "epoch": 0.9282751157337822, + "flos": 18843765857280.0, + "grad_norm": 1.7916470428314453, + "language_loss": 0.81119788, + "learning_rate": 5.364444471726592e-08, + "loss": 0.83165336, + "num_input_tokens_seen": 166869795, + "step": 7720, + "time_per_iteration": 2.7981221675872803 + }, + { + "auxiliary_loss_clip": 0.01157119, + "auxiliary_loss_mlp": 0.01021667, + "balance_loss_clip": 1.04461753, + "balance_loss_mlp": 1.01359606, + "epoch": 0.9283953586244214, + "flos": 25556834476800.0, + "grad_norm": 3.0506603860138517, + "language_loss": 0.80290508, + "learning_rate": 5.346538652060939e-08, + "loss": 0.8246929, + "num_input_tokens_seen": 166891150, + "step": 7721, + "time_per_iteration": 2.7158684730529785 + }, + { + "auxiliary_loss_clip": 0.01145259, + "auxiliary_loss_mlp": 0.0102566, + "balance_loss_clip": 1.04432607, + "balance_loss_mlp": 1.01859367, + "epoch": 0.9285156015150604, + "flos": 18223480869120.0, + "grad_norm": 1.7947405329074433, + "language_loss": 0.70182681, + "learning_rate": 5.3286623611705994e-08, + "loss": 0.72353601, + "num_input_tokens_seen": 166909195, + "step": 7722, + "time_per_iteration": 2.7509677410125732 + }, + { + "auxiliary_loss_clip": 0.01055136, + "auxiliary_loss_mlp": 0.01001272, + "balance_loss_clip": 1.00451219, + "balance_loss_mlp": 0.99993724, + "epoch": 0.9286358444056995, + "flos": 66400017690240.0, + "grad_norm": 0.8114041038720153, + "language_loss": 0.60621321, + "learning_rate": 5.3108156017673824e-08, + "loss": 0.62677729, + "num_input_tokens_seen": 166970955, + "step": 7723, + "time_per_iteration": 3.3352696895599365 + }, + { + "auxiliary_loss_clip": 0.01148613, + "auxiliary_loss_mlp": 0.01025702, + "balance_loss_clip": 1.04526663, + "balance_loss_mlp": 1.01713681, + "epoch": 0.9287560872963386, + "flos": 22345630594560.0, + "grad_norm": 1.641072018803293, + "language_loss": 0.71681136, + "learning_rate": 5.2929983765586775e-08, + "loss": 0.73855454, + "num_input_tokens_seen": 166989735, + "step": 7724, + "time_per_iteration": 4.604909181594849 + }, + { + "auxiliary_loss_clip": 0.0117072, + "auxiliary_loss_mlp": 0.01033874, + "balance_loss_clip": 1.04828537, + "balance_loss_mlp": 1.02563035, + "epoch": 0.9288763301869777, + "flos": 25700225569920.0, + "grad_norm": 1.7994056800217906, + "language_loss": 0.62392473, + "learning_rate": 5.275210688247278e-08, + "loss": 0.6459707, + "num_input_tokens_seen": 167010060, + "step": 7725, + "time_per_iteration": 2.6977028846740723 + }, + { + "auxiliary_loss_clip": 0.0113113, + "auxiliary_loss_mlp": 0.01024569, + "balance_loss_clip": 1.04266167, + "balance_loss_mlp": 1.01717246, + "epoch": 0.9289965730776167, + "flos": 12312046028160.0, + "grad_norm": 5.233909486262779, + "language_loss": 0.8503741, + "learning_rate": 5.257452539531604e-08, + "loss": 0.87193108, + "num_input_tokens_seen": 167027130, + "step": 7726, + "time_per_iteration": 2.9070773124694824 + }, + { + "auxiliary_loss_clip": 0.01158525, + "auxiliary_loss_mlp": 0.01025602, + "balance_loss_clip": 1.04342866, + "balance_loss_mlp": 1.0183301, + "epoch": 0.9291168159682559, + "flos": 26685973486080.0, + "grad_norm": 1.6439500128722468, + "language_loss": 0.68293822, + "learning_rate": 5.2397239331055445e-08, + "loss": 0.70477951, + "num_input_tokens_seen": 167049130, + "step": 7727, + "time_per_iteration": 3.7103354930877686 + }, + { + "auxiliary_loss_clip": 0.01146309, + "auxiliary_loss_mlp": 0.01030988, + "balance_loss_clip": 1.04411483, + "balance_loss_mlp": 1.02314353, + "epoch": 0.929237058858895, + "flos": 14538256179840.0, + "grad_norm": 2.046240881676623, + "language_loss": 0.81043363, + "learning_rate": 5.2220248716585036e-08, + "loss": 0.83220661, + "num_input_tokens_seen": 167066810, + "step": 7728, + "time_per_iteration": 2.738413095474243 + }, + { + "auxiliary_loss_clip": 0.01152151, + "auxiliary_loss_mlp": 0.01032786, + "balance_loss_clip": 1.04439628, + "balance_loss_mlp": 1.02480245, + "epoch": 0.929357301749534, + "flos": 23835456023040.0, + "grad_norm": 2.3193677587004076, + "language_loss": 0.75654948, + "learning_rate": 5.204355357875445e-08, + "loss": 0.77839887, + "num_input_tokens_seen": 167085155, + "step": 7729, + "time_per_iteration": 2.7558224201202393 + }, + { + "auxiliary_loss_clip": 0.01144799, + "auxiliary_loss_mlp": 0.01027056, + "balance_loss_clip": 1.04176235, + "balance_loss_mlp": 1.01868701, + "epoch": 0.9294775446401732, + "flos": 12969319046400.0, + "grad_norm": 7.315673415982597, + "language_loss": 0.69845974, + "learning_rate": 5.1867153944367584e-08, + "loss": 0.72017825, + "num_input_tokens_seen": 167101545, + "step": 7730, + "time_per_iteration": 3.736067056655884 + }, + { + "auxiliary_loss_clip": 0.01147048, + "auxiliary_loss_mlp": 0.01028119, + "balance_loss_clip": 1.04556274, + "balance_loss_mlp": 1.02037644, + "epoch": 0.9295977875308122, + "flos": 26211809024640.0, + "grad_norm": 1.6512751513451525, + "language_loss": 0.73245806, + "learning_rate": 5.16910498401848e-08, + "loss": 0.75420976, + "num_input_tokens_seen": 167120995, + "step": 7731, + "time_per_iteration": 2.9207584857940674 + }, + { + "auxiliary_loss_clip": 0.01167159, + "auxiliary_loss_mlp": 0.01027464, + "balance_loss_clip": 1.04793131, + "balance_loss_mlp": 1.0200727, + "epoch": 0.9297180304214513, + "flos": 16472297105280.0, + "grad_norm": 2.6537980443728824, + "language_loss": 0.83386207, + "learning_rate": 5.151524129292073e-08, + "loss": 0.85580826, + "num_input_tokens_seen": 167138890, + "step": 7732, + "time_per_iteration": 2.605825185775757 + }, + { + "auxiliary_loss_clip": 0.01157718, + "auxiliary_loss_mlp": 0.01028241, + "balance_loss_clip": 1.0463593, + "balance_loss_mlp": 1.02043271, + "epoch": 0.9298382733120905, + "flos": 24060436859520.0, + "grad_norm": 1.9572452660439805, + "language_loss": 0.66553402, + "learning_rate": 5.1339728329245155e-08, + "loss": 0.68739367, + "num_input_tokens_seen": 167159455, + "step": 7733, + "time_per_iteration": 2.764850378036499 + }, + { + "auxiliary_loss_clip": 0.01172011, + "auxiliary_loss_mlp": 0.01033083, + "balance_loss_clip": 1.0479176, + "balance_loss_mlp": 1.02496445, + "epoch": 0.9299585162027295, + "flos": 22127652910080.0, + "grad_norm": 2.892759689112872, + "language_loss": 0.79134566, + "learning_rate": 5.116451097578367e-08, + "loss": 0.81339651, + "num_input_tokens_seen": 167178495, + "step": 7734, + "time_per_iteration": 2.635488510131836 + }, + { + "auxiliary_loss_clip": 0.0113974, + "auxiliary_loss_mlp": 0.01029877, + "balance_loss_clip": 1.04279661, + "balance_loss_mlp": 1.0228374, + "epoch": 0.9300787590933686, + "flos": 21471780522240.0, + "grad_norm": 1.6459896162687124, + "language_loss": 0.74487364, + "learning_rate": 5.0989589259115895e-08, + "loss": 0.76656979, + "num_input_tokens_seen": 167199380, + "step": 7735, + "time_per_iteration": 2.8255598545074463 + }, + { + "auxiliary_loss_clip": 0.01156508, + "auxiliary_loss_mlp": 0.01028925, + "balance_loss_clip": 1.04296684, + "balance_loss_mlp": 1.02004409, + "epoch": 0.9301990019840077, + "flos": 17779588594560.0, + "grad_norm": 1.9623974380260367, + "language_loss": 0.71811229, + "learning_rate": 5.081496320577816e-08, + "loss": 0.73996663, + "num_input_tokens_seen": 167216500, + "step": 7736, + "time_per_iteration": 2.6557445526123047 + }, + { + "auxiliary_loss_clip": 0.01041275, + "auxiliary_loss_mlp": 0.01003023, + "balance_loss_clip": 1.00639749, + "balance_loss_mlp": 1.00143719, + "epoch": 0.9303192448746468, + "flos": 58896122307840.0, + "grad_norm": 0.9155022363702463, + "language_loss": 0.61216497, + "learning_rate": 5.0640632842260835e-08, + "loss": 0.63260794, + "num_input_tokens_seen": 167276760, + "step": 7737, + "time_per_iteration": 3.3567147254943848 + }, + { + "auxiliary_loss_clip": 0.01141011, + "auxiliary_loss_mlp": 0.00885582, + "balance_loss_clip": 1.04517663, + "balance_loss_mlp": 1.00057757, + "epoch": 0.9304394877652858, + "flos": 57663522172800.0, + "grad_norm": 1.5838386837523413, + "language_loss": 0.72631723, + "learning_rate": 5.0466598195009426e-08, + "loss": 0.74658316, + "num_input_tokens_seen": 167303630, + "step": 7738, + "time_per_iteration": 3.094954013824463 + }, + { + "auxiliary_loss_clip": 0.01134699, + "auxiliary_loss_mlp": 0.0103088, + "balance_loss_clip": 1.04289722, + "balance_loss_mlp": 1.02346492, + "epoch": 0.930559730655925, + "flos": 20996143603200.0, + "grad_norm": 1.939049005958199, + "language_loss": 0.70297229, + "learning_rate": 5.0292859290425036e-08, + "loss": 0.72462815, + "num_input_tokens_seen": 167321500, + "step": 7739, + "time_per_iteration": 2.820643663406372 + }, + { + "auxiliary_loss_clip": 0.01166042, + "auxiliary_loss_mlp": 0.01023479, + "balance_loss_clip": 1.04551435, + "balance_loss_mlp": 1.01615703, + "epoch": 0.9306799735465641, + "flos": 23258264376960.0, + "grad_norm": 1.927337301771195, + "language_loss": 0.7768234, + "learning_rate": 5.011941615486348e-08, + "loss": 0.79871863, + "num_input_tokens_seen": 167340615, + "step": 7740, + "time_per_iteration": 2.6487200260162354 + }, + { + "auxiliary_loss_clip": 0.01167633, + "auxiliary_loss_mlp": 0.01030156, + "balance_loss_clip": 1.04657674, + "balance_loss_mlp": 1.02280426, + "epoch": 0.9308002164372031, + "flos": 15231547560960.0, + "grad_norm": 2.206001594990777, + "language_loss": 0.84644723, + "learning_rate": 4.994626881463659e-08, + "loss": 0.86842513, + "num_input_tokens_seen": 167356870, + "step": 7741, + "time_per_iteration": 2.583533525466919 + }, + { + "auxiliary_loss_clip": 0.01117378, + "auxiliary_loss_mlp": 0.01027859, + "balance_loss_clip": 1.0398941, + "balance_loss_mlp": 1.02055418, + "epoch": 0.9309204593278423, + "flos": 30847481539200.0, + "grad_norm": 1.9090561763558649, + "language_loss": 0.70795339, + "learning_rate": 4.9773417296009814e-08, + "loss": 0.72940576, + "num_input_tokens_seen": 167378390, + "step": 7742, + "time_per_iteration": 2.8837392330169678 + }, + { + "auxiliary_loss_clip": 0.01164701, + "auxiliary_loss_mlp": 0.01026119, + "balance_loss_clip": 1.04684532, + "balance_loss_mlp": 1.01841164, + "epoch": 0.9310407022184813, + "flos": 23037269950080.0, + "grad_norm": 1.7605492344779852, + "language_loss": 0.65004486, + "learning_rate": 4.960086162520527e-08, + "loss": 0.67195308, + "num_input_tokens_seen": 167398480, + "step": 7743, + "time_per_iteration": 2.717402935028076 + }, + { + "auxiliary_loss_clip": 0.01126568, + "auxiliary_loss_mlp": 0.01031183, + "balance_loss_clip": 1.04229093, + "balance_loss_mlp": 1.02380419, + "epoch": 0.9311609451091204, + "flos": 22127976132480.0, + "grad_norm": 2.1892059124020955, + "language_loss": 0.82373291, + "learning_rate": 4.942860182839936e-08, + "loss": 0.84531045, + "num_input_tokens_seen": 167416825, + "step": 7744, + "time_per_iteration": 2.8554537296295166 + }, + { + "auxiliary_loss_clip": 0.01148421, + "auxiliary_loss_mlp": 0.01029933, + "balance_loss_clip": 1.0438962, + "balance_loss_mlp": 1.0220232, + "epoch": 0.9312811879997596, + "flos": 21099206701440.0, + "grad_norm": 1.7944497362270986, + "language_loss": 0.79749244, + "learning_rate": 4.925663793172341e-08, + "loss": 0.81927598, + "num_input_tokens_seen": 167434785, + "step": 7745, + "time_per_iteration": 2.812192440032959 + }, + { + "auxiliary_loss_clip": 0.0104075, + "auxiliary_loss_mlp": 0.00876793, + "balance_loss_clip": 1.00568628, + "balance_loss_mlp": 1.00113547, + "epoch": 0.9314014308903986, + "flos": 67148179096320.0, + "grad_norm": 0.7807242783983449, + "language_loss": 0.56541133, + "learning_rate": 4.908496996126477e-08, + "loss": 0.5845868, + "num_input_tokens_seen": 167498245, + "step": 7746, + "time_per_iteration": 3.3359386920928955 + }, + { + "auxiliary_loss_clip": 0.01156888, + "auxiliary_loss_mlp": 0.01031205, + "balance_loss_clip": 1.0489347, + "balance_loss_mlp": 1.02318239, + "epoch": 0.9315216737810377, + "flos": 22565583527040.0, + "grad_norm": 1.8267341447831376, + "language_loss": 0.76333112, + "learning_rate": 4.89135979430646e-08, + "loss": 0.78521204, + "num_input_tokens_seen": 167518290, + "step": 7747, + "time_per_iteration": 2.743058919906616 + }, + { + "auxiliary_loss_clip": 0.01169072, + "auxiliary_loss_mlp": 0.01026623, + "balance_loss_clip": 1.04818821, + "balance_loss_mlp": 1.01842785, + "epoch": 0.9316419166716768, + "flos": 23984054588160.0, + "grad_norm": 1.7221823668987644, + "language_loss": 0.85508865, + "learning_rate": 4.874252190312078e-08, + "loss": 0.87704557, + "num_input_tokens_seen": 167538675, + "step": 7748, + "time_per_iteration": 2.729076385498047 + }, + { + "auxiliary_loss_clip": 0.0115265, + "auxiliary_loss_mlp": 0.01029839, + "balance_loss_clip": 1.04479742, + "balance_loss_mlp": 1.02169728, + "epoch": 0.9317621595623159, + "flos": 30230464688640.0, + "grad_norm": 1.5952192318827543, + "language_loss": 0.64702392, + "learning_rate": 4.857174186738477e-08, + "loss": 0.66884881, + "num_input_tokens_seen": 167562025, + "step": 7749, + "time_per_iteration": 2.805048942565918 + }, + { + "auxiliary_loss_clip": 0.01172371, + "auxiliary_loss_mlp": 0.01031678, + "balance_loss_clip": 1.05043364, + "balance_loss_mlp": 1.02363718, + "epoch": 0.931882402452955, + "flos": 15742735966080.0, + "grad_norm": 2.1568663967980175, + "language_loss": 0.73156214, + "learning_rate": 4.840125786176408e-08, + "loss": 0.75360262, + "num_input_tokens_seen": 167578230, + "step": 7750, + "time_per_iteration": 3.664116859436035 + }, + { + "auxiliary_loss_clip": 0.01146436, + "auxiliary_loss_mlp": 0.01027096, + "balance_loss_clip": 1.04359031, + "balance_loss_mlp": 1.01981831, + "epoch": 0.932002645343594, + "flos": 28366521154560.0, + "grad_norm": 1.7483320367975033, + "language_loss": 0.77414984, + "learning_rate": 4.823106991212067e-08, + "loss": 0.79588515, + "num_input_tokens_seen": 167597470, + "step": 7751, + "time_per_iteration": 3.8158175945281982 + }, + { + "auxiliary_loss_clip": 0.01159756, + "auxiliary_loss_mlp": 0.01027987, + "balance_loss_clip": 1.04449475, + "balance_loss_mlp": 1.02085233, + "epoch": 0.9321228882342332, + "flos": 15341146934400.0, + "grad_norm": 1.923671360753545, + "language_loss": 0.83667707, + "learning_rate": 4.806117804427212e-08, + "loss": 0.85855448, + "num_input_tokens_seen": 167615405, + "step": 7752, + "time_per_iteration": 3.653618335723877 + }, + { + "auxiliary_loss_clip": 0.0115397, + "auxiliary_loss_mlp": 0.01034499, + "balance_loss_clip": 1.04452085, + "balance_loss_mlp": 1.02633357, + "epoch": 0.9322431311248722, + "flos": 17895365107200.0, + "grad_norm": 2.378618421982233, + "language_loss": 0.64130926, + "learning_rate": 4.7891582283990926e-08, + "loss": 0.66319394, + "num_input_tokens_seen": 167634130, + "step": 7753, + "time_per_iteration": 2.7517998218536377 + }, + { + "auxiliary_loss_clip": 0.01130454, + "auxiliary_loss_mlp": 0.01021783, + "balance_loss_clip": 1.04088974, + "balance_loss_mlp": 1.01419222, + "epoch": 0.9323633740155113, + "flos": 24169713010560.0, + "grad_norm": 1.7008692370546032, + "language_loss": 0.72984397, + "learning_rate": 4.772228265700473e-08, + "loss": 0.75136638, + "num_input_tokens_seen": 167654990, + "step": 7754, + "time_per_iteration": 2.8001224994659424 + }, + { + "auxiliary_loss_clip": 0.01162173, + "auxiliary_loss_mlp": 0.01030547, + "balance_loss_clip": 1.04528797, + "balance_loss_mlp": 1.02266479, + "epoch": 0.9324836169061504, + "flos": 15043482927360.0, + "grad_norm": 1.9501496984259703, + "language_loss": 0.75593621, + "learning_rate": 4.75532791889961e-08, + "loss": 0.77786338, + "num_input_tokens_seen": 167671690, + "step": 7755, + "time_per_iteration": 2.779355049133301 + }, + { + "auxiliary_loss_clip": 0.01156252, + "auxiliary_loss_mlp": 0.01029676, + "balance_loss_clip": 1.04290974, + "balance_loss_mlp": 1.02165377, + "epoch": 0.9326038597967895, + "flos": 18624890332800.0, + "grad_norm": 1.887422132872404, + "language_loss": 0.65624237, + "learning_rate": 4.738457190560252e-08, + "loss": 0.67810166, + "num_input_tokens_seen": 167690800, + "step": 7756, + "time_per_iteration": 2.65189266204834 + }, + { + "auxiliary_loss_clip": 0.01128922, + "auxiliary_loss_mlp": 0.01030528, + "balance_loss_clip": 1.04381013, + "balance_loss_mlp": 1.02265406, + "epoch": 0.9327241026874286, + "flos": 18952646958720.0, + "grad_norm": 2.5240507386611597, + "language_loss": 0.78438359, + "learning_rate": 4.721616083241664e-08, + "loss": 0.80597806, + "num_input_tokens_seen": 167709055, + "step": 7757, + "time_per_iteration": 3.7402591705322266 + }, + { + "auxiliary_loss_clip": 0.01155817, + "auxiliary_loss_mlp": 0.01027443, + "balance_loss_clip": 1.0445435, + "balance_loss_mlp": 1.0196408, + "epoch": 0.9328443455780677, + "flos": 29570282668800.0, + "grad_norm": 1.611775152584153, + "language_loss": 0.77474767, + "learning_rate": 4.7048045994986684e-08, + "loss": 0.79658031, + "num_input_tokens_seen": 167729915, + "step": 7758, + "time_per_iteration": 2.8208279609680176 + }, + { + "auxiliary_loss_clip": 0.01157707, + "auxiliary_loss_mlp": 0.01031621, + "balance_loss_clip": 1.04735959, + "balance_loss_mlp": 1.02394962, + "epoch": 0.9329645884687068, + "flos": 30081722469120.0, + "grad_norm": 2.123250270338549, + "language_loss": 0.91167241, + "learning_rate": 4.688022741881559e-08, + "loss": 0.93356568, + "num_input_tokens_seen": 167750440, + "step": 7759, + "time_per_iteration": 2.6982381343841553 + }, + { + "auxiliary_loss_clip": 0.01154314, + "auxiliary_loss_mlp": 0.01026448, + "balance_loss_clip": 1.04435122, + "balance_loss_mlp": 1.01905441, + "epoch": 0.9330848313593458, + "flos": 21867982513920.0, + "grad_norm": 1.625848619614779, + "language_loss": 0.75280827, + "learning_rate": 4.671270512936076e-08, + "loss": 0.77461588, + "num_input_tokens_seen": 167769600, + "step": 7760, + "time_per_iteration": 2.764188289642334 + }, + { + "auxiliary_loss_clip": 0.01133349, + "auxiliary_loss_mlp": 0.010219, + "balance_loss_clip": 1.04148054, + "balance_loss_mlp": 1.01508391, + "epoch": 0.933205074249985, + "flos": 22127221946880.0, + "grad_norm": 1.76785641244908, + "language_loss": 0.82828867, + "learning_rate": 4.6545479152035884e-08, + "loss": 0.84984112, + "num_input_tokens_seen": 167788770, + "step": 7761, + "time_per_iteration": 2.826786994934082 + }, + { + "auxiliary_loss_clip": 0.01160521, + "auxiliary_loss_mlp": 0.01028902, + "balance_loss_clip": 1.04606652, + "balance_loss_mlp": 1.02165961, + "epoch": 0.9333253171406241, + "flos": 15341254675200.0, + "grad_norm": 5.472177247363793, + "language_loss": 0.76139486, + "learning_rate": 4.637854951220821e-08, + "loss": 0.78328907, + "num_input_tokens_seen": 167805555, + "step": 7762, + "time_per_iteration": 2.677725315093994 + }, + { + "auxiliary_loss_clip": 0.01131593, + "auxiliary_loss_mlp": 0.01027059, + "balance_loss_clip": 1.04025412, + "balance_loss_mlp": 1.01961994, + "epoch": 0.9334455600312631, + "flos": 15706142985600.0, + "grad_norm": 1.858481166056405, + "language_loss": 0.75222141, + "learning_rate": 4.621191623520171e-08, + "loss": 0.77380788, + "num_input_tokens_seen": 167823985, + "step": 7763, + "time_per_iteration": 2.7495105266571045 + }, + { + "auxiliary_loss_clip": 0.0112691, + "auxiliary_loss_mlp": 0.01026967, + "balance_loss_clip": 1.04213667, + "balance_loss_mlp": 1.01911104, + "epoch": 0.9335658029219023, + "flos": 22163563532160.0, + "grad_norm": 2.3042148439988805, + "language_loss": 0.84526527, + "learning_rate": 4.604557934629372e-08, + "loss": 0.866804, + "num_input_tokens_seen": 167843060, + "step": 7764, + "time_per_iteration": 2.929870843887329 + }, + { + "auxiliary_loss_clip": 0.01144298, + "auxiliary_loss_mlp": 0.01028404, + "balance_loss_clip": 1.04453039, + "balance_loss_mlp": 1.02113807, + "epoch": 0.9336860458125413, + "flos": 20266833859200.0, + "grad_norm": 10.2971279254338, + "language_loss": 0.80461782, + "learning_rate": 4.587953887071805e-08, + "loss": 0.82634485, + "num_input_tokens_seen": 167862880, + "step": 7765, + "time_per_iteration": 2.808717727661133 + }, + { + "auxiliary_loss_clip": 0.01144063, + "auxiliary_loss_mlp": 0.01026448, + "balance_loss_clip": 1.0391922, + "balance_loss_mlp": 1.01849675, + "epoch": 0.9338062887031804, + "flos": 20919689504640.0, + "grad_norm": 1.74881589035283, + "language_loss": 0.85742801, + "learning_rate": 4.5713794833662554e-08, + "loss": 0.87913316, + "num_input_tokens_seen": 167882095, + "step": 7766, + "time_per_iteration": 2.758901834487915 + }, + { + "auxiliary_loss_clip": 0.011685, + "auxiliary_loss_mlp": 0.01029053, + "balance_loss_clip": 1.04580188, + "balance_loss_mlp": 1.02085698, + "epoch": 0.9339265315938196, + "flos": 23221635482880.0, + "grad_norm": 1.8899308896788907, + "language_loss": 0.63342118, + "learning_rate": 4.5548347260270236e-08, + "loss": 0.6553967, + "num_input_tokens_seen": 167901385, + "step": 7767, + "time_per_iteration": 2.7036893367767334 + }, + { + "auxiliary_loss_clip": 0.01134905, + "auxiliary_loss_mlp": 0.01027328, + "balance_loss_clip": 1.0430578, + "balance_loss_mlp": 1.02032423, + "epoch": 0.9340467744844586, + "flos": 22820261932800.0, + "grad_norm": 1.8313998812745382, + "language_loss": 0.69350171, + "learning_rate": 4.538319617564012e-08, + "loss": 0.71512401, + "num_input_tokens_seen": 167920405, + "step": 7768, + "time_per_iteration": 2.8435862064361572 + }, + { + "auxiliary_loss_clip": 0.01146815, + "auxiliary_loss_mlp": 0.01027303, + "balance_loss_clip": 1.04173994, + "balance_loss_mlp": 1.01869631, + "epoch": 0.9341670173750977, + "flos": 23660428026240.0, + "grad_norm": 1.9952429008509316, + "language_loss": 0.74417996, + "learning_rate": 4.521834160482485e-08, + "loss": 0.76592112, + "num_input_tokens_seen": 167939145, + "step": 7769, + "time_per_iteration": 2.822293996810913 + }, + { + "auxiliary_loss_clip": 0.01161718, + "auxiliary_loss_mlp": 0.01023674, + "balance_loss_clip": 1.04622889, + "balance_loss_mlp": 1.01649117, + "epoch": 0.9342872602657368, + "flos": 24824256595200.0, + "grad_norm": 1.5304425760725404, + "language_loss": 0.81918454, + "learning_rate": 4.5053783572832846e-08, + "loss": 0.84103847, + "num_input_tokens_seen": 167959325, + "step": 7770, + "time_per_iteration": 2.7525041103363037 + }, + { + "auxiliary_loss_clip": 0.01159812, + "auxiliary_loss_mlp": 0.01026963, + "balance_loss_clip": 1.04601991, + "balance_loss_mlp": 1.0188446, + "epoch": 0.9344075031563759, + "flos": 25771831332480.0, + "grad_norm": 1.5830927895736446, + "language_loss": 0.76504171, + "learning_rate": 4.488952210462771e-08, + "loss": 0.7869094, + "num_input_tokens_seen": 167979530, + "step": 7771, + "time_per_iteration": 2.788525342941284 + }, + { + "auxiliary_loss_clip": 0.01169438, + "auxiliary_loss_mlp": 0.01027467, + "balance_loss_clip": 1.0486964, + "balance_loss_mlp": 1.01976311, + "epoch": 0.9345277460470149, + "flos": 25551303782400.0, + "grad_norm": 2.0298804423557075, + "language_loss": 0.85647839, + "learning_rate": 4.4725557225127495e-08, + "loss": 0.87844753, + "num_input_tokens_seen": 167997870, + "step": 7772, + "time_per_iteration": 2.7780792713165283 + }, + { + "auxiliary_loss_clip": 0.01158825, + "auxiliary_loss_mlp": 0.01023593, + "balance_loss_clip": 1.04525173, + "balance_loss_mlp": 1.01646447, + "epoch": 0.9346479889376541, + "flos": 34313112432000.0, + "grad_norm": 1.5065804707315937, + "language_loss": 0.79570842, + "learning_rate": 4.456188895920565e-08, + "loss": 0.81753266, + "num_input_tokens_seen": 168019625, + "step": 7773, + "time_per_iteration": 2.9016923904418945 + }, + { + "auxiliary_loss_clip": 0.01168649, + "auxiliary_loss_mlp": 0.01024538, + "balance_loss_clip": 1.04649079, + "balance_loss_mlp": 1.01675367, + "epoch": 0.9347682318282932, + "flos": 19093739581440.0, + "grad_norm": 3.0989344021184655, + "language_loss": 0.85162556, + "learning_rate": 4.439851733169031e-08, + "loss": 0.87355745, + "num_input_tokens_seen": 168037415, + "step": 7774, + "time_per_iteration": 2.71516752243042 + }, + { + "auxiliary_loss_clip": 0.01139061, + "auxiliary_loss_mlp": 0.01025371, + "balance_loss_clip": 1.0420742, + "balance_loss_mlp": 1.01817083, + "epoch": 0.9348884747189322, + "flos": 26249587153920.0, + "grad_norm": 2.2566629655276538, + "language_loss": 0.69521391, + "learning_rate": 4.4235442367365204e-08, + "loss": 0.71685815, + "num_input_tokens_seen": 168057725, + "step": 7775, + "time_per_iteration": 2.8137269020080566 + }, + { + "auxiliary_loss_clip": 0.01146078, + "auxiliary_loss_mlp": 0.01030134, + "balance_loss_clip": 1.04167974, + "balance_loss_mlp": 1.02232623, + "epoch": 0.9350087176095714, + "flos": 18333080242560.0, + "grad_norm": 2.7285469373947753, + "language_loss": 0.79269969, + "learning_rate": 4.4072664090968545e-08, + "loss": 0.81446183, + "num_input_tokens_seen": 168076110, + "step": 7776, + "time_per_iteration": 3.7546095848083496 + }, + { + "auxiliary_loss_clip": 0.01149185, + "auxiliary_loss_mlp": 0.01028614, + "balance_loss_clip": 1.04082048, + "balance_loss_mlp": 1.02119946, + "epoch": 0.9351289605002104, + "flos": 19318253541120.0, + "grad_norm": 1.8453474833253884, + "language_loss": 0.84576821, + "learning_rate": 4.391018252719347e-08, + "loss": 0.8675462, + "num_input_tokens_seen": 168095905, + "step": 7777, + "time_per_iteration": 2.8356873989105225 + }, + { + "auxiliary_loss_clip": 0.01150342, + "auxiliary_loss_mlp": 0.01023711, + "balance_loss_clip": 1.04159355, + "balance_loss_mlp": 1.01568782, + "epoch": 0.9352492033908495, + "flos": 18799990156800.0, + "grad_norm": 2.8406699104264432, + "language_loss": 0.69162989, + "learning_rate": 4.374799770068849e-08, + "loss": 0.71337044, + "num_input_tokens_seen": 168112580, + "step": 7778, + "time_per_iteration": 2.730087995529175 + }, + { + "auxiliary_loss_clip": 0.01158432, + "auxiliary_loss_mlp": 0.01028065, + "balance_loss_clip": 1.04734588, + "balance_loss_mlp": 1.02004814, + "epoch": 0.9353694462814887, + "flos": 29530134241920.0, + "grad_norm": 1.8635859369241707, + "language_loss": 0.74806577, + "learning_rate": 4.358610963605658e-08, + "loss": 0.76993072, + "num_input_tokens_seen": 168133030, + "step": 7779, + "time_per_iteration": 3.5080301761627197 + }, + { + "auxiliary_loss_clip": 0.01173159, + "auxiliary_loss_mlp": 0.0102245, + "balance_loss_clip": 1.0489645, + "balance_loss_mlp": 1.01461244, + "epoch": 0.9354896891721277, + "flos": 30665450390400.0, + "grad_norm": 2.12204767817985, + "language_loss": 0.68314183, + "learning_rate": 4.342451835785677e-08, + "loss": 0.70509791, + "num_input_tokens_seen": 168153940, + "step": 7780, + "time_per_iteration": 2.5513193607330322 + }, + { + "auxiliary_loss_clip": 0.01144961, + "auxiliary_loss_mlp": 0.0102573, + "balance_loss_clip": 1.04204273, + "balance_loss_mlp": 1.01848173, + "epoch": 0.9356099320627668, + "flos": 19463907191040.0, + "grad_norm": 1.595459103987431, + "language_loss": 0.75395399, + "learning_rate": 4.3263223890601665e-08, + "loss": 0.77566093, + "num_input_tokens_seen": 168172650, + "step": 7781, + "time_per_iteration": 2.58290433883667 + }, + { + "auxiliary_loss_clip": 0.01156585, + "auxiliary_loss_mlp": 0.00885254, + "balance_loss_clip": 1.04661298, + "balance_loss_mlp": 1.0006721, + "epoch": 0.9357301749534058, + "flos": 19098156954240.0, + "grad_norm": 1.6868209485707502, + "language_loss": 0.79502404, + "learning_rate": 4.31022262587597e-08, + "loss": 0.81544244, + "num_input_tokens_seen": 168191325, + "step": 7782, + "time_per_iteration": 3.3079304695129395 + }, + { + "auxiliary_loss_clip": 0.01160877, + "auxiliary_loss_mlp": 0.01029672, + "balance_loss_clip": 1.04844308, + "balance_loss_mlp": 1.02148819, + "epoch": 0.935850417844045, + "flos": 23550361776000.0, + "grad_norm": 1.5521426629826647, + "language_loss": 0.65986061, + "learning_rate": 4.2941525486754225e-08, + "loss": 0.68176615, + "num_input_tokens_seen": 168211645, + "step": 7783, + "time_per_iteration": 2.592782974243164 + }, + { + "auxiliary_loss_clip": 0.0113464, + "auxiliary_loss_mlp": 0.01029242, + "balance_loss_clip": 1.04210973, + "balance_loss_mlp": 1.02199125, + "epoch": 0.935970660734684, + "flos": 18588333265920.0, + "grad_norm": 1.8497757353874813, + "language_loss": 0.79483712, + "learning_rate": 4.278112159896286e-08, + "loss": 0.81647599, + "num_input_tokens_seen": 168229485, + "step": 7784, + "time_per_iteration": 2.5895798206329346 + }, + { + "auxiliary_loss_clip": 0.01139564, + "auxiliary_loss_mlp": 0.01023997, + "balance_loss_clip": 1.03760433, + "balance_loss_mlp": 1.01672792, + "epoch": 0.9360909036253231, + "flos": 20631255292800.0, + "grad_norm": 2.60694313791522, + "language_loss": 0.6763109, + "learning_rate": 4.2621014619719896e-08, + "loss": 0.69794655, + "num_input_tokens_seen": 168247250, + "step": 7785, + "time_per_iteration": 2.8116815090179443 + }, + { + "auxiliary_loss_clip": 0.01043415, + "auxiliary_loss_mlp": 0.01001674, + "balance_loss_clip": 1.00541413, + "balance_loss_mlp": 1.00022602, + "epoch": 0.9362111465159623, + "flos": 61791421052160.0, + "grad_norm": 0.7173869470472938, + "language_loss": 0.58642173, + "learning_rate": 4.246120457331215e-08, + "loss": 0.60687268, + "num_input_tokens_seen": 168309425, + "step": 7786, + "time_per_iteration": 3.339784622192383 + }, + { + "auxiliary_loss_clip": 0.01142644, + "auxiliary_loss_mlp": 0.01034292, + "balance_loss_clip": 1.04574573, + "balance_loss_mlp": 1.02629948, + "epoch": 0.9363313894066013, + "flos": 24170395368960.0, + "grad_norm": 2.0735644789728322, + "language_loss": 0.72076422, + "learning_rate": 4.2301691483983325e-08, + "loss": 0.74253356, + "num_input_tokens_seen": 168329545, + "step": 7787, + "time_per_iteration": 2.7938179969787598 + }, + { + "auxiliary_loss_clip": 0.01162914, + "auxiliary_loss_mlp": 0.01025631, + "balance_loss_clip": 1.04696155, + "balance_loss_mlp": 1.01793623, + "epoch": 0.9364516322972404, + "flos": 20120354196480.0, + "grad_norm": 1.5684428721287338, + "language_loss": 0.75916946, + "learning_rate": 4.214247537593163e-08, + "loss": 0.78105497, + "num_input_tokens_seen": 168348795, + "step": 7788, + "time_per_iteration": 2.890300750732422 + }, + { + "auxiliary_loss_clip": 0.01148748, + "auxiliary_loss_mlp": 0.0102535, + "balance_loss_clip": 1.04196882, + "balance_loss_mlp": 1.01844501, + "epoch": 0.9365718751878795, + "flos": 20703758895360.0, + "grad_norm": 2.1701317991439657, + "language_loss": 0.80556154, + "learning_rate": 4.1983556273309293e-08, + "loss": 0.82730246, + "num_input_tokens_seen": 168367545, + "step": 7789, + "time_per_iteration": 2.8272128105163574 + }, + { + "auxiliary_loss_clip": 0.01170001, + "auxiliary_loss_mlp": 0.01028138, + "balance_loss_clip": 1.04717445, + "balance_loss_mlp": 1.01953745, + "epoch": 0.9366921180785186, + "flos": 18655270260480.0, + "grad_norm": 3.4698021428729593, + "language_loss": 0.68651265, + "learning_rate": 4.182493420022526e-08, + "loss": 0.70849407, + "num_input_tokens_seen": 168383215, + "step": 7790, + "time_per_iteration": 2.6257543563842773 + }, + { + "auxiliary_loss_clip": 0.01129784, + "auxiliary_loss_mlp": 0.01026145, + "balance_loss_clip": 1.04073513, + "balance_loss_mlp": 1.01821733, + "epoch": 0.9368123609691577, + "flos": 25774955815680.0, + "grad_norm": 1.7691236766109129, + "language_loss": 0.78635573, + "learning_rate": 4.166660918074139e-08, + "loss": 0.80791497, + "num_input_tokens_seen": 168403120, + "step": 7791, + "time_per_iteration": 2.7107889652252197 + }, + { + "auxiliary_loss_clip": 0.0113609, + "auxiliary_loss_mlp": 0.01030073, + "balance_loss_clip": 1.04236865, + "balance_loss_mlp": 1.02263999, + "epoch": 0.9369326038597968, + "flos": 25553386771200.0, + "grad_norm": 1.385117157416611, + "language_loss": 0.73584676, + "learning_rate": 4.15085812388758e-08, + "loss": 0.7575084, + "num_input_tokens_seen": 168425340, + "step": 7792, + "time_per_iteration": 2.764021158218384 + }, + { + "auxiliary_loss_clip": 0.0114912, + "auxiliary_loss_mlp": 0.01030368, + "balance_loss_clip": 1.04464161, + "balance_loss_mlp": 1.02202952, + "epoch": 0.9370528467504359, + "flos": 23220019370880.0, + "grad_norm": 2.0873315903987093, + "language_loss": 0.7852214, + "learning_rate": 4.135085039860153e-08, + "loss": 0.80701625, + "num_input_tokens_seen": 168444740, + "step": 7793, + "time_per_iteration": 2.845499277114868 + }, + { + "auxiliary_loss_clip": 0.01147437, + "auxiliary_loss_mlp": 0.0102673, + "balance_loss_clip": 1.0457896, + "balance_loss_mlp": 1.01817632, + "epoch": 0.9371730896410749, + "flos": 24967468120320.0, + "grad_norm": 2.625912440770605, + "language_loss": 0.78591514, + "learning_rate": 4.1193416683845906e-08, + "loss": 0.80765682, + "num_input_tokens_seen": 168463670, + "step": 7794, + "time_per_iteration": 2.7394628524780273 + }, + { + "auxiliary_loss_clip": 0.01140876, + "auxiliary_loss_mlp": 0.01021982, + "balance_loss_clip": 1.04285026, + "balance_loss_mlp": 1.0145551, + "epoch": 0.9372933325317141, + "flos": 15553091134080.0, + "grad_norm": 4.449312144387183, + "language_loss": 0.8336606, + "learning_rate": 4.103628011849136e-08, + "loss": 0.85528922, + "num_input_tokens_seen": 168479030, + "step": 7795, + "time_per_iteration": 2.86855149269104 + }, + { + "auxiliary_loss_clip": 0.01150538, + "auxiliary_loss_mlp": 0.01026886, + "balance_loss_clip": 1.0455153, + "balance_loss_mlp": 1.01840377, + "epoch": 0.9374135754223532, + "flos": 21871861182720.0, + "grad_norm": 1.9147330667404234, + "language_loss": 0.75903112, + "learning_rate": 4.0879440726375506e-08, + "loss": 0.78080535, + "num_input_tokens_seen": 168496815, + "step": 7796, + "time_per_iteration": 2.7990915775299072 + }, + { + "auxiliary_loss_clip": 0.01144127, + "auxiliary_loss_mlp": 0.0102637, + "balance_loss_clip": 1.03908932, + "balance_loss_mlp": 1.01795959, + "epoch": 0.9375338183129922, + "flos": 22631048064000.0, + "grad_norm": 2.5835642120678703, + "language_loss": 0.56218052, + "learning_rate": 4.0722898531291074e-08, + "loss": 0.58388549, + "num_input_tokens_seen": 168514055, + "step": 7797, + "time_per_iteration": 2.9611082077026367 + }, + { + "auxiliary_loss_clip": 0.01155717, + "auxiliary_loss_mlp": 0.01025197, + "balance_loss_clip": 1.04552102, + "balance_loss_mlp": 1.0173645, + "epoch": 0.9376540612036314, + "flos": 26104292640000.0, + "grad_norm": 4.1073986693496485, + "language_loss": 0.76833689, + "learning_rate": 4.0566653556985295e-08, + "loss": 0.79014599, + "num_input_tokens_seen": 168534600, + "step": 7798, + "time_per_iteration": 2.812196969985962 + }, + { + "auxiliary_loss_clip": 0.01115746, + "auxiliary_loss_mlp": 0.01030692, + "balance_loss_clip": 1.04061687, + "balance_loss_mlp": 1.02232385, + "epoch": 0.9377743040942704, + "flos": 19717580016000.0, + "grad_norm": 14.195470270377788, + "language_loss": 0.81481159, + "learning_rate": 4.0410705827159886e-08, + "loss": 0.83627594, + "num_input_tokens_seen": 168551895, + "step": 7799, + "time_per_iteration": 2.9966795444488525 + }, + { + "auxiliary_loss_clip": 0.01145642, + "auxiliary_loss_mlp": 0.01029176, + "balance_loss_clip": 1.04136705, + "balance_loss_mlp": 1.02087283, + "epoch": 0.9378945469849095, + "flos": 15267530010240.0, + "grad_norm": 2.049503537781993, + "language_loss": 0.71314347, + "learning_rate": 4.0255055365472356e-08, + "loss": 0.73489165, + "num_input_tokens_seen": 168569990, + "step": 7800, + "time_per_iteration": 2.848933458328247 + }, + { + "auxiliary_loss_clip": 0.01120181, + "auxiliary_loss_mlp": 0.01032953, + "balance_loss_clip": 1.03605688, + "balance_loss_mlp": 1.02513289, + "epoch": 0.9380147898755486, + "flos": 20591394174720.0, + "grad_norm": 3.4577127350907206, + "language_loss": 0.74783945, + "learning_rate": 4.009970219553471e-08, + "loss": 0.76937073, + "num_input_tokens_seen": 168586940, + "step": 7801, + "time_per_iteration": 2.9042060375213623 + }, + { + "auxiliary_loss_clip": 0.0116358, + "auxiliary_loss_mlp": 0.01032359, + "balance_loss_clip": 1.0459404, + "balance_loss_mlp": 1.02388346, + "epoch": 0.9381350327661877, + "flos": 26281116316800.0, + "grad_norm": 2.71822869394266, + "language_loss": 0.76709366, + "learning_rate": 3.99446463409141e-08, + "loss": 0.78905302, + "num_input_tokens_seen": 168604795, + "step": 7802, + "time_per_iteration": 4.690422058105469 + }, + { + "auxiliary_loss_clip": 0.01165059, + "auxiliary_loss_mlp": 0.01028597, + "balance_loss_clip": 1.04496491, + "balance_loss_mlp": 1.02077079, + "epoch": 0.9382552756568268, + "flos": 23586344225280.0, + "grad_norm": 2.1988645963797024, + "language_loss": 0.68582082, + "learning_rate": 3.978988782513215e-08, + "loss": 0.70775729, + "num_input_tokens_seen": 168622290, + "step": 7803, + "time_per_iteration": 2.7139792442321777 + }, + { + "auxiliary_loss_clip": 0.01164489, + "auxiliary_loss_mlp": 0.01025677, + "balance_loss_clip": 1.04707384, + "balance_loss_mlp": 1.0178659, + "epoch": 0.9383755185474659, + "flos": 28438809275520.0, + "grad_norm": 1.9036277847867944, + "language_loss": 0.76452887, + "learning_rate": 3.963542667166586e-08, + "loss": 0.78643054, + "num_input_tokens_seen": 168642395, + "step": 7804, + "time_per_iteration": 3.704569101333618 + }, + { + "auxiliary_loss_clip": 0.01131959, + "auxiliary_loss_mlp": 0.01025371, + "balance_loss_clip": 1.04469585, + "balance_loss_mlp": 1.01842082, + "epoch": 0.938495761438105, + "flos": 20449583280000.0, + "grad_norm": 1.7036065633129898, + "language_loss": 0.68205488, + "learning_rate": 3.9481262903946486e-08, + "loss": 0.70362818, + "num_input_tokens_seen": 168661840, + "step": 7805, + "time_per_iteration": 3.0095784664154053 + }, + { + "auxiliary_loss_clip": 0.01036309, + "auxiliary_loss_mlp": 0.01001402, + "balance_loss_clip": 1.00580096, + "balance_loss_mlp": 0.99978095, + "epoch": 0.938616004328744, + "flos": 69302711658240.0, + "grad_norm": 0.7784628265935776, + "language_loss": 0.54440802, + "learning_rate": 3.932739654536066e-08, + "loss": 0.56478518, + "num_input_tokens_seen": 168724540, + "step": 7806, + "time_per_iteration": 3.3941519260406494 + }, + { + "auxiliary_loss_clip": 0.01157698, + "auxiliary_loss_mlp": 0.01019158, + "balance_loss_clip": 1.04605794, + "balance_loss_mlp": 1.01204705, + "epoch": 0.9387362472193832, + "flos": 18911636605440.0, + "grad_norm": 2.957909049595549, + "language_loss": 0.74006021, + "learning_rate": 3.917382761925014e-08, + "loss": 0.76182872, + "num_input_tokens_seen": 168740375, + "step": 7807, + "time_per_iteration": 2.652167797088623 + }, + { + "auxiliary_loss_clip": 0.01156259, + "auxiliary_loss_mlp": 0.01030047, + "balance_loss_clip": 1.04563332, + "balance_loss_mlp": 1.02242327, + "epoch": 0.9388564901100223, + "flos": 26501967089280.0, + "grad_norm": 1.7083014171727562, + "language_loss": 0.79270738, + "learning_rate": 3.9020556148910754e-08, + "loss": 0.81457043, + "num_input_tokens_seen": 168759730, + "step": 7808, + "time_per_iteration": 3.684765577316284 + }, + { + "auxiliary_loss_clip": 0.01047455, + "auxiliary_loss_mlp": 0.01001753, + "balance_loss_clip": 1.00604033, + "balance_loss_mlp": 1.0002985, + "epoch": 0.9389767330006613, + "flos": 58941083157120.0, + "grad_norm": 0.7152847056358815, + "language_loss": 0.56678575, + "learning_rate": 3.8867582157593895e-08, + "loss": 0.58727783, + "num_input_tokens_seen": 168813935, + "step": 7809, + "time_per_iteration": 3.2357828617095947 + }, + { + "auxiliary_loss_clip": 0.01157866, + "auxiliary_loss_mlp": 0.01029887, + "balance_loss_clip": 1.04779005, + "balance_loss_mlp": 1.02228189, + "epoch": 0.9390969758913005, + "flos": 31102554994560.0, + "grad_norm": 1.714619187961966, + "language_loss": 0.76715171, + "learning_rate": 3.871490566850544e-08, + "loss": 0.78902918, + "num_input_tokens_seen": 168838145, + "step": 7810, + "time_per_iteration": 2.907594680786133 + }, + { + "auxiliary_loss_clip": 0.01144514, + "auxiliary_loss_mlp": 0.0102685, + "balance_loss_clip": 1.0431993, + "balance_loss_mlp": 1.01855946, + "epoch": 0.9392172187819395, + "flos": 22419391173120.0, + "grad_norm": 1.6481203095471806, + "language_loss": 0.70712388, + "learning_rate": 3.856252670480642e-08, + "loss": 0.72883749, + "num_input_tokens_seen": 168856805, + "step": 7811, + "time_per_iteration": 2.8140504360198975 + }, + { + "auxiliary_loss_clip": 0.01143158, + "auxiliary_loss_mlp": 0.01029304, + "balance_loss_clip": 1.0387814, + "balance_loss_mlp": 1.02164149, + "epoch": 0.9393374616725786, + "flos": 19719483436800.0, + "grad_norm": 1.7290177451651552, + "language_loss": 0.81731141, + "learning_rate": 3.841044528961279e-08, + "loss": 0.83903605, + "num_input_tokens_seen": 168874600, + "step": 7812, + "time_per_iteration": 2.779477119445801 + }, + { + "auxiliary_loss_clip": 0.01170597, + "auxiliary_loss_mlp": 0.01021931, + "balance_loss_clip": 1.04701221, + "balance_loss_mlp": 1.01434898, + "epoch": 0.9394577045632178, + "flos": 24170215800960.0, + "grad_norm": 1.8889698087473852, + "language_loss": 0.78809756, + "learning_rate": 3.825866144599477e-08, + "loss": 0.81002283, + "num_input_tokens_seen": 168893655, + "step": 7813, + "time_per_iteration": 2.7397210597991943 + }, + { + "auxiliary_loss_clip": 0.01147813, + "auxiliary_loss_mlp": 0.01023116, + "balance_loss_clip": 1.04175019, + "balance_loss_mlp": 1.01530147, + "epoch": 0.9395779474538568, + "flos": 19023929498880.0, + "grad_norm": 2.4226807713022147, + "language_loss": 0.75572199, + "learning_rate": 3.8107175196978145e-08, + "loss": 0.77743125, + "num_input_tokens_seen": 168909960, + "step": 7814, + "time_per_iteration": 2.6854655742645264 + }, + { + "auxiliary_loss_clip": 0.01137451, + "auxiliary_loss_mlp": 0.01028124, + "balance_loss_clip": 1.04231441, + "balance_loss_mlp": 1.02093267, + "epoch": 0.9396981903444959, + "flos": 14319129260160.0, + "grad_norm": 1.7227973629830895, + "language_loss": 0.76636255, + "learning_rate": 3.7955986565542996e-08, + "loss": 0.78801823, + "num_input_tokens_seen": 168928040, + "step": 7815, + "time_per_iteration": 2.8292994499206543 + }, + { + "auxiliary_loss_clip": 0.01138865, + "auxiliary_loss_mlp": 0.01027346, + "balance_loss_clip": 1.04141235, + "balance_loss_mlp": 1.01964235, + "epoch": 0.9398184332351349, + "flos": 34787564202240.0, + "grad_norm": 1.71104367298893, + "language_loss": 0.68153763, + "learning_rate": 3.780509557462497e-08, + "loss": 0.7031998, + "num_input_tokens_seen": 168948240, + "step": 7816, + "time_per_iteration": 2.9297878742218018 + }, + { + "auxiliary_loss_clip": 0.01148224, + "auxiliary_loss_mlp": 0.01032495, + "balance_loss_clip": 1.04257679, + "balance_loss_mlp": 1.02497339, + "epoch": 0.9399386761257741, + "flos": 25372253462400.0, + "grad_norm": 2.7883504920919853, + "language_loss": 0.75450456, + "learning_rate": 3.765450224711375e-08, + "loss": 0.77631176, + "num_input_tokens_seen": 168968745, + "step": 7817, + "time_per_iteration": 2.8104538917541504 + }, + { + "auxiliary_loss_clip": 0.01142425, + "auxiliary_loss_mlp": 0.01025189, + "balance_loss_clip": 1.04336119, + "balance_loss_mlp": 1.0177263, + "epoch": 0.9400589190164131, + "flos": 27304965584640.0, + "grad_norm": 1.6896111381735521, + "language_loss": 0.79755414, + "learning_rate": 3.750420660585396e-08, + "loss": 0.81923026, + "num_input_tokens_seen": 168990685, + "step": 7818, + "time_per_iteration": 2.793888807296753 + }, + { + "auxiliary_loss_clip": 0.01167746, + "auxiliary_loss_mlp": 0.01025928, + "balance_loss_clip": 1.04759407, + "balance_loss_mlp": 1.01831317, + "epoch": 0.9401791619070522, + "flos": 23399859790080.0, + "grad_norm": 1.6728684217431933, + "language_loss": 0.79274327, + "learning_rate": 3.735420867364603e-08, + "loss": 0.81468004, + "num_input_tokens_seen": 169011665, + "step": 7819, + "time_per_iteration": 2.6987576484680176 + }, + { + "auxiliary_loss_clip": 0.01111016, + "auxiliary_loss_mlp": 0.01020196, + "balance_loss_clip": 1.03579497, + "balance_loss_mlp": 1.01308501, + "epoch": 0.9402994047976914, + "flos": 35881403120640.0, + "grad_norm": 1.5391715884923634, + "language_loss": 0.61743784, + "learning_rate": 3.7204508473244186e-08, + "loss": 0.6387499, + "num_input_tokens_seen": 169035290, + "step": 7820, + "time_per_iteration": 3.0375707149505615 + }, + { + "auxiliary_loss_clip": 0.01107509, + "auxiliary_loss_mlp": 0.01023403, + "balance_loss_clip": 1.03956127, + "balance_loss_mlp": 1.01599443, + "epoch": 0.9404196476883304, + "flos": 22236821320320.0, + "grad_norm": 1.6111685239798728, + "language_loss": 0.69469935, + "learning_rate": 3.7055106027357395e-08, + "loss": 0.71600848, + "num_input_tokens_seen": 169055155, + "step": 7821, + "time_per_iteration": 2.9824187755584717 + }, + { + "auxiliary_loss_clip": 0.01155813, + "auxiliary_loss_mlp": 0.01024999, + "balance_loss_clip": 1.04648781, + "balance_loss_mlp": 1.01769137, + "epoch": 0.9405398905789695, + "flos": 18915802583040.0, + "grad_norm": 2.087021677942555, + "language_loss": 0.71789598, + "learning_rate": 3.690600135865063e-08, + "loss": 0.73970401, + "num_input_tokens_seen": 169072080, + "step": 7822, + "time_per_iteration": 2.7587575912475586 + }, + { + "auxiliary_loss_clip": 0.01026327, + "auxiliary_loss_mlp": 0.01000946, + "balance_loss_clip": 1.0062685, + "balance_loss_mlp": 0.99952728, + "epoch": 0.9406601334696086, + "flos": 70274130048000.0, + "grad_norm": 0.8611319148098536, + "language_loss": 0.58146143, + "learning_rate": 3.675719448974246e-08, + "loss": 0.60173416, + "num_input_tokens_seen": 169137170, + "step": 7823, + "time_per_iteration": 3.415146827697754 + }, + { + "auxiliary_loss_clip": 0.01128086, + "auxiliary_loss_mlp": 0.00886109, + "balance_loss_clip": 1.04258657, + "balance_loss_mlp": 1.00062275, + "epoch": 0.9407803763602477, + "flos": 22165071903360.0, + "grad_norm": 1.9869893665904603, + "language_loss": 0.59851569, + "learning_rate": 3.6608685443207054e-08, + "loss": 0.61865759, + "num_input_tokens_seen": 169156320, + "step": 7824, + "time_per_iteration": 2.910057306289673 + }, + { + "auxiliary_loss_clip": 0.01140918, + "auxiliary_loss_mlp": 0.01024742, + "balance_loss_clip": 1.04373193, + "balance_loss_mlp": 1.01716638, + "epoch": 0.9409006192508867, + "flos": 18879496911360.0, + "grad_norm": 3.4942903077399707, + "language_loss": 0.66373855, + "learning_rate": 3.646047424157306e-08, + "loss": 0.68539512, + "num_input_tokens_seen": 169173295, + "step": 7825, + "time_per_iteration": 2.800605058670044 + }, + { + "auxiliary_loss_clip": 0.01153653, + "auxiliary_loss_mlp": 0.01028066, + "balance_loss_clip": 1.04712176, + "balance_loss_mlp": 1.01901817, + "epoch": 0.9410208621415259, + "flos": 23368258800000.0, + "grad_norm": 3.4737975565733854, + "language_loss": 0.68355531, + "learning_rate": 3.631256090732382e-08, + "loss": 0.70537245, + "num_input_tokens_seen": 169193755, + "step": 7826, + "time_per_iteration": 2.7626562118530273 + }, + { + "auxiliary_loss_clip": 0.01142425, + "auxiliary_loss_mlp": 0.01026341, + "balance_loss_clip": 1.04628134, + "balance_loss_mlp": 1.01911664, + "epoch": 0.941141105032165, + "flos": 22742227635840.0, + "grad_norm": 1.6741053825186536, + "language_loss": 0.82796061, + "learning_rate": 3.6164945462897833e-08, + "loss": 0.8496483, + "num_input_tokens_seen": 169213045, + "step": 7827, + "time_per_iteration": 3.7790234088897705 + }, + { + "auxiliary_loss_clip": 0.01156382, + "auxiliary_loss_mlp": 0.00884846, + "balance_loss_clip": 1.04561543, + "balance_loss_mlp": 1.00062418, + "epoch": 0.941261347922804, + "flos": 20704908130560.0, + "grad_norm": 1.6911087838356198, + "language_loss": 0.75530899, + "learning_rate": 3.6017627930687856e-08, + "loss": 0.77572131, + "num_input_tokens_seen": 169232870, + "step": 7828, + "time_per_iteration": 3.8301286697387695 + }, + { + "auxiliary_loss_clip": 0.01123079, + "auxiliary_loss_mlp": 0.01022841, + "balance_loss_clip": 1.03830302, + "balance_loss_mlp": 1.01552153, + "epoch": 0.9413815908134432, + "flos": 19421998997760.0, + "grad_norm": 2.510100078303686, + "language_loss": 0.77339125, + "learning_rate": 3.587060833304267e-08, + "loss": 0.79485041, + "num_input_tokens_seen": 169251060, + "step": 7829, + "time_per_iteration": 3.800568103790283 + }, + { + "auxiliary_loss_clip": 0.01165466, + "auxiliary_loss_mlp": 0.01032425, + "balance_loss_clip": 1.0495975, + "balance_loss_mlp": 1.02517116, + "epoch": 0.9415018337040822, + "flos": 17493452853120.0, + "grad_norm": 2.366999258069712, + "language_loss": 0.6425395, + "learning_rate": 3.5723886692264225e-08, + "loss": 0.66451836, + "num_input_tokens_seen": 169268600, + "step": 7830, + "time_per_iteration": 2.6475725173950195 + }, + { + "auxiliary_loss_clip": 0.01144683, + "auxiliary_loss_mlp": 0.0102896, + "balance_loss_clip": 1.04092216, + "balance_loss_mlp": 1.02154565, + "epoch": 0.9416220765947213, + "flos": 31831613343360.0, + "grad_norm": 2.3790935497024615, + "language_loss": 0.61757731, + "learning_rate": 3.557746303061071e-08, + "loss": 0.63931382, + "num_input_tokens_seen": 169290355, + "step": 7831, + "time_per_iteration": 2.8759188652038574 + }, + { + "auxiliary_loss_clip": 0.0114649, + "auxiliary_loss_mlp": 0.01026014, + "balance_loss_clip": 1.04125619, + "balance_loss_mlp": 1.01852179, + "epoch": 0.9417423194853605, + "flos": 23511973115520.0, + "grad_norm": 1.6244624698782386, + "language_loss": 0.72534776, + "learning_rate": 3.543133737029391e-08, + "loss": 0.74707276, + "num_input_tokens_seen": 169310865, + "step": 7832, + "time_per_iteration": 2.7497591972351074 + }, + { + "auxiliary_loss_clip": 0.01160577, + "auxiliary_loss_mlp": 0.0103169, + "balance_loss_clip": 1.04479754, + "balance_loss_mlp": 1.02385151, + "epoch": 0.9418625623759995, + "flos": 23915106432000.0, + "grad_norm": 1.6873248228444904, + "language_loss": 0.69179434, + "learning_rate": 3.5285509733481214e-08, + "loss": 0.71371698, + "num_input_tokens_seen": 169330590, + "step": 7833, + "time_per_iteration": 3.6127400398254395 + }, + { + "auxiliary_loss_clip": 0.01154068, + "auxiliary_loss_mlp": 0.01029941, + "balance_loss_clip": 1.04405189, + "balance_loss_mlp": 1.02157855, + "epoch": 0.9419828052666386, + "flos": 18076965292800.0, + "grad_norm": 1.8477529371624393, + "language_loss": 0.76515734, + "learning_rate": 3.513998014229469e-08, + "loss": 0.7869975, + "num_input_tokens_seen": 169349540, + "step": 7834, + "time_per_iteration": 2.6630916595458984 + }, + { + "auxiliary_loss_clip": 0.011406, + "auxiliary_loss_mlp": 0.0102704, + "balance_loss_clip": 1.0442791, + "balance_loss_mlp": 1.01963151, + "epoch": 0.9421030481572777, + "flos": 17712328377600.0, + "grad_norm": 2.3508396539905783, + "language_loss": 0.86499161, + "learning_rate": 3.499474861881069e-08, + "loss": 0.88666809, + "num_input_tokens_seen": 169366765, + "step": 7835, + "time_per_iteration": 2.739558458328247 + }, + { + "auxiliary_loss_clip": 0.01120177, + "auxiliary_loss_mlp": 0.01026386, + "balance_loss_clip": 1.04016483, + "balance_loss_mlp": 1.01858413, + "epoch": 0.9422232910479168, + "flos": 20194114775040.0, + "grad_norm": 2.302010079203908, + "language_loss": 0.68489456, + "learning_rate": 3.4849815185061136e-08, + "loss": 0.70636022, + "num_input_tokens_seen": 169386655, + "step": 7836, + "time_per_iteration": 2.9039740562438965 + }, + { + "auxiliary_loss_clip": 0.0115664, + "auxiliary_loss_mlp": 0.01024854, + "balance_loss_clip": 1.04322875, + "balance_loss_mlp": 1.0175643, + "epoch": 0.9423435339385559, + "flos": 18442571875200.0, + "grad_norm": 1.9494745417226422, + "language_loss": 0.75962877, + "learning_rate": 3.470517986303223e-08, + "loss": 0.78144372, + "num_input_tokens_seen": 169405640, + "step": 7837, + "time_per_iteration": 2.6551151275634766 + }, + { + "auxiliary_loss_clip": 0.01137849, + "auxiliary_loss_mlp": 0.01027093, + "balance_loss_clip": 1.04368186, + "balance_loss_mlp": 1.01964831, + "epoch": 0.942463776829195, + "flos": 20080636732800.0, + "grad_norm": 1.970797038713921, + "language_loss": 0.79308337, + "learning_rate": 3.4560842674664856e-08, + "loss": 0.81473279, + "num_input_tokens_seen": 169424155, + "step": 7838, + "time_per_iteration": 2.843878984451294 + }, + { + "auxiliary_loss_clip": 0.01159668, + "auxiliary_loss_mlp": 0.01029797, + "balance_loss_clip": 1.04332852, + "balance_loss_mlp": 1.02222121, + "epoch": 0.9425840197198341, + "flos": 22636255536000.0, + "grad_norm": 1.8096409931027573, + "language_loss": 0.75132287, + "learning_rate": 3.441680364185506e-08, + "loss": 0.77321756, + "num_input_tokens_seen": 169444025, + "step": 7839, + "time_per_iteration": 2.74723744392395 + }, + { + "auxiliary_loss_clip": 0.01150206, + "auxiliary_loss_mlp": 0.01025923, + "balance_loss_clip": 1.04600048, + "balance_loss_mlp": 1.01817989, + "epoch": 0.9427042626104731, + "flos": 19937892084480.0, + "grad_norm": 2.2535890323823105, + "language_loss": 0.74431479, + "learning_rate": 3.427306278645314e-08, + "loss": 0.76607609, + "num_input_tokens_seen": 169462480, + "step": 7840, + "time_per_iteration": 2.7200400829315186 + }, + { + "auxiliary_loss_clip": 0.01129242, + "auxiliary_loss_mlp": 0.01025641, + "balance_loss_clip": 1.04197669, + "balance_loss_mlp": 1.01835096, + "epoch": 0.9428245055011123, + "flos": 22856998567680.0, + "grad_norm": 1.8027940101707234, + "language_loss": 0.72965354, + "learning_rate": 3.4129620130264767e-08, + "loss": 0.7512024, + "num_input_tokens_seen": 169480840, + "step": 7841, + "time_per_iteration": 2.8198657035827637 + }, + { + "auxiliary_loss_clip": 0.01153759, + "auxiliary_loss_mlp": 0.00885536, + "balance_loss_clip": 1.04655337, + "balance_loss_mlp": 1.00064337, + "epoch": 0.9429447483917514, + "flos": 20951757371520.0, + "grad_norm": 2.141592212989152, + "language_loss": 0.77828932, + "learning_rate": 3.398647569505009e-08, + "loss": 0.79868227, + "num_input_tokens_seen": 169498265, + "step": 7842, + "time_per_iteration": 2.8626677989959717 + }, + { + "auxiliary_loss_clip": 0.01142703, + "auxiliary_loss_mlp": 0.01024947, + "balance_loss_clip": 1.04242384, + "balance_loss_mlp": 1.01659608, + "epoch": 0.9430649912823904, + "flos": 18843658116480.0, + "grad_norm": 2.41093481746743, + "language_loss": 0.75188553, + "learning_rate": 3.384362950252373e-08, + "loss": 0.77356195, + "num_input_tokens_seen": 169515235, + "step": 7843, + "time_per_iteration": 2.8502464294433594 + }, + { + "auxiliary_loss_clip": 0.01147386, + "auxiliary_loss_mlp": 0.01030317, + "balance_loss_clip": 1.04274166, + "balance_loss_mlp": 1.02284837, + "epoch": 0.9431852341730296, + "flos": 32556038837760.0, + "grad_norm": 1.784363052237616, + "language_loss": 0.56884146, + "learning_rate": 3.3701081574355473e-08, + "loss": 0.59061849, + "num_input_tokens_seen": 169537195, + "step": 7844, + "time_per_iteration": 2.7918357849121094 + }, + { + "auxiliary_loss_clip": 0.01039263, + "auxiliary_loss_mlp": 0.01000874, + "balance_loss_clip": 1.00681651, + "balance_loss_mlp": 0.99952108, + "epoch": 0.9433054770636686, + "flos": 66904490252160.0, + "grad_norm": 0.6398033811674333, + "language_loss": 0.51660705, + "learning_rate": 3.3558831932169796e-08, + "loss": 0.5370084, + "num_input_tokens_seen": 169605865, + "step": 7845, + "time_per_iteration": 3.330307722091675 + }, + { + "auxiliary_loss_clip": 0.0115568, + "auxiliary_loss_mlp": 0.01022733, + "balance_loss_clip": 1.04399049, + "balance_loss_mlp": 1.01502585, + "epoch": 0.9434257199543077, + "flos": 26140346916480.0, + "grad_norm": 2.2310541701594473, + "language_loss": 0.88643646, + "learning_rate": 3.341688059754588e-08, + "loss": 0.90822053, + "num_input_tokens_seen": 169621520, + "step": 7846, + "time_per_iteration": 2.699158191680908 + }, + { + "auxiliary_loss_clip": 0.01145222, + "auxiliary_loss_mlp": 0.00885769, + "balance_loss_clip": 1.04146075, + "balance_loss_mlp": 1.00054526, + "epoch": 0.9435459628449467, + "flos": 25003486483200.0, + "grad_norm": 2.0736858981121093, + "language_loss": 0.78182375, + "learning_rate": 3.327522759201762e-08, + "loss": 0.80213362, + "num_input_tokens_seen": 169641390, + "step": 7847, + "time_per_iteration": 2.8613474369049072 + }, + { + "auxiliary_loss_clip": 0.01138582, + "auxiliary_loss_mlp": 0.01027006, + "balance_loss_clip": 1.04325914, + "balance_loss_mlp": 1.01951349, + "epoch": 0.9436662057355859, + "flos": 22163240309760.0, + "grad_norm": 2.1838598025331284, + "language_loss": 0.66878223, + "learning_rate": 3.313387293707359e-08, + "loss": 0.69043803, + "num_input_tokens_seen": 169660095, + "step": 7848, + "time_per_iteration": 2.7644500732421875 + }, + { + "auxiliary_loss_clip": 0.01134557, + "auxiliary_loss_mlp": 0.0102877, + "balance_loss_clip": 1.0432471, + "balance_loss_mlp": 1.02108955, + "epoch": 0.943786448626225, + "flos": 20118522602880.0, + "grad_norm": 1.8349508098373317, + "language_loss": 0.6832369, + "learning_rate": 3.29928166541571e-08, + "loss": 0.70487016, + "num_input_tokens_seen": 169679050, + "step": 7849, + "time_per_iteration": 2.8300702571868896 + }, + { + "auxiliary_loss_clip": 0.01143096, + "auxiliary_loss_mlp": 0.01029606, + "balance_loss_clip": 1.04363978, + "balance_loss_mlp": 1.02154183, + "epoch": 0.943906691516864, + "flos": 22090808534400.0, + "grad_norm": 2.109309965626557, + "language_loss": 0.80365711, + "learning_rate": 3.2852058764666346e-08, + "loss": 0.82538414, + "num_input_tokens_seen": 169698150, + "step": 7850, + "time_per_iteration": 2.760979652404785 + }, + { + "auxiliary_loss_clip": 0.01131865, + "auxiliary_loss_mlp": 0.01032479, + "balance_loss_clip": 1.04627299, + "balance_loss_mlp": 1.02445602, + "epoch": 0.9440269344075032, + "flos": 35298501212160.0, + "grad_norm": 1.9415713449174083, + "language_loss": 0.68792272, + "learning_rate": 3.2711599289954264e-08, + "loss": 0.70956618, + "num_input_tokens_seen": 169722185, + "step": 7851, + "time_per_iteration": 2.9236741065979004 + }, + { + "auxiliary_loss_clip": 0.01121735, + "auxiliary_loss_mlp": 0.01030807, + "balance_loss_clip": 1.04171276, + "balance_loss_mlp": 1.02299881, + "epoch": 0.9441471772981422, + "flos": 19238136255360.0, + "grad_norm": 1.8263248970039216, + "language_loss": 0.77678216, + "learning_rate": 3.257143825132847e-08, + "loss": 0.7983076, + "num_input_tokens_seen": 169740355, + "step": 7852, + "time_per_iteration": 2.831282377243042 + }, + { + "auxiliary_loss_clip": 0.01148106, + "auxiliary_loss_mlp": 0.0103167, + "balance_loss_clip": 1.04178977, + "balance_loss_mlp": 1.02455592, + "epoch": 0.9442674201887813, + "flos": 25739799379200.0, + "grad_norm": 1.7874175197562956, + "language_loss": 0.75810647, + "learning_rate": 3.243157567005106e-08, + "loss": 0.77990425, + "num_input_tokens_seen": 169758535, + "step": 7853, + "time_per_iteration": 3.6910200119018555 + }, + { + "auxiliary_loss_clip": 0.0117485, + "auxiliary_loss_mlp": 0.01028026, + "balance_loss_clip": 1.05182743, + "balance_loss_mlp": 1.01957393, + "epoch": 0.9443876630794205, + "flos": 15523321737600.0, + "grad_norm": 3.2031602419060996, + "language_loss": 0.63818061, + "learning_rate": 3.2292011567339296e-08, + "loss": 0.66020936, + "num_input_tokens_seen": 169776340, + "step": 7854, + "time_per_iteration": 3.557457685470581 + }, + { + "auxiliary_loss_clip": 0.01159781, + "auxiliary_loss_mlp": 0.00885339, + "balance_loss_clip": 1.04450333, + "balance_loss_mlp": 1.00056088, + "epoch": 0.9445079059700595, + "flos": 13400821128960.0, + "grad_norm": 2.0782180358478097, + "language_loss": 0.55659282, + "learning_rate": 3.21527459643649e-08, + "loss": 0.57704401, + "num_input_tokens_seen": 169793225, + "step": 7855, + "time_per_iteration": 3.6193082332611084 + }, + { + "auxiliary_loss_clip": 0.01160639, + "auxiliary_loss_mlp": 0.0103077, + "balance_loss_clip": 1.04553473, + "balance_loss_mlp": 1.02248514, + "epoch": 0.9446281488606986, + "flos": 23659242877440.0, + "grad_norm": 2.4574555973632757, + "language_loss": 0.74294257, + "learning_rate": 3.2013778882254536e-08, + "loss": 0.76485664, + "num_input_tokens_seen": 169812020, + "step": 7856, + "time_per_iteration": 2.730152130126953 + }, + { + "auxiliary_loss_clip": 0.01149848, + "auxiliary_loss_mlp": 0.01026646, + "balance_loss_clip": 1.04330039, + "balance_loss_mlp": 1.01920104, + "epoch": 0.9447483917513377, + "flos": 25557337267200.0, + "grad_norm": 1.829184902766577, + "language_loss": 0.75576055, + "learning_rate": 3.1875110342088676e-08, + "loss": 0.77752542, + "num_input_tokens_seen": 169833470, + "step": 7857, + "time_per_iteration": 2.7250783443450928 + }, + { + "auxiliary_loss_clip": 0.01142198, + "auxiliary_loss_mlp": 0.01027778, + "balance_loss_clip": 1.043998, + "balance_loss_mlp": 1.01990438, + "epoch": 0.9448686346419768, + "flos": 24535463247360.0, + "grad_norm": 1.6334117307468625, + "language_loss": 0.65497792, + "learning_rate": 3.1736740364904035e-08, + "loss": 0.67667764, + "num_input_tokens_seen": 169854000, + "step": 7858, + "time_per_iteration": 3.0224573612213135 + }, + { + "auxiliary_loss_clip": 0.01128071, + "auxiliary_loss_mlp": 0.00885754, + "balance_loss_clip": 1.04069281, + "balance_loss_mlp": 1.00048661, + "epoch": 0.9449888775326158, + "flos": 14721256995840.0, + "grad_norm": 2.4323988966178676, + "language_loss": 0.77223516, + "learning_rate": 3.159866897169094e-08, + "loss": 0.79237342, + "num_input_tokens_seen": 169872200, + "step": 7859, + "time_per_iteration": 3.8061749935150146 + }, + { + "auxiliary_loss_clip": 0.01142522, + "auxiliary_loss_mlp": 0.01030192, + "balance_loss_clip": 1.04706097, + "balance_loss_mlp": 1.02193046, + "epoch": 0.945109120423255, + "flos": 15447873219840.0, + "grad_norm": 2.1899582711264776, + "language_loss": 0.75833553, + "learning_rate": 3.146089618339487e-08, + "loss": 0.78006268, + "num_input_tokens_seen": 169889055, + "step": 7860, + "time_per_iteration": 2.846524715423584 + }, + { + "auxiliary_loss_clip": 0.01140898, + "auxiliary_loss_mlp": 0.01029166, + "balance_loss_clip": 1.04325914, + "balance_loss_mlp": 1.0215131, + "epoch": 0.9452293633138941, + "flos": 25448097029760.0, + "grad_norm": 2.155991798747751, + "language_loss": 0.68119478, + "learning_rate": 3.132342202091554e-08, + "loss": 0.70289546, + "num_input_tokens_seen": 169909280, + "step": 7861, + "time_per_iteration": 2.931276798248291 + }, + { + "auxiliary_loss_clip": 0.01167571, + "auxiliary_loss_mlp": 0.01026572, + "balance_loss_clip": 1.04475141, + "balance_loss_mlp": 1.01822698, + "epoch": 0.9453496062045331, + "flos": 21215342350080.0, + "grad_norm": 2.3592586756942024, + "language_loss": 0.68770301, + "learning_rate": 3.1186246505107595e-08, + "loss": 0.70964444, + "num_input_tokens_seen": 169928420, + "step": 7862, + "time_per_iteration": 2.7099504470825195 + }, + { + "auxiliary_loss_clip": 0.01161851, + "auxiliary_loss_mlp": 0.01027045, + "balance_loss_clip": 1.0486809, + "balance_loss_mlp": 1.01927876, + "epoch": 0.9454698490951723, + "flos": 20010898477440.0, + "grad_norm": 7.934224778032011, + "language_loss": 0.83818662, + "learning_rate": 3.104936965678084e-08, + "loss": 0.86007553, + "num_input_tokens_seen": 169946750, + "step": 7863, + "time_per_iteration": 2.789257764816284 + }, + { + "auxiliary_loss_clip": 0.0115845, + "auxiliary_loss_mlp": 0.01030768, + "balance_loss_clip": 1.04404044, + "balance_loss_mlp": 1.02280509, + "epoch": 0.9455900919858113, + "flos": 21069652786560.0, + "grad_norm": 2.3190090442650217, + "language_loss": 0.81749988, + "learning_rate": 3.091279149669956e-08, + "loss": 0.83939207, + "num_input_tokens_seen": 169965540, + "step": 7864, + "time_per_iteration": 2.6829254627227783 + }, + { + "auxiliary_loss_clip": 0.01157555, + "auxiliary_loss_mlp": 0.00885754, + "balance_loss_clip": 1.04446685, + "balance_loss_mlp": 1.00054789, + "epoch": 0.9457103348764504, + "flos": 20740854666240.0, + "grad_norm": 1.8551087964322013, + "language_loss": 0.73552859, + "learning_rate": 3.0776512045581624e-08, + "loss": 0.75596166, + "num_input_tokens_seen": 169984330, + "step": 7865, + "time_per_iteration": 2.7216796875 + }, + { + "auxiliary_loss_clip": 0.01145435, + "auxiliary_loss_mlp": 0.01026468, + "balance_loss_clip": 1.04554641, + "balance_loss_mlp": 1.01805162, + "epoch": 0.9458305777670896, + "flos": 21428363957760.0, + "grad_norm": 1.8731654461323377, + "language_loss": 0.7780543, + "learning_rate": 3.0640531324101384e-08, + "loss": 0.79977334, + "num_input_tokens_seen": 170002095, + "step": 7866, + "time_per_iteration": 2.7250449657440186 + }, + { + "auxiliary_loss_clip": 0.01164831, + "auxiliary_loss_mlp": 0.01030412, + "balance_loss_clip": 1.05172694, + "balance_loss_mlp": 1.0222168, + "epoch": 0.9459508206577286, + "flos": 20011185786240.0, + "grad_norm": 1.695387061943152, + "language_loss": 0.75905359, + "learning_rate": 3.0504849352886554e-08, + "loss": 0.7810061, + "num_input_tokens_seen": 170020240, + "step": 7867, + "time_per_iteration": 2.6690773963928223 + }, + { + "auxiliary_loss_clip": 0.01158242, + "auxiliary_loss_mlp": 0.01032998, + "balance_loss_clip": 1.04787898, + "balance_loss_mlp": 1.02511227, + "epoch": 0.9460710635483677, + "flos": 12166428291840.0, + "grad_norm": 2.6020660814034664, + "language_loss": 0.71721029, + "learning_rate": 3.036946615252023e-08, + "loss": 0.73912269, + "num_input_tokens_seen": 170035770, + "step": 7868, + "time_per_iteration": 2.669644355773926 + }, + { + "auxiliary_loss_clip": 0.01146932, + "auxiliary_loss_mlp": 0.01036358, + "balance_loss_clip": 1.04407454, + "balance_loss_mlp": 1.02828741, + "epoch": 0.9461913064390068, + "flos": 34276196229120.0, + "grad_norm": 2.186105183640427, + "language_loss": 0.66572922, + "learning_rate": 3.0234381743539984e-08, + "loss": 0.68756211, + "num_input_tokens_seen": 170053385, + "step": 7869, + "time_per_iteration": 2.833509683609009 + }, + { + "auxiliary_loss_clip": 0.0115391, + "auxiliary_loss_mlp": 0.01022885, + "balance_loss_clip": 1.04299927, + "balance_loss_mlp": 1.01568484, + "epoch": 0.9463115493296459, + "flos": 19463763536640.0, + "grad_norm": 1.9755487991038143, + "language_loss": 0.80104816, + "learning_rate": 3.0099596146437863e-08, + "loss": 0.82281607, + "num_input_tokens_seen": 170070490, + "step": 7870, + "time_per_iteration": 2.6869242191314697 + }, + { + "auxiliary_loss_clip": 0.01054963, + "auxiliary_loss_mlp": 0.01002102, + "balance_loss_clip": 1.00447965, + "balance_loss_mlp": 1.00075459, + "epoch": 0.946431792220285, + "flos": 70570824387840.0, + "grad_norm": 0.775833010279687, + "language_loss": 0.60061061, + "learning_rate": 2.996510938166086e-08, + "loss": 0.62118125, + "num_input_tokens_seen": 170133465, + "step": 7871, + "time_per_iteration": 3.3135998249053955 + }, + { + "auxiliary_loss_clip": 0.01158576, + "auxiliary_loss_mlp": 0.01033498, + "balance_loss_clip": 1.04698884, + "balance_loss_mlp": 1.02606535, + "epoch": 0.9465520351109241, + "flos": 18947906363520.0, + "grad_norm": 5.101960452164773, + "language_loss": 0.73382092, + "learning_rate": 2.983092146960997e-08, + "loss": 0.75574172, + "num_input_tokens_seen": 170150810, + "step": 7872, + "time_per_iteration": 2.743974447250366 + }, + { + "auxiliary_loss_clip": 0.01149567, + "auxiliary_loss_mlp": 0.01026613, + "balance_loss_clip": 1.04255605, + "balance_loss_mlp": 1.01828599, + "epoch": 0.9466722780015632, + "flos": 19135647774720.0, + "grad_norm": 2.2893732617209697, + "language_loss": 0.8027159, + "learning_rate": 2.9697032430642256e-08, + "loss": 0.82447767, + "num_input_tokens_seen": 170169025, + "step": 7873, + "time_per_iteration": 2.805893898010254 + }, + { + "auxiliary_loss_clip": 0.01162301, + "auxiliary_loss_mlp": 0.01022819, + "balance_loss_clip": 1.04370403, + "balance_loss_mlp": 1.01584792, + "epoch": 0.9467925208922022, + "flos": 17237912520960.0, + "grad_norm": 2.388119338591531, + "language_loss": 0.74312103, + "learning_rate": 2.9563442285067906e-08, + "loss": 0.76497221, + "num_input_tokens_seen": 170186070, + "step": 7874, + "time_per_iteration": 2.6617157459259033 + }, + { + "auxiliary_loss_clip": 0.01161001, + "auxiliary_loss_mlp": 0.01028484, + "balance_loss_clip": 1.04705644, + "balance_loss_mlp": 1.02098024, + "epoch": 0.9469127637828414, + "flos": 29169016859520.0, + "grad_norm": 2.2058874712660193, + "language_loss": 0.7962538, + "learning_rate": 2.943015105315294e-08, + "loss": 0.81814867, + "num_input_tokens_seen": 170206265, + "step": 7875, + "time_per_iteration": 2.8777854442596436 + }, + { + "auxiliary_loss_clip": 0.01133113, + "auxiliary_loss_mlp": 0.0102607, + "balance_loss_clip": 1.0403347, + "balance_loss_mlp": 1.01740348, + "epoch": 0.9470330066734804, + "flos": 26030460234240.0, + "grad_norm": 2.457581383722607, + "language_loss": 0.66659975, + "learning_rate": 2.929715875511718e-08, + "loss": 0.68819153, + "num_input_tokens_seen": 170225300, + "step": 7876, + "time_per_iteration": 2.9135892391204834 + }, + { + "auxiliary_loss_clip": 0.01156801, + "auxiliary_loss_mlp": 0.01028457, + "balance_loss_clip": 1.04140282, + "balance_loss_mlp": 1.02076781, + "epoch": 0.9471532495641195, + "flos": 23440906056960.0, + "grad_norm": 1.8147846867621997, + "language_loss": 0.69862056, + "learning_rate": 2.9164465411135375e-08, + "loss": 0.72047317, + "num_input_tokens_seen": 170245070, + "step": 7877, + "time_per_iteration": 2.7174232006073 + }, + { + "auxiliary_loss_clip": 0.01159196, + "auxiliary_loss_mlp": 0.01028077, + "balance_loss_clip": 1.0478493, + "balance_loss_mlp": 1.02044809, + "epoch": 0.9472734924547586, + "flos": 15815850099840.0, + "grad_norm": 1.8908173080470443, + "language_loss": 0.80852383, + "learning_rate": 2.9032071041337426e-08, + "loss": 0.83039659, + "num_input_tokens_seen": 170263305, + "step": 7878, + "time_per_iteration": 2.6874046325683594 + }, + { + "auxiliary_loss_clip": 0.01142713, + "auxiliary_loss_mlp": 0.01026882, + "balance_loss_clip": 1.04418707, + "balance_loss_mlp": 1.01892138, + "epoch": 0.9473937353453977, + "flos": 11181793697280.0, + "grad_norm": 1.6931147361974097, + "language_loss": 0.7292906, + "learning_rate": 2.889997566580704e-08, + "loss": 0.75098658, + "num_input_tokens_seen": 170281460, + "step": 7879, + "time_per_iteration": 3.681964159011841 + }, + { + "auxiliary_loss_clip": 0.01170792, + "auxiliary_loss_mlp": 0.01027429, + "balance_loss_clip": 1.04730117, + "balance_loss_mlp": 1.01956093, + "epoch": 0.9475139782360368, + "flos": 25775530433280.0, + "grad_norm": 2.0421459658173506, + "language_loss": 0.70617485, + "learning_rate": 2.8768179304583086e-08, + "loss": 0.72815704, + "num_input_tokens_seen": 170303515, + "step": 7880, + "time_per_iteration": 3.570350408554077 + }, + { + "auxiliary_loss_clip": 0.01140826, + "auxiliary_loss_mlp": 0.01022362, + "balance_loss_clip": 1.04606974, + "balance_loss_mlp": 1.01543641, + "epoch": 0.9476342211266758, + "flos": 22820046451200.0, + "grad_norm": 1.5797507112424192, + "language_loss": 0.7341634, + "learning_rate": 2.8636681977659117e-08, + "loss": 0.7557953, + "num_input_tokens_seen": 170323165, + "step": 7881, + "time_per_iteration": 3.708209991455078 + }, + { + "auxiliary_loss_clip": 0.01130333, + "auxiliary_loss_mlp": 0.01025283, + "balance_loss_clip": 1.04437375, + "balance_loss_mlp": 1.01750493, + "epoch": 0.947754464017315, + "flos": 20193611984640.0, + "grad_norm": 2.6889422873091027, + "language_loss": 0.77785313, + "learning_rate": 2.850548370498318e-08, + "loss": 0.79940927, + "num_input_tokens_seen": 170341005, + "step": 7882, + "time_per_iteration": 2.8476943969726562 + }, + { + "auxiliary_loss_clip": 0.01157392, + "auxiliary_loss_mlp": 0.01026194, + "balance_loss_clip": 1.04373753, + "balance_loss_mlp": 1.01858211, + "epoch": 0.9478747069079541, + "flos": 24717925359360.0, + "grad_norm": 1.5593993714798293, + "language_loss": 0.71398681, + "learning_rate": 2.8374584506457798e-08, + "loss": 0.73582268, + "num_input_tokens_seen": 170362280, + "step": 7883, + "time_per_iteration": 2.6937451362609863 + }, + { + "auxiliary_loss_clip": 0.01148324, + "auxiliary_loss_mlp": 0.01026579, + "balance_loss_clip": 1.04636574, + "balance_loss_mlp": 1.01856852, + "epoch": 0.9479949497985931, + "flos": 21361355136000.0, + "grad_norm": 2.6394402489083824, + "language_loss": 0.67319739, + "learning_rate": 2.824398440193998e-08, + "loss": 0.69494647, + "num_input_tokens_seen": 170381080, + "step": 7884, + "time_per_iteration": 2.6964099407196045 + }, + { + "auxiliary_loss_clip": 0.01121166, + "auxiliary_loss_mlp": 0.01026918, + "balance_loss_clip": 1.03892887, + "balance_loss_mlp": 1.01875222, + "epoch": 0.9481151926892323, + "flos": 18148606968960.0, + "grad_norm": 2.896180303947816, + "language_loss": 0.71523678, + "learning_rate": 2.811368341124232e-08, + "loss": 0.73671758, + "num_input_tokens_seen": 170400150, + "step": 7885, + "time_per_iteration": 3.663517475128174 + }, + { + "auxiliary_loss_clip": 0.01149415, + "auxiliary_loss_mlp": 0.01033366, + "balance_loss_clip": 1.04441237, + "balance_loss_mlp": 1.02603459, + "epoch": 0.9482354355798713, + "flos": 22128012046080.0, + "grad_norm": 2.2685391816163696, + "language_loss": 0.6827963, + "learning_rate": 2.7983681554131222e-08, + "loss": 0.70462412, + "num_input_tokens_seen": 170420410, + "step": 7886, + "time_per_iteration": 2.7680556774139404 + }, + { + "auxiliary_loss_clip": 0.01148242, + "auxiliary_loss_mlp": 0.01026929, + "balance_loss_clip": 1.04296923, + "balance_loss_mlp": 1.01926386, + "epoch": 0.9483556784705104, + "flos": 19063072344960.0, + "grad_norm": 2.430974220264433, + "language_loss": 0.70290363, + "learning_rate": 2.7853978850327365e-08, + "loss": 0.72465533, + "num_input_tokens_seen": 170439580, + "step": 7887, + "time_per_iteration": 2.7750532627105713 + }, + { + "auxiliary_loss_clip": 0.01141087, + "auxiliary_loss_mlp": 0.01028385, + "balance_loss_clip": 1.04673171, + "balance_loss_mlp": 1.02079988, + "epoch": 0.9484759213611496, + "flos": 25777110631680.0, + "grad_norm": 1.8789688222160574, + "language_loss": 0.87263, + "learning_rate": 2.7724575319507225e-08, + "loss": 0.89432466, + "num_input_tokens_seen": 170459290, + "step": 7888, + "time_per_iteration": 2.786654233932495 + }, + { + "auxiliary_loss_clip": 0.01155436, + "auxiliary_loss_mlp": 0.01025114, + "balance_loss_clip": 1.04179668, + "balance_loss_mlp": 1.01797044, + "epoch": 0.9485961642517886, + "flos": 20667740532480.0, + "grad_norm": 1.9769278407576982, + "language_loss": 0.77089566, + "learning_rate": 2.759547098130044e-08, + "loss": 0.79270124, + "num_input_tokens_seen": 170478020, + "step": 7889, + "time_per_iteration": 2.714797019958496 + }, + { + "auxiliary_loss_clip": 0.01164833, + "auxiliary_loss_mlp": 0.01028019, + "balance_loss_clip": 1.04585683, + "balance_loss_mlp": 1.0204761, + "epoch": 0.9487164071424277, + "flos": 22674069578880.0, + "grad_norm": 2.0483339346831393, + "language_loss": 0.76912969, + "learning_rate": 2.746666585529267e-08, + "loss": 0.79105818, + "num_input_tokens_seen": 170498295, + "step": 7890, + "time_per_iteration": 2.626673698425293 + }, + { + "auxiliary_loss_clip": 0.01152843, + "auxiliary_loss_mlp": 0.0102787, + "balance_loss_clip": 1.04466236, + "balance_loss_mlp": 1.0202111, + "epoch": 0.9488366500330668, + "flos": 38726461716480.0, + "grad_norm": 2.1599147502402216, + "language_loss": 0.74074841, + "learning_rate": 2.73381599610234e-08, + "loss": 0.7625556, + "num_input_tokens_seen": 170518695, + "step": 7891, + "time_per_iteration": 2.830014944076538 + }, + { + "auxiliary_loss_clip": 0.01153922, + "auxiliary_loss_mlp": 0.01025541, + "balance_loss_clip": 1.04298568, + "balance_loss_mlp": 1.01775694, + "epoch": 0.9489568929237059, + "flos": 27890920149120.0, + "grad_norm": 1.8887911916179116, + "language_loss": 0.71643728, + "learning_rate": 2.7209953317987033e-08, + "loss": 0.73823196, + "num_input_tokens_seen": 170539735, + "step": 7892, + "time_per_iteration": 2.672273874282837 + }, + { + "auxiliary_loss_clip": 0.01160949, + "auxiliary_loss_mlp": 0.01025009, + "balance_loss_clip": 1.0462451, + "balance_loss_mlp": 1.01759434, + "epoch": 0.9490771358143449, + "flos": 33580642291200.0, + "grad_norm": 2.1583181580098723, + "language_loss": 0.77890396, + "learning_rate": 2.7082045945631793e-08, + "loss": 0.80076355, + "num_input_tokens_seen": 170561950, + "step": 7893, + "time_per_iteration": 2.8261003494262695 + }, + { + "auxiliary_loss_clip": 0.01132252, + "auxiliary_loss_mlp": 0.01027326, + "balance_loss_clip": 1.04173565, + "balance_loss_mlp": 1.01982832, + "epoch": 0.9491973787049841, + "flos": 14793796512000.0, + "grad_norm": 1.960664873987772, + "language_loss": 0.69385159, + "learning_rate": 2.6954437863361712e-08, + "loss": 0.71544737, + "num_input_tokens_seen": 170579865, + "step": 7894, + "time_per_iteration": 2.745450258255005 + }, + { + "auxiliary_loss_clip": 0.01124944, + "auxiliary_loss_mlp": 0.01022448, + "balance_loss_clip": 1.04125643, + "balance_loss_mlp": 1.01543272, + "epoch": 0.9493176215956232, + "flos": 25332535998720.0, + "grad_norm": 2.1253844196972844, + "language_loss": 0.70777184, + "learning_rate": 2.6827129090534862e-08, + "loss": 0.72924578, + "num_input_tokens_seen": 170600165, + "step": 7895, + "time_per_iteration": 2.9446656703948975 + }, + { + "auxiliary_loss_clip": 0.01145562, + "auxiliary_loss_mlp": 0.01028399, + "balance_loss_clip": 1.04372275, + "balance_loss_mlp": 1.02045333, + "epoch": 0.9494378644862622, + "flos": 21029971236480.0, + "grad_norm": 1.8557028052371707, + "language_loss": 0.77966791, + "learning_rate": 2.670011964646335e-08, + "loss": 0.80140758, + "num_input_tokens_seen": 170618845, + "step": 7896, + "time_per_iteration": 2.7423884868621826 + }, + { + "auxiliary_loss_clip": 0.01114092, + "auxiliary_loss_mlp": 0.01029087, + "balance_loss_clip": 1.03340936, + "balance_loss_mlp": 1.02095127, + "epoch": 0.9495581073769014, + "flos": 15195134148480.0, + "grad_norm": 3.2786457402660245, + "language_loss": 0.68154317, + "learning_rate": 2.657340955041487e-08, + "loss": 0.70297498, + "num_input_tokens_seen": 170637620, + "step": 7897, + "time_per_iteration": 2.9453279972076416 + }, + { + "auxiliary_loss_clip": 0.01147937, + "auxiliary_loss_mlp": 0.01032857, + "balance_loss_clip": 1.04685712, + "balance_loss_mlp": 1.02468514, + "epoch": 0.9496783502675404, + "flos": 28616566705920.0, + "grad_norm": 2.8093194976032905, + "language_loss": 0.71522474, + "learning_rate": 2.6446998821611167e-08, + "loss": 0.73703271, + "num_input_tokens_seen": 170657815, + "step": 7898, + "time_per_iteration": 2.7952802181243896 + }, + { + "auxiliary_loss_clip": 0.01134382, + "auxiliary_loss_mlp": 0.01027721, + "balance_loss_clip": 1.04290724, + "balance_loss_mlp": 1.01988649, + "epoch": 0.9497985931581795, + "flos": 14866874732160.0, + "grad_norm": 2.352133959891056, + "language_loss": 0.71915317, + "learning_rate": 2.6320887479228228e-08, + "loss": 0.74077415, + "num_input_tokens_seen": 170674415, + "step": 7899, + "time_per_iteration": 2.8541433811187744 + }, + { + "auxiliary_loss_clip": 0.01151164, + "auxiliary_loss_mlp": 0.01027691, + "balance_loss_clip": 1.04346204, + "balance_loss_mlp": 1.01988852, + "epoch": 0.9499188360488187, + "flos": 27193319136000.0, + "grad_norm": 2.3435253751446288, + "language_loss": 0.72658896, + "learning_rate": 2.619507554239786e-08, + "loss": 0.7483775, + "num_input_tokens_seen": 170692975, + "step": 7900, + "time_per_iteration": 2.8625011444091797 + }, + { + "auxiliary_loss_clip": 0.01148004, + "auxiliary_loss_mlp": 0.01028252, + "balance_loss_clip": 1.04449129, + "balance_loss_mlp": 1.02074838, + "epoch": 0.9500390789394577, + "flos": 24316479982080.0, + "grad_norm": 1.6661262874240024, + "language_loss": 0.69751966, + "learning_rate": 2.606956303020502e-08, + "loss": 0.71928227, + "num_input_tokens_seen": 170713780, + "step": 7901, + "time_per_iteration": 2.7862300872802734 + }, + { + "auxiliary_loss_clip": 0.01158198, + "auxiliary_loss_mlp": 0.01028374, + "balance_loss_clip": 1.04630625, + "balance_loss_mlp": 1.02051258, + "epoch": 0.9501593218300968, + "flos": 14354752573440.0, + "grad_norm": 1.9199902133009537, + "language_loss": 0.83798099, + "learning_rate": 2.5944349961690036e-08, + "loss": 0.85984677, + "num_input_tokens_seen": 170730800, + "step": 7902, + "time_per_iteration": 2.6716623306274414 + }, + { + "auxiliary_loss_clip": 0.01140854, + "auxiliary_loss_mlp": 0.01027335, + "balance_loss_clip": 1.04362285, + "balance_loss_mlp": 1.01963425, + "epoch": 0.9502795647207359, + "flos": 38728113742080.0, + "grad_norm": 1.9174153936881564, + "language_loss": 0.73160946, + "learning_rate": 2.581943635584749e-08, + "loss": 0.75329137, + "num_input_tokens_seen": 170753630, + "step": 7903, + "time_per_iteration": 2.910691738128662 + }, + { + "auxiliary_loss_clip": 0.01140567, + "auxiliary_loss_mlp": 0.01026956, + "balance_loss_clip": 1.04184544, + "balance_loss_mlp": 1.01952672, + "epoch": 0.950399807611375, + "flos": 40808023799040.0, + "grad_norm": 1.6000741193071149, + "language_loss": 0.6557771, + "learning_rate": 2.569482223162689e-08, + "loss": 0.67745233, + "num_input_tokens_seen": 170777605, + "step": 7904, + "time_per_iteration": 2.9578325748443604 + }, + { + "auxiliary_loss_clip": 0.01160701, + "auxiliary_loss_mlp": 0.01030376, + "balance_loss_clip": 1.04494619, + "balance_loss_mlp": 1.02218628, + "epoch": 0.950520050502014, + "flos": 23440403266560.0, + "grad_norm": 3.0207549409690104, + "language_loss": 0.72316384, + "learning_rate": 2.5570507607932e-08, + "loss": 0.74507463, + "num_input_tokens_seen": 170797520, + "step": 7905, + "time_per_iteration": 3.5936715602874756 + }, + { + "auxiliary_loss_clip": 0.01164882, + "auxiliary_loss_mlp": 0.01027936, + "balance_loss_clip": 1.04698348, + "balance_loss_mlp": 1.02005053, + "epoch": 0.9506402933926532, + "flos": 17783718658560.0, + "grad_norm": 8.515354103098367, + "language_loss": 0.63856268, + "learning_rate": 2.54464925036213e-08, + "loss": 0.66049087, + "num_input_tokens_seen": 170814810, + "step": 7906, + "time_per_iteration": 2.7826128005981445 + }, + { + "auxiliary_loss_clip": 0.01156368, + "auxiliary_loss_mlp": 0.01027899, + "balance_loss_clip": 1.04515433, + "balance_loss_mlp": 1.01996541, + "epoch": 0.9507605362832923, + "flos": 32561928668160.0, + "grad_norm": 1.9256723068111297, + "language_loss": 0.60967928, + "learning_rate": 2.532277693750773e-08, + "loss": 0.63152194, + "num_input_tokens_seen": 170835735, + "step": 7907, + "time_per_iteration": 4.64526104927063 + }, + { + "auxiliary_loss_clip": 0.01127018, + "auxiliary_loss_mlp": 0.01027842, + "balance_loss_clip": 1.0446806, + "balance_loss_mlp": 1.01998019, + "epoch": 0.9508807791739313, + "flos": 19602054898560.0, + "grad_norm": 18.904553494309177, + "language_loss": 0.75804406, + "learning_rate": 2.5199360928358948e-08, + "loss": 0.77959269, + "num_input_tokens_seen": 170852970, + "step": 7908, + "time_per_iteration": 2.842751979827881 + }, + { + "auxiliary_loss_clip": 0.011478, + "auxiliary_loss_mlp": 0.00884797, + "balance_loss_clip": 1.04203987, + "balance_loss_mlp": 1.00059724, + "epoch": 0.9510010220645704, + "flos": 21471852349440.0, + "grad_norm": 1.6131466614336796, + "language_loss": 0.86928934, + "learning_rate": 2.507624449489665e-08, + "loss": 0.8896153, + "num_input_tokens_seen": 170871600, + "step": 7909, + "time_per_iteration": 2.73111629486084 + }, + { + "auxiliary_loss_clip": 0.01152322, + "auxiliary_loss_mlp": 0.01030601, + "balance_loss_clip": 1.0477984, + "balance_loss_mlp": 1.02351952, + "epoch": 0.9511212649552095, + "flos": 18879999701760.0, + "grad_norm": 1.8176592929759425, + "language_loss": 0.65159845, + "learning_rate": 2.495342765579811e-08, + "loss": 0.6734277, + "num_input_tokens_seen": 170890260, + "step": 7910, + "time_per_iteration": 2.8291592597961426 + }, + { + "auxiliary_loss_clip": 0.01127878, + "auxiliary_loss_mlp": 0.01025647, + "balance_loss_clip": 1.04335296, + "balance_loss_mlp": 1.0181849, + "epoch": 0.9512415078458486, + "flos": 20810521094400.0, + "grad_norm": 1.936480951792102, + "language_loss": 0.71155977, + "learning_rate": 2.4830910429693984e-08, + "loss": 0.73309505, + "num_input_tokens_seen": 170910220, + "step": 7911, + "time_per_iteration": 3.7638659477233887 + }, + { + "auxiliary_loss_clip": 0.01167686, + "auxiliary_loss_mlp": 0.01026259, + "balance_loss_clip": 1.04617977, + "balance_loss_mlp": 1.01878464, + "epoch": 0.9513617507364877, + "flos": 18369565482240.0, + "grad_norm": 2.7814534288658086, + "language_loss": 0.79876232, + "learning_rate": 2.470869283517052e-08, + "loss": 0.82070178, + "num_input_tokens_seen": 170928255, + "step": 7912, + "time_per_iteration": 2.599518299102783 + }, + { + "auxiliary_loss_clip": 0.01152685, + "auxiliary_loss_mlp": 0.01028051, + "balance_loss_clip": 1.04388118, + "balance_loss_mlp": 1.020401, + "epoch": 0.9514819936271268, + "flos": 25010166412800.0, + "grad_norm": 1.621776633035954, + "language_loss": 0.77128923, + "learning_rate": 2.458677489076777e-08, + "loss": 0.7930966, + "num_input_tokens_seen": 170949265, + "step": 7913, + "time_per_iteration": 2.773514986038208 + }, + { + "auxiliary_loss_clip": 0.01148431, + "auxiliary_loss_mlp": 0.0102491, + "balance_loss_clip": 1.04236603, + "balance_loss_mlp": 1.01729274, + "epoch": 0.9516022365177659, + "flos": 18662129758080.0, + "grad_norm": 1.8016381059997295, + "language_loss": 0.82832301, + "learning_rate": 2.446515661498072e-08, + "loss": 0.85005641, + "num_input_tokens_seen": 170968595, + "step": 7914, + "time_per_iteration": 2.679718017578125 + }, + { + "auxiliary_loss_clip": 0.01115611, + "auxiliary_loss_mlp": 0.01030672, + "balance_loss_clip": 1.03791404, + "balance_loss_mlp": 1.02369833, + "epoch": 0.9517224794084049, + "flos": 25372109808000.0, + "grad_norm": 2.442259747588414, + "language_loss": 0.74264002, + "learning_rate": 2.434383802625861e-08, + "loss": 0.76410288, + "num_input_tokens_seen": 170987550, + "step": 7915, + "time_per_iteration": 2.8999581336975098 + }, + { + "auxiliary_loss_clip": 0.01139079, + "auxiliary_loss_mlp": 0.01024693, + "balance_loss_clip": 1.04144025, + "balance_loss_mlp": 1.0172658, + "epoch": 0.9518427222990441, + "flos": 21470918595840.0, + "grad_norm": 1.9605095991184749, + "language_loss": 0.7387141, + "learning_rate": 2.4222819143005168e-08, + "loss": 0.76035184, + "num_input_tokens_seen": 171007145, + "step": 7916, + "time_per_iteration": 2.7729995250701904 + }, + { + "auxiliary_loss_clip": 0.01166175, + "auxiliary_loss_mlp": 0.01024849, + "balance_loss_clip": 1.04746199, + "balance_loss_mlp": 1.01735044, + "epoch": 0.9519629651896832, + "flos": 21033634423680.0, + "grad_norm": 1.9566303735626347, + "language_loss": 0.80683899, + "learning_rate": 2.4102099983579706e-08, + "loss": 0.8287493, + "num_input_tokens_seen": 171026295, + "step": 7917, + "time_per_iteration": 2.731086015701294 + }, + { + "auxiliary_loss_clip": 0.0116044, + "auxiliary_loss_mlp": 0.01030357, + "balance_loss_clip": 1.0461148, + "balance_loss_mlp": 1.02132654, + "epoch": 0.9520832080803222, + "flos": 21689219502720.0, + "grad_norm": 1.6664202149617438, + "language_loss": 0.7727586, + "learning_rate": 2.3981680566294236e-08, + "loss": 0.79466659, + "num_input_tokens_seen": 171045895, + "step": 7918, + "time_per_iteration": 2.696143388748169 + }, + { + "auxiliary_loss_clip": 0.01166874, + "auxiliary_loss_mlp": 0.0102852, + "balance_loss_clip": 1.04829955, + "balance_loss_mlp": 1.0212481, + "epoch": 0.9522034509709614, + "flos": 23145289125120.0, + "grad_norm": 2.2750160696173367, + "language_loss": 0.73390961, + "learning_rate": 2.3861560909416822e-08, + "loss": 0.75586355, + "num_input_tokens_seen": 171065445, + "step": 7919, + "time_per_iteration": 2.692227840423584 + }, + { + "auxiliary_loss_clip": 0.0112215, + "auxiliary_loss_mlp": 0.01025853, + "balance_loss_clip": 1.04346645, + "balance_loss_mlp": 1.01800883, + "epoch": 0.9523236938616004, + "flos": 24679428958080.0, + "grad_norm": 1.7391049893982742, + "language_loss": 0.82344013, + "learning_rate": 2.3741741031169325e-08, + "loss": 0.84492016, + "num_input_tokens_seen": 171085015, + "step": 7920, + "time_per_iteration": 2.9755945205688477 + }, + { + "auxiliary_loss_clip": 0.01123387, + "auxiliary_loss_mlp": 0.01026522, + "balance_loss_clip": 1.03916502, + "balance_loss_mlp": 1.01901746, + "epoch": 0.9524439367522395, + "flos": 22672309812480.0, + "grad_norm": 1.681970295222739, + "language_loss": 0.71577716, + "learning_rate": 2.3622220949728544e-08, + "loss": 0.7372762, + "num_input_tokens_seen": 171103900, + "step": 7921, + "time_per_iteration": 2.8324179649353027 + }, + { + "auxiliary_loss_clip": 0.01149509, + "auxiliary_loss_mlp": 0.01029548, + "balance_loss_clip": 1.04166031, + "balance_loss_mlp": 1.0209235, + "epoch": 0.9525641796428787, + "flos": 34055525024640.0, + "grad_norm": 2.695084296209094, + "language_loss": 0.61759055, + "learning_rate": 2.3503000683225526e-08, + "loss": 0.63938111, + "num_input_tokens_seen": 171121615, + "step": 7922, + "time_per_iteration": 2.8512558937072754 + }, + { + "auxiliary_loss_clip": 0.01169038, + "auxiliary_loss_mlp": 0.0102578, + "balance_loss_clip": 1.04703474, + "balance_loss_mlp": 1.0177213, + "epoch": 0.9526844225335177, + "flos": 16727083251840.0, + "grad_norm": 4.738789659540824, + "language_loss": 0.84125412, + "learning_rate": 2.3384080249745585e-08, + "loss": 0.86320221, + "num_input_tokens_seen": 171139505, + "step": 7923, + "time_per_iteration": 2.5798254013061523 + }, + { + "auxiliary_loss_clip": 0.01132525, + "auxiliary_loss_mlp": 0.01026978, + "balance_loss_clip": 1.04060388, + "balance_loss_mlp": 1.01990271, + "epoch": 0.9528046654241568, + "flos": 36939367330560.0, + "grad_norm": 2.229758694347636, + "language_loss": 0.82892966, + "learning_rate": 2.3265459667329178e-08, + "loss": 0.85052466, + "num_input_tokens_seen": 171158995, + "step": 7924, + "time_per_iteration": 2.919663429260254 + }, + { + "auxiliary_loss_clip": 0.01151424, + "auxiliary_loss_mlp": 0.01028933, + "balance_loss_clip": 1.04498601, + "balance_loss_mlp": 1.02132797, + "epoch": 0.9529249083147959, + "flos": 18255010032000.0, + "grad_norm": 2.4471957850300208, + "language_loss": 0.86395121, + "learning_rate": 2.31471389539708e-08, + "loss": 0.88575482, + "num_input_tokens_seen": 171176120, + "step": 7925, + "time_per_iteration": 2.6854135990142822 + }, + { + "auxiliary_loss_clip": 0.011611, + "auxiliary_loss_mlp": 0.00884984, + "balance_loss_clip": 1.04726315, + "balance_loss_mlp": 1.00055122, + "epoch": 0.953045151205435, + "flos": 28658438985600.0, + "grad_norm": 2.8323783786200685, + "language_loss": 0.72750473, + "learning_rate": 2.3029118127619872e-08, + "loss": 0.74796557, + "num_input_tokens_seen": 171195835, + "step": 7926, + "time_per_iteration": 2.7749881744384766 + }, + { + "auxiliary_loss_clip": 0.01142991, + "auxiliary_loss_mlp": 0.01028088, + "balance_loss_clip": 1.0438441, + "balance_loss_mlp": 1.02036309, + "epoch": 0.953165394096074, + "flos": 21835232288640.0, + "grad_norm": 2.0554453687717626, + "language_loss": 0.86666298, + "learning_rate": 2.2911397206179628e-08, + "loss": 0.88837373, + "num_input_tokens_seen": 171212585, + "step": 7927, + "time_per_iteration": 2.6953072547912598 + }, + { + "auxiliary_loss_clip": 0.01167399, + "auxiliary_loss_mlp": 0.01029028, + "balance_loss_clip": 1.04817176, + "balance_loss_mlp": 1.02138352, + "epoch": 0.9532856369867132, + "flos": 19975059682560.0, + "grad_norm": 2.6662137796093974, + "language_loss": 0.6270892, + "learning_rate": 2.279397620750845e-08, + "loss": 0.64905345, + "num_input_tokens_seen": 171231630, + "step": 7928, + "time_per_iteration": 2.6400837898254395 + }, + { + "auxiliary_loss_clip": 0.01145754, + "auxiliary_loss_mlp": 0.01024077, + "balance_loss_clip": 1.04241776, + "balance_loss_mlp": 1.01718712, + "epoch": 0.9534058798773523, + "flos": 15049588239360.0, + "grad_norm": 2.312766658861932, + "language_loss": 0.78793502, + "learning_rate": 2.2676855149419195e-08, + "loss": 0.80963331, + "num_input_tokens_seen": 171248800, + "step": 7929, + "time_per_iteration": 2.664207696914673 + }, + { + "auxiliary_loss_clip": 0.01142292, + "auxiliary_loss_mlp": 0.01023482, + "balance_loss_clip": 1.04492044, + "balance_loss_mlp": 1.01607323, + "epoch": 0.9535261227679913, + "flos": 17602800831360.0, + "grad_norm": 2.4616684081493516, + "language_loss": 0.75279891, + "learning_rate": 2.2560034049678988e-08, + "loss": 0.77445662, + "num_input_tokens_seen": 171263150, + "step": 7930, + "time_per_iteration": 2.693126678466797 + }, + { + "auxiliary_loss_clip": 0.01171459, + "auxiliary_loss_mlp": 0.01030921, + "balance_loss_clip": 1.04779816, + "balance_loss_mlp": 1.02316666, + "epoch": 0.9536463656586305, + "flos": 23142954741120.0, + "grad_norm": 3.23100657202206, + "language_loss": 0.75548065, + "learning_rate": 2.2443512926008988e-08, + "loss": 0.77750444, + "num_input_tokens_seen": 171282480, + "step": 7931, + "time_per_iteration": 3.5047168731689453 + }, + { + "auxiliary_loss_clip": 0.01143309, + "auxiliary_loss_mlp": 0.01025851, + "balance_loss_clip": 1.04118311, + "balance_loss_mlp": 1.01848078, + "epoch": 0.9537666085492695, + "flos": 18625033987200.0, + "grad_norm": 2.356195838105719, + "language_loss": 0.69594121, + "learning_rate": 2.2327291796085946e-08, + "loss": 0.71763277, + "num_input_tokens_seen": 171300840, + "step": 7932, + "time_per_iteration": 2.7283682823181152 + }, + { + "auxiliary_loss_clip": 0.011693, + "auxiliary_loss_mlp": 0.0102815, + "balance_loss_clip": 1.04742682, + "balance_loss_mlp": 1.01994216, + "epoch": 0.9538868514399086, + "flos": 18989347680000.0, + "grad_norm": 2.6252468449245563, + "language_loss": 0.77488494, + "learning_rate": 2.2211370677540197e-08, + "loss": 0.79685944, + "num_input_tokens_seen": 171317365, + "step": 7933, + "time_per_iteration": 4.565019845962524 + }, + { + "auxiliary_loss_clip": 0.01169895, + "auxiliary_loss_mlp": 0.01027971, + "balance_loss_clip": 1.04641271, + "balance_loss_mlp": 1.02013934, + "epoch": 0.9540070943305478, + "flos": 16800556521600.0, + "grad_norm": 2.5083268066169198, + "language_loss": 0.78170669, + "learning_rate": 2.2095749587957012e-08, + "loss": 0.80368531, + "num_input_tokens_seen": 171335270, + "step": 7934, + "time_per_iteration": 2.6346633434295654 + }, + { + "auxiliary_loss_clip": 0.01144044, + "auxiliary_loss_mlp": 0.01025942, + "balance_loss_clip": 1.03943157, + "balance_loss_mlp": 1.01815736, + "epoch": 0.9541273372211868, + "flos": 20156911263360.0, + "grad_norm": 1.7614299375182103, + "language_loss": 0.69445086, + "learning_rate": 2.1980428544876138e-08, + "loss": 0.71615064, + "num_input_tokens_seen": 171353910, + "step": 7935, + "time_per_iteration": 2.6966004371643066 + }, + { + "auxiliary_loss_clip": 0.011229, + "auxiliary_loss_mlp": 0.01031139, + "balance_loss_clip": 1.03699994, + "balance_loss_mlp": 1.02313364, + "epoch": 0.9542475801118259, + "flos": 26725511381760.0, + "grad_norm": 1.6863143734112505, + "language_loss": 0.74284625, + "learning_rate": 2.1865407565791584e-08, + "loss": 0.76438665, + "num_input_tokens_seen": 171375480, + "step": 7936, + "time_per_iteration": 2.8841562271118164 + }, + { + "auxiliary_loss_clip": 0.01145802, + "auxiliary_loss_mlp": 0.01026906, + "balance_loss_clip": 1.04030573, + "balance_loss_mlp": 1.01902604, + "epoch": 0.954367823002465, + "flos": 23330911633920.0, + "grad_norm": 2.169916460880819, + "language_loss": 0.77425987, + "learning_rate": 2.175068666815183e-08, + "loss": 0.79598701, + "num_input_tokens_seen": 171396320, + "step": 7937, + "time_per_iteration": 3.742231607437134 + }, + { + "auxiliary_loss_clip": 0.01139319, + "auxiliary_loss_mlp": 0.01028064, + "balance_loss_clip": 1.04275179, + "balance_loss_mlp": 1.02054465, + "epoch": 0.9544880658931041, + "flos": 14902713527040.0, + "grad_norm": 2.18284310138628, + "language_loss": 0.78733778, + "learning_rate": 2.163626586935985e-08, + "loss": 0.80901158, + "num_input_tokens_seen": 171412860, + "step": 7938, + "time_per_iteration": 2.7605700492858887 + }, + { + "auxiliary_loss_clip": 0.01153624, + "auxiliary_loss_mlp": 0.01025487, + "balance_loss_clip": 1.04310536, + "balance_loss_mlp": 1.01764929, + "epoch": 0.9546083087837431, + "flos": 29095902725760.0, + "grad_norm": 1.9963767764935905, + "language_loss": 0.63228464, + "learning_rate": 2.1522145186773755e-08, + "loss": 0.6540758, + "num_input_tokens_seen": 171431780, + "step": 7939, + "time_per_iteration": 2.7503318786621094 + }, + { + "auxiliary_loss_clip": 0.01149115, + "auxiliary_loss_mlp": 0.01029119, + "balance_loss_clip": 1.04450655, + "balance_loss_mlp": 1.0217638, + "epoch": 0.9547285516743822, + "flos": 21142335957120.0, + "grad_norm": 1.6979385236241793, + "language_loss": 0.85554504, + "learning_rate": 2.140832463770481e-08, + "loss": 0.87732732, + "num_input_tokens_seen": 171450975, + "step": 7940, + "time_per_iteration": 2.7115097045898438 + }, + { + "auxiliary_loss_clip": 0.01151239, + "auxiliary_loss_mlp": 0.01027176, + "balance_loss_clip": 1.04218245, + "balance_loss_mlp": 1.01942754, + "epoch": 0.9548487945650214, + "flos": 27490157130240.0, + "grad_norm": 2.1562950585191687, + "language_loss": 0.7598784, + "learning_rate": 2.129480423941987e-08, + "loss": 0.78166258, + "num_input_tokens_seen": 171467645, + "step": 7941, + "time_per_iteration": 2.856386184692383 + }, + { + "auxiliary_loss_clip": 0.01141977, + "auxiliary_loss_mlp": 0.01025931, + "balance_loss_clip": 1.04283857, + "balance_loss_mlp": 1.01893902, + "epoch": 0.9549690374556604, + "flos": 22273198819200.0, + "grad_norm": 1.644727408041112, + "language_loss": 0.80234754, + "learning_rate": 2.1181584009140052e-08, + "loss": 0.82402658, + "num_input_tokens_seen": 171487185, + "step": 7942, + "time_per_iteration": 2.7382116317749023 + }, + { + "auxiliary_loss_clip": 0.01145418, + "auxiliary_loss_mlp": 0.01033482, + "balance_loss_clip": 1.04204464, + "balance_loss_mlp": 1.02609682, + "epoch": 0.9550892803462995, + "flos": 17595294888960.0, + "grad_norm": 1.9324969424504608, + "language_loss": 0.83537358, + "learning_rate": 2.10686639640405e-08, + "loss": 0.85716259, + "num_input_tokens_seen": 171501275, + "step": 7943, + "time_per_iteration": 2.7910590171813965 + }, + { + "auxiliary_loss_clip": 0.01153376, + "auxiliary_loss_mlp": 0.01022897, + "balance_loss_clip": 1.04553699, + "balance_loss_mlp": 1.0156852, + "epoch": 0.9552095232369386, + "flos": 24353144789760.0, + "grad_norm": 1.5994591703180827, + "language_loss": 0.81223255, + "learning_rate": 2.0956044121251294e-08, + "loss": 0.83399522, + "num_input_tokens_seen": 171520060, + "step": 7944, + "time_per_iteration": 2.7767767906188965 + }, + { + "auxiliary_loss_clip": 0.01141065, + "auxiliary_loss_mlp": 0.01028233, + "balance_loss_clip": 1.04638696, + "balance_loss_mlp": 1.02017474, + "epoch": 0.9553297661275777, + "flos": 22746860490240.0, + "grad_norm": 1.684880466592922, + "language_loss": 0.80997241, + "learning_rate": 2.084372449785654e-08, + "loss": 0.8316654, + "num_input_tokens_seen": 171539895, + "step": 7945, + "time_per_iteration": 2.7794952392578125 + }, + { + "auxiliary_loss_clip": 0.01142961, + "auxiliary_loss_mlp": 0.01033268, + "balance_loss_clip": 1.04119182, + "balance_loss_mlp": 1.02556658, + "epoch": 0.9554500090182168, + "flos": 15413866018560.0, + "grad_norm": 1.6906830046453272, + "language_loss": 0.68986797, + "learning_rate": 2.0731705110895282e-08, + "loss": 0.71163023, + "num_input_tokens_seen": 171557385, + "step": 7946, + "time_per_iteration": 2.8501272201538086 + }, + { + "auxiliary_loss_clip": 0.01161814, + "auxiliary_loss_mlp": 0.01030101, + "balance_loss_clip": 1.04819083, + "balance_loss_mlp": 1.02175605, + "epoch": 0.9555702519088559, + "flos": 23513517400320.0, + "grad_norm": 1.8645614634695409, + "language_loss": 0.86594939, + "learning_rate": 2.0619985977360587e-08, + "loss": 0.88786852, + "num_input_tokens_seen": 171575705, + "step": 7947, + "time_per_iteration": 2.8823769092559814 + }, + { + "auxiliary_loss_clip": 0.01133173, + "auxiliary_loss_mlp": 0.01029325, + "balance_loss_clip": 1.03944254, + "balance_loss_mlp": 1.02205968, + "epoch": 0.955690494799495, + "flos": 22962072827520.0, + "grad_norm": 1.664867081225043, + "language_loss": 0.76667303, + "learning_rate": 2.0508567114200237e-08, + "loss": 0.78829795, + "num_input_tokens_seen": 171595620, + "step": 7948, + "time_per_iteration": 2.8477425575256348 + }, + { + "auxiliary_loss_clip": 0.01151249, + "auxiliary_loss_mlp": 0.01025669, + "balance_loss_clip": 1.0431267, + "balance_loss_mlp": 1.01792943, + "epoch": 0.955810737690134, + "flos": 26031250333440.0, + "grad_norm": 1.9495718855627695, + "language_loss": 0.78480136, + "learning_rate": 2.0397448538316485e-08, + "loss": 0.80657059, + "num_input_tokens_seen": 171616660, + "step": 7949, + "time_per_iteration": 2.753476142883301 + }, + { + "auxiliary_loss_clip": 0.01133557, + "auxiliary_loss_mlp": 0.01023463, + "balance_loss_clip": 1.04129517, + "balance_loss_mlp": 1.01648068, + "epoch": 0.9559309805807732, + "flos": 20849951249280.0, + "grad_norm": 2.268661442312164, + "language_loss": 0.66940606, + "learning_rate": 2.028663026656563e-08, + "loss": 0.69097614, + "num_input_tokens_seen": 171635515, + "step": 7950, + "time_per_iteration": 2.753089189529419 + }, + { + "auxiliary_loss_clip": 0.01166031, + "auxiliary_loss_mlp": 0.00885822, + "balance_loss_clip": 1.04603004, + "balance_loss_mlp": 1.00059247, + "epoch": 0.9560512234714122, + "flos": 21578219498880.0, + "grad_norm": 2.092124243084657, + "language_loss": 0.71697003, + "learning_rate": 2.0176112315758885e-08, + "loss": 0.73748857, + "num_input_tokens_seen": 171653305, + "step": 7951, + "time_per_iteration": 2.6489617824554443 + }, + { + "auxiliary_loss_clip": 0.01136402, + "auxiliary_loss_mlp": 0.0102391, + "balance_loss_clip": 1.04185486, + "balance_loss_mlp": 1.01575065, + "epoch": 0.9561714663620513, + "flos": 17450144029440.0, + "grad_norm": 2.9277406572174502, + "language_loss": 0.69301486, + "learning_rate": 2.0065894702661957e-08, + "loss": 0.71461797, + "num_input_tokens_seen": 171669980, + "step": 7952, + "time_per_iteration": 2.7793641090393066 + }, + { + "auxiliary_loss_clip": 0.01130656, + "auxiliary_loss_mlp": 0.00885578, + "balance_loss_clip": 1.03931785, + "balance_loss_mlp": 1.00053549, + "epoch": 0.9562917092526905, + "flos": 26098510550400.0, + "grad_norm": 1.7355065095929967, + "language_loss": 0.77838856, + "learning_rate": 1.9955977443994577e-08, + "loss": 0.7985509, + "num_input_tokens_seen": 171689970, + "step": 7953, + "time_per_iteration": 2.8948676586151123 + }, + { + "auxiliary_loss_clip": 0.01147242, + "auxiliary_loss_mlp": 0.01029748, + "balance_loss_clip": 1.04358757, + "balance_loss_mlp": 1.02076006, + "epoch": 0.9564119521433295, + "flos": 24096742531200.0, + "grad_norm": 1.926244308018758, + "language_loss": 0.6198746, + "learning_rate": 1.9846360556430965e-08, + "loss": 0.6416446, + "num_input_tokens_seen": 171708270, + "step": 7954, + "time_per_iteration": 2.7315917015075684 + }, + { + "auxiliary_loss_clip": 0.01166625, + "auxiliary_loss_mlp": 0.01029505, + "balance_loss_clip": 1.04615581, + "balance_loss_mlp": 1.02167904, + "epoch": 0.9565321950339686, + "flos": 32008903896960.0, + "grad_norm": 2.3769194222228722, + "language_loss": 0.61728275, + "learning_rate": 1.973704405660004e-08, + "loss": 0.63924408, + "num_input_tokens_seen": 171729385, + "step": 7955, + "time_per_iteration": 2.7513034343719482 + }, + { + "auxiliary_loss_clip": 0.01118832, + "auxiliary_loss_mlp": 0.01023754, + "balance_loss_clip": 1.03916764, + "balance_loss_mlp": 1.01635146, + "epoch": 0.9566524379246077, + "flos": 23588642695680.0, + "grad_norm": 1.6103977893607182, + "language_loss": 0.78005445, + "learning_rate": 1.9628027961085203e-08, + "loss": 0.80148029, + "num_input_tokens_seen": 171752615, + "step": 7956, + "time_per_iteration": 3.873039484024048 + }, + { + "auxiliary_loss_clip": 0.01126226, + "auxiliary_loss_mlp": 0.01026153, + "balance_loss_clip": 1.0386734, + "balance_loss_mlp": 1.01815987, + "epoch": 0.9567726808152468, + "flos": 38067716240640.0, + "grad_norm": 1.8875103323410178, + "language_loss": 0.83815545, + "learning_rate": 1.9519312286423894e-08, + "loss": 0.85967922, + "num_input_tokens_seen": 171775810, + "step": 7957, + "time_per_iteration": 2.9323456287384033 + }, + { + "auxiliary_loss_clip": 0.01156066, + "auxiliary_loss_mlp": 0.01027869, + "balance_loss_clip": 1.04551768, + "balance_loss_mlp": 1.0202961, + "epoch": 0.9568929237058859, + "flos": 22744059229440.0, + "grad_norm": 1.6783819386933088, + "language_loss": 0.78022766, + "learning_rate": 1.9410897049108255e-08, + "loss": 0.80206704, + "num_input_tokens_seen": 171795090, + "step": 7958, + "time_per_iteration": 3.7488503456115723 + }, + { + "auxiliary_loss_clip": 0.01174692, + "auxiliary_loss_mlp": 0.01028611, + "balance_loss_clip": 1.05046296, + "balance_loss_mlp": 1.02079105, + "epoch": 0.957013166596525, + "flos": 23841633162240.0, + "grad_norm": 2.4216589835359916, + "language_loss": 0.91176862, + "learning_rate": 1.9302782265584905e-08, + "loss": 0.93380165, + "num_input_tokens_seen": 171815755, + "step": 7959, + "time_per_iteration": 3.693263530731201 + }, + { + "auxiliary_loss_clip": 0.0111835, + "auxiliary_loss_mlp": 0.01025862, + "balance_loss_clip": 1.04089463, + "balance_loss_mlp": 1.01853907, + "epoch": 0.9571334094871641, + "flos": 17639286071040.0, + "grad_norm": 2.113419789030168, + "language_loss": 0.86850947, + "learning_rate": 1.9194967952254282e-08, + "loss": 0.88995153, + "num_input_tokens_seen": 171834330, + "step": 7960, + "time_per_iteration": 2.8568499088287354 + }, + { + "auxiliary_loss_clip": 0.01160826, + "auxiliary_loss_mlp": 0.01033069, + "balance_loss_clip": 1.04894483, + "balance_loss_mlp": 1.02520704, + "epoch": 0.9572536523778031, + "flos": 15369623441280.0, + "grad_norm": 2.5952196492733575, + "language_loss": 0.80411851, + "learning_rate": 1.9087454125472635e-08, + "loss": 0.82605743, + "num_input_tokens_seen": 171848805, + "step": 7961, + "time_per_iteration": 2.7763514518737793 + }, + { + "auxiliary_loss_clip": 0.01171531, + "auxiliary_loss_mlp": 0.01027929, + "balance_loss_clip": 1.04964662, + "balance_loss_mlp": 1.02035928, + "epoch": 0.9573738952684423, + "flos": 24969838417920.0, + "grad_norm": 4.6446081562563855, + "language_loss": 0.78250384, + "learning_rate": 1.8980240801548696e-08, + "loss": 0.80449843, + "num_input_tokens_seen": 171867995, + "step": 7962, + "time_per_iteration": 3.6884548664093018 + }, + { + "auxiliary_loss_clip": 0.0114772, + "auxiliary_loss_mlp": 0.01028964, + "balance_loss_clip": 1.04653072, + "balance_loss_mlp": 1.02079272, + "epoch": 0.9574941381590814, + "flos": 25769461034880.0, + "grad_norm": 1.7171317856446207, + "language_loss": 0.74113053, + "learning_rate": 1.8873327996747458e-08, + "loss": 0.76289737, + "num_input_tokens_seen": 171886495, + "step": 7963, + "time_per_iteration": 2.8049533367156982 + }, + { + "auxiliary_loss_clip": 0.01162354, + "auxiliary_loss_mlp": 0.01025881, + "balance_loss_clip": 1.04579556, + "balance_loss_mlp": 1.01823699, + "epoch": 0.9576143810497204, + "flos": 32307178435200.0, + "grad_norm": 1.9236739248345507, + "language_loss": 0.65882033, + "learning_rate": 1.8766715727287053e-08, + "loss": 0.68070269, + "num_input_tokens_seen": 171908200, + "step": 7964, + "time_per_iteration": 2.7680530548095703 + }, + { + "auxiliary_loss_clip": 0.01163127, + "auxiliary_loss_mlp": 0.00886336, + "balance_loss_clip": 1.04565239, + "balance_loss_mlp": 1.00068796, + "epoch": 0.9577346239403596, + "flos": 27745733376000.0, + "grad_norm": 1.6068824695686907, + "language_loss": 0.79544449, + "learning_rate": 1.8660404009340546e-08, + "loss": 0.81593907, + "num_input_tokens_seen": 171928650, + "step": 7965, + "time_per_iteration": 2.8575663566589355 + }, + { + "auxiliary_loss_clip": 0.01051805, + "auxiliary_loss_mlp": 0.01001295, + "balance_loss_clip": 1.00516844, + "balance_loss_mlp": 0.9999544, + "epoch": 0.9578548668309986, + "flos": 57468313710720.0, + "grad_norm": 0.8681202579829703, + "language_loss": 0.59530699, + "learning_rate": 1.8554392859035485e-08, + "loss": 0.61583799, + "num_input_tokens_seen": 171986400, + "step": 7966, + "time_per_iteration": 3.24652099609375 + }, + { + "auxiliary_loss_clip": 0.0111698, + "auxiliary_loss_mlp": 0.01020749, + "balance_loss_clip": 1.03798652, + "balance_loss_mlp": 1.01310146, + "epoch": 0.9579751097216377, + "flos": 19756040503680.0, + "grad_norm": 1.8973150972646906, + "language_loss": 0.78746939, + "learning_rate": 1.8448682292453444e-08, + "loss": 0.80884671, + "num_input_tokens_seen": 172005475, + "step": 7967, + "time_per_iteration": 2.8665847778320312 + }, + { + "auxiliary_loss_clip": 0.01167203, + "auxiliary_loss_mlp": 0.01028823, + "balance_loss_clip": 1.04601622, + "balance_loss_mlp": 1.02114046, + "epoch": 0.9580953526122769, + "flos": 18041270152320.0, + "grad_norm": 1.6934698494416092, + "language_loss": 0.65752304, + "learning_rate": 1.8343272325631154e-08, + "loss": 0.67948335, + "num_input_tokens_seen": 172024420, + "step": 7968, + "time_per_iteration": 2.714926242828369 + }, + { + "auxiliary_loss_clip": 0.01114879, + "auxiliary_loss_mlp": 0.00885971, + "balance_loss_clip": 1.03856349, + "balance_loss_mlp": 1.00052142, + "epoch": 0.9582155955029159, + "flos": 24270154416000.0, + "grad_norm": 2.2005084024952093, + "language_loss": 0.77727628, + "learning_rate": 1.8238162974558492e-08, + "loss": 0.79728484, + "num_input_tokens_seen": 172038350, + "step": 7969, + "time_per_iteration": 2.9944167137145996 + }, + { + "auxiliary_loss_clip": 0.01150205, + "auxiliary_loss_mlp": 0.01028436, + "balance_loss_clip": 1.04836202, + "balance_loss_mlp": 1.02138221, + "epoch": 0.958335838393555, + "flos": 22783309816320.0, + "grad_norm": 2.0153959750160153, + "language_loss": 0.74472487, + "learning_rate": 1.8133354255181144e-08, + "loss": 0.76651132, + "num_input_tokens_seen": 172058665, + "step": 7970, + "time_per_iteration": 2.805081605911255 + }, + { + "auxiliary_loss_clip": 0.01152898, + "auxiliary_loss_mlp": 0.01027268, + "balance_loss_clip": 1.04335141, + "balance_loss_mlp": 1.01970434, + "epoch": 0.958456081284194, + "flos": 16911484698240.0, + "grad_norm": 2.6016745425778818, + "language_loss": 0.74164647, + "learning_rate": 1.802884618339795e-08, + "loss": 0.76344812, + "num_input_tokens_seen": 172077470, + "step": 7971, + "time_per_iteration": 2.763378620147705 + }, + { + "auxiliary_loss_clip": 0.01162596, + "auxiliary_loss_mlp": 0.01033352, + "balance_loss_clip": 1.04900706, + "balance_loss_mlp": 1.02531171, + "epoch": 0.9585763241748332, + "flos": 19974951941760.0, + "grad_norm": 3.4914309414328875, + "language_loss": 0.80709958, + "learning_rate": 1.7924638775062894e-08, + "loss": 0.829059, + "num_input_tokens_seen": 172096590, + "step": 7972, + "time_per_iteration": 2.6933960914611816 + }, + { + "auxiliary_loss_clip": 0.01136319, + "auxiliary_loss_mlp": 0.01026808, + "balance_loss_clip": 1.0457778, + "balance_loss_mlp": 1.01913667, + "epoch": 0.9586965670654722, + "flos": 21395649646080.0, + "grad_norm": 2.067808365487906, + "language_loss": 0.81729007, + "learning_rate": 1.7820732045984444e-08, + "loss": 0.83892137, + "num_input_tokens_seen": 172116735, + "step": 7973, + "time_per_iteration": 2.950850009918213 + }, + { + "auxiliary_loss_clip": 0.01158188, + "auxiliary_loss_mlp": 0.01036543, + "balance_loss_clip": 1.04576409, + "balance_loss_mlp": 1.02829957, + "epoch": 0.9588168099561113, + "flos": 21435115714560.0, + "grad_norm": 1.860666064787955, + "language_loss": 0.74068642, + "learning_rate": 1.7717126011924655e-08, + "loss": 0.76263374, + "num_input_tokens_seen": 172138320, + "step": 7974, + "time_per_iteration": 2.752581834793091 + }, + { + "auxiliary_loss_clip": 0.01121568, + "auxiliary_loss_mlp": 0.01029299, + "balance_loss_clip": 1.03590453, + "balance_loss_mlp": 1.02119243, + "epoch": 0.9589370528467505, + "flos": 11763761852160.0, + "grad_norm": 2.4845411398012294, + "language_loss": 0.76460356, + "learning_rate": 1.7613820688600957e-08, + "loss": 0.78611219, + "num_input_tokens_seen": 172154225, + "step": 7975, + "time_per_iteration": 2.8664658069610596 + }, + { + "auxiliary_loss_clip": 0.0114687, + "auxiliary_loss_mlp": 0.01029461, + "balance_loss_clip": 1.04375625, + "balance_loss_mlp": 1.02183414, + "epoch": 0.9590572957373895, + "flos": 23441516588160.0, + "grad_norm": 1.7446075995425727, + "language_loss": 0.78160548, + "learning_rate": 1.7510816091684588e-08, + "loss": 0.80336881, + "num_input_tokens_seen": 172174150, + "step": 7976, + "time_per_iteration": 2.7633237838745117 + }, + { + "auxiliary_loss_clip": 0.01150885, + "auxiliary_loss_mlp": 0.01029687, + "balance_loss_clip": 1.04611313, + "balance_loss_mlp": 1.02082968, + "epoch": 0.9591775386280286, + "flos": 22528272274560.0, + "grad_norm": 2.684759298312644, + "language_loss": 0.78342128, + "learning_rate": 1.740811223680083e-08, + "loss": 0.80522698, + "num_input_tokens_seen": 172191005, + "step": 7977, + "time_per_iteration": 2.8092353343963623 + }, + { + "auxiliary_loss_clip": 0.01170554, + "auxiliary_loss_mlp": 0.01028792, + "balance_loss_clip": 1.04912782, + "balance_loss_mlp": 1.02122533, + "epoch": 0.9592977815186677, + "flos": 18186959715840.0, + "grad_norm": 2.506578729684537, + "language_loss": 0.74062103, + "learning_rate": 1.7305709139530334e-08, + "loss": 0.76261449, + "num_input_tokens_seen": 172209785, + "step": 7978, + "time_per_iteration": 2.664719343185425 + }, + { + "auxiliary_loss_clip": 0.0115087, + "auxiliary_loss_mlp": 0.010268, + "balance_loss_clip": 1.0426861, + "balance_loss_mlp": 1.01835966, + "epoch": 0.9594180244093068, + "flos": 16537797555840.0, + "grad_norm": 2.673014329695423, + "language_loss": 0.7436409, + "learning_rate": 1.7203606815407334e-08, + "loss": 0.76541758, + "num_input_tokens_seen": 172224380, + "step": 7979, + "time_per_iteration": 2.7710890769958496 + }, + { + "auxiliary_loss_clip": 0.01153435, + "auxiliary_loss_mlp": 0.01032307, + "balance_loss_clip": 1.04506755, + "balance_loss_mlp": 1.02454615, + "epoch": 0.9595382672999458, + "flos": 20554334317440.0, + "grad_norm": 1.6780404516701284, + "language_loss": 0.79623401, + "learning_rate": 1.7101805279920557e-08, + "loss": 0.81809139, + "num_input_tokens_seen": 172242540, + "step": 7980, + "time_per_iteration": 2.8465704917907715 + }, + { + "auxiliary_loss_clip": 0.01170236, + "auxiliary_loss_mlp": 0.01033967, + "balance_loss_clip": 1.04835498, + "balance_loss_mlp": 1.02590823, + "epoch": 0.959658510190585, + "flos": 22638266697600.0, + "grad_norm": 2.195496750516444, + "language_loss": 0.80871481, + "learning_rate": 1.7000304548513643e-08, + "loss": 0.8307569, + "num_input_tokens_seen": 172262645, + "step": 7981, + "time_per_iteration": 2.721292495727539 + }, + { + "auxiliary_loss_clip": 0.01135503, + "auxiliary_loss_mlp": 0.01025195, + "balance_loss_clip": 1.04046047, + "balance_loss_mlp": 1.01675487, + "epoch": 0.9597787530812241, + "flos": 19135252725120.0, + "grad_norm": 2.1248299691753143, + "language_loss": 0.82685137, + "learning_rate": 1.6899104636583394e-08, + "loss": 0.84845829, + "num_input_tokens_seen": 172280695, + "step": 7982, + "time_per_iteration": 3.6561989784240723 + }, + { + "auxiliary_loss_clip": 0.01051951, + "auxiliary_loss_mlp": 0.01002091, + "balance_loss_clip": 1.00503516, + "balance_loss_mlp": 1.00068426, + "epoch": 0.9598989959718631, + "flos": 60098124055680.0, + "grad_norm": 0.7296168402263064, + "language_loss": 0.61964822, + "learning_rate": 1.6798205559482638e-08, + "loss": 0.64018857, + "num_input_tokens_seen": 172343075, + "step": 7983, + "time_per_iteration": 3.4421958923339844 + }, + { + "auxiliary_loss_clip": 0.01139073, + "auxiliary_loss_mlp": 0.01027504, + "balance_loss_clip": 1.04302478, + "balance_loss_mlp": 1.01990473, + "epoch": 0.9600192388625023, + "flos": 20886795624960.0, + "grad_norm": 1.7071332145277036, + "language_loss": 0.76480675, + "learning_rate": 1.669760733251713e-08, + "loss": 0.78647256, + "num_input_tokens_seen": 172361950, + "step": 7984, + "time_per_iteration": 3.7913401126861572 + }, + { + "auxiliary_loss_clip": 0.01119712, + "auxiliary_loss_mlp": 0.01029262, + "balance_loss_clip": 1.03998661, + "balance_loss_mlp": 1.02182341, + "epoch": 0.9601394817531413, + "flos": 20445740524800.0, + "grad_norm": 1.5622656871255205, + "language_loss": 0.82402933, + "learning_rate": 1.659730997094755e-08, + "loss": 0.84551901, + "num_input_tokens_seen": 172380440, + "step": 7985, + "time_per_iteration": 3.8845129013061523 + }, + { + "auxiliary_loss_clip": 0.01152133, + "auxiliary_loss_mlp": 0.01024444, + "balance_loss_clip": 1.04443002, + "balance_loss_mlp": 1.01674867, + "epoch": 0.9602597246437804, + "flos": 21507152440320.0, + "grad_norm": 1.8877445746986643, + "language_loss": 0.6243211, + "learning_rate": 1.6497313489989283e-08, + "loss": 0.64608681, + "num_input_tokens_seen": 172400265, + "step": 7986, + "time_per_iteration": 2.8367702960968018 + }, + { + "auxiliary_loss_clip": 0.01126099, + "auxiliary_loss_mlp": 0.01024515, + "balance_loss_clip": 1.03438139, + "balance_loss_mlp": 1.01629543, + "epoch": 0.9603799675344196, + "flos": 29935099152000.0, + "grad_norm": 1.96567622285896, + "language_loss": 0.70148289, + "learning_rate": 1.639761790481131e-08, + "loss": 0.72298896, + "num_input_tokens_seen": 172421145, + "step": 7987, + "time_per_iteration": 3.1029586791992188 + }, + { + "auxiliary_loss_clip": 0.0115259, + "auxiliary_loss_mlp": 0.01024449, + "balance_loss_clip": 1.04588449, + "balance_loss_mlp": 1.01674831, + "epoch": 0.9605002104250586, + "flos": 28001525103360.0, + "grad_norm": 1.9107953567760085, + "language_loss": 0.79081303, + "learning_rate": 1.6298223230537754e-08, + "loss": 0.81258345, + "num_input_tokens_seen": 172438945, + "step": 7988, + "time_per_iteration": 2.984647035598755 + }, + { + "auxiliary_loss_clip": 0.01150405, + "auxiliary_loss_mlp": 0.00885741, + "balance_loss_clip": 1.04472542, + "balance_loss_mlp": 1.00052905, + "epoch": 0.9606204533156977, + "flos": 35590490870400.0, + "grad_norm": 2.163691128858216, + "language_loss": 0.69490677, + "learning_rate": 1.619912948224611e-08, + "loss": 0.71526825, + "num_input_tokens_seen": 172460150, + "step": 7989, + "time_per_iteration": 3.9432735443115234 + }, + { + "auxiliary_loss_clip": 0.01135564, + "auxiliary_loss_mlp": 0.01028151, + "balance_loss_clip": 1.04414868, + "balance_loss_mlp": 1.0199554, + "epoch": 0.9607406962063368, + "flos": 26574614346240.0, + "grad_norm": 2.4279369244840687, + "language_loss": 0.61205757, + "learning_rate": 1.6100336674969682e-08, + "loss": 0.63369471, + "num_input_tokens_seen": 172478990, + "step": 7990, + "time_per_iteration": 2.8331105709075928 + }, + { + "auxiliary_loss_clip": 0.01123479, + "auxiliary_loss_mlp": 0.01031841, + "balance_loss_clip": 1.03773069, + "balance_loss_mlp": 1.02390182, + "epoch": 0.9608609390969759, + "flos": 25331781813120.0, + "grad_norm": 1.7145544264362984, + "language_loss": 0.75906819, + "learning_rate": 1.600184482369449e-08, + "loss": 0.78062141, + "num_input_tokens_seen": 172498905, + "step": 7991, + "time_per_iteration": 2.961977243423462 + }, + { + "auxiliary_loss_clip": 0.01143276, + "auxiliary_loss_mlp": 0.01027183, + "balance_loss_clip": 1.04244566, + "balance_loss_mlp": 1.01942301, + "epoch": 0.960981181987615, + "flos": 21069114082560.0, + "grad_norm": 2.8745429960655953, + "language_loss": 0.89042437, + "learning_rate": 1.5903653943362126e-08, + "loss": 0.91212898, + "num_input_tokens_seen": 172517900, + "step": 7992, + "time_per_iteration": 2.741088628768921 + }, + { + "auxiliary_loss_clip": 0.01150157, + "auxiliary_loss_mlp": 0.01027208, + "balance_loss_clip": 1.04442418, + "balance_loss_mlp": 1.01976907, + "epoch": 0.9611014248782541, + "flos": 17823256554240.0, + "grad_norm": 2.771368706584334, + "language_loss": 0.77216768, + "learning_rate": 1.580576404886802e-08, + "loss": 0.79394138, + "num_input_tokens_seen": 172536430, + "step": 7993, + "time_per_iteration": 2.6980812549591064 + }, + { + "auxiliary_loss_clip": 0.01159139, + "auxiliary_loss_mlp": 0.01024881, + "balance_loss_clip": 1.0453043, + "balance_loss_mlp": 1.01790738, + "epoch": 0.9612216677688932, + "flos": 19354631040000.0, + "grad_norm": 2.0923022706585335, + "language_loss": 0.79750216, + "learning_rate": 1.570817515506162e-08, + "loss": 0.81934237, + "num_input_tokens_seen": 172555120, + "step": 7994, + "time_per_iteration": 2.7085189819335938 + }, + { + "auxiliary_loss_clip": 0.01167443, + "auxiliary_loss_mlp": 0.01032625, + "balance_loss_clip": 1.04759836, + "balance_loss_mlp": 1.02530551, + "epoch": 0.9613419106595322, + "flos": 15808739207040.0, + "grad_norm": 2.0554181033746106, + "language_loss": 0.81001359, + "learning_rate": 1.561088727674753e-08, + "loss": 0.83201432, + "num_input_tokens_seen": 172569330, + "step": 7995, + "time_per_iteration": 2.7713146209716797 + }, + { + "auxiliary_loss_clip": 0.01143259, + "auxiliary_loss_mlp": 0.01026777, + "balance_loss_clip": 1.04238153, + "balance_loss_mlp": 1.01797295, + "epoch": 0.9614621535501714, + "flos": 25702488126720.0, + "grad_norm": 2.127390776181393, + "language_loss": 0.7122246, + "learning_rate": 1.551390042868417e-08, + "loss": 0.73392498, + "num_input_tokens_seen": 172591100, + "step": 7996, + "time_per_iteration": 2.845867395401001 + }, + { + "auxiliary_loss_clip": 0.0115669, + "auxiliary_loss_mlp": 0.01027892, + "balance_loss_clip": 1.04517567, + "balance_loss_mlp": 1.02035809, + "epoch": 0.9615823964408104, + "flos": 17819054663040.0, + "grad_norm": 1.8328073145649233, + "language_loss": 0.70907164, + "learning_rate": 1.5417214625584207e-08, + "loss": 0.73091745, + "num_input_tokens_seen": 172608755, + "step": 7997, + "time_per_iteration": 2.6418278217315674 + }, + { + "auxiliary_loss_clip": 0.01154659, + "auxiliary_loss_mlp": 0.01026256, + "balance_loss_clip": 1.04404855, + "balance_loss_mlp": 1.01860833, + "epoch": 0.9617026393314495, + "flos": 20190020624640.0, + "grad_norm": 1.8431465509327813, + "language_loss": 0.85094202, + "learning_rate": 1.5320829882114806e-08, + "loss": 0.87275112, + "num_input_tokens_seen": 172626830, + "step": 7998, + "time_per_iteration": 2.6570565700531006 + }, + { + "auxiliary_loss_clip": 0.01166297, + "auxiliary_loss_mlp": 0.01029012, + "balance_loss_clip": 1.04454553, + "balance_loss_mlp": 1.02098894, + "epoch": 0.9618228822220887, + "flos": 20267013427200.0, + "grad_norm": 4.113680464745845, + "language_loss": 0.79110795, + "learning_rate": 1.5224746212897378e-08, + "loss": 0.813061, + "num_input_tokens_seen": 172646125, + "step": 7999, + "time_per_iteration": 2.6141085624694824 + }, + { + "auxiliary_loss_clip": 0.01163262, + "auxiliary_loss_mlp": 0.01034405, + "balance_loss_clip": 1.04459846, + "balance_loss_mlp": 1.02615547, + "epoch": 0.9619431251127277, + "flos": 21031300039680.0, + "grad_norm": 2.386377630455994, + "language_loss": 0.77317393, + "learning_rate": 1.512896363250804e-08, + "loss": 0.79515058, + "num_input_tokens_seen": 172666235, + "step": 8000, + "time_per_iteration": 2.641005516052246 + }, + { + "auxiliary_loss_clip": 0.01160335, + "auxiliary_loss_mlp": 0.01021999, + "balance_loss_clip": 1.04353428, + "balance_loss_mlp": 1.01492977, + "epoch": 0.9620633680033668, + "flos": 22382654538240.0, + "grad_norm": 2.0066743594291734, + "language_loss": 0.75382757, + "learning_rate": 1.503348215547673e-08, + "loss": 0.77565086, + "num_input_tokens_seen": 172687325, + "step": 8001, + "time_per_iteration": 2.629539728164673 + }, + { + "auxiliary_loss_clip": 0.01146465, + "auxiliary_loss_mlp": 0.01027178, + "balance_loss_clip": 1.04407144, + "balance_loss_mlp": 1.0196178, + "epoch": 0.962183610894006, + "flos": 18471730740480.0, + "grad_norm": 2.0223732535627654, + "language_loss": 0.80691826, + "learning_rate": 1.4938301796288078e-08, + "loss": 0.82865477, + "num_input_tokens_seen": 172703895, + "step": 8002, + "time_per_iteration": 2.7099342346191406 + }, + { + "auxiliary_loss_clip": 0.01168587, + "auxiliary_loss_mlp": 0.01029212, + "balance_loss_clip": 1.04730749, + "balance_loss_mlp": 1.0217731, + "epoch": 0.962303853784645, + "flos": 18435245500800.0, + "grad_norm": 2.4339280574698026, + "language_loss": 0.81803703, + "learning_rate": 1.4843422569380537e-08, + "loss": 0.84001499, + "num_input_tokens_seen": 172720650, + "step": 8003, + "time_per_iteration": 2.6300957202911377 + }, + { + "auxiliary_loss_clip": 0.01121479, + "auxiliary_loss_mlp": 0.0102829, + "balance_loss_clip": 1.03904498, + "balance_loss_mlp": 1.02027655, + "epoch": 0.9624240966752841, + "flos": 26391074826240.0, + "grad_norm": 1.6704810732830493, + "language_loss": 0.82604718, + "learning_rate": 1.4748844489147483e-08, + "loss": 0.84754491, + "num_input_tokens_seen": 172737640, + "step": 8004, + "time_per_iteration": 2.946932315826416 + }, + { + "auxiliary_loss_clip": 0.01147825, + "auxiliary_loss_mlp": 0.01026357, + "balance_loss_clip": 1.04135537, + "balance_loss_mlp": 1.01894188, + "epoch": 0.9625443395659231, + "flos": 14647675985280.0, + "grad_norm": 1.827716016519358, + "language_loss": 0.70711404, + "learning_rate": 1.4654567569936326e-08, + "loss": 0.72885585, + "num_input_tokens_seen": 172755215, + "step": 8005, + "time_per_iteration": 2.715479612350464 + }, + { + "auxiliary_loss_clip": 0.01126765, + "auxiliary_loss_mlp": 0.01031068, + "balance_loss_clip": 1.04161787, + "balance_loss_mlp": 1.0235579, + "epoch": 0.9626645824565623, + "flos": 18367626147840.0, + "grad_norm": 1.9822391359812066, + "language_loss": 0.82706124, + "learning_rate": 1.456059182604874e-08, + "loss": 0.84863961, + "num_input_tokens_seen": 172774020, + "step": 8006, + "time_per_iteration": 2.832174301147461 + }, + { + "auxiliary_loss_clip": 0.01170667, + "auxiliary_loss_mlp": 0.01028627, + "balance_loss_clip": 1.04842472, + "balance_loss_mlp": 1.02124476, + "epoch": 0.9627848253472013, + "flos": 16580424021120.0, + "grad_norm": 2.0174077943884936, + "language_loss": 0.76164436, + "learning_rate": 1.4466917271740653e-08, + "loss": 0.78363729, + "num_input_tokens_seen": 172792220, + "step": 8007, + "time_per_iteration": 2.9273955821990967 + }, + { + "auxiliary_loss_clip": 0.01145357, + "auxiliary_loss_mlp": 0.01026453, + "balance_loss_clip": 1.04272914, + "balance_loss_mlp": 1.01763749, + "epoch": 0.9629050682378404, + "flos": 20886867452160.0, + "grad_norm": 5.631730471623854, + "language_loss": 0.67432511, + "learning_rate": 1.4373543921222697e-08, + "loss": 0.69604325, + "num_input_tokens_seen": 172811805, + "step": 8008, + "time_per_iteration": 3.7496769428253174 + }, + { + "auxiliary_loss_clip": 0.01147558, + "auxiliary_loss_mlp": 0.01025644, + "balance_loss_clip": 1.04380178, + "balance_loss_mlp": 1.01777017, + "epoch": 0.9630253111284796, + "flos": 17019252478080.0, + "grad_norm": 1.9255650823693231, + "language_loss": 0.77952373, + "learning_rate": 1.428047178865932e-08, + "loss": 0.8012557, + "num_input_tokens_seen": 172828595, + "step": 8009, + "time_per_iteration": 2.8167457580566406 + }, + { + "auxiliary_loss_clip": 0.01145609, + "auxiliary_loss_mlp": 0.01032393, + "balance_loss_clip": 1.04064798, + "balance_loss_mlp": 1.02449584, + "epoch": 0.9631455540191186, + "flos": 20338942412160.0, + "grad_norm": 1.6774200761883213, + "language_loss": 0.74420029, + "learning_rate": 1.4187700888169451e-08, + "loss": 0.7659803, + "num_input_tokens_seen": 172847770, + "step": 8010, + "time_per_iteration": 3.721350908279419 + }, + { + "auxiliary_loss_clip": 0.01049381, + "auxiliary_loss_mlp": 0.01000413, + "balance_loss_clip": 1.00567913, + "balance_loss_mlp": 0.9991495, + "epoch": 0.9632657969097577, + "flos": 65956700033280.0, + "grad_norm": 0.7505833825358119, + "language_loss": 0.56957543, + "learning_rate": 1.40952312338265e-08, + "loss": 0.59007335, + "num_input_tokens_seen": 172912415, + "step": 8011, + "time_per_iteration": 4.201587200164795 + }, + { + "auxiliary_loss_clip": 0.01140205, + "auxiliary_loss_mlp": 0.01024525, + "balance_loss_clip": 1.04104805, + "balance_loss_mlp": 1.01669335, + "epoch": 0.9633860398003968, + "flos": 44419523823360.0, + "grad_norm": 1.9812042835621149, + "language_loss": 0.68745148, + "learning_rate": 1.4003062839657909e-08, + "loss": 0.70909882, + "num_input_tokens_seen": 172934895, + "step": 8012, + "time_per_iteration": 3.020918369293213 + }, + { + "auxiliary_loss_clip": 0.01134335, + "auxiliary_loss_mlp": 0.01024611, + "balance_loss_clip": 1.04301357, + "balance_loss_mlp": 1.01710629, + "epoch": 0.9635062826910359, + "flos": 24827704300800.0, + "grad_norm": 1.583889185317871, + "language_loss": 0.79870462, + "learning_rate": 1.391119571964583e-08, + "loss": 0.82029408, + "num_input_tokens_seen": 172955835, + "step": 8013, + "time_per_iteration": 2.863131523132324 + }, + { + "auxiliary_loss_clip": 0.01157329, + "auxiliary_loss_mlp": 0.01027867, + "balance_loss_clip": 1.04608321, + "balance_loss_mlp": 1.0209235, + "epoch": 0.9636265255816749, + "flos": 15961360095360.0, + "grad_norm": 1.91483825442669, + "language_loss": 0.72781461, + "learning_rate": 1.3819629887726225e-08, + "loss": 0.74966657, + "num_input_tokens_seen": 172973925, + "step": 8014, + "time_per_iteration": 3.5828981399536133 + }, + { + "auxiliary_loss_clip": 0.01155179, + "auxiliary_loss_mlp": 0.01030882, + "balance_loss_clip": 1.04714251, + "balance_loss_mlp": 1.02309752, + "epoch": 0.9637467684723141, + "flos": 22601781457920.0, + "grad_norm": 1.7104444658932672, + "language_loss": 0.76219261, + "learning_rate": 1.3728365357789317e-08, + "loss": 0.78405321, + "num_input_tokens_seen": 172993290, + "step": 8015, + "time_per_iteration": 2.719064950942993 + }, + { + "auxiliary_loss_clip": 0.01105171, + "auxiliary_loss_mlp": 0.01027998, + "balance_loss_clip": 1.03919315, + "balance_loss_mlp": 1.02034187, + "epoch": 0.9638670113629532, + "flos": 17565812801280.0, + "grad_norm": 2.4545617057817735, + "language_loss": 0.76800537, + "learning_rate": 1.3637402143680254e-08, + "loss": 0.78933704, + "num_input_tokens_seen": 173008190, + "step": 8016, + "time_per_iteration": 2.792959690093994 + }, + { + "auxiliary_loss_clip": 0.01029552, + "auxiliary_loss_mlp": 0.01003721, + "balance_loss_clip": 1.00667942, + "balance_loss_mlp": 1.00230241, + "epoch": 0.9639872542535922, + "flos": 55072139379840.0, + "grad_norm": 0.7229472466460695, + "language_loss": 0.55028015, + "learning_rate": 1.3546740259197998e-08, + "loss": 0.57061285, + "num_input_tokens_seen": 173061000, + "step": 8017, + "time_per_iteration": 3.244410514831543 + }, + { + "auxiliary_loss_clip": 0.01152717, + "auxiliary_loss_mlp": 0.01029526, + "balance_loss_clip": 1.04577911, + "balance_loss_mlp": 1.02181351, + "epoch": 0.9641074971442314, + "flos": 24134484746880.0, + "grad_norm": 5.081068109475002, + "language_loss": 0.69954598, + "learning_rate": 1.3456379718095989e-08, + "loss": 0.72136843, + "num_input_tokens_seen": 173081415, + "step": 8018, + "time_per_iteration": 2.7689754962921143 + }, + { + "auxiliary_loss_clip": 0.0104288, + "auxiliary_loss_mlp": 0.01002207, + "balance_loss_clip": 1.00528574, + "balance_loss_mlp": 1.00064588, + "epoch": 0.9642277400348704, + "flos": 66747416077440.0, + "grad_norm": 0.8412505784173353, + "language_loss": 0.62008333, + "learning_rate": 1.3366320534081487e-08, + "loss": 0.64053422, + "num_input_tokens_seen": 173144095, + "step": 8019, + "time_per_iteration": 3.296351432800293 + }, + { + "auxiliary_loss_clip": 0.01159061, + "auxiliary_loss_mlp": 0.01027708, + "balance_loss_clip": 1.04544759, + "balance_loss_mlp": 1.01982212, + "epoch": 0.9643479829255095, + "flos": 30920272450560.0, + "grad_norm": 2.64789577213312, + "language_loss": 0.7616818, + "learning_rate": 1.3276562720816675e-08, + "loss": 0.78354955, + "num_input_tokens_seen": 173165605, + "step": 8020, + "time_per_iteration": 2.742572784423828 + }, + { + "auxiliary_loss_clip": 0.01168309, + "auxiliary_loss_mlp": 0.01027036, + "balance_loss_clip": 1.04607487, + "balance_loss_mlp": 1.01891804, + "epoch": 0.9644682258161487, + "flos": 20048245643520.0, + "grad_norm": 2.1673277834364733, + "language_loss": 0.8261627, + "learning_rate": 1.3187106291917549e-08, + "loss": 0.84811616, + "num_input_tokens_seen": 173182595, + "step": 8021, + "time_per_iteration": 2.636868953704834 + }, + { + "auxiliary_loss_clip": 0.01152959, + "auxiliary_loss_mlp": 0.01025996, + "balance_loss_clip": 1.04489446, + "balance_loss_mlp": 1.01784158, + "epoch": 0.9645884687067877, + "flos": 21178713456000.0, + "grad_norm": 1.7623273722877244, + "language_loss": 0.70278859, + "learning_rate": 1.309795126095503e-08, + "loss": 0.72457814, + "num_input_tokens_seen": 173200895, + "step": 8022, + "time_per_iteration": 2.7483086585998535 + }, + { + "auxiliary_loss_clip": 0.01104594, + "auxiliary_loss_mlp": 0.01027613, + "balance_loss_clip": 1.03528035, + "balance_loss_mlp": 1.02001035, + "epoch": 0.9647087115974268, + "flos": 18945967029120.0, + "grad_norm": 2.1647529223750523, + "language_loss": 0.80677199, + "learning_rate": 1.3009097641453192e-08, + "loss": 0.82809407, + "num_input_tokens_seen": 173218745, + "step": 8023, + "time_per_iteration": 2.90091872215271 + }, + { + "auxiliary_loss_clip": 0.01146132, + "auxiliary_loss_mlp": 0.01023432, + "balance_loss_clip": 1.04427838, + "balance_loss_mlp": 1.01603532, + "epoch": 0.9648289544880659, + "flos": 16545088016640.0, + "grad_norm": 1.6725820242093095, + "language_loss": 0.76146007, + "learning_rate": 1.2920545446891474e-08, + "loss": 0.78315568, + "num_input_tokens_seen": 173235465, + "step": 8024, + "time_per_iteration": 2.7017719745635986 + }, + { + "auxiliary_loss_clip": 0.01152112, + "auxiliary_loss_mlp": 0.01025631, + "balance_loss_clip": 1.04698741, + "balance_loss_mlp": 1.01768601, + "epoch": 0.964949197378705, + "flos": 24057527857920.0, + "grad_norm": 1.670838611397477, + "language_loss": 0.71003842, + "learning_rate": 1.2832294690703127e-08, + "loss": 0.73181593, + "num_input_tokens_seen": 173254440, + "step": 8025, + "time_per_iteration": 2.7355241775512695 + }, + { + "auxiliary_loss_clip": 0.01157992, + "auxiliary_loss_mlp": 0.01027378, + "balance_loss_clip": 1.04690289, + "balance_loss_mlp": 1.0193311, + "epoch": 0.965069440269344, + "flos": 23365565280000.0, + "grad_norm": 1.8547491092294082, + "language_loss": 0.77357042, + "learning_rate": 1.2744345386275668e-08, + "loss": 0.79542416, + "num_input_tokens_seen": 173273980, + "step": 8026, + "time_per_iteration": 2.672391891479492 + }, + { + "auxiliary_loss_clip": 0.01147666, + "auxiliary_loss_mlp": 0.01028049, + "balance_loss_clip": 1.04697967, + "balance_loss_mlp": 1.02053297, + "epoch": 0.9651896831599832, + "flos": 25374875155200.0, + "grad_norm": 2.0005751946010095, + "language_loss": 0.78785074, + "learning_rate": 1.265669754695109e-08, + "loss": 0.80960786, + "num_input_tokens_seen": 173293550, + "step": 8027, + "time_per_iteration": 2.762322425842285 + }, + { + "auxiliary_loss_clip": 0.01120258, + "auxiliary_loss_mlp": 0.01027063, + "balance_loss_clip": 1.03727424, + "balance_loss_mlp": 1.01921916, + "epoch": 0.9653099260506223, + "flos": 22272875596800.0, + "grad_norm": 1.8044377573085641, + "language_loss": 0.81596911, + "learning_rate": 1.2569351186025201e-08, + "loss": 0.83744234, + "num_input_tokens_seen": 173312005, + "step": 8028, + "time_per_iteration": 2.875493049621582 + }, + { + "auxiliary_loss_clip": 0.01127455, + "auxiliary_loss_mlp": 0.01025712, + "balance_loss_clip": 1.03980613, + "balance_loss_mlp": 1.01811218, + "epoch": 0.9654301689412613, + "flos": 26760847386240.0, + "grad_norm": 1.5368015948058427, + "language_loss": 0.75568968, + "learning_rate": 1.2482306316748737e-08, + "loss": 0.77722132, + "num_input_tokens_seen": 173332450, + "step": 8029, + "time_per_iteration": 2.796171188354492 + }, + { + "auxiliary_loss_clip": 0.01163892, + "auxiliary_loss_mlp": 0.01023639, + "balance_loss_clip": 1.04410052, + "balance_loss_mlp": 1.01572323, + "epoch": 0.9655504118319005, + "flos": 17412689122560.0, + "grad_norm": 2.6980758471655704, + "language_loss": 0.7842024, + "learning_rate": 1.2395562952326021e-08, + "loss": 0.80607772, + "num_input_tokens_seen": 173349610, + "step": 8030, + "time_per_iteration": 2.608603000640869 + }, + { + "auxiliary_loss_clip": 0.01160086, + "auxiliary_loss_mlp": 0.0102676, + "balance_loss_clip": 1.04700184, + "balance_loss_mlp": 1.01863003, + "epoch": 0.9656706547225395, + "flos": 22126970551680.0, + "grad_norm": 2.1236774845276076, + "language_loss": 0.81640804, + "learning_rate": 1.2309121105916309e-08, + "loss": 0.83827657, + "num_input_tokens_seen": 173367900, + "step": 8031, + "time_per_iteration": 2.7279248237609863 + }, + { + "auxiliary_loss_clip": 0.01163697, + "auxiliary_loss_mlp": 0.01027188, + "balance_loss_clip": 1.04718387, + "balance_loss_mlp": 1.01929617, + "epoch": 0.9657908976131786, + "flos": 37049289926400.0, + "grad_norm": 2.1927225627901783, + "language_loss": 0.69427025, + "learning_rate": 1.222298079063222e-08, + "loss": 0.71617913, + "num_input_tokens_seen": 173389040, + "step": 8032, + "time_per_iteration": 2.8323051929473877 + }, + { + "auxiliary_loss_clip": 0.01158265, + "auxiliary_loss_mlp": 0.01023384, + "balance_loss_clip": 1.04530072, + "balance_loss_mlp": 1.01634431, + "epoch": 0.9659111405038178, + "flos": 24389809597440.0, + "grad_norm": 2.0578153295660364, + "language_loss": 0.72321635, + "learning_rate": 1.2137142019541524e-08, + "loss": 0.74503285, + "num_input_tokens_seen": 173407595, + "step": 8033, + "time_per_iteration": 2.705932855606079 + }, + { + "auxiliary_loss_clip": 0.01148257, + "auxiliary_loss_mlp": 0.01028859, + "balance_loss_clip": 1.04481864, + "balance_loss_mlp": 1.02085972, + "epoch": 0.9660313833944568, + "flos": 25009412227200.0, + "grad_norm": 1.8419643942714572, + "language_loss": 0.73472881, + "learning_rate": 1.2051604805666027e-08, + "loss": 0.75650001, + "num_input_tokens_seen": 173424720, + "step": 8034, + "time_per_iteration": 3.5828044414520264 + }, + { + "auxiliary_loss_clip": 0.01166904, + "auxiliary_loss_mlp": 0.00885807, + "balance_loss_clip": 1.04621613, + "balance_loss_mlp": 1.00065398, + "epoch": 0.9661516262850959, + "flos": 11801575895040.0, + "grad_norm": 2.204650469485695, + "language_loss": 0.78051162, + "learning_rate": 1.196636916198135e-08, + "loss": 0.80103874, + "num_input_tokens_seen": 173442260, + "step": 8035, + "time_per_iteration": 2.619025468826294 + }, + { + "auxiliary_loss_clip": 0.01169764, + "auxiliary_loss_mlp": 0.01023705, + "balance_loss_clip": 1.04699624, + "balance_loss_mlp": 1.01615882, + "epoch": 0.9662718691757349, + "flos": 20047778766720.0, + "grad_norm": 1.967201175965134, + "language_loss": 0.76748633, + "learning_rate": 1.1881435101418036e-08, + "loss": 0.78942108, + "num_input_tokens_seen": 173461675, + "step": 8036, + "time_per_iteration": 2.7457656860351562 + }, + { + "auxiliary_loss_clip": 0.01046159, + "auxiliary_loss_mlp": 0.01003666, + "balance_loss_clip": 1.00712931, + "balance_loss_mlp": 1.00216377, + "epoch": 0.9663921120663741, + "flos": 68027703517440.0, + "grad_norm": 0.7251558654130021, + "language_loss": 0.65563405, + "learning_rate": 1.1796802636860003e-08, + "loss": 0.67613226, + "num_input_tokens_seen": 173530205, + "step": 8037, + "time_per_iteration": 5.17301344871521 + }, + { + "auxiliary_loss_clip": 0.01168679, + "auxiliary_loss_mlp": 0.01027926, + "balance_loss_clip": 1.045856, + "balance_loss_mlp": 1.01999259, + "epoch": 0.9665123549570132, + "flos": 26322916769280.0, + "grad_norm": 1.9071837694839036, + "language_loss": 0.73715013, + "learning_rate": 1.1712471781146316e-08, + "loss": 0.75911617, + "num_input_tokens_seen": 173549540, + "step": 8038, + "time_per_iteration": 2.6898348331451416 + }, + { + "auxiliary_loss_clip": 0.01167328, + "auxiliary_loss_mlp": 0.01026084, + "balance_loss_clip": 1.04619014, + "balance_loss_mlp": 1.01833498, + "epoch": 0.9666325978476522, + "flos": 43941121557120.0, + "grad_norm": 1.8177813761638009, + "language_loss": 0.6710757, + "learning_rate": 1.1628442547069628e-08, + "loss": 0.69300973, + "num_input_tokens_seen": 173571740, + "step": 8039, + "time_per_iteration": 2.8438222408294678 + }, + { + "auxiliary_loss_clip": 0.01162022, + "auxiliary_loss_mlp": 0.00886523, + "balance_loss_clip": 1.04551744, + "balance_loss_mlp": 1.00055027, + "epoch": 0.9667528407382914, + "flos": 21543422198400.0, + "grad_norm": 2.602176391997981, + "language_loss": 0.77190191, + "learning_rate": 1.1544714947377521e-08, + "loss": 0.79238737, + "num_input_tokens_seen": 173589425, + "step": 8040, + "time_per_iteration": 2.6661536693573 + }, + { + "auxiliary_loss_clip": 0.0117384, + "auxiliary_loss_mlp": 0.01027397, + "balance_loss_clip": 1.04985309, + "balance_loss_mlp": 1.01933825, + "epoch": 0.9668730836289304, + "flos": 23878585278720.0, + "grad_norm": 2.516939626211888, + "language_loss": 0.69591194, + "learning_rate": 1.1461288994770945e-08, + "loss": 0.7179243, + "num_input_tokens_seen": 173608500, + "step": 8041, + "time_per_iteration": 3.5607106685638428 + }, + { + "auxiliary_loss_clip": 0.01172276, + "auxiliary_loss_mlp": 0.01026207, + "balance_loss_clip": 1.04718423, + "balance_loss_mlp": 1.01795173, + "epoch": 0.9669933265195695, + "flos": 28293011971200.0, + "grad_norm": 1.8796489128650864, + "language_loss": 0.77057922, + "learning_rate": 1.1378164701906002e-08, + "loss": 0.79256409, + "num_input_tokens_seen": 173630265, + "step": 8042, + "time_per_iteration": 2.762047052383423 + }, + { + "auxiliary_loss_clip": 0.01172398, + "auxiliary_loss_mlp": 0.0102487, + "balance_loss_clip": 1.04786134, + "balance_loss_mlp": 1.01615596, + "epoch": 0.9671135694102087, + "flos": 22454763091200.0, + "grad_norm": 1.700391165362716, + "language_loss": 0.66892225, + "learning_rate": 1.1295342081392156e-08, + "loss": 0.6908949, + "num_input_tokens_seen": 173649625, + "step": 8043, + "time_per_iteration": 2.6069371700286865 + }, + { + "auxiliary_loss_clip": 0.01142431, + "auxiliary_loss_mlp": 0.0102784, + "balance_loss_clip": 1.04337096, + "balance_loss_mlp": 1.02029681, + "epoch": 0.9672338123008477, + "flos": 20155941596160.0, + "grad_norm": 1.6710638012815997, + "language_loss": 0.69397891, + "learning_rate": 1.1212821145793804e-08, + "loss": 0.71568161, + "num_input_tokens_seen": 173669240, + "step": 8044, + "time_per_iteration": 2.6829349994659424 + }, + { + "auxiliary_loss_clip": 0.01149511, + "auxiliary_loss_mlp": 0.01035328, + "balance_loss_clip": 1.04271472, + "balance_loss_mlp": 1.02715051, + "epoch": 0.9673540551914868, + "flos": 16977487939200.0, + "grad_norm": 2.6296220037559785, + "language_loss": 0.79253548, + "learning_rate": 1.1130601907629156e-08, + "loss": 0.81438386, + "num_input_tokens_seen": 173686970, + "step": 8045, + "time_per_iteration": 2.78544020652771 + }, + { + "auxiliary_loss_clip": 0.01051053, + "auxiliary_loss_mlp": 0.01001322, + "balance_loss_clip": 1.00443792, + "balance_loss_mlp": 0.99985617, + "epoch": 0.9674742980821259, + "flos": 61892903952000.0, + "grad_norm": 0.9470061427135749, + "language_loss": 0.64811265, + "learning_rate": 1.1048684379370899e-08, + "loss": 0.66863638, + "num_input_tokens_seen": 173747655, + "step": 8046, + "time_per_iteration": 3.228282928466797 + }, + { + "auxiliary_loss_clip": 0.0113983, + "auxiliary_loss_mlp": 0.01020135, + "balance_loss_clip": 1.0430057, + "balance_loss_mlp": 1.01277387, + "epoch": 0.967594540972765, + "flos": 18697824898560.0, + "grad_norm": 1.8564493583523036, + "language_loss": 0.74267888, + "learning_rate": 1.0967068573445759e-08, + "loss": 0.76427853, + "num_input_tokens_seen": 173765140, + "step": 8047, + "time_per_iteration": 2.739323854446411 + }, + { + "auxiliary_loss_clip": 0.01145814, + "auxiliary_loss_mlp": 0.01026191, + "balance_loss_clip": 1.04219925, + "balance_loss_mlp": 1.01825166, + "epoch": 0.967714783863404, + "flos": 20777411733120.0, + "grad_norm": 2.5307038483912834, + "language_loss": 0.65360749, + "learning_rate": 1.0885754502234945e-08, + "loss": 0.67532754, + "num_input_tokens_seen": 173784800, + "step": 8048, + "time_per_iteration": 2.702632188796997 + }, + { + "auxiliary_loss_clip": 0.01131681, + "auxiliary_loss_mlp": 0.01029659, + "balance_loss_clip": 1.04285431, + "balance_loss_mlp": 1.02223766, + "epoch": 0.9678350267540432, + "flos": 23185473465600.0, + "grad_norm": 2.1697338708473133, + "language_loss": 0.78195792, + "learning_rate": 1.08047421780737e-08, + "loss": 0.80357134, + "num_input_tokens_seen": 173803990, + "step": 8049, + "time_per_iteration": 2.7576608657836914 + }, + { + "auxiliary_loss_clip": 0.01154438, + "auxiliary_loss_mlp": 0.00886039, + "balance_loss_clip": 1.0435307, + "balance_loss_mlp": 1.00056434, + "epoch": 0.9679552696446823, + "flos": 21726063878400.0, + "grad_norm": 1.9844606896888681, + "language_loss": 0.73860514, + "learning_rate": 1.0724031613251305e-08, + "loss": 0.7590099, + "num_input_tokens_seen": 173821890, + "step": 8050, + "time_per_iteration": 2.7755069732666016 + }, + { + "auxiliary_loss_clip": 0.01165029, + "auxiliary_loss_mlp": 0.01022069, + "balance_loss_clip": 1.04686451, + "balance_loss_mlp": 1.01347387, + "epoch": 0.9680755125353213, + "flos": 26869046129280.0, + "grad_norm": 2.159736221963737, + "language_loss": 0.66936678, + "learning_rate": 1.0643622820011744e-08, + "loss": 0.69123781, + "num_input_tokens_seen": 173842945, + "step": 8051, + "time_per_iteration": 2.7917540073394775 + }, + { + "auxiliary_loss_clip": 0.01172233, + "auxiliary_loss_mlp": 0.01035465, + "balance_loss_clip": 1.04810548, + "balance_loss_mlp": 1.02711725, + "epoch": 0.9681957554259605, + "flos": 28325008010880.0, + "grad_norm": 2.6764584471357344, + "language_loss": 0.67881954, + "learning_rate": 1.0563515810552814e-08, + "loss": 0.7008965, + "num_input_tokens_seen": 173859915, + "step": 8052, + "time_per_iteration": 2.6702582836151123 + }, + { + "auxiliary_loss_clip": 0.01172249, + "auxiliary_loss_mlp": 0.01029955, + "balance_loss_clip": 1.05017054, + "balance_loss_mlp": 1.02220082, + "epoch": 0.9683159983165995, + "flos": 20557674282240.0, + "grad_norm": 1.6755780807298197, + "language_loss": 0.73319757, + "learning_rate": 1.0483710597026795e-08, + "loss": 0.75521964, + "num_input_tokens_seen": 173879775, + "step": 8053, + "time_per_iteration": 2.632723093032837 + }, + { + "auxiliary_loss_clip": 0.01138858, + "auxiliary_loss_mlp": 0.01031179, + "balance_loss_clip": 1.04221487, + "balance_loss_mlp": 1.02350807, + "epoch": 0.9684362412072386, + "flos": 24207958016640.0, + "grad_norm": 2.1223622795548125, + "language_loss": 0.74000287, + "learning_rate": 1.0404207191540227e-08, + "loss": 0.76170319, + "num_input_tokens_seen": 173900230, + "step": 8054, + "time_per_iteration": 2.849550724029541 + }, + { + "auxiliary_loss_clip": 0.01165453, + "auxiliary_loss_mlp": 0.01028169, + "balance_loss_clip": 1.04453087, + "balance_loss_mlp": 1.02039075, + "epoch": 0.9685564840978778, + "flos": 22346241125760.0, + "grad_norm": 1.9210544146585646, + "language_loss": 0.74414492, + "learning_rate": 1.0325005606153236e-08, + "loss": 0.76608121, + "num_input_tokens_seen": 173919690, + "step": 8055, + "time_per_iteration": 2.6454432010650635 + }, + { + "auxiliary_loss_clip": 0.01124766, + "auxiliary_loss_mlp": 0.01029348, + "balance_loss_clip": 1.04029059, + "balance_loss_mlp": 1.02161753, + "epoch": 0.9686767269885168, + "flos": 14386389477120.0, + "grad_norm": 2.969817190669767, + "language_loss": 0.79147971, + "learning_rate": 1.0246105852881104e-08, + "loss": 0.81302083, + "num_input_tokens_seen": 173934790, + "step": 8056, + "time_per_iteration": 2.9975297451019287 + }, + { + "auxiliary_loss_clip": 0.01170412, + "auxiliary_loss_mlp": 0.01033657, + "balance_loss_clip": 1.04640341, + "balance_loss_mlp": 1.02552748, + "epoch": 0.9687969698791559, + "flos": 21287630471040.0, + "grad_norm": 1.8413064939158048, + "language_loss": 0.7841903, + "learning_rate": 1.0167507943692476e-08, + "loss": 0.80623108, + "num_input_tokens_seen": 173953875, + "step": 8057, + "time_per_iteration": 2.710644006729126 + }, + { + "auxiliary_loss_clip": 0.01160172, + "auxiliary_loss_mlp": 0.01028028, + "balance_loss_clip": 1.04809976, + "balance_loss_mlp": 1.01961136, + "epoch": 0.968917212769795, + "flos": 19828328624640.0, + "grad_norm": 2.04579912881618, + "language_loss": 0.70966971, + "learning_rate": 1.008921189051093e-08, + "loss": 0.73155165, + "num_input_tokens_seen": 173971220, + "step": 8058, + "time_per_iteration": 2.784271240234375 + }, + { + "auxiliary_loss_clip": 0.01171502, + "auxiliary_loss_mlp": 0.01028847, + "balance_loss_clip": 1.05037141, + "balance_loss_mlp": 1.02088332, + "epoch": 0.9690374556604341, + "flos": 21681749473920.0, + "grad_norm": 4.286190246658306, + "language_loss": 0.77061605, + "learning_rate": 1.0011217705213848e-08, + "loss": 0.79261953, + "num_input_tokens_seen": 173989095, + "step": 8059, + "time_per_iteration": 2.647634506225586 + }, + { + "auxiliary_loss_clip": 0.01156383, + "auxiliary_loss_mlp": 0.01025768, + "balance_loss_clip": 1.0466826, + "balance_loss_mlp": 1.01867199, + "epoch": 0.9691576985510731, + "flos": 32635437851520.0, + "grad_norm": 1.8242358855824217, + "language_loss": 0.74343431, + "learning_rate": 9.933525399632658e-09, + "loss": 0.76525581, + "num_input_tokens_seen": 174007330, + "step": 8060, + "time_per_iteration": 3.6301722526550293 + }, + { + "auxiliary_loss_clip": 0.01145953, + "auxiliary_loss_mlp": 0.01023693, + "balance_loss_clip": 1.0444243, + "balance_loss_mlp": 1.01525331, + "epoch": 0.9692779414417123, + "flos": 35663174040960.0, + "grad_norm": 1.767331651595602, + "language_loss": 0.65042132, + "learning_rate": 9.856134985553488e-09, + "loss": 0.67211777, + "num_input_tokens_seen": 174027055, + "step": 8061, + "time_per_iteration": 2.8682243824005127 + }, + { + "auxiliary_loss_clip": 0.01167695, + "auxiliary_loss_mlp": 0.01027985, + "balance_loss_clip": 1.04632258, + "balance_loss_mlp": 1.02063, + "epoch": 0.9693981843323514, + "flos": 28366952117760.0, + "grad_norm": 1.54531897218019, + "language_loss": 0.73648632, + "learning_rate": 9.77904647471628e-09, + "loss": 0.75844312, + "num_input_tokens_seen": 174050235, + "step": 8062, + "time_per_iteration": 2.7296814918518066 + }, + { + "auxiliary_loss_clip": 0.01121012, + "auxiliary_loss_mlp": 0.01027037, + "balance_loss_clip": 1.03931177, + "balance_loss_mlp": 1.01905608, + "epoch": 0.9695184272229904, + "flos": 23622865378560.0, + "grad_norm": 1.474905392749536, + "language_loss": 0.74027216, + "learning_rate": 9.702259878815454e-09, + "loss": 0.76175267, + "num_input_tokens_seen": 174070560, + "step": 8063, + "time_per_iteration": 3.855506658554077 + }, + { + "auxiliary_loss_clip": 0.01161678, + "auxiliary_loss_mlp": 0.01035556, + "balance_loss_clip": 1.04631615, + "balance_loss_mlp": 1.0273962, + "epoch": 0.9696386701136296, + "flos": 23294677789440.0, + "grad_norm": 2.6030402383309696, + "language_loss": 0.74206662, + "learning_rate": 9.625775209499254e-09, + "loss": 0.76403898, + "num_input_tokens_seen": 174090565, + "step": 8064, + "time_per_iteration": 3.632282018661499 + }, + { + "auxiliary_loss_clip": 0.01132951, + "auxiliary_loss_mlp": 0.01033456, + "balance_loss_clip": 1.039891, + "balance_loss_mlp": 1.02573109, + "epoch": 0.9697589130042686, + "flos": 15121876360320.0, + "grad_norm": 2.2065807318774975, + "language_loss": 0.74018645, + "learning_rate": 9.549592478370172e-09, + "loss": 0.76185048, + "num_input_tokens_seen": 174108745, + "step": 8065, + "time_per_iteration": 2.773029327392578 + }, + { + "auxiliary_loss_clip": 0.01160418, + "auxiliary_loss_mlp": 0.01024304, + "balance_loss_clip": 1.04477763, + "balance_loss_mlp": 1.0169487, + "epoch": 0.9698791558949077, + "flos": 18879532824960.0, + "grad_norm": 1.6618836581662573, + "language_loss": 0.79180551, + "learning_rate": 9.473711696985632e-09, + "loss": 0.81365269, + "num_input_tokens_seen": 174128075, + "step": 8066, + "time_per_iteration": 2.7305102348327637 + }, + { + "auxiliary_loss_clip": 0.01141413, + "auxiliary_loss_mlp": 0.01023356, + "balance_loss_clip": 1.04384017, + "balance_loss_mlp": 1.01524997, + "epoch": 0.9699993987855468, + "flos": 17931455297280.0, + "grad_norm": 2.0046983795194238, + "language_loss": 0.75930655, + "learning_rate": 9.398132876856201e-09, + "loss": 0.78095424, + "num_input_tokens_seen": 174147040, + "step": 8067, + "time_per_iteration": 3.6760222911834717 + }, + { + "auxiliary_loss_clip": 0.01035772, + "auxiliary_loss_mlp": 0.01001411, + "balance_loss_clip": 1.00833225, + "balance_loss_mlp": 0.99995631, + "epoch": 0.9701196416761859, + "flos": 67182186297600.0, + "grad_norm": 0.774380417221623, + "language_loss": 0.60793924, + "learning_rate": 9.322856029447379e-09, + "loss": 0.62831116, + "num_input_tokens_seen": 174208225, + "step": 8068, + "time_per_iteration": 3.3669309616088867 + }, + { + "auxiliary_loss_clip": 0.0116512, + "auxiliary_loss_mlp": 0.010247, + "balance_loss_clip": 1.04583645, + "balance_loss_mlp": 1.01716638, + "epoch": 0.970239884566825, + "flos": 24277804012800.0, + "grad_norm": 1.9791023010384434, + "language_loss": 0.80410123, + "learning_rate": 9.247881166178695e-09, + "loss": 0.82599938, + "num_input_tokens_seen": 174226935, + "step": 8069, + "time_per_iteration": 2.783116340637207 + }, + { + "auxiliary_loss_clip": 0.01147953, + "auxiliary_loss_mlp": 0.01028991, + "balance_loss_clip": 1.04460359, + "balance_loss_mlp": 1.02112949, + "epoch": 0.970360127457464, + "flos": 25301689194240.0, + "grad_norm": 2.3416019327020012, + "language_loss": 0.76893032, + "learning_rate": 9.173208298423274e-09, + "loss": 0.79069978, + "num_input_tokens_seen": 174248140, + "step": 8070, + "time_per_iteration": 2.931835174560547 + }, + { + "auxiliary_loss_clip": 0.01118868, + "auxiliary_loss_mlp": 0.00885505, + "balance_loss_clip": 1.04148173, + "balance_loss_mlp": 1.00056672, + "epoch": 0.9704803703481032, + "flos": 29572473398400.0, + "grad_norm": 1.584987592219495, + "language_loss": 0.75955606, + "learning_rate": 9.09883743750961e-09, + "loss": 0.77959973, + "num_input_tokens_seen": 174271030, + "step": 8071, + "time_per_iteration": 2.932417631149292 + }, + { + "auxiliary_loss_clip": 0.01146593, + "auxiliary_loss_mlp": 0.01025443, + "balance_loss_clip": 1.04391432, + "balance_loss_mlp": 1.01764631, + "epoch": 0.9706006132387422, + "flos": 17380046638080.0, + "grad_norm": 2.2418301491135533, + "language_loss": 0.83528244, + "learning_rate": 9.024768594719124e-09, + "loss": 0.85700285, + "num_input_tokens_seen": 174289410, + "step": 8072, + "time_per_iteration": 2.786682367324829 + }, + { + "auxiliary_loss_clip": 0.01140324, + "auxiliary_loss_mlp": 0.0102204, + "balance_loss_clip": 1.04455781, + "balance_loss_mlp": 1.01416612, + "epoch": 0.9707208561293813, + "flos": 18186421011840.0, + "grad_norm": 2.0178434843929387, + "language_loss": 0.73076564, + "learning_rate": 8.95100178128816e-09, + "loss": 0.75238925, + "num_input_tokens_seen": 174308550, + "step": 8073, + "time_per_iteration": 2.731717586517334 + }, + { + "auxiliary_loss_clip": 0.01150661, + "auxiliary_loss_mlp": 0.01026735, + "balance_loss_clip": 1.04426098, + "balance_loss_mlp": 1.01912332, + "epoch": 0.9708410990200205, + "flos": 31248388212480.0, + "grad_norm": 2.193621340194644, + "language_loss": 0.69903582, + "learning_rate": 8.877537008407321e-09, + "loss": 0.72080976, + "num_input_tokens_seen": 174328600, + "step": 8074, + "time_per_iteration": 2.767646312713623 + }, + { + "auxiliary_loss_clip": 0.01152568, + "auxiliary_loss_mlp": 0.01027582, + "balance_loss_clip": 1.04457903, + "balance_loss_mlp": 1.01999986, + "epoch": 0.9709613419106595, + "flos": 30554450386560.0, + "grad_norm": 1.51050067916555, + "language_loss": 0.68885529, + "learning_rate": 8.804374287221028e-09, + "loss": 0.71065676, + "num_input_tokens_seen": 174349835, + "step": 8075, + "time_per_iteration": 2.8131325244903564 + }, + { + "auxiliary_loss_clip": 0.01132187, + "auxiliary_loss_mlp": 0.01027424, + "balance_loss_clip": 1.03892314, + "balance_loss_mlp": 1.02007508, + "epoch": 0.9710815848012986, + "flos": 23730166281600.0, + "grad_norm": 1.6981501425646366, + "language_loss": 0.84661686, + "learning_rate": 8.731513628827958e-09, + "loss": 0.868213, + "num_input_tokens_seen": 174369200, + "step": 8076, + "time_per_iteration": 2.804502248764038 + }, + { + "auxiliary_loss_clip": 0.01159839, + "auxiliary_loss_mlp": 0.01026401, + "balance_loss_clip": 1.04535067, + "balance_loss_mlp": 1.0191772, + "epoch": 0.9712018276919377, + "flos": 23761875012480.0, + "grad_norm": 1.855784865609696, + "language_loss": 0.82861507, + "learning_rate": 8.658955044280825e-09, + "loss": 0.85047746, + "num_input_tokens_seen": 174388125, + "step": 8077, + "time_per_iteration": 2.7177274227142334 + }, + { + "auxiliary_loss_clip": 0.01155844, + "auxiliary_loss_mlp": 0.01026466, + "balance_loss_clip": 1.04615211, + "balance_loss_mlp": 1.01891136, + "epoch": 0.9713220705825768, + "flos": 23330983461120.0, + "grad_norm": 1.5360453899105684, + "language_loss": 0.77266979, + "learning_rate": 8.586698544587268e-09, + "loss": 0.79449284, + "num_input_tokens_seen": 174409735, + "step": 8078, + "time_per_iteration": 2.6764352321624756 + }, + { + "auxiliary_loss_clip": 0.01143278, + "auxiliary_loss_mlp": 0.01029363, + "balance_loss_clip": 1.04413676, + "balance_loss_mlp": 1.02122664, + "epoch": 0.9714423134732159, + "flos": 22200946611840.0, + "grad_norm": 1.8955740261360223, + "language_loss": 0.73855036, + "learning_rate": 8.514744140707853e-09, + "loss": 0.76027679, + "num_input_tokens_seen": 174428875, + "step": 8079, + "time_per_iteration": 2.7746493816375732 + }, + { + "auxiliary_loss_clip": 0.01166191, + "auxiliary_loss_mlp": 0.01027626, + "balance_loss_clip": 1.04596376, + "balance_loss_mlp": 1.0197407, + "epoch": 0.971562556363855, + "flos": 20229917656320.0, + "grad_norm": 1.6962093893360664, + "language_loss": 0.76157337, + "learning_rate": 8.443091843558515e-09, + "loss": 0.78351152, + "num_input_tokens_seen": 174447960, + "step": 8080, + "time_per_iteration": 2.646347761154175 + }, + { + "auxiliary_loss_clip": 0.01142936, + "auxiliary_loss_mlp": 0.01029237, + "balance_loss_clip": 1.04290795, + "balance_loss_mlp": 1.02058208, + "epoch": 0.9716827992544941, + "flos": 24970197553920.0, + "grad_norm": 2.1128261691505084, + "language_loss": 0.64708555, + "learning_rate": 8.37174166400878e-09, + "loss": 0.66880727, + "num_input_tokens_seen": 174463535, + "step": 8081, + "time_per_iteration": 2.774951934814453 + }, + { + "auxiliary_loss_clip": 0.01170828, + "auxiliary_loss_mlp": 0.01025093, + "balance_loss_clip": 1.04953718, + "balance_loss_mlp": 1.01720703, + "epoch": 0.9718030421451331, + "flos": 24681476033280.0, + "grad_norm": 2.2796232477305978, + "language_loss": 0.85419708, + "learning_rate": 8.300693612881992e-09, + "loss": 0.87615633, + "num_input_tokens_seen": 174483600, + "step": 8082, + "time_per_iteration": 2.6580240726470947 + }, + { + "auxiliary_loss_clip": 0.01157308, + "auxiliary_loss_mlp": 0.00885332, + "balance_loss_clip": 1.04523993, + "balance_loss_mlp": 1.00054383, + "epoch": 0.9719232850357723, + "flos": 22090700793600.0, + "grad_norm": 1.949659214950869, + "language_loss": 0.81480801, + "learning_rate": 8.22994770095664e-09, + "loss": 0.8352344, + "num_input_tokens_seen": 174502175, + "step": 8083, + "time_per_iteration": 2.722048044204712 + }, + { + "auxiliary_loss_clip": 0.01149757, + "auxiliary_loss_mlp": 0.01032136, + "balance_loss_clip": 1.04815722, + "balance_loss_mlp": 1.02427411, + "epoch": 0.9720435279264114, + "flos": 23656908493440.0, + "grad_norm": 2.6878838558265463, + "language_loss": 0.75646186, + "learning_rate": 8.159503938964585e-09, + "loss": 0.77828085, + "num_input_tokens_seen": 174519495, + "step": 8084, + "time_per_iteration": 2.914134979248047 + }, + { + "auxiliary_loss_clip": 0.01131278, + "auxiliary_loss_mlp": 0.01027085, + "balance_loss_clip": 1.0412755, + "balance_loss_mlp": 1.01963997, + "epoch": 0.9721637708170504, + "flos": 28365910623360.0, + "grad_norm": 1.7748770546612502, + "language_loss": 0.70333123, + "learning_rate": 8.089362337592164e-09, + "loss": 0.72491491, + "num_input_tokens_seen": 174543120, + "step": 8085, + "time_per_iteration": 2.8803975582122803 + }, + { + "auxiliary_loss_clip": 0.01146187, + "auxiliary_loss_mlp": 0.01022621, + "balance_loss_clip": 1.04380989, + "balance_loss_mlp": 1.01503038, + "epoch": 0.9722840137076896, + "flos": 29130807767040.0, + "grad_norm": 1.5975119664494342, + "language_loss": 0.7219556, + "learning_rate": 8.019522907479536e-09, + "loss": 0.74364364, + "num_input_tokens_seen": 174563480, + "step": 8086, + "time_per_iteration": 3.7215652465820312 + }, + { + "auxiliary_loss_clip": 0.0116394, + "auxiliary_loss_mlp": 0.01022752, + "balance_loss_clip": 1.04716229, + "balance_loss_mlp": 1.0153966, + "epoch": 0.9724042565983286, + "flos": 19243954258560.0, + "grad_norm": 2.1589182078309843, + "language_loss": 0.77600646, + "learning_rate": 7.949985659221558e-09, + "loss": 0.79787344, + "num_input_tokens_seen": 174580745, + "step": 8087, + "time_per_iteration": 2.7053027153015137 + }, + { + "auxiliary_loss_clip": 0.01153991, + "auxiliary_loss_mlp": 0.01022862, + "balance_loss_clip": 1.046525, + "balance_loss_mlp": 1.01553977, + "epoch": 0.9725244994889677, + "flos": 23039676161280.0, + "grad_norm": 1.9966066946340866, + "language_loss": 0.79063785, + "learning_rate": 7.880750603366904e-09, + "loss": 0.81240642, + "num_input_tokens_seen": 174599615, + "step": 8088, + "time_per_iteration": 2.691256523132324 + }, + { + "auxiliary_loss_clip": 0.01150058, + "auxiliary_loss_mlp": 0.01027102, + "balance_loss_clip": 1.04309821, + "balance_loss_mlp": 1.01899576, + "epoch": 0.9726447423796069, + "flos": 23367468700800.0, + "grad_norm": 4.21001458378461, + "language_loss": 0.7941916, + "learning_rate": 7.811817750418282e-09, + "loss": 0.81596321, + "num_input_tokens_seen": 174618375, + "step": 8089, + "time_per_iteration": 4.697148323059082 + }, + { + "auxiliary_loss_clip": 0.01140026, + "auxiliary_loss_mlp": 0.01027167, + "balance_loss_clip": 1.04547536, + "balance_loss_mlp": 1.0187211, + "epoch": 0.9727649852702459, + "flos": 26541648639360.0, + "grad_norm": 1.8797549390821127, + "language_loss": 0.80087864, + "learning_rate": 7.743187110833105e-09, + "loss": 0.82255059, + "num_input_tokens_seen": 174641135, + "step": 8090, + "time_per_iteration": 2.876478433609009 + }, + { + "auxiliary_loss_clip": 0.01150013, + "auxiliary_loss_mlp": 0.01019693, + "balance_loss_clip": 1.04152238, + "balance_loss_mlp": 1.01214075, + "epoch": 0.972885228160885, + "flos": 20522338277760.0, + "grad_norm": 1.4482325846530408, + "language_loss": 0.80569398, + "learning_rate": 7.674858695022602e-09, + "loss": 0.82739103, + "num_input_tokens_seen": 174659490, + "step": 8091, + "time_per_iteration": 2.7199766635894775 + }, + { + "auxiliary_loss_clip": 0.01171834, + "auxiliary_loss_mlp": 0.01023968, + "balance_loss_clip": 1.04936576, + "balance_loss_mlp": 1.01565266, + "epoch": 0.9730054710515241, + "flos": 17566064196480.0, + "grad_norm": 3.8197291561046844, + "language_loss": 0.76088595, + "learning_rate": 7.606832513351591e-09, + "loss": 0.78284401, + "num_input_tokens_seen": 174677440, + "step": 8092, + "time_per_iteration": 2.6769704818725586 + }, + { + "auxiliary_loss_clip": 0.01054986, + "auxiliary_loss_mlp": 0.00876789, + "balance_loss_clip": 1.0044663, + "balance_loss_mlp": 1.00127971, + "epoch": 0.9731257139421632, + "flos": 68972010117120.0, + "grad_norm": 0.8317456401591986, + "language_loss": 0.63944429, + "learning_rate": 7.539108576140264e-09, + "loss": 0.65876204, + "num_input_tokens_seen": 174741550, + "step": 8093, + "time_per_iteration": 4.310396909713745 + }, + { + "auxiliary_loss_clip": 0.01128362, + "auxiliary_loss_mlp": 0.01024529, + "balance_loss_clip": 1.04070091, + "balance_loss_mlp": 1.01716232, + "epoch": 0.9732459568328022, + "flos": 18478841633280.0, + "grad_norm": 1.9362477228420045, + "language_loss": 0.70450735, + "learning_rate": 7.471686893661732e-09, + "loss": 0.72603625, + "num_input_tokens_seen": 174759845, + "step": 8094, + "time_per_iteration": 2.8757665157318115 + }, + { + "auxiliary_loss_clip": 0.01146106, + "auxiliary_loss_mlp": 0.01027853, + "balance_loss_clip": 1.04386878, + "balance_loss_mlp": 1.01989579, + "epoch": 0.9733661997234414, + "flos": 20883886623360.0, + "grad_norm": 1.8141775625963696, + "language_loss": 0.6442951, + "learning_rate": 7.4045674761442636e-09, + "loss": 0.6660347, + "num_input_tokens_seen": 174777175, + "step": 8095, + "time_per_iteration": 2.9158174991607666 + }, + { + "auxiliary_loss_clip": 0.01166694, + "auxiliary_loss_mlp": 0.00885604, + "balance_loss_clip": 1.04681742, + "balance_loss_mlp": 1.00058115, + "epoch": 0.9734864426140805, + "flos": 23766795175680.0, + "grad_norm": 1.7073450587546442, + "language_loss": 0.7429989, + "learning_rate": 7.337750333769488e-09, + "loss": 0.76352191, + "num_input_tokens_seen": 174796980, + "step": 8096, + "time_per_iteration": 2.703904151916504 + }, + { + "auxiliary_loss_clip": 0.01152416, + "auxiliary_loss_mlp": 0.01024451, + "balance_loss_clip": 1.04049754, + "balance_loss_mlp": 1.01616621, + "epoch": 0.9736066855047195, + "flos": 35042422176000.0, + "grad_norm": 1.6390523842592633, + "language_loss": 0.72628534, + "learning_rate": 7.2712354766737425e-09, + "loss": 0.74805397, + "num_input_tokens_seen": 174817310, + "step": 8097, + "time_per_iteration": 2.9025421142578125 + }, + { + "auxiliary_loss_clip": 0.01131171, + "auxiliary_loss_mlp": 0.01029065, + "balance_loss_clip": 1.04464507, + "balance_loss_mlp": 1.02114916, + "epoch": 0.9737269283953586, + "flos": 20410620001920.0, + "grad_norm": 1.43415320527125, + "language_loss": 0.80856192, + "learning_rate": 7.2050229149469565e-09, + "loss": 0.83016431, + "num_input_tokens_seen": 174837320, + "step": 8098, + "time_per_iteration": 2.804335594177246 + }, + { + "auxiliary_loss_clip": 0.01140757, + "auxiliary_loss_mlp": 0.01030449, + "balance_loss_clip": 1.03979433, + "balance_loss_mlp": 1.0225513, + "epoch": 0.9738471712859977, + "flos": 28911680847360.0, + "grad_norm": 1.716500874223675, + "language_loss": 0.63600314, + "learning_rate": 7.139112658633984e-09, + "loss": 0.6577152, + "num_input_tokens_seen": 174857470, + "step": 8099, + "time_per_iteration": 2.8636550903320312 + }, + { + "auxiliary_loss_clip": 0.01139445, + "auxiliary_loss_mlp": 0.01027055, + "balance_loss_clip": 1.04558814, + "balance_loss_mlp": 1.0190258, + "epoch": 0.9739674141766368, + "flos": 27782326356480.0, + "grad_norm": 3.695686028881741, + "language_loss": 0.70420104, + "learning_rate": 7.073504717733048e-09, + "loss": 0.72586608, + "num_input_tokens_seen": 174877035, + "step": 8100, + "time_per_iteration": 2.852544069290161 + }, + { + "auxiliary_loss_clip": 0.01027115, + "auxiliary_loss_mlp": 0.01002542, + "balance_loss_clip": 1.00713825, + "balance_loss_mlp": 1.00108802, + "epoch": 0.9740876570672758, + "flos": 68863057188480.0, + "grad_norm": 0.7362675484835126, + "language_loss": 0.57174563, + "learning_rate": 7.008199102196855e-09, + "loss": 0.59204221, + "num_input_tokens_seen": 174938460, + "step": 8101, + "time_per_iteration": 3.3037960529327393 + }, + { + "auxiliary_loss_clip": 0.01041167, + "auxiliary_loss_mlp": 0.01000747, + "balance_loss_clip": 1.00908315, + "balance_loss_mlp": 0.99941808, + "epoch": 0.974207899957915, + "flos": 58236622646400.0, + "grad_norm": 0.797930058358525, + "language_loss": 0.58981073, + "learning_rate": 6.9431958219321464e-09, + "loss": 0.61022997, + "num_input_tokens_seen": 174994625, + "step": 8102, + "time_per_iteration": 3.2539732456207275 + }, + { + "auxiliary_loss_clip": 0.01146284, + "auxiliary_loss_mlp": 0.01026825, + "balance_loss_clip": 1.04177749, + "balance_loss_mlp": 1.01907611, + "epoch": 0.9743281428485541, + "flos": 22600057605120.0, + "grad_norm": 1.5289837688961876, + "language_loss": 0.77822471, + "learning_rate": 6.878494886800146e-09, + "loss": 0.79995573, + "num_input_tokens_seen": 175015400, + "step": 8103, + "time_per_iteration": 2.8022778034210205 + }, + { + "auxiliary_loss_clip": 0.01143545, + "auxiliary_loss_mlp": 0.01029747, + "balance_loss_clip": 1.04476666, + "balance_loss_mlp": 1.02162862, + "epoch": 0.9744483857391931, + "flos": 20008815488640.0, + "grad_norm": 1.9345963336246785, + "language_loss": 0.76183558, + "learning_rate": 6.814096306615669e-09, + "loss": 0.7835685, + "num_input_tokens_seen": 175033540, + "step": 8104, + "time_per_iteration": 2.7434821128845215 + }, + { + "auxiliary_loss_clip": 0.01155618, + "auxiliary_loss_mlp": 0.01032466, + "balance_loss_clip": 1.04339194, + "balance_loss_mlp": 1.024652, + "epoch": 0.9745686286298323, + "flos": 17675268520320.0, + "grad_norm": 2.0448777471875577, + "language_loss": 0.65009487, + "learning_rate": 6.750000091148011e-09, + "loss": 0.67197573, + "num_input_tokens_seen": 175050835, + "step": 8105, + "time_per_iteration": 2.729536294937134 + }, + { + "auxiliary_loss_clip": 0.01170974, + "auxiliary_loss_mlp": 0.01028462, + "balance_loss_clip": 1.05017221, + "balance_loss_mlp": 1.02078176, + "epoch": 0.9746888715204713, + "flos": 29460252332160.0, + "grad_norm": 2.103367590957574, + "language_loss": 0.72408795, + "learning_rate": 6.686206250120729e-09, + "loss": 0.74608237, + "num_input_tokens_seen": 175072330, + "step": 8106, + "time_per_iteration": 2.693439483642578 + }, + { + "auxiliary_loss_clip": 0.01143467, + "auxiliary_loss_mlp": 0.0102376, + "balance_loss_clip": 1.04032028, + "balance_loss_mlp": 1.01593351, + "epoch": 0.9748091144111104, + "flos": 18479308510080.0, + "grad_norm": 1.7315041402216398, + "language_loss": 0.74626267, + "learning_rate": 6.622714793210749e-09, + "loss": 0.76793492, + "num_input_tokens_seen": 175091250, + "step": 8107, + "time_per_iteration": 2.844813823699951 + }, + { + "auxiliary_loss_clip": 0.01167637, + "auxiliary_loss_mlp": 0.01025352, + "balance_loss_clip": 1.04584789, + "balance_loss_mlp": 1.01783562, + "epoch": 0.9749293573017496, + "flos": 20665154753280.0, + "grad_norm": 1.6433295241274846, + "language_loss": 0.78628957, + "learning_rate": 6.559525730050364e-09, + "loss": 0.80821943, + "num_input_tokens_seen": 175111350, + "step": 8108, + "time_per_iteration": 2.586888551712036 + }, + { + "auxiliary_loss_clip": 0.01143785, + "auxiliary_loss_mlp": 0.01022675, + "balance_loss_clip": 1.04470205, + "balance_loss_mlp": 1.01517642, + "epoch": 0.9750496001923886, + "flos": 18478590238080.0, + "grad_norm": 2.1701315039028253, + "language_loss": 0.76197523, + "learning_rate": 6.496639070224574e-09, + "loss": 0.78363979, + "num_input_tokens_seen": 175129835, + "step": 8109, + "time_per_iteration": 2.7718307971954346 + }, + { + "auxiliary_loss_clip": 0.0116082, + "auxiliary_loss_mlp": 0.01026329, + "balance_loss_clip": 1.04573488, + "balance_loss_mlp": 1.01898599, + "epoch": 0.9751698430830277, + "flos": 19572967860480.0, + "grad_norm": 2.35592068361339, + "language_loss": 0.83678842, + "learning_rate": 6.4340548232739714e-09, + "loss": 0.85865992, + "num_input_tokens_seen": 175146035, + "step": 8110, + "time_per_iteration": 2.7201554775238037 + }, + { + "auxiliary_loss_clip": 0.01144207, + "auxiliary_loss_mlp": 0.01030047, + "balance_loss_clip": 1.04254246, + "balance_loss_mlp": 1.02221489, + "epoch": 0.9752900859736668, + "flos": 23550325862400.0, + "grad_norm": 1.6155971842864327, + "language_loss": 0.79098892, + "learning_rate": 6.371772998692071e-09, + "loss": 0.81273144, + "num_input_tokens_seen": 175165290, + "step": 8111, + "time_per_iteration": 2.8373475074768066 + }, + { + "auxiliary_loss_clip": 0.01141624, + "auxiliary_loss_mlp": 0.01026388, + "balance_loss_clip": 1.0398792, + "balance_loss_mlp": 1.01908088, + "epoch": 0.9754103288643059, + "flos": 20303211358080.0, + "grad_norm": 2.7718421979593164, + "language_loss": 0.64924777, + "learning_rate": 6.309793605927094e-09, + "loss": 0.67092788, + "num_input_tokens_seen": 175183610, + "step": 8112, + "time_per_iteration": 3.7710540294647217 + }, + { + "auxiliary_loss_clip": 0.01147109, + "auxiliary_loss_mlp": 0.01026253, + "balance_loss_clip": 1.04404259, + "balance_loss_mlp": 1.0187037, + "epoch": 0.975530571754945, + "flos": 19350680544000.0, + "grad_norm": 1.9883839038903246, + "language_loss": 0.80044532, + "learning_rate": 6.248116654381297e-09, + "loss": 0.82217896, + "num_input_tokens_seen": 175202080, + "step": 8113, + "time_per_iteration": 2.7622122764587402 + }, + { + "auxiliary_loss_clip": 0.0115213, + "auxiliary_loss_mlp": 0.01023923, + "balance_loss_clip": 1.04147983, + "balance_loss_mlp": 1.01662171, + "epoch": 0.9756508146455841, + "flos": 23583399310080.0, + "grad_norm": 2.0185475615472197, + "language_loss": 0.72381157, + "learning_rate": 6.186742153410751e-09, + "loss": 0.74557215, + "num_input_tokens_seen": 175221575, + "step": 8114, + "time_per_iteration": 2.8655340671539307 + }, + { + "auxiliary_loss_clip": 0.01146734, + "auxiliary_loss_mlp": 0.01033865, + "balance_loss_clip": 1.04334235, + "balance_loss_mlp": 1.02529955, + "epoch": 0.9757710575362232, + "flos": 22966921163520.0, + "grad_norm": 2.674758389196144, + "language_loss": 0.87477875, + "learning_rate": 6.125670112326453e-09, + "loss": 0.89658463, + "num_input_tokens_seen": 175240835, + "step": 8115, + "time_per_iteration": 4.650297403335571 + }, + { + "auxiliary_loss_clip": 0.01159082, + "auxiliary_loss_mlp": 0.01027256, + "balance_loss_clip": 1.0438273, + "balance_loss_mlp": 1.01958442, + "epoch": 0.9758913004268622, + "flos": 27966009530880.0, + "grad_norm": 1.9864382392304392, + "language_loss": 0.70211273, + "learning_rate": 6.064900540392548e-09, + "loss": 0.72397614, + "num_input_tokens_seen": 175262930, + "step": 8116, + "time_per_iteration": 2.806351900100708 + }, + { + "auxiliary_loss_clip": 0.011416, + "auxiliary_loss_mlp": 0.01030344, + "balance_loss_clip": 1.0435847, + "balance_loss_mlp": 1.02284884, + "epoch": 0.9760115433175014, + "flos": 22200156512640.0, + "grad_norm": 1.913639621686685, + "language_loss": 0.79105902, + "learning_rate": 6.0044334468278835e-09, + "loss": 0.81277847, + "num_input_tokens_seen": 175282275, + "step": 8117, + "time_per_iteration": 2.7639267444610596 + }, + { + "auxiliary_loss_clip": 0.01132483, + "auxiliary_loss_mlp": 0.01028433, + "balance_loss_clip": 1.04186034, + "balance_loss_mlp": 1.02030277, + "epoch": 0.9761317862081405, + "flos": 26250736389120.0, + "grad_norm": 1.948161859194908, + "language_loss": 0.71572942, + "learning_rate": 5.944268840805345e-09, + "loss": 0.73733854, + "num_input_tokens_seen": 175303020, + "step": 8118, + "time_per_iteration": 2.848384141921997 + }, + { + "auxiliary_loss_clip": 0.01136459, + "auxiliary_loss_mlp": 0.0102509, + "balance_loss_clip": 1.04127598, + "balance_loss_mlp": 1.01730025, + "epoch": 0.9762520290987795, + "flos": 26575440359040.0, + "grad_norm": 2.556513985770166, + "language_loss": 0.64184749, + "learning_rate": 5.88440673145163e-09, + "loss": 0.663463, + "num_input_tokens_seen": 175324070, + "step": 8119, + "time_per_iteration": 3.702033042907715 + }, + { + "auxiliary_loss_clip": 0.01157847, + "auxiliary_loss_mlp": 0.01022865, + "balance_loss_clip": 1.04646361, + "balance_loss_mlp": 1.01509905, + "epoch": 0.9763722719894187, + "flos": 18005036307840.0, + "grad_norm": 2.4688579015565155, + "language_loss": 0.82839036, + "learning_rate": 5.824847127848142e-09, + "loss": 0.85019755, + "num_input_tokens_seen": 175342595, + "step": 8120, + "time_per_iteration": 2.665916681289673 + }, + { + "auxiliary_loss_clip": 0.01135183, + "auxiliary_loss_mlp": 0.01025091, + "balance_loss_clip": 1.04408681, + "balance_loss_mlp": 1.01730084, + "epoch": 0.9764925148800577, + "flos": 22455660931200.0, + "grad_norm": 1.9184599951455485, + "language_loss": 0.78790736, + "learning_rate": 5.765590039029433e-09, + "loss": 0.80951011, + "num_input_tokens_seen": 175361915, + "step": 8121, + "time_per_iteration": 2.8018391132354736 + }, + { + "auxiliary_loss_clip": 0.01168143, + "auxiliary_loss_mlp": 0.0102331, + "balance_loss_clip": 1.048383, + "balance_loss_mlp": 1.01620257, + "epoch": 0.9766127577706968, + "flos": 36757084786560.0, + "grad_norm": 1.6756528435502969, + "language_loss": 0.71005422, + "learning_rate": 5.706635473985422e-09, + "loss": 0.73196876, + "num_input_tokens_seen": 175385785, + "step": 8122, + "time_per_iteration": 2.7818121910095215 + }, + { + "auxiliary_loss_clip": 0.01157568, + "auxiliary_loss_mlp": 0.01029895, + "balance_loss_clip": 1.04537487, + "balance_loss_mlp": 1.02184582, + "epoch": 0.976733000661336, + "flos": 22309971367680.0, + "grad_norm": 1.6911722712980661, + "language_loss": 0.8522051, + "learning_rate": 5.6479834416591764e-09, + "loss": 0.8740797, + "num_input_tokens_seen": 175405145, + "step": 8123, + "time_per_iteration": 2.6270058155059814 + }, + { + "auxiliary_loss_clip": 0.01157847, + "auxiliary_loss_mlp": 0.0088623, + "balance_loss_clip": 1.04578567, + "balance_loss_mlp": 1.00057769, + "epoch": 0.976853243551975, + "flos": 25810938264960.0, + "grad_norm": 1.8823209621896861, + "language_loss": 0.68896484, + "learning_rate": 5.589633950947803e-09, + "loss": 0.70940566, + "num_input_tokens_seen": 175422645, + "step": 8124, + "time_per_iteration": 2.7205650806427 + }, + { + "auxiliary_loss_clip": 0.0114934, + "auxiliary_loss_mlp": 0.01032851, + "balance_loss_clip": 1.04517817, + "balance_loss_mlp": 1.02436972, + "epoch": 0.9769734864426141, + "flos": 21397445326080.0, + "grad_norm": 2.0727004208343165, + "language_loss": 0.69462848, + "learning_rate": 5.5315870107035535e-09, + "loss": 0.71645045, + "num_input_tokens_seen": 175440695, + "step": 8125, + "time_per_iteration": 2.7807512283325195 + }, + { + "auxiliary_loss_clip": 0.01144795, + "auxiliary_loss_mlp": 0.01023245, + "balance_loss_clip": 1.04665327, + "balance_loss_mlp": 1.01580906, + "epoch": 0.9770937293332532, + "flos": 13990977584640.0, + "grad_norm": 1.8711568171937176, + "language_loss": 0.7878443, + "learning_rate": 5.473842629731607e-09, + "loss": 0.80952466, + "num_input_tokens_seen": 175459195, + "step": 8126, + "time_per_iteration": 2.7535693645477295 + }, + { + "auxiliary_loss_clip": 0.01156791, + "auxiliary_loss_mlp": 0.00886163, + "balance_loss_clip": 1.04247952, + "balance_loss_mlp": 1.00065756, + "epoch": 0.9772139722238923, + "flos": 17931994001280.0, + "grad_norm": 3.665056277800547, + "language_loss": 0.77935481, + "learning_rate": 5.416400816792066e-09, + "loss": 0.7997843, + "num_input_tokens_seen": 175476710, + "step": 8127, + "time_per_iteration": 2.7483556270599365 + }, + { + "auxiliary_loss_clip": 0.01167088, + "auxiliary_loss_mlp": 0.01021003, + "balance_loss_clip": 1.04617929, + "balance_loss_mlp": 1.01324213, + "epoch": 0.9773342151145313, + "flos": 20446171488000.0, + "grad_norm": 2.32752188631873, + "language_loss": 0.78517121, + "learning_rate": 5.359261580598407e-09, + "loss": 0.80705214, + "num_input_tokens_seen": 175492550, + "step": 8128, + "time_per_iteration": 2.5991508960723877 + }, + { + "auxiliary_loss_clip": 0.01162385, + "auxiliary_loss_mlp": 0.01027515, + "balance_loss_clip": 1.04678285, + "balance_loss_mlp": 1.01962948, + "epoch": 0.9774544580051704, + "flos": 11837306949120.0, + "grad_norm": 2.62633927753931, + "language_loss": 0.7835052, + "learning_rate": 5.302424929819027e-09, + "loss": 0.80540419, + "num_input_tokens_seen": 175506560, + "step": 8129, + "time_per_iteration": 2.8068957328796387 + }, + { + "auxiliary_loss_clip": 0.01159861, + "auxiliary_loss_mlp": 0.01023862, + "balance_loss_clip": 1.04208899, + "balance_loss_mlp": 1.0162921, + "epoch": 0.9775747008958096, + "flos": 13479932833920.0, + "grad_norm": 2.5298286569239643, + "language_loss": 0.73342866, + "learning_rate": 5.24589087307592e-09, + "loss": 0.75526583, + "num_input_tokens_seen": 175524180, + "step": 8130, + "time_per_iteration": 2.7547049522399902 + }, + { + "auxiliary_loss_clip": 0.0116845, + "auxiliary_loss_mlp": 0.01023445, + "balance_loss_clip": 1.0451268, + "balance_loss_mlp": 1.01556563, + "epoch": 0.9776949437864486, + "flos": 59532314042880.0, + "grad_norm": 1.4123128373905975, + "language_loss": 0.64977539, + "learning_rate": 5.189659418944891e-09, + "loss": 0.67169428, + "num_input_tokens_seen": 175554355, + "step": 8131, + "time_per_iteration": 3.070455312728882 + }, + { + "auxiliary_loss_clip": 0.01165614, + "auxiliary_loss_mlp": 0.0103591, + "balance_loss_clip": 1.04507303, + "balance_loss_mlp": 1.02822971, + "epoch": 0.9778151866770877, + "flos": 21178605715200.0, + "grad_norm": 1.7615893855358256, + "language_loss": 0.78235227, + "learning_rate": 5.133730575956674e-09, + "loss": 0.80436748, + "num_input_tokens_seen": 175574025, + "step": 8132, + "time_per_iteration": 2.69136118888855 + }, + { + "auxiliary_loss_clip": 0.01152557, + "auxiliary_loss_mlp": 0.0102389, + "balance_loss_clip": 1.04465187, + "balance_loss_mlp": 1.01637352, + "epoch": 0.9779354295677268, + "flos": 20886795624960.0, + "grad_norm": 1.9809928660563016, + "language_loss": 0.71741855, + "learning_rate": 5.0781043525953696e-09, + "loss": 0.73918307, + "num_input_tokens_seen": 175592090, + "step": 8133, + "time_per_iteration": 2.7600910663604736 + }, + { + "auxiliary_loss_clip": 0.01143932, + "auxiliary_loss_mlp": 0.01028494, + "balance_loss_clip": 1.04648614, + "balance_loss_mlp": 1.0207603, + "epoch": 0.9780556724583659, + "flos": 23440618748160.0, + "grad_norm": 1.7849093185538327, + "language_loss": 0.73877686, + "learning_rate": 5.0227807572995605e-09, + "loss": 0.76050115, + "num_input_tokens_seen": 175614065, + "step": 8134, + "time_per_iteration": 2.7855887413024902 + }, + { + "auxiliary_loss_clip": 0.01151916, + "auxiliary_loss_mlp": 0.01024538, + "balance_loss_clip": 1.04350495, + "balance_loss_mlp": 1.01686144, + "epoch": 0.9781759153490049, + "flos": 20923244951040.0, + "grad_norm": 2.090951207476109, + "language_loss": 0.67303443, + "learning_rate": 4.967759798461646e-09, + "loss": 0.69479895, + "num_input_tokens_seen": 175632410, + "step": 8135, + "time_per_iteration": 2.812638759613037 + }, + { + "auxiliary_loss_clip": 0.01167076, + "auxiliary_loss_mlp": 0.0102607, + "balance_loss_clip": 1.04761922, + "balance_loss_mlp": 1.01862586, + "epoch": 0.9782961582396441, + "flos": 28293191539200.0, + "grad_norm": 2.027011937630224, + "language_loss": 0.75126338, + "learning_rate": 4.913041484428282e-09, + "loss": 0.77319491, + "num_input_tokens_seen": 175652885, + "step": 8136, + "time_per_iteration": 2.756645441055298 + }, + { + "auxiliary_loss_clip": 0.01158145, + "auxiliary_loss_mlp": 0.01020266, + "balance_loss_clip": 1.04396939, + "balance_loss_mlp": 1.013358, + "epoch": 0.9784164011302832, + "flos": 25552955808000.0, + "grad_norm": 1.8597792244478157, + "language_loss": 0.74067044, + "learning_rate": 4.858625823500384e-09, + "loss": 0.76245463, + "num_input_tokens_seen": 175670585, + "step": 8137, + "time_per_iteration": 2.7442984580993652 + }, + { + "auxiliary_loss_clip": 0.01162846, + "auxiliary_loss_mlp": 0.01025379, + "balance_loss_clip": 1.04506588, + "balance_loss_mlp": 1.01776111, + "epoch": 0.9785366440209222, + "flos": 29965945956480.0, + "grad_norm": 1.952723748081468, + "language_loss": 0.73611403, + "learning_rate": 4.80451282393246e-09, + "loss": 0.7579962, + "num_input_tokens_seen": 175690570, + "step": 8138, + "time_per_iteration": 2.8595499992370605 + }, + { + "auxiliary_loss_clip": 0.01145308, + "auxiliary_loss_mlp": 0.0102858, + "balance_loss_clip": 1.04563785, + "balance_loss_mlp": 1.02095628, + "epoch": 0.9786568869115614, + "flos": 32343591847680.0, + "grad_norm": 2.2791525776976296, + "language_loss": 0.67507118, + "learning_rate": 4.750702493933722e-09, + "loss": 0.69681007, + "num_input_tokens_seen": 175710455, + "step": 8139, + "time_per_iteration": 3.793450117111206 + }, + { + "auxiliary_loss_clip": 0.01148094, + "auxiliary_loss_mlp": 0.00885843, + "balance_loss_clip": 1.04623652, + "balance_loss_mlp": 1.00064278, + "epoch": 0.9787771298022004, + "flos": 23331414424320.0, + "grad_norm": 1.8283242158126385, + "language_loss": 0.85188818, + "learning_rate": 4.697194841666974e-09, + "loss": 0.87222755, + "num_input_tokens_seen": 175729380, + "step": 8140, + "time_per_iteration": 2.781062126159668 + }, + { + "auxiliary_loss_clip": 0.01159569, + "auxiliary_loss_mlp": 0.01030362, + "balance_loss_clip": 1.04460895, + "balance_loss_mlp": 1.02227044, + "epoch": 0.9788973726928395, + "flos": 21468548298240.0, + "grad_norm": 1.6892664780341906, + "language_loss": 0.82008034, + "learning_rate": 4.6439898752492764e-09, + "loss": 0.84197962, + "num_input_tokens_seen": 175749520, + "step": 8141, + "time_per_iteration": 4.677395582199097 + }, + { + "auxiliary_loss_clip": 0.01050977, + "auxiliary_loss_mlp": 0.00876858, + "balance_loss_clip": 1.00562358, + "balance_loss_mlp": 1.00120687, + "epoch": 0.9790176155834787, + "flos": 68897459439360.0, + "grad_norm": 0.7466468125767076, + "language_loss": 0.6368885, + "learning_rate": 4.591087602751731e-09, + "loss": 0.65616691, + "num_input_tokens_seen": 175811380, + "step": 8142, + "time_per_iteration": 3.388230800628662 + }, + { + "auxiliary_loss_clip": 0.01157245, + "auxiliary_loss_mlp": 0.01028304, + "balance_loss_clip": 1.04569423, + "balance_loss_mlp": 1.02043033, + "epoch": 0.9791378584741177, + "flos": 21430877909760.0, + "grad_norm": 1.8469554176334786, + "language_loss": 0.72104657, + "learning_rate": 4.538488032199916e-09, + "loss": 0.74290198, + "num_input_tokens_seen": 175829480, + "step": 8143, + "time_per_iteration": 2.6769256591796875 + }, + { + "auxiliary_loss_clip": 0.01164012, + "auxiliary_loss_mlp": 0.01023048, + "balance_loss_clip": 1.04582977, + "balance_loss_mlp": 1.01540685, + "epoch": 0.9792581013647568, + "flos": 20153032594560.0, + "grad_norm": 1.9509922555002024, + "language_loss": 0.69068915, + "learning_rate": 4.486191171572784e-09, + "loss": 0.7125597, + "num_input_tokens_seen": 175846750, + "step": 8144, + "time_per_iteration": 2.676146984100342 + }, + { + "auxiliary_loss_clip": 0.01163516, + "auxiliary_loss_mlp": 0.01026285, + "balance_loss_clip": 1.04719782, + "balance_loss_mlp": 1.01891172, + "epoch": 0.9793783442553959, + "flos": 23728191033600.0, + "grad_norm": 1.5967174028958306, + "language_loss": 0.77599478, + "learning_rate": 4.434197028803766e-09, + "loss": 0.79789281, + "num_input_tokens_seen": 175865975, + "step": 8145, + "time_per_iteration": 3.6401824951171875 + }, + { + "auxiliary_loss_clip": 0.01150352, + "auxiliary_loss_mlp": 0.01026027, + "balance_loss_clip": 1.046188, + "balance_loss_mlp": 1.01849902, + "epoch": 0.979498587146035, + "flos": 23038742407680.0, + "grad_norm": 2.3076698465579275, + "language_loss": 0.8205663, + "learning_rate": 4.3825056117805514e-09, + "loss": 0.8423301, + "num_input_tokens_seen": 175881860, + "step": 8146, + "time_per_iteration": 2.9036455154418945 + }, + { + "auxiliary_loss_clip": 0.01168954, + "auxiliary_loss_mlp": 0.01020464, + "balance_loss_clip": 1.04564118, + "balance_loss_mlp": 1.01328516, + "epoch": 0.979618830036674, + "flos": 14318841951360.0, + "grad_norm": 2.4118775289581382, + "language_loss": 0.79412031, + "learning_rate": 4.331116928344425e-09, + "loss": 0.81601453, + "num_input_tokens_seen": 175898175, + "step": 8147, + "time_per_iteration": 2.646634817123413 + }, + { + "auxiliary_loss_clip": 0.0115746, + "auxiliary_loss_mlp": 0.00885941, + "balance_loss_clip": 1.04548943, + "balance_loss_mlp": 1.00054169, + "epoch": 0.9797390729273132, + "flos": 16727514215040.0, + "grad_norm": 1.9942201999125748, + "language_loss": 0.62673891, + "learning_rate": 4.28003098629115e-09, + "loss": 0.64717293, + "num_input_tokens_seen": 175914310, + "step": 8148, + "time_per_iteration": 2.846686363220215 + }, + { + "auxiliary_loss_clip": 0.01136861, + "auxiliary_loss_mlp": 0.01025333, + "balance_loss_clip": 1.03740752, + "balance_loss_mlp": 1.01737559, + "epoch": 0.9798593158179523, + "flos": 24532661986560.0, + "grad_norm": 1.7949068277182814, + "language_loss": 0.78571349, + "learning_rate": 4.229247793370305e-09, + "loss": 0.80733544, + "num_input_tokens_seen": 175933435, + "step": 8149, + "time_per_iteration": 2.870833396911621 + }, + { + "auxiliary_loss_clip": 0.0117002, + "auxiliary_loss_mlp": 0.01028525, + "balance_loss_clip": 1.04791975, + "balance_loss_mlp": 1.02081823, + "epoch": 0.9799795587085913, + "flos": 27308808339840.0, + "grad_norm": 3.522176097154761, + "language_loss": 0.70562232, + "learning_rate": 4.178767357285951e-09, + "loss": 0.72760773, + "num_input_tokens_seen": 175955065, + "step": 8150, + "time_per_iteration": 2.6944947242736816 + }, + { + "auxiliary_loss_clip": 0.0116246, + "auxiliary_loss_mlp": 0.00885616, + "balance_loss_clip": 1.04858041, + "balance_loss_mlp": 1.00063527, + "epoch": 0.9800998015992305, + "flos": 26286575184000.0, + "grad_norm": 1.9608456009189248, + "language_loss": 0.71434021, + "learning_rate": 4.128589685695516e-09, + "loss": 0.73482102, + "num_input_tokens_seen": 175975490, + "step": 8151, + "time_per_iteration": 2.850271224975586 + }, + { + "auxiliary_loss_clip": 0.01168698, + "auxiliary_loss_mlp": 0.01026961, + "balance_loss_clip": 1.04713917, + "balance_loss_mlp": 1.01920033, + "epoch": 0.9802200444898695, + "flos": 16723635546240.0, + "grad_norm": 1.908971943408864, + "language_loss": 0.84244126, + "learning_rate": 4.078714786211135e-09, + "loss": 0.86439788, + "num_input_tokens_seen": 175991340, + "step": 8152, + "time_per_iteration": 2.6660735607147217 + }, + { + "auxiliary_loss_clip": 0.0115664, + "auxiliary_loss_mlp": 0.01023909, + "balance_loss_clip": 1.0456773, + "balance_loss_mlp": 1.01633954, + "epoch": 0.9803402873805086, + "flos": 24900459298560.0, + "grad_norm": 2.050185675795304, + "language_loss": 0.76354945, + "learning_rate": 4.029142666398977e-09, + "loss": 0.78535491, + "num_input_tokens_seen": 176011505, + "step": 8153, + "time_per_iteration": 2.7479355335235596 + }, + { + "auxiliary_loss_clip": 0.01167483, + "auxiliary_loss_mlp": 0.01023456, + "balance_loss_clip": 1.04793143, + "balance_loss_mlp": 1.01573145, + "epoch": 0.9804605302711478, + "flos": 22564937082240.0, + "grad_norm": 1.7919849803511385, + "language_loss": 0.79882473, + "learning_rate": 3.979873333778805e-09, + "loss": 0.82073414, + "num_input_tokens_seen": 176029680, + "step": 8154, + "time_per_iteration": 2.660216808319092 + }, + { + "auxiliary_loss_clip": 0.01154617, + "auxiliary_loss_mlp": 0.0102306, + "balance_loss_clip": 1.04595757, + "balance_loss_mlp": 1.01553226, + "epoch": 0.9805807731617868, + "flos": 38905368382080.0, + "grad_norm": 1.8467650651167082, + "language_loss": 0.74160522, + "learning_rate": 3.930906795824862e-09, + "loss": 0.76338202, + "num_input_tokens_seen": 176050355, + "step": 8155, + "time_per_iteration": 2.957648515701294 + }, + { + "auxiliary_loss_clip": 0.01154232, + "auxiliary_loss_mlp": 0.01027596, + "balance_loss_clip": 1.04321957, + "balance_loss_mlp": 1.0201931, + "epoch": 0.9807010160524259, + "flos": 17821999578240.0, + "grad_norm": 4.019567167414915, + "language_loss": 0.76828039, + "learning_rate": 3.882243059965207e-09, + "loss": 0.79009867, + "num_input_tokens_seen": 176068070, + "step": 8156, + "time_per_iteration": 2.7539608478546143 + }, + { + "auxiliary_loss_clip": 0.01150559, + "auxiliary_loss_mlp": 0.01024697, + "balance_loss_clip": 1.04161406, + "balance_loss_mlp": 1.01658487, + "epoch": 0.980821258943065, + "flos": 13552975140480.0, + "grad_norm": 5.503805677993342, + "language_loss": 0.65616179, + "learning_rate": 3.833882133582156e-09, + "loss": 0.67791438, + "num_input_tokens_seen": 176083730, + "step": 8157, + "time_per_iteration": 2.631713628768921 + }, + { + "auxiliary_loss_clip": 0.01160361, + "auxiliary_loss_mlp": 0.010255, + "balance_loss_clip": 1.04470098, + "balance_loss_mlp": 1.0180434, + "epoch": 0.9809415018337041, + "flos": 21689794120320.0, + "grad_norm": 2.3748221706995403, + "language_loss": 0.78348863, + "learning_rate": 3.785824024012285e-09, + "loss": 0.80534726, + "num_input_tokens_seen": 176102730, + "step": 8158, + "time_per_iteration": 2.7959251403808594 + }, + { + "auxiliary_loss_clip": 0.01146742, + "auxiliary_loss_mlp": 0.01024033, + "balance_loss_clip": 1.04704189, + "balance_loss_mlp": 1.01681137, + "epoch": 0.9810617447243432, + "flos": 23294857357440.0, + "grad_norm": 1.5508584028597523, + "language_loss": 0.78719556, + "learning_rate": 3.738068738545541e-09, + "loss": 0.80890322, + "num_input_tokens_seen": 176121815, + "step": 8159, + "time_per_iteration": 2.802603006362915 + }, + { + "auxiliary_loss_clip": 0.0116297, + "auxiliary_loss_mlp": 0.0103235, + "balance_loss_clip": 1.0454123, + "balance_loss_mlp": 1.02435958, + "epoch": 0.9811819876149822, + "flos": 18332038748160.0, + "grad_norm": 2.501151646109076, + "language_loss": 0.78324986, + "learning_rate": 3.6906162844265733e-09, + "loss": 0.80520308, + "num_input_tokens_seen": 176138900, + "step": 8160, + "time_per_iteration": 2.769470691680908 + }, + { + "auxiliary_loss_clip": 0.01143735, + "auxiliary_loss_mlp": 0.01024339, + "balance_loss_clip": 1.0427959, + "balance_loss_mlp": 1.01531529, + "epoch": 0.9813022305056214, + "flos": 22601961025920.0, + "grad_norm": 1.9323578890298925, + "language_loss": 0.70699298, + "learning_rate": 3.643466668853845e-09, + "loss": 0.7286737, + "num_input_tokens_seen": 176156925, + "step": 8161, + "time_per_iteration": 2.7179348468780518 + }, + { + "auxiliary_loss_clip": 0.01151158, + "auxiliary_loss_mlp": 0.01021081, + "balance_loss_clip": 1.0429281, + "balance_loss_mlp": 1.01366603, + "epoch": 0.9814224733962604, + "flos": 25413335642880.0, + "grad_norm": 1.9904666008318646, + "language_loss": 0.75200701, + "learning_rate": 3.59661989898008e-09, + "loss": 0.77372944, + "num_input_tokens_seen": 176177980, + "step": 8162, + "time_per_iteration": 2.851284980773926 + }, + { + "auxiliary_loss_clip": 0.01130417, + "auxiliary_loss_mlp": 0.01024208, + "balance_loss_clip": 1.04340124, + "balance_loss_mlp": 1.01679039, + "epoch": 0.9815427162868995, + "flos": 25007185584000.0, + "grad_norm": 1.7594907474213404, + "language_loss": 0.76463866, + "learning_rate": 3.5500759819115934e-09, + "loss": 0.78618491, + "num_input_tokens_seen": 176198345, + "step": 8163, + "time_per_iteration": 2.83589243888855 + }, + { + "auxiliary_loss_clip": 0.01168974, + "auxiliary_loss_mlp": 0.01028827, + "balance_loss_clip": 1.04867673, + "balance_loss_mlp": 1.02073264, + "epoch": 0.9816629591775387, + "flos": 20662604887680.0, + "grad_norm": 1.891213599359792, + "language_loss": 0.81439048, + "learning_rate": 3.5038349247094034e-09, + "loss": 0.83636844, + "num_input_tokens_seen": 176215605, + "step": 8164, + "time_per_iteration": 2.7881786823272705 + }, + { + "auxiliary_loss_clip": 0.01149448, + "auxiliary_loss_mlp": 0.01025926, + "balance_loss_clip": 1.04232812, + "balance_loss_mlp": 1.01874352, + "epoch": 0.9817832020681777, + "flos": 17712220636800.0, + "grad_norm": 2.4273464766039385, + "language_loss": 0.77620465, + "learning_rate": 3.4578967343878994e-09, + "loss": 0.79795837, + "num_input_tokens_seen": 176231810, + "step": 8165, + "time_per_iteration": 3.7009811401367188 + }, + { + "auxiliary_loss_clip": 0.01151274, + "auxiliary_loss_mlp": 0.01025658, + "balance_loss_clip": 1.04677129, + "balance_loss_mlp": 1.01818395, + "epoch": 0.9819034449588168, + "flos": 22530032040960.0, + "grad_norm": 1.7218304629396786, + "language_loss": 0.81035209, + "learning_rate": 3.4122614179161733e-09, + "loss": 0.83212137, + "num_input_tokens_seen": 176251770, + "step": 8166, + "time_per_iteration": 2.731295347213745 + }, + { + "auxiliary_loss_clip": 0.01125876, + "auxiliary_loss_mlp": 0.0102326, + "balance_loss_clip": 1.03776348, + "balance_loss_mlp": 1.01596403, + "epoch": 0.9820236878494559, + "flos": 20011221699840.0, + "grad_norm": 2.3036462692777078, + "language_loss": 0.78059864, + "learning_rate": 3.36692898221691e-09, + "loss": 0.80209005, + "num_input_tokens_seen": 176270135, + "step": 8167, + "time_per_iteration": 4.5783421993255615 + }, + { + "auxiliary_loss_clip": 0.0115788, + "auxiliary_loss_mlp": 0.01024942, + "balance_loss_clip": 1.04434717, + "balance_loss_mlp": 1.01726747, + "epoch": 0.982143930740095, + "flos": 18807316531200.0, + "grad_norm": 1.6101028153246795, + "language_loss": 0.73184365, + "learning_rate": 3.3218994341668305e-09, + "loss": 0.75367188, + "num_input_tokens_seen": 176289065, + "step": 8168, + "time_per_iteration": 2.764256000518799 + }, + { + "auxiliary_loss_clip": 0.01168215, + "auxiliary_loss_mlp": 0.0102868, + "balance_loss_clip": 1.0493356, + "balance_loss_mlp": 1.02111328, + "epoch": 0.982264173630734, + "flos": 26578026138240.0, + "grad_norm": 1.855635917372128, + "language_loss": 0.75603288, + "learning_rate": 3.2771727805971373e-09, + "loss": 0.77800184, + "num_input_tokens_seen": 176310450, + "step": 8169, + "time_per_iteration": 2.7233569622039795 + }, + { + "auxiliary_loss_clip": 0.01124768, + "auxiliary_loss_mlp": 0.01024958, + "balance_loss_clip": 1.03856277, + "balance_loss_mlp": 1.01722097, + "epoch": 0.9823844165213732, + "flos": 22014462176640.0, + "grad_norm": 1.9137881683959237, + "language_loss": 0.7721234, + "learning_rate": 3.232749028292847e-09, + "loss": 0.79362071, + "num_input_tokens_seen": 176327415, + "step": 8170, + "time_per_iteration": 2.8643226623535156 + }, + { + "auxiliary_loss_clip": 0.01169342, + "auxiliary_loss_mlp": 0.01023884, + "balance_loss_clip": 1.04602981, + "balance_loss_mlp": 1.01615596, + "epoch": 0.9825046594120123, + "flos": 21908166854400.0, + "grad_norm": 1.8728659770464744, + "language_loss": 0.88415551, + "learning_rate": 3.188628183992792e-09, + "loss": 0.90608776, + "num_input_tokens_seen": 176347680, + "step": 8171, + "time_per_iteration": 3.637049913406372 + }, + { + "auxiliary_loss_clip": 0.01051727, + "auxiliary_loss_mlp": 0.01002016, + "balance_loss_clip": 1.0047884, + "balance_loss_mlp": 1.00054932, + "epoch": 0.9826249023026513, + "flos": 59494610718720.0, + "grad_norm": 0.7921041509613723, + "language_loss": 0.62476003, + "learning_rate": 3.1448102543902844e-09, + "loss": 0.64529741, + "num_input_tokens_seen": 176411595, + "step": 8172, + "time_per_iteration": 3.255570650100708 + }, + { + "auxiliary_loss_clip": 0.0114262, + "auxiliary_loss_mlp": 0.01024477, + "balance_loss_clip": 1.04249787, + "balance_loss_mlp": 1.01686215, + "epoch": 0.9827451451932905, + "flos": 16071031296000.0, + "grad_norm": 1.9261192698291847, + "language_loss": 0.67740047, + "learning_rate": 3.1012952461324515e-09, + "loss": 0.69907147, + "num_input_tokens_seen": 176430570, + "step": 8173, + "time_per_iteration": 2.7765755653381348 + }, + { + "auxiliary_loss_clip": 0.01156981, + "auxiliary_loss_mlp": 0.01029241, + "balance_loss_clip": 1.04838181, + "balance_loss_mlp": 1.02212691, + "epoch": 0.9828653880839295, + "flos": 20262775622400.0, + "grad_norm": 2.148626483990965, + "language_loss": 0.73953283, + "learning_rate": 3.0580831658204575e-09, + "loss": 0.76139504, + "num_input_tokens_seen": 176448150, + "step": 8174, + "time_per_iteration": 2.6986236572265625 + }, + { + "auxiliary_loss_clip": 0.01155591, + "auxiliary_loss_mlp": 0.01023343, + "balance_loss_clip": 1.04521728, + "balance_loss_mlp": 1.01604104, + "epoch": 0.9829856309745686, + "flos": 21616141282560.0, + "grad_norm": 1.9168435045146626, + "language_loss": 0.78178149, + "learning_rate": 3.015174020009281e-09, + "loss": 0.80357081, + "num_input_tokens_seen": 176467475, + "step": 8175, + "time_per_iteration": 2.838873863220215 + }, + { + "auxiliary_loss_clip": 0.0114202, + "auxiliary_loss_mlp": 0.01026065, + "balance_loss_clip": 1.04139423, + "balance_loss_mlp": 1.01884127, + "epoch": 0.9831058738652078, + "flos": 23764209396480.0, + "grad_norm": 2.1359220438394484, + "language_loss": 0.7511096, + "learning_rate": 2.9725678152086043e-09, + "loss": 0.77279049, + "num_input_tokens_seen": 176486045, + "step": 8176, + "time_per_iteration": 2.880112409591675 + }, + { + "auxiliary_loss_clip": 0.01141571, + "auxiliary_loss_mlp": 0.01024088, + "balance_loss_clip": 1.04314494, + "balance_loss_mlp": 1.0159874, + "epoch": 0.9832261167558468, + "flos": 11320911072000.0, + "grad_norm": 14.755118352668633, + "language_loss": 0.81815517, + "learning_rate": 2.930264557881257e-09, + "loss": 0.8398118, + "num_input_tokens_seen": 176501230, + "step": 8177, + "time_per_iteration": 2.6853818893432617 + }, + { + "auxiliary_loss_clip": 0.01055198, + "auxiliary_loss_mlp": 0.01001181, + "balance_loss_clip": 1.00456274, + "balance_loss_mlp": 0.99984562, + "epoch": 0.9833463596464859, + "flos": 60000304343040.0, + "grad_norm": 0.840470552767975, + "language_loss": 0.58208728, + "learning_rate": 2.8882642544452163e-09, + "loss": 0.602651, + "num_input_tokens_seen": 176565955, + "step": 8178, + "time_per_iteration": 3.242426633834839 + }, + { + "auxiliary_loss_clip": 0.01138613, + "auxiliary_loss_mlp": 0.01027391, + "balance_loss_clip": 1.04104114, + "balance_loss_mlp": 1.01984525, + "epoch": 0.983466602537125, + "flos": 13626699805440.0, + "grad_norm": 3.3605212882922033, + "language_loss": 0.74601364, + "learning_rate": 2.8465669112716083e-09, + "loss": 0.76767367, + "num_input_tokens_seen": 176583480, + "step": 8179, + "time_per_iteration": 2.8061132431030273 + }, + { + "auxiliary_loss_clip": 0.01162317, + "auxiliary_loss_mlp": 0.00886164, + "balance_loss_clip": 1.04665542, + "balance_loss_mlp": 1.00055754, + "epoch": 0.9835868454277641, + "flos": 22926844563840.0, + "grad_norm": 2.0583877097987155, + "language_loss": 0.77100343, + "learning_rate": 2.8051725346858177e-09, + "loss": 0.79148823, + "num_input_tokens_seen": 176603740, + "step": 8180, + "time_per_iteration": 2.823312520980835 + }, + { + "auxiliary_loss_clip": 0.01169319, + "auxiliary_loss_mlp": 0.01025718, + "balance_loss_clip": 1.0461154, + "balance_loss_mlp": 1.01779306, + "epoch": 0.9837070883184031, + "flos": 27673409341440.0, + "grad_norm": 2.105548421828895, + "language_loss": 0.71033949, + "learning_rate": 2.7640811309674883e-09, + "loss": 0.73228979, + "num_input_tokens_seen": 176623240, + "step": 8181, + "time_per_iteration": 2.6939663887023926 + }, + { + "auxiliary_loss_clip": 0.01132091, + "auxiliary_loss_mlp": 0.01027698, + "balance_loss_clip": 1.04219627, + "balance_loss_mlp": 1.02016664, + "epoch": 0.9838273312090423, + "flos": 29241951425280.0, + "grad_norm": 1.64754505548404, + "language_loss": 0.80883026, + "learning_rate": 2.7232927063498557e-09, + "loss": 0.83042812, + "num_input_tokens_seen": 176643615, + "step": 8182, + "time_per_iteration": 2.863746404647827 + }, + { + "auxiliary_loss_clip": 0.01160985, + "auxiliary_loss_mlp": 0.01029016, + "balance_loss_clip": 1.04722941, + "balance_loss_mlp": 1.02114773, + "epoch": 0.9839475740996814, + "flos": 40110207304320.0, + "grad_norm": 1.76601243645177, + "language_loss": 0.68828076, + "learning_rate": 2.682807267020859e-09, + "loss": 0.71018076, + "num_input_tokens_seen": 176666375, + "step": 8183, + "time_per_iteration": 2.871027946472168 + }, + { + "auxiliary_loss_clip": 0.0115897, + "auxiliary_loss_mlp": 0.01025122, + "balance_loss_clip": 1.0468657, + "balance_loss_mlp": 1.01724219, + "epoch": 0.9840678169903204, + "flos": 24169389788160.0, + "grad_norm": 1.5772139228101403, + "language_loss": 0.62468445, + "learning_rate": 2.642624819121808e-09, + "loss": 0.64652538, + "num_input_tokens_seen": 176686525, + "step": 8184, + "time_per_iteration": 2.7504725456237793 + }, + { + "auxiliary_loss_clip": 0.01147256, + "auxiliary_loss_mlp": 0.01023139, + "balance_loss_clip": 1.04612875, + "balance_loss_mlp": 1.01609325, + "epoch": 0.9841880598809596, + "flos": 14684484447360.0, + "grad_norm": 2.546307735360777, + "language_loss": 0.61396646, + "learning_rate": 2.6027453687487154e-09, + "loss": 0.63567048, + "num_input_tokens_seen": 176703615, + "step": 8185, + "time_per_iteration": 2.742823839187622 + }, + { + "auxiliary_loss_clip": 0.01149061, + "auxiliary_loss_mlp": 0.01028337, + "balance_loss_clip": 1.04414701, + "balance_loss_mlp": 1.02060676, + "epoch": 0.9843083027715986, + "flos": 22344768668160.0, + "grad_norm": 2.683935644476791, + "language_loss": 0.53504306, + "learning_rate": 2.5631689219509643e-09, + "loss": 0.55681705, + "num_input_tokens_seen": 176722295, + "step": 8186, + "time_per_iteration": 2.679067373275757 + }, + { + "auxiliary_loss_clip": 0.01141848, + "auxiliary_loss_mlp": 0.01022779, + "balance_loss_clip": 1.04490376, + "balance_loss_mlp": 1.01562047, + "epoch": 0.9844285456622377, + "flos": 21800111765760.0, + "grad_norm": 1.7911840518120068, + "language_loss": 0.83395529, + "learning_rate": 2.523895484732197e-09, + "loss": 0.85560155, + "num_input_tokens_seen": 176741750, + "step": 8187, + "time_per_iteration": 2.77199387550354 + }, + { + "auxiliary_loss_clip": 0.01164606, + "auxiliary_loss_mlp": 0.01027836, + "balance_loss_clip": 1.04583466, + "balance_loss_mlp": 1.019521, + "epoch": 0.9845487885528769, + "flos": 18035380321920.0, + "grad_norm": 1.8055792963082187, + "language_loss": 0.74759436, + "learning_rate": 2.4849250630505357e-09, + "loss": 0.76951879, + "num_input_tokens_seen": 176759995, + "step": 8188, + "time_per_iteration": 2.743060350418091 + }, + { + "auxiliary_loss_clip": 0.01102857, + "auxiliary_loss_mlp": 0.01030076, + "balance_loss_clip": 1.03942621, + "balance_loss_mlp": 1.02276289, + "epoch": 0.9846690314435159, + "flos": 25228610974080.0, + "grad_norm": 4.048190488842716, + "language_loss": 0.7305038, + "learning_rate": 2.4462576628172528e-09, + "loss": 0.75183308, + "num_input_tokens_seen": 176778625, + "step": 8189, + "time_per_iteration": 3.043469190597534 + }, + { + "auxiliary_loss_clip": 0.01157153, + "auxiliary_loss_mlp": 0.01035839, + "balance_loss_clip": 1.04777789, + "balance_loss_mlp": 1.02810264, + "epoch": 0.984789274334155, + "flos": 18552171248640.0, + "grad_norm": 2.576772514889621, + "language_loss": 0.73953432, + "learning_rate": 2.407893289898766e-09, + "loss": 0.76146424, + "num_input_tokens_seen": 176797655, + "step": 8190, + "time_per_iteration": 2.6802077293395996 + }, + { + "auxiliary_loss_clip": 0.01131597, + "auxiliary_loss_mlp": 0.0102971, + "balance_loss_clip": 1.03998995, + "balance_loss_mlp": 1.02163315, + "epoch": 0.984909517224794, + "flos": 27345437233920.0, + "grad_norm": 1.8495117415370075, + "language_loss": 0.83799821, + "learning_rate": 2.3698319501144202e-09, + "loss": 0.85961127, + "num_input_tokens_seen": 176818640, + "step": 8191, + "time_per_iteration": 3.7838051319122314 + }, + { + "auxiliary_loss_clip": 0.01156858, + "auxiliary_loss_mlp": 0.01031421, + "balance_loss_clip": 1.04584241, + "balance_loss_mlp": 1.0231483, + "epoch": 0.9850297601154332, + "flos": 18734058743040.0, + "grad_norm": 1.8138002279995793, + "language_loss": 0.73707712, + "learning_rate": 2.3320736492382644e-09, + "loss": 0.75896001, + "num_input_tokens_seen": 176837475, + "step": 8192, + "time_per_iteration": 2.6902337074279785 + }, + { + "auxiliary_loss_clip": 0.01167073, + "auxiliary_loss_mlp": 0.01028645, + "balance_loss_clip": 1.04769921, + "balance_loss_mlp": 1.02115297, + "epoch": 0.9851500030060723, + "flos": 22308247514880.0, + "grad_norm": 1.6412871503841857, + "language_loss": 0.68073118, + "learning_rate": 2.29461839299816e-09, + "loss": 0.70268834, + "num_input_tokens_seen": 176857190, + "step": 8193, + "time_per_iteration": 4.660129070281982 + }, + { + "auxiliary_loss_clip": 0.01132496, + "auxiliary_loss_mlp": 0.01029615, + "balance_loss_clip": 1.04173052, + "balance_loss_mlp": 1.02239728, + "epoch": 0.9852702458967113, + "flos": 26353691746560.0, + "grad_norm": 1.716911185668347, + "language_loss": 0.80057269, + "learning_rate": 2.257466187076229e-09, + "loss": 0.8221938, + "num_input_tokens_seen": 176876395, + "step": 8194, + "time_per_iteration": 2.8113832473754883 + }, + { + "auxiliary_loss_clip": 0.01165177, + "auxiliary_loss_mlp": 0.00885876, + "balance_loss_clip": 1.04649329, + "balance_loss_mlp": 1.00052464, + "epoch": 0.9853904887873505, + "flos": 20883599314560.0, + "grad_norm": 1.6036055915789855, + "language_loss": 0.70895094, + "learning_rate": 2.2206170371081854e-09, + "loss": 0.72946143, + "num_input_tokens_seen": 176894980, + "step": 8195, + "time_per_iteration": 2.709063768386841 + }, + { + "auxiliary_loss_clip": 0.01150031, + "auxiliary_loss_mlp": 0.01035939, + "balance_loss_clip": 1.04317522, + "balance_loss_mlp": 1.02812481, + "epoch": 0.9855107316779895, + "flos": 25263444188160.0, + "grad_norm": 1.6770263537506709, + "language_loss": 0.84857047, + "learning_rate": 2.1840709486842247e-09, + "loss": 0.87043017, + "num_input_tokens_seen": 176914600, + "step": 8196, + "time_per_iteration": 3.6746816635131836 + }, + { + "auxiliary_loss_clip": 0.01145199, + "auxiliary_loss_mlp": 0.01029324, + "balance_loss_clip": 1.0436058, + "balance_loss_mlp": 1.0210743, + "epoch": 0.9856309745686286, + "flos": 19062102677760.0, + "grad_norm": 2.2724954806403486, + "language_loss": 0.79282492, + "learning_rate": 2.1478279273481335e-09, + "loss": 0.81457019, + "num_input_tokens_seen": 176933085, + "step": 8197, + "time_per_iteration": 2.7269644737243652 + }, + { + "auxiliary_loss_clip": 0.01159752, + "auxiliary_loss_mlp": 0.0102631, + "balance_loss_clip": 1.04843068, + "balance_loss_mlp": 1.01873422, + "epoch": 0.9857512174592677, + "flos": 34130758060800.0, + "grad_norm": 2.063025242029517, + "language_loss": 0.80260623, + "learning_rate": 2.1118879785981815e-09, + "loss": 0.82446688, + "num_input_tokens_seen": 176953225, + "step": 8198, + "time_per_iteration": 2.7882180213928223 + }, + { + "auxiliary_loss_clip": 0.01150602, + "auxiliary_loss_mlp": 0.01023636, + "balance_loss_clip": 1.04492581, + "balance_loss_mlp": 1.01601315, + "epoch": 0.9858714603499068, + "flos": 25994693266560.0, + "grad_norm": 1.8979282321378748, + "language_loss": 0.79310447, + "learning_rate": 2.0762511078862288e-09, + "loss": 0.81484687, + "num_input_tokens_seen": 176973570, + "step": 8199, + "time_per_iteration": 2.734984874725342 + }, + { + "auxiliary_loss_clip": 0.01147578, + "auxiliary_loss_mlp": 0.01028244, + "balance_loss_clip": 1.04275107, + "balance_loss_mlp": 1.02032304, + "epoch": 0.9859917032405459, + "flos": 23696230907520.0, + "grad_norm": 2.3334363265466567, + "language_loss": 0.64955854, + "learning_rate": 2.0409173206186183e-09, + "loss": 0.67131674, + "num_input_tokens_seen": 176992810, + "step": 8200, + "time_per_iteration": 2.776698112487793 + }, + { + "auxiliary_loss_clip": 0.0113917, + "auxiliary_loss_mlp": 0.01026424, + "balance_loss_clip": 1.04579115, + "balance_loss_mlp": 1.01936984, + "epoch": 0.986111946131185, + "flos": 19938287134080.0, + "grad_norm": 1.899956468075688, + "language_loss": 0.87097752, + "learning_rate": 2.0058866221550617e-09, + "loss": 0.89263356, + "num_input_tokens_seen": 177011050, + "step": 8201, + "time_per_iteration": 2.8963568210601807 + }, + { + "auxiliary_loss_clip": 0.01169987, + "auxiliary_loss_mlp": 0.01026095, + "balance_loss_clip": 1.04654765, + "balance_loss_mlp": 1.01813173, + "epoch": 0.9862321890218241, + "flos": 19828831415040.0, + "grad_norm": 2.0842315869783086, + "language_loss": 0.75070298, + "learning_rate": 1.971159017809976e-09, + "loss": 0.77266383, + "num_input_tokens_seen": 177029340, + "step": 8202, + "time_per_iteration": 2.7414164543151855 + }, + { + "auxiliary_loss_clip": 0.01159041, + "auxiliary_loss_mlp": 0.01025392, + "balance_loss_clip": 1.04715037, + "balance_loss_mlp": 1.01786041, + "epoch": 0.9863524319124631, + "flos": 21652051904640.0, + "grad_norm": 2.3029105036896538, + "language_loss": 0.77550107, + "learning_rate": 1.93673451285159e-09, + "loss": 0.7973454, + "num_input_tokens_seen": 177048390, + "step": 8203, + "time_per_iteration": 2.731980562210083 + }, + { + "auxiliary_loss_clip": 0.0104696, + "auxiliary_loss_mlp": 0.01001817, + "balance_loss_clip": 1.0046221, + "balance_loss_mlp": 1.00029719, + "epoch": 0.9864726748031023, + "flos": 52769977920000.0, + "grad_norm": 0.7414218886870468, + "language_loss": 0.56512439, + "learning_rate": 1.9026131125019495e-09, + "loss": 0.58561218, + "num_input_tokens_seen": 177105760, + "step": 8204, + "time_per_iteration": 3.2543962001800537 + }, + { + "auxiliary_loss_clip": 0.01155189, + "auxiliary_loss_mlp": 0.0102271, + "balance_loss_clip": 1.04492879, + "balance_loss_mlp": 1.01550364, + "epoch": 0.9865929176937414, + "flos": 23364631526400.0, + "grad_norm": 1.7055019659860624, + "language_loss": 0.86935413, + "learning_rate": 1.8687948219371363e-09, + "loss": 0.89113307, + "num_input_tokens_seen": 177124985, + "step": 8205, + "time_per_iteration": 2.7264132499694824 + }, + { + "auxiliary_loss_clip": 0.01170745, + "auxiliary_loss_mlp": 0.01028163, + "balance_loss_clip": 1.04580748, + "balance_loss_mlp": 1.01965189, + "epoch": 0.9867131605843804, + "flos": 21616679986560.0, + "grad_norm": 2.1787007793340174, + "language_loss": 0.88404936, + "learning_rate": 1.835279646287491e-09, + "loss": 0.90603846, + "num_input_tokens_seen": 177142995, + "step": 8206, + "time_per_iteration": 2.6620664596557617 + }, + { + "auxiliary_loss_clip": 0.0116716, + "auxiliary_loss_mlp": 0.01041149, + "balance_loss_clip": 1.0482794, + "balance_loss_mlp": 1.03296483, + "epoch": 0.9868334034750196, + "flos": 22271403139200.0, + "grad_norm": 1.7042864463439602, + "language_loss": 0.76597238, + "learning_rate": 1.8020675906371685e-09, + "loss": 0.78805548, + "num_input_tokens_seen": 177162390, + "step": 8207, + "time_per_iteration": 2.7643885612487793 + }, + { + "auxiliary_loss_clip": 0.01127997, + "auxiliary_loss_mlp": 0.01027429, + "balance_loss_clip": 1.0409627, + "balance_loss_mlp": 1.01954305, + "epoch": 0.9869536463656586, + "flos": 25809573548160.0, + "grad_norm": 1.8758083990939003, + "language_loss": 0.75310367, + "learning_rate": 1.7691586600243612e-09, + "loss": 0.77465796, + "num_input_tokens_seen": 177181290, + "step": 8208, + "time_per_iteration": 2.8837013244628906 + }, + { + "auxiliary_loss_clip": 0.01145439, + "auxiliary_loss_mlp": 0.01021897, + "balance_loss_clip": 1.04538643, + "balance_loss_mlp": 1.01417232, + "epoch": 0.9870738892562977, + "flos": 16398500613120.0, + "grad_norm": 2.8928058677402295, + "language_loss": 0.86698025, + "learning_rate": 1.7365528594415202e-09, + "loss": 0.88865364, + "num_input_tokens_seen": 177195360, + "step": 8209, + "time_per_iteration": 2.773414134979248 + }, + { + "auxiliary_loss_clip": 0.01165095, + "auxiliary_loss_mlp": 0.00885826, + "balance_loss_clip": 1.04637659, + "balance_loss_mlp": 1.00057077, + "epoch": 0.9871941321469369, + "flos": 35481358373760.0, + "grad_norm": 1.8059030274112151, + "language_loss": 0.67639709, + "learning_rate": 1.7042501938346888e-09, + "loss": 0.69690633, + "num_input_tokens_seen": 177218090, + "step": 8210, + "time_per_iteration": 2.855776309967041 + }, + { + "auxiliary_loss_clip": 0.01136162, + "auxiliary_loss_mlp": 0.01027141, + "balance_loss_clip": 1.03854632, + "balance_loss_mlp": 1.02027786, + "epoch": 0.9873143750375759, + "flos": 21434217874560.0, + "grad_norm": 1.8374639070812802, + "language_loss": 0.7635808, + "learning_rate": 1.6722506681043913e-09, + "loss": 0.78521383, + "num_input_tokens_seen": 177237050, + "step": 8211, + "time_per_iteration": 2.8331105709075928 + }, + { + "auxiliary_loss_clip": 0.01153757, + "auxiliary_loss_mlp": 0.01029762, + "balance_loss_clip": 1.04496908, + "balance_loss_mlp": 1.0224545, + "epoch": 0.987434617928215, + "flos": 16326499800960.0, + "grad_norm": 2.2349965444599973, + "language_loss": 0.69155955, + "learning_rate": 1.640554287104745e-09, + "loss": 0.71339476, + "num_input_tokens_seen": 177255325, + "step": 8212, + "time_per_iteration": 2.722172498703003 + }, + { + "auxiliary_loss_clip": 0.01141298, + "auxiliary_loss_mlp": 0.01031932, + "balance_loss_clip": 1.0387969, + "balance_loss_mlp": 1.02318215, + "epoch": 0.9875548608188541, + "flos": 17851984456320.0, + "grad_norm": 1.989603960838646, + "language_loss": 0.805152, + "learning_rate": 1.609161055644348e-09, + "loss": 0.82688427, + "num_input_tokens_seen": 177271250, + "step": 8213, + "time_per_iteration": 2.826180934906006 + }, + { + "auxiliary_loss_clip": 0.01163015, + "auxiliary_loss_mlp": 0.01021894, + "balance_loss_clip": 1.04358602, + "balance_loss_mlp": 1.01407361, + "epoch": 0.9876751037094932, + "flos": 26132876887680.0, + "grad_norm": 3.199006676858439, + "language_loss": 0.68061954, + "learning_rate": 1.5780709784849467e-09, + "loss": 0.70246863, + "num_input_tokens_seen": 177288270, + "step": 8214, + "time_per_iteration": 2.732172966003418 + }, + { + "auxiliary_loss_clip": 0.01126644, + "auxiliary_loss_mlp": 0.01026077, + "balance_loss_clip": 1.04481435, + "balance_loss_mlp": 1.01792359, + "epoch": 0.9877953466001322, + "flos": 15991344973440.0, + "grad_norm": 2.214444157155429, + "language_loss": 0.82280374, + "learning_rate": 1.5472840603436565e-09, + "loss": 0.84433091, + "num_input_tokens_seen": 177305500, + "step": 8215, + "time_per_iteration": 2.811066150665283 + }, + { + "auxiliary_loss_clip": 0.01145813, + "auxiliary_loss_mlp": 0.01025705, + "balance_loss_clip": 1.04586244, + "balance_loss_mlp": 1.01830208, + "epoch": 0.9879155894907714, + "flos": 18806777827200.0, + "grad_norm": 2.69952354679353, + "language_loss": 0.78025091, + "learning_rate": 1.5168003058900757e-09, + "loss": 0.80196607, + "num_input_tokens_seen": 177323500, + "step": 8216, + "time_per_iteration": 2.6856930255889893 + }, + { + "auxiliary_loss_clip": 0.01139725, + "auxiliary_loss_mlp": 0.01027535, + "balance_loss_clip": 1.04156113, + "balance_loss_mlp": 1.02029932, + "epoch": 0.9880358323814105, + "flos": 22382044007040.0, + "grad_norm": 1.984752209069816, + "language_loss": 0.92175651, + "learning_rate": 1.4866197197491715e-09, + "loss": 0.94342911, + "num_input_tokens_seen": 177342860, + "step": 8217, + "time_per_iteration": 3.737863540649414 + }, + { + "auxiliary_loss_clip": 0.01163519, + "auxiliary_loss_mlp": 0.00886642, + "balance_loss_clip": 1.04736245, + "balance_loss_mlp": 1.00059855, + "epoch": 0.9881560752720495, + "flos": 15668831733120.0, + "grad_norm": 3.1662332435431013, + "language_loss": 0.78654408, + "learning_rate": 1.4567423064988371e-09, + "loss": 0.8070457, + "num_input_tokens_seen": 177360210, + "step": 8218, + "time_per_iteration": 2.752932071685791 + }, + { + "auxiliary_loss_clip": 0.01169017, + "auxiliary_loss_mlp": 0.01029968, + "balance_loss_clip": 1.04624939, + "balance_loss_mlp": 1.02235603, + "epoch": 0.9882763181626887, + "flos": 21500113374720.0, + "grad_norm": 2.0764291456325044, + "language_loss": 0.77743262, + "learning_rate": 1.4271680706718913e-09, + "loss": 0.79942238, + "num_input_tokens_seen": 177377885, + "step": 8219, + "time_per_iteration": 4.536207914352417 + }, + { + "auxiliary_loss_clip": 0.01162711, + "auxiliary_loss_mlp": 0.0102943, + "balance_loss_clip": 1.04862845, + "balance_loss_mlp": 1.02103198, + "epoch": 0.9883965610533277, + "flos": 28034598551040.0, + "grad_norm": 1.7514924784995909, + "language_loss": 0.82677823, + "learning_rate": 1.3978970167543013e-09, + "loss": 0.84869969, + "num_input_tokens_seen": 177398065, + "step": 8220, + "time_per_iteration": 2.7750139236450195 + }, + { + "auxiliary_loss_clip": 0.01142312, + "auxiliary_loss_mlp": 0.01030639, + "balance_loss_clip": 1.04349887, + "balance_loss_mlp": 1.02265215, + "epoch": 0.9885168039439668, + "flos": 14098601710080.0, + "grad_norm": 2.0422898374938936, + "language_loss": 0.77021146, + "learning_rate": 1.3689291491867372e-09, + "loss": 0.79194105, + "num_input_tokens_seen": 177416380, + "step": 8221, + "time_per_iteration": 2.764402151107788 + }, + { + "auxiliary_loss_clip": 0.01170391, + "auxiliary_loss_mlp": 0.01025672, + "balance_loss_clip": 1.04738879, + "balance_loss_mlp": 1.01749444, + "epoch": 0.988637046834606, + "flos": 26432013352320.0, + "grad_norm": 1.8224467021139101, + "language_loss": 0.73572946, + "learning_rate": 1.3402644723636836e-09, + "loss": 0.75769007, + "num_input_tokens_seen": 177438410, + "step": 8222, + "time_per_iteration": 2.6623637676239014 + }, + { + "auxiliary_loss_clip": 0.01149374, + "auxiliary_loss_mlp": 0.01029717, + "balance_loss_clip": 1.04769468, + "balance_loss_mlp": 1.02218902, + "epoch": 0.988757289725245, + "flos": 25229113764480.0, + "grad_norm": 2.0858443431947773, + "language_loss": 0.83276069, + "learning_rate": 1.311902990633218e-09, + "loss": 0.85455161, + "num_input_tokens_seen": 177457375, + "step": 8223, + "time_per_iteration": 3.8495254516601562 + }, + { + "auxiliary_loss_clip": 0.01141304, + "auxiliary_loss_mlp": 0.01025504, + "balance_loss_clip": 1.03917134, + "balance_loss_mlp": 1.01794624, + "epoch": 0.9888775326158841, + "flos": 26359042872960.0, + "grad_norm": 1.549707358601602, + "language_loss": 0.71289903, + "learning_rate": 1.2838447082978987e-09, + "loss": 0.73456705, + "num_input_tokens_seen": 177478530, + "step": 8224, + "time_per_iteration": 2.833354949951172 + }, + { + "auxiliary_loss_clip": 0.01155289, + "auxiliary_loss_mlp": 0.01026146, + "balance_loss_clip": 1.04354429, + "balance_loss_mlp": 1.0182718, + "epoch": 0.9889977755065231, + "flos": 24316120846080.0, + "grad_norm": 2.546824769749074, + "language_loss": 0.83321536, + "learning_rate": 1.2560896296143208e-09, + "loss": 0.8550297, + "num_input_tokens_seen": 177496995, + "step": 8225, + "time_per_iteration": 2.8502047061920166 + }, + { + "auxiliary_loss_clip": 0.01168042, + "auxiliary_loss_mlp": 0.010293, + "balance_loss_clip": 1.04629683, + "balance_loss_mlp": 1.0216167, + "epoch": 0.9891180183971623, + "flos": 18951066760320.0, + "grad_norm": 2.481852171661861, + "language_loss": 0.8230387, + "learning_rate": 1.2286377587926722e-09, + "loss": 0.84501207, + "num_input_tokens_seen": 177513785, + "step": 8226, + "time_per_iteration": 2.7540040016174316 + }, + { + "auxiliary_loss_clip": 0.01168124, + "auxiliary_loss_mlp": 0.01021677, + "balance_loss_clip": 1.04565978, + "balance_loss_mlp": 1.01436996, + "epoch": 0.9892382612878013, + "flos": 26176580760960.0, + "grad_norm": 2.0143666981009583, + "language_loss": 0.74882925, + "learning_rate": 1.2014890999973992e-09, + "loss": 0.77072728, + "num_input_tokens_seen": 177530705, + "step": 8227, + "time_per_iteration": 2.8960342407226562 + }, + { + "auxiliary_loss_clip": 0.01165718, + "auxiliary_loss_mlp": 0.01023958, + "balance_loss_clip": 1.04420948, + "balance_loss_mlp": 1.01671565, + "epoch": 0.9893585041784404, + "flos": 25449605400960.0, + "grad_norm": 1.5658766858698925, + "language_loss": 0.78487957, + "learning_rate": 1.1746436573472073e-09, + "loss": 0.80677629, + "num_input_tokens_seen": 177552440, + "step": 8228, + "time_per_iteration": 2.7204575538635254 + }, + { + "auxiliary_loss_clip": 0.01157798, + "auxiliary_loss_mlp": 0.01027986, + "balance_loss_clip": 1.04473686, + "balance_loss_mlp": 1.02028823, + "epoch": 0.9894787470690796, + "flos": 20189302352640.0, + "grad_norm": 2.03169844823351, + "language_loss": 0.69155532, + "learning_rate": 1.1481014349141726e-09, + "loss": 0.71341318, + "num_input_tokens_seen": 177569660, + "step": 8229, + "time_per_iteration": 2.8483920097351074 + }, + { + "auxiliary_loss_clip": 0.01152587, + "auxiliary_loss_mlp": 0.01029829, + "balance_loss_clip": 1.04503012, + "balance_loss_mlp": 1.02175891, + "epoch": 0.9895989899597186, + "flos": 24644308435200.0, + "grad_norm": 2.252354105123092, + "language_loss": 0.84688735, + "learning_rate": 1.121862436724852e-09, + "loss": 0.86871159, + "num_input_tokens_seen": 177588500, + "step": 8230, + "time_per_iteration": 2.759221315383911 + }, + { + "auxiliary_loss_clip": 0.01158907, + "auxiliary_loss_mlp": 0.0102401, + "balance_loss_clip": 1.04643679, + "balance_loss_mlp": 1.01647043, + "epoch": 0.9897192328503577, + "flos": 21799034357760.0, + "grad_norm": 1.7892393988892603, + "language_loss": 0.69866264, + "learning_rate": 1.0959266667598388e-09, + "loss": 0.72049183, + "num_input_tokens_seen": 177607315, + "step": 8231, + "time_per_iteration": 2.6866986751556396 + }, + { + "auxiliary_loss_clip": 0.01142328, + "auxiliary_loss_mlp": 0.01030784, + "balance_loss_clip": 1.04385662, + "balance_loss_mlp": 1.02212346, + "epoch": 0.9898394757409968, + "flos": 21325229032320.0, + "grad_norm": 1.9729112060099285, + "language_loss": 0.74398792, + "learning_rate": 1.0702941289533196e-09, + "loss": 0.76571906, + "num_input_tokens_seen": 177625990, + "step": 8232, + "time_per_iteration": 2.775374412536621 + }, + { + "auxiliary_loss_clip": 0.01137572, + "auxiliary_loss_mlp": 0.01026901, + "balance_loss_clip": 1.04217601, + "balance_loss_mlp": 1.0191462, + "epoch": 0.9899597186316359, + "flos": 18545024442240.0, + "grad_norm": 2.493162627343286, + "language_loss": 0.88690376, + "learning_rate": 1.0449648271939615e-09, + "loss": 0.90854847, + "num_input_tokens_seen": 177642335, + "step": 8233, + "time_per_iteration": 3.1581342220306396 + }, + { + "auxiliary_loss_clip": 0.01134949, + "auxiliary_loss_mlp": 0.00885453, + "balance_loss_clip": 1.04419851, + "balance_loss_mlp": 1.00056958, + "epoch": 0.990079961522275, + "flos": 23766723348480.0, + "grad_norm": 1.5311601362804528, + "language_loss": 0.72865129, + "learning_rate": 1.0199387653240243e-09, + "loss": 0.74885535, + "num_input_tokens_seen": 177662025, + "step": 8234, + "time_per_iteration": 3.0437967777252197 + }, + { + "auxiliary_loss_clip": 0.01143444, + "auxiliary_loss_mlp": 0.01027407, + "balance_loss_clip": 1.04479361, + "balance_loss_mlp": 1.01996839, + "epoch": 0.9902002044129141, + "flos": 16399182971520.0, + "grad_norm": 1.5960388230752716, + "language_loss": 0.70561993, + "learning_rate": 9.952159471400267e-10, + "loss": 0.72732842, + "num_input_tokens_seen": 177679065, + "step": 8235, + "time_per_iteration": 2.731485366821289 + }, + { + "auxiliary_loss_clip": 0.01153968, + "auxiliary_loss_mlp": 0.00885354, + "balance_loss_clip": 1.04564166, + "balance_loss_mlp": 1.00058103, + "epoch": 0.9903204473035532, + "flos": 22559657783040.0, + "grad_norm": 1.9886646270500625, + "language_loss": 0.84182048, + "learning_rate": 9.707963763923022e-10, + "loss": 0.86221373, + "num_input_tokens_seen": 177698115, + "step": 8236, + "time_per_iteration": 2.820355176925659 + }, + { + "auxiliary_loss_clip": 0.01146095, + "auxiliary_loss_mlp": 0.01028399, + "balance_loss_clip": 1.0399164, + "balance_loss_mlp": 1.0206238, + "epoch": 0.9904406901941922, + "flos": 16144001775360.0, + "grad_norm": 1.667477395843247, + "language_loss": 0.78873932, + "learning_rate": 9.466800567854427e-10, + "loss": 0.81048429, + "num_input_tokens_seen": 177716715, + "step": 8237, + "time_per_iteration": 2.754495859146118 + }, + { + "auxiliary_loss_clip": 0.01133091, + "auxiliary_loss_mlp": 0.01033808, + "balance_loss_clip": 1.0416379, + "balance_loss_mlp": 1.02509999, + "epoch": 0.9905609330848314, + "flos": 26651499408000.0, + "grad_norm": 1.9663449298587232, + "language_loss": 0.67976379, + "learning_rate": 9.228669919778553e-10, + "loss": 0.70143282, + "num_input_tokens_seen": 177735640, + "step": 8238, + "time_per_iteration": 2.9059505462646484 + }, + { + "auxiliary_loss_clip": 0.01136907, + "auxiliary_loss_mlp": 0.01030255, + "balance_loss_clip": 1.04291129, + "balance_loss_mlp": 1.02239335, + "epoch": 0.9906811759754705, + "flos": 23111820627840.0, + "grad_norm": 4.678859747557513, + "language_loss": 0.797589, + "learning_rate": 8.993571855817617e-10, + "loss": 0.8192606, + "num_input_tokens_seen": 177754470, + "step": 8239, + "time_per_iteration": 2.7762539386749268 + }, + { + "auxiliary_loss_clip": 0.01156739, + "auxiliary_loss_mlp": 0.01023658, + "balance_loss_clip": 1.04453659, + "balance_loss_mlp": 1.01593363, + "epoch": 0.9908014188661095, + "flos": 22090593052800.0, + "grad_norm": 1.9166253874675694, + "language_loss": 0.75134933, + "learning_rate": 8.761506411638642e-10, + "loss": 0.77315331, + "num_input_tokens_seen": 177773935, + "step": 8240, + "time_per_iteration": 2.8598906993865967 + }, + { + "auxiliary_loss_clip": 0.01149963, + "auxiliary_loss_mlp": 0.01026559, + "balance_loss_clip": 1.04623485, + "balance_loss_mlp": 1.01907825, + "epoch": 0.9909216617567487, + "flos": 19242948677760.0, + "grad_norm": 1.9609131464067162, + "language_loss": 0.73587602, + "learning_rate": 8.53247362244236e-10, + "loss": 0.75764126, + "num_input_tokens_seen": 177792745, + "step": 8241, + "time_per_iteration": 2.8014581203460693 + }, + { + "auxiliary_loss_clip": 0.01150182, + "auxiliary_loss_mlp": 0.01023769, + "balance_loss_clip": 1.04517841, + "balance_loss_mlp": 1.01571655, + "epoch": 0.9910419046473877, + "flos": 23621213352960.0, + "grad_norm": 1.818866378187455, + "language_loss": 0.68361735, + "learning_rate": 8.306473522976532e-10, + "loss": 0.7053569, + "num_input_tokens_seen": 177812150, + "step": 8242, + "time_per_iteration": 2.810197353363037 + }, + { + "auxiliary_loss_clip": 0.01167978, + "auxiliary_loss_mlp": 0.01029608, + "balance_loss_clip": 1.0471313, + "balance_loss_mlp": 1.02150178, + "epoch": 0.9911621475380268, + "flos": 22711380831360.0, + "grad_norm": 1.7274526730336348, + "language_loss": 0.71665311, + "learning_rate": 8.083506147522623e-10, + "loss": 0.73862898, + "num_input_tokens_seen": 177831545, + "step": 8243, + "time_per_iteration": 3.6186399459838867 + }, + { + "auxiliary_loss_clip": 0.0115343, + "auxiliary_loss_mlp": 0.010354, + "balance_loss_clip": 1.0436579, + "balance_loss_mlp": 1.02800024, + "epoch": 0.991282390428666, + "flos": 13516956777600.0, + "grad_norm": 2.2278497611361496, + "language_loss": 0.85407567, + "learning_rate": 7.863571529906909e-10, + "loss": 0.87596399, + "num_input_tokens_seen": 177847130, + "step": 8244, + "time_per_iteration": 2.6964876651763916 + }, + { + "auxiliary_loss_clip": 0.0105155, + "auxiliary_loss_mlp": 0.0100199, + "balance_loss_clip": 1.00510287, + "balance_loss_mlp": 1.00051141, + "epoch": 0.991402633319305, + "flos": 61830492071040.0, + "grad_norm": 0.8229896729374744, + "language_loss": 0.59636664, + "learning_rate": 7.646669703489372e-10, + "loss": 0.61690199, + "num_input_tokens_seen": 177911440, + "step": 8245, + "time_per_iteration": 4.438992023468018 + }, + { + "auxiliary_loss_clip": 0.01098646, + "auxiliary_loss_mlp": 0.0102538, + "balance_loss_clip": 1.0325067, + "balance_loss_mlp": 1.01797664, + "epoch": 0.9915228762099441, + "flos": 18770148933120.0, + "grad_norm": 2.639924353799587, + "language_loss": 0.5696131, + "learning_rate": 7.432800701177023e-10, + "loss": 0.59085333, + "num_input_tokens_seen": 177929440, + "step": 8246, + "time_per_iteration": 3.100931167602539 + }, + { + "auxiliary_loss_clip": 0.01046058, + "auxiliary_loss_mlp": 0.01001653, + "balance_loss_clip": 1.00579286, + "balance_loss_mlp": 1.00019825, + "epoch": 0.9916431191005832, + "flos": 65936660244480.0, + "grad_norm": 0.8040676864162017, + "language_loss": 0.5779174, + "learning_rate": 7.221964555415017e-10, + "loss": 0.59839451, + "num_input_tokens_seen": 177989100, + "step": 8247, + "time_per_iteration": 3.455082654953003 + }, + { + "auxiliary_loss_clip": 0.01148388, + "auxiliary_loss_mlp": 0.01021335, + "balance_loss_clip": 1.04636025, + "balance_loss_mlp": 1.01428986, + "epoch": 0.9917633619912223, + "flos": 16581573256320.0, + "grad_norm": 1.7847701263413167, + "language_loss": 0.75110424, + "learning_rate": 7.01416129818222e-10, + "loss": 0.77280146, + "num_input_tokens_seen": 178006720, + "step": 8248, + "time_per_iteration": 3.9131836891174316 + }, + { + "auxiliary_loss_clip": 0.01147981, + "auxiliary_loss_mlp": 0.01024634, + "balance_loss_clip": 1.04362249, + "balance_loss_mlp": 1.01695657, + "epoch": 0.9918836048818613, + "flos": 25411108999680.0, + "grad_norm": 1.9474367115426683, + "language_loss": 0.58549809, + "learning_rate": 6.809390961006745e-10, + "loss": 0.60722423, + "num_input_tokens_seen": 178026850, + "step": 8249, + "time_per_iteration": 2.9033210277557373 + }, + { + "auxiliary_loss_clip": 0.01152305, + "auxiliary_loss_mlp": 0.0103081, + "balance_loss_clip": 1.04623342, + "balance_loss_mlp": 1.02291262, + "epoch": 0.9920038477725005, + "flos": 25046867134080.0, + "grad_norm": 1.9251471633657549, + "language_loss": 0.68633717, + "learning_rate": 6.607653574948191e-10, + "loss": 0.70816833, + "num_input_tokens_seen": 178047630, + "step": 8250, + "time_per_iteration": 2.8623669147491455 + }, + { + "auxiliary_loss_clip": 0.01150829, + "auxiliary_loss_mlp": 0.01027482, + "balance_loss_clip": 1.0420084, + "balance_loss_mlp": 1.01983464, + "epoch": 0.9921240906631396, + "flos": 21829773421440.0, + "grad_norm": 2.024630298102595, + "language_loss": 0.81788671, + "learning_rate": 6.408949170613187e-10, + "loss": 0.83966976, + "num_input_tokens_seen": 178066895, + "step": 8251, + "time_per_iteration": 2.8615224361419678 + }, + { + "auxiliary_loss_clip": 0.0114873, + "auxiliary_loss_mlp": 0.01027844, + "balance_loss_clip": 1.04368031, + "balance_loss_mlp": 1.02009547, + "epoch": 0.9922443335537786, + "flos": 24864225454080.0, + "grad_norm": 2.1052414687876975, + "language_loss": 0.82074618, + "learning_rate": 6.213277778144288e-10, + "loss": 0.84251195, + "num_input_tokens_seen": 178088540, + "step": 8252, + "time_per_iteration": 2.8451266288757324 + }, + { + "auxiliary_loss_clip": 0.01116595, + "auxiliary_loss_mlp": 0.0103196, + "balance_loss_clip": 1.03972793, + "balance_loss_mlp": 1.02362728, + "epoch": 0.9923645764444178, + "flos": 21613088626560.0, + "grad_norm": 1.952949015280277, + "language_loss": 0.66757643, + "learning_rate": 6.020639427224416e-10, + "loss": 0.689062, + "num_input_tokens_seen": 178106185, + "step": 8253, + "time_per_iteration": 2.923607110977173 + }, + { + "auxiliary_loss_clip": 0.01148903, + "auxiliary_loss_mlp": 0.01036316, + "balance_loss_clip": 1.04528904, + "balance_loss_mlp": 1.02868652, + "epoch": 0.9924848193350568, + "flos": 25001798544000.0, + "grad_norm": 2.1169001993575636, + "language_loss": 0.72394168, + "learning_rate": 5.831034147076864e-10, + "loss": 0.74579382, + "num_input_tokens_seen": 178123435, + "step": 8254, + "time_per_iteration": 2.7606310844421387 + }, + { + "auxiliary_loss_clip": 0.01048031, + "auxiliary_loss_mlp": 0.01001312, + "balance_loss_clip": 1.00473523, + "balance_loss_mlp": 0.99994165, + "epoch": 0.9926050622256959, + "flos": 68912543151360.0, + "grad_norm": 0.687934408906857, + "language_loss": 0.55749261, + "learning_rate": 5.644461966463065e-10, + "loss": 0.57798612, + "num_input_tokens_seen": 178191045, + "step": 8255, + "time_per_iteration": 3.350386619567871 + }, + { + "auxiliary_loss_clip": 0.01151043, + "auxiliary_loss_mlp": 0.01024608, + "balance_loss_clip": 1.04579103, + "balance_loss_mlp": 1.01763475, + "epoch": 0.9927253051163349, + "flos": 20923675914240.0, + "grad_norm": 5.465924438564209, + "language_loss": 0.75932771, + "learning_rate": 5.460922913687049e-10, + "loss": 0.78108424, + "num_input_tokens_seen": 178210135, + "step": 8256, + "time_per_iteration": 2.797229051589966 + }, + { + "auxiliary_loss_clip": 0.01130089, + "auxiliary_loss_mlp": 0.00886004, + "balance_loss_clip": 1.04002547, + "balance_loss_mlp": 1.00054753, + "epoch": 0.9928455480069741, + "flos": 22308211601280.0, + "grad_norm": 2.394415125387388, + "language_loss": 0.75474834, + "learning_rate": 5.280417016593208e-10, + "loss": 0.77490926, + "num_input_tokens_seen": 178229925, + "step": 8257, + "time_per_iteration": 2.8958041667938232 + }, + { + "auxiliary_loss_clip": 0.01158166, + "auxiliary_loss_mlp": 0.00885571, + "balance_loss_clip": 1.04941607, + "balance_loss_mlp": 1.00070333, + "epoch": 0.9929657908976132, + "flos": 17383889393280.0, + "grad_norm": 1.6244906339719902, + "language_loss": 0.74909508, + "learning_rate": 5.102944302559642e-10, + "loss": 0.76953244, + "num_input_tokens_seen": 178247420, + "step": 8258, + "time_per_iteration": 2.6881957054138184 + }, + { + "auxiliary_loss_clip": 0.01130108, + "auxiliary_loss_mlp": 0.01028859, + "balance_loss_clip": 1.04237831, + "balance_loss_mlp": 1.02056766, + "epoch": 0.9930860337882522, + "flos": 22674680110080.0, + "grad_norm": 2.3795913043109134, + "language_loss": 0.7968294, + "learning_rate": 4.9285047985137e-10, + "loss": 0.8184191, + "num_input_tokens_seen": 178266840, + "step": 8259, + "time_per_iteration": 3.0733964443206787 + }, + { + "auxiliary_loss_clip": 0.01164922, + "auxiliary_loss_mlp": 0.01028732, + "balance_loss_clip": 1.04802442, + "balance_loss_mlp": 1.02085817, + "epoch": 0.9932062766788914, + "flos": 28147789284480.0, + "grad_norm": 1.5746612975208094, + "language_loss": 0.74557877, + "learning_rate": 4.757098530916436e-10, + "loss": 0.7675153, + "num_input_tokens_seen": 178287285, + "step": 8260, + "time_per_iteration": 2.7687876224517822 + }, + { + "auxiliary_loss_clip": 0.01161628, + "auxiliary_loss_mlp": 0.01028262, + "balance_loss_clip": 1.04717553, + "balance_loss_mlp": 1.01982188, + "epoch": 0.9933265195695304, + "flos": 20156659868160.0, + "grad_norm": 2.775166619362736, + "language_loss": 0.77099454, + "learning_rate": 4.5887255257670563e-10, + "loss": 0.79289347, + "num_input_tokens_seen": 178304325, + "step": 8261, + "time_per_iteration": 2.751415252685547 + }, + { + "auxiliary_loss_clip": 0.01167353, + "auxiliary_loss_mlp": 0.01028108, + "balance_loss_clip": 1.04548168, + "balance_loss_mlp": 1.02027559, + "epoch": 0.9934467624601695, + "flos": 21362037494400.0, + "grad_norm": 2.0656399395699787, + "language_loss": 0.76975346, + "learning_rate": 4.4233858086117906e-10, + "loss": 0.79170811, + "num_input_tokens_seen": 178322850, + "step": 8262, + "time_per_iteration": 2.680957078933716 + }, + { + "auxiliary_loss_clip": 0.01127878, + "auxiliary_loss_mlp": 0.01028389, + "balance_loss_clip": 1.04503858, + "balance_loss_mlp": 1.02024698, + "epoch": 0.9935670053508087, + "flos": 19756040503680.0, + "grad_norm": 2.500050268072667, + "language_loss": 0.67322356, + "learning_rate": 4.261079404528356e-10, + "loss": 0.69478619, + "num_input_tokens_seen": 178342330, + "step": 8263, + "time_per_iteration": 2.863107681274414 + }, + { + "auxiliary_loss_clip": 0.01153649, + "auxiliary_loss_mlp": 0.01032644, + "balance_loss_clip": 1.04264009, + "balance_loss_mlp": 1.02472866, + "epoch": 0.9936872482414477, + "flos": 21978838863360.0, + "grad_norm": 1.8795224073068588, + "language_loss": 0.69366819, + "learning_rate": 4.1018063381437205e-10, + "loss": 0.71553111, + "num_input_tokens_seen": 178362715, + "step": 8264, + "time_per_iteration": 2.710148811340332 + }, + { + "auxiliary_loss_clip": 0.01049087, + "auxiliary_loss_mlp": 0.0100135, + "balance_loss_clip": 1.00762331, + "balance_loss_mlp": 0.99994385, + "epoch": 0.9938074911320868, + "flos": 69810667839360.0, + "grad_norm": 0.865815488189497, + "language_loss": 0.61109239, + "learning_rate": 3.9455666336141167e-10, + "loss": 0.6315968, + "num_input_tokens_seen": 178426495, + "step": 8265, + "time_per_iteration": 3.384077787399292 + }, + { + "auxiliary_loss_clip": 0.01166441, + "auxiliary_loss_mlp": 0.01025068, + "balance_loss_clip": 1.04627061, + "balance_loss_mlp": 1.01733148, + "epoch": 0.9939277340227259, + "flos": 15084170058240.0, + "grad_norm": 2.4721580666908043, + "language_loss": 0.83086634, + "learning_rate": 3.7923603146450267e-10, + "loss": 0.85278147, + "num_input_tokens_seen": 178442555, + "step": 8266, + "time_per_iteration": 2.871288299560547 + }, + { + "auxiliary_loss_clip": 0.01139085, + "auxiliary_loss_mlp": 0.01025163, + "balance_loss_clip": 1.03958762, + "balance_loss_mlp": 1.0175221, + "epoch": 0.994047976913365, + "flos": 17712364291200.0, + "grad_norm": 2.127996054947201, + "language_loss": 0.8089515, + "learning_rate": 3.642187404473418e-10, + "loss": 0.83059394, + "num_input_tokens_seen": 178460715, + "step": 8267, + "time_per_iteration": 2.7983248233795166 + }, + { + "auxiliary_loss_clip": 0.01161585, + "auxiliary_loss_mlp": 0.0102847, + "balance_loss_clip": 1.04636669, + "balance_loss_mlp": 1.02039361, + "epoch": 0.994168219804004, + "flos": 19171558396800.0, + "grad_norm": 1.9027677504365652, + "language_loss": 0.85870743, + "learning_rate": 3.495047925885508e-10, + "loss": 0.88060796, + "num_input_tokens_seen": 178479050, + "step": 8268, + "time_per_iteration": 2.7460343837738037 + }, + { + "auxiliary_loss_clip": 0.01148868, + "auxiliary_loss_mlp": 0.01025696, + "balance_loss_clip": 1.04349482, + "balance_loss_mlp": 1.01778054, + "epoch": 0.9942884626946432, + "flos": 17851589406720.0, + "grad_norm": 2.066049866755759, + "language_loss": 0.82667416, + "learning_rate": 3.350941901199e-10, + "loss": 0.84841979, + "num_input_tokens_seen": 178495970, + "step": 8269, + "time_per_iteration": 3.9619786739349365 + }, + { + "auxiliary_loss_clip": 0.01153037, + "auxiliary_loss_mlp": 0.01030831, + "balance_loss_clip": 1.04410243, + "balance_loss_mlp": 1.02260292, + "epoch": 0.9944087055852823, + "flos": 18796578364800.0, + "grad_norm": 3.046604364836254, + "language_loss": 0.83205521, + "learning_rate": 3.2098693522764066e-10, + "loss": 0.85389394, + "num_input_tokens_seen": 178509170, + "step": 8270, + "time_per_iteration": 2.7460837364196777 + }, + { + "auxiliary_loss_clip": 0.01155051, + "auxiliary_loss_mlp": 0.00885954, + "balance_loss_clip": 1.04415715, + "balance_loss_mlp": 1.00054526, + "epoch": 0.9945289484759213, + "flos": 20996969616000.0, + "grad_norm": 2.5531388661293724, + "language_loss": 0.81164277, + "learning_rate": 3.071830300516165e-10, + "loss": 0.83205283, + "num_input_tokens_seen": 178527000, + "step": 8271, + "time_per_iteration": 4.764764785766602 + }, + { + "auxiliary_loss_clip": 0.01164597, + "auxiliary_loss_mlp": 0.01025709, + "balance_loss_clip": 1.04575777, + "balance_loss_mlp": 1.01727533, + "epoch": 0.9946491913665605, + "flos": 14756952136320.0, + "grad_norm": 1.96908969816263, + "language_loss": 0.70752943, + "learning_rate": 2.9368247668615234e-10, + "loss": 0.72943252, + "num_input_tokens_seen": 178545590, + "step": 8272, + "time_per_iteration": 2.7652535438537598 + }, + { + "auxiliary_loss_clip": 0.01174514, + "auxiliary_loss_mlp": 0.01028502, + "balance_loss_clip": 1.04944789, + "balance_loss_mlp": 1.02056217, + "epoch": 0.9947694342571995, + "flos": 12669931186560.0, + "grad_norm": 2.3232935416480864, + "language_loss": 0.61624908, + "learning_rate": 2.804852771789434e-10, + "loss": 0.6382792, + "num_input_tokens_seen": 178558890, + "step": 8273, + "time_per_iteration": 2.6476194858551025 + }, + { + "auxiliary_loss_clip": 0.01164245, + "auxiliary_loss_mlp": 0.01028386, + "balance_loss_clip": 1.04378581, + "balance_loss_mlp": 1.02116179, + "epoch": 0.9948896771478386, + "flos": 18843442634880.0, + "grad_norm": 1.6871609093604965, + "language_loss": 0.55919731, + "learning_rate": 2.675914335321661e-10, + "loss": 0.58112359, + "num_input_tokens_seen": 178577645, + "step": 8274, + "time_per_iteration": 3.6345767974853516 + }, + { + "auxiliary_loss_clip": 0.01162433, + "auxiliary_loss_mlp": 0.01029859, + "balance_loss_clip": 1.04528308, + "balance_loss_mlp": 1.02157092, + "epoch": 0.9950099200384778, + "flos": 24900207903360.0, + "grad_norm": 2.5859216104021168, + "language_loss": 0.79342258, + "learning_rate": 2.550009477018111e-10, + "loss": 0.81534547, + "num_input_tokens_seen": 178596415, + "step": 8275, + "time_per_iteration": 2.69636607170105 + }, + { + "auxiliary_loss_clip": 0.01150071, + "auxiliary_loss_mlp": 0.0088586, + "balance_loss_clip": 1.04535913, + "balance_loss_mlp": 1.00058961, + "epoch": 0.9951301629291168, + "flos": 23733613987200.0, + "grad_norm": 1.8902397038918304, + "language_loss": 0.6282298, + "learning_rate": 2.4271382159790634e-10, + "loss": 0.64858913, + "num_input_tokens_seen": 178613845, + "step": 8276, + "time_per_iteration": 2.8334076404571533 + }, + { + "auxiliary_loss_clip": 0.01127294, + "auxiliary_loss_mlp": 0.01031176, + "balance_loss_clip": 1.043468, + "balance_loss_mlp": 1.02299833, + "epoch": 0.9952504058197559, + "flos": 22236893147520.0, + "grad_norm": 1.7908072463924332, + "language_loss": 0.85949111, + "learning_rate": 2.3073005708429406e-10, + "loss": 0.88107586, + "num_input_tokens_seen": 178633490, + "step": 8277, + "time_per_iteration": 2.9308512210845947 + }, + { + "auxiliary_loss_clip": 0.01137837, + "auxiliary_loss_mlp": 0.01022556, + "balance_loss_clip": 1.04691124, + "balance_loss_mlp": 1.01540351, + "epoch": 0.995370648710395, + "flos": 21211032718080.0, + "grad_norm": 1.8079506783260009, + "language_loss": 0.71754992, + "learning_rate": 2.190496559788535e-10, + "loss": 0.73915386, + "num_input_tokens_seen": 178651775, + "step": 8278, + "time_per_iteration": 2.7714521884918213 + }, + { + "auxiliary_loss_clip": 0.01149559, + "auxiliary_loss_mlp": 0.01032035, + "balance_loss_clip": 1.04535449, + "balance_loss_mlp": 1.02390504, + "epoch": 0.9954908916010341, + "flos": 14866731077760.0, + "grad_norm": 3.9870487203746836, + "language_loss": 0.76761198, + "learning_rate": 2.0767262005372265e-10, + "loss": 0.78942788, + "num_input_tokens_seen": 178669290, + "step": 8279, + "time_per_iteration": 2.8248183727264404 + }, + { + "auxiliary_loss_clip": 0.01137942, + "auxiliary_loss_mlp": 0.01029056, + "balance_loss_clip": 1.04169655, + "balance_loss_mlp": 1.02184951, + "epoch": 0.9956111344916732, + "flos": 19208259118080.0, + "grad_norm": 1.953896309809742, + "language_loss": 0.75371325, + "learning_rate": 1.965989510346322e-10, + "loss": 0.77538323, + "num_input_tokens_seen": 178688410, + "step": 8280, + "time_per_iteration": 2.7307746410369873 + }, + { + "auxiliary_loss_clip": 0.01126471, + "auxiliary_loss_mlp": 0.01027167, + "balance_loss_clip": 1.04211128, + "balance_loss_mlp": 1.01938891, + "epoch": 0.9957313773823123, + "flos": 20047060494720.0, + "grad_norm": 1.989597323263211, + "language_loss": 0.71107399, + "learning_rate": 1.8582865060134955e-10, + "loss": 0.7326104, + "num_input_tokens_seen": 178706600, + "step": 8281, + "time_per_iteration": 2.8330719470977783 + }, + { + "auxiliary_loss_clip": 0.01055016, + "auxiliary_loss_mlp": 0.01001325, + "balance_loss_clip": 1.00438321, + "balance_loss_mlp": 0.99994212, + "epoch": 0.9958516202729514, + "flos": 57483253768320.0, + "grad_norm": 1.0037738293538736, + "language_loss": 0.55768746, + "learning_rate": 1.7536172038790098e-10, + "loss": 0.57825083, + "num_input_tokens_seen": 178766910, + "step": 8282, + "time_per_iteration": 3.28952693939209 + }, + { + "auxiliary_loss_clip": 0.01148957, + "auxiliary_loss_mlp": 0.01023853, + "balance_loss_clip": 1.04495049, + "balance_loss_mlp": 1.01636696, + "epoch": 0.9959718631635904, + "flos": 27782900974080.0, + "grad_norm": 1.998428192418251, + "language_loss": 0.69616818, + "learning_rate": 1.651981619819054e-10, + "loss": 0.71789622, + "num_input_tokens_seen": 178784060, + "step": 8283, + "time_per_iteration": 2.8381619453430176 + }, + { + "auxiliary_loss_clip": 0.01134321, + "auxiliary_loss_mlp": 0.01025862, + "balance_loss_clip": 1.04219413, + "balance_loss_mlp": 1.01817894, + "epoch": 0.9960921060542296, + "flos": 24024095274240.0, + "grad_norm": 2.67199200693585, + "language_loss": 0.70717436, + "learning_rate": 1.5533797692546257e-10, + "loss": 0.72877622, + "num_input_tokens_seen": 178802795, + "step": 8284, + "time_per_iteration": 2.8963239192962646 + }, + { + "auxiliary_loss_clip": 0.01155978, + "auxiliary_loss_mlp": 0.01031506, + "balance_loss_clip": 1.04462028, + "balance_loss_mlp": 1.02341783, + "epoch": 0.9962123489448687, + "flos": 18697393935360.0, + "grad_norm": 2.011997689332027, + "language_loss": 0.84344482, + "learning_rate": 1.4578116671404296e-10, + "loss": 0.86531973, + "num_input_tokens_seen": 178821075, + "step": 8285, + "time_per_iteration": 2.713792085647583 + }, + { + "auxiliary_loss_clip": 0.01156551, + "auxiliary_loss_mlp": 0.01028512, + "balance_loss_clip": 1.04779148, + "balance_loss_mlp": 1.02041769, + "epoch": 0.9963325918355077, + "flos": 20010754823040.0, + "grad_norm": 3.120761439920452, + "language_loss": 0.71326959, + "learning_rate": 1.3652773279759777e-10, + "loss": 0.7351203, + "num_input_tokens_seen": 178837725, + "step": 8286, + "time_per_iteration": 2.7617762088775635 + }, + { + "auxiliary_loss_clip": 0.01156533, + "auxiliary_loss_mlp": 0.0102788, + "balance_loss_clip": 1.04354548, + "balance_loss_mlp": 1.0199492, + "epoch": 0.9964528347261468, + "flos": 33108488991360.0, + "grad_norm": 1.6808900195954395, + "language_loss": 0.63134539, + "learning_rate": 1.2757767657989305e-10, + "loss": 0.65318948, + "num_input_tokens_seen": 178861515, + "step": 8287, + "time_per_iteration": 2.8157384395599365 + }, + { + "auxiliary_loss_clip": 0.01159426, + "auxiliary_loss_mlp": 0.01026417, + "balance_loss_clip": 1.04714549, + "balance_loss_mlp": 1.01846576, + "epoch": 0.9965730776167859, + "flos": 23109342589440.0, + "grad_norm": 1.8315837176267786, + "language_loss": 0.87069267, + "learning_rate": 1.1893099941850948e-10, + "loss": 0.89255106, + "num_input_tokens_seen": 178880410, + "step": 8288, + "time_per_iteration": 2.835456132888794 + }, + { + "auxiliary_loss_clip": 0.01151339, + "auxiliary_loss_mlp": 0.0102525, + "balance_loss_clip": 1.0413568, + "balance_loss_mlp": 1.01754308, + "epoch": 0.996693320507425, + "flos": 22965843755520.0, + "grad_norm": 2.0447075906958463, + "language_loss": 0.77465028, + "learning_rate": 1.105877026252866e-10, + "loss": 0.79641616, + "num_input_tokens_seen": 178898740, + "step": 8289, + "time_per_iteration": 2.720250368118286 + }, + { + "auxiliary_loss_clip": 0.01169419, + "auxiliary_loss_mlp": 0.01031152, + "balance_loss_clip": 1.04607165, + "balance_loss_mlp": 1.02301002, + "epoch": 0.996813563398064, + "flos": 13222740476160.0, + "grad_norm": 1.959726039132904, + "language_loss": 0.72003579, + "learning_rate": 1.0254778746565663e-10, + "loss": 0.74204153, + "num_input_tokens_seen": 178914015, + "step": 8290, + "time_per_iteration": 2.695803642272949 + }, + { + "auxiliary_loss_clip": 0.01140031, + "auxiliary_loss_mlp": 0.01029416, + "balance_loss_clip": 1.04470217, + "balance_loss_mlp": 1.02164352, + "epoch": 0.9969338062887032, + "flos": 14647855553280.0, + "grad_norm": 2.280160227975361, + "language_loss": 0.73407567, + "learning_rate": 9.481125515953259e-11, + "loss": 0.75577021, + "num_input_tokens_seen": 178932075, + "step": 8291, + "time_per_iteration": 2.7274770736694336 + }, + { + "auxiliary_loss_clip": 0.01130406, + "auxiliary_loss_mlp": 0.01028098, + "balance_loss_clip": 1.03870487, + "balance_loss_mlp": 1.02018845, + "epoch": 0.9970540491793423, + "flos": 25735741142400.0, + "grad_norm": 3.044377461333165, + "language_loss": 0.79746085, + "learning_rate": 8.737810688064228e-11, + "loss": 0.8190459, + "num_input_tokens_seen": 178951910, + "step": 8292, + "time_per_iteration": 2.892287015914917 + }, + { + "auxiliary_loss_clip": 0.01134618, + "auxiliary_loss_mlp": 0.01033022, + "balance_loss_clip": 1.04353952, + "balance_loss_mlp": 1.02422476, + "epoch": 0.9971742920699813, + "flos": 21470236237440.0, + "grad_norm": 2.4160409825588856, + "language_loss": 0.79042065, + "learning_rate": 8.024834375608414e-11, + "loss": 0.81209707, + "num_input_tokens_seen": 178970500, + "step": 8293, + "time_per_iteration": 2.788886547088623 + }, + { + "auxiliary_loss_clip": 0.01054999, + "auxiliary_loss_mlp": 0.01001696, + "balance_loss_clip": 1.0044539, + "balance_loss_mlp": 1.00037849, + "epoch": 0.9972945349606205, + "flos": 72211223629440.0, + "grad_norm": 0.8306826620921471, + "language_loss": 0.62870383, + "learning_rate": 7.342196686788149e-11, + "loss": 0.64927077, + "num_input_tokens_seen": 179023665, + "step": 8294, + "time_per_iteration": 3.1165895462036133 + }, + { + "auxiliary_loss_clip": 0.01148354, + "auxiliary_loss_mlp": 0.01028461, + "balance_loss_clip": 1.04782438, + "balance_loss_mlp": 1.02042055, + "epoch": 0.9974147778512595, + "flos": 19678293515520.0, + "grad_norm": 2.118026282174375, + "language_loss": 0.69039637, + "learning_rate": 6.689897725142834e-11, + "loss": 0.71216452, + "num_input_tokens_seen": 179043140, + "step": 8295, + "time_per_iteration": 3.6858391761779785 + }, + { + "auxiliary_loss_clip": 0.01151883, + "auxiliary_loss_mlp": 0.01033075, + "balance_loss_clip": 1.0437156, + "balance_loss_mlp": 1.02563024, + "epoch": 0.9975350207418986, + "flos": 15960821391360.0, + "grad_norm": 2.0042288574443696, + "language_loss": 0.88343388, + "learning_rate": 6.067937589615545e-11, + "loss": 0.90528345, + "num_input_tokens_seen": 179061215, + "step": 8296, + "time_per_iteration": 2.78049373626709 + }, + { + "auxiliary_loss_clip": 0.01048155, + "auxiliary_loss_mlp": 0.01001704, + "balance_loss_clip": 1.00647879, + "balance_loss_mlp": 1.0002017, + "epoch": 0.9976552636325378, + "flos": 59961879768960.0, + "grad_norm": 0.7534744205427, + "language_loss": 0.57649899, + "learning_rate": 5.476316374575241e-11, + "loss": 0.59699756, + "num_input_tokens_seen": 179124700, + "step": 8297, + "time_per_iteration": 4.220259666442871 + }, + { + "auxiliary_loss_clip": 0.01171492, + "auxiliary_loss_mlp": 0.01040671, + "balance_loss_clip": 1.04898989, + "balance_loss_mlp": 1.03281212, + "epoch": 0.9977755065231768, + "flos": 22487872452480.0, + "grad_norm": 2.055650185344528, + "language_loss": 0.72561646, + "learning_rate": 4.9150341697723476e-11, + "loss": 0.74773812, + "num_input_tokens_seen": 179144590, + "step": 8298, + "time_per_iteration": 2.7503809928894043 + }, + { + "auxiliary_loss_clip": 0.01147098, + "auxiliary_loss_mlp": 0.01026628, + "balance_loss_clip": 1.04514217, + "balance_loss_mlp": 1.01883161, + "epoch": 0.9978957494138159, + "flos": 26030280666240.0, + "grad_norm": 1.6477548720459767, + "language_loss": 0.66569448, + "learning_rate": 4.384091060338768e-11, + "loss": 0.68743169, + "num_input_tokens_seen": 179165060, + "step": 8299, + "time_per_iteration": 2.7813189029693604 + }, + { + "auxiliary_loss_clip": 0.01155571, + "auxiliary_loss_mlp": 0.01028329, + "balance_loss_clip": 1.04471755, + "balance_loss_mlp": 1.02100348, + "epoch": 0.998015992304455, + "flos": 22637835734400.0, + "grad_norm": 2.691993986832115, + "language_loss": 0.73734581, + "learning_rate": 3.883487126810081e-11, + "loss": 0.75918484, + "num_input_tokens_seen": 179184320, + "step": 8300, + "time_per_iteration": 3.68453049659729 + }, + { + "auxiliary_loss_clip": 0.01149525, + "auxiliary_loss_mlp": 0.01024593, + "balance_loss_clip": 1.04257929, + "balance_loss_mlp": 1.01703525, + "epoch": 0.9981362351950941, + "flos": 18223444955520.0, + "grad_norm": 1.6395365948191882, + "language_loss": 0.79661131, + "learning_rate": 3.41322244516995e-11, + "loss": 0.81835252, + "num_input_tokens_seen": 179202265, + "step": 8301, + "time_per_iteration": 2.6635076999664307 + }, + { + "auxiliary_loss_clip": 0.01115768, + "auxiliary_loss_mlp": 0.01024877, + "balance_loss_clip": 1.04060578, + "balance_loss_mlp": 1.01746249, + "epoch": 0.9982564780857331, + "flos": 33474095573760.0, + "grad_norm": 1.5998464867199123, + "language_loss": 0.62742376, + "learning_rate": 2.9732970866946925e-11, + "loss": 0.64883024, + "num_input_tokens_seen": 179222145, + "step": 8302, + "time_per_iteration": 3.0004684925079346 + }, + { + "auxiliary_loss_clip": 0.01127215, + "auxiliary_loss_mlp": 0.01029294, + "balance_loss_clip": 1.03888392, + "balance_loss_mlp": 1.02023387, + "epoch": 0.9983767209763723, + "flos": 15523465392000.0, + "grad_norm": 3.350075737643154, + "language_loss": 0.77970159, + "learning_rate": 2.563711118175327e-11, + "loss": 0.80126673, + "num_input_tokens_seen": 179239030, + "step": 8303, + "time_per_iteration": 2.918919324874878 + }, + { + "auxiliary_loss_clip": 0.01133309, + "auxiliary_loss_mlp": 0.0102771, + "balance_loss_clip": 1.04176188, + "balance_loss_mlp": 1.01990795, + "epoch": 0.9984969638670114, + "flos": 19974377324160.0, + "grad_norm": 3.967875609074684, + "language_loss": 0.83745122, + "learning_rate": 2.184464601717728e-11, + "loss": 0.85906136, + "num_input_tokens_seen": 179257345, + "step": 8304, + "time_per_iteration": 2.863217830657959 + }, + { + "auxiliary_loss_clip": 0.01164043, + "auxiliary_loss_mlp": 0.01030042, + "balance_loss_clip": 1.04880369, + "balance_loss_mlp": 1.02164412, + "epoch": 0.9986172067576504, + "flos": 20375750874240.0, + "grad_norm": 2.7000613066266994, + "language_loss": 0.77240074, + "learning_rate": 1.8355575948758585e-11, + "loss": 0.79434156, + "num_input_tokens_seen": 179275330, + "step": 8305, + "time_per_iteration": 2.7671444416046143 + }, + { + "auxiliary_loss_clip": 0.01145999, + "auxiliary_loss_mlp": 0.01026812, + "balance_loss_clip": 1.0408746, + "balance_loss_mlp": 1.01823473, + "epoch": 0.9987374496482896, + "flos": 23727903724800.0, + "grad_norm": 2.1370368669632436, + "language_loss": 0.73631322, + "learning_rate": 1.5169901505407424e-11, + "loss": 0.75804132, + "num_input_tokens_seen": 179292395, + "step": 8306, + "time_per_iteration": 2.8392343521118164 + }, + { + "auxiliary_loss_clip": 0.01150065, + "auxiliary_loss_mlp": 0.01029169, + "balance_loss_clip": 1.04475677, + "balance_loss_mlp": 1.02196908, + "epoch": 0.9988576925389286, + "flos": 25044029959680.0, + "grad_norm": 1.724964495654509, + "language_loss": 0.74068618, + "learning_rate": 1.228762317073695e-11, + "loss": 0.76247847, + "num_input_tokens_seen": 179311225, + "step": 8307, + "time_per_iteration": 2.8467578887939453 + }, + { + "auxiliary_loss_clip": 0.01147718, + "auxiliary_loss_mlp": 0.01027062, + "balance_loss_clip": 1.04470778, + "balance_loss_mlp": 1.01965022, + "epoch": 0.9989779354295677, + "flos": 31285627637760.0, + "grad_norm": 2.118443867560775, + "language_loss": 0.78996348, + "learning_rate": 9.70874138195299e-12, + "loss": 0.81171125, + "num_input_tokens_seen": 179333135, + "step": 8308, + "time_per_iteration": 2.827049732208252 + }, + { + "auxiliary_loss_clip": 0.01168442, + "auxiliary_loss_mlp": 0.01027369, + "balance_loss_clip": 1.04611933, + "balance_loss_mlp": 1.01944757, + "epoch": 0.9990981783202069, + "flos": 19573398823680.0, + "grad_norm": 1.9174465705134207, + "language_loss": 0.74714279, + "learning_rate": 7.433256530076093e-12, + "loss": 0.7691009, + "num_input_tokens_seen": 179353090, + "step": 8309, + "time_per_iteration": 2.720804214477539 + }, + { + "auxiliary_loss_clip": 0.01133249, + "auxiliary_loss_mlp": 0.01025851, + "balance_loss_clip": 1.03913736, + "balance_loss_mlp": 1.01839161, + "epoch": 0.9992184212108459, + "flos": 17199667514880.0, + "grad_norm": 2.293728697746123, + "language_loss": 0.75799048, + "learning_rate": 5.46116896038562e-12, + "loss": 0.77958143, + "num_input_tokens_seen": 179367500, + "step": 8310, + "time_per_iteration": 2.8322885036468506 + }, + { + "auxiliary_loss_clip": 0.01147181, + "auxiliary_loss_mlp": 0.01032046, + "balance_loss_clip": 1.04447389, + "balance_loss_mlp": 1.02425599, + "epoch": 0.999338664101485, + "flos": 46497853681920.0, + "grad_norm": 1.8990546939993191, + "language_loss": 0.61852503, + "learning_rate": 3.792478972197699e-12, + "loss": 0.64031732, + "num_input_tokens_seen": 179388085, + "step": 8311, + "time_per_iteration": 2.903895378112793 + }, + { + "auxiliary_loss_clip": 0.01165043, + "auxiliary_loss_mlp": 0.01025442, + "balance_loss_clip": 1.04424989, + "balance_loss_mlp": 1.01776791, + "epoch": 0.9994589069921241, + "flos": 15158253859200.0, + "grad_norm": 2.3118318109743328, + "language_loss": 0.69593728, + "learning_rate": 2.4271868181990895e-12, + "loss": 0.71784216, + "num_input_tokens_seen": 179405250, + "step": 8312, + "time_per_iteration": 2.695883274078369 + }, + { + "auxiliary_loss_clip": 0.01160043, + "auxiliary_loss_mlp": 0.01020632, + "balance_loss_clip": 1.04461682, + "balance_loss_mlp": 1.01294279, + "epoch": 0.9995791498827632, + "flos": 12531460256640.0, + "grad_norm": 2.0781870283025556, + "language_loss": 0.8113817, + "learning_rate": 1.3652927060014973e-12, + "loss": 0.83318835, + "num_input_tokens_seen": 179420845, + "step": 8313, + "time_per_iteration": 2.651263952255249 + }, + { + "auxiliary_loss_clip": 0.01142681, + "auxiliary_loss_mlp": 0.01029366, + "balance_loss_clip": 1.04508209, + "balance_loss_mlp": 1.02142692, + "epoch": 0.9996993927734023, + "flos": 19245175320960.0, + "grad_norm": 2.4589534914090323, + "language_loss": 0.6353575, + "learning_rate": 6.067967965872612e-13, + "loss": 0.65707791, + "num_input_tokens_seen": 179440455, + "step": 8314, + "time_per_iteration": 2.7024452686309814 + }, + { + "auxiliary_loss_clip": 0.01144714, + "auxiliary_loss_mlp": 0.0102971, + "balance_loss_clip": 1.04780793, + "balance_loss_mlp": 1.02211297, + "epoch": 0.9998196356640414, + "flos": 62952804518400.0, + "grad_norm": 1.6224188542254347, + "language_loss": 0.76996589, + "learning_rate": 1.5169920497548615e-13, + "loss": 0.79171014, + "num_input_tokens_seen": 179465075, + "step": 8315, + "time_per_iteration": 3.071489095687866 + }, + { + "auxiliary_loss_clip": 0.01107609, + "auxiliary_loss_mlp": 0.01014138, + "balance_loss_clip": 1.02432823, + "balance_loss_mlp": 1.00961685, + "epoch": 0.9999398785546805, + "flos": 50922375073920.0, + "grad_norm": 1.1498405571635673, + "language_loss": 0.54985368, + "learning_rate": 0.0, + "loss": 0.57107115, + "num_input_tokens_seen": 179513955, + "step": 8316, + "time_per_iteration": 3.108785390853882 + }, + { + "epoch": 0.9999398785546805, + "num_input_tokens_seen": 179513955, + "step": 8316, + "total_flos": 6.996749092776837e+17, + "train_loss": 0.7895319424766712, + "train_runtime": 26258.4167, + "train_samples_per_second": 12.668, + "train_steps_per_second": 0.317 + } + ], + "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_adv_competesmoe/training_args.bin b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..1604fc8fd9a8fcf8967868a80f9dd3fde1ab382b --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_adv_competesmoe/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d03fb5c72ed40a58b38d0e99fccf33f63e1f6c5670296c86a016b5f8e83c27f3 +size 7992