Upload 12 files

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

config.json

@@ -0,0 +1,33 @@
+{
+  "_name_or_path": "google/mt5-small",
+  "architectures": [
+    "MT5ForConditionalGeneration"
+  ],
+  "classifier_dropout": 0.0,
+  "d_ff": 1024,
+  "d_kv": 64,
+  "d_model": 512,
+  "decoder_start_token_id": 0,
+  "dense_act_fn": "gelu_new",
+  "dropout_rate": 0.1,
+  "eos_token_id": 1,
+  "feed_forward_proj": "gated-gelu",
+  "initializer_factor": 1.0,
+  "is_encoder_decoder": true,
+  "is_gated_act": true,
+  "layer_norm_epsilon": 1e-06,
+  "max_length": 1024,
+  "model_type": "mt5",
+  "num_decoder_layers": 8,
+  "num_heads": 6,
+  "num_layers": 8,
+  "pad_token_id": 0,
+  "relative_attention_max_distance": 128,
+  "relative_attention_num_buckets": 32,
+  "tie_word_embeddings": false,
+  "tokenizer_class": "T5Tokenizer",
+  "torch_dtype": "float32",
+  "transformers_version": "4.37.2",
+  "use_cache": true,
+  "vocab_size": 250112
+}

generation_config.json

@@ -0,0 +1,7 @@
+{
+  "decoder_start_token_id": 0,
+  "eos_token_id": 1,
+  "max_length": 1024,
+  "pad_token_id": 0,
+  "transformers_version": "4.37.2"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c95ad1a49dd4a6a975f7124006822d255d8197df7cbc7862b7824a2a9e00db8b
+size 1200729512

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08f17f9b17033787f46b303f16f8fd40f6777f416f5e5658486ab0de566c4a05
+size 2401574330

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:307bbe15658713574c0b996b26e29048ee2e31bf283d936219e28dd7975d7832
+size 14244

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:35c5df2030891c0dfdaeabdb7935c129f217c8ddcc79ea0dd94e98567de8cdfb
+size 1064

special_tokens_map.json

@@ -0,0 +1,23 @@
+{
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<pad>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

spiece.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef78f86560d809067d12bac6c09f19a462cb3af3f54d2b8acbba26e1433125d6
+size 4309802

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:722bf80470f2ae6a7cfeaaaf876fb7ef9ad33b2e9a514ded463498f4a8bbd6be
+size 16330661

tokenizer_config.json

@@ -0,0 +1,38 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [],
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "</s>",
+  "extra_ids": 0,
+  "legacy": true,
+  "model_max_length": 1024,
+  "pad_token": "<pad>",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "T5Tokenizer",
+  "unk_token": "<unk>"
+}

trainer_state.json

@@ -0,0 +1,2461 @@
+{
+  "best_metric": 26.5388,
+  "best_model_checkpoint": "output/checkpoint-200000",
+  "epoch": 14.498006524102935,
+  "eval_steps": 50000,
+  "global_step": 200000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9951673311586325e-05,
+      "loss": 12.9031,
+      "step": 500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9903346623172648e-05,
+      "loss": 4.995,
+      "step": 1000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.985501993475897e-05,
+      "loss": 4.013,
+      "step": 1500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9806693246345298e-05,
+      "loss": 3.7433,
+      "step": 2000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9758366557931618e-05,
+      "loss": 3.5754,
+      "step": 2500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9710039869517944e-05,
+      "loss": 3.4371,
+      "step": 3000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9661713181104268e-05,
+      "loss": 3.3418,
+      "step": 3500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.961338649269059e-05,
+      "loss": 3.2642,
+      "step": 4000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9565059804276914e-05,
+      "loss": 3.2133,
+      "step": 4500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9516733115863237e-05,
+      "loss": 3.1533,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.946840642744956e-05,
+      "loss": 3.1304,
+      "step": 5500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9420079739035884e-05,
+      "loss": 3.0707,
+      "step": 6000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9371753050622207e-05,
+      "loss": 3.0091,
+      "step": 6500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.932342636220853e-05,
+      "loss": 2.9992,
+      "step": 7000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.9275099673794853e-05,
+      "loss": 2.9653,
+      "step": 7500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.922677298538118e-05,
+      "loss": 2.9463,
+      "step": 8000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.91784462969675e-05,
+      "loss": 2.9304,
+      "step": 8500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9130119608553827e-05,
+      "loss": 2.8825,
+      "step": 9000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.908179292014015e-05,
+      "loss": 2.863,
+      "step": 9500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9033466231726473e-05,
+      "loss": 2.8367,
+      "step": 10000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.8985139543312796e-05,
+      "loss": 2.8211,
+      "step": 10500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.893681285489912e-05,
+      "loss": 2.8159,
+      "step": 11000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.8888486166485443e-05,
+      "loss": 2.7894,
+      "step": 11500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.8840159478071766e-05,
+      "loss": 2.7596,
+      "step": 12000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8791832789658093e-05,
+      "loss": 2.7576,
+      "step": 12500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.8743506101244412e-05,
+      "loss": 2.7496,
+      "step": 13000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.869517941283074e-05,
+      "loss": 2.728,
+      "step": 13500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8646852724417062e-05,
+      "loss": 2.6927,
+      "step": 14000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.8598526036003382e-05,
+      "loss": 2.7037,
+      "step": 14500
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.855019934758971e-05,
+      "loss": 2.6919,
+      "step": 15000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.8501872659176032e-05,
+      "loss": 2.66,
+      "step": 15500
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.8453545970762355e-05,
+      "loss": 2.6546,
+      "step": 16000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.840521928234868e-05,
+      "loss": 2.6415,
+      "step": 16500
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.8356892593935e-05,
+      "loss": 2.6423,
+      "step": 17000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8308565905521325e-05,
+      "loss": 2.625,
+      "step": 17500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8260239217107648e-05,
+      "loss": 2.6041,
+      "step": 18000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.8211912528693975e-05,
+      "loss": 2.5971,
+      "step": 18500
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.8163585840280295e-05,
+      "loss": 2.5993,
+      "step": 19000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.811525915186662e-05,
+      "loss": 2.5707,
+      "step": 19500
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.8066932463452944e-05,
+      "loss": 2.5933,
+      "step": 20000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.8018605775039268e-05,
+      "loss": 2.5669,
+      "step": 20500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.797027908662559e-05,
+      "loss": 2.5681,
+      "step": 21000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.7921952398211914e-05,
+      "loss": 2.5648,
+      "step": 21500
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.7873625709798237e-05,
+      "loss": 2.5403,
+      "step": 22000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.782529902138456e-05,
+      "loss": 2.5314,
+      "step": 22500
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.7776972332970884e-05,
+      "loss": 2.5094,
+      "step": 23000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.7728645644557207e-05,
+      "loss": 2.537,
+      "step": 23500
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.768031895614353e-05,
+      "loss": 2.5389,
+      "step": 24000
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.7631992267729857e-05,
+      "loss": 2.5243,
+      "step": 24500
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.7583665579316177e-05,
+      "loss": 2.5064,
+      "step": 25000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.7535338890902503e-05,
+      "loss": 2.5106,
+      "step": 25500
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.7487012202488827e-05,
+      "loss": 2.4955,
+      "step": 26000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.743868551407515e-05,
+      "loss": 2.488,
+      "step": 26500
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7390358825661473e-05,
+      "loss": 2.4869,
+      "step": 27000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7342032137247796e-05,
+      "loss": 2.4838,
+      "step": 27500
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.729370544883412e-05,
+      "loss": 2.4756,
+      "step": 28000
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.7245378760420443e-05,
+      "loss": 2.4512,
+      "step": 28500
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.719705207200677e-05,
+      "loss": 2.4456,
+      "step": 29000
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.714872538359309e-05,
+      "loss": 2.4485,
+      "step": 29500
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.7100398695179416e-05,
+      "loss": 2.4304,
+      "step": 30000
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.705207200676574e-05,
+      "loss": 2.4417,
+      "step": 30500
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.7003745318352062e-05,
+      "loss": 2.43,
+      "step": 31000
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.6955418629938385e-05,
+      "loss": 2.443,
+      "step": 31500
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.690709194152471e-05,
+      "loss": 2.4195,
+      "step": 32000
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.6858765253111032e-05,
+      "loss": 2.4103,
+      "step": 32500
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.6810438564697355e-05,
+      "loss": 2.4133,
+      "step": 33000
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.676211187628368e-05,
+      "loss": 2.4059,
+      "step": 33500
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.671378518787e-05,
+      "loss": 2.4059,
+      "step": 34000
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.6665458499456325e-05,
+      "loss": 2.4109,
+      "step": 34500
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.661713181104265e-05,
+      "loss": 2.4025,
+      "step": 35000
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.656880512262897e-05,
+      "loss": 2.397,
+      "step": 35500
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.6520478434215298e-05,
+      "loss": 2.3958,
+      "step": 36000
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.647215174580162e-05,
+      "loss": 2.3681,
+      "step": 36500
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.6423825057387944e-05,
+      "loss": 2.4033,
+      "step": 37000
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.6375498368974268e-05,
+      "loss": 2.3801,
+      "step": 37500
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.632717168056059e-05,
+      "loss": 2.3898,
+      "step": 38000
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.6278844992146914e-05,
+      "loss": 2.3613,
+      "step": 38500
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.6230518303733237e-05,
+      "loss": 2.3879,
+      "step": 39000
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.6182191615319564e-05,
+      "loss": 2.362,
+      "step": 39500
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.6133864926905884e-05,
+      "loss": 2.3625,
+      "step": 40000
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.608553823849221e-05,
+      "loss": 2.3635,
+      "step": 40500
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.6037211550078534e-05,
+      "loss": 2.3615,
+      "step": 41000
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.5988884861664853e-05,
+      "loss": 2.3327,
+      "step": 41500
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.594055817325118e-05,
+      "loss": 2.3424,
+      "step": 42000
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.5892231484837503e-05,
+      "loss": 2.3462,
+      "step": 42500
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.5843904796423827e-05,
+      "loss": 2.3395,
+      "step": 43000
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.579557810801015e-05,
+      "loss": 2.3287,
+      "step": 43500
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.5747251419596473e-05,
+      "loss": 2.3408,
+      "step": 44000
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.5698924731182796e-05,
+      "loss": 2.3126,
+      "step": 44500
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.565059804276912e-05,
+      "loss": 2.3234,
+      "step": 45000
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.5602271354355446e-05,
+      "loss": 2.324,
+      "step": 45500
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.5553944665941766e-05,
+      "loss": 2.3156,
+      "step": 46000
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.5505617977528093e-05,
+      "loss": 2.3264,
+      "step": 46500
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.5457291289114416e-05,
+      "loss": 2.3086,
+      "step": 47000
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.540896460070074e-05,
+      "loss": 2.3095,
+      "step": 47500
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 1.5360637912287062e-05,
+      "loss": 2.3015,
+      "step": 48000
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 1.5312311223873385e-05,
+      "loss": 2.302,
+      "step": 48500
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 1.526398453545971e-05,
+      "loss": 2.3164,
+      "step": 49000
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 1.5215657847046032e-05,
+      "loss": 2.3067,
+      "step": 49500
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 1.5167331158632357e-05,
+      "loss": 2.2705,
+      "step": 50000
+    },
+    {
+      "epoch": 3.62,
+      "eval_bleu": 23.2646,
+      "eval_gen_len": 26.1331,
+      "eval_loss": 1.9072902202606201,
+      "eval_runtime": 5550.0392,
+      "eval_samples_per_second": 9.942,
+      "eval_steps_per_second": 1.243,
+      "step": 50000
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1.5119004470218678e-05,
+      "loss": 2.2769,
+      "step": 50500
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1.5070677781805003e-05,
+      "loss": 2.3089,
+      "step": 51000
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.5022351093391328e-05,
+      "loss": 2.2817,
+      "step": 51500
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.497402440497765e-05,
+      "loss": 2.2757,
+      "step": 52000
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 1.4925697716563973e-05,
+      "loss": 2.287,
+      "step": 52500
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 1.4877371028150298e-05,
+      "loss": 2.289,
+      "step": 53000
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 1.482904433973662e-05,
+      "loss": 2.268,
+      "step": 53500
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.4780717651322944e-05,
+      "loss": 2.2766,
+      "step": 54000
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 1.473239096290927e-05,
+      "loss": 2.2775,
+      "step": 54500
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 1.4684064274495591e-05,
+      "loss": 2.2795,
+      "step": 55000
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 1.4635737586081916e-05,
+      "loss": 2.2683,
+      "step": 55500
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 1.4587410897668239e-05,
+      "loss": 2.2551,
+      "step": 56000
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 1.4539084209254562e-05,
+      "loss": 2.2511,
+      "step": 56500
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 1.4490757520840885e-05,
+      "loss": 2.258,
+      "step": 57000
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.444243083242721e-05,
+      "loss": 2.2593,
+      "step": 57500
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1.4394104144013532e-05,
+      "loss": 2.2481,
+      "step": 58000
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 1.4345777455599857e-05,
+      "loss": 2.2485,
+      "step": 58500
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 1.429745076718618e-05,
+      "loss": 2.2594,
+      "step": 59000
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 1.4249124078772503e-05,
+      "loss": 2.2593,
+      "step": 59500
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 1.4200797390358827e-05,
+      "loss": 2.242,
+      "step": 60000
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 1.4152470701945151e-05,
+      "loss": 2.2294,
+      "step": 60500
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 1.4104144013531473e-05,
+      "loss": 2.2517,
+      "step": 61000
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 1.4055817325117798e-05,
+      "loss": 2.2243,
+      "step": 61500
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 1.4007490636704121e-05,
+      "loss": 2.2484,
+      "step": 62000
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 1.3959163948290444e-05,
+      "loss": 2.228,
+      "step": 62500
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 1.3910837259876768e-05,
+      "loss": 2.2314,
+      "step": 63000
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1.3862510571463093e-05,
+      "loss": 2.2317,
+      "step": 63500
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 1.3814183883049414e-05,
+      "loss": 2.2187,
+      "step": 64000
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 1.3765857194635739e-05,
+      "loss": 2.2325,
+      "step": 64500
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1.3717530506222064e-05,
+      "loss": 2.2221,
+      "step": 65000
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1.3669203817808385e-05,
+      "loss": 2.2218,
+      "step": 65500
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1.3620877129394709e-05,
+      "loss": 2.2156,
+      "step": 66000
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.3572550440981034e-05,
+      "loss": 2.2232,
+      "step": 66500
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 1.3524223752567355e-05,
+      "loss": 2.2084,
+      "step": 67000
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 1.347589706415368e-05,
+      "loss": 2.2374,
+      "step": 67500
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 1.3427570375740005e-05,
+      "loss": 2.2046,
+      "step": 68000
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.3379243687326327e-05,
+      "loss": 2.2109,
+      "step": 68500
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 1.3330916998912651e-05,
+      "loss": 2.2115,
+      "step": 69000
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 1.3282590310498975e-05,
+      "loss": 2.2053,
+      "step": 69500
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 1.3234263622085298e-05,
+      "loss": 2.1956,
+      "step": 70000
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 1.3185936933671621e-05,
+      "loss": 2.2031,
+      "step": 70500
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 1.3137610245257946e-05,
+      "loss": 2.1959,
+      "step": 71000
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 1.3089283556844268e-05,
+      "loss": 2.2023,
+      "step": 71500
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 1.3040956868430593e-05,
+      "loss": 2.197,
+      "step": 72000
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 1.2992630180016916e-05,
+      "loss": 2.212,
+      "step": 72500
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 1.2944303491603239e-05,
+      "loss": 2.204,
+      "step": 73000
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 1.2895976803189562e-05,
+      "loss": 2.1871,
+      "step": 73500
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 1.2847650114775887e-05,
+      "loss": 2.1852,
+      "step": 74000
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 1.2799323426362209e-05,
+      "loss": 2.1804,
+      "step": 74500
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 1.2750996737948534e-05,
+      "loss": 2.1949,
+      "step": 75000
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 1.2702670049534857e-05,
+      "loss": 2.1662,
+      "step": 75500
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 1.265434336112118e-05,
+      "loss": 2.1871,
+      "step": 76000
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 1.2606016672707503e-05,
+      "loss": 2.1841,
+      "step": 76500
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 1.2557689984293828e-05,
+      "loss": 2.1719,
+      "step": 77000
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 1.250936329588015e-05,
+      "loss": 2.1755,
+      "step": 77500
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 1.2461036607466475e-05,
+      "loss": 2.1589,
+      "step": 78000
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 1.24127099190528e-05,
+      "loss": 2.1785,
+      "step": 78500
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 1.2364383230639121e-05,
+      "loss": 2.1931,
+      "step": 79000
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 1.2316056542225444e-05,
+      "loss": 2.1827,
+      "step": 79500
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 1.226772985381177e-05,
+      "loss": 2.1628,
+      "step": 80000
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 1.2219403165398091e-05,
+      "loss": 2.1702,
+      "step": 80500
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 1.2171076476984416e-05,
+      "loss": 2.161,
+      "step": 81000
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 1.212274978857074e-05,
+      "loss": 2.172,
+      "step": 81500
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 1.2074423100157062e-05,
+      "loss": 2.1731,
+      "step": 82000
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 1.2026096411743387e-05,
+      "loss": 2.1582,
+      "step": 82500
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 1.197776972332971e-05,
+      "loss": 2.1739,
+      "step": 83000
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 1.1929443034916034e-05,
+      "loss": 2.1447,
+      "step": 83500
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 1.1881116346502357e-05,
+      "loss": 2.159,
+      "step": 84000
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 1.1832789658088682e-05,
+      "loss": 2.17,
+      "step": 84500
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 1.1784462969675003e-05,
+      "loss": 2.1441,
+      "step": 85000
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 1.1736136281261328e-05,
+      "loss": 2.1447,
+      "step": 85500
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 1.1687809592847651e-05,
+      "loss": 2.1529,
+      "step": 86000
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 1.1639482904433975e-05,
+      "loss": 2.1558,
+      "step": 86500
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 1.1591156216020298e-05,
+      "loss": 2.1356,
+      "step": 87000
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 1.1542829527606623e-05,
+      "loss": 2.1593,
+      "step": 87500
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 1.1494502839192944e-05,
+      "loss": 2.1467,
+      "step": 88000
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 1.144617615077927e-05,
+      "loss": 2.1533,
+      "step": 88500
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 1.1397849462365593e-05,
+      "loss": 2.1261,
+      "step": 89000
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 1.1349522773951916e-05,
+      "loss": 2.1519,
+      "step": 89500
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 1.1301196085538239e-05,
+      "loss": 2.1513,
+      "step": 90000
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 1.1252869397124564e-05,
+      "loss": 2.1404,
+      "step": 90500
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 1.1204542708710885e-05,
+      "loss": 2.1427,
+      "step": 91000
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 1.115621602029721e-05,
+      "loss": 2.138,
+      "step": 91500
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 1.1107889331883535e-05,
+      "loss": 2.1382,
+      "step": 92000
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 1.1059562643469857e-05,
+      "loss": 2.1344,
+      "step": 92500
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 1.101123595505618e-05,
+      "loss": 2.1474,
+      "step": 93000
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 1.0962909266642505e-05,
+      "loss": 2.1197,
+      "step": 93500
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 1.0914582578228827e-05,
+      "loss": 2.146,
+      "step": 94000
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 1.0866255889815151e-05,
+      "loss": 2.1313,
+      "step": 94500
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 1.0817929201401476e-05,
+      "loss": 2.1374,
+      "step": 95000
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 1.0769602512987798e-05,
+      "loss": 2.1329,
+      "step": 95500
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 1.0721275824574123e-05,
+      "loss": 2.132,
+      "step": 96000
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 1.0672949136160446e-05,
+      "loss": 2.1142,
+      "step": 96500
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 1.062462244774677e-05,
+      "loss": 2.1338,
+      "step": 97000
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 1.0576295759333093e-05,
+      "loss": 2.1137,
+      "step": 97500
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 1.0527969070919417e-05,
+      "loss": 2.1221,
+      "step": 98000
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 1.0479642382505739e-05,
+      "loss": 2.1382,
+      "step": 98500
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 1.0431315694092064e-05,
+      "loss": 2.1207,
+      "step": 99000
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 1.0382989005678387e-05,
+      "loss": 2.1163,
+      "step": 99500
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 1.033466231726471e-05,
+      "loss": 2.1263,
+      "step": 100000
+    },
+    {
+      "epoch": 7.25,
+      "eval_bleu": 25.2752,
+      "eval_gen_len": 24.8098,
+      "eval_loss": 1.8044791221618652,
+      "eval_runtime": 4767.4438,
+      "eval_samples_per_second": 11.574,
+      "eval_steps_per_second": 1.447,
+      "step": 100000
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 1.0286335628851034e-05,
+      "loss": 2.0977,
+      "step": 100500
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 1.0238008940437359e-05,
+      "loss": 2.1114,
+      "step": 101000
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 1.018968225202368e-05,
+      "loss": 2.1208,
+      "step": 101500
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 1.0141355563610005e-05,
+      "loss": 2.1006,
+      "step": 102000
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 1.0093028875196328e-05,
+      "loss": 2.1091,
+      "step": 102500
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 1.0044702186782651e-05,
+      "loss": 2.1347,
+      "step": 103000
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 9.996375498368975e-06,
+      "loss": 2.1106,
+      "step": 103500
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 9.948048809955298e-06,
+      "loss": 2.1071,
+      "step": 104000
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 9.899722121541621e-06,
+      "loss": 2.0954,
+      "step": 104500
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 9.851395433127946e-06,
+      "loss": 2.1072,
+      "step": 105000
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 9.80306874471427e-06,
+      "loss": 2.1051,
+      "step": 105500
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 9.754742056300593e-06,
+      "loss": 2.1066,
+      "step": 106000
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 9.706415367886916e-06,
+      "loss": 2.1122,
+      "step": 106500
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 9.658088679473239e-06,
+      "loss": 2.0952,
+      "step": 107000
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 9.609761991059562e-06,
+      "loss": 2.085,
+      "step": 107500
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 9.561435302645887e-06,
+      "loss": 2.1076,
+      "step": 108000
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 9.51310861423221e-06,
+      "loss": 2.1236,
+      "step": 108500
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 9.464781925818534e-06,
+      "loss": 2.0932,
+      "step": 109000
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 9.416455237404859e-06,
+      "loss": 2.0892,
+      "step": 109500
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 9.368128548991182e-06,
+      "loss": 2.0778,
+      "step": 110000
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 9.319801860577505e-06,
+      "loss": 2.1035,
+      "step": 110500
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 9.271475172163828e-06,
+      "loss": 2.0876,
+      "step": 111000
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 9.223148483750151e-06,
+      "loss": 2.0869,
+      "step": 111500
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 9.174821795336475e-06,
+      "loss": 2.0972,
+      "step": 112000
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 9.1264951069228e-06,
+      "loss": 2.0948,
+      "step": 112500
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 9.078168418509123e-06,
+      "loss": 2.0742,
+      "step": 113000
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 9.029841730095446e-06,
+      "loss": 2.085,
+      "step": 113500
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 8.98151504168177e-06,
+      "loss": 2.0885,
+      "step": 114000
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 8.933188353268093e-06,
+      "loss": 2.0667,
+      "step": 114500
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 8.884861664854416e-06,
+      "loss": 2.1004,
+      "step": 115000
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 8.83653497644074e-06,
+      "loss": 2.099,
+      "step": 115500
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 8.788208288027064e-06,
+      "loss": 2.0988,
+      "step": 116000
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 8.739881599613387e-06,
+      "loss": 2.0861,
+      "step": 116500
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 8.69155491119971e-06,
+      "loss": 2.0722,
+      "step": 117000
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 8.643228222786034e-06,
+      "loss": 2.0951,
+      "step": 117500
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 8.594901534372357e-06,
+      "loss": 2.0915,
+      "step": 118000
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 8.546574845958682e-06,
+      "loss": 2.099,
+      "step": 118500
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 8.498248157545005e-06,
+      "loss": 2.0805,
+      "step": 119000
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 8.449921469131328e-06,
+      "loss": 2.0857,
+      "step": 119500
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 8.401594780717651e-06,
+      "loss": 2.0849,
+      "step": 120000
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 8.353268092303975e-06,
+      "loss": 2.0936,
+      "step": 120500
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 8.304941403890298e-06,
+      "loss": 2.0781,
+      "step": 121000
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 8.256614715476623e-06,
+      "loss": 2.0824,
+      "step": 121500
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 8.208288027062946e-06,
+      "loss": 2.0824,
+      "step": 122000
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 8.15996133864927e-06,
+      "loss": 2.0822,
+      "step": 122500
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 8.111634650235594e-06,
+      "loss": 2.0873,
+      "step": 123000
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 8.063307961821917e-06,
+      "loss": 2.0679,
+      "step": 123500
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 8.01498127340824e-06,
+      "loss": 2.077,
+      "step": 124000
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 7.966654584994564e-06,
+      "loss": 2.0701,
+      "step": 124500
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 7.918327896580887e-06,
+      "loss": 2.0566,
+      "step": 125000
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 7.87000120816721e-06,
+      "loss": 2.0904,
+      "step": 125500
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 7.821674519753535e-06,
+      "loss": 2.0787,
+      "step": 126000
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 7.773347831339859e-06,
+      "loss": 2.0832,
+      "step": 126500
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 7.725021142926182e-06,
+      "loss": 2.0615,
+      "step": 127000
+    },
+    {
+      "epoch": 9.24,
+      "learning_rate": 7.676694454512505e-06,
+      "loss": 2.0707,
+      "step": 127500
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 7.628367766098828e-06,
+      "loss": 2.0756,
+      "step": 128000
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 7.5800410776851515e-06,
+      "loss": 2.0687,
+      "step": 128500
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 7.531714389271476e-06,
+      "loss": 2.0657,
+      "step": 129000
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 7.4833877008578e-06,
+      "loss": 2.0559,
+      "step": 129500
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 7.435061012444123e-06,
+      "loss": 2.0727,
+      "step": 130000
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 7.386734324030447e-06,
+      "loss": 2.0532,
+      "step": 130500
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 7.33840763561677e-06,
+      "loss": 2.0783,
+      "step": 131000
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 7.290080947203093e-06,
+      "loss": 2.0657,
+      "step": 131500
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 7.2417542587894175e-06,
+      "loss": 2.0554,
+      "step": 132000
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 7.193427570375741e-06,
+      "loss": 2.0649,
+      "step": 132500
+    },
+    {
+      "epoch": 9.64,
+      "learning_rate": 7.145100881962064e-06,
+      "loss": 2.0642,
+      "step": 133000
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 2.0416,
+      "step": 133500
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 7.048447505134711e-06,
+      "loss": 2.0413,
+      "step": 134000
+    },
+    {
+      "epoch": 9.75,
+      "learning_rate": 7.0001208167210345e-06,
+      "loss": 2.0675,
+      "step": 134500
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 6.9517941283073586e-06,
+      "loss": 2.034,
+      "step": 135000
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 6.903467439893682e-06,
+      "loss": 2.0619,
+      "step": 135500
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 6.855140751480005e-06,
+      "loss": 2.0584,
+      "step": 136000
+    },
+    {
+      "epoch": 9.89,
+      "learning_rate": 6.806814063066329e-06,
+      "loss": 2.0691,
+      "step": 136500
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 6.758487374652652e-06,
+      "loss": 2.0519,
+      "step": 137000
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 6.7101606862389756e-06,
+      "loss": 2.0675,
+      "step": 137500
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 6.6618339978253e-06,
+      "loss": 2.0675,
+      "step": 138000
+    },
+    {
+      "epoch": 10.04,
+      "learning_rate": 6.613507309411623e-06,
+      "loss": 2.049,
+      "step": 138500
+    },
+    {
+      "epoch": 10.08,
+      "learning_rate": 6.565180620997946e-06,
+      "loss": 2.0531,
+      "step": 139000
+    },
+    {
+      "epoch": 10.11,
+      "learning_rate": 6.51685393258427e-06,
+      "loss": 2.0482,
+      "step": 139500
+    },
+    {
+      "epoch": 10.15,
+      "learning_rate": 6.468527244170593e-06,
+      "loss": 2.0685,
+      "step": 140000
+    },
+    {
+      "epoch": 10.18,
+      "learning_rate": 6.420200555756917e-06,
+      "loss": 2.0612,
+      "step": 140500
+    },
+    {
+      "epoch": 10.22,
+      "learning_rate": 6.3718738673432416e-06,
+      "loss": 2.0456,
+      "step": 141000
+    },
+    {
+      "epoch": 10.26,
+      "learning_rate": 6.323547178929564e-06,
+      "loss": 2.0311,
+      "step": 141500
+    },
+    {
+      "epoch": 10.29,
+      "learning_rate": 6.275220490515887e-06,
+      "loss": 2.0582,
+      "step": 142000
+    },
+    {
+      "epoch": 10.33,
+      "learning_rate": 6.226893802102212e-06,
+      "loss": 2.0563,
+      "step": 142500
+    },
+    {
+      "epoch": 10.37,
+      "learning_rate": 6.178567113688535e-06,
+      "loss": 2.0671,
+      "step": 143000
+    },
+    {
+      "epoch": 10.4,
+      "learning_rate": 6.1302404252748586e-06,
+      "loss": 2.0511,
+      "step": 143500
+    },
+    {
+      "epoch": 10.44,
+      "learning_rate": 6.081913736861183e-06,
+      "loss": 2.0436,
+      "step": 144000
+    },
+    {
+      "epoch": 10.47,
+      "learning_rate": 6.033587048447506e-06,
+      "loss": 2.0459,
+      "step": 144500
+    },
+    {
+      "epoch": 10.51,
+      "learning_rate": 5.985260360033829e-06,
+      "loss": 2.0462,
+      "step": 145000
+    },
+    {
+      "epoch": 10.55,
+      "learning_rate": 5.936933671620153e-06,
+      "loss": 2.0258,
+      "step": 145500
+    },
+    {
+      "epoch": 10.58,
+      "learning_rate": 5.888606983206476e-06,
+      "loss": 2.0513,
+      "step": 146000
+    },
+    {
+      "epoch": 10.62,
+      "learning_rate": 5.8402802947928e-06,
+      "loss": 2.0566,
+      "step": 146500
+    },
+    {
+      "epoch": 10.66,
+      "learning_rate": 5.791953606379124e-06,
+      "loss": 2.0485,
+      "step": 147000
+    },
+    {
+      "epoch": 10.69,
+      "learning_rate": 5.743626917965447e-06,
+      "loss": 2.0351,
+      "step": 147500
+    },
+    {
+      "epoch": 10.73,
+      "learning_rate": 5.69530022955177e-06,
+      "loss": 2.0449,
+      "step": 148000
+    },
+    {
+      "epoch": 10.76,
+      "learning_rate": 5.646973541138094e-06,
+      "loss": 2.0635,
+      "step": 148500
+    },
+    {
+      "epoch": 10.8,
+      "learning_rate": 5.5986468527244175e-06,
+      "loss": 2.0325,
+      "step": 149000
+    },
+    {
+      "epoch": 10.84,
+      "learning_rate": 5.550320164310741e-06,
+      "loss": 2.0581,
+      "step": 149500
+    },
+    {
+      "epoch": 10.87,
+      "learning_rate": 5.501993475897065e-06,
+      "loss": 2.0223,
+      "step": 150000
+    },
+    {
+      "epoch": 10.87,
+      "eval_bleu": 26.2063,
+      "eval_gen_len": 24.6095,
+      "eval_loss": 1.7641891241073608,
+      "eval_runtime": 4661.5713,
+      "eval_samples_per_second": 11.837,
+      "eval_steps_per_second": 1.48,
+      "step": 150000
+    },
+    {
+      "epoch": 10.91,
+      "learning_rate": 5.453666787483388e-06,
+      "loss": 2.0329,
+      "step": 150500
+    },
+    {
+      "epoch": 10.95,
+      "learning_rate": 5.405340099069711e-06,
+      "loss": 2.0588,
+      "step": 151000
+    },
+    {
+      "epoch": 10.98,
+      "learning_rate": 5.357013410656035e-06,
+      "loss": 2.0424,
+      "step": 151500
+    },
+    {
+      "epoch": 11.02,
+      "learning_rate": 5.3086867222423586e-06,
+      "loss": 2.0397,
+      "step": 152000
+    },
+    {
+      "epoch": 11.05,
+      "learning_rate": 5.260360033828682e-06,
+      "loss": 2.0162,
+      "step": 152500
+    },
+    {
+      "epoch": 11.09,
+      "learning_rate": 5.212033345415006e-06,
+      "loss": 2.0444,
+      "step": 153000
+    },
+    {
+      "epoch": 11.13,
+      "learning_rate": 5.163706657001329e-06,
+      "loss": 2.0421,
+      "step": 153500
+    },
+    {
+      "epoch": 11.16,
+      "learning_rate": 5.115379968587652e-06,
+      "loss": 2.0251,
+      "step": 154000
+    },
+    {
+      "epoch": 11.2,
+      "learning_rate": 5.067053280173977e-06,
+      "loss": 2.0599,
+      "step": 154500
+    },
+    {
+      "epoch": 11.24,
+      "learning_rate": 5.0187265917603005e-06,
+      "loss": 2.0569,
+      "step": 155000
+    },
+    {
+      "epoch": 11.27,
+      "learning_rate": 4.970399903346624e-06,
+      "loss": 2.0439,
+      "step": 155500
+    },
+    {
+      "epoch": 11.31,
+      "learning_rate": 4.922073214932947e-06,
+      "loss": 2.0327,
+      "step": 156000
+    },
+    {
+      "epoch": 11.34,
+      "learning_rate": 4.873746526519271e-06,
+      "loss": 2.0487,
+      "step": 156500
+    },
+    {
+      "epoch": 11.38,
+      "learning_rate": 4.825419838105594e-06,
+      "loss": 2.0561,
+      "step": 157000
+    },
+    {
+      "epoch": 11.42,
+      "learning_rate": 4.7770931496919175e-06,
+      "loss": 2.0341,
+      "step": 157500
+    },
+    {
+      "epoch": 11.45,
+      "learning_rate": 4.7287664612782416e-06,
+      "loss": 2.0343,
+      "step": 158000
+    },
+    {
+      "epoch": 11.49,
+      "learning_rate": 4.680439772864565e-06,
+      "loss": 2.0393,
+      "step": 158500
+    },
+    {
+      "epoch": 11.53,
+      "learning_rate": 4.632113084450888e-06,
+      "loss": 2.0165,
+      "step": 159000
+    },
+    {
+      "epoch": 11.56,
+      "learning_rate": 4.583786396037212e-06,
+      "loss": 2.0432,
+      "step": 159500
+    },
+    {
+      "epoch": 11.6,
+      "learning_rate": 4.535459707623535e-06,
+      "loss": 2.0254,
+      "step": 160000
+    },
+    {
+      "epoch": 11.63,
+      "learning_rate": 4.4871330192098586e-06,
+      "loss": 2.0027,
+      "step": 160500
+    },
+    {
+      "epoch": 11.67,
+      "learning_rate": 4.438806330796183e-06,
+      "loss": 2.0255,
+      "step": 161000
+    },
+    {
+      "epoch": 11.71,
+      "learning_rate": 4.390479642382507e-06,
+      "loss": 2.0347,
+      "step": 161500
+    },
+    {
+      "epoch": 11.74,
+      "learning_rate": 4.342152953968829e-06,
+      "loss": 2.0265,
+      "step": 162000
+    },
+    {
+      "epoch": 11.78,
+      "learning_rate": 4.293826265555153e-06,
+      "loss": 2.0319,
+      "step": 162500
+    },
+    {
+      "epoch": 11.82,
+      "learning_rate": 4.245499577141477e-06,
+      "loss": 2.0332,
+      "step": 163000
+    },
+    {
+      "epoch": 11.85,
+      "learning_rate": 4.1971728887278005e-06,
+      "loss": 2.0418,
+      "step": 163500
+    },
+    {
+      "epoch": 11.89,
+      "learning_rate": 4.148846200314124e-06,
+      "loss": 2.0511,
+      "step": 164000
+    },
+    {
+      "epoch": 11.92,
+      "learning_rate": 4.100519511900448e-06,
+      "loss": 2.0353,
+      "step": 164500
+    },
+    {
+      "epoch": 11.96,
+      "learning_rate": 4.052192823486771e-06,
+      "loss": 2.0393,
+      "step": 165000
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 4.003866135073094e-06,
+      "loss": 2.016,
+      "step": 165500
+    },
+    {
+      "epoch": 12.03,
+      "learning_rate": 3.955539446659418e-06,
+      "loss": 2.0318,
+      "step": 166000
+    },
+    {
+      "epoch": 12.07,
+      "learning_rate": 3.9072127582457416e-06,
+      "loss": 2.0342,
+      "step": 166500
+    },
+    {
+      "epoch": 12.11,
+      "learning_rate": 3.858886069832065e-06,
+      "loss": 1.9975,
+      "step": 167000
+    },
+    {
+      "epoch": 12.14,
+      "learning_rate": 3.810559381418389e-06,
+      "loss": 2.0222,
+      "step": 167500
+    },
+    {
+      "epoch": 12.18,
+      "learning_rate": 3.762232693004712e-06,
+      "loss": 2.0455,
+      "step": 168000
+    },
+    {
+      "epoch": 12.21,
+      "learning_rate": 3.7139060045910358e-06,
+      "loss": 2.028,
+      "step": 168500
+    },
+    {
+      "epoch": 12.25,
+      "learning_rate": 3.6655793161773594e-06,
+      "loss": 2.0267,
+      "step": 169000
+    },
+    {
+      "epoch": 12.29,
+      "learning_rate": 3.6172526277636827e-06,
+      "loss": 2.0249,
+      "step": 169500
+    },
+    {
+      "epoch": 12.32,
+      "learning_rate": 3.5689259393500063e-06,
+      "loss": 2.0451,
+      "step": 170000
+    },
+    {
+      "epoch": 12.36,
+      "learning_rate": 3.52059925093633e-06,
+      "loss": 2.0133,
+      "step": 170500
+    },
+    {
+      "epoch": 12.4,
+      "learning_rate": 3.472272562522653e-06,
+      "loss": 2.0377,
+      "step": 171000
+    },
+    {
+      "epoch": 12.43,
+      "learning_rate": 3.423945874108977e-06,
+      "loss": 2.024,
+      "step": 171500
+    },
+    {
+      "epoch": 12.47,
+      "learning_rate": 3.375619185695301e-06,
+      "loss": 2.0368,
+      "step": 172000
+    },
+    {
+      "epoch": 12.5,
+      "learning_rate": 3.327292497281624e-06,
+      "loss": 2.0463,
+      "step": 172500
+    },
+    {
+      "epoch": 12.54,
+      "learning_rate": 3.278965808867948e-06,
+      "loss": 2.0359,
+      "step": 173000
+    },
+    {
+      "epoch": 12.58,
+      "learning_rate": 3.230639120454271e-06,
+      "loss": 2.0283,
+      "step": 173500
+    },
+    {
+      "epoch": 12.61,
+      "learning_rate": 3.1823124320405947e-06,
+      "loss": 2.035,
+      "step": 174000
+    },
+    {
+      "epoch": 12.65,
+      "learning_rate": 3.1339857436269183e-06,
+      "loss": 2.0163,
+      "step": 174500
+    },
+    {
+      "epoch": 12.69,
+      "learning_rate": 3.0856590552132416e-06,
+      "loss": 2.0234,
+      "step": 175000
+    },
+    {
+      "epoch": 12.72,
+      "learning_rate": 3.0373323667995652e-06,
+      "loss": 2.0273,
+      "step": 175500
+    },
+    {
+      "epoch": 12.76,
+      "learning_rate": 2.989005678385889e-06,
+      "loss": 2.0323,
+      "step": 176000
+    },
+    {
+      "epoch": 12.79,
+      "learning_rate": 2.940678989972212e-06,
+      "loss": 2.019,
+      "step": 176500
+    },
+    {
+      "epoch": 12.83,
+      "learning_rate": 2.8923523015585358e-06,
+      "loss": 2.0145,
+      "step": 177000
+    },
+    {
+      "epoch": 12.87,
+      "learning_rate": 2.8440256131448594e-06,
+      "loss": 2.0078,
+      "step": 177500
+    },
+    {
+      "epoch": 12.9,
+      "learning_rate": 2.7956989247311827e-06,
+      "loss": 2.0162,
+      "step": 178000
+    },
+    {
+      "epoch": 12.94,
+      "learning_rate": 2.7473722363175063e-06,
+      "loss": 2.0227,
+      "step": 178500
+    },
+    {
+      "epoch": 12.98,
+      "learning_rate": 2.6990455479038304e-06,
+      "loss": 2.0291,
+      "step": 179000
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 2.650718859490153e-06,
+      "loss": 2.0248,
+      "step": 179500
+    },
+    {
+      "epoch": 13.05,
+      "learning_rate": 2.6023921710764773e-06,
+      "loss": 1.996,
+      "step": 180000
+    },
+    {
+      "epoch": 13.08,
+      "learning_rate": 2.554065482662801e-06,
+      "loss": 2.012,
+      "step": 180500
+    },
+    {
+      "epoch": 13.12,
+      "learning_rate": 2.505738794249124e-06,
+      "loss": 2.002,
+      "step": 181000
+    },
+    {
+      "epoch": 13.16,
+      "learning_rate": 2.457412105835448e-06,
+      "loss": 2.0191,
+      "step": 181500
+    },
+    {
+      "epoch": 13.19,
+      "learning_rate": 2.4090854174217715e-06,
+      "loss": 2.0173,
+      "step": 182000
+    },
+    {
+      "epoch": 13.23,
+      "learning_rate": 2.360758729008095e-06,
+      "loss": 2.0347,
+      "step": 182500
+    },
+    {
+      "epoch": 13.27,
+      "learning_rate": 2.3124320405944183e-06,
+      "loss": 2.0357,
+      "step": 183000
+    },
+    {
+      "epoch": 13.3,
+      "learning_rate": 2.264105352180742e-06,
+      "loss": 2.0414,
+      "step": 183500
+    },
+    {
+      "epoch": 13.34,
+      "learning_rate": 2.2157786637670657e-06,
+      "loss": 2.036,
+      "step": 184000
+    },
+    {
+      "epoch": 13.37,
+      "learning_rate": 2.167451975353389e-06,
+      "loss": 2.0063,
+      "step": 184500
+    },
+    {
+      "epoch": 13.41,
+      "learning_rate": 2.119125286939713e-06,
+      "loss": 2.0062,
+      "step": 185000
+    },
+    {
+      "epoch": 13.45,
+      "learning_rate": 2.070798598526036e-06,
+      "loss": 2.0191,
+      "step": 185500
+    },
+    {
+      "epoch": 13.48,
+      "learning_rate": 2.02247191011236e-06,
+      "loss": 2.0148,
+      "step": 186000
+    },
+    {
+      "epoch": 13.52,
+      "learning_rate": 1.9741452216986835e-06,
+      "loss": 2.0332,
+      "step": 186500
+    },
+    {
+      "epoch": 13.56,
+      "learning_rate": 1.9258185332850067e-06,
+      "loss": 2.0137,
+      "step": 187000
+    },
+    {
+      "epoch": 13.59,
+      "learning_rate": 1.8774918448713304e-06,
+      "loss": 2.0341,
+      "step": 187500
+    },
+    {
+      "epoch": 13.63,
+      "learning_rate": 1.829165156457654e-06,
+      "loss": 1.9992,
+      "step": 188000
+    },
+    {
+      "epoch": 13.66,
+      "learning_rate": 1.7808384680439775e-06,
+      "loss": 2.0233,
+      "step": 188500
+    },
+    {
+      "epoch": 13.7,
+      "learning_rate": 1.732511779630301e-06,
+      "loss": 2.0065,
+      "step": 189000
+    },
+    {
+      "epoch": 13.74,
+      "learning_rate": 1.6841850912166246e-06,
+      "loss": 2.024,
+      "step": 189500
+    },
+    {
+      "epoch": 13.77,
+      "learning_rate": 1.635858402802948e-06,
+      "loss": 2.0306,
+      "step": 190000
+    },
+    {
+      "epoch": 13.81,
+      "learning_rate": 1.5875317143892715e-06,
+      "loss": 2.0454,
+      "step": 190500
+    },
+    {
+      "epoch": 13.85,
+      "learning_rate": 1.5392050259755951e-06,
+      "loss": 2.0143,
+      "step": 191000
+    },
+    {
+      "epoch": 13.88,
+      "learning_rate": 1.4908783375619188e-06,
+      "loss": 2.0186,
+      "step": 191500
+    },
+    {
+      "epoch": 13.92,
+      "learning_rate": 1.4425516491482422e-06,
+      "loss": 2.0109,
+      "step": 192000
+    },
+    {
+      "epoch": 13.95,
+      "learning_rate": 1.3942249607345657e-06,
+      "loss": 2.0091,
+      "step": 192500
+    },
+    {
+      "epoch": 13.99,
+      "learning_rate": 1.3458982723208893e-06,
+      "loss": 2.0223,
+      "step": 193000
+    },
+    {
+      "epoch": 14.03,
+      "learning_rate": 1.2975715839072127e-06,
+      "loss": 1.9994,
+      "step": 193500
+    },
+    {
+      "epoch": 14.06,
+      "learning_rate": 1.2492448954935364e-06,
+      "loss": 2.0055,
+      "step": 194000
+    },
+    {
+      "epoch": 14.1,
+      "learning_rate": 1.20091820707986e-06,
+      "loss": 2.0195,
+      "step": 194500
+    },
+    {
+      "epoch": 14.14,
+      "learning_rate": 1.1525915186661835e-06,
+      "loss": 2.0082,
+      "step": 195000
+    },
+    {
+      "epoch": 14.17,
+      "learning_rate": 1.1042648302525072e-06,
+      "loss": 2.0337,
+      "step": 195500
+    },
+    {
+      "epoch": 14.21,
+      "learning_rate": 1.0559381418388306e-06,
+      "loss": 2.0057,
+      "step": 196000
+    },
+    {
+      "epoch": 14.24,
+      "learning_rate": 1.007611453425154e-06,
+      "loss": 2.0224,
+      "step": 196500
+    },
+    {
+      "epoch": 14.28,
+      "learning_rate": 9.592847650114777e-07,
+      "loss": 2.031,
+      "step": 197000
+    },
+    {
+      "epoch": 14.32,
+      "learning_rate": 9.109580765978012e-07,
+      "loss": 2.0261,
+      "step": 197500
+    },
+    {
+      "epoch": 14.35,
+      "learning_rate": 8.626313881841247e-07,
+      "loss": 2.0288,
+      "step": 198000
+    },
+    {
+      "epoch": 14.39,
+      "learning_rate": 8.143046997704483e-07,
+      "loss": 2.0266,
+      "step": 198500
+    },
+    {
+      "epoch": 14.43,
+      "learning_rate": 7.659780113567718e-07,
+      "loss": 2.0164,
+      "step": 199000
+    },
+    {
+      "epoch": 14.46,
+      "learning_rate": 7.176513229430953e-07,
+      "loss": 2.0295,
+      "step": 199500
+    },
+    {
+      "epoch": 14.5,
+      "learning_rate": 6.69324634529419e-07,
+      "loss": 2.0177,
+      "step": 200000
+    },
+    {
+      "epoch": 14.5,
+      "eval_bleu": 26.5388,
+      "eval_gen_len": 24.6306,
+      "eval_loss": 1.7493975162506104,
+      "eval_runtime": 4655.3081,
+      "eval_samples_per_second": 11.853,
+      "eval_steps_per_second": 1.482,
+      "step": 200000
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 206925,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 15,
+  "save_steps": 50000,
+  "total_flos": 2.0776464395563008e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:596e6df7ff73946fbdc2cbdb9d67e12494a368115bae324f52a496aacf7624ac
+size 4856