Hugging Face's logo

Language-Media-Lab
/
byt5-small-ain-jpn-mt

Translation
Transformers
PyTorch
Ainu (Japan)
Japanese
t5
text2text-generation
text-generation-inference
Model card Files
xet
 Community
2
koki commited on
Commit
75ddb1b
·
1 Parent(s): cbb4902

first commit from cli

Browse files
Files changed (9) hide show
  1. checkpoint-262976/config.json +29 -0
  2. checkpoint-262976/optimizer.pt +3 -0
  3. checkpoint-262976/pytorch_model.bin +3 -0
  4. checkpoint-262976/rng_state.pth +3 -0
  5. checkpoint-262976/scheduler.pt +3 -0
  6. checkpoint-262976/special_tokens_map.json +1 -0
  7. checkpoint-262976/tokenizer_config.json +1 -0
  8. checkpoint-262976/trainer_state.json +3236 -0
  9. checkpoint-262976/training_args.bin +3 -0
checkpoint-262976/config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "google/byt5-small",
3
+ "architectures": [
4
+ "T5ForConditionalGeneration"
5
+ ],
6
+ "d_ff": 3584,
7
+ "d_kv": 64,
8
+ "d_model": 1472,
9
+ "decoder_start_token_id": 0,
10
+ "dropout_rate": 0.1,
11
+ "eos_token_id": 1,
12
+ "feed_forward_proj": "gated-gelu",
13
+ "gradient_checkpointing": false,
14
+ "initializer_factor": 1.0,
15
+ "is_encoder_decoder": true,
16
+ "layer_norm_epsilon": 1e-06,
17
+ "model_type": "t5",
18
+ "num_decoder_layers": 4,
19
+ "num_heads": 6,
20
+ "num_layers": 12,
21
+ "pad_token_id": 0,
22
+ "relative_attention_num_buckets": 32,
23
+ "tie_word_embeddings": false,
24
+ "tokenizer_class": "ByT5Tokenizer",
25
+ "torch_dtype": "float32",
26
+ "transformers_version": "4.16.0.dev0",
27
+ "use_cache": true,
28
+ "vocab_size": 384
29
+ }
checkpoint-262976/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:97d123fb4f044311e907c6b6e58b5194248799f20ebb8b7d3fc1680466781645
3
+ size 2397201945
checkpoint-262976/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3debdeffcda5edfc1795c612546e6f06b59f5ae61f9a87fec0c51551a4536d06
3
+ size 1198627501
checkpoint-262976/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e7ddc8c6d0140b28cd9ba7f0c9be40bf813a51db9150835fcdd596b9ab6645a
3
+ size 14503
checkpoint-262976/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:364075e50132c9d27f7daf75d52a7900659ee3abaede7969ebad8ea28e99ebf8
3
+ size 623
checkpoint-262976/special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"eos_token": {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "unk_token": {"content": "<unk>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "pad_token": {"content": "<pad>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "additional_special_tokens": ["<extra_id_0>", "<extra_id_1>", "<extra_id_2>", "<extra_id_3>", "<extra_id_4>", "<extra_id_5>", "<extra_id_6>", "<extra_id_7>", "<extra_id_8>", "<extra_id_9>", "<extra_id_10>", "<extra_id_11>", "<extra_id_12>", "<extra_id_13>", "<extra_id_14>", "<extra_id_15>", "<extra_id_16>", "<extra_id_17>", "<extra_id_18>", "<extra_id_19>", "<extra_id_20>", "<extra_id_21>", "<extra_id_22>", "<extra_id_23>", "<extra_id_24>", "<extra_id_25>", "<extra_id_26>", "<extra_id_27>", "<extra_id_28>", "<extra_id_29>", "<extra_id_30>", "<extra_id_31>", "<extra_id_32>", "<extra_id_33>", "<extra_id_34>", "<extra_id_35>", "<extra_id_36>", "<extra_id_37>", "<extra_id_38>", "<extra_id_39>", "<extra_id_40>", "<extra_id_41>", "<extra_id_42>", "<extra_id_43>", "<extra_id_44>", "<extra_id_45>", "<extra_id_46>", "<extra_id_47>", "<extra_id_48>", "<extra_id_49>", "<extra_id_50>", "<extra_id_51>", "<extra_id_52>", "<extra_id_53>", "<extra_id_54>", "<extra_id_55>", "<extra_id_56>", "<extra_id_57>", "<extra_id_58>", "<extra_id_59>", "<extra_id_60>", "<extra_id_61>", "<extra_id_62>", "<extra_id_63>", "<extra_id_64>", "<extra_id_65>", "<extra_id_66>", "<extra_id_67>", "<extra_id_68>", "<extra_id_69>", "<extra_id_70>", "<extra_id_71>", "<extra_id_72>", "<extra_id_73>", "<extra_id_74>", "<extra_id_75>", "<extra_id_76>", "<extra_id_77>", "<extra_id_78>", "<extra_id_79>", "<extra_id_80>", "<extra_id_81>", "<extra_id_82>", "<extra_id_83>", "<extra_id_84>", "<extra_id_85>", "<extra_id_86>", "<extra_id_87>", "<extra_id_88>", "<extra_id_89>", "<extra_id_90>", "<extra_id_91>", "<extra_id_92>", "<extra_id_93>", "<extra_id_94>", "<extra_id_95>", "<extra_id_96>", "<extra_id_97>", "<extra_id_98>", "<extra_id_99>", "<extra_id_100>", "<extra_id_101>", "<extra_id_102>", "<extra_id_103>", "<extra_id_104>", "<extra_id_105>", "<extra_id_106>", "<extra_id_107>", "<extra_id_108>", "<extra_id_109>", "<extra_id_110>", "<extra_id_111>", "<extra_id_112>", "<extra_id_113>", "<extra_id_114>", "<extra_id_115>", "<extra_id_116>", "<extra_id_117>", "<extra_id_118>", "<extra_id_119>", "<extra_id_120>", "<extra_id_121>", "<extra_id_122>", "<extra_id_123>", "<extra_id_124>"]}
checkpoint-262976/tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"eos_token": {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "unk_token": {"content": "<unk>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "pad_token": {"content": "<pad>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "extra_ids": 125, "additional_special_tokens": ["<extra_id_0>", "<extra_id_1>", "<extra_id_2>", "<extra_id_3>", "<extra_id_4>", "<extra_id_5>", "<extra_id_6>", "<extra_id_7>", "<extra_id_8>", "<extra_id_9>", "<extra_id_10>", "<extra_id_11>", "<extra_id_12>", "<extra_id_13>", "<extra_id_14>", "<extra_id_15>", "<extra_id_16>", "<extra_id_17>", "<extra_id_18>", "<extra_id_19>", "<extra_id_20>", "<extra_id_21>", "<extra_id_22>", "<extra_id_23>", "<extra_id_24>", "<extra_id_25>", "<extra_id_26>", "<extra_id_27>", "<extra_id_28>", "<extra_id_29>", "<extra_id_30>", "<extra_id_31>", "<extra_id_32>", "<extra_id_33>", "<extra_id_34>", "<extra_id_35>", "<extra_id_36>", "<extra_id_37>", "<extra_id_38>", "<extra_id_39>", "<extra_id_40>", "<extra_id_41>", "<extra_id_42>", "<extra_id_43>", "<extra_id_44>", "<extra_id_45>", "<extra_id_46>", "<extra_id_47>", "<extra_id_48>", "<extra_id_49>", "<extra_id_50>", "<extra_id_51>", "<extra_id_52>", "<extra_id_53>", "<extra_id_54>", "<extra_id_55>", "<extra_id_56>", "<extra_id_57>", "<extra_id_58>", "<extra_id_59>", "<extra_id_60>", "<extra_id_61>", "<extra_id_62>", "<extra_id_63>", "<extra_id_64>", "<extra_id_65>", "<extra_id_66>", "<extra_id_67>", "<extra_id_68>", "<extra_id_69>", "<extra_id_70>", "<extra_id_71>", "<extra_id_72>", "<extra_id_73>", "<extra_id_74>", "<extra_id_75>", "<extra_id_76>", "<extra_id_77>", "<extra_id_78>", "<extra_id_79>", "<extra_id_80>", "<extra_id_81>", "<extra_id_82>", "<extra_id_83>", "<extra_id_84>", "<extra_id_85>", "<extra_id_86>", "<extra_id_87>", "<extra_id_88>", "<extra_id_89>", "<extra_id_90>", "<extra_id_91>", "<extra_id_92>", "<extra_id_93>", "<extra_id_94>", "<extra_id_95>", "<extra_id_96>", "<extra_id_97>", "<extra_id_98>", "<extra_id_99>", "<extra_id_100>", "<extra_id_101>", "<extra_id_102>", "<extra_id_103>", "<extra_id_104>", "<extra_id_105>", "<extra_id_106>", "<extra_id_107>", "<extra_id_108>", "<extra_id_109>", "<extra_id_110>", "<extra_id_111>", "<extra_id_112>", "<extra_id_113>", "<extra_id_114>", "<extra_id_115>", "<extra_id_116>", "<extra_id_117>", "<extra_id_118>", "<extra_id_119>", "<extra_id_120>", "<extra_id_121>", "<extra_id_122>", "<extra_id_123>", "<extra_id_124>"], "special_tokens_map_file": "/root/.cache/huggingface/transformers/3715c3de7f5eda8f5c9274c55ea8e8632fe6e4cdfdf8e9066ae173b956cd9b80.063895353d5ef9e19a25220cb616c43abc5e84a2f11b1ffb71c29e097572a109", "tokenizer_file": null, "name_or_path": "google/byt5-small", "tokenizer_class": "ByT5Tokenizer"}
checkpoint-262976/trainer_state.json ADDED
@@ -0,0 +1,3236 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 7.0,
5
+ "global_step": 262976,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.01,
12
+ "learning_rate": 0.00019998960218803238,
13
+ "loss": 1.7263,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.03,
18
+ "learning_rate": 0.00019997920437606474,
19
+ "loss": 1.3861,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.04,
24
+ "learning_rate": 0.0001999688065640971,
25
+ "loss": 1.2883,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.05,
30
+ "learning_rate": 0.00019995840875212947,
31
+ "loss": 1.2357,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.07,
36
+ "learning_rate": 0.00019994801094016184,
37
+ "loss": 1.2031,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.08,
42
+ "learning_rate": 0.0001999376131281942,
43
+ "loss": 1.1688,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.09,
48
+ "learning_rate": 0.00019992721531622657,
49
+ "loss": 1.1213,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.11,
54
+ "learning_rate": 0.00019991681750425894,
55
+ "loss": 1.1088,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.12,
60
+ "learning_rate": 0.0001999064196922913,
61
+ "loss": 1.1039,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.13,
66
+ "learning_rate": 0.00019989602188032367,
67
+ "loss": 1.0876,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.15,
72
+ "learning_rate": 0.00019988562406835603,
73
+ "loss": 1.0792,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.16,
78
+ "learning_rate": 0.00019987522625638843,
79
+ "loss": 1.0654,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.17,
84
+ "learning_rate": 0.0001998648284444208,
85
+ "loss": 1.0561,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.19,
90
+ "learning_rate": 0.00019985443063245316,
91
+ "loss": 1.0406,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.2,
96
+ "learning_rate": 0.00019984403282048553,
97
+ "loss": 1.0274,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.21,
102
+ "learning_rate": 0.00019983363500851792,
103
+ "loss": 1.0271,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.23,
108
+ "learning_rate": 0.00019982323719655028,
109
+ "loss": 1.0028,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.24,
114
+ "learning_rate": 0.00019981283938458265,
115
+ "loss": 0.9957,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.25,
120
+ "learning_rate": 0.00019980244157261502,
121
+ "loss": 0.9892,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.27,
126
+ "learning_rate": 0.00019979204376064738,
127
+ "loss": 0.9524,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.28,
132
+ "learning_rate": 0.00019978164594867975,
133
+ "loss": 0.9618,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.29,
138
+ "learning_rate": 0.00019977124813671211,
139
+ "loss": 0.9499,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.31,
144
+ "learning_rate": 0.00019976085032474448,
145
+ "loss": 0.929,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.32,
150
+ "learning_rate": 0.00019975045251277685,
151
+ "loss": 0.9172,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.33,
156
+ "learning_rate": 0.0001997400547008092,
157
+ "loss": 0.917,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.35,
162
+ "learning_rate": 0.00019972965688884158,
163
+ "loss": 0.8794,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.36,
168
+ "learning_rate": 0.00019971925907687394,
169
+ "loss": 0.8818,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.37,
174
+ "learning_rate": 0.0001997088612649063,
175
+ "loss": 0.8747,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.39,
180
+ "learning_rate": 0.00019969846345293868,
181
+ "loss": 0.8592,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.4,
186
+ "learning_rate": 0.00019968806564097104,
187
+ "loss": 0.8561,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.41,
192
+ "learning_rate": 0.0001996776678290034,
193
+ "loss": 0.8538,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.43,
198
+ "learning_rate": 0.00019966727001703577,
199
+ "loss": 0.8396,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.44,
204
+ "learning_rate": 0.00019965687220506814,
205
+ "loss": 0.8225,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.45,
210
+ "learning_rate": 0.0001996464743931005,
211
+ "loss": 0.8246,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.47,
216
+ "learning_rate": 0.00019963607658113287,
217
+ "loss": 0.8278,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.48,
222
+ "learning_rate": 0.00019962567876916524,
223
+ "loss": 0.8071,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.49,
228
+ "learning_rate": 0.0001996152809571976,
229
+ "loss": 0.7961,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.51,
234
+ "learning_rate": 0.00019960488314522997,
235
+ "loss": 0.7681,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.52,
240
+ "learning_rate": 0.00019959448533326236,
241
+ "loss": 0.7806,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.53,
246
+ "learning_rate": 0.00019958408752129473,
247
+ "loss": 0.754,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.55,
252
+ "learning_rate": 0.0001995736897093271,
253
+ "loss": 0.7402,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.56,
258
+ "learning_rate": 0.00019956329189735946,
259
+ "loss": 0.7455,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.57,
264
+ "learning_rate": 0.00019955289408539183,
265
+ "loss": 0.7404,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.59,
270
+ "learning_rate": 0.00019954249627342422,
271
+ "loss": 0.7372,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.6,
276
+ "learning_rate": 0.00019953209846145658,
277
+ "loss": 0.7389,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.61,
282
+ "learning_rate": 0.00019952170064948895,
283
+ "loss": 0.7227,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.63,
288
+ "learning_rate": 0.00019951130283752132,
289
+ "loss": 0.7496,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.64,
294
+ "learning_rate": 0.00019950090502555368,
295
+ "loss": 0.7116,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.65,
300
+ "learning_rate": 0.00019949050721358605,
301
+ "loss": 0.7025,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.67,
306
+ "learning_rate": 0.00019948010940161841,
307
+ "loss": 0.6993,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.68,
312
+ "learning_rate": 0.00019946971158965078,
313
+ "loss": 0.6999,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 0.69,
318
+ "learning_rate": 0.00019945931377768315,
319
+ "loss": 0.6792,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 0.71,
324
+ "learning_rate": 0.0001994489159657155,
325
+ "loss": 0.6936,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 0.72,
330
+ "learning_rate": 0.00019943851815374788,
331
+ "loss": 0.6906,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 0.73,
336
+ "learning_rate": 0.00019942812034178024,
337
+ "loss": 0.6717,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 0.75,
342
+ "learning_rate": 0.0001994177225298126,
343
+ "loss": 0.6746,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 0.76,
348
+ "learning_rate": 0.00019940732471784498,
349
+ "loss": 0.674,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 0.77,
354
+ "learning_rate": 0.00019939692690587734,
355
+ "loss": 0.6781,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 0.79,
360
+ "learning_rate": 0.0001993865290939097,
361
+ "loss": 0.6897,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 0.8,
366
+ "learning_rate": 0.00019937613128194207,
367
+ "loss": 0.6656,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 0.81,
372
+ "learning_rate": 0.00019936573346997444,
373
+ "loss": 0.6772,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 0.83,
378
+ "learning_rate": 0.0001993553356580068,
379
+ "loss": 0.6434,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 0.84,
384
+ "learning_rate": 0.00019934493784603917,
385
+ "loss": 0.6644,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 0.85,
390
+ "learning_rate": 0.00019933454003407156,
391
+ "loss": 0.6459,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 0.87,
396
+ "learning_rate": 0.00019932414222210393,
397
+ "loss": 0.6428,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 0.88,
402
+ "learning_rate": 0.0001993137444101363,
403
+ "loss": 0.6313,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 0.89,
408
+ "learning_rate": 0.00019930334659816866,
409
+ "loss": 0.641,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 0.91,
414
+ "learning_rate": 0.00019929294878620103,
415
+ "loss": 0.6335,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 0.92,
420
+ "learning_rate": 0.0001992825509742334,
421
+ "loss": 0.6468,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 0.93,
426
+ "learning_rate": 0.00019927215316226576,
427
+ "loss": 0.6108,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 0.94,
432
+ "learning_rate": 0.00019926175535029815,
433
+ "loss": 0.64,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 0.96,
438
+ "learning_rate": 0.00019925135753833052,
439
+ "loss": 0.6094,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 0.97,
444
+ "learning_rate": 0.00019924095972636289,
445
+ "loss": 0.6081,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 0.98,
450
+ "learning_rate": 0.00019923056191439525,
451
+ "loss": 0.609,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 1.0,
456
+ "learning_rate": 0.00019922016410242762,
457
+ "loss": 0.6386,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 1.0,
462
+ "eval_bleu": 0.0649,
463
+ "eval_gen_len": 18.2942,
464
+ "eval_loss": 0.5461284518241882,
465
+ "eval_runtime": 1943.184,
466
+ "eval_samples_per_second": 4.834,
467
+ "eval_steps_per_second": 2.417,
468
+ "step": 37568
469
+ },
470
+ {
471
+ "epoch": 1.01,
472
+ "learning_rate": 0.00019920976629045998,
473
+ "loss": 0.5822,
474
+ "step": 38000
475
+ },
476
+ {
477
+ "epoch": 1.02,
478
+ "learning_rate": 0.00019919936847849235,
479
+ "loss": 0.5875,
480
+ "step": 38500
481
+ },
482
+ {
483
+ "epoch": 1.04,
484
+ "learning_rate": 0.00019918897066652471,
485
+ "loss": 0.5772,
486
+ "step": 39000
487
+ },
488
+ {
489
+ "epoch": 1.05,
490
+ "learning_rate": 0.00019917857285455708,
491
+ "loss": 0.5732,
492
+ "step": 39500
493
+ },
494
+ {
495
+ "epoch": 1.06,
496
+ "learning_rate": 0.00019916817504258945,
497
+ "loss": 0.5781,
498
+ "step": 40000
499
+ },
500
+ {
501
+ "epoch": 1.08,
502
+ "learning_rate": 0.0001991577772306218,
503
+ "loss": 0.5841,
504
+ "step": 40500
505
+ },
506
+ {
507
+ "epoch": 1.09,
508
+ "learning_rate": 0.00019914737941865418,
509
+ "loss": 0.5801,
510
+ "step": 41000
511
+ },
512
+ {
513
+ "epoch": 1.1,
514
+ "learning_rate": 0.00019913698160668654,
515
+ "loss": 0.57,
516
+ "step": 41500
517
+ },
518
+ {
519
+ "epoch": 1.12,
520
+ "learning_rate": 0.0001991265837947189,
521
+ "loss": 0.5653,
522
+ "step": 42000
523
+ },
524
+ {
525
+ "epoch": 1.13,
526
+ "learning_rate": 0.00019911618598275128,
527
+ "loss": 0.5885,
528
+ "step": 42500
529
+ },
530
+ {
531
+ "epoch": 1.14,
532
+ "learning_rate": 0.00019910578817078364,
533
+ "loss": 0.5605,
534
+ "step": 43000
535
+ },
536
+ {
537
+ "epoch": 1.16,
538
+ "learning_rate": 0.000199095390358816,
539
+ "loss": 0.5882,
540
+ "step": 43500
541
+ },
542
+ {
543
+ "epoch": 1.17,
544
+ "learning_rate": 0.00019908499254684837,
545
+ "loss": 0.5702,
546
+ "step": 44000
547
+ },
548
+ {
549
+ "epoch": 1.18,
550
+ "learning_rate": 0.00019907459473488074,
551
+ "loss": 0.5631,
552
+ "step": 44500
553
+ },
554
+ {
555
+ "epoch": 1.2,
556
+ "learning_rate": 0.00019906419692291313,
557
+ "loss": 0.5557,
558
+ "step": 45000
559
+ },
560
+ {
561
+ "epoch": 1.21,
562
+ "learning_rate": 0.0001990537991109455,
563
+ "loss": 0.5603,
564
+ "step": 45500
565
+ },
566
+ {
567
+ "epoch": 1.22,
568
+ "learning_rate": 0.00019904340129897787,
569
+ "loss": 0.5578,
570
+ "step": 46000
571
+ },
572
+ {
573
+ "epoch": 1.24,
574
+ "learning_rate": 0.00019903300348701023,
575
+ "loss": 0.5589,
576
+ "step": 46500
577
+ },
578
+ {
579
+ "epoch": 1.25,
580
+ "learning_rate": 0.0001990226056750426,
581
+ "loss": 0.5489,
582
+ "step": 47000
583
+ },
584
+ {
585
+ "epoch": 1.26,
586
+ "learning_rate": 0.00019901220786307496,
587
+ "loss": 0.5662,
588
+ "step": 47500
589
+ },
590
+ {
591
+ "epoch": 1.28,
592
+ "learning_rate": 0.00019900181005110733,
593
+ "loss": 0.5345,
594
+ "step": 48000
595
+ },
596
+ {
597
+ "epoch": 1.29,
598
+ "learning_rate": 0.0001989914122391397,
599
+ "loss": 0.5629,
600
+ "step": 48500
601
+ },
602
+ {
603
+ "epoch": 1.3,
604
+ "learning_rate": 0.0001989810144271721,
605
+ "loss": 0.5505,
606
+ "step": 49000
607
+ },
608
+ {
609
+ "epoch": 1.32,
610
+ "learning_rate": 0.00019897061661520445,
611
+ "loss": 0.5443,
612
+ "step": 49500
613
+ },
614
+ {
615
+ "epoch": 1.33,
616
+ "learning_rate": 0.00019896021880323682,
617
+ "loss": 0.5426,
618
+ "step": 50000
619
+ },
620
+ {
621
+ "epoch": 1.34,
622
+ "learning_rate": 0.00019894982099126919,
623
+ "loss": 0.5445,
624
+ "step": 50500
625
+ },
626
+ {
627
+ "epoch": 1.36,
628
+ "learning_rate": 0.00019893942317930155,
629
+ "loss": 0.5402,
630
+ "step": 51000
631
+ },
632
+ {
633
+ "epoch": 1.37,
634
+ "learning_rate": 0.00019892902536733392,
635
+ "loss": 0.5744,
636
+ "step": 51500
637
+ },
638
+ {
639
+ "epoch": 1.38,
640
+ "learning_rate": 0.00019891862755536628,
641
+ "loss": 0.5523,
642
+ "step": 52000
643
+ },
644
+ {
645
+ "epoch": 1.4,
646
+ "learning_rate": 0.00019890822974339865,
647
+ "loss": 0.5395,
648
+ "step": 52500
649
+ },
650
+ {
651
+ "epoch": 1.41,
652
+ "learning_rate": 0.00019889783193143102,
653
+ "loss": 0.5469,
654
+ "step": 53000
655
+ },
656
+ {
657
+ "epoch": 1.42,
658
+ "learning_rate": 0.00019888743411946338,
659
+ "loss": 0.5314,
660
+ "step": 53500
661
+ },
662
+ {
663
+ "epoch": 1.44,
664
+ "learning_rate": 0.00019887703630749575,
665
+ "loss": 0.5437,
666
+ "step": 54000
667
+ },
668
+ {
669
+ "epoch": 1.45,
670
+ "learning_rate": 0.0001988666384955281,
671
+ "loss": 0.5466,
672
+ "step": 54500
673
+ },
674
+ {
675
+ "epoch": 1.46,
676
+ "learning_rate": 0.00019885624068356048,
677
+ "loss": 0.5351,
678
+ "step": 55000
679
+ },
680
+ {
681
+ "epoch": 1.48,
682
+ "learning_rate": 0.00019884584287159284,
683
+ "loss": 0.5495,
684
+ "step": 55500
685
+ },
686
+ {
687
+ "epoch": 1.49,
688
+ "learning_rate": 0.0001988354450596252,
689
+ "loss": 0.5189,
690
+ "step": 56000
691
+ },
692
+ {
693
+ "epoch": 1.5,
694
+ "learning_rate": 0.00019882504724765758,
695
+ "loss": 0.5568,
696
+ "step": 56500
697
+ },
698
+ {
699
+ "epoch": 1.52,
700
+ "learning_rate": 0.00019881464943568994,
701
+ "loss": 0.523,
702
+ "step": 57000
703
+ },
704
+ {
705
+ "epoch": 1.53,
706
+ "learning_rate": 0.00019880425162372234,
707
+ "loss": 0.5247,
708
+ "step": 57500
709
+ },
710
+ {
711
+ "epoch": 1.54,
712
+ "learning_rate": 0.0001987938538117547,
713
+ "loss": 0.5274,
714
+ "step": 58000
715
+ },
716
+ {
717
+ "epoch": 1.56,
718
+ "learning_rate": 0.00019878345599978707,
719
+ "loss": 0.538,
720
+ "step": 58500
721
+ },
722
+ {
723
+ "epoch": 1.57,
724
+ "learning_rate": 0.00019877305818781943,
725
+ "loss": 0.5319,
726
+ "step": 59000
727
+ },
728
+ {
729
+ "epoch": 1.58,
730
+ "learning_rate": 0.0001987626603758518,
731
+ "loss": 0.5145,
732
+ "step": 59500
733
+ },
734
+ {
735
+ "epoch": 1.6,
736
+ "learning_rate": 0.00019875226256388417,
737
+ "loss": 0.534,
738
+ "step": 60000
739
+ },
740
+ {
741
+ "epoch": 1.61,
742
+ "learning_rate": 0.00019874186475191653,
743
+ "loss": 0.5331,
744
+ "step": 60500
745
+ },
746
+ {
747
+ "epoch": 1.62,
748
+ "learning_rate": 0.0001987314669399489,
749
+ "loss": 0.5081,
750
+ "step": 61000
751
+ },
752
+ {
753
+ "epoch": 1.64,
754
+ "learning_rate": 0.00019872106912798126,
755
+ "loss": 0.5151,
756
+ "step": 61500
757
+ },
758
+ {
759
+ "epoch": 1.65,
760
+ "learning_rate": 0.00019871067131601363,
761
+ "loss": 0.5192,
762
+ "step": 62000
763
+ },
764
+ {
765
+ "epoch": 1.66,
766
+ "learning_rate": 0.000198700273504046,
767
+ "loss": 0.4988,
768
+ "step": 62500
769
+ },
770
+ {
771
+ "epoch": 1.68,
772
+ "learning_rate": 0.0001986898756920784,
773
+ "loss": 0.5186,
774
+ "step": 63000
775
+ },
776
+ {
777
+ "epoch": 1.69,
778
+ "learning_rate": 0.00019867947788011075,
779
+ "loss": 0.5215,
780
+ "step": 63500
781
+ },
782
+ {
783
+ "epoch": 1.7,
784
+ "learning_rate": 0.00019866908006814312,
785
+ "loss": 0.5082,
786
+ "step": 64000
787
+ },
788
+ {
789
+ "epoch": 1.72,
790
+ "learning_rate": 0.00019865868225617549,
791
+ "loss": 0.5186,
792
+ "step": 64500
793
+ },
794
+ {
795
+ "epoch": 1.73,
796
+ "learning_rate": 0.00019864828444420785,
797
+ "loss": 0.5192,
798
+ "step": 65000
799
+ },
800
+ {
801
+ "epoch": 1.74,
802
+ "learning_rate": 0.00019863788663224022,
803
+ "loss": 0.5033,
804
+ "step": 65500
805
+ },
806
+ {
807
+ "epoch": 1.76,
808
+ "learning_rate": 0.00019862748882027258,
809
+ "loss": 0.5196,
810
+ "step": 66000
811
+ },
812
+ {
813
+ "epoch": 1.77,
814
+ "learning_rate": 0.00019861709100830495,
815
+ "loss": 0.5259,
816
+ "step": 66500
817
+ },
818
+ {
819
+ "epoch": 1.78,
820
+ "learning_rate": 0.00019860669319633732,
821
+ "loss": 0.5013,
822
+ "step": 67000
823
+ },
824
+ {
825
+ "epoch": 1.8,
826
+ "learning_rate": 0.00019859629538436968,
827
+ "loss": 0.497,
828
+ "step": 67500
829
+ },
830
+ {
831
+ "epoch": 1.81,
832
+ "learning_rate": 0.00019858589757240205,
833
+ "loss": 0.5142,
834
+ "step": 68000
835
+ },
836
+ {
837
+ "epoch": 1.82,
838
+ "learning_rate": 0.0001985754997604344,
839
+ "loss": 0.4933,
840
+ "step": 68500
841
+ },
842
+ {
843
+ "epoch": 1.84,
844
+ "learning_rate": 0.00019856510194846678,
845
+ "loss": 0.5157,
846
+ "step": 69000
847
+ },
848
+ {
849
+ "epoch": 1.85,
850
+ "learning_rate": 0.00019855470413649915,
851
+ "loss": 0.5244,
852
+ "step": 69500
853
+ },
854
+ {
855
+ "epoch": 1.86,
856
+ "learning_rate": 0.0001985443063245315,
857
+ "loss": 0.4957,
858
+ "step": 70000
859
+ },
860
+ {
861
+ "epoch": 1.88,
862
+ "learning_rate": 0.0001985339085125639,
863
+ "loss": 0.4948,
864
+ "step": 70500
865
+ },
866
+ {
867
+ "epoch": 1.89,
868
+ "learning_rate": 0.00019852351070059627,
869
+ "loss": 0.497,
870
+ "step": 71000
871
+ },
872
+ {
873
+ "epoch": 1.9,
874
+ "learning_rate": 0.00019851311288862864,
875
+ "loss": 0.4975,
876
+ "step": 71500
877
+ },
878
+ {
879
+ "epoch": 1.92,
880
+ "learning_rate": 0.000198502715076661,
881
+ "loss": 0.4931,
882
+ "step": 72000
883
+ },
884
+ {
885
+ "epoch": 1.93,
886
+ "learning_rate": 0.00019849231726469337,
887
+ "loss": 0.4886,
888
+ "step": 72500
889
+ },
890
+ {
891
+ "epoch": 1.94,
892
+ "learning_rate": 0.00019848191945272573,
893
+ "loss": 0.4972,
894
+ "step": 73000
895
+ },
896
+ {
897
+ "epoch": 1.96,
898
+ "learning_rate": 0.0001984715216407581,
899
+ "loss": 0.4861,
900
+ "step": 73500
901
+ },
902
+ {
903
+ "epoch": 1.97,
904
+ "learning_rate": 0.00019846112382879047,
905
+ "loss": 0.5027,
906
+ "step": 74000
907
+ },
908
+ {
909
+ "epoch": 1.98,
910
+ "learning_rate": 0.00019845072601682283,
911
+ "loss": 0.4816,
912
+ "step": 74500
913
+ },
914
+ {
915
+ "epoch": 2.0,
916
+ "learning_rate": 0.0001984403282048552,
917
+ "loss": 0.489,
918
+ "step": 75000
919
+ },
920
+ {
921
+ "epoch": 2.0,
922
+ "eval_bleu": 0.0787,
923
+ "eval_gen_len": 18.4245,
924
+ "eval_loss": 0.4464058578014374,
925
+ "eval_runtime": 1938.9705,
926
+ "eval_samples_per_second": 4.844,
927
+ "eval_steps_per_second": 2.422,
928
+ "step": 75136
929
+ },
930
+ {
931
+ "epoch": 2.01,
932
+ "learning_rate": 0.00019842993039288756,
933
+ "loss": 0.4753,
934
+ "step": 75500
935
+ },
936
+ {
937
+ "epoch": 2.02,
938
+ "learning_rate": 0.00019841953258091993,
939
+ "loss": 0.4532,
940
+ "step": 76000
941
+ },
942
+ {
943
+ "epoch": 2.04,
944
+ "learning_rate": 0.00019840913476895232,
945
+ "loss": 0.4509,
946
+ "step": 76500
947
+ },
948
+ {
949
+ "epoch": 2.05,
950
+ "learning_rate": 0.0001983987369569847,
951
+ "loss": 0.4569,
952
+ "step": 77000
953
+ },
954
+ {
955
+ "epoch": 2.06,
956
+ "learning_rate": 0.00019838833914501705,
957
+ "loss": 0.4517,
958
+ "step": 77500
959
+ },
960
+ {
961
+ "epoch": 2.08,
962
+ "learning_rate": 0.00019837794133304942,
963
+ "loss": 0.432,
964
+ "step": 78000
965
+ },
966
+ {
967
+ "epoch": 2.09,
968
+ "learning_rate": 0.00019836754352108179,
969
+ "loss": 0.4639,
970
+ "step": 78500
971
+ },
972
+ {
973
+ "epoch": 2.1,
974
+ "learning_rate": 0.00019835714570911415,
975
+ "loss": 0.4425,
976
+ "step": 79000
977
+ },
978
+ {
979
+ "epoch": 2.12,
980
+ "learning_rate": 0.00019834674789714652,
981
+ "loss": 0.4403,
982
+ "step": 79500
983
+ },
984
+ {
985
+ "epoch": 2.13,
986
+ "learning_rate": 0.00019833635008517888,
987
+ "loss": 0.4653,
988
+ "step": 80000
989
+ },
990
+ {
991
+ "epoch": 2.14,
992
+ "learning_rate": 0.00019832595227321125,
993
+ "loss": 0.4584,
994
+ "step": 80500
995
+ },
996
+ {
997
+ "epoch": 2.16,
998
+ "learning_rate": 0.00019831555446124362,
999
+ "loss": 0.4514,
1000
+ "step": 81000
1001
+ },
1002
+ {
1003
+ "epoch": 2.17,
1004
+ "learning_rate": 0.00019830515664927598,
1005
+ "loss": 0.4542,
1006
+ "step": 81500
1007
+ },
1008
+ {
1009
+ "epoch": 2.18,
1010
+ "learning_rate": 0.00019829475883730835,
1011
+ "loss": 0.4473,
1012
+ "step": 82000
1013
+ },
1014
+ {
1015
+ "epoch": 2.2,
1016
+ "learning_rate": 0.00019828436102534071,
1017
+ "loss": 0.4642,
1018
+ "step": 82500
1019
+ },
1020
+ {
1021
+ "epoch": 2.21,
1022
+ "learning_rate": 0.0001982739632133731,
1023
+ "loss": 0.4382,
1024
+ "step": 83000
1025
+ },
1026
+ {
1027
+ "epoch": 2.22,
1028
+ "learning_rate": 0.00019826356540140547,
1029
+ "loss": 0.4541,
1030
+ "step": 83500
1031
+ },
1032
+ {
1033
+ "epoch": 2.24,
1034
+ "learning_rate": 0.00019825316758943784,
1035
+ "loss": 0.4621,
1036
+ "step": 84000
1037
+ },
1038
+ {
1039
+ "epoch": 2.25,
1040
+ "learning_rate": 0.0001982427697774702,
1041
+ "loss": 0.4585,
1042
+ "step": 84500
1043
+ },
1044
+ {
1045
+ "epoch": 2.26,
1046
+ "learning_rate": 0.00019823237196550257,
1047
+ "loss": 0.4595,
1048
+ "step": 85000
1049
+ },
1050
+ {
1051
+ "epoch": 2.28,
1052
+ "learning_rate": 0.00019822197415353494,
1053
+ "loss": 0.4554,
1054
+ "step": 85500
1055
+ },
1056
+ {
1057
+ "epoch": 2.29,
1058
+ "learning_rate": 0.0001982115763415673,
1059
+ "loss": 0.4545,
1060
+ "step": 86000
1061
+ },
1062
+ {
1063
+ "epoch": 2.3,
1064
+ "learning_rate": 0.00019820117852959967,
1065
+ "loss": 0.448,
1066
+ "step": 86500
1067
+ },
1068
+ {
1069
+ "epoch": 2.32,
1070
+ "learning_rate": 0.00019819078071763203,
1071
+ "loss": 0.4582,
1072
+ "step": 87000
1073
+ },
1074
+ {
1075
+ "epoch": 2.33,
1076
+ "learning_rate": 0.0001981803829056644,
1077
+ "loss": 0.4362,
1078
+ "step": 87500
1079
+ },
1080
+ {
1081
+ "epoch": 2.34,
1082
+ "learning_rate": 0.00019816998509369677,
1083
+ "loss": 0.4386,
1084
+ "step": 88000
1085
+ },
1086
+ {
1087
+ "epoch": 2.36,
1088
+ "learning_rate": 0.00019815958728172913,
1089
+ "loss": 0.4437,
1090
+ "step": 88500
1091
+ },
1092
+ {
1093
+ "epoch": 2.37,
1094
+ "learning_rate": 0.0001981491894697615,
1095
+ "loss": 0.4485,
1096
+ "step": 89000
1097
+ },
1098
+ {
1099
+ "epoch": 2.38,
1100
+ "learning_rate": 0.00019813879165779386,
1101
+ "loss": 0.4489,
1102
+ "step": 89500
1103
+ },
1104
+ {
1105
+ "epoch": 2.4,
1106
+ "learning_rate": 0.00019812839384582626,
1107
+ "loss": 0.4528,
1108
+ "step": 90000
1109
+ },
1110
+ {
1111
+ "epoch": 2.41,
1112
+ "learning_rate": 0.00019811799603385862,
1113
+ "loss": 0.4558,
1114
+ "step": 90500
1115
+ },
1116
+ {
1117
+ "epoch": 2.42,
1118
+ "learning_rate": 0.000198107598221891,
1119
+ "loss": 0.4408,
1120
+ "step": 91000
1121
+ },
1122
+ {
1123
+ "epoch": 2.44,
1124
+ "learning_rate": 0.00019809720040992335,
1125
+ "loss": 0.4403,
1126
+ "step": 91500
1127
+ },
1128
+ {
1129
+ "epoch": 2.45,
1130
+ "learning_rate": 0.00019808680259795572,
1131
+ "loss": 0.4336,
1132
+ "step": 92000
1133
+ },
1134
+ {
1135
+ "epoch": 2.46,
1136
+ "learning_rate": 0.00019807640478598809,
1137
+ "loss": 0.4327,
1138
+ "step": 92500
1139
+ },
1140
+ {
1141
+ "epoch": 2.48,
1142
+ "learning_rate": 0.00019806600697402045,
1143
+ "loss": 0.4551,
1144
+ "step": 93000
1145
+ },
1146
+ {
1147
+ "epoch": 2.49,
1148
+ "learning_rate": 0.00019805560916205282,
1149
+ "loss": 0.45,
1150
+ "step": 93500
1151
+ },
1152
+ {
1153
+ "epoch": 2.5,
1154
+ "learning_rate": 0.00019804521135008518,
1155
+ "loss": 0.4313,
1156
+ "step": 94000
1157
+ },
1158
+ {
1159
+ "epoch": 2.52,
1160
+ "learning_rate": 0.00019803481353811755,
1161
+ "loss": 0.4422,
1162
+ "step": 94500
1163
+ },
1164
+ {
1165
+ "epoch": 2.53,
1166
+ "learning_rate": 0.00019802441572614992,
1167
+ "loss": 0.4528,
1168
+ "step": 95000
1169
+ },
1170
+ {
1171
+ "epoch": 2.54,
1172
+ "learning_rate": 0.00019801401791418228,
1173
+ "loss": 0.4507,
1174
+ "step": 95500
1175
+ },
1176
+ {
1177
+ "epoch": 2.56,
1178
+ "learning_rate": 0.00019800362010221468,
1179
+ "loss": 0.4476,
1180
+ "step": 96000
1181
+ },
1182
+ {
1183
+ "epoch": 2.57,
1184
+ "learning_rate": 0.00019799322229024704,
1185
+ "loss": 0.4405,
1186
+ "step": 96500
1187
+ },
1188
+ {
1189
+ "epoch": 2.58,
1190
+ "learning_rate": 0.0001979828244782794,
1191
+ "loss": 0.425,
1192
+ "step": 97000
1193
+ },
1194
+ {
1195
+ "epoch": 2.6,
1196
+ "learning_rate": 0.00019797242666631177,
1197
+ "loss": 0.4496,
1198
+ "step": 97500
1199
+ },
1200
+ {
1201
+ "epoch": 2.61,
1202
+ "learning_rate": 0.00019796202885434414,
1203
+ "loss": 0.4452,
1204
+ "step": 98000
1205
+ },
1206
+ {
1207
+ "epoch": 2.62,
1208
+ "learning_rate": 0.0001979516310423765,
1209
+ "loss": 0.4269,
1210
+ "step": 98500
1211
+ },
1212
+ {
1213
+ "epoch": 2.64,
1214
+ "learning_rate": 0.00019794123323040887,
1215
+ "loss": 0.4406,
1216
+ "step": 99000
1217
+ },
1218
+ {
1219
+ "epoch": 2.65,
1220
+ "learning_rate": 0.00019793083541844124,
1221
+ "loss": 0.4477,
1222
+ "step": 99500
1223
+ },
1224
+ {
1225
+ "epoch": 2.66,
1226
+ "learning_rate": 0.0001979204376064736,
1227
+ "loss": 0.4266,
1228
+ "step": 100000
1229
+ },
1230
+ {
1231
+ "epoch": 2.68,
1232
+ "learning_rate": 0.00019791003979450597,
1233
+ "loss": 0.4395,
1234
+ "step": 100500
1235
+ },
1236
+ {
1237
+ "epoch": 2.69,
1238
+ "learning_rate": 0.00019789964198253833,
1239
+ "loss": 0.4569,
1240
+ "step": 101000
1241
+ },
1242
+ {
1243
+ "epoch": 2.7,
1244
+ "learning_rate": 0.0001978892441705707,
1245
+ "loss": 0.4366,
1246
+ "step": 101500
1247
+ },
1248
+ {
1249
+ "epoch": 2.72,
1250
+ "learning_rate": 0.00019787884635860307,
1251
+ "loss": 0.4321,
1252
+ "step": 102000
1253
+ },
1254
+ {
1255
+ "epoch": 2.73,
1256
+ "learning_rate": 0.00019786844854663543,
1257
+ "loss": 0.439,
1258
+ "step": 102500
1259
+ },
1260
+ {
1261
+ "epoch": 2.74,
1262
+ "learning_rate": 0.0001978580507346678,
1263
+ "loss": 0.4426,
1264
+ "step": 103000
1265
+ },
1266
+ {
1267
+ "epoch": 2.76,
1268
+ "learning_rate": 0.00019784765292270016,
1269
+ "loss": 0.4348,
1270
+ "step": 103500
1271
+ },
1272
+ {
1273
+ "epoch": 2.77,
1274
+ "learning_rate": 0.00019783725511073256,
1275
+ "loss": 0.4365,
1276
+ "step": 104000
1277
+ },
1278
+ {
1279
+ "epoch": 2.78,
1280
+ "learning_rate": 0.00019782685729876492,
1281
+ "loss": 0.4439,
1282
+ "step": 104500
1283
+ },
1284
+ {
1285
+ "epoch": 2.79,
1286
+ "learning_rate": 0.0001978164594867973,
1287
+ "loss": 0.4375,
1288
+ "step": 105000
1289
+ },
1290
+ {
1291
+ "epoch": 2.81,
1292
+ "learning_rate": 0.00019780606167482965,
1293
+ "loss": 0.4506,
1294
+ "step": 105500
1295
+ },
1296
+ {
1297
+ "epoch": 2.82,
1298
+ "learning_rate": 0.00019779566386286202,
1299
+ "loss": 0.44,
1300
+ "step": 106000
1301
+ },
1302
+ {
1303
+ "epoch": 2.83,
1304
+ "learning_rate": 0.0001977852660508944,
1305
+ "loss": 0.4203,
1306
+ "step": 106500
1307
+ },
1308
+ {
1309
+ "epoch": 2.85,
1310
+ "learning_rate": 0.00019777486823892675,
1311
+ "loss": 0.4346,
1312
+ "step": 107000
1313
+ },
1314
+ {
1315
+ "epoch": 2.86,
1316
+ "learning_rate": 0.00019776447042695912,
1317
+ "loss": 0.4432,
1318
+ "step": 107500
1319
+ },
1320
+ {
1321
+ "epoch": 2.87,
1322
+ "learning_rate": 0.00019775407261499148,
1323
+ "loss": 0.4415,
1324
+ "step": 108000
1325
+ },
1326
+ {
1327
+ "epoch": 2.89,
1328
+ "learning_rate": 0.00019774367480302388,
1329
+ "loss": 0.4418,
1330
+ "step": 108500
1331
+ },
1332
+ {
1333
+ "epoch": 2.9,
1334
+ "learning_rate": 0.00019773327699105624,
1335
+ "loss": 0.441,
1336
+ "step": 109000
1337
+ },
1338
+ {
1339
+ "epoch": 2.91,
1340
+ "learning_rate": 0.0001977228791790886,
1341
+ "loss": 0.4352,
1342
+ "step": 109500
1343
+ },
1344
+ {
1345
+ "epoch": 2.93,
1346
+ "learning_rate": 0.00019771248136712098,
1347
+ "loss": 0.4348,
1348
+ "step": 110000
1349
+ },
1350
+ {
1351
+ "epoch": 2.94,
1352
+ "learning_rate": 0.00019770208355515334,
1353
+ "loss": 0.4143,
1354
+ "step": 110500
1355
+ },
1356
+ {
1357
+ "epoch": 2.95,
1358
+ "learning_rate": 0.0001976916857431857,
1359
+ "loss": 0.4189,
1360
+ "step": 111000
1361
+ },
1362
+ {
1363
+ "epoch": 2.97,
1364
+ "learning_rate": 0.00019768128793121807,
1365
+ "loss": 0.4232,
1366
+ "step": 111500
1367
+ },
1368
+ {
1369
+ "epoch": 2.98,
1370
+ "learning_rate": 0.00019767089011925044,
1371
+ "loss": 0.4272,
1372
+ "step": 112000
1373
+ },
1374
+ {
1375
+ "epoch": 2.99,
1376
+ "learning_rate": 0.0001976604923072828,
1377
+ "loss": 0.4389,
1378
+ "step": 112500
1379
+ },
1380
+ {
1381
+ "epoch": 3.0,
1382
+ "eval_bleu": 0.0877,
1383
+ "eval_gen_len": 18.3533,
1384
+ "eval_loss": 0.4100983440876007,
1385
+ "eval_runtime": 1952.3856,
1386
+ "eval_samples_per_second": 4.811,
1387
+ "eval_steps_per_second": 2.406,
1388
+ "step": 112704
1389
+ },
1390
+ {
1391
+ "epoch": 3.01,
1392
+ "learning_rate": 0.00019765009449531517,
1393
+ "loss": 0.4031,
1394
+ "step": 113000
1395
+ },
1396
+ {
1397
+ "epoch": 3.02,
1398
+ "learning_rate": 0.00019763969668334754,
1399
+ "loss": 0.3936,
1400
+ "step": 113500
1401
+ },
1402
+ {
1403
+ "epoch": 3.03,
1404
+ "learning_rate": 0.0001976292988713799,
1405
+ "loss": 0.3902,
1406
+ "step": 114000
1407
+ },
1408
+ {
1409
+ "epoch": 3.05,
1410
+ "learning_rate": 0.00019761890105941227,
1411
+ "loss": 0.3927,
1412
+ "step": 114500
1413
+ },
1414
+ {
1415
+ "epoch": 3.06,
1416
+ "learning_rate": 0.00019760850324744463,
1417
+ "loss": 0.3834,
1418
+ "step": 115000
1419
+ },
1420
+ {
1421
+ "epoch": 3.07,
1422
+ "learning_rate": 0.000197598105435477,
1423
+ "loss": 0.3901,
1424
+ "step": 115500
1425
+ },
1426
+ {
1427
+ "epoch": 3.09,
1428
+ "learning_rate": 0.00019758770762350937,
1429
+ "loss": 0.3836,
1430
+ "step": 116000
1431
+ },
1432
+ {
1433
+ "epoch": 3.1,
1434
+ "learning_rate": 0.00019757730981154173,
1435
+ "loss": 0.395,
1436
+ "step": 116500
1437
+ },
1438
+ {
1439
+ "epoch": 3.11,
1440
+ "learning_rate": 0.0001975669119995741,
1441
+ "loss": 0.3999,
1442
+ "step": 117000
1443
+ },
1444
+ {
1445
+ "epoch": 3.13,
1446
+ "learning_rate": 0.0001975565141876065,
1447
+ "loss": 0.3829,
1448
+ "step": 117500
1449
+ },
1450
+ {
1451
+ "epoch": 3.14,
1452
+ "learning_rate": 0.00019754611637563886,
1453
+ "loss": 0.3975,
1454
+ "step": 118000
1455
+ },
1456
+ {
1457
+ "epoch": 3.15,
1458
+ "learning_rate": 0.00019753571856367122,
1459
+ "loss": 0.3784,
1460
+ "step": 118500
1461
+ },
1462
+ {
1463
+ "epoch": 3.17,
1464
+ "learning_rate": 0.0001975253207517036,
1465
+ "loss": 0.4048,
1466
+ "step": 119000
1467
+ },
1468
+ {
1469
+ "epoch": 3.18,
1470
+ "learning_rate": 0.00019751492293973596,
1471
+ "loss": 0.3909,
1472
+ "step": 119500
1473
+ },
1474
+ {
1475
+ "epoch": 3.19,
1476
+ "learning_rate": 0.00019750452512776832,
1477
+ "loss": 0.4005,
1478
+ "step": 120000
1479
+ },
1480
+ {
1481
+ "epoch": 3.21,
1482
+ "learning_rate": 0.0001974941273158007,
1483
+ "loss": 0.3976,
1484
+ "step": 120500
1485
+ },
1486
+ {
1487
+ "epoch": 3.22,
1488
+ "learning_rate": 0.00019748372950383305,
1489
+ "loss": 0.3971,
1490
+ "step": 121000
1491
+ },
1492
+ {
1493
+ "epoch": 3.23,
1494
+ "learning_rate": 0.00019747333169186545,
1495
+ "loss": 0.3968,
1496
+ "step": 121500
1497
+ },
1498
+ {
1499
+ "epoch": 3.25,
1500
+ "learning_rate": 0.0001974629338798978,
1501
+ "loss": 0.3856,
1502
+ "step": 122000
1503
+ },
1504
+ {
1505
+ "epoch": 3.26,
1506
+ "learning_rate": 0.00019745253606793018,
1507
+ "loss": 0.3833,
1508
+ "step": 122500
1509
+ },
1510
+ {
1511
+ "epoch": 3.27,
1512
+ "learning_rate": 0.00019744213825596254,
1513
+ "loss": 0.4037,
1514
+ "step": 123000
1515
+ },
1516
+ {
1517
+ "epoch": 3.29,
1518
+ "learning_rate": 0.0001974317404439949,
1519
+ "loss": 0.3821,
1520
+ "step": 123500
1521
+ },
1522
+ {
1523
+ "epoch": 3.3,
1524
+ "learning_rate": 0.00019742134263202728,
1525
+ "loss": 0.3986,
1526
+ "step": 124000
1527
+ },
1528
+ {
1529
+ "epoch": 3.31,
1530
+ "learning_rate": 0.00019741094482005964,
1531
+ "loss": 0.3838,
1532
+ "step": 124500
1533
+ },
1534
+ {
1535
+ "epoch": 3.33,
1536
+ "learning_rate": 0.000197400547008092,
1537
+ "loss": 0.3827,
1538
+ "step": 125000
1539
+ },
1540
+ {
1541
+ "epoch": 3.34,
1542
+ "learning_rate": 0.00019739014919612437,
1543
+ "loss": 0.3914,
1544
+ "step": 125500
1545
+ },
1546
+ {
1547
+ "epoch": 3.35,
1548
+ "learning_rate": 0.00019737975138415674,
1549
+ "loss": 0.3923,
1550
+ "step": 126000
1551
+ },
1552
+ {
1553
+ "epoch": 3.37,
1554
+ "learning_rate": 0.0001973693535721891,
1555
+ "loss": 0.3982,
1556
+ "step": 126500
1557
+ },
1558
+ {
1559
+ "epoch": 3.38,
1560
+ "learning_rate": 0.00019735895576022147,
1561
+ "loss": 0.3915,
1562
+ "step": 127000
1563
+ },
1564
+ {
1565
+ "epoch": 3.39,
1566
+ "learning_rate": 0.00019734855794825384,
1567
+ "loss": 0.4022,
1568
+ "step": 127500
1569
+ },
1570
+ {
1571
+ "epoch": 3.41,
1572
+ "learning_rate": 0.0001973381601362862,
1573
+ "loss": 0.3888,
1574
+ "step": 128000
1575
+ },
1576
+ {
1577
+ "epoch": 3.42,
1578
+ "learning_rate": 0.00019732776232431857,
1579
+ "loss": 0.3855,
1580
+ "step": 128500
1581
+ },
1582
+ {
1583
+ "epoch": 3.43,
1584
+ "learning_rate": 0.00019731736451235093,
1585
+ "loss": 0.3947,
1586
+ "step": 129000
1587
+ },
1588
+ {
1589
+ "epoch": 3.45,
1590
+ "learning_rate": 0.0001973069667003833,
1591
+ "loss": 0.3877,
1592
+ "step": 129500
1593
+ },
1594
+ {
1595
+ "epoch": 3.46,
1596
+ "learning_rate": 0.00019729656888841567,
1597
+ "loss": 0.3921,
1598
+ "step": 130000
1599
+ },
1600
+ {
1601
+ "epoch": 3.47,
1602
+ "learning_rate": 0.00019728617107644803,
1603
+ "loss": 0.3994,
1604
+ "step": 130500
1605
+ },
1606
+ {
1607
+ "epoch": 3.49,
1608
+ "learning_rate": 0.00019727577326448043,
1609
+ "loss": 0.3949,
1610
+ "step": 131000
1611
+ },
1612
+ {
1613
+ "epoch": 3.5,
1614
+ "learning_rate": 0.0001972653754525128,
1615
+ "loss": 0.3901,
1616
+ "step": 131500
1617
+ },
1618
+ {
1619
+ "epoch": 3.51,
1620
+ "learning_rate": 0.00019725497764054516,
1621
+ "loss": 0.3954,
1622
+ "step": 132000
1623
+ },
1624
+ {
1625
+ "epoch": 3.53,
1626
+ "learning_rate": 0.00019724457982857752,
1627
+ "loss": 0.3962,
1628
+ "step": 132500
1629
+ },
1630
+ {
1631
+ "epoch": 3.54,
1632
+ "learning_rate": 0.0001972341820166099,
1633
+ "loss": 0.405,
1634
+ "step": 133000
1635
+ },
1636
+ {
1637
+ "epoch": 3.55,
1638
+ "learning_rate": 0.00019722378420464226,
1639
+ "loss": 0.3985,
1640
+ "step": 133500
1641
+ },
1642
+ {
1643
+ "epoch": 3.57,
1644
+ "learning_rate": 0.00019721338639267465,
1645
+ "loss": 0.3763,
1646
+ "step": 134000
1647
+ },
1648
+ {
1649
+ "epoch": 3.58,
1650
+ "learning_rate": 0.00019720298858070701,
1651
+ "loss": 0.4068,
1652
+ "step": 134500
1653
+ },
1654
+ {
1655
+ "epoch": 3.59,
1656
+ "learning_rate": 0.00019719259076873938,
1657
+ "loss": 0.3885,
1658
+ "step": 135000
1659
+ },
1660
+ {
1661
+ "epoch": 3.61,
1662
+ "learning_rate": 0.00019718219295677175,
1663
+ "loss": 0.3833,
1664
+ "step": 135500
1665
+ },
1666
+ {
1667
+ "epoch": 3.62,
1668
+ "learning_rate": 0.0001971717951448041,
1669
+ "loss": 0.3864,
1670
+ "step": 136000
1671
+ },
1672
+ {
1673
+ "epoch": 3.63,
1674
+ "learning_rate": 0.00019716139733283648,
1675
+ "loss": 0.3765,
1676
+ "step": 136500
1677
+ },
1678
+ {
1679
+ "epoch": 3.65,
1680
+ "learning_rate": 0.00019715099952086884,
1681
+ "loss": 0.3988,
1682
+ "step": 137000
1683
+ },
1684
+ {
1685
+ "epoch": 3.66,
1686
+ "learning_rate": 0.0001971406017089012,
1687
+ "loss": 0.3867,
1688
+ "step": 137500
1689
+ },
1690
+ {
1691
+ "epoch": 3.67,
1692
+ "learning_rate": 0.00019713020389693358,
1693
+ "loss": 0.3901,
1694
+ "step": 138000
1695
+ },
1696
+ {
1697
+ "epoch": 3.69,
1698
+ "learning_rate": 0.00019711980608496594,
1699
+ "loss": 0.3822,
1700
+ "step": 138500
1701
+ },
1702
+ {
1703
+ "epoch": 3.7,
1704
+ "learning_rate": 0.0001971094082729983,
1705
+ "loss": 0.3864,
1706
+ "step": 139000
1707
+ },
1708
+ {
1709
+ "epoch": 3.71,
1710
+ "learning_rate": 0.00019709901046103067,
1711
+ "loss": 0.3886,
1712
+ "step": 139500
1713
+ },
1714
+ {
1715
+ "epoch": 3.73,
1716
+ "learning_rate": 0.00019708861264906304,
1717
+ "loss": 0.3715,
1718
+ "step": 140000
1719
+ },
1720
+ {
1721
+ "epoch": 3.74,
1722
+ "learning_rate": 0.0001970782148370954,
1723
+ "loss": 0.388,
1724
+ "step": 140500
1725
+ },
1726
+ {
1727
+ "epoch": 3.75,
1728
+ "learning_rate": 0.00019706781702512777,
1729
+ "loss": 0.3941,
1730
+ "step": 141000
1731
+ },
1732
+ {
1733
+ "epoch": 3.77,
1734
+ "learning_rate": 0.00019705741921316014,
1735
+ "loss": 0.3922,
1736
+ "step": 141500
1737
+ },
1738
+ {
1739
+ "epoch": 3.78,
1740
+ "learning_rate": 0.0001970470214011925,
1741
+ "loss": 0.3918,
1742
+ "step": 142000
1743
+ },
1744
+ {
1745
+ "epoch": 3.79,
1746
+ "learning_rate": 0.00019703662358922487,
1747
+ "loss": 0.3898,
1748
+ "step": 142500
1749
+ },
1750
+ {
1751
+ "epoch": 3.81,
1752
+ "learning_rate": 0.00019702622577725724,
1753
+ "loss": 0.384,
1754
+ "step": 143000
1755
+ },
1756
+ {
1757
+ "epoch": 3.82,
1758
+ "learning_rate": 0.0001970158279652896,
1759
+ "loss": 0.3875,
1760
+ "step": 143500
1761
+ },
1762
+ {
1763
+ "epoch": 3.83,
1764
+ "learning_rate": 0.00019700543015332197,
1765
+ "loss": 0.3893,
1766
+ "step": 144000
1767
+ },
1768
+ {
1769
+ "epoch": 3.85,
1770
+ "learning_rate": 0.00019699503234135433,
1771
+ "loss": 0.3883,
1772
+ "step": 144500
1773
+ },
1774
+ {
1775
+ "epoch": 3.86,
1776
+ "learning_rate": 0.00019698463452938673,
1777
+ "loss": 0.3866,
1778
+ "step": 145000
1779
+ },
1780
+ {
1781
+ "epoch": 3.87,
1782
+ "learning_rate": 0.0001969742367174191,
1783
+ "loss": 0.3901,
1784
+ "step": 145500
1785
+ },
1786
+ {
1787
+ "epoch": 3.89,
1788
+ "learning_rate": 0.00019696383890545146,
1789
+ "loss": 0.4022,
1790
+ "step": 146000
1791
+ },
1792
+ {
1793
+ "epoch": 3.9,
1794
+ "learning_rate": 0.00019695344109348382,
1795
+ "loss": 0.4021,
1796
+ "step": 146500
1797
+ },
1798
+ {
1799
+ "epoch": 3.91,
1800
+ "learning_rate": 0.00019694304328151622,
1801
+ "loss": 0.3873,
1802
+ "step": 147000
1803
+ },
1804
+ {
1805
+ "epoch": 3.93,
1806
+ "learning_rate": 0.00019693264546954858,
1807
+ "loss": 0.3844,
1808
+ "step": 147500
1809
+ },
1810
+ {
1811
+ "epoch": 3.94,
1812
+ "learning_rate": 0.00019692224765758095,
1813
+ "loss": 0.3824,
1814
+ "step": 148000
1815
+ },
1816
+ {
1817
+ "epoch": 3.95,
1818
+ "learning_rate": 0.00019691184984561331,
1819
+ "loss": 0.392,
1820
+ "step": 148500
1821
+ },
1822
+ {
1823
+ "epoch": 3.97,
1824
+ "learning_rate": 0.00019690145203364568,
1825
+ "loss": 0.407,
1826
+ "step": 149000
1827
+ },
1828
+ {
1829
+ "epoch": 3.98,
1830
+ "learning_rate": 0.00019689105422167805,
1831
+ "loss": 0.3864,
1832
+ "step": 149500
1833
+ },
1834
+ {
1835
+ "epoch": 3.99,
1836
+ "learning_rate": 0.0001968806564097104,
1837
+ "loss": 0.3812,
1838
+ "step": 150000
1839
+ },
1840
+ {
1841
+ "epoch": 4.0,
1842
+ "eval_bleu": 0.09,
1843
+ "eval_gen_len": 18.3779,
1844
+ "eval_loss": 0.39251887798309326,
1845
+ "eval_runtime": 1926.9186,
1846
+ "eval_samples_per_second": 4.875,
1847
+ "eval_steps_per_second": 2.438,
1848
+ "step": 150272
1849
+ },
1850
+ {
1851
+ "epoch": 4.01,
1852
+ "learning_rate": 0.00019687025859774278,
1853
+ "loss": 0.366,
1854
+ "step": 150500
1855
+ },
1856
+ {
1857
+ "epoch": 4.02,
1858
+ "learning_rate": 0.00019685986078577514,
1859
+ "loss": 0.3288,
1860
+ "step": 151000
1861
+ },
1862
+ {
1863
+ "epoch": 4.03,
1864
+ "learning_rate": 0.0001968494629738075,
1865
+ "loss": 0.3431,
1866
+ "step": 151500
1867
+ },
1868
+ {
1869
+ "epoch": 4.05,
1870
+ "learning_rate": 0.00019683906516183988,
1871
+ "loss": 0.3462,
1872
+ "step": 152000
1873
+ },
1874
+ {
1875
+ "epoch": 4.06,
1876
+ "learning_rate": 0.00019682866734987224,
1877
+ "loss": 0.3496,
1878
+ "step": 152500
1879
+ },
1880
+ {
1881
+ "epoch": 4.07,
1882
+ "learning_rate": 0.0001968182695379046,
1883
+ "loss": 0.3385,
1884
+ "step": 153000
1885
+ },
1886
+ {
1887
+ "epoch": 4.09,
1888
+ "learning_rate": 0.00019680787172593697,
1889
+ "loss": 0.3384,
1890
+ "step": 153500
1891
+ },
1892
+ {
1893
+ "epoch": 4.1,
1894
+ "learning_rate": 0.00019679747391396934,
1895
+ "loss": 0.3389,
1896
+ "step": 154000
1897
+ },
1898
+ {
1899
+ "epoch": 4.11,
1900
+ "learning_rate": 0.0001967870761020017,
1901
+ "loss": 0.3292,
1902
+ "step": 154500
1903
+ },
1904
+ {
1905
+ "epoch": 4.13,
1906
+ "learning_rate": 0.00019677667829003407,
1907
+ "loss": 0.337,
1908
+ "step": 155000
1909
+ },
1910
+ {
1911
+ "epoch": 4.14,
1912
+ "learning_rate": 0.00019676628047806644,
1913
+ "loss": 0.3476,
1914
+ "step": 155500
1915
+ },
1916
+ {
1917
+ "epoch": 4.15,
1918
+ "learning_rate": 0.0001967558826660988,
1919
+ "loss": 0.3404,
1920
+ "step": 156000
1921
+ },
1922
+ {
1923
+ "epoch": 4.17,
1924
+ "learning_rate": 0.00019674548485413117,
1925
+ "loss": 0.3466,
1926
+ "step": 156500
1927
+ },
1928
+ {
1929
+ "epoch": 4.18,
1930
+ "learning_rate": 0.00019673508704216354,
1931
+ "loss": 0.3427,
1932
+ "step": 157000
1933
+ },
1934
+ {
1935
+ "epoch": 4.19,
1936
+ "learning_rate": 0.0001967246892301959,
1937
+ "loss": 0.3435,
1938
+ "step": 157500
1939
+ },
1940
+ {
1941
+ "epoch": 4.21,
1942
+ "learning_rate": 0.00019671429141822827,
1943
+ "loss": 0.3516,
1944
+ "step": 158000
1945
+ },
1946
+ {
1947
+ "epoch": 4.22,
1948
+ "learning_rate": 0.00019670389360626066,
1949
+ "loss": 0.3553,
1950
+ "step": 158500
1951
+ },
1952
+ {
1953
+ "epoch": 4.23,
1954
+ "learning_rate": 0.00019669349579429303,
1955
+ "loss": 0.3511,
1956
+ "step": 159000
1957
+ },
1958
+ {
1959
+ "epoch": 4.25,
1960
+ "learning_rate": 0.0001966830979823254,
1961
+ "loss": 0.3376,
1962
+ "step": 159500
1963
+ },
1964
+ {
1965
+ "epoch": 4.26,
1966
+ "learning_rate": 0.00019667270017035776,
1967
+ "loss": 0.3475,
1968
+ "step": 160000
1969
+ },
1970
+ {
1971
+ "epoch": 4.27,
1972
+ "learning_rate": 0.00019666230235839015,
1973
+ "loss": 0.3548,
1974
+ "step": 160500
1975
+ },
1976
+ {
1977
+ "epoch": 4.29,
1978
+ "learning_rate": 0.00019665190454642252,
1979
+ "loss": 0.3595,
1980
+ "step": 161000
1981
+ },
1982
+ {
1983
+ "epoch": 4.3,
1984
+ "learning_rate": 0.00019664150673445488,
1985
+ "loss": 0.339,
1986
+ "step": 161500
1987
+ },
1988
+ {
1989
+ "epoch": 4.31,
1990
+ "learning_rate": 0.00019663110892248725,
1991
+ "loss": 0.3568,
1992
+ "step": 162000
1993
+ },
1994
+ {
1995
+ "epoch": 4.33,
1996
+ "learning_rate": 0.00019662071111051962,
1997
+ "loss": 0.3578,
1998
+ "step": 162500
1999
+ },
2000
+ {
2001
+ "epoch": 4.34,
2002
+ "learning_rate": 0.00019661031329855198,
2003
+ "loss": 0.3465,
2004
+ "step": 163000
2005
+ },
2006
+ {
2007
+ "epoch": 4.35,
2008
+ "learning_rate": 0.00019659991548658435,
2009
+ "loss": 0.3601,
2010
+ "step": 163500
2011
+ },
2012
+ {
2013
+ "epoch": 4.37,
2014
+ "learning_rate": 0.0001965895176746167,
2015
+ "loss": 0.3594,
2016
+ "step": 164000
2017
+ },
2018
+ {
2019
+ "epoch": 4.38,
2020
+ "learning_rate": 0.00019657911986264908,
2021
+ "loss": 0.3442,
2022
+ "step": 164500
2023
+ },
2024
+ {
2025
+ "epoch": 4.39,
2026
+ "learning_rate": 0.00019656872205068144,
2027
+ "loss": 0.3395,
2028
+ "step": 165000
2029
+ },
2030
+ {
2031
+ "epoch": 4.41,
2032
+ "learning_rate": 0.0001965583242387138,
2033
+ "loss": 0.3419,
2034
+ "step": 165500
2035
+ },
2036
+ {
2037
+ "epoch": 4.42,
2038
+ "learning_rate": 0.00019654792642674618,
2039
+ "loss": 0.3438,
2040
+ "step": 166000
2041
+ },
2042
+ {
2043
+ "epoch": 4.43,
2044
+ "learning_rate": 0.00019653752861477854,
2045
+ "loss": 0.3533,
2046
+ "step": 166500
2047
+ },
2048
+ {
2049
+ "epoch": 4.45,
2050
+ "learning_rate": 0.0001965271308028109,
2051
+ "loss": 0.3508,
2052
+ "step": 167000
2053
+ },
2054
+ {
2055
+ "epoch": 4.46,
2056
+ "learning_rate": 0.00019651673299084327,
2057
+ "loss": 0.3589,
2058
+ "step": 167500
2059
+ },
2060
+ {
2061
+ "epoch": 4.47,
2062
+ "learning_rate": 0.00019650633517887564,
2063
+ "loss": 0.3492,
2064
+ "step": 168000
2065
+ },
2066
+ {
2067
+ "epoch": 4.49,
2068
+ "learning_rate": 0.000196495937366908,
2069
+ "loss": 0.3466,
2070
+ "step": 168500
2071
+ },
2072
+ {
2073
+ "epoch": 4.5,
2074
+ "learning_rate": 0.00019648553955494037,
2075
+ "loss": 0.3509,
2076
+ "step": 169000
2077
+ },
2078
+ {
2079
+ "epoch": 4.51,
2080
+ "learning_rate": 0.00019647514174297274,
2081
+ "loss": 0.3584,
2082
+ "step": 169500
2083
+ },
2084
+ {
2085
+ "epoch": 4.53,
2086
+ "learning_rate": 0.0001964647439310051,
2087
+ "loss": 0.3563,
2088
+ "step": 170000
2089
+ },
2090
+ {
2091
+ "epoch": 4.54,
2092
+ "learning_rate": 0.00019645434611903747,
2093
+ "loss": 0.3431,
2094
+ "step": 170500
2095
+ },
2096
+ {
2097
+ "epoch": 4.55,
2098
+ "learning_rate": 0.00019644394830706984,
2099
+ "loss": 0.3626,
2100
+ "step": 171000
2101
+ },
2102
+ {
2103
+ "epoch": 4.57,
2104
+ "learning_rate": 0.0001964335504951022,
2105
+ "loss": 0.3505,
2106
+ "step": 171500
2107
+ },
2108
+ {
2109
+ "epoch": 4.58,
2110
+ "learning_rate": 0.0001964231526831346,
2111
+ "loss": 0.3638,
2112
+ "step": 172000
2113
+ },
2114
+ {
2115
+ "epoch": 4.59,
2116
+ "learning_rate": 0.00019641275487116696,
2117
+ "loss": 0.3485,
2118
+ "step": 172500
2119
+ },
2120
+ {
2121
+ "epoch": 4.6,
2122
+ "learning_rate": 0.00019640235705919933,
2123
+ "loss": 0.3613,
2124
+ "step": 173000
2125
+ },
2126
+ {
2127
+ "epoch": 4.62,
2128
+ "learning_rate": 0.0001963919592472317,
2129
+ "loss": 0.3576,
2130
+ "step": 173500
2131
+ },
2132
+ {
2133
+ "epoch": 4.63,
2134
+ "learning_rate": 0.00019638156143526406,
2135
+ "loss": 0.3556,
2136
+ "step": 174000
2137
+ },
2138
+ {
2139
+ "epoch": 4.64,
2140
+ "learning_rate": 0.00019637116362329645,
2141
+ "loss": 0.3552,
2142
+ "step": 174500
2143
+ },
2144
+ {
2145
+ "epoch": 4.66,
2146
+ "learning_rate": 0.00019636076581132882,
2147
+ "loss": 0.3521,
2148
+ "step": 175000
2149
+ },
2150
+ {
2151
+ "epoch": 4.67,
2152
+ "learning_rate": 0.00019635036799936118,
2153
+ "loss": 0.3629,
2154
+ "step": 175500
2155
+ },
2156
+ {
2157
+ "epoch": 4.68,
2158
+ "learning_rate": 0.00019633997018739355,
2159
+ "loss": 0.3416,
2160
+ "step": 176000
2161
+ },
2162
+ {
2163
+ "epoch": 4.7,
2164
+ "learning_rate": 0.00019632957237542592,
2165
+ "loss": 0.3583,
2166
+ "step": 176500
2167
+ },
2168
+ {
2169
+ "epoch": 4.71,
2170
+ "learning_rate": 0.00019631917456345828,
2171
+ "loss": 0.3697,
2172
+ "step": 177000
2173
+ },
2174
+ {
2175
+ "epoch": 4.72,
2176
+ "learning_rate": 0.00019630877675149065,
2177
+ "loss": 0.3403,
2178
+ "step": 177500
2179
+ },
2180
+ {
2181
+ "epoch": 4.74,
2182
+ "learning_rate": 0.000196298378939523,
2183
+ "loss": 0.3639,
2184
+ "step": 178000
2185
+ },
2186
+ {
2187
+ "epoch": 4.75,
2188
+ "learning_rate": 0.00019628798112755538,
2189
+ "loss": 0.3585,
2190
+ "step": 178500
2191
+ },
2192
+ {
2193
+ "epoch": 4.76,
2194
+ "learning_rate": 0.00019627758331558774,
2195
+ "loss": 0.3365,
2196
+ "step": 179000
2197
+ },
2198
+ {
2199
+ "epoch": 4.78,
2200
+ "learning_rate": 0.0001962671855036201,
2201
+ "loss": 0.3497,
2202
+ "step": 179500
2203
+ },
2204
+ {
2205
+ "epoch": 4.79,
2206
+ "learning_rate": 0.00019625678769165248,
2207
+ "loss": 0.3505,
2208
+ "step": 180000
2209
+ },
2210
+ {
2211
+ "epoch": 4.8,
2212
+ "learning_rate": 0.00019624638987968484,
2213
+ "loss": 0.3592,
2214
+ "step": 180500
2215
+ },
2216
+ {
2217
+ "epoch": 4.82,
2218
+ "learning_rate": 0.0001962359920677172,
2219
+ "loss": 0.3585,
2220
+ "step": 181000
2221
+ },
2222
+ {
2223
+ "epoch": 4.83,
2224
+ "learning_rate": 0.00019622559425574957,
2225
+ "loss": 0.355,
2226
+ "step": 181500
2227
+ },
2228
+ {
2229
+ "epoch": 4.84,
2230
+ "learning_rate": 0.00019621519644378194,
2231
+ "loss": 0.3688,
2232
+ "step": 182000
2233
+ },
2234
+ {
2235
+ "epoch": 4.86,
2236
+ "learning_rate": 0.0001962047986318143,
2237
+ "loss": 0.3431,
2238
+ "step": 182500
2239
+ },
2240
+ {
2241
+ "epoch": 4.87,
2242
+ "learning_rate": 0.00019619440081984667,
2243
+ "loss": 0.3733,
2244
+ "step": 183000
2245
+ },
2246
+ {
2247
+ "epoch": 4.88,
2248
+ "learning_rate": 0.00019618400300787904,
2249
+ "loss": 0.3615,
2250
+ "step": 183500
2251
+ },
2252
+ {
2253
+ "epoch": 4.9,
2254
+ "learning_rate": 0.0001961736051959114,
2255
+ "loss": 0.3455,
2256
+ "step": 184000
2257
+ },
2258
+ {
2259
+ "epoch": 4.91,
2260
+ "learning_rate": 0.00019616320738394377,
2261
+ "loss": 0.3517,
2262
+ "step": 184500
2263
+ },
2264
+ {
2265
+ "epoch": 4.92,
2266
+ "learning_rate": 0.00019615280957197614,
2267
+ "loss": 0.3566,
2268
+ "step": 185000
2269
+ },
2270
+ {
2271
+ "epoch": 4.94,
2272
+ "learning_rate": 0.00019614241176000853,
2273
+ "loss": 0.3476,
2274
+ "step": 185500
2275
+ },
2276
+ {
2277
+ "epoch": 4.95,
2278
+ "learning_rate": 0.0001961320139480409,
2279
+ "loss": 0.3465,
2280
+ "step": 186000
2281
+ },
2282
+ {
2283
+ "epoch": 4.96,
2284
+ "learning_rate": 0.00019612161613607326,
2285
+ "loss": 0.3526,
2286
+ "step": 186500
2287
+ },
2288
+ {
2289
+ "epoch": 4.98,
2290
+ "learning_rate": 0.00019611121832410563,
2291
+ "loss": 0.3492,
2292
+ "step": 187000
2293
+ },
2294
+ {
2295
+ "epoch": 4.99,
2296
+ "learning_rate": 0.000196100820512138,
2297
+ "loss": 0.3724,
2298
+ "step": 187500
2299
+ },
2300
+ {
2301
+ "epoch": 5.0,
2302
+ "eval_bleu": 0.0937,
2303
+ "eval_gen_len": 18.3788,
2304
+ "eval_loss": 0.37562912702560425,
2305
+ "eval_runtime": 1978.9757,
2306
+ "eval_samples_per_second": 4.746,
2307
+ "eval_steps_per_second": 2.373,
2308
+ "step": 187840
2309
+ },
2310
+ {
2311
+ "epoch": 5.0,
2312
+ "learning_rate": 0.00019609042270017039,
2313
+ "loss": 0.3439,
2314
+ "step": 188000
2315
+ },
2316
+ {
2317
+ "epoch": 5.02,
2318
+ "learning_rate": 0.00019608002488820275,
2319
+ "loss": 0.3028,
2320
+ "step": 188500
2321
+ },
2322
+ {
2323
+ "epoch": 5.03,
2324
+ "learning_rate": 0.00019606962707623512,
2325
+ "loss": 0.3119,
2326
+ "step": 189000
2327
+ },
2328
+ {
2329
+ "epoch": 5.04,
2330
+ "learning_rate": 0.00019605922926426748,
2331
+ "loss": 0.3091,
2332
+ "step": 189500
2333
+ },
2334
+ {
2335
+ "epoch": 5.06,
2336
+ "learning_rate": 0.00019604883145229985,
2337
+ "loss": 0.3061,
2338
+ "step": 190000
2339
+ },
2340
+ {
2341
+ "epoch": 5.07,
2342
+ "learning_rate": 0.00019603843364033222,
2343
+ "loss": 0.3135,
2344
+ "step": 190500
2345
+ },
2346
+ {
2347
+ "epoch": 5.08,
2348
+ "learning_rate": 0.00019602803582836458,
2349
+ "loss": 0.3028,
2350
+ "step": 191000
2351
+ },
2352
+ {
2353
+ "epoch": 5.1,
2354
+ "learning_rate": 0.00019601763801639695,
2355
+ "loss": 0.3112,
2356
+ "step": 191500
2357
+ },
2358
+ {
2359
+ "epoch": 5.11,
2360
+ "learning_rate": 0.0001960072402044293,
2361
+ "loss": 0.3101,
2362
+ "step": 192000
2363
+ },
2364
+ {
2365
+ "epoch": 5.12,
2366
+ "learning_rate": 0.00019599684239246168,
2367
+ "loss": 0.3264,
2368
+ "step": 192500
2369
+ },
2370
+ {
2371
+ "epoch": 5.14,
2372
+ "learning_rate": 0.00019598644458049405,
2373
+ "loss": 0.3087,
2374
+ "step": 193000
2375
+ },
2376
+ {
2377
+ "epoch": 5.15,
2378
+ "learning_rate": 0.0001959760467685264,
2379
+ "loss": 0.3197,
2380
+ "step": 193500
2381
+ },
2382
+ {
2383
+ "epoch": 5.16,
2384
+ "learning_rate": 0.00019596564895655878,
2385
+ "loss": 0.3071,
2386
+ "step": 194000
2387
+ },
2388
+ {
2389
+ "epoch": 5.18,
2390
+ "learning_rate": 0.00019595525114459114,
2391
+ "loss": 0.3106,
2392
+ "step": 194500
2393
+ },
2394
+ {
2395
+ "epoch": 5.19,
2396
+ "learning_rate": 0.0001959448533326235,
2397
+ "loss": 0.3187,
2398
+ "step": 195000
2399
+ },
2400
+ {
2401
+ "epoch": 5.2,
2402
+ "learning_rate": 0.00019593445552065587,
2403
+ "loss": 0.3206,
2404
+ "step": 195500
2405
+ },
2406
+ {
2407
+ "epoch": 5.22,
2408
+ "learning_rate": 0.00019592405770868824,
2409
+ "loss": 0.3081,
2410
+ "step": 196000
2411
+ },
2412
+ {
2413
+ "epoch": 5.23,
2414
+ "learning_rate": 0.0001959136598967206,
2415
+ "loss": 0.3191,
2416
+ "step": 196500
2417
+ },
2418
+ {
2419
+ "epoch": 5.24,
2420
+ "learning_rate": 0.00019590326208475297,
2421
+ "loss": 0.3177,
2422
+ "step": 197000
2423
+ },
2424
+ {
2425
+ "epoch": 5.26,
2426
+ "learning_rate": 0.00019589286427278534,
2427
+ "loss": 0.3197,
2428
+ "step": 197500
2429
+ },
2430
+ {
2431
+ "epoch": 5.27,
2432
+ "learning_rate": 0.00019588246646081773,
2433
+ "loss": 0.3189,
2434
+ "step": 198000
2435
+ },
2436
+ {
2437
+ "epoch": 5.28,
2438
+ "learning_rate": 0.0001958720686488501,
2439
+ "loss": 0.3172,
2440
+ "step": 198500
2441
+ },
2442
+ {
2443
+ "epoch": 5.3,
2444
+ "learning_rate": 0.00019586167083688246,
2445
+ "loss": 0.3183,
2446
+ "step": 199000
2447
+ },
2448
+ {
2449
+ "epoch": 5.31,
2450
+ "learning_rate": 0.00019585127302491483,
2451
+ "loss": 0.3166,
2452
+ "step": 199500
2453
+ },
2454
+ {
2455
+ "epoch": 5.32,
2456
+ "learning_rate": 0.0001958408752129472,
2457
+ "loss": 0.3154,
2458
+ "step": 200000
2459
+ },
2460
+ {
2461
+ "epoch": 5.34,
2462
+ "learning_rate": 0.00019583047740097956,
2463
+ "loss": 0.3263,
2464
+ "step": 200500
2465
+ },
2466
+ {
2467
+ "epoch": 5.35,
2468
+ "learning_rate": 0.00019582007958901193,
2469
+ "loss": 0.3001,
2470
+ "step": 201000
2471
+ },
2472
+ {
2473
+ "epoch": 5.36,
2474
+ "learning_rate": 0.00019580968177704432,
2475
+ "loss": 0.3107,
2476
+ "step": 201500
2477
+ },
2478
+ {
2479
+ "epoch": 5.38,
2480
+ "learning_rate": 0.00019579928396507669,
2481
+ "loss": 0.3084,
2482
+ "step": 202000
2483
+ },
2484
+ {
2485
+ "epoch": 5.39,
2486
+ "learning_rate": 0.00019578888615310905,
2487
+ "loss": 0.3189,
2488
+ "step": 202500
2489
+ },
2490
+ {
2491
+ "epoch": 5.4,
2492
+ "learning_rate": 0.00019577848834114142,
2493
+ "loss": 0.3198,
2494
+ "step": 203000
2495
+ },
2496
+ {
2497
+ "epoch": 5.42,
2498
+ "learning_rate": 0.00019576809052917378,
2499
+ "loss": 0.3187,
2500
+ "step": 203500
2501
+ },
2502
+ {
2503
+ "epoch": 5.43,
2504
+ "learning_rate": 0.00019575769271720615,
2505
+ "loss": 0.3137,
2506
+ "step": 204000
2507
+ },
2508
+ {
2509
+ "epoch": 5.44,
2510
+ "learning_rate": 0.00019574729490523852,
2511
+ "loss": 0.3232,
2512
+ "step": 204500
2513
+ },
2514
+ {
2515
+ "epoch": 5.46,
2516
+ "learning_rate": 0.00019573689709327088,
2517
+ "loss": 0.3193,
2518
+ "step": 205000
2519
+ },
2520
+ {
2521
+ "epoch": 5.47,
2522
+ "learning_rate": 0.00019572649928130325,
2523
+ "loss": 0.3097,
2524
+ "step": 205500
2525
+ },
2526
+ {
2527
+ "epoch": 5.48,
2528
+ "learning_rate": 0.00019571610146933561,
2529
+ "loss": 0.3185,
2530
+ "step": 206000
2531
+ },
2532
+ {
2533
+ "epoch": 5.5,
2534
+ "learning_rate": 0.00019570570365736798,
2535
+ "loss": 0.3319,
2536
+ "step": 206500
2537
+ },
2538
+ {
2539
+ "epoch": 5.51,
2540
+ "learning_rate": 0.00019569530584540035,
2541
+ "loss": 0.328,
2542
+ "step": 207000
2543
+ },
2544
+ {
2545
+ "epoch": 5.52,
2546
+ "learning_rate": 0.0001956849080334327,
2547
+ "loss": 0.3228,
2548
+ "step": 207500
2549
+ },
2550
+ {
2551
+ "epoch": 5.54,
2552
+ "learning_rate": 0.00019567451022146508,
2553
+ "loss": 0.3299,
2554
+ "step": 208000
2555
+ },
2556
+ {
2557
+ "epoch": 5.55,
2558
+ "learning_rate": 0.00019566411240949744,
2559
+ "loss": 0.3154,
2560
+ "step": 208500
2561
+ },
2562
+ {
2563
+ "epoch": 5.56,
2564
+ "learning_rate": 0.0001956537145975298,
2565
+ "loss": 0.3183,
2566
+ "step": 209000
2567
+ },
2568
+ {
2569
+ "epoch": 5.58,
2570
+ "learning_rate": 0.00019564331678556218,
2571
+ "loss": 0.3235,
2572
+ "step": 209500
2573
+ },
2574
+ {
2575
+ "epoch": 5.59,
2576
+ "learning_rate": 0.00019563291897359454,
2577
+ "loss": 0.3179,
2578
+ "step": 210000
2579
+ },
2580
+ {
2581
+ "epoch": 5.6,
2582
+ "learning_rate": 0.0001956225211616269,
2583
+ "loss": 0.3211,
2584
+ "step": 210500
2585
+ },
2586
+ {
2587
+ "epoch": 5.62,
2588
+ "learning_rate": 0.0001956121233496593,
2589
+ "loss": 0.315,
2590
+ "step": 211000
2591
+ },
2592
+ {
2593
+ "epoch": 5.63,
2594
+ "learning_rate": 0.00019560172553769167,
2595
+ "loss": 0.3191,
2596
+ "step": 211500
2597
+ },
2598
+ {
2599
+ "epoch": 5.64,
2600
+ "learning_rate": 0.00019559132772572403,
2601
+ "loss": 0.3341,
2602
+ "step": 212000
2603
+ },
2604
+ {
2605
+ "epoch": 5.66,
2606
+ "learning_rate": 0.0001955809299137564,
2607
+ "loss": 0.3229,
2608
+ "step": 212500
2609
+ },
2610
+ {
2611
+ "epoch": 5.67,
2612
+ "learning_rate": 0.00019557053210178876,
2613
+ "loss": 0.3277,
2614
+ "step": 213000
2615
+ },
2616
+ {
2617
+ "epoch": 5.68,
2618
+ "learning_rate": 0.00019556013428982113,
2619
+ "loss": 0.3233,
2620
+ "step": 213500
2621
+ },
2622
+ {
2623
+ "epoch": 5.7,
2624
+ "learning_rate": 0.0001955497364778535,
2625
+ "loss": 0.3302,
2626
+ "step": 214000
2627
+ },
2628
+ {
2629
+ "epoch": 5.71,
2630
+ "learning_rate": 0.00019553933866588586,
2631
+ "loss": 0.324,
2632
+ "step": 214500
2633
+ },
2634
+ {
2635
+ "epoch": 5.72,
2636
+ "learning_rate": 0.00019552894085391823,
2637
+ "loss": 0.3315,
2638
+ "step": 215000
2639
+ },
2640
+ {
2641
+ "epoch": 5.74,
2642
+ "learning_rate": 0.00019551854304195062,
2643
+ "loss": 0.3342,
2644
+ "step": 215500
2645
+ },
2646
+ {
2647
+ "epoch": 5.75,
2648
+ "learning_rate": 0.000195508145229983,
2649
+ "loss": 0.3189,
2650
+ "step": 216000
2651
+ },
2652
+ {
2653
+ "epoch": 5.76,
2654
+ "learning_rate": 0.00019549774741801535,
2655
+ "loss": 0.3323,
2656
+ "step": 216500
2657
+ },
2658
+ {
2659
+ "epoch": 5.78,
2660
+ "learning_rate": 0.00019548734960604772,
2661
+ "loss": 0.3163,
2662
+ "step": 217000
2663
+ },
2664
+ {
2665
+ "epoch": 5.79,
2666
+ "learning_rate": 0.00019547695179408008,
2667
+ "loss": 0.3265,
2668
+ "step": 217500
2669
+ },
2670
+ {
2671
+ "epoch": 5.8,
2672
+ "learning_rate": 0.00019546655398211245,
2673
+ "loss": 0.3341,
2674
+ "step": 218000
2675
+ },
2676
+ {
2677
+ "epoch": 5.82,
2678
+ "learning_rate": 0.00019545615617014482,
2679
+ "loss": 0.3283,
2680
+ "step": 218500
2681
+ },
2682
+ {
2683
+ "epoch": 5.83,
2684
+ "learning_rate": 0.00019544575835817718,
2685
+ "loss": 0.3181,
2686
+ "step": 219000
2687
+ },
2688
+ {
2689
+ "epoch": 5.84,
2690
+ "learning_rate": 0.00019543536054620955,
2691
+ "loss": 0.298,
2692
+ "step": 219500
2693
+ },
2694
+ {
2695
+ "epoch": 5.86,
2696
+ "learning_rate": 0.00019542496273424191,
2697
+ "loss": 0.323,
2698
+ "step": 220000
2699
+ },
2700
+ {
2701
+ "epoch": 5.87,
2702
+ "learning_rate": 0.00019541456492227428,
2703
+ "loss": 0.325,
2704
+ "step": 220500
2705
+ },
2706
+ {
2707
+ "epoch": 5.88,
2708
+ "learning_rate": 0.00019540416711030665,
2709
+ "loss": 0.3159,
2710
+ "step": 221000
2711
+ },
2712
+ {
2713
+ "epoch": 5.9,
2714
+ "learning_rate": 0.000195393769298339,
2715
+ "loss": 0.3243,
2716
+ "step": 221500
2717
+ },
2718
+ {
2719
+ "epoch": 5.91,
2720
+ "learning_rate": 0.00019538337148637138,
2721
+ "loss": 0.3195,
2722
+ "step": 222000
2723
+ },
2724
+ {
2725
+ "epoch": 5.92,
2726
+ "learning_rate": 0.00019537297367440374,
2727
+ "loss": 0.3318,
2728
+ "step": 222500
2729
+ },
2730
+ {
2731
+ "epoch": 5.94,
2732
+ "learning_rate": 0.0001953625758624361,
2733
+ "loss": 0.3332,
2734
+ "step": 223000
2735
+ },
2736
+ {
2737
+ "epoch": 5.95,
2738
+ "learning_rate": 0.0001953521780504685,
2739
+ "loss": 0.338,
2740
+ "step": 223500
2741
+ },
2742
+ {
2743
+ "epoch": 5.96,
2744
+ "learning_rate": 0.00019534178023850087,
2745
+ "loss": 0.3242,
2746
+ "step": 224000
2747
+ },
2748
+ {
2749
+ "epoch": 5.98,
2750
+ "learning_rate": 0.00019533138242653323,
2751
+ "loss": 0.3194,
2752
+ "step": 224500
2753
+ },
2754
+ {
2755
+ "epoch": 5.99,
2756
+ "learning_rate": 0.0001953209846145656,
2757
+ "loss": 0.3195,
2758
+ "step": 225000
2759
+ },
2760
+ {
2761
+ "epoch": 6.0,
2762
+ "eval_bleu": 0.0981,
2763
+ "eval_gen_len": 18.3471,
2764
+ "eval_loss": 0.37137025594711304,
2765
+ "eval_runtime": 1997.3462,
2766
+ "eval_samples_per_second": 4.703,
2767
+ "eval_steps_per_second": 2.352,
2768
+ "step": 225408
2769
+ },
2770
+ {
2771
+ "epoch": 6.0,
2772
+ "learning_rate": 0.00019531058680259797,
2773
+ "loss": 0.3295,
2774
+ "step": 225500
2775
+ },
2776
+ {
2777
+ "epoch": 6.02,
2778
+ "learning_rate": 0.00019530018899063033,
2779
+ "loss": 0.27,
2780
+ "step": 226000
2781
+ },
2782
+ {
2783
+ "epoch": 6.03,
2784
+ "learning_rate": 0.0001952897911786627,
2785
+ "loss": 0.2778,
2786
+ "step": 226500
2787
+ },
2788
+ {
2789
+ "epoch": 6.04,
2790
+ "learning_rate": 0.00019527939336669506,
2791
+ "loss": 0.2797,
2792
+ "step": 227000
2793
+ },
2794
+ {
2795
+ "epoch": 6.06,
2796
+ "learning_rate": 0.00019526899555472743,
2797
+ "loss": 0.2667,
2798
+ "step": 227500
2799
+ },
2800
+ {
2801
+ "epoch": 6.07,
2802
+ "learning_rate": 0.0001952585977427598,
2803
+ "loss": 0.2811,
2804
+ "step": 228000
2805
+ },
2806
+ {
2807
+ "epoch": 6.08,
2808
+ "learning_rate": 0.00019524819993079216,
2809
+ "loss": 0.2732,
2810
+ "step": 228500
2811
+ },
2812
+ {
2813
+ "epoch": 6.1,
2814
+ "learning_rate": 0.00019523780211882456,
2815
+ "loss": 0.2763,
2816
+ "step": 229000
2817
+ },
2818
+ {
2819
+ "epoch": 6.11,
2820
+ "learning_rate": 0.00019522740430685692,
2821
+ "loss": 0.2905,
2822
+ "step": 229500
2823
+ },
2824
+ {
2825
+ "epoch": 6.12,
2826
+ "learning_rate": 0.0001952170064948893,
2827
+ "loss": 0.2811,
2828
+ "step": 230000
2829
+ },
2830
+ {
2831
+ "epoch": 6.14,
2832
+ "learning_rate": 0.00019520660868292165,
2833
+ "loss": 0.2916,
2834
+ "step": 230500
2835
+ },
2836
+ {
2837
+ "epoch": 6.15,
2838
+ "learning_rate": 0.00019519621087095402,
2839
+ "loss": 0.2845,
2840
+ "step": 231000
2841
+ },
2842
+ {
2843
+ "epoch": 6.16,
2844
+ "learning_rate": 0.00019518581305898638,
2845
+ "loss": 0.288,
2846
+ "step": 231500
2847
+ },
2848
+ {
2849
+ "epoch": 6.18,
2850
+ "learning_rate": 0.00019517541524701875,
2851
+ "loss": 0.2876,
2852
+ "step": 232000
2853
+ },
2854
+ {
2855
+ "epoch": 6.19,
2856
+ "learning_rate": 0.00019516501743505112,
2857
+ "loss": 0.2796,
2858
+ "step": 232500
2859
+ },
2860
+ {
2861
+ "epoch": 6.2,
2862
+ "learning_rate": 0.00019515461962308348,
2863
+ "loss": 0.2823,
2864
+ "step": 233000
2865
+ },
2866
+ {
2867
+ "epoch": 6.22,
2868
+ "learning_rate": 0.00019514422181111585,
2869
+ "loss": 0.303,
2870
+ "step": 233500
2871
+ },
2872
+ {
2873
+ "epoch": 6.23,
2874
+ "learning_rate": 0.00019513382399914821,
2875
+ "loss": 0.2855,
2876
+ "step": 234000
2877
+ },
2878
+ {
2879
+ "epoch": 6.24,
2880
+ "learning_rate": 0.00019512342618718058,
2881
+ "loss": 0.2981,
2882
+ "step": 234500
2883
+ },
2884
+ {
2885
+ "epoch": 6.26,
2886
+ "learning_rate": 0.00019511302837521295,
2887
+ "loss": 0.286,
2888
+ "step": 235000
2889
+ },
2890
+ {
2891
+ "epoch": 6.27,
2892
+ "learning_rate": 0.0001951026305632453,
2893
+ "loss": 0.2718,
2894
+ "step": 235500
2895
+ },
2896
+ {
2897
+ "epoch": 6.28,
2898
+ "learning_rate": 0.00019509223275127768,
2899
+ "loss": 0.2918,
2900
+ "step": 236000
2901
+ },
2902
+ {
2903
+ "epoch": 6.3,
2904
+ "learning_rate": 0.00019508183493931007,
2905
+ "loss": 0.2779,
2906
+ "step": 236500
2907
+ },
2908
+ {
2909
+ "epoch": 6.31,
2910
+ "learning_rate": 0.00019507143712734244,
2911
+ "loss": 0.2986,
2912
+ "step": 237000
2913
+ },
2914
+ {
2915
+ "epoch": 6.32,
2916
+ "learning_rate": 0.0001950610393153748,
2917
+ "loss": 0.2819,
2918
+ "step": 237500
2919
+ },
2920
+ {
2921
+ "epoch": 6.34,
2922
+ "learning_rate": 0.00019505064150340717,
2923
+ "loss": 0.2828,
2924
+ "step": 238000
2925
+ },
2926
+ {
2927
+ "epoch": 6.35,
2928
+ "learning_rate": 0.00019504024369143953,
2929
+ "loss": 0.2901,
2930
+ "step": 238500
2931
+ },
2932
+ {
2933
+ "epoch": 6.36,
2934
+ "learning_rate": 0.0001950298458794719,
2935
+ "loss": 0.2908,
2936
+ "step": 239000
2937
+ },
2938
+ {
2939
+ "epoch": 6.38,
2940
+ "learning_rate": 0.00019501944806750427,
2941
+ "loss": 0.2803,
2942
+ "step": 239500
2943
+ },
2944
+ {
2945
+ "epoch": 6.39,
2946
+ "learning_rate": 0.00019500905025553663,
2947
+ "loss": 0.3002,
2948
+ "step": 240000
2949
+ },
2950
+ {
2951
+ "epoch": 6.4,
2952
+ "learning_rate": 0.000194998652443569,
2953
+ "loss": 0.2918,
2954
+ "step": 240500
2955
+ },
2956
+ {
2957
+ "epoch": 6.42,
2958
+ "learning_rate": 0.00019498825463160136,
2959
+ "loss": 0.2911,
2960
+ "step": 241000
2961
+ },
2962
+ {
2963
+ "epoch": 6.43,
2964
+ "learning_rate": 0.00019497785681963373,
2965
+ "loss": 0.2897,
2966
+ "step": 241500
2967
+ },
2968
+ {
2969
+ "epoch": 6.44,
2970
+ "learning_rate": 0.0001949674590076661,
2971
+ "loss": 0.2784,
2972
+ "step": 242000
2973
+ },
2974
+ {
2975
+ "epoch": 6.45,
2976
+ "learning_rate": 0.0001949570611956985,
2977
+ "loss": 0.2749,
2978
+ "step": 242500
2979
+ },
2980
+ {
2981
+ "epoch": 6.47,
2982
+ "learning_rate": 0.00019494666338373086,
2983
+ "loss": 0.3055,
2984
+ "step": 243000
2985
+ },
2986
+ {
2987
+ "epoch": 6.48,
2988
+ "learning_rate": 0.00019493626557176322,
2989
+ "loss": 0.2917,
2990
+ "step": 243500
2991
+ },
2992
+ {
2993
+ "epoch": 6.49,
2994
+ "learning_rate": 0.0001949258677597956,
2995
+ "loss": 0.2856,
2996
+ "step": 244000
2997
+ },
2998
+ {
2999
+ "epoch": 6.51,
3000
+ "learning_rate": 0.00019491546994782795,
3001
+ "loss": 0.2957,
3002
+ "step": 244500
3003
+ },
3004
+ {
3005
+ "epoch": 6.52,
3006
+ "learning_rate": 0.00019490507213586032,
3007
+ "loss": 0.2897,
3008
+ "step": 245000
3009
+ },
3010
+ {
3011
+ "epoch": 6.53,
3012
+ "learning_rate": 0.00019489467432389268,
3013
+ "loss": 0.2983,
3014
+ "step": 245500
3015
+ },
3016
+ {
3017
+ "epoch": 6.55,
3018
+ "learning_rate": 0.00019488427651192505,
3019
+ "loss": 0.2911,
3020
+ "step": 246000
3021
+ },
3022
+ {
3023
+ "epoch": 6.56,
3024
+ "learning_rate": 0.00019487387869995742,
3025
+ "loss": 0.2985,
3026
+ "step": 246500
3027
+ },
3028
+ {
3029
+ "epoch": 6.57,
3030
+ "learning_rate": 0.00019486348088798978,
3031
+ "loss": 0.2989,
3032
+ "step": 247000
3033
+ },
3034
+ {
3035
+ "epoch": 6.59,
3036
+ "learning_rate": 0.00019485308307602215,
3037
+ "loss": 0.295,
3038
+ "step": 247500
3039
+ },
3040
+ {
3041
+ "epoch": 6.6,
3042
+ "learning_rate": 0.00019484268526405451,
3043
+ "loss": 0.2949,
3044
+ "step": 248000
3045
+ },
3046
+ {
3047
+ "epoch": 6.61,
3048
+ "learning_rate": 0.00019483228745208688,
3049
+ "loss": 0.2943,
3050
+ "step": 248500
3051
+ },
3052
+ {
3053
+ "epoch": 6.63,
3054
+ "learning_rate": 0.00019482188964011927,
3055
+ "loss": 0.303,
3056
+ "step": 249000
3057
+ },
3058
+ {
3059
+ "epoch": 6.64,
3060
+ "learning_rate": 0.00019481149182815164,
3061
+ "loss": 0.297,
3062
+ "step": 249500
3063
+ },
3064
+ {
3065
+ "epoch": 6.65,
3066
+ "learning_rate": 0.000194801094016184,
3067
+ "loss": 0.3019,
3068
+ "step": 250000
3069
+ },
3070
+ {
3071
+ "epoch": 6.67,
3072
+ "learning_rate": 0.00019479069620421637,
3073
+ "loss": 0.3005,
3074
+ "step": 250500
3075
+ },
3076
+ {
3077
+ "epoch": 6.68,
3078
+ "learning_rate": 0.00019478029839224874,
3079
+ "loss": 0.294,
3080
+ "step": 251000
3081
+ },
3082
+ {
3083
+ "epoch": 6.69,
3084
+ "learning_rate": 0.0001947699005802811,
3085
+ "loss": 0.2977,
3086
+ "step": 251500
3087
+ },
3088
+ {
3089
+ "epoch": 6.71,
3090
+ "learning_rate": 0.00019475950276831347,
3091
+ "loss": 0.294,
3092
+ "step": 252000
3093
+ },
3094
+ {
3095
+ "epoch": 6.72,
3096
+ "learning_rate": 0.00019474910495634584,
3097
+ "loss": 0.2967,
3098
+ "step": 252500
3099
+ },
3100
+ {
3101
+ "epoch": 6.73,
3102
+ "learning_rate": 0.0001947387071443782,
3103
+ "loss": 0.2987,
3104
+ "step": 253000
3105
+ },
3106
+ {
3107
+ "epoch": 6.75,
3108
+ "learning_rate": 0.00019472830933241057,
3109
+ "loss": 0.2929,
3110
+ "step": 253500
3111
+ },
3112
+ {
3113
+ "epoch": 6.76,
3114
+ "learning_rate": 0.00019471791152044293,
3115
+ "loss": 0.2994,
3116
+ "step": 254000
3117
+ },
3118
+ {
3119
+ "epoch": 6.77,
3120
+ "learning_rate": 0.0001947075137084753,
3121
+ "loss": 0.3023,
3122
+ "step": 254500
3123
+ },
3124
+ {
3125
+ "epoch": 6.79,
3126
+ "learning_rate": 0.00019469711589650766,
3127
+ "loss": 0.2961,
3128
+ "step": 255000
3129
+ },
3130
+ {
3131
+ "epoch": 6.8,
3132
+ "learning_rate": 0.00019468671808454003,
3133
+ "loss": 0.2927,
3134
+ "step": 255500
3135
+ },
3136
+ {
3137
+ "epoch": 6.81,
3138
+ "learning_rate": 0.0001946763202725724,
3139
+ "loss": 0.295,
3140
+ "step": 256000
3141
+ },
3142
+ {
3143
+ "epoch": 6.83,
3144
+ "learning_rate": 0.0001946659224606048,
3145
+ "loss": 0.2967,
3146
+ "step": 256500
3147
+ },
3148
+ {
3149
+ "epoch": 6.84,
3150
+ "learning_rate": 0.00019465552464863716,
3151
+ "loss": 0.3147,
3152
+ "step": 257000
3153
+ },
3154
+ {
3155
+ "epoch": 6.85,
3156
+ "learning_rate": 0.00019464512683666952,
3157
+ "loss": 0.3064,
3158
+ "step": 257500
3159
+ },
3160
+ {
3161
+ "epoch": 6.87,
3162
+ "learning_rate": 0.0001946347290247019,
3163
+ "loss": 0.2907,
3164
+ "step": 258000
3165
+ },
3166
+ {
3167
+ "epoch": 6.88,
3168
+ "learning_rate": 0.00019462433121273425,
3169
+ "loss": 0.3076,
3170
+ "step": 258500
3171
+ },
3172
+ {
3173
+ "epoch": 6.89,
3174
+ "learning_rate": 0.00019461393340076662,
3175
+ "loss": 0.288,
3176
+ "step": 259000
3177
+ },
3178
+ {
3179
+ "epoch": 6.91,
3180
+ "learning_rate": 0.00019460353558879899,
3181
+ "loss": 0.3036,
3182
+ "step": 259500
3183
+ },
3184
+ {
3185
+ "epoch": 6.92,
3186
+ "learning_rate": 0.00019459313777683135,
3187
+ "loss": 0.3001,
3188
+ "step": 260000
3189
+ },
3190
+ {
3191
+ "epoch": 6.93,
3192
+ "learning_rate": 0.00019458273996486372,
3193
+ "loss": 0.2986,
3194
+ "step": 260500
3195
+ },
3196
+ {
3197
+ "epoch": 6.95,
3198
+ "learning_rate": 0.00019457234215289608,
3199
+ "loss": 0.2964,
3200
+ "step": 261000
3201
+ },
3202
+ {
3203
+ "epoch": 6.96,
3204
+ "learning_rate": 0.00019456194434092845,
3205
+ "loss": 0.2935,
3206
+ "step": 261500
3207
+ },
3208
+ {
3209
+ "epoch": 6.97,
3210
+ "learning_rate": 0.00019455154652896084,
3211
+ "loss": 0.3099,
3212
+ "step": 262000
3213
+ },
3214
+ {
3215
+ "epoch": 6.99,
3216
+ "learning_rate": 0.0001945411487169932,
3217
+ "loss": 0.2951,
3218
+ "step": 262500
3219
+ },
3220
+ {
3221
+ "epoch": 7.0,
3222
+ "eval_bleu": 0.1004,
3223
+ "eval_gen_len": 18.389,
3224
+ "eval_loss": 0.37594369053840637,
3225
+ "eval_runtime": 1995.597,
3226
+ "eval_samples_per_second": 4.707,
3227
+ "eval_steps_per_second": 2.354,
3228
+ "step": 262976
3229
+ }
3230
+ ],
3231
+ "max_steps": 9617408,
3232
+ "num_train_epochs": 256,
3233
+ "total_flos": 4.369756739154739e+16,
3234
+ "trial_name": null,
3235
+ "trial_params": null
3236
+ }
checkpoint-262976/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:96e4b36dc34e56d0b31306cd5cea9659aa12a0b2a579e060030638791132a5da
3
+ size 3119