Hugging Face's logo

Propicto
/
t2p-t5-large-orfeo

Translation
Transformers
PyTorch
French
t5
text2text-generation
orfeo
pictograms
text-generation-inference
Inference Endpoints
Model card Files Community
cecilemacaire commited on
Commit
f815f2e
•
1 Parent(s): b3ddb09

Upload 12 files

Browse files
Files changed (12) hide show
  1. config.json +60 -0
  2. optimizer.pt +3 -0
  3. pytorch_model.bin +3 -0
  4. rng_state.pth +3 -0
  5. scaler.pt +3 -0
  6. scheduler.pt +3 -0
  7. special_tokens_map.json +107 -0
  8. spiece.model +3 -0
  9. tokenizer.json +0 -0
  10. tokenizer_config.json +111 -0
  11. trainer_state.json +3884 -0
  12. training_args.bin +3 -0
config.json ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/gpfsstore/rech/czj/uef37or/pretrained_models/t5-base/",
3
+ "architectures": [
4
+ "T5ForConditionalGeneration"
5
+ ],
6
+ "d_ff": 3072,
7
+ "d_kv": 64,
8
+ "d_model": 768,
9
+ "decoder_start_token_id": 0,
10
+ "dense_act_fn": "relu",
11
+ "dropout_rate": 0.1,
12
+ "eos_token_id": 1,
13
+ "feed_forward_proj": "relu",
14
+ "initializer_factor": 1.0,
15
+ "is_encoder_decoder": true,
16
+ "is_gated_act": false,
17
+ "layer_norm_epsilon": 1e-06,
18
+ "model_type": "t5",
19
+ "n_positions": 512,
20
+ "num_decoder_layers": 12,
21
+ "num_heads": 12,
22
+ "num_layers": 12,
23
+ "output_past": true,
24
+ "pad_token_id": 0,
25
+ "relative_attention_max_distance": 128,
26
+ "relative_attention_num_buckets": 32,
27
+ "task_specific_params": {
28
+ "summarization": {
29
+ "early_stopping": true,
30
+ "length_penalty": 2.0,
31
+ "max_length": 200,
32
+ "min_length": 30,
33
+ "no_repeat_ngram_size": 3,
34
+ "num_beams": 4,
35
+ "prefix": "summarize: "
36
+ },
37
+ "translation_en_to_de": {
38
+ "early_stopping": true,
39
+ "max_length": 300,
40
+ "num_beams": 4,
41
+ "prefix": "translate English to German: "
42
+ },
43
+ "translation_en_to_fr": {
44
+ "early_stopping": true,
45
+ "max_length": 300,
46
+ "num_beams": 4,
47
+ "prefix": "translate English to French: "
48
+ },
49
+ "translation_en_to_ro": {
50
+ "early_stopping": true,
51
+ "max_length": 300,
52
+ "num_beams": 4,
53
+ "prefix": "translate English to Romanian: "
54
+ }
55
+ },
56
+ "torch_dtype": "float32",
57
+ "transformers_version": "4.21.3",
58
+ "use_cache": true,
59
+ "vocab_size": 32128
60
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f9752e5c98e7e5bab86c37312be48ce0f16717025703a07b5315cc56738d611
3
+ size 1783379645
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af66f74ced5778c978617d3454409d943f595ed11eb9a7fa51ebebc8ab545d55
3
+ size 891700799
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1433c95a8b8ef0f1c5dc82caf5dc6109902d8faba5cd31052451013e9378451d
3
+ size 14503
scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:104e0b3aa035245f9d59d4fab2a9e7ecd5db723787e6e831ff1bac7ee7825264
3
+ size 559
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:21015cf3abd54d71cd0beb32ef630982e733d9485d4a47119ca8c4c79d4e1c30
3
+ size 623
special_tokens_map.json ADDED
@@ -0,0 +1,107 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<extra_id_0>",
4
+ "<extra_id_1>",
5
+ "<extra_id_2>",
6
+ "<extra_id_3>",
7
+ "<extra_id_4>",
8
+ "<extra_id_5>",
9
+ "<extra_id_6>",
10
+ "<extra_id_7>",
11
+ "<extra_id_8>",
12
+ "<extra_id_9>",
13
+ "<extra_id_10>",
14
+ "<extra_id_11>",
15
+ "<extra_id_12>",
16
+ "<extra_id_13>",
17
+ "<extra_id_14>",
18
+ "<extra_id_15>",
19
+ "<extra_id_16>",
20
+ "<extra_id_17>",
21
+ "<extra_id_18>",
22
+ "<extra_id_19>",
23
+ "<extra_id_20>",
24
+ "<extra_id_21>",
25
+ "<extra_id_22>",
26
+ "<extra_id_23>",
27
+ "<extra_id_24>",
28
+ "<extra_id_25>",
29
+ "<extra_id_26>",
30
+ "<extra_id_27>",
31
+ "<extra_id_28>",
32
+ "<extra_id_29>",
33
+ "<extra_id_30>",
34
+ "<extra_id_31>",
35
+ "<extra_id_32>",
36
+ "<extra_id_33>",
37
+ "<extra_id_34>",
38
+ "<extra_id_35>",
39
+ "<extra_id_36>",
40
+ "<extra_id_37>",
41
+ "<extra_id_38>",
42
+ "<extra_id_39>",
43
+ "<extra_id_40>",
44
+ "<extra_id_41>",
45
+ "<extra_id_42>",
46
+ "<extra_id_43>",
47
+ "<extra_id_44>",
48
+ "<extra_id_45>",
49
+ "<extra_id_46>",
50
+ "<extra_id_47>",
51
+ "<extra_id_48>",
52
+ "<extra_id_49>",
53
+ "<extra_id_50>",
54
+ "<extra_id_51>",
55
+ "<extra_id_52>",
56
+ "<extra_id_53>",
57
+ "<extra_id_54>",
58
+ "<extra_id_55>",
59
+ "<extra_id_56>",
60
+ "<extra_id_57>",
61
+ "<extra_id_58>",
62
+ "<extra_id_59>",
63
+ "<extra_id_60>",
64
+ "<extra_id_61>",
65
+ "<extra_id_62>",
66
+ "<extra_id_63>",
67
+ "<extra_id_64>",
68
+ "<extra_id_65>",
69
+ "<extra_id_66>",
70
+ "<extra_id_67>",
71
+ "<extra_id_68>",
72
+ "<extra_id_69>",
73
+ "<extra_id_70>",
74
+ "<extra_id_71>",
75
+ "<extra_id_72>",
76
+ "<extra_id_73>",
77
+ "<extra_id_74>",
78
+ "<extra_id_75>",
79
+ "<extra_id_76>",
80
+ "<extra_id_77>",
81
+ "<extra_id_78>",
82
+ "<extra_id_79>",
83
+ "<extra_id_80>",
84
+ "<extra_id_81>",
85
+ "<extra_id_82>",
86
+ "<extra_id_83>",
87
+ "<extra_id_84>",
88
+ "<extra_id_85>",
89
+ "<extra_id_86>",
90
+ "<extra_id_87>",
91
+ "<extra_id_88>",
92
+ "<extra_id_89>",
93
+ "<extra_id_90>",
94
+ "<extra_id_91>",
95
+ "<extra_id_92>",
96
+ "<extra_id_93>",
97
+ "<extra_id_94>",
98
+ "<extra_id_95>",
99
+ "<extra_id_96>",
100
+ "<extra_id_97>",
101
+ "<extra_id_98>",
102
+ "<extra_id_99>"
103
+ ],
104
+ "eos_token": "</s>",
105
+ "pad_token": "<pad>",
106
+ "unk_token": "<unk>"
107
+ }
spiece.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d60acb128cf7b7f2536e8f38a5b18a05535c9e14c7a355904270e15b0945ea86
3
+ size 791656
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,111 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<extra_id_0>",
4
+ "<extra_id_1>",
5
+ "<extra_id_2>",
6
+ "<extra_id_3>",
7
+ "<extra_id_4>",
8
+ "<extra_id_5>",
9
+ "<extra_id_6>",
10
+ "<extra_id_7>",
11
+ "<extra_id_8>",
12
+ "<extra_id_9>",
13
+ "<extra_id_10>",
14
+ "<extra_id_11>",
15
+ "<extra_id_12>",
16
+ "<extra_id_13>",
17
+ "<extra_id_14>",
18
+ "<extra_id_15>",
19
+ "<extra_id_16>",
20
+ "<extra_id_17>",
21
+ "<extra_id_18>",
22
+ "<extra_id_19>",
23
+ "<extra_id_20>",
24
+ "<extra_id_21>",
25
+ "<extra_id_22>",
26
+ "<extra_id_23>",
27
+ "<extra_id_24>",
28
+ "<extra_id_25>",
29
+ "<extra_id_26>",
30
+ "<extra_id_27>",
31
+ "<extra_id_28>",
32
+ "<extra_id_29>",
33
+ "<extra_id_30>",
34
+ "<extra_id_31>",
35
+ "<extra_id_32>",
36
+ "<extra_id_33>",
37
+ "<extra_id_34>",
38
+ "<extra_id_35>",
39
+ "<extra_id_36>",
40
+ "<extra_id_37>",
41
+ "<extra_id_38>",
42
+ "<extra_id_39>",
43
+ "<extra_id_40>",
44
+ "<extra_id_41>",
45
+ "<extra_id_42>",
46
+ "<extra_id_43>",
47
+ "<extra_id_44>",
48
+ "<extra_id_45>",
49
+ "<extra_id_46>",
50
+ "<extra_id_47>",
51
+ "<extra_id_48>",
52
+ "<extra_id_49>",
53
+ "<extra_id_50>",
54
+ "<extra_id_51>",
55
+ "<extra_id_52>",
56
+ "<extra_id_53>",
57
+ "<extra_id_54>",
58
+ "<extra_id_55>",
59
+ "<extra_id_56>",
60
+ "<extra_id_57>",
61
+ "<extra_id_58>",
62
+ "<extra_id_59>",
63
+ "<extra_id_60>",
64
+ "<extra_id_61>",
65
+ "<extra_id_62>",
66
+ "<extra_id_63>",
67
+ "<extra_id_64>",
68
+ "<extra_id_65>",
69
+ "<extra_id_66>",
70
+ "<extra_id_67>",
71
+ "<extra_id_68>",
72
+ "<extra_id_69>",
73
+ "<extra_id_70>",
74
+ "<extra_id_71>",
75
+ "<extra_id_72>",
76
+ "<extra_id_73>",
77
+ "<extra_id_74>",
78
+ "<extra_id_75>",
79
+ "<extra_id_76>",
80
+ "<extra_id_77>",
81
+ "<extra_id_78>",
82
+ "<extra_id_79>",
83
+ "<extra_id_80>",
84
+ "<extra_id_81>",
85
+ "<extra_id_82>",
86
+ "<extra_id_83>",
87
+ "<extra_id_84>",
88
+ "<extra_id_85>",
89
+ "<extra_id_86>",
90
+ "<extra_id_87>",
91
+ "<extra_id_88>",
92
+ "<extra_id_89>",
93
+ "<extra_id_90>",
94
+ "<extra_id_91>",
95
+ "<extra_id_92>",
96
+ "<extra_id_93>",
97
+ "<extra_id_94>",
98
+ "<extra_id_95>",
99
+ "<extra_id_96>",
100
+ "<extra_id_97>",
101
+ "<extra_id_98>",
102
+ "<extra_id_99>"
103
+ ],
104
+ "eos_token": "</s>",
105
+ "extra_ids": 100,
106
+ "name_or_path": "/gpfsstore/rech/czj/uef37or/pretrained_models/t5-base/",
107
+ "pad_token": "<pad>",
108
+ "special_tokens_map_file": null,
109
+ "tokenizer_class": "T5Tokenizer",
110
+ "unk_token": "<unk>"
111
+ }
trainer_state.json ADDED
@@ -0,0 +1,3884 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.16182634234428406,
3
+ "best_model_checkpoint": "checkpoints_orfeo/checkpoint-260316",
4
+ "epoch": 40.0,
5
+ "global_step": 289240,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.07,
12
+ "learning_rate": 1.996570322223759e-05,
13
+ "loss": 2.1923,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.14,
18
+ "learning_rate": 1.9931129857557737e-05,
19
+ "loss": 1.3226,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.21,
24
+ "learning_rate": 1.989655649287789e-05,
25
+ "loss": 1.1003,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.28,
30
+ "learning_rate": 1.9861983128198038e-05,
31
+ "loss": 0.9671,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.35,
36
+ "learning_rate": 1.9827409763518186e-05,
37
+ "loss": 0.8916,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.41,
42
+ "learning_rate": 1.9792836398838338e-05,
43
+ "loss": 0.8332,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.48,
48
+ "learning_rate": 1.9758263034158486e-05,
49
+ "loss": 0.7772,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.55,
54
+ "learning_rate": 1.9723689669478635e-05,
55
+ "loss": 0.734,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.62,
60
+ "learning_rate": 1.9689116304798787e-05,
61
+ "loss": 0.7019,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.69,
66
+ "learning_rate": 1.9654542940118935e-05,
67
+ "loss": 0.6743,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.76,
72
+ "learning_rate": 1.9619969575439084e-05,
73
+ "loss": 0.6475,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.83,
78
+ "learning_rate": 1.9585396210759232e-05,
79
+ "loss": 0.6189,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.9,
84
+ "learning_rate": 1.9550822846079384e-05,
85
+ "loss": 0.6036,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.97,
90
+ "learning_rate": 1.9516249481399533e-05,
91
+ "loss": 0.587,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 1.0,
96
+ "eval_bleu": 60.5819,
97
+ "eval_gen_len": 12.8913,
98
+ "eval_loss": 0.4583694636821747,
99
+ "eval_runtime": 327.4166,
100
+ "eval_samples_per_second": 87.949,
101
+ "eval_steps_per_second": 2.749,
102
+ "step": 7231
103
+ },
104
+ {
105
+ "epoch": 1.04,
106
+ "learning_rate": 1.948167611671968e-05,
107
+ "loss": 0.5662,
108
+ "step": 7500
109
+ },
110
+ {
111
+ "epoch": 1.11,
112
+ "learning_rate": 1.944710275203983e-05,
113
+ "loss": 0.5361,
114
+ "step": 8000
115
+ },
116
+ {
117
+ "epoch": 1.18,
118
+ "learning_rate": 1.9412529387359978e-05,
119
+ "loss": 0.5269,
120
+ "step": 8500
121
+ },
122
+ {
123
+ "epoch": 1.24,
124
+ "learning_rate": 1.937795602268013e-05,
125
+ "loss": 0.5157,
126
+ "step": 9000
127
+ },
128
+ {
129
+ "epoch": 1.31,
130
+ "learning_rate": 1.934338265800028e-05,
131
+ "loss": 0.5041,
132
+ "step": 9500
133
+ },
134
+ {
135
+ "epoch": 1.38,
136
+ "learning_rate": 1.9308809293320427e-05,
137
+ "loss": 0.4851,
138
+ "step": 10000
139
+ },
140
+ {
141
+ "epoch": 1.45,
142
+ "learning_rate": 1.9274305075369937e-05,
143
+ "loss": 0.4798,
144
+ "step": 10500
145
+ },
146
+ {
147
+ "epoch": 1.52,
148
+ "learning_rate": 1.9239731710690085e-05,
149
+ "loss": 0.4697,
150
+ "step": 11000
151
+ },
152
+ {
153
+ "epoch": 1.59,
154
+ "learning_rate": 1.9205158346010234e-05,
155
+ "loss": 0.4657,
156
+ "step": 11500
157
+ },
158
+ {
159
+ "epoch": 1.66,
160
+ "learning_rate": 1.9170584981330386e-05,
161
+ "loss": 0.4555,
162
+ "step": 12000
163
+ },
164
+ {
165
+ "epoch": 1.73,
166
+ "learning_rate": 1.9136011616650534e-05,
167
+ "loss": 0.4485,
168
+ "step": 12500
169
+ },
170
+ {
171
+ "epoch": 1.8,
172
+ "learning_rate": 1.9101507398700044e-05,
173
+ "loss": 0.4446,
174
+ "step": 13000
175
+ },
176
+ {
177
+ "epoch": 1.87,
178
+ "learning_rate": 1.9067003180749554e-05,
179
+ "loss": 0.4311,
180
+ "step": 13500
181
+ },
182
+ {
183
+ "epoch": 1.94,
184
+ "learning_rate": 1.9032429816069702e-05,
185
+ "loss": 0.4213,
186
+ "step": 14000
187
+ },
188
+ {
189
+ "epoch": 2.0,
190
+ "eval_bleu": 66.0413,
191
+ "eval_gen_len": 12.9749,
192
+ "eval_loss": 0.33632054924964905,
193
+ "eval_runtime": 328.8185,
194
+ "eval_samples_per_second": 87.574,
195
+ "eval_steps_per_second": 2.737,
196
+ "step": 14462
197
+ },
198
+ {
199
+ "epoch": 2.01,
200
+ "learning_rate": 1.899785645138985e-05,
201
+ "loss": 0.4164,
202
+ "step": 14500
203
+ },
204
+ {
205
+ "epoch": 2.07,
206
+ "learning_rate": 1.896328308671e-05,
207
+ "loss": 0.4019,
208
+ "step": 15000
209
+ },
210
+ {
211
+ "epoch": 2.14,
212
+ "learning_rate": 1.892870972203015e-05,
213
+ "loss": 0.3962,
214
+ "step": 15500
215
+ },
216
+ {
217
+ "epoch": 2.21,
218
+ "learning_rate": 1.88941363573503e-05,
219
+ "loss": 0.392,
220
+ "step": 16000
221
+ },
222
+ {
223
+ "epoch": 2.28,
224
+ "learning_rate": 1.8859562992670448e-05,
225
+ "loss": 0.3899,
226
+ "step": 16500
227
+ },
228
+ {
229
+ "epoch": 2.35,
230
+ "learning_rate": 1.8824989627990596e-05,
231
+ "loss": 0.3876,
232
+ "step": 17000
233
+ },
234
+ {
235
+ "epoch": 2.42,
236
+ "learning_rate": 1.879041626331075e-05,
237
+ "loss": 0.3774,
238
+ "step": 17500
239
+ },
240
+ {
241
+ "epoch": 2.49,
242
+ "learning_rate": 1.8755842898630897e-05,
243
+ "loss": 0.3744,
244
+ "step": 18000
245
+ },
246
+ {
247
+ "epoch": 2.56,
248
+ "learning_rate": 1.8721269533951045e-05,
249
+ "loss": 0.3724,
250
+ "step": 18500
251
+ },
252
+ {
253
+ "epoch": 2.63,
254
+ "learning_rate": 1.8686696169271194e-05,
255
+ "loss": 0.3614,
256
+ "step": 19000
257
+ },
258
+ {
259
+ "epoch": 2.7,
260
+ "learning_rate": 1.8652191951320704e-05,
261
+ "loss": 0.3665,
262
+ "step": 19500
263
+ },
264
+ {
265
+ "epoch": 2.77,
266
+ "learning_rate": 1.8617618586640852e-05,
267
+ "loss": 0.3581,
268
+ "step": 20000
269
+ },
270
+ {
271
+ "epoch": 2.84,
272
+ "learning_rate": 1.8583045221961004e-05,
273
+ "loss": 0.3495,
274
+ "step": 20500
275
+ },
276
+ {
277
+ "epoch": 2.9,
278
+ "learning_rate": 1.8548471857281153e-05,
279
+ "loss": 0.3487,
280
+ "step": 21000
281
+ },
282
+ {
283
+ "epoch": 2.97,
284
+ "learning_rate": 1.8513967639330662e-05,
285
+ "loss": 0.354,
286
+ "step": 21500
287
+ },
288
+ {
289
+ "epoch": 3.0,
290
+ "eval_bleu": 68.1423,
291
+ "eval_gen_len": 12.9805,
292
+ "eval_loss": 0.2864578068256378,
293
+ "eval_runtime": 328.1672,
294
+ "eval_samples_per_second": 87.748,
295
+ "eval_steps_per_second": 2.743,
296
+ "step": 21693
297
+ },
298
+ {
299
+ "epoch": 3.04,
300
+ "learning_rate": 1.847939427465081e-05,
301
+ "loss": 0.3412,
302
+ "step": 22000
303
+ },
304
+ {
305
+ "epoch": 3.11,
306
+ "learning_rate": 1.844482090997096e-05,
307
+ "loss": 0.3407,
308
+ "step": 22500
309
+ },
310
+ {
311
+ "epoch": 3.18,
312
+ "learning_rate": 1.8410247545291108e-05,
313
+ "loss": 0.3336,
314
+ "step": 23000
315
+ },
316
+ {
317
+ "epoch": 3.25,
318
+ "learning_rate": 1.8375743327340618e-05,
319
+ "loss": 0.33,
320
+ "step": 23500
321
+ },
322
+ {
323
+ "epoch": 3.32,
324
+ "learning_rate": 1.8341169962660766e-05,
325
+ "loss": 0.3222,
326
+ "step": 24000
327
+ },
328
+ {
329
+ "epoch": 3.39,
330
+ "learning_rate": 1.8306596597980918e-05,
331
+ "loss": 0.3209,
332
+ "step": 24500
333
+ },
334
+ {
335
+ "epoch": 3.46,
336
+ "learning_rate": 1.8272023233301066e-05,
337
+ "loss": 0.3221,
338
+ "step": 25000
339
+ },
340
+ {
341
+ "epoch": 3.53,
342
+ "learning_rate": 1.8237449868621215e-05,
343
+ "loss": 0.3194,
344
+ "step": 25500
345
+ },
346
+ {
347
+ "epoch": 3.6,
348
+ "learning_rate": 1.8202876503941363e-05,
349
+ "loss": 0.309,
350
+ "step": 26000
351
+ },
352
+ {
353
+ "epoch": 3.66,
354
+ "learning_rate": 1.8168303139261515e-05,
355
+ "loss": 0.3149,
356
+ "step": 26500
357
+ },
358
+ {
359
+ "epoch": 3.73,
360
+ "learning_rate": 1.8133729774581664e-05,
361
+ "loss": 0.3069,
362
+ "step": 27000
363
+ },
364
+ {
365
+ "epoch": 3.8,
366
+ "learning_rate": 1.8099225556631174e-05,
367
+ "loss": 0.3158,
368
+ "step": 27500
369
+ },
370
+ {
371
+ "epoch": 3.87,
372
+ "learning_rate": 1.8064652191951322e-05,
373
+ "loss": 0.3039,
374
+ "step": 28000
375
+ },
376
+ {
377
+ "epoch": 3.94,
378
+ "learning_rate": 1.803007882727147e-05,
379
+ "loss": 0.3092,
380
+ "step": 28500
381
+ },
382
+ {
383
+ "epoch": 4.0,
384
+ "eval_bleu": 69.4372,
385
+ "eval_gen_len": 12.9939,
386
+ "eval_loss": 0.2564955949783325,
387
+ "eval_runtime": 327.0452,
388
+ "eval_samples_per_second": 88.049,
389
+ "eval_steps_per_second": 2.752,
390
+ "step": 28924
391
+ },
392
+ {
393
+ "epoch": 4.01,
394
+ "learning_rate": 1.799550546259162e-05,
395
+ "loss": 0.3004,
396
+ "step": 29000
397
+ },
398
+ {
399
+ "epoch": 4.08,
400
+ "learning_rate": 1.796100124464113e-05,
401
+ "loss": 0.2942,
402
+ "step": 29500
403
+ },
404
+ {
405
+ "epoch": 4.15,
406
+ "learning_rate": 1.7926427879961277e-05,
407
+ "loss": 0.2906,
408
+ "step": 30000
409
+ },
410
+ {
411
+ "epoch": 4.22,
412
+ "learning_rate": 1.789185451528143e-05,
413
+ "loss": 0.2907,
414
+ "step": 30500
415
+ },
416
+ {
417
+ "epoch": 4.29,
418
+ "learning_rate": 1.7857281150601578e-05,
419
+ "loss": 0.29,
420
+ "step": 31000
421
+ },
422
+ {
423
+ "epoch": 4.36,
424
+ "learning_rate": 1.7822776932651088e-05,
425
+ "loss": 0.2871,
426
+ "step": 31500
427
+ },
428
+ {
429
+ "epoch": 4.43,
430
+ "learning_rate": 1.7788272714700597e-05,
431
+ "loss": 0.2906,
432
+ "step": 32000
433
+ },
434
+ {
435
+ "epoch": 4.49,
436
+ "learning_rate": 1.7753699350020746e-05,
437
+ "loss": 0.287,
438
+ "step": 32500
439
+ },
440
+ {
441
+ "epoch": 4.56,
442
+ "learning_rate": 1.7719125985340894e-05,
443
+ "loss": 0.2797,
444
+ "step": 33000
445
+ },
446
+ {
447
+ "epoch": 4.63,
448
+ "learning_rate": 1.7684552620661043e-05,
449
+ "loss": 0.2864,
450
+ "step": 33500
451
+ },
452
+ {
453
+ "epoch": 4.7,
454
+ "learning_rate": 1.7649979255981195e-05,
455
+ "loss": 0.2841,
456
+ "step": 34000
457
+ },
458
+ {
459
+ "epoch": 4.77,
460
+ "learning_rate": 1.7615405891301343e-05,
461
+ "loss": 0.2781,
462
+ "step": 34500
463
+ },
464
+ {
465
+ "epoch": 4.84,
466
+ "learning_rate": 1.7580901673350853e-05,
467
+ "loss": 0.2803,
468
+ "step": 35000
469
+ },
470
+ {
471
+ "epoch": 4.91,
472
+ "learning_rate": 1.7546328308671e-05,
473
+ "loss": 0.2729,
474
+ "step": 35500
475
+ },
476
+ {
477
+ "epoch": 4.98,
478
+ "learning_rate": 1.751175494399115e-05,
479
+ "loss": 0.2717,
480
+ "step": 36000
481
+ },
482
+ {
483
+ "epoch": 5.0,
484
+ "eval_bleu": 70.3702,
485
+ "eval_gen_len": 12.957,
486
+ "eval_loss": 0.23622238636016846,
487
+ "eval_runtime": 329.1892,
488
+ "eval_samples_per_second": 87.476,
489
+ "eval_steps_per_second": 2.734,
490
+ "step": 36155
491
+ },
492
+ {
493
+ "epoch": 5.05,
494
+ "learning_rate": 1.74771815793113e-05,
495
+ "loss": 0.2668,
496
+ "step": 36500
497
+ },
498
+ {
499
+ "epoch": 5.12,
500
+ "learning_rate": 1.744260821463145e-05,
501
+ "loss": 0.2651,
502
+ "step": 37000
503
+ },
504
+ {
505
+ "epoch": 5.19,
506
+ "learning_rate": 1.74080348499516e-05,
507
+ "loss": 0.2647,
508
+ "step": 37500
509
+ },
510
+ {
511
+ "epoch": 5.26,
512
+ "learning_rate": 1.7373461485271747e-05,
513
+ "loss": 0.2618,
514
+ "step": 38000
515
+ },
516
+ {
517
+ "epoch": 5.32,
518
+ "learning_rate": 1.7338888120591896e-05,
519
+ "loss": 0.265,
520
+ "step": 38500
521
+ },
522
+ {
523
+ "epoch": 5.39,
524
+ "learning_rate": 1.7304383902641406e-05,
525
+ "loss": 0.2614,
526
+ "step": 39000
527
+ },
528
+ {
529
+ "epoch": 5.46,
530
+ "learning_rate": 1.7269810537961554e-05,
531
+ "loss": 0.2531,
532
+ "step": 39500
533
+ },
534
+ {
535
+ "epoch": 5.53,
536
+ "learning_rate": 1.7235237173281706e-05,
537
+ "loss": 0.2583,
538
+ "step": 40000
539
+ },
540
+ {
541
+ "epoch": 5.6,
542
+ "learning_rate": 1.7200663808601855e-05,
543
+ "loss": 0.2631,
544
+ "step": 40500
545
+ },
546
+ {
547
+ "epoch": 5.67,
548
+ "learning_rate": 1.7166159590651364e-05,
549
+ "loss": 0.2573,
550
+ "step": 41000
551
+ },
552
+ {
553
+ "epoch": 5.74,
554
+ "learning_rate": 1.7131655372700874e-05,
555
+ "loss": 0.2562,
556
+ "step": 41500
557
+ },
558
+ {
559
+ "epoch": 5.81,
560
+ "learning_rate": 1.7097082008021023e-05,
561
+ "loss": 0.2546,
562
+ "step": 42000
563
+ },
564
+ {
565
+ "epoch": 5.88,
566
+ "learning_rate": 1.706250864334117e-05,
567
+ "loss": 0.2527,
568
+ "step": 42500
569
+ },
570
+ {
571
+ "epoch": 5.95,
572
+ "learning_rate": 1.702793527866132e-05,
573
+ "loss": 0.2542,
574
+ "step": 43000
575
+ },
576
+ {
577
+ "epoch": 6.0,
578
+ "eval_bleu": 70.9448,
579
+ "eval_gen_len": 12.945,
580
+ "eval_loss": 0.22083553671836853,
581
+ "eval_runtime": 327.8279,
582
+ "eval_samples_per_second": 87.839,
583
+ "eval_steps_per_second": 2.745,
584
+ "step": 43386
585
+ },
586
+ {
587
+ "epoch": 6.02,
588
+ "learning_rate": 1.699336191398147e-05,
589
+ "loss": 0.2567,
590
+ "step": 43500
591
+ },
592
+ {
593
+ "epoch": 6.08,
594
+ "learning_rate": 1.695878854930162e-05,
595
+ "loss": 0.2443,
596
+ "step": 44000
597
+ },
598
+ {
599
+ "epoch": 6.15,
600
+ "learning_rate": 1.692421518462177e-05,
601
+ "loss": 0.2423,
602
+ "step": 44500
603
+ },
604
+ {
605
+ "epoch": 6.22,
606
+ "learning_rate": 1.6889641819941917e-05,
607
+ "loss": 0.2476,
608
+ "step": 45000
609
+ },
610
+ {
611
+ "epoch": 6.29,
612
+ "learning_rate": 1.6855068455262065e-05,
613
+ "loss": 0.2449,
614
+ "step": 45500
615
+ },
616
+ {
617
+ "epoch": 6.36,
618
+ "learning_rate": 1.6820495090582217e-05,
619
+ "loss": 0.2435,
620
+ "step": 46000
621
+ },
622
+ {
623
+ "epoch": 6.43,
624
+ "learning_rate": 1.6785921725902366e-05,
625
+ "loss": 0.2403,
626
+ "step": 46500
627
+ },
628
+ {
629
+ "epoch": 6.5,
630
+ "learning_rate": 1.6751348361222514e-05,
631
+ "loss": 0.2384,
632
+ "step": 47000
633
+ },
634
+ {
635
+ "epoch": 6.57,
636
+ "learning_rate": 1.6716844143272024e-05,
637
+ "loss": 0.2365,
638
+ "step": 47500
639
+ },
640
+ {
641
+ "epoch": 6.64,
642
+ "learning_rate": 1.6682270778592173e-05,
643
+ "loss": 0.2385,
644
+ "step": 48000
645
+ },
646
+ {
647
+ "epoch": 6.71,
648
+ "learning_rate": 1.664769741391232e-05,
649
+ "loss": 0.2373,
650
+ "step": 48500
651
+ },
652
+ {
653
+ "epoch": 6.78,
654
+ "learning_rate": 1.6613124049232473e-05,
655
+ "loss": 0.2432,
656
+ "step": 49000
657
+ },
658
+ {
659
+ "epoch": 6.85,
660
+ "learning_rate": 1.6578619831281983e-05,
661
+ "loss": 0.2375,
662
+ "step": 49500
663
+ },
664
+ {
665
+ "epoch": 6.91,
666
+ "learning_rate": 1.654404646660213e-05,
667
+ "loss": 0.234,
668
+ "step": 50000
669
+ },
670
+ {
671
+ "epoch": 6.98,
672
+ "learning_rate": 1.650947310192228e-05,
673
+ "loss": 0.2344,
674
+ "step": 50500
675
+ },
676
+ {
677
+ "epoch": 7.0,
678
+ "eval_bleu": 71.317,
679
+ "eval_gen_len": 12.9746,
680
+ "eval_loss": 0.21209098398685455,
681
+ "eval_runtime": 332.4569,
682
+ "eval_samples_per_second": 86.616,
683
+ "eval_steps_per_second": 2.707,
684
+ "step": 50617
685
+ },
686
+ {
687
+ "epoch": 7.05,
688
+ "learning_rate": 1.6474899737242428e-05,
689
+ "loss": 0.2278,
690
+ "step": 51000
691
+ },
692
+ {
693
+ "epoch": 7.12,
694
+ "learning_rate": 1.6440395519291938e-05,
695
+ "loss": 0.2249,
696
+ "step": 51500
697
+ },
698
+ {
699
+ "epoch": 7.19,
700
+ "learning_rate": 1.6405891301341448e-05,
701
+ "loss": 0.2326,
702
+ "step": 52000
703
+ },
704
+ {
705
+ "epoch": 7.26,
706
+ "learning_rate": 1.6371317936661596e-05,
707
+ "loss": 0.2287,
708
+ "step": 52500
709
+ },
710
+ {
711
+ "epoch": 7.33,
712
+ "learning_rate": 1.6336744571981745e-05,
713
+ "loss": 0.2276,
714
+ "step": 53000
715
+ },
716
+ {
717
+ "epoch": 7.4,
718
+ "learning_rate": 1.6302171207301897e-05,
719
+ "loss": 0.2279,
720
+ "step": 53500
721
+ },
722
+ {
723
+ "epoch": 7.47,
724
+ "learning_rate": 1.6267597842622045e-05,
725
+ "loss": 0.2242,
726
+ "step": 54000
727
+ },
728
+ {
729
+ "epoch": 7.54,
730
+ "learning_rate": 1.6233024477942194e-05,
731
+ "loss": 0.2269,
732
+ "step": 54500
733
+ },
734
+ {
735
+ "epoch": 7.61,
736
+ "learning_rate": 1.6198451113262342e-05,
737
+ "loss": 0.2243,
738
+ "step": 55000
739
+ },
740
+ {
741
+ "epoch": 7.68,
742
+ "learning_rate": 1.6163946895311852e-05,
743
+ "loss": 0.2221,
744
+ "step": 55500
745
+ },
746
+ {
747
+ "epoch": 7.74,
748
+ "learning_rate": 1.6129373530632e-05,
749
+ "loss": 0.2268,
750
+ "step": 56000
751
+ },
752
+ {
753
+ "epoch": 7.81,
754
+ "learning_rate": 1.609486931268151e-05,
755
+ "loss": 0.2189,
756
+ "step": 56500
757
+ },
758
+ {
759
+ "epoch": 7.88,
760
+ "learning_rate": 1.6060295948001662e-05,
761
+ "loss": 0.2221,
762
+ "step": 57000
763
+ },
764
+ {
765
+ "epoch": 7.95,
766
+ "learning_rate": 1.602579173005117e-05,
767
+ "loss": 0.2206,
768
+ "step": 57500
769
+ },
770
+ {
771
+ "epoch": 8.0,
772
+ "eval_bleu": 71.7397,
773
+ "eval_gen_len": 12.9546,
774
+ "eval_loss": 0.2024153172969818,
775
+ "eval_runtime": 331.9482,
776
+ "eval_samples_per_second": 86.748,
777
+ "eval_steps_per_second": 2.711,
778
+ "step": 57848
779
+ },
780
+ {
781
+ "epoch": 8.02,
782
+ "learning_rate": 1.599121836537132e-05,
783
+ "loss": 0.2218,
784
+ "step": 58000
785
+ },
786
+ {
787
+ "epoch": 8.09,
788
+ "learning_rate": 1.595664500069147e-05,
789
+ "loss": 0.2122,
790
+ "step": 58500
791
+ },
792
+ {
793
+ "epoch": 8.16,
794
+ "learning_rate": 1.5922071636011617e-05,
795
+ "loss": 0.214,
796
+ "step": 59000
797
+ },
798
+ {
799
+ "epoch": 8.23,
800
+ "learning_rate": 1.5887498271331766e-05,
801
+ "loss": 0.2145,
802
+ "step": 59500
803
+ },
804
+ {
805
+ "epoch": 8.3,
806
+ "learning_rate": 1.5852924906651918e-05,
807
+ "loss": 0.2138,
808
+ "step": 60000
809
+ },
810
+ {
811
+ "epoch": 8.37,
812
+ "learning_rate": 1.5818351541972066e-05,
813
+ "loss": 0.2157,
814
+ "step": 60500
815
+ },
816
+ {
817
+ "epoch": 8.44,
818
+ "learning_rate": 1.5783778177292215e-05,
819
+ "loss": 0.211,
820
+ "step": 61000
821
+ },
822
+ {
823
+ "epoch": 8.51,
824
+ "learning_rate": 1.5749204812612363e-05,
825
+ "loss": 0.2078,
826
+ "step": 61500
827
+ },
828
+ {
829
+ "epoch": 8.57,
830
+ "learning_rate": 1.5714631447932512e-05,
831
+ "loss": 0.2137,
832
+ "step": 62000
833
+ },
834
+ {
835
+ "epoch": 8.64,
836
+ "learning_rate": 1.5680058083252664e-05,
837
+ "loss": 0.2164,
838
+ "step": 62500
839
+ },
840
+ {
841
+ "epoch": 8.71,
842
+ "learning_rate": 1.5645484718572812e-05,
843
+ "loss": 0.2141,
844
+ "step": 63000
845
+ },
846
+ {
847
+ "epoch": 8.78,
848
+ "learning_rate": 1.5610980500622322e-05,
849
+ "loss": 0.2139,
850
+ "step": 63500
851
+ },
852
+ {
853
+ "epoch": 8.85,
854
+ "learning_rate": 1.557640713594247e-05,
855
+ "loss": 0.2121,
856
+ "step": 64000
857
+ },
858
+ {
859
+ "epoch": 8.92,
860
+ "learning_rate": 1.554183377126262e-05,
861
+ "loss": 0.2151,
862
+ "step": 64500
863
+ },
864
+ {
865
+ "epoch": 8.99,
866
+ "learning_rate": 1.550726040658277e-05,
867
+ "loss": 0.2117,
868
+ "step": 65000
869
+ },
870
+ {
871
+ "epoch": 9.0,
872
+ "eval_bleu": 72.143,
873
+ "eval_gen_len": 12.9467,
874
+ "eval_loss": 0.19509540498256683,
875
+ "eval_runtime": 327.4667,
876
+ "eval_samples_per_second": 87.936,
877
+ "eval_steps_per_second": 2.748,
878
+ "step": 65079
879
+ },
880
+ {
881
+ "epoch": 9.06,
882
+ "learning_rate": 1.547268704190292e-05,
883
+ "loss": 0.2024,
884
+ "step": 65500
885
+ },
886
+ {
887
+ "epoch": 9.13,
888
+ "learning_rate": 1.5438113677223068e-05,
889
+ "loss": 0.2026,
890
+ "step": 66000
891
+ },
892
+ {
893
+ "epoch": 9.2,
894
+ "learning_rate": 1.5403540312543216e-05,
895
+ "loss": 0.2066,
896
+ "step": 66500
897
+ },
898
+ {
899
+ "epoch": 9.27,
900
+ "learning_rate": 1.5368966947863365e-05,
901
+ "loss": 0.1989,
902
+ "step": 67000
903
+ },
904
+ {
905
+ "epoch": 9.33,
906
+ "learning_rate": 1.5334462729912875e-05,
907
+ "loss": 0.2057,
908
+ "step": 67500
909
+ },
910
+ {
911
+ "epoch": 9.4,
912
+ "learning_rate": 1.5299889365233026e-05,
913
+ "loss": 0.2047,
914
+ "step": 68000
915
+ },
916
+ {
917
+ "epoch": 9.47,
918
+ "learning_rate": 1.5265316000553175e-05,
919
+ "loss": 0.2022,
920
+ "step": 68500
921
+ },
922
+ {
923
+ "epoch": 9.54,
924
+ "learning_rate": 1.5230811782602683e-05,
925
+ "loss": 0.2055,
926
+ "step": 69000
927
+ },
928
+ {
929
+ "epoch": 9.61,
930
+ "learning_rate": 1.5196238417922833e-05,
931
+ "loss": 0.2064,
932
+ "step": 69500
933
+ },
934
+ {
935
+ "epoch": 9.68,
936
+ "learning_rate": 1.5161665053242982e-05,
937
+ "loss": 0.2041,
938
+ "step": 70000
939
+ },
940
+ {
941
+ "epoch": 9.75,
942
+ "learning_rate": 1.5127091688563132e-05,
943
+ "loss": 0.2,
944
+ "step": 70500
945
+ },
946
+ {
947
+ "epoch": 9.82,
948
+ "learning_rate": 1.509251832388328e-05,
949
+ "loss": 0.1993,
950
+ "step": 71000
951
+ },
952
+ {
953
+ "epoch": 9.89,
954
+ "learning_rate": 1.505794495920343e-05,
955
+ "loss": 0.2008,
956
+ "step": 71500
957
+ },
958
+ {
959
+ "epoch": 9.96,
960
+ "learning_rate": 1.5023371594523579e-05,
961
+ "loss": 0.2018,
962
+ "step": 72000
963
+ },
964
+ {
965
+ "epoch": 10.0,
966
+ "eval_bleu": 72.2777,
967
+ "eval_gen_len": 12.9741,
968
+ "eval_loss": 0.19072672724723816,
969
+ "eval_runtime": 327.3978,
970
+ "eval_samples_per_second": 87.954,
971
+ "eval_steps_per_second": 2.749,
972
+ "step": 72310
973
+ },
974
+ {
975
+ "epoch": 10.03,
976
+ "learning_rate": 1.4988798229843731e-05,
977
+ "loss": 0.1979,
978
+ "step": 72500
979
+ },
980
+ {
981
+ "epoch": 10.1,
982
+ "learning_rate": 1.4954294011893237e-05,
983
+ "loss": 0.1957,
984
+ "step": 73000
985
+ },
986
+ {
987
+ "epoch": 10.16,
988
+ "learning_rate": 1.4919720647213388e-05,
989
+ "loss": 0.1945,
990
+ "step": 73500
991
+ },
992
+ {
993
+ "epoch": 10.23,
994
+ "learning_rate": 1.4885147282533536e-05,
995
+ "loss": 0.1945,
996
+ "step": 74000
997
+ },
998
+ {
999
+ "epoch": 10.3,
1000
+ "learning_rate": 1.4850573917853686e-05,
1001
+ "loss": 0.1933,
1002
+ "step": 74500
1003
+ },
1004
+ {
1005
+ "epoch": 10.37,
1006
+ "learning_rate": 1.4816069699903194e-05,
1007
+ "loss": 0.19,
1008
+ "step": 75000
1009
+ },
1010
+ {
1011
+ "epoch": 10.44,
1012
+ "learning_rate": 1.4781496335223345e-05,
1013
+ "loss": 0.1929,
1014
+ "step": 75500
1015
+ },
1016
+ {
1017
+ "epoch": 10.51,
1018
+ "learning_rate": 1.4746922970543493e-05,
1019
+ "loss": 0.1957,
1020
+ "step": 76000
1021
+ },
1022
+ {
1023
+ "epoch": 10.58,
1024
+ "learning_rate": 1.4712349605863643e-05,
1025
+ "loss": 0.1944,
1026
+ "step": 76500
1027
+ },
1028
+ {
1029
+ "epoch": 10.65,
1030
+ "learning_rate": 1.4677776241183795e-05,
1031
+ "loss": 0.1984,
1032
+ "step": 77000
1033
+ },
1034
+ {
1035
+ "epoch": 10.72,
1036
+ "learning_rate": 1.4643202876503944e-05,
1037
+ "loss": 0.1901,
1038
+ "step": 77500
1039
+ },
1040
+ {
1041
+ "epoch": 10.79,
1042
+ "learning_rate": 1.4608629511824092e-05,
1043
+ "loss": 0.1935,
1044
+ "step": 78000
1045
+ },
1046
+ {
1047
+ "epoch": 10.86,
1048
+ "learning_rate": 1.4574056147144242e-05,
1049
+ "loss": 0.1921,
1050
+ "step": 78500
1051
+ },
1052
+ {
1053
+ "epoch": 10.93,
1054
+ "learning_rate": 1.4539551929193749e-05,
1055
+ "loss": 0.1909,
1056
+ "step": 79000
1057
+ },
1058
+ {
1059
+ "epoch": 10.99,
1060
+ "learning_rate": 1.45049785645139e-05,
1061
+ "loss": 0.1944,
1062
+ "step": 79500
1063
+ },
1064
+ {
1065
+ "epoch": 11.0,
1066
+ "eval_bleu": 72.448,
1067
+ "eval_gen_len": 12.9678,
1068
+ "eval_loss": 0.18560351431369781,
1069
+ "eval_runtime": 331.6458,
1070
+ "eval_samples_per_second": 86.828,
1071
+ "eval_steps_per_second": 2.714,
1072
+ "step": 79541
1073
+ },
1074
+ {
1075
+ "epoch": 11.06,
1076
+ "learning_rate": 1.447040519983405e-05,
1077
+ "loss": 0.1887,
1078
+ "step": 80000
1079
+ },
1080
+ {
1081
+ "epoch": 11.13,
1082
+ "learning_rate": 1.44358318351542e-05,
1083
+ "loss": 0.1842,
1084
+ "step": 80500
1085
+ },
1086
+ {
1087
+ "epoch": 11.2,
1088
+ "learning_rate": 1.4401327617203706e-05,
1089
+ "loss": 0.1891,
1090
+ "step": 81000
1091
+ },
1092
+ {
1093
+ "epoch": 11.27,
1094
+ "learning_rate": 1.4366754252523856e-05,
1095
+ "loss": 0.1866,
1096
+ "step": 81500
1097
+ },
1098
+ {
1099
+ "epoch": 11.34,
1100
+ "learning_rate": 1.4332180887844008e-05,
1101
+ "loss": 0.1851,
1102
+ "step": 82000
1103
+ },
1104
+ {
1105
+ "epoch": 11.41,
1106
+ "learning_rate": 1.4297607523164156e-05,
1107
+ "loss": 0.185,
1108
+ "step": 82500
1109
+ },
1110
+ {
1111
+ "epoch": 11.48,
1112
+ "learning_rate": 1.4263172451943024e-05,
1113
+ "loss": 0.1872,
1114
+ "step": 83000
1115
+ },
1116
+ {
1117
+ "epoch": 11.55,
1118
+ "learning_rate": 1.4228599087263172e-05,
1119
+ "loss": 0.1847,
1120
+ "step": 83500
1121
+ },
1122
+ {
1123
+ "epoch": 11.62,
1124
+ "learning_rate": 1.4194025722583323e-05,
1125
+ "loss": 0.1838,
1126
+ "step": 84000
1127
+ },
1128
+ {
1129
+ "epoch": 11.69,
1130
+ "learning_rate": 1.4159452357903471e-05,
1131
+ "loss": 0.1883,
1132
+ "step": 84500
1133
+ },
1134
+ {
1135
+ "epoch": 11.75,
1136
+ "learning_rate": 1.4124948139952981e-05,
1137
+ "loss": 0.1832,
1138
+ "step": 85000
1139
+ },
1140
+ {
1141
+ "epoch": 11.82,
1142
+ "learning_rate": 1.409037477527313e-05,
1143
+ "loss": 0.1876,
1144
+ "step": 85500
1145
+ },
1146
+ {
1147
+ "epoch": 11.89,
1148
+ "learning_rate": 1.405580141059328e-05,
1149
+ "loss": 0.1861,
1150
+ "step": 86000
1151
+ },
1152
+ {
1153
+ "epoch": 11.96,
1154
+ "learning_rate": 1.4021228045913428e-05,
1155
+ "loss": 0.1856,
1156
+ "step": 86500
1157
+ },
1158
+ {
1159
+ "epoch": 12.0,
1160
+ "eval_bleu": 72.6473,
1161
+ "eval_gen_len": 12.9764,
1162
+ "eval_loss": 0.18210774660110474,
1163
+ "eval_runtime": 326.793,
1164
+ "eval_samples_per_second": 88.117,
1165
+ "eval_steps_per_second": 2.754,
1166
+ "step": 86772
1167
+ },
1168
+ {
1169
+ "epoch": 12.03,
1170
+ "learning_rate": 1.3986654681233578e-05,
1171
+ "loss": 0.1826,
1172
+ "step": 87000
1173
+ },
1174
+ {
1175
+ "epoch": 12.1,
1176
+ "learning_rate": 1.3952081316553727e-05,
1177
+ "loss": 0.1785,
1178
+ "step": 87500
1179
+ },
1180
+ {
1181
+ "epoch": 12.17,
1182
+ "learning_rate": 1.3917507951873877e-05,
1183
+ "loss": 0.1791,
1184
+ "step": 88000
1185
+ },
1186
+ {
1187
+ "epoch": 12.24,
1188
+ "learning_rate": 1.3882934587194025e-05,
1189
+ "loss": 0.1777,
1190
+ "step": 88500
1191
+ },
1192
+ {
1193
+ "epoch": 12.31,
1194
+ "learning_rate": 1.3848361222514177e-05,
1195
+ "loss": 0.1786,
1196
+ "step": 89000
1197
+ },
1198
+ {
1199
+ "epoch": 12.38,
1200
+ "learning_rate": 1.3813857004563684e-05,
1201
+ "loss": 0.1792,
1202
+ "step": 89500
1203
+ },
1204
+ {
1205
+ "epoch": 12.45,
1206
+ "learning_rate": 1.3779283639883834e-05,
1207
+ "loss": 0.179,
1208
+ "step": 90000
1209
+ },
1210
+ {
1211
+ "epoch": 12.52,
1212
+ "learning_rate": 1.3744779421933344e-05,
1213
+ "loss": 0.1777,
1214
+ "step": 90500
1215
+ },
1216
+ {
1217
+ "epoch": 12.58,
1218
+ "learning_rate": 1.3710206057253492e-05,
1219
+ "loss": 0.1816,
1220
+ "step": 91000
1221
+ },
1222
+ {
1223
+ "epoch": 12.65,
1224
+ "learning_rate": 1.367563269257364e-05,
1225
+ "loss": 0.1772,
1226
+ "step": 91500
1227
+ },
1228
+ {
1229
+ "epoch": 12.72,
1230
+ "learning_rate": 1.3641059327893791e-05,
1231
+ "loss": 0.1778,
1232
+ "step": 92000
1233
+ },
1234
+ {
1235
+ "epoch": 12.79,
1236
+ "learning_rate": 1.360648596321394e-05,
1237
+ "loss": 0.182,
1238
+ "step": 92500
1239
+ },
1240
+ {
1241
+ "epoch": 12.86,
1242
+ "learning_rate": 1.357198174526345e-05,
1243
+ "loss": 0.1805,
1244
+ "step": 93000
1245
+ },
1246
+ {
1247
+ "epoch": 12.93,
1248
+ "learning_rate": 1.35374083805836e-05,
1249
+ "loss": 0.1781,
1250
+ "step": 93500
1251
+ },
1252
+ {
1253
+ "epoch": 13.0,
1254
+ "learning_rate": 1.3502835015903748e-05,
1255
+ "loss": 0.1816,
1256
+ "step": 94000
1257
+ },
1258
+ {
1259
+ "epoch": 13.0,
1260
+ "eval_bleu": 72.7148,
1261
+ "eval_gen_len": 12.9977,
1262
+ "eval_loss": 0.17964386940002441,
1263
+ "eval_runtime": 331.7781,
1264
+ "eval_samples_per_second": 86.793,
1265
+ "eval_steps_per_second": 2.713,
1266
+ "step": 94003
1267
+ },
1268
+ {
1269
+ "epoch": 13.07,
1270
+ "learning_rate": 1.3468261651223898e-05,
1271
+ "loss": 0.1737,
1272
+ "step": 94500
1273
+ },
1274
+ {
1275
+ "epoch": 13.14,
1276
+ "learning_rate": 1.3433757433273406e-05,
1277
+ "loss": 0.1732,
1278
+ "step": 95000
1279
+ },
1280
+ {
1281
+ "epoch": 13.21,
1282
+ "learning_rate": 1.3399184068593556e-05,
1283
+ "loss": 0.1714,
1284
+ "step": 95500
1285
+ },
1286
+ {
1287
+ "epoch": 13.28,
1288
+ "learning_rate": 1.3364610703913705e-05,
1289
+ "loss": 0.1744,
1290
+ "step": 96000
1291
+ },
1292
+ {
1293
+ "epoch": 13.35,
1294
+ "learning_rate": 1.3330037339233855e-05,
1295
+ "loss": 0.1734,
1296
+ "step": 96500
1297
+ },
1298
+ {
1299
+ "epoch": 13.41,
1300
+ "learning_rate": 1.3295463974554004e-05,
1301
+ "loss": 0.1759,
1302
+ "step": 97000
1303
+ },
1304
+ {
1305
+ "epoch": 13.48,
1306
+ "learning_rate": 1.3260890609874154e-05,
1307
+ "loss": 0.177,
1308
+ "step": 97500
1309
+ },
1310
+ {
1311
+ "epoch": 13.55,
1312
+ "learning_rate": 1.3226317245194304e-05,
1313
+ "loss": 0.175,
1314
+ "step": 98000
1315
+ },
1316
+ {
1317
+ "epoch": 13.62,
1318
+ "learning_rate": 1.3191743880514454e-05,
1319
+ "loss": 0.1714,
1320
+ "step": 98500
1321
+ },
1322
+ {
1323
+ "epoch": 13.69,
1324
+ "learning_rate": 1.315723966256396e-05,
1325
+ "loss": 0.1735,
1326
+ "step": 99000
1327
+ },
1328
+ {
1329
+ "epoch": 13.76,
1330
+ "learning_rate": 1.312266629788411e-05,
1331
+ "loss": 0.1706,
1332
+ "step": 99500
1333
+ },
1334
+ {
1335
+ "epoch": 13.83,
1336
+ "learning_rate": 1.308809293320426e-05,
1337
+ "loss": 0.1725,
1338
+ "step": 100000
1339
+ },
1340
+ {
1341
+ "epoch": 13.9,
1342
+ "learning_rate": 1.3053519568524411e-05,
1343
+ "loss": 0.1697,
1344
+ "step": 100500
1345
+ },
1346
+ {
1347
+ "epoch": 13.97,
1348
+ "learning_rate": 1.301894620384456e-05,
1349
+ "loss": 0.1735,
1350
+ "step": 101000
1351
+ },
1352
+ {
1353
+ "epoch": 14.0,
1354
+ "eval_bleu": 72.9217,
1355
+ "eval_gen_len": 12.9687,
1356
+ "eval_loss": 0.17671316862106323,
1357
+ "eval_runtime": 331.3829,
1358
+ "eval_samples_per_second": 86.896,
1359
+ "eval_steps_per_second": 2.716,
1360
+ "step": 101234
1361
+ },
1362
+ {
1363
+ "epoch": 14.04,
1364
+ "learning_rate": 1.2984441985894068e-05,
1365
+ "loss": 0.1672,
1366
+ "step": 101500
1367
+ },
1368
+ {
1369
+ "epoch": 14.11,
1370
+ "learning_rate": 1.2949868621214216e-05,
1371
+ "loss": 0.1681,
1372
+ "step": 102000
1373
+ },
1374
+ {
1375
+ "epoch": 14.18,
1376
+ "learning_rate": 1.2915295256534366e-05,
1377
+ "loss": 0.1679,
1378
+ "step": 102500
1379
+ },
1380
+ {
1381
+ "epoch": 14.24,
1382
+ "learning_rate": 1.2880721891854517e-05,
1383
+ "loss": 0.1656,
1384
+ "step": 103000
1385
+ },
1386
+ {
1387
+ "epoch": 14.31,
1388
+ "learning_rate": 1.2846217673904025e-05,
1389
+ "loss": 0.1677,
1390
+ "step": 103500
1391
+ },
1392
+ {
1393
+ "epoch": 14.38,
1394
+ "learning_rate": 1.2811644309224173e-05,
1395
+ "loss": 0.1675,
1396
+ "step": 104000
1397
+ },
1398
+ {
1399
+ "epoch": 14.45,
1400
+ "learning_rate": 1.2777070944544323e-05,
1401
+ "loss": 0.1693,
1402
+ "step": 104500
1403
+ },
1404
+ {
1405
+ "epoch": 14.52,
1406
+ "learning_rate": 1.2742497579864475e-05,
1407
+ "loss": 0.1662,
1408
+ "step": 105000
1409
+ },
1410
+ {
1411
+ "epoch": 14.59,
1412
+ "learning_rate": 1.2708062508643341e-05,
1413
+ "loss": 0.1666,
1414
+ "step": 105500
1415
+ },
1416
+ {
1417
+ "epoch": 14.66,
1418
+ "learning_rate": 1.2673489143963491e-05,
1419
+ "loss": 0.1711,
1420
+ "step": 106000
1421
+ },
1422
+ {
1423
+ "epoch": 14.73,
1424
+ "learning_rate": 1.263891577928364e-05,
1425
+ "loss": 0.1676,
1426
+ "step": 106500
1427
+ },
1428
+ {
1429
+ "epoch": 14.8,
1430
+ "learning_rate": 1.260434241460379e-05,
1431
+ "loss": 0.1665,
1432
+ "step": 107000
1433
+ },
1434
+ {
1435
+ "epoch": 14.87,
1436
+ "learning_rate": 1.2569769049923939e-05,
1437
+ "loss": 0.1715,
1438
+ "step": 107500
1439
+ },
1440
+ {
1441
+ "epoch": 14.94,
1442
+ "learning_rate": 1.2535195685244089e-05,
1443
+ "loss": 0.1673,
1444
+ "step": 108000
1445
+ },
1446
+ {
1447
+ "epoch": 15.0,
1448
+ "eval_bleu": 73.0674,
1449
+ "eval_gen_len": 12.9771,
1450
+ "eval_loss": 0.17453685402870178,
1451
+ "eval_runtime": 331.4528,
1452
+ "eval_samples_per_second": 86.878,
1453
+ "eval_steps_per_second": 2.715,
1454
+ "step": 108465
1455
+ },
1456
+ {
1457
+ "epoch": 15.0,
1458
+ "learning_rate": 1.2500622320564237e-05,
1459
+ "loss": 0.1662,
1460
+ "step": 108500
1461
+ },
1462
+ {
1463
+ "epoch": 15.07,
1464
+ "learning_rate": 1.2466048955884387e-05,
1465
+ "loss": 0.1629,
1466
+ "step": 109000
1467
+ },
1468
+ {
1469
+ "epoch": 15.14,
1470
+ "learning_rate": 1.2431475591204536e-05,
1471
+ "loss": 0.1616,
1472
+ "step": 109500
1473
+ },
1474
+ {
1475
+ "epoch": 15.21,
1476
+ "learning_rate": 1.2396971373254046e-05,
1477
+ "loss": 0.1626,
1478
+ "step": 110000
1479
+ },
1480
+ {
1481
+ "epoch": 15.28,
1482
+ "learning_rate": 1.2362398008574194e-05,
1483
+ "loss": 0.16,
1484
+ "step": 110500
1485
+ },
1486
+ {
1487
+ "epoch": 15.35,
1488
+ "learning_rate": 1.2327824643894344e-05,
1489
+ "loss": 0.1635,
1490
+ "step": 111000
1491
+ },
1492
+ {
1493
+ "epoch": 15.42,
1494
+ "learning_rate": 1.2293251279214493e-05,
1495
+ "loss": 0.161,
1496
+ "step": 111500
1497
+ },
1498
+ {
1499
+ "epoch": 15.49,
1500
+ "learning_rate": 1.2258747061264003e-05,
1501
+ "loss": 0.1618,
1502
+ "step": 112000
1503
+ },
1504
+ {
1505
+ "epoch": 15.56,
1506
+ "learning_rate": 1.2224173696584151e-05,
1507
+ "loss": 0.1667,
1508
+ "step": 112500
1509
+ },
1510
+ {
1511
+ "epoch": 15.63,
1512
+ "learning_rate": 1.2189600331904301e-05,
1513
+ "loss": 0.1609,
1514
+ "step": 113000
1515
+ },
1516
+ {
1517
+ "epoch": 15.7,
1518
+ "learning_rate": 1.215502696722445e-05,
1519
+ "loss": 0.1653,
1520
+ "step": 113500
1521
+ },
1522
+ {
1523
+ "epoch": 15.77,
1524
+ "learning_rate": 1.212052274927396e-05,
1525
+ "loss": 0.1619,
1526
+ "step": 114000
1527
+ },
1528
+ {
1529
+ "epoch": 15.83,
1530
+ "learning_rate": 1.2085949384594108e-05,
1531
+ "loss": 0.1632,
1532
+ "step": 114500
1533
+ },
1534
+ {
1535
+ "epoch": 15.9,
1536
+ "learning_rate": 1.2051376019914258e-05,
1537
+ "loss": 0.1682,
1538
+ "step": 115000
1539
+ },
1540
+ {
1541
+ "epoch": 15.97,
1542
+ "learning_rate": 1.2016802655234407e-05,
1543
+ "loss": 0.1603,
1544
+ "step": 115500
1545
+ },
1546
+ {
1547
+ "epoch": 16.0,
1548
+ "eval_bleu": 73.0685,
1549
+ "eval_gen_len": 13.0001,
1550
+ "eval_loss": 0.1714019477367401,
1551
+ "eval_runtime": 331.5243,
1552
+ "eval_samples_per_second": 86.859,
1553
+ "eval_steps_per_second": 2.715,
1554
+ "step": 115696
1555
+ },
1556
+ {
1557
+ "epoch": 16.04,
1558
+ "learning_rate": 1.1982298437283917e-05,
1559
+ "loss": 0.158,
1560
+ "step": 116000
1561
+ },
1562
+ {
1563
+ "epoch": 16.11,
1564
+ "learning_rate": 1.1947725072604067e-05,
1565
+ "loss": 0.1612,
1566
+ "step": 116500
1567
+ },
1568
+ {
1569
+ "epoch": 16.18,
1570
+ "learning_rate": 1.1913151707924215e-05,
1571
+ "loss": 0.1541,
1572
+ "step": 117000
1573
+ },
1574
+ {
1575
+ "epoch": 16.25,
1576
+ "learning_rate": 1.1878578343244364e-05,
1577
+ "loss": 0.1585,
1578
+ "step": 117500
1579
+ },
1580
+ {
1581
+ "epoch": 16.32,
1582
+ "learning_rate": 1.1844074125293874e-05,
1583
+ "loss": 0.1608,
1584
+ "step": 118000
1585
+ },
1586
+ {
1587
+ "epoch": 16.39,
1588
+ "learning_rate": 1.1809500760614024e-05,
1589
+ "loss": 0.1593,
1590
+ "step": 118500
1591
+ },
1592
+ {
1593
+ "epoch": 16.46,
1594
+ "learning_rate": 1.1774927395934172e-05,
1595
+ "loss": 0.156,
1596
+ "step": 119000
1597
+ },
1598
+ {
1599
+ "epoch": 16.53,
1600
+ "learning_rate": 1.1740354031254323e-05,
1601
+ "loss": 0.1573,
1602
+ "step": 119500
1603
+ },
1604
+ {
1605
+ "epoch": 16.6,
1606
+ "learning_rate": 1.1705780666574471e-05,
1607
+ "loss": 0.1584,
1608
+ "step": 120000
1609
+ },
1610
+ {
1611
+ "epoch": 16.66,
1612
+ "learning_rate": 1.1671276448623981e-05,
1613
+ "loss": 0.1595,
1614
+ "step": 120500
1615
+ },
1616
+ {
1617
+ "epoch": 16.73,
1618
+ "learning_rate": 1.1636772230673489e-05,
1619
+ "loss": 0.1604,
1620
+ "step": 121000
1621
+ },
1622
+ {
1623
+ "epoch": 16.8,
1624
+ "learning_rate": 1.160219886599364e-05,
1625
+ "loss": 0.1608,
1626
+ "step": 121500
1627
+ },
1628
+ {
1629
+ "epoch": 16.87,
1630
+ "learning_rate": 1.1567625501313788e-05,
1631
+ "loss": 0.1545,
1632
+ "step": 122000
1633
+ },
1634
+ {
1635
+ "epoch": 16.94,
1636
+ "learning_rate": 1.1533052136633938e-05,
1637
+ "loss": 0.1599,
1638
+ "step": 122500
1639
+ },
1640
+ {
1641
+ "epoch": 17.0,
1642
+ "eval_bleu": 73.1736,
1643
+ "eval_gen_len": 13.0044,
1644
+ "eval_loss": 0.17083990573883057,
1645
+ "eval_runtime": 331.5218,
1646
+ "eval_samples_per_second": 86.86,
1647
+ "eval_steps_per_second": 2.715,
1648
+ "step": 122927
1649
+ },
1650
+ {
1651
+ "epoch": 17.01,
1652
+ "learning_rate": 1.1498478771954086e-05,
1653
+ "loss": 0.1588,
1654
+ "step": 123000
1655
+ },
1656
+ {
1657
+ "epoch": 17.08,
1658
+ "learning_rate": 1.1463905407274237e-05,
1659
+ "loss": 0.1534,
1660
+ "step": 123500
1661
+ },
1662
+ {
1663
+ "epoch": 17.15,
1664
+ "learning_rate": 1.1429332042594385e-05,
1665
+ "loss": 0.1555,
1666
+ "step": 124000
1667
+ },
1668
+ {
1669
+ "epoch": 17.22,
1670
+ "learning_rate": 1.1394758677914535e-05,
1671
+ "loss": 0.1545,
1672
+ "step": 124500
1673
+ },
1674
+ {
1675
+ "epoch": 17.29,
1676
+ "learning_rate": 1.1360185313234684e-05,
1677
+ "loss": 0.1522,
1678
+ "step": 125000
1679
+ },
1680
+ {
1681
+ "epoch": 17.36,
1682
+ "learning_rate": 1.1325750242013553e-05,
1683
+ "loss": 0.1581,
1684
+ "step": 125500
1685
+ },
1686
+ {
1687
+ "epoch": 17.42,
1688
+ "learning_rate": 1.1291176877333703e-05,
1689
+ "loss": 0.1571,
1690
+ "step": 126000
1691
+ },
1692
+ {
1693
+ "epoch": 17.49,
1694
+ "learning_rate": 1.1256603512653852e-05,
1695
+ "loss": 0.1521,
1696
+ "step": 126500
1697
+ },
1698
+ {
1699
+ "epoch": 17.56,
1700
+ "learning_rate": 1.1222030147974002e-05,
1701
+ "loss": 0.1536,
1702
+ "step": 127000
1703
+ },
1704
+ {
1705
+ "epoch": 17.63,
1706
+ "learning_rate": 1.118745678329415e-05,
1707
+ "loss": 0.1565,
1708
+ "step": 127500
1709
+ },
1710
+ {
1711
+ "epoch": 17.7,
1712
+ "learning_rate": 1.1152883418614299e-05,
1713
+ "loss": 0.1516,
1714
+ "step": 128000
1715
+ },
1716
+ {
1717
+ "epoch": 17.77,
1718
+ "learning_rate": 1.1118310053934449e-05,
1719
+ "loss": 0.1525,
1720
+ "step": 128500
1721
+ },
1722
+ {
1723
+ "epoch": 17.84,
1724
+ "learning_rate": 1.1083736689254598e-05,
1725
+ "loss": 0.1543,
1726
+ "step": 129000
1727
+ },
1728
+ {
1729
+ "epoch": 17.91,
1730
+ "learning_rate": 1.1049232471304107e-05,
1731
+ "loss": 0.1526,
1732
+ "step": 129500
1733
+ },
1734
+ {
1735
+ "epoch": 17.98,
1736
+ "learning_rate": 1.1014659106624258e-05,
1737
+ "loss": 0.1547,
1738
+ "step": 130000
1739
+ },
1740
+ {
1741
+ "epoch": 18.0,
1742
+ "eval_bleu": 73.1982,
1743
+ "eval_gen_len": 12.9891,
1744
+ "eval_loss": 0.16940416395664215,
1745
+ "eval_runtime": 331.787,
1746
+ "eval_samples_per_second": 86.791,
1747
+ "eval_steps_per_second": 2.713,
1748
+ "step": 130158
1749
+ },
1750
+ {
1751
+ "epoch": 18.05,
1752
+ "learning_rate": 1.0980085741944406e-05,
1753
+ "loss": 0.1542,
1754
+ "step": 130500
1755
+ },
1756
+ {
1757
+ "epoch": 18.12,
1758
+ "learning_rate": 1.0945512377264556e-05,
1759
+ "loss": 0.1527,
1760
+ "step": 131000
1761
+ },
1762
+ {
1763
+ "epoch": 18.19,
1764
+ "learning_rate": 1.0911008159314064e-05,
1765
+ "loss": 0.148,
1766
+ "step": 131500
1767
+ },
1768
+ {
1769
+ "epoch": 18.25,
1770
+ "learning_rate": 1.0876503941363574e-05,
1771
+ "loss": 0.148,
1772
+ "step": 132000
1773
+ },
1774
+ {
1775
+ "epoch": 18.32,
1776
+ "learning_rate": 1.0841930576683723e-05,
1777
+ "loss": 0.1474,
1778
+ "step": 132500
1779
+ },
1780
+ {
1781
+ "epoch": 18.39,
1782
+ "learning_rate": 1.0807357212003873e-05,
1783
+ "loss": 0.1525,
1784
+ "step": 133000
1785
+ },
1786
+ {
1787
+ "epoch": 18.46,
1788
+ "learning_rate": 1.0772783847324021e-05,
1789
+ "loss": 0.1502,
1790
+ "step": 133500
1791
+ },
1792
+ {
1793
+ "epoch": 18.53,
1794
+ "learning_rate": 1.0738210482644172e-05,
1795
+ "loss": 0.1532,
1796
+ "step": 134000
1797
+ },
1798
+ {
1799
+ "epoch": 18.6,
1800
+ "learning_rate": 1.070363711796432e-05,
1801
+ "loss": 0.1517,
1802
+ "step": 134500
1803
+ },
1804
+ {
1805
+ "epoch": 18.67,
1806
+ "learning_rate": 1.066906375328447e-05,
1807
+ "loss": 0.1456,
1808
+ "step": 135000
1809
+ },
1810
+ {
1811
+ "epoch": 18.74,
1812
+ "learning_rate": 1.0634490388604619e-05,
1813
+ "loss": 0.154,
1814
+ "step": 135500
1815
+ },
1816
+ {
1817
+ "epoch": 18.81,
1818
+ "learning_rate": 1.0599986170654129e-05,
1819
+ "loss": 0.151,
1820
+ "step": 136000
1821
+ },
1822
+ {
1823
+ "epoch": 18.88,
1824
+ "learning_rate": 1.0565481952703638e-05,
1825
+ "loss": 0.152,
1826
+ "step": 136500
1827
+ },
1828
+ {
1829
+ "epoch": 18.95,
1830
+ "learning_rate": 1.0530977734753147e-05,
1831
+ "loss": 0.1505,
1832
+ "step": 137000
1833
+ },
1834
+ {
1835
+ "epoch": 19.0,
1836
+ "eval_bleu": 73.2671,
1837
+ "eval_gen_len": 12.9711,
1838
+ "eval_loss": 0.16765068471431732,
1839
+ "eval_runtime": 325.5863,
1840
+ "eval_samples_per_second": 88.444,
1841
+ "eval_steps_per_second": 2.764,
1842
+ "step": 137389
1843
+ },
1844
+ {
1845
+ "epoch": 19.02,
1846
+ "learning_rate": 1.0496404370073297e-05,
1847
+ "loss": 0.1522,
1848
+ "step": 137500
1849
+ },
1850
+ {
1851
+ "epoch": 19.08,
1852
+ "learning_rate": 1.0461969298852166e-05,
1853
+ "loss": 0.1462,
1854
+ "step": 138000
1855
+ },
1856
+ {
1857
+ "epoch": 19.15,
1858
+ "learning_rate": 1.0427395934172315e-05,
1859
+ "loss": 0.145,
1860
+ "step": 138500
1861
+ },
1862
+ {
1863
+ "epoch": 19.22,
1864
+ "learning_rate": 1.0392822569492465e-05,
1865
+ "loss": 0.1479,
1866
+ "step": 139000
1867
+ },
1868
+ {
1869
+ "epoch": 19.29,
1870
+ "learning_rate": 1.0358318351541975e-05,
1871
+ "loss": 0.1457,
1872
+ "step": 139500
1873
+ },
1874
+ {
1875
+ "epoch": 19.36,
1876
+ "learning_rate": 1.0323744986862123e-05,
1877
+ "loss": 0.1502,
1878
+ "step": 140000
1879
+ },
1880
+ {
1881
+ "epoch": 19.43,
1882
+ "learning_rate": 1.0289171622182272e-05,
1883
+ "loss": 0.1474,
1884
+ "step": 140500
1885
+ },
1886
+ {
1887
+ "epoch": 19.5,
1888
+ "learning_rate": 1.0254598257502422e-05,
1889
+ "loss": 0.1467,
1890
+ "step": 141000
1891
+ },
1892
+ {
1893
+ "epoch": 19.57,
1894
+ "learning_rate": 1.022002489282257e-05,
1895
+ "loss": 0.1514,
1896
+ "step": 141500
1897
+ },
1898
+ {
1899
+ "epoch": 19.64,
1900
+ "learning_rate": 1.018545152814272e-05,
1901
+ "loss": 0.1507,
1902
+ "step": 142000
1903
+ },
1904
+ {
1905
+ "epoch": 19.71,
1906
+ "learning_rate": 1.0150878163462869e-05,
1907
+ "loss": 0.1489,
1908
+ "step": 142500
1909
+ },
1910
+ {
1911
+ "epoch": 19.78,
1912
+ "learning_rate": 1.0116373945512379e-05,
1913
+ "loss": 0.1471,
1914
+ "step": 143000
1915
+ },
1916
+ {
1917
+ "epoch": 19.85,
1918
+ "learning_rate": 1.0081800580832529e-05,
1919
+ "loss": 0.1455,
1920
+ "step": 143500
1921
+ },
1922
+ {
1923
+ "epoch": 19.91,
1924
+ "learning_rate": 1.0047227216152677e-05,
1925
+ "loss": 0.1459,
1926
+ "step": 144000
1927
+ },
1928
+ {
1929
+ "epoch": 19.98,
1930
+ "learning_rate": 1.0012653851472826e-05,
1931
+ "loss": 0.1464,
1932
+ "step": 144500
1933
+ },
1934
+ {
1935
+ "epoch": 20.0,
1936
+ "eval_bleu": 73.3543,
1937
+ "eval_gen_len": 12.9918,
1938
+ "eval_loss": 0.16689570248126984,
1939
+ "eval_runtime": 325.9649,
1940
+ "eval_samples_per_second": 88.341,
1941
+ "eval_steps_per_second": 2.761,
1942
+ "step": 144620
1943
+ },
1944
+ {
1945
+ "epoch": 20.05,
1946
+ "learning_rate": 9.978080486792976e-06,
1947
+ "loss": 0.1458,
1948
+ "step": 145000
1949
+ },
1950
+ {
1951
+ "epoch": 20.12,
1952
+ "learning_rate": 9.943507122113125e-06,
1953
+ "loss": 0.1418,
1954
+ "step": 145500
1955
+ },
1956
+ {
1957
+ "epoch": 20.19,
1958
+ "learning_rate": 9.908933757433275e-06,
1959
+ "loss": 0.1446,
1960
+ "step": 146000
1961
+ },
1962
+ {
1963
+ "epoch": 20.26,
1964
+ "learning_rate": 9.874360392753423e-06,
1965
+ "loss": 0.1473,
1966
+ "step": 146500
1967
+ },
1968
+ {
1969
+ "epoch": 20.33,
1970
+ "learning_rate": 9.839787028073573e-06,
1971
+ "loss": 0.1471,
1972
+ "step": 147000
1973
+ },
1974
+ {
1975
+ "epoch": 20.4,
1976
+ "learning_rate": 9.805213663393722e-06,
1977
+ "loss": 0.1411,
1978
+ "step": 147500
1979
+ },
1980
+ {
1981
+ "epoch": 20.47,
1982
+ "learning_rate": 9.770640298713872e-06,
1983
+ "loss": 0.1439,
1984
+ "step": 148000
1985
+ },
1986
+ {
1987
+ "epoch": 20.54,
1988
+ "learning_rate": 9.73606693403402e-06,
1989
+ "loss": 0.1452,
1990
+ "step": 148500
1991
+ },
1992
+ {
1993
+ "epoch": 20.61,
1994
+ "learning_rate": 9.70149356935417e-06,
1995
+ "loss": 0.1422,
1996
+ "step": 149000
1997
+ },
1998
+ {
1999
+ "epoch": 20.67,
2000
+ "learning_rate": 9.66692020467432e-06,
2001
+ "loss": 0.1427,
2002
+ "step": 149500
2003
+ },
2004
+ {
2005
+ "epoch": 20.74,
2006
+ "learning_rate": 9.63234683999447e-06,
2007
+ "loss": 0.1423,
2008
+ "step": 150000
2009
+ },
2010
+ {
2011
+ "epoch": 20.81,
2012
+ "learning_rate": 9.597773475314618e-06,
2013
+ "loss": 0.1445,
2014
+ "step": 150500
2015
+ },
2016
+ {
2017
+ "epoch": 20.88,
2018
+ "learning_rate": 9.563269257364128e-06,
2019
+ "loss": 0.1465,
2020
+ "step": 151000
2021
+ },
2022
+ {
2023
+ "epoch": 20.95,
2024
+ "learning_rate": 9.528695892684276e-06,
2025
+ "loss": 0.1449,
2026
+ "step": 151500
2027
+ },
2028
+ {
2029
+ "epoch": 21.0,
2030
+ "eval_bleu": 73.3506,
2031
+ "eval_gen_len": 12.9916,
2032
+ "eval_loss": 0.16551125049591064,
2033
+ "eval_runtime": 331.0362,
2034
+ "eval_samples_per_second": 86.987,
2035
+ "eval_steps_per_second": 2.719,
2036
+ "step": 151851
2037
+ },
2038
+ {
2039
+ "epoch": 21.02,
2040
+ "learning_rate": 9.494122528004426e-06,
2041
+ "loss": 0.1462,
2042
+ "step": 152000
2043
+ },
2044
+ {
2045
+ "epoch": 21.09,
2046
+ "learning_rate": 9.459549163324575e-06,
2047
+ "loss": 0.1398,
2048
+ "step": 152500
2049
+ },
2050
+ {
2051
+ "epoch": 21.16,
2052
+ "learning_rate": 9.424975798644725e-06,
2053
+ "loss": 0.1408,
2054
+ "step": 153000
2055
+ },
2056
+ {
2057
+ "epoch": 21.23,
2058
+ "learning_rate": 9.390402433964875e-06,
2059
+ "loss": 0.1415,
2060
+ "step": 153500
2061
+ },
2062
+ {
2063
+ "epoch": 21.3,
2064
+ "learning_rate": 9.355829069285024e-06,
2065
+ "loss": 0.1426,
2066
+ "step": 154000
2067
+ },
2068
+ {
2069
+ "epoch": 21.37,
2070
+ "learning_rate": 9.321255704605174e-06,
2071
+ "loss": 0.1402,
2072
+ "step": 154500
2073
+ },
2074
+ {
2075
+ "epoch": 21.44,
2076
+ "learning_rate": 9.286682339925322e-06,
2077
+ "loss": 0.1406,
2078
+ "step": 155000
2079
+ },
2080
+ {
2081
+ "epoch": 21.5,
2082
+ "learning_rate": 9.252108975245473e-06,
2083
+ "loss": 0.1375,
2084
+ "step": 155500
2085
+ },
2086
+ {
2087
+ "epoch": 21.57,
2088
+ "learning_rate": 9.217535610565621e-06,
2089
+ "loss": 0.1429,
2090
+ "step": 156000
2091
+ },
2092
+ {
2093
+ "epoch": 21.64,
2094
+ "learning_rate": 9.18296224588577e-06,
2095
+ "loss": 0.1426,
2096
+ "step": 156500
2097
+ },
2098
+ {
2099
+ "epoch": 21.71,
2100
+ "learning_rate": 9.14838888120592e-06,
2101
+ "loss": 0.1433,
2102
+ "step": 157000
2103
+ },
2104
+ {
2105
+ "epoch": 21.78,
2106
+ "learning_rate": 9.113815516526068e-06,
2107
+ "loss": 0.1414,
2108
+ "step": 157500
2109
+ },
2110
+ {
2111
+ "epoch": 21.85,
2112
+ "learning_rate": 9.079242151846218e-06,
2113
+ "loss": 0.1469,
2114
+ "step": 158000
2115
+ },
2116
+ {
2117
+ "epoch": 21.92,
2118
+ "learning_rate": 9.044668787166367e-06,
2119
+ "loss": 0.1397,
2120
+ "step": 158500
2121
+ },
2122
+ {
2123
+ "epoch": 21.99,
2124
+ "learning_rate": 9.010164569215877e-06,
2125
+ "loss": 0.1389,
2126
+ "step": 159000
2127
+ },
2128
+ {
2129
+ "epoch": 22.0,
2130
+ "eval_bleu": 73.4685,
2131
+ "eval_gen_len": 13.01,
2132
+ "eval_loss": 0.16494113206863403,
2133
+ "eval_runtime": 326.5923,
2134
+ "eval_samples_per_second": 88.171,
2135
+ "eval_steps_per_second": 2.756,
2136
+ "step": 159082
2137
+ },
2138
+ {
2139
+ "epoch": 22.06,
2140
+ "learning_rate": 8.975591204536025e-06,
2141
+ "loss": 0.1405,
2142
+ "step": 159500
2143
+ },
2144
+ {
2145
+ "epoch": 22.13,
2146
+ "learning_rate": 8.941017839856175e-06,
2147
+ "loss": 0.1378,
2148
+ "step": 160000
2149
+ },
2150
+ {
2151
+ "epoch": 22.2,
2152
+ "learning_rate": 8.906444475176324e-06,
2153
+ "loss": 0.137,
2154
+ "step": 160500
2155
+ },
2156
+ {
2157
+ "epoch": 22.27,
2158
+ "learning_rate": 8.871940257225834e-06,
2159
+ "loss": 0.1416,
2160
+ "step": 161000
2161
+ },
2162
+ {
2163
+ "epoch": 22.33,
2164
+ "learning_rate": 8.837366892545984e-06,
2165
+ "loss": 0.138,
2166
+ "step": 161500
2167
+ },
2168
+ {
2169
+ "epoch": 22.4,
2170
+ "learning_rate": 8.802793527866132e-06,
2171
+ "loss": 0.1389,
2172
+ "step": 162000
2173
+ },
2174
+ {
2175
+ "epoch": 22.47,
2176
+ "learning_rate": 8.768220163186283e-06,
2177
+ "loss": 0.1422,
2178
+ "step": 162500
2179
+ },
2180
+ {
2181
+ "epoch": 22.54,
2182
+ "learning_rate": 8.73371594523579e-06,
2183
+ "loss": 0.1354,
2184
+ "step": 163000
2185
+ },
2186
+ {
2187
+ "epoch": 22.61,
2188
+ "learning_rate": 8.699142580555941e-06,
2189
+ "loss": 0.1404,
2190
+ "step": 163500
2191
+ },
2192
+ {
2193
+ "epoch": 22.68,
2194
+ "learning_rate": 8.66456921587609e-06,
2195
+ "loss": 0.1416,
2196
+ "step": 164000
2197
+ },
2198
+ {
2199
+ "epoch": 22.75,
2200
+ "learning_rate": 8.6300649979256e-06,
2201
+ "loss": 0.1395,
2202
+ "step": 164500
2203
+ },
2204
+ {
2205
+ "epoch": 22.82,
2206
+ "learning_rate": 8.595491633245748e-06,
2207
+ "loss": 0.1405,
2208
+ "step": 165000
2209
+ },
2210
+ {
2211
+ "epoch": 22.89,
2212
+ "learning_rate": 8.560918268565898e-06,
2213
+ "loss": 0.1377,
2214
+ "step": 165500
2215
+ },
2216
+ {
2217
+ "epoch": 22.96,
2218
+ "learning_rate": 8.526344903886046e-06,
2219
+ "loss": 0.1362,
2220
+ "step": 166000
2221
+ },
2222
+ {
2223
+ "epoch": 23.0,
2224
+ "eval_bleu": 73.4517,
2225
+ "eval_gen_len": 13.0042,
2226
+ "eval_loss": 0.16470502316951752,
2227
+ "eval_runtime": 326.27,
2228
+ "eval_samples_per_second": 88.258,
2229
+ "eval_steps_per_second": 2.758,
2230
+ "step": 166313
2231
+ },
2232
+ {
2233
+ "epoch": 23.03,
2234
+ "learning_rate": 8.491771539206197e-06,
2235
+ "loss": 0.1374,
2236
+ "step": 166500
2237
+ },
2238
+ {
2239
+ "epoch": 23.1,
2240
+ "learning_rate": 8.457198174526345e-06,
2241
+ "loss": 0.1373,
2242
+ "step": 167000
2243
+ },
2244
+ {
2245
+ "epoch": 23.16,
2246
+ "learning_rate": 8.422624809846495e-06,
2247
+ "loss": 0.1361,
2248
+ "step": 167500
2249
+ },
2250
+ {
2251
+ "epoch": 23.23,
2252
+ "learning_rate": 8.388051445166644e-06,
2253
+ "loss": 0.1333,
2254
+ "step": 168000
2255
+ },
2256
+ {
2257
+ "epoch": 23.3,
2258
+ "learning_rate": 8.353547227216154e-06,
2259
+ "loss": 0.1369,
2260
+ "step": 168500
2261
+ },
2262
+ {
2263
+ "epoch": 23.37,
2264
+ "learning_rate": 8.318973862536302e-06,
2265
+ "loss": 0.1369,
2266
+ "step": 169000
2267
+ },
2268
+ {
2269
+ "epoch": 23.44,
2270
+ "learning_rate": 8.284400497856452e-06,
2271
+ "loss": 0.1364,
2272
+ "step": 169500
2273
+ },
2274
+ {
2275
+ "epoch": 23.51,
2276
+ "learning_rate": 8.2498271331766e-06,
2277
+ "loss": 0.1369,
2278
+ "step": 170000
2279
+ },
2280
+ {
2281
+ "epoch": 23.58,
2282
+ "learning_rate": 8.215253768496751e-06,
2283
+ "loss": 0.1362,
2284
+ "step": 170500
2285
+ },
2286
+ {
2287
+ "epoch": 23.65,
2288
+ "learning_rate": 8.180749550546259e-06,
2289
+ "loss": 0.1386,
2290
+ "step": 171000
2291
+ },
2292
+ {
2293
+ "epoch": 23.72,
2294
+ "learning_rate": 8.14617618586641e-06,
2295
+ "loss": 0.1386,
2296
+ "step": 171500
2297
+ },
2298
+ {
2299
+ "epoch": 23.79,
2300
+ "learning_rate": 8.111602821186558e-06,
2301
+ "loss": 0.1352,
2302
+ "step": 172000
2303
+ },
2304
+ {
2305
+ "epoch": 23.86,
2306
+ "learning_rate": 8.077029456506708e-06,
2307
+ "loss": 0.1367,
2308
+ "step": 172500
2309
+ },
2310
+ {
2311
+ "epoch": 23.92,
2312
+ "learning_rate": 8.042525238556218e-06,
2313
+ "loss": 0.1355,
2314
+ "step": 173000
2315
+ },
2316
+ {
2317
+ "epoch": 23.99,
2318
+ "learning_rate": 8.008021020605726e-06,
2319
+ "loss": 0.1322,
2320
+ "step": 173500
2321
+ },
2322
+ {
2323
+ "epoch": 24.0,
2324
+ "eval_bleu": 73.5427,
2325
+ "eval_gen_len": 13.025,
2326
+ "eval_loss": 0.16514846682548523,
2327
+ "eval_runtime": 326.8973,
2328
+ "eval_samples_per_second": 88.089,
2329
+ "eval_steps_per_second": 2.753,
2330
+ "step": 173544
2331
+ },
2332
+ {
2333
+ "epoch": 24.06,
2334
+ "learning_rate": 7.973447655925876e-06,
2335
+ "loss": 0.132,
2336
+ "step": 174000
2337
+ },
2338
+ {
2339
+ "epoch": 24.13,
2340
+ "learning_rate": 7.938874291246024e-06,
2341
+ "loss": 0.1332,
2342
+ "step": 174500
2343
+ },
2344
+ {
2345
+ "epoch": 24.2,
2346
+ "learning_rate": 7.904300926566175e-06,
2347
+ "loss": 0.1351,
2348
+ "step": 175000
2349
+ },
2350
+ {
2351
+ "epoch": 24.27,
2352
+ "learning_rate": 7.869727561886323e-06,
2353
+ "loss": 0.135,
2354
+ "step": 175500
2355
+ },
2356
+ {
2357
+ "epoch": 24.34,
2358
+ "learning_rate": 7.835154197206473e-06,
2359
+ "loss": 0.1354,
2360
+ "step": 176000
2361
+ },
2362
+ {
2363
+ "epoch": 24.41,
2364
+ "learning_rate": 7.800580832526622e-06,
2365
+ "loss": 0.1351,
2366
+ "step": 176500
2367
+ },
2368
+ {
2369
+ "epoch": 24.48,
2370
+ "learning_rate": 7.766007467846772e-06,
2371
+ "loss": 0.132,
2372
+ "step": 177000
2373
+ },
2374
+ {
2375
+ "epoch": 24.55,
2376
+ "learning_rate": 7.73150324989628e-06,
2377
+ "loss": 0.1317,
2378
+ "step": 177500
2379
+ },
2380
+ {
2381
+ "epoch": 24.62,
2382
+ "learning_rate": 7.69692988521643e-06,
2383
+ "loss": 0.1366,
2384
+ "step": 178000
2385
+ },
2386
+ {
2387
+ "epoch": 24.69,
2388
+ "learning_rate": 7.662356520536579e-06,
2389
+ "loss": 0.1328,
2390
+ "step": 178500
2391
+ },
2392
+ {
2393
+ "epoch": 24.75,
2394
+ "learning_rate": 7.627783155856728e-06,
2395
+ "loss": 0.1348,
2396
+ "step": 179000
2397
+ },
2398
+ {
2399
+ "epoch": 24.82,
2400
+ "learning_rate": 7.5932097911768775e-06,
2401
+ "loss": 0.1345,
2402
+ "step": 179500
2403
+ },
2404
+ {
2405
+ "epoch": 24.89,
2406
+ "learning_rate": 7.558705573226387e-06,
2407
+ "loss": 0.1331,
2408
+ "step": 180000
2409
+ },
2410
+ {
2411
+ "epoch": 24.96,
2412
+ "learning_rate": 7.524132208546537e-06,
2413
+ "loss": 0.1337,
2414
+ "step": 180500
2415
+ },
2416
+ {
2417
+ "epoch": 25.0,
2418
+ "eval_bleu": 73.5595,
2419
+ "eval_gen_len": 13.0306,
2420
+ "eval_loss": 0.16346952319145203,
2421
+ "eval_runtime": 325.4712,
2422
+ "eval_samples_per_second": 88.475,
2423
+ "eval_steps_per_second": 2.765,
2424
+ "step": 180775
2425
+ },
2426
+ {
2427
+ "epoch": 25.03,
2428
+ "learning_rate": 7.489558843866685e-06,
2429
+ "loss": 0.1342,
2430
+ "step": 181000
2431
+ },
2432
+ {
2433
+ "epoch": 25.1,
2434
+ "learning_rate": 7.4549854791868344e-06,
2435
+ "loss": 0.1325,
2436
+ "step": 181500
2437
+ },
2438
+ {
2439
+ "epoch": 25.17,
2440
+ "learning_rate": 7.420481261236344e-06,
2441
+ "loss": 0.1297,
2442
+ "step": 182000
2443
+ },
2444
+ {
2445
+ "epoch": 25.24,
2446
+ "learning_rate": 7.385907896556494e-06,
2447
+ "loss": 0.1344,
2448
+ "step": 182500
2449
+ },
2450
+ {
2451
+ "epoch": 25.31,
2452
+ "learning_rate": 7.351334531876643e-06,
2453
+ "loss": 0.133,
2454
+ "step": 183000
2455
+ },
2456
+ {
2457
+ "epoch": 25.38,
2458
+ "learning_rate": 7.316761167196792e-06,
2459
+ "loss": 0.1311,
2460
+ "step": 183500
2461
+ },
2462
+ {
2463
+ "epoch": 25.45,
2464
+ "learning_rate": 7.282256949246301e-06,
2465
+ "loss": 0.1341,
2466
+ "step": 184000
2467
+ },
2468
+ {
2469
+ "epoch": 25.52,
2470
+ "learning_rate": 7.2476835845664506e-06,
2471
+ "loss": 0.1312,
2472
+ "step": 184500
2473
+ },
2474
+ {
2475
+ "epoch": 25.58,
2476
+ "learning_rate": 7.21317936661596e-06,
2477
+ "loss": 0.132,
2478
+ "step": 185000
2479
+ },
2480
+ {
2481
+ "epoch": 25.65,
2482
+ "learning_rate": 7.178606001936109e-06,
2483
+ "loss": 0.13,
2484
+ "step": 185500
2485
+ },
2486
+ {
2487
+ "epoch": 25.72,
2488
+ "learning_rate": 7.144032637256258e-06,
2489
+ "loss": 0.134,
2490
+ "step": 186000
2491
+ },
2492
+ {
2493
+ "epoch": 25.79,
2494
+ "learning_rate": 7.1094592725764075e-06,
2495
+ "loss": 0.1295,
2496
+ "step": 186500
2497
+ },
2498
+ {
2499
+ "epoch": 25.86,
2500
+ "learning_rate": 7.074885907896557e-06,
2501
+ "loss": 0.1332,
2502
+ "step": 187000
2503
+ },
2504
+ {
2505
+ "epoch": 25.93,
2506
+ "learning_rate": 7.040312543216706e-06,
2507
+ "loss": 0.133,
2508
+ "step": 187500
2509
+ },
2510
+ {
2511
+ "epoch": 26.0,
2512
+ "learning_rate": 7.0057391785368556e-06,
2513
+ "loss": 0.1308,
2514
+ "step": 188000
2515
+ },
2516
+ {
2517
+ "epoch": 26.0,
2518
+ "eval_bleu": 73.5529,
2519
+ "eval_gen_len": 13.0183,
2520
+ "eval_loss": 0.16344179213047028,
2521
+ "eval_runtime": 323.8431,
2522
+ "eval_samples_per_second": 88.92,
2523
+ "eval_steps_per_second": 2.779,
2524
+ "step": 188006
2525
+ },
2526
+ {
2527
+ "epoch": 26.07,
2528
+ "learning_rate": 6.971165813857005e-06,
2529
+ "loss": 0.1298,
2530
+ "step": 188500
2531
+ },
2532
+ {
2533
+ "epoch": 26.14,
2534
+ "learning_rate": 6.936661595906514e-06,
2535
+ "loss": 0.1277,
2536
+ "step": 189000
2537
+ },
2538
+ {
2539
+ "epoch": 26.21,
2540
+ "learning_rate": 6.902157377956024e-06,
2541
+ "loss": 0.1334,
2542
+ "step": 189500
2543
+ },
2544
+ {
2545
+ "epoch": 26.28,
2546
+ "learning_rate": 6.867584013276173e-06,
2547
+ "loss": 0.1289,
2548
+ "step": 190000
2549
+ },
2550
+ {
2551
+ "epoch": 26.34,
2552
+ "learning_rate": 6.833010648596322e-06,
2553
+ "loss": 0.1301,
2554
+ "step": 190500
2555
+ },
2556
+ {
2557
+ "epoch": 26.41,
2558
+ "learning_rate": 6.798437283916472e-06,
2559
+ "loss": 0.1318,
2560
+ "step": 191000
2561
+ },
2562
+ {
2563
+ "epoch": 26.48,
2564
+ "learning_rate": 6.763863919236621e-06,
2565
+ "loss": 0.1311,
2566
+ "step": 191500
2567
+ },
2568
+ {
2569
+ "epoch": 26.55,
2570
+ "learning_rate": 6.72929055455677e-06,
2571
+ "loss": 0.1314,
2572
+ "step": 192000
2573
+ },
2574
+ {
2575
+ "epoch": 26.62,
2576
+ "learning_rate": 6.694717189876919e-06,
2577
+ "loss": 0.131,
2578
+ "step": 192500
2579
+ },
2580
+ {
2581
+ "epoch": 26.69,
2582
+ "learning_rate": 6.660143825197068e-06,
2583
+ "loss": 0.1297,
2584
+ "step": 193000
2585
+ },
2586
+ {
2587
+ "epoch": 26.76,
2588
+ "learning_rate": 6.625639607246578e-06,
2589
+ "loss": 0.1284,
2590
+ "step": 193500
2591
+ },
2592
+ {
2593
+ "epoch": 26.83,
2594
+ "learning_rate": 6.591066242566727e-06,
2595
+ "loss": 0.1309,
2596
+ "step": 194000
2597
+ },
2598
+ {
2599
+ "epoch": 26.9,
2600
+ "learning_rate": 6.556562024616236e-06,
2601
+ "loss": 0.1281,
2602
+ "step": 194500
2603
+ },
2604
+ {
2605
+ "epoch": 26.97,
2606
+ "learning_rate": 6.521988659936386e-06,
2607
+ "loss": 0.1309,
2608
+ "step": 195000
2609
+ },
2610
+ {
2611
+ "epoch": 27.0,
2612
+ "eval_bleu": 73.6151,
2613
+ "eval_gen_len": 13.0011,
2614
+ "eval_loss": 0.16237156093120575,
2615
+ "eval_runtime": 322.9735,
2616
+ "eval_samples_per_second": 89.159,
2617
+ "eval_steps_per_second": 2.787,
2618
+ "step": 195237
2619
+ },
2620
+ {
2621
+ "epoch": 27.04,
2622
+ "learning_rate": 6.487415295256535e-06,
2623
+ "loss": 0.1252,
2624
+ "step": 195500
2625
+ },
2626
+ {
2627
+ "epoch": 27.11,
2628
+ "learning_rate": 6.452841930576684e-06,
2629
+ "loss": 0.1296,
2630
+ "step": 196000
2631
+ },
2632
+ {
2633
+ "epoch": 27.17,
2634
+ "learning_rate": 6.418268565896834e-06,
2635
+ "loss": 0.129,
2636
+ "step": 196500
2637
+ },
2638
+ {
2639
+ "epoch": 27.24,
2640
+ "learning_rate": 6.383695201216983e-06,
2641
+ "loss": 0.1249,
2642
+ "step": 197000
2643
+ },
2644
+ {
2645
+ "epoch": 27.31,
2646
+ "learning_rate": 6.349121836537132e-06,
2647
+ "loss": 0.1292,
2648
+ "step": 197500
2649
+ },
2650
+ {
2651
+ "epoch": 27.38,
2652
+ "learning_rate": 6.314548471857282e-06,
2653
+ "loss": 0.1278,
2654
+ "step": 198000
2655
+ },
2656
+ {
2657
+ "epoch": 27.45,
2658
+ "learning_rate": 6.279975107177431e-06,
2659
+ "loss": 0.1271,
2660
+ "step": 198500
2661
+ },
2662
+ {
2663
+ "epoch": 27.52,
2664
+ "learning_rate": 6.24547088922694e-06,
2665
+ "loss": 0.1298,
2666
+ "step": 199000
2667
+ },
2668
+ {
2669
+ "epoch": 27.59,
2670
+ "learning_rate": 6.210897524547089e-06,
2671
+ "loss": 0.1299,
2672
+ "step": 199500
2673
+ },
2674
+ {
2675
+ "epoch": 27.66,
2676
+ "learning_rate": 6.176324159867239e-06,
2677
+ "loss": 0.1283,
2678
+ "step": 200000
2679
+ },
2680
+ {
2681
+ "epoch": 27.73,
2682
+ "learning_rate": 6.141750795187388e-06,
2683
+ "loss": 0.1283,
2684
+ "step": 200500
2685
+ },
2686
+ {
2687
+ "epoch": 27.8,
2688
+ "learning_rate": 6.107246577236897e-06,
2689
+ "loss": 0.1307,
2690
+ "step": 201000
2691
+ },
2692
+ {
2693
+ "epoch": 27.87,
2694
+ "learning_rate": 6.072673212557046e-06,
2695
+ "loss": 0.1317,
2696
+ "step": 201500
2697
+ },
2698
+ {
2699
+ "epoch": 27.94,
2700
+ "learning_rate": 6.038099847877196e-06,
2701
+ "loss": 0.1269,
2702
+ "step": 202000
2703
+ },
2704
+ {
2705
+ "epoch": 28.0,
2706
+ "eval_bleu": 73.6409,
2707
+ "eval_gen_len": 13.0209,
2708
+ "eval_loss": 0.1626293659210205,
2709
+ "eval_runtime": 324.7844,
2710
+ "eval_samples_per_second": 88.662,
2711
+ "eval_steps_per_second": 2.771,
2712
+ "step": 202468
2713
+ },
2714
+ {
2715
+ "epoch": 28.0,
2716
+ "learning_rate": 6.003595629926705e-06,
2717
+ "loss": 0.1262,
2718
+ "step": 202500
2719
+ },
2720
+ {
2721
+ "epoch": 28.07,
2722
+ "learning_rate": 5.969091411976214e-06,
2723
+ "loss": 0.1246,
2724
+ "step": 203000
2725
+ },
2726
+ {
2727
+ "epoch": 28.14,
2728
+ "learning_rate": 5.934518047296364e-06,
2729
+ "loss": 0.1253,
2730
+ "step": 203500
2731
+ },
2732
+ {
2733
+ "epoch": 28.21,
2734
+ "learning_rate": 5.899944682616513e-06,
2735
+ "loss": 0.1229,
2736
+ "step": 204000
2737
+ },
2738
+ {
2739
+ "epoch": 28.28,
2740
+ "learning_rate": 5.865371317936662e-06,
2741
+ "loss": 0.1286,
2742
+ "step": 204500
2743
+ },
2744
+ {
2745
+ "epoch": 28.35,
2746
+ "learning_rate": 5.830797953256812e-06,
2747
+ "loss": 0.123,
2748
+ "step": 205000
2749
+ },
2750
+ {
2751
+ "epoch": 28.42,
2752
+ "learning_rate": 5.796224588576961e-06,
2753
+ "loss": 0.1266,
2754
+ "step": 205500
2755
+ },
2756
+ {
2757
+ "epoch": 28.49,
2758
+ "learning_rate": 5.76165122389711e-06,
2759
+ "loss": 0.1268,
2760
+ "step": 206000
2761
+ },
2762
+ {
2763
+ "epoch": 28.56,
2764
+ "learning_rate": 5.72707785921726e-06,
2765
+ "loss": 0.1268,
2766
+ "step": 206500
2767
+ },
2768
+ {
2769
+ "epoch": 28.63,
2770
+ "learning_rate": 5.692504494537408e-06,
2771
+ "loss": 0.1273,
2772
+ "step": 207000
2773
+ },
2774
+ {
2775
+ "epoch": 28.7,
2776
+ "learning_rate": 5.6579311298575576e-06,
2777
+ "loss": 0.1286,
2778
+ "step": 207500
2779
+ },
2780
+ {
2781
+ "epoch": 28.77,
2782
+ "learning_rate": 5.623496058636427e-06,
2783
+ "loss": 0.1281,
2784
+ "step": 208000
2785
+ },
2786
+ {
2787
+ "epoch": 28.83,
2788
+ "learning_rate": 5.588922693956576e-06,
2789
+ "loss": 0.1289,
2790
+ "step": 208500
2791
+ },
2792
+ {
2793
+ "epoch": 28.9,
2794
+ "learning_rate": 5.554349329276726e-06,
2795
+ "loss": 0.1285,
2796
+ "step": 209000
2797
+ },
2798
+ {
2799
+ "epoch": 28.97,
2800
+ "learning_rate": 5.5198451113262355e-06,
2801
+ "loss": 0.1239,
2802
+ "step": 209500
2803
+ },
2804
+ {
2805
+ "epoch": 29.0,
2806
+ "eval_bleu": 73.6486,
2807
+ "eval_gen_len": 13.023,
2808
+ "eval_loss": 0.1618330031633377,
2809
+ "eval_runtime": 321.6597,
2810
+ "eval_samples_per_second": 89.523,
2811
+ "eval_steps_per_second": 2.798,
2812
+ "step": 209699
2813
+ },
2814
+ {
2815
+ "epoch": 29.04,
2816
+ "learning_rate": 5.485271746646385e-06,
2817
+ "loss": 0.1327,
2818
+ "step": 210000
2819
+ },
2820
+ {
2821
+ "epoch": 29.11,
2822
+ "learning_rate": 5.450698381966534e-06,
2823
+ "loss": 0.126,
2824
+ "step": 210500
2825
+ },
2826
+ {
2827
+ "epoch": 29.18,
2828
+ "learning_rate": 5.416125017286683e-06,
2829
+ "loss": 0.1253,
2830
+ "step": 211000
2831
+ },
2832
+ {
2833
+ "epoch": 29.25,
2834
+ "learning_rate": 5.381551652606832e-06,
2835
+ "loss": 0.1265,
2836
+ "step": 211500
2837
+ },
2838
+ {
2839
+ "epoch": 29.32,
2840
+ "learning_rate": 5.346978287926981e-06,
2841
+ "loss": 0.1222,
2842
+ "step": 212000
2843
+ },
2844
+ {
2845
+ "epoch": 29.39,
2846
+ "learning_rate": 5.312404923247131e-06,
2847
+ "loss": 0.1258,
2848
+ "step": 212500
2849
+ },
2850
+ {
2851
+ "epoch": 29.46,
2852
+ "learning_rate": 5.27783155856728e-06,
2853
+ "loss": 0.1254,
2854
+ "step": 213000
2855
+ },
2856
+ {
2857
+ "epoch": 29.53,
2858
+ "learning_rate": 5.243258193887429e-06,
2859
+ "loss": 0.1257,
2860
+ "step": 213500
2861
+ },
2862
+ {
2863
+ "epoch": 29.59,
2864
+ "learning_rate": 5.208684829207579e-06,
2865
+ "loss": 0.1256,
2866
+ "step": 214000
2867
+ },
2868
+ {
2869
+ "epoch": 29.66,
2870
+ "learning_rate": 5.174111464527728e-06,
2871
+ "loss": 0.1245,
2872
+ "step": 214500
2873
+ },
2874
+ {
2875
+ "epoch": 29.73,
2876
+ "learning_rate": 5.139538099847877e-06,
2877
+ "loss": 0.1237,
2878
+ "step": 215000
2879
+ },
2880
+ {
2881
+ "epoch": 29.8,
2882
+ "learning_rate": 5.104964735168027e-06,
2883
+ "loss": 0.1243,
2884
+ "step": 215500
2885
+ },
2886
+ {
2887
+ "epoch": 29.87,
2888
+ "learning_rate": 5.070391370488177e-06,
2889
+ "loss": 0.1262,
2890
+ "step": 216000
2891
+ },
2892
+ {
2893
+ "epoch": 29.94,
2894
+ "learning_rate": 5.035818005808326e-06,
2895
+ "loss": 0.1216,
2896
+ "step": 216500
2897
+ },
2898
+ {
2899
+ "epoch": 30.0,
2900
+ "eval_bleu": 73.7267,
2901
+ "eval_gen_len": 13.0134,
2902
+ "eval_loss": 0.16207487881183624,
2903
+ "eval_runtime": 321.9244,
2904
+ "eval_samples_per_second": 89.45,
2905
+ "eval_steps_per_second": 2.796,
2906
+ "step": 216930
2907
+ },
2908
+ {
2909
+ "epoch": 30.01,
2910
+ "learning_rate": 5.001313787857834e-06,
2911
+ "loss": 0.1246,
2912
+ "step": 217000
2913
+ },
2914
+ {
2915
+ "epoch": 30.08,
2916
+ "learning_rate": 4.9667404231779845e-06,
2917
+ "loss": 0.1237,
2918
+ "step": 217500
2919
+ },
2920
+ {
2921
+ "epoch": 30.15,
2922
+ "learning_rate": 4.932167058498134e-06,
2923
+ "loss": 0.1242,
2924
+ "step": 218000
2925
+ },
2926
+ {
2927
+ "epoch": 30.22,
2928
+ "learning_rate": 4.897593693818283e-06,
2929
+ "loss": 0.1228,
2930
+ "step": 218500
2931
+ },
2932
+ {
2933
+ "epoch": 30.29,
2934
+ "learning_rate": 4.863020329138432e-06,
2935
+ "loss": 0.1239,
2936
+ "step": 219000
2937
+ },
2938
+ {
2939
+ "epoch": 30.36,
2940
+ "learning_rate": 4.828446964458581e-06,
2941
+ "loss": 0.1227,
2942
+ "step": 219500
2943
+ },
2944
+ {
2945
+ "epoch": 30.42,
2946
+ "learning_rate": 4.793873599778731e-06,
2947
+ "loss": 0.1218,
2948
+ "step": 220000
2949
+ },
2950
+ {
2951
+ "epoch": 30.49,
2952
+ "learning_rate": 4.7593002350988805e-06,
2953
+ "loss": 0.123,
2954
+ "step": 220500
2955
+ },
2956
+ {
2957
+ "epoch": 30.56,
2958
+ "learning_rate": 4.72472687041903e-06,
2959
+ "loss": 0.1272,
2960
+ "step": 221000
2961
+ },
2962
+ {
2963
+ "epoch": 30.63,
2964
+ "learning_rate": 4.690222652468539e-06,
2965
+ "loss": 0.1236,
2966
+ "step": 221500
2967
+ },
2968
+ {
2969
+ "epoch": 30.7,
2970
+ "learning_rate": 4.655649287788688e-06,
2971
+ "loss": 0.1246,
2972
+ "step": 222000
2973
+ },
2974
+ {
2975
+ "epoch": 30.77,
2976
+ "learning_rate": 4.6210759231088375e-06,
2977
+ "loss": 0.1239,
2978
+ "step": 222500
2979
+ },
2980
+ {
2981
+ "epoch": 30.84,
2982
+ "learning_rate": 4.586502558428987e-06,
2983
+ "loss": 0.1227,
2984
+ "step": 223000
2985
+ },
2986
+ {
2987
+ "epoch": 30.91,
2988
+ "learning_rate": 4.551998340478496e-06,
2989
+ "loss": 0.1228,
2990
+ "step": 223500
2991
+ },
2992
+ {
2993
+ "epoch": 30.98,
2994
+ "learning_rate": 4.517494122528005e-06,
2995
+ "loss": 0.1276,
2996
+ "step": 224000
2997
+ },
2998
+ {
2999
+ "epoch": 31.0,
3000
+ "eval_bleu": 73.7362,
3001
+ "eval_gen_len": 13.0236,
3002
+ "eval_loss": 0.16207154095172882,
3003
+ "eval_runtime": 322.2078,
3004
+ "eval_samples_per_second": 89.371,
3005
+ "eval_steps_per_second": 2.793,
3006
+ "step": 224161
3007
+ },
3008
+ {
3009
+ "epoch": 31.05,
3010
+ "learning_rate": 4.482920757848154e-06,
3011
+ "loss": 0.1211,
3012
+ "step": 224500
3013
+ },
3014
+ {
3015
+ "epoch": 31.12,
3016
+ "learning_rate": 4.4483473931683034e-06,
3017
+ "loss": 0.1211,
3018
+ "step": 225000
3019
+ },
3020
+ {
3021
+ "epoch": 31.19,
3022
+ "learning_rate": 4.413774028488453e-06,
3023
+ "loss": 0.1214,
3024
+ "step": 225500
3025
+ },
3026
+ {
3027
+ "epoch": 31.25,
3028
+ "learning_rate": 4.379200663808602e-06,
3029
+ "loss": 0.1215,
3030
+ "step": 226000
3031
+ },
3032
+ {
3033
+ "epoch": 31.32,
3034
+ "learning_rate": 4.3446272991287515e-06,
3035
+ "loss": 0.1232,
3036
+ "step": 226500
3037
+ },
3038
+ {
3039
+ "epoch": 31.39,
3040
+ "learning_rate": 4.310053934448901e-06,
3041
+ "loss": 0.1258,
3042
+ "step": 227000
3043
+ },
3044
+ {
3045
+ "epoch": 31.46,
3046
+ "learning_rate": 4.27548056976905e-06,
3047
+ "loss": 0.1229,
3048
+ "step": 227500
3049
+ },
3050
+ {
3051
+ "epoch": 31.53,
3052
+ "learning_rate": 4.240976351818559e-06,
3053
+ "loss": 0.1222,
3054
+ "step": 228000
3055
+ },
3056
+ {
3057
+ "epoch": 31.6,
3058
+ "learning_rate": 4.2064029871387084e-06,
3059
+ "loss": 0.1267,
3060
+ "step": 228500
3061
+ },
3062
+ {
3063
+ "epoch": 31.67,
3064
+ "learning_rate": 4.171829622458858e-06,
3065
+ "loss": 0.1212,
3066
+ "step": 229000
3067
+ },
3068
+ {
3069
+ "epoch": 31.74,
3070
+ "learning_rate": 4.137256257779008e-06,
3071
+ "loss": 0.124,
3072
+ "step": 229500
3073
+ },
3074
+ {
3075
+ "epoch": 31.81,
3076
+ "learning_rate": 4.102752039828516e-06,
3077
+ "loss": 0.1232,
3078
+ "step": 230000
3079
+ },
3080
+ {
3081
+ "epoch": 31.88,
3082
+ "learning_rate": 4.068178675148665e-06,
3083
+ "loss": 0.1228,
3084
+ "step": 230500
3085
+ },
3086
+ {
3087
+ "epoch": 31.95,
3088
+ "learning_rate": 4.033674457198175e-06,
3089
+ "loss": 0.1207,
3090
+ "step": 231000
3091
+ },
3092
+ {
3093
+ "epoch": 32.0,
3094
+ "eval_bleu": 73.7607,
3095
+ "eval_gen_len": 13.0179,
3096
+ "eval_loss": 0.162560373544693,
3097
+ "eval_runtime": 322.4139,
3098
+ "eval_samples_per_second": 89.314,
3099
+ "eval_steps_per_second": 2.791,
3100
+ "step": 231392
3101
+ },
3102
+ {
3103
+ "epoch": 32.01,
3104
+ "learning_rate": 3.9991010925183246e-06,
3105
+ "loss": 0.1249,
3106
+ "step": 231500
3107
+ },
3108
+ {
3109
+ "epoch": 32.08,
3110
+ "learning_rate": 3.9645968745678335e-06,
3111
+ "loss": 0.1206,
3112
+ "step": 232000
3113
+ },
3114
+ {
3115
+ "epoch": 32.15,
3116
+ "learning_rate": 3.930023509887982e-06,
3117
+ "loss": 0.12,
3118
+ "step": 232500
3119
+ },
3120
+ {
3121
+ "epoch": 32.22,
3122
+ "learning_rate": 3.895450145208132e-06,
3123
+ "loss": 0.1209,
3124
+ "step": 233000
3125
+ },
3126
+ {
3127
+ "epoch": 32.29,
3128
+ "learning_rate": 3.8608767805282815e-06,
3129
+ "loss": 0.1229,
3130
+ "step": 233500
3131
+ },
3132
+ {
3133
+ "epoch": 32.36,
3134
+ "learning_rate": 3.826303415848431e-06,
3135
+ "loss": 0.12,
3136
+ "step": 234000
3137
+ },
3138
+ {
3139
+ "epoch": 32.43,
3140
+ "learning_rate": 3.79173005116858e-06,
3141
+ "loss": 0.1222,
3142
+ "step": 234500
3143
+ },
3144
+ {
3145
+ "epoch": 32.5,
3146
+ "learning_rate": 3.757156686488729e-06,
3147
+ "loss": 0.1248,
3148
+ "step": 235000
3149
+ },
3150
+ {
3151
+ "epoch": 32.57,
3152
+ "learning_rate": 3.7225833218088785e-06,
3153
+ "loss": 0.1222,
3154
+ "step": 235500
3155
+ },
3156
+ {
3157
+ "epoch": 32.64,
3158
+ "learning_rate": 3.688009957129028e-06,
3159
+ "loss": 0.1216,
3160
+ "step": 236000
3161
+ },
3162
+ {
3163
+ "epoch": 32.71,
3164
+ "learning_rate": 3.6534365924491776e-06,
3165
+ "loss": 0.1235,
3166
+ "step": 236500
3167
+ },
3168
+ {
3169
+ "epoch": 32.78,
3170
+ "learning_rate": 3.618863227769327e-06,
3171
+ "loss": 0.1203,
3172
+ "step": 237000
3173
+ },
3174
+ {
3175
+ "epoch": 32.84,
3176
+ "learning_rate": 3.5842898630894762e-06,
3177
+ "loss": 0.1212,
3178
+ "step": 237500
3179
+ },
3180
+ {
3181
+ "epoch": 32.91,
3182
+ "learning_rate": 3.549785645138985e-06,
3183
+ "loss": 0.1208,
3184
+ "step": 238000
3185
+ },
3186
+ {
3187
+ "epoch": 32.98,
3188
+ "learning_rate": 3.5152122804591345e-06,
3189
+ "loss": 0.1217,
3190
+ "step": 238500
3191
+ },
3192
+ {
3193
+ "epoch": 33.0,
3194
+ "eval_bleu": 73.75,
3195
+ "eval_gen_len": 13.0163,
3196
+ "eval_loss": 0.1623145490884781,
3197
+ "eval_runtime": 322.1607,
3198
+ "eval_samples_per_second": 89.384,
3199
+ "eval_steps_per_second": 2.794,
3200
+ "step": 238623
3201
+ },
3202
+ {
3203
+ "epoch": 33.05,
3204
+ "learning_rate": 3.4806389157792843e-06,
3205
+ "loss": 0.1208,
3206
+ "step": 239000
3207
+ },
3208
+ {
3209
+ "epoch": 33.12,
3210
+ "learning_rate": 3.4460655510994336e-06,
3211
+ "loss": 0.1198,
3212
+ "step": 239500
3213
+ },
3214
+ {
3215
+ "epoch": 33.19,
3216
+ "learning_rate": 3.4114921864195825e-06,
3217
+ "loss": 0.1207,
3218
+ "step": 240000
3219
+ },
3220
+ {
3221
+ "epoch": 33.26,
3222
+ "learning_rate": 3.3769879684690915e-06,
3223
+ "loss": 0.1217,
3224
+ "step": 240500
3225
+ },
3226
+ {
3227
+ "epoch": 33.33,
3228
+ "learning_rate": 3.3424146037892413e-06,
3229
+ "loss": 0.1196,
3230
+ "step": 241000
3231
+ },
3232
+ {
3233
+ "epoch": 33.4,
3234
+ "learning_rate": 3.3078412391093906e-06,
3235
+ "loss": 0.12,
3236
+ "step": 241500
3237
+ },
3238
+ {
3239
+ "epoch": 33.47,
3240
+ "learning_rate": 3.27326787442954e-06,
3241
+ "loss": 0.1219,
3242
+ "step": 242000
3243
+ },
3244
+ {
3245
+ "epoch": 33.54,
3246
+ "learning_rate": 3.2387636564790485e-06,
3247
+ "loss": 0.1213,
3248
+ "step": 242500
3249
+ },
3250
+ {
3251
+ "epoch": 33.61,
3252
+ "learning_rate": 3.204190291799198e-06,
3253
+ "loss": 0.1208,
3254
+ "step": 243000
3255
+ },
3256
+ {
3257
+ "epoch": 33.67,
3258
+ "learning_rate": 3.1696169271193476e-06,
3259
+ "loss": 0.1206,
3260
+ "step": 243500
3261
+ },
3262
+ {
3263
+ "epoch": 33.74,
3264
+ "learning_rate": 3.135043562439497e-06,
3265
+ "loss": 0.1209,
3266
+ "step": 244000
3267
+ },
3268
+ {
3269
+ "epoch": 33.81,
3270
+ "learning_rate": 3.100539344489006e-06,
3271
+ "loss": 0.1187,
3272
+ "step": 244500
3273
+ },
3274
+ {
3275
+ "epoch": 33.88,
3276
+ "learning_rate": 3.0660351265385153e-06,
3277
+ "loss": 0.1184,
3278
+ "step": 245000
3279
+ },
3280
+ {
3281
+ "epoch": 33.95,
3282
+ "learning_rate": 3.0314617618586646e-06,
3283
+ "loss": 0.1194,
3284
+ "step": 245500
3285
+ },
3286
+ {
3287
+ "epoch": 34.0,
3288
+ "eval_bleu": 73.7724,
3289
+ "eval_gen_len": 13.0249,
3290
+ "eval_loss": 0.16194060444831848,
3291
+ "eval_runtime": 321.7496,
3292
+ "eval_samples_per_second": 89.498,
3293
+ "eval_steps_per_second": 2.797,
3294
+ "step": 245854
3295
+ },
3296
+ {
3297
+ "epoch": 34.02,
3298
+ "learning_rate": 2.996888397178814e-06,
3299
+ "loss": 0.1224,
3300
+ "step": 246000
3301
+ },
3302
+ {
3303
+ "epoch": 34.09,
3304
+ "learning_rate": 2.962315032498963e-06,
3305
+ "loss": 0.1187,
3306
+ "step": 246500
3307
+ },
3308
+ {
3309
+ "epoch": 34.16,
3310
+ "learning_rate": 2.927741667819112e-06,
3311
+ "loss": 0.12,
3312
+ "step": 247000
3313
+ },
3314
+ {
3315
+ "epoch": 34.23,
3316
+ "learning_rate": 2.8931683031392615e-06,
3317
+ "loss": 0.1197,
3318
+ "step": 247500
3319
+ },
3320
+ {
3321
+ "epoch": 34.3,
3322
+ "learning_rate": 2.858664085188771e-06,
3323
+ "loss": 0.1199,
3324
+ "step": 248000
3325
+ },
3326
+ {
3327
+ "epoch": 34.37,
3328
+ "learning_rate": 2.8240907205089203e-06,
3329
+ "loss": 0.1198,
3330
+ "step": 248500
3331
+ },
3332
+ {
3333
+ "epoch": 34.44,
3334
+ "learning_rate": 2.7895173558290696e-06,
3335
+ "loss": 0.1193,
3336
+ "step": 249000
3337
+ },
3338
+ {
3339
+ "epoch": 34.5,
3340
+ "learning_rate": 2.754943991149219e-06,
3341
+ "loss": 0.117,
3342
+ "step": 249500
3343
+ },
3344
+ {
3345
+ "epoch": 34.57,
3346
+ "learning_rate": 2.720370626469368e-06,
3347
+ "loss": 0.1213,
3348
+ "step": 250000
3349
+ },
3350
+ {
3351
+ "epoch": 34.64,
3352
+ "learning_rate": 2.685797261789518e-06,
3353
+ "loss": 0.1165,
3354
+ "step": 250500
3355
+ },
3356
+ {
3357
+ "epoch": 34.71,
3358
+ "learning_rate": 2.651223897109667e-06,
3359
+ "loss": 0.12,
3360
+ "step": 251000
3361
+ },
3362
+ {
3363
+ "epoch": 34.78,
3364
+ "learning_rate": 2.6166505324298163e-06,
3365
+ "loss": 0.1215,
3366
+ "step": 251500
3367
+ },
3368
+ {
3369
+ "epoch": 34.85,
3370
+ "learning_rate": 2.5820771677499656e-06,
3371
+ "loss": 0.1183,
3372
+ "step": 252000
3373
+ },
3374
+ {
3375
+ "epoch": 34.92,
3376
+ "learning_rate": 2.5475729497994746e-06,
3377
+ "loss": 0.1175,
3378
+ "step": 252500
3379
+ },
3380
+ {
3381
+ "epoch": 34.99,
3382
+ "learning_rate": 2.5129995851196243e-06,
3383
+ "loss": 0.1229,
3384
+ "step": 253000
3385
+ },
3386
+ {
3387
+ "epoch": 35.0,
3388
+ "eval_bleu": 73.7548,
3389
+ "eval_gen_len": 13.0268,
3390
+ "eval_loss": 0.16187956929206848,
3391
+ "eval_runtime": 322.0456,
3392
+ "eval_samples_per_second": 89.416,
3393
+ "eval_steps_per_second": 2.795,
3394
+ "step": 253085
3395
+ },
3396
+ {
3397
+ "epoch": 35.06,
3398
+ "learning_rate": 2.4784262204397733e-06,
3399
+ "loss": 0.117,
3400
+ "step": 253500
3401
+ },
3402
+ {
3403
+ "epoch": 35.13,
3404
+ "learning_rate": 2.443852855759923e-06,
3405
+ "loss": 0.1198,
3406
+ "step": 254000
3407
+ },
3408
+ {
3409
+ "epoch": 35.2,
3410
+ "learning_rate": 2.409348637809432e-06,
3411
+ "loss": 0.1191,
3412
+ "step": 254500
3413
+ },
3414
+ {
3415
+ "epoch": 35.26,
3416
+ "learning_rate": 2.3747752731295813e-06,
3417
+ "loss": 0.1171,
3418
+ "step": 255000
3419
+ },
3420
+ {
3421
+ "epoch": 35.33,
3422
+ "learning_rate": 2.3402710551790903e-06,
3423
+ "loss": 0.1169,
3424
+ "step": 255500
3425
+ },
3426
+ {
3427
+ "epoch": 35.4,
3428
+ "learning_rate": 2.3056976904992396e-06,
3429
+ "loss": 0.1202,
3430
+ "step": 256000
3431
+ },
3432
+ {
3433
+ "epoch": 35.47,
3434
+ "learning_rate": 2.271124325819389e-06,
3435
+ "loss": 0.1196,
3436
+ "step": 256500
3437
+ },
3438
+ {
3439
+ "epoch": 35.54,
3440
+ "learning_rate": 2.2365509611395383e-06,
3441
+ "loss": 0.1194,
3442
+ "step": 257000
3443
+ },
3444
+ {
3445
+ "epoch": 35.61,
3446
+ "learning_rate": 2.2019775964596876e-06,
3447
+ "loss": 0.1221,
3448
+ "step": 257500
3449
+ },
3450
+ {
3451
+ "epoch": 35.68,
3452
+ "learning_rate": 2.1674733785091966e-06,
3453
+ "loss": 0.1198,
3454
+ "step": 258000
3455
+ },
3456
+ {
3457
+ "epoch": 35.75,
3458
+ "learning_rate": 2.132900013829346e-06,
3459
+ "loss": 0.1185,
3460
+ "step": 258500
3461
+ },
3462
+ {
3463
+ "epoch": 35.82,
3464
+ "learning_rate": 2.0983266491494957e-06,
3465
+ "loss": 0.1188,
3466
+ "step": 259000
3467
+ },
3468
+ {
3469
+ "epoch": 35.89,
3470
+ "learning_rate": 2.0637532844696446e-06,
3471
+ "loss": 0.1199,
3472
+ "step": 259500
3473
+ },
3474
+ {
3475
+ "epoch": 35.96,
3476
+ "learning_rate": 2.029179919789794e-06,
3477
+ "loss": 0.1158,
3478
+ "step": 260000
3479
+ },
3480
+ {
3481
+ "epoch": 36.0,
3482
+ "eval_bleu": 73.7699,
3483
+ "eval_gen_len": 13.0219,
3484
+ "eval_loss": 0.16182634234428406,
3485
+ "eval_runtime": 322.0872,
3486
+ "eval_samples_per_second": 89.404,
3487
+ "eval_steps_per_second": 2.794,
3488
+ "step": 260316
3489
+ },
3490
+ {
3491
+ "epoch": 36.03,
3492
+ "learning_rate": 1.9946065551099433e-06,
3493
+ "loss": 0.121,
3494
+ "step": 260500
3495
+ },
3496
+ {
3497
+ "epoch": 36.09,
3498
+ "learning_rate": 1.9601023371594522e-06,
3499
+ "loss": 0.119,
3500
+ "step": 261000
3501
+ },
3502
+ {
3503
+ "epoch": 36.16,
3504
+ "learning_rate": 1.925528972479602e-06,
3505
+ "loss": 0.1198,
3506
+ "step": 261500
3507
+ },
3508
+ {
3509
+ "epoch": 36.23,
3510
+ "learning_rate": 1.8909556077997511e-06,
3511
+ "loss": 0.1204,
3512
+ "step": 262000
3513
+ },
3514
+ {
3515
+ "epoch": 36.3,
3516
+ "learning_rate": 1.8563822431199005e-06,
3517
+ "loss": 0.1191,
3518
+ "step": 262500
3519
+ },
3520
+ {
3521
+ "epoch": 36.37,
3522
+ "learning_rate": 1.82180887844005e-06,
3523
+ "loss": 0.1149,
3524
+ "step": 263000
3525
+ },
3526
+ {
3527
+ "epoch": 36.44,
3528
+ "learning_rate": 1.7872355137601993e-06,
3529
+ "loss": 0.1165,
3530
+ "step": 263500
3531
+ },
3532
+ {
3533
+ "epoch": 36.51,
3534
+ "learning_rate": 1.7526621490803487e-06,
3535
+ "loss": 0.1192,
3536
+ "step": 264000
3537
+ },
3538
+ {
3539
+ "epoch": 36.58,
3540
+ "learning_rate": 1.7180887844004978e-06,
3541
+ "loss": 0.1206,
3542
+ "step": 264500
3543
+ },
3544
+ {
3545
+ "epoch": 36.65,
3546
+ "learning_rate": 1.6835154197206474e-06,
3547
+ "loss": 0.12,
3548
+ "step": 265000
3549
+ },
3550
+ {
3551
+ "epoch": 36.72,
3552
+ "learning_rate": 1.6490112017701565e-06,
3553
+ "loss": 0.1159,
3554
+ "step": 265500
3555
+ },
3556
+ {
3557
+ "epoch": 36.79,
3558
+ "learning_rate": 1.6144378370903059e-06,
3559
+ "loss": 0.1205,
3560
+ "step": 266000
3561
+ },
3562
+ {
3563
+ "epoch": 36.86,
3564
+ "learning_rate": 1.579864472410455e-06,
3565
+ "loss": 0.1174,
3566
+ "step": 266500
3567
+ },
3568
+ {
3569
+ "epoch": 36.92,
3570
+ "learning_rate": 1.5452911077306045e-06,
3571
+ "loss": 0.1177,
3572
+ "step": 267000
3573
+ },
3574
+ {
3575
+ "epoch": 36.99,
3576
+ "learning_rate": 1.5107868897801133e-06,
3577
+ "loss": 0.117,
3578
+ "step": 267500
3579
+ },
3580
+ {
3581
+ "epoch": 37.0,
3582
+ "eval_bleu": 73.7699,
3583
+ "eval_gen_len": 13.0218,
3584
+ "eval_loss": 0.16194874048233032,
3585
+ "eval_runtime": 324.2894,
3586
+ "eval_samples_per_second": 88.797,
3587
+ "eval_steps_per_second": 2.775,
3588
+ "step": 267547
3589
+ },
3590
+ {
3591
+ "epoch": 37.06,
3592
+ "learning_rate": 1.4762135251002629e-06,
3593
+ "loss": 0.1176,
3594
+ "step": 268000
3595
+ },
3596
+ {
3597
+ "epoch": 37.13,
3598
+ "learning_rate": 1.4416401604204122e-06,
3599
+ "loss": 0.1191,
3600
+ "step": 268500
3601
+ },
3602
+ {
3603
+ "epoch": 37.2,
3604
+ "learning_rate": 1.4070667957405615e-06,
3605
+ "loss": 0.1161,
3606
+ "step": 269000
3607
+ },
3608
+ {
3609
+ "epoch": 37.27,
3610
+ "learning_rate": 1.3725625777900705e-06,
3611
+ "loss": 0.1156,
3612
+ "step": 269500
3613
+ },
3614
+ {
3615
+ "epoch": 37.34,
3616
+ "learning_rate": 1.33798921311022e-06,
3617
+ "loss": 0.116,
3618
+ "step": 270000
3619
+ },
3620
+ {
3621
+ "epoch": 37.41,
3622
+ "learning_rate": 1.3034158484303694e-06,
3623
+ "loss": 0.1189,
3624
+ "step": 270500
3625
+ },
3626
+ {
3627
+ "epoch": 37.48,
3628
+ "learning_rate": 1.2688424837505187e-06,
3629
+ "loss": 0.1176,
3630
+ "step": 271000
3631
+ },
3632
+ {
3633
+ "epoch": 37.55,
3634
+ "learning_rate": 1.2343382658000277e-06,
3635
+ "loss": 0.1162,
3636
+ "step": 271500
3637
+ },
3638
+ {
3639
+ "epoch": 37.62,
3640
+ "learning_rate": 1.1997649011201772e-06,
3641
+ "loss": 0.1211,
3642
+ "step": 272000
3643
+ },
3644
+ {
3645
+ "epoch": 37.68,
3646
+ "learning_rate": 1.1651915364403264e-06,
3647
+ "loss": 0.1158,
3648
+ "step": 272500
3649
+ },
3650
+ {
3651
+ "epoch": 37.75,
3652
+ "learning_rate": 1.130618171760476e-06,
3653
+ "loss": 0.12,
3654
+ "step": 273000
3655
+ },
3656
+ {
3657
+ "epoch": 37.82,
3658
+ "learning_rate": 1.0960448070806252e-06,
3659
+ "loss": 0.1176,
3660
+ "step": 273500
3661
+ },
3662
+ {
3663
+ "epoch": 37.89,
3664
+ "learning_rate": 1.0615405891301342e-06,
3665
+ "loss": 0.1195,
3666
+ "step": 274000
3667
+ },
3668
+ {
3669
+ "epoch": 37.96,
3670
+ "learning_rate": 1.0269672244502835e-06,
3671
+ "loss": 0.1181,
3672
+ "step": 274500
3673
+ },
3674
+ {
3675
+ "epoch": 38.0,
3676
+ "eval_bleu": 73.7582,
3677
+ "eval_gen_len": 13.0304,
3678
+ "eval_loss": 0.16226525604724884,
3679
+ "eval_runtime": 323.3851,
3680
+ "eval_samples_per_second": 89.046,
3681
+ "eval_steps_per_second": 2.783,
3682
+ "step": 274778
3683
+ },
3684
+ {
3685
+ "epoch": 38.03,
3686
+ "learning_rate": 9.923938597704329e-07,
3687
+ "loss": 0.1179,
3688
+ "step": 275000
3689
+ },
3690
+ {
3691
+ "epoch": 38.1,
3692
+ "learning_rate": 9.578204950905822e-07,
3693
+ "loss": 0.1173,
3694
+ "step": 275500
3695
+ },
3696
+ {
3697
+ "epoch": 38.17,
3698
+ "learning_rate": 9.233162771400914e-07,
3699
+ "loss": 0.1181,
3700
+ "step": 276000
3701
+ },
3702
+ {
3703
+ "epoch": 38.24,
3704
+ "learning_rate": 8.887429124602406e-07,
3705
+ "loss": 0.118,
3706
+ "step": 276500
3707
+ },
3708
+ {
3709
+ "epoch": 38.31,
3710
+ "learning_rate": 8.541695477803901e-07,
3711
+ "loss": 0.1171,
3712
+ "step": 277000
3713
+ },
3714
+ {
3715
+ "epoch": 38.38,
3716
+ "learning_rate": 8.195961831005395e-07,
3717
+ "loss": 0.1189,
3718
+ "step": 277500
3719
+ },
3720
+ {
3721
+ "epoch": 38.45,
3722
+ "learning_rate": 7.850919651500484e-07,
3723
+ "loss": 0.1182,
3724
+ "step": 278000
3725
+ },
3726
+ {
3727
+ "epoch": 38.51,
3728
+ "learning_rate": 7.505186004701978e-07,
3729
+ "loss": 0.1168,
3730
+ "step": 278500
3731
+ },
3732
+ {
3733
+ "epoch": 38.58,
3734
+ "learning_rate": 7.159452357903473e-07,
3735
+ "loss": 0.1165,
3736
+ "step": 279000
3737
+ },
3738
+ {
3739
+ "epoch": 38.65,
3740
+ "learning_rate": 6.813718711104965e-07,
3741
+ "loss": 0.1151,
3742
+ "step": 279500
3743
+ },
3744
+ {
3745
+ "epoch": 38.72,
3746
+ "learning_rate": 6.468676531600056e-07,
3747
+ "loss": 0.1172,
3748
+ "step": 280000
3749
+ },
3750
+ {
3751
+ "epoch": 38.79,
3752
+ "learning_rate": 6.122942884801549e-07,
3753
+ "loss": 0.1174,
3754
+ "step": 280500
3755
+ },
3756
+ {
3757
+ "epoch": 38.86,
3758
+ "learning_rate": 5.777209238003043e-07,
3759
+ "loss": 0.1209,
3760
+ "step": 281000
3761
+ },
3762
+ {
3763
+ "epoch": 38.93,
3764
+ "learning_rate": 5.431475591204537e-07,
3765
+ "loss": 0.1193,
3766
+ "step": 281500
3767
+ },
3768
+ {
3769
+ "epoch": 39.0,
3770
+ "learning_rate": 5.08574194440603e-07,
3771
+ "loss": 0.1132,
3772
+ "step": 282000
3773
+ },
3774
+ {
3775
+ "epoch": 39.0,
3776
+ "eval_bleu": 73.7861,
3777
+ "eval_gen_len": 13.0281,
3778
+ "eval_loss": 0.16211672127246857,
3779
+ "eval_runtime": 325.6515,
3780
+ "eval_samples_per_second": 88.426,
3781
+ "eval_steps_per_second": 2.764,
3782
+ "step": 282009
3783
+ },
3784
+ {
3785
+ "epoch": 39.07,
3786
+ "learning_rate": 4.740699764901121e-07,
3787
+ "loss": 0.1193,
3788
+ "step": 282500
3789
+ },
3790
+ {
3791
+ "epoch": 39.14,
3792
+ "learning_rate": 4.395657585396211e-07,
3793
+ "loss": 0.1177,
3794
+ "step": 283000
3795
+ },
3796
+ {
3797
+ "epoch": 39.21,
3798
+ "learning_rate": 4.0499239385977044e-07,
3799
+ "loss": 0.1142,
3800
+ "step": 283500
3801
+ },
3802
+ {
3803
+ "epoch": 39.28,
3804
+ "learning_rate": 3.704190291799198e-07,
3805
+ "loss": 0.1159,
3806
+ "step": 284000
3807
+ },
3808
+ {
3809
+ "epoch": 39.34,
3810
+ "learning_rate": 3.358456645000692e-07,
3811
+ "loss": 0.1151,
3812
+ "step": 284500
3813
+ },
3814
+ {
3815
+ "epoch": 39.41,
3816
+ "learning_rate": 3.0127229982021856e-07,
3817
+ "loss": 0.1155,
3818
+ "step": 285000
3819
+ },
3820
+ {
3821
+ "epoch": 39.48,
3822
+ "learning_rate": 2.666989351403679e-07,
3823
+ "loss": 0.1189,
3824
+ "step": 285500
3825
+ },
3826
+ {
3827
+ "epoch": 39.55,
3828
+ "learning_rate": 2.3212557046051725e-07,
3829
+ "loss": 0.1179,
3830
+ "step": 286000
3831
+ },
3832
+ {
3833
+ "epoch": 39.62,
3834
+ "learning_rate": 1.975522057806666e-07,
3835
+ "loss": 0.1169,
3836
+ "step": 286500
3837
+ },
3838
+ {
3839
+ "epoch": 39.69,
3840
+ "learning_rate": 1.6297884110081592e-07,
3841
+ "loss": 0.1164,
3842
+ "step": 287000
3843
+ },
3844
+ {
3845
+ "epoch": 39.76,
3846
+ "learning_rate": 1.284054764209653e-07,
3847
+ "loss": 0.1182,
3848
+ "step": 287500
3849
+ },
3850
+ {
3851
+ "epoch": 39.83,
3852
+ "learning_rate": 9.383211174111465e-08,
3853
+ "loss": 0.1179,
3854
+ "step": 288000
3855
+ },
3856
+ {
3857
+ "epoch": 39.9,
3858
+ "learning_rate": 5.925874706126401e-08,
3859
+ "loss": 0.1164,
3860
+ "step": 288500
3861
+ },
3862
+ {
3863
+ "epoch": 39.97,
3864
+ "learning_rate": 2.475452911077306e-08,
3865
+ "loss": 0.1199,
3866
+ "step": 289000
3867
+ },
3868
+ {
3869
+ "epoch": 40.0,
3870
+ "eval_bleu": 73.7743,
3871
+ "eval_gen_len": 13.0286,
3872
+ "eval_loss": 0.16223154962062836,
3873
+ "eval_runtime": 325.8877,
3874
+ "eval_samples_per_second": 88.362,
3875
+ "eval_steps_per_second": 2.762,
3876
+ "step": 289240
3877
+ }
3878
+ ],
3879
+ "max_steps": 289240,
3880
+ "num_train_epochs": 40,
3881
+ "total_flos": 6.189779025355162e+17,
3882
+ "trial_name": null,
3883
+ "trial_params": null
3884
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b7da800360637ae0d5a82d97070a77f7f00255283b65c2dd43fbe6feda5c9d16
3
+ size 3439