Hugging Face's logo

metricspace
/
DataPrivacyComplianceCheck-3B-V0.9

Text2Text Generation
Transformers
PyTorch
Safetensors
t5
Inference Endpoints
text-generation-inference
Model card Files Community
1
Wachu2005 commited on
Commit
b629bb2
•
1 Parent(s): cf511d7

Enable Git LFS

Browse files
Files changed (12) hide show
  1. config.json +61 -0
  2. generation_config.json +7 -0
  3. optimizer.pt +3 -0
  4. pytorch_model.bin +3 -0
  5. rng_state.pth +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 +112 -0
  11. trainer_state.json +3025 -0
  12. training_args.bin +3 -0
config.json ADDED
@@ -0,0 +1,61 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/media/metricspace/Data1/LLamaT5/TLAN-T5/checkpoint-4695",
3
+ "architectures": [
4
+ "T5ForConditionalGeneration"
5
+ ],
6
+ "d_ff": 5120,
7
+ "d_kv": 64,
8
+ "d_model": 2048,
9
+ "decoder_start_token_id": 0,
10
+ "dense_act_fn": "gelu_new",
11
+ "dropout_rate": 0.1,
12
+ "eos_token_id": 1,
13
+ "feed_forward_proj": "gated-gelu",
14
+ "initializer_factor": 1.0,
15
+ "is_encoder_decoder": true,
16
+ "is_gated_act": true,
17
+ "layer_norm_epsilon": 1e-06,
18
+ "model_type": "t5",
19
+ "n_positions": 512,
20
+ "num_decoder_layers": 24,
21
+ "num_heads": 32,
22
+ "num_layers": 24,
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
+ "tie_word_embeddings": false,
57
+ "torch_dtype": "bfloat16",
58
+ "transformers_version": "4.31.0",
59
+ "use_cache": true,
60
+ "vocab_size": 32128
61
+ }
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "decoder_start_token_id": 0,
4
+ "eos_token_id": 1,
5
+ "pad_token_id": 0,
6
+ "transformers_version": "4.31.0"
7
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b4e0d5d23b84bc57105163d651cdd28294bcff69de9aefe8356fa44fdd47d543
3
+ size 11399369081
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3541f5594b3b8fc6baac79ac5610fe6dc69dd437a77a4fd212d1123221ecb765
3
+ size 5699709261
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0fc06166a1576c0167039cc219a62a1036dee1e101c8e4c8e7cc6a0fd2506f8b
3
+ size 14575
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a7b4102175339dcc236ea501f6cb1a2a6702bdb8eca67eab578ba3d384a35a54
3
+ size 627
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 +1,112 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "clean_up_tokenization_spaces": true,
105
+ "eos_token": "</s>",
106
+ "extra_ids": 100,
107
+ "model_max_length": 512,
108
+ "pad_token": "<pad>",
109
+ "sp_model_kwargs": {},
110
+ "tokenizer_class": "T5Tokenizer",
111
+ "unk_token": "<unk>"
112
+ }
trainer_state.json ADDED
@@ -0,0 +1,3025 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.011560924351215363,
3
+ "best_model_checkpoint": "/media/metricspace/Data1/LLamaT5/TLAN-T5/checkpoint-4078",
4
+ "epoch": 14.964143426294822,
5
+ "global_step": 4695,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.03,
12
+ "learning_rate": 4.989350372736955e-05,
13
+ "loss": 39.1826,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.06,
18
+ "learning_rate": 4.9787007454739086e-05,
19
+ "loss": 29.1838,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.1,
24
+ "learning_rate": 4.968051118210863e-05,
25
+ "loss": 20.9138,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.13,
30
+ "learning_rate": 4.957401490947817e-05,
31
+ "loss": 8.3513,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.16,
36
+ "learning_rate": 4.9467518636847715e-05,
37
+ "loss": 4.4736,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.19,
42
+ "learning_rate": 4.936102236421725e-05,
43
+ "loss": 3.3849,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.22,
48
+ "learning_rate": 4.925452609158679e-05,
49
+ "loss": 2.2067,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.25,
54
+ "learning_rate": 4.9148029818956336e-05,
55
+ "loss": 0.5818,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.29,
60
+ "learning_rate": 4.904153354632588e-05,
61
+ "loss": 0.1404,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.32,
66
+ "learning_rate": 4.893503727369542e-05,
67
+ "loss": 0.0717,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.35,
72
+ "learning_rate": 4.8828541001064965e-05,
73
+ "loss": 0.0578,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.38,
78
+ "learning_rate": 4.872204472843451e-05,
79
+ "loss": 0.0405,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.41,
84
+ "learning_rate": 4.861554845580405e-05,
85
+ "loss": 0.0468,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.45,
90
+ "learning_rate": 4.8509052183173594e-05,
91
+ "loss": 0.0277,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.48,
96
+ "learning_rate": 4.840255591054313e-05,
97
+ "loss": 0.0327,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.51,
102
+ "learning_rate": 4.829605963791268e-05,
103
+ "loss": 0.0297,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.54,
108
+ "learning_rate": 4.818956336528222e-05,
109
+ "loss": 0.0242,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.57,
114
+ "learning_rate": 4.8083067092651754e-05,
115
+ "loss": 0.0453,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.61,
120
+ "learning_rate": 4.79765708200213e-05,
121
+ "loss": 0.0239,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.64,
126
+ "learning_rate": 4.7870074547390844e-05,
127
+ "loss": 0.0451,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.67,
132
+ "learning_rate": 4.776357827476038e-05,
133
+ "loss": 0.0186,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.7,
138
+ "learning_rate": 4.765708200212993e-05,
139
+ "loss": 0.0367,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.73,
144
+ "learning_rate": 4.755058572949947e-05,
145
+ "loss": 0.0152,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.76,
150
+ "learning_rate": 4.744408945686901e-05,
151
+ "loss": 0.0207,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.8,
156
+ "learning_rate": 4.7337593184238556e-05,
157
+ "loss": 0.016,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.83,
162
+ "learning_rate": 4.7231096911608095e-05,
163
+ "loss": 0.0235,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.86,
168
+ "learning_rate": 4.712460063897764e-05,
169
+ "loss": 0.0196,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.89,
174
+ "learning_rate": 4.7018104366347185e-05,
175
+ "loss": 0.0216,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.92,
180
+ "learning_rate": 4.691160809371672e-05,
181
+ "loss": 0.0278,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.96,
186
+ "learning_rate": 4.680511182108626e-05,
187
+ "loss": 0.0297,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.99,
192
+ "learning_rate": 4.669861554845581e-05,
193
+ "loss": 0.0215,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 1.0,
198
+ "eval_gen_len": 19.0,
199
+ "eval_loss": 0.015359184704720974,
200
+ "eval_rouge1": 31.9011,
201
+ "eval_rouge2": 22.9103,
202
+ "eval_rougeL": 31.716,
203
+ "eval_rougeLsum": 31.8299,
204
+ "eval_runtime": 717.4413,
205
+ "eval_samples_per_second": 0.389,
206
+ "eval_steps_per_second": 0.195,
207
+ "step": 313
208
+ },
209
+ {
210
+ "epoch": 1.02,
211
+ "learning_rate": 4.6592119275825345e-05,
212
+ "loss": 0.0171,
213
+ "step": 320
214
+ },
215
+ {
216
+ "epoch": 1.05,
217
+ "learning_rate": 4.648562300319489e-05,
218
+ "loss": 0.0242,
219
+ "step": 330
220
+ },
221
+ {
222
+ "epoch": 1.08,
223
+ "learning_rate": 4.637912673056443e-05,
224
+ "loss": 0.0216,
225
+ "step": 340
226
+ },
227
+ {
228
+ "epoch": 1.12,
229
+ "learning_rate": 4.6272630457933974e-05,
230
+ "loss": 0.0135,
231
+ "step": 350
232
+ },
233
+ {
234
+ "epoch": 1.15,
235
+ "learning_rate": 4.616613418530352e-05,
236
+ "loss": 0.0183,
237
+ "step": 360
238
+ },
239
+ {
240
+ "epoch": 1.18,
241
+ "learning_rate": 4.605963791267306e-05,
242
+ "loss": 0.0234,
243
+ "step": 370
244
+ },
245
+ {
246
+ "epoch": 1.21,
247
+ "learning_rate": 4.59531416400426e-05,
248
+ "loss": 0.03,
249
+ "step": 380
250
+ },
251
+ {
252
+ "epoch": 1.24,
253
+ "learning_rate": 4.584664536741215e-05,
254
+ "loss": 0.021,
255
+ "step": 390
256
+ },
257
+ {
258
+ "epoch": 1.27,
259
+ "learning_rate": 4.5740149094781686e-05,
260
+ "loss": 0.0212,
261
+ "step": 400
262
+ },
263
+ {
264
+ "epoch": 1.31,
265
+ "learning_rate": 4.563365282215123e-05,
266
+ "loss": 0.0121,
267
+ "step": 410
268
+ },
269
+ {
270
+ "epoch": 1.34,
271
+ "learning_rate": 4.552715654952077e-05,
272
+ "loss": 0.0148,
273
+ "step": 420
274
+ },
275
+ {
276
+ "epoch": 1.37,
277
+ "learning_rate": 4.542066027689031e-05,
278
+ "loss": 0.0155,
279
+ "step": 430
280
+ },
281
+ {
282
+ "epoch": 1.4,
283
+ "learning_rate": 4.531416400425985e-05,
284
+ "loss": 0.0213,
285
+ "step": 440
286
+ },
287
+ {
288
+ "epoch": 1.43,
289
+ "learning_rate": 4.520766773162939e-05,
290
+ "loss": 0.0193,
291
+ "step": 450
292
+ },
293
+ {
294
+ "epoch": 1.47,
295
+ "learning_rate": 4.5101171458998936e-05,
296
+ "loss": 0.0191,
297
+ "step": 460
298
+ },
299
+ {
300
+ "epoch": 1.5,
301
+ "learning_rate": 4.499467518636848e-05,
302
+ "loss": 0.0314,
303
+ "step": 470
304
+ },
305
+ {
306
+ "epoch": 1.53,
307
+ "learning_rate": 4.488817891373802e-05,
308
+ "loss": 0.0185,
309
+ "step": 480
310
+ },
311
+ {
312
+ "epoch": 1.56,
313
+ "learning_rate": 4.4781682641107565e-05,
314
+ "loss": 0.0152,
315
+ "step": 490
316
+ },
317
+ {
318
+ "epoch": 1.59,
319
+ "learning_rate": 4.46751863684771e-05,
320
+ "loss": 0.0361,
321
+ "step": 500
322
+ },
323
+ {
324
+ "epoch": 1.63,
325
+ "learning_rate": 4.456869009584665e-05,
326
+ "loss": 0.0189,
327
+ "step": 510
328
+ },
329
+ {
330
+ "epoch": 1.66,
331
+ "learning_rate": 4.4462193823216194e-05,
332
+ "loss": 0.0149,
333
+ "step": 520
334
+ },
335
+ {
336
+ "epoch": 1.69,
337
+ "learning_rate": 4.435569755058573e-05,
338
+ "loss": 0.0236,
339
+ "step": 530
340
+ },
341
+ {
342
+ "epoch": 1.72,
343
+ "learning_rate": 4.424920127795527e-05,
344
+ "loss": 0.0328,
345
+ "step": 540
346
+ },
347
+ {
348
+ "epoch": 1.75,
349
+ "learning_rate": 4.4142705005324815e-05,
350
+ "loss": 0.0182,
351
+ "step": 550
352
+ },
353
+ {
354
+ "epoch": 1.78,
355
+ "learning_rate": 4.4036208732694354e-05,
356
+ "loss": 0.015,
357
+ "step": 560
358
+ },
359
+ {
360
+ "epoch": 1.82,
361
+ "learning_rate": 4.39297124600639e-05,
362
+ "loss": 0.0123,
363
+ "step": 570
364
+ },
365
+ {
366
+ "epoch": 1.85,
367
+ "learning_rate": 4.3823216187433444e-05,
368
+ "loss": 0.0175,
369
+ "step": 580
370
+ },
371
+ {
372
+ "epoch": 1.88,
373
+ "learning_rate": 4.371671991480298e-05,
374
+ "loss": 0.0164,
375
+ "step": 590
376
+ },
377
+ {
378
+ "epoch": 1.91,
379
+ "learning_rate": 4.361022364217253e-05,
380
+ "loss": 0.0156,
381
+ "step": 600
382
+ },
383
+ {
384
+ "epoch": 1.94,
385
+ "learning_rate": 4.3503727369542066e-05,
386
+ "loss": 0.0109,
387
+ "step": 610
388
+ },
389
+ {
390
+ "epoch": 1.98,
391
+ "learning_rate": 4.339723109691161e-05,
392
+ "loss": 0.0186,
393
+ "step": 620
394
+ },
395
+ {
396
+ "epoch": 2.0,
397
+ "eval_gen_len": 17.795698924731184,
398
+ "eval_loss": 0.013456961140036583,
399
+ "eval_rouge1": 39.398,
400
+ "eval_rouge2": 31.4448,
401
+ "eval_rougeL": 39.0922,
402
+ "eval_rougeLsum": 39.237,
403
+ "eval_runtime": 716.8614,
404
+ "eval_samples_per_second": 0.389,
405
+ "eval_steps_per_second": 0.195,
406
+ "step": 627
407
+ },
408
+ {
409
+ "epoch": 2.01,
410
+ "learning_rate": 4.3290734824281156e-05,
411
+ "loss": 0.0119,
412
+ "step": 630
413
+ },
414
+ {
415
+ "epoch": 2.04,
416
+ "learning_rate": 4.3184238551650695e-05,
417
+ "loss": 0.0169,
418
+ "step": 640
419
+ },
420
+ {
421
+ "epoch": 2.07,
422
+ "learning_rate": 4.307774227902024e-05,
423
+ "loss": 0.0157,
424
+ "step": 650
425
+ },
426
+ {
427
+ "epoch": 2.1,
428
+ "learning_rate": 4.297124600638978e-05,
429
+ "loss": 0.0246,
430
+ "step": 660
431
+ },
432
+ {
433
+ "epoch": 2.14,
434
+ "learning_rate": 4.2864749733759316e-05,
435
+ "loss": 0.0389,
436
+ "step": 670
437
+ },
438
+ {
439
+ "epoch": 2.17,
440
+ "learning_rate": 4.275825346112886e-05,
441
+ "loss": 0.02,
442
+ "step": 680
443
+ },
444
+ {
445
+ "epoch": 2.2,
446
+ "learning_rate": 4.265175718849841e-05,
447
+ "loss": 0.0176,
448
+ "step": 690
449
+ },
450
+ {
451
+ "epoch": 2.23,
452
+ "learning_rate": 4.2545260915867945e-05,
453
+ "loss": 0.0147,
454
+ "step": 700
455
+ },
456
+ {
457
+ "epoch": 2.26,
458
+ "learning_rate": 4.243876464323749e-05,
459
+ "loss": 0.0153,
460
+ "step": 710
461
+ },
462
+ {
463
+ "epoch": 2.29,
464
+ "learning_rate": 4.233226837060703e-05,
465
+ "loss": 0.0134,
466
+ "step": 720
467
+ },
468
+ {
469
+ "epoch": 2.33,
470
+ "learning_rate": 4.2225772097976574e-05,
471
+ "loss": 0.0119,
472
+ "step": 730
473
+ },
474
+ {
475
+ "epoch": 2.36,
476
+ "learning_rate": 4.211927582534612e-05,
477
+ "loss": 0.0256,
478
+ "step": 740
479
+ },
480
+ {
481
+ "epoch": 2.39,
482
+ "learning_rate": 4.201277955271566e-05,
483
+ "loss": 0.0196,
484
+ "step": 750
485
+ },
486
+ {
487
+ "epoch": 2.42,
488
+ "learning_rate": 4.19062832800852e-05,
489
+ "loss": 0.0149,
490
+ "step": 760
491
+ },
492
+ {
493
+ "epoch": 2.45,
494
+ "learning_rate": 4.179978700745474e-05,
495
+ "loss": 0.0204,
496
+ "step": 770
497
+ },
498
+ {
499
+ "epoch": 2.49,
500
+ "learning_rate": 4.169329073482428e-05,
501
+ "loss": 0.0182,
502
+ "step": 780
503
+ },
504
+ {
505
+ "epoch": 2.52,
506
+ "learning_rate": 4.1586794462193824e-05,
507
+ "loss": 0.0142,
508
+ "step": 790
509
+ },
510
+ {
511
+ "epoch": 2.55,
512
+ "learning_rate": 4.148029818956336e-05,
513
+ "loss": 0.0095,
514
+ "step": 800
515
+ },
516
+ {
517
+ "epoch": 2.58,
518
+ "learning_rate": 4.137380191693291e-05,
519
+ "loss": 0.0108,
520
+ "step": 810
521
+ },
522
+ {
523
+ "epoch": 2.61,
524
+ "learning_rate": 4.126730564430245e-05,
525
+ "loss": 0.0123,
526
+ "step": 820
527
+ },
528
+ {
529
+ "epoch": 2.65,
530
+ "learning_rate": 4.116080937167199e-05,
531
+ "loss": 0.0226,
532
+ "step": 830
533
+ },
534
+ {
535
+ "epoch": 2.68,
536
+ "learning_rate": 4.1054313099041536e-05,
537
+ "loss": 0.0224,
538
+ "step": 840
539
+ },
540
+ {
541
+ "epoch": 2.71,
542
+ "learning_rate": 4.094781682641108e-05,
543
+ "loss": 0.0147,
544
+ "step": 850
545
+ },
546
+ {
547
+ "epoch": 2.74,
548
+ "learning_rate": 4.084132055378062e-05,
549
+ "loss": 0.0158,
550
+ "step": 860
551
+ },
552
+ {
553
+ "epoch": 2.77,
554
+ "learning_rate": 4.0734824281150165e-05,
555
+ "loss": 0.0118,
556
+ "step": 870
557
+ },
558
+ {
559
+ "epoch": 2.8,
560
+ "learning_rate": 4.06283280085197e-05,
561
+ "loss": 0.0272,
562
+ "step": 880
563
+ },
564
+ {
565
+ "epoch": 2.84,
566
+ "learning_rate": 4.052183173588925e-05,
567
+ "loss": 0.0148,
568
+ "step": 890
569
+ },
570
+ {
571
+ "epoch": 2.87,
572
+ "learning_rate": 4.041533546325879e-05,
573
+ "loss": 0.0314,
574
+ "step": 900
575
+ },
576
+ {
577
+ "epoch": 2.9,
578
+ "learning_rate": 4.0308839190628325e-05,
579
+ "loss": 0.016,
580
+ "step": 910
581
+ },
582
+ {
583
+ "epoch": 2.93,
584
+ "learning_rate": 4.020234291799787e-05,
585
+ "loss": 0.0146,
586
+ "step": 920
587
+ },
588
+ {
589
+ "epoch": 2.96,
590
+ "learning_rate": 4.0095846645367415e-05,
591
+ "loss": 0.0159,
592
+ "step": 930
593
+ },
594
+ {
595
+ "epoch": 3.0,
596
+ "learning_rate": 3.9989350372736954e-05,
597
+ "loss": 0.0166,
598
+ "step": 940
599
+ },
600
+ {
601
+ "epoch": 3.0,
602
+ "eval_gen_len": 16.892473118279568,
603
+ "eval_loss": 0.01282462291419506,
604
+ "eval_rouge1": 44.2185,
605
+ "eval_rouge2": 36.802,
606
+ "eval_rougeL": 44.0029,
607
+ "eval_rougeLsum": 44.0496,
608
+ "eval_runtime": 709.8988,
609
+ "eval_samples_per_second": 0.393,
610
+ "eval_steps_per_second": 0.197,
611
+ "step": 941
612
+ },
613
+ {
614
+ "epoch": 3.03,
615
+ "learning_rate": 3.98828541001065e-05,
616
+ "loss": 0.011,
617
+ "step": 950
618
+ },
619
+ {
620
+ "epoch": 3.06,
621
+ "learning_rate": 3.9776357827476044e-05,
622
+ "loss": 0.0136,
623
+ "step": 960
624
+ },
625
+ {
626
+ "epoch": 3.09,
627
+ "learning_rate": 3.966986155484558e-05,
628
+ "loss": 0.0165,
629
+ "step": 970
630
+ },
631
+ {
632
+ "epoch": 3.12,
633
+ "learning_rate": 3.956336528221513e-05,
634
+ "loss": 0.0132,
635
+ "step": 980
636
+ },
637
+ {
638
+ "epoch": 3.16,
639
+ "learning_rate": 3.9456869009584666e-05,
640
+ "loss": 0.0184,
641
+ "step": 990
642
+ },
643
+ {
644
+ "epoch": 3.19,
645
+ "learning_rate": 3.935037273695421e-05,
646
+ "loss": 0.0303,
647
+ "step": 1000
648
+ },
649
+ {
650
+ "epoch": 3.22,
651
+ "learning_rate": 3.9243876464323756e-05,
652
+ "loss": 0.0202,
653
+ "step": 1010
654
+ },
655
+ {
656
+ "epoch": 3.25,
657
+ "learning_rate": 3.913738019169329e-05,
658
+ "loss": 0.0159,
659
+ "step": 1020
660
+ },
661
+ {
662
+ "epoch": 3.28,
663
+ "learning_rate": 3.903088391906283e-05,
664
+ "loss": 0.0246,
665
+ "step": 1030
666
+ },
667
+ {
668
+ "epoch": 3.31,
669
+ "learning_rate": 3.892438764643238e-05,
670
+ "loss": 0.0108,
671
+ "step": 1040
672
+ },
673
+ {
674
+ "epoch": 3.35,
675
+ "learning_rate": 3.8817891373801916e-05,
676
+ "loss": 0.0167,
677
+ "step": 1050
678
+ },
679
+ {
680
+ "epoch": 3.38,
681
+ "learning_rate": 3.871139510117146e-05,
682
+ "loss": 0.0134,
683
+ "step": 1060
684
+ },
685
+ {
686
+ "epoch": 3.41,
687
+ "learning_rate": 3.8604898828541e-05,
688
+ "loss": 0.0122,
689
+ "step": 1070
690
+ },
691
+ {
692
+ "epoch": 3.44,
693
+ "learning_rate": 3.8498402555910545e-05,
694
+ "loss": 0.0101,
695
+ "step": 1080
696
+ },
697
+ {
698
+ "epoch": 3.47,
699
+ "learning_rate": 3.839190628328009e-05,
700
+ "loss": 0.0118,
701
+ "step": 1090
702
+ },
703
+ {
704
+ "epoch": 3.51,
705
+ "learning_rate": 3.828541001064963e-05,
706
+ "loss": 0.0139,
707
+ "step": 1100
708
+ },
709
+ {
710
+ "epoch": 3.54,
711
+ "learning_rate": 3.8178913738019174e-05,
712
+ "loss": 0.0175,
713
+ "step": 1110
714
+ },
715
+ {
716
+ "epoch": 3.57,
717
+ "learning_rate": 3.807241746538872e-05,
718
+ "loss": 0.0156,
719
+ "step": 1120
720
+ },
721
+ {
722
+ "epoch": 3.6,
723
+ "learning_rate": 3.796592119275825e-05,
724
+ "loss": 0.0235,
725
+ "step": 1130
726
+ },
727
+ {
728
+ "epoch": 3.63,
729
+ "learning_rate": 3.7859424920127795e-05,
730
+ "loss": 0.0124,
731
+ "step": 1140
732
+ },
733
+ {
734
+ "epoch": 3.67,
735
+ "learning_rate": 3.775292864749734e-05,
736
+ "loss": 0.0109,
737
+ "step": 1150
738
+ },
739
+ {
740
+ "epoch": 3.7,
741
+ "learning_rate": 3.764643237486688e-05,
742
+ "loss": 0.0189,
743
+ "step": 1160
744
+ },
745
+ {
746
+ "epoch": 3.73,
747
+ "learning_rate": 3.7539936102236424e-05,
748
+ "loss": 0.0134,
749
+ "step": 1170
750
+ },
751
+ {
752
+ "epoch": 3.76,
753
+ "learning_rate": 3.743343982960596e-05,
754
+ "loss": 0.0314,
755
+ "step": 1180
756
+ },
757
+ {
758
+ "epoch": 3.79,
759
+ "learning_rate": 3.732694355697551e-05,
760
+ "loss": 0.0377,
761
+ "step": 1190
762
+ },
763
+ {
764
+ "epoch": 3.82,
765
+ "learning_rate": 3.722044728434505e-05,
766
+ "loss": 0.015,
767
+ "step": 1200
768
+ },
769
+ {
770
+ "epoch": 3.86,
771
+ "learning_rate": 3.711395101171459e-05,
772
+ "loss": 0.0129,
773
+ "step": 1210
774
+ },
775
+ {
776
+ "epoch": 3.89,
777
+ "learning_rate": 3.7007454739084136e-05,
778
+ "loss": 0.0154,
779
+ "step": 1220
780
+ },
781
+ {
782
+ "epoch": 3.92,
783
+ "learning_rate": 3.6900958466453675e-05,
784
+ "loss": 0.0117,
785
+ "step": 1230
786
+ },
787
+ {
788
+ "epoch": 3.95,
789
+ "learning_rate": 3.679446219382322e-05,
790
+ "loss": 0.0125,
791
+ "step": 1240
792
+ },
793
+ {
794
+ "epoch": 3.98,
795
+ "learning_rate": 3.668796592119276e-05,
796
+ "loss": 0.0103,
797
+ "step": 1250
798
+ },
799
+ {
800
+ "epoch": 4.0,
801
+ "eval_gen_len": 14.734767025089607,
802
+ "eval_loss": 0.012266965582966805,
803
+ "eval_rouge1": 54.8551,
804
+ "eval_rouge2": 49.2036,
805
+ "eval_rougeL": 54.5865,
806
+ "eval_rougeLsum": 54.8397,
807
+ "eval_runtime": 674.9391,
808
+ "eval_samples_per_second": 0.413,
809
+ "eval_steps_per_second": 0.207,
810
+ "step": 1255
811
+ },
812
+ {
813
+ "epoch": 4.02,
814
+ "learning_rate": 3.65814696485623e-05,
815
+ "loss": 0.0177,
816
+ "step": 1260
817
+ },
818
+ {
819
+ "epoch": 4.05,
820
+ "learning_rate": 3.647497337593184e-05,
821
+ "loss": 0.0189,
822
+ "step": 1270
823
+ },
824
+ {
825
+ "epoch": 4.08,
826
+ "learning_rate": 3.636847710330139e-05,
827
+ "loss": 0.0113,
828
+ "step": 1280
829
+ },
830
+ {
831
+ "epoch": 4.11,
832
+ "learning_rate": 3.6261980830670925e-05,
833
+ "loss": 0.0277,
834
+ "step": 1290
835
+ },
836
+ {
837
+ "epoch": 4.14,
838
+ "learning_rate": 3.615548455804047e-05,
839
+ "loss": 0.0236,
840
+ "step": 1300
841
+ },
842
+ {
843
+ "epoch": 4.18,
844
+ "learning_rate": 3.6048988285410015e-05,
845
+ "loss": 0.0198,
846
+ "step": 1310
847
+ },
848
+ {
849
+ "epoch": 4.21,
850
+ "learning_rate": 3.5942492012779554e-05,
851
+ "loss": 0.0155,
852
+ "step": 1320
853
+ },
854
+ {
855
+ "epoch": 4.24,
856
+ "learning_rate": 3.58359957401491e-05,
857
+ "loss": 0.0138,
858
+ "step": 1330
859
+ },
860
+ {
861
+ "epoch": 4.27,
862
+ "learning_rate": 3.572949946751864e-05,
863
+ "loss": 0.0154,
864
+ "step": 1340
865
+ },
866
+ {
867
+ "epoch": 4.3,
868
+ "learning_rate": 3.562300319488818e-05,
869
+ "loss": 0.0106,
870
+ "step": 1350
871
+ },
872
+ {
873
+ "epoch": 4.33,
874
+ "learning_rate": 3.551650692225773e-05,
875
+ "loss": 0.0168,
876
+ "step": 1360
877
+ },
878
+ {
879
+ "epoch": 4.37,
880
+ "learning_rate": 3.541001064962726e-05,
881
+ "loss": 0.0114,
882
+ "step": 1370
883
+ },
884
+ {
885
+ "epoch": 4.4,
886
+ "learning_rate": 3.5303514376996804e-05,
887
+ "loss": 0.0119,
888
+ "step": 1380
889
+ },
890
+ {
891
+ "epoch": 4.43,
892
+ "learning_rate": 3.519701810436635e-05,
893
+ "loss": 0.0168,
894
+ "step": 1390
895
+ },
896
+ {
897
+ "epoch": 4.46,
898
+ "learning_rate": 3.509052183173589e-05,
899
+ "loss": 0.012,
900
+ "step": 1400
901
+ },
902
+ {
903
+ "epoch": 4.49,
904
+ "learning_rate": 3.498402555910543e-05,
905
+ "loss": 0.011,
906
+ "step": 1410
907
+ },
908
+ {
909
+ "epoch": 4.53,
910
+ "learning_rate": 3.487752928647498e-05,
911
+ "loss": 0.014,
912
+ "step": 1420
913
+ },
914
+ {
915
+ "epoch": 4.56,
916
+ "learning_rate": 3.4771033013844516e-05,
917
+ "loss": 0.0151,
918
+ "step": 1430
919
+ },
920
+ {
921
+ "epoch": 4.59,
922
+ "learning_rate": 3.466453674121406e-05,
923
+ "loss": 0.0148,
924
+ "step": 1440
925
+ },
926
+ {
927
+ "epoch": 4.62,
928
+ "learning_rate": 3.45580404685836e-05,
929
+ "loss": 0.0142,
930
+ "step": 1450
931
+ },
932
+ {
933
+ "epoch": 4.65,
934
+ "learning_rate": 3.4451544195953145e-05,
935
+ "loss": 0.0183,
936
+ "step": 1460
937
+ },
938
+ {
939
+ "epoch": 4.69,
940
+ "learning_rate": 3.434504792332269e-05,
941
+ "loss": 0.0249,
942
+ "step": 1470
943
+ },
944
+ {
945
+ "epoch": 4.72,
946
+ "learning_rate": 3.423855165069223e-05,
947
+ "loss": 0.0147,
948
+ "step": 1480
949
+ },
950
+ {
951
+ "epoch": 4.75,
952
+ "learning_rate": 3.413205537806177e-05,
953
+ "loss": 0.0205,
954
+ "step": 1490
955
+ },
956
+ {
957
+ "epoch": 4.78,
958
+ "learning_rate": 3.402555910543131e-05,
959
+ "loss": 0.0298,
960
+ "step": 1500
961
+ },
962
+ {
963
+ "epoch": 4.81,
964
+ "learning_rate": 3.391906283280085e-05,
965
+ "loss": 0.009,
966
+ "step": 1510
967
+ },
968
+ {
969
+ "epoch": 4.84,
970
+ "learning_rate": 3.3812566560170395e-05,
971
+ "loss": 0.0167,
972
+ "step": 1520
973
+ },
974
+ {
975
+ "epoch": 4.88,
976
+ "learning_rate": 3.3706070287539934e-05,
977
+ "loss": 0.0108,
978
+ "step": 1530
979
+ },
980
+ {
981
+ "epoch": 4.91,
982
+ "learning_rate": 3.359957401490948e-05,
983
+ "loss": 0.0126,
984
+ "step": 1540
985
+ },
986
+ {
987
+ "epoch": 4.94,
988
+ "learning_rate": 3.3493077742279024e-05,
989
+ "loss": 0.0201,
990
+ "step": 1550
991
+ },
992
+ {
993
+ "epoch": 4.97,
994
+ "learning_rate": 3.338658146964856e-05,
995
+ "loss": 0.0132,
996
+ "step": 1560
997
+ },
998
+ {
999
+ "epoch": 5.0,
1000
+ "eval_gen_len": 13.17921146953405,
1001
+ "eval_loss": 0.012017679400742054,
1002
+ "eval_rouge1": 61.0563,
1003
+ "eval_rouge2": 56.3505,
1004
+ "eval_rougeL": 60.775,
1005
+ "eval_rougeLsum": 60.928,
1006
+ "eval_runtime": 625.9017,
1007
+ "eval_samples_per_second": 0.446,
1008
+ "eval_steps_per_second": 0.224,
1009
+ "step": 1568
1010
+ },
1011
+ {
1012
+ "epoch": 5.0,
1013
+ "learning_rate": 3.328008519701811e-05,
1014
+ "loss": 0.0088,
1015
+ "step": 1570
1016
+ },
1017
+ {
1018
+ "epoch": 5.04,
1019
+ "learning_rate": 3.317358892438765e-05,
1020
+ "loss": 0.012,
1021
+ "step": 1580
1022
+ },
1023
+ {
1024
+ "epoch": 5.07,
1025
+ "learning_rate": 3.306709265175719e-05,
1026
+ "loss": 0.0094,
1027
+ "step": 1590
1028
+ },
1029
+ {
1030
+ "epoch": 5.1,
1031
+ "learning_rate": 3.2960596379126736e-05,
1032
+ "loss": 0.0263,
1033
+ "step": 1600
1034
+ },
1035
+ {
1036
+ "epoch": 5.13,
1037
+ "learning_rate": 3.2854100106496274e-05,
1038
+ "loss": 0.0139,
1039
+ "step": 1610
1040
+ },
1041
+ {
1042
+ "epoch": 5.16,
1043
+ "learning_rate": 3.274760383386581e-05,
1044
+ "loss": 0.0111,
1045
+ "step": 1620
1046
+ },
1047
+ {
1048
+ "epoch": 5.2,
1049
+ "learning_rate": 3.264110756123536e-05,
1050
+ "loss": 0.0099,
1051
+ "step": 1630
1052
+ },
1053
+ {
1054
+ "epoch": 5.23,
1055
+ "learning_rate": 3.2534611288604896e-05,
1056
+ "loss": 0.0284,
1057
+ "step": 1640
1058
+ },
1059
+ {
1060
+ "epoch": 5.26,
1061
+ "learning_rate": 3.242811501597444e-05,
1062
+ "loss": 0.023,
1063
+ "step": 1650
1064
+ },
1065
+ {
1066
+ "epoch": 5.29,
1067
+ "learning_rate": 3.2321618743343987e-05,
1068
+ "loss": 0.0077,
1069
+ "step": 1660
1070
+ },
1071
+ {
1072
+ "epoch": 5.32,
1073
+ "learning_rate": 3.2215122470713525e-05,
1074
+ "loss": 0.0167,
1075
+ "step": 1670
1076
+ },
1077
+ {
1078
+ "epoch": 5.35,
1079
+ "learning_rate": 3.210862619808307e-05,
1080
+ "loss": 0.0096,
1081
+ "step": 1680
1082
+ },
1083
+ {
1084
+ "epoch": 5.39,
1085
+ "learning_rate": 3.2002129925452615e-05,
1086
+ "loss": 0.0197,
1087
+ "step": 1690
1088
+ },
1089
+ {
1090
+ "epoch": 5.42,
1091
+ "learning_rate": 3.1895633652822154e-05,
1092
+ "loss": 0.0105,
1093
+ "step": 1700
1094
+ },
1095
+ {
1096
+ "epoch": 5.45,
1097
+ "learning_rate": 3.17891373801917e-05,
1098
+ "loss": 0.0308,
1099
+ "step": 1710
1100
+ },
1101
+ {
1102
+ "epoch": 5.48,
1103
+ "learning_rate": 3.168264110756124e-05,
1104
+ "loss": 0.0118,
1105
+ "step": 1720
1106
+ },
1107
+ {
1108
+ "epoch": 5.51,
1109
+ "learning_rate": 3.1576144834930775e-05,
1110
+ "loss": 0.0128,
1111
+ "step": 1730
1112
+ },
1113
+ {
1114
+ "epoch": 5.55,
1115
+ "learning_rate": 3.146964856230032e-05,
1116
+ "loss": 0.0142,
1117
+ "step": 1740
1118
+ },
1119
+ {
1120
+ "epoch": 5.58,
1121
+ "learning_rate": 3.136315228966986e-05,
1122
+ "loss": 0.0086,
1123
+ "step": 1750
1124
+ },
1125
+ {
1126
+ "epoch": 5.61,
1127
+ "learning_rate": 3.1256656017039404e-05,
1128
+ "loss": 0.0189,
1129
+ "step": 1760
1130
+ },
1131
+ {
1132
+ "epoch": 5.64,
1133
+ "learning_rate": 3.115015974440895e-05,
1134
+ "loss": 0.0137,
1135
+ "step": 1770
1136
+ },
1137
+ {
1138
+ "epoch": 5.67,
1139
+ "learning_rate": 3.104366347177849e-05,
1140
+ "loss": 0.0117,
1141
+ "step": 1780
1142
+ },
1143
+ {
1144
+ "epoch": 5.71,
1145
+ "learning_rate": 3.093716719914803e-05,
1146
+ "loss": 0.0112,
1147
+ "step": 1790
1148
+ },
1149
+ {
1150
+ "epoch": 5.74,
1151
+ "learning_rate": 3.083067092651757e-05,
1152
+ "loss": 0.017,
1153
+ "step": 1800
1154
+ },
1155
+ {
1156
+ "epoch": 5.77,
1157
+ "learning_rate": 3.0724174653887116e-05,
1158
+ "loss": 0.0089,
1159
+ "step": 1810
1160
+ },
1161
+ {
1162
+ "epoch": 5.8,
1163
+ "learning_rate": 3.061767838125666e-05,
1164
+ "loss": 0.0354,
1165
+ "step": 1820
1166
+ },
1167
+ {
1168
+ "epoch": 5.83,
1169
+ "learning_rate": 3.0511182108626203e-05,
1170
+ "loss": 0.0173,
1171
+ "step": 1830
1172
+ },
1173
+ {
1174
+ "epoch": 5.86,
1175
+ "learning_rate": 3.0404685835995745e-05,
1176
+ "loss": 0.0132,
1177
+ "step": 1840
1178
+ },
1179
+ {
1180
+ "epoch": 5.9,
1181
+ "learning_rate": 3.029818956336528e-05,
1182
+ "loss": 0.0101,
1183
+ "step": 1850
1184
+ },
1185
+ {
1186
+ "epoch": 5.93,
1187
+ "learning_rate": 3.0191693290734825e-05,
1188
+ "loss": 0.0172,
1189
+ "step": 1860
1190
+ },
1191
+ {
1192
+ "epoch": 5.96,
1193
+ "learning_rate": 3.0085197018104367e-05,
1194
+ "loss": 0.0109,
1195
+ "step": 1870
1196
+ },
1197
+ {
1198
+ "epoch": 5.99,
1199
+ "learning_rate": 2.997870074547391e-05,
1200
+ "loss": 0.0179,
1201
+ "step": 1880
1202
+ },
1203
+ {
1204
+ "epoch": 6.0,
1205
+ "eval_gen_len": 12.727598566308243,
1206
+ "eval_loss": 0.011885586194694042,
1207
+ "eval_rouge1": 63.4215,
1208
+ "eval_rouge2": 59.2435,
1209
+ "eval_rougeL": 63.1198,
1210
+ "eval_rougeLsum": 63.3059,
1211
+ "eval_runtime": 611.8658,
1212
+ "eval_samples_per_second": 0.456,
1213
+ "eval_steps_per_second": 0.229,
1214
+ "step": 1882
1215
+ },
1216
+ {
1217
+ "epoch": 6.02,
1218
+ "learning_rate": 2.987220447284345e-05,
1219
+ "loss": 0.0157,
1220
+ "step": 1890
1221
+ },
1222
+ {
1223
+ "epoch": 6.06,
1224
+ "learning_rate": 2.9765708200212995e-05,
1225
+ "loss": 0.0173,
1226
+ "step": 1900
1227
+ },
1228
+ {
1229
+ "epoch": 6.09,
1230
+ "learning_rate": 2.9659211927582537e-05,
1231
+ "loss": 0.0132,
1232
+ "step": 1910
1233
+ },
1234
+ {
1235
+ "epoch": 6.12,
1236
+ "learning_rate": 2.955271565495208e-05,
1237
+ "loss": 0.0253,
1238
+ "step": 1920
1239
+ },
1240
+ {
1241
+ "epoch": 6.15,
1242
+ "learning_rate": 2.944621938232162e-05,
1243
+ "loss": 0.0158,
1244
+ "step": 1930
1245
+ },
1246
+ {
1247
+ "epoch": 6.18,
1248
+ "learning_rate": 2.9339723109691166e-05,
1249
+ "loss": 0.0112,
1250
+ "step": 1940
1251
+ },
1252
+ {
1253
+ "epoch": 6.22,
1254
+ "learning_rate": 2.9233226837060707e-05,
1255
+ "loss": 0.0097,
1256
+ "step": 1950
1257
+ },
1258
+ {
1259
+ "epoch": 6.25,
1260
+ "learning_rate": 2.912673056443025e-05,
1261
+ "loss": 0.0101,
1262
+ "step": 1960
1263
+ },
1264
+ {
1265
+ "epoch": 6.28,
1266
+ "learning_rate": 2.9020234291799788e-05,
1267
+ "loss": 0.0197,
1268
+ "step": 1970
1269
+ },
1270
+ {
1271
+ "epoch": 6.31,
1272
+ "learning_rate": 2.891373801916933e-05,
1273
+ "loss": 0.0138,
1274
+ "step": 1980
1275
+ },
1276
+ {
1277
+ "epoch": 6.34,
1278
+ "learning_rate": 2.880724174653887e-05,
1279
+ "loss": 0.0123,
1280
+ "step": 1990
1281
+ },
1282
+ {
1283
+ "epoch": 6.37,
1284
+ "learning_rate": 2.8700745473908413e-05,
1285
+ "loss": 0.0187,
1286
+ "step": 2000
1287
+ },
1288
+ {
1289
+ "epoch": 6.41,
1290
+ "learning_rate": 2.8594249201277955e-05,
1291
+ "loss": 0.0164,
1292
+ "step": 2010
1293
+ },
1294
+ {
1295
+ "epoch": 6.44,
1296
+ "learning_rate": 2.84877529286475e-05,
1297
+ "loss": 0.0097,
1298
+ "step": 2020
1299
+ },
1300
+ {
1301
+ "epoch": 6.47,
1302
+ "learning_rate": 2.838125665601704e-05,
1303
+ "loss": 0.017,
1304
+ "step": 2030
1305
+ },
1306
+ {
1307
+ "epoch": 6.5,
1308
+ "learning_rate": 2.8274760383386583e-05,
1309
+ "loss": 0.0072,
1310
+ "step": 2040
1311
+ },
1312
+ {
1313
+ "epoch": 6.53,
1314
+ "learning_rate": 2.8168264110756125e-05,
1315
+ "loss": 0.0173,
1316
+ "step": 2050
1317
+ },
1318
+ {
1319
+ "epoch": 6.57,
1320
+ "learning_rate": 2.806176783812567e-05,
1321
+ "loss": 0.0196,
1322
+ "step": 2060
1323
+ },
1324
+ {
1325
+ "epoch": 6.6,
1326
+ "learning_rate": 2.7955271565495212e-05,
1327
+ "loss": 0.0255,
1328
+ "step": 2070
1329
+ },
1330
+ {
1331
+ "epoch": 6.63,
1332
+ "learning_rate": 2.7848775292864754e-05,
1333
+ "loss": 0.019,
1334
+ "step": 2080
1335
+ },
1336
+ {
1337
+ "epoch": 6.66,
1338
+ "learning_rate": 2.7742279020234292e-05,
1339
+ "loss": 0.0103,
1340
+ "step": 2090
1341
+ },
1342
+ {
1343
+ "epoch": 6.69,
1344
+ "learning_rate": 2.7635782747603834e-05,
1345
+ "loss": 0.0173,
1346
+ "step": 2100
1347
+ },
1348
+ {
1349
+ "epoch": 6.73,
1350
+ "learning_rate": 2.7529286474973375e-05,
1351
+ "loss": 0.0113,
1352
+ "step": 2110
1353
+ },
1354
+ {
1355
+ "epoch": 6.76,
1356
+ "learning_rate": 2.7422790202342917e-05,
1357
+ "loss": 0.0189,
1358
+ "step": 2120
1359
+ },
1360
+ {
1361
+ "epoch": 6.79,
1362
+ "learning_rate": 2.7316293929712462e-05,
1363
+ "loss": 0.0149,
1364
+ "step": 2130
1365
+ },
1366
+ {
1367
+ "epoch": 6.82,
1368
+ "learning_rate": 2.7209797657082004e-05,
1369
+ "loss": 0.0128,
1370
+ "step": 2140
1371
+ },
1372
+ {
1373
+ "epoch": 6.85,
1374
+ "learning_rate": 2.7103301384451546e-05,
1375
+ "loss": 0.0181,
1376
+ "step": 2150
1377
+ },
1378
+ {
1379
+ "epoch": 6.88,
1380
+ "learning_rate": 2.6996805111821088e-05,
1381
+ "loss": 0.0097,
1382
+ "step": 2160
1383
+ },
1384
+ {
1385
+ "epoch": 6.92,
1386
+ "learning_rate": 2.6890308839190633e-05,
1387
+ "loss": 0.013,
1388
+ "step": 2170
1389
+ },
1390
+ {
1391
+ "epoch": 6.95,
1392
+ "learning_rate": 2.6783812566560174e-05,
1393
+ "loss": 0.014,
1394
+ "step": 2180
1395
+ },
1396
+ {
1397
+ "epoch": 6.98,
1398
+ "learning_rate": 2.6677316293929716e-05,
1399
+ "loss": 0.0115,
1400
+ "step": 2190
1401
+ },
1402
+ {
1403
+ "epoch": 7.0,
1404
+ "eval_gen_len": 12.92831541218638,
1405
+ "eval_loss": 0.011721653863787651,
1406
+ "eval_rouge1": 62.5874,
1407
+ "eval_rouge2": 58.2782,
1408
+ "eval_rougeL": 62.2116,
1409
+ "eval_rougeLsum": 62.3846,
1410
+ "eval_runtime": 618.8742,
1411
+ "eval_samples_per_second": 0.451,
1412
+ "eval_steps_per_second": 0.226,
1413
+ "step": 2196
1414
+ },
1415
+ {
1416
+ "epoch": 7.01,
1417
+ "learning_rate": 2.6570820021299255e-05,
1418
+ "loss": 0.0124,
1419
+ "step": 2200
1420
+ },
1421
+ {
1422
+ "epoch": 7.04,
1423
+ "learning_rate": 2.6464323748668796e-05,
1424
+ "loss": 0.0166,
1425
+ "step": 2210
1426
+ },
1427
+ {
1428
+ "epoch": 7.08,
1429
+ "learning_rate": 2.6357827476038338e-05,
1430
+ "loss": 0.0099,
1431
+ "step": 2220
1432
+ },
1433
+ {
1434
+ "epoch": 7.11,
1435
+ "learning_rate": 2.625133120340788e-05,
1436
+ "loss": 0.0113,
1437
+ "step": 2230
1438
+ },
1439
+ {
1440
+ "epoch": 7.14,
1441
+ "learning_rate": 2.6144834930777425e-05,
1442
+ "loss": 0.0174,
1443
+ "step": 2240
1444
+ },
1445
+ {
1446
+ "epoch": 7.17,
1447
+ "learning_rate": 2.6038338658146967e-05,
1448
+ "loss": 0.0095,
1449
+ "step": 2250
1450
+ },
1451
+ {
1452
+ "epoch": 7.2,
1453
+ "learning_rate": 2.593184238551651e-05,
1454
+ "loss": 0.0068,
1455
+ "step": 2260
1456
+ },
1457
+ {
1458
+ "epoch": 7.24,
1459
+ "learning_rate": 2.582534611288605e-05,
1460
+ "loss": 0.0291,
1461
+ "step": 2270
1462
+ },
1463
+ {
1464
+ "epoch": 7.27,
1465
+ "learning_rate": 2.5718849840255592e-05,
1466
+ "loss": 0.0171,
1467
+ "step": 2280
1468
+ },
1469
+ {
1470
+ "epoch": 7.3,
1471
+ "learning_rate": 2.5612353567625137e-05,
1472
+ "loss": 0.0183,
1473
+ "step": 2290
1474
+ },
1475
+ {
1476
+ "epoch": 7.33,
1477
+ "learning_rate": 2.550585729499468e-05,
1478
+ "loss": 0.0105,
1479
+ "step": 2300
1480
+ },
1481
+ {
1482
+ "epoch": 7.36,
1483
+ "learning_rate": 2.539936102236422e-05,
1484
+ "loss": 0.0227,
1485
+ "step": 2310
1486
+ },
1487
+ {
1488
+ "epoch": 7.39,
1489
+ "learning_rate": 2.529286474973376e-05,
1490
+ "loss": 0.0311,
1491
+ "step": 2320
1492
+ },
1493
+ {
1494
+ "epoch": 7.43,
1495
+ "learning_rate": 2.51863684771033e-05,
1496
+ "loss": 0.0133,
1497
+ "step": 2330
1498
+ },
1499
+ {
1500
+ "epoch": 7.46,
1501
+ "learning_rate": 2.5079872204472842e-05,
1502
+ "loss": 0.0102,
1503
+ "step": 2340
1504
+ },
1505
+ {
1506
+ "epoch": 7.49,
1507
+ "learning_rate": 2.4973375931842384e-05,
1508
+ "loss": 0.0189,
1509
+ "step": 2350
1510
+ },
1511
+ {
1512
+ "epoch": 7.52,
1513
+ "learning_rate": 2.486687965921193e-05,
1514
+ "loss": 0.0157,
1515
+ "step": 2360
1516
+ },
1517
+ {
1518
+ "epoch": 7.55,
1519
+ "learning_rate": 2.476038338658147e-05,
1520
+ "loss": 0.0114,
1521
+ "step": 2370
1522
+ },
1523
+ {
1524
+ "epoch": 7.59,
1525
+ "learning_rate": 2.4653887113951013e-05,
1526
+ "loss": 0.0131,
1527
+ "step": 2380
1528
+ },
1529
+ {
1530
+ "epoch": 7.62,
1531
+ "learning_rate": 2.4547390841320554e-05,
1532
+ "loss": 0.0144,
1533
+ "step": 2390
1534
+ },
1535
+ {
1536
+ "epoch": 7.65,
1537
+ "learning_rate": 2.44408945686901e-05,
1538
+ "loss": 0.0107,
1539
+ "step": 2400
1540
+ },
1541
+ {
1542
+ "epoch": 7.68,
1543
+ "learning_rate": 2.4334398296059638e-05,
1544
+ "loss": 0.0089,
1545
+ "step": 2410
1546
+ },
1547
+ {
1548
+ "epoch": 7.71,
1549
+ "learning_rate": 2.422790202342918e-05,
1550
+ "loss": 0.0119,
1551
+ "step": 2420
1552
+ },
1553
+ {
1554
+ "epoch": 7.75,
1555
+ "learning_rate": 2.412140575079872e-05,
1556
+ "loss": 0.0136,
1557
+ "step": 2430
1558
+ },
1559
+ {
1560
+ "epoch": 7.78,
1561
+ "learning_rate": 2.4014909478168267e-05,
1562
+ "loss": 0.026,
1563
+ "step": 2440
1564
+ },
1565
+ {
1566
+ "epoch": 7.81,
1567
+ "learning_rate": 2.390841320553781e-05,
1568
+ "loss": 0.0147,
1569
+ "step": 2450
1570
+ },
1571
+ {
1572
+ "epoch": 7.84,
1573
+ "learning_rate": 2.380191693290735e-05,
1574
+ "loss": 0.0114,
1575
+ "step": 2460
1576
+ },
1577
+ {
1578
+ "epoch": 7.87,
1579
+ "learning_rate": 2.3695420660276892e-05,
1580
+ "loss": 0.0122,
1581
+ "step": 2470
1582
+ },
1583
+ {
1584
+ "epoch": 7.9,
1585
+ "learning_rate": 2.3588924387646434e-05,
1586
+ "loss": 0.0115,
1587
+ "step": 2480
1588
+ },
1589
+ {
1590
+ "epoch": 7.94,
1591
+ "learning_rate": 2.3482428115015975e-05,
1592
+ "loss": 0.016,
1593
+ "step": 2490
1594
+ },
1595
+ {
1596
+ "epoch": 7.97,
1597
+ "learning_rate": 2.3375931842385517e-05,
1598
+ "loss": 0.0177,
1599
+ "step": 2500
1600
+ },
1601
+ {
1602
+ "epoch": 8.0,
1603
+ "learning_rate": 2.326943556975506e-05,
1604
+ "loss": 0.0097,
1605
+ "step": 2510
1606
+ },
1607
+ {
1608
+ "epoch": 8.0,
1609
+ "eval_gen_len": 12.827956989247312,
1610
+ "eval_loss": 0.0116655882447958,
1611
+ "eval_rouge1": 63.06,
1612
+ "eval_rouge2": 58.8818,
1613
+ "eval_rougeL": 62.7484,
1614
+ "eval_rougeLsum": 62.8736,
1615
+ "eval_runtime": 618.883,
1616
+ "eval_samples_per_second": 0.451,
1617
+ "eval_steps_per_second": 0.226,
1618
+ "step": 2510
1619
+ },
1620
+ {
1621
+ "epoch": 8.03,
1622
+ "learning_rate": 2.3162939297124604e-05,
1623
+ "loss": 0.0122,
1624
+ "step": 2520
1625
+ },
1626
+ {
1627
+ "epoch": 8.06,
1628
+ "learning_rate": 2.3056443024494142e-05,
1629
+ "loss": 0.0125,
1630
+ "step": 2530
1631
+ },
1632
+ {
1633
+ "epoch": 8.1,
1634
+ "learning_rate": 2.2949946751863684e-05,
1635
+ "loss": 0.013,
1636
+ "step": 2540
1637
+ },
1638
+ {
1639
+ "epoch": 8.13,
1640
+ "learning_rate": 2.284345047923323e-05,
1641
+ "loss": 0.0138,
1642
+ "step": 2550
1643
+ },
1644
+ {
1645
+ "epoch": 8.16,
1646
+ "learning_rate": 2.273695420660277e-05,
1647
+ "loss": 0.017,
1648
+ "step": 2560
1649
+ },
1650
+ {
1651
+ "epoch": 8.19,
1652
+ "learning_rate": 2.2630457933972313e-05,
1653
+ "loss": 0.0199,
1654
+ "step": 2570
1655
+ },
1656
+ {
1657
+ "epoch": 8.22,
1658
+ "learning_rate": 2.2523961661341854e-05,
1659
+ "loss": 0.0157,
1660
+ "step": 2580
1661
+ },
1662
+ {
1663
+ "epoch": 8.25,
1664
+ "learning_rate": 2.2417465388711396e-05,
1665
+ "loss": 0.0092,
1666
+ "step": 2590
1667
+ },
1668
+ {
1669
+ "epoch": 8.29,
1670
+ "learning_rate": 2.2310969116080938e-05,
1671
+ "loss": 0.011,
1672
+ "step": 2600
1673
+ },
1674
+ {
1675
+ "epoch": 8.32,
1676
+ "learning_rate": 2.220447284345048e-05,
1677
+ "loss": 0.012,
1678
+ "step": 2610
1679
+ },
1680
+ {
1681
+ "epoch": 8.35,
1682
+ "learning_rate": 2.209797657082002e-05,
1683
+ "loss": 0.0155,
1684
+ "step": 2620
1685
+ },
1686
+ {
1687
+ "epoch": 8.38,
1688
+ "learning_rate": 2.1991480298189567e-05,
1689
+ "loss": 0.0132,
1690
+ "step": 2630
1691
+ },
1692
+ {
1693
+ "epoch": 8.41,
1694
+ "learning_rate": 2.188498402555911e-05,
1695
+ "loss": 0.0119,
1696
+ "step": 2640
1697
+ },
1698
+ {
1699
+ "epoch": 8.45,
1700
+ "learning_rate": 2.1778487752928647e-05,
1701
+ "loss": 0.0147,
1702
+ "step": 2650
1703
+ },
1704
+ {
1705
+ "epoch": 8.48,
1706
+ "learning_rate": 2.167199148029819e-05,
1707
+ "loss": 0.0152,
1708
+ "step": 2660
1709
+ },
1710
+ {
1711
+ "epoch": 8.51,
1712
+ "learning_rate": 2.1565495207667734e-05,
1713
+ "loss": 0.0158,
1714
+ "step": 2670
1715
+ },
1716
+ {
1717
+ "epoch": 8.54,
1718
+ "learning_rate": 2.1458998935037275e-05,
1719
+ "loss": 0.017,
1720
+ "step": 2680
1721
+ },
1722
+ {
1723
+ "epoch": 8.57,
1724
+ "learning_rate": 2.1352502662406817e-05,
1725
+ "loss": 0.0143,
1726
+ "step": 2690
1727
+ },
1728
+ {
1729
+ "epoch": 8.61,
1730
+ "learning_rate": 2.124600638977636e-05,
1731
+ "loss": 0.015,
1732
+ "step": 2700
1733
+ },
1734
+ {
1735
+ "epoch": 8.64,
1736
+ "learning_rate": 2.11395101171459e-05,
1737
+ "loss": 0.0113,
1738
+ "step": 2710
1739
+ },
1740
+ {
1741
+ "epoch": 8.67,
1742
+ "learning_rate": 2.1033013844515442e-05,
1743
+ "loss": 0.0131,
1744
+ "step": 2720
1745
+ },
1746
+ {
1747
+ "epoch": 8.7,
1748
+ "learning_rate": 2.0926517571884984e-05,
1749
+ "loss": 0.0109,
1750
+ "step": 2730
1751
+ },
1752
+ {
1753
+ "epoch": 8.73,
1754
+ "learning_rate": 2.0820021299254526e-05,
1755
+ "loss": 0.0132,
1756
+ "step": 2740
1757
+ },
1758
+ {
1759
+ "epoch": 8.76,
1760
+ "learning_rate": 2.071352502662407e-05,
1761
+ "loss": 0.0219,
1762
+ "step": 2750
1763
+ },
1764
+ {
1765
+ "epoch": 8.8,
1766
+ "learning_rate": 2.0607028753993613e-05,
1767
+ "loss": 0.0108,
1768
+ "step": 2760
1769
+ },
1770
+ {
1771
+ "epoch": 8.83,
1772
+ "learning_rate": 2.050053248136315e-05,
1773
+ "loss": 0.0143,
1774
+ "step": 2770
1775
+ },
1776
+ {
1777
+ "epoch": 8.86,
1778
+ "learning_rate": 2.0394036208732696e-05,
1779
+ "loss": 0.0114,
1780
+ "step": 2780
1781
+ },
1782
+ {
1783
+ "epoch": 8.89,
1784
+ "learning_rate": 2.0287539936102238e-05,
1785
+ "loss": 0.0136,
1786
+ "step": 2790
1787
+ },
1788
+ {
1789
+ "epoch": 8.92,
1790
+ "learning_rate": 2.018104366347178e-05,
1791
+ "loss": 0.0114,
1792
+ "step": 2800
1793
+ },
1794
+ {
1795
+ "epoch": 8.96,
1796
+ "learning_rate": 2.007454739084132e-05,
1797
+ "loss": 0.0379,
1798
+ "step": 2810
1799
+ },
1800
+ {
1801
+ "epoch": 8.99,
1802
+ "learning_rate": 1.9968051118210863e-05,
1803
+ "loss": 0.0172,
1804
+ "step": 2820
1805
+ },
1806
+ {
1807
+ "epoch": 9.0,
1808
+ "eval_gen_len": 12.827956989247312,
1809
+ "eval_loss": 0.011623449623584747,
1810
+ "eval_rouge1": 63.4646,
1811
+ "eval_rouge2": 59.358,
1812
+ "eval_rougeL": 63.1834,
1813
+ "eval_rougeLsum": 63.3962,
1814
+ "eval_runtime": 618.8791,
1815
+ "eval_samples_per_second": 0.451,
1816
+ "eval_steps_per_second": 0.226,
1817
+ "step": 2823
1818
+ },
1819
+ {
1820
+ "epoch": 9.02,
1821
+ "learning_rate": 1.9861554845580405e-05,
1822
+ "loss": 0.0179,
1823
+ "step": 2830
1824
+ },
1825
+ {
1826
+ "epoch": 9.05,
1827
+ "learning_rate": 1.9755058572949947e-05,
1828
+ "loss": 0.0116,
1829
+ "step": 2840
1830
+ },
1831
+ {
1832
+ "epoch": 9.08,
1833
+ "learning_rate": 1.964856230031949e-05,
1834
+ "loss": 0.0108,
1835
+ "step": 2850
1836
+ },
1837
+ {
1838
+ "epoch": 9.12,
1839
+ "learning_rate": 1.9542066027689034e-05,
1840
+ "loss": 0.0126,
1841
+ "step": 2860
1842
+ },
1843
+ {
1844
+ "epoch": 9.15,
1845
+ "learning_rate": 1.9435569755058575e-05,
1846
+ "loss": 0.0092,
1847
+ "step": 2870
1848
+ },
1849
+ {
1850
+ "epoch": 9.18,
1851
+ "learning_rate": 1.9329073482428117e-05,
1852
+ "loss": 0.0163,
1853
+ "step": 2880
1854
+ },
1855
+ {
1856
+ "epoch": 9.21,
1857
+ "learning_rate": 1.9222577209797655e-05,
1858
+ "loss": 0.0162,
1859
+ "step": 2890
1860
+ },
1861
+ {
1862
+ "epoch": 9.24,
1863
+ "learning_rate": 1.91160809371672e-05,
1864
+ "loss": 0.0161,
1865
+ "step": 2900
1866
+ },
1867
+ {
1868
+ "epoch": 9.27,
1869
+ "learning_rate": 1.9009584664536742e-05,
1870
+ "loss": 0.0335,
1871
+ "step": 2910
1872
+ },
1873
+ {
1874
+ "epoch": 9.31,
1875
+ "learning_rate": 1.8903088391906284e-05,
1876
+ "loss": 0.0087,
1877
+ "step": 2920
1878
+ },
1879
+ {
1880
+ "epoch": 9.34,
1881
+ "learning_rate": 1.8796592119275826e-05,
1882
+ "loss": 0.0163,
1883
+ "step": 2930
1884
+ },
1885
+ {
1886
+ "epoch": 9.37,
1887
+ "learning_rate": 1.869009584664537e-05,
1888
+ "loss": 0.01,
1889
+ "step": 2940
1890
+ },
1891
+ {
1892
+ "epoch": 9.4,
1893
+ "learning_rate": 1.858359957401491e-05,
1894
+ "loss": 0.0215,
1895
+ "step": 2950
1896
+ },
1897
+ {
1898
+ "epoch": 9.43,
1899
+ "learning_rate": 1.847710330138445e-05,
1900
+ "loss": 0.0137,
1901
+ "step": 2960
1902
+ },
1903
+ {
1904
+ "epoch": 9.47,
1905
+ "learning_rate": 1.8370607028753993e-05,
1906
+ "loss": 0.0208,
1907
+ "step": 2970
1908
+ },
1909
+ {
1910
+ "epoch": 9.5,
1911
+ "learning_rate": 1.8264110756123538e-05,
1912
+ "loss": 0.026,
1913
+ "step": 2980
1914
+ },
1915
+ {
1916
+ "epoch": 9.53,
1917
+ "learning_rate": 1.815761448349308e-05,
1918
+ "loss": 0.0109,
1919
+ "step": 2990
1920
+ },
1921
+ {
1922
+ "epoch": 9.56,
1923
+ "learning_rate": 1.805111821086262e-05,
1924
+ "loss": 0.0122,
1925
+ "step": 3000
1926
+ },
1927
+ {
1928
+ "epoch": 9.59,
1929
+ "learning_rate": 1.7944621938232163e-05,
1930
+ "loss": 0.0109,
1931
+ "step": 3010
1932
+ },
1933
+ {
1934
+ "epoch": 9.63,
1935
+ "learning_rate": 1.7838125665601705e-05,
1936
+ "loss": 0.0203,
1937
+ "step": 3020
1938
+ },
1939
+ {
1940
+ "epoch": 9.66,
1941
+ "learning_rate": 1.7731629392971247e-05,
1942
+ "loss": 0.0081,
1943
+ "step": 3030
1944
+ },
1945
+ {
1946
+ "epoch": 9.69,
1947
+ "learning_rate": 1.762513312034079e-05,
1948
+ "loss": 0.0166,
1949
+ "step": 3040
1950
+ },
1951
+ {
1952
+ "epoch": 9.72,
1953
+ "learning_rate": 1.7518636847710333e-05,
1954
+ "loss": 0.0131,
1955
+ "step": 3050
1956
+ },
1957
+ {
1958
+ "epoch": 9.75,
1959
+ "learning_rate": 1.7412140575079875e-05,
1960
+ "loss": 0.0159,
1961
+ "step": 3060
1962
+ },
1963
+ {
1964
+ "epoch": 9.78,
1965
+ "learning_rate": 1.7305644302449414e-05,
1966
+ "loss": 0.0089,
1967
+ "step": 3070
1968
+ },
1969
+ {
1970
+ "epoch": 9.82,
1971
+ "learning_rate": 1.7199148029818955e-05,
1972
+ "loss": 0.0112,
1973
+ "step": 3080
1974
+ },
1975
+ {
1976
+ "epoch": 9.85,
1977
+ "learning_rate": 1.70926517571885e-05,
1978
+ "loss": 0.0165,
1979
+ "step": 3090
1980
+ },
1981
+ {
1982
+ "epoch": 9.88,
1983
+ "learning_rate": 1.6986155484558042e-05,
1984
+ "loss": 0.0094,
1985
+ "step": 3100
1986
+ },
1987
+ {
1988
+ "epoch": 9.91,
1989
+ "learning_rate": 1.6879659211927584e-05,
1990
+ "loss": 0.0118,
1991
+ "step": 3110
1992
+ },
1993
+ {
1994
+ "epoch": 9.94,
1995
+ "learning_rate": 1.6773162939297126e-05,
1996
+ "loss": 0.0125,
1997
+ "step": 3120
1998
+ },
1999
+ {
2000
+ "epoch": 9.98,
2001
+ "learning_rate": 1.6666666666666667e-05,
2002
+ "loss": 0.0122,
2003
+ "step": 3130
2004
+ },
2005
+ {
2006
+ "epoch": 10.0,
2007
+ "eval_gen_len": 12.526881720430108,
2008
+ "eval_loss": 0.011583242565393448,
2009
+ "eval_rouge1": 63.9777,
2010
+ "eval_rouge2": 60.058,
2011
+ "eval_rougeL": 63.7162,
2012
+ "eval_rougeLsum": 63.8393,
2013
+ "eval_runtime": 615.3638,
2014
+ "eval_samples_per_second": 0.453,
2015
+ "eval_steps_per_second": 0.228,
2016
+ "step": 3137
2017
+ },
2018
+ {
2019
+ "epoch": 10.01,
2020
+ "learning_rate": 1.656017039403621e-05,
2021
+ "loss": 0.0102,
2022
+ "step": 3140
2023
+ },
2024
+ {
2025
+ "epoch": 10.04,
2026
+ "learning_rate": 1.645367412140575e-05,
2027
+ "loss": 0.0139,
2028
+ "step": 3150
2029
+ },
2030
+ {
2031
+ "epoch": 10.07,
2032
+ "learning_rate": 1.6347177848775293e-05,
2033
+ "loss": 0.0146,
2034
+ "step": 3160
2035
+ },
2036
+ {
2037
+ "epoch": 10.1,
2038
+ "learning_rate": 1.6240681576144838e-05,
2039
+ "loss": 0.022,
2040
+ "step": 3170
2041
+ },
2042
+ {
2043
+ "epoch": 10.14,
2044
+ "learning_rate": 1.6134185303514376e-05,
2045
+ "loss": 0.0159,
2046
+ "step": 3180
2047
+ },
2048
+ {
2049
+ "epoch": 10.17,
2050
+ "learning_rate": 1.6027689030883918e-05,
2051
+ "loss": 0.0124,
2052
+ "step": 3190
2053
+ },
2054
+ {
2055
+ "epoch": 10.2,
2056
+ "learning_rate": 1.5921192758253463e-05,
2057
+ "loss": 0.0144,
2058
+ "step": 3200
2059
+ },
2060
+ {
2061
+ "epoch": 10.23,
2062
+ "learning_rate": 1.5814696485623005e-05,
2063
+ "loss": 0.0145,
2064
+ "step": 3210
2065
+ },
2066
+ {
2067
+ "epoch": 10.26,
2068
+ "learning_rate": 1.5708200212992547e-05,
2069
+ "loss": 0.0134,
2070
+ "step": 3220
2071
+ },
2072
+ {
2073
+ "epoch": 10.29,
2074
+ "learning_rate": 1.560170394036209e-05,
2075
+ "loss": 0.0152,
2076
+ "step": 3230
2077
+ },
2078
+ {
2079
+ "epoch": 10.33,
2080
+ "learning_rate": 1.549520766773163e-05,
2081
+ "loss": 0.0105,
2082
+ "step": 3240
2083
+ },
2084
+ {
2085
+ "epoch": 10.36,
2086
+ "learning_rate": 1.5388711395101172e-05,
2087
+ "loss": 0.0124,
2088
+ "step": 3250
2089
+ },
2090
+ {
2091
+ "epoch": 10.39,
2092
+ "learning_rate": 1.5282215122470714e-05,
2093
+ "loss": 0.0179,
2094
+ "step": 3260
2095
+ },
2096
+ {
2097
+ "epoch": 10.42,
2098
+ "learning_rate": 1.5175718849840257e-05,
2099
+ "loss": 0.0236,
2100
+ "step": 3270
2101
+ },
2102
+ {
2103
+ "epoch": 10.45,
2104
+ "learning_rate": 1.5069222577209799e-05,
2105
+ "loss": 0.0083,
2106
+ "step": 3280
2107
+ },
2108
+ {
2109
+ "epoch": 10.49,
2110
+ "learning_rate": 1.4962726304579342e-05,
2111
+ "loss": 0.0232,
2112
+ "step": 3290
2113
+ },
2114
+ {
2115
+ "epoch": 10.52,
2116
+ "learning_rate": 1.485623003194888e-05,
2117
+ "loss": 0.0089,
2118
+ "step": 3300
2119
+ },
2120
+ {
2121
+ "epoch": 10.55,
2122
+ "learning_rate": 1.4749733759318424e-05,
2123
+ "loss": 0.0116,
2124
+ "step": 3310
2125
+ },
2126
+ {
2127
+ "epoch": 10.58,
2128
+ "learning_rate": 1.4643237486687966e-05,
2129
+ "loss": 0.0106,
2130
+ "step": 3320
2131
+ },
2132
+ {
2133
+ "epoch": 10.61,
2134
+ "learning_rate": 1.453674121405751e-05,
2135
+ "loss": 0.0327,
2136
+ "step": 3330
2137
+ },
2138
+ {
2139
+ "epoch": 10.65,
2140
+ "learning_rate": 1.4430244941427051e-05,
2141
+ "loss": 0.0074,
2142
+ "step": 3340
2143
+ },
2144
+ {
2145
+ "epoch": 10.68,
2146
+ "learning_rate": 1.4323748668796594e-05,
2147
+ "loss": 0.0113,
2148
+ "step": 3350
2149
+ },
2150
+ {
2151
+ "epoch": 10.71,
2152
+ "learning_rate": 1.4217252396166134e-05,
2153
+ "loss": 0.0139,
2154
+ "step": 3360
2155
+ },
2156
+ {
2157
+ "epoch": 10.74,
2158
+ "learning_rate": 1.4110756123535676e-05,
2159
+ "loss": 0.0106,
2160
+ "step": 3370
2161
+ },
2162
+ {
2163
+ "epoch": 10.77,
2164
+ "learning_rate": 1.4004259850905218e-05,
2165
+ "loss": 0.011,
2166
+ "step": 3380
2167
+ },
2168
+ {
2169
+ "epoch": 10.8,
2170
+ "learning_rate": 1.3897763578274761e-05,
2171
+ "loss": 0.0161,
2172
+ "step": 3390
2173
+ },
2174
+ {
2175
+ "epoch": 10.84,
2176
+ "learning_rate": 1.3791267305644303e-05,
2177
+ "loss": 0.0156,
2178
+ "step": 3400
2179
+ },
2180
+ {
2181
+ "epoch": 10.87,
2182
+ "learning_rate": 1.3684771033013847e-05,
2183
+ "loss": 0.0076,
2184
+ "step": 3410
2185
+ },
2186
+ {
2187
+ "epoch": 10.9,
2188
+ "learning_rate": 1.3578274760383387e-05,
2189
+ "loss": 0.0111,
2190
+ "step": 3420
2191
+ },
2192
+ {
2193
+ "epoch": 10.93,
2194
+ "learning_rate": 1.3471778487752928e-05,
2195
+ "loss": 0.0159,
2196
+ "step": 3430
2197
+ },
2198
+ {
2199
+ "epoch": 10.96,
2200
+ "learning_rate": 1.3365282215122472e-05,
2201
+ "loss": 0.0138,
2202
+ "step": 3440
2203
+ },
2204
+ {
2205
+ "epoch": 11.0,
2206
+ "learning_rate": 1.3258785942492014e-05,
2207
+ "loss": 0.0132,
2208
+ "step": 3450
2209
+ },
2210
+ {
2211
+ "epoch": 11.0,
2212
+ "eval_gen_len": 12.577060931899641,
2213
+ "eval_loss": 0.011564880609512329,
2214
+ "eval_rouge1": 63.7793,
2215
+ "eval_rouge2": 59.7984,
2216
+ "eval_rougeL": 63.5653,
2217
+ "eval_rougeLsum": 63.7034,
2218
+ "eval_runtime": 615.3689,
2219
+ "eval_samples_per_second": 0.453,
2220
+ "eval_steps_per_second": 0.228,
2221
+ "step": 3451
2222
+ },
2223
+ {
2224
+ "epoch": 11.03,
2225
+ "learning_rate": 1.3152289669861555e-05,
2226
+ "loss": 0.0129,
2227
+ "step": 3460
2228
+ },
2229
+ {
2230
+ "epoch": 11.06,
2231
+ "learning_rate": 1.3045793397231099e-05,
2232
+ "loss": 0.0227,
2233
+ "step": 3470
2234
+ },
2235
+ {
2236
+ "epoch": 11.09,
2237
+ "learning_rate": 1.2939297124600639e-05,
2238
+ "loss": 0.0112,
2239
+ "step": 3480
2240
+ },
2241
+ {
2242
+ "epoch": 11.12,
2243
+ "learning_rate": 1.283280085197018e-05,
2244
+ "loss": 0.0147,
2245
+ "step": 3490
2246
+ },
2247
+ {
2248
+ "epoch": 11.16,
2249
+ "learning_rate": 1.2726304579339724e-05,
2250
+ "loss": 0.0113,
2251
+ "step": 3500
2252
+ },
2253
+ {
2254
+ "epoch": 11.19,
2255
+ "learning_rate": 1.2619808306709266e-05,
2256
+ "loss": 0.0123,
2257
+ "step": 3510
2258
+ },
2259
+ {
2260
+ "epoch": 11.22,
2261
+ "learning_rate": 1.2513312034078809e-05,
2262
+ "loss": 0.0191,
2263
+ "step": 3520
2264
+ },
2265
+ {
2266
+ "epoch": 11.25,
2267
+ "learning_rate": 1.240681576144835e-05,
2268
+ "loss": 0.0128,
2269
+ "step": 3530
2270
+ },
2271
+ {
2272
+ "epoch": 11.28,
2273
+ "learning_rate": 1.2300319488817893e-05,
2274
+ "loss": 0.0101,
2275
+ "step": 3540
2276
+ },
2277
+ {
2278
+ "epoch": 11.31,
2279
+ "learning_rate": 1.2193823216187434e-05,
2280
+ "loss": 0.0129,
2281
+ "step": 3550
2282
+ },
2283
+ {
2284
+ "epoch": 11.35,
2285
+ "learning_rate": 1.2087326943556976e-05,
2286
+ "loss": 0.0093,
2287
+ "step": 3560
2288
+ },
2289
+ {
2290
+ "epoch": 11.38,
2291
+ "learning_rate": 1.1980830670926518e-05,
2292
+ "loss": 0.0137,
2293
+ "step": 3570
2294
+ },
2295
+ {
2296
+ "epoch": 11.41,
2297
+ "learning_rate": 1.1874334398296061e-05,
2298
+ "loss": 0.0114,
2299
+ "step": 3580
2300
+ },
2301
+ {
2302
+ "epoch": 11.44,
2303
+ "learning_rate": 1.1767838125665601e-05,
2304
+ "loss": 0.0164,
2305
+ "step": 3590
2306
+ },
2307
+ {
2308
+ "epoch": 11.47,
2309
+ "learning_rate": 1.1661341853035145e-05,
2310
+ "loss": 0.0207,
2311
+ "step": 3600
2312
+ },
2313
+ {
2314
+ "epoch": 11.51,
2315
+ "learning_rate": 1.1554845580404687e-05,
2316
+ "loss": 0.0109,
2317
+ "step": 3610
2318
+ },
2319
+ {
2320
+ "epoch": 11.54,
2321
+ "learning_rate": 1.1448349307774228e-05,
2322
+ "loss": 0.0169,
2323
+ "step": 3620
2324
+ },
2325
+ {
2326
+ "epoch": 11.57,
2327
+ "learning_rate": 1.134185303514377e-05,
2328
+ "loss": 0.0102,
2329
+ "step": 3630
2330
+ },
2331
+ {
2332
+ "epoch": 11.6,
2333
+ "learning_rate": 1.1235356762513314e-05,
2334
+ "loss": 0.0248,
2335
+ "step": 3640
2336
+ },
2337
+ {
2338
+ "epoch": 11.63,
2339
+ "learning_rate": 1.1128860489882854e-05,
2340
+ "loss": 0.0112,
2341
+ "step": 3650
2342
+ },
2343
+ {
2344
+ "epoch": 11.67,
2345
+ "learning_rate": 1.1022364217252397e-05,
2346
+ "loss": 0.02,
2347
+ "step": 3660
2348
+ },
2349
+ {
2350
+ "epoch": 11.7,
2351
+ "learning_rate": 1.0915867944621939e-05,
2352
+ "loss": 0.0127,
2353
+ "step": 3670
2354
+ },
2355
+ {
2356
+ "epoch": 11.73,
2357
+ "learning_rate": 1.080937167199148e-05,
2358
+ "loss": 0.0147,
2359
+ "step": 3680
2360
+ },
2361
+ {
2362
+ "epoch": 11.76,
2363
+ "learning_rate": 1.0702875399361024e-05,
2364
+ "loss": 0.014,
2365
+ "step": 3690
2366
+ },
2367
+ {
2368
+ "epoch": 11.79,
2369
+ "learning_rate": 1.0596379126730564e-05,
2370
+ "loss": 0.0079,
2371
+ "step": 3700
2372
+ },
2373
+ {
2374
+ "epoch": 11.82,
2375
+ "learning_rate": 1.0489882854100107e-05,
2376
+ "loss": 0.0134,
2377
+ "step": 3710
2378
+ },
2379
+ {
2380
+ "epoch": 11.86,
2381
+ "learning_rate": 1.038338658146965e-05,
2382
+ "loss": 0.0188,
2383
+ "step": 3720
2384
+ },
2385
+ {
2386
+ "epoch": 11.89,
2387
+ "learning_rate": 1.0276890308839191e-05,
2388
+ "loss": 0.0087,
2389
+ "step": 3730
2390
+ },
2391
+ {
2392
+ "epoch": 11.92,
2393
+ "learning_rate": 1.0170394036208733e-05,
2394
+ "loss": 0.0104,
2395
+ "step": 3740
2396
+ },
2397
+ {
2398
+ "epoch": 11.95,
2399
+ "learning_rate": 1.0063897763578276e-05,
2400
+ "loss": 0.0132,
2401
+ "step": 3750
2402
+ },
2403
+ {
2404
+ "epoch": 11.98,
2405
+ "learning_rate": 9.957401490947816e-06,
2406
+ "loss": 0.0242,
2407
+ "step": 3760
2408
+ },
2409
+ {
2410
+ "epoch": 12.0,
2411
+ "eval_gen_len": 12.67741935483871,
2412
+ "eval_loss": 0.011580224148929119,
2413
+ "eval_rouge1": 63.3071,
2414
+ "eval_rouge2": 59.216,
2415
+ "eval_rougeL": 62.9553,
2416
+ "eval_rougeLsum": 63.1625,
2417
+ "eval_runtime": 618.8937,
2418
+ "eval_samples_per_second": 0.451,
2419
+ "eval_steps_per_second": 0.226,
2420
+ "step": 3765
2421
+ },
2422
+ {
2423
+ "epoch": 12.02,
2424
+ "learning_rate": 9.85090521831736e-06,
2425
+ "loss": 0.0084,
2426
+ "step": 3770
2427
+ },
2428
+ {
2429
+ "epoch": 12.05,
2430
+ "learning_rate": 9.744408945686901e-06,
2431
+ "loss": 0.0106,
2432
+ "step": 3780
2433
+ },
2434
+ {
2435
+ "epoch": 12.08,
2436
+ "learning_rate": 9.637912673056443e-06,
2437
+ "loss": 0.0106,
2438
+ "step": 3790
2439
+ },
2440
+ {
2441
+ "epoch": 12.11,
2442
+ "learning_rate": 9.531416400425985e-06,
2443
+ "loss": 0.0111,
2444
+ "step": 3800
2445
+ },
2446
+ {
2447
+ "epoch": 12.14,
2448
+ "learning_rate": 9.424920127795528e-06,
2449
+ "loss": 0.0118,
2450
+ "step": 3810
2451
+ },
2452
+ {
2453
+ "epoch": 12.18,
2454
+ "learning_rate": 9.318423855165068e-06,
2455
+ "loss": 0.0114,
2456
+ "step": 3820
2457
+ },
2458
+ {
2459
+ "epoch": 12.21,
2460
+ "learning_rate": 9.211927582534612e-06,
2461
+ "loss": 0.0134,
2462
+ "step": 3830
2463
+ },
2464
+ {
2465
+ "epoch": 12.24,
2466
+ "learning_rate": 9.105431309904154e-06,
2467
+ "loss": 0.0119,
2468
+ "step": 3840
2469
+ },
2470
+ {
2471
+ "epoch": 12.27,
2472
+ "learning_rate": 8.998935037273695e-06,
2473
+ "loss": 0.0097,
2474
+ "step": 3850
2475
+ },
2476
+ {
2477
+ "epoch": 12.3,
2478
+ "learning_rate": 8.892438764643237e-06,
2479
+ "loss": 0.0167,
2480
+ "step": 3860
2481
+ },
2482
+ {
2483
+ "epoch": 12.33,
2484
+ "learning_rate": 8.78594249201278e-06,
2485
+ "loss": 0.0281,
2486
+ "step": 3870
2487
+ },
2488
+ {
2489
+ "epoch": 12.37,
2490
+ "learning_rate": 8.679446219382322e-06,
2491
+ "loss": 0.0084,
2492
+ "step": 3880
2493
+ },
2494
+ {
2495
+ "epoch": 12.4,
2496
+ "learning_rate": 8.572949946751864e-06,
2497
+ "loss": 0.0105,
2498
+ "step": 3890
2499
+ },
2500
+ {
2501
+ "epoch": 12.43,
2502
+ "learning_rate": 8.466453674121406e-06,
2503
+ "loss": 0.0146,
2504
+ "step": 3900
2505
+ },
2506
+ {
2507
+ "epoch": 12.46,
2508
+ "learning_rate": 8.359957401490947e-06,
2509
+ "loss": 0.011,
2510
+ "step": 3910
2511
+ },
2512
+ {
2513
+ "epoch": 12.49,
2514
+ "learning_rate": 8.253461128860491e-06,
2515
+ "loss": 0.0077,
2516
+ "step": 3920
2517
+ },
2518
+ {
2519
+ "epoch": 12.53,
2520
+ "learning_rate": 8.146964856230033e-06,
2521
+ "loss": 0.0278,
2522
+ "step": 3930
2523
+ },
2524
+ {
2525
+ "epoch": 12.56,
2526
+ "learning_rate": 8.040468583599574e-06,
2527
+ "loss": 0.0151,
2528
+ "step": 3940
2529
+ },
2530
+ {
2531
+ "epoch": 12.59,
2532
+ "learning_rate": 7.933972310969116e-06,
2533
+ "loss": 0.0124,
2534
+ "step": 3950
2535
+ },
2536
+ {
2537
+ "epoch": 12.62,
2538
+ "learning_rate": 7.82747603833866e-06,
2539
+ "loss": 0.0265,
2540
+ "step": 3960
2541
+ },
2542
+ {
2543
+ "epoch": 12.65,
2544
+ "learning_rate": 7.7209797657082e-06,
2545
+ "loss": 0.014,
2546
+ "step": 3970
2547
+ },
2548
+ {
2549
+ "epoch": 12.69,
2550
+ "learning_rate": 7.614483493077742e-06,
2551
+ "loss": 0.0118,
2552
+ "step": 3980
2553
+ },
2554
+ {
2555
+ "epoch": 12.72,
2556
+ "learning_rate": 7.507987220447285e-06,
2557
+ "loss": 0.0163,
2558
+ "step": 3990
2559
+ },
2560
+ {
2561
+ "epoch": 12.75,
2562
+ "learning_rate": 7.4014909478168266e-06,
2563
+ "loss": 0.0199,
2564
+ "step": 4000
2565
+ },
2566
+ {
2567
+ "epoch": 12.78,
2568
+ "learning_rate": 7.294994675186369e-06,
2569
+ "loss": 0.0082,
2570
+ "step": 4010
2571
+ },
2572
+ {
2573
+ "epoch": 12.81,
2574
+ "learning_rate": 7.188498402555911e-06,
2575
+ "loss": 0.0067,
2576
+ "step": 4020
2577
+ },
2578
+ {
2579
+ "epoch": 12.84,
2580
+ "learning_rate": 7.082002129925453e-06,
2581
+ "loss": 0.015,
2582
+ "step": 4030
2583
+ },
2584
+ {
2585
+ "epoch": 12.88,
2586
+ "learning_rate": 6.975505857294995e-06,
2587
+ "loss": 0.0156,
2588
+ "step": 4040
2589
+ },
2590
+ {
2591
+ "epoch": 12.91,
2592
+ "learning_rate": 6.869009584664538e-06,
2593
+ "loss": 0.0218,
2594
+ "step": 4050
2595
+ },
2596
+ {
2597
+ "epoch": 12.94,
2598
+ "learning_rate": 6.762513312034079e-06,
2599
+ "loss": 0.0165,
2600
+ "step": 4060
2601
+ },
2602
+ {
2603
+ "epoch": 12.97,
2604
+ "learning_rate": 6.656017039403621e-06,
2605
+ "loss": 0.0178,
2606
+ "step": 4070
2607
+ },
2608
+ {
2609
+ "epoch": 13.0,
2610
+ "eval_gen_len": 12.67741935483871,
2611
+ "eval_loss": 0.011560924351215363,
2612
+ "eval_rouge1": 63.3607,
2613
+ "eval_rouge2": 59.1699,
2614
+ "eval_rougeL": 63.0136,
2615
+ "eval_rougeLsum": 63.1274,
2616
+ "eval_runtime": 618.877,
2617
+ "eval_samples_per_second": 0.451,
2618
+ "eval_steps_per_second": 0.226,
2619
+ "step": 4078
2620
+ },
2621
+ {
2622
+ "epoch": 13.0,
2623
+ "learning_rate": 6.549520766773164e-06,
2624
+ "loss": 0.0092,
2625
+ "step": 4080
2626
+ },
2627
+ {
2628
+ "epoch": 13.04,
2629
+ "learning_rate": 6.443024494142705e-06,
2630
+ "loss": 0.0084,
2631
+ "step": 4090
2632
+ },
2633
+ {
2634
+ "epoch": 13.07,
2635
+ "learning_rate": 6.336528221512247e-06,
2636
+ "loss": 0.0149,
2637
+ "step": 4100
2638
+ },
2639
+ {
2640
+ "epoch": 13.1,
2641
+ "learning_rate": 6.230031948881789e-06,
2642
+ "loss": 0.0138,
2643
+ "step": 4110
2644
+ },
2645
+ {
2646
+ "epoch": 13.13,
2647
+ "learning_rate": 6.123535676251332e-06,
2648
+ "loss": 0.0128,
2649
+ "step": 4120
2650
+ },
2651
+ {
2652
+ "epoch": 13.16,
2653
+ "learning_rate": 6.0170394036208735e-06,
2654
+ "loss": 0.0124,
2655
+ "step": 4130
2656
+ },
2657
+ {
2658
+ "epoch": 13.2,
2659
+ "learning_rate": 5.910543130990415e-06,
2660
+ "loss": 0.0122,
2661
+ "step": 4140
2662
+ },
2663
+ {
2664
+ "epoch": 13.23,
2665
+ "learning_rate": 5.804046858359958e-06,
2666
+ "loss": 0.0175,
2667
+ "step": 4150
2668
+ },
2669
+ {
2670
+ "epoch": 13.26,
2671
+ "learning_rate": 5.6975505857295e-06,
2672
+ "loss": 0.0142,
2673
+ "step": 4160
2674
+ },
2675
+ {
2676
+ "epoch": 13.29,
2677
+ "learning_rate": 5.591054313099041e-06,
2678
+ "loss": 0.009,
2679
+ "step": 4170
2680
+ },
2681
+ {
2682
+ "epoch": 13.32,
2683
+ "learning_rate": 5.484558040468584e-06,
2684
+ "loss": 0.0129,
2685
+ "step": 4180
2686
+ },
2687
+ {
2688
+ "epoch": 13.35,
2689
+ "learning_rate": 5.378061767838126e-06,
2690
+ "loss": 0.0086,
2691
+ "step": 4190
2692
+ },
2693
+ {
2694
+ "epoch": 13.39,
2695
+ "learning_rate": 5.2715654952076674e-06,
2696
+ "loss": 0.0109,
2697
+ "step": 4200
2698
+ },
2699
+ {
2700
+ "epoch": 13.42,
2701
+ "learning_rate": 5.16506922257721e-06,
2702
+ "loss": 0.0127,
2703
+ "step": 4210
2704
+ },
2705
+ {
2706
+ "epoch": 13.45,
2707
+ "learning_rate": 5.058572949946752e-06,
2708
+ "loss": 0.0235,
2709
+ "step": 4220
2710
+ },
2711
+ {
2712
+ "epoch": 13.48,
2713
+ "learning_rate": 4.952076677316294e-06,
2714
+ "loss": 0.0101,
2715
+ "step": 4230
2716
+ },
2717
+ {
2718
+ "epoch": 13.51,
2719
+ "learning_rate": 4.845580404685836e-06,
2720
+ "loss": 0.0177,
2721
+ "step": 4240
2722
+ },
2723
+ {
2724
+ "epoch": 13.55,
2725
+ "learning_rate": 4.739084132055379e-06,
2726
+ "loss": 0.0242,
2727
+ "step": 4250
2728
+ },
2729
+ {
2730
+ "epoch": 13.58,
2731
+ "learning_rate": 4.6325878594249205e-06,
2732
+ "loss": 0.0157,
2733
+ "step": 4260
2734
+ },
2735
+ {
2736
+ "epoch": 13.61,
2737
+ "learning_rate": 4.526091586794463e-06,
2738
+ "loss": 0.015,
2739
+ "step": 4270
2740
+ },
2741
+ {
2742
+ "epoch": 13.64,
2743
+ "learning_rate": 4.419595314164005e-06,
2744
+ "loss": 0.0139,
2745
+ "step": 4280
2746
+ },
2747
+ {
2748
+ "epoch": 13.67,
2749
+ "learning_rate": 4.3130990415335465e-06,
2750
+ "loss": 0.0126,
2751
+ "step": 4290
2752
+ },
2753
+ {
2754
+ "epoch": 13.71,
2755
+ "learning_rate": 4.206602768903089e-06,
2756
+ "loss": 0.0137,
2757
+ "step": 4300
2758
+ },
2759
+ {
2760
+ "epoch": 13.74,
2761
+ "learning_rate": 4.100106496272631e-06,
2762
+ "loss": 0.0071,
2763
+ "step": 4310
2764
+ },
2765
+ {
2766
+ "epoch": 13.77,
2767
+ "learning_rate": 3.993610223642173e-06,
2768
+ "loss": 0.0201,
2769
+ "step": 4320
2770
+ },
2771
+ {
2772
+ "epoch": 13.8,
2773
+ "learning_rate": 3.887113951011715e-06,
2774
+ "loss": 0.0188,
2775
+ "step": 4330
2776
+ },
2777
+ {
2778
+ "epoch": 13.83,
2779
+ "learning_rate": 3.780617678381257e-06,
2780
+ "loss": 0.0081,
2781
+ "step": 4340
2782
+ },
2783
+ {
2784
+ "epoch": 13.86,
2785
+ "learning_rate": 3.6741214057507987e-06,
2786
+ "loss": 0.0154,
2787
+ "step": 4350
2788
+ },
2789
+ {
2790
+ "epoch": 13.9,
2791
+ "learning_rate": 3.5676251331203413e-06,
2792
+ "loss": 0.0159,
2793
+ "step": 4360
2794
+ },
2795
+ {
2796
+ "epoch": 13.93,
2797
+ "learning_rate": 3.461128860489883e-06,
2798
+ "loss": 0.0128,
2799
+ "step": 4370
2800
+ },
2801
+ {
2802
+ "epoch": 13.96,
2803
+ "learning_rate": 3.354632587859425e-06,
2804
+ "loss": 0.0144,
2805
+ "step": 4380
2806
+ },
2807
+ {
2808
+ "epoch": 13.99,
2809
+ "learning_rate": 3.2481363152289674e-06,
2810
+ "loss": 0.0225,
2811
+ "step": 4390
2812
+ },
2813
+ {
2814
+ "epoch": 14.0,
2815
+ "eval_gen_len": 12.627240143369175,
2816
+ "eval_loss": 0.011567563749849796,
2817
+ "eval_rouge1": 63.5224,
2818
+ "eval_rouge2": 59.4962,
2819
+ "eval_rougeL": 63.22,
2820
+ "eval_rougeLsum": 63.4236,
2821
+ "eval_runtime": 615.3823,
2822
+ "eval_samples_per_second": 0.453,
2823
+ "eval_steps_per_second": 0.228,
2824
+ "step": 4392
2825
+ },
2826
+ {
2827
+ "epoch": 14.02,
2828
+ "learning_rate": 3.141640042598509e-06,
2829
+ "loss": 0.0107,
2830
+ "step": 4400
2831
+ },
2832
+ {
2833
+ "epoch": 14.06,
2834
+ "learning_rate": 3.0351437699680513e-06,
2835
+ "loss": 0.0207,
2836
+ "step": 4410
2837
+ },
2838
+ {
2839
+ "epoch": 14.09,
2840
+ "learning_rate": 2.9286474973375935e-06,
2841
+ "loss": 0.0132,
2842
+ "step": 4420
2843
+ },
2844
+ {
2845
+ "epoch": 14.12,
2846
+ "learning_rate": 2.8221512247071357e-06,
2847
+ "loss": 0.0142,
2848
+ "step": 4430
2849
+ },
2850
+ {
2851
+ "epoch": 14.15,
2852
+ "learning_rate": 2.7156549520766774e-06,
2853
+ "loss": 0.0139,
2854
+ "step": 4440
2855
+ },
2856
+ {
2857
+ "epoch": 14.18,
2858
+ "learning_rate": 2.6091586794462196e-06,
2859
+ "loss": 0.0127,
2860
+ "step": 4450
2861
+ },
2862
+ {
2863
+ "epoch": 14.22,
2864
+ "learning_rate": 2.5026624068157617e-06,
2865
+ "loss": 0.0136,
2866
+ "step": 4460
2867
+ },
2868
+ {
2869
+ "epoch": 14.25,
2870
+ "learning_rate": 2.3961661341853035e-06,
2871
+ "loss": 0.0086,
2872
+ "step": 4470
2873
+ },
2874
+ {
2875
+ "epoch": 14.28,
2876
+ "learning_rate": 2.2896698615548457e-06,
2877
+ "loss": 0.0104,
2878
+ "step": 4480
2879
+ },
2880
+ {
2881
+ "epoch": 14.31,
2882
+ "learning_rate": 2.183173588924388e-06,
2883
+ "loss": 0.0205,
2884
+ "step": 4490
2885
+ },
2886
+ {
2887
+ "epoch": 14.34,
2888
+ "learning_rate": 2.0766773162939296e-06,
2889
+ "loss": 0.0158,
2890
+ "step": 4500
2891
+ },
2892
+ {
2893
+ "epoch": 14.37,
2894
+ "learning_rate": 1.9701810436634718e-06,
2895
+ "loss": 0.0114,
2896
+ "step": 4510
2897
+ },
2898
+ {
2899
+ "epoch": 14.41,
2900
+ "learning_rate": 1.8636847710330141e-06,
2901
+ "loss": 0.0296,
2902
+ "step": 4520
2903
+ },
2904
+ {
2905
+ "epoch": 14.44,
2906
+ "learning_rate": 1.7571884984025559e-06,
2907
+ "loss": 0.021,
2908
+ "step": 4530
2909
+ },
2910
+ {
2911
+ "epoch": 14.47,
2912
+ "learning_rate": 1.650692225772098e-06,
2913
+ "loss": 0.0123,
2914
+ "step": 4540
2915
+ },
2916
+ {
2917
+ "epoch": 14.5,
2918
+ "learning_rate": 1.54419595314164e-06,
2919
+ "loss": 0.0178,
2920
+ "step": 4550
2921
+ },
2922
+ {
2923
+ "epoch": 14.53,
2924
+ "learning_rate": 1.4376996805111822e-06,
2925
+ "loss": 0.0116,
2926
+ "step": 4560
2927
+ },
2928
+ {
2929
+ "epoch": 14.57,
2930
+ "learning_rate": 1.3312034078807244e-06,
2931
+ "loss": 0.0101,
2932
+ "step": 4570
2933
+ },
2934
+ {
2935
+ "epoch": 14.6,
2936
+ "learning_rate": 1.2247071352502663e-06,
2937
+ "loss": 0.0166,
2938
+ "step": 4580
2939
+ },
2940
+ {
2941
+ "epoch": 14.63,
2942
+ "learning_rate": 1.1182108626198083e-06,
2943
+ "loss": 0.0136,
2944
+ "step": 4590
2945
+ },
2946
+ {
2947
+ "epoch": 14.66,
2948
+ "learning_rate": 1.0117145899893504e-06,
2949
+ "loss": 0.0161,
2950
+ "step": 4600
2951
+ },
2952
+ {
2953
+ "epoch": 14.69,
2954
+ "learning_rate": 9.052183173588925e-07,
2955
+ "loss": 0.0082,
2956
+ "step": 4610
2957
+ },
2958
+ {
2959
+ "epoch": 14.73,
2960
+ "learning_rate": 7.987220447284345e-07,
2961
+ "loss": 0.0159,
2962
+ "step": 4620
2963
+ },
2964
+ {
2965
+ "epoch": 14.76,
2966
+ "learning_rate": 6.922257720979766e-07,
2967
+ "loss": 0.0093,
2968
+ "step": 4630
2969
+ },
2970
+ {
2971
+ "epoch": 14.79,
2972
+ "learning_rate": 5.857294994675187e-07,
2973
+ "loss": 0.0096,
2974
+ "step": 4640
2975
+ },
2976
+ {
2977
+ "epoch": 14.82,
2978
+ "learning_rate": 4.792332268370607e-07,
2979
+ "loss": 0.0135,
2980
+ "step": 4650
2981
+ },
2982
+ {
2983
+ "epoch": 14.85,
2984
+ "learning_rate": 3.727369542066028e-07,
2985
+ "loss": 0.0123,
2986
+ "step": 4660
2987
+ },
2988
+ {
2989
+ "epoch": 14.88,
2990
+ "learning_rate": 2.6624068157614484e-07,
2991
+ "loss": 0.007,
2992
+ "step": 4670
2993
+ },
2994
+ {
2995
+ "epoch": 14.92,
2996
+ "learning_rate": 1.597444089456869e-07,
2997
+ "loss": 0.0184,
2998
+ "step": 4680
2999
+ },
3000
+ {
3001
+ "epoch": 14.95,
3002
+ "learning_rate": 5.3248136315228974e-08,
3003
+ "loss": 0.0204,
3004
+ "step": 4690
3005
+ },
3006
+ {
3007
+ "epoch": 14.96,
3008
+ "eval_gen_len": 12.577060931899641,
3009
+ "eval_loss": 0.01156473346054554,
3010
+ "eval_rouge1": 63.2308,
3011
+ "eval_rouge2": 59.1657,
3012
+ "eval_rougeL": 62.9973,
3013
+ "eval_rougeLsum": 63.1556,
3014
+ "eval_runtime": 615.4381,
3015
+ "eval_samples_per_second": 0.453,
3016
+ "eval_steps_per_second": 0.227,
3017
+ "step": 4695
3018
+ }
3019
+ ],
3020
+ "max_steps": 4695,
3021
+ "num_train_epochs": 15,
3022
+ "total_flos": 3.211052520088535e+17,
3023
+ "trial_name": null,
3024
+ "trial_params": null
3025
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:22e8716922f51317411205c42896c0e346c5f64f2782ee0b7659686624bfa284
3
+ size 4155