roy commited on
Commit
59e423f
1 Parent(s): bfce93f
.gitattributes CHANGED
@@ -32,3 +32,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
32
  *.zip filter=lfs diff=lfs merge=lfs -text
33
  *.zst filter=lfs diff=lfs merge=lfs -text
34
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
32
  *.zip filter=lfs diff=lfs merge=lfs -text
33
  *.zst filter=lfs diff=lfs merge=lfs -text
34
  *tfevents* filter=lfs diff=lfs merge=lfs -text
35
+ pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
all_results.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "eval_gen_len": 20.0,
4
+ "eval_loss": 1.933449387550354,
5
+ "eval_rouge1": 22.3137,
6
+ "eval_rouge2": 5.8376,
7
+ "eval_rougeL": 22.1264,
8
+ "eval_rougeLsum": 22.1274,
9
+ "eval_runtime": 163.5249,
10
+ "eval_samples": 7008,
11
+ "eval_samples_per_second": 42.856,
12
+ "eval_steps_per_second": 2.678,
13
+ "train_loss": 1.5038130920392954,
14
+ "train_runtime": 13804.3068,
15
+ "train_samples": 56760,
16
+ "train_samples_per_second": 41.118,
17
+ "train_steps_per_second": 2.57
18
+ }
config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "gogamza/kobart-base-v2",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "gelu",
5
+ "add_bias_logits": false,
6
+ "add_final_layer_norm": false,
7
+ "architectures": [
8
+ "BartForConditionalGeneration"
9
+ ],
10
+ "attention_dropout": 0.0,
11
+ "author": "Heewon Jeon(madjakarta@gmail.com)",
12
+ "bos_token_id": 1,
13
+ "classif_dropout": 0.1,
14
+ "classifier_dropout": 0.1,
15
+ "d_model": 768,
16
+ "decoder_attention_heads": 16,
17
+ "decoder_ffn_dim": 3072,
18
+ "decoder_layerdrop": 0.0,
19
+ "decoder_layers": 6,
20
+ "decoder_start_token_id": 1,
21
+ "do_blenderbot_90_layernorm": false,
22
+ "dropout": 0.1,
23
+ "encoder_attention_heads": 16,
24
+ "encoder_ffn_dim": 3072,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 6,
27
+ "eos_token_id": 1,
28
+ "extra_pos_embeddings": 2,
29
+ "force_bos_token_to_be_generated": false,
30
+ "forced_eos_token_id": 1,
31
+ "gradient_checkpointing": false,
32
+ "id2label": {
33
+ "0": "NEGATIVE",
34
+ "1": "POSITIVE"
35
+ },
36
+ "init_std": 0.02,
37
+ "is_encoder_decoder": true,
38
+ "kobart_version": 2.0,
39
+ "label2id": {
40
+ "NEGATIVE": 0,
41
+ "POSITIVE": 1
42
+ },
43
+ "max_position_embeddings": 1026,
44
+ "model_type": "bart",
45
+ "normalize_before": false,
46
+ "normalize_embedding": true,
47
+ "num_hidden_layers": 6,
48
+ "pad_token_id": 3,
49
+ "scale_embedding": false,
50
+ "static_position_embeddings": false,
51
+ "tokenizer_class": "PreTrainedTokenizerFast",
52
+ "torch_dtype": "float32",
53
+ "transformers_version": "4.25.0",
54
+ "use_cache": true,
55
+ "vocab_size": 30000
56
+ }
eval_results.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "eval_gen_len": 20.0,
4
+ "eval_loss": 1.933449387550354,
5
+ "eval_rouge1": 22.3137,
6
+ "eval_rouge2": 5.8376,
7
+ "eval_rougeL": 22.1264,
8
+ "eval_rougeLsum": 22.1274,
9
+ "eval_runtime": 163.5249,
10
+ "eval_samples": 7008,
11
+ "eval_samples_per_second": 42.856,
12
+ "eval_steps_per_second": 2.678
13
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:50b60bc88908f3b3fab51f1e924033f49a8c9055d01c6148a53ad4cf0dc47454
3
+ size 495652819
special_tokens_map.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "</s>",
3
+ "eos_token": "</s>",
4
+ "mask_token": "<mask>",
5
+ "pad_token": "<pad>",
6
+ "unk_token": "<unk>"
7
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_max_length": 1000000000000000019884624838656,
3
+ "name_or_path": "gogamza/kobart-base-v2",
4
+ "special_tokens_map_file": "/opt/ml/input/Summarization/checkpoint-34000/models--gogamza--kobart-base-v2/snapshots/f9f2ec35d3c32a1ecc7a3281f9626b7ec1913fed/special_tokens_map.json",
5
+ "tokenizer_class": "PreTrainedTokenizerFast",
6
+ "use_fast": true
7
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "train_loss": 1.5038130920392954,
4
+ "train_runtime": 13804.3068,
5
+ "train_samples": 56760,
6
+ "train_samples_per_second": 41.118,
7
+ "train_steps_per_second": 2.57
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,2370 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 1.933449387550354,
3
+ "best_model_checkpoint": "/opt/ml/input/Summarization/checkpoint-10000",
4
+ "epoch": 10.0,
5
+ "global_step": 35480,
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": 9.603605964719648e-06,
13
+ "loss": 2.9628,
14
+ "step": 100
15
+ },
16
+ {
17
+ "epoch": 0.06,
18
+ "learning_rate": 2.651882655109716e-05,
19
+ "loss": 2.3592,
20
+ "step": 200
21
+ },
22
+ {
23
+ "epoch": 0.08,
24
+ "learning_rate": 4.9653319145399344e-05,
25
+ "loss": 2.2458,
26
+ "step": 300
27
+ },
28
+ {
29
+ "epoch": 0.11,
30
+ "learning_rate": 3.5272614557174996e-05,
31
+ "loss": 2.2573,
32
+ "step": 400
33
+ },
34
+ {
35
+ "epoch": 0.14,
36
+ "learning_rate": 1.1805441557799242e-05,
37
+ "loss": 2.2079,
38
+ "step": 500
39
+ },
40
+ {
41
+ "epoch": 0.17,
42
+ "learning_rate": 1.1082060287457063e-07,
43
+ "loss": 2.1759,
44
+ "step": 600
45
+ },
46
+ {
47
+ "epoch": 0.2,
48
+ "learning_rate": 4.722474358076239e-05,
49
+ "loss": 2.1902,
50
+ "step": 700
51
+ },
52
+ {
53
+ "epoch": 0.23,
54
+ "learning_rate": 3.921546108807772e-05,
55
+ "loss": 2.1833,
56
+ "step": 800
57
+ },
58
+ {
59
+ "epoch": 0.25,
60
+ "learning_rate": 2.786089912498746e-05,
61
+ "loss": 2.1655,
62
+ "step": 900
63
+ },
64
+ {
65
+ "epoch": 0.28,
66
+ "learning_rate": 1.5842522332451548e-05,
67
+ "loss": 2.1736,
68
+ "step": 1000
69
+ },
70
+ {
71
+ "epoch": 0.31,
72
+ "learning_rate": 5.998560182965376e-06,
73
+ "loss": 2.1269,
74
+ "step": 1100
75
+ },
76
+ {
77
+ "epoch": 0.34,
78
+ "learning_rate": 6.537378853569256e-07,
79
+ "loss": 2.1282,
80
+ "step": 1200
81
+ },
82
+ {
83
+ "epoch": 0.37,
84
+ "learning_rate": 4.975624045467085e-05,
85
+ "loss": 2.1398,
86
+ "step": 1300
87
+ },
88
+ {
89
+ "epoch": 0.39,
90
+ "learning_rate": 4.817047144722464e-05,
91
+ "loss": 2.1608,
92
+ "step": 1400
93
+ },
94
+ {
95
+ "epoch": 0.42,
96
+ "learning_rate": 4.519891933012913e-05,
97
+ "loss": 2.1441,
98
+ "step": 1500
99
+ },
100
+ {
101
+ "epoch": 0.45,
102
+ "learning_rate": 4.1019691489417214e-05,
103
+ "loss": 2.1726,
104
+ "step": 1600
105
+ },
106
+ {
107
+ "epoch": 0.48,
108
+ "learning_rate": 3.588328036709013e-05,
109
+ "loss": 2.08,
110
+ "step": 1700
111
+ },
112
+ {
113
+ "epoch": 0.51,
114
+ "learning_rate": 3.009754957197535e-05,
115
+ "loss": 2.1441,
116
+ "step": 1800
117
+ },
118
+ {
119
+ "epoch": 0.54,
120
+ "learning_rate": 2.4009281306639542e-05,
121
+ "loss": 2.121,
122
+ "step": 1900
123
+ },
124
+ {
125
+ "epoch": 0.56,
126
+ "learning_rate": 1.798339111133779e-05,
127
+ "loss": 2.115,
128
+ "step": 2000
129
+ },
130
+ {
131
+ "epoch": 0.56,
132
+ "eval_gen_len": 20.0,
133
+ "eval_loss": 1.985213279724121,
134
+ "eval_rouge1": 21.6401,
135
+ "eval_rouge2": 5.6588,
136
+ "eval_rougeL": 21.4503,
137
+ "eval_rougeLsum": 21.4438,
138
+ "eval_runtime": 165.0934,
139
+ "eval_samples_per_second": 42.449,
140
+ "eval_steps_per_second": 2.653,
141
+ "step": 2000
142
+ },
143
+ {
144
+ "epoch": 0.59,
145
+ "learning_rate": 1.2381055741151043e-05,
146
+ "loss": 2.0857,
147
+ "step": 2100
148
+ },
149
+ {
150
+ "epoch": 0.62,
151
+ "learning_rate": 7.538065136536039e-06,
152
+ "loss": 2.0659,
153
+ "step": 2200
154
+ },
155
+ {
156
+ "epoch": 0.65,
157
+ "learning_rate": 3.7446960159146312e-06,
158
+ "loss": 2.0782,
159
+ "step": 2300
160
+ },
161
+ {
162
+ "epoch": 0.68,
163
+ "learning_rate": 1.2283134167178716e-06,
164
+ "loss": 2.0432,
165
+ "step": 2400
166
+ },
167
+ {
168
+ "epoch": 0.7,
169
+ "learning_rate": 1.397430049077326e-07,
170
+ "loss": 2.0695,
171
+ "step": 2500
172
+ },
173
+ {
174
+ "epoch": 0.73,
175
+ "learning_rate": 4.9888693982206995e-05,
176
+ "loss": 2.1082,
177
+ "step": 2600
178
+ },
179
+ {
180
+ "epoch": 0.76,
181
+ "learning_rate": 4.9413728617356034e-05,
182
+ "loss": 2.1176,
183
+ "step": 2700
184
+ },
185
+ {
186
+ "epoch": 0.79,
187
+ "learning_rate": 4.857231009627857e-05,
188
+ "loss": 2.1424,
189
+ "step": 2800
190
+ },
191
+ {
192
+ "epoch": 0.82,
193
+ "learning_rate": 4.737709413672616e-05,
194
+ "loss": 2.1241,
195
+ "step": 2900
196
+ },
197
+ {
198
+ "epoch": 0.85,
199
+ "learning_rate": 4.5846057899240565e-05,
200
+ "loss": 2.115,
201
+ "step": 3000
202
+ },
203
+ {
204
+ "epoch": 0.87,
205
+ "learning_rate": 4.400222959392605e-05,
206
+ "loss": 2.0883,
207
+ "step": 3100
208
+ },
209
+ {
210
+ "epoch": 0.9,
211
+ "learning_rate": 4.187334211473467e-05,
212
+ "loss": 2.1088,
213
+ "step": 3200
214
+ },
215
+ {
216
+ "epoch": 0.93,
217
+ "learning_rate": 3.949141591092745e-05,
218
+ "loss": 2.0855,
219
+ "step": 3300
220
+ },
221
+ {
222
+ "epoch": 0.96,
223
+ "learning_rate": 3.6892277369712686e-05,
224
+ "loss": 2.0811,
225
+ "step": 3400
226
+ },
227
+ {
228
+ "epoch": 0.99,
229
+ "learning_rate": 3.411501995403511e-05,
230
+ "loss": 2.0632,
231
+ "step": 3500
232
+ },
233
+ {
234
+ "epoch": 1.01,
235
+ "learning_rate": 3.1201416200504895e-05,
236
+ "loss": 1.9746,
237
+ "step": 3600
238
+ },
239
+ {
240
+ "epoch": 1.04,
241
+ "learning_rate": 2.8195289421565e-05,
242
+ "loss": 1.8776,
243
+ "step": 3700
244
+ },
245
+ {
246
+ "epoch": 1.07,
247
+ "learning_rate": 2.5141854562081277e-05,
248
+ "loss": 1.878,
249
+ "step": 3800
250
+ },
251
+ {
252
+ "epoch": 1.1,
253
+ "learning_rate": 2.208703812448537e-05,
254
+ "loss": 1.9068,
255
+ "step": 3900
256
+ },
257
+ {
258
+ "epoch": 1.13,
259
+ "learning_rate": 1.9076787391429783e-05,
260
+ "loss": 1.8943,
261
+ "step": 4000
262
+ },
263
+ {
264
+ "epoch": 1.13,
265
+ "eval_gen_len": 20.0,
266
+ "eval_loss": 1.964289665222168,
267
+ "eval_rouge1": 21.3418,
268
+ "eval_rouge2": 5.6223,
269
+ "eval_rougeL": 21.1288,
270
+ "eval_rougeLsum": 21.1143,
271
+ "eval_runtime": 162.0352,
272
+ "eval_samples_per_second": 43.25,
273
+ "eval_steps_per_second": 2.703,
274
+ "step": 4000
275
+ },
276
+ {
277
+ "epoch": 1.16,
278
+ "learning_rate": 1.6156379335888783e-05,
279
+ "loss": 1.8808,
280
+ "step": 4100
281
+ },
282
+ {
283
+ "epoch": 1.18,
284
+ "learning_rate": 1.3369739613341079e-05,
285
+ "loss": 1.8741,
286
+ "step": 4200
287
+ },
288
+ {
289
+ "epoch": 1.21,
290
+ "learning_rate": 1.0758781879025238e-05,
291
+ "loss": 1.86,
292
+ "step": 4300
293
+ },
294
+ {
295
+ "epoch": 1.24,
296
+ "learning_rate": 8.362777367551548e-06,
297
+ "loss": 1.8857,
298
+ "step": 4400
299
+ },
300
+ {
301
+ "epoch": 1.27,
302
+ "learning_rate": 6.2177642169797586e-06,
303
+ "loss": 1.8757,
304
+ "step": 4500
305
+ },
306
+ {
307
+ "epoch": 1.3,
308
+ "learning_rate": 4.356005421678249e-06,
309
+ "loss": 1.8509,
310
+ "step": 4600
311
+ },
312
+ {
313
+ "epoch": 1.32,
314
+ "learning_rate": 2.8055035668582452e-06,
315
+ "loss": 1.8304,
316
+ "step": 4700
317
+ },
318
+ {
319
+ "epoch": 1.35,
320
+ "learning_rate": 1.589579643627363e-06,
321
+ "loss": 1.8435,
322
+ "step": 4800
323
+ },
324
+ {
325
+ "epoch": 1.38,
326
+ "learning_rate": 7.265222795761219e-07,
327
+ "loss": 1.888,
328
+ "step": 4900
329
+ },
330
+ {
331
+ "epoch": 1.41,
332
+ "learning_rate": 2.2931266079587017e-07,
333
+ "loss": 1.8392,
334
+ "step": 5000
335
+ },
336
+ {
337
+ "epoch": 1.44,
338
+ "learning_rate": 4.99986426423885e-05,
339
+ "loss": 1.8821,
340
+ "step": 5100
341
+ },
342
+ {
343
+ "epoch": 1.47,
344
+ "learning_rate": 4.9935733363756125e-05,
345
+ "loss": 1.9121,
346
+ "step": 5200
347
+ },
348
+ {
349
+ "epoch": 1.49,
350
+ "learning_rate": 4.9779159803948603e-05,
351
+ "loss": 1.8862,
352
+ "step": 5300
353
+ },
354
+ {
355
+ "epoch": 1.52,
356
+ "learning_rate": 4.952951127049067e-05,
357
+ "loss": 1.9367,
358
+ "step": 5400
359
+ },
360
+ {
361
+ "epoch": 1.55,
362
+ "learning_rate": 4.9187727384091566e-05,
363
+ "loss": 1.9389,
364
+ "step": 5500
365
+ },
366
+ {
367
+ "epoch": 1.58,
368
+ "learning_rate": 4.875509454212479e-05,
369
+ "loss": 1.917,
370
+ "step": 5600
371
+ },
372
+ {
373
+ "epoch": 1.61,
374
+ "learning_rate": 4.823324107691887e-05,
375
+ "loss": 1.9325,
376
+ "step": 5700
377
+ },
378
+ {
379
+ "epoch": 1.63,
380
+ "learning_rate": 4.762413112708219e-05,
381
+ "loss": 1.939,
382
+ "step": 5800
383
+ },
384
+ {
385
+ "epoch": 1.66,
386
+ "learning_rate": 4.6930057244928707e-05,
387
+ "loss": 1.9489,
388
+ "step": 5900
389
+ },
390
+ {
391
+ "epoch": 1.69,
392
+ "learning_rate": 4.6153631767828776e-05,
393
+ "loss": 1.9567,
394
+ "step": 6000
395
+ },
396
+ {
397
+ "epoch": 1.69,
398
+ "eval_gen_len": 20.0,
399
+ "eval_loss": 1.9752851724624634,
400
+ "eval_rouge1": 21.8045,
401
+ "eval_rouge2": 5.6422,
402
+ "eval_rougeL": 21.628,
403
+ "eval_rougeLsum": 21.6131,
404
+ "eval_runtime": 162.1464,
405
+ "eval_samples_per_second": 43.22,
406
+ "eval_steps_per_second": 2.701,
407
+ "step": 6000
408
+ },
409
+ {
410
+ "epoch": 1.72,
411
+ "learning_rate": 4.529777698596109e-05,
412
+ "loss": 1.9347,
413
+ "step": 6100
414
+ },
415
+ {
416
+ "epoch": 1.75,
417
+ "learning_rate": 4.4365714143472324e-05,
418
+ "loss": 1.9325,
419
+ "step": 6200
420
+ },
421
+ {
422
+ "epoch": 1.78,
423
+ "learning_rate": 4.3360951314441554e-05,
424
+ "loss": 1.9184,
425
+ "step": 6300
426
+ },
427
+ {
428
+ "epoch": 1.8,
429
+ "learning_rate": 4.228727019928149e-05,
430
+ "loss": 1.919,
431
+ "step": 6400
432
+ },
433
+ {
434
+ "epoch": 1.83,
435
+ "learning_rate": 4.114871189127209e-05,
436
+ "loss": 1.9375,
437
+ "step": 6500
438
+ },
439
+ {
440
+ "epoch": 1.86,
441
+ "learning_rate": 3.994956166679789e-05,
442
+ "loss": 1.9609,
443
+ "step": 6600
444
+ },
445
+ {
446
+ "epoch": 1.89,
447
+ "learning_rate": 3.869433285653533e-05,
448
+ "loss": 1.9409,
449
+ "step": 6700
450
+ },
451
+ {
452
+ "epoch": 1.92,
453
+ "learning_rate": 3.73877498582952e-05,
454
+ "loss": 1.9153,
455
+ "step": 6800
456
+ },
457
+ {
458
+ "epoch": 1.94,
459
+ "learning_rate": 3.603473035545605e-05,
460
+ "loss": 1.918,
461
+ "step": 6900
462
+ },
463
+ {
464
+ "epoch": 1.97,
465
+ "learning_rate": 3.4640366807914365e-05,
466
+ "loss": 1.9425,
467
+ "step": 7000
468
+ },
469
+ {
470
+ "epoch": 2.0,
471
+ "learning_rate": 3.320990728521537e-05,
472
+ "loss": 1.918,
473
+ "step": 7100
474
+ },
475
+ {
476
+ "epoch": 2.03,
477
+ "learning_rate": 3.174873571400432e-05,
478
+ "loss": 1.7143,
479
+ "step": 7200
480
+ },
481
+ {
482
+ "epoch": 2.06,
483
+ "learning_rate": 3.0262351614142278e-05,
484
+ "loss": 1.7251,
485
+ "step": 7300
486
+ },
487
+ {
488
+ "epoch": 2.09,
489
+ "learning_rate": 2.875634939975517e-05,
490
+ "loss": 1.7188,
491
+ "step": 7400
492
+ },
493
+ {
494
+ "epoch": 2.11,
495
+ "learning_rate": 2.7236397323122172e-05,
496
+ "loss": 1.732,
497
+ "step": 7500
498
+ },
499
+ {
500
+ "epoch": 2.14,
501
+ "learning_rate": 2.570821614065411e-05,
502
+ "loss": 1.7565,
503
+ "step": 7600
504
+ },
505
+ {
506
+ "epoch": 2.17,
507
+ "learning_rate": 2.4177557581258026e-05,
508
+ "loss": 1.7159,
509
+ "step": 7700
510
+ },
511
+ {
512
+ "epoch": 2.2,
513
+ "learning_rate": 2.2650182698128346e-05,
514
+ "loss": 1.7191,
515
+ "step": 7800
516
+ },
517
+ {
518
+ "epoch": 2.23,
519
+ "learning_rate": 2.1131840185443426e-05,
520
+ "loss": 1.7215,
521
+ "step": 7900
522
+ },
523
+ {
524
+ "epoch": 2.25,
525
+ "learning_rate": 1.962824474157885e-05,
526
+ "loss": 1.7129,
527
+ "step": 8000
528
+ },
529
+ {
530
+ "epoch": 2.25,
531
+ "eval_gen_len": 20.0,
532
+ "eval_loss": 1.9584242105484009,
533
+ "eval_rouge1": 22.117,
534
+ "eval_rouge2": 5.7798,
535
+ "eval_rougeL": 21.9114,
536
+ "eval_rougeLsum": 21.8886,
537
+ "eval_runtime": 162.6289,
538
+ "eval_samples_per_second": 43.092,
539
+ "eval_steps_per_second": 2.693,
540
+ "step": 8000
541
+ },
542
+ {
543
+ "epoch": 2.28,
544
+ "learning_rate": 1.8145055560273016e-05,
545
+ "loss": 1.6991,
546
+ "step": 8100
547
+ },
548
+ {
549
+ "epoch": 2.31,
550
+ "learning_rate": 1.6687855030699798e-05,
551
+ "loss": 1.7131,
552
+ "step": 8200
553
+ },
554
+ {
555
+ "epoch": 2.34,
556
+ "learning_rate": 1.52621277266162e-05,
557
+ "loss": 1.7091,
558
+ "step": 8300
559
+ },
560
+ {
561
+ "epoch": 2.37,
562
+ "learning_rate": 1.3873239763665138e-05,
563
+ "loss": 1.7208,
564
+ "step": 8400
565
+ },
566
+ {
567
+ "epoch": 2.4,
568
+ "learning_rate": 1.2526418602527839e-05,
569
+ "loss": 1.7373,
570
+ "step": 8500
571
+ },
572
+ {
573
+ "epoch": 2.42,
574
+ "learning_rate": 1.1226733373942214e-05,
575
+ "loss": 1.6905,
576
+ "step": 8600
577
+ },
578
+ {
579
+ "epoch": 2.45,
580
+ "learning_rate": 9.979075799639335e-06,
581
+ "loss": 1.7103,
582
+ "step": 8700
583
+ },
584
+ {
585
+ "epoch": 2.48,
586
+ "learning_rate": 8.788141781007439e-06,
587
+ "loss": 1.7381,
588
+ "step": 8800
589
+ },
590
+ {
591
+ "epoch": 2.51,
592
+ "learning_rate": 7.658413724779386e-06,
593
+ "loss": 1.7005,
594
+ "step": 8900
595
+ },
596
+ {
597
+ "epoch": 2.54,
598
+ "learning_rate": 6.594143672265889e-06,
599
+ "loss": 1.692,
600
+ "step": 9000
601
+ },
602
+ {
603
+ "epoch": 2.56,
604
+ "learning_rate": 5.599337295632133e-06,
605
+ "loss": 1.6929,
606
+ "step": 9100
607
+ },
608
+ {
609
+ "epoch": 2.59,
610
+ "learning_rate": 4.677738821452162e-06,
611
+ "loss": 1.7245,
612
+ "step": 9200
613
+ },
614
+ {
615
+ "epoch": 2.62,
616
+ "learning_rate": 3.832816938285592e-06,
617
+ "loss": 1.6818,
618
+ "step": 9300
619
+ },
620
+ {
621
+ "epoch": 2.65,
622
+ "learning_rate": 3.0677517413170523e-06,
623
+ "loss": 1.7,
624
+ "step": 9400
625
+ },
626
+ {
627
+ "epoch": 2.68,
628
+ "learning_rate": 2.385422763195976e-06,
629
+ "loss": 1.6961,
630
+ "step": 9500
631
+ },
632
+ {
633
+ "epoch": 2.71,
634
+ "learning_rate": 1.7883981361259272e-06,
635
+ "loss": 1.7067,
636
+ "step": 9600
637
+ },
638
+ {
639
+ "epoch": 2.73,
640
+ "learning_rate": 1.2789249259948616e-06,
641
+ "loss": 1.6975,
642
+ "step": 9700
643
+ },
644
+ {
645
+ "epoch": 2.76,
646
+ "learning_rate": 8.589206749264998e-07,
647
+ "loss": 1.7217,
648
+ "step": 9800
649
+ },
650
+ {
651
+ "epoch": 2.79,
652
+ "learning_rate": 5.29966184084788e-07,
653
+ "loss": 1.7213,
654
+ "step": 9900
655
+ },
656
+ {
657
+ "epoch": 2.82,
658
+ "learning_rate": 2.932995638952263e-07,
659
+ "loss": 1.7072,
660
+ "step": 10000
661
+ },
662
+ {
663
+ "epoch": 2.82,
664
+ "eval_gen_len": 20.0,
665
+ "eval_loss": 1.933449387550354,
666
+ "eval_rouge1": 22.3137,
667
+ "eval_rouge2": 5.8376,
668
+ "eval_rougeL": 22.1264,
669
+ "eval_rougeLsum": 22.1274,
670
+ "eval_runtime": 162.4524,
671
+ "eval_samples_per_second": 43.139,
672
+ "eval_steps_per_second": 2.696,
673
+ "step": 10000
674
+ },
675
+ {
676
+ "epoch": 2.85,
677
+ "learning_rate": 1.5078315448915942e-07,
678
+ "loss": 1.6934,
679
+ "step": 10100
680
+ },
681
+ {
682
+ "epoch": 2.87,
683
+ "learning_rate": 1.0007514846990862e-07,
684
+ "loss": 1.6952,
685
+ "step": 10200
686
+ },
687
+ {
688
+ "epoch": 2.9,
689
+ "learning_rate": 4.998917939712543e-05,
690
+ "loss": 1.757,
691
+ "step": 10300
692
+ },
693
+ {
694
+ "epoch": 2.93,
695
+ "learning_rate": 4.995490569865264e-05,
696
+ "loss": 1.7591,
697
+ "step": 10400
698
+ },
699
+ {
700
+ "epoch": 2.96,
701
+ "learning_rate": 4.989719237466243e-05,
702
+ "loss": 1.7812,
703
+ "step": 10500
704
+ },
705
+ {
706
+ "epoch": 2.99,
707
+ "learning_rate": 4.9816093742915244e-05,
708
+ "loss": 1.7639,
709
+ "step": 10600
710
+ },
711
+ {
712
+ "epoch": 3.02,
713
+ "learning_rate": 4.9711686130604656e-05,
714
+ "loss": 1.7264,
715
+ "step": 10700
716
+ },
717
+ {
718
+ "epoch": 3.04,
719
+ "learning_rate": 4.9584067802520866e-05,
720
+ "loss": 1.6739,
721
+ "step": 10800
722
+ },
723
+ {
724
+ "epoch": 3.07,
725
+ "learning_rate": 4.943497980862275e-05,
726
+ "loss": 1.6629,
727
+ "step": 10900
728
+ },
729
+ {
730
+ "epoch": 3.1,
731
+ "learning_rate": 4.926155082525451e-05,
732
+ "loss": 1.6756,
733
+ "step": 11000
734
+ },
735
+ {
736
+ "epoch": 3.13,
737
+ "learning_rate": 4.906533477770676e-05,
738
+ "loss": 1.6823,
739
+ "step": 11100
740
+ },
741
+ {
742
+ "epoch": 3.16,
743
+ "learning_rate": 4.884651633765546e-05,
744
+ "loss": 1.6432,
745
+ "step": 11200
746
+ },
747
+ {
748
+ "epoch": 3.18,
749
+ "learning_rate": 4.860530144935725e-05,
750
+ "loss": 1.6803,
751
+ "step": 11300
752
+ },
753
+ {
754
+ "epoch": 3.21,
755
+ "learning_rate": 4.834191713582197e-05,
756
+ "loss": 1.6985,
757
+ "step": 11400
758
+ },
759
+ {
760
+ "epoch": 3.24,
761
+ "learning_rate": 4.8056611285146466e-05,
762
+ "loss": 1.7087,
763
+ "step": 11500
764
+ },
765
+ {
766
+ "epoch": 3.27,
767
+ "learning_rate": 4.7749652417211076e-05,
768
+ "loss": 1.7142,
769
+ "step": 11600
770
+ },
771
+ {
772
+ "epoch": 3.3,
773
+ "learning_rate": 4.742132943095807e-05,
774
+ "loss": 1.7135,
775
+ "step": 11700
776
+ },
777
+ {
778
+ "epoch": 3.33,
779
+ "learning_rate": 4.70719513324901e-05,
780
+ "loss": 1.7136,
781
+ "step": 11800
782
+ },
783
+ {
784
+ "epoch": 3.35,
785
+ "learning_rate": 4.670184694424442e-05,
786
+ "loss": 1.7098,
787
+ "step": 11900
788
+ },
789
+ {
790
+ "epoch": 3.38,
791
+ "learning_rate": 4.63113645955168e-05,
792
+ "loss": 1.7184,
793
+ "step": 12000
794
+ },
795
+ {
796
+ "epoch": 3.38,
797
+ "eval_gen_len": 20.0,
798
+ "eval_loss": 2.0014004707336426,
799
+ "eval_rouge1": 21.452,
800
+ "eval_rouge2": 5.494,
801
+ "eval_rougeL": 21.3143,
802
+ "eval_rougeLsum": 21.3158,
803
+ "eval_runtime": 162.8289,
804
+ "eval_samples_per_second": 43.039,
805
+ "eval_steps_per_second": 2.69,
806
+ "step": 12000
807
+ },
808
+ {
809
+ "epoch": 3.41,
810
+ "learning_rate": 4.590087179462613e-05,
811
+ "loss": 1.7121,
812
+ "step": 12100
813
+ },
814
+ {
815
+ "epoch": 3.44,
816
+ "learning_rate": 4.547075488302845e-05,
817
+ "loss": 1.7064,
818
+ "step": 12200
819
+ },
820
+ {
821
+ "epoch": 3.47,
822
+ "learning_rate": 4.502141867170594e-05,
823
+ "loss": 1.7182,
824
+ "step": 12300
825
+ },
826
+ {
827
+ "epoch": 3.49,
828
+ "learning_rate": 4.455328606017294e-05,
829
+ "loss": 1.7126,
830
+ "step": 12400
831
+ },
832
+ {
833
+ "epoch": 3.52,
834
+ "learning_rate": 4.4066797638457747e-05,
835
+ "loss": 1.7135,
836
+ "step": 12500
837
+ },
838
+ {
839
+ "epoch": 3.55,
840
+ "learning_rate": 4.3562411272434736e-05,
841
+ "loss": 1.7224,
842
+ "step": 12600
843
+ },
844
+ {
845
+ "epoch": 3.58,
846
+ "learning_rate": 4.304060167289698e-05,
847
+ "loss": 1.7393,
848
+ "step": 12700
849
+ },
850
+ {
851
+ "epoch": 3.61,
852
+ "learning_rate": 4.250185994877508e-05,
853
+ "loss": 1.7153,
854
+ "step": 12800
855
+ },
856
+ {
857
+ "epoch": 3.64,
858
+ "learning_rate": 4.194669314492266e-05,
859
+ "loss": 1.6978,
860
+ "step": 12900
861
+ },
862
+ {
863
+ "epoch": 3.66,
864
+ "learning_rate": 4.137562376490343e-05,
865
+ "loss": 1.7238,
866
+ "step": 13000
867
+ },
868
+ {
869
+ "epoch": 3.69,
870
+ "learning_rate": 4.07891892792292e-05,
871
+ "loss": 1.73,
872
+ "step": 13100
873
+ },
874
+ {
875
+ "epoch": 3.72,
876
+ "learning_rate": 4.0187941619511394e-05,
877
+ "loss": 1.7077,
878
+ "step": 13200
879
+ },
880
+ {
881
+ "epoch": 3.75,
882
+ "learning_rate": 3.957244665900235e-05,
883
+ "loss": 1.696,
884
+ "step": 13300
885
+ },
886
+ {
887
+ "epoch": 3.78,
888
+ "learning_rate": 3.8943283680015245e-05,
889
+ "loss": 1.7197,
890
+ "step": 13400
891
+ },
892
+ {
893
+ "epoch": 3.8,
894
+ "learning_rate": 3.830104482872382e-05,
895
+ "loss": 1.755,
896
+ "step": 13500
897
+ },
898
+ {
899
+ "epoch": 3.83,
900
+ "learning_rate": 3.764633455785513e-05,
901
+ "loss": 1.7267,
902
+ "step": 13600
903
+ },
904
+ {
905
+ "epoch": 3.86,
906
+ "learning_rate": 3.697976905779984e-05,
907
+ "loss": 1.7325,
908
+ "step": 13700
909
+ },
910
+ {
911
+ "epoch": 3.89,
912
+ "learning_rate": 3.630197567667534e-05,
913
+ "loss": 1.7396,
914
+ "step": 13800
915
+ },
916
+ {
917
+ "epoch": 3.92,
918
+ "learning_rate": 3.5613592329887714e-05,
919
+ "loss": 1.7154,
920
+ "step": 13900
921
+ },
922
+ {
923
+ "epoch": 3.95,
924
+ "learning_rate": 3.491526689974802e-05,
925
+ "loss": 1.7372,
926
+ "step": 14000
927
+ },
928
+ {
929
+ "epoch": 3.95,
930
+ "eval_gen_len": 20.0,
931
+ "eval_loss": 1.9770143032073975,
932
+ "eval_rouge1": 22.273,
933
+ "eval_rouge2": 5.7637,
934
+ "eval_rougeL": 22.1,
935
+ "eval_rougeLsum": 22.0813,
936
+ "eval_runtime": 163.6202,
937
+ "eval_samples_per_second": 42.831,
938
+ "eval_steps_per_second": 2.677,
939
+ "step": 14000
940
+ },
941
+ {
942
+ "epoch": 3.97,
943
+ "learning_rate": 3.42076566257082e-05,
944
+ "loss": 1.756,
945
+ "step": 14100
946
+ },
947
+ {
948
+ "epoch": 4.0,
949
+ "learning_rate": 3.349142748579026e-05,
950
+ "loss": 1.7312,
951
+ "step": 14200
952
+ },
953
+ {
954
+ "epoch": 4.03,
955
+ "learning_rate": 3.276725356979111e-05,
956
+ "loss": 1.4431,
957
+ "step": 14300
958
+ },
959
+ {
960
+ "epoch": 4.06,
961
+ "learning_rate": 3.2035816444852827e-05,
962
+ "loss": 1.4696,
963
+ "step": 14400
964
+ },
965
+ {
966
+ "epoch": 4.09,
967
+ "learning_rate": 3.1297804513995484e-05,
968
+ "loss": 1.4713,
969
+ "step": 14500
970
+ },
971
+ {
972
+ "epoch": 4.11,
973
+ "learning_rate": 3.055391236821639e-05,
974
+ "loss": 1.4552,
975
+ "step": 14600
976
+ },
977
+ {
978
+ "epoch": 4.14,
979
+ "learning_rate": 2.9804840132765305e-05,
980
+ "loss": 1.4596,
981
+ "step": 14700
982
+ },
983
+ {
984
+ "epoch": 4.17,
985
+ "learning_rate": 2.905129280821106e-05,
986
+ "loss": 1.4558,
987
+ "step": 14800
988
+ },
989
+ {
990
+ "epoch": 4.2,
991
+ "learning_rate": 2.8293979606919723e-05,
992
+ "loss": 1.4785,
993
+ "step": 14900
994
+ },
995
+ {
996
+ "epoch": 4.23,
997
+ "learning_rate": 2.7533613285568734e-05,
998
+ "loss": 1.4652,
999
+ "step": 15000
1000
+ },
1001
+ {
1002
+ "epoch": 4.26,
1003
+ "learning_rate": 2.6778545727506706e-05,
1004
+ "loss": 1.4952,
1005
+ "step": 15100
1006
+ },
1007
+ {
1008
+ "epoch": 4.28,
1009
+ "learning_rate": 2.6014234894273854e-05,
1010
+ "loss": 1.4794,
1011
+ "step": 15200
1012
+ },
1013
+ {
1014
+ "epoch": 4.31,
1015
+ "learning_rate": 2.5249016556908847e-05,
1016
+ "loss": 1.4764,
1017
+ "step": 15300
1018
+ },
1019
+ {
1020
+ "epoch": 4.34,
1021
+ "learning_rate": 2.448361091212649e-05,
1022
+ "loss": 1.4685,
1023
+ "step": 15400
1024
+ },
1025
+ {
1026
+ "epoch": 4.37,
1027
+ "learning_rate": 2.3718738332928784e-05,
1028
+ "loss": 1.4858,
1029
+ "step": 15500
1030
+ },
1031
+ {
1032
+ "epoch": 4.4,
1033
+ "learning_rate": 2.295511869061505e-05,
1034
+ "loss": 1.4953,
1035
+ "step": 15600
1036
+ },
1037
+ {
1038
+ "epoch": 4.43,
1039
+ "learning_rate": 2.219347067726429e-05,
1040
+ "loss": 1.4836,
1041
+ "step": 15700
1042
+ },
1043
+ {
1044
+ "epoch": 4.45,
1045
+ "learning_rate": 2.143451112932731e-05,
1046
+ "loss": 1.4926,
1047
+ "step": 15800
1048
+ },
1049
+ {
1050
+ "epoch": 4.48,
1051
+ "learning_rate": 2.0678954352965396e-05,
1052
+ "loss": 1.4652,
1053
+ "step": 15900
1054
+ },
1055
+ {
1056
+ "epoch": 4.51,
1057
+ "learning_rate": 1.992751145177032e-05,
1058
+ "loss": 1.4724,
1059
+ "step": 16000
1060
+ },
1061
+ {
1062
+ "epoch": 4.51,
1063
+ "eval_gen_len": 20.0,
1064
+ "eval_loss": 2.019371509552002,
1065
+ "eval_rouge1": 22.0979,
1066
+ "eval_rouge2": 5.7144,
1067
+ "eval_rougeL": 21.9045,
1068
+ "eval_rougeLsum": 21.8765,
1069
+ "eval_runtime": 162.8312,
1070
+ "eval_samples_per_second": 43.038,
1071
+ "eval_steps_per_second": 2.69,
1072
+ "step": 16000
1073
+ },
1074
+ {
1075
+ "epoch": 4.54,
1076
+ "learning_rate": 1.9180889657498532e-05,
1077
+ "loss": 1.4723,
1078
+ "step": 16100
1079
+ },
1080
+ {
1081
+ "epoch": 4.57,
1082
+ "learning_rate": 1.843979166444942e-05,
1083
+ "loss": 1.475,
1084
+ "step": 16200
1085
+ },
1086
+ {
1087
+ "epoch": 4.59,
1088
+ "learning_rate": 1.770491496811398e-05,
1089
+ "loss": 1.4657,
1090
+ "step": 16300
1091
+ },
1092
+ {
1093
+ "epoch": 4.62,
1094
+ "learning_rate": 1.6976951208716527e-05,
1095
+ "loss": 1.466,
1096
+ "step": 16400
1097
+ },
1098
+ {
1099
+ "epoch": 4.65,
1100
+ "learning_rate": 1.625658552026706e-05,
1101
+ "loss": 1.4649,
1102
+ "step": 16500
1103
+ },
1104
+ {
1105
+ "epoch": 4.68,
1106
+ "learning_rate": 1.554449588573719e-05,
1107
+ "loss": 1.4792,
1108
+ "step": 16600
1109
+ },
1110
+ {
1111
+ "epoch": 4.71,
1112
+ "learning_rate": 1.4841352498966237e-05,
1113
+ "loss": 1.4477,
1114
+ "step": 16700
1115
+ },
1116
+ {
1117
+ "epoch": 4.74,
1118
+ "learning_rate": 1.4147817133898276e-05,
1119
+ "loss": 1.4744,
1120
+ "step": 16800
1121
+ },
1122
+ {
1123
+ "epoch": 4.76,
1124
+ "learning_rate": 1.346454252174365e-05,
1125
+ "loss": 1.4915,
1126
+ "step": 16900
1127
+ },
1128
+ {
1129
+ "epoch": 4.79,
1130
+ "learning_rate": 1.2792171736651217e-05,
1131
+ "loss": 1.4767,
1132
+ "step": 17000
1133
+ },
1134
+ {
1135
+ "epoch": 4.82,
1136
+ "learning_rate": 1.213133759046946e-05,
1137
+ "loss": 1.483,
1138
+ "step": 17100
1139
+ },
1140
+ {
1141
+ "epoch": 4.85,
1142
+ "learning_rate": 1.148908659327833e-05,
1143
+ "loss": 1.4579,
1144
+ "step": 17200
1145
+ },
1146
+ {
1147
+ "epoch": 4.88,
1148
+ "learning_rate": 1.0853049469504679e-05,
1149
+ "loss": 1.4791,
1150
+ "step": 17300
1151
+ },
1152
+ {
1153
+ "epoch": 4.9,
1154
+ "learning_rate": 1.0230374018646024e-05,
1155
+ "loss": 1.4593,
1156
+ "step": 17400
1157
+ },
1158
+ {
1159
+ "epoch": 4.93,
1160
+ "learning_rate": 9.621646281041709e-06,
1161
+ "loss": 1.4629,
1162
+ "step": 17500
1163
+ },
1164
+ {
1165
+ "epoch": 4.96,
1166
+ "learning_rate": 9.027439169931561e-06,
1167
+ "loss": 1.4677,
1168
+ "step": 17600
1169
+ },
1170
+ {
1171
+ "epoch": 4.99,
1172
+ "learning_rate": 8.448311932250029e-06,
1173
+ "loss": 1.4382,
1174
+ "step": 17700
1175
+ },
1176
+ {
1177
+ "epoch": 5.02,
1178
+ "learning_rate": 7.884809622282473e-06,
1179
+ "loss": 1.369,
1180
+ "step": 17800
1181
+ },
1182
+ {
1183
+ "epoch": 5.05,
1184
+ "learning_rate": 7.337462588679232e-06,
1185
+ "loss": 1.3049,
1186
+ "step": 17900
1187
+ },
1188
+ {
1189
+ "epoch": 5.07,
1190
+ "learning_rate": 6.806785975309991e-06,
1191
+ "loss": 1.3297,
1192
+ "step": 18000
1193
+ },
1194
+ {
1195
+ "epoch": 5.07,
1196
+ "eval_gen_len": 20.0,
1197
+ "eval_loss": 2.037019729614258,
1198
+ "eval_rouge1": 21.9,
1199
+ "eval_rouge2": 5.533,
1200
+ "eval_rougeL": 21.7042,
1201
+ "eval_rougeLsum": 21.7021,
1202
+ "eval_runtime": 162.394,
1203
+ "eval_samples_per_second": 43.154,
1204
+ "eval_steps_per_second": 2.697,
1205
+ "step": 18000
1206
+ },
1207
+ {
1208
+ "epoch": 5.1,
1209
+ "learning_rate": 6.293279236428415e-06,
1210
+ "loss": 1.3197,
1211
+ "step": 18100
1212
+ },
1213
+ {
1214
+ "epoch": 5.13,
1215
+ "learning_rate": 5.7974256666032834e-06,
1216
+ "loss": 1.2993,
1217
+ "step": 18200
1218
+ },
1219
+ {
1220
+ "epoch": 5.16,
1221
+ "learning_rate": 5.31969194585855e-06,
1222
+ "loss": 1.3202,
1223
+ "step": 18300
1224
+ },
1225
+ {
1226
+ "epoch": 5.19,
1227
+ "learning_rate": 4.8605277004504475e-06,
1228
+ "loss": 1.2897,
1229
+ "step": 18400
1230
+ },
1231
+ {
1232
+ "epoch": 5.21,
1233
+ "learning_rate": 4.42036507969501e-06,
1234
+ "loss": 1.3157,
1235
+ "step": 18500
1236
+ },
1237
+ {
1238
+ "epoch": 5.24,
1239
+ "learning_rate": 3.999618349244242e-06,
1240
+ "loss": 1.3208,
1241
+ "step": 18600
1242
+ },
1243
+ {
1244
+ "epoch": 5.27,
1245
+ "learning_rate": 3.5986835011937814e-06,
1246
+ "loss": 1.3238,
1247
+ "step": 18700
1248
+ },
1249
+ {
1250
+ "epoch": 5.3,
1251
+ "learning_rate": 3.2179378813890276e-06,
1252
+ "loss": 1.3285,
1253
+ "step": 18800
1254
+ },
1255
+ {
1256
+ "epoch": 5.33,
1257
+ "learning_rate": 2.857739834280396e-06,
1258
+ "loss": 1.3075,
1259
+ "step": 18900
1260
+ },
1261
+ {
1262
+ "epoch": 5.36,
1263
+ "learning_rate": 2.5184283656621174e-06,
1264
+ "loss": 1.3187,
1265
+ "step": 19000
1266
+ },
1267
+ {
1268
+ "epoch": 5.38,
1269
+ "learning_rate": 2.2003228236118087e-06,
1270
+ "loss": 1.3019,
1271
+ "step": 19100
1272
+ },
1273
+ {
1274
+ "epoch": 5.41,
1275
+ "learning_rate": 1.9037225979312562e-06,
1276
+ "loss": 1.3156,
1277
+ "step": 19200
1278
+ },
1279
+ {
1280
+ "epoch": 5.44,
1281
+ "learning_rate": 1.631546296469755e-06,
1282
+ "loss": 1.2991,
1283
+ "step": 19300
1284
+ },
1285
+ {
1286
+ "epoch": 5.47,
1287
+ "learning_rate": 1.3785520064985022e-06,
1288
+ "loss": 1.2969,
1289
+ "step": 19400
1290
+ },
1291
+ {
1292
+ "epoch": 5.5,
1293
+ "learning_rate": 1.1478364548263003e-06,
1294
+ "loss": 1.3149,
1295
+ "step": 19500
1296
+ },
1297
+ {
1298
+ "epoch": 5.52,
1299
+ "learning_rate": 9.396167828548325e-07,
1300
+ "loss": 1.2895,
1301
+ "step": 19600
1302
+ },
1303
+ {
1304
+ "epoch": 5.55,
1305
+ "learning_rate": 7.540889596512168e-07,
1306
+ "loss": 1.3152,
1307
+ "step": 19700
1308
+ },
1309
+ {
1310
+ "epoch": 5.58,
1311
+ "learning_rate": 5.914275975087762e-07,
1312
+ "loss": 1.3272,
1313
+ "step": 19800
1314
+ },
1315
+ {
1316
+ "epoch": 5.61,
1317
+ "learning_rate": 4.517857876080481e-07,
1318
+ "loss": 1.3142,
1319
+ "step": 19900
1320
+ },
1321
+ {
1322
+ "epoch": 5.64,
1323
+ "learning_rate": 3.352949559327476e-07,
1324
+ "loss": 1.3186,
1325
+ "step": 20000
1326
+ },
1327
+ {
1328
+ "epoch": 5.64,
1329
+ "eval_gen_len": 20.0,
1330
+ "eval_loss": 2.039311170578003,
1331
+ "eval_rouge1": 21.585,
1332
+ "eval_rouge2": 5.4626,
1333
+ "eval_rougeL": 21.4273,
1334
+ "eval_rougeLsum": 21.4329,
1335
+ "eval_runtime": 162.2232,
1336
+ "eval_samples_per_second": 43.2,
1337
+ "eval_steps_per_second": 2.7,
1338
+ "step": 20000
1339
+ },
1340
+ {
1341
+ "epoch": 5.67,
1342
+ "learning_rate": 2.4206473957621197e-07,
1343
+ "loss": 1.3001,
1344
+ "step": 20100
1345
+ },
1346
+ {
1347
+ "epoch": 5.69,
1348
+ "learning_rate": 1.721828835548135e-07,
1349
+ "loss": 1.3033,
1350
+ "step": 20200
1351
+ },
1352
+ {
1353
+ "epoch": 5.72,
1354
+ "learning_rate": 1.2571515822542734e-07,
1355
+ "loss": 1.3053,
1356
+ "step": 20300
1357
+ },
1358
+ {
1359
+ "epoch": 5.75,
1360
+ "learning_rate": 1.0270529738465452e-07,
1361
+ "loss": 1.3123,
1362
+ "step": 20400
1363
+ },
1364
+ {
1365
+ "epoch": 5.78,
1366
+ "learning_rate": 4.9999206248096876e-05,
1367
+ "loss": 1.3399,
1368
+ "step": 20500
1369
+ },
1370
+ {
1371
+ "epoch": 5.81,
1372
+ "learning_rate": 4.999321815444586e-05,
1373
+ "loss": 1.3654,
1374
+ "step": 20600
1375
+ },
1376
+ {
1377
+ "epoch": 5.83,
1378
+ "learning_rate": 4.9981360794562834e-05,
1379
+ "loss": 1.3834,
1380
+ "step": 20700
1381
+ },
1382
+ {
1383
+ "epoch": 5.86,
1384
+ "learning_rate": 4.996363695854496e-05,
1385
+ "loss": 1.4284,
1386
+ "step": 20800
1387
+ },
1388
+ {
1389
+ "epoch": 5.89,
1390
+ "learning_rate": 4.994005081690109e-05,
1391
+ "loss": 1.4216,
1392
+ "step": 20900
1393
+ },
1394
+ {
1395
+ "epoch": 5.92,
1396
+ "learning_rate": 4.991060791957044e-05,
1397
+ "loss": 1.4262,
1398
+ "step": 21000
1399
+ },
1400
+ {
1401
+ "epoch": 5.95,
1402
+ "learning_rate": 4.987531519461667e-05,
1403
+ "loss": 1.4331,
1404
+ "step": 21100
1405
+ },
1406
+ {
1407
+ "epoch": 5.98,
1408
+ "learning_rate": 4.983418094659765e-05,
1409
+ "loss": 1.4425,
1410
+ "step": 21200
1411
+ },
1412
+ {
1413
+ "epoch": 6.0,
1414
+ "learning_rate": 4.978721485461138e-05,
1415
+ "loss": 1.4284,
1416
+ "step": 21300
1417
+ },
1418
+ {
1419
+ "epoch": 6.03,
1420
+ "learning_rate": 4.973554067339494e-05,
1421
+ "loss": 1.3686,
1422
+ "step": 21400
1423
+ },
1424
+ {
1425
+ "epoch": 6.06,
1426
+ "learning_rate": 4.967706145407849e-05,
1427
+ "loss": 1.3545,
1428
+ "step": 21500
1429
+ },
1430
+ {
1431
+ "epoch": 6.09,
1432
+ "learning_rate": 4.961278736181136e-05,
1433
+ "loss": 1.3899,
1434
+ "step": 21600
1435
+ },
1436
+ {
1437
+ "epoch": 6.12,
1438
+ "learning_rate": 4.954273352061473e-05,
1439
+ "loss": 1.3801,
1440
+ "step": 21700
1441
+ },
1442
+ {
1443
+ "epoch": 6.14,
1444
+ "learning_rate": 4.946691641451411e-05,
1445
+ "loss": 1.3902,
1446
+ "step": 21800
1447
+ },
1448
+ {
1449
+ "epoch": 6.17,
1450
+ "learning_rate": 4.938535388366059e-05,
1451
+ "loss": 1.3836,
1452
+ "step": 21900
1453
+ },
1454
+ {
1455
+ "epoch": 6.2,
1456
+ "learning_rate": 4.9298065120132974e-05,
1457
+ "loss": 1.3839,
1458
+ "step": 22000
1459
+ },
1460
+ {
1461
+ "epoch": 6.2,
1462
+ "eval_gen_len": 20.0,
1463
+ "eval_loss": 2.1140034198760986,
1464
+ "eval_rouge1": 20.6468,
1465
+ "eval_rouge2": 5.2318,
1466
+ "eval_rougeL": 20.4445,
1467
+ "eval_rougeLsum": 20.4398,
1468
+ "eval_runtime": 162.2862,
1469
+ "eval_samples_per_second": 43.183,
1470
+ "eval_steps_per_second": 2.699,
1471
+ "step": 22000
1472
+ },
1473
+ {
1474
+ "epoch": 6.23,
1475
+ "learning_rate": 4.920507066342175e-05,
1476
+ "loss": 1.4005,
1477
+ "step": 22100
1478
+ },
1479
+ {
1480
+ "epoch": 6.26,
1481
+ "learning_rate": 4.910639239559606e-05,
1482
+ "loss": 1.3887,
1483
+ "step": 22200
1484
+ },
1485
+ {
1486
+ "epoch": 6.29,
1487
+ "learning_rate": 4.900205353615477e-05,
1488
+ "loss": 1.4081,
1489
+ "step": 22300
1490
+ },
1491
+ {
1492
+ "epoch": 6.31,
1493
+ "learning_rate": 4.889207863656274e-05,
1494
+ "loss": 1.4239,
1495
+ "step": 22400
1496
+ },
1497
+ {
1498
+ "epoch": 6.34,
1499
+ "learning_rate": 4.87764935744738e-05,
1500
+ "loss": 1.4332,
1501
+ "step": 22500
1502
+ },
1503
+ {
1504
+ "epoch": 6.37,
1505
+ "learning_rate": 4.865532554764157e-05,
1506
+ "loss": 1.4264,
1507
+ "step": 22600
1508
+ },
1509
+ {
1510
+ "epoch": 6.4,
1511
+ "learning_rate": 4.852860306751969e-05,
1512
+ "loss": 1.4465,
1513
+ "step": 22700
1514
+ },
1515
+ {
1516
+ "epoch": 6.43,
1517
+ "learning_rate": 4.839635595255296e-05,
1518
+ "loss": 1.4339,
1519
+ "step": 22800
1520
+ },
1521
+ {
1522
+ "epoch": 6.45,
1523
+ "learning_rate": 4.825861532116087e-05,
1524
+ "loss": 1.4463,
1525
+ "step": 22900
1526
+ },
1527
+ {
1528
+ "epoch": 6.48,
1529
+ "learning_rate": 4.811541358441531e-05,
1530
+ "loss": 1.447,
1531
+ "step": 23000
1532
+ },
1533
+ {
1534
+ "epoch": 6.51,
1535
+ "learning_rate": 4.796678443841405e-05,
1536
+ "loss": 1.4343,
1537
+ "step": 23100
1538
+ },
1539
+ {
1540
+ "epoch": 6.54,
1541
+ "learning_rate": 4.78127628563519e-05,
1542
+ "loss": 1.4502,
1543
+ "step": 23200
1544
+ },
1545
+ {
1546
+ "epoch": 6.57,
1547
+ "learning_rate": 4.765338508029135e-05,
1548
+ "loss": 1.4459,
1549
+ "step": 23300
1550
+ },
1551
+ {
1552
+ "epoch": 6.6,
1553
+ "learning_rate": 4.748868861263457e-05,
1554
+ "loss": 1.4448,
1555
+ "step": 23400
1556
+ },
1557
+ {
1558
+ "epoch": 6.62,
1559
+ "learning_rate": 4.731871220729898e-05,
1560
+ "loss": 1.4377,
1561
+ "step": 23500
1562
+ },
1563
+ {
1564
+ "epoch": 6.65,
1565
+ "learning_rate": 4.714349586059819e-05,
1566
+ "loss": 1.4549,
1567
+ "step": 23600
1568
+ },
1569
+ {
1570
+ "epoch": 6.68,
1571
+ "learning_rate": 4.696308080183071e-05,
1572
+ "loss": 1.4393,
1573
+ "step": 23700
1574
+ },
1575
+ {
1576
+ "epoch": 6.71,
1577
+ "learning_rate": 4.677750948357847e-05,
1578
+ "loss": 1.4678,
1579
+ "step": 23800
1580
+ },
1581
+ {
1582
+ "epoch": 6.74,
1583
+ "learning_rate": 4.65868255717175e-05,
1584
+ "loss": 1.4629,
1585
+ "step": 23900
1586
+ },
1587
+ {
1588
+ "epoch": 6.76,
1589
+ "learning_rate": 4.639107393514314e-05,
1590
+ "loss": 1.4605,
1591
+ "step": 24000
1592
+ },
1593
+ {
1594
+ "epoch": 6.76,
1595
+ "eval_gen_len": 20.0,
1596
+ "eval_loss": 2.088627815246582,
1597
+ "eval_rouge1": 21.2304,
1598
+ "eval_rouge2": 5.2346,
1599
+ "eval_rougeL": 21.0611,
1600
+ "eval_rougeLsum": 21.0649,
1601
+ "eval_runtime": 163.3508,
1602
+ "eval_samples_per_second": 42.902,
1603
+ "eval_steps_per_second": 2.681,
1604
+ "step": 24000
1605
+ },
1606
+ {
1607
+ "epoch": 6.79,
1608
+ "learning_rate": 4.619030063521217e-05,
1609
+ "loss": 1.4447,
1610
+ "step": 24100
1611
+ },
1612
+ {
1613
+ "epoch": 6.82,
1614
+ "learning_rate": 4.5984552914904304e-05,
1615
+ "loss": 1.4733,
1616
+ "step": 24200
1617
+ },
1618
+ {
1619
+ "epoch": 6.85,
1620
+ "learning_rate": 4.5773879187705715e-05,
1621
+ "loss": 1.4607,
1622
+ "step": 24300
1623
+ },
1624
+ {
1625
+ "epoch": 6.88,
1626
+ "learning_rate": 4.555832902621708e-05,
1627
+ "loss": 1.4659,
1628
+ "step": 24400
1629
+ },
1630
+ {
1631
+ "epoch": 6.91,
1632
+ "learning_rate": 4.533795315048888e-05,
1633
+ "loss": 1.482,
1634
+ "step": 24500
1635
+ },
1636
+ {
1637
+ "epoch": 6.93,
1638
+ "learning_rate": 4.511280341608673e-05,
1639
+ "loss": 1.4645,
1640
+ "step": 24600
1641
+ },
1642
+ {
1643
+ "epoch": 6.96,
1644
+ "learning_rate": 4.488293280188952e-05,
1645
+ "loss": 1.4645,
1646
+ "step": 24700
1647
+ },
1648
+ {
1649
+ "epoch": 6.99,
1650
+ "learning_rate": 4.464839539762315e-05,
1651
+ "loss": 1.4573,
1652
+ "step": 24800
1653
+ },
1654
+ {
1655
+ "epoch": 7.02,
1656
+ "learning_rate": 4.4409246391133016e-05,
1657
+ "loss": 1.3075,
1658
+ "step": 24900
1659
+ },
1660
+ {
1661
+ "epoch": 7.05,
1662
+ "learning_rate": 4.416554205539801e-05,
1663
+ "loss": 1.2172,
1664
+ "step": 25000
1665
+ },
1666
+ {
1667
+ "epoch": 7.07,
1668
+ "learning_rate": 4.391733973528915e-05,
1669
+ "loss": 1.1984,
1670
+ "step": 25100
1671
+ },
1672
+ {
1673
+ "epoch": 7.1,
1674
+ "learning_rate": 4.3664697834076095e-05,
1675
+ "loss": 1.2395,
1676
+ "step": 25200
1677
+ },
1678
+ {
1679
+ "epoch": 7.13,
1680
+ "learning_rate": 4.340767579968456e-05,
1681
+ "loss": 1.2372,
1682
+ "step": 25300
1683
+ },
1684
+ {
1685
+ "epoch": 7.16,
1686
+ "learning_rate": 4.314896870880448e-05,
1687
+ "loss": 1.2366,
1688
+ "step": 25400
1689
+ },
1690
+ {
1691
+ "epoch": 7.19,
1692
+ "learning_rate": 4.288341113416578e-05,
1693
+ "loss": 1.2374,
1694
+ "step": 25500
1695
+ },
1696
+ {
1697
+ "epoch": 7.22,
1698
+ "learning_rate": 4.261365726708706e-05,
1699
+ "loss": 1.2235,
1700
+ "step": 25600
1701
+ },
1702
+ {
1703
+ "epoch": 7.24,
1704
+ "learning_rate": 4.2339770582026725e-05,
1705
+ "loss": 1.2633,
1706
+ "step": 25700
1707
+ },
1708
+ {
1709
+ "epoch": 7.27,
1710
+ "learning_rate": 4.206181552591627e-05,
1711
+ "loss": 1.2347,
1712
+ "step": 25800
1713
+ },
1714
+ {
1715
+ "epoch": 7.3,
1716
+ "learning_rate": 4.1779857502995634e-05,
1717
+ "loss": 1.2533,
1718
+ "step": 25900
1719
+ },
1720
+ {
1721
+ "epoch": 7.33,
1722
+ "learning_rate": 4.1493962859423225e-05,
1723
+ "loss": 1.2594,
1724
+ "step": 26000
1725
+ },
1726
+ {
1727
+ "epoch": 7.33,
1728
+ "eval_gen_len": 20.0,
1729
+ "eval_loss": 2.1528804302215576,
1730
+ "eval_rouge1": 21.6885,
1731
+ "eval_rouge2": 5.3904,
1732
+ "eval_rougeL": 21.4125,
1733
+ "eval_rougeLsum": 21.4342,
1734
+ "eval_runtime": 162.2811,
1735
+ "eval_samples_per_second": 43.184,
1736
+ "eval_steps_per_second": 2.699,
1737
+ "step": 26000
1738
+ },
1739
+ {
1740
+ "epoch": 7.36,
1741
+ "learning_rate": 4.120419886766432e-05,
1742
+ "loss": 1.2742,
1743
+ "step": 26100
1744
+ },
1745
+ {
1746
+ "epoch": 7.38,
1747
+ "learning_rate": 4.091063371066154e-05,
1748
+ "loss": 1.246,
1749
+ "step": 26200
1750
+ },
1751
+ {
1752
+ "epoch": 7.41,
1753
+ "learning_rate": 4.061632768309558e-05,
1754
+ "loss": 1.2496,
1755
+ "step": 26300
1756
+ },
1757
+ {
1758
+ "epoch": 7.44,
1759
+ "learning_rate": 4.03154045781127e-05,
1760
+ "loss": 1.2608,
1761
+ "step": 26400
1762
+ },
1763
+ {
1764
+ "epoch": 7.47,
1765
+ "learning_rate": 4.001088944570764e-05,
1766
+ "loss": 1.2696,
1767
+ "step": 26500
1768
+ },
1769
+ {
1770
+ "epoch": 7.5,
1771
+ "learning_rate": 3.9702853939841514e-05,
1772
+ "loss": 1.2624,
1773
+ "step": 26600
1774
+ },
1775
+ {
1776
+ "epoch": 7.53,
1777
+ "learning_rate": 3.939137054283725e-05,
1778
+ "loss": 1.2809,
1779
+ "step": 26700
1780
+ },
1781
+ {
1782
+ "epoch": 7.55,
1783
+ "learning_rate": 3.9076512548324085e-05,
1784
+ "loss": 1.2533,
1785
+ "step": 26800
1786
+ },
1787
+ {
1788
+ "epoch": 7.58,
1789
+ "learning_rate": 3.875835404399126e-05,
1790
+ "loss": 1.2811,
1791
+ "step": 26900
1792
+ },
1793
+ {
1794
+ "epoch": 7.61,
1795
+ "learning_rate": 3.843696989415477e-05,
1796
+ "loss": 1.2851,
1797
+ "step": 27000
1798
+ },
1799
+ {
1800
+ "epoch": 7.64,
1801
+ "learning_rate": 3.811243572214143e-05,
1802
+ "loss": 1.2661,
1803
+ "step": 27100
1804
+ },
1805
+ {
1806
+ "epoch": 7.67,
1807
+ "learning_rate": 3.7784827892494295e-05,
1808
+ "loss": 1.2661,
1809
+ "step": 27200
1810
+ },
1811
+ {
1812
+ "epoch": 7.69,
1813
+ "learning_rate": 3.745422349300373e-05,
1814
+ "loss": 1.2525,
1815
+ "step": 27300
1816
+ },
1817
+ {
1818
+ "epoch": 7.72,
1819
+ "learning_rate": 3.712070031656822e-05,
1820
+ "loss": 1.2864,
1821
+ "step": 27400
1822
+ },
1823
+ {
1824
+ "epoch": 7.75,
1825
+ "learning_rate": 3.6784336842889355e-05,
1826
+ "loss": 1.2801,
1827
+ "step": 27500
1828
+ },
1829
+ {
1830
+ "epoch": 7.78,
1831
+ "learning_rate": 3.64452122200051e-05,
1832
+ "loss": 1.2888,
1833
+ "step": 27600
1834
+ },
1835
+ {
1836
+ "epoch": 7.81,
1837
+ "learning_rate": 3.6103406245665894e-05,
1838
+ "loss": 1.2825,
1839
+ "step": 27700
1840
+ },
1841
+ {
1842
+ "epoch": 7.84,
1843
+ "learning_rate": 3.575899934855785e-05,
1844
+ "loss": 1.2791,
1845
+ "step": 27800
1846
+ },
1847
+ {
1848
+ "epoch": 7.86,
1849
+ "learning_rate": 3.5412072569377407e-05,
1850
+ "loss": 1.2704,
1851
+ "step": 27900
1852
+ },
1853
+ {
1854
+ "epoch": 7.89,
1855
+ "learning_rate": 3.506270754176212e-05,
1856
+ "loss": 1.2809,
1857
+ "step": 28000
1858
+ },
1859
+ {
1860
+ "epoch": 7.89,
1861
+ "eval_gen_len": 20.0,
1862
+ "eval_loss": 2.155640125274658,
1863
+ "eval_rouge1": 20.7036,
1864
+ "eval_rouge2": 5.3801,
1865
+ "eval_rougeL": 20.5265,
1866
+ "eval_rougeLsum": 20.5209,
1867
+ "eval_runtime": 163.3415,
1868
+ "eval_samples_per_second": 42.904,
1869
+ "eval_steps_per_second": 2.681,
1870
+ "step": 28000
1871
+ },
1872
+ {
1873
+ "epoch": 7.92,
1874
+ "learning_rate": 3.471098647308181e-05,
1875
+ "loss": 1.2758,
1876
+ "step": 28100
1877
+ },
1878
+ {
1879
+ "epoch": 7.95,
1880
+ "learning_rate": 3.4356992125094747e-05,
1881
+ "loss": 1.2923,
1882
+ "step": 28200
1883
+ },
1884
+ {
1885
+ "epoch": 7.98,
1886
+ "learning_rate": 3.40008077944734e-05,
1887
+ "loss": 1.2823,
1888
+ "step": 28300
1889
+ },
1890
+ {
1891
+ "epoch": 8.0,
1892
+ "learning_rate": 3.3642517293204305e-05,
1893
+ "loss": 1.265,
1894
+ "step": 28400
1895
+ },
1896
+ {
1897
+ "epoch": 8.03,
1898
+ "learning_rate": 3.328220492886667e-05,
1899
+ "loss": 1.0506,
1900
+ "step": 28500
1901
+ },
1902
+ {
1903
+ "epoch": 8.06,
1904
+ "learning_rate": 3.291995548479439e-05,
1905
+ "loss": 1.0599,
1906
+ "step": 28600
1907
+ },
1908
+ {
1909
+ "epoch": 8.09,
1910
+ "learning_rate": 3.255585420012606e-05,
1911
+ "loss": 1.07,
1912
+ "step": 28700
1913
+ },
1914
+ {
1915
+ "epoch": 8.12,
1916
+ "learning_rate": 3.2189986749747835e-05,
1917
+ "loss": 1.0685,
1918
+ "step": 28800
1919
+ },
1920
+ {
1921
+ "epoch": 8.15,
1922
+ "learning_rate": 3.182243922413361e-05,
1923
+ "loss": 1.088,
1924
+ "step": 28900
1925
+ },
1926
+ {
1927
+ "epoch": 8.17,
1928
+ "learning_rate": 3.145329810908756e-05,
1929
+ "loss": 1.0686,
1930
+ "step": 29000
1931
+ },
1932
+ {
1933
+ "epoch": 8.2,
1934
+ "learning_rate": 3.1082650265393565e-05,
1935
+ "loss": 1.0841,
1936
+ "step": 29100
1937
+ },
1938
+ {
1939
+ "epoch": 8.23,
1940
+ "learning_rate": 3.071058290837644e-05,
1941
+ "loss": 1.0808,
1942
+ "step": 29200
1943
+ },
1944
+ {
1945
+ "epoch": 8.26,
1946
+ "learning_rate": 3.0337183587379695e-05,
1947
+ "loss": 1.0797,
1948
+ "step": 29300
1949
+ },
1950
+ {
1951
+ "epoch": 8.29,
1952
+ "learning_rate": 2.9962540165164753e-05,
1953
+ "loss": 1.0824,
1954
+ "step": 29400
1955
+ },
1956
+ {
1957
+ "epoch": 8.31,
1958
+ "learning_rate": 2.958674079723637e-05,
1959
+ "loss": 1.0767,
1960
+ "step": 29500
1961
+ },
1962
+ {
1963
+ "epoch": 8.34,
1964
+ "learning_rate": 2.9209873911099242e-05,
1965
+ "loss": 1.0872,
1966
+ "step": 29600
1967
+ },
1968
+ {
1969
+ "epoch": 8.37,
1970
+ "learning_rate": 2.8832028185450523e-05,
1971
+ "loss": 1.0726,
1972
+ "step": 29700
1973
+ },
1974
+ {
1975
+ "epoch": 8.4,
1976
+ "learning_rate": 2.845329252931329e-05,
1977
+ "loss": 1.0763,
1978
+ "step": 29800
1979
+ },
1980
+ {
1981
+ "epoch": 8.43,
1982
+ "learning_rate": 2.8073756061115794e-05,
1983
+ "loss": 1.1097,
1984
+ "step": 29900
1985
+ },
1986
+ {
1987
+ "epoch": 8.46,
1988
+ "learning_rate": 2.7693508087721433e-05,
1989
+ "loss": 1.0819,
1990
+ "step": 30000
1991
+ },
1992
+ {
1993
+ "epoch": 8.46,
1994
+ "eval_gen_len": 20.0,
1995
+ "eval_loss": 2.23087477684021,
1996
+ "eval_rouge1": 20.9622,
1997
+ "eval_rouge2": 5.3019,
1998
+ "eval_rougeL": 20.7446,
1999
+ "eval_rougeLsum": 20.7429,
2000
+ "eval_runtime": 162.5566,
2001
+ "eval_samples_per_second": 43.111,
2002
+ "eval_steps_per_second": 2.694,
2003
+ "step": 30000
2004
+ },
2005
+ {
2006
+ "epoch": 8.48,
2007
+ "learning_rate": 2.731263808341441e-05,
2008
+ "loss": 1.1006,
2009
+ "step": 30100
2010
+ },
2011
+ {
2012
+ "epoch": 8.51,
2013
+ "learning_rate": 2.6931235668845973e-05,
2014
+ "loss": 1.0997,
2015
+ "step": 30200
2016
+ },
2017
+ {
2018
+ "epoch": 8.54,
2019
+ "learning_rate": 2.65493905899462e-05,
2020
+ "loss": 1.1084,
2021
+ "step": 30300
2022
+ },
2023
+ {
2024
+ "epoch": 8.57,
2025
+ "learning_rate": 2.6167192696806356e-05,
2026
+ "loss": 1.0907,
2027
+ "step": 30400
2028
+ },
2029
+ {
2030
+ "epoch": 8.6,
2031
+ "learning_rate": 2.5784731922536687e-05,
2032
+ "loss": 1.0974,
2033
+ "step": 30500
2034
+ },
2035
+ {
2036
+ "epoch": 8.62,
2037
+ "learning_rate": 2.540592515888669e-05,
2038
+ "loss": 1.1108,
2039
+ "step": 30600
2040
+ },
2041
+ {
2042
+ "epoch": 8.65,
2043
+ "learning_rate": 2.5023209030687986e-05,
2044
+ "loss": 1.1082,
2045
+ "step": 30700
2046
+ },
2047
+ {
2048
+ "epoch": 8.68,
2049
+ "learning_rate": 2.4640499206540774e-05,
2050
+ "loss": 1.0964,
2051
+ "step": 30800
2052
+ },
2053
+ {
2054
+ "epoch": 8.71,
2055
+ "learning_rate": 2.425788574001527e-05,
2056
+ "loss": 1.1008,
2057
+ "step": 30900
2058
+ },
2059
+ {
2060
+ "epoch": 8.74,
2061
+ "learning_rate": 2.3875458662008246e-05,
2062
+ "loss": 1.0956,
2063
+ "step": 31000
2064
+ },
2065
+ {
2066
+ "epoch": 8.77,
2067
+ "learning_rate": 2.3493307959558313e-05,
2068
+ "loss": 1.1051,
2069
+ "step": 31100
2070
+ },
2071
+ {
2072
+ "epoch": 8.79,
2073
+ "learning_rate": 2.3111523554671516e-05,
2074
+ "loss": 1.1033,
2075
+ "step": 31200
2076
+ },
2077
+ {
2078
+ "epoch": 8.82,
2079
+ "learning_rate": 2.2730195283162194e-05,
2080
+ "loss": 1.1003,
2081
+ "step": 31300
2082
+ },
2083
+ {
2084
+ "epoch": 8.85,
2085
+ "learning_rate": 2.2349412873514212e-05,
2086
+ "loss": 1.1031,
2087
+ "step": 31400
2088
+ },
2089
+ {
2090
+ "epoch": 8.88,
2091
+ "learning_rate": 2.1969265925767334e-05,
2092
+ "loss": 1.1195,
2093
+ "step": 31500
2094
+ },
2095
+ {
2096
+ "epoch": 8.91,
2097
+ "learning_rate": 2.158984389043391e-05,
2098
+ "loss": 1.0978,
2099
+ "step": 31600
2100
+ },
2101
+ {
2102
+ "epoch": 8.93,
2103
+ "learning_rate": 2.1211236047450704e-05,
2104
+ "loss": 1.1105,
2105
+ "step": 31700
2106
+ },
2107
+ {
2108
+ "epoch": 8.96,
2109
+ "learning_rate": 2.08335314851709e-05,
2110
+ "loss": 1.1088,
2111
+ "step": 31800
2112
+ },
2113
+ {
2114
+ "epoch": 8.99,
2115
+ "learning_rate": 2.0456819079401094e-05,
2116
+ "loss": 1.1241,
2117
+ "step": 31900
2118
+ },
2119
+ {
2120
+ "epoch": 9.02,
2121
+ "learning_rate": 2.008118747248843e-05,
2122
+ "loss": 0.9814,
2123
+ "step": 32000
2124
+ },
2125
+ {
2126
+ "epoch": 9.02,
2127
+ "eval_gen_len": 20.0,
2128
+ "eval_loss": 2.2810895442962646,
2129
+ "eval_rouge1": 20.6296,
2130
+ "eval_rouge2": 5.1936,
2131
+ "eval_rougeL": 20.4403,
2132
+ "eval_rougeLsum": 20.4304,
2133
+ "eval_runtime": 162.301,
2134
+ "eval_samples_per_second": 43.179,
2135
+ "eval_steps_per_second": 2.699,
2136
+ "step": 32000
2137
+ },
2138
+ {
2139
+ "epoch": 9.05,
2140
+ "learning_rate": 1.9706725052462575e-05,
2141
+ "loss": 0.9353,
2142
+ "step": 32100
2143
+ },
2144
+ {
2145
+ "epoch": 9.08,
2146
+ "learning_rate": 1.933351993223755e-05,
2147
+ "loss": 0.9372,
2148
+ "step": 32200
2149
+ },
2150
+ {
2151
+ "epoch": 9.1,
2152
+ "learning_rate": 1.8961659928878366e-05,
2153
+ "loss": 0.9601,
2154
+ "step": 32300
2155
+ },
2156
+ {
2157
+ "epoch": 9.13,
2158
+ "learning_rate": 1.859123254293715e-05,
2159
+ "loss": 0.9461,
2160
+ "step": 32400
2161
+ },
2162
+ {
2163
+ "epoch": 9.16,
2164
+ "learning_rate": 1.8222324937863884e-05,
2165
+ "loss": 0.9575,
2166
+ "step": 32500
2167
+ },
2168
+ {
2169
+ "epoch": 9.19,
2170
+ "learning_rate": 1.7858688692966636e-05,
2171
+ "loss": 0.951,
2172
+ "step": 32600
2173
+ },
2174
+ {
2175
+ "epoch": 9.22,
2176
+ "learning_rate": 1.74930633324332e-05,
2177
+ "loss": 0.9606,
2178
+ "step": 32700
2179
+ },
2180
+ {
2181
+ "epoch": 9.24,
2182
+ "learning_rate": 1.712921615757434e-05,
2183
+ "loss": 0.9316,
2184
+ "step": 32800
2185
+ },
2186
+ {
2187
+ "epoch": 9.27,
2188
+ "learning_rate": 1.676723278348288e-05,
2189
+ "loss": 0.9447,
2190
+ "step": 32900
2191
+ },
2192
+ {
2193
+ "epoch": 9.3,
2194
+ "learning_rate": 1.640719838668979e-05,
2195
+ "loss": 0.9591,
2196
+ "step": 33000
2197
+ },
2198
+ {
2199
+ "epoch": 9.33,
2200
+ "learning_rate": 1.604919768512179e-05,
2201
+ "loss": 0.9629,
2202
+ "step": 33100
2203
+ },
2204
+ {
2205
+ "epoch": 9.36,
2206
+ "learning_rate": 1.569331491816673e-05,
2207
+ "loss": 0.9541,
2208
+ "step": 33200
2209
+ },
2210
+ {
2211
+ "epoch": 9.39,
2212
+ "learning_rate": 1.53396338268517e-05,
2213
+ "loss": 0.9409,
2214
+ "step": 33300
2215
+ },
2216
+ {
2217
+ "epoch": 9.41,
2218
+ "learning_rate": 1.4988237634138258e-05,
2219
+ "loss": 0.9517,
2220
+ "step": 33400
2221
+ },
2222
+ {
2223
+ "epoch": 9.44,
2224
+ "learning_rate": 1.4639209025339731e-05,
2225
+ "loss": 0.9569,
2226
+ "step": 33500
2227
+ },
2228
+ {
2229
+ "epoch": 9.47,
2230
+ "learning_rate": 1.4292630128664853e-05,
2231
+ "loss": 0.9421,
2232
+ "step": 33600
2233
+ },
2234
+ {
2235
+ "epoch": 9.5,
2236
+ "learning_rate": 1.3948582495892665e-05,
2237
+ "loss": 0.9455,
2238
+ "step": 33700
2239
+ },
2240
+ {
2241
+ "epoch": 9.53,
2242
+ "learning_rate": 1.3610548242500389e-05,
2243
+ "loss": 0.9553,
2244
+ "step": 33800
2245
+ },
2246
+ {
2247
+ "epoch": 9.55,
2248
+ "learning_rate": 1.3271778070102909e-05,
2249
+ "loss": 0.9668,
2250
+ "step": 33900
2251
+ },
2252
+ {
2253
+ "epoch": 9.58,
2254
+ "learning_rate": 1.2935779373295224e-05,
2255
+ "loss": 0.9403,
2256
+ "step": 34000
2257
+ },
2258
+ {
2259
+ "epoch": 9.58,
2260
+ "eval_gen_len": 20.0,
2261
+ "eval_loss": 2.29062819480896,
2262
+ "eval_rouge1": 21.0701,
2263
+ "eval_rouge2": 5.1666,
2264
+ "eval_rougeL": 20.8587,
2265
+ "eval_rougeLsum": 20.8463,
2266
+ "eval_runtime": 162.5034,
2267
+ "eval_samples_per_second": 43.125,
2268
+ "eval_steps_per_second": 2.695,
2269
+ "step": 34000
2270
+ },
2271
+ {
2272
+ "epoch": 9.61,
2273
+ "learning_rate": 1.2602631214281477e-05,
2274
+ "loss": 0.9553,
2275
+ "step": 34100
2276
+ },
2277
+ {
2278
+ "epoch": 9.64,
2279
+ "learning_rate": 1.2272411984519708e-05,
2280
+ "loss": 0.9575,
2281
+ "step": 34200
2282
+ },
2283
+ {
2284
+ "epoch": 9.67,
2285
+ "learning_rate": 1.1945199386276041e-05,
2286
+ "loss": 0.9545,
2287
+ "step": 34300
2288
+ },
2289
+ {
2290
+ "epoch": 9.7,
2291
+ "learning_rate": 1.1621070414340846e-05,
2292
+ "loss": 0.9406,
2293
+ "step": 34400
2294
+ },
2295
+ {
2296
+ "epoch": 9.72,
2297
+ "learning_rate": 1.1300101337911568e-05,
2298
+ "loss": 0.9689,
2299
+ "step": 34500
2300
+ },
2301
+ {
2302
+ "epoch": 9.75,
2303
+ "learning_rate": 1.0982367682646119e-05,
2304
+ "loss": 0.9576,
2305
+ "step": 34600
2306
+ },
2307
+ {
2308
+ "epoch": 9.78,
2309
+ "learning_rate": 1.0667944212891339e-05,
2310
+ "loss": 0.9547,
2311
+ "step": 34700
2312
+ },
2313
+ {
2314
+ "epoch": 9.81,
2315
+ "learning_rate": 1.0356904914090589e-05,
2316
+ "loss": 0.9619,
2317
+ "step": 34800
2318
+ },
2319
+ {
2320
+ "epoch": 9.84,
2321
+ "learning_rate": 1.0049322975374548e-05,
2322
+ "loss": 0.9578,
2323
+ "step": 34900
2324
+ },
2325
+ {
2326
+ "epoch": 9.86,
2327
+ "learning_rate": 9.745270772339474e-06,
2328
+ "loss": 0.9509,
2329
+ "step": 35000
2330
+ },
2331
+ {
2332
+ "epoch": 9.89,
2333
+ "learning_rate": 9.447806300054016e-06,
2334
+ "loss": 0.9593,
2335
+ "step": 35100
2336
+ },
2337
+ {
2338
+ "epoch": 9.92,
2339
+ "learning_rate": 9.15099028916227e-06,
2340
+ "loss": 0.9576,
2341
+ "step": 35200
2342
+ },
2343
+ {
2344
+ "epoch": 9.95,
2345
+ "learning_rate": 8.857915396206473e-06,
2346
+ "loss": 0.9653,
2347
+ "step": 35300
2348
+ },
2349
+ {
2350
+ "epoch": 9.98,
2351
+ "learning_rate": 8.56865058320117e-06,
2352
+ "loss": 0.958,
2353
+ "step": 35400
2354
+ },
2355
+ {
2356
+ "epoch": 10.0,
2357
+ "step": 35480,
2358
+ "total_flos": 2.3762843604025344e+17,
2359
+ "train_loss": 1.5038130920392954,
2360
+ "train_runtime": 13804.3068,
2361
+ "train_samples_per_second": 41.118,
2362
+ "train_steps_per_second": 2.57
2363
+ }
2364
+ ],
2365
+ "max_steps": 35480,
2366
+ "num_train_epochs": 10,
2367
+ "total_flos": 2.3762843604025344e+17,
2368
+ "trial_name": null,
2369
+ "trial_params": null
2370
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9cdba313b31bdac3438861022cd20b0533cea266378d23d59149777ab8970935
3
+ size 3567