nour4286 commited on
Commit
c5feb94
1 Parent(s): 55d235d

Training in progress, epoch 21, checkpoint

Browse files
checkpoint-78603/config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "nour4286/ArabartModel",
3
+ "activation_dropout": 0.1,
4
+ "activation_function": "gelu",
5
+ "add_bias_logits": false,
6
+ "add_final_layer_norm": true,
7
+ "architectures": [
8
+ "MBartForConditionalGeneration"
9
+ ],
10
+ "attention_dropout": 0.1,
11
+ "bos_token_id": 0,
12
+ "classif_dropout": 0.1,
13
+ "classifier_dropout": 0.0,
14
+ "d_model": 768,
15
+ "decoder_attention_heads": 12,
16
+ "decoder_ffn_dim": 3072,
17
+ "decoder_layerdrop": 0.0,
18
+ "decoder_layers": 6,
19
+ "decoder_start_token_id": 2,
20
+ "do_blenderbot_90_layernorm": false,
21
+ "dropout": 0.1,
22
+ "early_stopping": true,
23
+ "encoder_attention_heads": 12,
24
+ "encoder_ffn_dim": 3072,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 6,
27
+ "eos_token_id": 2,
28
+ "extra_pos_embeddings": 2,
29
+ "force_bos_token_to_be_generated": false,
30
+ "forced_eos_token_id": 2,
31
+ "gradient_checkpointing": false,
32
+ "id2label": {
33
+ "0": "LABEL_0",
34
+ "1": "LABEL_1",
35
+ "2": "LABEL_2"
36
+ },
37
+ "init_std": 0.02,
38
+ "is_encoder_decoder": true,
39
+ "label2id": {
40
+ "LABEL_0": 0,
41
+ "LABEL_1": 1,
42
+ "LABEL_2": 2
43
+ },
44
+ "max_position_embeddings": 1024,
45
+ "model_type": "mbart",
46
+ "no_repeat_ngram_size": 3,
47
+ "normalize_before": true,
48
+ "normalize_embedding": true,
49
+ "num_beams": 4,
50
+ "num_hidden_layers": 6,
51
+ "pad_token_id": 1,
52
+ "scale_embedding": false,
53
+ "static_position_embeddings": false,
54
+ "tokenizer_class": "BarthezTokenizer",
55
+ "torch_dtype": "float32",
56
+ "transformers_version": "4.42.3",
57
+ "use_cache": true,
58
+ "vocab_size": 50002
59
+ }
checkpoint-78603/generation_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 0,
3
+ "decoder_start_token_id": 2,
4
+ "early_stopping": true,
5
+ "eos_token_id": 2,
6
+ "forced_eos_token_id": 2,
7
+ "max_new_tokens": 100,
8
+ "no_repeat_ngram_size": 3,
9
+ "num_beams": 3,
10
+ "pad_token_id": 1,
11
+ "transformers_version": "4.42.3"
12
+ }
checkpoint-78603/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c798c971ed58000041ccc5301830cb3603c2f768916682f71f8f538a5a093fe8
3
+ size 557116312
checkpoint-78603/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:25f42064ec39fd7a95974e25a784596a0b7dd3bb02cdc264df2ec63515c4bb71
3
+ size 1113991930
checkpoint-78603/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:68e934ccfdcf84eec0f84804fc72582de32cdc438aa8591d0af65dfaec5ef1f2
3
+ size 14244
checkpoint-78603/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:31736b5da93dc254a73b262a1f5d72b4227ce40205a9ffce9b159c5e29bca98e
3
+ size 1064
checkpoint-78603/sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cbb59d772bc9bb2da5dc4a73a00c61c0912c6d2596aad970fa2cd3d69898b245
3
+ size 1315170
checkpoint-78603/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
checkpoint-78603/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-78603/tokenizer_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "50001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": true,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "mask_token": "<mask>",
49
+ "model_max_length": 1000000000000000019884624838656,
50
+ "pad_token": "<pad>",
51
+ "sep_token": "</s>",
52
+ "sp_model_kwargs": {},
53
+ "tokenizer_class": "BarthezTokenizer",
54
+ "unk_token": "<unk>"
55
+ }
checkpoint-78603/trainer_state.json ADDED
@@ -0,0 +1,1451 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 29.9002,
3
+ "best_model_checkpoint": "/content/drive/MyDrive/ArabartModel/checkpoint-54128",
4
+ "epoch": 21.0,
5
+ "eval_steps": 500,
6
+ "global_step": 78603,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.32722513089005234,
13
+ "grad_norm": 1.563672423362732,
14
+ "learning_rate": 4.6727748691099475e-05,
15
+ "loss": 0.0195,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 0.6544502617801047,
20
+ "grad_norm": 1.7397737503051758,
21
+ "learning_rate": 4.3455497382198955e-05,
22
+ "loss": 0.0244,
23
+ "step": 1000
24
+ },
25
+ {
26
+ "epoch": 0.981675392670157,
27
+ "grad_norm": 3.006633758544922,
28
+ "learning_rate": 4.018324607329843e-05,
29
+ "loss": 0.0293,
30
+ "step": 1500
31
+ },
32
+ {
33
+ "epoch": 1.0,
34
+ "eval_loss": 5.626201629638672,
35
+ "eval_rouge1": 25.2705,
36
+ "eval_rouge2": 9.9865,
37
+ "eval_rougeL": 21.8678,
38
+ "eval_runtime": 372.5145,
39
+ "eval_samples_per_second": 5.892,
40
+ "eval_steps_per_second": 0.591,
41
+ "step": 1528
42
+ },
43
+ {
44
+ "epoch": 1.3089005235602094,
45
+ "grad_norm": 3.4732136726379395,
46
+ "learning_rate": 4.3455497382198955e-05,
47
+ "loss": 0.0456,
48
+ "step": 2000
49
+ },
50
+ {
51
+ "epoch": 1.6361256544502618,
52
+ "grad_norm": 2.2575156688690186,
53
+ "learning_rate": 4.181937172774869e-05,
54
+ "loss": 0.0619,
55
+ "step": 2500
56
+ },
57
+ {
58
+ "epoch": 1.9633507853403143,
59
+ "grad_norm": 6.030032157897949,
60
+ "learning_rate": 4.018324607329843e-05,
61
+ "loss": 0.0993,
62
+ "step": 3000
63
+ },
64
+ {
65
+ "epoch": 2.0,
66
+ "eval_loss": 5.542174339294434,
67
+ "eval_rouge1": 25.0744,
68
+ "eval_rouge2": 9.8534,
69
+ "eval_rougeL": 21.6992,
70
+ "eval_runtime": 411.9676,
71
+ "eval_samples_per_second": 5.328,
72
+ "eval_steps_per_second": 0.534,
73
+ "step": 3056
74
+ },
75
+ {
76
+ "epoch": 2.2905759162303667,
77
+ "grad_norm": 6.401626110076904,
78
+ "learning_rate": 3.8547120418848174e-05,
79
+ "loss": 0.21,
80
+ "step": 3500
81
+ },
82
+ {
83
+ "epoch": 2.6178010471204187,
84
+ "grad_norm": 5.806931495666504,
85
+ "learning_rate": 3.691099476439791e-05,
86
+ "loss": 0.3595,
87
+ "step": 4000
88
+ },
89
+ {
90
+ "epoch": 2.945026178010471,
91
+ "grad_norm": 5.840355396270752,
92
+ "learning_rate": 3.5274869109947647e-05,
93
+ "loss": 0.6436,
94
+ "step": 4500
95
+ },
96
+ {
97
+ "epoch": 3.0,
98
+ "eval_loss": 3.927168369293213,
99
+ "eval_rouge1": 25.5914,
100
+ "eval_rouge2": 10.2904,
101
+ "eval_rougeL": 22.3089,
102
+ "eval_runtime": 412.3529,
103
+ "eval_samples_per_second": 5.323,
104
+ "eval_steps_per_second": 0.534,
105
+ "step": 4584
106
+ },
107
+ {
108
+ "epoch": 3.2722513089005236,
109
+ "grad_norm": 5.320337772369385,
110
+ "learning_rate": 3.3638743455497386e-05,
111
+ "loss": 0.7617,
112
+ "step": 5000
113
+ },
114
+ {
115
+ "epoch": 3.599476439790576,
116
+ "grad_norm": 5.708410739898682,
117
+ "learning_rate": 3.2002617801047126e-05,
118
+ "loss": 0.9186,
119
+ "step": 5500
120
+ },
121
+ {
122
+ "epoch": 3.9267015706806285,
123
+ "grad_norm": 6.043923377990723,
124
+ "learning_rate": 3.036649214659686e-05,
125
+ "loss": 1.112,
126
+ "step": 6000
127
+ },
128
+ {
129
+ "epoch": 4.0,
130
+ "eval_loss": 3.2077620029449463,
131
+ "eval_rouge1": 26.5258,
132
+ "eval_rouge2": 11.116,
133
+ "eval_rougeL": 23.3381,
134
+ "eval_runtime": 397.8337,
135
+ "eval_samples_per_second": 5.517,
136
+ "eval_steps_per_second": 0.553,
137
+ "step": 6112
138
+ },
139
+ {
140
+ "epoch": 4.2539267015706805,
141
+ "grad_norm": 5.179211139678955,
142
+ "learning_rate": 2.87303664921466e-05,
143
+ "loss": 1.0997,
144
+ "step": 6500
145
+ },
146
+ {
147
+ "epoch": 4.581151832460733,
148
+ "grad_norm": 4.997713565826416,
149
+ "learning_rate": 2.709424083769634e-05,
150
+ "loss": 1.0736,
151
+ "step": 7000
152
+ },
153
+ {
154
+ "epoch": 4.908376963350785,
155
+ "grad_norm": 5.441833972930908,
156
+ "learning_rate": 2.545811518324607e-05,
157
+ "loss": 1.0866,
158
+ "step": 7500
159
+ },
160
+ {
161
+ "epoch": 5.0,
162
+ "eval_loss": 3.3156800270080566,
163
+ "eval_rouge1": 26.4275,
164
+ "eval_rouge2": 10.8484,
165
+ "eval_rougeL": 23.0503,
166
+ "eval_runtime": 402.2407,
167
+ "eval_samples_per_second": 5.457,
168
+ "eval_steps_per_second": 0.547,
169
+ "step": 7640
170
+ },
171
+ {
172
+ "epoch": 5.2356020942408374,
173
+ "grad_norm": 4.417330741882324,
174
+ "learning_rate": 2.382198952879581e-05,
175
+ "loss": 0.9343,
176
+ "step": 8000
177
+ },
178
+ {
179
+ "epoch": 5.56282722513089,
180
+ "grad_norm": 5.149570465087891,
181
+ "learning_rate": 2.218586387434555e-05,
182
+ "loss": 0.9473,
183
+ "step": 8500
184
+ },
185
+ {
186
+ "epoch": 5.890052356020942,
187
+ "grad_norm": 5.305272102355957,
188
+ "learning_rate": 2.054973821989529e-05,
189
+ "loss": 0.9589,
190
+ "step": 9000
191
+ },
192
+ {
193
+ "epoch": 6.0,
194
+ "eval_loss": 3.433619737625122,
195
+ "eval_rouge1": 26.3479,
196
+ "eval_rouge2": 10.8544,
197
+ "eval_rougeL": 23.0206,
198
+ "eval_runtime": 385.2891,
199
+ "eval_samples_per_second": 5.697,
200
+ "eval_steps_per_second": 0.571,
201
+ "step": 9168
202
+ },
203
+ {
204
+ "epoch": 6.217277486910994,
205
+ "grad_norm": 5.96689510345459,
206
+ "learning_rate": 1.8913612565445027e-05,
207
+ "loss": 0.8437,
208
+ "step": 9500
209
+ },
210
+ {
211
+ "epoch": 6.544502617801047,
212
+ "grad_norm": 5.179261207580566,
213
+ "learning_rate": 1.7277486910994763e-05,
214
+ "loss": 0.8582,
215
+ "step": 10000
216
+ },
217
+ {
218
+ "epoch": 6.871727748691099,
219
+ "grad_norm": 4.92213773727417,
220
+ "learning_rate": 1.5641361256544503e-05,
221
+ "loss": 0.8669,
222
+ "step": 10500
223
+ },
224
+ {
225
+ "epoch": 7.0,
226
+ "eval_loss": 3.551990032196045,
227
+ "eval_rouge1": 26.0878,
228
+ "eval_rouge2": 10.6444,
229
+ "eval_rougeL": 22.7671,
230
+ "eval_runtime": 413.1448,
231
+ "eval_samples_per_second": 5.313,
232
+ "eval_steps_per_second": 0.533,
233
+ "step": 10696
234
+ },
235
+ {
236
+ "epoch": 7.198952879581152,
237
+ "grad_norm": 4.538914203643799,
238
+ "learning_rate": 1.4005235602094241e-05,
239
+ "loss": 0.7803,
240
+ "step": 11000
241
+ },
242
+ {
243
+ "epoch": 7.526178010471204,
244
+ "grad_norm": 5.420250415802002,
245
+ "learning_rate": 1.236910994764398e-05,
246
+ "loss": 0.7777,
247
+ "step": 11500
248
+ },
249
+ {
250
+ "epoch": 7.853403141361256,
251
+ "grad_norm": 4.503458023071289,
252
+ "learning_rate": 1.0732984293193717e-05,
253
+ "loss": 0.7894,
254
+ "step": 12000
255
+ },
256
+ {
257
+ "epoch": 8.0,
258
+ "eval_loss": 3.652216672897339,
259
+ "eval_rouge1": 26.1239,
260
+ "eval_rouge2": 10.689,
261
+ "eval_rougeL": 22.8303,
262
+ "eval_runtime": 389.5703,
263
+ "eval_samples_per_second": 5.634,
264
+ "eval_steps_per_second": 0.565,
265
+ "step": 12224
266
+ },
267
+ {
268
+ "epoch": 8.180628272251308,
269
+ "grad_norm": 4.9211812019348145,
270
+ "learning_rate": 9.096858638743457e-06,
271
+ "loss": 0.7505,
272
+ "step": 12500
273
+ },
274
+ {
275
+ "epoch": 8.507853403141361,
276
+ "grad_norm": 4.948258876800537,
277
+ "learning_rate": 7.4607329842931935e-06,
278
+ "loss": 0.731,
279
+ "step": 13000
280
+ },
281
+ {
282
+ "epoch": 8.835078534031414,
283
+ "grad_norm": 4.717357158660889,
284
+ "learning_rate": 5.824607329842932e-06,
285
+ "loss": 0.7331,
286
+ "step": 13500
287
+ },
288
+ {
289
+ "epoch": 9.0,
290
+ "eval_loss": 3.7317800521850586,
291
+ "eval_rouge1": 26.2374,
292
+ "eval_rouge2": 10.7166,
293
+ "eval_rougeL": 22.9298,
294
+ "eval_runtime": 385.8895,
295
+ "eval_samples_per_second": 5.688,
296
+ "eval_steps_per_second": 0.57,
297
+ "step": 13752
298
+ },
299
+ {
300
+ "epoch": 9.162303664921467,
301
+ "grad_norm": 4.790928363800049,
302
+ "learning_rate": 4.18848167539267e-06,
303
+ "loss": 0.7187,
304
+ "step": 14000
305
+ },
306
+ {
307
+ "epoch": 9.489528795811518,
308
+ "grad_norm": 4.424731254577637,
309
+ "learning_rate": 2.5523560209424085e-06,
310
+ "loss": 0.6988,
311
+ "step": 14500
312
+ },
313
+ {
314
+ "epoch": 9.81675392670157,
315
+ "grad_norm": 4.868335723876953,
316
+ "learning_rate": 9.162303664921465e-07,
317
+ "loss": 0.7027,
318
+ "step": 15000
319
+ },
320
+ {
321
+ "epoch": 10.0,
322
+ "eval_loss": 3.772279739379883,
323
+ "eval_rouge1": 26.1242,
324
+ "eval_rouge2": 10.7689,
325
+ "eval_rougeL": 22.8692,
326
+ "eval_runtime": 387.9211,
327
+ "eval_samples_per_second": 5.658,
328
+ "eval_steps_per_second": 0.567,
329
+ "step": 15280
330
+ },
331
+ {
332
+ "epoch": 4.141063318193962,
333
+ "grad_norm": 4.127313613891602,
334
+ "learning_rate": 3.619645560602013e-05,
335
+ "loss": 2.3856,
336
+ "step": 15500
337
+ },
338
+ {
339
+ "epoch": 4.2746460058776385,
340
+ "grad_norm": 3.7080917358398438,
341
+ "learning_rate": 3.575117998040787e-05,
342
+ "loss": 2.2343,
343
+ "step": 16000
344
+ },
345
+ {
346
+ "epoch": 4.408228693561314,
347
+ "grad_norm": 3.771442413330078,
348
+ "learning_rate": 3.530590435479562e-05,
349
+ "loss": 2.2498,
350
+ "step": 16500
351
+ },
352
+ {
353
+ "epoch": 4.54181138124499,
354
+ "grad_norm": 3.3883867263793945,
355
+ "learning_rate": 3.486062872918336e-05,
356
+ "loss": 2.2734,
357
+ "step": 17000
358
+ },
359
+ {
360
+ "epoch": 4.675394068928667,
361
+ "grad_norm": 3.82869029045105,
362
+ "learning_rate": 3.441535310357111e-05,
363
+ "loss": 2.2957,
364
+ "step": 17500
365
+ },
366
+ {
367
+ "epoch": 4.808976756612343,
368
+ "grad_norm": 3.480701208114624,
369
+ "learning_rate": 3.397007747795886e-05,
370
+ "loss": 2.255,
371
+ "step": 18000
372
+ },
373
+ {
374
+ "epoch": 4.94255944429602,
375
+ "grad_norm": 3.433276891708374,
376
+ "learning_rate": 3.3524801852346606e-05,
377
+ "loss": 2.2527,
378
+ "step": 18500
379
+ },
380
+ {
381
+ "epoch": 5.0,
382
+ "eval_loss": 2.624432325363159,
383
+ "eval_rouge1": 27.5622,
384
+ "eval_rouge2": 11.8188,
385
+ "eval_rougeL": 24.2339,
386
+ "eval_runtime": 862.1361,
387
+ "eval_samples_per_second": 5.439,
388
+ "eval_steps_per_second": 0.544,
389
+ "step": 18715
390
+ },
391
+ {
392
+ "epoch": 5.0761421319796955,
393
+ "grad_norm": 3.8290796279907227,
394
+ "learning_rate": 3.307952622673435e-05,
395
+ "loss": 2.1614,
396
+ "step": 19000
397
+ },
398
+ {
399
+ "epoch": 5.209724819663371,
400
+ "grad_norm": 4.148732662200928,
401
+ "learning_rate": 3.2634250601122095e-05,
402
+ "loss": 2.028,
403
+ "step": 19500
404
+ },
405
+ {
406
+ "epoch": 5.343307507347048,
407
+ "grad_norm": 3.9491941928863525,
408
+ "learning_rate": 3.218897497550984e-05,
409
+ "loss": 2.0413,
410
+ "step": 20000
411
+ },
412
+ {
413
+ "epoch": 5.476890195030724,
414
+ "grad_norm": 4.000278472900391,
415
+ "learning_rate": 3.174369934989759e-05,
416
+ "loss": 2.1058,
417
+ "step": 20500
418
+ },
419
+ {
420
+ "epoch": 5.6104728827144,
421
+ "grad_norm": 4.109215259552002,
422
+ "learning_rate": 3.129842372428534e-05,
423
+ "loss": 2.0563,
424
+ "step": 21000
425
+ },
426
+ {
427
+ "epoch": 5.744055570398077,
428
+ "grad_norm": 3.3794937133789062,
429
+ "learning_rate": 3.085314809867308e-05,
430
+ "loss": 2.0737,
431
+ "step": 21500
432
+ },
433
+ {
434
+ "epoch": 5.8776382580817526,
435
+ "grad_norm": 3.326864719390869,
436
+ "learning_rate": 3.0407872473060827e-05,
437
+ "loss": 2.085,
438
+ "step": 22000
439
+ },
440
+ {
441
+ "epoch": 6.0,
442
+ "eval_loss": 2.6439783573150635,
443
+ "eval_rouge1": 27.1127,
444
+ "eval_rouge2": 11.6364,
445
+ "eval_rougeL": 23.8039,
446
+ "eval_runtime": 879.7933,
447
+ "eval_samples_per_second": 5.33,
448
+ "eval_steps_per_second": 0.533,
449
+ "step": 22458
450
+ },
451
+ {
452
+ "epoch": 6.011220945765428,
453
+ "grad_norm": 3.5264782905578613,
454
+ "learning_rate": 2.9962596847448572e-05,
455
+ "loss": 1.8559,
456
+ "step": 22500
457
+ },
458
+ {
459
+ "epoch": 6.144803633449105,
460
+ "grad_norm": 3.7577388286590576,
461
+ "learning_rate": 2.951732122183632e-05,
462
+ "loss": 1.8982,
463
+ "step": 23000
464
+ },
465
+ {
466
+ "epoch": 6.278386321132781,
467
+ "grad_norm": 3.2084107398986816,
468
+ "learning_rate": 2.907204559622406e-05,
469
+ "loss": 1.8801,
470
+ "step": 23500
471
+ },
472
+ {
473
+ "epoch": 6.411969008816458,
474
+ "grad_norm": 4.275125503540039,
475
+ "learning_rate": 2.8626769970611812e-05,
476
+ "loss": 1.9144,
477
+ "step": 24000
478
+ },
479
+ {
480
+ "epoch": 6.545551696500134,
481
+ "grad_norm": 3.5503015518188477,
482
+ "learning_rate": 2.8181494344999553e-05,
483
+ "loss": 1.9229,
484
+ "step": 24500
485
+ },
486
+ {
487
+ "epoch": 6.67913438418381,
488
+ "grad_norm": 3.489950656890869,
489
+ "learning_rate": 2.77362187193873e-05,
490
+ "loss": 1.9033,
491
+ "step": 25000
492
+ },
493
+ {
494
+ "epoch": 6.812717071867486,
495
+ "grad_norm": 3.5493781566619873,
496
+ "learning_rate": 2.7290943093775046e-05,
497
+ "loss": 1.9058,
498
+ "step": 25500
499
+ },
500
+ {
501
+ "epoch": 6.946299759551162,
502
+ "grad_norm": 3.603605031967163,
503
+ "learning_rate": 2.6845667468162793e-05,
504
+ "loss": 1.8774,
505
+ "step": 26000
506
+ },
507
+ {
508
+ "epoch": 7.0,
509
+ "eval_loss": 2.670841693878174,
510
+ "eval_rouge1": 27.7286,
511
+ "eval_rouge2": 11.9572,
512
+ "eval_rougeL": 24.3316,
513
+ "eval_runtime": 903.5292,
514
+ "eval_samples_per_second": 5.19,
515
+ "eval_steps_per_second": 0.519,
516
+ "step": 26201
517
+ },
518
+ {
519
+ "epoch": 7.079882447234838,
520
+ "grad_norm": 3.9646942615509033,
521
+ "learning_rate": 2.640039184255054e-05,
522
+ "loss": 1.8442,
523
+ "step": 26500
524
+ },
525
+ {
526
+ "epoch": 7.213465134918515,
527
+ "grad_norm": 4.019280433654785,
528
+ "learning_rate": 2.5955116216938286e-05,
529
+ "loss": 1.729,
530
+ "step": 27000
531
+ },
532
+ {
533
+ "epoch": 7.347047822602191,
534
+ "grad_norm": 3.9131412506103516,
535
+ "learning_rate": 2.5509840591326034e-05,
536
+ "loss": 1.7656,
537
+ "step": 27500
538
+ },
539
+ {
540
+ "epoch": 7.480630510285867,
541
+ "grad_norm": 3.336970806121826,
542
+ "learning_rate": 2.5064564965713778e-05,
543
+ "loss": 1.7733,
544
+ "step": 28000
545
+ },
546
+ {
547
+ "epoch": 7.614213197969543,
548
+ "grad_norm": 3.7517452239990234,
549
+ "learning_rate": 2.4619289340101523e-05,
550
+ "loss": 1.7842,
551
+ "step": 28500
552
+ },
553
+ {
554
+ "epoch": 7.747795885653219,
555
+ "grad_norm": 3.8601882457733154,
556
+ "learning_rate": 2.4174013714489267e-05,
557
+ "loss": 1.7876,
558
+ "step": 29000
559
+ },
560
+ {
561
+ "epoch": 7.881378573336896,
562
+ "grad_norm": 3.928703784942627,
563
+ "learning_rate": 2.372873808887702e-05,
564
+ "loss": 1.7826,
565
+ "step": 29500
566
+ },
567
+ {
568
+ "epoch": 8.0,
569
+ "eval_loss": 2.7165374755859375,
570
+ "eval_rouge1": 27.5923,
571
+ "eval_rouge2": 11.9,
572
+ "eval_rougeL": 24.2586,
573
+ "eval_runtime": 898.7001,
574
+ "eval_samples_per_second": 5.218,
575
+ "eval_steps_per_second": 0.522,
576
+ "step": 29944
577
+ },
578
+ {
579
+ "epoch": 9.151921903599757,
580
+ "grad_norm": 4.411174774169922,
581
+ "learning_rate": 1.949359365466748e-05,
582
+ "loss": 1.3756,
583
+ "step": 30000
584
+ },
585
+ {
586
+ "epoch": 9.304453935326418,
587
+ "grad_norm": 4.3166985511779785,
588
+ "learning_rate": 1.8985153548911936e-05,
589
+ "loss": 1.4059,
590
+ "step": 30500
591
+ },
592
+ {
593
+ "epoch": 9.45698596705308,
594
+ "grad_norm": 3.6181671619415283,
595
+ "learning_rate": 1.8476713443156397e-05,
596
+ "loss": 1.4086,
597
+ "step": 31000
598
+ },
599
+ {
600
+ "epoch": 9.609517998779744,
601
+ "grad_norm": 4.077247142791748,
602
+ "learning_rate": 1.7968273337400855e-05,
603
+ "loss": 1.417,
604
+ "step": 31500
605
+ },
606
+ {
607
+ "epoch": 9.762050030506407,
608
+ "grad_norm": 3.4715030193328857,
609
+ "learning_rate": 1.7459833231645312e-05,
610
+ "loss": 1.4021,
611
+ "step": 32000
612
+ },
613
+ {
614
+ "epoch": 9.91458206223307,
615
+ "grad_norm": 3.7810068130493164,
616
+ "learning_rate": 1.6951393125889773e-05,
617
+ "loss": 1.4047,
618
+ "step": 32500
619
+ },
620
+ {
621
+ "epoch": 10.0,
622
+ "eval_loss": 2.808703899383545,
623
+ "eval_rouge1": 27.8772,
624
+ "eval_rouge2": 12.0303,
625
+ "eval_rougeL": 24.5034,
626
+ "eval_runtime": 743.946,
627
+ "eval_samples_per_second": 5.877,
628
+ "eval_steps_per_second": 0.589,
629
+ "step": 32780
630
+ },
631
+ {
632
+ "epoch": 10.06711409395973,
633
+ "grad_norm": 3.981015682220459,
634
+ "learning_rate": 1.644295302013423e-05,
635
+ "loss": 1.3691,
636
+ "step": 33000
637
+ },
638
+ {
639
+ "epoch": 10.219646125686394,
640
+ "grad_norm": 3.829094648361206,
641
+ "learning_rate": 1.5934512914378688e-05,
642
+ "loss": 1.3323,
643
+ "step": 33500
644
+ },
645
+ {
646
+ "epoch": 10.372178157413057,
647
+ "grad_norm": 3.636704206466675,
648
+ "learning_rate": 1.5426072808623145e-05,
649
+ "loss": 1.3337,
650
+ "step": 34000
651
+ },
652
+ {
653
+ "epoch": 10.52471018913972,
654
+ "grad_norm": 4.167761325836182,
655
+ "learning_rate": 1.4917632702867604e-05,
656
+ "loss": 1.3512,
657
+ "step": 34500
658
+ },
659
+ {
660
+ "epoch": 10.677242220866383,
661
+ "grad_norm": 3.829911470413208,
662
+ "learning_rate": 1.4409192597112062e-05,
663
+ "loss": 1.3453,
664
+ "step": 35000
665
+ },
666
+ {
667
+ "epoch": 10.829774252593044,
668
+ "grad_norm": 4.00424861907959,
669
+ "learning_rate": 1.390075249135652e-05,
670
+ "loss": 1.3308,
671
+ "step": 35500
672
+ },
673
+ {
674
+ "epoch": 10.982306284319707,
675
+ "grad_norm": 4.3669657707214355,
676
+ "learning_rate": 1.3392312385600977e-05,
677
+ "loss": 1.3781,
678
+ "step": 36000
679
+ },
680
+ {
681
+ "epoch": 11.0,
682
+ "eval_loss": 2.769470453262329,
683
+ "eval_rouge1": 28.0814,
684
+ "eval_rouge2": 12.3353,
685
+ "eval_rougeL": 24.7091,
686
+ "eval_runtime": 745.2741,
687
+ "eval_samples_per_second": 5.866,
688
+ "eval_steps_per_second": 0.588,
689
+ "step": 36058
690
+ },
691
+ {
692
+ "epoch": 11.13483831604637,
693
+ "grad_norm": 3.7836453914642334,
694
+ "learning_rate": 1.2883872279845436e-05,
695
+ "loss": 1.275,
696
+ "step": 36500
697
+ },
698
+ {
699
+ "epoch": 11.287370347773033,
700
+ "grad_norm": 4.0213751792907715,
701
+ "learning_rate": 1.2375432174089893e-05,
702
+ "loss": 1.2572,
703
+ "step": 37000
704
+ },
705
+ {
706
+ "epoch": 11.439902379499696,
707
+ "grad_norm": 3.626115083694458,
708
+ "learning_rate": 1.186699206833435e-05,
709
+ "loss": 1.2687,
710
+ "step": 37500
711
+ },
712
+ {
713
+ "epoch": 11.592434411226357,
714
+ "grad_norm": 4.4103007316589355,
715
+ "learning_rate": 1.1358551962578808e-05,
716
+ "loss": 1.2807,
717
+ "step": 38000
718
+ },
719
+ {
720
+ "epoch": 11.74496644295302,
721
+ "grad_norm": 3.681063413619995,
722
+ "learning_rate": 1.0850111856823267e-05,
723
+ "loss": 1.2986,
724
+ "step": 38500
725
+ },
726
+ {
727
+ "epoch": 11.897498474679683,
728
+ "grad_norm": 4.0368170738220215,
729
+ "learning_rate": 1.0341671751067725e-05,
730
+ "loss": 1.2784,
731
+ "step": 39000
732
+ },
733
+ {
734
+ "epoch": 12.0,
735
+ "eval_loss": 2.83615779876709,
736
+ "eval_rouge1": 28.2406,
737
+ "eval_rouge2": 12.197,
738
+ "eval_rougeL": 24.7367,
739
+ "eval_runtime": 733.2323,
740
+ "eval_samples_per_second": 5.963,
741
+ "eval_steps_per_second": 0.597,
742
+ "step": 39336
743
+ },
744
+ {
745
+ "epoch": 12.050030506406346,
746
+ "grad_norm": 4.6332173347473145,
747
+ "learning_rate": 9.833231645312182e-06,
748
+ "loss": 1.2071,
749
+ "step": 39500
750
+ },
751
+ {
752
+ "epoch": 12.202562538133009,
753
+ "grad_norm": 3.7481937408447266,
754
+ "learning_rate": 9.324791539556641e-06,
755
+ "loss": 1.2085,
756
+ "step": 40000
757
+ },
758
+ {
759
+ "epoch": 12.35509456985967,
760
+ "grad_norm": 4.701822757720947,
761
+ "learning_rate": 8.816351433801099e-06,
762
+ "loss": 1.2225,
763
+ "step": 40500
764
+ },
765
+ {
766
+ "epoch": 12.507626601586333,
767
+ "grad_norm": 4.165287494659424,
768
+ "learning_rate": 8.307911328045556e-06,
769
+ "loss": 1.2187,
770
+ "step": 41000
771
+ },
772
+ {
773
+ "epoch": 12.660158633312996,
774
+ "grad_norm": 4.279840469360352,
775
+ "learning_rate": 7.799471222290014e-06,
776
+ "loss": 1.2157,
777
+ "step": 41500
778
+ },
779
+ {
780
+ "epoch": 12.812690665039659,
781
+ "grad_norm": 4.762269973754883,
782
+ "learning_rate": 7.291031116534472e-06,
783
+ "loss": 1.2277,
784
+ "step": 42000
785
+ },
786
+ {
787
+ "epoch": 12.965222696766322,
788
+ "grad_norm": 4.380380153656006,
789
+ "learning_rate": 6.782591010778931e-06,
790
+ "loss": 1.2371,
791
+ "step": 42500
792
+ },
793
+ {
794
+ "epoch": 13.0,
795
+ "eval_loss": 2.921924114227295,
796
+ "eval_rouge1": 27.8556,
797
+ "eval_rouge2": 12.033,
798
+ "eval_rougeL": 24.4386,
799
+ "eval_runtime": 791.9797,
800
+ "eval_samples_per_second": 5.52,
801
+ "eval_steps_per_second": 0.553,
802
+ "step": 42614
803
+ },
804
+ {
805
+ "epoch": 12.710611882944132,
806
+ "grad_norm": 4.347652435302734,
807
+ "learning_rate": 7.631293723519558e-06,
808
+ "loss": 1.6632,
809
+ "step": 43000
810
+ },
811
+ {
812
+ "epoch": 12.858409695536507,
813
+ "grad_norm": 4.261178493499756,
814
+ "learning_rate": 7.138634348211647e-06,
815
+ "loss": 1.627,
816
+ "step": 43500
817
+ },
818
+ {
819
+ "epoch": 13.0,
820
+ "eval_loss": 3.073051929473877,
821
+ "eval_rouge1": 29.5215,
822
+ "eval_rouge2": 13.1839,
823
+ "eval_rougeL": 25.617,
824
+ "eval_runtime": 744.9289,
825
+ "eval_samples_per_second": 6.049,
826
+ "eval_steps_per_second": 0.605,
827
+ "step": 43979
828
+ },
829
+ {
830
+ "epoch": 13.00620750812888,
831
+ "grad_norm": 5.098820686340332,
832
+ "learning_rate": 6.645974972903735e-06,
833
+ "loss": 1.5807,
834
+ "step": 44000
835
+ },
836
+ {
837
+ "epoch": 13.154005320721254,
838
+ "grad_norm": 3.9083046913146973,
839
+ "learning_rate": 6.153315597595822e-06,
840
+ "loss": 1.562,
841
+ "step": 44500
842
+ },
843
+ {
844
+ "epoch": 13.301803133313626,
845
+ "grad_norm": 4.274121284484863,
846
+ "learning_rate": 5.66065622228791e-06,
847
+ "loss": 1.5721,
848
+ "step": 45000
849
+ },
850
+ {
851
+ "epoch": 13.449600945906,
852
+ "grad_norm": 4.222695827484131,
853
+ "learning_rate": 5.167996846979999e-06,
854
+ "loss": 1.5415,
855
+ "step": 45500
856
+ },
857
+ {
858
+ "epoch": 13.597398758498374,
859
+ "grad_norm": 4.467597484588623,
860
+ "learning_rate": 4.675337471672086e-06,
861
+ "loss": 1.5718,
862
+ "step": 46000
863
+ },
864
+ {
865
+ "epoch": 13.745196571090748,
866
+ "grad_norm": 4.109827995300293,
867
+ "learning_rate": 4.182678096364174e-06,
868
+ "loss": 1.5285,
869
+ "step": 46500
870
+ },
871
+ {
872
+ "epoch": 13.89299438368312,
873
+ "grad_norm": 4.74594783782959,
874
+ "learning_rate": 3.690018721056262e-06,
875
+ "loss": 1.5215,
876
+ "step": 47000
877
+ },
878
+ {
879
+ "epoch": 14.0,
880
+ "eval_loss": 3.0527069568634033,
881
+ "eval_rouge1": 29.7103,
882
+ "eval_rouge2": 13.2821,
883
+ "eval_rougeL": 25.7681,
884
+ "eval_runtime": 741.3618,
885
+ "eval_samples_per_second": 6.078,
886
+ "eval_steps_per_second": 0.608,
887
+ "step": 47362
888
+ },
889
+ {
890
+ "epoch": 14.040792196275495,
891
+ "grad_norm": 4.313952922821045,
892
+ "learning_rate": 1.4898019509311262e-05,
893
+ "loss": 1.4804,
894
+ "step": 47500
895
+ },
896
+ {
897
+ "epoch": 14.18859000886787,
898
+ "grad_norm": 4.262555122375488,
899
+ "learning_rate": 1.4528524977830329e-05,
900
+ "loss": 1.534,
901
+ "step": 48000
902
+ },
903
+ {
904
+ "epoch": 14.336387821460242,
905
+ "grad_norm": 4.611397743225098,
906
+ "learning_rate": 1.4159030446349395e-05,
907
+ "loss": 1.5168,
908
+ "step": 48500
909
+ },
910
+ {
911
+ "epoch": 14.484185634052617,
912
+ "grad_norm": 4.340530872344971,
913
+ "learning_rate": 1.378953591486846e-05,
914
+ "loss": 1.5254,
915
+ "step": 49000
916
+ },
917
+ {
918
+ "epoch": 14.63198344664499,
919
+ "grad_norm": 4.0280375480651855,
920
+ "learning_rate": 1.3420041383387528e-05,
921
+ "loss": 1.5317,
922
+ "step": 49500
923
+ },
924
+ {
925
+ "epoch": 14.779781259237364,
926
+ "grad_norm": 4.776717662811279,
927
+ "learning_rate": 1.3050546851906593e-05,
928
+ "loss": 1.5312,
929
+ "step": 50000
930
+ },
931
+ {
932
+ "epoch": 14.927579071829737,
933
+ "grad_norm": 4.440686225891113,
934
+ "learning_rate": 1.2681052320425657e-05,
935
+ "loss": 1.5348,
936
+ "step": 50500
937
+ },
938
+ {
939
+ "epoch": 15.0,
940
+ "eval_loss": 3.036102056503296,
941
+ "eval_rouge1": 29.8805,
942
+ "eval_rouge2": 13.4576,
943
+ "eval_rougeL": 25.9729,
944
+ "eval_runtime": 726.153,
945
+ "eval_samples_per_second": 6.205,
946
+ "eval_steps_per_second": 0.621,
947
+ "step": 50745
948
+ },
949
+ {
950
+ "epoch": 15.075376884422111,
951
+ "grad_norm": 4.464206695556641,
952
+ "learning_rate": 1.2311557788944725e-05,
953
+ "loss": 1.4799,
954
+ "step": 51000
955
+ },
956
+ {
957
+ "epoch": 15.223174697014484,
958
+ "grad_norm": 4.814591884613037,
959
+ "learning_rate": 1.194206325746379e-05,
960
+ "loss": 1.4313,
961
+ "step": 51500
962
+ },
963
+ {
964
+ "epoch": 15.370972509606858,
965
+ "grad_norm": 3.8840887546539307,
966
+ "learning_rate": 1.1572568725982856e-05,
967
+ "loss": 1.458,
968
+ "step": 52000
969
+ },
970
+ {
971
+ "epoch": 15.51877032219923,
972
+ "grad_norm": 4.051957130432129,
973
+ "learning_rate": 1.1203074194501923e-05,
974
+ "loss": 1.4723,
975
+ "step": 52500
976
+ },
977
+ {
978
+ "epoch": 15.666568134791605,
979
+ "grad_norm": 4.178130626678467,
980
+ "learning_rate": 1.0833579663020987e-05,
981
+ "loss": 1.4597,
982
+ "step": 53000
983
+ },
984
+ {
985
+ "epoch": 15.814365947383978,
986
+ "grad_norm": 4.45131254196167,
987
+ "learning_rate": 1.0464085131540054e-05,
988
+ "loss": 1.4332,
989
+ "step": 53500
990
+ },
991
+ {
992
+ "epoch": 15.962163759976352,
993
+ "grad_norm": 4.57073974609375,
994
+ "learning_rate": 1.009459060005912e-05,
995
+ "loss": 1.4437,
996
+ "step": 54000
997
+ },
998
+ {
999
+ "epoch": 16.0,
1000
+ "eval_loss": 3.0735461711883545,
1001
+ "eval_rouge1": 29.9002,
1002
+ "eval_rouge2": 13.5403,
1003
+ "eval_rougeL": 26.0194,
1004
+ "eval_runtime": 722.6691,
1005
+ "eval_samples_per_second": 6.235,
1006
+ "eval_steps_per_second": 0.624,
1007
+ "step": 54128
1008
+ },
1009
+ {
1010
+ "epoch": 16.109961572568725,
1011
+ "grad_norm": 4.4488444328308105,
1012
+ "learning_rate": 9.725096068578185e-06,
1013
+ "loss": 1.3652,
1014
+ "step": 54500
1015
+ },
1016
+ {
1017
+ "epoch": 16.2577593851611,
1018
+ "grad_norm": 4.422569751739502,
1019
+ "learning_rate": 9.355601537097251e-06,
1020
+ "loss": 1.3663,
1021
+ "step": 55000
1022
+ },
1023
+ {
1024
+ "epoch": 16.405557197753474,
1025
+ "grad_norm": 4.383566379547119,
1026
+ "learning_rate": 8.986107005616317e-06,
1027
+ "loss": 1.3829,
1028
+ "step": 55500
1029
+ },
1030
+ {
1031
+ "epoch": 16.553355010345847,
1032
+ "grad_norm": 3.979015827178955,
1033
+ "learning_rate": 8.616612474135382e-06,
1034
+ "loss": 1.3908,
1035
+ "step": 56000
1036
+ },
1037
+ {
1038
+ "epoch": 16.70115282293822,
1039
+ "grad_norm": 4.212048530578613,
1040
+ "learning_rate": 8.247117942654448e-06,
1041
+ "loss": 1.3937,
1042
+ "step": 56500
1043
+ },
1044
+ {
1045
+ "epoch": 16.848950635530596,
1046
+ "grad_norm": 5.7812581062316895,
1047
+ "learning_rate": 7.877623411173515e-06,
1048
+ "loss": 1.3842,
1049
+ "step": 57000
1050
+ },
1051
+ {
1052
+ "epoch": 16.99674844812297,
1053
+ "grad_norm": 4.071883678436279,
1054
+ "learning_rate": 7.50812887969258e-06,
1055
+ "loss": 1.3967,
1056
+ "step": 57500
1057
+ },
1058
+ {
1059
+ "epoch": 17.0,
1060
+ "eval_loss": 3.1532084941864014,
1061
+ "eval_rouge1": 29.8426,
1062
+ "eval_rouge2": 13.3464,
1063
+ "eval_rougeL": 25.9413,
1064
+ "eval_runtime": 715.7403,
1065
+ "eval_samples_per_second": 6.296,
1066
+ "eval_steps_per_second": 0.63,
1067
+ "step": 57511
1068
+ },
1069
+ {
1070
+ "epoch": 17.14454626071534,
1071
+ "grad_norm": 4.053774833679199,
1072
+ "learning_rate": 7.138634348211647e-06,
1073
+ "loss": 1.32,
1074
+ "step": 58000
1075
+ },
1076
+ {
1077
+ "epoch": 17.292344073307714,
1078
+ "grad_norm": 4.921830654144287,
1079
+ "learning_rate": 6.769139816730713e-06,
1080
+ "loss": 1.3326,
1081
+ "step": 58500
1082
+ },
1083
+ {
1084
+ "epoch": 17.44014188590009,
1085
+ "grad_norm": 4.038568496704102,
1086
+ "learning_rate": 6.399645285249779e-06,
1087
+ "loss": 1.3259,
1088
+ "step": 59000
1089
+ },
1090
+ {
1091
+ "epoch": 17.587939698492463,
1092
+ "grad_norm": 4.252479553222656,
1093
+ "learning_rate": 6.030150753768844e-06,
1094
+ "loss": 1.344,
1095
+ "step": 59500
1096
+ },
1097
+ {
1098
+ "epoch": 17.735737511084835,
1099
+ "grad_norm": 3.9074063301086426,
1100
+ "learning_rate": 5.66065622228791e-06,
1101
+ "loss": 1.3381,
1102
+ "step": 60000
1103
+ },
1104
+ {
1105
+ "epoch": 17.883535323677208,
1106
+ "grad_norm": 4.212071418762207,
1107
+ "learning_rate": 5.291161690806976e-06,
1108
+ "loss": 1.3383,
1109
+ "step": 60500
1110
+ },
1111
+ {
1112
+ "epoch": 18.0,
1113
+ "eval_loss": 3.163597822189331,
1114
+ "eval_rouge1": 29.8838,
1115
+ "eval_rouge2": 13.3291,
1116
+ "eval_rougeL": 25.9562,
1117
+ "eval_runtime": 755.5618,
1118
+ "eval_samples_per_second": 5.964,
1119
+ "eval_steps_per_second": 0.597,
1120
+ "step": 60894
1121
+ },
1122
+ {
1123
+ "epoch": 18.031333136269584,
1124
+ "grad_norm": 3.875527858734131,
1125
+ "learning_rate": 4.921667159326042e-06,
1126
+ "loss": 1.3315,
1127
+ "step": 61000
1128
+ },
1129
+ {
1130
+ "epoch": 18.179130948861957,
1131
+ "grad_norm": 4.280786991119385,
1132
+ "learning_rate": 4.552172627845109e-06,
1133
+ "loss": 1.2882,
1134
+ "step": 61500
1135
+ },
1136
+ {
1137
+ "epoch": 18.32692876145433,
1138
+ "grad_norm": 5.005005836486816,
1139
+ "learning_rate": 4.182678096364174e-06,
1140
+ "loss": 1.2949,
1141
+ "step": 62000
1142
+ },
1143
+ {
1144
+ "epoch": 18.474726574046706,
1145
+ "grad_norm": 4.365475177764893,
1146
+ "learning_rate": 3.81318356488324e-06,
1147
+ "loss": 1.2823,
1148
+ "step": 62500
1149
+ },
1150
+ {
1151
+ "epoch": 18.62252438663908,
1152
+ "grad_norm": 3.963200807571411,
1153
+ "learning_rate": 3.443689033402306e-06,
1154
+ "loss": 1.3049,
1155
+ "step": 63000
1156
+ },
1157
+ {
1158
+ "epoch": 18.77032219923145,
1159
+ "grad_norm": 4.525120735168457,
1160
+ "learning_rate": 3.074194501921372e-06,
1161
+ "loss": 1.3062,
1162
+ "step": 63500
1163
+ },
1164
+ {
1165
+ "epoch": 18.918120011823824,
1166
+ "grad_norm": 5.0892205238342285,
1167
+ "learning_rate": 2.7046999704404375e-06,
1168
+ "loss": 1.3109,
1169
+ "step": 64000
1170
+ },
1171
+ {
1172
+ "epoch": 19.0,
1173
+ "eval_loss": 3.1824729442596436,
1174
+ "eval_rouge1": 29.814,
1175
+ "eval_rouge2": 13.3444,
1176
+ "eval_rougeL": 25.9306,
1177
+ "eval_runtime": 755.0118,
1178
+ "eval_samples_per_second": 5.968,
1179
+ "eval_steps_per_second": 0.597,
1180
+ "step": 64277
1181
+ },
1182
+ {
1183
+ "epoch": 17.23216671119423,
1184
+ "grad_norm": 4.376983165740967,
1185
+ "learning_rate": 1.5619124362073598e-05,
1186
+ "loss": 2.0195,
1187
+ "step": 64500
1188
+ },
1189
+ {
1190
+ "epoch": 17.365749398877906,
1191
+ "grad_norm": 4.5946478843688965,
1192
+ "learning_rate": 1.5350523771152298e-05,
1193
+ "loss": 1.7778,
1194
+ "step": 65000
1195
+ },
1196
+ {
1197
+ "epoch": 17.49933208656158,
1198
+ "grad_norm": 4.226222515106201,
1199
+ "learning_rate": 1.5081923180230998e-05,
1200
+ "loss": 1.7337,
1201
+ "step": 65500
1202
+ },
1203
+ {
1204
+ "epoch": 17.632914774245258,
1205
+ "grad_norm": 4.022632598876953,
1206
+ "learning_rate": 1.4813322589309698e-05,
1207
+ "loss": 1.7263,
1208
+ "step": 66000
1209
+ },
1210
+ {
1211
+ "epoch": 17.766497461928935,
1212
+ "grad_norm": 150.4783172607422,
1213
+ "learning_rate": 1.4544721998388396e-05,
1214
+ "loss": 1.7239,
1215
+ "step": 66500
1216
+ },
1217
+ {
1218
+ "epoch": 17.90008014961261,
1219
+ "grad_norm": 4.1645026206970215,
1220
+ "learning_rate": 1.4276121407467096e-05,
1221
+ "loss": 1.7483,
1222
+ "step": 67000
1223
+ },
1224
+ {
1225
+ "epoch": 18.0,
1226
+ "eval_loss": 2.929865837097168,
1227
+ "eval_rouge1": 25.8395,
1228
+ "eval_rouge2": 10.2436,
1229
+ "eval_rougeL": 22.6099,
1230
+ "eval_runtime": 793.8225,
1231
+ "eval_samples_per_second": 5.907,
1232
+ "eval_steps_per_second": 0.591,
1233
+ "step": 67374
1234
+ },
1235
+ {
1236
+ "epoch": 18.033662837296287,
1237
+ "grad_norm": 4.0770721435546875,
1238
+ "learning_rate": 1.4007520816545796e-05,
1239
+ "loss": 1.6939,
1240
+ "step": 67500
1241
+ },
1242
+ {
1243
+ "epoch": 18.16724552497996,
1244
+ "grad_norm": 3.565656900405884,
1245
+ "learning_rate": 1.3738920225624496e-05,
1246
+ "loss": 1.6307,
1247
+ "step": 68000
1248
+ },
1249
+ {
1250
+ "epoch": 18.30082821266364,
1251
+ "grad_norm": 4.281002998352051,
1252
+ "learning_rate": 1.3470319634703196e-05,
1253
+ "loss": 1.6194,
1254
+ "step": 68500
1255
+ },
1256
+ {
1257
+ "epoch": 18.434410900347316,
1258
+ "grad_norm": 3.6802380084991455,
1259
+ "learning_rate": 1.3201719043781896e-05,
1260
+ "loss": 1.6239,
1261
+ "step": 69000
1262
+ },
1263
+ {
1264
+ "epoch": 18.56799358803099,
1265
+ "grad_norm": 4.384114742279053,
1266
+ "learning_rate": 1.2933118452860596e-05,
1267
+ "loss": 1.6396,
1268
+ "step": 69500
1269
+ },
1270
+ {
1271
+ "epoch": 18.701576275714668,
1272
+ "grad_norm": 3.82995343208313,
1273
+ "learning_rate": 1.2664517861939296e-05,
1274
+ "loss": 1.6364,
1275
+ "step": 70000
1276
+ },
1277
+ {
1278
+ "epoch": 18.835158963398342,
1279
+ "grad_norm": 3.700697898864746,
1280
+ "learning_rate": 1.2395917271017998e-05,
1281
+ "loss": 1.6401,
1282
+ "step": 70500
1283
+ },
1284
+ {
1285
+ "epoch": 18.96874165108202,
1286
+ "grad_norm": 4.294804096221924,
1287
+ "learning_rate": 1.2127316680096696e-05,
1288
+ "loss": 1.6425,
1289
+ "step": 71000
1290
+ },
1291
+ {
1292
+ "epoch": 19.0,
1293
+ "eval_loss": 3.020838499069214,
1294
+ "eval_rouge1": 25.7158,
1295
+ "eval_rouge2": 10.2414,
1296
+ "eval_rougeL": 22.5002,
1297
+ "eval_runtime": 811.7808,
1298
+ "eval_samples_per_second": 5.776,
1299
+ "eval_steps_per_second": 0.578,
1300
+ "step": 71117
1301
+ },
1302
+ {
1303
+ "epoch": 19.102324338765698,
1304
+ "grad_norm": 3.450388193130493,
1305
+ "learning_rate": 1.646426619709758e-05,
1306
+ "loss": 2.1375,
1307
+ "step": 71500
1308
+ },
1309
+ {
1310
+ "epoch": 19.23590702644937,
1311
+ "grad_norm": 3.6473982334136963,
1312
+ "learning_rate": 1.5705264436213492e-05,
1313
+ "loss": 2.0095,
1314
+ "step": 72000
1315
+ },
1316
+ {
1317
+ "epoch": 19.36948971413305,
1318
+ "grad_norm": 3.5623624324798584,
1319
+ "learning_rate": 1.4946262675329408e-05,
1320
+ "loss": 1.9427,
1321
+ "step": 72500
1322
+ },
1323
+ {
1324
+ "epoch": 19.503072401816723,
1325
+ "grad_norm": 3.6597957611083984,
1326
+ "learning_rate": 1.4187260914445322e-05,
1327
+ "loss": 1.9029,
1328
+ "step": 73000
1329
+ },
1330
+ {
1331
+ "epoch": 19.6366550895004,
1332
+ "grad_norm": 3.924410104751587,
1333
+ "learning_rate": 1.3428259153561237e-05,
1334
+ "loss": 1.8831,
1335
+ "step": 73500
1336
+ },
1337
+ {
1338
+ "epoch": 19.77023777718408,
1339
+ "grad_norm": 3.5461361408233643,
1340
+ "learning_rate": 1.2669257392677152e-05,
1341
+ "loss": 1.8919,
1342
+ "step": 74000
1343
+ },
1344
+ {
1345
+ "epoch": 19.903820464867753,
1346
+ "grad_norm": 3.7633907794952393,
1347
+ "learning_rate": 1.1910255631793067e-05,
1348
+ "loss": 1.8611,
1349
+ "step": 74500
1350
+ },
1351
+ {
1352
+ "epoch": 20.0,
1353
+ "eval_loss": 2.5265896320343018,
1354
+ "eval_rouge1": 27.4886,
1355
+ "eval_rouge2": 11.6109,
1356
+ "eval_rougeL": 24.0414,
1357
+ "eval_runtime": 795.0309,
1358
+ "eval_samples_per_second": 5.898,
1359
+ "eval_steps_per_second": 0.59,
1360
+ "step": 74860
1361
+ },
1362
+ {
1363
+ "epoch": 20.03740315255143,
1364
+ "grad_norm": 3.6295604705810547,
1365
+ "learning_rate": 1.115125387090898e-05,
1366
+ "loss": 1.8162,
1367
+ "step": 75000
1368
+ },
1369
+ {
1370
+ "epoch": 20.170985840235105,
1371
+ "grad_norm": 3.809011697769165,
1372
+ "learning_rate": 1.0392252110024895e-05,
1373
+ "loss": 1.7436,
1374
+ "step": 75500
1375
+ },
1376
+ {
1377
+ "epoch": 20.304568527918782,
1378
+ "grad_norm": 3.8579320907592773,
1379
+ "learning_rate": 9.63325034914081e-06,
1380
+ "loss": 1.7372,
1381
+ "step": 76000
1382
+ },
1383
+ {
1384
+ "epoch": 20.438151215602456,
1385
+ "grad_norm": 4.437480926513672,
1386
+ "learning_rate": 8.874248588256725e-06,
1387
+ "loss": 1.7583,
1388
+ "step": 76500
1389
+ },
1390
+ {
1391
+ "epoch": 20.571733903286134,
1392
+ "grad_norm": 3.335023880004883,
1393
+ "learning_rate": 8.115246827372639e-06,
1394
+ "loss": 1.7373,
1395
+ "step": 77000
1396
+ },
1397
+ {
1398
+ "epoch": 20.70531659096981,
1399
+ "grad_norm": 3.7852025032043457,
1400
+ "learning_rate": 7.356245066488554e-06,
1401
+ "loss": 1.7559,
1402
+ "step": 77500
1403
+ },
1404
+ {
1405
+ "epoch": 20.838899278653486,
1406
+ "grad_norm": 3.836491823196411,
1407
+ "learning_rate": 6.59724330560447e-06,
1408
+ "loss": 1.7193,
1409
+ "step": 78000
1410
+ },
1411
+ {
1412
+ "epoch": 20.972481966337163,
1413
+ "grad_norm": 3.511979341506958,
1414
+ "learning_rate": 5.838241544720384e-06,
1415
+ "loss": 1.741,
1416
+ "step": 78500
1417
+ },
1418
+ {
1419
+ "epoch": 21.0,
1420
+ "eval_loss": 2.5569753646850586,
1421
+ "eval_rouge1": 27.3556,
1422
+ "eval_rouge2": 11.6363,
1423
+ "eval_rougeL": 24.0256,
1424
+ "eval_runtime": 807.3562,
1425
+ "eval_samples_per_second": 5.808,
1426
+ "eval_steps_per_second": 0.581,
1427
+ "step": 78603
1428
+ }
1429
+ ],
1430
+ "logging_steps": 500,
1431
+ "max_steps": 82346,
1432
+ "num_input_tokens_seen": 0,
1433
+ "num_train_epochs": 22,
1434
+ "save_steps": 500,
1435
+ "stateful_callbacks": {
1436
+ "TrainerControl": {
1437
+ "args": {
1438
+ "should_epoch_stop": false,
1439
+ "should_evaluate": false,
1440
+ "should_log": false,
1441
+ "should_save": true,
1442
+ "should_training_stop": false
1443
+ },
1444
+ "attributes": {}
1445
+ }
1446
+ },
1447
+ "total_flos": 4.003847093358305e+17,
1448
+ "train_batch_size": 10,
1449
+ "trial_name": null,
1450
+ "trial_params": null
1451
+ }
checkpoint-78603/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:36e66a5bcd0958ab230e9fbbd79edbd311edcdb14487846a14ab08602e8738ad
3
+ size 7032