Training in progress, epoch 0
Browse files- .gitignore +1 -0
- config.json +56 -0
- last-checkpoint/config.json +56 -0
- last-checkpoint/optimizer.pt +3 -0
- last-checkpoint/pytorch_model.bin +3 -0
- last-checkpoint/rng_state.pth +3 -0
- last-checkpoint/scheduler.pt +3 -0
- last-checkpoint/special_tokens_map.json +110 -0
- last-checkpoint/spiece.model +3 -0
- last-checkpoint/tokenizer_config.json +122 -0
- last-checkpoint/trainer_state.json +28 -0
- last-checkpoint/training_args.bin +3 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +110 -0
- spiece.model +3 -0
- tokenizer_config.json +122 -0
- training_args.bin +3 -0
.gitignore
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
checkpoint-*/
|
config.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "google/pegasus-multi_news",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"activation_function": "relu",
|
| 5 |
+
"add_bias_logits": false,
|
| 6 |
+
"add_final_layer_norm": true,
|
| 7 |
+
"architectures": [
|
| 8 |
+
"PegasusForConditionalGeneration"
|
| 9 |
+
],
|
| 10 |
+
"attention_dropout": 0.1,
|
| 11 |
+
"bos_token_id": 0,
|
| 12 |
+
"classif_dropout": 0.0,
|
| 13 |
+
"classifier_dropout": 0.0,
|
| 14 |
+
"d_model": 1024,
|
| 15 |
+
"decoder_attention_heads": 16,
|
| 16 |
+
"decoder_ffn_dim": 4096,
|
| 17 |
+
"decoder_layerdrop": 0.0,
|
| 18 |
+
"decoder_layers": 16,
|
| 19 |
+
"decoder_start_token_id": 0,
|
| 20 |
+
"dropout": 0.1,
|
| 21 |
+
"encoder_attention_heads": 16,
|
| 22 |
+
"encoder_ffn_dim": 4096,
|
| 23 |
+
"encoder_layerdrop": 0.0,
|
| 24 |
+
"encoder_layers": 16,
|
| 25 |
+
"eos_token_id": 1,
|
| 26 |
+
"extra_pos_embeddings": 1,
|
| 27 |
+
"forced_eos_token_id": 1,
|
| 28 |
+
"id2label": {
|
| 29 |
+
"0": "LABEL_0",
|
| 30 |
+
"1": "LABEL_1",
|
| 31 |
+
"2": "LABEL_2"
|
| 32 |
+
},
|
| 33 |
+
"init_std": 0.02,
|
| 34 |
+
"is_encoder_decoder": true,
|
| 35 |
+
"label2id": {
|
| 36 |
+
"LABEL_0": 0,
|
| 37 |
+
"LABEL_1": 1,
|
| 38 |
+
"LABEL_2": 2
|
| 39 |
+
},
|
| 40 |
+
"length_penalty": 0.8,
|
| 41 |
+
"max_length": 256,
|
| 42 |
+
"max_position_embeddings": 1024,
|
| 43 |
+
"min_length": 32,
|
| 44 |
+
"model_type": "pegasus",
|
| 45 |
+
"normalize_before": true,
|
| 46 |
+
"normalize_embedding": false,
|
| 47 |
+
"num_beams": 8,
|
| 48 |
+
"num_hidden_layers": 16,
|
| 49 |
+
"pad_token_id": 0,
|
| 50 |
+
"scale_embedding": true,
|
| 51 |
+
"static_position_embeddings": true,
|
| 52 |
+
"torch_dtype": "float32",
|
| 53 |
+
"transformers_version": "4.21.3",
|
| 54 |
+
"use_cache": true,
|
| 55 |
+
"vocab_size": 96103
|
| 56 |
+
}
|
last-checkpoint/config.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "google/pegasus-multi_news",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"activation_function": "relu",
|
| 5 |
+
"add_bias_logits": false,
|
| 6 |
+
"add_final_layer_norm": true,
|
| 7 |
+
"architectures": [
|
| 8 |
+
"PegasusForConditionalGeneration"
|
| 9 |
+
],
|
| 10 |
+
"attention_dropout": 0.1,
|
| 11 |
+
"bos_token_id": 0,
|
| 12 |
+
"classif_dropout": 0.0,
|
| 13 |
+
"classifier_dropout": 0.0,
|
| 14 |
+
"d_model": 1024,
|
| 15 |
+
"decoder_attention_heads": 16,
|
| 16 |
+
"decoder_ffn_dim": 4096,
|
| 17 |
+
"decoder_layerdrop": 0.0,
|
| 18 |
+
"decoder_layers": 16,
|
| 19 |
+
"decoder_start_token_id": 0,
|
| 20 |
+
"dropout": 0.1,
|
| 21 |
+
"encoder_attention_heads": 16,
|
| 22 |
+
"encoder_ffn_dim": 4096,
|
| 23 |
+
"encoder_layerdrop": 0.0,
|
| 24 |
+
"encoder_layers": 16,
|
| 25 |
+
"eos_token_id": 1,
|
| 26 |
+
"extra_pos_embeddings": 1,
|
| 27 |
+
"forced_eos_token_id": 1,
|
| 28 |
+
"id2label": {
|
| 29 |
+
"0": "LABEL_0",
|
| 30 |
+
"1": "LABEL_1",
|
| 31 |
+
"2": "LABEL_2"
|
| 32 |
+
},
|
| 33 |
+
"init_std": 0.02,
|
| 34 |
+
"is_encoder_decoder": true,
|
| 35 |
+
"label2id": {
|
| 36 |
+
"LABEL_0": 0,
|
| 37 |
+
"LABEL_1": 1,
|
| 38 |
+
"LABEL_2": 2
|
| 39 |
+
},
|
| 40 |
+
"length_penalty": 0.8,
|
| 41 |
+
"max_length": 256,
|
| 42 |
+
"max_position_embeddings": 1024,
|
| 43 |
+
"min_length": 32,
|
| 44 |
+
"model_type": "pegasus",
|
| 45 |
+
"normalize_before": true,
|
| 46 |
+
"normalize_embedding": false,
|
| 47 |
+
"num_beams": 8,
|
| 48 |
+
"num_hidden_layers": 16,
|
| 49 |
+
"pad_token_id": 0,
|
| 50 |
+
"scale_embedding": true,
|
| 51 |
+
"static_position_embeddings": true,
|
| 52 |
+
"torch_dtype": "float32",
|
| 53 |
+
"transformers_version": "4.21.3",
|
| 54 |
+
"use_cache": true,
|
| 55 |
+
"vocab_size": 96103
|
| 56 |
+
}
|
last-checkpoint/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:49435d42545218e02ee2a0fbeb8767a93b93b1b71a0a5df10399893c38e468a7
|
| 3 |
+
size 4549979235
|
last-checkpoint/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:85496f3b10fe45c0b541485a306b2663ae79d5d3db48b3dbac33200f1b225f94
|
| 3 |
+
size 2283790449
|
last-checkpoint/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f09b94cd083ef4afcb2f02d00def27deaeadf52c489f3a2209402791c62e43d5
|
| 3 |
+
size 13547
|
last-checkpoint/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2092dc5378c143bd0e8b624b136aa9e3c6ff6d2d063045cbd2b9e54e077c83ae
|
| 3 |
+
size 623
|
last-checkpoint/special_tokens_map.json
ADDED
|
@@ -0,0 +1,110 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<mask_1>",
|
| 4 |
+
"<unk_2>",
|
| 5 |
+
"<unk_3>",
|
| 6 |
+
"<unk_4>",
|
| 7 |
+
"<unk_5>",
|
| 8 |
+
"<unk_6>",
|
| 9 |
+
"<unk_7>",
|
| 10 |
+
"<unk_8>",
|
| 11 |
+
"<unk_9>",
|
| 12 |
+
"<unk_10>",
|
| 13 |
+
"<unk_11>",
|
| 14 |
+
"<unk_12>",
|
| 15 |
+
"<unk_13>",
|
| 16 |
+
"<unk_14>",
|
| 17 |
+
"<unk_15>",
|
| 18 |
+
"<unk_16>",
|
| 19 |
+
"<unk_17>",
|
| 20 |
+
"<unk_18>",
|
| 21 |
+
"<unk_19>",
|
| 22 |
+
"<unk_20>",
|
| 23 |
+
"<unk_21>",
|
| 24 |
+
"<unk_22>",
|
| 25 |
+
"<unk_23>",
|
| 26 |
+
"<unk_24>",
|
| 27 |
+
"<unk_25>",
|
| 28 |
+
"<unk_26>",
|
| 29 |
+
"<unk_27>",
|
| 30 |
+
"<unk_28>",
|
| 31 |
+
"<unk_29>",
|
| 32 |
+
"<unk_30>",
|
| 33 |
+
"<unk_31>",
|
| 34 |
+
"<unk_32>",
|
| 35 |
+
"<unk_33>",
|
| 36 |
+
"<unk_34>",
|
| 37 |
+
"<unk_35>",
|
| 38 |
+
"<unk_36>",
|
| 39 |
+
"<unk_37>",
|
| 40 |
+
"<unk_38>",
|
| 41 |
+
"<unk_39>",
|
| 42 |
+
"<unk_40>",
|
| 43 |
+
"<unk_41>",
|
| 44 |
+
"<unk_42>",
|
| 45 |
+
"<unk_43>",
|
| 46 |
+
"<unk_44>",
|
| 47 |
+
"<unk_45>",
|
| 48 |
+
"<unk_46>",
|
| 49 |
+
"<unk_47>",
|
| 50 |
+
"<unk_48>",
|
| 51 |
+
"<unk_49>",
|
| 52 |
+
"<unk_50>",
|
| 53 |
+
"<unk_51>",
|
| 54 |
+
"<unk_52>",
|
| 55 |
+
"<unk_53>",
|
| 56 |
+
"<unk_54>",
|
| 57 |
+
"<unk_55>",
|
| 58 |
+
"<unk_56>",
|
| 59 |
+
"<unk_57>",
|
| 60 |
+
"<unk_58>",
|
| 61 |
+
"<unk_59>",
|
| 62 |
+
"<unk_60>",
|
| 63 |
+
"<unk_61>",
|
| 64 |
+
"<unk_62>",
|
| 65 |
+
"<unk_63>",
|
| 66 |
+
"<unk_64>",
|
| 67 |
+
"<unk_65>",
|
| 68 |
+
"<unk_66>",
|
| 69 |
+
"<unk_67>",
|
| 70 |
+
"<unk_68>",
|
| 71 |
+
"<unk_69>",
|
| 72 |
+
"<unk_70>",
|
| 73 |
+
"<unk_71>",
|
| 74 |
+
"<unk_72>",
|
| 75 |
+
"<unk_73>",
|
| 76 |
+
"<unk_74>",
|
| 77 |
+
"<unk_75>",
|
| 78 |
+
"<unk_76>",
|
| 79 |
+
"<unk_77>",
|
| 80 |
+
"<unk_78>",
|
| 81 |
+
"<unk_79>",
|
| 82 |
+
"<unk_80>",
|
| 83 |
+
"<unk_81>",
|
| 84 |
+
"<unk_82>",
|
| 85 |
+
"<unk_83>",
|
| 86 |
+
"<unk_84>",
|
| 87 |
+
"<unk_85>",
|
| 88 |
+
"<unk_86>",
|
| 89 |
+
"<unk_87>",
|
| 90 |
+
"<unk_88>",
|
| 91 |
+
"<unk_89>",
|
| 92 |
+
"<unk_90>",
|
| 93 |
+
"<unk_91>",
|
| 94 |
+
"<unk_92>",
|
| 95 |
+
"<unk_93>",
|
| 96 |
+
"<unk_94>",
|
| 97 |
+
"<unk_95>",
|
| 98 |
+
"<unk_96>",
|
| 99 |
+
"<unk_97>",
|
| 100 |
+
"<unk_98>",
|
| 101 |
+
"<unk_99>",
|
| 102 |
+
"<unk_100>",
|
| 103 |
+
"<unk_101>",
|
| 104 |
+
"<unk_102>"
|
| 105 |
+
],
|
| 106 |
+
"eos_token": "</s>",
|
| 107 |
+
"mask_token": "<mask_2>",
|
| 108 |
+
"pad_token": "<pad>",
|
| 109 |
+
"unk_token": "<unk>"
|
| 110 |
+
}
|
last-checkpoint/spiece.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0015189ef36359283fec8b93cf6d9ce51bca37eb1101defc68a53b394913b96c
|
| 3 |
+
size 1912529
|
last-checkpoint/tokenizer_config.json
ADDED
|
@@ -0,0 +1,122 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<mask_1>",
|
| 4 |
+
"<unk_2>",
|
| 5 |
+
"<unk_3>",
|
| 6 |
+
"<unk_4>",
|
| 7 |
+
"<unk_5>",
|
| 8 |
+
"<unk_6>",
|
| 9 |
+
"<unk_7>",
|
| 10 |
+
"<unk_8>",
|
| 11 |
+
"<unk_9>",
|
| 12 |
+
"<unk_10>",
|
| 13 |
+
"<unk_11>",
|
| 14 |
+
"<unk_12>",
|
| 15 |
+
"<unk_13>",
|
| 16 |
+
"<unk_14>",
|
| 17 |
+
"<unk_15>",
|
| 18 |
+
"<unk_16>",
|
| 19 |
+
"<unk_17>",
|
| 20 |
+
"<unk_18>",
|
| 21 |
+
"<unk_19>",
|
| 22 |
+
"<unk_20>",
|
| 23 |
+
"<unk_21>",
|
| 24 |
+
"<unk_22>",
|
| 25 |
+
"<unk_23>",
|
| 26 |
+
"<unk_24>",
|
| 27 |
+
"<unk_25>",
|
| 28 |
+
"<unk_26>",
|
| 29 |
+
"<unk_27>",
|
| 30 |
+
"<unk_28>",
|
| 31 |
+
"<unk_29>",
|
| 32 |
+
"<unk_30>",
|
| 33 |
+
"<unk_31>",
|
| 34 |
+
"<unk_32>",
|
| 35 |
+
"<unk_33>",
|
| 36 |
+
"<unk_34>",
|
| 37 |
+
"<unk_35>",
|
| 38 |
+
"<unk_36>",
|
| 39 |
+
"<unk_37>",
|
| 40 |
+
"<unk_38>",
|
| 41 |
+
"<unk_39>",
|
| 42 |
+
"<unk_40>",
|
| 43 |
+
"<unk_41>",
|
| 44 |
+
"<unk_42>",
|
| 45 |
+
"<unk_43>",
|
| 46 |
+
"<unk_44>",
|
| 47 |
+
"<unk_45>",
|
| 48 |
+
"<unk_46>",
|
| 49 |
+
"<unk_47>",
|
| 50 |
+
"<unk_48>",
|
| 51 |
+
"<unk_49>",
|
| 52 |
+
"<unk_50>",
|
| 53 |
+
"<unk_51>",
|
| 54 |
+
"<unk_52>",
|
| 55 |
+
"<unk_53>",
|
| 56 |
+
"<unk_54>",
|
| 57 |
+
"<unk_55>",
|
| 58 |
+
"<unk_56>",
|
| 59 |
+
"<unk_57>",
|
| 60 |
+
"<unk_58>",
|
| 61 |
+
"<unk_59>",
|
| 62 |
+
"<unk_60>",
|
| 63 |
+
"<unk_61>",
|
| 64 |
+
"<unk_62>",
|
| 65 |
+
"<unk_63>",
|
| 66 |
+
"<unk_64>",
|
| 67 |
+
"<unk_65>",
|
| 68 |
+
"<unk_66>",
|
| 69 |
+
"<unk_67>",
|
| 70 |
+
"<unk_68>",
|
| 71 |
+
"<unk_69>",
|
| 72 |
+
"<unk_70>",
|
| 73 |
+
"<unk_71>",
|
| 74 |
+
"<unk_72>",
|
| 75 |
+
"<unk_73>",
|
| 76 |
+
"<unk_74>",
|
| 77 |
+
"<unk_75>",
|
| 78 |
+
"<unk_76>",
|
| 79 |
+
"<unk_77>",
|
| 80 |
+
"<unk_78>",
|
| 81 |
+
"<unk_79>",
|
| 82 |
+
"<unk_80>",
|
| 83 |
+
"<unk_81>",
|
| 84 |
+
"<unk_82>",
|
| 85 |
+
"<unk_83>",
|
| 86 |
+
"<unk_84>",
|
| 87 |
+
"<unk_85>",
|
| 88 |
+
"<unk_86>",
|
| 89 |
+
"<unk_87>",
|
| 90 |
+
"<unk_88>",
|
| 91 |
+
"<unk_89>",
|
| 92 |
+
"<unk_90>",
|
| 93 |
+
"<unk_91>",
|
| 94 |
+
"<unk_92>",
|
| 95 |
+
"<unk_93>",
|
| 96 |
+
"<unk_94>",
|
| 97 |
+
"<unk_95>",
|
| 98 |
+
"<unk_96>",
|
| 99 |
+
"<unk_97>",
|
| 100 |
+
"<unk_98>",
|
| 101 |
+
"<unk_99>",
|
| 102 |
+
"<unk_100>",
|
| 103 |
+
"<unk_101>",
|
| 104 |
+
"<unk_102>"
|
| 105 |
+
],
|
| 106 |
+
"batch_size": null,
|
| 107 |
+
"batched": true,
|
| 108 |
+
"eos_token": "</s>",
|
| 109 |
+
"full_tokenizer_file": null,
|
| 110 |
+
"mask_token": "<mask_2>",
|
| 111 |
+
"mask_token_sent": "<mask_1>",
|
| 112 |
+
"model_max_length": 1024,
|
| 113 |
+
"name_or_path": "google/pegasus-multi_news",
|
| 114 |
+
"offset": 103,
|
| 115 |
+
"pad_token": "<pad>",
|
| 116 |
+
"padding": true,
|
| 117 |
+
"sp_model_kwargs": {},
|
| 118 |
+
"special_tokens_map_file": null,
|
| 119 |
+
"tokenizer_class": "PegasusTokenizer",
|
| 120 |
+
"truncation": true,
|
| 121 |
+
"unk_token": "<unk>"
|
| 122 |
+
}
|
last-checkpoint/trainer_state.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9988331388564761,
|
| 5 |
+
"global_step": 107,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 4.0000000000000003e-07,
|
| 13 |
+
"loss": 8.06,
|
| 14 |
+
"step": 1
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.0,
|
| 18 |
+
"learning_rate": 1.304878048780488e-05,
|
| 19 |
+
"loss": 6.7693,
|
| 20 |
+
"step": 107
|
| 21 |
+
}
|
| 22 |
+
],
|
| 23 |
+
"max_steps": 214,
|
| 24 |
+
"num_train_epochs": 2,
|
| 25 |
+
"total_flos": 4910232797528064.0,
|
| 26 |
+
"trial_name": null,
|
| 27 |
+
"trial_params": null
|
| 28 |
+
}
|
last-checkpoint/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7afb5e1a3fd58d1d79a665850eb630a8235ae4c7c337aafde96679f57f2a4459
|
| 3 |
+
size 3439
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:85496f3b10fe45c0b541485a306b2663ae79d5d3db48b3dbac33200f1b225f94
|
| 3 |
+
size 2283790449
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,110 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<mask_1>",
|
| 4 |
+
"<unk_2>",
|
| 5 |
+
"<unk_3>",
|
| 6 |
+
"<unk_4>",
|
| 7 |
+
"<unk_5>",
|
| 8 |
+
"<unk_6>",
|
| 9 |
+
"<unk_7>",
|
| 10 |
+
"<unk_8>",
|
| 11 |
+
"<unk_9>",
|
| 12 |
+
"<unk_10>",
|
| 13 |
+
"<unk_11>",
|
| 14 |
+
"<unk_12>",
|
| 15 |
+
"<unk_13>",
|
| 16 |
+
"<unk_14>",
|
| 17 |
+
"<unk_15>",
|
| 18 |
+
"<unk_16>",
|
| 19 |
+
"<unk_17>",
|
| 20 |
+
"<unk_18>",
|
| 21 |
+
"<unk_19>",
|
| 22 |
+
"<unk_20>",
|
| 23 |
+
"<unk_21>",
|
| 24 |
+
"<unk_22>",
|
| 25 |
+
"<unk_23>",
|
| 26 |
+
"<unk_24>",
|
| 27 |
+
"<unk_25>",
|
| 28 |
+
"<unk_26>",
|
| 29 |
+
"<unk_27>",
|
| 30 |
+
"<unk_28>",
|
| 31 |
+
"<unk_29>",
|
| 32 |
+
"<unk_30>",
|
| 33 |
+
"<unk_31>",
|
| 34 |
+
"<unk_32>",
|
| 35 |
+
"<unk_33>",
|
| 36 |
+
"<unk_34>",
|
| 37 |
+
"<unk_35>",
|
| 38 |
+
"<unk_36>",
|
| 39 |
+
"<unk_37>",
|
| 40 |
+
"<unk_38>",
|
| 41 |
+
"<unk_39>",
|
| 42 |
+
"<unk_40>",
|
| 43 |
+
"<unk_41>",
|
| 44 |
+
"<unk_42>",
|
| 45 |
+
"<unk_43>",
|
| 46 |
+
"<unk_44>",
|
| 47 |
+
"<unk_45>",
|
| 48 |
+
"<unk_46>",
|
| 49 |
+
"<unk_47>",
|
| 50 |
+
"<unk_48>",
|
| 51 |
+
"<unk_49>",
|
| 52 |
+
"<unk_50>",
|
| 53 |
+
"<unk_51>",
|
| 54 |
+
"<unk_52>",
|
| 55 |
+
"<unk_53>",
|
| 56 |
+
"<unk_54>",
|
| 57 |
+
"<unk_55>",
|
| 58 |
+
"<unk_56>",
|
| 59 |
+
"<unk_57>",
|
| 60 |
+
"<unk_58>",
|
| 61 |
+
"<unk_59>",
|
| 62 |
+
"<unk_60>",
|
| 63 |
+
"<unk_61>",
|
| 64 |
+
"<unk_62>",
|
| 65 |
+
"<unk_63>",
|
| 66 |
+
"<unk_64>",
|
| 67 |
+
"<unk_65>",
|
| 68 |
+
"<unk_66>",
|
| 69 |
+
"<unk_67>",
|
| 70 |
+
"<unk_68>",
|
| 71 |
+
"<unk_69>",
|
| 72 |
+
"<unk_70>",
|
| 73 |
+
"<unk_71>",
|
| 74 |
+
"<unk_72>",
|
| 75 |
+
"<unk_73>",
|
| 76 |
+
"<unk_74>",
|
| 77 |
+
"<unk_75>",
|
| 78 |
+
"<unk_76>",
|
| 79 |
+
"<unk_77>",
|
| 80 |
+
"<unk_78>",
|
| 81 |
+
"<unk_79>",
|
| 82 |
+
"<unk_80>",
|
| 83 |
+
"<unk_81>",
|
| 84 |
+
"<unk_82>",
|
| 85 |
+
"<unk_83>",
|
| 86 |
+
"<unk_84>",
|
| 87 |
+
"<unk_85>",
|
| 88 |
+
"<unk_86>",
|
| 89 |
+
"<unk_87>",
|
| 90 |
+
"<unk_88>",
|
| 91 |
+
"<unk_89>",
|
| 92 |
+
"<unk_90>",
|
| 93 |
+
"<unk_91>",
|
| 94 |
+
"<unk_92>",
|
| 95 |
+
"<unk_93>",
|
| 96 |
+
"<unk_94>",
|
| 97 |
+
"<unk_95>",
|
| 98 |
+
"<unk_96>",
|
| 99 |
+
"<unk_97>",
|
| 100 |
+
"<unk_98>",
|
| 101 |
+
"<unk_99>",
|
| 102 |
+
"<unk_100>",
|
| 103 |
+
"<unk_101>",
|
| 104 |
+
"<unk_102>"
|
| 105 |
+
],
|
| 106 |
+
"eos_token": "</s>",
|
| 107 |
+
"mask_token": "<mask_2>",
|
| 108 |
+
"pad_token": "<pad>",
|
| 109 |
+
"unk_token": "<unk>"
|
| 110 |
+
}
|
spiece.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0015189ef36359283fec8b93cf6d9ce51bca37eb1101defc68a53b394913b96c
|
| 3 |
+
size 1912529
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,122 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<mask_1>",
|
| 4 |
+
"<unk_2>",
|
| 5 |
+
"<unk_3>",
|
| 6 |
+
"<unk_4>",
|
| 7 |
+
"<unk_5>",
|
| 8 |
+
"<unk_6>",
|
| 9 |
+
"<unk_7>",
|
| 10 |
+
"<unk_8>",
|
| 11 |
+
"<unk_9>",
|
| 12 |
+
"<unk_10>",
|
| 13 |
+
"<unk_11>",
|
| 14 |
+
"<unk_12>",
|
| 15 |
+
"<unk_13>",
|
| 16 |
+
"<unk_14>",
|
| 17 |
+
"<unk_15>",
|
| 18 |
+
"<unk_16>",
|
| 19 |
+
"<unk_17>",
|
| 20 |
+
"<unk_18>",
|
| 21 |
+
"<unk_19>",
|
| 22 |
+
"<unk_20>",
|
| 23 |
+
"<unk_21>",
|
| 24 |
+
"<unk_22>",
|
| 25 |
+
"<unk_23>",
|
| 26 |
+
"<unk_24>",
|
| 27 |
+
"<unk_25>",
|
| 28 |
+
"<unk_26>",
|
| 29 |
+
"<unk_27>",
|
| 30 |
+
"<unk_28>",
|
| 31 |
+
"<unk_29>",
|
| 32 |
+
"<unk_30>",
|
| 33 |
+
"<unk_31>",
|
| 34 |
+
"<unk_32>",
|
| 35 |
+
"<unk_33>",
|
| 36 |
+
"<unk_34>",
|
| 37 |
+
"<unk_35>",
|
| 38 |
+
"<unk_36>",
|
| 39 |
+
"<unk_37>",
|
| 40 |
+
"<unk_38>",
|
| 41 |
+
"<unk_39>",
|
| 42 |
+
"<unk_40>",
|
| 43 |
+
"<unk_41>",
|
| 44 |
+
"<unk_42>",
|
| 45 |
+
"<unk_43>",
|
| 46 |
+
"<unk_44>",
|
| 47 |
+
"<unk_45>",
|
| 48 |
+
"<unk_46>",
|
| 49 |
+
"<unk_47>",
|
| 50 |
+
"<unk_48>",
|
| 51 |
+
"<unk_49>",
|
| 52 |
+
"<unk_50>",
|
| 53 |
+
"<unk_51>",
|
| 54 |
+
"<unk_52>",
|
| 55 |
+
"<unk_53>",
|
| 56 |
+
"<unk_54>",
|
| 57 |
+
"<unk_55>",
|
| 58 |
+
"<unk_56>",
|
| 59 |
+
"<unk_57>",
|
| 60 |
+
"<unk_58>",
|
| 61 |
+
"<unk_59>",
|
| 62 |
+
"<unk_60>",
|
| 63 |
+
"<unk_61>",
|
| 64 |
+
"<unk_62>",
|
| 65 |
+
"<unk_63>",
|
| 66 |
+
"<unk_64>",
|
| 67 |
+
"<unk_65>",
|
| 68 |
+
"<unk_66>",
|
| 69 |
+
"<unk_67>",
|
| 70 |
+
"<unk_68>",
|
| 71 |
+
"<unk_69>",
|
| 72 |
+
"<unk_70>",
|
| 73 |
+
"<unk_71>",
|
| 74 |
+
"<unk_72>",
|
| 75 |
+
"<unk_73>",
|
| 76 |
+
"<unk_74>",
|
| 77 |
+
"<unk_75>",
|
| 78 |
+
"<unk_76>",
|
| 79 |
+
"<unk_77>",
|
| 80 |
+
"<unk_78>",
|
| 81 |
+
"<unk_79>",
|
| 82 |
+
"<unk_80>",
|
| 83 |
+
"<unk_81>",
|
| 84 |
+
"<unk_82>",
|
| 85 |
+
"<unk_83>",
|
| 86 |
+
"<unk_84>",
|
| 87 |
+
"<unk_85>",
|
| 88 |
+
"<unk_86>",
|
| 89 |
+
"<unk_87>",
|
| 90 |
+
"<unk_88>",
|
| 91 |
+
"<unk_89>",
|
| 92 |
+
"<unk_90>",
|
| 93 |
+
"<unk_91>",
|
| 94 |
+
"<unk_92>",
|
| 95 |
+
"<unk_93>",
|
| 96 |
+
"<unk_94>",
|
| 97 |
+
"<unk_95>",
|
| 98 |
+
"<unk_96>",
|
| 99 |
+
"<unk_97>",
|
| 100 |
+
"<unk_98>",
|
| 101 |
+
"<unk_99>",
|
| 102 |
+
"<unk_100>",
|
| 103 |
+
"<unk_101>",
|
| 104 |
+
"<unk_102>"
|
| 105 |
+
],
|
| 106 |
+
"batch_size": null,
|
| 107 |
+
"batched": true,
|
| 108 |
+
"eos_token": "</s>",
|
| 109 |
+
"full_tokenizer_file": null,
|
| 110 |
+
"mask_token": "<mask_2>",
|
| 111 |
+
"mask_token_sent": "<mask_1>",
|
| 112 |
+
"model_max_length": 1024,
|
| 113 |
+
"name_or_path": "google/pegasus-multi_news",
|
| 114 |
+
"offset": 103,
|
| 115 |
+
"pad_token": "<pad>",
|
| 116 |
+
"padding": true,
|
| 117 |
+
"sp_model_kwargs": {},
|
| 118 |
+
"special_tokens_map_file": null,
|
| 119 |
+
"tokenizer_class": "PegasusTokenizer",
|
| 120 |
+
"truncation": true,
|
| 121 |
+
"unk_token": "<unk>"
|
| 122 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7afb5e1a3fd58d1d79a665850eb630a8235ae4c7c337aafde96679f57f2a4459
|
| 3 |
+
size 3439
|