Training in progress, step 500
Browse files- config.json +28 -48
- pytorch_model.bin +2 -2
- runs/Apr15_22-00-01_132-145-132-14/1681596027.2840343/events.out.tfevents.1681596027.132-145-132-14.68350.1 +3 -0
- runs/Apr15_22-00-01_132-145-132-14/events.out.tfevents.1681596027.132-145-132-14.68350.0 +3 -0
- special_tokens_map.json +105 -10
- spiece.model +3 -0
- tokenizer.json +0 -0
- tokenizer_config.json +110 -8
- training_args.bin +1 -1
config.json
CHANGED
@@ -1,32 +1,29 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "
|
3 |
"activation_dropout": 0.1,
|
4 |
-
"activation_function": "
|
5 |
"add_bias_logits": false,
|
6 |
-
"add_final_layer_norm":
|
7 |
"architectures": [
|
8 |
-
"
|
9 |
],
|
10 |
"attention_dropout": 0.1,
|
11 |
"bos_token_id": 0,
|
12 |
-
"classif_dropout": 0.
|
13 |
-
"
|
14 |
-
"
|
15 |
-
"
|
16 |
-
"decoder_ffn_dim": 3072,
|
17 |
"decoder_layerdrop": 0.0,
|
18 |
-
"decoder_layers":
|
19 |
-
"decoder_start_token_id":
|
20 |
"dropout": 0.1,
|
21 |
-
"
|
22 |
-
"
|
23 |
-
"encoder_ffn_dim": 3072,
|
24 |
"encoder_layerdrop": 0.0,
|
25 |
-
"encoder_layers":
|
26 |
-
"eos_token_id":
|
27 |
-
"
|
28 |
-
"forced_eos_token_id":
|
29 |
-
"gradient_checkpointing": false,
|
30 |
"id2label": {
|
31 |
"0": "LABEL_0",
|
32 |
"1": "LABEL_1",
|
@@ -39,37 +36,20 @@
|
|
39 |
"LABEL_1": 1,
|
40 |
"LABEL_2": 2
|
41 |
},
|
|
|
|
|
42 |
"max_position_embeddings": 1024,
|
43 |
-
"
|
44 |
-
"
|
45 |
-
"normalize_before":
|
46 |
-
"normalize_embedding":
|
47 |
-
"num_beams":
|
48 |
-
"num_hidden_layers":
|
49 |
-
"pad_token_id":
|
50 |
-
"scale_embedding":
|
51 |
-
"
|
52 |
-
"summarization": {
|
53 |
-
"length_penalty": 1.0,
|
54 |
-
"max_length": 128,
|
55 |
-
"min_length": 12,
|
56 |
-
"num_beams": 4
|
57 |
-
},
|
58 |
-
"summarization_cnn": {
|
59 |
-
"length_penalty": 2.0,
|
60 |
-
"max_length": 142,
|
61 |
-
"min_length": 56,
|
62 |
-
"num_beams": 4
|
63 |
-
},
|
64 |
-
"summarization_xsum": {
|
65 |
-
"length_penalty": 1.0,
|
66 |
-
"max_length": 62,
|
67 |
-
"min_length": 11,
|
68 |
-
"num_beams": 6
|
69 |
-
}
|
70 |
-
},
|
71 |
"torch_dtype": "float32",
|
72 |
"transformers_version": "4.28.1",
|
73 |
"use_cache": true,
|
74 |
-
"vocab_size":
|
75 |
}
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "google/pegasus-arxiv",
|
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 |
+
"d_model": 1024,
|
14 |
+
"decoder_attention_heads": 16,
|
15 |
+
"decoder_ffn_dim": 4096,
|
|
|
16 |
"decoder_layerdrop": 0.0,
|
17 |
+
"decoder_layers": 16,
|
18 |
+
"decoder_start_token_id": 0,
|
19 |
"dropout": 0.1,
|
20 |
+
"encoder_attention_heads": 16,
|
21 |
+
"encoder_ffn_dim": 4096,
|
|
|
22 |
"encoder_layerdrop": 0.0,
|
23 |
+
"encoder_layers": 16,
|
24 |
+
"eos_token_id": 1,
|
25 |
+
"extra_pos_embeddings": 1,
|
26 |
+
"forced_eos_token_id": 1,
|
|
|
27 |
"id2label": {
|
28 |
"0": "LABEL_0",
|
29 |
"1": "LABEL_1",
|
|
|
36 |
"LABEL_1": 1,
|
37 |
"LABEL_2": 2
|
38 |
},
|
39 |
+
"length_penalty": 0.8,
|
40 |
+
"max_length": 256,
|
41 |
"max_position_embeddings": 1024,
|
42 |
+
"min_length": 32,
|
43 |
+
"model_type": "pegasus",
|
44 |
+
"normalize_before": true,
|
45 |
+
"normalize_embedding": false,
|
46 |
+
"num_beams": 8,
|
47 |
+
"num_hidden_layers": 16,
|
48 |
+
"pad_token_id": 0,
|
49 |
+
"scale_embedding": true,
|
50 |
+
"static_position_embeddings": true,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
51 |
"torch_dtype": "float32",
|
52 |
"transformers_version": "4.28.1",
|
53 |
"use_cache": true,
|
54 |
+
"vocab_size": 96103
|
55 |
}
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cce4c09fa99917b04c212c5fcf7dc599b34a76e3a4cdfa1bfd1a342bedb5b3f1
|
3 |
+
size 2283800049
|
runs/Apr15_22-00-01_132-145-132-14/1681596027.2840343/events.out.tfevents.1681596027.132-145-132-14.68350.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:160de2075506acc5dd9708294309de4bf7f23a232f0e9151990357838bec4b03
|
3 |
+
size 6094
|
runs/Apr15_22-00-01_132-145-132-14/events.out.tfevents.1681596027.132-145-132-14.68350.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a8a12f5fe5db7cfa44d8f9b2c25d01a4c080df49f46752f1979d9635a19ef990
|
3 |
+
size 4873
|
special_tokens_map.json
CHANGED
@@ -1,15 +1,110 @@
|
|
1 |
{
|
2 |
-
"
|
3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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 |
}
|
|
|
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.json
CHANGED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
CHANGED
@@ -1,15 +1,117 @@
|
|
1 |
{
|
2 |
-
"
|
3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
4 |
"clean_up_tokenization_spaces": true,
|
5 |
-
"cls_token": "<s>",
|
6 |
"eos_token": "</s>",
|
7 |
-
"
|
8 |
-
"mask_token": "<
|
|
|
9 |
"model_max_length": 1024,
|
|
|
10 |
"pad_token": "<pad>",
|
11 |
-
"
|
12 |
-
"tokenizer_class": "
|
13 |
-
"trim_offsets": true,
|
14 |
"unk_token": "<unk>"
|
15 |
}
|
|
|
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 |
"clean_up_tokenization_spaces": true,
|
|
|
107 |
"eos_token": "</s>",
|
108 |
+
"full_tokenizer_file": null,
|
109 |
+
"mask_token": "<mask_2>",
|
110 |
+
"mask_token_sent": "<mask_1>",
|
111 |
"model_max_length": 1024,
|
112 |
+
"offset": 103,
|
113 |
"pad_token": "<pad>",
|
114 |
+
"sp_model_kwargs": {},
|
115 |
+
"tokenizer_class": "PegasusTokenizer",
|
|
|
116 |
"unk_token": "<unk>"
|
117 |
}
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 3695
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:dd3961288c257336014d0f9c7bc8263657f093286cf8c3744f874345b06bbae7
|
3 |
size 3695
|