Upload 12 files
Browse files- config.json +31 -0
- generation_config.json +6 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state_0.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +108 -0
- spiece.model +3 -0
- tokenizer.json +0 -0
- tokenizer_config.json +113 -0
- trainer_state.json +1738 -0
- training_args.bin +3 -0
config.json
ADDED
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "/home/patrick/hugging_face/t5/t5-v1_1-base",
|
3 |
+
"architectures": [
|
4 |
+
"T5ForConditionalGeneration"
|
5 |
+
],
|
6 |
+
"d_ff": 2048,
|
7 |
+
"d_kv": 64,
|
8 |
+
"d_model": 768,
|
9 |
+
"decoder_start_token_id": 2,
|
10 |
+
"dense_act_fn": "gelu_new",
|
11 |
+
"dropout_rate": 0.1,
|
12 |
+
"eos_token_id": 3,
|
13 |
+
"feed_forward_proj": "gated-gelu",
|
14 |
+
"initializer_factor": 1.0,
|
15 |
+
"is_encoder_decoder": true,
|
16 |
+
"is_gated_act": true,
|
17 |
+
"layer_norm_epsilon": 1e-06,
|
18 |
+
"model_type": "t5",
|
19 |
+
"num_decoder_layers": 12,
|
20 |
+
"num_heads": 12,
|
21 |
+
"num_layers": 12,
|
22 |
+
"output_past": true,
|
23 |
+
"pad_token_id": 0,
|
24 |
+
"relative_attention_max_distance": 128,
|
25 |
+
"relative_attention_num_buckets": 32,
|
26 |
+
"tie_word_embeddings": false,
|
27 |
+
"torch_dtype": "float32",
|
28 |
+
"transformers_version": "4.27.0.dev0",
|
29 |
+
"use_cache": true,
|
30 |
+
"vocab_size": 32100
|
31 |
+
}
|
generation_config.json
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"decoder_start_token_id": 2,
|
3 |
+
"eos_token_id": 3,
|
4 |
+
"pad_token_id": 0,
|
5 |
+
"transformers_version": "4.27.0.dev0"
|
6 |
+
}
|
optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b16c95c61840cbbe54a30ae0e087f3cbd6174776cb1c76e91a5679fa69090ff5
|
3 |
+
size 1980446085
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:48bd864639eec9191c8c85143c2142848a6794486d885911e2cb2d92477161c8
|
3 |
+
size 990236853
|
rng_state_0.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:99f75dfea40d7e90412835f0e7db3bc5b2df07f8fc16c850a7462ab37be91251
|
3 |
+
size 14583
|
scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e70b5945654ca35a76139b5908295396538ad29607a8bc18ef8b73cb68670f6a
|
3 |
+
size 627
|
special_tokens_map.json
ADDED
@@ -0,0 +1,108 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"additional_special_tokens": [
|
3 |
+
"<extra_id_0>",
|
4 |
+
"<extra_id_1>",
|
5 |
+
"<extra_id_2>",
|
6 |
+
"<extra_id_3>",
|
7 |
+
"<extra_id_4>",
|
8 |
+
"<extra_id_5>",
|
9 |
+
"<extra_id_6>",
|
10 |
+
"<extra_id_7>",
|
11 |
+
"<extra_id_8>",
|
12 |
+
"<extra_id_9>",
|
13 |
+
"<extra_id_10>",
|
14 |
+
"<extra_id_11>",
|
15 |
+
"<extra_id_12>",
|
16 |
+
"<extra_id_13>",
|
17 |
+
"<extra_id_14>",
|
18 |
+
"<extra_id_15>",
|
19 |
+
"<extra_id_16>",
|
20 |
+
"<extra_id_17>",
|
21 |
+
"<extra_id_18>",
|
22 |
+
"<extra_id_19>",
|
23 |
+
"<extra_id_20>",
|
24 |
+
"<extra_id_21>",
|
25 |
+
"<extra_id_22>",
|
26 |
+
"<extra_id_23>",
|
27 |
+
"<extra_id_24>",
|
28 |
+
"<extra_id_25>",
|
29 |
+
"<extra_id_26>",
|
30 |
+
"<extra_id_27>",
|
31 |
+
"<extra_id_28>",
|
32 |
+
"<extra_id_29>",
|
33 |
+
"<extra_id_30>",
|
34 |
+
"<extra_id_31>",
|
35 |
+
"<extra_id_32>",
|
36 |
+
"<extra_id_33>",
|
37 |
+
"<extra_id_34>",
|
38 |
+
"<extra_id_35>",
|
39 |
+
"<extra_id_36>",
|
40 |
+
"<extra_id_37>",
|
41 |
+
"<extra_id_38>",
|
42 |
+
"<extra_id_39>",
|
43 |
+
"<extra_id_40>",
|
44 |
+
"<extra_id_41>",
|
45 |
+
"<extra_id_42>",
|
46 |
+
"<extra_id_43>",
|
47 |
+
"<extra_id_44>",
|
48 |
+
"<extra_id_45>",
|
49 |
+
"<extra_id_46>",
|
50 |
+
"<extra_id_47>",
|
51 |
+
"<extra_id_48>",
|
52 |
+
"<extra_id_49>",
|
53 |
+
"<extra_id_50>",
|
54 |
+
"<extra_id_51>",
|
55 |
+
"<extra_id_52>",
|
56 |
+
"<extra_id_53>",
|
57 |
+
"<extra_id_54>",
|
58 |
+
"<extra_id_55>",
|
59 |
+
"<extra_id_56>",
|
60 |
+
"<extra_id_57>",
|
61 |
+
"<extra_id_58>",
|
62 |
+
"<extra_id_59>",
|
63 |
+
"<extra_id_60>",
|
64 |
+
"<extra_id_61>",
|
65 |
+
"<extra_id_62>",
|
66 |
+
"<extra_id_63>",
|
67 |
+
"<extra_id_64>",
|
68 |
+
"<extra_id_65>",
|
69 |
+
"<extra_id_66>",
|
70 |
+
"<extra_id_67>",
|
71 |
+
"<extra_id_68>",
|
72 |
+
"<extra_id_69>",
|
73 |
+
"<extra_id_70>",
|
74 |
+
"<extra_id_71>",
|
75 |
+
"<extra_id_72>",
|
76 |
+
"<extra_id_73>",
|
77 |
+
"<extra_id_74>",
|
78 |
+
"<extra_id_75>",
|
79 |
+
"<extra_id_76>",
|
80 |
+
"<extra_id_77>",
|
81 |
+
"<extra_id_78>",
|
82 |
+
"<extra_id_79>",
|
83 |
+
"<extra_id_80>",
|
84 |
+
"<extra_id_81>",
|
85 |
+
"<extra_id_82>",
|
86 |
+
"<extra_id_83>",
|
87 |
+
"<extra_id_84>",
|
88 |
+
"<extra_id_85>",
|
89 |
+
"<extra_id_86>",
|
90 |
+
"<extra_id_87>",
|
91 |
+
"<extra_id_88>",
|
92 |
+
"<extra_id_89>",
|
93 |
+
"<extra_id_90>",
|
94 |
+
"<extra_id_91>",
|
95 |
+
"<extra_id_92>",
|
96 |
+
"<extra_id_93>",
|
97 |
+
"<extra_id_94>",
|
98 |
+
"<extra_id_95>",
|
99 |
+
"<extra_id_96>",
|
100 |
+
"<extra_id_97>",
|
101 |
+
"<extra_id_98>",
|
102 |
+
"<extra_id_99>"
|
103 |
+
],
|
104 |
+
"bos_token": "<s>",
|
105 |
+
"eos_token": "</s>",
|
106 |
+
"pad_token": "<pad>",
|
107 |
+
"unk_token": "<unk>"
|
108 |
+
}
|
spiece.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:270da11f1d731cd2234015d5c233dba7c5b5996a466fbf74e24f7ea332924b93
|
3 |
+
size 882013
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1,113 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"additional_special_tokens": [
|
3 |
+
"<extra_id_0>",
|
4 |
+
"<extra_id_1>",
|
5 |
+
"<extra_id_2>",
|
6 |
+
"<extra_id_3>",
|
7 |
+
"<extra_id_4>",
|
8 |
+
"<extra_id_5>",
|
9 |
+
"<extra_id_6>",
|
10 |
+
"<extra_id_7>",
|
11 |
+
"<extra_id_8>",
|
12 |
+
"<extra_id_9>",
|
13 |
+
"<extra_id_10>",
|
14 |
+
"<extra_id_11>",
|
15 |
+
"<extra_id_12>",
|
16 |
+
"<extra_id_13>",
|
17 |
+
"<extra_id_14>",
|
18 |
+
"<extra_id_15>",
|
19 |
+
"<extra_id_16>",
|
20 |
+
"<extra_id_17>",
|
21 |
+
"<extra_id_18>",
|
22 |
+
"<extra_id_19>",
|
23 |
+
"<extra_id_20>",
|
24 |
+
"<extra_id_21>",
|
25 |
+
"<extra_id_22>",
|
26 |
+
"<extra_id_23>",
|
27 |
+
"<extra_id_24>",
|
28 |
+
"<extra_id_25>",
|
29 |
+
"<extra_id_26>",
|
30 |
+
"<extra_id_27>",
|
31 |
+
"<extra_id_28>",
|
32 |
+
"<extra_id_29>",
|
33 |
+
"<extra_id_30>",
|
34 |
+
"<extra_id_31>",
|
35 |
+
"<extra_id_32>",
|
36 |
+
"<extra_id_33>",
|
37 |
+
"<extra_id_34>",
|
38 |
+
"<extra_id_35>",
|
39 |
+
"<extra_id_36>",
|
40 |
+
"<extra_id_37>",
|
41 |
+
"<extra_id_38>",
|
42 |
+
"<extra_id_39>",
|
43 |
+
"<extra_id_40>",
|
44 |
+
"<extra_id_41>",
|
45 |
+
"<extra_id_42>",
|
46 |
+
"<extra_id_43>",
|
47 |
+
"<extra_id_44>",
|
48 |
+
"<extra_id_45>",
|
49 |
+
"<extra_id_46>",
|
50 |
+
"<extra_id_47>",
|
51 |
+
"<extra_id_48>",
|
52 |
+
"<extra_id_49>",
|
53 |
+
"<extra_id_50>",
|
54 |
+
"<extra_id_51>",
|
55 |
+
"<extra_id_52>",
|
56 |
+
"<extra_id_53>",
|
57 |
+
"<extra_id_54>",
|
58 |
+
"<extra_id_55>",
|
59 |
+
"<extra_id_56>",
|
60 |
+
"<extra_id_57>",
|
61 |
+
"<extra_id_58>",
|
62 |
+
"<extra_id_59>",
|
63 |
+
"<extra_id_60>",
|
64 |
+
"<extra_id_61>",
|
65 |
+
"<extra_id_62>",
|
66 |
+
"<extra_id_63>",
|
67 |
+
"<extra_id_64>",
|
68 |
+
"<extra_id_65>",
|
69 |
+
"<extra_id_66>",
|
70 |
+
"<extra_id_67>",
|
71 |
+
"<extra_id_68>",
|
72 |
+
"<extra_id_69>",
|
73 |
+
"<extra_id_70>",
|
74 |
+
"<extra_id_71>",
|
75 |
+
"<extra_id_72>",
|
76 |
+
"<extra_id_73>",
|
77 |
+
"<extra_id_74>",
|
78 |
+
"<extra_id_75>",
|
79 |
+
"<extra_id_76>",
|
80 |
+
"<extra_id_77>",
|
81 |
+
"<extra_id_78>",
|
82 |
+
"<extra_id_79>",
|
83 |
+
"<extra_id_80>",
|
84 |
+
"<extra_id_81>",
|
85 |
+
"<extra_id_82>",
|
86 |
+
"<extra_id_83>",
|
87 |
+
"<extra_id_84>",
|
88 |
+
"<extra_id_85>",
|
89 |
+
"<extra_id_86>",
|
90 |
+
"<extra_id_87>",
|
91 |
+
"<extra_id_88>",
|
92 |
+
"<extra_id_89>",
|
93 |
+
"<extra_id_90>",
|
94 |
+
"<extra_id_91>",
|
95 |
+
"<extra_id_92>",
|
96 |
+
"<extra_id_93>",
|
97 |
+
"<extra_id_94>",
|
98 |
+
"<extra_id_95>",
|
99 |
+
"<extra_id_96>",
|
100 |
+
"<extra_id_97>",
|
101 |
+
"<extra_id_98>",
|
102 |
+
"<extra_id_99>"
|
103 |
+
],
|
104 |
+
"bos_token": "<s>",
|
105 |
+
"eos_token": "</s>",
|
106 |
+
"extra_ids": 100,
|
107 |
+
"model_max_length": 1000000000000000019884624838656,
|
108 |
+
"pad_token": "<pad>",
|
109 |
+
"sp_model_kwargs": {},
|
110 |
+
"special_tokens_map_file": "./T5Tokenizer/special_tokens_map.json",
|
111 |
+
"tokenizer_class": "T5Tokenizer",
|
112 |
+
"unk_token": "<unk>"
|
113 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,1738 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 5.739096280610677,
|
5 |
+
"global_step": 21000,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.27,
|
12 |
+
"learning_rate": 0.00015,
|
13 |
+
"loss": 46.3847,
|
14 |
+
"step": 1000
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 0.27,
|
18 |
+
"eval_validation_chrf_beta": 2,
|
19 |
+
"eval_validation_chrf_char_order": 6,
|
20 |
+
"eval_validation_chrf_score": 6.927323312955705,
|
21 |
+
"eval_validation_chrf_word_order": 2,
|
22 |
+
"eval_validation_loss": 16.81111717224121,
|
23 |
+
"eval_validation_meteor": 0.06701502044580518,
|
24 |
+
"eval_validation_runtime": 40.1943,
|
25 |
+
"eval_validation_samples_per_second": 24.805,
|
26 |
+
"eval_validation_scarebleu_bp": 1.0,
|
27 |
+
"eval_validation_scarebleu_counts": [
|
28 |
+
2501,
|
29 |
+
140,
|
30 |
+
21,
|
31 |
+
3
|
32 |
+
],
|
33 |
+
"eval_validation_scarebleu_precisions": [
|
34 |
+
9.256106587712805,
|
35 |
+
0.5373248896564958,
|
36 |
+
0.08352889702080267,
|
37 |
+
0.01236603462489695
|
38 |
+
],
|
39 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
40 |
+
"eval_validation_scarebleu_score": 0.26772142473423427,
|
41 |
+
"eval_validation_scarebleu_sys_len": 27020,
|
42 |
+
"eval_validation_scarebleu_totals": [
|
43 |
+
27020,
|
44 |
+
26055,
|
45 |
+
25141,
|
46 |
+
24260
|
47 |
+
],
|
48 |
+
"eval_validation_steps_per_second": 0.398,
|
49 |
+
"eval_validation_ter_num_edits": 24081,
|
50 |
+
"eval_validation_ter_ref_length": 16729.0,
|
51 |
+
"eval_validation_ter_score": 143.9476358419511,
|
52 |
+
"step": 1000
|
53 |
+
},
|
54 |
+
{
|
55 |
+
"epoch": 0.27,
|
56 |
+
"eval_test_chrf_beta": 2,
|
57 |
+
"eval_test_chrf_char_order": 6,
|
58 |
+
"eval_test_chrf_score": 6.879629526302676,
|
59 |
+
"eval_test_chrf_word_order": 2,
|
60 |
+
"eval_test_loss": 16.940214157104492,
|
61 |
+
"eval_test_meteor": 0.06578125012212357,
|
62 |
+
"eval_test_runtime": 43.5255,
|
63 |
+
"eval_test_samples_per_second": 23.251,
|
64 |
+
"eval_test_scarebleu_bp": 1.0,
|
65 |
+
"eval_test_scarebleu_counts": [
|
66 |
+
2620,
|
67 |
+
181,
|
68 |
+
21,
|
69 |
+
1
|
70 |
+
],
|
71 |
+
"eval_test_scarebleu_precisions": [
|
72 |
+
8.875639418679494,
|
73 |
+
0.6341087443946188,
|
74 |
+
0.07605113533480606,
|
75 |
+
0.0037420948246828577
|
76 |
+
],
|
77 |
+
"eval_test_scarebleu_ref_len": 20238,
|
78 |
+
"eval_test_scarebleu_score": 0.20005341558923276,
|
79 |
+
"eval_test_scarebleu_sys_len": 29519,
|
80 |
+
"eval_test_scarebleu_totals": [
|
81 |
+
29519,
|
82 |
+
28544,
|
83 |
+
27613,
|
84 |
+
26723
|
85 |
+
],
|
86 |
+
"eval_test_steps_per_second": 0.368,
|
87 |
+
"eval_test_ter_num_edits": 25915,
|
88 |
+
"eval_test_ter_ref_length": 17367.0,
|
89 |
+
"eval_test_ter_score": 149.21978464904703,
|
90 |
+
"step": 1000
|
91 |
+
},
|
92 |
+
{
|
93 |
+
"epoch": 0.55,
|
94 |
+
"learning_rate": 0.0003,
|
95 |
+
"loss": 19.9672,
|
96 |
+
"step": 2000
|
97 |
+
},
|
98 |
+
{
|
99 |
+
"epoch": 0.55,
|
100 |
+
"eval_validation_chrf_beta": 2,
|
101 |
+
"eval_validation_chrf_char_order": 6,
|
102 |
+
"eval_validation_chrf_score": 6.608664701301832,
|
103 |
+
"eval_validation_chrf_word_order": 2,
|
104 |
+
"eval_validation_loss": 9.209909439086914,
|
105 |
+
"eval_validation_meteor": 0.06402750263419245,
|
106 |
+
"eval_validation_runtime": 39.5005,
|
107 |
+
"eval_validation_samples_per_second": 25.24,
|
108 |
+
"eval_validation_scarebleu_bp": 0.7877829309778815,
|
109 |
+
"eval_validation_scarebleu_counts": [
|
110 |
+
2144,
|
111 |
+
200,
|
112 |
+
35,
|
113 |
+
2
|
114 |
+
],
|
115 |
+
"eval_validation_scarebleu_precisions": [
|
116 |
+
13.677830940988835,
|
117 |
+
1.352356481168436,
|
118 |
+
0.2506624650862995,
|
119 |
+
0.015055706112616682
|
120 |
+
],
|
121 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
122 |
+
"eval_validation_scarebleu_score": 0.4049312368014021,
|
123 |
+
"eval_validation_scarebleu_sys_len": 15675,
|
124 |
+
"eval_validation_scarebleu_totals": [
|
125 |
+
15675,
|
126 |
+
14789,
|
127 |
+
13963,
|
128 |
+
13284
|
129 |
+
],
|
130 |
+
"eval_validation_steps_per_second": 0.405,
|
131 |
+
"eval_validation_ter_num_edits": 19406,
|
132 |
+
"eval_validation_ter_ref_length": 16729.0,
|
133 |
+
"eval_validation_ter_score": 116.00215195170065,
|
134 |
+
"step": 2000
|
135 |
+
},
|
136 |
+
{
|
137 |
+
"epoch": 0.55,
|
138 |
+
"eval_test_chrf_beta": 2,
|
139 |
+
"eval_test_chrf_char_order": 6,
|
140 |
+
"eval_test_chrf_score": 6.872536195708606,
|
141 |
+
"eval_test_chrf_word_order": 2,
|
142 |
+
"eval_test_loss": 9.260172843933105,
|
143 |
+
"eval_test_meteor": 0.06539736281948932,
|
144 |
+
"eval_test_runtime": 42.6967,
|
145 |
+
"eval_test_samples_per_second": 23.702,
|
146 |
+
"eval_test_scarebleu_bp": 0.8237885182366754,
|
147 |
+
"eval_test_scarebleu_counts": [
|
148 |
+
2294,
|
149 |
+
254,
|
150 |
+
51,
|
151 |
+
9
|
152 |
+
],
|
153 |
+
"eval_test_scarebleu_precisions": [
|
154 |
+
13.532326569136385,
|
155 |
+
1.582258767831558,
|
156 |
+
0.33537186821858356,
|
157 |
+
0.06213324128408699
|
158 |
+
],
|
159 |
+
"eval_test_scarebleu_ref_len": 20238,
|
160 |
+
"eval_test_scarebleu_score": 0.6732725214400193,
|
161 |
+
"eval_test_scarebleu_sys_len": 16952,
|
162 |
+
"eval_test_scarebleu_totals": [
|
163 |
+
16952,
|
164 |
+
16053,
|
165 |
+
15207,
|
166 |
+
14485
|
167 |
+
],
|
168 |
+
"eval_test_steps_per_second": 0.375,
|
169 |
+
"eval_test_ter_num_edits": 20720,
|
170 |
+
"eval_test_ter_ref_length": 17367.0,
|
171 |
+
"eval_test_ter_score": 119.30673115679163,
|
172 |
+
"step": 2000
|
173 |
+
},
|
174 |
+
{
|
175 |
+
"epoch": 0.82,
|
176 |
+
"learning_rate": 0.0002999986101989992,
|
177 |
+
"loss": 8.7409,
|
178 |
+
"step": 3000
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 0.82,
|
182 |
+
"eval_validation_chrf_beta": 2,
|
183 |
+
"eval_validation_chrf_char_order": 6,
|
184 |
+
"eval_validation_chrf_score": 13.23574850542808,
|
185 |
+
"eval_validation_chrf_word_order": 2,
|
186 |
+
"eval_validation_loss": 6.500675201416016,
|
187 |
+
"eval_validation_meteor": 0.1196379582028911,
|
188 |
+
"eval_validation_runtime": 40.0578,
|
189 |
+
"eval_validation_samples_per_second": 24.889,
|
190 |
+
"eval_validation_scarebleu_bp": 1.0,
|
191 |
+
"eval_validation_scarebleu_counts": [
|
192 |
+
4007,
|
193 |
+
528,
|
194 |
+
99,
|
195 |
+
14
|
196 |
+
],
|
197 |
+
"eval_validation_scarebleu_precisions": [
|
198 |
+
12.693635758862102,
|
199 |
+
1.7262799973844243,
|
200 |
+
0.334256195556756,
|
201 |
+
0.04885538805136795
|
202 |
+
],
|
203 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
204 |
+
"eval_validation_scarebleu_score": 0.7734322925143389,
|
205 |
+
"eval_validation_scarebleu_sys_len": 31567,
|
206 |
+
"eval_validation_scarebleu_totals": [
|
207 |
+
31567,
|
208 |
+
30586,
|
209 |
+
29618,
|
210 |
+
28656
|
211 |
+
],
|
212 |
+
"eval_validation_steps_per_second": 0.399,
|
213 |
+
"eval_validation_ter_num_edits": 22118,
|
214 |
+
"eval_validation_ter_ref_length": 16729.0,
|
215 |
+
"eval_validation_ter_score": 132.21352142985234,
|
216 |
+
"step": 3000
|
217 |
+
},
|
218 |
+
{
|
219 |
+
"epoch": 0.82,
|
220 |
+
"eval_test_chrf_beta": 2,
|
221 |
+
"eval_test_chrf_char_order": 6,
|
222 |
+
"eval_test_chrf_score": 13.044816904787435,
|
223 |
+
"eval_test_chrf_word_order": 2,
|
224 |
+
"eval_test_loss": 6.5189080238342285,
|
225 |
+
"eval_test_meteor": 0.11759299622619233,
|
226 |
+
"eval_test_runtime": 39.9506,
|
227 |
+
"eval_test_samples_per_second": 25.331,
|
228 |
+
"eval_test_scarebleu_bp": 1.0,
|
229 |
+
"eval_test_scarebleu_counts": [
|
230 |
+
4139,
|
231 |
+
590,
|
232 |
+
121,
|
233 |
+
25
|
234 |
+
],
|
235 |
+
"eval_test_scarebleu_precisions": [
|
236 |
+
12.661364331599877,
|
237 |
+
1.8612574529164958,
|
238 |
+
0.39385456675997654,
|
239 |
+
0.08403361344537816
|
240 |
+
],
|
241 |
+
"eval_test_scarebleu_ref_len": 20238,
|
242 |
+
"eval_test_scarebleu_score": 0.9397643584535246,
|
243 |
+
"eval_test_scarebleu_sys_len": 32690,
|
244 |
+
"eval_test_scarebleu_totals": [
|
245 |
+
32690,
|
246 |
+
31699,
|
247 |
+
30722,
|
248 |
+
29750
|
249 |
+
],
|
250 |
+
"eval_test_steps_per_second": 0.4,
|
251 |
+
"eval_test_ter_num_edits": 23506,
|
252 |
+
"eval_test_ter_ref_length": 17367.0,
|
253 |
+
"eval_test_ter_score": 135.34864973800887,
|
254 |
+
"step": 3000
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 1.09,
|
258 |
+
"learning_rate": 0.00029999444082175083,
|
259 |
+
"loss": 5.8284,
|
260 |
+
"step": 4000
|
261 |
+
},
|
262 |
+
{
|
263 |
+
"epoch": 1.09,
|
264 |
+
"eval_validation_chrf_beta": 2,
|
265 |
+
"eval_validation_chrf_char_order": 6,
|
266 |
+
"eval_validation_chrf_score": 15.697678446835747,
|
267 |
+
"eval_validation_chrf_word_order": 2,
|
268 |
+
"eval_validation_loss": 5.613649368286133,
|
269 |
+
"eval_validation_meteor": 0.14151503184835326,
|
270 |
+
"eval_validation_runtime": 37.324,
|
271 |
+
"eval_validation_samples_per_second": 26.712,
|
272 |
+
"eval_validation_scarebleu_bp": 1.0,
|
273 |
+
"eval_validation_scarebleu_counts": [
|
274 |
+
4311,
|
275 |
+
658,
|
276 |
+
141,
|
277 |
+
30
|
278 |
+
],
|
279 |
+
"eval_validation_scarebleu_precisions": [
|
280 |
+
21.276280722534793,
|
281 |
+
3.413925495486147,
|
282 |
+
0.7710395362826051,
|
283 |
+
0.17339035949601203
|
284 |
+
],
|
285 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
286 |
+
"eval_validation_scarebleu_score": 1.7652768928036136,
|
287 |
+
"eval_validation_scarebleu_sys_len": 20262,
|
288 |
+
"eval_validation_scarebleu_totals": [
|
289 |
+
20262,
|
290 |
+
19274,
|
291 |
+
18287,
|
292 |
+
17302
|
293 |
+
],
|
294 |
+
"eval_validation_steps_per_second": 0.429,
|
295 |
+
"eval_validation_ter_num_edits": 18183,
|
296 |
+
"eval_validation_ter_ref_length": 16729.0,
|
297 |
+
"eval_validation_ter_score": 108.69149381313885,
|
298 |
+
"step": 4000
|
299 |
+
},
|
300 |
+
{
|
301 |
+
"epoch": 1.09,
|
302 |
+
"eval_test_chrf_beta": 2,
|
303 |
+
"eval_test_chrf_char_order": 6,
|
304 |
+
"eval_test_chrf_score": 15.824281404133039,
|
305 |
+
"eval_test_chrf_word_order": 2,
|
306 |
+
"eval_test_loss": 5.644526481628418,
|
307 |
+
"eval_test_meteor": 0.14262643610521655,
|
308 |
+
"eval_test_runtime": 37.2514,
|
309 |
+
"eval_test_samples_per_second": 27.167,
|
310 |
+
"eval_test_scarebleu_bp": 1.0,
|
311 |
+
"eval_test_scarebleu_counts": [
|
312 |
+
4480,
|
313 |
+
750,
|
314 |
+
151,
|
315 |
+
21
|
316 |
+
],
|
317 |
+
"eval_test_scarebleu_precisions": [
|
318 |
+
21.725425537073857,
|
319 |
+
3.8230196758079313,
|
320 |
+
0.8109995166228047,
|
321 |
+
0.11913541725761616
|
322 |
+
],
|
323 |
+
"eval_test_scarebleu_ref_len": 20238,
|
324 |
+
"eval_test_scarebleu_score": 1.683096735458044,
|
325 |
+
"eval_test_scarebleu_sys_len": 20621,
|
326 |
+
"eval_test_scarebleu_totals": [
|
327 |
+
20621,
|
328 |
+
19618,
|
329 |
+
18619,
|
330 |
+
17627
|
331 |
+
],
|
332 |
+
"eval_test_steps_per_second": 0.43,
|
333 |
+
"eval_test_ter_num_edits": 18915,
|
334 |
+
"eval_test_ter_ref_length": 17367.0,
|
335 |
+
"eval_test_ter_score": 108.91345655553637,
|
336 |
+
"step": 4000
|
337 |
+
},
|
338 |
+
{
|
339 |
+
"epoch": 1.37,
|
340 |
+
"learning_rate": 0.0002999874919455162,
|
341 |
+
"loss": 4.9666,
|
342 |
+
"step": 5000
|
343 |
+
},
|
344 |
+
{
|
345 |
+
"epoch": 1.37,
|
346 |
+
"eval_validation_chrf_beta": 2,
|
347 |
+
"eval_validation_chrf_char_order": 6,
|
348 |
+
"eval_validation_chrf_score": 19.94602423909563,
|
349 |
+
"eval_validation_chrf_word_order": 2,
|
350 |
+
"eval_validation_loss": 5.06654691696167,
|
351 |
+
"eval_validation_meteor": 0.16677714234198524,
|
352 |
+
"eval_validation_runtime": 35.8549,
|
353 |
+
"eval_validation_samples_per_second": 27.806,
|
354 |
+
"eval_validation_scarebleu_bp": 1.0,
|
355 |
+
"eval_validation_scarebleu_counts": [
|
356 |
+
5239,
|
357 |
+
885,
|
358 |
+
180,
|
359 |
+
35
|
360 |
+
],
|
361 |
+
"eval_validation_scarebleu_precisions": [
|
362 |
+
20.370154360589446,
|
363 |
+
3.579807458943451,
|
364 |
+
0.7586933614330874,
|
365 |
+
0.1539815222173339
|
366 |
+
],
|
367 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
368 |
+
"eval_validation_scarebleu_score": 1.70843012306825,
|
369 |
+
"eval_validation_scarebleu_sys_len": 25719,
|
370 |
+
"eval_validation_scarebleu_totals": [
|
371 |
+
25719,
|
372 |
+
24722,
|
373 |
+
23725,
|
374 |
+
22730
|
375 |
+
],
|
376 |
+
"eval_validation_steps_per_second": 0.446,
|
377 |
+
"eval_validation_ter_num_edits": 20131,
|
378 |
+
"eval_validation_ter_ref_length": 16729.0,
|
379 |
+
"eval_validation_ter_score": 120.3359435710443,
|
380 |
+
"step": 5000
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 1.37,
|
384 |
+
"eval_test_chrf_beta": 2,
|
385 |
+
"eval_test_chrf_char_order": 6,
|
386 |
+
"eval_test_chrf_score": 19.73223665848102,
|
387 |
+
"eval_test_chrf_word_order": 2,
|
388 |
+
"eval_test_loss": 5.101978302001953,
|
389 |
+
"eval_test_meteor": 0.16941489704708929,
|
390 |
+
"eval_test_runtime": 40.2976,
|
391 |
+
"eval_test_samples_per_second": 25.113,
|
392 |
+
"eval_test_scarebleu_bp": 1.0,
|
393 |
+
"eval_test_scarebleu_counts": [
|
394 |
+
5479,
|
395 |
+
997,
|
396 |
+
226,
|
397 |
+
45
|
398 |
+
],
|
399 |
+
"eval_test_scarebleu_precisions": [
|
400 |
+
19.717853672580702,
|
401 |
+
3.723622782446312,
|
402 |
+
0.8771929824561403,
|
403 |
+
0.18179614592170645
|
404 |
+
],
|
405 |
+
"eval_test_scarebleu_ref_len": 20238,
|
406 |
+
"eval_test_scarebleu_score": 1.8498064298140473,
|
407 |
+
"eval_test_scarebleu_sys_len": 27787,
|
408 |
+
"eval_test_scarebleu_totals": [
|
409 |
+
27787,
|
410 |
+
26775,
|
411 |
+
25764,
|
412 |
+
24753
|
413 |
+
],
|
414 |
+
"eval_test_steps_per_second": 0.397,
|
415 |
+
"eval_test_ter_num_edits": 21806,
|
416 |
+
"eval_test_ter_ref_length": 17367.0,
|
417 |
+
"eval_test_ter_score": 125.55997005815627,
|
418 |
+
"step": 5000
|
419 |
+
},
|
420 |
+
{
|
421 |
+
"epoch": 1.64,
|
422 |
+
"learning_rate": 0.00029997776369906286,
|
423 |
+
"loss": 4.4311,
|
424 |
+
"step": 6000
|
425 |
+
},
|
426 |
+
{
|
427 |
+
"epoch": 1.64,
|
428 |
+
"eval_validation_chrf_beta": 2,
|
429 |
+
"eval_validation_chrf_char_order": 6,
|
430 |
+
"eval_validation_chrf_score": 24.59127986581791,
|
431 |
+
"eval_validation_chrf_word_order": 2,
|
432 |
+
"eval_validation_loss": 4.524135112762451,
|
433 |
+
"eval_validation_meteor": 0.2102906308603433,
|
434 |
+
"eval_validation_runtime": 31.5423,
|
435 |
+
"eval_validation_samples_per_second": 31.608,
|
436 |
+
"eval_validation_scarebleu_bp": 1.0,
|
437 |
+
"eval_validation_scarebleu_counts": [
|
438 |
+
5964,
|
439 |
+
1335,
|
440 |
+
396,
|
441 |
+
121
|
442 |
+
],
|
443 |
+
"eval_validation_scarebleu_precisions": [
|
444 |
+
28.030267424918925,
|
445 |
+
6.5828402366863905,
|
446 |
+
2.053622361665716,
|
447 |
+
0.6617084108060811
|
448 |
+
],
|
449 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
450 |
+
"eval_validation_scarebleu_score": 3.979302446849127,
|
451 |
+
"eval_validation_scarebleu_sys_len": 21277,
|
452 |
+
"eval_validation_scarebleu_totals": [
|
453 |
+
21277,
|
454 |
+
20280,
|
455 |
+
19283,
|
456 |
+
18286
|
457 |
+
],
|
458 |
+
"eval_validation_steps_per_second": 0.507,
|
459 |
+
"eval_validation_ter_num_edits": 17139,
|
460 |
+
"eval_validation_ter_ref_length": 16729.0,
|
461 |
+
"eval_validation_ter_score": 102.450833881284,
|
462 |
+
"step": 6000
|
463 |
+
},
|
464 |
+
{
|
465 |
+
"epoch": 1.64,
|
466 |
+
"eval_test_chrf_beta": 2,
|
467 |
+
"eval_test_chrf_char_order": 6,
|
468 |
+
"eval_test_chrf_score": 24.41919902051296,
|
469 |
+
"eval_test_chrf_word_order": 2,
|
470 |
+
"eval_test_loss": 4.5522613525390625,
|
471 |
+
"eval_test_meteor": 0.2072945229916662,
|
472 |
+
"eval_test_runtime": 34.2408,
|
473 |
+
"eval_test_samples_per_second": 29.555,
|
474 |
+
"eval_test_scarebleu_bp": 1.0,
|
475 |
+
"eval_test_scarebleu_counts": [
|
476 |
+
6159,
|
477 |
+
1354,
|
478 |
+
372,
|
479 |
+
84
|
480 |
+
],
|
481 |
+
"eval_test_scarebleu_precisions": [
|
482 |
+
27.975109011627907,
|
483 |
+
6.446391163587888,
|
484 |
+
1.8607442977190876,
|
485 |
+
0.44252449689179224
|
486 |
+
],
|
487 |
+
"eval_test_scarebleu_ref_len": 20238,
|
488 |
+
"eval_test_scarebleu_score": 3.4908252929483026,
|
489 |
+
"eval_test_scarebleu_sys_len": 22016,
|
490 |
+
"eval_test_scarebleu_totals": [
|
491 |
+
22016,
|
492 |
+
21004,
|
493 |
+
19992,
|
494 |
+
18982
|
495 |
+
],
|
496 |
+
"eval_test_steps_per_second": 0.467,
|
497 |
+
"eval_test_ter_num_edits": 17741,
|
498 |
+
"eval_test_ter_ref_length": 17367.0,
|
499 |
+
"eval_test_ter_score": 102.15350952956757,
|
500 |
+
"step": 6000
|
501 |
+
},
|
502 |
+
{
|
503 |
+
"epoch": 1.91,
|
504 |
+
"learning_rate": 0.00029996525626266166,
|
505 |
+
"loss": 3.8914,
|
506 |
+
"step": 7000
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 1.91,
|
510 |
+
"eval_validation_chrf_beta": 2,
|
511 |
+
"eval_validation_chrf_char_order": 6,
|
512 |
+
"eval_validation_chrf_score": 28.001456752933812,
|
513 |
+
"eval_validation_chrf_word_order": 2,
|
514 |
+
"eval_validation_loss": 4.09087610244751,
|
515 |
+
"eval_validation_meteor": 0.246925162242717,
|
516 |
+
"eval_validation_runtime": 25.3477,
|
517 |
+
"eval_validation_samples_per_second": 39.333,
|
518 |
+
"eval_validation_scarebleu_bp": 0.9367877834918877,
|
519 |
+
"eval_validation_scarebleu_counts": [
|
520 |
+
6414,
|
521 |
+
1774,
|
522 |
+
615,
|
523 |
+
220
|
524 |
+
],
|
525 |
+
"eval_validation_scarebleu_precisions": [
|
526 |
+
35.19534679543459,
|
527 |
+
10.297190619921059,
|
528 |
+
3.7888122227698373,
|
529 |
+
1.4439485429246521
|
530 |
+
],
|
531 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
532 |
+
"eval_validation_scarebleu_score": 6.251092400807247,
|
533 |
+
"eval_validation_scarebleu_sys_len": 18224,
|
534 |
+
"eval_validation_scarebleu_totals": [
|
535 |
+
18224,
|
536 |
+
17228,
|
537 |
+
16232,
|
538 |
+
15236
|
539 |
+
],
|
540 |
+
"eval_validation_steps_per_second": 0.631,
|
541 |
+
"eval_validation_ter_num_edits": 14680,
|
542 |
+
"eval_validation_ter_ref_length": 16729.0,
|
543 |
+
"eval_validation_ter_score": 87.75180823719289,
|
544 |
+
"step": 7000
|
545 |
+
},
|
546 |
+
{
|
547 |
+
"epoch": 1.91,
|
548 |
+
"eval_test_chrf_beta": 2,
|
549 |
+
"eval_test_chrf_char_order": 6,
|
550 |
+
"eval_test_chrf_score": 27.737380160414922,
|
551 |
+
"eval_test_chrf_word_order": 2,
|
552 |
+
"eval_test_loss": 4.100297451019287,
|
553 |
+
"eval_test_meteor": 0.23872126072037225,
|
554 |
+
"eval_test_runtime": 27.793,
|
555 |
+
"eval_test_samples_per_second": 36.412,
|
556 |
+
"eval_test_scarebleu_bp": 0.9496588509867531,
|
557 |
+
"eval_test_scarebleu_counts": [
|
558 |
+
6585,
|
559 |
+
1745,
|
560 |
+
575,
|
561 |
+
175
|
562 |
+
],
|
563 |
+
"eval_test_scarebleu_precisions": [
|
564 |
+
34.21845770110164,
|
565 |
+
9.571083808688021,
|
566 |
+
3.3391405342624854,
|
567 |
+
1.0796471096304523
|
568 |
+
],
|
569 |
+
"eval_test_scarebleu_ref_len": 20238,
|
570 |
+
"eval_test_scarebleu_score": 5.566755396990573,
|
571 |
+
"eval_test_scarebleu_sys_len": 19244,
|
572 |
+
"eval_test_scarebleu_totals": [
|
573 |
+
19244,
|
574 |
+
18232,
|
575 |
+
17220,
|
576 |
+
16209
|
577 |
+
],
|
578 |
+
"eval_test_steps_per_second": 0.576,
|
579 |
+
"eval_test_ter_num_edits": 15506,
|
580 |
+
"eval_test_ter_ref_length": 17367.0,
|
581 |
+
"eval_test_ter_score": 89.28427477399666,
|
582 |
+
"step": 7000
|
583 |
+
},
|
584 |
+
{
|
585 |
+
"epoch": 2.19,
|
586 |
+
"learning_rate": 0.000299949969868084,
|
587 |
+
"loss": 3.4344,
|
588 |
+
"step": 8000
|
589 |
+
},
|
590 |
+
{
|
591 |
+
"epoch": 2.19,
|
592 |
+
"eval_validation_chrf_beta": 2,
|
593 |
+
"eval_validation_chrf_char_order": 6,
|
594 |
+
"eval_validation_chrf_score": 31.29398951552832,
|
595 |
+
"eval_validation_chrf_word_order": 2,
|
596 |
+
"eval_validation_loss": 3.7306714057922363,
|
597 |
+
"eval_validation_meteor": 0.27820026125205044,
|
598 |
+
"eval_validation_runtime": 18.376,
|
599 |
+
"eval_validation_samples_per_second": 54.256,
|
600 |
+
"eval_validation_scarebleu_bp": 0.9889681301158993,
|
601 |
+
"eval_validation_scarebleu_counts": [
|
602 |
+
7069,
|
603 |
+
2158,
|
604 |
+
789,
|
605 |
+
302
|
606 |
+
],
|
607 |
+
"eval_validation_scarebleu_precisions": [
|
608 |
+
36.81579084422686,
|
609 |
+
11.854537464293562,
|
610 |
+
4.585343174289533,
|
611 |
+
1.8630475015422578
|
612 |
+
],
|
613 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
614 |
+
"eval_validation_scarebleu_score": 7.727889909602072,
|
615 |
+
"eval_validation_scarebleu_sys_len": 19201,
|
616 |
+
"eval_validation_scarebleu_totals": [
|
617 |
+
19201,
|
618 |
+
18204,
|
619 |
+
17207,
|
620 |
+
16210
|
621 |
+
],
|
622 |
+
"eval_validation_steps_per_second": 0.871,
|
623 |
+
"eval_validation_ter_num_edits": 14351,
|
624 |
+
"eval_validation_ter_ref_length": 16729.0,
|
625 |
+
"eval_validation_ter_score": 85.78516348855281,
|
626 |
+
"step": 8000
|
627 |
+
},
|
628 |
+
{
|
629 |
+
"epoch": 2.19,
|
630 |
+
"eval_test_chrf_beta": 2,
|
631 |
+
"eval_test_chrf_char_order": 6,
|
632 |
+
"eval_test_chrf_score": 31.913351815650685,
|
633 |
+
"eval_test_chrf_word_order": 2,
|
634 |
+
"eval_test_loss": 3.728149890899658,
|
635 |
+
"eval_test_meteor": 0.2797376842090434,
|
636 |
+
"eval_test_runtime": 22.9959,
|
637 |
+
"eval_test_samples_per_second": 44.008,
|
638 |
+
"eval_test_scarebleu_bp": 0.9872701879268588,
|
639 |
+
"eval_test_scarebleu_counts": [
|
640 |
+
7428,
|
641 |
+
2283,
|
642 |
+
851,
|
643 |
+
319
|
644 |
+
],
|
645 |
+
"eval_test_scarebleu_precisions": [
|
646 |
+
37.17345611049945,
|
647 |
+
12.034791776489193,
|
648 |
+
4.738835059583472,
|
649 |
+
1.882450135725245
|
650 |
+
],
|
651 |
+
"eval_test_scarebleu_ref_len": 20238,
|
652 |
+
"eval_test_scarebleu_score": 7.846982277886522,
|
653 |
+
"eval_test_scarebleu_sys_len": 19982,
|
654 |
+
"eval_test_scarebleu_totals": [
|
655 |
+
19982,
|
656 |
+
18970,
|
657 |
+
17958,
|
658 |
+
16946
|
659 |
+
],
|
660 |
+
"eval_test_steps_per_second": 0.696,
|
661 |
+
"eval_test_ter_num_edits": 14894,
|
662 |
+
"eval_test_ter_ref_length": 17367.0,
|
663 |
+
"eval_test_ter_score": 85.76035008924973,
|
664 |
+
"step": 8000
|
665 |
+
},
|
666 |
+
{
|
667 |
+
"epoch": 2.46,
|
668 |
+
"learning_rate": 0.0002999319047985972,
|
669 |
+
"loss": 3.1209,
|
670 |
+
"step": 9000
|
671 |
+
},
|
672 |
+
{
|
673 |
+
"epoch": 2.46,
|
674 |
+
"eval_validation_chrf_beta": 2,
|
675 |
+
"eval_validation_chrf_char_order": 6,
|
676 |
+
"eval_validation_chrf_score": 33.55090391489323,
|
677 |
+
"eval_validation_chrf_word_order": 2,
|
678 |
+
"eval_validation_loss": 3.4695987701416016,
|
679 |
+
"eval_validation_meteor": 0.2976359335029889,
|
680 |
+
"eval_validation_runtime": 21.7643,
|
681 |
+
"eval_validation_samples_per_second": 45.809,
|
682 |
+
"eval_validation_scarebleu_bp": 1.0,
|
683 |
+
"eval_validation_scarebleu_counts": [
|
684 |
+
7463,
|
685 |
+
2467,
|
686 |
+
1011,
|
687 |
+
418
|
688 |
+
],
|
689 |
+
"eval_validation_scarebleu_precisions": [
|
690 |
+
37.90056370930882,
|
691 |
+
13.196747619557078,
|
692 |
+
5.712832683505679,
|
693 |
+
2.502994011976048
|
694 |
+
],
|
695 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
696 |
+
"eval_validation_scarebleu_score": 9.196148464160226,
|
697 |
+
"eval_validation_scarebleu_sys_len": 19691,
|
698 |
+
"eval_validation_scarebleu_totals": [
|
699 |
+
19691,
|
700 |
+
18694,
|
701 |
+
17697,
|
702 |
+
16700
|
703 |
+
],
|
704 |
+
"eval_validation_steps_per_second": 0.735,
|
705 |
+
"eval_validation_ter_num_edits": 14205,
|
706 |
+
"eval_validation_ter_ref_length": 16729.0,
|
707 |
+
"eval_validation_ter_score": 84.9124275210712,
|
708 |
+
"step": 9000
|
709 |
+
},
|
710 |
+
{
|
711 |
+
"epoch": 2.46,
|
712 |
+
"eval_test_chrf_beta": 2,
|
713 |
+
"eval_test_chrf_char_order": 6,
|
714 |
+
"eval_test_chrf_score": 33.9716211383833,
|
715 |
+
"eval_test_chrf_word_order": 2,
|
716 |
+
"eval_test_loss": 3.4928810596466064,
|
717 |
+
"eval_test_meteor": 0.3022870093526768,
|
718 |
+
"eval_test_runtime": 25.0697,
|
719 |
+
"eval_test_samples_per_second": 40.367,
|
720 |
+
"eval_test_scarebleu_bp": 1.0,
|
721 |
+
"eval_test_scarebleu_counts": [
|
722 |
+
7869,
|
723 |
+
2633,
|
724 |
+
1054,
|
725 |
+
427
|
726 |
+
],
|
727 |
+
"eval_test_scarebleu_precisions": [
|
728 |
+
38.3741343996879,
|
729 |
+
13.506720016415308,
|
730 |
+
5.702846012336328,
|
731 |
+
2.444190040068689
|
732 |
+
],
|
733 |
+
"eval_test_scarebleu_ref_len": 20238,
|
734 |
+
"eval_test_scarebleu_score": 9.219423736514582,
|
735 |
+
"eval_test_scarebleu_sys_len": 20506,
|
736 |
+
"eval_test_scarebleu_totals": [
|
737 |
+
20506,
|
738 |
+
19494,
|
739 |
+
18482,
|
740 |
+
17470
|
741 |
+
],
|
742 |
+
"eval_test_steps_per_second": 0.638,
|
743 |
+
"eval_test_ter_num_edits": 14739,
|
744 |
+
"eval_test_ter_ref_length": 17367.0,
|
745 |
+
"eval_test_ter_score": 84.86785282432199,
|
746 |
+
"step": 9000
|
747 |
+
},
|
748 |
+
{
|
749 |
+
"epoch": 2.73,
|
750 |
+
"learning_rate": 0.00029991106138895916,
|
751 |
+
"loss": 2.9118,
|
752 |
+
"step": 10000
|
753 |
+
},
|
754 |
+
{
|
755 |
+
"epoch": 2.73,
|
756 |
+
"eval_validation_chrf_beta": 2,
|
757 |
+
"eval_validation_chrf_char_order": 6,
|
758 |
+
"eval_validation_chrf_score": 34.57697226346535,
|
759 |
+
"eval_validation_chrf_word_order": 2,
|
760 |
+
"eval_validation_loss": 3.2827696800231934,
|
761 |
+
"eval_validation_meteor": 0.30655372675156295,
|
762 |
+
"eval_validation_runtime": 21.2242,
|
763 |
+
"eval_validation_samples_per_second": 46.975,
|
764 |
+
"eval_validation_scarebleu_bp": 1.0,
|
765 |
+
"eval_validation_scarebleu_counts": [
|
766 |
+
7614,
|
767 |
+
2604,
|
768 |
+
1072,
|
769 |
+
445
|
770 |
+
],
|
771 |
+
"eval_validation_scarebleu_precisions": [
|
772 |
+
39.094269870609985,
|
773 |
+
14.091671627252557,
|
774 |
+
6.132021507836632,
|
775 |
+
2.6994237185319987
|
776 |
+
],
|
777 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
778 |
+
"eval_validation_scarebleu_score": 9.772092910936896,
|
779 |
+
"eval_validation_scarebleu_sys_len": 19476,
|
780 |
+
"eval_validation_scarebleu_totals": [
|
781 |
+
19476,
|
782 |
+
18479,
|
783 |
+
17482,
|
784 |
+
16485
|
785 |
+
],
|
786 |
+
"eval_validation_steps_per_second": 0.754,
|
787 |
+
"eval_validation_ter_num_edits": 13943,
|
788 |
+
"eval_validation_ter_ref_length": 16729.0,
|
789 |
+
"eval_validation_ter_score": 83.34628489449459,
|
790 |
+
"step": 10000
|
791 |
+
},
|
792 |
+
{
|
793 |
+
"epoch": 2.73,
|
794 |
+
"eval_test_chrf_beta": 2,
|
795 |
+
"eval_test_chrf_char_order": 6,
|
796 |
+
"eval_test_chrf_score": 35.254931037556965,
|
797 |
+
"eval_test_chrf_word_order": 2,
|
798 |
+
"eval_test_loss": 3.296231985092163,
|
799 |
+
"eval_test_meteor": 0.3139488805447367,
|
800 |
+
"eval_test_runtime": 23.0481,
|
801 |
+
"eval_test_samples_per_second": 43.908,
|
802 |
+
"eval_test_scarebleu_bp": 0.9980710741204407,
|
803 |
+
"eval_test_scarebleu_counts": [
|
804 |
+
8108,
|
805 |
+
2823,
|
806 |
+
1184,
|
807 |
+
513
|
808 |
+
],
|
809 |
+
"eval_test_scarebleu_precisions": [
|
810 |
+
40.14060101985247,
|
811 |
+
14.713086985980091,
|
812 |
+
6.514442916093535,
|
813 |
+
2.9889879391714733
|
814 |
+
],
|
815 |
+
"eval_test_scarebleu_ref_len": 20238,
|
816 |
+
"eval_test_scarebleu_score": 10.335553726245266,
|
817 |
+
"eval_test_scarebleu_sys_len": 20199,
|
818 |
+
"eval_test_scarebleu_totals": [
|
819 |
+
20199,
|
820 |
+
19187,
|
821 |
+
18175,
|
822 |
+
17163
|
823 |
+
],
|
824 |
+
"eval_test_steps_per_second": 0.694,
|
825 |
+
"eval_test_ter_num_edits": 14363,
|
826 |
+
"eval_test_ter_ref_length": 17367.0,
|
827 |
+
"eval_test_ter_score": 82.70282720101342,
|
828 |
+
"step": 10000
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 3.01,
|
832 |
+
"learning_rate": 0.0002998874400254125,
|
833 |
+
"loss": 2.7563,
|
834 |
+
"step": 11000
|
835 |
+
},
|
836 |
+
{
|
837 |
+
"epoch": 3.01,
|
838 |
+
"eval_validation_chrf_beta": 2,
|
839 |
+
"eval_validation_chrf_char_order": 6,
|
840 |
+
"eval_validation_chrf_score": 35.922822688896666,
|
841 |
+
"eval_validation_chrf_word_order": 2,
|
842 |
+
"eval_validation_loss": 3.1191513538360596,
|
843 |
+
"eval_validation_meteor": 0.3225733936623777,
|
844 |
+
"eval_validation_runtime": 18.9973,
|
845 |
+
"eval_validation_samples_per_second": 52.481,
|
846 |
+
"eval_validation_scarebleu_bp": 0.9987114400155332,
|
847 |
+
"eval_validation_scarebleu_counts": [
|
848 |
+
7893,
|
849 |
+
2765,
|
850 |
+
1154,
|
851 |
+
506
|
852 |
+
],
|
853 |
+
"eval_validation_scarebleu_precisions": [
|
854 |
+
40.70864923410181,
|
855 |
+
15.033710308829926,
|
856 |
+
6.634090255820638,
|
857 |
+
3.085742163678497
|
858 |
+
],
|
859 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
860 |
+
"eval_validation_scarebleu_score": 10.56607027026474,
|
861 |
+
"eval_validation_scarebleu_sys_len": 19389,
|
862 |
+
"eval_validation_scarebleu_totals": [
|
863 |
+
19389,
|
864 |
+
18392,
|
865 |
+
17395,
|
866 |
+
16398
|
867 |
+
],
|
868 |
+
"eval_validation_steps_per_second": 0.842,
|
869 |
+
"eval_validation_ter_num_edits": 13631,
|
870 |
+
"eval_validation_ter_ref_length": 16729.0,
|
871 |
+
"eval_validation_ter_score": 81.4812600872736,
|
872 |
+
"step": 11000
|
873 |
+
},
|
874 |
+
{
|
875 |
+
"epoch": 3.01,
|
876 |
+
"eval_test_chrf_beta": 2,
|
877 |
+
"eval_test_chrf_char_order": 6,
|
878 |
+
"eval_test_chrf_score": 36.3162560210105,
|
879 |
+
"eval_test_chrf_word_order": 2,
|
880 |
+
"eval_test_loss": 3.134953737258911,
|
881 |
+
"eval_test_meteor": 0.3246997818548813,
|
882 |
+
"eval_test_runtime": 22.6178,
|
883 |
+
"eval_test_samples_per_second": 44.744,
|
884 |
+
"eval_test_scarebleu_bp": 1.0,
|
885 |
+
"eval_test_scarebleu_counts": [
|
886 |
+
8246,
|
887 |
+
2958,
|
888 |
+
1263,
|
889 |
+
563
|
890 |
+
],
|
891 |
+
"eval_test_scarebleu_precisions": [
|
892 |
+
40.70088845014808,
|
893 |
+
15.36783042394015,
|
894 |
+
6.925860934415442,
|
895 |
+
3.2686948444031585
|
896 |
+
],
|
897 |
+
"eval_test_scarebleu_ref_len": 20238,
|
898 |
+
"eval_test_scarebleu_score": 10.90852794539211,
|
899 |
+
"eval_test_scarebleu_sys_len": 20260,
|
900 |
+
"eval_test_scarebleu_totals": [
|
901 |
+
20260,
|
902 |
+
19248,
|
903 |
+
18236,
|
904 |
+
17224
|
905 |
+
],
|
906 |
+
"eval_test_steps_per_second": 0.707,
|
907 |
+
"eval_test_ter_num_edits": 14265,
|
908 |
+
"eval_test_ter_ref_length": 17367.0,
|
909 |
+
"eval_test_ter_score": 82.13853860770428,
|
910 |
+
"step": 11000
|
911 |
+
},
|
912 |
+
{
|
913 |
+
"epoch": 3.28,
|
914 |
+
"learning_rate": 0.0002998610411456772,
|
915 |
+
"loss": 2.5851,
|
916 |
+
"step": 12000
|
917 |
+
},
|
918 |
+
{
|
919 |
+
"epoch": 3.28,
|
920 |
+
"eval_validation_chrf_beta": 2,
|
921 |
+
"eval_validation_chrf_char_order": 6,
|
922 |
+
"eval_validation_chrf_score": 36.76732059324029,
|
923 |
+
"eval_validation_chrf_word_order": 2,
|
924 |
+
"eval_validation_loss": 2.992377996444702,
|
925 |
+
"eval_validation_meteor": 0.3322433302283302,
|
926 |
+
"eval_validation_runtime": 19.5358,
|
927 |
+
"eval_validation_samples_per_second": 51.034,
|
928 |
+
"eval_validation_scarebleu_bp": 1.0,
|
929 |
+
"eval_validation_scarebleu_counts": [
|
930 |
+
8059,
|
931 |
+
2918,
|
932 |
+
1246,
|
933 |
+
534
|
934 |
+
],
|
935 |
+
"eval_validation_scarebleu_precisions": [
|
936 |
+
41.24782475176579,
|
937 |
+
15.738093953939917,
|
938 |
+
7.102143182854538,
|
939 |
+
3.2269760696156635
|
940 |
+
],
|
941 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
942 |
+
"eval_validation_scarebleu_score": 11.044208276358312,
|
943 |
+
"eval_validation_scarebleu_sys_len": 19538,
|
944 |
+
"eval_validation_scarebleu_totals": [
|
945 |
+
19538,
|
946 |
+
18541,
|
947 |
+
17544,
|
948 |
+
16548
|
949 |
+
],
|
950 |
+
"eval_validation_steps_per_second": 0.819,
|
951 |
+
"eval_validation_ter_num_edits": 13514,
|
952 |
+
"eval_validation_ter_ref_length": 16729.0,
|
953 |
+
"eval_validation_ter_score": 80.78187578456573,
|
954 |
+
"step": 12000
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 3.28,
|
958 |
+
"eval_test_chrf_beta": 2,
|
959 |
+
"eval_test_chrf_char_order": 6,
|
960 |
+
"eval_test_chrf_score": 37.26551462594263,
|
961 |
+
"eval_test_chrf_word_order": 2,
|
962 |
+
"eval_test_loss": 3.0072576999664307,
|
963 |
+
"eval_test_meteor": 0.333414935567908,
|
964 |
+
"eval_test_runtime": 23.446,
|
965 |
+
"eval_test_samples_per_second": 43.163,
|
966 |
+
"eval_test_scarebleu_bp": 1.0,
|
967 |
+
"eval_test_scarebleu_counts": [
|
968 |
+
8465,
|
969 |
+
3091,
|
970 |
+
1327,
|
971 |
+
584
|
972 |
+
],
|
973 |
+
"eval_test_scarebleu_precisions": [
|
974 |
+
41.44836703716398,
|
975 |
+
15.923960640873732,
|
976 |
+
7.212348497200935,
|
977 |
+
3.3588313107494105
|
978 |
+
],
|
979 |
+
"eval_test_scarebleu_ref_len": 20238,
|
980 |
+
"eval_test_scarebleu_score": 11.244906644276366,
|
981 |
+
"eval_test_scarebleu_sys_len": 20423,
|
982 |
+
"eval_test_scarebleu_totals": [
|
983 |
+
20423,
|
984 |
+
19411,
|
985 |
+
18399,
|
986 |
+
17387
|
987 |
+
],
|
988 |
+
"eval_test_steps_per_second": 0.682,
|
989 |
+
"eval_test_ter_num_edits": 14135,
|
990 |
+
"eval_test_ter_ref_length": 17367.0,
|
991 |
+
"eval_test_ter_score": 81.38999251453907,
|
992 |
+
"step": 12000
|
993 |
+
},
|
994 |
+
{
|
995 |
+
"epoch": 3.55,
|
996 |
+
"learning_rate": 0.00029983186523894237,
|
997 |
+
"loss": 2.4949,
|
998 |
+
"step": 13000
|
999 |
+
},
|
1000 |
+
{
|
1001 |
+
"epoch": 3.55,
|
1002 |
+
"eval_validation_chrf_beta": 2,
|
1003 |
+
"eval_validation_chrf_char_order": 6,
|
1004 |
+
"eval_validation_chrf_score": 37.97775302751643,
|
1005 |
+
"eval_validation_chrf_word_order": 2,
|
1006 |
+
"eval_validation_loss": 2.9026131629943848,
|
1007 |
+
"eval_validation_meteor": 0.34218163692855424,
|
1008 |
+
"eval_validation_runtime": 18.3627,
|
1009 |
+
"eval_validation_samples_per_second": 54.295,
|
1010 |
+
"eval_validation_scarebleu_bp": 1.0,
|
1011 |
+
"eval_validation_scarebleu_counts": [
|
1012 |
+
8292,
|
1013 |
+
3082,
|
1014 |
+
1350,
|
1015 |
+
611
|
1016 |
+
],
|
1017 |
+
"eval_validation_scarebleu_precisions": [
|
1018 |
+
42.28886168910649,
|
1019 |
+
16.56009886626189,
|
1020 |
+
7.664357897127285,
|
1021 |
+
3.6769573328518987
|
1022 |
+
],
|
1023 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
1024 |
+
"eval_validation_scarebleu_score": 11.852591640319904,
|
1025 |
+
"eval_validation_scarebleu_sys_len": 19608,
|
1026 |
+
"eval_validation_scarebleu_totals": [
|
1027 |
+
19608,
|
1028 |
+
18611,
|
1029 |
+
17614,
|
1030 |
+
16617
|
1031 |
+
],
|
1032 |
+
"eval_validation_steps_per_second": 0.871,
|
1033 |
+
"eval_validation_ter_num_edits": 13331,
|
1034 |
+
"eval_validation_ter_ref_length": 16729.0,
|
1035 |
+
"eval_validation_ter_score": 79.68796700340725,
|
1036 |
+
"step": 13000
|
1037 |
+
},
|
1038 |
+
{
|
1039 |
+
"epoch": 3.55,
|
1040 |
+
"eval_test_chrf_beta": 2,
|
1041 |
+
"eval_test_chrf_char_order": 6,
|
1042 |
+
"eval_test_chrf_score": 38.64688755179878,
|
1043 |
+
"eval_test_chrf_word_order": 2,
|
1044 |
+
"eval_test_loss": 2.9025168418884277,
|
1045 |
+
"eval_test_meteor": 0.3488289320983426,
|
1046 |
+
"eval_test_runtime": 23.8956,
|
1047 |
+
"eval_test_samples_per_second": 42.351,
|
1048 |
+
"eval_test_scarebleu_bp": 1.0,
|
1049 |
+
"eval_test_scarebleu_counts": [
|
1050 |
+
8783,
|
1051 |
+
3333,
|
1052 |
+
1490,
|
1053 |
+
682
|
1054 |
+
],
|
1055 |
+
"eval_test_scarebleu_precisions": [
|
1056 |
+
42.666990527082824,
|
1057 |
+
17.02855975067695,
|
1058 |
+
8.027584720650827,
|
1059 |
+
3.8862613254316485
|
1060 |
+
],
|
1061 |
+
"eval_test_scarebleu_ref_len": 20238,
|
1062 |
+
"eval_test_scarebleu_score": 12.270060576985161,
|
1063 |
+
"eval_test_scarebleu_sys_len": 20585,
|
1064 |
+
"eval_test_scarebleu_totals": [
|
1065 |
+
20585,
|
1066 |
+
19573,
|
1067 |
+
18561,
|
1068 |
+
17549
|
1069 |
+
],
|
1070 |
+
"eval_test_steps_per_second": 0.67,
|
1071 |
+
"eval_test_ter_num_edits": 13913,
|
1072 |
+
"eval_test_ter_ref_length": 17367.0,
|
1073 |
+
"eval_test_ter_score": 80.11170610928772,
|
1074 |
+
"step": 13000
|
1075 |
+
},
|
1076 |
+
{
|
1077 |
+
"epoch": 3.83,
|
1078 |
+
"learning_rate": 0.0002997999128458575,
|
1079 |
+
"loss": 2.4153,
|
1080 |
+
"step": 14000
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 3.83,
|
1084 |
+
"eval_validation_chrf_beta": 2,
|
1085 |
+
"eval_validation_chrf_char_order": 6,
|
1086 |
+
"eval_validation_chrf_score": 38.410061864122575,
|
1087 |
+
"eval_validation_chrf_word_order": 2,
|
1088 |
+
"eval_validation_loss": 2.802281141281128,
|
1089 |
+
"eval_validation_meteor": 0.3482320817009387,
|
1090 |
+
"eval_validation_runtime": 20.1679,
|
1091 |
+
"eval_validation_samples_per_second": 49.435,
|
1092 |
+
"eval_validation_scarebleu_bp": 1.0,
|
1093 |
+
"eval_validation_scarebleu_counts": [
|
1094 |
+
8386,
|
1095 |
+
3169,
|
1096 |
+
1413,
|
1097 |
+
644
|
1098 |
+
],
|
1099 |
+
"eval_validation_scarebleu_precisions": [
|
1100 |
+
42.807554874936194,
|
1101 |
+
17.044048835583283,
|
1102 |
+
8.030234144123664,
|
1103 |
+
3.879751792276643
|
1104 |
+
],
|
1105 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
1106 |
+
"eval_validation_scarebleu_score": 12.278811597675354,
|
1107 |
+
"eval_validation_scarebleu_sys_len": 19590,
|
1108 |
+
"eval_validation_scarebleu_totals": [
|
1109 |
+
19590,
|
1110 |
+
18593,
|
1111 |
+
17596,
|
1112 |
+
16599
|
1113 |
+
],
|
1114 |
+
"eval_validation_steps_per_second": 0.793,
|
1115 |
+
"eval_validation_ter_num_edits": 13285,
|
1116 |
+
"eval_validation_ter_ref_length": 16729.0,
|
1117 |
+
"eval_validation_ter_score": 79.41299539721442,
|
1118 |
+
"step": 14000
|
1119 |
+
},
|
1120 |
+
{
|
1121 |
+
"epoch": 3.83,
|
1122 |
+
"eval_test_chrf_beta": 2,
|
1123 |
+
"eval_test_chrf_char_order": 6,
|
1124 |
+
"eval_test_chrf_score": 39.18817112230258,
|
1125 |
+
"eval_test_chrf_word_order": 2,
|
1126 |
+
"eval_test_loss": 2.8074586391448975,
|
1127 |
+
"eval_test_meteor": 0.35206418879072193,
|
1128 |
+
"eval_test_runtime": 22.9097,
|
1129 |
+
"eval_test_samples_per_second": 44.173,
|
1130 |
+
"eval_test_scarebleu_bp": 1.0,
|
1131 |
+
"eval_test_scarebleu_counts": [
|
1132 |
+
8797,
|
1133 |
+
3403,
|
1134 |
+
1547,
|
1135 |
+
713
|
1136 |
+
],
|
1137 |
+
"eval_test_scarebleu_precisions": [
|
1138 |
+
42.6273198623831,
|
1139 |
+
17.34012738853503,
|
1140 |
+
8.311395261376457,
|
1141 |
+
4.0509061985114485
|
1142 |
+
],
|
1143 |
+
"eval_test_scarebleu_ref_len": 20238,
|
1144 |
+
"eval_test_scarebleu_score": 12.560056952808758,
|
1145 |
+
"eval_test_scarebleu_sys_len": 20637,
|
1146 |
+
"eval_test_scarebleu_totals": [
|
1147 |
+
20637,
|
1148 |
+
19625,
|
1149 |
+
18613,
|
1150 |
+
17601
|
1151 |
+
],
|
1152 |
+
"eval_test_steps_per_second": 0.698,
|
1153 |
+
"eval_test_ter_num_edits": 13882,
|
1154 |
+
"eval_test_ter_ref_length": 17367.0,
|
1155 |
+
"eval_test_ter_score": 79.93320665630218,
|
1156 |
+
"step": 14000
|
1157 |
+
},
|
1158 |
+
{
|
1159 |
+
"epoch": 4.1,
|
1160 |
+
"learning_rate": 0.000299765184558522,
|
1161 |
+
"loss": 2.331,
|
1162 |
+
"step": 15000
|
1163 |
+
},
|
1164 |
+
{
|
1165 |
+
"epoch": 4.1,
|
1166 |
+
"eval_validation_chrf_beta": 2,
|
1167 |
+
"eval_validation_chrf_char_order": 6,
|
1168 |
+
"eval_validation_chrf_score": 39.13729411872558,
|
1169 |
+
"eval_validation_chrf_word_order": 2,
|
1170 |
+
"eval_validation_loss": 2.7163193225860596,
|
1171 |
+
"eval_validation_meteor": 0.354395671848595,
|
1172 |
+
"eval_validation_runtime": 19.2961,
|
1173 |
+
"eval_validation_samples_per_second": 51.668,
|
1174 |
+
"eval_validation_scarebleu_bp": 1.0,
|
1175 |
+
"eval_validation_scarebleu_counts": [
|
1176 |
+
8486,
|
1177 |
+
3263,
|
1178 |
+
1446,
|
1179 |
+
664
|
1180 |
+
],
|
1181 |
+
"eval_validation_scarebleu_precisions": [
|
1182 |
+
43.388894569996936,
|
1183 |
+
17.5798717741501,
|
1184 |
+
8.23274880437258,
|
1185 |
+
4.007967646526227
|
1186 |
+
],
|
1187 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
1188 |
+
"eval_validation_scarebleu_score": 12.595512408378779,
|
1189 |
+
"eval_validation_scarebleu_sys_len": 19558,
|
1190 |
+
"eval_validation_scarebleu_totals": [
|
1191 |
+
19558,
|
1192 |
+
18561,
|
1193 |
+
17564,
|
1194 |
+
16567
|
1195 |
+
],
|
1196 |
+
"eval_validation_steps_per_second": 0.829,
|
1197 |
+
"eval_validation_ter_num_edits": 13087,
|
1198 |
+
"eval_validation_ter_ref_length": 16729.0,
|
1199 |
+
"eval_validation_ter_score": 78.22942196186263,
|
1200 |
+
"step": 15000
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"epoch": 4.1,
|
1204 |
+
"eval_test_chrf_beta": 2,
|
1205 |
+
"eval_test_chrf_char_order": 6,
|
1206 |
+
"eval_test_chrf_score": 39.693895702702456,
|
1207 |
+
"eval_test_chrf_word_order": 2,
|
1208 |
+
"eval_test_loss": 2.730414390563965,
|
1209 |
+
"eval_test_meteor": 0.35872327155724143,
|
1210 |
+
"eval_test_runtime": 23.3127,
|
1211 |
+
"eval_test_samples_per_second": 43.41,
|
1212 |
+
"eval_test_scarebleu_bp": 1.0,
|
1213 |
+
"eval_test_scarebleu_counts": [
|
1214 |
+
8979,
|
1215 |
+
3479,
|
1216 |
+
1580,
|
1217 |
+
745
|
1218 |
+
],
|
1219 |
+
"eval_test_scarebleu_precisions": [
|
1220 |
+
43.72321776392676,
|
1221 |
+
17.819094447859044,
|
1222 |
+
8.535004321521175,
|
1223 |
+
4.257142857142857
|
1224 |
+
],
|
1225 |
+
"eval_test_scarebleu_ref_len": 20238,
|
1226 |
+
"eval_test_scarebleu_score": 12.971193222930435,
|
1227 |
+
"eval_test_scarebleu_sys_len": 20536,
|
1228 |
+
"eval_test_scarebleu_totals": [
|
1229 |
+
20536,
|
1230 |
+
19524,
|
1231 |
+
18512,
|
1232 |
+
17500
|
1233 |
+
],
|
1234 |
+
"eval_test_steps_per_second": 0.686,
|
1235 |
+
"eval_test_ter_num_edits": 13688,
|
1236 |
+
"eval_test_ter_ref_length": 17367.0,
|
1237 |
+
"eval_test_ter_score": 78.81614556342488,
|
1238 |
+
"step": 15000
|
1239 |
+
},
|
1240 |
+
{
|
1241 |
+
"epoch": 4.37,
|
1242 |
+
"learning_rate": 0.00029972768102047483,
|
1243 |
+
"loss": 2.2516,
|
1244 |
+
"step": 16000
|
1245 |
+
},
|
1246 |
+
{
|
1247 |
+
"epoch": 4.37,
|
1248 |
+
"eval_validation_chrf_beta": 2,
|
1249 |
+
"eval_validation_chrf_char_order": 6,
|
1250 |
+
"eval_validation_chrf_score": 39.53731351221887,
|
1251 |
+
"eval_validation_chrf_word_order": 2,
|
1252 |
+
"eval_validation_loss": 2.65919828414917,
|
1253 |
+
"eval_validation_meteor": 0.35875663369715244,
|
1254 |
+
"eval_validation_runtime": 18.3314,
|
1255 |
+
"eval_validation_samples_per_second": 54.387,
|
1256 |
+
"eval_validation_scarebleu_bp": 1.0,
|
1257 |
+
"eval_validation_scarebleu_counts": [
|
1258 |
+
8559,
|
1259 |
+
3305,
|
1260 |
+
1499,
|
1261 |
+
688
|
1262 |
+
],
|
1263 |
+
"eval_validation_scarebleu_precisions": [
|
1264 |
+
43.7442502299908,
|
1265 |
+
17.79848133986752,
|
1266 |
+
8.530616890507625,
|
1267 |
+
4.150829562594269
|
1268 |
+
],
|
1269 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
1270 |
+
"eval_validation_scarebleu_score": 12.885605055056248,
|
1271 |
+
"eval_validation_scarebleu_sys_len": 19566,
|
1272 |
+
"eval_validation_scarebleu_totals": [
|
1273 |
+
19566,
|
1274 |
+
18569,
|
1275 |
+
17572,
|
1276 |
+
16575
|
1277 |
+
],
|
1278 |
+
"eval_validation_steps_per_second": 0.873,
|
1279 |
+
"eval_validation_ter_num_edits": 13032,
|
1280 |
+
"eval_validation_ter_ref_length": 16729.0,
|
1281 |
+
"eval_validation_ter_score": 77.9006515631538,
|
1282 |
+
"step": 16000
|
1283 |
+
},
|
1284 |
+
{
|
1285 |
+
"epoch": 4.37,
|
1286 |
+
"eval_test_chrf_beta": 2,
|
1287 |
+
"eval_test_chrf_char_order": 6,
|
1288 |
+
"eval_test_chrf_score": 40.0170544194821,
|
1289 |
+
"eval_test_chrf_word_order": 2,
|
1290 |
+
"eval_test_loss": 2.6628172397613525,
|
1291 |
+
"eval_test_meteor": 0.364117400692827,
|
1292 |
+
"eval_test_runtime": 20.9827,
|
1293 |
+
"eval_test_samples_per_second": 48.23,
|
1294 |
+
"eval_test_scarebleu_bp": 1.0,
|
1295 |
+
"eval_test_scarebleu_counts": [
|
1296 |
+
9039,
|
1297 |
+
3538,
|
1298 |
+
1634,
|
1299 |
+
772
|
1300 |
+
],
|
1301 |
+
"eval_test_scarebleu_precisions": [
|
1302 |
+
44.11420204978038,
|
1303 |
+
18.164082554677073,
|
1304 |
+
8.848694898732806,
|
1305 |
+
4.423054887131889
|
1306 |
+
],
|
1307 |
+
"eval_test_scarebleu_ref_len": 20238,
|
1308 |
+
"eval_test_scarebleu_score": 13.307557800740135,
|
1309 |
+
"eval_test_scarebleu_sys_len": 20490,
|
1310 |
+
"eval_test_scarebleu_totals": [
|
1311 |
+
20490,
|
1312 |
+
19478,
|
1313 |
+
18466,
|
1314 |
+
17454
|
1315 |
+
],
|
1316 |
+
"eval_test_steps_per_second": 0.763,
|
1317 |
+
"eval_test_ter_num_edits": 13615,
|
1318 |
+
"eval_test_ter_ref_length": 17367.0,
|
1319 |
+
"eval_test_ter_score": 78.39580814187828,
|
1320 |
+
"step": 16000
|
1321 |
+
},
|
1322 |
+
{
|
1323 |
+
"epoch": 4.65,
|
1324 |
+
"learning_rate": 0.00029968740292668196,
|
1325 |
+
"loss": 2.213,
|
1326 |
+
"step": 17000
|
1327 |
+
},
|
1328 |
+
{
|
1329 |
+
"epoch": 4.65,
|
1330 |
+
"eval_validation_chrf_beta": 2,
|
1331 |
+
"eval_validation_chrf_char_order": 6,
|
1332 |
+
"eval_validation_chrf_score": 40.099777569436895,
|
1333 |
+
"eval_validation_chrf_word_order": 2,
|
1334 |
+
"eval_validation_loss": 2.612889051437378,
|
1335 |
+
"eval_validation_meteor": 0.3648782076482497,
|
1336 |
+
"eval_validation_runtime": 18.3903,
|
1337 |
+
"eval_validation_samples_per_second": 54.213,
|
1338 |
+
"eval_validation_scarebleu_bp": 1.0,
|
1339 |
+
"eval_validation_scarebleu_counts": [
|
1340 |
+
8663,
|
1341 |
+
3436,
|
1342 |
+
1573,
|
1343 |
+
730
|
1344 |
+
],
|
1345 |
+
"eval_validation_scarebleu_precisions": [
|
1346 |
+
43.93002028397566,
|
1347 |
+
18.351759867542594,
|
1348 |
+
8.873970438903307,
|
1349 |
+
4.363679837408093
|
1350 |
+
],
|
1351 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
1352 |
+
"eval_validation_scarebleu_score": 13.292372233680728,
|
1353 |
+
"eval_validation_scarebleu_sys_len": 19720,
|
1354 |
+
"eval_validation_scarebleu_totals": [
|
1355 |
+
19720,
|
1356 |
+
18723,
|
1357 |
+
17726,
|
1358 |
+
16729
|
1359 |
+
],
|
1360 |
+
"eval_validation_steps_per_second": 0.87,
|
1361 |
+
"eval_validation_ter_num_edits": 12962,
|
1362 |
+
"eval_validation_ter_ref_length": 16729.0,
|
1363 |
+
"eval_validation_ter_score": 77.48221651025166,
|
1364 |
+
"step": 17000
|
1365 |
+
},
|
1366 |
+
{
|
1367 |
+
"epoch": 4.65,
|
1368 |
+
"eval_test_chrf_beta": 2,
|
1369 |
+
"eval_test_chrf_char_order": 6,
|
1370 |
+
"eval_test_chrf_score": 40.94861131064201,
|
1371 |
+
"eval_test_chrf_word_order": 2,
|
1372 |
+
"eval_test_loss": 2.625495195388794,
|
1373 |
+
"eval_test_meteor": 0.37167048896927946,
|
1374 |
+
"eval_test_runtime": 23.8134,
|
1375 |
+
"eval_test_samples_per_second": 42.497,
|
1376 |
+
"eval_test_scarebleu_bp": 1.0,
|
1377 |
+
"eval_test_scarebleu_counts": [
|
1378 |
+
9193,
|
1379 |
+
3711,
|
1380 |
+
1745,
|
1381 |
+
853
|
1382 |
+
],
|
1383 |
+
"eval_test_scarebleu_precisions": [
|
1384 |
+
44.511693216481866,
|
1385 |
+
18.894149992362916,
|
1386 |
+
9.367115787213484,
|
1387 |
+
4.841914060282681
|
1388 |
+
],
|
1389 |
+
"eval_test_scarebleu_ref_len": 20238,
|
1390 |
+
"eval_test_scarebleu_score": 13.975138151336937,
|
1391 |
+
"eval_test_scarebleu_sys_len": 20653,
|
1392 |
+
"eval_test_scarebleu_totals": [
|
1393 |
+
20653,
|
1394 |
+
19641,
|
1395 |
+
18629,
|
1396 |
+
17617
|
1397 |
+
],
|
1398 |
+
"eval_test_steps_per_second": 0.672,
|
1399 |
+
"eval_test_ter_num_edits": 13520,
|
1400 |
+
"eval_test_ter_ref_length": 17367.0,
|
1401 |
+
"eval_test_ter_score": 77.84879368918062,
|
1402 |
+
"step": 17000
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 4.92,
|
1406 |
+
"learning_rate": 0.00029964435102352384,
|
1407 |
+
"loss": 2.172,
|
1408 |
+
"step": 18000
|
1409 |
+
},
|
1410 |
+
{
|
1411 |
+
"epoch": 4.92,
|
1412 |
+
"eval_validation_chrf_beta": 2,
|
1413 |
+
"eval_validation_chrf_char_order": 6,
|
1414 |
+
"eval_validation_chrf_score": 40.57571982774164,
|
1415 |
+
"eval_validation_chrf_word_order": 2,
|
1416 |
+
"eval_validation_loss": 2.5481536388397217,
|
1417 |
+
"eval_validation_meteor": 0.3734694399709547,
|
1418 |
+
"eval_validation_runtime": 19.2312,
|
1419 |
+
"eval_validation_samples_per_second": 51.843,
|
1420 |
+
"eval_validation_scarebleu_bp": 1.0,
|
1421 |
+
"eval_validation_scarebleu_counts": [
|
1422 |
+
8738,
|
1423 |
+
3503,
|
1424 |
+
1625,
|
1425 |
+
771
|
1426 |
+
],
|
1427 |
+
"eval_validation_scarebleu_precisions": [
|
1428 |
+
44.85165794066317,
|
1429 |
+
18.950500405734378,
|
1430 |
+
9.292086001829826,
|
1431 |
+
4.675277424049481
|
1432 |
+
],
|
1433 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
1434 |
+
"eval_validation_scarebleu_score": 13.862129372426606,
|
1435 |
+
"eval_validation_scarebleu_sys_len": 19482,
|
1436 |
+
"eval_validation_scarebleu_totals": [
|
1437 |
+
19482,
|
1438 |
+
18485,
|
1439 |
+
17488,
|
1440 |
+
16491
|
1441 |
+
],
|
1442 |
+
"eval_validation_steps_per_second": 0.832,
|
1443 |
+
"eval_validation_ter_num_edits": 12785,
|
1444 |
+
"eval_validation_ter_ref_length": 16729.0,
|
1445 |
+
"eval_validation_ter_score": 76.42417359077052,
|
1446 |
+
"step": 18000
|
1447 |
+
},
|
1448 |
+
{
|
1449 |
+
"epoch": 4.92,
|
1450 |
+
"eval_test_chrf_beta": 2,
|
1451 |
+
"eval_test_chrf_char_order": 6,
|
1452 |
+
"eval_test_chrf_score": 41.3542278861197,
|
1453 |
+
"eval_test_chrf_word_order": 2,
|
1454 |
+
"eval_test_loss": 2.5593228340148926,
|
1455 |
+
"eval_test_meteor": 0.3781022025701054,
|
1456 |
+
"eval_test_runtime": 21.5553,
|
1457 |
+
"eval_test_samples_per_second": 46.949,
|
1458 |
+
"eval_test_scarebleu_bp": 1.0,
|
1459 |
+
"eval_test_scarebleu_counts": [
|
1460 |
+
9240,
|
1461 |
+
3803,
|
1462 |
+
1804,
|
1463 |
+
862
|
1464 |
+
],
|
1465 |
+
"eval_test_scarebleu_precisions": [
|
1466 |
+
45.42772861356932,
|
1467 |
+
19.676117549668874,
|
1468 |
+
9.84931207687268,
|
1469 |
+
4.981507165973185
|
1470 |
+
],
|
1471 |
+
"eval_test_scarebleu_ref_len": 20238,
|
1472 |
+
"eval_test_scarebleu_score": 14.471274188368865,
|
1473 |
+
"eval_test_scarebleu_sys_len": 20340,
|
1474 |
+
"eval_test_scarebleu_totals": [
|
1475 |
+
20340,
|
1476 |
+
19328,
|
1477 |
+
18316,
|
1478 |
+
17304
|
1479 |
+
],
|
1480 |
+
"eval_test_steps_per_second": 0.742,
|
1481 |
+
"eval_test_ter_num_edits": 13336,
|
1482 |
+
"eval_test_ter_ref_length": 17367.0,
|
1483 |
+
"eval_test_ter_score": 76.78931306500834,
|
1484 |
+
"step": 18000
|
1485 |
+
},
|
1486 |
+
{
|
1487 |
+
"epoch": 5.19,
|
1488 |
+
"learning_rate": 0.0002995985261087815,
|
1489 |
+
"loss": 2.1011,
|
1490 |
+
"step": 19000
|
1491 |
+
},
|
1492 |
+
{
|
1493 |
+
"epoch": 5.19,
|
1494 |
+
"eval_validation_chrf_beta": 2,
|
1495 |
+
"eval_validation_chrf_char_order": 6,
|
1496 |
+
"eval_validation_chrf_score": 40.91683022147655,
|
1497 |
+
"eval_validation_chrf_word_order": 2,
|
1498 |
+
"eval_validation_loss": 2.509413957595825,
|
1499 |
+
"eval_validation_meteor": 0.3741422146344069,
|
1500 |
+
"eval_validation_runtime": 19.6815,
|
1501 |
+
"eval_validation_samples_per_second": 50.657,
|
1502 |
+
"eval_validation_scarebleu_bp": 1.0,
|
1503 |
+
"eval_validation_scarebleu_counts": [
|
1504 |
+
8797,
|
1505 |
+
3539,
|
1506 |
+
1630,
|
1507 |
+
777
|
1508 |
+
],
|
1509 |
+
"eval_validation_scarebleu_precisions": [
|
1510 |
+
44.709290506200446,
|
1511 |
+
18.94641040740939,
|
1512 |
+
9.218414206537721,
|
1513 |
+
4.656877434821696
|
1514 |
+
],
|
1515 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
1516 |
+
"eval_validation_scarebleu_score": 13.809213245335025,
|
1517 |
+
"eval_validation_scarebleu_sys_len": 19676,
|
1518 |
+
"eval_validation_scarebleu_totals": [
|
1519 |
+
19676,
|
1520 |
+
18679,
|
1521 |
+
17682,
|
1522 |
+
16685
|
1523 |
+
],
|
1524 |
+
"eval_validation_steps_per_second": 0.813,
|
1525 |
+
"eval_validation_ter_num_edits": 12842,
|
1526 |
+
"eval_validation_ter_ref_length": 16729.0,
|
1527 |
+
"eval_validation_ter_score": 76.76489927670512,
|
1528 |
+
"step": 19000
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 5.19,
|
1532 |
+
"eval_test_chrf_beta": 2,
|
1533 |
+
"eval_test_chrf_char_order": 6,
|
1534 |
+
"eval_test_chrf_score": 41.61549355265803,
|
1535 |
+
"eval_test_chrf_word_order": 2,
|
1536 |
+
"eval_test_loss": 2.518157482147217,
|
1537 |
+
"eval_test_meteor": 0.3797284608780533,
|
1538 |
+
"eval_test_runtime": 19.8347,
|
1539 |
+
"eval_test_samples_per_second": 51.022,
|
1540 |
+
"eval_test_scarebleu_bp": 1.0,
|
1541 |
+
"eval_test_scarebleu_counts": [
|
1542 |
+
9292,
|
1543 |
+
3832,
|
1544 |
+
1832,
|
1545 |
+
898
|
1546 |
+
],
|
1547 |
+
"eval_test_scarebleu_precisions": [
|
1548 |
+
45.2297507788162,
|
1549 |
+
19.61908662707352,
|
1550 |
+
9.892008639308855,
|
1551 |
+
5.129083847384053
|
1552 |
+
],
|
1553 |
+
"eval_test_scarebleu_ref_len": 20238,
|
1554 |
+
"eval_test_scarebleu_score": 14.566553770170765,
|
1555 |
+
"eval_test_scarebleu_sys_len": 20544,
|
1556 |
+
"eval_test_scarebleu_totals": [
|
1557 |
+
20544,
|
1558 |
+
19532,
|
1559 |
+
18520,
|
1560 |
+
17508
|
1561 |
+
],
|
1562 |
+
"eval_test_steps_per_second": 0.807,
|
1563 |
+
"eval_test_ter_num_edits": 13324,
|
1564 |
+
"eval_test_ter_ref_length": 17367.0,
|
1565 |
+
"eval_test_ter_score": 76.72021650256234,
|
1566 |
+
"step": 19000
|
1567 |
+
},
|
1568 |
+
{
|
1569 |
+
"epoch": 5.47,
|
1570 |
+
"learning_rate": 0.0002995499290316219,
|
1571 |
+
"loss": 2.0657,
|
1572 |
+
"step": 20000
|
1573 |
+
},
|
1574 |
+
{
|
1575 |
+
"epoch": 5.47,
|
1576 |
+
"eval_validation_chrf_beta": 2,
|
1577 |
+
"eval_validation_chrf_char_order": 6,
|
1578 |
+
"eval_validation_chrf_score": 41.094856625774376,
|
1579 |
+
"eval_validation_chrf_word_order": 2,
|
1580 |
+
"eval_validation_loss": 2.466625452041626,
|
1581 |
+
"eval_validation_meteor": 0.378875561890472,
|
1582 |
+
"eval_validation_runtime": 18.4931,
|
1583 |
+
"eval_validation_samples_per_second": 53.912,
|
1584 |
+
"eval_validation_scarebleu_bp": 1.0,
|
1585 |
+
"eval_validation_scarebleu_counts": [
|
1586 |
+
8851,
|
1587 |
+
3584,
|
1588 |
+
1652,
|
1589 |
+
766
|
1590 |
+
],
|
1591 |
+
"eval_validation_scarebleu_precisions": [
|
1592 |
+
44.70653601373876,
|
1593 |
+
19.062815807669804,
|
1594 |
+
9.278813749719165,
|
1595 |
+
4.557624799190814
|
1596 |
+
],
|
1597 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
1598 |
+
"eval_validation_scarebleu_score": 13.778351975109262,
|
1599 |
+
"eval_validation_scarebleu_sys_len": 19798,
|
1600 |
+
"eval_validation_scarebleu_totals": [
|
1601 |
+
19798,
|
1602 |
+
18801,
|
1603 |
+
17804,
|
1604 |
+
16807
|
1605 |
+
],
|
1606 |
+
"eval_validation_steps_per_second": 0.865,
|
1607 |
+
"eval_validation_ter_num_edits": 12843,
|
1608 |
+
"eval_validation_ter_ref_length": 16729.0,
|
1609 |
+
"eval_validation_ter_score": 76.77087692031802,
|
1610 |
+
"step": 20000
|
1611 |
+
},
|
1612 |
+
{
|
1613 |
+
"epoch": 5.47,
|
1614 |
+
"eval_test_chrf_beta": 2,
|
1615 |
+
"eval_test_chrf_char_order": 6,
|
1616 |
+
"eval_test_chrf_score": 42.01510354031067,
|
1617 |
+
"eval_test_chrf_word_order": 2,
|
1618 |
+
"eval_test_loss": 2.4802513122558594,
|
1619 |
+
"eval_test_meteor": 0.3865124149670556,
|
1620 |
+
"eval_test_runtime": 21.1754,
|
1621 |
+
"eval_test_samples_per_second": 47.791,
|
1622 |
+
"eval_test_scarebleu_bp": 1.0,
|
1623 |
+
"eval_test_scarebleu_counts": [
|
1624 |
+
9418,
|
1625 |
+
3919,
|
1626 |
+
1907,
|
1627 |
+
937
|
1628 |
+
],
|
1629 |
+
"eval_test_scarebleu_precisions": [
|
1630 |
+
45.62542389303362,
|
1631 |
+
19.964340295466123,
|
1632 |
+
10.242775808357504,
|
1633 |
+
5.322049301374531
|
1634 |
+
],
|
1635 |
+
"eval_test_scarebleu_ref_len": 20238,
|
1636 |
+
"eval_test_scarebleu_score": 14.927586937806508,
|
1637 |
+
"eval_test_scarebleu_sys_len": 20642,
|
1638 |
+
"eval_test_scarebleu_totals": [
|
1639 |
+
20642,
|
1640 |
+
19630,
|
1641 |
+
18618,
|
1642 |
+
17606
|
1643 |
+
],
|
1644 |
+
"eval_test_steps_per_second": 0.756,
|
1645 |
+
"eval_test_ter_num_edits": 13273,
|
1646 |
+
"eval_test_ter_ref_length": 17367.0,
|
1647 |
+
"eval_test_ter_score": 76.42655611216675,
|
1648 |
+
"step": 20000
|
1649 |
+
},
|
1650 |
+
{
|
1651 |
+
"epoch": 5.74,
|
1652 |
+
"learning_rate": 0.00029949856069258176,
|
1653 |
+
"loss": 2.042,
|
1654 |
+
"step": 21000
|
1655 |
+
},
|
1656 |
+
{
|
1657 |
+
"epoch": 5.74,
|
1658 |
+
"eval_validation_chrf_beta": 2,
|
1659 |
+
"eval_validation_chrf_char_order": 6,
|
1660 |
+
"eval_validation_chrf_score": 41.49472756887711,
|
1661 |
+
"eval_validation_chrf_word_order": 2,
|
1662 |
+
"eval_validation_loss": 2.428771734237671,
|
1663 |
+
"eval_validation_meteor": 0.3799987462633123,
|
1664 |
+
"eval_validation_runtime": 19.422,
|
1665 |
+
"eval_validation_samples_per_second": 51.333,
|
1666 |
+
"eval_validation_scarebleu_bp": 1.0,
|
1667 |
+
"eval_validation_scarebleu_counts": [
|
1668 |
+
8875,
|
1669 |
+
3637,
|
1670 |
+
1710,
|
1671 |
+
824
|
1672 |
+
],
|
1673 |
+
"eval_validation_scarebleu_precisions": [
|
1674 |
+
45.40570960810396,
|
1675 |
+
19.607526012183946,
|
1676 |
+
9.742479489516864,
|
1677 |
+
4.9773482331621866
|
1678 |
+
],
|
1679 |
+
"eval_validation_scarebleu_ref_len": 19414,
|
1680 |
+
"eval_validation_scarebleu_score": 14.41452044509331,
|
1681 |
+
"eval_validation_scarebleu_sys_len": 19546,
|
1682 |
+
"eval_validation_scarebleu_totals": [
|
1683 |
+
19546,
|
1684 |
+
18549,
|
1685 |
+
17552,
|
1686 |
+
16555
|
1687 |
+
],
|
1688 |
+
"eval_validation_steps_per_second": 0.824,
|
1689 |
+
"eval_validation_ter_num_edits": 12677,
|
1690 |
+
"eval_validation_ter_ref_length": 16729.0,
|
1691 |
+
"eval_validation_ter_score": 75.77858808057863,
|
1692 |
+
"step": 21000
|
1693 |
+
},
|
1694 |
+
{
|
1695 |
+
"epoch": 5.74,
|
1696 |
+
"eval_test_chrf_beta": 2,
|
1697 |
+
"eval_test_chrf_char_order": 6,
|
1698 |
+
"eval_test_chrf_score": 42.150596446775204,
|
1699 |
+
"eval_test_chrf_word_order": 2,
|
1700 |
+
"eval_test_loss": 2.435882091522217,
|
1701 |
+
"eval_test_meteor": 0.3867430131575975,
|
1702 |
+
"eval_test_runtime": 21.5564,
|
1703 |
+
"eval_test_samples_per_second": 46.947,
|
1704 |
+
"eval_test_scarebleu_bp": 1.0,
|
1705 |
+
"eval_test_scarebleu_counts": [
|
1706 |
+
9407,
|
1707 |
+
3916,
|
1708 |
+
1868,
|
1709 |
+
903
|
1710 |
+
],
|
1711 |
+
"eval_test_scarebleu_precisions": [
|
1712 |
+
45.89899975603806,
|
1713 |
+
20.099573987578914,
|
1714 |
+
10.113150343782145,
|
1715 |
+
5.172117532504726
|
1716 |
+
],
|
1717 |
+
"eval_test_scarebleu_ref_len": 20238,
|
1718 |
+
"eval_test_scarebleu_score": 14.821297704758008,
|
1719 |
+
"eval_test_scarebleu_sys_len": 20495,
|
1720 |
+
"eval_test_scarebleu_totals": [
|
1721 |
+
20495,
|
1722 |
+
19483,
|
1723 |
+
18471,
|
1724 |
+
17459
|
1725 |
+
],
|
1726 |
+
"eval_test_steps_per_second": 0.742,
|
1727 |
+
"eval_test_ter_num_edits": 13178,
|
1728 |
+
"eval_test_ter_ref_length": 17367.0,
|
1729 |
+
"eval_test_ter_score": 75.87954165946911,
|
1730 |
+
"step": 21000
|
1731 |
+
}
|
1732 |
+
],
|
1733 |
+
"max_steps": 731800,
|
1734 |
+
"num_train_epochs": 200,
|
1735 |
+
"total_flos": 1.4694529965249004e+18,
|
1736 |
+
"trial_name": null,
|
1737 |
+
"trial_params": null
|
1738 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d993f7ffbd6981369c4ac5f4a1b13c7a1af645c6c0726aa348bdbf89ddaaf4b3
|
3 |
+
size 3835
|