Hugging Face's logo

ethansimrm
/
opus_big_enfr_FT_wce

Transformers
PyTorch
marian
text2text-generation
Model card Files
xet
Community
1
Ethan Sim commited on
Commit
a35e093
·
1 Parent(s): 37d2a2c

stage best wce model

Browse files
Files changed (23) hide show
  1. checkpoint-48000/config.json +46 -0
  2. checkpoint-48000/generation_config.json +15 -0
  3. checkpoint-48000/optimizer.pt +3 -0
  4. checkpoint-48000/pytorch_model.bin +3 -0
  5. checkpoint-48000/rng_state.pth +3 -0
  6. checkpoint-48000/scaler.pt +3 -0
  7. checkpoint-48000/scheduler.pt +3 -0
  8. checkpoint-48000/source.spm +0 -0
  9. checkpoint-48000/special_tokens_map.json +5 -0
  10. checkpoint-48000/target.spm +0 -0
  11. checkpoint-48000/tokenizer_config.json +12 -0
  12. checkpoint-48000/trainer_state.json +112 -0
  13. checkpoint-48000/training_args.bin +3 -0
  14. checkpoint-48000/vocab.json +0 -0
  15. config.json +46 -0
  16. generation_config.json +15 -0
  17. pytorch_model.bin +3 -0
  18. source.spm +0 -0
  19. special_tokens_map.json +5 -0
  20. target.spm +0 -0
  21. tokenizer_config.json +12 -0
  22. training_args.bin +3 -0
  23. vocab.json +0 -0
checkpoint-48000/config.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "Helsinki-NLP/opus-mt-tc-big-en-fr",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "relu",
5
+ "architectures": [
6
+ "MarianMTModel"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "bad_words_ids": [
10
+ [
11
+ 53016
12
+ ]
13
+ ],
14
+ "bos_token_id": 0,
15
+ "classifier_dropout": 0.0,
16
+ "d_model": 1024,
17
+ "decoder_attention_heads": 16,
18
+ "decoder_ffn_dim": 4096,
19
+ "decoder_layerdrop": 0.0,
20
+ "decoder_layers": 6,
21
+ "decoder_start_token_id": 53016,
22
+ "decoder_vocab_size": 53017,
23
+ "dropout": 0.1,
24
+ "encoder_attention_heads": 16,
25
+ "encoder_ffn_dim": 4096,
26
+ "encoder_layerdrop": 0.0,
27
+ "encoder_layers": 6,
28
+ "eos_token_id": 43311,
29
+ "forced_eos_token_id": 43311,
30
+ "init_std": 0.02,
31
+ "is_encoder_decoder": true,
32
+ "max_length": 512,
33
+ "max_position_embeddings": 1024,
34
+ "model_type": "marian",
35
+ "normalize_embedding": false,
36
+ "num_beams": 4,
37
+ "num_hidden_layers": 6,
38
+ "pad_token_id": 53016,
39
+ "scale_embedding": true,
40
+ "share_encoder_decoder_embeddings": true,
41
+ "static_position_embeddings": true,
42
+ "torch_dtype": "float32",
43
+ "transformers_version": "4.28.0",
44
+ "use_cache": true,
45
+ "vocab_size": 53017
46
+ }
checkpoint-48000/generation_config.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bad_words_ids": [
3
+ [
4
+ 53016
5
+ ]
6
+ ],
7
+ "bos_token_id": 0,
8
+ "decoder_start_token_id": 53016,
9
+ "eos_token_id": 43311,
10
+ "forced_eos_token_id": 43311,
11
+ "max_length": 512,
12
+ "num_beams": 4,
13
+ "pad_token_id": 53016,
14
+ "transformers_version": "4.28.0"
15
+ }
checkpoint-48000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:feb944dd3694845a5a827e8355ddb3eb7119afc806e43fb93c003121f511b0c2
3
+ size 1845322757
checkpoint-48000/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:033a7f99e720b4906a371665a4a87d95d5707e9f03a48714c113d2e1db26282d
3
+ size 922885701
checkpoint-48000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0229a5f2115e7147f52ccec11492de81225a00cd1c6cf89a19ef407cf5ad3904
3
+ size 14575
checkpoint-48000/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2543dae2fad4aef2bd6aa05e0f4050d32e1386326a137c134e69671f720671c0
3
+ size 557
checkpoint-48000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0d5fb0ff60520aa2fb2cdc484ebac963ffe6c641d211ecec8e2ae7d9d5e8dd56
3
+ size 627
checkpoint-48000/source.spm ADDED
Binary file (802 kB). View file
 
checkpoint-48000/special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "eos_token": "</s>",
3
+ "pad_token": "<pad>",
4
+ "unk_token": "<unk>"
5
+ }
checkpoint-48000/target.spm ADDED
Binary file (820 kB). View file
 
checkpoint-48000/tokenizer_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "clean_up_tokenization_spaces": true,
3
+ "eos_token": "</s>",
4
+ "model_max_length": 512,
5
+ "pad_token": "<pad>",
6
+ "separate_vocabs": false,
7
+ "source_lang": "en",
8
+ "sp_model_kwargs": {},
9
+ "target_lang": "fr",
10
+ "tokenizer_class": "MarianTokenizer",
11
+ "unk_token": "<unk>"
12
+ }
checkpoint-48000/trainer_state.json ADDED
@@ -0,0 +1,112 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 43.4061,
3
+ "best_model_checkpoint": "opus_big_enfr_FT_wce/checkpoint-48000",
4
+ "epoch": 1.1829944547134936,
5
+ "global_step": 48000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.2,
12
+ "learning_rate": 1.9753635243376465e-05,
13
+ "loss": 0.1935,
14
+ "step": 8000
15
+ },
16
+ {
17
+ "epoch": 0.2,
18
+ "eval_bleu": 42.2657,
19
+ "eval_gen_len": 33.6098,
20
+ "eval_loss": 0.09848988056182861,
21
+ "eval_runtime": 153.6166,
22
+ "eval_samples_per_second": 6.79,
23
+ "eval_steps_per_second": 0.43,
24
+ "step": 8000
25
+ },
26
+ {
27
+ "epoch": 0.39,
28
+ "learning_rate": 1.950723967960567e-05,
29
+ "loss": 0.162,
30
+ "step": 16000
31
+ },
32
+ {
33
+ "epoch": 0.39,
34
+ "eval_bleu": 42.8204,
35
+ "eval_gen_len": 34.3816,
36
+ "eval_loss": 0.09720870107412338,
37
+ "eval_runtime": 158.3909,
38
+ "eval_samples_per_second": 6.585,
39
+ "eval_steps_per_second": 0.417,
40
+ "step": 16000
41
+ },
42
+ {
43
+ "epoch": 0.59,
44
+ "learning_rate": 1.9260874922982132e-05,
45
+ "loss": 0.1575,
46
+ "step": 24000
47
+ },
48
+ {
49
+ "epoch": 0.59,
50
+ "eval_bleu": 42.6214,
51
+ "eval_gen_len": 34.3087,
52
+ "eval_loss": 0.0975893884897232,
53
+ "eval_runtime": 129.8641,
54
+ "eval_samples_per_second": 8.031,
55
+ "eval_steps_per_second": 0.508,
56
+ "step": 24000
57
+ },
58
+ {
59
+ "epoch": 0.79,
60
+ "learning_rate": 1.901447935921134e-05,
61
+ "loss": 0.1533,
62
+ "step": 32000
63
+ },
64
+ {
65
+ "epoch": 0.79,
66
+ "eval_bleu": 42.7095,
67
+ "eval_gen_len": 34.1035,
68
+ "eval_loss": 0.09651403874158859,
69
+ "eval_runtime": 109.2582,
70
+ "eval_samples_per_second": 9.546,
71
+ "eval_steps_per_second": 0.604,
72
+ "step": 32000
73
+ },
74
+ {
75
+ "epoch": 0.99,
76
+ "learning_rate": 1.876814540973506e-05,
77
+ "loss": 0.1517,
78
+ "step": 40000
79
+ },
80
+ {
81
+ "epoch": 0.99,
82
+ "eval_bleu": 42.6552,
83
+ "eval_gen_len": 33.7267,
84
+ "eval_loss": 0.0962938442826271,
85
+ "eval_runtime": 121.4941,
86
+ "eval_samples_per_second": 8.585,
87
+ "eval_steps_per_second": 0.543,
88
+ "step": 40000
89
+ },
90
+ {
91
+ "epoch": 1.18,
92
+ "learning_rate": 1.852181146025878e-05,
93
+ "loss": 0.1352,
94
+ "step": 48000
95
+ },
96
+ {
97
+ "epoch": 1.18,
98
+ "eval_bleu": 43.4061,
99
+ "eval_gen_len": 33.8447,
100
+ "eval_loss": 0.09698126465082169,
101
+ "eval_runtime": 107.0828,
102
+ "eval_samples_per_second": 9.74,
103
+ "eval_steps_per_second": 0.616,
104
+ "step": 48000
105
+ }
106
+ ],
107
+ "max_steps": 649200,
108
+ "num_train_epochs": 16,
109
+ "total_flos": 1.5506756153848627e+17,
110
+ "trial_name": null,
111
+ "trial_params": null
112
+ }
checkpoint-48000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7271f16142689129ff74e2fbdfff484e58a6b4d54b9c292458687c830a3e99ee
3
+ size 3771
checkpoint-48000/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
config.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "Helsinki-NLP/opus-mt-tc-big-en-fr",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "relu",
5
+ "architectures": [
6
+ "MarianMTModel"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "bad_words_ids": [
10
+ [
11
+ 53016
12
+ ]
13
+ ],
14
+ "bos_token_id": 0,
15
+ "classifier_dropout": 0.0,
16
+ "d_model": 1024,
17
+ "decoder_attention_heads": 16,
18
+ "decoder_ffn_dim": 4096,
19
+ "decoder_layerdrop": 0.0,
20
+ "decoder_layers": 6,
21
+ "decoder_start_token_id": 53016,
22
+ "decoder_vocab_size": 53017,
23
+ "dropout": 0.1,
24
+ "encoder_attention_heads": 16,
25
+ "encoder_ffn_dim": 4096,
26
+ "encoder_layerdrop": 0.0,
27
+ "encoder_layers": 6,
28
+ "eos_token_id": 43311,
29
+ "forced_eos_token_id": 43311,
30
+ "init_std": 0.02,
31
+ "is_encoder_decoder": true,
32
+ "max_length": 512,
33
+ "max_position_embeddings": 1024,
34
+ "model_type": "marian",
35
+ "normalize_embedding": false,
36
+ "num_beams": 4,
37
+ "num_hidden_layers": 6,
38
+ "pad_token_id": 53016,
39
+ "scale_embedding": true,
40
+ "share_encoder_decoder_embeddings": true,
41
+ "static_position_embeddings": true,
42
+ "torch_dtype": "float32",
43
+ "transformers_version": "4.28.0",
44
+ "use_cache": true,
45
+ "vocab_size": 53017
46
+ }
generation_config.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bad_words_ids": [
3
+ [
4
+ 53016
5
+ ]
6
+ ],
7
+ "bos_token_id": 0,
8
+ "decoder_start_token_id": 53016,
9
+ "eos_token_id": 43311,
10
+ "forced_eos_token_id": 43311,
11
+ "max_length": 512,
12
+ "num_beams": 4,
13
+ "pad_token_id": 53016,
14
+ "transformers_version": "4.28.0"
15
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:033a7f99e720b4906a371665a4a87d95d5707e9f03a48714c113d2e1db26282d
3
+ size 922885701
source.spm ADDED
Binary file (802 kB). View file
 
special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "eos_token": "</s>",
3
+ "pad_token": "<pad>",
4
+ "unk_token": "<unk>"
5
+ }
target.spm ADDED
Binary file (820 kB). View file
 
tokenizer_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "clean_up_tokenization_spaces": true,
3
+ "eos_token": "</s>",
4
+ "model_max_length": 512,
5
+ "pad_token": "<pad>",
6
+ "separate_vocabs": false,
7
+ "source_lang": "en",
8
+ "sp_model_kwargs": {},
9
+ "target_lang": "fr",
10
+ "tokenizer_class": "MarianTokenizer",
11
+ "unk_token": "<unk>"
12
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7271f16142689129ff74e2fbdfff484e58a6b4d54b9c292458687c830a3e99ee
3
+ size 3771
vocab.json ADDED
The diff for this file is too large to render. See raw diff