Upload 14 files
Browse files- config.json +59 -0
- gitattributes +27 -0
- optimizer.pt +3 -0
- pytorch_model_2.bin +3 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- source.spm +0 -0
- special_tokens_map.json +1 -0
- target.spm +0 -0
- tokenizer_config.json +1 -0
- trainer_state.json +568 -0
- training_args.bin +3 -0
- vocab.json +0 -0
config.json
ADDED
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "Helsinki-NLP/opus-mt-en-mul",
|
3 |
+
"activation_dropout": 0.0,
|
4 |
+
"activation_function": "swish",
|
5 |
+
"add_bias_logits": false,
|
6 |
+
"add_final_layer_norm": false,
|
7 |
+
"architectures": [
|
8 |
+
"MarianMTModel"
|
9 |
+
],
|
10 |
+
"attention_dropout": 0.0,
|
11 |
+
"bad_words_ids": [
|
12 |
+
[
|
13 |
+
64109
|
14 |
+
]
|
15 |
+
],
|
16 |
+
"bos_token_id": 0,
|
17 |
+
"classif_dropout": 0.0,
|
18 |
+
"classifier_dropout": 0.0,
|
19 |
+
"d_model": 512,
|
20 |
+
"decoder_attention_heads": 8,
|
21 |
+
"decoder_ffn_dim": 2048,
|
22 |
+
"decoder_layerdrop": 0.0,
|
23 |
+
"decoder_layers": 6,
|
24 |
+
"decoder_start_token_id": 64109,
|
25 |
+
"dropout": 0.1,
|
26 |
+
"encoder_attention_heads": 8,
|
27 |
+
"encoder_ffn_dim": 2048,
|
28 |
+
"encoder_layerdrop": 0.0,
|
29 |
+
"encoder_layers": 6,
|
30 |
+
"eos_token_id": 0,
|
31 |
+
"extra_pos_embeddings": 64110,
|
32 |
+
"forced_eos_token_id": 0,
|
33 |
+
"id2label": {
|
34 |
+
"0": "LABEL_0",
|
35 |
+
"1": "LABEL_1",
|
36 |
+
"2": "LABEL_2"
|
37 |
+
},
|
38 |
+
"init_std": 0.02,
|
39 |
+
"is_encoder_decoder": true,
|
40 |
+
"label2id": {
|
41 |
+
"LABEL_0": 0,
|
42 |
+
"LABEL_1": 1,
|
43 |
+
"LABEL_2": 2
|
44 |
+
},
|
45 |
+
"max_length": 512,
|
46 |
+
"max_position_embeddings": 512,
|
47 |
+
"model_type": "marian",
|
48 |
+
"normalize_before": false,
|
49 |
+
"normalize_embedding": false,
|
50 |
+
"num_beams": 4,
|
51 |
+
"num_hidden_layers": 6,
|
52 |
+
"pad_token_id": 64109,
|
53 |
+
"scale_embedding": true,
|
54 |
+
"static_position_embeddings": true,
|
55 |
+
"torch_dtype": "float32",
|
56 |
+
"transformers_version": "4.12.5",
|
57 |
+
"use_cache": true,
|
58 |
+
"vocab_size": 64110
|
59 |
+
}
|
gitattributes
ADDED
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
*.7z filter=lfs diff=lfs merge=lfs -text
|
2 |
+
*.arrow filter=lfs diff=lfs merge=lfs -text
|
3 |
+
*.bin filter=lfs diff=lfs merge=lfs -text
|
4 |
+
*.bin.* filter=lfs diff=lfs merge=lfs -text
|
5 |
+
*.bz2 filter=lfs diff=lfs merge=lfs -text
|
6 |
+
*.ftz filter=lfs diff=lfs merge=lfs -text
|
7 |
+
*.gz filter=lfs diff=lfs merge=lfs -text
|
8 |
+
*.h5 filter=lfs diff=lfs merge=lfs -text
|
9 |
+
*.joblib filter=lfs diff=lfs merge=lfs -text
|
10 |
+
*.lfs.* filter=lfs diff=lfs merge=lfs -text
|
11 |
+
*.model filter=lfs diff=lfs merge=lfs -text
|
12 |
+
*.msgpack filter=lfs diff=lfs merge=lfs -text
|
13 |
+
*.onnx filter=lfs diff=lfs merge=lfs -text
|
14 |
+
*.ot filter=lfs diff=lfs merge=lfs -text
|
15 |
+
*.parquet filter=lfs diff=lfs merge=lfs -text
|
16 |
+
*.pb filter=lfs diff=lfs merge=lfs -text
|
17 |
+
*.pt filter=lfs diff=lfs merge=lfs -text
|
18 |
+
*.pth filter=lfs diff=lfs merge=lfs -text
|
19 |
+
*.rar filter=lfs diff=lfs merge=lfs -text
|
20 |
+
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
21 |
+
*.tar.* filter=lfs diff=lfs merge=lfs -text
|
22 |
+
*.tflite filter=lfs diff=lfs merge=lfs -text
|
23 |
+
*.tgz filter=lfs diff=lfs merge=lfs -text
|
24 |
+
*.xz filter=lfs diff=lfs merge=lfs -text
|
25 |
+
*.zip filter=lfs diff=lfs merge=lfs -text
|
26 |
+
*.zstandard filter=lfs diff=lfs merge=lfs -text
|
27 |
+
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c1b58990362d560dc143ec21ffc2594548ba9a7d152cb22bbafeb55a47f8f076
|
3 |
+
size 615849421
|
pytorch_model_2.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e16ff360577ef55de6e5eea1d5a9f63f18393a7839064417f9fb505306f13115
|
3 |
+
size 308202053
|
rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7ea20fe982488ed1ee798b65fe83f9c294a8d53cc2f5565eda25c95b541a203f
|
3 |
+
size 14503
|
scaler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:789c0c0371c0e442e481bd0868b60806543c4f9a8dc2e03b323f9a51e1810dbf
|
3 |
+
size 559
|
scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:470515f31c22590b249db77ae8849b7dbee64c6df4b901cf9437edd165b7a16a
|
3 |
+
size 623
|
source.spm
ADDED
Binary file (790 kB). View file
|
|
special_tokens_map.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>"}
|
target.spm
ADDED
Binary file (707 kB). View file
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"source_lang": "eng", "target_lang": "mul", "unk_token": "<unk>", "eos_token": "</s>", "pad_token": "<pad>", "model_max_length": 512, "sp_model_kwargs": {}, "special_tokens_map_file": null, "tokenizer_file": null, "name_or_path": "Helsinki-NLP/opus-mt-en-mul", "tokenizer_class": "MarianTokenizer"}
|
trainer_state.json
ADDED
@@ -0,0 +1,568 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 2.6507720947265625,
|
3 |
+
"best_model_checkpoint": "output-en-mul/checkpoint-975",
|
4 |
+
"epoch": 3.882320064584239,
|
5 |
+
"global_step": 975,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.1,
|
12 |
+
"eval_BLEU_ach": 1.8026,
|
13 |
+
"eval_BLEU_lgg": 2.4121,
|
14 |
+
"eval_BLEU_lug": 14.4386,
|
15 |
+
"eval_BLEU_mean": 4.5453,
|
16 |
+
"eval_BLEU_nyn": 3.307,
|
17 |
+
"eval_BLEU_teo": 0.7663,
|
18 |
+
"eval_loss": 4.095067024230957,
|
19 |
+
"eval_runtime": 117.7968,
|
20 |
+
"eval_samples_per_second": 21.223,
|
21 |
+
"eval_steps_per_second": 1.333,
|
22 |
+
"step": 25
|
23 |
+
},
|
24 |
+
{
|
25 |
+
"epoch": 0.2,
|
26 |
+
"eval_BLEU_ach": 6.6301,
|
27 |
+
"eval_BLEU_lgg": 3.6686,
|
28 |
+
"eval_BLEU_lug": 17.863,
|
29 |
+
"eval_BLEU_mean": 7.7115,
|
30 |
+
"eval_BLEU_nyn": 7.0341,
|
31 |
+
"eval_BLEU_teo": 3.3617,
|
32 |
+
"eval_loss": 3.4971461296081543,
|
33 |
+
"eval_runtime": 126.0103,
|
34 |
+
"eval_samples_per_second": 19.84,
|
35 |
+
"eval_steps_per_second": 1.246,
|
36 |
+
"step": 50
|
37 |
+
},
|
38 |
+
{
|
39 |
+
"epoch": 0.3,
|
40 |
+
"eval_BLEU_ach": 9.2766,
|
41 |
+
"eval_BLEU_lgg": 5.9959,
|
42 |
+
"eval_BLEU_lug": 18.5448,
|
43 |
+
"eval_BLEU_mean": 10.2227,
|
44 |
+
"eval_BLEU_nyn": 9.4904,
|
45 |
+
"eval_BLEU_teo": 7.8061,
|
46 |
+
"eval_loss": 3.2395877838134766,
|
47 |
+
"eval_runtime": 126.2158,
|
48 |
+
"eval_samples_per_second": 19.807,
|
49 |
+
"eval_steps_per_second": 1.244,
|
50 |
+
"step": 75
|
51 |
+
},
|
52 |
+
{
|
53 |
+
"epoch": 0.4,
|
54 |
+
"eval_BLEU_ach": 11.7826,
|
55 |
+
"eval_BLEU_lgg": 9.205,
|
56 |
+
"eval_BLEU_lug": 20.2742,
|
57 |
+
"eval_BLEU_mean": 12.501,
|
58 |
+
"eval_BLEU_nyn": 10.2936,
|
59 |
+
"eval_BLEU_teo": 10.9496,
|
60 |
+
"eval_loss": 3.0855562686920166,
|
61 |
+
"eval_runtime": 115.1293,
|
62 |
+
"eval_samples_per_second": 21.715,
|
63 |
+
"eval_steps_per_second": 1.364,
|
64 |
+
"step": 100
|
65 |
+
},
|
66 |
+
{
|
67 |
+
"epoch": 0.5,
|
68 |
+
"eval_BLEU_ach": 12.2884,
|
69 |
+
"eval_BLEU_lgg": 10.8685,
|
70 |
+
"eval_BLEU_lug": 20.0989,
|
71 |
+
"eval_BLEU_mean": 13.331,
|
72 |
+
"eval_BLEU_nyn": 10.8764,
|
73 |
+
"eval_BLEU_teo": 12.5227,
|
74 |
+
"eval_loss": 2.9902637004852295,
|
75 |
+
"eval_runtime": 112.0904,
|
76 |
+
"eval_samples_per_second": 22.303,
|
77 |
+
"eval_steps_per_second": 1.401,
|
78 |
+
"step": 125
|
79 |
+
},
|
80 |
+
{
|
81 |
+
"epoch": 0.6,
|
82 |
+
"eval_BLEU_ach": 12.9297,
|
83 |
+
"eval_BLEU_lgg": 12.0138,
|
84 |
+
"eval_BLEU_lug": 21.6123,
|
85 |
+
"eval_BLEU_mean": 14.4131,
|
86 |
+
"eval_BLEU_nyn": 11.2431,
|
87 |
+
"eval_BLEU_teo": 14.2664,
|
88 |
+
"eval_loss": 2.9314682483673096,
|
89 |
+
"eval_runtime": 114.1492,
|
90 |
+
"eval_samples_per_second": 21.901,
|
91 |
+
"eval_steps_per_second": 1.375,
|
92 |
+
"step": 150
|
93 |
+
},
|
94 |
+
{
|
95 |
+
"epoch": 0.7,
|
96 |
+
"eval_BLEU_ach": 13.8035,
|
97 |
+
"eval_BLEU_lgg": 13.3415,
|
98 |
+
"eval_BLEU_lug": 21.8656,
|
99 |
+
"eval_BLEU_mean": 15.1199,
|
100 |
+
"eval_BLEU_nyn": 11.9199,
|
101 |
+
"eval_BLEU_teo": 14.669,
|
102 |
+
"eval_loss": 2.8768787384033203,
|
103 |
+
"eval_runtime": 110.271,
|
104 |
+
"eval_samples_per_second": 22.671,
|
105 |
+
"eval_steps_per_second": 1.424,
|
106 |
+
"step": 175
|
107 |
+
},
|
108 |
+
{
|
109 |
+
"epoch": 0.79,
|
110 |
+
"eval_BLEU_ach": 13.8195,
|
111 |
+
"eval_BLEU_lgg": 13.478,
|
112 |
+
"eval_BLEU_lug": 21.4511,
|
113 |
+
"eval_BLEU_mean": 15.3648,
|
114 |
+
"eval_BLEU_nyn": 12.8941,
|
115 |
+
"eval_BLEU_teo": 15.1812,
|
116 |
+
"eval_loss": 2.8409996032714844,
|
117 |
+
"eval_runtime": 114.4231,
|
118 |
+
"eval_samples_per_second": 21.849,
|
119 |
+
"eval_steps_per_second": 1.372,
|
120 |
+
"step": 200
|
121 |
+
},
|
122 |
+
{
|
123 |
+
"epoch": 0.89,
|
124 |
+
"eval_BLEU_ach": 14.981,
|
125 |
+
"eval_BLEU_lgg": 14.1773,
|
126 |
+
"eval_BLEU_lug": 22.9603,
|
127 |
+
"eval_BLEU_mean": 16.2302,
|
128 |
+
"eval_BLEU_nyn": 12.8695,
|
129 |
+
"eval_BLEU_teo": 16.1631,
|
130 |
+
"eval_loss": 2.815749168395996,
|
131 |
+
"eval_runtime": 112.6187,
|
132 |
+
"eval_samples_per_second": 22.199,
|
133 |
+
"eval_steps_per_second": 1.394,
|
134 |
+
"step": 225
|
135 |
+
},
|
136 |
+
{
|
137 |
+
"epoch": 0.99,
|
138 |
+
"eval_BLEU_ach": 16.184,
|
139 |
+
"eval_BLEU_lgg": 15.1618,
|
140 |
+
"eval_BLEU_lug": 22.062,
|
141 |
+
"eval_BLEU_mean": 16.6458,
|
142 |
+
"eval_BLEU_nyn": 13.7415,
|
143 |
+
"eval_BLEU_teo": 16.0795,
|
144 |
+
"eval_loss": 2.785487174987793,
|
145 |
+
"eval_runtime": 111.0998,
|
146 |
+
"eval_samples_per_second": 22.502,
|
147 |
+
"eval_steps_per_second": 1.413,
|
148 |
+
"step": 250
|
149 |
+
},
|
150 |
+
{
|
151 |
+
"epoch": 1.1,
|
152 |
+
"eval_BLEU_ach": 16.0982,
|
153 |
+
"eval_BLEU_lgg": 15.3667,
|
154 |
+
"eval_BLEU_lug": 23.0175,
|
155 |
+
"eval_BLEU_mean": 17.0507,
|
156 |
+
"eval_BLEU_nyn": 13.6528,
|
157 |
+
"eval_BLEU_teo": 17.1184,
|
158 |
+
"eval_loss": 2.7754287719726562,
|
159 |
+
"eval_runtime": 109.9249,
|
160 |
+
"eval_samples_per_second": 22.743,
|
161 |
+
"eval_steps_per_second": 1.428,
|
162 |
+
"step": 275
|
163 |
+
},
|
164 |
+
{
|
165 |
+
"epoch": 1.19,
|
166 |
+
"eval_BLEU_ach": 15.5565,
|
167 |
+
"eval_BLEU_lgg": 15.718,
|
168 |
+
"eval_BLEU_lug": 23.1352,
|
169 |
+
"eval_BLEU_mean": 16.9913,
|
170 |
+
"eval_BLEU_nyn": 13.8518,
|
171 |
+
"eval_BLEU_teo": 16.695,
|
172 |
+
"eval_loss": 2.7620601654052734,
|
173 |
+
"eval_runtime": 110.9898,
|
174 |
+
"eval_samples_per_second": 22.525,
|
175 |
+
"eval_steps_per_second": 1.415,
|
176 |
+
"step": 300
|
177 |
+
},
|
178 |
+
{
|
179 |
+
"epoch": 1.29,
|
180 |
+
"eval_BLEU_ach": 16.1387,
|
181 |
+
"eval_BLEU_lgg": 16.7057,
|
182 |
+
"eval_BLEU_lug": 23.6614,
|
183 |
+
"eval_BLEU_mean": 17.5894,
|
184 |
+
"eval_BLEU_nyn": 13.8801,
|
185 |
+
"eval_BLEU_teo": 17.561,
|
186 |
+
"eval_loss": 2.7422778606414795,
|
187 |
+
"eval_runtime": 108.6905,
|
188 |
+
"eval_samples_per_second": 23.001,
|
189 |
+
"eval_steps_per_second": 1.444,
|
190 |
+
"step": 325
|
191 |
+
},
|
192 |
+
{
|
193 |
+
"epoch": 1.39,
|
194 |
+
"eval_BLEU_ach": 16.0125,
|
195 |
+
"eval_BLEU_lgg": 16.7273,
|
196 |
+
"eval_BLEU_lug": 23.3442,
|
197 |
+
"eval_BLEU_mean": 17.4938,
|
198 |
+
"eval_BLEU_nyn": 13.8987,
|
199 |
+
"eval_BLEU_teo": 17.4862,
|
200 |
+
"eval_loss": 2.7281692028045654,
|
201 |
+
"eval_runtime": 110.7001,
|
202 |
+
"eval_samples_per_second": 22.584,
|
203 |
+
"eval_steps_per_second": 1.418,
|
204 |
+
"step": 350
|
205 |
+
},
|
206 |
+
{
|
207 |
+
"epoch": 1.49,
|
208 |
+
"eval_BLEU_ach": 15.9115,
|
209 |
+
"eval_BLEU_lgg": 16.5585,
|
210 |
+
"eval_BLEU_lug": 24.5961,
|
211 |
+
"eval_BLEU_mean": 18.0068,
|
212 |
+
"eval_BLEU_nyn": 14.7126,
|
213 |
+
"eval_BLEU_teo": 18.2553,
|
214 |
+
"eval_loss": 2.7274107933044434,
|
215 |
+
"eval_runtime": 108.8652,
|
216 |
+
"eval_samples_per_second": 22.964,
|
217 |
+
"eval_steps_per_second": 1.442,
|
218 |
+
"step": 375
|
219 |
+
},
|
220 |
+
{
|
221 |
+
"epoch": 1.59,
|
222 |
+
"eval_BLEU_ach": 16.5652,
|
223 |
+
"eval_BLEU_lgg": 16.9672,
|
224 |
+
"eval_BLEU_lug": 23.9997,
|
225 |
+
"eval_BLEU_mean": 18.1261,
|
226 |
+
"eval_BLEU_nyn": 14.5361,
|
227 |
+
"eval_BLEU_teo": 18.5623,
|
228 |
+
"eval_loss": 2.7145676612854004,
|
229 |
+
"eval_runtime": 111.5209,
|
230 |
+
"eval_samples_per_second": 22.417,
|
231 |
+
"eval_steps_per_second": 1.408,
|
232 |
+
"step": 400
|
233 |
+
},
|
234 |
+
{
|
235 |
+
"epoch": 1.69,
|
236 |
+
"eval_BLEU_ach": 16.9721,
|
237 |
+
"eval_BLEU_lgg": 17.0032,
|
238 |
+
"eval_BLEU_lug": 24.5043,
|
239 |
+
"eval_BLEU_mean": 18.3546,
|
240 |
+
"eval_BLEU_nyn": 14.2281,
|
241 |
+
"eval_BLEU_teo": 19.0653,
|
242 |
+
"eval_loss": 2.7048418521881104,
|
243 |
+
"eval_runtime": 110.0323,
|
244 |
+
"eval_samples_per_second": 22.721,
|
245 |
+
"eval_steps_per_second": 1.427,
|
246 |
+
"step": 425
|
247 |
+
},
|
248 |
+
{
|
249 |
+
"epoch": 1.79,
|
250 |
+
"eval_BLEU_ach": 16.1836,
|
251 |
+
"eval_BLEU_lgg": 17.1272,
|
252 |
+
"eval_BLEU_lug": 24.5073,
|
253 |
+
"eval_BLEU_mean": 18.0436,
|
254 |
+
"eval_BLEU_nyn": 14.385,
|
255 |
+
"eval_BLEU_teo": 18.0147,
|
256 |
+
"eval_loss": 2.6982274055480957,
|
257 |
+
"eval_runtime": 109.4462,
|
258 |
+
"eval_samples_per_second": 22.842,
|
259 |
+
"eval_steps_per_second": 1.434,
|
260 |
+
"step": 450
|
261 |
+
},
|
262 |
+
{
|
263 |
+
"epoch": 1.89,
|
264 |
+
"eval_BLEU_ach": 16.1255,
|
265 |
+
"eval_BLEU_lgg": 17.1414,
|
266 |
+
"eval_BLEU_lug": 25.317,
|
267 |
+
"eval_BLEU_mean": 18.2627,
|
268 |
+
"eval_BLEU_nyn": 14.5828,
|
269 |
+
"eval_BLEU_teo": 18.1469,
|
270 |
+
"eval_loss": 2.689755916595459,
|
271 |
+
"eval_runtime": 119.3336,
|
272 |
+
"eval_samples_per_second": 20.95,
|
273 |
+
"eval_steps_per_second": 1.316,
|
274 |
+
"step": 475
|
275 |
+
},
|
276 |
+
{
|
277 |
+
"epoch": 1.99,
|
278 |
+
"learning_rate": 0.000250996015936255,
|
279 |
+
"loss": 2.9449,
|
280 |
+
"step": 500
|
281 |
+
},
|
282 |
+
{
|
283 |
+
"epoch": 1.99,
|
284 |
+
"eval_BLEU_ach": 15.5985,
|
285 |
+
"eval_BLEU_lgg": 17.9639,
|
286 |
+
"eval_BLEU_lug": 24.749,
|
287 |
+
"eval_BLEU_mean": 18.3452,
|
288 |
+
"eval_BLEU_nyn": 13.8982,
|
289 |
+
"eval_BLEU_teo": 19.5164,
|
290 |
+
"eval_loss": 2.6801609992980957,
|
291 |
+
"eval_runtime": 119.8601,
|
292 |
+
"eval_samples_per_second": 20.858,
|
293 |
+
"eval_steps_per_second": 1.31,
|
294 |
+
"step": 500
|
295 |
+
},
|
296 |
+
{
|
297 |
+
"epoch": 2.09,
|
298 |
+
"eval_BLEU_ach": 16.5254,
|
299 |
+
"eval_BLEU_lgg": 18.2852,
|
300 |
+
"eval_BLEU_lug": 24.2641,
|
301 |
+
"eval_BLEU_mean": 18.5989,
|
302 |
+
"eval_BLEU_nyn": 14.5069,
|
303 |
+
"eval_BLEU_teo": 19.4127,
|
304 |
+
"eval_loss": 2.681849956512451,
|
305 |
+
"eval_runtime": 117.9942,
|
306 |
+
"eval_samples_per_second": 21.187,
|
307 |
+
"eval_steps_per_second": 1.331,
|
308 |
+
"step": 525
|
309 |
+
},
|
310 |
+
{
|
311 |
+
"epoch": 2.19,
|
312 |
+
"eval_BLEU_ach": 16.299,
|
313 |
+
"eval_BLEU_lgg": 17.7158,
|
314 |
+
"eval_BLEU_lug": 24.1465,
|
315 |
+
"eval_BLEU_mean": 18.3291,
|
316 |
+
"eval_BLEU_nyn": 13.8427,
|
317 |
+
"eval_BLEU_teo": 19.6415,
|
318 |
+
"eval_loss": 2.6819944381713867,
|
319 |
+
"eval_runtime": 117.8806,
|
320 |
+
"eval_samples_per_second": 21.208,
|
321 |
+
"eval_steps_per_second": 1.332,
|
322 |
+
"step": 550
|
323 |
+
},
|
324 |
+
{
|
325 |
+
"epoch": 2.29,
|
326 |
+
"eval_BLEU_ach": 16.7809,
|
327 |
+
"eval_BLEU_lgg": 17.4699,
|
328 |
+
"eval_BLEU_lug": 25.1645,
|
329 |
+
"eval_BLEU_mean": 18.6296,
|
330 |
+
"eval_BLEU_nyn": 14.751,
|
331 |
+
"eval_BLEU_teo": 18.9815,
|
332 |
+
"eval_loss": 2.676717758178711,
|
333 |
+
"eval_runtime": 121.9425,
|
334 |
+
"eval_samples_per_second": 20.501,
|
335 |
+
"eval_steps_per_second": 1.287,
|
336 |
+
"step": 575
|
337 |
+
},
|
338 |
+
{
|
339 |
+
"epoch": 2.39,
|
340 |
+
"eval_BLEU_ach": 16.32,
|
341 |
+
"eval_BLEU_lgg": 17.7798,
|
342 |
+
"eval_BLEU_lug": 25.1038,
|
343 |
+
"eval_BLEU_mean": 18.7062,
|
344 |
+
"eval_BLEU_nyn": 14.4702,
|
345 |
+
"eval_BLEU_teo": 19.8571,
|
346 |
+
"eval_loss": 2.673452854156494,
|
347 |
+
"eval_runtime": 124.1156,
|
348 |
+
"eval_samples_per_second": 20.143,
|
349 |
+
"eval_steps_per_second": 1.265,
|
350 |
+
"step": 600
|
351 |
+
},
|
352 |
+
{
|
353 |
+
"epoch": 2.49,
|
354 |
+
"eval_BLEU_ach": 16.4085,
|
355 |
+
"eval_BLEU_lgg": 17.313,
|
356 |
+
"eval_BLEU_lug": 24.7636,
|
357 |
+
"eval_BLEU_mean": 18.6532,
|
358 |
+
"eval_BLEU_nyn": 14.7807,
|
359 |
+
"eval_BLEU_teo": 20.0003,
|
360 |
+
"eval_loss": 2.6720378398895264,
|
361 |
+
"eval_runtime": 129.4859,
|
362 |
+
"eval_samples_per_second": 19.307,
|
363 |
+
"eval_steps_per_second": 1.212,
|
364 |
+
"step": 625
|
365 |
+
},
|
366 |
+
{
|
367 |
+
"epoch": 2.59,
|
368 |
+
"eval_BLEU_ach": 17.9293,
|
369 |
+
"eval_BLEU_lgg": 18.0515,
|
370 |
+
"eval_BLEU_lug": 25.257,
|
371 |
+
"eval_BLEU_mean": 19.2304,
|
372 |
+
"eval_BLEU_nyn": 15.0062,
|
373 |
+
"eval_BLEU_teo": 19.9081,
|
374 |
+
"eval_loss": 2.6651768684387207,
|
375 |
+
"eval_runtime": 123.349,
|
376 |
+
"eval_samples_per_second": 20.268,
|
377 |
+
"eval_steps_per_second": 1.273,
|
378 |
+
"step": 650
|
379 |
+
},
|
380 |
+
{
|
381 |
+
"epoch": 2.69,
|
382 |
+
"eval_BLEU_ach": 17.6407,
|
383 |
+
"eval_BLEU_lgg": 18.0205,
|
384 |
+
"eval_BLEU_lug": 24.7637,
|
385 |
+
"eval_BLEU_mean": 19.1218,
|
386 |
+
"eval_BLEU_nyn": 14.5182,
|
387 |
+
"eval_BLEU_teo": 20.6661,
|
388 |
+
"eval_loss": 2.6660828590393066,
|
389 |
+
"eval_runtime": 128.8489,
|
390 |
+
"eval_samples_per_second": 19.403,
|
391 |
+
"eval_steps_per_second": 1.218,
|
392 |
+
"step": 675
|
393 |
+
},
|
394 |
+
{
|
395 |
+
"epoch": 2.79,
|
396 |
+
"eval_BLEU_ach": 17.3567,
|
397 |
+
"eval_BLEU_lgg": 18.3187,
|
398 |
+
"eval_BLEU_lug": 24.6553,
|
399 |
+
"eval_BLEU_mean": 18.7991,
|
400 |
+
"eval_BLEU_nyn": 14.4547,
|
401 |
+
"eval_BLEU_teo": 19.2101,
|
402 |
+
"eval_loss": 2.658956527709961,
|
403 |
+
"eval_runtime": 126.8237,
|
404 |
+
"eval_samples_per_second": 19.712,
|
405 |
+
"eval_steps_per_second": 1.238,
|
406 |
+
"step": 700
|
407 |
+
},
|
408 |
+
{
|
409 |
+
"epoch": 2.89,
|
410 |
+
"eval_BLEU_ach": 17.3235,
|
411 |
+
"eval_BLEU_lgg": 18.6854,
|
412 |
+
"eval_BLEU_lug": 25.0742,
|
413 |
+
"eval_BLEU_mean": 19.1268,
|
414 |
+
"eval_BLEU_nyn": 14.8488,
|
415 |
+
"eval_BLEU_teo": 19.7021,
|
416 |
+
"eval_loss": 2.6642343997955322,
|
417 |
+
"eval_runtime": 118.3036,
|
418 |
+
"eval_samples_per_second": 21.132,
|
419 |
+
"eval_steps_per_second": 1.327,
|
420 |
+
"step": 725
|
421 |
+
},
|
422 |
+
{
|
423 |
+
"epoch": 2.99,
|
424 |
+
"eval_BLEU_ach": 17.2203,
|
425 |
+
"eval_BLEU_lgg": 18.0574,
|
426 |
+
"eval_BLEU_lug": 25.487,
|
427 |
+
"eval_BLEU_mean": 19.0773,
|
428 |
+
"eval_BLEU_nyn": 14.7458,
|
429 |
+
"eval_BLEU_teo": 19.8762,
|
430 |
+
"eval_loss": 2.654205799102783,
|
431 |
+
"eval_runtime": 124.2827,
|
432 |
+
"eval_samples_per_second": 20.115,
|
433 |
+
"eval_steps_per_second": 1.263,
|
434 |
+
"step": 750
|
435 |
+
},
|
436 |
+
{
|
437 |
+
"epoch": 3.09,
|
438 |
+
"eval_BLEU_ach": 17.2162,
|
439 |
+
"eval_BLEU_lgg": 18.6608,
|
440 |
+
"eval_BLEU_lug": 25.0911,
|
441 |
+
"eval_BLEU_mean": 19.2459,
|
442 |
+
"eval_BLEU_nyn": 14.8809,
|
443 |
+
"eval_BLEU_teo": 20.3803,
|
444 |
+
"eval_loss": 2.6576144695281982,
|
445 |
+
"eval_runtime": 120.8965,
|
446 |
+
"eval_samples_per_second": 20.679,
|
447 |
+
"eval_steps_per_second": 1.299,
|
448 |
+
"step": 775
|
449 |
+
},
|
450 |
+
{
|
451 |
+
"epoch": 3.19,
|
452 |
+
"eval_BLEU_ach": 17.2713,
|
453 |
+
"eval_BLEU_lgg": 18.5656,
|
454 |
+
"eval_BLEU_lug": 25.0043,
|
455 |
+
"eval_BLEU_mean": 19.155,
|
456 |
+
"eval_BLEU_nyn": 14.6028,
|
457 |
+
"eval_BLEU_teo": 20.3311,
|
458 |
+
"eval_loss": 2.661299705505371,
|
459 |
+
"eval_runtime": 127.8339,
|
460 |
+
"eval_samples_per_second": 19.557,
|
461 |
+
"eval_steps_per_second": 1.228,
|
462 |
+
"step": 800
|
463 |
+
},
|
464 |
+
{
|
465 |
+
"epoch": 3.29,
|
466 |
+
"eval_BLEU_ach": 17.7839,
|
467 |
+
"eval_BLEU_lgg": 18.5926,
|
468 |
+
"eval_BLEU_lug": 25.1088,
|
469 |
+
"eval_BLEU_mean": 19.183,
|
470 |
+
"eval_BLEU_nyn": 14.4439,
|
471 |
+
"eval_BLEU_teo": 19.9857,
|
472 |
+
"eval_loss": 2.6544294357299805,
|
473 |
+
"eval_runtime": 128.4239,
|
474 |
+
"eval_samples_per_second": 19.467,
|
475 |
+
"eval_steps_per_second": 1.223,
|
476 |
+
"step": 825
|
477 |
+
},
|
478 |
+
{
|
479 |
+
"epoch": 3.39,
|
480 |
+
"eval_BLEU_ach": 17.2002,
|
481 |
+
"eval_BLEU_lgg": 18.8921,
|
482 |
+
"eval_BLEU_lug": 25.2989,
|
483 |
+
"eval_BLEU_mean": 19.3189,
|
484 |
+
"eval_BLEU_nyn": 15.0877,
|
485 |
+
"eval_BLEU_teo": 20.1156,
|
486 |
+
"eval_loss": 2.6571733951568604,
|
487 |
+
"eval_runtime": 127.0177,
|
488 |
+
"eval_samples_per_second": 19.682,
|
489 |
+
"eval_steps_per_second": 1.236,
|
490 |
+
"step": 850
|
491 |
+
},
|
492 |
+
{
|
493 |
+
"epoch": 3.48,
|
494 |
+
"eval_BLEU_ach": 17.4507,
|
495 |
+
"eval_BLEU_lgg": 19.0401,
|
496 |
+
"eval_BLEU_lug": 25.3251,
|
497 |
+
"eval_BLEU_mean": 19.2048,
|
498 |
+
"eval_BLEU_nyn": 14.5218,
|
499 |
+
"eval_BLEU_teo": 19.6862,
|
500 |
+
"eval_loss": 2.660137414932251,
|
501 |
+
"eval_runtime": 123.8715,
|
502 |
+
"eval_samples_per_second": 20.182,
|
503 |
+
"eval_steps_per_second": 1.267,
|
504 |
+
"step": 875
|
505 |
+
},
|
506 |
+
{
|
507 |
+
"epoch": 3.58,
|
508 |
+
"eval_BLEU_ach": 17.274,
|
509 |
+
"eval_BLEU_lgg": 18.9203,
|
510 |
+
"eval_BLEU_lug": 25.7546,
|
511 |
+
"eval_BLEU_mean": 19.4384,
|
512 |
+
"eval_BLEU_nyn": 14.8943,
|
513 |
+
"eval_BLEU_teo": 20.349,
|
514 |
+
"eval_loss": 2.6526899337768555,
|
515 |
+
"eval_runtime": 128.7667,
|
516 |
+
"eval_samples_per_second": 19.415,
|
517 |
+
"eval_steps_per_second": 1.219,
|
518 |
+
"step": 900
|
519 |
+
},
|
520 |
+
{
|
521 |
+
"epoch": 3.68,
|
522 |
+
"eval_BLEU_ach": 17.0936,
|
523 |
+
"eval_BLEU_lgg": 18.8419,
|
524 |
+
"eval_BLEU_lug": 25.4369,
|
525 |
+
"eval_BLEU_mean": 19.3561,
|
526 |
+
"eval_BLEU_nyn": 14.8679,
|
527 |
+
"eval_BLEU_teo": 20.5401,
|
528 |
+
"eval_loss": 2.653111219406128,
|
529 |
+
"eval_runtime": 125.2153,
|
530 |
+
"eval_samples_per_second": 19.966,
|
531 |
+
"eval_steps_per_second": 1.254,
|
532 |
+
"step": 925
|
533 |
+
},
|
534 |
+
{
|
535 |
+
"epoch": 3.78,
|
536 |
+
"eval_BLEU_ach": 17.6214,
|
537 |
+
"eval_BLEU_lgg": 18.6578,
|
538 |
+
"eval_BLEU_lug": 25.5293,
|
539 |
+
"eval_BLEU_mean": 19.5211,
|
540 |
+
"eval_BLEU_nyn": 14.9188,
|
541 |
+
"eval_BLEU_teo": 20.8783,
|
542 |
+
"eval_loss": 2.6521167755126953,
|
543 |
+
"eval_runtime": 121.3634,
|
544 |
+
"eval_samples_per_second": 20.599,
|
545 |
+
"eval_steps_per_second": 1.294,
|
546 |
+
"step": 950
|
547 |
+
},
|
548 |
+
{
|
549 |
+
"epoch": 3.88,
|
550 |
+
"eval_BLEU_ach": 17.1886,
|
551 |
+
"eval_BLEU_lgg": 18.6225,
|
552 |
+
"eval_BLEU_lug": 25.2939,
|
553 |
+
"eval_BLEU_mean": 19.3689,
|
554 |
+
"eval_BLEU_nyn": 15.3009,
|
555 |
+
"eval_BLEU_teo": 20.4388,
|
556 |
+
"eval_loss": 2.6507720947265625,
|
557 |
+
"eval_runtime": 121.9609,
|
558 |
+
"eval_samples_per_second": 20.498,
|
559 |
+
"eval_steps_per_second": 1.287,
|
560 |
+
"step": 975
|
561 |
+
}
|
562 |
+
],
|
563 |
+
"max_steps": 1004,
|
564 |
+
"num_train_epochs": 4,
|
565 |
+
"total_flos": 1.0817461893464064e+17,
|
566 |
+
"trial_name": null,
|
567 |
+
"trial_params": null
|
568 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4ea6c112e3cf25652c44a1429c8c834292bca7a36e529627f5b6a6579843e26d
|
3 |
+
size 2927
|
vocab.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|