Hugging Face's logo

calisolo
/
OFA_huge_NICE_captioning

Transformers
PyTorch
ofa
Inference Endpoints
Model card Files Community
calisolo commited on
Commit
0ccc8bd
•
1 Parent(s): 99ad9eb

Upload 17 files

Browse files

upload ensemble candidates

Files changed (17) hide show
  1. candidate1_trainLess/config.json +53 -0
  2. candidate1_trainLess/pytorch_model.bin +3 -0
  3. candidate1_trainLess/train_args.json +32 -0
  4. candidate1_trainLess/trainer_state.json +121 -0
  5. candidate1_trainLess/training_args.bin +3 -0
  6. candidate2_short/config.json +53 -0
  7. candidate2_short/pytorch_model.bin +3 -0
  8. candidate2_short/train_args.json +32 -0
  9. candidate2_short/trainer_state.json +73 -0
  10. candidate2_short/training_args.bin +3 -0
  11. candidate2_short/vocab.json +0 -0
  12. candidate3_lastcoin/config.json +53 -0
  13. candidate3_lastcoin/pytorch_model.bin +3 -0
  14. candidate3_lastcoin/train_args.json +32 -0
  15. candidate3_lastcoin/trainer_state.json +73 -0
  16. candidate3_lastcoin/training_args.bin +3 -0
  17. candidate3_lastcoin/vocab.json +0 -0
candidate1_trainLess/config.json ADDED
@@ -0,0 +1,53 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "../drive/MyDrive/NICE/model/cap_best/OFA-huge-caption",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "gelu",
5
+ "add_type_embedding": true,
6
+ "architectures": [
7
+ "OFAModelForCaption"
8
+ ],
9
+ "attention_dropout": 0.0,
10
+ "attn_scale_factor": 2.0,
11
+ "bos_token_id": 0,
12
+ "classifier_dropout": 0.0,
13
+ "code_image_size": 128,
14
+ "code_layernorm_embedding": true,
15
+ "d_model": 1280,
16
+ "decoder_attention_heads": 16,
17
+ "decoder_drop_path_rate": 0.0,
18
+ "decoder_ffn_dim": 5120,
19
+ "decoder_layerdrop": 0.0,
20
+ "decoder_layers": 12,
21
+ "decoder_normalize_before": true,
22
+ "decoder_start_token_id": 0,
23
+ "dropout": 0.1,
24
+ "encoder_attention_heads": 16,
25
+ "encoder_drop_path_rate": 0.0,
26
+ "encoder_ffn_dim": 5120,
27
+ "encoder_layerdrop": 0.0,
28
+ "encoder_layers": 24,
29
+ "encoder_normalize_before": true,
30
+ "entangle_position_embedding": false,
31
+ "eos_token_id": 2,
32
+ "forced_eos_token_id": 2,
33
+ "image_bucket_size": 42,
34
+ "init_std": 0.02,
35
+ "is_encoder_decoder": true,
36
+ "layernorm_embedding": true,
37
+ "max_position_embeddings": 1024,
38
+ "model_type": "ofa",
39
+ "normformer": true,
40
+ "num_hidden_layers": 24,
41
+ "pad_token_id": 1,
42
+ "patch_layernorm_embedding": true,
43
+ "resnet_drop_path_rate": 0.0,
44
+ "resnet_model_path": null,
45
+ "resnet_type": "resnet152",
46
+ "scale_embedding": false,
47
+ "share_decoder_input_output_embed": true,
48
+ "token_bucket_size": 256,
49
+ "torch_dtype": "float32",
50
+ "transformers_version": "4.20.0",
51
+ "use_cache": false,
52
+ "vocab_size": 59457
53
+ }
candidate1_trainLess/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a0bb830b18a2245756810d13732f645c138512924eff5862e896e65069178457
3
+ size 3785732668
candidate1_trainLess/train_args.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "output_dir": "../drive/MyDrive/NICE/model/0430_train_less",
3
+ "model_name_or_path": "../drive/MyDrive/NICE/model/cap_best/OFA-huge-caption",
4
+ "train_caption_file": "./data/train_caption_catAdded.jsonl",
5
+ "train_image_file": "../drive/MyDrive/NICE/train_image.tsv",
6
+ "test_caption_file": null,
7
+ "test_image_file": null,
8
+ "freeze_encoder": false,
9
+ "freeze_word_embed": false,
10
+ "num_train_epochs": 20,
11
+ "max_steps": -1,
12
+ "per_device_train_batch_size": 12,
13
+ "per_device_eval_batch_size": 256,
14
+ "learning_rate": 5e-05,
15
+ "max_seq_length": 150,
16
+ "logging_steps": 500,
17
+ "save_steps": 5000,
18
+ "save_total_limit": 1,
19
+ "lr_scheduler_type": "cosine",
20
+ "warmup_steps": 400,
21
+ "warmup_ratio": 0,
22
+ "gradient_accumulation_steps": 1,
23
+ "optim": "adamw_torch",
24
+ "seed": 69,
25
+ "fp16": false,
26
+ "no_cuda": false,
27
+ "dataloader_num_workers": 8,
28
+ "save_strategy": "steps",
29
+ "weight_decay": 0.01,
30
+ "max_grad_norm": 1.0,
31
+ "remove_unused_columns": false
32
+ }
candidate1_trainLess/trainer_state.json ADDED
@@ -0,0 +1,121 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 20.0,
5
+ "global_step": 8340,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 1.2,
12
+ "learning_rate": 4.9980433547419305e-05,
13
+ "loss": 2.3482,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 2.4,
18
+ "learning_rate": 4.929881824859985e-05,
19
+ "loss": 1.0043,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 3.6,
24
+ "learning_rate": 4.7669293205603196e-05,
25
+ "loss": 0.6055,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 4.8,
30
+ "learning_rate": 4.515542705171981e-05,
31
+ "loss": 0.3829,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 6.0,
36
+ "learning_rate": 4.1855287034436555e-05,
37
+ "loss": 0.2724,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 7.19,
42
+ "learning_rate": 3.789761336027403e-05,
43
+ "loss": 0.1855,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 8.39,
48
+ "learning_rate": 3.3436796971278526e-05,
49
+ "loss": 0.1394,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 9.59,
54
+ "learning_rate": 2.864685667276201e-05,
55
+ "loss": 0.0952,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 10.79,
60
+ "learning_rate": 2.3714650567721016e-05,
61
+ "loss": 0.0658,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 11.99,
66
+ "learning_rate": 1.883258662346596e-05,
67
+ "loss": 0.0426,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 13.19,
72
+ "learning_rate": 1.4191116735110007e-05,
73
+ "loss": 0.0287,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 14.39,
78
+ "learning_rate": 9.971307096468203e-06,
79
+ "loss": 0.0209,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 15.59,
84
+ "learning_rate": 6.337774712125597e-06,
85
+ "loss": 0.0166,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 16.79,
90
+ "learning_rate": 3.4322656010722542e-06,
91
+ "loss": 0.0146,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 17.99,
96
+ "learning_rate": 1.3681252095757852e-06,
97
+ "loss": 0.0129,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 19.18,
102
+ "learning_rate": 2.258767454083155e-07,
103
+ "loss": 0.0116,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 20.0,
108
+ "step": 8340,
109
+ "total_flos": 7.012455568878336e+16,
110
+ "train_loss": 0.31504503674358486,
111
+ "train_runtime": 11228.5117,
112
+ "train_samples_per_second": 8.906,
113
+ "train_steps_per_second": 0.743
114
+ }
115
+ ],
116
+ "max_steps": 8340,
117
+ "num_train_epochs": 20,
118
+ "total_flos": 7.012455568878336e+16,
119
+ "trial_name": null,
120
+ "trial_params": null
121
+ }
candidate1_trainLess/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aef4da31986be3bb58055a93adcca4d9c1a53a688bf4dbe78583eca0d4749451
3
+ size 3323
candidate2_short/config.json ADDED
@@ -0,0 +1,53 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "../drive/MyDrive/NICE/model/cap_best/OFA-huge-caption",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "gelu",
5
+ "add_type_embedding": true,
6
+ "architectures": [
7
+ "OFAModelForCaption"
8
+ ],
9
+ "attention_dropout": 0.0,
10
+ "attn_scale_factor": 2.0,
11
+ "bos_token_id": 0,
12
+ "classifier_dropout": 0.0,
13
+ "code_image_size": 128,
14
+ "code_layernorm_embedding": true,
15
+ "d_model": 1280,
16
+ "decoder_attention_heads": 16,
17
+ "decoder_drop_path_rate": 0.0,
18
+ "decoder_ffn_dim": 5120,
19
+ "decoder_layerdrop": 0.0,
20
+ "decoder_layers": 12,
21
+ "decoder_normalize_before": true,
22
+ "decoder_start_token_id": 0,
23
+ "dropout": 0.1,
24
+ "encoder_attention_heads": 16,
25
+ "encoder_drop_path_rate": 0.0,
26
+ "encoder_ffn_dim": 5120,
27
+ "encoder_layerdrop": 0.0,
28
+ "encoder_layers": 24,
29
+ "encoder_normalize_before": true,
30
+ "entangle_position_embedding": false,
31
+ "eos_token_id": 2,
32
+ "forced_eos_token_id": 2,
33
+ "image_bucket_size": 42,
34
+ "init_std": 0.02,
35
+ "is_encoder_decoder": true,
36
+ "layernorm_embedding": true,
37
+ "max_position_embeddings": 1024,
38
+ "model_type": "ofa",
39
+ "normformer": true,
40
+ "num_hidden_layers": 24,
41
+ "pad_token_id": 1,
42
+ "patch_layernorm_embedding": true,
43
+ "resnet_drop_path_rate": 0.0,
44
+ "resnet_model_path": null,
45
+ "resnet_type": "resnet152",
46
+ "scale_embedding": false,
47
+ "share_decoder_input_output_embed": true,
48
+ "token_bucket_size": 256,
49
+ "torch_dtype": "float32",
50
+ "transformers_version": "4.20.0",
51
+ "use_cache": false,
52
+ "vocab_size": 59457
53
+ }
candidate2_short/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f60a8d12f9f32fe721f822cfb663c88b2136a2c5cab1c9c74bfd06e45db99181
3
+ size 3785732668
candidate2_short/train_args.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "output_dir": "../drive/MyDrive/NICE/model/0431_short",
3
+ "model_name_or_path": "../drive/MyDrive/NICE/model/cap_best/OFA-huge-caption",
4
+ "train_caption_file": "./data/train_caption_catAdded.jsonl",
5
+ "train_image_file": "../drive/MyDrive/NICE/train_image.tsv",
6
+ "test_caption_file": null,
7
+ "test_image_file": null,
8
+ "freeze_encoder": false,
9
+ "freeze_word_embed": false,
10
+ "num_train_epochs": 10,
11
+ "max_steps": -1,
12
+ "per_device_train_batch_size": 12,
13
+ "per_device_eval_batch_size": 256,
14
+ "learning_rate": 5e-05,
15
+ "max_seq_length": 150,
16
+ "logging_steps": 500,
17
+ "save_steps": 5000,
18
+ "save_total_limit": 1,
19
+ "lr_scheduler_type": "cosine",
20
+ "warmup_steps": 400,
21
+ "warmup_ratio": 0,
22
+ "gradient_accumulation_steps": 1,
23
+ "optim": "adamw_torch",
24
+ "seed": 42,
25
+ "fp16": false,
26
+ "no_cuda": false,
27
+ "dataloader_num_workers": 8,
28
+ "save_strategy": "steps",
29
+ "weight_decay": 0.01,
30
+ "max_grad_norm": 1.0,
31
+ "remove_unused_columns": false
32
+ }
candidate2_short/trainer_state.json ADDED
@@ -0,0 +1,73 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 10.0,
5
+ "global_step": 4170,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 1.2,
12
+ "learning_rate": 4.991324875275794e-05,
13
+ "loss": 2.3265,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 2.4,
18
+ "learning_rate": 4.69397052278032e-05,
19
+ "loss": 0.9988,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 3.6,
24
+ "learning_rate": 4.021214822246739e-05,
25
+ "loss": 0.5942,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 4.8,
30
+ "learning_rate": 3.088170257757369e-05,
31
+ "loss": 0.3763,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 6.0,
36
+ "learning_rate": 2.054486279317156e-05,
37
+ "loss": 0.2525,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 7.19,
42
+ "learning_rate": 1.0970323365940444e-05,
43
+ "loss": 0.1775,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 8.39,
48
+ "learning_rate": 3.7963446874883848e-06,
49
+ "loss": 0.1421,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 9.59,
54
+ "learning_rate": 2.504369903934739e-07,
55
+ "loss": 0.1256,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 10.0,
60
+ "step": 4170,
61
+ "total_flos": 3.505921644059904e+16,
62
+ "train_loss": 0.6037493255212725,
63
+ "train_runtime": 5563.3864,
64
+ "train_samples_per_second": 8.987,
65
+ "train_steps_per_second": 0.75
66
+ }
67
+ ],
68
+ "max_steps": 4170,
69
+ "num_train_epochs": 10,
70
+ "total_flos": 3.505921644059904e+16,
71
+ "trial_name": null,
72
+ "trial_params": null
73
+ }
candidate2_short/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:73b30b79dc6c1da0485ab3520e5a52e9b968d1e9f4f07b283e2b72a70ffbd6a9
3
+ size 3323
candidate2_short/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
candidate3_lastcoin/config.json ADDED
@@ -0,0 +1,53 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "../drive/MyDrive/NICE/model/cap_best/OFA-huge-caption",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "gelu",
5
+ "add_type_embedding": true,
6
+ "architectures": [
7
+ "OFAModelForCaption"
8
+ ],
9
+ "attention_dropout": 0.0,
10
+ "attn_scale_factor": 2.0,
11
+ "bos_token_id": 0,
12
+ "classifier_dropout": 0.0,
13
+ "code_image_size": 128,
14
+ "code_layernorm_embedding": true,
15
+ "d_model": 1280,
16
+ "decoder_attention_heads": 16,
17
+ "decoder_drop_path_rate": 0.0,
18
+ "decoder_ffn_dim": 5120,
19
+ "decoder_layerdrop": 0.0,
20
+ "decoder_layers": 12,
21
+ "decoder_normalize_before": true,
22
+ "decoder_start_token_id": 0,
23
+ "dropout": 0.1,
24
+ "encoder_attention_heads": 16,
25
+ "encoder_drop_path_rate": 0.0,
26
+ "encoder_ffn_dim": 5120,
27
+ "encoder_layerdrop": 0.0,
28
+ "encoder_layers": 24,
29
+ "encoder_normalize_before": true,
30
+ "entangle_position_embedding": false,
31
+ "eos_token_id": 2,
32
+ "forced_eos_token_id": 2,
33
+ "image_bucket_size": 42,
34
+ "init_std": 0.02,
35
+ "is_encoder_decoder": true,
36
+ "layernorm_embedding": true,
37
+ "max_position_embeddings": 1024,
38
+ "model_type": "ofa",
39
+ "normformer": true,
40
+ "num_hidden_layers": 24,
41
+ "pad_token_id": 1,
42
+ "patch_layernorm_embedding": true,
43
+ "resnet_drop_path_rate": 0.0,
44
+ "resnet_model_path": null,
45
+ "resnet_type": "resnet152",
46
+ "scale_embedding": false,
47
+ "share_decoder_input_output_embed": true,
48
+ "token_bucket_size": 256,
49
+ "torch_dtype": "float32",
50
+ "transformers_version": "4.20.0",
51
+ "use_cache": false,
52
+ "vocab_size": 59457
53
+ }
candidate3_lastcoin/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3fb69e2919f84af0cf4fd120e661745718252c064a1c7082853833c76a57c45f
3
+ size 3785732668
candidate3_lastcoin/train_args.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "output_dir": "../drive/MyDrive/NICE/model/0501_lastcoin",
3
+ "model_name_or_path": "../drive/MyDrive/NICE/model/cap_best/OFA-huge-caption",
4
+ "train_caption_file": "./data/train_caption_catAdded_2.jsonl",
5
+ "train_image_file": "../drive/MyDrive/NICE/train_image.tsv",
6
+ "test_caption_file": null,
7
+ "test_image_file": null,
8
+ "freeze_encoder": false,
9
+ "freeze_word_embed": false,
10
+ "num_train_epochs": 20,
11
+ "max_steps": -1,
12
+ "per_device_train_batch_size": 12,
13
+ "per_device_eval_batch_size": 256,
14
+ "learning_rate": 5e-05,
15
+ "max_seq_length": 150,
16
+ "logging_steps": 500,
17
+ "save_steps": 5000,
18
+ "save_total_limit": 1,
19
+ "lr_scheduler_type": "cosine",
20
+ "warmup_steps": 400,
21
+ "warmup_ratio": 0,
22
+ "gradient_accumulation_steps": 1,
23
+ "optim": "adamw_torch",
24
+ "seed": 69,
25
+ "fp16": false,
26
+ "no_cuda": false,
27
+ "dataloader_num_workers": 8,
28
+ "save_strategy": "steps",
29
+ "weight_decay": 0.01,
30
+ "max_grad_norm": 1.0,
31
+ "remove_unused_columns": false
32
+ }
candidate3_lastcoin/trainer_state.json ADDED
@@ -0,0 +1,73 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 10.0,
5
+ "global_step": 4170,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 1.2,
12
+ "learning_rate": 4.991324875275794e-05,
13
+ "loss": 2.3493,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 2.4,
18
+ "learning_rate": 4.69397052278032e-05,
19
+ "loss": 1.0029,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 3.6,
24
+ "learning_rate": 4.021214822246739e-05,
25
+ "loss": 0.6028,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 4.8,
30
+ "learning_rate": 3.088170257757369e-05,
31
+ "loss": 0.3698,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 6.0,
36
+ "learning_rate": 2.054486279317156e-05,
37
+ "loss": 0.2563,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 7.19,
42
+ "learning_rate": 1.0970323365940444e-05,
43
+ "loss": 0.1784,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 8.39,
48
+ "learning_rate": 3.7963446874883848e-06,
49
+ "loss": 0.1485,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 9.59,
54
+ "learning_rate": 2.504369903934739e-07,
55
+ "loss": 0.1288,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 10.0,
60
+ "step": 4170,
61
+ "total_flos": 3.58495435939584e+16,
62
+ "train_loss": 0.6090772413807235,
63
+ "train_runtime": 5626.7954,
64
+ "train_samples_per_second": 8.886,
65
+ "train_steps_per_second": 0.741
66
+ }
67
+ ],
68
+ "max_steps": 4170,
69
+ "num_train_epochs": 10,
70
+ "total_flos": 3.58495435939584e+16,
71
+ "trial_name": null,
72
+ "trial_params": null
73
+ }
candidate3_lastcoin/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6411fc2e50975b7ff683bd365499d303bb30d72c5b46fac83aaf64baff82b43e
3
+ size 3323
candidate3_lastcoin/vocab.json ADDED
The diff for this file is too large to render. See raw diff