Spaces:
Runtime error
Runtime error
Added distilbert checkpoint
Browse files- checkpoint-49092/config.json +43 -0
- checkpoint-49092/optimizer.pt +3 -0
- checkpoint-49092/pytorch_model.bin +3 -0
- checkpoint-49092/rng_state.pth +3 -0
- checkpoint-49092/scheduler.pt +3 -0
- checkpoint-49092/special_tokens_map.json +1 -0
- checkpoint-49092/tokenizer.json +0 -0
- checkpoint-49092/tokenizer_config.json +1 -0
- checkpoint-49092/trainer_state.json +648 -0
- checkpoint-49092/training_args.bin +3 -0
- checkpoint-49092/vocab.txt +0 -0
checkpoint-49092/config.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "distilbert-base-uncased",
|
| 3 |
+
"activation": "gelu",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"DistilBertForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_dropout": 0.1,
|
| 8 |
+
"dim": 768,
|
| 9 |
+
"dropout": 0.1,
|
| 10 |
+
"hidden_dim": 3072,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "Action",
|
| 13 |
+
"1": "Drama",
|
| 14 |
+
"2": "Romance",
|
| 15 |
+
"3": "Comedy",
|
| 16 |
+
"4": "Thriller",
|
| 17 |
+
"5": "Crime",
|
| 18 |
+
"6": "Horror"
|
| 19 |
+
},
|
| 20 |
+
"initializer_range": 0.02,
|
| 21 |
+
"label2id": {
|
| 22 |
+
"Action": 0,
|
| 23 |
+
"Comedy": 3,
|
| 24 |
+
"Crime": 5,
|
| 25 |
+
"Drama": 1,
|
| 26 |
+
"Horror": 6,
|
| 27 |
+
"Romance": 2,
|
| 28 |
+
"Thriller": 4
|
| 29 |
+
},
|
| 30 |
+
"max_position_embeddings": 512,
|
| 31 |
+
"model_type": "distilbert",
|
| 32 |
+
"n_heads": 12,
|
| 33 |
+
"n_layers": 6,
|
| 34 |
+
"pad_token_id": 0,
|
| 35 |
+
"problem_type": "multi_label_classification",
|
| 36 |
+
"qa_dropout": 0.1,
|
| 37 |
+
"seq_classif_dropout": 0.2,
|
| 38 |
+
"sinusoidal_pos_embds": false,
|
| 39 |
+
"tie_weights_": true,
|
| 40 |
+
"torch_dtype": "float32",
|
| 41 |
+
"transformers_version": "4.19.4",
|
| 42 |
+
"vocab_size": 30522
|
| 43 |
+
}
|
checkpoint-49092/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0e18e027dc0def079755e0a6794d675a631ed4bced4b9d6e6d311fc052077722
|
| 3 |
+
size 535734463
|
checkpoint-49092/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:52650ba951e2eee97620a7d1a7ff75a4808ad9f795e87179941300e448493ac6
|
| 3 |
+
size 267871927
|
checkpoint-49092/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4c4e592e745087898cbf63d6b5e46f638c3acda1d07536b96aa3db6acab4d9bb
|
| 3 |
+
size 14649
|
checkpoint-49092/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e324b0cd2d6b0940da219ac40ab939d917236474e9a58aa2c91c71bce00b75d5
|
| 3 |
+
size 623
|
checkpoint-49092/special_tokens_map.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
| 1 |
+
{"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}
|
checkpoint-49092/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
checkpoint-49092/tokenizer_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
| 1 |
+
{"do_lower_case": true, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "distilbert-base-uncased", "tokenizer_class": "DistilBertTokenizer"}
|
checkpoint-49092/trainer_state.json
ADDED
|
@@ -0,0 +1,648 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.5930562407498767,
|
| 3 |
+
"best_model_checkpoint": "bert-finetuned-sem_eval-english/checkpoint-49092",
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
+
"global_step": 49092,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.04,
|
| 12 |
+
"learning_rate": 1.9796300822944678e-05,
|
| 13 |
+
"loss": 0.472,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.08,
|
| 18 |
+
"learning_rate": 1.9592601645889353e-05,
|
| 19 |
+
"loss": 0.4302,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.12,
|
| 24 |
+
"learning_rate": 1.938890246883403e-05,
|
| 25 |
+
"loss": 0.4104,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.16,
|
| 30 |
+
"learning_rate": 1.9185203291778702e-05,
|
| 31 |
+
"loss": 0.4044,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.2,
|
| 36 |
+
"learning_rate": 1.8981504114723378e-05,
|
| 37 |
+
"loss": 0.3954,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.24,
|
| 42 |
+
"learning_rate": 1.8777804937668054e-05,
|
| 43 |
+
"loss": 0.3921,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.29,
|
| 48 |
+
"learning_rate": 1.857410576061273e-05,
|
| 49 |
+
"loss": 0.393,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.33,
|
| 54 |
+
"learning_rate": 1.8370406583557402e-05,
|
| 55 |
+
"loss": 0.3993,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.37,
|
| 60 |
+
"learning_rate": 1.816670740650208e-05,
|
| 61 |
+
"loss": 0.4029,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.41,
|
| 66 |
+
"learning_rate": 1.7963008229446754e-05,
|
| 67 |
+
"loss": 0.4048,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.45,
|
| 72 |
+
"learning_rate": 1.775930905239143e-05,
|
| 73 |
+
"loss": 0.3931,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.49,
|
| 78 |
+
"learning_rate": 1.7555609875336106e-05,
|
| 79 |
+
"loss": 0.3879,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.53,
|
| 84 |
+
"learning_rate": 1.735191069828078e-05,
|
| 85 |
+
"loss": 0.3907,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.57,
|
| 90 |
+
"learning_rate": 1.7148211521225454e-05,
|
| 91 |
+
"loss": 0.3909,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.61,
|
| 96 |
+
"learning_rate": 1.6944512344170133e-05,
|
| 97 |
+
"loss": 0.3854,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.65,
|
| 102 |
+
"learning_rate": 1.6740813167114806e-05,
|
| 103 |
+
"loss": 0.3854,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.69,
|
| 108 |
+
"learning_rate": 1.6537113990059482e-05,
|
| 109 |
+
"loss": 0.3899,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.73,
|
| 114 |
+
"learning_rate": 1.6333414813004154e-05,
|
| 115 |
+
"loss": 0.3855,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.77,
|
| 120 |
+
"learning_rate": 1.6129715635948834e-05,
|
| 121 |
+
"loss": 0.387,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.81,
|
| 126 |
+
"learning_rate": 1.5926016458893506e-05,
|
| 127 |
+
"loss": 0.388,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.86,
|
| 132 |
+
"learning_rate": 1.5722317281838182e-05,
|
| 133 |
+
"loss": 0.3754,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.9,
|
| 138 |
+
"learning_rate": 1.5518618104782858e-05,
|
| 139 |
+
"loss": 0.3844,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.94,
|
| 144 |
+
"learning_rate": 1.5314918927727534e-05,
|
| 145 |
+
"loss": 0.3768,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.98,
|
| 150 |
+
"learning_rate": 1.5111219750672208e-05,
|
| 151 |
+
"loss": 0.3763,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.0,
|
| 156 |
+
"eval_accuracy": 0.3124490793547336,
|
| 157 |
+
"eval_f1": 0.5500132135306554,
|
| 158 |
+
"eval_loss": 0.360363245010376,
|
| 159 |
+
"eval_roc_auc": 0.7014207843050106,
|
| 160 |
+
"eval_runtime": 1041.5255,
|
| 161 |
+
"eval_samples_per_second": 11.785,
|
| 162 |
+
"eval_steps_per_second": 1.474,
|
| 163 |
+
"step": 12273
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 1.02,
|
| 167 |
+
"learning_rate": 1.4907520573616886e-05,
|
| 168 |
+
"loss": 0.376,
|
| 169 |
+
"step": 12500
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 1.06,
|
| 173 |
+
"learning_rate": 1.470382139656156e-05,
|
| 174 |
+
"loss": 0.3613,
|
| 175 |
+
"step": 13000
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.1,
|
| 179 |
+
"learning_rate": 1.4500122219506234e-05,
|
| 180 |
+
"loss": 0.3644,
|
| 181 |
+
"step": 13500
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 1.14,
|
| 185 |
+
"learning_rate": 1.4296423042450908e-05,
|
| 186 |
+
"loss": 0.3535,
|
| 187 |
+
"step": 14000
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 1.18,
|
| 191 |
+
"learning_rate": 1.4092723865395586e-05,
|
| 192 |
+
"loss": 0.3588,
|
| 193 |
+
"step": 14500
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 1.22,
|
| 197 |
+
"learning_rate": 1.388902468834026e-05,
|
| 198 |
+
"loss": 0.3505,
|
| 199 |
+
"step": 15000
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.26,
|
| 203 |
+
"learning_rate": 1.3685325511284934e-05,
|
| 204 |
+
"loss": 0.3488,
|
| 205 |
+
"step": 15500
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 1.3,
|
| 209 |
+
"learning_rate": 1.3481626334229612e-05,
|
| 210 |
+
"loss": 0.3513,
|
| 211 |
+
"step": 16000
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 1.34,
|
| 215 |
+
"learning_rate": 1.3277927157174286e-05,
|
| 216 |
+
"loss": 0.3536,
|
| 217 |
+
"step": 16500
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 1.39,
|
| 221 |
+
"learning_rate": 1.307422798011896e-05,
|
| 222 |
+
"loss": 0.3627,
|
| 223 |
+
"step": 17000
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 1.43,
|
| 227 |
+
"learning_rate": 1.2870528803063638e-05,
|
| 228 |
+
"loss": 0.3439,
|
| 229 |
+
"step": 17500
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 1.47,
|
| 233 |
+
"learning_rate": 1.2666829626008312e-05,
|
| 234 |
+
"loss": 0.3598,
|
| 235 |
+
"step": 18000
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 1.51,
|
| 239 |
+
"learning_rate": 1.2463130448952986e-05,
|
| 240 |
+
"loss": 0.3524,
|
| 241 |
+
"step": 18500
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.55,
|
| 245 |
+
"learning_rate": 1.2259431271897662e-05,
|
| 246 |
+
"loss": 0.3533,
|
| 247 |
+
"step": 19000
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.59,
|
| 251 |
+
"learning_rate": 1.2055732094842338e-05,
|
| 252 |
+
"loss": 0.3422,
|
| 253 |
+
"step": 19500
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 1.63,
|
| 257 |
+
"learning_rate": 1.1852032917787012e-05,
|
| 258 |
+
"loss": 0.3578,
|
| 259 |
+
"step": 20000
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 1.67,
|
| 263 |
+
"learning_rate": 1.1648333740731688e-05,
|
| 264 |
+
"loss": 0.3554,
|
| 265 |
+
"step": 20500
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 1.71,
|
| 269 |
+
"learning_rate": 1.1444634563676364e-05,
|
| 270 |
+
"loss": 0.3448,
|
| 271 |
+
"step": 21000
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 1.75,
|
| 275 |
+
"learning_rate": 1.124093538662104e-05,
|
| 276 |
+
"loss": 0.3539,
|
| 277 |
+
"step": 21500
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.79,
|
| 281 |
+
"learning_rate": 1.1037236209565714e-05,
|
| 282 |
+
"loss": 0.3556,
|
| 283 |
+
"step": 22000
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 1.83,
|
| 287 |
+
"learning_rate": 1.083353703251039e-05,
|
| 288 |
+
"loss": 0.3545,
|
| 289 |
+
"step": 22500
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.87,
|
| 293 |
+
"learning_rate": 1.0629837855455066e-05,
|
| 294 |
+
"loss": 0.3588,
|
| 295 |
+
"step": 23000
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 1.91,
|
| 299 |
+
"learning_rate": 1.042613867839974e-05,
|
| 300 |
+
"loss": 0.3519,
|
| 301 |
+
"step": 23500
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.96,
|
| 305 |
+
"learning_rate": 1.0222439501344414e-05,
|
| 306 |
+
"loss": 0.3533,
|
| 307 |
+
"step": 24000
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 2.0,
|
| 311 |
+
"learning_rate": 1.0018740324289092e-05,
|
| 312 |
+
"loss": 0.349,
|
| 313 |
+
"step": 24500
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 2.0,
|
| 317 |
+
"eval_accuracy": 0.3259736027374939,
|
| 318 |
+
"eval_f1": 0.5744935216000517,
|
| 319 |
+
"eval_loss": 0.3546680212020874,
|
| 320 |
+
"eval_roc_auc": 0.7161739733306818,
|
| 321 |
+
"eval_runtime": 1043.4528,
|
| 322 |
+
"eval_samples_per_second": 11.763,
|
| 323 |
+
"eval_steps_per_second": 1.471,
|
| 324 |
+
"step": 24546
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 2.04,
|
| 328 |
+
"learning_rate": 9.815041147233766e-06,
|
| 329 |
+
"loss": 0.3248,
|
| 330 |
+
"step": 25000
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 2.08,
|
| 334 |
+
"learning_rate": 9.611341970178442e-06,
|
| 335 |
+
"loss": 0.3291,
|
| 336 |
+
"step": 25500
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 2.12,
|
| 340 |
+
"learning_rate": 9.407642793123116e-06,
|
| 341 |
+
"loss": 0.3279,
|
| 342 |
+
"step": 26000
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 2.16,
|
| 346 |
+
"learning_rate": 9.203943616067792e-06,
|
| 347 |
+
"loss": 0.3212,
|
| 348 |
+
"step": 26500
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 2.2,
|
| 352 |
+
"learning_rate": 9.000244439012466e-06,
|
| 353 |
+
"loss": 0.3233,
|
| 354 |
+
"step": 27000
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.24,
|
| 358 |
+
"learning_rate": 8.796545261957142e-06,
|
| 359 |
+
"loss": 0.3171,
|
| 360 |
+
"step": 27500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 2.28,
|
| 364 |
+
"learning_rate": 8.592846084901818e-06,
|
| 365 |
+
"loss": 0.324,
|
| 366 |
+
"step": 28000
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 2.32,
|
| 370 |
+
"learning_rate": 8.389146907846492e-06,
|
| 371 |
+
"loss": 0.3279,
|
| 372 |
+
"step": 28500
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 2.36,
|
| 376 |
+
"learning_rate": 8.185447730791168e-06,
|
| 377 |
+
"loss": 0.3263,
|
| 378 |
+
"step": 29000
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 2.4,
|
| 382 |
+
"learning_rate": 7.981748553735843e-06,
|
| 383 |
+
"loss": 0.3171,
|
| 384 |
+
"step": 29500
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 2.44,
|
| 388 |
+
"learning_rate": 7.778049376680518e-06,
|
| 389 |
+
"loss": 0.3259,
|
| 390 |
+
"step": 30000
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 2.49,
|
| 394 |
+
"learning_rate": 7.574350199625194e-06,
|
| 395 |
+
"loss": 0.321,
|
| 396 |
+
"step": 30500
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 2.53,
|
| 400 |
+
"learning_rate": 7.370651022569869e-06,
|
| 401 |
+
"loss": 0.3155,
|
| 402 |
+
"step": 31000
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 2.57,
|
| 406 |
+
"learning_rate": 7.1669518455145444e-06,
|
| 407 |
+
"loss": 0.3317,
|
| 408 |
+
"step": 31500
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 2.61,
|
| 412 |
+
"learning_rate": 6.96325266845922e-06,
|
| 413 |
+
"loss": 0.3195,
|
| 414 |
+
"step": 32000
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 2.65,
|
| 418 |
+
"learning_rate": 6.759553491403895e-06,
|
| 419 |
+
"loss": 0.3215,
|
| 420 |
+
"step": 32500
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 2.69,
|
| 424 |
+
"learning_rate": 6.555854314348571e-06,
|
| 425 |
+
"loss": 0.3188,
|
| 426 |
+
"step": 33000
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 2.73,
|
| 430 |
+
"learning_rate": 6.3521551372932455e-06,
|
| 431 |
+
"loss": 0.3157,
|
| 432 |
+
"step": 33500
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 2.77,
|
| 436 |
+
"learning_rate": 6.148455960237921e-06,
|
| 437 |
+
"loss": 0.318,
|
| 438 |
+
"step": 34000
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 2.81,
|
| 442 |
+
"learning_rate": 5.944756783182597e-06,
|
| 443 |
+
"loss": 0.325,
|
| 444 |
+
"step": 34500
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 2.85,
|
| 448 |
+
"learning_rate": 5.7410576061272715e-06,
|
| 449 |
+
"loss": 0.3213,
|
| 450 |
+
"step": 35000
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.89,
|
| 454 |
+
"learning_rate": 5.537358429071947e-06,
|
| 455 |
+
"loss": 0.3306,
|
| 456 |
+
"step": 35500
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 2.93,
|
| 460 |
+
"learning_rate": 5.333659252016622e-06,
|
| 461 |
+
"loss": 0.3254,
|
| 462 |
+
"step": 36000
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 2.97,
|
| 466 |
+
"learning_rate": 5.1299600749612975e-06,
|
| 467 |
+
"loss": 0.3179,
|
| 468 |
+
"step": 36500
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 3.0,
|
| 472 |
+
"eval_accuracy": 0.3271956982238879,
|
| 473 |
+
"eval_f1": 0.5859622450272393,
|
| 474 |
+
"eval_loss": 0.3562159240245819,
|
| 475 |
+
"eval_roc_auc": 0.7241606691764836,
|
| 476 |
+
"eval_runtime": 1044.5243,
|
| 477 |
+
"eval_samples_per_second": 11.751,
|
| 478 |
+
"eval_steps_per_second": 1.47,
|
| 479 |
+
"step": 36819
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 3.01,
|
| 483 |
+
"learning_rate": 4.9262608979059726e-06,
|
| 484 |
+
"loss": 0.3184,
|
| 485 |
+
"step": 37000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 3.06,
|
| 489 |
+
"learning_rate": 4.722561720850648e-06,
|
| 490 |
+
"loss": 0.2906,
|
| 491 |
+
"step": 37500
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 3.1,
|
| 495 |
+
"learning_rate": 4.5188625437953235e-06,
|
| 496 |
+
"loss": 0.2902,
|
| 497 |
+
"step": 38000
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 3.14,
|
| 501 |
+
"learning_rate": 4.3151633667399986e-06,
|
| 502 |
+
"loss": 0.3049,
|
| 503 |
+
"step": 38500
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 3.18,
|
| 507 |
+
"learning_rate": 4.1114641896846745e-06,
|
| 508 |
+
"loss": 0.2982,
|
| 509 |
+
"step": 39000
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 3.22,
|
| 513 |
+
"learning_rate": 3.9077650126293495e-06,
|
| 514 |
+
"loss": 0.3011,
|
| 515 |
+
"step": 39500
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 3.26,
|
| 519 |
+
"learning_rate": 3.7040658355740246e-06,
|
| 520 |
+
"loss": 0.2966,
|
| 521 |
+
"step": 40000
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 3.3,
|
| 525 |
+
"learning_rate": 3.5003666585186996e-06,
|
| 526 |
+
"loss": 0.298,
|
| 527 |
+
"step": 40500
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 3.34,
|
| 531 |
+
"learning_rate": 3.2966674814633755e-06,
|
| 532 |
+
"loss": 0.2937,
|
| 533 |
+
"step": 41000
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 3.38,
|
| 537 |
+
"learning_rate": 3.0929683044080506e-06,
|
| 538 |
+
"loss": 0.3008,
|
| 539 |
+
"step": 41500
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 3.42,
|
| 543 |
+
"learning_rate": 2.8892691273527256e-06,
|
| 544 |
+
"loss": 0.3002,
|
| 545 |
+
"step": 42000
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 3.46,
|
| 549 |
+
"learning_rate": 2.6855699502974007e-06,
|
| 550 |
+
"loss": 0.2881,
|
| 551 |
+
"step": 42500
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 3.5,
|
| 555 |
+
"learning_rate": 2.481870773242076e-06,
|
| 556 |
+
"loss": 0.2956,
|
| 557 |
+
"step": 43000
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 3.54,
|
| 561 |
+
"learning_rate": 2.2781715961867516e-06,
|
| 562 |
+
"loss": 0.2945,
|
| 563 |
+
"step": 43500
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 3.59,
|
| 567 |
+
"learning_rate": 2.074472419131427e-06,
|
| 568 |
+
"loss": 0.3016,
|
| 569 |
+
"step": 44000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 3.63,
|
| 573 |
+
"learning_rate": 1.8707732420761022e-06,
|
| 574 |
+
"loss": 0.2977,
|
| 575 |
+
"step": 44500
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 3.67,
|
| 579 |
+
"learning_rate": 1.6670740650207777e-06,
|
| 580 |
+
"loss": 0.2958,
|
| 581 |
+
"step": 45000
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 3.71,
|
| 585 |
+
"learning_rate": 1.4633748879654527e-06,
|
| 586 |
+
"loss": 0.2957,
|
| 587 |
+
"step": 45500
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 3.75,
|
| 591 |
+
"learning_rate": 1.259675710910128e-06,
|
| 592 |
+
"loss": 0.2996,
|
| 593 |
+
"step": 46000
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 3.79,
|
| 597 |
+
"learning_rate": 1.0559765338548034e-06,
|
| 598 |
+
"loss": 0.2937,
|
| 599 |
+
"step": 46500
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 3.83,
|
| 603 |
+
"learning_rate": 8.522773567994785e-07,
|
| 604 |
+
"loss": 0.2956,
|
| 605 |
+
"step": 47000
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 3.87,
|
| 609 |
+
"learning_rate": 6.485781797441539e-07,
|
| 610 |
+
"loss": 0.298,
|
| 611 |
+
"step": 47500
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 3.91,
|
| 615 |
+
"learning_rate": 4.4487900268882914e-07,
|
| 616 |
+
"loss": 0.2876,
|
| 617 |
+
"step": 48000
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 3.95,
|
| 621 |
+
"learning_rate": 2.4117982563350446e-07,
|
| 622 |
+
"loss": 0.3056,
|
| 623 |
+
"step": 48500
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 3.99,
|
| 627 |
+
"learning_rate": 3.748064857817975e-08,
|
| 628 |
+
"loss": 0.2951,
|
| 629 |
+
"step": 49000
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 4.0,
|
| 633 |
+
"eval_accuracy": 0.3271142251914616,
|
| 634 |
+
"eval_f1": 0.5930562407498767,
|
| 635 |
+
"eval_loss": 0.3666437268257141,
|
| 636 |
+
"eval_roc_auc": 0.7306438427437314,
|
| 637 |
+
"eval_runtime": 1044.3283,
|
| 638 |
+
"eval_samples_per_second": 11.753,
|
| 639 |
+
"eval_steps_per_second": 1.47,
|
| 640 |
+
"step": 49092
|
| 641 |
+
}
|
| 642 |
+
],
|
| 643 |
+
"max_steps": 49092,
|
| 644 |
+
"num_train_epochs": 4,
|
| 645 |
+
"total_flos": 2.601467760667853e+16,
|
| 646 |
+
"trial_name": null,
|
| 647 |
+
"trial_params": null
|
| 648 |
+
}
|
checkpoint-49092/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4f92413d1ebf6625cbccb44c4ea982afbdd832ba88b2b8eacfea946f554c2422
|
| 3 |
+
size 3183
|
checkpoint-49092/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|