Training in progress, epoch 1
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- model.safetensors +1 -1
- run-4/checkpoint-550/config.json +54 -0
- run-4/checkpoint-550/model.safetensors +3 -0
- run-4/checkpoint-550/optimizer.pt +3 -0
- run-4/checkpoint-550/rng_state.pth +3 -0
- run-4/checkpoint-550/scheduler.pt +3 -0
- run-4/checkpoint-550/special_tokens_map.json +37 -0
- run-4/checkpoint-550/tokenizer.json +0 -0
- run-4/checkpoint-550/tokenizer_config.json +64 -0
- run-4/checkpoint-550/trainer_state.json +50 -0
- run-4/checkpoint-550/training_args.bin +3 -0
- run-4/checkpoint-550/vocab.txt +0 -0
- run-5/checkpoint-138/config.json +54 -0
- run-5/checkpoint-138/model.safetensors +3 -0
- run-5/checkpoint-138/optimizer.pt +3 -0
- run-5/checkpoint-138/rng_state.pth +3 -0
- run-5/checkpoint-138/scheduler.pt +3 -0
- run-5/checkpoint-138/special_tokens_map.json +37 -0
- run-5/checkpoint-138/tokenizer.json +0 -0
- run-5/checkpoint-138/tokenizer_config.json +64 -0
- run-5/checkpoint-138/trainer_state.json +35 -0
- run-5/checkpoint-138/training_args.bin +3 -0
- run-5/checkpoint-138/vocab.txt +0 -0
- run-5/checkpoint-276/config.json +54 -0
- run-5/checkpoint-276/model.safetensors +3 -0
- run-5/checkpoint-276/optimizer.pt +3 -0
- run-5/checkpoint-276/rng_state.pth +3 -0
- run-5/checkpoint-276/scheduler.pt +3 -0
- run-5/checkpoint-276/special_tokens_map.json +37 -0
- run-5/checkpoint-276/tokenizer.json +0 -0
- run-5/checkpoint-276/tokenizer_config.json +64 -0
- run-5/checkpoint-276/trainer_state.json +44 -0
- run-5/checkpoint-276/training_args.bin +3 -0
- run-5/checkpoint-276/vocab.txt +0 -0
- run-5/checkpoint-414/config.json +54 -0
- run-5/checkpoint-414/model.safetensors +3 -0
- run-5/checkpoint-414/optimizer.pt +3 -0
- run-5/checkpoint-414/rng_state.pth +3 -0
- run-5/checkpoint-414/scheduler.pt +3 -0
- run-5/checkpoint-414/special_tokens_map.json +37 -0
- run-5/checkpoint-414/tokenizer.json +0 -0
- run-5/checkpoint-414/tokenizer_config.json +64 -0
- run-5/checkpoint-414/trainer_state.json +53 -0
- run-5/checkpoint-414/training_args.bin +3 -0
- run-5/checkpoint-414/vocab.txt +0 -0
- run-5/checkpoint-552/config.json +54 -0
- run-5/checkpoint-552/model.safetensors +3 -0
- run-5/checkpoint-552/optimizer.pt +3 -0
- run-5/checkpoint-552/rng_state.pth +3 -0
- run-5/checkpoint-552/scheduler.pt +3 -0
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 116795404
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ae4329c58459811b44a9a227b6d43e3a829df3ddedbbd8643370a9863ed33507
|
3 |
size 116795404
|
run-4/checkpoint-550/config.json
ADDED
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "cointegrated/rubert-tiny2",
|
3 |
+
"architectures": [
|
4 |
+
"BertForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"emb_size": 312,
|
9 |
+
"gradient_checkpointing": false,
|
10 |
+
"hidden_act": "gelu",
|
11 |
+
"hidden_dropout_prob": 0.1,
|
12 |
+
"hidden_size": 312,
|
13 |
+
"id2label": {
|
14 |
+
"0": "LABEL_0",
|
15 |
+
"1": "LABEL_1",
|
16 |
+
"2": "LABEL_2",
|
17 |
+
"3": "LABEL_3",
|
18 |
+
"4": "LABEL_4",
|
19 |
+
"5": "LABEL_5",
|
20 |
+
"6": "LABEL_6",
|
21 |
+
"7": "LABEL_7",
|
22 |
+
"8": "LABEL_8",
|
23 |
+
"9": "LABEL_9",
|
24 |
+
"10": "LABEL_10"
|
25 |
+
},
|
26 |
+
"initializer_range": 0.02,
|
27 |
+
"intermediate_size": 600,
|
28 |
+
"label2id": {
|
29 |
+
"LABEL_0": 0,
|
30 |
+
"LABEL_1": 1,
|
31 |
+
"LABEL_10": 10,
|
32 |
+
"LABEL_2": 2,
|
33 |
+
"LABEL_3": 3,
|
34 |
+
"LABEL_4": 4,
|
35 |
+
"LABEL_5": 5,
|
36 |
+
"LABEL_6": 6,
|
37 |
+
"LABEL_7": 7,
|
38 |
+
"LABEL_8": 8,
|
39 |
+
"LABEL_9": 9
|
40 |
+
},
|
41 |
+
"layer_norm_eps": 1e-12,
|
42 |
+
"max_position_embeddings": 2048,
|
43 |
+
"model_type": "bert",
|
44 |
+
"num_attention_heads": 12,
|
45 |
+
"num_hidden_layers": 3,
|
46 |
+
"pad_token_id": 0,
|
47 |
+
"position_embedding_type": "absolute",
|
48 |
+
"problem_type": "single_label_classification",
|
49 |
+
"torch_dtype": "float32",
|
50 |
+
"transformers_version": "4.36.2",
|
51 |
+
"type_vocab_size": 2,
|
52 |
+
"use_cache": true,
|
53 |
+
"vocab_size": 83828
|
54 |
+
}
|
run-4/checkpoint-550/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f110a6c10b21ce91fb3766d4f1e4f41d5e47ecc1808dee0c572974da54669084
|
3 |
+
size 116795404
|
run-4/checkpoint-550/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e9b838c5ec4ba537f3556b5ad8a7b921e8e6f90255fb0bb8f99319e6442c1f1c
|
3 |
+
size 233621882
|
run-4/checkpoint-550/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:232552168fdef8353d2cc61956beebf7a66a42df694b6ee5f5152ad3e1fac9a8
|
3 |
+
size 14054
|
run-4/checkpoint-550/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:617d3a6963d8e8b5255276b673fdbbc3814a44a8d14514d3a8588764d3db82af
|
3 |
+
size 1064
|
run-4/checkpoint-550/special_tokens_map.json
ADDED
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": {
|
3 |
+
"content": "[CLS]",
|
4 |
+
"lstrip": false,
|
5 |
+
"normalized": false,
|
6 |
+
"rstrip": false,
|
7 |
+
"single_word": false
|
8 |
+
},
|
9 |
+
"mask_token": {
|
10 |
+
"content": "[MASK]",
|
11 |
+
"lstrip": false,
|
12 |
+
"normalized": false,
|
13 |
+
"rstrip": false,
|
14 |
+
"single_word": false
|
15 |
+
},
|
16 |
+
"pad_token": {
|
17 |
+
"content": "[PAD]",
|
18 |
+
"lstrip": false,
|
19 |
+
"normalized": false,
|
20 |
+
"rstrip": false,
|
21 |
+
"single_word": false
|
22 |
+
},
|
23 |
+
"sep_token": {
|
24 |
+
"content": "[SEP]",
|
25 |
+
"lstrip": false,
|
26 |
+
"normalized": false,
|
27 |
+
"rstrip": false,
|
28 |
+
"single_word": false
|
29 |
+
},
|
30 |
+
"unk_token": {
|
31 |
+
"content": "[UNK]",
|
32 |
+
"lstrip": false,
|
33 |
+
"normalized": false,
|
34 |
+
"rstrip": false,
|
35 |
+
"single_word": false
|
36 |
+
}
|
37 |
+
}
|
run-4/checkpoint-550/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
run-4/checkpoint-550/tokenizer_config.json
ADDED
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "[PAD]",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"1": {
|
12 |
+
"content": "[UNK]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"2": {
|
20 |
+
"content": "[CLS]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"3": {
|
28 |
+
"content": "[SEP]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"4": {
|
36 |
+
"content": "[MASK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"clean_up_tokenization_spaces": true,
|
45 |
+
"cls_token": "[CLS]",
|
46 |
+
"do_basic_tokenize": true,
|
47 |
+
"do_lower_case": false,
|
48 |
+
"mask_token": "[MASK]",
|
49 |
+
"max_length": 512,
|
50 |
+
"model_max_length": 2048,
|
51 |
+
"never_split": null,
|
52 |
+
"pad_to_multiple_of": null,
|
53 |
+
"pad_token": "[PAD]",
|
54 |
+
"pad_token_type_id": 0,
|
55 |
+
"padding_side": "right",
|
56 |
+
"sep_token": "[SEP]",
|
57 |
+
"stride": 0,
|
58 |
+
"strip_accents": null,
|
59 |
+
"tokenize_chinese_chars": true,
|
60 |
+
"tokenizer_class": "BertTokenizer",
|
61 |
+
"truncation_side": "right",
|
62 |
+
"truncation_strategy": "longest_first",
|
63 |
+
"unk_token": "[UNK]"
|
64 |
+
}
|
run-4/checkpoint-550/trainer_state.json
ADDED
@@ -0,0 +1,50 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.8448979591836735,
|
3 |
+
"best_model_checkpoint": "rubert-tiny2-finetuned-classification\\run-4\\checkpoint-550",
|
4 |
+
"epoch": 2.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 550,
|
7 |
+
"is_hyper_param_search": true,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 1.0,
|
13 |
+
"eval_accuracy": 0.7183673469387755,
|
14 |
+
"eval_loss": 1.1960268020629883,
|
15 |
+
"eval_runtime": 1.24,
|
16 |
+
"eval_samples_per_second": 197.581,
|
17 |
+
"eval_steps_per_second": 12.903,
|
18 |
+
"step": 275
|
19 |
+
},
|
20 |
+
{
|
21 |
+
"epoch": 1.82,
|
22 |
+
"learning_rate": 3.000048616821857e-06,
|
23 |
+
"loss": 1.4724,
|
24 |
+
"step": 500
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"epoch": 2.0,
|
28 |
+
"eval_accuracy": 0.8448979591836735,
|
29 |
+
"eval_loss": 0.9476840496063232,
|
30 |
+
"eval_runtime": 2.417,
|
31 |
+
"eval_samples_per_second": 101.365,
|
32 |
+
"eval_steps_per_second": 6.62,
|
33 |
+
"step": 550
|
34 |
+
}
|
35 |
+
],
|
36 |
+
"logging_steps": 500,
|
37 |
+
"max_steps": 550,
|
38 |
+
"num_input_tokens_seen": 0,
|
39 |
+
"num_train_epochs": 2,
|
40 |
+
"save_steps": 500,
|
41 |
+
"total_flos": 2461491933600.0,
|
42 |
+
"train_batch_size": 8,
|
43 |
+
"trial_name": null,
|
44 |
+
"trial_params": {
|
45 |
+
"learning_rate": 3.3000534785040425e-05,
|
46 |
+
"num_train_epochs": 2,
|
47 |
+
"per_device_train_batch_size": 8,
|
48 |
+
"seed": 26
|
49 |
+
}
|
50 |
+
}
|
run-4/checkpoint-550/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f2208f174fb9536d0ea5aa6aab3c7daab8b14170de6a8d88fd72a8d35b061ea0
|
3 |
+
size 4728
|
run-4/checkpoint-550/vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|
run-5/checkpoint-138/config.json
ADDED
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "cointegrated/rubert-tiny2",
|
3 |
+
"architectures": [
|
4 |
+
"BertForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"emb_size": 312,
|
9 |
+
"gradient_checkpointing": false,
|
10 |
+
"hidden_act": "gelu",
|
11 |
+
"hidden_dropout_prob": 0.1,
|
12 |
+
"hidden_size": 312,
|
13 |
+
"id2label": {
|
14 |
+
"0": "LABEL_0",
|
15 |
+
"1": "LABEL_1",
|
16 |
+
"2": "LABEL_2",
|
17 |
+
"3": "LABEL_3",
|
18 |
+
"4": "LABEL_4",
|
19 |
+
"5": "LABEL_5",
|
20 |
+
"6": "LABEL_6",
|
21 |
+
"7": "LABEL_7",
|
22 |
+
"8": "LABEL_8",
|
23 |
+
"9": "LABEL_9",
|
24 |
+
"10": "LABEL_10"
|
25 |
+
},
|
26 |
+
"initializer_range": 0.02,
|
27 |
+
"intermediate_size": 600,
|
28 |
+
"label2id": {
|
29 |
+
"LABEL_0": 0,
|
30 |
+
"LABEL_1": 1,
|
31 |
+
"LABEL_10": 10,
|
32 |
+
"LABEL_2": 2,
|
33 |
+
"LABEL_3": 3,
|
34 |
+
"LABEL_4": 4,
|
35 |
+
"LABEL_5": 5,
|
36 |
+
"LABEL_6": 6,
|
37 |
+
"LABEL_7": 7,
|
38 |
+
"LABEL_8": 8,
|
39 |
+
"LABEL_9": 9
|
40 |
+
},
|
41 |
+
"layer_norm_eps": 1e-12,
|
42 |
+
"max_position_embeddings": 2048,
|
43 |
+
"model_type": "bert",
|
44 |
+
"num_attention_heads": 12,
|
45 |
+
"num_hidden_layers": 3,
|
46 |
+
"pad_token_id": 0,
|
47 |
+
"position_embedding_type": "absolute",
|
48 |
+
"problem_type": "single_label_classification",
|
49 |
+
"torch_dtype": "float32",
|
50 |
+
"transformers_version": "4.36.2",
|
51 |
+
"type_vocab_size": 2,
|
52 |
+
"use_cache": true,
|
53 |
+
"vocab_size": 83828
|
54 |
+
}
|
run-5/checkpoint-138/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fc6e9147b3404d987fa629b8aca60ca180b65b26c074dc17657e02d66ee33b5b
|
3 |
+
size 116795404
|
run-5/checkpoint-138/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a6ad430a29ab1f2682312e287630c7d9df3fd8834af583f7c3594d73d1d1c479
|
3 |
+
size 233621882
|
run-5/checkpoint-138/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a0fabfcdc4965cb73474d6ce1ffdbaf797ca96e698284098edc68418398e7f54
|
3 |
+
size 13990
|
run-5/checkpoint-138/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:dc55130e4fccf18638e33dd8409950e68fd6d3ee4e2325c1db5331879cf5f887
|
3 |
+
size 1064
|
run-5/checkpoint-138/special_tokens_map.json
ADDED
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": {
|
3 |
+
"content": "[CLS]",
|
4 |
+
"lstrip": false,
|
5 |
+
"normalized": false,
|
6 |
+
"rstrip": false,
|
7 |
+
"single_word": false
|
8 |
+
},
|
9 |
+
"mask_token": {
|
10 |
+
"content": "[MASK]",
|
11 |
+
"lstrip": false,
|
12 |
+
"normalized": false,
|
13 |
+
"rstrip": false,
|
14 |
+
"single_word": false
|
15 |
+
},
|
16 |
+
"pad_token": {
|
17 |
+
"content": "[PAD]",
|
18 |
+
"lstrip": false,
|
19 |
+
"normalized": false,
|
20 |
+
"rstrip": false,
|
21 |
+
"single_word": false
|
22 |
+
},
|
23 |
+
"sep_token": {
|
24 |
+
"content": "[SEP]",
|
25 |
+
"lstrip": false,
|
26 |
+
"normalized": false,
|
27 |
+
"rstrip": false,
|
28 |
+
"single_word": false
|
29 |
+
},
|
30 |
+
"unk_token": {
|
31 |
+
"content": "[UNK]",
|
32 |
+
"lstrip": false,
|
33 |
+
"normalized": false,
|
34 |
+
"rstrip": false,
|
35 |
+
"single_word": false
|
36 |
+
}
|
37 |
+
}
|
run-5/checkpoint-138/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
run-5/checkpoint-138/tokenizer_config.json
ADDED
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "[PAD]",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"1": {
|
12 |
+
"content": "[UNK]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"2": {
|
20 |
+
"content": "[CLS]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"3": {
|
28 |
+
"content": "[SEP]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"4": {
|
36 |
+
"content": "[MASK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"clean_up_tokenization_spaces": true,
|
45 |
+
"cls_token": "[CLS]",
|
46 |
+
"do_basic_tokenize": true,
|
47 |
+
"do_lower_case": false,
|
48 |
+
"mask_token": "[MASK]",
|
49 |
+
"max_length": 512,
|
50 |
+
"model_max_length": 2048,
|
51 |
+
"never_split": null,
|
52 |
+
"pad_to_multiple_of": null,
|
53 |
+
"pad_token": "[PAD]",
|
54 |
+
"pad_token_type_id": 0,
|
55 |
+
"padding_side": "right",
|
56 |
+
"sep_token": "[SEP]",
|
57 |
+
"stride": 0,
|
58 |
+
"strip_accents": null,
|
59 |
+
"tokenize_chinese_chars": true,
|
60 |
+
"tokenizer_class": "BertTokenizer",
|
61 |
+
"truncation_side": "right",
|
62 |
+
"truncation_strategy": "longest_first",
|
63 |
+
"unk_token": "[UNK]"
|
64 |
+
}
|
run-5/checkpoint-138/trainer_state.json
ADDED
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.8612244897959184,
|
3 |
+
"best_model_checkpoint": "rubert-tiny2-finetuned-classification\\run-5\\checkpoint-138",
|
4 |
+
"epoch": 1.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 138,
|
7 |
+
"is_hyper_param_search": true,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 1.0,
|
13 |
+
"eval_accuracy": 0.8612244897959184,
|
14 |
+
"eval_loss": 0.6846103668212891,
|
15 |
+
"eval_runtime": 1.159,
|
16 |
+
"eval_samples_per_second": 211.389,
|
17 |
+
"eval_steps_per_second": 13.805,
|
18 |
+
"step": 138
|
19 |
+
}
|
20 |
+
],
|
21 |
+
"logging_steps": 500,
|
22 |
+
"max_steps": 552,
|
23 |
+
"num_input_tokens_seen": 0,
|
24 |
+
"num_train_epochs": 4,
|
25 |
+
"save_steps": 500,
|
26 |
+
"total_flos": 0,
|
27 |
+
"train_batch_size": 16,
|
28 |
+
"trial_name": null,
|
29 |
+
"trial_params": {
|
30 |
+
"learning_rate": 9.67338621768383e-05,
|
31 |
+
"num_train_epochs": 4,
|
32 |
+
"per_device_train_batch_size": 16,
|
33 |
+
"seed": 1
|
34 |
+
}
|
35 |
+
}
|
run-5/checkpoint-138/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:824c486ca80c908e88fb61bcd59fcffc329cebb90080363918d9dccfafd4e1a8
|
3 |
+
size 4728
|
run-5/checkpoint-138/vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|
run-5/checkpoint-276/config.json
ADDED
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "cointegrated/rubert-tiny2",
|
3 |
+
"architectures": [
|
4 |
+
"BertForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"emb_size": 312,
|
9 |
+
"gradient_checkpointing": false,
|
10 |
+
"hidden_act": "gelu",
|
11 |
+
"hidden_dropout_prob": 0.1,
|
12 |
+
"hidden_size": 312,
|
13 |
+
"id2label": {
|
14 |
+
"0": "LABEL_0",
|
15 |
+
"1": "LABEL_1",
|
16 |
+
"2": "LABEL_2",
|
17 |
+
"3": "LABEL_3",
|
18 |
+
"4": "LABEL_4",
|
19 |
+
"5": "LABEL_5",
|
20 |
+
"6": "LABEL_6",
|
21 |
+
"7": "LABEL_7",
|
22 |
+
"8": "LABEL_8",
|
23 |
+
"9": "LABEL_9",
|
24 |
+
"10": "LABEL_10"
|
25 |
+
},
|
26 |
+
"initializer_range": 0.02,
|
27 |
+
"intermediate_size": 600,
|
28 |
+
"label2id": {
|
29 |
+
"LABEL_0": 0,
|
30 |
+
"LABEL_1": 1,
|
31 |
+
"LABEL_10": 10,
|
32 |
+
"LABEL_2": 2,
|
33 |
+
"LABEL_3": 3,
|
34 |
+
"LABEL_4": 4,
|
35 |
+
"LABEL_5": 5,
|
36 |
+
"LABEL_6": 6,
|
37 |
+
"LABEL_7": 7,
|
38 |
+
"LABEL_8": 8,
|
39 |
+
"LABEL_9": 9
|
40 |
+
},
|
41 |
+
"layer_norm_eps": 1e-12,
|
42 |
+
"max_position_embeddings": 2048,
|
43 |
+
"model_type": "bert",
|
44 |
+
"num_attention_heads": 12,
|
45 |
+
"num_hidden_layers": 3,
|
46 |
+
"pad_token_id": 0,
|
47 |
+
"position_embedding_type": "absolute",
|
48 |
+
"problem_type": "single_label_classification",
|
49 |
+
"torch_dtype": "float32",
|
50 |
+
"transformers_version": "4.36.2",
|
51 |
+
"type_vocab_size": 2,
|
52 |
+
"use_cache": true,
|
53 |
+
"vocab_size": 83828
|
54 |
+
}
|
run-5/checkpoint-276/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c8ba45611a4933866b788d99193f631d33b54063c6c7226de07af1bf2556fbad
|
3 |
+
size 116795404
|
run-5/checkpoint-276/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5c824daf70d5ce2e5cb0dcd3af04c4d3717b0ffa543eeef9f24d4e96e354fff9
|
3 |
+
size 233621882
|
run-5/checkpoint-276/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6ad16b4fcd85936d4e51d3496b96350c749191cd1673cbad6e1111cad4be674c
|
3 |
+
size 13990
|
run-5/checkpoint-276/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:40d6268d92ce968a350ddc735bfb65dd9d1801d3b116a5c74e00b5d8fbcd608a
|
3 |
+
size 1064
|
run-5/checkpoint-276/special_tokens_map.json
ADDED
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": {
|
3 |
+
"content": "[CLS]",
|
4 |
+
"lstrip": false,
|
5 |
+
"normalized": false,
|
6 |
+
"rstrip": false,
|
7 |
+
"single_word": false
|
8 |
+
},
|
9 |
+
"mask_token": {
|
10 |
+
"content": "[MASK]",
|
11 |
+
"lstrip": false,
|
12 |
+
"normalized": false,
|
13 |
+
"rstrip": false,
|
14 |
+
"single_word": false
|
15 |
+
},
|
16 |
+
"pad_token": {
|
17 |
+
"content": "[PAD]",
|
18 |
+
"lstrip": false,
|
19 |
+
"normalized": false,
|
20 |
+
"rstrip": false,
|
21 |
+
"single_word": false
|
22 |
+
},
|
23 |
+
"sep_token": {
|
24 |
+
"content": "[SEP]",
|
25 |
+
"lstrip": false,
|
26 |
+
"normalized": false,
|
27 |
+
"rstrip": false,
|
28 |
+
"single_word": false
|
29 |
+
},
|
30 |
+
"unk_token": {
|
31 |
+
"content": "[UNK]",
|
32 |
+
"lstrip": false,
|
33 |
+
"normalized": false,
|
34 |
+
"rstrip": false,
|
35 |
+
"single_word": false
|
36 |
+
}
|
37 |
+
}
|
run-5/checkpoint-276/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
run-5/checkpoint-276/tokenizer_config.json
ADDED
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "[PAD]",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"1": {
|
12 |
+
"content": "[UNK]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"2": {
|
20 |
+
"content": "[CLS]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"3": {
|
28 |
+
"content": "[SEP]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"4": {
|
36 |
+
"content": "[MASK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"clean_up_tokenization_spaces": true,
|
45 |
+
"cls_token": "[CLS]",
|
46 |
+
"do_basic_tokenize": true,
|
47 |
+
"do_lower_case": false,
|
48 |
+
"mask_token": "[MASK]",
|
49 |
+
"max_length": 512,
|
50 |
+
"model_max_length": 2048,
|
51 |
+
"never_split": null,
|
52 |
+
"pad_to_multiple_of": null,
|
53 |
+
"pad_token": "[PAD]",
|
54 |
+
"pad_token_type_id": 0,
|
55 |
+
"padding_side": "right",
|
56 |
+
"sep_token": "[SEP]",
|
57 |
+
"stride": 0,
|
58 |
+
"strip_accents": null,
|
59 |
+
"tokenize_chinese_chars": true,
|
60 |
+
"tokenizer_class": "BertTokenizer",
|
61 |
+
"truncation_side": "right",
|
62 |
+
"truncation_strategy": "longest_first",
|
63 |
+
"unk_token": "[UNK]"
|
64 |
+
}
|
run-5/checkpoint-276/trainer_state.json
ADDED
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.8979591836734694,
|
3 |
+
"best_model_checkpoint": "rubert-tiny2-finetuned-classification\\run-5\\checkpoint-276",
|
4 |
+
"epoch": 2.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 276,
|
7 |
+
"is_hyper_param_search": true,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 1.0,
|
13 |
+
"eval_accuracy": 0.8612244897959184,
|
14 |
+
"eval_loss": 0.6846103668212891,
|
15 |
+
"eval_runtime": 1.159,
|
16 |
+
"eval_samples_per_second": 211.389,
|
17 |
+
"eval_steps_per_second": 13.805,
|
18 |
+
"step": 138
|
19 |
+
},
|
20 |
+
{
|
21 |
+
"epoch": 2.0,
|
22 |
+
"eval_accuracy": 0.8979591836734694,
|
23 |
+
"eval_loss": 0.4278050363063812,
|
24 |
+
"eval_runtime": 1.189,
|
25 |
+
"eval_samples_per_second": 206.056,
|
26 |
+
"eval_steps_per_second": 13.457,
|
27 |
+
"step": 276
|
28 |
+
}
|
29 |
+
],
|
30 |
+
"logging_steps": 500,
|
31 |
+
"max_steps": 552,
|
32 |
+
"num_input_tokens_seen": 0,
|
33 |
+
"num_train_epochs": 4,
|
34 |
+
"save_steps": 500,
|
35 |
+
"total_flos": 0,
|
36 |
+
"train_batch_size": 16,
|
37 |
+
"trial_name": null,
|
38 |
+
"trial_params": {
|
39 |
+
"learning_rate": 9.67338621768383e-05,
|
40 |
+
"num_train_epochs": 4,
|
41 |
+
"per_device_train_batch_size": 16,
|
42 |
+
"seed": 1
|
43 |
+
}
|
44 |
+
}
|
run-5/checkpoint-276/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:824c486ca80c908e88fb61bcd59fcffc329cebb90080363918d9dccfafd4e1a8
|
3 |
+
size 4728
|
run-5/checkpoint-276/vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|
run-5/checkpoint-414/config.json
ADDED
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "cointegrated/rubert-tiny2",
|
3 |
+
"architectures": [
|
4 |
+
"BertForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"emb_size": 312,
|
9 |
+
"gradient_checkpointing": false,
|
10 |
+
"hidden_act": "gelu",
|
11 |
+
"hidden_dropout_prob": 0.1,
|
12 |
+
"hidden_size": 312,
|
13 |
+
"id2label": {
|
14 |
+
"0": "LABEL_0",
|
15 |
+
"1": "LABEL_1",
|
16 |
+
"2": "LABEL_2",
|
17 |
+
"3": "LABEL_3",
|
18 |
+
"4": "LABEL_4",
|
19 |
+
"5": "LABEL_5",
|
20 |
+
"6": "LABEL_6",
|
21 |
+
"7": "LABEL_7",
|
22 |
+
"8": "LABEL_8",
|
23 |
+
"9": "LABEL_9",
|
24 |
+
"10": "LABEL_10"
|
25 |
+
},
|
26 |
+
"initializer_range": 0.02,
|
27 |
+
"intermediate_size": 600,
|
28 |
+
"label2id": {
|
29 |
+
"LABEL_0": 0,
|
30 |
+
"LABEL_1": 1,
|
31 |
+
"LABEL_10": 10,
|
32 |
+
"LABEL_2": 2,
|
33 |
+
"LABEL_3": 3,
|
34 |
+
"LABEL_4": 4,
|
35 |
+
"LABEL_5": 5,
|
36 |
+
"LABEL_6": 6,
|
37 |
+
"LABEL_7": 7,
|
38 |
+
"LABEL_8": 8,
|
39 |
+
"LABEL_9": 9
|
40 |
+
},
|
41 |
+
"layer_norm_eps": 1e-12,
|
42 |
+
"max_position_embeddings": 2048,
|
43 |
+
"model_type": "bert",
|
44 |
+
"num_attention_heads": 12,
|
45 |
+
"num_hidden_layers": 3,
|
46 |
+
"pad_token_id": 0,
|
47 |
+
"position_embedding_type": "absolute",
|
48 |
+
"problem_type": "single_label_classification",
|
49 |
+
"torch_dtype": "float32",
|
50 |
+
"transformers_version": "4.36.2",
|
51 |
+
"type_vocab_size": 2,
|
52 |
+
"use_cache": true,
|
53 |
+
"vocab_size": 83828
|
54 |
+
}
|
run-5/checkpoint-414/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c1e2569c0e34103c0b48d25b314d20e24f611f610c4ab8d7c242f615bc8fb01e
|
3 |
+
size 116795404
|
run-5/checkpoint-414/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e3f2ec21438687c83d055dcdeef0ab33a9bbe53a7331c3ec8bcc58f5261188e5
|
3 |
+
size 233621882
|
run-5/checkpoint-414/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:dd004404062d796af66351de090915b0b7da25e13d45c895d83537b5d6a19459
|
3 |
+
size 13990
|
run-5/checkpoint-414/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ba8ccb9abb713d354337bb7407dd9f5fb864c88a2270456931e9bb2d7effe2dd
|
3 |
+
size 1064
|
run-5/checkpoint-414/special_tokens_map.json
ADDED
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": {
|
3 |
+
"content": "[CLS]",
|
4 |
+
"lstrip": false,
|
5 |
+
"normalized": false,
|
6 |
+
"rstrip": false,
|
7 |
+
"single_word": false
|
8 |
+
},
|
9 |
+
"mask_token": {
|
10 |
+
"content": "[MASK]",
|
11 |
+
"lstrip": false,
|
12 |
+
"normalized": false,
|
13 |
+
"rstrip": false,
|
14 |
+
"single_word": false
|
15 |
+
},
|
16 |
+
"pad_token": {
|
17 |
+
"content": "[PAD]",
|
18 |
+
"lstrip": false,
|
19 |
+
"normalized": false,
|
20 |
+
"rstrip": false,
|
21 |
+
"single_word": false
|
22 |
+
},
|
23 |
+
"sep_token": {
|
24 |
+
"content": "[SEP]",
|
25 |
+
"lstrip": false,
|
26 |
+
"normalized": false,
|
27 |
+
"rstrip": false,
|
28 |
+
"single_word": false
|
29 |
+
},
|
30 |
+
"unk_token": {
|
31 |
+
"content": "[UNK]",
|
32 |
+
"lstrip": false,
|
33 |
+
"normalized": false,
|
34 |
+
"rstrip": false,
|
35 |
+
"single_word": false
|
36 |
+
}
|
37 |
+
}
|
run-5/checkpoint-414/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
run-5/checkpoint-414/tokenizer_config.json
ADDED
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "[PAD]",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"1": {
|
12 |
+
"content": "[UNK]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"2": {
|
20 |
+
"content": "[CLS]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"3": {
|
28 |
+
"content": "[SEP]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"4": {
|
36 |
+
"content": "[MASK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"clean_up_tokenization_spaces": true,
|
45 |
+
"cls_token": "[CLS]",
|
46 |
+
"do_basic_tokenize": true,
|
47 |
+
"do_lower_case": false,
|
48 |
+
"mask_token": "[MASK]",
|
49 |
+
"max_length": 512,
|
50 |
+
"model_max_length": 2048,
|
51 |
+
"never_split": null,
|
52 |
+
"pad_to_multiple_of": null,
|
53 |
+
"pad_token": "[PAD]",
|
54 |
+
"pad_token_type_id": 0,
|
55 |
+
"padding_side": "right",
|
56 |
+
"sep_token": "[SEP]",
|
57 |
+
"stride": 0,
|
58 |
+
"strip_accents": null,
|
59 |
+
"tokenize_chinese_chars": true,
|
60 |
+
"tokenizer_class": "BertTokenizer",
|
61 |
+
"truncation_side": "right",
|
62 |
+
"truncation_strategy": "longest_first",
|
63 |
+
"unk_token": "[UNK]"
|
64 |
+
}
|
run-5/checkpoint-414/trainer_state.json
ADDED
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.9102040816326531,
|
3 |
+
"best_model_checkpoint": "rubert-tiny2-finetuned-classification\\run-5\\checkpoint-414",
|
4 |
+
"epoch": 3.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 414,
|
7 |
+
"is_hyper_param_search": true,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 1.0,
|
13 |
+
"eval_accuracy": 0.8612244897959184,
|
14 |
+
"eval_loss": 0.6846103668212891,
|
15 |
+
"eval_runtime": 1.159,
|
16 |
+
"eval_samples_per_second": 211.389,
|
17 |
+
"eval_steps_per_second": 13.805,
|
18 |
+
"step": 138
|
19 |
+
},
|
20 |
+
{
|
21 |
+
"epoch": 2.0,
|
22 |
+
"eval_accuracy": 0.8979591836734694,
|
23 |
+
"eval_loss": 0.4278050363063812,
|
24 |
+
"eval_runtime": 1.189,
|
25 |
+
"eval_samples_per_second": 206.056,
|
26 |
+
"eval_steps_per_second": 13.457,
|
27 |
+
"step": 276
|
28 |
+
},
|
29 |
+
{
|
30 |
+
"epoch": 3.0,
|
31 |
+
"eval_accuracy": 0.9102040816326531,
|
32 |
+
"eval_loss": 0.36468449234962463,
|
33 |
+
"eval_runtime": 1.2525,
|
34 |
+
"eval_samples_per_second": 195.604,
|
35 |
+
"eval_steps_per_second": 12.774,
|
36 |
+
"step": 414
|
37 |
+
}
|
38 |
+
],
|
39 |
+
"logging_steps": 500,
|
40 |
+
"max_steps": 552,
|
41 |
+
"num_input_tokens_seen": 0,
|
42 |
+
"num_train_epochs": 4,
|
43 |
+
"save_steps": 500,
|
44 |
+
"total_flos": 0,
|
45 |
+
"train_batch_size": 16,
|
46 |
+
"trial_name": null,
|
47 |
+
"trial_params": {
|
48 |
+
"learning_rate": 9.67338621768383e-05,
|
49 |
+
"num_train_epochs": 4,
|
50 |
+
"per_device_train_batch_size": 16,
|
51 |
+
"seed": 1
|
52 |
+
}
|
53 |
+
}
|
run-5/checkpoint-414/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:824c486ca80c908e88fb61bcd59fcffc329cebb90080363918d9dccfafd4e1a8
|
3 |
+
size 4728
|
run-5/checkpoint-414/vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|
run-5/checkpoint-552/config.json
ADDED
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "cointegrated/rubert-tiny2",
|
3 |
+
"architectures": [
|
4 |
+
"BertForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"emb_size": 312,
|
9 |
+
"gradient_checkpointing": false,
|
10 |
+
"hidden_act": "gelu",
|
11 |
+
"hidden_dropout_prob": 0.1,
|
12 |
+
"hidden_size": 312,
|
13 |
+
"id2label": {
|
14 |
+
"0": "LABEL_0",
|
15 |
+
"1": "LABEL_1",
|
16 |
+
"2": "LABEL_2",
|
17 |
+
"3": "LABEL_3",
|
18 |
+
"4": "LABEL_4",
|
19 |
+
"5": "LABEL_5",
|
20 |
+
"6": "LABEL_6",
|
21 |
+
"7": "LABEL_7",
|
22 |
+
"8": "LABEL_8",
|
23 |
+
"9": "LABEL_9",
|
24 |
+
"10": "LABEL_10"
|
25 |
+
},
|
26 |
+
"initializer_range": 0.02,
|
27 |
+
"intermediate_size": 600,
|
28 |
+
"label2id": {
|
29 |
+
"LABEL_0": 0,
|
30 |
+
"LABEL_1": 1,
|
31 |
+
"LABEL_10": 10,
|
32 |
+
"LABEL_2": 2,
|
33 |
+
"LABEL_3": 3,
|
34 |
+
"LABEL_4": 4,
|
35 |
+
"LABEL_5": 5,
|
36 |
+
"LABEL_6": 6,
|
37 |
+
"LABEL_7": 7,
|
38 |
+
"LABEL_8": 8,
|
39 |
+
"LABEL_9": 9
|
40 |
+
},
|
41 |
+
"layer_norm_eps": 1e-12,
|
42 |
+
"max_position_embeddings": 2048,
|
43 |
+
"model_type": "bert",
|
44 |
+
"num_attention_heads": 12,
|
45 |
+
"num_hidden_layers": 3,
|
46 |
+
"pad_token_id": 0,
|
47 |
+
"position_embedding_type": "absolute",
|
48 |
+
"problem_type": "single_label_classification",
|
49 |
+
"torch_dtype": "float32",
|
50 |
+
"transformers_version": "4.36.2",
|
51 |
+
"type_vocab_size": 2,
|
52 |
+
"use_cache": true,
|
53 |
+
"vocab_size": 83828
|
54 |
+
}
|
run-5/checkpoint-552/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fd4007c85d4ae6c08117d0ba9dfa81c0acf03929b46e6bc1596127b4eb06a0f6
|
3 |
+
size 116795404
|
run-5/checkpoint-552/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e9ce35306b702c395ad003772223670aaa65756e70943ad6cace20a261e14a1e
|
3 |
+
size 233621882
|
run-5/checkpoint-552/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3ffa35aebbbb73516319ab2a6e27b7c86c2e8dbcb15589d859bc25ab4078b43f
|
3 |
+
size 13990
|
run-5/checkpoint-552/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9b57e64ad1e87409614096baa5ee2d70d7b64d526b7b8b5fc0ce76ae59754307
|
3 |
+
size 1064
|