Upload folder using huggingface_hub
Browse files- README.md +36 -0
- checkpoint-150/config.json +44 -0
- checkpoint-150/model.safetensors +3 -0
- checkpoint-150/optimizer.pt +3 -0
- checkpoint-150/rng_state.pth +3 -0
- checkpoint-150/scheduler.pt +3 -0
- checkpoint-150/trainer_state.json +201 -0
- checkpoint-150/training_args.bin +3 -0
- config.json +44 -0
- model.safetensors +3 -0
- runs/Aug18_21-09-43_r-vpingale07-phi-3-transaction-categories-9a0vqm8c-18b68-bg15o/events.out.tfevents.1724015384.r-vpingale07-phi-3-transaction-categories-9a0vqm8c-18b68-bg15o.183.0 +2 -2
- runs/Aug18_21-09-43_r-vpingale07-phi-3-transaction-categories-9a0vqm8c-18b68-bg15o/events.out.tfevents.1724016371.r-vpingale07-phi-3-transaction-categories-9a0vqm8c-18b68-bg15o.183.1 +3 -0
- special_tokens_map.json +7 -0
- tokenizer.json +0 -0
- tokenizer_config.json +55 -0
- training_args.bin +3 -0
- training_params.json +30 -0
- vocab.txt +0 -0
README.md
ADDED
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
|
2 |
+
---
|
3 |
+
tags:
|
4 |
+
- autotrain
|
5 |
+
- text-classification
|
6 |
+
base_model: google-bert/bert-base-uncased
|
7 |
+
widget:
|
8 |
+
- text: "I love AutoTrain"
|
9 |
+
---
|
10 |
+
|
11 |
+
# Model Trained Using AutoTrain
|
12 |
+
|
13 |
+
- Problem type: Text Classification
|
14 |
+
|
15 |
+
## Validation Metrics
|
16 |
+
loss: 0.29163089394569397
|
17 |
+
|
18 |
+
f1_macro: 0.7727998858935957
|
19 |
+
|
20 |
+
f1_micro: 0.92
|
21 |
+
|
22 |
+
f1_weighted: 0.9113598630723149
|
23 |
+
|
24 |
+
precision_macro: 0.7648809523809524
|
25 |
+
|
26 |
+
precision_micro: 0.92
|
27 |
+
|
28 |
+
precision_weighted: 0.9053571428571427
|
29 |
+
|
30 |
+
recall_macro: 0.7833333333333332
|
31 |
+
|
32 |
+
recall_micro: 0.92
|
33 |
+
|
34 |
+
recall_weighted: 0.92
|
35 |
+
|
36 |
+
accuracy: 0.92
|
checkpoint-150/config.json
ADDED
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "google-bert/bert-base-uncased",
|
3 |
+
"_num_labels": 6,
|
4 |
+
"architectures": [
|
5 |
+
"BertForSequenceClassification"
|
6 |
+
],
|
7 |
+
"attention_probs_dropout_prob": 0.1,
|
8 |
+
"classifier_dropout": null,
|
9 |
+
"gradient_checkpointing": false,
|
10 |
+
"hidden_act": "gelu",
|
11 |
+
"hidden_dropout_prob": 0.1,
|
12 |
+
"hidden_size": 768,
|
13 |
+
"id2label": {
|
14 |
+
"0": "Entertainment",
|
15 |
+
"1": "Food",
|
16 |
+
"2": "Housing",
|
17 |
+
"3": "Medical",
|
18 |
+
"4": "Miscellaneous",
|
19 |
+
"5": "Travel"
|
20 |
+
},
|
21 |
+
"initializer_range": 0.02,
|
22 |
+
"intermediate_size": 3072,
|
23 |
+
"label2id": {
|
24 |
+
"Entertainment": 0,
|
25 |
+
"Food": 1,
|
26 |
+
"Housing": 2,
|
27 |
+
"Medical": 3,
|
28 |
+
"Miscellaneous": 4,
|
29 |
+
"Travel": 5
|
30 |
+
},
|
31 |
+
"layer_norm_eps": 1e-12,
|
32 |
+
"max_position_embeddings": 512,
|
33 |
+
"model_type": "bert",
|
34 |
+
"num_attention_heads": 12,
|
35 |
+
"num_hidden_layers": 12,
|
36 |
+
"pad_token_id": 0,
|
37 |
+
"position_embedding_type": "absolute",
|
38 |
+
"problem_type": "single_label_classification",
|
39 |
+
"torch_dtype": "float32",
|
40 |
+
"transformers_version": "4.44.0",
|
41 |
+
"type_vocab_size": 2,
|
42 |
+
"use_cache": true,
|
43 |
+
"vocab_size": 30522
|
44 |
+
}
|
checkpoint-150/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8b80cf61dc51924bcde8910ef314f709a88811306928475cdb9bd6776ade1bb5
|
3 |
+
size 437970952
|
checkpoint-150/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:10ae67a1886ffd443673f84a6a2a2451203609691ed283ec25e8586c1a7a9f8f
|
3 |
+
size 876057338
|
checkpoint-150/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b53c99a2ea19505de0d4e18f71e3b435e94fad506b2f76c61e4575c00e2ea2ca
|
3 |
+
size 13990
|
checkpoint-150/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1edb25f446c7d2fc8a07e3635e302199dbca58668ddad584ef7fbc0b6b5406b6
|
3 |
+
size 1064
|
checkpoint-150/trainer_state.json
ADDED
@@ -0,0 +1,201 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.29163089394569397,
|
3 |
+
"best_model_checkpoint": "autotrain-2p12r-cw2s6/checkpoint-150",
|
4 |
+
"epoch": 3.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 150,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.2,
|
13 |
+
"grad_norm": 13.35822582244873,
|
14 |
+
"learning_rate": 1.9999999999999998e-05,
|
15 |
+
"loss": 1.8569,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.4,
|
20 |
+
"grad_norm": 10.923691749572754,
|
21 |
+
"learning_rate": 2.8888888888888888e-05,
|
22 |
+
"loss": 1.6203,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.6,
|
27 |
+
"grad_norm": 6.453679084777832,
|
28 |
+
"learning_rate": 2.6666666666666667e-05,
|
29 |
+
"loss": 1.2797,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.8,
|
34 |
+
"grad_norm": 10.624079704284668,
|
35 |
+
"learning_rate": 2.4444444444444445e-05,
|
36 |
+
"loss": 1.1935,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 1.0,
|
41 |
+
"grad_norm": 10.568455696105957,
|
42 |
+
"learning_rate": 2.222222222222222e-05,
|
43 |
+
"loss": 0.9807,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 1.0,
|
48 |
+
"eval_accuracy": 0.82,
|
49 |
+
"eval_f1_macro": 0.6733605498937164,
|
50 |
+
"eval_f1_micro": 0.82,
|
51 |
+
"eval_f1_weighted": 0.789679718695989,
|
52 |
+
"eval_loss": 0.8373663425445557,
|
53 |
+
"eval_precision_macro": 0.6934624017957352,
|
54 |
+
"eval_precision_micro": 0.82,
|
55 |
+
"eval_precision_weighted": 0.8152104377104377,
|
56 |
+
"eval_recall_macro": 0.7000000000000001,
|
57 |
+
"eval_recall_micro": 0.82,
|
58 |
+
"eval_recall_weighted": 0.82,
|
59 |
+
"eval_runtime": 19.5504,
|
60 |
+
"eval_samples_per_second": 5.115,
|
61 |
+
"eval_steps_per_second": 1.279,
|
62 |
+
"step": 50
|
63 |
+
},
|
64 |
+
{
|
65 |
+
"epoch": 1.2,
|
66 |
+
"grad_norm": 6.850324630737305,
|
67 |
+
"learning_rate": 1.9999999999999998e-05,
|
68 |
+
"loss": 0.8111,
|
69 |
+
"step": 60
|
70 |
+
},
|
71 |
+
{
|
72 |
+
"epoch": 1.4,
|
73 |
+
"grad_norm": 6.868027210235596,
|
74 |
+
"learning_rate": 1.7777777777777777e-05,
|
75 |
+
"loss": 0.721,
|
76 |
+
"step": 70
|
77 |
+
},
|
78 |
+
{
|
79 |
+
"epoch": 1.6,
|
80 |
+
"grad_norm": 5.674863815307617,
|
81 |
+
"learning_rate": 1.5555555555555555e-05,
|
82 |
+
"loss": 0.5653,
|
83 |
+
"step": 80
|
84 |
+
},
|
85 |
+
{
|
86 |
+
"epoch": 1.8,
|
87 |
+
"grad_norm": 4.8254852294921875,
|
88 |
+
"learning_rate": 1.3333333333333333e-05,
|
89 |
+
"loss": 0.5168,
|
90 |
+
"step": 90
|
91 |
+
},
|
92 |
+
{
|
93 |
+
"epoch": 2.0,
|
94 |
+
"grad_norm": 9.87863826751709,
|
95 |
+
"learning_rate": 1.111111111111111e-05,
|
96 |
+
"loss": 0.4505,
|
97 |
+
"step": 100
|
98 |
+
},
|
99 |
+
{
|
100 |
+
"epoch": 2.0,
|
101 |
+
"eval_accuracy": 0.92,
|
102 |
+
"eval_f1_macro": 0.7727998858935957,
|
103 |
+
"eval_f1_micro": 0.92,
|
104 |
+
"eval_f1_weighted": 0.9113598630723149,
|
105 |
+
"eval_loss": 0.3886770009994507,
|
106 |
+
"eval_precision_macro": 0.7648809523809524,
|
107 |
+
"eval_precision_micro": 0.92,
|
108 |
+
"eval_precision_weighted": 0.9053571428571427,
|
109 |
+
"eval_recall_macro": 0.7833333333333332,
|
110 |
+
"eval_recall_micro": 0.92,
|
111 |
+
"eval_recall_weighted": 0.92,
|
112 |
+
"eval_runtime": 20.1168,
|
113 |
+
"eval_samples_per_second": 4.971,
|
114 |
+
"eval_steps_per_second": 1.243,
|
115 |
+
"step": 100
|
116 |
+
},
|
117 |
+
{
|
118 |
+
"epoch": 2.2,
|
119 |
+
"grad_norm": 2.6550135612487793,
|
120 |
+
"learning_rate": 8.888888888888888e-06,
|
121 |
+
"loss": 0.347,
|
122 |
+
"step": 110
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 2.4,
|
126 |
+
"grad_norm": 5.567917346954346,
|
127 |
+
"learning_rate": 6.666666666666667e-06,
|
128 |
+
"loss": 0.3921,
|
129 |
+
"step": 120
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 2.6,
|
133 |
+
"grad_norm": 11.323433876037598,
|
134 |
+
"learning_rate": 4.444444444444444e-06,
|
135 |
+
"loss": 0.3246,
|
136 |
+
"step": 130
|
137 |
+
},
|
138 |
+
{
|
139 |
+
"epoch": 2.8,
|
140 |
+
"grad_norm": 5.023998260498047,
|
141 |
+
"learning_rate": 2.222222222222222e-06,
|
142 |
+
"loss": 0.3217,
|
143 |
+
"step": 140
|
144 |
+
},
|
145 |
+
{
|
146 |
+
"epoch": 3.0,
|
147 |
+
"grad_norm": 2.515089750289917,
|
148 |
+
"learning_rate": 0.0,
|
149 |
+
"loss": 0.2423,
|
150 |
+
"step": 150
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 3.0,
|
154 |
+
"eval_accuracy": 0.92,
|
155 |
+
"eval_f1_macro": 0.7727998858935957,
|
156 |
+
"eval_f1_micro": 0.92,
|
157 |
+
"eval_f1_weighted": 0.9113598630723149,
|
158 |
+
"eval_loss": 0.29163089394569397,
|
159 |
+
"eval_precision_macro": 0.7648809523809524,
|
160 |
+
"eval_precision_micro": 0.92,
|
161 |
+
"eval_precision_weighted": 0.9053571428571427,
|
162 |
+
"eval_recall_macro": 0.7833333333333332,
|
163 |
+
"eval_recall_micro": 0.92,
|
164 |
+
"eval_recall_weighted": 0.92,
|
165 |
+
"eval_runtime": 19.9884,
|
166 |
+
"eval_samples_per_second": 5.003,
|
167 |
+
"eval_steps_per_second": 1.251,
|
168 |
+
"step": 150
|
169 |
+
}
|
170 |
+
],
|
171 |
+
"logging_steps": 10,
|
172 |
+
"max_steps": 150,
|
173 |
+
"num_input_tokens_seen": 0,
|
174 |
+
"num_train_epochs": 3,
|
175 |
+
"save_steps": 500,
|
176 |
+
"stateful_callbacks": {
|
177 |
+
"EarlyStoppingCallback": {
|
178 |
+
"args": {
|
179 |
+
"early_stopping_patience": 5,
|
180 |
+
"early_stopping_threshold": 0.01
|
181 |
+
},
|
182 |
+
"attributes": {
|
183 |
+
"early_stopping_patience_counter": 0
|
184 |
+
}
|
185 |
+
},
|
186 |
+
"TrainerControl": {
|
187 |
+
"args": {
|
188 |
+
"should_epoch_stop": false,
|
189 |
+
"should_evaluate": false,
|
190 |
+
"should_log": false,
|
191 |
+
"should_save": true,
|
192 |
+
"should_training_stop": true
|
193 |
+
},
|
194 |
+
"attributes": {}
|
195 |
+
}
|
196 |
+
},
|
197 |
+
"total_flos": 78936151449600.0,
|
198 |
+
"train_batch_size": 2,
|
199 |
+
"trial_name": null,
|
200 |
+
"trial_params": null
|
201 |
+
}
|
checkpoint-150/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b52150b188249c22f02c127cccbcc0d9ad437b777179e20f20bc5f13869936d9
|
3 |
+
size 5240
|
config.json
ADDED
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "google-bert/bert-base-uncased",
|
3 |
+
"_num_labels": 6,
|
4 |
+
"architectures": [
|
5 |
+
"BertForSequenceClassification"
|
6 |
+
],
|
7 |
+
"attention_probs_dropout_prob": 0.1,
|
8 |
+
"classifier_dropout": null,
|
9 |
+
"gradient_checkpointing": false,
|
10 |
+
"hidden_act": "gelu",
|
11 |
+
"hidden_dropout_prob": 0.1,
|
12 |
+
"hidden_size": 768,
|
13 |
+
"id2label": {
|
14 |
+
"0": "Entertainment",
|
15 |
+
"1": "Food",
|
16 |
+
"2": "Housing",
|
17 |
+
"3": "Medical",
|
18 |
+
"4": "Miscellaneous",
|
19 |
+
"5": "Travel"
|
20 |
+
},
|
21 |
+
"initializer_range": 0.02,
|
22 |
+
"intermediate_size": 3072,
|
23 |
+
"label2id": {
|
24 |
+
"Entertainment": 0,
|
25 |
+
"Food": 1,
|
26 |
+
"Housing": 2,
|
27 |
+
"Medical": 3,
|
28 |
+
"Miscellaneous": 4,
|
29 |
+
"Travel": 5
|
30 |
+
},
|
31 |
+
"layer_norm_eps": 1e-12,
|
32 |
+
"max_position_embeddings": 512,
|
33 |
+
"model_type": "bert",
|
34 |
+
"num_attention_heads": 12,
|
35 |
+
"num_hidden_layers": 12,
|
36 |
+
"pad_token_id": 0,
|
37 |
+
"position_embedding_type": "absolute",
|
38 |
+
"problem_type": "single_label_classification",
|
39 |
+
"torch_dtype": "float32",
|
40 |
+
"transformers_version": "4.44.0",
|
41 |
+
"type_vocab_size": 2,
|
42 |
+
"use_cache": true,
|
43 |
+
"vocab_size": 30522
|
44 |
+
}
|
model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8b80cf61dc51924bcde8910ef314f709a88811306928475cdb9bd6776ade1bb5
|
3 |
+
size 437970952
|
runs/Aug18_21-09-43_r-vpingale07-phi-3-transaction-categories-9a0vqm8c-18b68-bg15o/events.out.tfevents.1724015384.r-vpingale07-phi-3-transaction-categories-9a0vqm8c-18b68-bg15o.183.0
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c25563fe8d0375a2bb7e7b6239ad6376706d6ff82c656a42716ba24a35fe9d63
|
3 |
+
size 11211
|
runs/Aug18_21-09-43_r-vpingale07-phi-3-transaction-categories-9a0vqm8c-18b68-bg15o/events.out.tfevents.1724016371.r-vpingale07-phi-3-transaction-categories-9a0vqm8c-18b68-bg15o.183.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c3d7f803ccb28a9a18c7d5d92acc896cd63404a2933942bb2068a39d5801f755
|
3 |
+
size 921
|
special_tokens_map.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": "[CLS]",
|
3 |
+
"mask_token": "[MASK]",
|
4 |
+
"pad_token": "[PAD]",
|
5 |
+
"sep_token": "[SEP]",
|
6 |
+
"unk_token": "[UNK]"
|
7 |
+
}
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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 |
+
"100": {
|
12 |
+
"content": "[UNK]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"101": {
|
20 |
+
"content": "[CLS]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"102": {
|
28 |
+
"content": "[SEP]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"103": {
|
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_lower_case": true,
|
47 |
+
"mask_token": "[MASK]",
|
48 |
+
"model_max_length": 512,
|
49 |
+
"pad_token": "[PAD]",
|
50 |
+
"sep_token": "[SEP]",
|
51 |
+
"strip_accents": null,
|
52 |
+
"tokenize_chinese_chars": true,
|
53 |
+
"tokenizer_class": "BertTokenizer",
|
54 |
+
"unk_token": "[UNK]"
|
55 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b52150b188249c22f02c127cccbcc0d9ad437b777179e20f20bc5f13869936d9
|
3 |
+
size 5240
|
training_params.json
ADDED
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"data_path": "autotrain-2p12r-cw2s6/autotrain-data",
|
3 |
+
"model": "google-bert/bert-base-uncased",
|
4 |
+
"lr": 3e-05,
|
5 |
+
"epochs": 3,
|
6 |
+
"max_seq_length": 128,
|
7 |
+
"batch_size": 2,
|
8 |
+
"warmup_ratio": 0.1,
|
9 |
+
"gradient_accumulation": 4,
|
10 |
+
"optimizer": "adamw_torch",
|
11 |
+
"scheduler": "linear",
|
12 |
+
"weight_decay": 0.0,
|
13 |
+
"max_grad_norm": 1.0,
|
14 |
+
"seed": 42,
|
15 |
+
"train_split": "train",
|
16 |
+
"valid_split": "validation",
|
17 |
+
"text_column": "autotrain_text",
|
18 |
+
"target_column": "autotrain_label",
|
19 |
+
"logging_steps": -1,
|
20 |
+
"project_name": "autotrain-2p12r-cw2s6",
|
21 |
+
"auto_find_batch_size": false,
|
22 |
+
"mixed_precision": "fp16",
|
23 |
+
"save_total_limit": 1,
|
24 |
+
"push_to_hub": true,
|
25 |
+
"eval_strategy": "epoch",
|
26 |
+
"username": "vpingale07",
|
27 |
+
"log": "tensorboard",
|
28 |
+
"early_stopping_patience": 5,
|
29 |
+
"early_stopping_threshold": 0.01
|
30 |
+
}
|
vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|