File size: 8,699 Bytes
d77b719
 
24a70a7
 
d77b719
 
 
 
 
 
 
24a70a7
d77b719
 
 
 
 
 
24a70a7
d77b719
24a70a7
d77b719
24a70a7
 
e389c19
24a70a7
 
d77b719
 
 
 
 
 
 
 
 
51abd45
 
092cc87
 
51abd45
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d77b719
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
24a70a7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d77b719
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
---
library_name: transformers
license: mit
base_model: intfloat/multilingual-e5-large-instruct
tags:
- generated_from_trainer
metrics:
- precision
- recall
- accuracy
model-index:
- name: multilingual-e5-large-instruct-edu-scorer-lr5e5-bs32
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# multilingual-e5-large-instruct-edu-scorer-lr5e5-bs32

This model is a fine-tuned version of [intfloat/multilingual-e5-large-instruct](https://huggingface.co/intfloat/multilingual-e5-large-instruct) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.0423
- Precision: 0.4783
- Recall: 0.3533
- F1 Macro: 0.3508
- Accuracy: 0.38

## Model description

More information needed

## Intended uses & limitations

More information needed

## Test results

Binary classification accuracy (threshold at label 3) ≈ 82.00%

Test Report:
```
              precision    recall  f1-score   support

           0       0.78      0.54      0.64       100
           1       0.31      0.34      0.33       100
           2       0.33      0.50      0.40       100
           3       0.29      0.47      0.36       100
           4       0.41      0.21      0.28       100
           5       0.75      0.06      0.11        50

    accuracy                           0.38       550
   macro avg       0.48      0.35      0.35       550
weighted avg       0.45      0.38      0.37       550
```

Confusion Matrix:
```
[[54 37  8  1  0  0]
 [13 34 40 12  1  0]
 [ 2 28 50 19  1  0]
 [ 0  9 35 47  8  1]
 [ 0  1 17 61 21  0]
 [ 0  0  3 24 20  3]]

Test metrics
```
  epoch                   =       20.0
  eval_accuracy           =       0.38
  eval_f1_macro           =     0.3508
  eval_loss               =     1.0423
  eval_precision          =     0.4783
  eval_recall             =     0.3533
  eval_runtime            = 0:00:05.99
  eval_samples_per_second =     91.782
  eval_steps_per_second   =      3.004
```

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 0.0003
- train_batch_size: 32
- eval_batch_size: 32
- seed: 0
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 20

### Training results

| Training Loss | Epoch   | Step  | Validation Loss | Precision | Recall | F1 Macro | Accuracy |
|:-------------:|:-------:|:-----:|:---------------:|:---------:|:------:|:--------:|:--------:|
| No log        | 0       | 0     | 4.0694          | 0.0587    | 0.1667 | 0.0869   | 0.3524   |
| 0.834         | 0.3368  | 1000  | 0.7770          | 0.4163    | 0.3268 | 0.3177   | 0.4876   |
| 0.8535        | 0.6736  | 2000  | 0.9632          | 0.3999    | 0.3404 | 0.3236   | 0.3578   |
| 0.7996        | 1.0104  | 3000  | 0.8151          | 0.4568    | 0.3298 | 0.3278   | 0.5098   |
| 0.8055        | 1.3473  | 4000  | 0.8134          | 0.4109    | 0.3380 | 0.3268   | 0.3946   |
| 0.7807        | 1.6841  | 5000  | 0.7303          | 0.4170    | 0.3698 | 0.3717   | 0.4824   |
| 0.7628        | 2.0209  | 6000  | 0.8066          | 0.4310    | 0.3406 | 0.3296   | 0.3928   |
| 0.7737        | 2.3577  | 7000  | 0.7125          | 0.4856    | 0.3456 | 0.3491   | 0.5032   |
| 0.7246        | 2.6945  | 8000  | 0.7268          | 0.5744    | 0.3593 | 0.3612   | 0.4676   |
| 0.7179        | 3.0313  | 9000  | 0.7028          | 0.4875    | 0.3668 | 0.3726   | 0.4916   |
| 0.715         | 3.3681  | 10000 | 0.6926          | 0.4094    | 0.3570 | 0.3583   | 0.5116   |
| 0.696         | 3.7050  | 11000 | 0.7401          | 0.4944    | 0.3689 | 0.3685   | 0.459    |
| 0.7221        | 4.0418  | 12000 | 0.6773          | 0.5800    | 0.3632 | 0.3707   | 0.5256   |
| 0.7425        | 4.3786  | 13000 | 0.7350          | 0.4280    | 0.3918 | 0.3939   | 0.489    |
| 0.7426        | 4.7154  | 14000 | 0.7564          | 0.4749    | 0.3715 | 0.3711   | 0.4656   |
| 0.6576        | 5.0522  | 15000 | 0.6928          | 0.4105    | 0.3515 | 0.3527   | 0.5278   |
| 0.7033        | 5.3890  | 16000 | 0.6894          | 0.4532    | 0.3863 | 0.3908   | 0.5138   |
| 0.6813        | 5.7258  | 17000 | 0.6655          | 0.4101    | 0.3506 | 0.3380   | 0.5216   |
| 0.6653        | 6.0626  | 18000 | 0.6813          | 0.4161    | 0.3429 | 0.3404   | 0.5248   |
| 0.7086        | 6.3995  | 19000 | 0.6744          | 0.5219    | 0.3909 | 0.4010   | 0.5076   |
| 0.6658        | 6.7363  | 20000 | 0.6524          | 0.5000    | 0.3788 | 0.3852   | 0.5226   |
| 0.6843        | 7.0731  | 21000 | 0.6561          | 0.4176    | 0.3527 | 0.3502   | 0.5256   |
| 0.6927        | 7.4099  | 22000 | 0.6783          | 0.4554    | 0.3810 | 0.3843   | 0.5006   |
| 0.6708        | 7.7467  | 23000 | 0.7420          | 0.3982    | 0.3270 | 0.3197   | 0.516    |
| 0.646         | 8.0835  | 24000 | 0.6684          | 0.4776    | 0.3820 | 0.3890   | 0.491    |
| 0.6577        | 8.4203  | 25000 | 0.6562          | 0.4946    | 0.3718 | 0.3783   | 0.5358   |
| 0.68          | 8.7572  | 26000 | 0.6628          | 0.4841    | 0.3940 | 0.4032   | 0.5168   |
| 0.6827        | 9.0940  | 27000 | 0.6508          | 0.4947    | 0.3715 | 0.3753   | 0.5104   |
| 0.6611        | 9.4308  | 28000 | 0.6434          | 0.5237    | 0.3862 | 0.3981   | 0.5378   |
| 0.6147        | 9.7676  | 29000 | 0.6597          | 0.4199    | 0.3584 | 0.3590   | 0.5376   |
| 0.638         | 10.1044 | 30000 | 0.6330          | 0.5038    | 0.3851 | 0.3902   | 0.5342   |
| 0.6286        | 10.4412 | 31000 | 0.6579          | 0.4508    | 0.3865 | 0.3926   | 0.5      |
| 0.6352        | 10.7780 | 32000 | 0.6586          | 0.4467    | 0.3917 | 0.3982   | 0.5096   |
| 0.6369        | 11.1149 | 33000 | 0.6365          | 0.4970    | 0.3846 | 0.3911   | 0.5248   |
| 0.624         | 11.4517 | 34000 | 0.7212          | 0.4463    | 0.3806 | 0.3796   | 0.457    |
| 0.6502        | 11.7885 | 35000 | 0.6419          | 0.4002    | 0.3771 | 0.3743   | 0.5274   |
| 0.6133        | 12.1253 | 36000 | 0.6765          | 0.3968    | 0.3506 | 0.3457   | 0.535    |
| 0.604         | 12.4621 | 37000 | 0.6361          | 0.4637    | 0.3833 | 0.3908   | 0.5334   |
| 0.6426        | 12.7989 | 38000 | 0.6376          | 0.4137    | 0.3631 | 0.3639   | 0.5352   |
| 0.6227        | 13.1357 | 39000 | 0.6637          | 0.4176    | 0.3579 | 0.3554   | 0.5406   |
| 0.6275        | 13.4725 | 40000 | 0.6446          | 0.4482    | 0.3989 | 0.4050   | 0.528    |
| 0.6545        | 13.8094 | 41000 | 0.6526          | 0.3960    | 0.3524 | 0.3479   | 0.534    |
| 0.5786        | 14.1462 | 42000 | 0.6280          | 0.4445    | 0.3847 | 0.3907   | 0.5358   |
| 0.6123        | 14.4830 | 43000 | 0.6351          | 0.4075    | 0.3799 | 0.3805   | 0.5234   |
| 0.5885        | 14.8198 | 44000 | 0.6633          | 0.4102    | 0.3775 | 0.3743   | 0.4986   |
| 0.6052        | 15.1566 | 45000 | 0.6437          | 0.4354    | 0.3940 | 0.3968   | 0.5226   |
| 0.6066        | 15.4934 | 46000 | 0.6305          | 0.4056    | 0.3724 | 0.3732   | 0.5382   |
| 0.6106        | 15.8302 | 47000 | 0.6317          | 0.4078    | 0.3802 | 0.3818   | 0.5298   |
| 0.5995        | 16.1671 | 48000 | 0.6644          | 0.4267    | 0.3933 | 0.3947   | 0.5078   |
| 0.6001        | 16.5039 | 49000 | 0.6294          | 0.4152    | 0.3830 | 0.3839   | 0.5424   |
| 0.5553        | 16.8407 | 50000 | 0.6293          | 0.4150    | 0.3829 | 0.3838   | 0.5448   |
| 0.5787        | 17.1775 | 51000 | 0.6284          | 0.4143    | 0.3813 | 0.3829   | 0.5404   |
| 0.5724        | 17.5143 | 52000 | 0.6255          | 0.4178    | 0.3810 | 0.3819   | 0.544    |
| 0.5558        | 17.8511 | 53000 | 0.6251          | 0.4187    | 0.3827 | 0.3840   | 0.539    |
| 0.533         | 18.1879 | 54000 | 0.6263          | 0.4175    | 0.3816 | 0.3824   | 0.538    |
| 0.5612        | 18.5248 | 55000 | 0.6302          | 0.4122    | 0.3870 | 0.3880   | 0.5382   |
| 0.5594        | 18.8616 | 56000 | 0.6230          | 0.4203    | 0.3800 | 0.3807   | 0.5402   |
| 0.565         | 19.1984 | 57000 | 0.6264          | 0.4117    | 0.3789 | 0.3799   | 0.5346   |
| 0.5533        | 19.5352 | 58000 | 0.6261          | 0.4153    | 0.3825 | 0.3837   | 0.537    |
| 0.5459        | 19.8720 | 59000 | 0.6289          | 0.4128    | 0.3846 | 0.3853   | 0.5342   |


### Framework versions

- Transformers 4.53.2
- Pytorch 2.7.1+cu126
- Datasets 4.0.0
- Tokenizers 0.21.2