Update README.md
Browse files
README.md
CHANGED
|
@@ -11,6 +11,10 @@ metrics:
|
|
| 11 |
model-index:
|
| 12 |
- name: xlm-roberta-large-azsci-topics
|
| 13 |
results: []
|
|
|
|
|
|
|
|
|
|
|
|
|
| 14 |
---
|
| 15 |
|
| 16 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -18,7 +22,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 18 |
|
| 19 |
# xlm-roberta-large-azsci-topics
|
| 20 |
|
| 21 |
-
This model is a fine-tuned version of [FacebookAI/xlm-roberta-large](https://huggingface.co/FacebookAI/xlm-roberta-large) on
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
- Loss: 0.4012
|
| 24 |
- Precision: 0.9115
|
|
@@ -26,18 +30,6 @@ It achieves the following results on the evaluation set:
|
|
| 26 |
- F1: 0.9121
|
| 27 |
- Accuracy: 0.9158
|
| 28 |
|
| 29 |
-
## Model description
|
| 30 |
-
|
| 31 |
-
More information needed
|
| 32 |
-
|
| 33 |
-
## Intended uses & limitations
|
| 34 |
-
|
| 35 |
-
More information needed
|
| 36 |
-
|
| 37 |
-
## Training and evaluation data
|
| 38 |
-
|
| 39 |
-
More information needed
|
| 40 |
-
|
| 41 |
## Training procedure
|
| 42 |
|
| 43 |
### Training hyperparameters
|
|
@@ -61,10 +53,39 @@ The following hyperparameters were used during training:
|
|
| 61 |
| 0.3135 | 4.0 | 1152 | 0.4177 | 0.9043 | 0.9080 | 0.9047 | 0.9080 |
|
| 62 |
| 0.3135 | 5.0 | 1440 | 0.4012 | 0.9115 | 0.9158 | 0.9121 | 0.9158 |
|
| 63 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 64 |
|
| 65 |
### Framework versions
|
| 66 |
|
| 67 |
- Transformers 4.38.2
|
| 68 |
- Pytorch 2.1.0+cu121
|
| 69 |
- Datasets 2.18.0
|
| 70 |
-
- Tokenizers 0.15.2
|
|
|
|
| 11 |
model-index:
|
| 12 |
- name: xlm-roberta-large-azsci-topics
|
| 13 |
results: []
|
| 14 |
+
datasets:
|
| 15 |
+
- hajili/azsci_topics
|
| 16 |
+
language:
|
| 17 |
+
- az
|
| 18 |
---
|
| 19 |
|
| 20 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 22 |
|
| 23 |
# xlm-roberta-large-azsci-topics
|
| 24 |
|
| 25 |
+
This model is a fine-tuned version of [FacebookAI/xlm-roberta-large](https://huggingface.co/FacebookAI/xlm-roberta-large) on [azsci_topics](https://huggingface.co/datasets/hajili/azsci_topics) dataset.
|
| 26 |
It achieves the following results on the evaluation set:
|
| 27 |
- Loss: 0.4012
|
| 28 |
- Precision: 0.9115
|
|
|
|
| 30 |
- F1: 0.9121
|
| 31 |
- Accuracy: 0.9158
|
| 32 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
## Training procedure
|
| 34 |
|
| 35 |
### Training hyperparameters
|
|
|
|
| 53 |
| 0.3135 | 4.0 | 1152 | 0.4177 | 0.9043 | 0.9080 | 0.9047 | 0.9080 |
|
| 54 |
| 0.3135 | 5.0 | 1440 | 0.4012 | 0.9115 | 0.9158 | 0.9121 | 0.9158 |
|
| 55 |
|
| 56 |
+
### Evaluation results
|
| 57 |
+
|
| 58 |
+
| Topic | Precision | Recall | F1 | Support |
|
| 59 |
+
|:-------------------|------------:|---------:|---------:|----------:|
|
| 60 |
+
| Aqrar elmlər | 0.846154 | 0.814815 | 0.830189 | 27 |
|
| 61 |
+
| Astronomiya | 0.666667 | 1 | 0.8 | 2 |
|
| 62 |
+
| Biologiya elmləri | 0.910891 | 0.87619 | 0.893204 | 105 |
|
| 63 |
+
| Coğrafiya | 0.888889 | 0.941176 | 0.914286 | 17 |
|
| 64 |
+
| Filologiya elmləri | 0.971098 | 0.96 | 0.965517 | 175 |
|
| 65 |
+
| Fizika | 0.769231 | 0.882353 | 0.821918 | 34 |
|
| 66 |
+
| Fəlsəfə | 0.875 | 0.5 | 0.636364 | 14 |
|
| 67 |
+
| Hüquq elmləri | 0.966667 | 1 | 0.983051 | 29 |
|
| 68 |
+
| Kimya | 0.855072 | 0.967213 | 0.907692 | 61 |
|
| 69 |
+
| Memarlıq | 0.714286 | 1 | 0.833333 | 5 |
|
| 70 |
+
| Mexanika | 0 | 0 | 0 | 4 |
|
| 71 |
+
| Pedaqogika | 0.958333 | 0.978723 | 0.968421 | 47 |
|
| 72 |
+
| Psixologiya | 0.944444 | 0.944444 | 0.944444 | 18 |
|
| 73 |
+
| Riyaziyyat | 0.921053 | 0.897436 | 0.909091 | 39 |
|
| 74 |
+
| Siyasi elmlər | 0.785714 | 0.88 | 0.830189 | 25 |
|
| 75 |
+
| Sosiologiya | 0.666667 | 1 | 0.8 | 4 |
|
| 76 |
+
| Sənətşünaslıq | 0.84 | 0.893617 | 0.865979 | 47 |
|
| 77 |
+
| Tarix | 0.933333 | 0.897436 | 0.915033 | 78 |
|
| 78 |
+
| Texnika elmləri | 0.894737 | 0.817308 | 0.854271 | 104 |
|
| 79 |
+
| Tibb elmləri | 0.935484 | 0.97973 | 0.957096 | 148 |
|
| 80 |
+
| Yer elmləri | 0.846154 | 0.846154 | 0.846154 | 13 |
|
| 81 |
+
| İqtisad elmləri | 0.973684 | 0.973684 | 0.973684 | 152 |
|
| 82 |
+
| Əczaçılıq elmləri | 0 | 0 | 0 | 4 |
|
| 83 |
+
| macro avg | 0.78972 | 0.828273 | 0.80217 | 1152 |
|
| 84 |
+
| weighted avg | 0.911546 | 0.915799 | 0.912067 | 1152 |
|
| 85 |
|
| 86 |
### Framework versions
|
| 87 |
|
| 88 |
- Transformers 4.38.2
|
| 89 |
- Pytorch 2.1.0+cu121
|
| 90 |
- Datasets 2.18.0
|
| 91 |
+
- Tokenizers 0.15.2
|