Model save

README.md

@@ -4,6 +4,11 @@ license: mit
 base_model: SCUT-DLVCLab/lilt-roberta-en-base
 tags:
 - generated_from_trainer
+metrics:
+- precision
+- recall
+- f1
+- accuracy
 model-index:
 - name: LiLTInvoiceCzech
   results: []
@@ -16,16 +21,11 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [SCUT-DLVCLab/lilt-roberta-en-base](https://huggingface.co/SCUT-DLVCLab/lilt-roberta-en-base) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- eval_loss: 0.0224
-- eval_precision: 0.9615
-- eval_recall: 0.9650
-- eval_f1: 0.9632
-- eval_accuracy: 0.9944
-- eval_runtime: 1.0854
-- eval_samples_per_second: 60.809
-- eval_steps_per_second: 30.405
-- epoch: 13.0
-- step: 559
+- Loss: 0.0399
+- Precision: 0.8785
+- Recall: 0.9074
+- F1: 0.8927
+- Accuracy: 0.9919
 
 ## Model description
 
@@ -45,18 +45,34 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 1e-05
-- train_batch_size: 16
+- train_batch_size: 4
 - eval_batch_size: 2
 - seed: 42
 - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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
+- lr_scheduler_warmup_steps: 0.1
+- num_epochs: 10
 - mixed_precision_training: Native AMP
 
+### Training results
+
+| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1     | Accuracy |
+|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
+| No log        | 1.0   | 46   | 0.0509          | 0.7973    | 0.8498 | 0.8227 | 0.9877   |
+| No log        | 2.0   | 92   | 0.0417          | 0.8521    | 0.9012 | 0.8760 | 0.9907   |
+| No log        | 3.0   | 138  | 0.0392          | 0.8651    | 0.8971 | 0.8808 | 0.9915   |
+| No log        | 4.0   | 184  | 0.0432          | 0.8261    | 0.8992 | 0.8611 | 0.9900   |
+| No log        | 5.0   | 230  | 0.0429          | 0.8671    | 0.8992 | 0.8828 | 0.9912   |
+| No log        | 6.0   | 276  | 0.0412          | 0.8661    | 0.9053 | 0.8853 | 0.9913   |
+| No log        | 7.0   | 322  | 0.0411          | 0.8685    | 0.8971 | 0.8826 | 0.9918   |
+| No log        | 8.0   | 368  | 0.0409          | 0.8661    | 0.9053 | 0.8853 | 0.9914   |
+| No log        | 9.0   | 414  | 0.0393          | 0.8802    | 0.9074 | 0.8936 | 0.9920   |
+| No log        | 10.0  | 460  | 0.0399          | 0.8785    | 0.9074 | 0.8927 | 0.9919   |
+
+
 ### Framework versions
 
-- Transformers 4.57.6
-- Pytorch 2.9.0+cu126
+- Transformers 5.0.0
+- Pytorch 2.10.0+cu128
 - Datasets 4.0.0
 - Tokenizers 0.22.2

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0fbf94069ba2151627e46279729b6effb239d431b3f641d72abf4d1947d30b53
+oid sha256:4c643bd8dcd0cfb0eaebe7e2e18c24a3a3144c65e7939682dd03c7ae0b716b54
 size 520801340