Update README.md

README.md

@@ -4,34 +4,100 @@ license: apache-2.0
 base_model: TomasFAV/Pix2StructCzechInvoice
 tags:
 - generated_from_trainer
+- invoice-processing
+- information-extraction
+- czech-language
+- document-ai
+- multimodal-model
+- generative-model
+- synthetic-data
+- layout-augmentation
 metrics:
 - f1
 model-index:
-- name: Pix2StructCzechInvoiceV1
+- name: Pix2StructCzechInvoice-V1
   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. -->
+# Pix2StructCzechInvoice (V1 – Synthetic + Random Layout)
 
-# Pix2StructCzechInvoiceV1
+This model is a fine-tuned version of [TomasFAV/Pix2StructCzechInvoice](https://huggingface.co/TomasFAV/Pix2StructCzechInvoice) for structured information extraction from Czech invoices.
 
-This model is a fine-tuned version of [TomasFAV/Pix2StructCzechInvoice](https://huggingface.co/TomasFAV/Pix2StructCzechInvoice) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.4679
-- F1: 0.6432
+- Loss: 0.4679  
+- F1: 0.6432  
+
+---
 
 ## Model description
 
-More information needed
+Pix2StructCzechInvoice (V1) extends the baseline generative model by introducing layout variability into the training data.
+
+Unlike token classification models, this model:
+- processes full document images  
+- generates structured outputs as text sequences  
+
+It is trained to extract key invoice fields:
+- supplier  
+- customer  
+- invoice number  
+- bank details  
+- totals  
+- dates  
+
+---
+
+## Training data
+
+The dataset consists of:
+
+- synthetically generated invoice images  
+- augmented variants with randomized layouts  
+- corresponding structured text outputs  
+
+Key properties:
+- variable layout structure  
+- visual diversity (spacing, positioning, formatting)  
+- consistent annotation format  
+- fully synthetic data  
+
+This introduces **layout variability in the visual domain**, which is crucial for generative multimodal models.
+
+---
+
+## Role in the pipeline
+
+This model corresponds to:
 
-## Intended uses & limitations
+**V1 – Synthetic templates + randomized layouts**
 
-More information needed
+It is used to:
+- evaluate the effect of layout variability on generative models  
+- compare against:
+  - V0 (fixed templates)  
+  - later hybrid and real-data stages (V2, V3)  
+- analyze robustness of end-to-end extraction  
+
+---
 
-## Training and evaluation data
+## Intended uses
+
+- End-to-end invoice extraction from images  
+- Document VQA-style tasks  
+- Research in generative document understanding  
+- Comparison with structured prediction models  
+
+---
 
-More information needed
+## Limitations
+
+- Still trained only on synthetic data  
+- Sensitive to output formatting inconsistencies  
+- Training instability (fluctuating F1 across epochs)  
+- Evaluation depends on string matching quality  
+- Less interpretable than token classification models  
+
+---
 
 ## Training procedure
 
@@ -48,6 +114,8 @@ The following hyperparameters were used during training:
 - num_epochs: 10
 - mixed_precision_training: Native AMP
 
+---
+
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | F1     |
@@ -63,10 +131,11 @@ The following hyperparameters were used during training:
 | 0.0393        | 9.0   | 675  | 0.4679          | 0.6432 |
 | 0.0392        | 10.0  | 750  | 0.5330          | 0.4931 |
 
+---
 
-### Framework versions
+## Framework versions
 
-- Transformers 5.0.0
-- Pytorch 2.10.0+cu128
-- Datasets 4.0.0
-- Tokenizers 0.22.2
+- Transformers 5.0.0  
+- PyTorch 2.10.0+cu128  
+- Datasets 4.0.0  
+- Tokenizers 0.22.2