Update README.md

README.md

@@ -28,7 +28,7 @@ It is tailored for multi-label [PhySH](https://physh.org/) tag recommendation fo
 - **Developed by:** swetlanas 
 - **Model type:** Transformer-based sequence classification with LoRA adapter (PEFT)
 - **Language(s) (NLP):** English
-- **License:** [More Information Needed]
+- **License:** apache-2.0
 - **Finetuned from model :** [thellert/physbert_uncased](https://huggingface.co/thellert/physbert_uncased)
 
 ### Model Sources [optional]
@@ -37,7 +37,7 @@ It is tailored for multi-label [PhySH](https://physh.org/) tag recommendation fo
 
 - **Repository:** [PhySH-Tank](https://github.com/swetlanas/PhySH-Tank)
 - **Base model :** [thellert/physbert_uncased](https://huggingface.co/thellert/physbert_uncased)
-- **Demo [optional]:** [More Information Needed]
+- **Demo :** [Coming soon]
 
 ## Uses
 
@@ -68,7 +68,7 @@ This model is designed to recommend Physics subject heading (PhySH) tags for APS
 
 Use the code snippet below to run inference on a title and abstract pair:
 
-```
+```python
 import torch
 from transformers import AutoTokenizer, AutoModelForSequenceClassification
 from peft import PeftModel
@@ -164,90 +164,68 @@ The model was fine-tuned using Hugging Face's `Trainer` and `peft` library with
 
 <!-- This should link to a Dataset Card if possible. -->
 
-[More Information Needed]
+Held-out test split from the ~50,000 APS Physical Review B abstracts (2016–2026), using Hugging Face Dataset `dataset.train_test_split(test_size=0.2, seed=0)` for consistency across all model tracks.
+
 
 #### Factors
 
 <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
 
-[More Information Needed]
+Evaluation was run across the full label space (~3,000 PhySH tags) with no per-subfield breakdown; performance is expected to be stronger on condensed matter / materials science / solid-state topics, which dominate the training distribution.
+
 
 #### Metrics
 
 <!-- These are the evaluation metrics being used, ideally with a description of why. -->
 
-[More Information Needed]
+Precision@5 (P@5) was used as the primary metric and the early-stopping / best-checkpoint criterion, since the task is to recommend the top 5 most relevant tags per paper. 
+Recall@5, NDCG@5, and Hit@5 were also tracked via a custom `compute_metrics` function in the HuggingFace `Trainer`.
+
 
 ### Results
 
-[More Information Needed]
+| Model | Precision@5 |
+|---|---|
+| TF-IDF + MultinomialNB (baseline) | 0.332 |
+| napkinXC (PLT) + frozen PhysBERT embeddings | 0.371 |
+| **LoRA fine-tuned PhysBERT (this model)** | **0.408** |
 
 #### Summary
 
+LoRA fine-tuning of PhysBERT outperforms both a classical TF-IDF baseline and a tree-based extreme multi-label classifier (napkinXC) using frozen PhysBERT embeddings, while updating only ~2.6M parameters (~2.27% of the base model's total).
 
 
-## Model Examination [optional]
-
-<!-- Relevant interpretability work for the model goes here -->
-
-[More Information Needed]
-
-## Environmental Impact
-
-<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
-
-Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
-
-- **Hardware Type:** RTX 3060 12GB graphics card 
-- **Hours used:** [More Information Needed]
-- **Cloud Provider:** [More Information Needed]
-- **Compute Region:** [More Information Needed]
-- **Carbon Emitted:** PyTorch, Hugging Face Transformers, PEFT
-
 ## Technical Specifications [optional]
 
 ### Model Architecture and Objective
 
-[More Information Needed]
+**Base architecture**: PhysBERT [thellert/physbert_uncased](https://huggingface.co/thellert/physbert_uncased), a BERT-style transformer encoder pretrained on physics (arXiv) text, with a sequence classification head for multi-label PhySH tag prediction.
+A LoRA (Low-Rank Adaptation) adapter (r=8, α=16, dropout=0.1) was applied to the `query` and `value` attention projection matrices, with the classification head fully fine-tuned. This updates ~2.6M parameters (~2.27% of the base model's total).
+The training objective is multi-label binary classification: each of the ~3,000 PhySH tags is treated as an independent binary target, optimized with `BCEWithLogitsLoss` (Binary Cross-Entropy with logits) over sigmoid outputs, rather than a single softmax over mutually exclusive classes.
 
 ### Compute Infrastructure
 
-[More Information Needed]
-
 #### Hardware
-
-[More Information Needed]
+RTX 3060 12GB GPU
 
 #### Software
+PyTorch, Hugging Face Transformers, PEFT
 
-[More Information Needed]
-
-## Citation [optional]
+## Citation
 
 <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
 
 **BibTeX:**
 
-[More Information Needed]
-
-**APA:**
-
-[More Information Needed]
-
-## Glossary [optional]
-
-<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
-
-[More Information Needed]
-
-## More Information [optional]
-
-[More Information Needed]
-
-## Model Card Authors [optional]
-
-[More Information Needed]
+```bibtex
+@misc{physh-tank,
+  author = {Swetlana S},
+  title = {PhySH-Tank: PhySH Tag Recommendation for PRB Abstracts},
+  year = {2026},
+  howpublished = {\url{https://github.com/swetlanas/PhySH-Tank}}
+}
+```
 
-## Model Card Contact
+## Model Card Authors
+[swetlanas](https://huggingface.co/swetlanas)
 
-[More Information Needed]