Upload folder using huggingface_hub

README.md

@@ -1,20 +1,23 @@
 ---
-base_model:
-- jealk/TTC-L2V-supervised-2
-language:
-- da
 library_name: model2vec
 license: mit
-model_name: andersborges/model2vecdk
+model_name: model2vecdk
 tags:
 - embeddings
 - static-embeddings
 - sentence-transformers
+base_model:
+  - jealk/TTC-L2V-supervised-2
+language:
+  - da
+datasets:
+  - DDSC/nordic-embedding-training-data
+repo_url: https://github.com/andersborges/dkmodel2vec
 ---
 
-# andersborges/model2vecdk Model Card
+# dkmodel2vec Model Card
 
-This [Model2Vec](https://github.com/MinishLab/model2vec) model is a distilled version of the ['jealk/TTC-L2V-supervised-2'](https://huggingface.co/['jealk/TTC-L2V-supervised-2']) Sentence Transformer. It uses static embeddings, allowing text embeddings to be computed orders of magnitude faster on both GPU and CPU. It is designed for applications where computational resources are limited or where real-time performance is critical. Model2Vec models are the smallest, fastest, and most performant static embedders available. The distilled models are up to 50 times smaller and 500 times faster than traditional Sentence Transformers.
+This [Model2Vec](https://github.com/MinishLab/model2vec) model is a distilled version of a [LLM2Vec](https://github.com/McGill-NLP/llm2vec) model. It uses static embeddings, allowing text embeddings to be computed orders of magnitude faster on both GPU and CPU. It is designed for applications where computational resources are limited or where real-time performance is critical. Model2Vec models are the smallest, fastest, and most performant static embedders available. The distilled models are up to 50 times smaller and 500 times faster than traditional Sentence Transformers.
 
 
 ## Installation
@@ -38,7 +41,7 @@ from model2vec import StaticModel
 model = StaticModel.from_pretrained("andersborges/model2vecdk")
 
 # Compute text embeddings
-embeddings = model.encode(["Example sentence"])
+embeddings = model.encode(["Jeg elsker kage"])
 ```
 
 ### Using Sentence Transformers
@@ -52,31 +55,55 @@ from sentence_transformers import SentenceTransformer
 model = SentenceTransformer("andersborges/model2vecdk")
 
 # Compute text embeddings
-embeddings = model.encode(["Example sentence"])
+embeddings = model.encode(["Jeg elsker kage"])
 ```
 
-### Distilling a Model2Vec model
+## How it works
 
-You can distill a Model2Vec model from a Sentence Transformer model using the `distill` method. First, install the `distill` extra with `pip install model2vec[distill]`. Then, run the following code:
+Model2vec creates a small, fast, and powerful model that outperforms other static embedding models by a large margin on all tasks we could find, while being much faster to create than traditional static embedding models such as GloVe. Best of all, you don't need any data to distill a model using Model2Vec.
 
-```python
-from model2vec.distill import distill
+It works by passing a vocabulary through a sentence transformer model, then reducing the dimensionality of the resulting embeddings using PCA, and finally weighting the embeddings using [SIF weighting](https://openreview.net/pdf?id=SyK00v5xx). During inference, we simply take the mean of all token embeddings occurring in a sentence.
+
+## Training
 
-# Distill a Sentence Transformer model, in this case the BAAI/bge-base-en-v1.5 model
-m2v_model = distill(model_name="BAAI/bge-base-en-v1.5", pca_dims=256)
+See [repo](https://github.com/andersborges/dkmodel2vec). The model was trained with the following commands: 
 
-# Save the model
-m2v_model.save_pretrained("m2v_model")
+```bash 
+# distill model 
+python scripts/hyperparams.py --output-dim 256 --sif-coefficient 0.0005 --strip-upper-case --strip-exotic --focus-pca --normalize-embeddings --vocab-size 150000
+
+# dump features
+python scripts/featurize.py --max-means 100000 --max-length 800
+
+#fine tune
+python scripts/finetune.py --model2vec-model-name scripts/models/dk-llm2vec-model2vec-dim256_sif0.0005_strip_upper_case_strip_exotic_focus_pca_normalize_embeddings --data-path features/features_100000_max_length_800 --lr 0.0001
 ```
 
-## How it works
+## Evaluation
+The model was evaluated on the 10% of unseen data from the DDSC/nordic-embedding-training-data which contains examples of triplets containing a query, a positive (relevant) document and a negative (not relevant) document. The model achieved the following performance: 
+
+accuracy
+
+model2vecdk : 0.867
+BM25: 0.882
+multilingual-e5-large-instruct: 0.963
+
+The model was also evaluated using the [Scandinavian Embedding Benchmark](https://kennethenevoldsen.github.io/scandinavian-embedding-benchmark/) and achieved the following performance: 
+
+| Rank | Model                          | Average Score | Average Rank | Angry Tweets | Bornholm Parallel | DKHate | Da Political Comments | DanFEVER | LCC   | Language Identification | Massive Intent | Massive Scenario | ScaLA | TV2Nord Retrieval | Twitterhjerne |
+|------|--------------------------------|---------------|--------------|--------------|--------------------|--------|------------------------|----------|-------|--------------------------|----------------|------------------|--------|---------------------|----------------|
+| 1    | TTC-L2V-supervised-2           | 0.68          | 4.75         | 67.09        | 54.59             | 69.00  | 45.84                 | 38.31    | 73.67 | 88.61                   | 74.80         | 78.35           | 53.04 | 92.79              | 85.02         |
+| 2    | multilingual-e5-large-instruct | 0.66          | 7.75         | 64.57        | 55.02             | 67.14  | 45.33                 | 39.52    | 70.60 | 82.48                   | 71.89         | 77.51           | 50.18 | 93.69              | 77.23         |
+| 3    | text-embedding-3-large         | 0.64          | 8.92         | 57.80        | 43.34             | 70.21  | 43.41                 | 39.61    | 58.07 | 79.74                   | 69.27         | 75.92           | 50.69 | 95.20              | 81.08         |
+| 42   | dfm-encoder-small-v1 (SimCSE)  | 0.42          | 33.54        | 51.92        | 40.82             | 60.00  | 35.25                 | 16.99    | 58.53 | 50.50                   | 47.92         | 52.95           | 51.36 | 22.28              | 20.02         |
+| 43   | **model2vecdk**                | 0.42          | 36.62        | 48.19        | 7.83              | 59.73  | 32.40                 | 26.04    | 47.67 | 63.97                   | 51.23         | 60.87           | 50.18 | 55.47              | 20.19         |
+| 44   | xlm-roberta-large              | 0.40          | 35.92        | 51.74        | 4.34              | 60.21  | 31.85                 | 10.62    | 48.73 | 81.29                   | 47.26         | 49.55           | 60.29 | 6.11               | 20.39         |
 
-Model2vec creates a small, fast, and powerful model that outperforms other static embedding models by a large margin on all tasks we could find, while being much faster to create than traditional static embedding models such as GloVe. Best of all, you don't need any data to distill a model using Model2Vec.
 
-It works by passing a vocabulary through a sentence transformer model, then reducing the dimensionality of the resulting embeddings using PCA, and finally weighting the embeddings using [SIF weighting](https://openreview.net/pdf?id=SyK00v5xx). During inference, we simply take the mean of all token embeddings occurring in a sentence.
 
 ## Additional Resources
 
+- [Repo used to finetune](https://github.com/andersborges/dkmodel2vec)
 - [Model2Vec Repo](https://github.com/MinishLab/model2vec)
 - [Model2Vec Base Models](https://huggingface.co/collections/minishlab/model2vec-base-models-66fd9dd9b7c3b3c0f25ca90e)
 - [Model2Vec Results](https://github.com/MinishLab/model2vec/tree/main/results)