Create README.md

README.md

@@ -0,0 +1,125 @@
+---
+license: mit
+datasets:
+- wikipedia
+language:
+- en
+metrics:
+- glue
+---
+# Model Card for SzegedAI/bert-medium-mlsm
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+This medium-sized BERT model was created using the [Masked Latent Semantic Modeling] (MLSM) pre-training objective, which is a sample efficient alternative for classic Masked Language Modeling (MLM).  
+During MLSM, the objective is to recover the latent semantic profile of the masked tokens, as opposed to recovering their exact identity.  
+The contextualized latent semantic profile during pre-training is determined by performing sparse coding of the hidden representation of an already pre-trained model (a base-sized BERT model in this particular case). 
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+- **Developed by:** SzegedAI
+- **Model type:** transformer encoder
+- **Language:** Engish
+- **License:** MIT
+
+### Model Sources
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [https://github.com/szegedai/MLSM](https://github.com/szegedai/MLSM)
+- **Paper:** [Masked Latent Semantic Modeling: an Efficient Pre-training Alternative to Masked Language Modeling](https://underline.io/events/395/posters/15279/poster/78046-masked-latent-semantic-modeling-an-efficient-pre-training-alternative-to-masked-language-modeling?tab=abstract+%26+voting)
+
+## How to Get Started with the Model
+
+The pre-trained model can be used in the usual manner, e.g., for fine tuning on a particular sequence classification task, invoke the code:
+
+```
+from transformers import AutoTokenizer, AutoModelForSequenceClassification
+
+tokenizer = AutoTokenizer.from_pretrained('SzegedAI/bert-medium-mlsm')
+model = AutoModelForSequenceClassification.from_pretrained('SzegedAI/bert-medium-mlsm')
+```
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+The model was pre-trained using a 2022 English Wikipedia dump pre-processed with [wiki-bert-pipeline](https://github.com/spyysalo/wiki-bert-pipeline).
+
+### Training Procedure 
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing
+
+
+#### Training Hyperparameters
+
+Pre-training was conducted with a batch size of 32 sequences and a gradient accumulation over 32 batches, resulting in an effective batch size of 1024.  
+A total of 300,000 update steps were performed using the AdamW optimizer with a linear learning rate scheduling having a peak learning rate of 1e-04.
+A maximum sequence length of 128 tokens was employed over the first 90% of the pre-training, while for the final 10% of the pre-training, the maximum sequence length was increased to 512 tokens.
+
+- **Training regime:** fp32
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+The model was evaluated on GLUE tasks and CoNLL2003 for named entity recognition.
+
+### Results
+The evaluation result after fine-tuning the given model on a wide range of tasks.  
+On each tasks 10 different fine-tuning were performed, during which the only difference was the random initialization of the task-specific classification head.
+Both the average and the standard deviation are displayed below on each tasks.
+
+| Dataset | Metric | Avg. | Std. |
+|---|---|---|---|
+| CoLA | Matthews correlation | 0.403 | 0.012 |
+| CoNLL2003 | F1 | 0.926  | 0.003 |
+| MNLI (matched) | Accuracy | 0.798 | 0.001 |
+| MNLI (mismatched) | Accuracy | 0.808 | 0.002 |
+| MRPC | Accuracy | 0.786 | 0.020 |
+| MRPC | F1 | 0.851 | 0.013 |
+| QNLI | Accuracy | 0.870 | 0.004 | 
+| QQP | Accuracy | 0.892 | 0.001 |
+| QQP | F1 | 0.855 | 0.001 |
+| RTE | Accuracy  | 0.571 | 0.011 |
+| SST2 | Accuracy | 0.905 | 0.004 |
+| STSB | Pearson correlation | 0.818 | 0.024 |
+| STSB | Spearman correlation | 0.820 | 0.021 |
+| WiC | Accuracy | 0.639 | 0.007 |
+| Average | --- | 0.7815 | --- |
+#### Summary
+
+This model was more sample efficient and reached practically the same average performance as an alternatively pre-trained language model of 2.5 times more parameter (of base size)
+that was pre-trained using the classical MLM objective.
+
+## 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 A6000
+- **Hours used:** 300
+- **Carbon Emitted:** 42 kg CO2 eq.
+
+## Citation
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+The pre-training objective is introduced in the ACL Findings paper _Masked Latent Semantic Modeling: an Efficient Pre-training Alternative to Masked Language Modeling_.
+
+**BibTeX:**
+
+
+