DigitalLearningGmbH
/

educa-ai-nemo-dpo

@@ -1,199 +1,101 @@
 ---
 library_name: transformers
-tags: []
 ---
-# Model Card for Model ID
-<!-- Provide a quick summary of what the model is/does. -->
 ## Model Details
 ### Model Description
-<!-- Provide a longer summary of what this model is. -->
-This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
-- **Developed by:** [More Information Needed]
-- **Funded by [optional]:** [More Information Needed]
-- **Shared by [optional]:** [More Information Needed]
-- **Model type:** [More Information Needed]
-- **Language(s) (NLP):** [More Information Needed]
-- **License:** [More Information Needed]
-- **Finetuned from model [optional]:** [More Information Needed]
-### Model Sources [optional]
-<!-- Provide the basic links for the model. -->
-- **Repository:** [More Information Needed]
-- **Paper [optional]:** [More Information Needed]
-- **Demo [optional]:** [More Information Needed]
 ## Uses
-<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
-### Direct Use
-<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
-[More Information Needed]
-### Downstream Use [optional]
-<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
-[More Information Needed]
-### Out-of-Scope Use
-<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
-[More Information Needed]
 ## Bias, Risks, and Limitations
-<!-- This section is meant to convey both technical and sociotechnical limitations. -->
-[More Information Needed]
-### Recommendations
-<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
-Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
 ## How to Get Started with the Model
-Use the code below to get started with the model.
-[More Information Needed]
 ## Training Details
-### Training Data
-<!-- This should link to a Dataset 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. -->
-[More Information Needed]
-### 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 [optional]
-[More Information Needed]
-#### Training Hyperparameters
-- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
-#### Speeds, Sizes, Times [optional]
-<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
-[More Information Needed]
 ## Evaluation
-<!-- This section describes the evaluation protocols and provides the results. -->
-### Testing Data, Factors & Metrics
-#### Testing Data
-<!-- This should link to a Dataset Card if possible. -->
-[More Information Needed]
-#### Factors
-<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
-[More Information Needed]
-#### Metrics
-<!-- These are the evaluation metrics being used, ideally with a description of why. -->
-[More Information Needed]
-### Results
-[More Information Needed]
-#### Summary
-## 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:** [More Information Needed]
-- **Hours used:** [More Information Needed]
-- **Cloud Provider:** [More Information Needed]
-- **Compute Region:** [More Information Needed]
-- **Carbon Emitted:** [More Information Needed]
-## Technical Specifications [optional]
-### Model Architecture and Objective
-[More Information Needed]
-### Compute Infrastructure
-[More Information Needed]
-#### Hardware
-[More Information Needed]
-#### Software
-[More Information Needed]
-## Citation [optional]
-<!-- 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]
 ## Model Card Contact
-[More Information Needed]

 ---
 library_name: transformers
+license: apache-2.0
+language:
+- en
+- fr
+- de
+- es
+- it
+- pt
+- ru
+- zh
+- ja
+base_model:
+- DigitalLearningGmbH/educa-ai-nemo-sft
 ---
+# Model Card for educa-ai-nemo-dpo
 ## Model Details
 ### Model Description
+`educa-ai-nemo-dpo` is the preference-aligned version of our SFT model [DigitalLearningGmbH/educa-ai-nemo-sft](https://huggingface.co/DigitalLearningGmbH/educa-ai-nemo-sft),
+using our internal dataset which contains a unique mix of German and English preference data covering a multitude of domains.
+In its creation we have paid special attention to data points that can improve performance in German, especially the educational field (text analysis, supporting students in completing textual tasks, ...).
+This is a preliminary release and subject to changes or updates.
+- **Developed by:** [Digital Learning GmbH](https://huggingface.co/DigitalLearningGmbH)
+- **Funded by [optional]:** [Digital Learning GmbH](https://huggingface.co/DigitalLearningGmbH)
+- **Shared by [optional]:** [Digital Learning GmbH](https://huggingface.co/DigitalLearningGmbH)
+- **Model type:** Transformer Decoder LLM
+- **Language(s) (NLP):** English, French, German, Spanish, Italian, Portuguese, Russian, Chinese, Japanese
+- **License:** [Apache License 2.0](https://choosealicense.com/licenses/apache-2.0/)
+- **Finetuned from model:** [DigitalLearningGmbH/educa-ai-nemo-sft](https://huggingface.co/DigitalLearningGmbH/educa-ai-nemo-sft)
 ## Uses
+As stated before, this is a preliminary release and we are still benchmarking the model as well as improving our datasets for possible further training.
+As such, we do not recommend using this model in a production setting yet and are looking forward to engaging with the community regarding possible downstream uses and improvements.
 ## Bias, Risks, and Limitations
+Refer to the [original model card](https://huggingface.co/mistralai/Mistral-Nemo-Instruct-2407) for an overview of the general risks associated with using this model.
+As this version is only fine-tuned using SFT without any preference alignment, the model may output harmful data. Use is at your own discretion, taking into account the potential risks.
 ## How to Get Started with the Model
+Refer to the [original model card](https://huggingface.co/mistralai/Mistral-Nemo-Instruct-2407) for code examples.
+Be aware that this model uses a slightly different chat template from the original: system prompts are placed before the first user prompt (before the first instance of `[INST]`).
+We include the updated template in the tokenizer config, so you can use `tokenizer.apply_chat_template`.
 ## Training Details
+Instead of standard sigmoid DPO Loss, we used [DPO-Positive](https://arxiv.org/abs/2402.13228) as we found it improved training stability and overall performance with our dataset.
+### Training Data
+The model has been trained on a mix of some publically-available and permissively-licensed data as well as a majority of unique internal datasets which we have created.
+Our data encompasses examples of a length up to 16384 tokens, further enhancing the model's long-context capability.
 ## Evaluation
+We ran all benchmarks using [lm-eval](https://github.com/EleutherAI/lm-evaluation-harness) with `--apply_chat_template`.
+For comparison, we performed the same benchmarks on the base model as well, in the exact same environment with the same parameters.
+### English Benchmarks
+| Benchmark | Mistral-Nemo-Instruct-2407 | educa-ai-nemo-dpo |
+| --- | --- | --- |
+| hellaswag (acc_norm) | 71.9% | **77.6%** |
+| winogrande (acc) | 69.8% | **75.2%** |
+| openbookqa (acc_norm) | 45.8% | **47.0%** |
+| commonsense_qa (acc) | 74.4% | **75.4%** |
+| truthfulqa_mc1 (acc) | 39.66% | **41.5%** |
+| mmlu (acc) | 64.9% | **66.5%** |
+| triviaqa (exact_match) | 12.3% | **23.99%** |
+| agieval (acc) | 36.6% | **39.1%** |
+| arc_challenge (acc_norm) | 52.5% | **54.4%** |
+| arc_easy (acc_norm) | 74.1% | **76.0%** |
+| piqa (acc_norm) | 78.9% | **81.5%** |
+| leaderboard_bbh (acc_norm) | 49.1% | **53.0%** |
+| leaderboard_gpqa (acc_norm) | **30.6%** | 29.4% |
+| leaderboard_ifeval (inst_level_loose_acc) | 72.8% | **75.1%** |
+| leaderboard_mmlu_pro (acc) | **35.1%** | 33.67% |
+| leaderboard_musr (acc_norm) | 39.3% | **40.2%** |
+### Multilingual Benchmarks
+... coming soon!
 ## Model Card Authors [optional]
+This model card was written by [Lennard Michael Strohmeyer](https://huggingface.co/LenDigLearn)
 ## Model Card Contact
+[Lennard Michael Strohmeyer](https://huggingface.co/LenDigLearn)