Update README.md

README.md

@@ -1,13 +1,13 @@
 ---
 license: llama3.1
 datasets:
-- avemio/GRAG-CPT-HESSIAN-AI
-- avemio/GRAG-SFT-ShareGPT-HESSIAN-AI
+- avemio/German_RAG-CPT-HESSIAN-AI
+- avemio/German_RAG-SFT-ShareGPT-HESSIAN-AI
 language:
 - en
 - de
 base_model:
-- avemio/GRAG-LLAMA-3.1-8B-CPT-HESSIAN-AI
+- avemio/German_RAG-LLAMA-3.1-8B-CPT-HESSIAN-AI
 pipeline_tag: question-answering
 tags:
 - German
@@ -19,25 +19,25 @@ tags:
 ---
 
 
-<img src="https://www.grag.ai/wp-content/uploads/2024/12/GRAG-ICON-TO-WORDLOGO-Animation_Loop-small-ezgif.com-video-to-gif-converter.gif" alt="GRAG Logo" width="400" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
+<img src="https://www.German_RAG.ai/wp-content/uploads/2024/12/German_RAG-ICON-TO-WORDLOGO-Animation_Loop-small-ezgif.com-video-to-gif-converter.gif" alt="German_RAG Logo" width="400" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
 
 
-# GRAG-LLAMA-3.1-8B-SFT-HESSIAN-AI
+# German_RAG-LLAMA-3.1-8B-SFT-HESSIAN-AI
 
 <!-- Provide a quick summary of what the model is/does. -->
 
-**GRAG** (**G**erman **R**etrieval **A**ugmented **G**eneration) models are designed for the German-speaking market, enabling innovation and AI solutions to drive German research collaboration in business-focused Generative AI by 2025
+**German_RAG** (**G**erman **R**etrieval **A**ugmented **G**eneration) models are designed for the German-speaking market, enabling innovation and AI solutions to drive German research collaboration in business-focused Generative AI by 2025
 
-Our GRAG-LLAMA-SFT model are trained on this **[GRAG-SFT](https://huggingface.co/datasets/avemio/GRAG-SFT-ShareGPT-HESSIAN-AI) dataset.**
+Our German_RAG-LLAMA-SFT model are trained on this **[German_RAG-SFT](https://huggingface.co/datasets/avemio/German_RAG-SFT-ShareGPT-HESSIAN-AI) dataset.**
 
 ## Model Details
 
 The core models released in this batch are the following: 
 | Size | Training Tokens | 
 |------|--------|
-| [GRAG-LLAMA-CPT](https://huggingface.co/avemio/GRAG-LLAMA-3.1-8B-CPT-HESSIAN-AI)   | 507.47 million |
-| [GRAG-LLAMA-SFT](https://huggingface.co/avemio/GRAG-LLAMA-3.1-8B-SFT-HESSIAN-AI) |  2.03 billion  |  
-| [GRAG-LLAMA-ORPO](https://huggingface.co/avemio/GRAG-LLAMA-3.1-8B-ORPO-HESSIAN-AI) |  2.0577 billion  | 
+| [German_RAG-LLAMA-CPT](https://huggingface.co/avemio/German_RAG-LLAMA-3.1-8B-CPT-HESSIAN-AI)   | 507.47 million |
+| [German_RAG-LLAMA-SFT](https://huggingface.co/avemio/German_RAG-LLAMA-3.1-8B-SFT-HESSIAN-AI) |  2.03 billion  |  
+| [German_RAG-LLAMA-ORPO](https://huggingface.co/avemio/German_RAG-LLAMA-3.1-8B-ORPO-HESSIAN-AI) |  2.0577 billion  | 
 ### Model Description
 
 <!-- Provide a longer summary of what this model is. -->
@@ -47,19 +47,19 @@ The core models released in this batch are the following:
 - **Model type:** a Transformer style autoregressive language model.
 - **Language(s) (NLP):** German, English
 - **License:** The code and model are released under Apache 2.0.
-- **Contact:** [grag@avemio.digital](mailto:grag@avemio.digital)
+- **Contact:** [German_RAG@avemio.digital](mailto:German_RAG@avemio.digital)
 
 
 ### Model Sources
 
 <!-- Provide the basic links for the model. -->
 
-- **Training Study:** [Training Study](https://avemio.digital/wp-content/uploads/2025/01/GRAG-TRAINING-STUDY-Advancing-German-Language-AI-with-hessian-AI.pdf)
+- **Training Study:** [Training Study](https://avemio.digital/wp-content/uploads/2025/01/German_RAG-TRAINING-STUDY-Advancing-German-Language-AI-with-hessian-AI.pdf)
 - **Repositories:** 
     - Training: [Colab-Notebook](https://colab.research.google.com/drive/18SH_aYLCnw1K7cRGOTTZ80y98V5Kquxb?usp=sharing)
     - Evaluation code: 
-        - [GRAG-LLM-HARD-BENCHMARK](https://github.com/avemio-digital/GRAG-LLM-HARD-BENCHMARK.git)
-        - [GRAG-LLM-EASY-BENCHMARK](https://github.com/avemio-digital/GRAG-LLM-EASY-BENCHMARK.git)
+        - [German_RAG-LLM-HARD-BENCHMARK](https://github.com/avemio-digital/German_RAG-LLM-HARD-BENCHMARK.git)
+        - [German_RAG-LLM-EASY-BENCHMARK](https://github.com/avemio-digital/German_RAG-LLM-EASY-BENCHMARK.git)
 -  **Technical blog post:**
 <!-- - **Press release:** TODO -->
 
@@ -73,7 +73,7 @@ Now, proceed as usual with HuggingFace:
 ```python
 from transformers import AutoModelForCausalLM, AutoTokenizer
  
-model_name = "avemio/GRAG-LLAMA-3.1-8B-SFT-HESSIAN-AI"
+model_name = "avemio/German_RAG-LLAMA-3.1-8B-SFT-HESSIAN-AI"
  
 model = AutoModelForCausalLM.from_pretrained(
     model_name,
@@ -135,7 +135,7 @@ Four evaluation metrics were employed across all subsets: language quality, over
 -   **Overall score:** This metric combined the results from the previous three metrics, offering a comprehensive evaluation of the model's capabilities across all subsets.
 
 
-| Metric                                    | [Vanila-llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct) | **[GRAG-LLAMA-SFT](https://huggingface.co/avemio/GRAG-LLAMA-3.1-8B-SFT-HESSIAN-AI)** | [GRAG-LLAMA-ORPO](https://huggingface.co/avemio/GRAG-LLAMA-3.1-8B-ORPO-HESSIAN-AI) | [GRAG-LLAMA-MERGED]| GPT-3.5-TURBO | 
+| Metric                                    | [Vanila-llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct) | **[German_RAG-LLAMA-SFT](https://huggingface.co/avemio/German_RAG-LLAMA-3.1-8B-SFT-HESSIAN-AI)** | [German_RAG-LLAMA-ORPO](https://huggingface.co/avemio/German_RAG-LLAMA-3.1-8B-ORPO-HESSIAN-AI) | [German_RAG-LLAMA-MERGED]| GPT-3.5-TURBO | 
 |------------------------------------------|---------------------------------------------------------------------------------|--------------------------------------------------------------------------------|------------------------------------------------------------------------------------------------|-----------------------------|----------------|
 | Average Language Quality             |87.78                                                                           |**88.93**                                                                           | 88.93                                                                                          |86.93                             |87.58                |
 | **OVERALL SCORES (weighted):**       |                                                                            |                                                                            |                                                                                           |                             |               |
@@ -149,7 +149,7 @@ Four evaluation metrics were employed across all subsets: language quality, over
 ## Model Details
 
 ### Data
-For training data details, please see the [GRAG-SFT-Dataset](https://huggingface.co/datasets/avemio/GRAG-SFT-ShareGPT-HESSIAN-AI) documentation.
+For training data details, please see the [German_RAG-SFT-Dataset](https://huggingface.co/datasets/avemio/German_RAG-SFT-ShareGPT-HESSIAN-AI) documentation.
 
 #### Description
 The SFT tasks represent a focused approach to enhance model capabilities through specialized RAG examples. Most of these tasks were developed using synthetically enhanced data derived from the German Wikipedia, accessed through Cohere's prepared dataset on HuggingFace (licensed CC-BY-SA 4.0). This data was structured in a training knowledge graph where Question-Answer nodes were connected to both relevant and irrelevant Context nodes from the same Wikipedia page, creating a rich and challenging network of relationships for training. The only exceptions are the function calling dataset, which was derived and extended from Salesforce's XLAM Function calling dataset by including function call results and final answer generation, and the reasoning task which synthetic generation was inspired by the Paper from Tencent ([“Scaling Synthetic Data Creation with 1,000,000,000 Personas”](https://arxiv.org/abs/2406.20094)), to generate a diverse set of reasoning tasks across various domains.
@@ -164,7 +164,7 @@ The implementation of these tasks within RAG systems can significantly improve o
 ### Architecture
 
 
-| Parameter            | GRAG-LLAMA-SFT                                                                                   |
+| Parameter            | German_RAG-LLAMA-SFT                                                                                   |
 |-----------------------|-----------------------------------------------------------------------------------------------|
 | **d_model**          | 3072                                                                                          |
 | **num heads**        | 32                                                                                            |
@@ -182,7 +182,7 @@ The implementation of these tasks within RAG systems can significantly improve o
 ### Hyperparameters 
 
 
-| Parameter                 | GRAG-LLAMA-SFT       |
+| Parameter                 | German_RAG-LLAMA-SFT       |
 |---------------------------|--------------------|
 | **warmup steps**          | 50                |
 | **peak LR**               | 5.0E-07           |
@@ -193,19 +193,19 @@ The implementation of these tasks within RAG systems can significantly improve o
 
 ## Environmental Impact
 
-GRAG-LLAMA-SFT, running on NVIDIA A100 with 40 GPUs for 7 days, has an approximate power consumption as follows:
+German_RAG-LLAMA-SFT, running on NVIDIA A100 with 40 GPUs for 7 days, has an approximate power consumption as follows:
 
 It's important to note that the actual power consumption may vary depending on the specific workload and operational conditions. For accurate power consumption measurements, using dedicated power monitoring tools is recommended.
 
 | Model          | GPU Type            | Power Consumption From GPUs | 
 |----------------|---------------------|-----------------------------|
-| GRAG-LLAMA-SFT   | A100 ([Hessian AI supercomputer](https://hessian.ai/de/)) | 0.02016 MWh                    |
+| German_RAG-LLAMA-SFT   | A100 ([Hessian AI supercomputer](https://hessian.ai/de/)) | 0.02016 MWh                    |
 ## Bias, Risks, and Limitations
 
 Like any base language model or fine-tuned model without safety filtering, it is relatively easy for a user to prompt these models to generate harmful and generally sensitive content.
 Such content can also be produced unintentionally, especially in the case of bias, so we recommend users consider the risks of applications of this technology.
 
-Otherwise, many facts from GRAG-LLAMA-SFT or any LLM will often not be true, so they should be checked.
+Otherwise, many facts from German_RAG-LLAMA-SFT or any LLM will often not be true, so they should be checked.
 
 
 
@@ -213,9 +213,9 @@ Otherwise, many facts from GRAG-LLAMA-SFT or any LLM will often not be true, so
 ## Model Card Contact
 
 
-For errors in this model card, please contact ([grag@avemio.digital](mailto:grag@avemio.digital)).
+For errors in this model card, please contact ([German_RAG@avemio.digital](mailto:German_RAG@avemio.digital)).
 
-## The GRAG AI Team
+## The German_RAG AI Team
 [Marcel Rosiak](https://de.linkedin.com/in/marcel-rosiak)
 [Soumya Paul](https://de.linkedin.com/in/soumya-paul-1636a68a)
 [Siavash Mollaebrahim](https://de.linkedin.com/in/siavash-mollaebrahim-4084b5153?trk=people-guest_people_search-card)