Update README.md
Browse files
README.md
CHANGED
|
@@ -21,10 +21,6 @@ tags:
|
|
| 21 |
- Reasoning
|
| 22 |
---
|
| 23 |
|
| 24 |
-
|
| 25 |
-
<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'"/>
|
| 26 |
-
|
| 27 |
-
|
| 28 |
# German-RAG-MISTRAL-7B-v3.0-ORPO-HESSIAN-AI
|
| 29 |
|
| 30 |
<!-- Provide a quick summary of what the model is/does. -->
|
|
@@ -253,6 +249,7 @@ It's important to note that the actual power consumption may vary depending on t
|
|
| 253 |
| Model | GPU Type | Power Consumption From GPUs |
|
| 254 |
|----------------|---------------------|-----------------------------|
|
| 255 |
| German-RAG-MISTRAL-ORPO | A100 ([Hessian AI supercomputer](https://hessian.ai/de/)) | 0.01843 MWh |
|
|
|
|
| 256 |
## Bias, Risks, and Limitations
|
| 257 |
|
| 258 |
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.
|
|
@@ -262,12 +259,6 @@ Otherwise, many facts from German-RAG-MISTRAL-ORPO or any LLM will often not be
|
|
| 262 |
|
| 263 |
|
| 264 |
|
| 265 |
-
|
| 266 |
-
## Model Card Contact
|
| 267 |
-
|
| 268 |
-
|
| 269 |
-
For errors in this model card, please contact ([[email protected]](mailto:[email protected])).
|
| 270 |
-
|
| 271 |
## The German-RAG AI Team
|
| 272 |
[Marcel Rosiak](https://de.linkedin.com/in/marcel-rosiak)
|
| 273 |
[Soumya Paul](https://de.linkedin.com/in/soumya-paul-1636a68a)
|
|
|
|
| 21 |
- Reasoning
|
| 22 |
---
|
| 23 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
# German-RAG-MISTRAL-7B-v3.0-ORPO-HESSIAN-AI
|
| 25 |
|
| 26 |
<!-- Provide a quick summary of what the model is/does. -->
|
|
|
|
| 249 |
| Model | GPU Type | Power Consumption From GPUs |
|
| 250 |
|----------------|---------------------|-----------------------------|
|
| 251 |
| German-RAG-MISTRAL-ORPO | A100 ([Hessian AI supercomputer](https://hessian.ai/de/)) | 0.01843 MWh |
|
| 252 |
+
|
| 253 |
## Bias, Risks, and Limitations
|
| 254 |
|
| 255 |
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.
|
|
|
|
| 259 |
|
| 260 |
|
| 261 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 262 |
## The German-RAG AI Team
|
| 263 |
[Marcel Rosiak](https://de.linkedin.com/in/marcel-rosiak)
|
| 264 |
[Soumya Paul](https://de.linkedin.com/in/soumya-paul-1636a68a)
|