Create README.md

README.md

@@ -0,0 +1,53 @@
+---
+license: other
+license_name: identity
+task_categories:
+- text-generation
+language:
+- en
+tags:
+- identity
+pretty_name: Identity
+size_categories:
+- 1K<n<10K
+---
+# Identity
+
+Do your models keep referring to themselves by the wrong name? This dataset is an attempt to train models to have a consistent identity.
+
+This dataset contains around 1000 short conversations where the user asks about the identity of the assistant. The assistant replies with its name and developer. You can customize the dataset for your own model.
+
+## Usage
+
+**Important!** Do not use an unmodified version of this dataset to train your models. Make sure to run the code below to generate dataset with the correct identity.
+
+```python
+# Replace with your model's identity
+MODEL_NAME = 'Assistant'
+MODEL_DEVELOPER = 'Developer'
+RESULT_DATASET_ID = '<your_username>/<repo_id>' # Where do you want to push the dataset?
+
+# Load the dataset
+from datasets import load_dataset
+
+dataset = load_dataset('mrfakename/identity', split='train')
+
+# Replace instances of "Assistant" and "Company" with identity
+def replace_identity(example):
+    messages = example['conversations']
+    for message in messages:
+        message['value'] = message['value'].replace('Assistant', MODEL_NAME).replace('Developer', MODEL_DEVELOPER)
+    return {'conversations': messages}
+
+# Apply the identity
+dataset = dataset.map(replace_identity)
+
+# Push to the hub
+dataset.push_to_hub(RESULT_DATASET_ID)
+```
+
+## License
+
+This dataset is free to use. If you redistribute a modified (or unmodified) version of this dataset, please include a link back to this original dataset. If you train a model on this dataset, attribution is not required but much appreciated.
+
+THE DATASET IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS DATASET INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS DATASET.