Create README.md

README.md

@@ -0,0 +1,35 @@
+
+
+# ⛏️ Andy-4-micro-base-LoRA ⛏️
+
+Andy-4-micro is a 1.7b version of Andy-4, it shows off how incredible the Andy-4 dataset can make models, even in tiny packages.
+
+
+## Model Details
+
+This is a Small version of Andy-4, which is 1.7b parameters, and will release in all facets around two weeks from the full Andy-4 release.
+
+This model used an optimal training environment for smaller models, as well as the same datasets as the full Andy-4 model.
+
+**Andy-4:** 7b or 8b parameters
+
+**Andy-4-micro:** 1.7b parameters.
+
+# Additional detail
+
+In preliminary testing of Andy-4-base and Andy-4-micro-base, Andy-4-micro-base showed to be only slightly worse than Andy-4, and sometimes on-par with non-building examples
+
+Andy-4-micro shows off its incredible ability to play when allowed to reason, this is because the model is able to walk through step by step of what it should do, instead of relying on what it "remembers" for what it should do
+
+In the preliminary testing, both non-reasoning and reasoning modes were used. Again, this was using Andy-4-micro-base and Andy-4-base, not the final versions.
+
+Using this LoRA to continue to train a model falls under the ruling of the [Andy License](LICENSE), as in the final model must include `Andy` in it somewhere, and credit to be posted to I, Sweaterdog, in the model repo.
+
+This is a base model, meaning it is not meant to play Minecraft with you, and is only for training updates to the model, as Mindcraft is not finished, we need to have a rolling update system
+
+```
+## License
+This model is licensed under the Andy 1.0 License.  
+Credit: https://huggingface.co/Sweaterdog  
+Acknowledgment: This work uses data and models created by @Sweaterdog.
+```