Update README.md

README.md

@@ -1,60 +1,59 @@
 ---
+datasets:
+- jondurbin/gutenberg-dpo-v0.1
+- Qwen/Qwen2.5-14B-Instruct
+- HuggingFaceH4/ultrafeedback_binarized
 base_model:
-- v000000/Qwen2.5-14B-Gutenberg-1e-Delta
-- v000000/qwen_results37
-- v000000/Qwen2.5-14B-Gutenberg-Instruct-Slerpeno
-- v000000/Qwen2.5-14B-Gutenberg-UltraLambda-Slerpeno
-- v000000/Qwen2.5-14B-Gutenberg-0.25e-Early-STACKED-0.37e
 - Qwen/Qwen2.5-14B-Instruct
-- v000000/qwen_results100
 - v000000/Qwen2.5-14B-Gutenberg-1e-Delta
-- v000000/Qwen2.5-14B-Gutenberg-0.25e-Early
 - tanliboy/lambda-qwen2.5-14b-dpo-test
-- Qwen/Qwen2.5-14B-Instruct
-- v000000/qwen_results37
-- v000000/Qwen2.5-14B-Gutenberg-0.25e-Early-STACKED-0.37e
-- v000000/qwen_results100
 library_name: transformers
 tags:
-- mergekit
-- merge
-
+- qwen
+- qwen2.5
+- finetune
+- dpo
+- qwen2
+- chat
+- conversational
+- instruct
+- storywriting
+- roleplay
+license: apache-2.0
+language:
+- en
+pipeline_tag: text-generation
 ---
-# merge
 
-This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
+# Qwen2.5-Lumen-14B
 
-## Merge Details
-### Merge Method
+* *Direct preference optimization finetuned for 3 epoch and mass checkpoint merged, Based on Qwen2.5-14B-Instruct.*
 
-This model was merged using the [Model Stock](https://arxiv.org/abs/2403.19522) merge method using [v000000/Qwen2.5-14B-Gutenberg-1e-Delta](https://huggingface.co/v000000/Qwen2.5-14B-Gutenberg-1e-Delta) as a base.
+-------------------------------------------------------------------------------
 
-### Models Merged
+## Training Notes
 
-The following models were included in the merge:
-* [v000000/Qwen2.5-14B-Gutenberg-1e-Delta](https://huggingface.co/v000000/Qwen2.5-14B-Gutenberg-1e-Delta) + [v000000/qwen_results37](https://huggingface.co/v000000/qwen_results37)
-* [v000000/Qwen2.5-14B-Gutenberg-Instruct-Slerpeno](https://huggingface.co/v000000/Qwen2.5-14B-Gutenberg-Instruct-Slerpeno)
-* [v000000/Qwen2.5-14B-Gutenberg-UltraLambda-Slerpeno](https://huggingface.co/v000000/Qwen2.5-14B-Gutenberg-UltraLambda-Slerpeno)
-* [v000000/Qwen2.5-14B-Gutenberg-0.25e-Early-STACKED-0.37e](https://huggingface.co/v000000/Qwen2.5-14B-Gutenberg-0.25e-Early-STACKED-0.37e)
-* [Qwen/Qwen2.5-14B-Instruct](https://huggingface.co/Qwen/Qwen2.5-14B-Instruct) + [v000000/qwen_results100](https://huggingface.co/v000000/qwen_results100)
-* [v000000/Qwen2.5-14B-Gutenberg-0.25e-Early](https://huggingface.co/v000000/Qwen2.5-14B-Gutenberg-0.25e-Early)
-* [tanliboy/lambda-qwen2.5-14b-dpo-test](https://huggingface.co/tanliboy/lambda-qwen2.5-14b-dpo-test)
-* [Qwen/Qwen2.5-14B-Instruct](https://huggingface.co/Qwen/Qwen2.5-14B-Instruct) + [v000000/qwen_results37](https://huggingface.co/v000000/qwen_results37)
-* [v000000/Qwen2.5-14B-Gutenberg-0.25e-Early-STACKED-0.37e](https://huggingface.co/v000000/Qwen2.5-14B-Gutenberg-0.25e-Early-STACKED-0.37e) + [v000000/qwen_results100](https://huggingface.co/v000000/qwen_results100)
+Trained [Qwen2.5-14B-Instruct] for 2 epochs on [jondurbin/gutenberg-dpo-v0.1] saving different checkpoints along the way.
+
+[Tanliboy](https://huggingface.co/tanliboy) trained [Qwen2.5-14B-Instruct] for 1 epoch on [HuggingFaceH4/ultrafeedback_binarized].
+
+## Merge
 
-### Configuration
+* Merged with a sophosympatheia <b>SLERP</b> *Ultrafeedback-Binarized DPO* and *Gutenberg DPO*
+* Merged with a sophosympatheia <b>SLERP</b> *Qwen2.5-14B-Instruct* and *Gutenberg DPO*
+* Merged all DPO <b>checkpoints</b> and <b>SLERP</b> variations with <b>MODEL_STOCK</b> to analyze geometric properties and get the best of all runs/merges.
 
-The following YAML configuration was used to produce this model:
+## Recipe
 
 ```yaml
 models:
   - model: v000000/Qwen2.5-14B-Gutenberg-1e-Delta
-  - model: v000000/Qwen2.5-14B-Gutenberg-0.25e-Early-STACKED-0.37e #(Qwen2.5-14B-Gutenberg-0.6e-Sequential)
+  - model: v000000/Qwen2.5-14B-Gutenberg-0.6e-Sequential
   - model: v000000/Qwen2.5-14B-Gutenberg-0.25e-Early
-  - model: v000000/Qwen2.5-14B-Gutenberg-1e-Delta+v000000/qwen_results37 #(Qwen2.5-14B-Gutenberg-2e-Sequential)
-  - model: Qwen/Qwen2.5-14B-Instruct+v000000/qwen_results37 #(Qwen2.5-14B-Gutenberg-0.37e-Early)
-  - model: v000000/Qwen2.5-14B-Gutenberg-0.25e-Early-STACKED-0.37e+v000000/qwen_results100 #(Qwen2.5-14B-Gutenberg-1e-Zeta)
-  - model: Qwen/Qwen2.5-14B-Instruct+v000000/qwen_results100 #(Qwen2.5-14B-Gutenberg-1e-Theta)
+  - model: v000000/Qwen2.5-14B-Gutenberg-2e-Sequential
+  - model: v000000/Qwen2.5-14B-Gutenberg-0.37e-Early
+  - model: v000000/Qwen2.5-14B-Gutenberg-2e-Zeta
+  - model: v000000/Qwen2.5-14B-Gutenberg-1e-Theta
   - model: tanliboy/lambda-qwen2.5-14b-dpo-test
   - model: v000000/Qwen2.5-14B-Gutenberg-1e-Delta
   - model: tanliboy/lambda-qwen2.5-14b-dpo-test
@@ -63,5 +62,24 @@ models:
 base_model: v000000/Qwen2.5-14B-Gutenberg-1e-Delta
 merge_method: model_stock
 dtype: bfloat16
-
 ```
+
+### Finetune and merge
+
+This is a merge and finetune of pre-trained language models.
+
+###  Models Merged
+
+[Arxiv 2403.19522](https://arxiv.org/abs/2403.19522)
+
+The following models were included in the merge:
+* v000000/Qwen2.5-14B-Gutenberg-1e-Delta
+* v000000/Qwen2.5-14B-Gutenberg-0.6e-Sequential
+* v000000/Qwen2.5-14B-Gutenberg-0.25e-Early
+* v000000/Qwen2.5-14B-Gutenberg-2e-Sequential
+* v000000/Qwen2.5-14B-Gutenberg-0.37e-Early
+* v000000/Qwen2.5-14B-Gutenberg-2e-Zeta
+* v000000/Qwen2.5-14B-Gutenberg-1e-Theta
+* v000000/Qwen2.5-14B-Gutenberg-UltraLambda-Slerpeno
+* v000000/Qwen2.5-14B-Gutenberg-Instruct-Slerpeno
+* tanliboy/lambda-qwen2.5-14b-dpo-test