OwenArli commited on
Commit
1c9d0e5
·
verified ·
1 Parent(s): 6275a12

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -1
README.md CHANGED
@@ -4,6 +4,8 @@ license: apache-2.0
4
  # ArliAI-RPMax-12B-v1.2
5
  =====================================
6
 
 
 
7
  ## RPMax Series Overview
8
 
9
  | [2B](https://huggingface.co/ArliAI/Gemma-2-2B-ArliAI-RPMax-v1.1) |
@@ -38,7 +40,7 @@ v1.2 update completely removes non-creative/RP examples in the dataset and is al
38
  * **Sequence Length**: 8192
39
  * **Training Duration**: Approximately 2 days on 2x3090Ti
40
  * **Epochs**: 1 epoch training for minimized repetition sickness
41
- * **QLORA**: 64-rank 128-alpha, resulting in ~2% trainable weights
42
  * **Learning Rate**: 0.00001
43
  * **Gradient accumulation**: Very low 32 for better learning.
44
 
 
4
  # ArliAI-RPMax-12B-v1.2
5
  =====================================
6
 
7
+ ## UPDATE: Merged it wrongly to base after LORA training. Working to reupload. The 8B version meanwhile is working fine.
8
+
9
  ## RPMax Series Overview
10
 
11
  | [2B](https://huggingface.co/ArliAI/Gemma-2-2B-ArliAI-RPMax-v1.1) |
 
40
  * **Sequence Length**: 8192
41
  * **Training Duration**: Approximately 2 days on 2x3090Ti
42
  * **Epochs**: 1 epoch training for minimized repetition sickness
43
+ * **LORA**: 64-rank 128-alpha, resulting in ~2% trainable weights
44
  * **Learning Rate**: 0.00001
45
  * **Gradient accumulation**: Very low 32 for better learning.
46