Create README.md

README.md

@@ -0,0 +1,18 @@
+## Model
+base_model : yanolja/KoSOLAR-10.7B-v0.2
+training objective: freeze, instruction Tuning 
+## Dataset
+공개 데이터 수집
+Deduplicating Training Data Makes Language Models Better 알고리즘 활용
+instruction version 1.4
+
+## Code
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
+
+model_name = "jjingyeom/freeze_KoSoLAR-10.7B-v0.2_1.4_dedup"
+model = AutoModelForCausalLM.from_pretrained(
+        model_name,
+)
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+