v7: r=96, alpha=192, lr=1e-6, max_len=1024, epoch=2 + DPO (README written by author)

README.md

@@ -28,7 +28,7 @@ This model has been optimized using DPO to align its responses with preferred ou
 **Stage 1 - SFT:**
 - **Dataset**: u-10bei/structured_data_with_cot_dataset_512_v2
 - **Max sequence length**: 1024
-- **Epochs**: 1
+- **Epochs**: 2
 - **Learning rate**: 1e-06
 - **LoRA**: r=96, alpha=192
 

adapter_config.json

@@ -33,13 +33,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
-    "o_proj",
-    "q_proj",
     "down_proj",
+    "q_proj",
     "up_proj",
     "k_proj",
-    "gate_proj"
+    "v_proj",
+    "gate_proj",
+    "o_proj"
   ],
   "target_parameters": null,
   "task_type": "CAUSAL_LM",

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c12a441c40d6d2a74058fb2a4c5354e9eb9e8184f16a61809ddadb49a12b3abd
+oid sha256:03e307618d3940106e51aedd1079b3c177b4018ff3748df585c2992b3582a7e0
 size 792791480