Upload LoRA adapter (README written by author)

README.md

@@ -35,8 +35,8 @@ while intermediate reasoning (Chain-of-Thought) is masked.
 - Method: QLoRA (4-bit)
 - Max sequence length: 512
 - Epochs: 2
-- Learning rate: 2e-05
-- LoRA: r=32, alpha=64
+- Learning rate: 2e-06
+- LoRA: r=64, alpha=128
 
 ## Usage
 

adapter_config.json

@@ -20,25 +20,25 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 64,
+  "lora_alpha": 128,
   "lora_bias": false,
-  "lora_dropout": 0.05,
+  "lora_dropout": 0.0,
   "megatron_config": null,
   "megatron_core": "megatron.core",
   "modules_to_save": null,
   "peft_type": "LORA",
   "peft_version": "0.18.1",
   "qalora_group_size": 16,
-  "r": 32,
+  "r": 64,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "gate_proj",
+    "q_proj",
+    "k_proj",
     "down_proj",
     "o_proj",
+    "gate_proj",
     "v_proj",
-    "q_proj",
-    "k_proj",
     "up_proj"
   ],
   "target_parameters": null,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:da34e792639157d8c4fc8eb3c2977d01bca0e596e2168befbcbdb567f4d247d8
-size 264308896
+oid sha256:376d63db2b557d45c6a8807513bb5e26d6032e67ab7d2a94b5531dce76471842
+size 528550256