Training in progress, epoch 1

adapter_config.json

@@ -17,7 +17,7 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 2048,
+  "lora_alpha": 1024,
   "lora_bias": false,
   "lora_dropout": 0,
   "megatron_config": null,
@@ -25,17 +25,17 @@
   "modules_to_save": null,
   "peft_type": "LORA",
   "qalora_group_size": 16,
-  "r": 1024,
+  "r": 512,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_proj",
-    "up_proj",
-    "gate_proj",
-    "k_proj",
     "v_proj",
     "o_proj",
-    "down_proj"
+    "q_proj",
+    "gate_proj",
+    "down_proj",
+    "up_proj",
+    "k_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:b7ff84e1f218c5d00667449a74c604e8032f3d5442038f84940bfc4f776d41ea
-size 4462793824
+oid sha256:05c4bf345d8acf3f0123aacf9a4053c2d407913f83e0dfca15ce0a4011166327
+size 2231423512

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c2244a6150dc4b01c456eb4f06b599dabe8698c7f43fb635579ae6998510109b
+oid sha256:f6438b892ae79a88939463ca25c200452f27d98f5d5ca0e273eb767a9c93584d
 size 5777