Training in progress, epoch 1

adapter_config.json

@@ -13,7 +13,7 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 32,
+  "lora_alpha": 256,
   "lora_dropout": 0.1,
   "megatron_config": null,
   "megatron_core": "megatron.core",
@@ -21,14 +21,14 @@
     "ln_f"
   ],
   "peft_type": "LORA",
-  "r": 32,
+  "r": 256,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
-    "k_proj",
     "q_proj",
-    "o_proj"
+    "v_proj",
+    "o_proj",
+    "k_proj"
   ],
   "task_type": null,
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:41d28824520a305e8365af4719edd1231faaf3d6a8410262b8ca53f222b4ea0f
-size 109086416
+oid sha256:5235467a8933e7266c5ef24ad27a398f2e5d73a29225f9662e02fda96ee665e0
+size 872450448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5eccc64710bbdc352c8dbdc8b532e8d6f1e026eee3ef088f7b093fedb52d57b4
+oid sha256:1d0210910462297dbecab254ac3c1de4e7f13d537650595acfe998b383a450ce
 size 4984