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 sha256:0d69199ae8b2ac047adfd247f267b47ba6365cb7dc6031e0bcffd39d6b606257
-size 109086416
+oid sha256:0c83fe5e09e380287de104c736f9423ffde554f8b9ecd25509d5649ee4432c19
+size 872450448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3c525a042cb50892e9f8f7648ef5d913a9fcb7cc8214a3bfb976a17a24145780
+oid sha256:6652210c5f8f6e3f7fa8c108025dcd5ab500a8b7cbc2e9acf80578fb70340563
 size 5048