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": [
-    "k_proj",
-    "v_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 sha256:6cb0c594e1e753ac615c83845beac2a099d0e097d17510920cb136d4fe4bfc00
-size 109086416
+oid sha256:be6576ff205b2d850534d9e7f1d0ffffc03da4fbfc618472e037522ba3e7f810
+size 872450448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6be034b2c756889ebe5b4e20632b86c66d63cb29d4582b79bf501f864a0f5e87
+oid sha256:355ee841c67cf4850bb0abf3a15a279af2cb276e4834b20d8ee009fb3809f036
 size 5048