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 sha256:51603034a35a6842c6db50a89306e50521a2ca848dda9eda5c2e79d14a50ebd9
-size 109086416
+oid sha256:4dc59fe25fde7a3979c4c22864cca546e3826fb72e75789d5976227d29ee351d
+size 872450448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:42acdddad94ab5540201ee1b941bc710f9b24629007062ae97cf5f7cc51c8222
+oid sha256:8c4d9c3354681d5d199580ad42ce4bf295f252cd85bb90e8e51c8a980fd5cd6e
 size 5048