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 sha256:290ef9bdf51380dee8ed5134f809e0b326736aec55d516d6be5844807740a6cc
-size 109086416
+oid sha256:4db85abcd36552c10af9f585dcb8aec15470e54a01038d6cc6be8aad615bb4b6
+size 872450448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:160d879f04be3d6d78bb8e46c63b4427773ccf191dd1c09b96943cc5da464b78
+oid sha256:e2fceef98cc102c80dbfe18e3059c8c84d45664601f0cdcc42cab395c063ae01
 size 5048