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 sha256:8e0e57f286cc0b35bfbb7fc4b24730e77ca4cc960323c8f637118bac7518a42d
-size 109086416
+oid sha256:c3c575b9dcf1a5137d5b8ebf2bcc1b44211b80a61757800c0d5b73518d05e2c3
+size 872450448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7f159a85b9e68142dad62c523d92f2fb31aac40241ce99b3a2e28aa65f339ae2
+oid sha256:77310af1a6d85908f65a83bfaec91be405c20510adcce3dfc77480c71f4a9fb8
 size 4984