Training in progress, epoch 1

adapter_config.json

@@ -13,7 +13,7 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 64,
+  "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",
-    "k_proj"
+    "v_proj",
+    "o_proj"
   ],
   "task_type": null,
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7bd4bab8db50bf8f8384e5e6bbd03145dd562b402aee72d24034d66e44cd49b6
-size 109086416
+oid sha256:99b95e4eb537607d22494d4498cce0d987ccfca5b38fbc8b7929f5590d976395
+size 872450448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:241eea6fbe44068bdb349499892e27b498dbfbc7c731ab73ed9889bca5950b5e
+oid sha256:813e8415e8486d0fb94a126f92c6b6f030f30f9189f8b00a0a1efff348525c71
 size 5048