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": [
-    "v_proj",
-    "k_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 sha256:02070b9f55764c9d7b872f8723bf78cbc87f8725d6993fc5c67cbbe5043a9eb8
-size 109086416
+oid sha256:23acbb205843efe03718bb8d639845fdf7f3a2350c7ca4b7613dc2593910df81
+size 872450448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9315068e7cede233b26a4044254cdf047088c9a62a232119fbe083ab6f8d655a
+oid sha256:00cbe93aba25afec42c03db1cb78fbe312a9a01e4bb2bad903d66f87e94e5863
 size 4984