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 sha256:949bd7206be8a58e1b5517a9a7ebaffd52aa3e2435befe41914cbbc16176bea8
-size 109086416
+oid sha256:5715de5e25845676be495b7feb9a3d76c0527eb03e55246e3f7692f8cec03831
+size 872450448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9a10e70d4353c29688d331d52560a56a86593ed77a706cd66e4dd77545318ed8
+oid sha256:8d05088ce3069b88fdba5f33147a6a4a39db581c8c1ea4f2bd4957fd074b91fd
 size 5048