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:ba6455681a056898fa54c5bcfee967bc2cecb9bf5295e755425adf8bc50c67b3
-size 109086416
+oid sha256:3bade5dd25ff22f8950347bd219dc83e16b2ca688d8c858cf1c58bb5facb64ba
+size 872450448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:17868a840381127e2f05fbdfcccfaa941c9855d02c2c529d3baa3b3dc7974316
+oid sha256:beb84b76d12791c7bb5d04b465733a21f6c2603cf8b5e3b56887bacd5d15966e
 size 5048