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 sha256:3e227136754044ea206627362fc1f3504df0b572b823b587b1bf5ea9f58d5f1d
-size 109086416
+oid sha256:cde1000cf69048244e5e62b10bb6b0dfcbf91e75c7c65bad1d789a84f02e69e5
+size 872450448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0fc3fd07eb7d4e14d9ca7e3e0f24cc6f6ba7a3969281ad72aae56badab394f4b
+oid sha256:b983ef60b9feda5d0d01610c01c98673a0b090dff4ee8e45584ae9b4e5da9244
 size 5048