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 sha256:14ccbd4c09ab0c6dd429b8b1d6093419b9936829696578f1f52b9f09d0452cbf
-size 109086416
+oid sha256:2ae5632af59dfbea90d244432437d0e5db32ebfdd3b2d5ac93eda3e8c1021364
+size 872450448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9624c83c628483e24780de53f082c80c5b67437bc3c17a83e44735364f3404a1
+oid sha256:c2c760815051eb2cc3f5060d6a7ab8849ed81e3292d2a2d877ebaa916e75d07b
 size 5048