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 sha256:9f3e26485037077c601e384acc3e722e66393fb5f3bc44743b0a681ebecbaadf
-size 109086416
+oid sha256:85807111bff4ee68aabe81c20d9959f2ead80e089e996a80c5c57b28c68cc5be
+size 872450448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:af576d1dd34817080021e1962c359b6eb51fc62de2945d760e984598391f824f
+oid sha256:c47ad3cd38c02232d6f08de96bc0346dfc4401b8ed8c59e5853ea6ea0a2e03ed
 size 5048