Training in progress, step 500

adapter_config.json

@@ -13,7 +13,7 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 64,
+  "lora_alpha": 128,
   "lora_bias": false,
   "lora_dropout": 0.0,
   "megatron_config": null,
@@ -21,17 +21,17 @@
   "modules_to_save": null,
   "peft_type": "LORA",
   "qalora_group_size": 16,
-  "r": 32,
+  "r": 64,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "gate_proj",
-    "down_proj",
-    "v_proj",
     "o_proj",
+    "q_proj",
     "k_proj",
+    "gate_proj",
     "up_proj",
-    "q_proj"
+    "down_proj",
+    "v_proj"
   ],
   "target_parameters": null,
   "task_type": "CAUSAL_LM",

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:de47021a005849dc0c29d9388cd230cddc3e7b276c00f0308ff793781b1bc37c
-size 1656902648
+oid sha256:3566a701422b3ed0b833b6bc2b591d68586b93772d288d12ca845acf28850fb4
+size 3313653480

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f86c2cea252e43f273c7c4e9a07f8e813014205e48b589994a9bd7b78e7dbf43
+oid sha256:c7d67b912fbfbf5bc0668a59a2bfa1be7c2d354e30c77e54f800f43dc84c5f3b
 size 6225