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 sha256:08b70864e4eac3bfb2de25add4801a49183ca5ba0b294382a32b5dc3882d96a7
-size 109086416
+oid sha256:059ef3c4c7ddd6d85e2bdab5d000f540d51eab0e465e268507bcc08c3d3fbe69
+size 872450448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6abb697a3f75972cb5d9b3cb066f3f1abb42f95fb037a4344a3352ffdaa9d858
+oid sha256:43a26e4bc6350fab2ebe3f424fca25f2ed80ff100437b1789a732bdc33640afa
 size 5048