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 sha256:1292388296f39df8dbeb05190207f073c3952f50ff248bb80e7d41df83491028
-size 109086416
+oid sha256:8c761e4975bf4d472deb3e63cad473e3c2537b820e79f41f71ea33779e746a29
+size 872450448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7da424d249e29af9233a8e30d73ba1b2136b36acabd56e1e83dca118cce12038
+oid sha256:adfe6109ba30810088b31f3930f8282f8e0d772886249804504c546d0c7dba27
 size 5048