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 sha256:2b0dc3cafb7b490bec171753a80fcb4581a6932302d0a1110bd8f0ca4646127f
-size 109086416
+oid sha256:cdfcbf0c16d5bb89847a434302c2fd79bb8c7ccad790255b8b3f05d2f0f1ed27
+size 872450448

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:98ad880a734a9ede96585e1b7c630f7ceb8d4e6d1e0f6ae9b363c9189cc51a1b
+oid sha256:ae7961bb2ea1d7844f6d18b4ff54f6961fef03371f9d1141be54d7426c7377ba
 size 5048