Training in progress, epoch 1

adapter_config.json

@@ -14,7 +14,7 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 1,
+  "lora_alpha": 8,
   "lora_dropout": 0.1,
   "megatron_config": null,
   "megatron_core": "megatron.core",
@@ -30,10 +30,10 @@
   "scheduler_verbose": true,
   "scheduler_verbose_interval": 100,
   "target_modules": [
-    "value",
-    "query",
     "key",
-    "dense"
+    "query",
+    "dense",
+    "value"
   ],
   "task_type": null,
   "total_steps": 1290.0,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3099178b152d3b2d4f94ccdfe3bad73acc0c7680d6dacfc38f4a724590a47e7d
+oid sha256:7358e2c6ad10088201d2cbef99fc54133d3ec06e764b6e26a38e8d3f8540dee1
 size 33056140

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7d42a0f262bf3f3f0619c289e1df43c2b7450fe833275a081c39cf12b776256a
+oid sha256:3e500ab6c715621e69f10dedc4599a5547c4ec54479c17e0ec846bb2c108f140
 size 5304