Training in progress, step 100

adapter_config.json

@@ -14,7 +14,7 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 128,
+  "lora_alpha": 256,
   "lora_dropout": 0.0,
   "megatron_config": null,
   "megatron_core": "megatron.core",
@@ -25,9 +25,9 @@
   "rank_pattern": null,
   "revision": null,
   "target_modules": [
+    "q_attn",
     "c_attn",
     "c_proj",
-    "q_attn",
     "c_fc"
   ],
   "target_r": 8,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4b00bc101c91604642230411bc965d385b08cf7f3a7890d20be5950c244c022a
+oid sha256:a1ea02e456d267c9b2c5a7eed6eac318ed9801cacf6e6c46b585e84bd1e8c5a8
 size 33364520

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dbe08dd6bd4645ebad579e3ebd09bdb35b6a0e8cd132361bd8fb925bfe40dcf4
+oid sha256:8524f783bfd2abf42e7960ccd272418a920b332aebb2407e47baeb49949cda7c
 size 5176