Training in progress, step 148

adapter_config.json

@@ -19,17 +19,17 @@
   "megatron_core": "megatron.core",
   "modules_to_save": null,
   "peft_type": "LORA",
-  "r": 256,
+  "r": 32,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
+    "gate_proj",
     "q_proj",
     "up_proj",
-    "down_proj",
     "o_proj",
-    "gate_proj",
     "v_proj",
-    "k_proj"
+    "k_proj",
+    "down_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cc5ec95f6910cbab01542ae8ebed03c66ccdf7490caed2703938373356ae275f
-size 2583745128
+oid sha256:80a9c082ef40e0ea3fbb85651c09b6576d3c6aff01f26058497e099bb7a4861c
+size 323014168

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7e68f2068a59cf80565d3ea3d1c38396f85e6fa43810c2e26106e97929a35096
+oid sha256:caf1da72f915c1d2500e75c822f39dac2c960444c4478987678336f52367d022
 size 6776