Training in progress, step 50

adapter_config.json

@@ -16,17 +16,17 @@
   "megatron_core": "megatron.core",
   "modules_to_save": null,
   "peft_type": "LORA",
-  "r": 8,
+  "r": 32,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_proj",
-    "k_proj",
-    "up_proj",
-    "v_proj",
     "down_proj",
+    "k_proj",
+    "o_proj",
+    "q_proj",
     "gate_proj",
-    "o_proj"
+    "up_proj",
+    "v_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1f2c54b3cff7229bba3a337321576ca3fbedcde46f10b6c700245830c01cb495
-size 80013120
+oid sha256:43756c87d3c353cd42eeb6185a59e755a36f9b382d2d1141fdd67b110a8adbc4
+size 319876032

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b8d9124138abd44af04b2c60a935bcab4ff5cdb3ea64e57559b87dc3f7e79065
+oid sha256:587385ca4b3a6c2778a0b2f3cca66b2116c2d78e99e16d5766eaff5ef6eeb893
 size 6776