full-state checkpoint 100-percent (step 200)

ckpt-100-percent/adapter_config.json

@@ -19,7 +19,7 @@
   "loftq_config": {},
   "lora_alpha": 256,
   "lora_bias": false,
-  "lora_dropout": 0.01,
+  "lora_dropout": 0.05,
   "megatron_config": null,
   "megatron_core": "megatron.core",
   "modules_to_save": null,
@@ -28,15 +28,7 @@
   "r": 128,
   "rank_pattern": {},
   "revision": null,
-  "target_modules": [
-    "up_proj",
-    "down_proj",
-    "v_proj",
-    "gate_proj",
-    "o_proj",
-    "k_proj",
-    "q_proj"
-  ],
+  "target_modules": "(?:.*?(?:language|text).*?(?:self_attn|attention|attn|mlp|feed_forward|ffn|dense).*?(?:k_proj|v_proj|q_proj|out_proj|fc1|fc2|o_proj|gate_proj|up_proj|down_proj).*?)|(?:\\bmodel\\.layers\\.[\\d]{1,}\\.(?:self_attn|attention|attn|mlp|feed_forward|ffn|dense)\\.(?:(?:k_proj|v_proj|q_proj|out_proj|fc1|fc2|o_proj|gate_proj|up_proj|down_proj)))",
   "target_parameters": null,
   "task_type": "CAUSAL_LM",
   "trainable_token_indices": null,

ckpt-100-percent/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9b80407612d781b6680d60112fa826717cd7136ad67569404f824ed41890e9c3
-size 1049331312
+oid sha256:15ebaafbc10258d89f43a2de191bc310a4a922382ad8627c200ce79fba1c7e29
+size 953751848

ckpt-100-percent/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:720322977874d798b7497ede5a4de34616347d4efe023f1d6ad7e85ef0e1b270
+oid sha256:c64a13540435de14d77f995cb21c2dbe480d0c9e3df79c3e35756e161003afc0
 size 7313