End of training

README.md

@@ -39,8 +39,8 @@ The following hyperparameters were used during training:
 - seed: 42
 - gradient_accumulation_steps: 8
 - total_train_batch_size: 8
-- optimizer: Use OptimizerNames.ADAMW_HF with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
-- lr_scheduler_type: linear
+- optimizer: Use OptimizerNames.ADAMW_8BIT with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
 - lr_scheduler_warmup_steps: 50
 - num_epochs: 3
 

adapter_config.json

@@ -15,19 +15,24 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 1,
+  "lora_alpha": 16,
   "lora_bias": false,
-  "lora_dropout": 0.05,
+  "lora_dropout": 0.1,
   "megatron_config": null,
   "megatron_core": "megatron.core",
   "modules_to_save": null,
   "peft_type": "LORA",
-  "r": 1,
+  "r": 4,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
+    "down_proj",
+    "q_proj",
+    "up_proj",
     "o_proj",
-    "down_proj"
+    "k_proj",
+    "v_proj",
+    "gate_proj"
   ],
   "task_type": null,
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b9ad98a8c2b4fde0a1f18a99a5240e99b6fa22974f359443bcc03ac699569024
-size 1389688
+oid sha256:2179834d8320c628e0ae4b2c4512210978c74c09297a77a34c10f3c143714b10
+size 21146656

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:df65e48ee9c7ac5c046a8d6f97a18e8363f796cde740e17aeab75fba83a487f5
+oid sha256:fa1a05de5cfdaf4d9888b30ba7a833c3852b88b37ba81bd059b899b6b8cfc2d0
 size 5304