Tatvajsh/Lllama_AHS_V_6.6

README.md

@@ -38,7 +38,7 @@ The following hyperparameters were used during training:
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- training_steps: 100
+- training_steps: 500
 - mixed_precision_training: Native AMP
 
 ### Training results

adapter_config.json

@@ -16,14 +16,14 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "base_layer",
-    "q_proj",
-    "v_proj",
+    "down_proj",
     "o_proj",
-    "gate_proj",
+    "v_proj",
     "up_proj",
+    "q_proj",
     "k_proj",
-    "down_proj"
+    "gate_proj",
+    "base_layer"
   ],
   "task_type": "CAUSAL_LM"
 }

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:18dafa6c0ee06c46901f52d9eed0b943e355197fe1341923551c2481c0391409
+oid sha256:26ae46498cc6ea63d753a209c8a9d49b8fc6d9e12e70aed0c6a4be18e45c5aef
 size 101834237

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1f9955b04800dc6c2cf924ada51ee03b2e139da1bb54bddd1445c63daa7185fc
+oid sha256:4fd357cd3434ca8ab3b16b6eee237402522b87a2e930ce73f350b77ea8380a45
 size 4027