jacobcd52/test_sft

README.md

@@ -27,7 +27,7 @@ print(output["generated_text"])
 
 ## Training procedure
 
-[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/training-saes/llm-finetuning/runs/18qgmcx1) 
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/training-saes/llm-finetuning/runs/tmjwqh8d) 
 
 
 This model was trained with SFT.

adapter_config.json

@@ -24,11 +24,11 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_proj",
-    "v_proj",
     "gate_proj",
     "o_proj",
     "k_proj",
+    "v_proj",
+    "q_proj",
     "up_proj",
     "down_proj"
   ],

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3c44ca2ecddade6179474af97d754260069cd531e2881842fec7fb9b33d5f389
+oid sha256:3646b67b9121177556c7faf419ddbbd244ab747c8c9fd53050606639920a4d83
 size 70430032

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0f21420a79b1a1e0eb2ebd3f5e02a8c14b53715a8a736117b8dfae69e7548e97
+oid sha256:35a0b7804baf28aa53bb7477d29117b1d130527223ff9308dd687e3e282f8a6f
 size 5624