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/p236tggb) 
+[<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) 
 
 
 This model was trained with SFT.

adapter_config.json

@@ -26,11 +26,11 @@
   "target_modules": [
     "q_proj",
     "v_proj",
+    "gate_proj",
     "o_proj",
     "k_proj",
     "up_proj",
-    "down_proj",
-    "gate_proj"
+    "down_proj"
   ],
   "task_type": "CAUSAL_LM",
   "trainable_token_indices": null,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:303ca34310435480b9311bbf05b507e894ed0d10d362068857caf2a056440e52
+oid sha256:3c44ca2ecddade6179474af97d754260069cd531e2881842fec7fb9b33d5f389
 size 70430032

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b21c8f7637654bda7c8eeb8e52e0a21d3206340d4f4fd21001852b317d1d6211
+oid sha256:0f21420a79b1a1e0eb2ebd3f5e02a8c14b53715a8a736117b8dfae69e7548e97
 size 5624