End of training

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/paolordls-upd/Crosslingual%20Lens/runs/bzixk4l9)
+[<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/paolordls-upd/Crosslingual%20Lens/runs/emil3s0q)
 
 This model was trained with SFT.
 
@@ -35,7 +35,7 @@ This model was trained with SFT.
 
 - TRL: 0.12.1
 - Transformers: 4.46.2
-- Pytorch: 2.5.0+cu121
+- Pytorch: 2.5.1+cu121
 - Datasets: 3.1.0
 - Tokenizers: 0.20.3
 

adapter_config.json

@@ -16,16 +16,16 @@
   "megatron_core": "megatron.core",
   "modules_to_save": null,
   "peft_type": "LORA",
-  "r": 8,
+  "r": 2048,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
     "k_proj",
+    "q_proj",
     "v_proj",
-    "gate_proj",
     "down_proj",
-    "up_proj",
-    "q_proj"
+    "gate_proj",
+    "up_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:308a69fbbc60d4e54027995749bf96ae0ee8f5f151db7bb92f76d57f91b9b2e2
-size 37953560
+oid sha256:a56ae26ef7444a17cd0ec54c15f4d3cb9ee38ce8648f37a25349e315edc2d729
+size 9705662728

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cc6b6b1f1a081d089da53ae8f4de4a13c2ad1018bd7c080b5846760d35be9482
+oid sha256:813ff732a1ba262a5c4b3686ecb091cd2380f8cc5c3c330ca864b5c50594f32a
 size 5496