Training in progress, step 100

README.md

@@ -4,8 +4,8 @@ library_name: transformers
 model_name: nmt_t1_v2
 tags:
 - generated_from_trainer
-- grpo
 - trl
+- grpo
 licence: license
 ---
 
@@ -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/oleg-dats/nmt/runs/sy16a2q6) 
+[<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/oleg-dats/nmt/runs/5clnlxvx) 
 
 
 This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://huggingface.co/papers/2402.03300).

adapter_config.json

@@ -27,13 +27,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "k_proj",
-    "up_proj",
     "down_proj",
-    "q_proj",
-    "v_proj",
     "o_proj",
-    "gate_proj"
+    "v_proj",
+    "gate_proj",
+    "q_proj",
+    "up_proj",
+    "k_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:feabe11f7d9785c32694a320d05819eaeb9b0525c28f42bb24ff077fa34f85d7
+oid sha256:eed8a4fd4be3b9887512802e0d95f06d46370f31f0f4818029aa5d9dd951a690
 size 1260191096

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8d6b2a3d930b9696bc3be25e19f9f6274becf3db99568495ffa084311d197a5a
+oid sha256:0a8c3e08db69f25a42364a6c75a9a8134ad5f91eadfde1f6f8f7162b9849cc5a
 size 6456