Training in progress, step 1000

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://nn-red.wandb.io/cellular-foundation-model/PertRL/runs/ratav7p8) 
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://nn-red.wandb.io/cellular-foundation-model/PertRL/runs/bg4qbxc9) 
 
 
 This model was trained with SFT.

adapter_config.json

@@ -24,13 +24,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
+    "q_proj",
+    "k_proj",
     "v_proj",
-    "gate_proj",
+    "o_proj",
     "up_proj",
     "down_proj",
-    "k_proj",
-    "q_proj",
-    "o_proj"
+    "gate_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:cb8e7deda6c1ff9dfa23a741113242fa908959558597c2455c41fbf1290ad32a
+oid sha256:e520b182fbf7de9d13df5f59c5b7b9ccc9f5d3817d3fbe341c83af02a6d5d4db
 size 1073863208

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b0bf1bdf43f69d24a8d258682570ea020f2f7e2cddbfcc0cf4573a0a8014b586
+oid sha256:f2ea7c91f0b4c77900b842a30137b7130f774607235358e27bed93e69407021d
 size 6737