Training in progress, step 100

adapter_config.json

@@ -23,230 +23,374 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "language_model.model.layers.22.self_attn.v_proj",
-    "language_model.model.layers.10.self_attn.k_proj",
-    "language_model.model.layers.26.self_attn.o_proj",
-    "language_model.model.layers.22.self_attn.k_proj",
-    "language_model.model.layers.23.mlp.up_proj",
-    "language_model.model.layers.4.mlp.gate_proj",
-    "language_model.model.layers.11.mlp.up_proj",
-    "language_model.model.layers.17.self_attn.q_proj",
-    "language_model.model.layers.11.mlp.down_proj",
-    "language_model.model.layers.3.self_attn.o_proj",
-    "language_model.model.layers.3.mlp.gate_proj",
-    "language_model.model.layers.20.self_attn.o_proj",
-    "language_model.model.layers.7.self_attn.v_proj",
-    "language_model.model.layers.26.mlp.gate_proj",
-    "language_model.model.layers.6.mlp.gate_proj",
-    "language_model.model.layers.16.self_attn.q_proj",
-    "language_model.model.layers.9.mlp.up_proj",
-    "language_model.model.layers.17.self_attn.k_proj",
-    "language_model.model.layers.8.self_attn.k_proj",
+    "language_model.model.layers.10.mlp.down_proj",
+    "language_model.model.layers.0.mlp.gate_proj",
+    "language_model.model.layers.31.self_attn.k_proj",
+    "language_model.model.layers.16.mlp.gate_proj",
+    "language_model.model.layers.1.mlp.gate_proj",
+    "language_model.model.layers.15.self_attn.o_proj",
+    "vision_tower.vision_model.encoder.layers.15.self_attn.out_proj",
+    "language_model.model.layers.12.self_attn.k_proj",
+    "language_model.model.layers.29.mlp.down_proj",
+    "language_model.model.layers.2.self_attn.k_proj",
+    "language_model.model.layers.13.self_attn.v_proj",
+    "language_model.model.layers.6.self_attn.q_proj",
+    "language_model.model.layers.23.mlp.down_proj",
+    "language_model.model.layers.2.mlp.down_proj",
     "language_model.model.layers.18.self_attn.o_proj",
-    "language_model.model.layers.17.mlp.up_proj",
+    "vision_tower.vision_model.encoder.layers.18.mlp.fc1",
+    "vision_tower.vision_model.encoder.layers.13.mlp.fc2",
+    "language_model.model.layers.17.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.13.self_attn.k_proj",
+    "language_model.model.layers.5.mlp.up_proj",
+    "language_model.model.layers.4.mlp.up_proj",
+    "language_model.model.layers.23.self_attn.k_proj",
+    "language_model.model.layers.18.mlp.up_proj",
+    "language_model.model.layers.5.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.16.mlp.fc2",
+    "language_model.model.layers.18.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.21.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.14.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.2.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.4.mlp.fc2",
+    "language_model.model.layers.23.self_attn.v_proj",
+    "language_model.model.layers.10.self_attn.o_proj",
+    "language_model.model.layers.30.mlp.gate_proj",
+    "language_model.model.layers.25.mlp.gate_proj",
+    "language_model.model.layers.0.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.15.mlp.fc1",
+    "vision_tower.vision_model.encoder.layers.20.mlp.fc1",
+    "language_model.model.layers.17.self_attn.q_proj",
+    "language_model.model.layers.26.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.12.self_attn.out_proj",
+    "language_model.model.layers.7.mlp.down_proj",
+    "language_model.model.layers.13.mlp.gate_proj",
+    "language_model.model.layers.27.self_attn.v_proj",
+    "language_model.model.layers.14.mlp.up_proj",
+    "vision_tower.vision_model.encoder.layers.20.self_attn.k_proj",
+    "language_model.model.layers.1.self_attn.v_proj",
+    "language_model.model.layers.30.self_attn.o_proj",
+    "vision_tower.vision_model.encoder.layers.19.self_attn.k_proj",
     "language_model.model.layers.29.self_attn.o_proj",
-    "language_model.model.layers.28.mlp.down_proj",
-    "language_model.model.layers.15.mlp.up_proj",
-    "language_model.model.layers.0.mlp.gate_proj",
-    "language_model.model.layers.9.mlp.gate_proj",
+    "language_model.model.layers.4.self_attn.q_proj",
+    "language_model.model.layers.12.self_attn.o_proj",
+    "language_model.model.layers.25.self_attn.o_proj",
+    "vision_tower.vision_model.encoder.layers.1.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.5.self_attn.out_proj",
+    "language_model.model.layers.5.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.14.mlp.fc1",
+    "vision_tower.vision_model.encoder.layers.12.self_attn.k_proj",
     "language_model.model.layers.19.mlp.down_proj",
-    "language_model.model.layers.6.self_attn.o_proj",
-    "language_model.model.layers.9.self_attn.o_proj",
-    "language_model.model.layers.5.self_attn.o_proj",
-    "language_model.model.layers.13.self_attn.o_proj",
-    "language_model.model.layers.31.mlp.up_proj",
-    "language_model.model.layers.17.self_attn.v_proj",
-    "language_model.model.layers.31.self_attn.v_proj",
-    "language_model.model.layers.6.mlp.down_proj",
-    "language_model.model.layers.19.self_attn.o_proj",
-    "language_model.model.layers.16.self_attn.o_proj",
+    "language_model.model.layers.3.self_attn.v_proj",
+    "language_model.model.layers.21.mlp.gate_proj",
+    "language_model.model.layers.17.mlp.gate_proj",
+    "language_model.model.layers.14.self_attn.k_proj",
+    "language_model.model.layers.2.mlp.up_proj",
+    "vision_tower.vision_model.encoder.layers.23.self_attn.k_proj",
+    "language_model.model.layers.28.mlp.gate_proj",
+    "language_model.model.layers.4.mlp.gate_proj",
+    "language_model.model.layers.7.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.7.mlp.fc2",
+    "vision_tower.vision_model.encoder.layers.4.mlp.fc1",
+    "language_model.model.layers.16.self_attn.q_proj",
+    "language_model.model.layers.22.self_attn.q_proj",
+    "language_model.model.layers.15.mlp.gate_proj",
+    "language_model.model.layers.18.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.6.mlp.fc1",
+    "vision_tower.vision_model.encoder.layers.8.self_attn.k_proj",
+    "language_model.model.layers.15.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.11.self_attn.k_proj",
     "language_model.model.layers.7.self_attn.k_proj",
-    "language_model.model.layers.18.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.8.self_attn.q_proj",
+    "language_model.model.layers.10.self_attn.k_proj",
+    "language_model.model.layers.8.mlp.down_proj",
+    "language_model.model.layers.12.mlp.down_proj",
+    "language_model.model.layers.1.self_attn.k_proj",
+    "language_model.model.layers.27.self_attn.k_proj",
+    "language_model.model.layers.27.mlp.gate_proj",
+    "language_model.model.layers.14.mlp.gate_proj",
     "language_model.model.layers.1.self_attn.q_proj",
-    "language_model.model.layers.2.mlp.gate_proj",
-    "language_model.model.layers.22.mlp.up_proj",
-    "language_model.model.layers.19.self_attn.q_proj",
-    "language_model.model.layers.16.self_attn.k_proj",
-    "language_model.model.layers.30.self_attn.k_proj",
-    "language_model.model.layers.2.self_attn.k_proj",
-    "language_model.model.layers.29.self_attn.v_proj",
+    "language_model.model.layers.28.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.12.self_attn.v_proj",
+    "language_model.model.layers.19.self_attn.v_proj",
+    "language_model.model.layers.8.self_attn.v_proj",
+    "language_model.model.layers.6.self_attn.o_proj",
+    "language_model.model.layers.8.mlp.gate_proj",
+    "language_model.model.layers.23.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.14.self_attn.out_proj",
+    "language_model.model.layers.11.self_attn.k_proj",
     "language_model.model.layers.26.mlp.down_proj",
-    "language_model.model.layers.0.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.3.self_attn.q_proj",
+    "language_model.model.layers.3.mlp.down_proj",
+    "language_model.model.layers.11.mlp.gate_proj",
+    "language_model.model.layers.22.self_attn.k_proj",
     "language_model.model.layers.7.mlp.gate_proj",
-    "language_model.model.layers.21.mlp.gate_proj",
-    "language_model.model.layers.30.self_attn.o_proj",
-    "language_model.model.layers.3.self_attn.k_proj",
-    "language_model.model.layers.5.self_attn.q_proj",
-    "language_model.model.layers.5.mlp.gate_proj",
-    "language_model.model.layers.1.self_attn.o_proj",
-    "language_model.model.layers.19.self_attn.v_proj",
-    "language_model.model.layers.4.mlp.up_proj",
-    "language_model.model.layers.8.mlp.down_proj",
-    "language_model.model.layers.25.mlp.up_proj",
-    "language_model.model.layers.26.self_attn.v_proj",
-    "language_model.model.layers.1.mlp.up_proj",
-    "language_model.model.layers.19.mlp.up_proj",
-    "language_model.model.layers.24.mlp.up_proj",
-    "language_model.model.layers.23.self_attn.o_proj",
-    "language_model.model.layers.27.self_attn.k_proj",
-    "language_model.model.layers.3.mlp.up_proj",
-    "language_model.model.layers.4.mlp.down_proj",
-    "language_model.model.layers.7.mlp.down_proj",
-    "language_model.model.layers.26.self_attn.q_proj",
-    "language_model.model.layers.0.self_attn.o_proj",
+    "vision_tower.vision_model.encoder.layers.23.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.10.mlp.fc1",
+    "language_model.model.layers.26.mlp.up_proj",
+    "vision_tower.vision_model.encoder.layers.10.self_attn.q_proj",
+    "language_model.model.layers.31.self_attn.o_proj",
+    "vision_tower.vision_model.encoder.layers.22.mlp.fc2",
+    "vision_tower.vision_model.encoder.layers.11.mlp.fc1",
+    "language_model.model.layers.19.mlp.gate_proj",
+    "vision_tower.vision_model.encoder.layers.2.mlp.fc2",
+    "language_model.model.layers.26.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.21.mlp.fc2",
+    "vision_tower.vision_model.encoder.layers.1.mlp.fc1",
     "language_model.model.layers.16.mlp.down_proj",
-    "language_model.model.layers.21.self_attn.q_proj",
-    "language_model.model.layers.18.self_attn.k_proj",
+    "language_model.model.layers.28.self_attn.o_proj",
+    "vision_tower.vision_model.encoder.layers.15.mlp.fc2",
+    "vision_tower.vision_model.encoder.layers.19.self_attn.q_proj",
+    "language_model.model.layers.2.mlp.gate_proj",
+    "language_model.model.layers.15.mlp.down_proj",
+    "language_model.model.layers.13.self_attn.q_proj",
+    "language_model.model.layers.17.mlp.up_proj",
+    "language_model.model.layers.5.self_attn.o_proj",
+    "vision_tower.vision_model.encoder.layers.18.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.7.mlp.fc1",
+    "vision_tower.vision_model.encoder.layers.14.mlp.fc2",
+    "vision_tower.vision_model.encoder.layers.8.mlp.fc2",
+    "language_model.model.layers.1.mlp.down_proj",
+    "language_model.model.layers.9.mlp.up_proj",
+    "vision_tower.vision_model.encoder.layers.5.self_attn.q_proj",
+    "language_model.model.layers.26.mlp.gate_proj",
     "language_model.model.layers.20.self_attn.k_proj",
-    "language_model.model.layers.23.self_attn.k_proj",
-    "language_model.model.layers.26.mlp.up_proj",
-    "language_model.model.layers.0.self_attn.v_proj",
-    "language_model.model.layers.3.self_attn.v_proj",
-    "language_model.model.layers.18.mlp.down_proj",
-    "language_model.model.layers.27.self_attn.q_proj",
-    "language_model.model.layers.29.mlp.up_proj",
-    "language_model.model.layers.31.self_attn.k_proj",
-    "language_model.model.layers.6.mlp.up_proj",
-    "language_model.model.layers.11.self_attn.q_proj",
-    "language_model.model.layers.31.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.20.self_attn.v_proj",
+    "language_model.model.layers.4.mlp.down_proj",
+    "vision_tower.vision_model.encoder.layers.9.mlp.fc2",
+    "language_model.model.layers.5.mlp.gate_proj",
+    "language_model.model.layers.30.mlp.up_proj",
+    "language_model.model.layers.17.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.2.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.4.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.17.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.7.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.8.mlp.fc1",
+    "language_model.model.layers.8.self_attn.o_proj",
+    "language_model.model.layers.5.mlp.down_proj",
+    "vision_tower.vision_model.encoder.layers.15.self_attn.k_proj",
+    "language_model.model.layers.19.mlp.up_proj",
+    "language_model.model.layers.11.mlp.down_proj",
+    "vision_tower.vision_model.encoder.layers.23.self_attn.out_proj",
+    "vision_tower.vision_model.encoder.layers.7.self_attn.v_proj",
+    "language_model.model.layers.27.self_attn.o_proj",
+    "vision_tower.vision_model.encoder.layers.20.mlp.fc2",
+    "language_model.model.layers.20.mlp.up_proj",
+    "vision_tower.vision_model.encoder.layers.0.self_attn.v_proj",
+    "language_model.model.layers.18.mlp.gate_proj",
+    "vision_tower.vision_model.encoder.layers.22.self_attn.k_proj",
     "language_model.model.layers.25.self_attn.k_proj",
-    "language_model.model.layers.24.self_attn.v_proj",
-    "language_model.model.layers.23.self_attn.v_proj",
-    "language_model.model.layers.7.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.23.mlp.fc1",
+    "vision_tower.vision_model.encoder.layers.9.mlp.fc1",
+    "language_model.model.layers.13.self_attn.k_proj",
+    "language_model.model.layers.14.self_attn.o_proj",
+    "language_model.model.layers.29.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.4.self_attn.v_proj",
     "language_model.model.layers.2.self_attn.v_proj",
-    "language_model.model.layers.31.mlp.gate_proj",
-    "language_model.model.layers.10.self_attn.o_proj",
-    "language_model.model.layers.18.mlp.up_proj",
-    "language_model.model.layers.11.self_attn.v_proj",
-    "language_model.model.layers.16.mlp.gate_proj",
-    "language_model.model.layers.20.self_attn.v_proj",
-    "language_model.model.layers.31.self_attn.o_proj",
-    "language_model.model.layers.6.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.2.self_attn.v_proj",
+    "language_model.model.layers.6.mlp.down_proj",
+    "vision_tower.vision_model.encoder.layers.13.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.9.self_attn.v_proj",
+    "language_model.model.layers.0.self_attn.q_proj",
+    "language_model.model.layers.12.self_attn.q_proj",
+    "language_model.model.layers.18.self_attn.v_proj",
+    "language_model.model.layers.6.self_attn.v_proj",
     "language_model.model.layers.17.self_attn.o_proj",
-    "language_model.model.layers.10.self_attn.q_proj",
-    "language_model.model.layers.23.mlp.down_proj",
-    "language_model.model.layers.8.self_attn.o_proj",
-    "language_model.model.layers.0.mlp.down_proj",
     "language_model.model.layers.22.self_attn.o_proj",
-    "language_model.model.layers.4.self_attn.k_proj",
-    "language_model.model.layers.14.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.17.mlp.fc2",
+    "vision_tower.vision_model.encoder.layers.4.self_attn.out_proj",
+    "language_model.model.layers.27.self_attn.q_proj",
+    "language_model.model.layers.25.self_attn.q_proj",
+    "language_model.model.layers.31.mlp.gate_proj",
+    "vision_tower.vision_model.encoder.layers.16.self_attn.q_proj",
+    "language_model.model.layers.30.self_attn.k_proj",
+    "language_model.model.layers.22.mlp.gate_proj",
+    "vision_tower.vision_model.encoder.layers.22.self_attn.v_proj",
     "language_model.model.layers.24.self_attn.o_proj",
-    "language_model.model.layers.25.self_attn.v_proj",
-    "language_model.model.layers.17.mlp.down_proj",
-    "language_model.model.layers.25.self_attn.o_proj",
-    "language_model.model.layers.2.self_attn.o_proj",
-    "language_model.model.layers.14.mlp.gate_proj",
-    "language_model.model.layers.25.mlp.gate_proj",
-    "language_model.model.layers.22.mlp.down_proj",
-    "language_model.model.layers.27.mlp.down_proj",
-    "language_model.model.layers.23.self_attn.q_proj",
-    "language_model.model.layers.21.self_attn.o_proj",
-    "language_model.model.layers.8.mlp.gate_proj",
-    "language_model.model.layers.5.self_attn.k_proj",
-    "language_model.model.layers.16.self_attn.v_proj",
-    "language_model.model.layers.29.self_attn.k_proj",
-    "language_model.model.layers.15.mlp.down_proj",
-    "language_model.model.layers.21.mlp.up_proj",
-    "language_model.model.layers.30.mlp.up_proj",
-    "language_model.model.layers.24.mlp.gate_proj",
-    "language_model.model.layers.11.mlp.gate_proj",
-    "language_model.model.layers.15.self_attn.q_proj",
-    "language_model.model.layers.24.mlp.down_proj",
-    "language_model.model.layers.28.self_attn.q_proj",
-    "language_model.model.layers.9.mlp.down_proj",
-    "language_model.model.layers.12.mlp.up_proj",
-    "language_model.model.layers.28.self_attn.o_proj",
-    "language_model.model.layers.27.mlp.gate_proj",
-    "language_model.model.layers.10.mlp.gate_proj",
-    "language_model.model.layers.19.self_attn.k_proj",
-    "language_model.model.layers.27.self_attn.v_proj",
-    "language_model.model.layers.29.mlp.down_proj",
-    "language_model.model.layers.8.self_attn.v_proj",
+    "language_model.model.layers.11.self_attn.q_proj",
+    "language_model.model.layers.3.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.6.self_attn.k_proj",
+    "language_model.model.layers.10.mlp.up_proj",
+    "language_model.model.layers.31.mlp.up_proj",
+    "vision_tower.vision_model.encoder.layers.11.self_attn.q_proj",
+    "language_model.model.layers.15.mlp.up_proj",
+    "vision_tower.vision_model.encoder.layers.23.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.5.mlp.fc2",
     "language_model.model.layers.8.mlp.up_proj",
-    "language_model.model.layers.21.self_attn.v_proj",
-    "language_model.model.layers.14.self_attn.o_proj",
-    "language_model.model.layers.2.mlp.down_proj",
-    "language_model.model.layers.12.self_attn.q_proj",
-    "language_model.model.layers.14.mlp.up_proj",
-    "language_model.model.layers.6.self_attn.v_proj",
-    "language_model.model.layers.11.self_attn.o_proj",
-    "language_model.model.layers.22.self_attn.q_proj",
-    "language_model.model.layers.0.self_attn.k_proj",
-    "language_model.model.layers.14.mlp.down_proj",
+    "language_model.model.layers.15.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.21.mlp.fc1",
+    "language_model.model.layers.10.self_attn.q_proj",
+    "language_model.model.layers.20.mlp.down_proj",
+    "language_model.model.layers.24.mlp.up_proj",
+    "language_model.model.layers.25.mlp.up_proj",
+    "language_model.model.layers.10.self_attn.v_proj",
+    "language_model.model.layers.0.mlp.up_proj",
+    "vision_tower.vision_model.encoder.layers.12.mlp.fc2",
+    "language_model.model.layers.30.self_attn.v_proj",
+    "language_model.model.layers.9.mlp.gate_proj",
     "language_model.model.layers.28.self_attn.v_proj",
-    "language_model.model.layers.4.self_attn.v_proj",
-    "language_model.model.layers.1.self_attn.v_proj",
-    "language_model.model.layers.5.mlp.down_proj",
-    "language_model.model.layers.15.self_attn.o_proj",
-    "language_model.model.layers.30.mlp.gate_proj",
-    "language_model.model.layers.4.self_attn.o_proj",
-    "language_model.model.layers.28.self_attn.k_proj",
-    "language_model.model.layers.12.self_attn.k_proj",
-    "language_model.model.layers.12.self_attn.o_proj",
+    "vision_tower.vision_model.encoder.layers.5.mlp.fc1",
+    "vision_tower.vision_model.encoder.layers.10.self_attn.k_proj",
     "language_model.model.layers.12.self_attn.v_proj",
-    "language_model.model.layers.15.self_attn.v_proj",
-    "language_model.model.layers.10.mlp.up_proj",
-    "language_model.model.layers.13.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.19.mlp.fc1",
     "language_model.model.layers.9.self_attn.v_proj",
-    "language_model.model.layers.19.mlp.gate_proj",
-    "language_model.model.layers.12.mlp.gate_proj",
+    "language_model.model.layers.3.self_attn.k_proj",
+    "language_model.model.layers.5.self_attn.k_proj",
+    "language_model.model.layers.26.self_attn.q_proj",
+    "language_model.model.layers.16.mlp.up_proj",
+    "language_model.model.layers.21.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.18.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.10.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.6.self_attn.q_proj",
+    "language_model.model.layers.4.self_attn.k_proj",
+    "language_model.model.layers.3.mlp.gate_proj",
+    "language_model.model.layers.18.mlp.down_proj",
+    "language_model.model.layers.11.mlp.up_proj",
+    "language_model.model.layers.26.self_attn.o_proj",
+    "language_model.model.layers.29.mlp.gate_proj",
+    "language_model.model.layers.29.mlp.up_proj",
+    "language_model.model.layers.22.self_attn.v_proj",
+    "language_model.model.layers.6.mlp.up_proj",
+    "vision_tower.vision_model.encoder.layers.3.mlp.fc1",
+    "language_model.model.layers.3.self_attn.o_proj",
+    "vision_tower.vision_model.encoder.layers.16.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.17.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.19.self_attn.out_proj",
+    "vision_tower.vision_model.encoder.layers.1.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.4.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.17.self_attn.k_proj",
+    "language_model.model.layers.27.mlp.up_proj",
+    "vision_tower.vision_model.encoder.layers.14.self_attn.v_proj",
+    "language_model.model.layers.20.mlp.gate_proj",
+    "vision_tower.vision_model.encoder.layers.16.mlp.fc1",
+    "vision_tower.vision_model.encoder.layers.10.mlp.fc2",
+    "vision_tower.vision_model.encoder.layers.1.self_attn.out_proj",
+    "vision_tower.vision_model.encoder.layers.2.mlp.fc1",
+    "language_model.model.layers.10.mlp.gate_proj",
+    "vision_tower.vision_model.encoder.layers.3.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.13.mlp.fc1",
+    "vision_tower.vision_model.encoder.layers.6.mlp.fc2",
+    "language_model.model.layers.20.self_attn.o_proj",
+    "language_model.model.layers.14.self_attn.v_proj",
+    "language_model.model.layers.0.mlp.down_proj",
+    "vision_tower.vision_model.encoder.layers.1.mlp.fc2",
+    "vision_tower.vision_model.encoder.layers.16.self_attn.k_proj",
+    "language_model.model.layers.0.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.22.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.18.self_attn.out_proj",
+    "language_model.model.layers.3.mlp.up_proj",
+    "language_model.model.layers.13.self_attn.o_proj",
     "language_model.model.layers.25.mlp.down_proj",
-    "language_model.model.layers.10.mlp.down_proj",
-    "language_model.model.layers.24.self_attn.k_proj",
+    "language_model.model.layers.1.self_attn.o_proj",
+    "language_model.model.layers.8.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.19.mlp.fc2",
+    "language_model.model.layers.2.self_attn.o_proj",
     "language_model.model.layers.30.mlp.down_proj",
-    "language_model.model.layers.20.self_attn.q_proj",
-    "language_model.model.layers.13.mlp.up_proj",
-    "language_model.model.layers.12.mlp.down_proj",
+    "language_model.model.layers.31.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.22.self_attn.out_proj",
+    "language_model.model.layers.17.mlp.down_proj",
+    "vision_tower.vision_model.encoder.layers.18.mlp.fc2",
     "language_model.model.layers.31.mlp.down_proj",
-    "language_model.model.layers.13.self_attn.q_proj",
-    "language_model.model.layers.14.self_attn.q_proj",
-    "language_model.model.layers.18.mlp.gate_proj",
-    "language_model.model.layers.6.self_attn.k_proj",
-    "language_model.model.layers.28.mlp.up_proj",
-    "language_model.model.layers.29.self_attn.q_proj",
-    "language_model.model.layers.1.self_attn.k_proj",
-    "language_model.model.layers.11.self_attn.k_proj",
-    "language_model.model.layers.23.mlp.gate_proj",
-    "language_model.model.layers.1.mlp.gate_proj",
-    "language_model.model.layers.30.self_attn.q_proj",
-    "language_model.model.layers.2.self_attn.q_proj",
-    "language_model.model.layers.3.self_attn.q_proj",
-    "language_model.model.layers.27.mlp.up_proj",
-    "language_model.model.layers.7.mlp.up_proj",
+    "vision_tower.vision_model.encoder.layers.10.self_attn.out_proj",
+    "vision_tower.vision_model.encoder.layers.6.self_attn.out_proj",
+    "language_model.model.layers.24.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.11.mlp.fc2",
     "language_model.model.layers.9.self_attn.k_proj",
-    "language_model.model.layers.21.mlp.down_proj",
-    "language_model.model.layers.13.self_attn.v_proj",
-    "language_model.model.layers.20.mlp.down_proj",
-    "language_model.model.layers.15.mlp.gate_proj",
-    "language_model.model.layers.27.self_attn.o_proj",
-    "language_model.model.layers.20.mlp.up_proj",
+    "vision_tower.vision_model.encoder.layers.0.self_attn.out_proj",
+    "language_model.model.layers.12.mlp.up_proj",
+    "vision_tower.vision_model.encoder.layers.18.self_attn.q_proj",
+    "language_model.model.layers.7.mlp.up_proj",
+    "vision_tower.vision_model.encoder.layers.7.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.5.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.17.self_attn.out_proj",
+    "vision_tower.vision_model.encoder.layers.19.self_attn.v_proj",
     "language_model.model.layers.7.self_attn.o_proj",
-    "language_model.model.layers.13.mlp.down_proj",
-    "language_model.model.layers.15.self_attn.k_proj",
-    "language_model.model.layers.30.self_attn.v_proj",
-    "language_model.model.layers.2.mlp.up_proj",
-    "language_model.model.layers.18.self_attn.q_proj",
-    "language_model.model.layers.1.mlp.down_proj",
-    "language_model.model.layers.5.self_attn.v_proj",
-    "language_model.model.layers.0.mlp.up_proj",
-    "language_model.model.layers.13.mlp.gate_proj",
+    "vision_tower.vision_model.encoder.layers.8.self_attn.out_proj",
+    "vision_tower.vision_model.encoder.layers.0.mlp.fc2",
+    "language_model.model.layers.29.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.13.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.14.self_attn.q_proj",
+    "language_model.model.layers.28.mlp.down_proj",
     "language_model.model.layers.24.self_attn.q_proj",
-    "language_model.model.layers.20.mlp.gate_proj",
-    "language_model.model.layers.22.mlp.gate_proj",
-    "language_model.model.layers.14.self_attn.v_proj",
-    "language_model.model.layers.29.mlp.gate_proj",
-    "language_model.model.layers.17.mlp.gate_proj",
+    "language_model.model.layers.12.mlp.gate_proj",
+    "language_model.model.layers.25.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.12.mlp.fc1",
     "language_model.model.layers.8.self_attn.q_proj",
-    "language_model.model.layers.3.mlp.down_proj",
-    "language_model.model.layers.9.self_attn.q_proj",
-    "language_model.model.layers.16.mlp.up_proj",
-    "language_model.model.layers.26.self_attn.k_proj",
-    "language_model.model.layers.5.mlp.up_proj",
-    "language_model.model.layers.10.self_attn.v_proj",
-    "language_model.model.layers.25.self_attn.q_proj",
+    "language_model.model.layers.20.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.20.self_attn.q_proj",
+    "language_model.model.layers.31.self_attn.v_proj",
+    "language_model.model.layers.24.mlp.down_proj",
+    "language_model.model.layers.30.self_attn.q_proj",
+    "language_model.model.layers.0.self_attn.o_proj",
+    "vision_tower.vision_model.encoder.layers.21.self_attn.out_proj",
+    "language_model.model.layers.13.mlp.down_proj",
+    "vision_tower.vision_model.encoder.layers.3.mlp.fc2",
+    "language_model.model.layers.9.self_attn.o_proj",
+    "language_model.model.layers.11.self_attn.o_proj",
+    "language_model.model.layers.23.mlp.gate_proj",
+    "vision_tower.vision_model.encoder.layers.9.self_attn.out_proj",
+    "language_model.model.layers.19.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.11.self_attn.out_proj",
+    "language_model.model.layers.24.mlp.gate_proj",
+    "vision_tower.vision_model.encoder.layers.5.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.21.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.22.mlp.fc1",
+    "language_model.model.layers.28.self_attn.q_proj",
+    "language_model.model.layers.1.mlp.up_proj",
+    "language_model.model.layers.21.mlp.up_proj",
+    "language_model.model.layers.27.mlp.down_proj",
+    "vision_tower.vision_model.encoder.layers.0.mlp.fc1",
+    "language_model.model.layers.20.self_attn.v_proj",
+    "language_model.model.layers.29.self_attn.k_proj",
+    "language_model.model.layers.4.self_attn.o_proj",
+    "vision_tower.vision_model.encoder.layers.9.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.13.self_attn.out_proj",
+    "vision_tower.vision_model.encoder.layers.9.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.20.self_attn.out_proj",
+    "language_model.model.layers.19.self_attn.o_proj",
+    "vision_tower.vision_model.encoder.layers.7.self_attn.out_proj",
+    "language_model.model.layers.24.self_attn.k_proj",
+    "language_model.model.layers.2.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.15.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.21.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.16.self_attn.out_proj",
+    "language_model.model.layers.4.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.11.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.8.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.3.self_attn.v_proj",
+    "language_model.model.layers.21.mlp.down_proj",
+    "language_model.model.layers.9.mlp.down_proj",
+    "language_model.model.layers.7.self_attn.q_proj",
+    "language_model.model.layers.16.self_attn.o_proj",
+    "language_model.model.layers.6.mlp.gate_proj",
+    "vision_tower.vision_model.encoder.layers.1.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.2.self_attn.out_proj",
+    "vision_tower.vision_model.encoder.layers.15.self_attn.v_proj",
+    "language_model.model.layers.14.mlp.down_proj",
+    "language_model.model.layers.21.self_attn.v_proj",
     "language_model.model.layers.21.self_attn.k_proj",
-    "language_model.model.layers.4.self_attn.q_proj",
-    "language_model.model.layers.28.mlp.gate_proj"
+    "vision_tower.vision_model.encoder.layers.0.self_attn.q_proj",
+    "vision_tower.vision_model.encoder.layers.3.self_attn.out_proj",
+    "language_model.model.layers.15.self_attn.q_proj",
+    "language_model.model.layers.19.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.6.self_attn.v_proj",
+    "language_model.model.layers.23.self_attn.o_proj",
+    "language_model.model.layers.16.self_attn.k_proj",
+    "language_model.model.layers.16.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.0.self_attn.k_proj",
+    "vision_tower.vision_model.encoder.layers.17.mlp.fc1",
+    "language_model.model.layers.23.mlp.up_proj",
+    "language_model.model.layers.6.self_attn.k_proj",
+    "language_model.model.layers.22.mlp.down_proj",
+    "language_model.model.layers.22.mlp.up_proj",
+    "language_model.model.layers.28.mlp.up_proj",
+    "language_model.model.layers.21.self_attn.o_proj",
+    "language_model.model.layers.14.self_attn.q_proj",
+    "language_model.model.layers.11.self_attn.v_proj",
+    "vision_tower.vision_model.encoder.layers.23.mlp.fc2",
+    "language_model.model.layers.9.self_attn.q_proj",
+    "language_model.model.layers.13.mlp.up_proj",
+    "vision_tower.vision_model.encoder.layers.12.self_attn.q_proj"
   ],
   "task_type": null,
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:26b071caec4e6cb2271c21ca2fade07025d97de3d13f65095a8ab55dd1fc1e4e
-size 671155888
+oid sha256:5f4814325a26d7be68124810dd17ea255ce0582a9be280eb2704cb68439ca341
+size 784448728

runs/Aug26_22-35-08_c8d12f21deff/events.out.tfevents.1724711723.c8d12f21deff.9122.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aff3b88017dbcf4b28803e58b403f4c4871ae8b81e79ef43ebf8c4f46debf091
+size 6337

runs/Aug27_00-39-25_c8d12f21deff/events.out.tfevents.1724719168.c8d12f21deff.9122.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1fa71ab72852d9ce95408c57467b007ec2db9e81bf1c1cda69ecadd5b8f4f73
+size 6603

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a9705c43870aab9b4f8f7430c4101dad026bcdc2e4f05126d3ac42f0e87f23b3
+oid sha256:b0cbf1c03190ad7d8d562cbd8a447169f5f8e343551f62a275e9fd3c9a4f9c0d
 size 5176