Training in progress, step 1000, checkpoint

checkpoint-1000/adapter_config.json

@@ -24,12 +24,12 @@
   "revision": null,
   "target_modules": [
     "up_proj",
-    "down_proj",
     "v_proj",
     "q_proj",
     "k_proj",
+    "o_proj",
     "gate_proj",
-    "o_proj"
+    "down_proj"
   ],
   "task_type": null,
   "use_dora": false,

checkpoint-1000/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:604a7712dc427d2a77cb50ba6c3d44da70994bd962a9c94159b4cae750f42aaa
+oid sha256:583b47b73ad945f64d1474c5df017c3798bf840a874e33f8d3b2e248c422fb46
 size 400616360

checkpoint-1000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:21e84e570501e757ddc950a0dc0009b596379db2c0e7e926194f4187d6fcf529
+oid sha256:66b160a0f5de2302d2f6bd9584ae8f101dbf0352881b873cb8359bc9bd1b14d4
 size 203945684

checkpoint-1000/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2f170124a40827acf047877d5bc75e0ec02eb7f5cad9efe5bdf237c40d7528bf
+oid sha256:49a7dae92c819aa704449c6dd0c24e323ff5d229baff9b5ababb166b4285eea1
 size 14308

checkpoint-1000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5bc2db573163ebfd1632bebc3d65513a05d4cee82b0b1637f8182262dcbd49f4
+oid sha256:9c6e566386e44f4a96260c39d65dda7afa6f21f00326a74209dab643f15ab7ae
 size 1064

checkpoint-1000/trainer_state.json

@@ -10,62 +10,62 @@
   "log_history": [
     {
       "epoch": 0.2333994631812347,
-      "grad_norm": 5.120560169219971,
-      "learning_rate": 8.987402315498223e-05,
-      "loss": 0.4681,
+      "grad_norm": 3.4083237648010254,
+      "learning_rate": 8.996514402527381e-05,
+      "loss": 0.2768,
       "step": 250
     },
     {
       "epoch": 0.2333994631812347,
-      "eval_loss": 0.3557259440422058,
-      "eval_runtime": 26.5914,
-      "eval_samples_per_second": 3.272,
-      "eval_steps_per_second": 3.272,
+      "eval_loss": 0.15054132044315338,
+      "eval_runtime": 39.385,
+      "eval_samples_per_second": 2.209,
+      "eval_steps_per_second": 2.209,
       "step": 250
     },
     {
       "epoch": 0.4667989263624694,
-      "grad_norm": 3.340378999710083,
-      "learning_rate": 5.828471682626175e-05,
-      "loss": 0.3022,
+      "grad_norm": 7.858141899108887,
+      "learning_rate": 5.843391598090062e-05,
+      "loss": 0.1092,
       "step": 500
     },
     {
       "epoch": 0.4667989263624694,
-      "eval_loss": 0.3018253445625305,
-      "eval_runtime": 26.6712,
-      "eval_samples_per_second": 3.262,
-      "eval_steps_per_second": 3.262,
+      "eval_loss": 0.11498060077428818,
+      "eval_runtime": 39.2923,
+      "eval_samples_per_second": 2.214,
+      "eval_steps_per_second": 2.214,
       "step": 500
     },
     {
       "epoch": 0.7001983895437041,
-      "grad_norm": 3.158240556716919,
-      "learning_rate": 2.2174321662025427e-05,
-      "loss": 0.2736,
+      "grad_norm": 2.502495765686035,
+      "learning_rate": 2.2300178981718183e-05,
+      "loss": 0.0945,
       "step": 750
     },
     {
       "epoch": 0.7001983895437041,
-      "eval_loss": 0.2686610519886017,
-      "eval_runtime": 26.6022,
-      "eval_samples_per_second": 3.27,
-      "eval_steps_per_second": 3.27,
+      "eval_loss": 0.10474376380443573,
+      "eval_runtime": 39.2665,
+      "eval_samples_per_second": 2.216,
+      "eval_steps_per_second": 2.216,
       "step": 750
     },
     {
       "epoch": 0.9335978527249388,
-      "grad_norm": 3.1761715412139893,
-      "learning_rate": 1.2487983905362933e-06,
-      "loss": 0.2525,
+      "grad_norm": 2.4208877086639404,
+      "learning_rate": 1.2826317302122581e-06,
+      "loss": 0.0862,
       "step": 1000
     },
     {
       "epoch": 0.9335978527249388,
-      "eval_loss": 0.259181410074234,
-      "eval_runtime": 26.6606,
-      "eval_samples_per_second": 3.263,
-      "eval_steps_per_second": 3.263,
+      "eval_loss": 0.10126293450593948,
+      "eval_runtime": 39.3834,
+      "eval_samples_per_second": 2.209,
+      "eval_steps_per_second": 2.209,
       "step": 1000
     }
   ],
@@ -86,7 +86,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.5404854903863296e+16,
+  "total_flos": 8.213169594641818e+16,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null