Training in progress, step 21, checkpoint

checkpoint-21/adapter_config.json

@@ -23,13 +23,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "down_proj",
-    "q_proj",
-    "up_proj",
     "k_proj",
-    "gate_proj",
     "o_proj",
-    "v_proj"
+    "v_proj",
+    "gate_proj",
+    "up_proj",
+    "down_proj",
+    "q_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

checkpoint-21/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:44e71a3c293f33d3d00ba5d07bf5e221f22eef99efabe5ba42d900420cbda86a
+oid sha256:675eddc86571193eb0870fe0e5c88ece07e4fa00ca342f5222e3d9c04722d273
 size 83945296

checkpoint-21/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0fd303d88d6316b401f13059794be833aabda5daa320f1db186bbc64962351e7
+oid sha256:b7b1563f73cd76301fbeba65c91e5e2c22bad6a25919a99713b66b8089d4442a
 size 168149074

checkpoint-21/trainer_state.json

@@ -10,205 +10,205 @@
   "log_history": [
     {
       "epoch": 0.0006200589055960316,
-      "grad_norm": 0.7733756303787231,
+      "grad_norm": 0.7731499075889587,
       "learning_rate": 4.000000000000001e-06,
       "loss": 1.166,
       "step": 1
     },
     {
       "epoch": 0.0012401178111920632,
-      "grad_norm": 0.34648507833480835,
+      "grad_norm": 0.346652090549469,
       "learning_rate": 8.000000000000001e-06,
       "loss": 0.9594,
       "step": 2
     },
     {
       "epoch": 0.0018601767167880949,
-      "grad_norm": 0.3973380923271179,
+      "grad_norm": 0.39937612414360046,
       "learning_rate": 1.2e-05,
-      "loss": 0.8281,
+      "loss": 0.828,
       "step": 3
     },
     {
       "epoch": 0.0018601767167880949,
-      "eval_loss": 1.0490500926971436,
-      "eval_runtime": 28.8129,
-      "eval_samples_per_second": 3.471,
-      "eval_steps_per_second": 3.471,
+      "eval_loss": 1.0491173267364502,
+      "eval_runtime": 26.0459,
+      "eval_samples_per_second": 3.839,
+      "eval_steps_per_second": 3.839,
       "step": 3
     },
     {
       "epoch": 0.0024802356223841263,
-      "grad_norm": 0.342887282371521,
+      "grad_norm": 0.3457990288734436,
       "learning_rate": 1.6000000000000003e-05,
-      "loss": 0.9881,
+      "loss": 0.988,
       "step": 4
     },
     {
       "epoch": 0.0031002945279801583,
-      "grad_norm": 0.46208909153938293,
+      "grad_norm": 0.513918399810791,
       "learning_rate": 2e-05,
       "loss": 0.8319,
       "step": 5
     },
     {
       "epoch": 0.0037203534335761897,
-      "grad_norm": 0.4722931981086731,
+      "grad_norm": 0.47683727741241455,
       "learning_rate": 1.9200000000000003e-05,
-      "loss": 0.9364,
+      "loss": 0.9366,
       "step": 6
     },
     {
       "epoch": 0.0037203534335761897,
-      "eval_loss": 1.0428661108016968,
-      "eval_runtime": 26.3402,
-      "eval_samples_per_second": 3.796,
-      "eval_steps_per_second": 3.796,
+      "eval_loss": 1.0431987047195435,
+      "eval_runtime": 26.2352,
+      "eval_samples_per_second": 3.812,
+      "eval_steps_per_second": 3.812,
       "step": 6
     },
     {
       "epoch": 0.004340412339172222,
-      "grad_norm": 0.4379231333732605,
+      "grad_norm": 0.4349649250507355,
       "learning_rate": 1.8400000000000003e-05,
-      "loss": 0.9073,
+      "loss": 0.9076,
       "step": 7
     },
     {
       "epoch": 0.004960471244768253,
-      "grad_norm": 0.5576643943786621,
+      "grad_norm": 0.5443339943885803,
       "learning_rate": 1.76e-05,
-      "loss": 0.9823,
+      "loss": 0.9829,
       "step": 8
     },
     {
       "epoch": 0.005580530150364285,
-      "grad_norm": 0.4029313325881958,
+      "grad_norm": 0.3990451395511627,
       "learning_rate": 1.6800000000000002e-05,
       "loss": 0.9045,
       "step": 9
     },
     {
       "epoch": 0.005580530150364285,
-      "eval_loss": 1.0290312767028809,
-      "eval_runtime": 26.3678,
-      "eval_samples_per_second": 3.793,
-      "eval_steps_per_second": 3.793,
+      "eval_loss": 1.0292819738388062,
+      "eval_runtime": 26.3477,
+      "eval_samples_per_second": 3.795,
+      "eval_steps_per_second": 3.795,
       "step": 9
     },
     {
       "epoch": 0.0062005890559603165,
-      "grad_norm": 0.4460880160331726,
+      "grad_norm": 0.44750896096229553,
       "learning_rate": 1.6000000000000003e-05,
-      "loss": 1.0763,
+      "loss": 1.0768,
       "step": 10
     },
     {
       "epoch": 0.0068206479615563476,
-      "grad_norm": 0.7060808539390564,
+      "grad_norm": 0.6865904331207275,
       "learning_rate": 1.5200000000000002e-05,
-      "loss": 0.8713,
+      "loss": 0.8717,
       "step": 11
     },
     {
       "epoch": 0.0074407068671523795,
-      "grad_norm": 0.6190696954727173,
+      "grad_norm": 0.548491895198822,
       "learning_rate": 1.4400000000000001e-05,
-      "loss": 1.1753,
+      "loss": 1.1758,
       "step": 12
     },
     {
       "epoch": 0.0074407068671523795,
-      "eval_loss": 1.0143623352050781,
-      "eval_runtime": 26.7829,
-      "eval_samples_per_second": 3.734,
-      "eval_steps_per_second": 3.734,
+      "eval_loss": 1.0153621435165405,
+      "eval_runtime": 26.5025,
+      "eval_samples_per_second": 3.773,
+      "eval_steps_per_second": 3.773,
       "step": 12
     },
     {
       "epoch": 0.008060765772748411,
-      "grad_norm": 0.5669320225715637,
+      "grad_norm": 0.5520397424697876,
       "learning_rate": 1.3600000000000002e-05,
-      "loss": 0.991,
+      "loss": 0.992,
       "step": 13
     },
     {
       "epoch": 0.008680824678344443,
-      "grad_norm": 0.49686795473098755,
+      "grad_norm": 0.381074458360672,
       "learning_rate": 1.2800000000000001e-05,
-      "loss": 0.8497,
+      "loss": 0.8504,
       "step": 14
     },
     {
       "epoch": 0.009300883583940475,
-      "grad_norm": 0.5733391642570496,
+      "grad_norm": 0.41114744544029236,
       "learning_rate": 1.2e-05,
-      "loss": 0.8679,
+      "loss": 0.8685,
       "step": 15
     },
     {
       "epoch": 0.009300883583940475,
-      "eval_loss": 0.9997775554656982,
-      "eval_runtime": 26.7666,
-      "eval_samples_per_second": 3.736,
-      "eval_steps_per_second": 3.736,
+      "eval_loss": 1.0006165504455566,
+      "eval_runtime": 26.4086,
+      "eval_samples_per_second": 3.787,
+      "eval_steps_per_second": 3.787,
       "step": 15
     },
     {
       "epoch": 0.009920942489536505,
-      "grad_norm": 0.5733532309532166,
+      "grad_norm": 0.5197786688804626,
       "learning_rate": 1.1200000000000001e-05,
-      "loss": 0.8173,
+      "loss": 0.8176,
       "step": 16
     },
     {
       "epoch": 0.010541001395132537,
-      "grad_norm": 0.375442773103714,
+      "grad_norm": 0.3905779719352722,
       "learning_rate": 1.04e-05,
-      "loss": 0.9054,
+      "loss": 0.9055,
       "step": 17
     },
     {
       "epoch": 0.01116106030072857,
-      "grad_norm": 0.4431408643722534,
+      "grad_norm": 0.43444159626960754,
       "learning_rate": 9.600000000000001e-06,
-      "loss": 0.8451,
+      "loss": 0.8452,
       "step": 18
     },
     {
       "epoch": 0.01116106030072857,
-      "eval_loss": 0.986146092414856,
-      "eval_runtime": 26.7104,
-      "eval_samples_per_second": 3.744,
-      "eval_steps_per_second": 3.744,
+      "eval_loss": 0.9866559505462646,
+      "eval_runtime": 26.3639,
+      "eval_samples_per_second": 3.793,
+      "eval_steps_per_second": 3.793,
       "step": 18
     },
     {
       "epoch": 0.011781119206324601,
-      "grad_norm": 0.5397400259971619,
+      "grad_norm": 0.5130135416984558,
       "learning_rate": 8.8e-06,
-      "loss": 0.8274,
+      "loss": 0.8276,
       "step": 19
     },
     {
       "epoch": 0.012401178111920633,
-      "grad_norm": 0.3542424440383911,
+      "grad_norm": 0.35828444361686707,
       "learning_rate": 8.000000000000001e-06,
-      "loss": 0.6586,
+      "loss": 0.6587,
       "step": 20
     },
     {
       "epoch": 0.013021237017516665,
-      "grad_norm": 0.45993754267692566,
+      "grad_norm": 0.4369220435619354,
       "learning_rate": 7.2000000000000005e-06,
-      "loss": 0.937,
+      "loss": 0.9372,
       "step": 21
     },
     {
       "epoch": 0.013021237017516665,
-      "eval_loss": 0.9746847748756409,
-      "eval_runtime": 26.6243,
-      "eval_samples_per_second": 3.756,
-      "eval_steps_per_second": 3.756,
+      "eval_loss": 0.9746317267417908,
+      "eval_runtime": 26.2985,
+      "eval_samples_per_second": 3.803,
+      "eval_steps_per_second": 3.803,
       "step": 21
     }
   ],

checkpoint-21/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7c433cc5009675ccadca19513c8ae659f957c78885071da55c2a8d32a98d517a
+oid sha256:cbe2fb31cf4c04f893e5299faa77267af76b8447f922d82c2d6c0200e1fe225f
 size 5688