Training in progress, step 27, checkpoint

checkpoint-27/adapter_config.json

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

checkpoint-27/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:98b524b6bafdcc9ffa1a11cc6de923d5236c626b9b82ca24cbbd6acfaf4d1cb7
+oid sha256:f19eb4cf0ea667dd0775291e9b80b3dab897a0a8abefa37b52dacba294331956
 size 137714904

checkpoint-27/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:69b6394bf4f68c9c590cb02d5458ddcd8347e8d3474ec43cc58bcbafcd9a745a
+oid sha256:27e31d7516a6b7165b2cd8b8da11e78fd19b108ab1b337d682c22d208b602a21
 size 275817586

checkpoint-27/trainer_state.json

@@ -10,263 +10,263 @@
   "log_history": [
     {
       "epoch": 0.0006200589055960316,
-      "grad_norm": 0.599152684211731,
+      "grad_norm": 0.6006524562835693,
       "learning_rate": 4.000000000000001e-06,
-      "loss": 1.4569,
+      "loss": 1.4363,
       "step": 1
     },
     {
       "epoch": 0.0012401178111920632,
-      "grad_norm": 0.41296300292015076,
+      "grad_norm": 0.4452102482318878,
       "learning_rate": 8.000000000000001e-06,
-      "loss": 1.1609,
+      "loss": 1.1641,
       "step": 2
     },
     {
       "epoch": 0.0018601767167880949,
-      "grad_norm": 0.3926275372505188,
+      "grad_norm": 0.418393075466156,
       "learning_rate": 1.2e-05,
-      "loss": 0.9938,
+      "loss": 1.0005,
       "step": 3
     },
     {
       "epoch": 0.0018601767167880949,
-      "eval_loss": 1.2600003480911255,
-      "eval_runtime": 49.3684,
-      "eval_samples_per_second": 2.026,
-      "eval_steps_per_second": 2.026,
+      "eval_loss": 1.2559138536453247,
+      "eval_runtime": 50.2632,
+      "eval_samples_per_second": 1.99,
+      "eval_steps_per_second": 1.99,
       "step": 3
     },
     {
       "epoch": 0.0024802356223841263,
-      "grad_norm": 0.4559924006462097,
+      "grad_norm": 0.4599858820438385,
       "learning_rate": 1.6000000000000003e-05,
-      "loss": 1.1897,
+      "loss": 1.1938,
       "step": 4
     },
     {
       "epoch": 0.0031002945279801583,
-      "grad_norm": 0.40210390090942383,
+      "grad_norm": 0.4313387870788574,
       "learning_rate": 2e-05,
-      "loss": 0.9835,
+      "loss": 0.9897,
       "step": 5
     },
     {
       "epoch": 0.0037203534335761897,
-      "grad_norm": 0.45071661472320557,
+      "grad_norm": 0.4715091586112976,
       "learning_rate": 1.9200000000000003e-05,
-      "loss": 1.112,
+      "loss": 1.1105,
       "step": 6
     },
     {
       "epoch": 0.0037203534335761897,
-      "eval_loss": 1.2538120746612549,
-      "eval_runtime": 49.623,
-      "eval_samples_per_second": 2.015,
-      "eval_steps_per_second": 2.015,
+      "eval_loss": 1.2492942810058594,
+      "eval_runtime": 49.6589,
+      "eval_samples_per_second": 2.014,
+      "eval_steps_per_second": 2.014,
       "step": 6
     },
     {
       "epoch": 0.004340412339172222,
-      "grad_norm": 0.44490283727645874,
+      "grad_norm": 0.4590495824813843,
       "learning_rate": 1.8400000000000003e-05,
-      "loss": 1.0809,
+      "loss": 1.0824,
       "step": 7
     },
     {
       "epoch": 0.004960471244768253,
-      "grad_norm": 0.6036495566368103,
+      "grad_norm": 0.6232957243919373,
       "learning_rate": 1.76e-05,
-      "loss": 1.2205,
+      "loss": 1.2176,
       "step": 8
     },
     {
       "epoch": 0.005580530150364285,
-      "grad_norm": 0.4536096751689911,
+      "grad_norm": 0.4820755422115326,
       "learning_rate": 1.6800000000000002e-05,
-      "loss": 1.0715,
+      "loss": 1.0774,
       "step": 9
     },
     {
       "epoch": 0.005580530150364285,
-      "eval_loss": 1.2402353286743164,
-      "eval_runtime": 49.777,
-      "eval_samples_per_second": 2.009,
-      "eval_steps_per_second": 2.009,
+      "eval_loss": 1.2356064319610596,
+      "eval_runtime": 49.9029,
+      "eval_samples_per_second": 2.004,
+      "eval_steps_per_second": 2.004,
       "step": 9
     },
     {
       "epoch": 0.0062005890559603165,
-      "grad_norm": 0.5870596170425415,
+      "grad_norm": 0.614038348197937,
       "learning_rate": 1.6000000000000003e-05,
-      "loss": 1.3001,
+      "loss": 1.2989,
       "step": 10
     },
     {
       "epoch": 0.0068206479615563476,
-      "grad_norm": 0.5543638467788696,
+      "grad_norm": 0.5838705897331238,
       "learning_rate": 1.5200000000000002e-05,
-      "loss": 1.0459,
+      "loss": 1.05,
       "step": 11
     },
     {
       "epoch": 0.0074407068671523795,
-      "grad_norm": 0.6657857894897461,
+      "grad_norm": 0.6922520995140076,
       "learning_rate": 1.4400000000000001e-05,
-      "loss": 1.4109,
+      "loss": 1.3979,
       "step": 12
     },
     {
       "epoch": 0.0074407068671523795,
-      "eval_loss": 1.2233326435089111,
-      "eval_runtime": 49.0096,
-      "eval_samples_per_second": 2.04,
-      "eval_steps_per_second": 2.04,
+      "eval_loss": 1.2183419466018677,
+      "eval_runtime": 50.5422,
+      "eval_samples_per_second": 1.979,
+      "eval_steps_per_second": 1.979,
       "step": 12
     },
     {
       "epoch": 0.008060765772748411,
-      "grad_norm": 0.5861024856567383,
+      "grad_norm": 0.6026104688644409,
       "learning_rate": 1.3600000000000002e-05,
-      "loss": 1.1955,
+      "loss": 1.1974,
       "step": 13
     },
     {
       "epoch": 0.008680824678344443,
-      "grad_norm": 0.436432808637619,
+      "grad_norm": 0.46571722626686096,
       "learning_rate": 1.2800000000000001e-05,
-      "loss": 0.9885,
+      "loss": 0.9946,
       "step": 14
     },
     {
       "epoch": 0.009300883583940475,
-      "grad_norm": 0.5418331623077393,
+      "grad_norm": 0.5651541948318481,
       "learning_rate": 1.2e-05,
-      "loss": 1.0504,
+      "loss": 1.0568,
       "step": 15
     },
     {
       "epoch": 0.009300883583940475,
-      "eval_loss": 1.2044346332550049,
-      "eval_runtime": 49.0412,
-      "eval_samples_per_second": 2.039,
-      "eval_steps_per_second": 2.039,
+      "eval_loss": 1.1998839378356934,
+      "eval_runtime": 50.192,
+      "eval_samples_per_second": 1.992,
+      "eval_steps_per_second": 1.992,
       "step": 15
     },
     {
       "epoch": 0.009920942489536505,
-      "grad_norm": 0.4635182023048401,
+      "grad_norm": 0.4988487958908081,
       "learning_rate": 1.1200000000000001e-05,
-      "loss": 0.9267,
+      "loss": 0.9318,
       "step": 16
     },
     {
       "epoch": 0.010541001395132537,
-      "grad_norm": 0.4401405155658722,
+      "grad_norm": 0.4823167324066162,
       "learning_rate": 1.04e-05,
-      "loss": 1.0594,
+      "loss": 1.0622,
       "step": 17
     },
     {
       "epoch": 0.01116106030072857,
-      "grad_norm": 0.49569207429885864,
+      "grad_norm": 0.5371574759483337,
       "learning_rate": 9.600000000000001e-06,
-      "loss": 0.9993,
+      "loss": 1.0033,
       "step": 18
     },
     {
       "epoch": 0.01116106030072857,
-      "eval_loss": 1.1875802278518677,
-      "eval_runtime": 49.6265,
-      "eval_samples_per_second": 2.015,
-      "eval_steps_per_second": 2.015,
+      "eval_loss": 1.1833301782608032,
+      "eval_runtime": 50.0653,
+      "eval_samples_per_second": 1.997,
+      "eval_steps_per_second": 1.997,
       "step": 18
     },
     {
       "epoch": 0.011781119206324601,
-      "grad_norm": 0.5453494191169739,
+      "grad_norm": 0.5795661807060242,
       "learning_rate": 8.8e-06,
-      "loss": 0.986,
+      "loss": 0.9854,
       "step": 19
     },
     {
       "epoch": 0.012401178111920633,
-      "grad_norm": 0.3823815584182739,
+      "grad_norm": 0.41762855648994446,
       "learning_rate": 8.000000000000001e-06,
-      "loss": 0.8053,
+      "loss": 0.8116,
       "step": 20
     },
     {
       "epoch": 0.013021237017516665,
-      "grad_norm": 0.551879346370697,
+      "grad_norm": 0.6034093499183655,
       "learning_rate": 7.2000000000000005e-06,
-      "loss": 1.1088,
+      "loss": 1.1081,
       "step": 21
     },
     {
       "epoch": 0.013021237017516665,
-      "eval_loss": 1.1738098859786987,
-      "eval_runtime": 49.0729,
-      "eval_samples_per_second": 2.038,
-      "eval_steps_per_second": 2.038,
+      "eval_loss": 1.1694728136062622,
+      "eval_runtime": 50.5031,
+      "eval_samples_per_second": 1.98,
+      "eval_steps_per_second": 1.98,
       "step": 21
     },
     {
       "epoch": 0.013641295923112695,
-      "grad_norm": 0.5134972333908081,
+      "grad_norm": 0.5518467426300049,
       "learning_rate": 6.4000000000000006e-06,
       "loss": 1.0851,
       "step": 22
     },
     {
       "epoch": 0.014261354828708727,
-      "grad_norm": 0.6215815544128418,
+      "grad_norm": 0.7419660687446594,
       "learning_rate": 5.600000000000001e-06,
-      "loss": 1.209,
+      "loss": 1.2086,
       "step": 23
     },
     {
       "epoch": 0.014881413734304759,
-      "grad_norm": 0.49390730261802673,
+      "grad_norm": 0.5225710272789001,
       "learning_rate": 4.800000000000001e-06,
-      "loss": 1.0424,
+      "loss": 1.044,
       "step": 24
     },
     {
       "epoch": 0.014881413734304759,
-      "eval_loss": 1.1632879972457886,
-      "eval_runtime": 49.3634,
-      "eval_samples_per_second": 2.026,
-      "eval_steps_per_second": 2.026,
+      "eval_loss": 1.1590291261672974,
+      "eval_runtime": 50.214,
+      "eval_samples_per_second": 1.991,
+      "eval_steps_per_second": 1.991,
       "step": 24
     },
     {
       "epoch": 0.01550147263990079,
-      "grad_norm": 0.4980444014072418,
+      "grad_norm": 0.539212167263031,
       "learning_rate": 4.000000000000001e-06,
-      "loss": 1.043,
+      "loss": 1.0468,
       "step": 25
     },
     {
       "epoch": 0.016121531545496823,
-      "grad_norm": 0.4858083724975586,
+      "grad_norm": 0.4953361451625824,
       "learning_rate": 3.2000000000000003e-06,
-      "loss": 0.9258,
+      "loss": 0.9279,
       "step": 26
     },
     {
       "epoch": 0.016741590451092855,
-      "grad_norm": 0.4682457149028778,
+      "grad_norm": 0.4864480793476105,
       "learning_rate": 2.4000000000000003e-06,
-      "loss": 0.9423,
+      "loss": 0.9413,
       "step": 27
     },
     {
       "epoch": 0.016741590451092855,
-      "eval_loss": 1.1565324068069458,
-      "eval_runtime": 49.7291,
-      "eval_samples_per_second": 2.011,
-      "eval_steps_per_second": 2.011,
+      "eval_loss": 1.1525880098342896,
+      "eval_runtime": 50.5444,
+      "eval_samples_per_second": 1.978,
+      "eval_steps_per_second": 1.978,
       "step": 27
     }
   ],
@@ -287,7 +287,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 5.708973277855334e+16,
+  "total_flos": 5.791479069566976e+16,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null

checkpoint-27/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bf98514c81a51106969007e868d7ab8fb036a6bedc2987ddadf4a06b2a43f04f
+oid sha256:522b553cd97c4d175d86850f43dd8230595c851fb1c776adff1402d853af53e9
 size 5752