Training in progress, step 450, checkpoint

last-checkpoint/adapter_config.json

@@ -22,8 +22,8 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "k_proj",
-    "q_proj"
+    "q_proj",
+    "k_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4de48b4cf0b79853ac74250265fc0fb5005b73f0afabe10787d5eca93366b80f
+oid sha256:2cd82b375151b5b5f9f0a58b8286fa2a7b7e910cea5c2d600b74e22d38bdbcca
 size 527048968

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ba857a8ce8386620cc7ae2730f87baa56d2b8d7e6ff77c08d0e9e683ea828925
+oid sha256:88e869796afac1a7a34f750f8422f24ec974478aafa79ce79e006a42fe78f21d
 size 1054135994

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:280ae9c0809eceeb464f568a46cee1cc6c997124b85d7c1db2a216ff1d3932aa
+oid sha256:2616293446c1c2f2c6b0e270c4bb50dd25782b075145433b30550e3bb3a1845c
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b5ff073b2d9654300593116ff77249d188bd5752ecbe3d0cb800964ea5d21c80
+oid sha256:da0b9a1e2fa3da24e9af8b74787d70ff4c95d9bc9b74eeab135df0350a00b462
 size 1064

last-checkpoint/trainer_state.json

@@ -1,464 +1,351 @@
 {
-  "best_metric": 0.7164609432220459,
+  "best_metric": 0.7168648838996887,
   "best_model_checkpoint": "./output/checkpoint-450",
-  "epoch": 26.08695652173913,
+  "epoch": 19.565217391304348,
   "eval_steps": 150,
-  "global_step": 600,
+  "global_step": 450,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.43478260869565216,
-      "grad_norm": 1.5021647214889526,
+      "grad_norm": 1.5021412372589111,
       "learning_rate": 3e-06,
-      "loss": 0.9064,
+      "loss": 0.9061,
       "step": 10
     },
     {
       "epoch": 0.8695652173913043,
-      "grad_norm": 1.6793277263641357,
+      "grad_norm": 1.687071442604065,
       "learning_rate": 6e-06,
-      "loss": 0.9024,
+      "loss": 0.9026,
       "step": 20
     },
     {
       "epoch": 1.3043478260869565,
-      "grad_norm": 1.7296056747436523,
+      "grad_norm": 1.7296651601791382,
       "learning_rate": 9e-06,
-      "loss": 0.9,
+      "loss": 0.9003,
       "step": 30
     },
     {
       "epoch": 1.7391304347826086,
-      "grad_norm": 1.4458602666854858,
+      "grad_norm": 1.4536513090133667,
       "learning_rate": 1.2e-05,
-      "loss": 0.9086,
+      "loss": 0.9093,
       "step": 40
     },
     {
       "epoch": 2.1739130434782608,
-      "grad_norm": 1.3516699075698853,
+      "grad_norm": 1.351776361465454,
       "learning_rate": 1.5e-05,
-      "loss": 0.8359,
+      "loss": 0.8358,
       "step": 50
     },
     {
       "epoch": 2.608695652173913,
-      "grad_norm": 2.03114914894104,
+      "grad_norm": 2.031055212020874,
       "learning_rate": 1.8e-05,
-      "loss": 0.8892,
+      "loss": 0.8894,
       "step": 60
     },
     {
       "epoch": 3.0434782608695654,
-      "grad_norm": 1.4920729398727417,
+      "grad_norm": 1.4920436143875122,
       "learning_rate": 2.1e-05,
-      "loss": 0.8913,
+      "loss": 0.8915,
       "step": 70
     },
     {
       "epoch": 3.4782608695652173,
-      "grad_norm": 1.7293280363082886,
+      "grad_norm": 1.729347586631775,
       "learning_rate": 2.4e-05,
-      "loss": 0.8234,
+      "loss": 0.8237,
       "step": 80
     },
     {
       "epoch": 3.9130434782608696,
-      "grad_norm": 1.4240705966949463,
+      "grad_norm": 1.4314346313476562,
       "learning_rate": 2.7000000000000002e-05,
       "loss": 0.8529,
       "step": 90
     },
     {
       "epoch": 4.3478260869565215,
-      "grad_norm": 1.3656291961669922,
+      "grad_norm": 1.3657989501953125,
       "learning_rate": 3e-05,
       "loss": 0.8649,
       "step": 100
     },
     {
       "epoch": 4.782608695652174,
-      "grad_norm": 2.198136806488037,
+      "grad_norm": 2.1989877223968506,
       "learning_rate": 2.999999702723963e-05,
       "loss": 0.8227,
       "step": 110
     },
     {
       "epoch": 5.217391304347826,
-      "grad_norm": 1.0726988315582275,
+      "grad_norm": 1.0724585056304932,
       "learning_rate": 2.9999988108959687e-05,
-      "loss": 0.765,
+      "loss": 0.7655,
       "step": 120
     },
     {
       "epoch": 5.6521739130434785,
-      "grad_norm": 1.5672078132629395,
+      "grad_norm": 1.5675039291381836,
       "learning_rate": 2.9999973245163716e-05,
       "loss": 0.7413,
       "step": 130
     },
     {
       "epoch": 6.086956521739131,
-      "grad_norm": 1.9071016311645508,
+      "grad_norm": 1.9068934917449951,
       "learning_rate": 2.99999524358576e-05,
-      "loss": 0.7655,
+      "loss": 0.7653,
       "step": 140
     },
     {
       "epoch": 6.521739130434782,
-      "grad_norm": 1.1219961643218994,
+      "grad_norm": 1.121341586112976,
       "learning_rate": 2.9999925681049593e-05,
       "loss": 0.7855,
       "step": 150
     },
     {
       "epoch": 6.521739130434782,
-      "eval_loss": 0.7960985898971558,
-      "eval_runtime": 0.5137,
-      "eval_samples_per_second": 19.465,
-      "eval_steps_per_second": 19.465,
+      "eval_loss": 0.7960079908370972,
+      "eval_runtime": 0.4677,
+      "eval_samples_per_second": 21.382,
+      "eval_steps_per_second": 21.382,
       "step": 150
     },
     {
       "epoch": 6.956521739130435,
-      "grad_norm": 1.5404783487319946,
+      "grad_norm": 1.5398368835449219,
       "learning_rate": 2.9999892980750297e-05,
-      "loss": 0.6588,
+      "loss": 0.6587,
       "step": 160
     },
     {
       "epoch": 7.391304347826087,
-      "grad_norm": 1.3464210033416748,
+      "grad_norm": 1.346014142036438,
       "learning_rate": 2.9999854334972675e-05,
       "loss": 0.739,
       "step": 170
     },
     {
       "epoch": 7.826086956521739,
-      "grad_norm": 1.7264870405197144,
+      "grad_norm": 1.7263554334640503,
       "learning_rate": 2.999980974373204e-05,
-      "loss": 0.7292,
+      "loss": 0.729,
       "step": 180
     },
     {
       "epoch": 8.26086956521739,
-      "grad_norm": 1.5393368005752563,
+      "grad_norm": 1.5398695468902588,
       "learning_rate": 2.9999759207046075e-05,
-      "loss": 0.6244,
+      "loss": 0.6246,
       "step": 190
     },
     {
       "epoch": 8.695652173913043,
-      "grad_norm": 1.7342770099639893,
+      "grad_norm": 1.7423065900802612,
       "learning_rate": 2.9999702724934804e-05,
-      "loss": 0.6764,
+      "loss": 0.6762,
       "step": 200
     },
     {
       "epoch": 9.130434782608695,
-      "grad_norm": 1.0416548252105713,
+      "grad_norm": 1.0418452024459839,
       "learning_rate": 2.999964029742062e-05,
-      "loss": 0.6521,
+      "loss": 0.6522,
       "step": 210
     },
     {
       "epoch": 9.565217391304348,
-      "grad_norm": 1.2193585634231567,
+      "grad_norm": 1.2121878862380981,
       "learning_rate": 2.9999571924528263e-05,
-      "loss": 0.5592,
+      "loss": 0.559,
       "step": 220
     },
     {
       "epoch": 10.0,
-      "grad_norm": 1.5286610126495361,
+      "grad_norm": 1.5282926559448242,
       "learning_rate": 2.9999497606284837e-05,
-      "loss": 0.756,
+      "loss": 0.7558,
       "step": 230
     },
     {
       "epoch": 10.434782608695652,
-      "grad_norm": 1.427451252937317,
+      "grad_norm": 1.4222936630249023,
       "learning_rate": 2.9999417342719796e-05,
-      "loss": 0.7118,
+      "loss": 0.7119,
       "step": 240
     },
     {
       "epoch": 10.869565217391305,
-      "grad_norm": 0.9794074296951294,
+      "grad_norm": 0.9751431941986084,
       "learning_rate": 2.9999331133864956e-05,
-      "loss": 0.5897,
+      "loss": 0.59,
       "step": 250
     },
     {
       "epoch": 11.304347826086957,
-      "grad_norm": 1.194082260131836,
+      "grad_norm": 1.194357991218567,
       "learning_rate": 2.9999238979754485e-05,
-      "loss": 0.6548,
+      "loss": 0.6547,
       "step": 260
     },
     {
       "epoch": 11.73913043478261,
-      "grad_norm": 1.0489437580108643,
+      "grad_norm": 1.049717903137207,
       "learning_rate": 2.999914088042492e-05,
       "loss": 0.6474,
       "step": 270
     },
     {
       "epoch": 12.173913043478262,
-      "grad_norm": 1.3105570077896118,
+      "grad_norm": 1.3123618364334106,
       "learning_rate": 2.9999036835915132e-05,
       "loss": 0.5936,
       "step": 280
     },
     {
       "epoch": 12.608695652173914,
-      "grad_norm": 1.0830979347229004,
+      "grad_norm": 1.0834040641784668,
       "learning_rate": 2.9998926846266365e-05,
       "loss": 0.6328,
       "step": 290
     },
     {
       "epoch": 13.043478260869565,
-      "grad_norm": 1.3943493366241455,
+      "grad_norm": 1.3875536918640137,
       "learning_rate": 2.9998810911522213e-05,
-      "loss": 0.5806,
+      "loss": 0.5807,
       "step": 300
     },
     {
       "epoch": 13.043478260869565,
-      "eval_loss": 0.7312111854553223,
-      "eval_runtime": 0.6239,
-      "eval_samples_per_second": 16.028,
-      "eval_steps_per_second": 16.028,
+      "eval_loss": 0.7317642569541931,
+      "eval_runtime": 0.6288,
+      "eval_samples_per_second": 15.902,
+      "eval_steps_per_second": 15.902,
       "step": 300
     },
     {
       "epoch": 13.478260869565217,
-      "grad_norm": 1.7821625471115112,
+      "grad_norm": 1.779242753982544,
       "learning_rate": 2.9998689031728636e-05,
-      "loss": 0.5143,
+      "loss": 0.5146,
       "step": 310
     },
     {
       "epoch": 13.91304347826087,
-      "grad_norm": 1.5245319604873657,
+      "grad_norm": 1.5307729244232178,
       "learning_rate": 2.9998561206933938e-05,
-      "loss": 0.6494,
+      "loss": 0.6497,
       "step": 320
     },
     {
       "epoch": 14.347826086956522,
-      "grad_norm": 1.479852557182312,
+      "grad_norm": 1.479602575302124,
       "learning_rate": 2.9998427437188786e-05,
-      "loss": 0.5744,
+      "loss": 0.5746,
       "step": 330
     },
     {
       "epoch": 14.782608695652174,
-      "grad_norm": 1.318290114402771,
+      "grad_norm": 1.3181294202804565,
       "learning_rate": 2.99982877225462e-05,
       "loss": 0.6014,
       "step": 340
     },
     {
       "epoch": 15.217391304347826,
-      "grad_norm": 0.9775345921516418,
+      "grad_norm": 0.9750956892967224,
       "learning_rate": 2.9998142063061564e-05,
       "loss": 0.4988,
       "step": 350
     },
     {
       "epoch": 15.652173913043478,
-      "grad_norm": 1.642491340637207,
+      "grad_norm": 1.6386020183563232,
       "learning_rate": 2.9997990458792603e-05,
-      "loss": 0.5628,
+      "loss": 0.5627,
       "step": 360
     },
     {
       "epoch": 16.08695652173913,
-      "grad_norm": 1.6377389430999756,
+      "grad_norm": 1.635014533996582,
       "learning_rate": 2.9997832909799417e-05,
-      "loss": 0.6675,
+      "loss": 0.6674,
       "step": 370
     },
     {
       "epoch": 16.52173913043478,
-      "grad_norm": 0.9537637829780579,
+      "grad_norm": 0.9534231424331665,
       "learning_rate": 2.9997669416144452e-05,
-      "loss": 0.5133,
+      "loss": 0.513,
       "step": 380
     },
     {
       "epoch": 16.956521739130434,
-      "grad_norm": 0.9352350234985352,
+      "grad_norm": 0.9356161952018738,
       "learning_rate": 2.999749997789251e-05,
-      "loss": 0.5792,
+      "loss": 0.5795,
       "step": 390
     },
     {
       "epoch": 17.391304347826086,
-      "grad_norm": 1.112601399421692,
+      "grad_norm": 1.112004280090332,
       "learning_rate": 2.9997324595110743e-05,
-      "loss": 0.5175,
+      "loss": 0.518,
       "step": 400
     },
     {
       "epoch": 17.82608695652174,
-      "grad_norm": 1.284747838973999,
+      "grad_norm": 1.2841087579727173,
       "learning_rate": 2.9997143267868683e-05,
-      "loss": 0.5879,
+      "loss": 0.5878,
       "step": 410
     },
     {
       "epoch": 18.26086956521739,
-      "grad_norm": 1.1622997522354126,
+      "grad_norm": 1.1624560356140137,
       "learning_rate": 2.9996955996238192e-05,
-      "loss": 0.5061,
+      "loss": 0.5056,
       "step": 420
     },
     {
       "epoch": 18.695652173913043,
-      "grad_norm": 1.1977813243865967,
+      "grad_norm": 1.2015905380249023,
       "learning_rate": 2.9996762780293503e-05,
-      "loss": 0.5318,
+      "loss": 0.5317,
       "step": 430
     },
     {
       "epoch": 19.130434782608695,
-      "grad_norm": 1.2117533683776855,
+      "grad_norm": 1.2141062021255493,
       "learning_rate": 2.9996563620111197e-05,
-      "loss": 0.5336,
+      "loss": 0.5335,
       "step": 440
     },
     {
       "epoch": 19.565217391304348,
-      "grad_norm": 1.4293084144592285,
+      "grad_norm": 1.4245619773864746,
       "learning_rate": 2.9996358515770218e-05,
-      "loss": 0.5678,
+      "loss": 0.5677,
       "step": 450
     },
     {
       "epoch": 19.565217391304348,
-      "eval_loss": 0.7164609432220459,
-      "eval_runtime": 0.3868,
-      "eval_samples_per_second": 25.854,
-      "eval_steps_per_second": 25.854,
+      "eval_loss": 0.7168648838996887,
+      "eval_runtime": 0.6094,
+      "eval_samples_per_second": 16.41,
+      "eval_steps_per_second": 16.41,
       "step": 450
-    },
-    {
-      "epoch": 20.0,
-      "grad_norm": 2.1205103397369385,
-      "learning_rate": 2.9996147467351856e-05,
-      "loss": 0.5147,
-      "step": 460
-    },
-    {
-      "epoch": 20.434782608695652,
-      "grad_norm": 1.27629554271698,
-      "learning_rate": 2.9995930474939773e-05,
-      "loss": 0.4782,
-      "step": 470
-    },
-    {
-      "epoch": 20.869565217391305,
-      "grad_norm": 1.4705679416656494,
-      "learning_rate": 2.9995707538619975e-05,
-      "loss": 0.5705,
-      "step": 480
-    },
-    {
-      "epoch": 21.304347826086957,
-      "grad_norm": 1.3248465061187744,
-      "learning_rate": 2.9995478658480822e-05,
-      "loss": 0.5161,
-      "step": 490
-    },
-    {
-      "epoch": 21.73913043478261,
-      "grad_norm": 1.2424507141113281,
-      "learning_rate": 2.9995243834613043e-05,
-      "loss": 0.5204,
-      "step": 500
-    },
-    {
-      "epoch": 22.17391304347826,
-      "grad_norm": 1.783718228340149,
-      "learning_rate": 2.9995003067109707e-05,
-      "loss": 0.4837,
-      "step": 510
-    },
-    {
-      "epoch": 22.608695652173914,
-      "grad_norm": 1.539941430091858,
-      "learning_rate": 2.9994756356066246e-05,
-      "loss": 0.5617,
-      "step": 520
-    },
-    {
-      "epoch": 23.043478260869566,
-      "grad_norm": 1.7417247295379639,
-      "learning_rate": 2.999450370158046e-05,
-      "loss": 0.4929,
-      "step": 530
-    },
-    {
-      "epoch": 23.47826086956522,
-      "grad_norm": 1.3086825609207153,
-      "learning_rate": 2.9994245103752478e-05,
-      "loss": 0.4383,
-      "step": 540
-    },
-    {
-      "epoch": 23.91304347826087,
-      "grad_norm": 1.2279788255691528,
-      "learning_rate": 2.999398056268481e-05,
-      "loss": 0.5266,
-      "step": 550
-    },
-    {
-      "epoch": 24.347826086956523,
-      "grad_norm": 1.40565025806427,
-      "learning_rate": 2.9993710078482306e-05,
-      "loss": 0.5207,
-      "step": 560
-    },
-    {
-      "epoch": 24.782608695652176,
-      "grad_norm": 0.9529893398284912,
-      "learning_rate": 2.9993433651252185e-05,
-      "loss": 0.4429,
-      "step": 570
-    },
-    {
-      "epoch": 25.217391304347824,
-      "grad_norm": 1.7137924432754517,
-      "learning_rate": 2.9993151281104006e-05,
-      "loss": 0.5325,
-      "step": 580
-    },
-    {
-      "epoch": 25.652173913043477,
-      "grad_norm": 1.1403567790985107,
-      "learning_rate": 2.9992862968149695e-05,
-      "loss": 0.4734,
-      "step": 590
-    },
-    {
-      "epoch": 26.08695652173913,
-      "grad_norm": 1.1702338457107544,
-      "learning_rate": 2.9992568712503533e-05,
-      "loss": 0.4609,
-      "step": 600
-    },
-    {
-      "epoch": 26.08695652173913,
-      "eval_loss": 0.7209551334381104,
-      "eval_runtime": 0.4345,
-      "eval_samples_per_second": 23.015,
-      "eval_steps_per_second": 23.015,
-      "step": 600
     }
   ],
   "logging_steps": 10,
@@ -478,7 +365,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.5363065839583232e+16,
+  "total_flos": 1.1591800814075904e+16,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0c84fd1341bca21f546fb34f0984c2737ed54e8dbbe2ce01666781675b951db1
+oid sha256:a4e26a07b48ab9f4eeb67e7d951dfa1f71b1fc6e6c955372cec1a3c50fa8b9a2
 size 5368