Training in progress, step 48, checkpoint

last-checkpoint/adapter_config.json

@@ -20,13 +20,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
+    "gate_proj",
+    "o_proj",
     "up_proj",
-    "down_proj",
-    "k_proj",
     "q_proj",
-    "o_proj",
-    "gate_proj"
+    "down_proj",
+    "v_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:bc3c5cf1f24acee131e5f54fcfba8706581af96f13f99cae67263bfe2643ce69
+oid sha256:5a32d389a9b3f151214705183b0c282d77e669cd08af614fc7aba8ca778c1316
 size 167832240

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6f4dca042f985414dcf6934d2473634a5fbc338c5b87609a091eb0a64316f60e
-size 335922386
+oid sha256:427776be467d9f16b7f0cdef32eea028e04f09d77dec3df57cbcdeddcf642247
+size 85723284

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ed20cb1e030baf9351dd9705735eda93875b2582f4380baf52cb1cb6e628a469
+oid sha256:e61ad4b53f154b31a93a0157080b7bfe2b2146a56957bcea44dc77c0516cf565
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0aa0f75a1f3e346be25756b578158b09a68943f0b9f1cfe29f97939687f864ef
+oid sha256:6a946ba862c9290cf73cbdcd016309ecac1eb5b3695453bc8a42a8a259a1c695
 size 1064

last-checkpoint/trainer_state.json

@@ -1,344 +1,452 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.015197889182058048,
+  "epoch": 0.040523427606585054,
   "eval_steps": 4,
-  "global_step": 36,
+  "global_step": 48,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.00042216358839050134,
-      "grad_norm": 6.515402317047119,
+      "epoch": 0.0008442380751371887,
+      "grad_norm": NaN,
       "learning_rate": 2e-05,
-      "loss": 6.2351,
+      "loss": 0.0,
       "step": 1
     },
     {
-      "epoch": 0.00042216358839050134,
-      "eval_loss": 7.163845539093018,
-      "eval_runtime": 59.6773,
-      "eval_samples_per_second": 8.362,
-      "eval_steps_per_second": 8.362,
+      "epoch": 0.0008442380751371887,
+      "eval_loss": NaN,
+      "eval_runtime": 40.4739,
+      "eval_samples_per_second": 12.329,
+      "eval_steps_per_second": 6.177,
       "step": 1
     },
     {
-      "epoch": 0.0008443271767810027,
-      "grad_norm": 6.853526592254639,
+      "epoch": 0.0016884761502743773,
+      "grad_norm": NaN,
       "learning_rate": 4e-05,
-      "loss": 6.2997,
+      "loss": 0.0,
       "step": 2
     },
     {
-      "epoch": 0.001266490765171504,
-      "grad_norm": 7.778280735015869,
+      "epoch": 0.002532714225411566,
+      "grad_norm": NaN,
       "learning_rate": 6e-05,
-      "loss": 7.021,
+      "loss": 0.0,
       "step": 3
     },
     {
-      "epoch": 0.0016886543535620053,
-      "grad_norm": 8.638328552246094,
+      "epoch": 0.0033769523005487546,
+      "grad_norm": NaN,
       "learning_rate": 8e-05,
-      "loss": 7.51,
+      "loss": 0.0,
       "step": 4
     },
     {
-      "epoch": 0.0016886543535620053,
-      "eval_loss": 6.921504974365234,
-      "eval_runtime": 59.6612,
-      "eval_samples_per_second": 8.364,
-      "eval_steps_per_second": 8.364,
+      "epoch": 0.0033769523005487546,
+      "eval_loss": NaN,
+      "eval_runtime": 39.7116,
+      "eval_samples_per_second": 12.566,
+      "eval_steps_per_second": 6.295,
       "step": 4
     },
     {
-      "epoch": 0.0021108179419525065,
-      "grad_norm": 8.384474754333496,
+      "epoch": 0.004221190375685944,
+      "grad_norm": NaN,
       "learning_rate": 0.0001,
-      "loss": 6.6173,
+      "loss": 0.0,
       "step": 5
     },
     {
-      "epoch": 0.002532981530343008,
-      "grad_norm": 6.948892116546631,
+      "epoch": 0.005065428450823132,
+      "grad_norm": NaN,
       "learning_rate": 0.00012,
-      "loss": 5.3785,
+      "loss": 0.0,
       "step": 6
     },
     {
-      "epoch": 0.0029551451187335093,
-      "grad_norm": 7.122462749481201,
+      "epoch": 0.0059096665259603205,
+      "grad_norm": NaN,
       "learning_rate": 0.00014,
-      "loss": 5.7491,
+      "loss": 0.0,
       "step": 7
     },
     {
-      "epoch": 0.0033773087071240107,
-      "grad_norm": 8.084318161010742,
+      "epoch": 0.006753904601097509,
+      "grad_norm": NaN,
       "learning_rate": 0.00016,
-      "loss": 4.7822,
+      "loss": 0.0,
       "step": 8
     },
     {
-      "epoch": 0.0033773087071240107,
-      "eval_loss": 4.463127613067627,
-      "eval_runtime": 59.3591,
-      "eval_samples_per_second": 8.406,
-      "eval_steps_per_second": 8.406,
+      "epoch": 0.006753904601097509,
+      "eval_loss": NaN,
+      "eval_runtime": 39.8847,
+      "eval_samples_per_second": 12.511,
+      "eval_steps_per_second": 6.268,
       "step": 8
     },
     {
-      "epoch": 0.003799472295514512,
-      "grad_norm": 7.751031875610352,
+      "epoch": 0.007598142676234698,
+      "grad_norm": NaN,
       "learning_rate": 0.00018,
-      "loss": 4.0501,
+      "loss": 0.0,
       "step": 9
     },
     {
-      "epoch": 0.004221635883905013,
-      "grad_norm": 9.506431579589844,
+      "epoch": 0.008442380751371888,
+      "grad_norm": NaN,
       "learning_rate": 0.0002,
-      "loss": 4.0302,
+      "loss": 0.0,
       "step": 10
     },
     {
-      "epoch": 0.004643799472295515,
-      "grad_norm": 7.7947001457214355,
+      "epoch": 0.009286618826509076,
+      "grad_norm": NaN,
       "learning_rate": 0.0001996917333733128,
-      "loss": 2.6523,
+      "loss": 0.0,
       "step": 11
     },
     {
-      "epoch": 0.005065963060686016,
-      "grad_norm": 7.803070068359375,
+      "epoch": 0.010130856901646263,
+      "grad_norm": NaN,
       "learning_rate": 0.00019876883405951377,
-      "loss": 1.5597,
+      "loss": 0.0,
       "step": 12
     },
     {
-      "epoch": 0.005065963060686016,
-      "eval_loss": 1.5647698640823364,
-      "eval_runtime": 59.7333,
-      "eval_samples_per_second": 8.354,
-      "eval_steps_per_second": 8.354,
+      "epoch": 0.010130856901646263,
+      "eval_loss": NaN,
+      "eval_runtime": 39.9436,
+      "eval_samples_per_second": 12.493,
+      "eval_steps_per_second": 6.259,
       "step": 12
     },
     {
-      "epoch": 0.005488126649076517,
-      "grad_norm": 7.336414813995361,
+      "epoch": 0.010975094976783452,
+      "grad_norm": NaN,
       "learning_rate": 0.00019723699203976766,
-      "loss": 1.2889,
+      "loss": 0.0,
       "step": 13
     },
     {
-      "epoch": 0.005910290237467019,
-      "grad_norm": 7.312011241912842,
+      "epoch": 0.011819333051920641,
+      "grad_norm": NaN,
       "learning_rate": 0.00019510565162951537,
-      "loss": 1.2851,
+      "loss": 0.0,
       "step": 14
     },
     {
-      "epoch": 0.0063324538258575196,
-      "grad_norm": 11.156982421875,
+      "epoch": 0.01266357112705783,
+      "grad_norm": NaN,
       "learning_rate": 0.0001923879532511287,
-      "loss": 0.9687,
+      "loss": 0.0,
       "step": 15
     },
     {
-      "epoch": 0.006754617414248021,
-      "grad_norm": 8.215885162353516,
+      "epoch": 0.013507809202195019,
+      "grad_norm": NaN,
       "learning_rate": 0.0001891006524188368,
-      "loss": 0.481,
+      "loss": 0.0,
       "step": 16
     },
     {
-      "epoch": 0.006754617414248021,
-      "eval_loss": 0.6843006014823914,
-      "eval_runtime": 59.5333,
-      "eval_samples_per_second": 8.382,
-      "eval_steps_per_second": 8.382,
+      "epoch": 0.013507809202195019,
+      "eval_loss": NaN,
+      "eval_runtime": 39.9577,
+      "eval_samples_per_second": 12.488,
+      "eval_steps_per_second": 6.257,
       "step": 16
     },
     {
-      "epoch": 0.007176781002638522,
-      "grad_norm": 4.760833263397217,
+      "epoch": 0.014352047277332207,
+      "grad_norm": NaN,
       "learning_rate": 0.00018526401643540922,
-      "loss": 0.4751,
+      "loss": 0.0,
       "step": 17
     },
     {
-      "epoch": 0.007598944591029024,
-      "grad_norm": 7.501655578613281,
+      "epoch": 0.015196285352469396,
+      "grad_norm": NaN,
       "learning_rate": 0.00018090169943749476,
-      "loss": 0.6609,
+      "loss": 0.0,
       "step": 18
     },
     {
-      "epoch": 0.008021108179419525,
-      "grad_norm": 6.412684917449951,
+      "epoch": 0.016040523427606587,
+      "grad_norm": NaN,
       "learning_rate": 0.0001760405965600031,
-      "loss": 0.6547,
+      "loss": 0.0,
       "step": 19
     },
     {
-      "epoch": 0.008443271767810026,
-      "grad_norm": 5.393802642822266,
+      "epoch": 0.016884761502743775,
+      "grad_norm": NaN,
       "learning_rate": 0.00017071067811865476,
-      "loss": 0.3143,
+      "loss": 0.0,
       "step": 20
     },
     {
-      "epoch": 0.008443271767810026,
-      "eval_loss": 0.530745804309845,
-      "eval_runtime": 59.4057,
-      "eval_samples_per_second": 8.4,
-      "eval_steps_per_second": 8.4,
+      "epoch": 0.016884761502743775,
+      "eval_loss": NaN,
+      "eval_runtime": 39.9571,
+      "eval_samples_per_second": 12.488,
+      "eval_steps_per_second": 6.257,
       "step": 20
     },
     {
-      "epoch": 0.008865435356200527,
-      "grad_norm": 6.753420352935791,
+      "epoch": 0.017728999577880964,
+      "grad_norm": NaN,
       "learning_rate": 0.00016494480483301836,
-      "loss": 0.7354,
+      "loss": 0.0,
       "step": 21
     },
     {
-      "epoch": 0.00928759894459103,
-      "grad_norm": 3.744786500930786,
+      "epoch": 0.018573237653018153,
+      "grad_norm": NaN,
       "learning_rate": 0.00015877852522924732,
-      "loss": 0.3945,
+      "loss": 0.0,
       "step": 22
     },
     {
-      "epoch": 0.00970976253298153,
-      "grad_norm": 3.2725045680999756,
+      "epoch": 0.019417475728155338,
+      "grad_norm": NaN,
       "learning_rate": 0.0001522498564715949,
-      "loss": 0.2881,
+      "loss": 0.0,
       "step": 23
     },
     {
-      "epoch": 0.010131926121372032,
-      "grad_norm": 2.0320169925689697,
+      "epoch": 0.020261713803292527,
+      "grad_norm": NaN,
       "learning_rate": 0.00014539904997395468,
-      "loss": 0.1861,
+      "loss": 0.0,
       "step": 24
     },
     {
-      "epoch": 0.010131926121372032,
-      "eval_loss": 0.4393819272518158,
-      "eval_runtime": 59.4549,
-      "eval_samples_per_second": 8.393,
-      "eval_steps_per_second": 8.393,
+      "epoch": 0.020261713803292527,
+      "eval_loss": NaN,
+      "eval_runtime": 39.9579,
+      "eval_samples_per_second": 12.488,
+      "eval_steps_per_second": 6.257,
       "step": 24
     },
     {
-      "epoch": 0.010554089709762533,
-      "grad_norm": 5.322549343109131,
+      "epoch": 0.021105951878429716,
+      "grad_norm": NaN,
       "learning_rate": 0.000138268343236509,
-      "loss": 0.7912,
+      "loss": 0.0,
       "step": 25
     },
     {
-      "epoch": 0.010976253298153034,
-      "grad_norm": 1.55814790725708,
+      "epoch": 0.021950189953566904,
+      "grad_norm": NaN,
       "learning_rate": 0.00013090169943749476,
-      "loss": 0.2046,
+      "loss": 0.0,
       "step": 26
     },
     {
-      "epoch": 0.011398416886543536,
-      "grad_norm": 2.2971291542053223,
+      "epoch": 0.022794428028704093,
+      "grad_norm": NaN,
       "learning_rate": 0.00012334453638559057,
-      "loss": 0.2717,
+      "loss": 0.0,
       "step": 27
     },
     {
-      "epoch": 0.011820580474934037,
-      "grad_norm": 1.6769708395004272,
+      "epoch": 0.023638666103841282,
+      "grad_norm": NaN,
       "learning_rate": 0.0001156434465040231,
-      "loss": 0.1889,
+      "loss": 0.0,
       "step": 28
     },
     {
-      "epoch": 0.011820580474934037,
-      "eval_loss": 0.43992501497268677,
-      "eval_runtime": 59.4329,
-      "eval_samples_per_second": 8.396,
-      "eval_steps_per_second": 8.396,
+      "epoch": 0.023638666103841282,
+      "eval_loss": NaN,
+      "eval_runtime": 39.9492,
+      "eval_samples_per_second": 12.491,
+      "eval_steps_per_second": 6.258,
       "step": 28
     },
     {
-      "epoch": 0.012242744063324538,
-      "grad_norm": 4.204702854156494,
+      "epoch": 0.02448290417897847,
+      "grad_norm": NaN,
       "learning_rate": 0.0001078459095727845,
-      "loss": 0.4801,
+      "loss": 0.0,
       "step": 29
     },
     {
-      "epoch": 0.012664907651715039,
-      "grad_norm": 3.246248483657837,
+      "epoch": 0.02532714225411566,
+      "grad_norm": NaN,
       "learning_rate": 0.0001,
-      "loss": 0.425,
+      "loss": 0.0,
       "step": 30
     },
     {
-      "epoch": 0.01308707124010554,
-      "grad_norm": 2.989017963409424,
+      "epoch": 0.026171380329252848,
+      "grad_norm": NaN,
       "learning_rate": 9.215409042721552e-05,
-      "loss": 0.46,
+      "loss": 0.0,
       "step": 31
     },
     {
-      "epoch": 0.013509234828496043,
-      "grad_norm": 2.7211666107177734,
+      "epoch": 0.027015618404390037,
+      "grad_norm": NaN,
       "learning_rate": 8.435655349597689e-05,
-      "loss": 0.2487,
+      "loss": 0.0,
       "step": 32
     },
     {
-      "epoch": 0.013509234828496043,
-      "eval_loss": 0.3397166132926941,
-      "eval_runtime": 59.4567,
-      "eval_samples_per_second": 8.393,
-      "eval_steps_per_second": 8.393,
+      "epoch": 0.027015618404390037,
+      "eval_loss": NaN,
+      "eval_runtime": 39.9472,
+      "eval_samples_per_second": 12.491,
+      "eval_steps_per_second": 6.258,
       "step": 32
     },
     {
-      "epoch": 0.013931398416886544,
-      "grad_norm": 1.7053347826004028,
+      "epoch": 0.027859856479527226,
+      "grad_norm": NaN,
       "learning_rate": 7.66554636144095e-05,
-      "loss": 0.251,
+      "loss": 0.0,
       "step": 33
     },
     {
-      "epoch": 0.014353562005277045,
-      "grad_norm": 2.717104434967041,
+      "epoch": 0.028704094554664415,
+      "grad_norm": NaN,
       "learning_rate": 6.909830056250527e-05,
-      "loss": 0.5596,
+      "loss": 0.0,
       "step": 34
     },
     {
-      "epoch": 0.014775725593667546,
-      "grad_norm": 1.9112799167633057,
+      "epoch": 0.029548332629801603,
+      "grad_norm": NaN,
       "learning_rate": 6.173165676349103e-05,
-      "loss": 0.2708,
+      "loss": 0.0,
       "step": 35
     },
     {
-      "epoch": 0.015197889182058048,
-      "grad_norm": 2.3562815189361572,
+      "epoch": 0.030392570704938792,
+      "grad_norm": NaN,
       "learning_rate": 5.4600950026045326e-05,
-      "loss": 0.4158,
+      "loss": 0.0,
       "step": 36
     },
     {
-      "epoch": 0.015197889182058048,
-      "eval_loss": 0.31511208415031433,
-      "eval_runtime": 59.4134,
-      "eval_samples_per_second": 8.399,
-      "eval_steps_per_second": 8.399,
+      "epoch": 0.030392570704938792,
+      "eval_loss": NaN,
+      "eval_runtime": 39.9504,
+      "eval_samples_per_second": 12.49,
+      "eval_steps_per_second": 6.258,
       "step": 36
+    },
+    {
+      "epoch": 0.03123680878007598,
+      "grad_norm": NaN,
+      "learning_rate": 4.7750143528405126e-05,
+      "loss": 0.0,
+      "step": 37
+    },
+    {
+      "epoch": 0.03208104685521317,
+      "grad_norm": NaN,
+      "learning_rate": 4.12214747707527e-05,
+      "loss": 0.0,
+      "step": 38
+    },
+    {
+      "epoch": 0.03292528493035036,
+      "grad_norm": NaN,
+      "learning_rate": 3.5055195166981645e-05,
+      "loss": 0.0,
+      "step": 39
+    },
+    {
+      "epoch": 0.03376952300548755,
+      "grad_norm": NaN,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 0.0,
+      "step": 40
+    },
+    {
+      "epoch": 0.03376952300548755,
+      "eval_loss": NaN,
+      "eval_runtime": 39.9467,
+      "eval_samples_per_second": 12.492,
+      "eval_steps_per_second": 6.258,
+      "step": 40
+    },
+    {
+      "epoch": 0.034613761080624736,
+      "grad_norm": NaN,
+      "learning_rate": 2.3959403439996907e-05,
+      "loss": 0.0,
+      "step": 41
+    },
+    {
+      "epoch": 0.03545799915576193,
+      "grad_norm": NaN,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 0.0,
+      "step": 42
+    },
+    {
+      "epoch": 0.03630223723089911,
+      "grad_norm": NaN,
+      "learning_rate": 1.4735983564590783e-05,
+      "loss": 0.0,
+      "step": 43
+    },
+    {
+      "epoch": 0.037146475306036306,
+      "grad_norm": NaN,
+      "learning_rate": 1.0899347581163221e-05,
+      "loss": 0.0,
+      "step": 44
+    },
+    {
+      "epoch": 0.037146475306036306,
+      "eval_loss": NaN,
+      "eval_runtime": 39.9428,
+      "eval_samples_per_second": 12.493,
+      "eval_steps_per_second": 6.259,
+      "step": 44
+    },
+    {
+      "epoch": 0.03799071338117349,
+      "grad_norm": NaN,
+      "learning_rate": 7.612046748871327e-06,
+      "loss": 0.0,
+      "step": 45
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": NaN,
+      "learning_rate": 4.8943483704846475e-06,
+      "loss": 0.0,
+      "step": 46
+    },
+    {
+      "epoch": 0.03967918953144787,
+      "grad_norm": NaN,
+      "learning_rate": 2.7630079602323442e-06,
+      "loss": 0.0,
+      "step": 47
+    },
+    {
+      "epoch": 0.040523427606585054,
+      "grad_norm": NaN,
+      "learning_rate": 1.231165940486234e-06,
+      "loss": 0.0,
+      "step": 48
+    },
+    {
+      "epoch": 0.040523427606585054,
+      "eval_loss": NaN,
+      "eval_runtime": 39.9611,
+      "eval_samples_per_second": 12.487,
+      "eval_steps_per_second": 6.256,
+      "step": 48
     }
   ],
   "logging_steps": 1,
@@ -358,8 +466,8 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.3353971104088064e+16,
-  "train_batch_size": 1,
+  "total_flos": 1.7805294805450752e+16,
+  "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null
 }

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3a3833977106fac48f5d71407b1a01067110c5fbf699342151524802ba318de1
+oid sha256:dfe6c9b320916ba73b5ecd7d507cc1843cb46082183e02f95c9c663f599d80e6
 size 6776