Upload folder using huggingface_hub

adapter_config.json

@@ -12,24 +12,24 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 128,
+  "lora_alpha": 256,
   "lora_bias": false,
-  "lora_dropout": 0,
+  "lora_dropout": 0.05,
   "megatron_config": null,
   "megatron_core": "megatron.core",
   "modules_to_save": null,
   "peft_type": "LORA",
-  "r": 64,
+  "r": 128,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
-    "gate_proj",
-    "k_proj",
-    "down_proj",
     "q_proj",
     "up_proj",
-    "o_proj"
+    "o_proj",
+    "v_proj",
+    "k_proj",
+    "gate_proj",
+    "down_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:310a72286497d50d5e6975572939b89cc2c30822edb5bf7327d465df0b83b07a
-size 479005064
+oid sha256:aa6c7de40a533ba2c4007a745595ab0f2ae9ee258e41c56770c1fd446b5c5664
+size 957942768

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b787d2ecf06ff6725d757284b065b1537c93c563f801a5cedf95dbd9a4aec45d
-size 958300235
+oid sha256:23717abd38a71cf0035162ad78af09ed8877c50c8910ff1c5069995befc406e7
+size 1916174411

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e5b517d1b8e2b0f837c8b00170b154961d4d989feba4326ac25583df7a55c57a
+oid sha256:a3d312928d0bb60518eb9856d5ab0ae1674bcb745294bf27f615cb6d07b0463e
 size 14645

trainer_state.json

@@ -11,702 +11,702 @@
   "log_history": [
     {
       "epoch": 0.05063291139240506,
-      "grad_norm": 29.55121612548828,
+      "grad_norm": 59.90552520751953,
       "learning_rate": 0.0,
       "loss": 3.0474,
       "step": 1
     },
     {
       "epoch": 0.10126582278481013,
-      "grad_norm": 47.9920539855957,
+      "grad_norm": 97.50157928466797,
       "learning_rate": 3.3333333333333333e-06,
       "loss": 3.2925,
       "step": 2
     },
     {
       "epoch": 0.1518987341772152,
-      "grad_norm": 21.540756225585938,
+      "grad_norm": 19.947616577148438,
       "learning_rate": 6.666666666666667e-06,
-      "loss": 3.201,
+      "loss": 3.0627,
       "step": 3
     },
     {
       "epoch": 0.20253164556962025,
-      "grad_norm": 8.517359733581543,
+      "grad_norm": 12.422538757324219,
       "learning_rate": 1e-05,
-      "loss": 2.6792,
+      "loss": 2.6282,
       "step": 4
     },
     {
       "epoch": 0.25316455696202533,
-      "grad_norm": 4.019173622131348,
+      "grad_norm": 7.736602783203125,
       "learning_rate": 1.3333333333333333e-05,
-      "loss": 2.2043,
+      "loss": 2.1838,
       "step": 5
     },
     {
       "epoch": 0.3037974683544304,
-      "grad_norm": 11.462793350219727,
+      "grad_norm": 11.91862964630127,
       "learning_rate": 1.6666666666666667e-05,
-      "loss": 2.8127,
+      "loss": 2.7703,
       "step": 6
     },
     {
       "epoch": 0.35443037974683544,
-      "grad_norm": 4.689273357391357,
+      "grad_norm": 9.939104080200195,
       "learning_rate": 2e-05,
-      "loss": 2.5712,
+      "loss": 2.5457,
       "step": 7
     },
     {
       "epoch": 0.4050632911392405,
-      "grad_norm": 4.675689697265625,
+      "grad_norm": 8.861398696899414,
       "learning_rate": 2.3333333333333336e-05,
-      "loss": 2.4222,
+      "loss": 2.3865,
       "step": 8
     },
     {
       "epoch": 0.45569620253164556,
-      "grad_norm": 4.813397407531738,
+      "grad_norm": 9.401021957397461,
       "learning_rate": 2.6666666666666667e-05,
-      "loss": 2.7645,
+      "loss": 2.7507,
       "step": 9
     },
     {
       "epoch": 0.5063291139240507,
-      "grad_norm": 4.817192077636719,
+      "grad_norm": 9.068034172058105,
       "learning_rate": 3e-05,
-      "loss": 2.5684,
+      "loss": 2.5836,
       "step": 10
     },
     {
       "epoch": 0.5569620253164557,
-      "grad_norm": 5.056623458862305,
+      "grad_norm": 9.445144653320312,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 2.7122,
+      "loss": 2.6978,
       "step": 11
     },
     {
       "epoch": 0.6075949367088608,
-      "grad_norm": 4.037632465362549,
+      "grad_norm": 67.43431854248047,
       "learning_rate": 3.6666666666666666e-05,
-      "loss": 2.7339,
+      "loss": 2.7532,
       "step": 12
     },
     {
       "epoch": 0.6582278481012658,
-      "grad_norm": 4.737704277038574,
+      "grad_norm": 10.308363914489746,
       "learning_rate": 4e-05,
-      "loss": 2.5438,
+      "loss": 2.5309,
       "step": 13
     },
     {
       "epoch": 0.7088607594936709,
-      "grad_norm": 4.448659896850586,
+      "grad_norm": 9.538652420043945,
       "learning_rate": 4.3333333333333334e-05,
-      "loss": 2.1418,
+      "loss": 2.1622,
       "step": 14
     },
     {
       "epoch": 0.759493670886076,
-      "grad_norm": 4.827819347381592,
+      "grad_norm": 10.145242691040039,
       "learning_rate": 4.666666666666667e-05,
-      "loss": 2.2931,
+      "loss": 2.3157,
       "step": 15
     },
     {
       "epoch": 0.810126582278481,
-      "grad_norm": 4.652594566345215,
+      "grad_norm": 12.203636169433594,
       "learning_rate": 5e-05,
-      "loss": 2.6527,
+      "loss": 2.7209,
       "step": 16
     },
     {
       "epoch": 0.8607594936708861,
-      "grad_norm": 3.7353029251098633,
+      "grad_norm": 8.771341323852539,
       "learning_rate": 4.998292650357558e-05,
-      "loss": 2.427,
+      "loss": 2.4761,
       "step": 17
     },
     {
       "epoch": 0.9113924050632911,
-      "grad_norm": 4.235123634338379,
+      "grad_norm": 9.946197509765625,
       "learning_rate": 4.993172933464471e-05,
-      "loss": 3.0115,
+      "loss": 2.9751,
       "step": 18
     },
     {
       "epoch": 0.9620253164556962,
-      "grad_norm": 4.2815046310424805,
+      "grad_norm": 7.850845813751221,
       "learning_rate": 4.984647842238185e-05,
-      "loss": 2.4574,
+      "loss": 2.4653,
       "step": 19
     },
     {
       "epoch": 1.0,
-      "grad_norm": 5.382309436798096,
+      "grad_norm": 10.167204856872559,
       "learning_rate": 4.972729020927865e-05,
-      "loss": 2.2266,
+      "loss": 2.3039,
       "step": 20
     },
     {
       "epoch": 1.0506329113924051,
-      "grad_norm": 3.046997308731079,
+      "grad_norm": 7.328862190246582,
       "learning_rate": 4.957432749209755e-05,
-      "loss": 2.1602,
+      "loss": 1.9503,
       "step": 21
     },
     {
       "epoch": 1.1012658227848102,
-      "grad_norm": 3.3037655353546143,
+      "grad_norm": 7.199270725250244,
       "learning_rate": 4.938779919951092e-05,
-      "loss": 2.0352,
+      "loss": 1.8009,
       "step": 22
     },
     {
       "epoch": 1.1518987341772151,
-      "grad_norm": 3.3994479179382324,
+      "grad_norm": 10.441041946411133,
       "learning_rate": 4.916796010672969e-05,
-      "loss": 2.0341,
+      "loss": 1.9009,
       "step": 23
     },
     {
       "epoch": 1.2025316455696202,
-      "grad_norm": 4.110241889953613,
+      "grad_norm": 9.012300491333008,
       "learning_rate": 4.891511048751102e-05,
-      "loss": 2.1729,
+      "loss": 1.9546,
       "step": 24
     },
     {
       "epoch": 1.2531645569620253,
-      "grad_norm": 3.1976218223571777,
+      "grad_norm": 8.55013656616211,
       "learning_rate": 4.862959570402049e-05,
-      "loss": 1.7632,
+      "loss": 1.5774,
       "step": 25
     },
     {
       "epoch": 1.3037974683544304,
-      "grad_norm": 4.051258563995361,
+      "grad_norm": 10.145879745483398,
       "learning_rate": 4.8311805735108894e-05,
-      "loss": 2.25,
+      "loss": 2.2121,
       "step": 26
     },
     {
       "epoch": 1.3544303797468356,
-      "grad_norm": 6.929507255554199,
+      "grad_norm": 13.984664916992188,
       "learning_rate": 4.796217464364808e-05,
-      "loss": 2.3207,
+      "loss": 2.0364,
       "step": 27
     },
     {
       "epoch": 1.4050632911392404,
-      "grad_norm": 4.213507652282715,
+      "grad_norm": 8.673234939575195,
       "learning_rate": 4.758117998365322e-05,
-      "loss": 2.1319,
+      "loss": 1.8328,
       "step": 28
     },
     {
       "epoch": 1.4556962025316456,
-      "grad_norm": 4.12835168838501,
+      "grad_norm": 7.145374298095703,
       "learning_rate": 4.716934214800155e-05,
-      "loss": 2.1621,
+      "loss": 1.8605,
       "step": 29
     },
     {
       "epoch": 1.5063291139240507,
-      "grad_norm": 4.359489917755127,
+      "grad_norm": 7.1981611251831055,
       "learning_rate": 4.672722365763821e-05,
-      "loss": 1.7892,
+      "loss": 1.539,
       "step": 30
     },
     {
       "epoch": 1.5569620253164556,
-      "grad_norm": 4.402414798736572,
+      "grad_norm": 7.796040058135986,
       "learning_rate": 4.625542839324036e-05,
-      "loss": 2.0661,
+      "loss": 1.6676,
       "step": 31
     },
     {
       "epoch": 1.6075949367088609,
-      "grad_norm": 4.111989498138428,
+      "grad_norm": 7.746204376220703,
       "learning_rate": 4.575460077038877e-05,
-      "loss": 2.3087,
+      "loss": 1.8508,
       "step": 32
     },
     {
       "epoch": 1.6582278481012658,
-      "grad_norm": 3.7056782245635986,
+      "grad_norm": 7.917386531829834,
       "learning_rate": 4.522542485937369e-05,
-      "loss": 1.7851,
+      "loss": 1.6083,
       "step": 33
     },
     {
       "epoch": 1.7088607594936709,
-      "grad_norm": 3.535465717315674,
+      "grad_norm": 7.896660804748535,
       "learning_rate": 4.4668623450837085e-05,
-      "loss": 1.9468,
+      "loss": 1.8305,
       "step": 34
     },
     {
       "epoch": 1.759493670886076,
-      "grad_norm": 4.073657989501953,
+      "grad_norm": 10.955219268798828,
       "learning_rate": 4.408495706852758e-05,
-      "loss": 2.2267,
+      "loss": 2.0006,
       "step": 35
     },
     {
       "epoch": 1.810126582278481,
-      "grad_norm": 6.199657917022705,
+      "grad_norm": 12.703990936279297,
       "learning_rate": 4.347522293051648e-05,
-      "loss": 2.0838,
+      "loss": 1.9769,
       "step": 36
     },
     {
       "epoch": 1.8607594936708862,
-      "grad_norm": 3.3513576984405518,
+      "grad_norm": 6.554587364196777,
       "learning_rate": 4.284025386029381e-05,
-      "loss": 1.7043,
+      "loss": 1.5205,
       "step": 37
     },
     {
       "epoch": 1.9113924050632911,
-      "grad_norm": 4.669745922088623,
+      "grad_norm": 9.378654479980469,
       "learning_rate": 4.218091714923157e-05,
-      "loss": 1.9935,
+      "loss": 2.112,
       "step": 38
     },
     {
       "epoch": 1.9620253164556962,
-      "grad_norm": 4.580756187438965,
+      "grad_norm": 8.477209091186523,
       "learning_rate": 4.149811337196807e-05,
-      "loss": 1.825,
+      "loss": 1.6478,
       "step": 39
     },
     {
       "epoch": 2.0,
-      "grad_norm": 4.372970104217529,
+      "grad_norm": 10.616765975952148,
       "learning_rate": 4.079277515633127e-05,
-      "loss": 2.303,
+      "loss": 2.3503,
       "step": 40
     },
     {
       "epoch": 2.050632911392405,
-      "grad_norm": 3.1565821170806885,
+      "grad_norm": 5.762049198150635,
       "learning_rate": 4.0065865909481417e-05,
-      "loss": 1.489,
+      "loss": 1.2976,
       "step": 41
     },
     {
       "epoch": 2.1012658227848102,
-      "grad_norm": 4.271334171295166,
+      "grad_norm": 7.1912736892700195,
       "learning_rate": 3.931837850201263e-05,
-      "loss": 1.8017,
+      "loss": 1.3291,
       "step": 42
     },
     {
       "epoch": 2.151898734177215,
-      "grad_norm": 3.5568835735321045,
+      "grad_norm": 7.300074577331543,
       "learning_rate": 3.855133391181124e-05,
-      "loss": 1.4802,
+      "loss": 1.0771,
       "step": 43
     },
     {
       "epoch": 2.2025316455696204,
-      "grad_norm": 4.271124362945557,
+      "grad_norm": 8.115863800048828,
       "learning_rate": 3.7765779829522675e-05,
-      "loss": 1.5782,
+      "loss": 1.1092,
       "step": 44
     },
     {
       "epoch": 2.2531645569620253,
-      "grad_norm": 6.451624870300293,
+      "grad_norm": 14.167332649230957,
       "learning_rate": 3.696278922753216e-05,
-      "loss": 1.7873,
+      "loss": 1.3089,
       "step": 45
     },
     {
       "epoch": 2.3037974683544302,
-      "grad_norm": 3.1512181758880615,
+      "grad_norm": 5.89024543762207,
       "learning_rate": 3.6143458894413465e-05,
-      "loss": 1.4847,
+      "loss": 1.2359,
       "step": 46
     },
     {
       "epoch": 2.3544303797468356,
-      "grad_norm": 5.191463470458984,
+      "grad_norm": 8.202404975891113,
       "learning_rate": 3.5308907936847594e-05,
-      "loss": 1.4307,
+      "loss": 1.0419,
       "step": 47
     },
     {
       "epoch": 2.4050632911392404,
-      "grad_norm": 3.847968816757202,
+      "grad_norm": 6.15331506729126,
       "learning_rate": 3.446027625105776e-05,
-      "loss": 1.558,
+      "loss": 1.2716,
       "step": 48
     },
     {
       "epoch": 2.4556962025316453,
-      "grad_norm": 4.588655948638916,
+      "grad_norm": 7.2445173263549805,
       "learning_rate": 3.3598722965848204e-05,
-      "loss": 1.55,
+      "loss": 1.2061,
       "step": 49
     },
     {
       "epoch": 2.5063291139240507,
-      "grad_norm": 4.686515808105469,
+      "grad_norm": 8.201730728149414,
       "learning_rate": 3.272542485937369e-05,
-      "loss": 1.3516,
+      "loss": 0.9834,
       "step": 50
     },
     {
       "epoch": 2.5569620253164556,
-      "grad_norm": 3.790144205093384,
+      "grad_norm": 6.665807723999023,
       "learning_rate": 3.1841574751802076e-05,
-      "loss": 1.3935,
+      "loss": 1.0482,
       "step": 51
     },
     {
       "epoch": 2.607594936708861,
-      "grad_norm": 5.291225910186768,
+      "grad_norm": 9.461847305297852,
       "learning_rate": 3.094837987606547e-05,
-      "loss": 1.3247,
+      "loss": 0.9212,
       "step": 52
     },
     {
       "epoch": 2.6582278481012658,
-      "grad_norm": 3.8952436447143555,
+      "grad_norm": 6.989381790161133,
       "learning_rate": 3.0047060228925256e-05,
-      "loss": 1.7195,
+      "loss": 1.158,
       "step": 53
     },
     {
       "epoch": 2.708860759493671,
-      "grad_norm": 5.9185261726379395,
+      "grad_norm": 10.334746360778809,
       "learning_rate": 2.913884690460325e-05,
-      "loss": 1.9431,
+      "loss": 1.5235,
       "step": 54
     },
     {
       "epoch": 2.759493670886076,
-      "grad_norm": 3.7539308071136475,
+      "grad_norm": 6.639229774475098,
       "learning_rate": 2.8224980413255086e-05,
-      "loss": 1.4235,
+      "loss": 1.0108,
       "step": 55
     },
     {
       "epoch": 2.810126582278481,
-      "grad_norm": 3.4502947330474854,
+      "grad_norm": 6.2803263664245605,
       "learning_rate": 2.7306708986582553e-05,
-      "loss": 1.2878,
+      "loss": 1.0466,
       "step": 56
     },
     {
       "epoch": 2.8607594936708862,
-      "grad_norm": 5.332718849182129,
+      "grad_norm": 8.546771049499512,
       "learning_rate": 2.638528687289925e-05,
-      "loss": 1.5786,
+      "loss": 1.0012,
       "step": 57
     },
     {
       "epoch": 2.911392405063291,
-      "grad_norm": 4.221064567565918,
+      "grad_norm": 7.191826820373535,
       "learning_rate": 2.5461972623978247e-05,
-      "loss": 1.3699,
+      "loss": 1.0746,
       "step": 58
     },
     {
       "epoch": 2.962025316455696,
-      "grad_norm": 5.887556076049805,
+      "grad_norm": 11.312370300292969,
       "learning_rate": 2.453802737602176e-05,
-      "loss": 1.8087,
+      "loss": 1.3323,
       "step": 59
     },
     {
       "epoch": 3.0,
-      "grad_norm": 3.3788065910339355,
+      "grad_norm": 4.749184608459473,
       "learning_rate": 2.361471312710075e-05,
-      "loss": 0.8745,
+      "loss": 0.5565,
       "step": 60
     },
     {
       "epoch": 3.050632911392405,
-      "grad_norm": 3.735031843185425,
+      "grad_norm": 5.846968650817871,
       "learning_rate": 2.2693291013417453e-05,
-      "loss": 1.2198,
+      "loss": 0.7122,
       "step": 61
     },
     {
       "epoch": 3.1012658227848102,
-      "grad_norm": 4.278702259063721,
+      "grad_norm": 5.533771514892578,
       "learning_rate": 2.1775019586744923e-05,
-      "loss": 1.1114,
+      "loss": 0.728,
       "step": 62
     },
     {
       "epoch": 3.151898734177215,
-      "grad_norm": 3.5521931648254395,
+      "grad_norm": 5.222585678100586,
       "learning_rate": 2.0861153095396748e-05,
-      "loss": 1.036,
+      "loss": 0.6109,
       "step": 63
     },
     {
       "epoch": 3.2025316455696204,
-      "grad_norm": 2.7933812141418457,
+      "grad_norm": 3.953166961669922,
       "learning_rate": 1.995293977107475e-05,
-      "loss": 0.7252,
+      "loss": 0.4257,
       "step": 64
     },
     {
       "epoch": 3.2531645569620253,
-      "grad_norm": 6.411159038543701,
+      "grad_norm": 5.500568866729736,
       "learning_rate": 1.9051620123934537e-05,
-      "loss": 0.9923,
+      "loss": 0.5261,
       "step": 65
     },
     {
       "epoch": 3.3037974683544302,
-      "grad_norm": 4.2825846672058105,
+      "grad_norm": 5.6277174949646,
       "learning_rate": 1.815842524819793e-05,
-      "loss": 1.1409,
+      "loss": 0.6001,
       "step": 66
     },
     {
       "epoch": 3.3544303797468356,
-      "grad_norm": 5.785189151763916,
+      "grad_norm": 8.257991790771484,
       "learning_rate": 1.7274575140626318e-05,
-      "loss": 1.1165,
+      "loss": 0.6239,
       "step": 67
     },
     {
       "epoch": 3.4050632911392404,
-      "grad_norm": 4.309160232543945,
+      "grad_norm": 6.649881839752197,
       "learning_rate": 1.6401277034151798e-05,
-      "loss": 1.2768,
+      "loss": 0.8054,
       "step": 68
     },
     {
       "epoch": 3.4556962025316453,
-      "grad_norm": 5.033167362213135,
+      "grad_norm": 7.573472499847412,
       "learning_rate": 1.5539723748942245e-05,
-      "loss": 1.2941,
+      "loss": 0.6508,
       "step": 69
     },
     {
       "epoch": 3.5063291139240507,
-      "grad_norm": 4.315425395965576,
+      "grad_norm": 6.748483180999756,
       "learning_rate": 1.4691092063152417e-05,
-      "loss": 0.9229,
+      "loss": 0.4997,
       "step": 70
     },
     {
       "epoch": 3.5569620253164556,
-      "grad_norm": 4.174651622772217,
+      "grad_norm": 6.085212707519531,
       "learning_rate": 1.3856541105586545e-05,
-      "loss": 1.0337,
+      "loss": 0.5069,
       "step": 71
     },
     {
       "epoch": 3.607594936708861,
-      "grad_norm": 4.93477201461792,
+      "grad_norm": 8.33370304107666,
       "learning_rate": 1.303721077246784e-05,
-      "loss": 0.9844,
+      "loss": 0.675,
       "step": 72
     },
     {
       "epoch": 3.6582278481012658,
-      "grad_norm": 5.448512554168701,
+      "grad_norm": 9.128050804138184,
       "learning_rate": 1.223422017047733e-05,
-      "loss": 1.0804,
+      "loss": 0.6149,
       "step": 73
     },
     {
       "epoch": 3.708860759493671,
-      "grad_norm": 4.748834609985352,
+      "grad_norm": 6.868535041809082,
       "learning_rate": 1.1448666088188764e-05,
-      "loss": 1.0972,
+      "loss": 0.5729,
       "step": 74
     },
     {
       "epoch": 3.759493670886076,
-      "grad_norm": 5.283952713012695,
+      "grad_norm": 10.637543678283691,
       "learning_rate": 1.068162149798737e-05,
-      "loss": 1.0916,
+      "loss": 0.6948,
       "step": 75
     },
     {
       "epoch": 3.810126582278481,
-      "grad_norm": 4.5018630027771,
+      "grad_norm": 5.958944320678711,
       "learning_rate": 9.934134090518593e-06,
-      "loss": 0.9075,
+      "loss": 0.4837,
       "step": 76
     },
     {
       "epoch": 3.8607594936708862,
-      "grad_norm": 5.909902572631836,
+      "grad_norm": 8.521463394165039,
       "learning_rate": 9.207224843668732e-06,
-      "loss": 1.314,
+      "loss": 0.7003,
       "step": 77
     },
     {
       "epoch": 3.911392405063291,
-      "grad_norm": 5.771544933319092,
+      "grad_norm": 7.978651523590088,
       "learning_rate": 8.50188662803194e-06,
-      "loss": 1.2385,
+      "loss": 0.5588,
       "step": 78
     },
     {
       "epoch": 3.962025316455696,
-      "grad_norm": 5.263690948486328,
+      "grad_norm": 6.872608184814453,
       "learning_rate": 7.819082850768434e-06,
-      "loss": 1.05,
+      "loss": 0.5483,
       "step": 79
     },
     {
       "epoch": 4.0,
-      "grad_norm": 4.816619873046875,
+      "grad_norm": 6.058796405792236,
       "learning_rate": 7.159746139706194e-06,
-      "loss": 1.0004,
+      "loss": 0.4585,
       "step": 80
     },
     {
       "epoch": 4.050632911392405,
-      "grad_norm": 4.460178375244141,
+      "grad_norm": 5.086803913116455,
       "learning_rate": 6.524777069483526e-06,
-      "loss": 0.8953,
+      "loss": 0.3461,
       "step": 81
     },
     {
       "epoch": 4.10126582278481,
-      "grad_norm": 4.612957954406738,
+      "grad_norm": 4.965209484100342,
       "learning_rate": 5.915042931472425e-06,
-      "loss": 1.0056,
+      "loss": 0.4123,
       "step": 82
     },
     {
       "epoch": 4.151898734177215,
-      "grad_norm": 5.433095932006836,
+      "grad_norm": 4.652761459350586,
       "learning_rate": 5.33137654916292e-06,
-      "loss": 0.7673,
+      "loss": 0.3055,
       "step": 83
     },
     {
       "epoch": 4.2025316455696204,
-      "grad_norm": 3.291714906692505,
+      "grad_norm": 3.384598731994629,
       "learning_rate": 4.7745751406263165e-06,
-      "loss": 0.6697,
+      "loss": 0.2181,
       "step": 84
     },
     {
       "epoch": 4.253164556962025,
-      "grad_norm": 3.8275110721588135,
+      "grad_norm": 4.2130818367004395,
       "learning_rate": 4.245399229611238e-06,
-      "loss": 0.942,
+      "loss": 0.3542,
       "step": 85
     },
     {
       "epoch": 4.30379746835443,
-      "grad_norm": 3.4578733444213867,
+      "grad_norm": 4.708261489868164,
       "learning_rate": 3.7445716067596503e-06,
-      "loss": 0.9534,
+      "loss": 0.4113,
       "step": 86
     },
     {
       "epoch": 4.3544303797468356,
-      "grad_norm": 4.556038856506348,
+      "grad_norm": 3.8586061000823975,
       "learning_rate": 3.2727763423617913e-06,
-      "loss": 0.6355,
+      "loss": 0.1785,
       "step": 87
     },
     {
       "epoch": 4.405063291139241,
-      "grad_norm": 3.973484754562378,
+      "grad_norm": 4.425046920776367,
       "learning_rate": 2.8306578519984527e-06,
-      "loss": 0.8523,
+      "loss": 0.3638,
       "step": 88
     },
     {
       "epoch": 4.455696202531645,
-      "grad_norm": 5.197175979614258,
+      "grad_norm": 4.765505790710449,
       "learning_rate": 2.418820016346779e-06,
-      "loss": 0.7797,
+      "loss": 0.2206,
       "step": 89
     },
     {
       "epoch": 4.506329113924051,
-      "grad_norm": 4.299957752227783,
+      "grad_norm": 4.627501964569092,
       "learning_rate": 2.0378253563519247e-06,
-      "loss": 0.9653,
+      "loss": 0.3092,
       "step": 90
     },
     {
       "epoch": 4.556962025316456,
-      "grad_norm": 4.691103458404541,
+      "grad_norm": 5.241892337799072,
       "learning_rate": 1.6881942648911076e-06,
-      "loss": 0.7476,
+      "loss": 0.3072,
       "step": 91
     },
     {
       "epoch": 4.6075949367088604,
-      "grad_norm": 3.614025354385376,
+      "grad_norm": 4.68987512588501,
       "learning_rate": 1.3704042959795132e-06,
-      "loss": 0.9732,
+      "loss": 0.4245,
       "step": 92
     },
     {
       "epoch": 4.658227848101266,
-      "grad_norm": 3.1150994300842285,
+      "grad_norm": 3.880495071411133,
       "learning_rate": 1.0848895124889818e-06,
-      "loss": 0.8307,
+      "loss": 0.3624,
       "step": 93
     },
     {
       "epoch": 4.708860759493671,
-      "grad_norm": 4.92384147644043,
+      "grad_norm": 5.35604190826416,
       "learning_rate": 8.320398932703144e-07,
-      "loss": 1.0243,
+      "loss": 0.3626,
       "step": 94
     },
     {
       "epoch": 4.759493670886076,
-      "grad_norm": 4.248635292053223,
+      "grad_norm": 4.486222267150879,
       "learning_rate": 6.122008004890851e-07,
-      "loss": 0.835,
+      "loss": 0.2896,
       "step": 95
     },
     {
       "epoch": 4.810126582278481,
-      "grad_norm": 4.580704212188721,
+      "grad_norm": 4.400012016296387,
       "learning_rate": 4.256725079024554e-07,
-      "loss": 0.6964,
+      "loss": 0.2279,
       "step": 96
     },
     {
       "epoch": 4.860759493670886,
-      "grad_norm": 4.322700023651123,
+      "grad_norm": 5.147950649261475,
       "learning_rate": 2.7270979072135104e-07,
-      "loss": 0.9184,
+      "loss": 0.3981,
       "step": 97
     },
     {
       "epoch": 4.911392405063291,
-      "grad_norm": 5.485646724700928,
+      "grad_norm": 5.160162448883057,
       "learning_rate": 1.5352157761815977e-07,
-      "loss": 0.9517,
+      "loss": 0.2999,
       "step": 98
     },
     {
       "epoch": 4.962025316455696,
-      "grad_norm": 3.603494167327881,
+      "grad_norm": 3.9566633701324463,
       "learning_rate": 6.827066535529946e-08,
-      "loss": 0.7875,
+      "loss": 0.2816,
       "step": 99
     },
     {
       "epoch": 5.0,
-      "grad_norm": 5.922054290771484,
+      "grad_norm": 6.496578693389893,
       "learning_rate": 1.7073496424427348e-08,
-      "loss": 0.5772,
+      "loss": 0.1943,
       "step": 100
     }
   ],
@@ -714,7 +714,7 @@
   "max_steps": 100,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 5,
-  "save_steps": 10,
+  "save_steps": 20,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {
@@ -727,7 +727,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.957146343615693e+16,
+  "total_flos": 3.0794717131554816e+16,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5d2bf5cbde8ca6f3d82294680caf16a11f0007e85940b9ae50a642637f810a67
+oid sha256:cec728dbde53c01fb89ffdc20ac760b3b4b343536639aa83df1e8f345df47df8
 size 6033