Upload folder using huggingface_hub

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8c5d70341558362e290bdb4f628c977061e53ea53a5516ddce8eb02f24d65aac
+oid sha256:1e4efbb90f4f79dc3857f34d206b10d38909a695d5d4deb4a12bb5119d0bd82d
 size 83724291

pytorch_adapter.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0375c1a4e4a3e3714b279cfb0629135bc9443a4f404d30ae00f387de9231f4ea
+oid sha256:b47d6ea2074c7e496272b090acb212f7f47946a07072c18f6b05f89e7c179dda
 size 39503700

pytorch_model_head.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6e457b3065b98478c19c66e764d122ab98bf768f9ef4e1dcc1ed82e9b5eff0d3
+oid sha256:6d4f8012f6d55002b4bd445921dc8ab161f0ff05a248afc28051977910a38a82
 size 774447414

tokenizer_config.json

@@ -842,7 +842,7 @@
     }
   },
   "bos_token": "[CLS]",
-  "clean_up_tokenization_spaces": true,
+  "clean_up_tokenization_spaces": false,
   "cls_token": "[CLS]",
   "do_lower_case": false,
   "eos_token": "[SEP]",

trainer_state.json

@@ -10,653 +10,653 @@
   "log_history": [
     {
       "epoch": 0.032,
-      "grad_norm": 3.7940785884857178,
+      "grad_norm": 2.9300780296325684,
       "learning_rate": 4.9466666666666665e-05,
-      "loss": 8.3828,
+      "loss": 8.4304,
       "step": 500
     },
     {
       "epoch": 0.064,
-      "grad_norm": 3.092071294784546,
+      "grad_norm": 2.708092212677002,
       "learning_rate": 4.8933333333333335e-05,
-      "loss": 6.3779,
+      "loss": 6.3815,
       "step": 1000
     },
     {
       "epoch": 0.096,
-      "grad_norm": 2.6804521083831787,
+      "grad_norm": 2.7160232067108154,
       "learning_rate": 4.8400000000000004e-05,
-      "loss": 5.7375,
+      "loss": 5.717,
       "step": 1500
     },
     {
       "epoch": 0.128,
-      "grad_norm": 2.5856428146362305,
+      "grad_norm": 2.5131046772003174,
       "learning_rate": 4.7866666666666674e-05,
-      "loss": 5.377,
+      "loss": 5.353,
       "step": 2000
     },
     {
       "epoch": 0.16,
-      "grad_norm": 2.9187934398651123,
+      "grad_norm": 2.8203608989715576,
       "learning_rate": 4.7333333333333336e-05,
-      "loss": 5.1249,
+      "loss": 5.0937,
       "step": 2500
     },
     {
       "epoch": 0.192,
-      "grad_norm": 2.5977468490600586,
+      "grad_norm": 2.6736185550689697,
       "learning_rate": 4.6800000000000006e-05,
-      "loss": 4.9347,
+      "loss": 4.905,
       "step": 3000
     },
     {
       "epoch": 0.224,
-      "grad_norm": 2.8521013259887695,
+      "grad_norm": 2.9173152446746826,
       "learning_rate": 4.626666666666667e-05,
-      "loss": 4.7731,
+      "loss": 4.7377,
       "step": 3500
     },
     {
       "epoch": 0.256,
-      "grad_norm": 3.0819144248962402,
+      "grad_norm": 2.861001968383789,
       "learning_rate": 4.573333333333333e-05,
-      "loss": 4.6407,
+      "loss": 4.606,
       "step": 4000
     },
     {
       "epoch": 0.288,
-      "grad_norm": 2.9283711910247803,
+      "grad_norm": 2.899388313293457,
       "learning_rate": 4.52e-05,
-      "loss": 4.5311,
+      "loss": 4.4928,
       "step": 4500
     },
     {
       "epoch": 0.32,
-      "grad_norm": 2.9473419189453125,
+      "grad_norm": 2.8987622261047363,
       "learning_rate": 4.466666666666667e-05,
-      "loss": 4.424,
+      "loss": 4.3843,
       "step": 5000
     },
     {
       "epoch": 0.352,
-      "grad_norm": 3.1425673961639404,
+      "grad_norm": 2.9992928504943848,
       "learning_rate": 4.413333333333334e-05,
-      "loss": 4.3445,
+      "loss": 4.3053,
       "step": 5500
     },
     {
       "epoch": 0.384,
-      "grad_norm": 2.8221523761749268,
+      "grad_norm": 2.944384813308716,
       "learning_rate": 4.36e-05,
-      "loss": 4.2723,
+      "loss": 4.2293,
       "step": 6000
     },
     {
       "epoch": 0.416,
-      "grad_norm": 3.374150276184082,
+      "grad_norm": 3.726702928543091,
       "learning_rate": 4.3066666666666665e-05,
-      "loss": 4.1767,
+      "loss": 4.1326,
       "step": 6500
     },
     {
       "epoch": 0.448,
-      "grad_norm": 3.098217487335205,
+      "grad_norm": 3.1758687496185303,
       "learning_rate": 4.2533333333333335e-05,
-      "loss": 4.1078,
+      "loss": 4.063,
       "step": 7000
     },
     {
       "epoch": 0.48,
-      "grad_norm": 3.4040091037750244,
+      "grad_norm": 3.0743587017059326,
       "learning_rate": 4.2e-05,
-      "loss": 4.0436,
+      "loss": 3.9998,
       "step": 7500
     },
     {
       "epoch": 0.512,
-      "grad_norm": 3.0540003776550293,
+      "grad_norm": 3.0619800090789795,
       "learning_rate": 4.146666666666667e-05,
-      "loss": 3.9936,
+      "loss": 3.9534,
       "step": 8000
     },
     {
       "epoch": 0.544,
-      "grad_norm": 3.0855867862701416,
+      "grad_norm": 3.1200902462005615,
       "learning_rate": 4.093333333333334e-05,
-      "loss": 3.9431,
+      "loss": 3.8978,
       "step": 8500
     },
     {
       "epoch": 0.576,
-      "grad_norm": 3.2475154399871826,
+      "grad_norm": 3.367762327194214,
       "learning_rate": 4.0400000000000006e-05,
-      "loss": 3.8871,
+      "loss": 3.8429,
       "step": 9000
     },
     {
       "epoch": 0.608,
-      "grad_norm": 3.1229586601257324,
+      "grad_norm": 3.1497597694396973,
       "learning_rate": 3.986666666666667e-05,
-      "loss": 3.8427,
+      "loss": 3.8007,
       "step": 9500
     },
     {
       "epoch": 0.64,
-      "grad_norm": 3.096017599105835,
+      "grad_norm": 3.31646728515625,
       "learning_rate": 3.933333333333333e-05,
-      "loss": 3.8009,
+      "loss": 3.7613,
       "step": 10000
     },
     {
       "epoch": 0.672,
-      "grad_norm": 3.2667548656463623,
+      "grad_norm": 3.294649839401245,
       "learning_rate": 3.88e-05,
-      "loss": 3.7488,
+      "loss": 3.7133,
       "step": 10500
     },
     {
       "epoch": 0.704,
-      "grad_norm": 3.1193501949310303,
+      "grad_norm": 3.1724936962127686,
       "learning_rate": 3.8266666666666664e-05,
-      "loss": 3.7168,
+      "loss": 3.6818,
       "step": 11000
     },
     {
       "epoch": 0.736,
-      "grad_norm": 3.8446643352508545,
+      "grad_norm": 3.0441651344299316,
       "learning_rate": 3.773333333333334e-05,
-      "loss": 3.6892,
+      "loss": 3.6576,
       "step": 11500
     },
     {
       "epoch": 0.768,
-      "grad_norm": 3.165462017059326,
+      "grad_norm": 3.160169839859009,
       "learning_rate": 3.72e-05,
-      "loss": 3.6588,
+      "loss": 3.6271,
       "step": 12000
     },
     {
       "epoch": 0.8,
-      "grad_norm": 3.333284378051758,
+      "grad_norm": 3.3029627799987793,
       "learning_rate": 3.6666666666666666e-05,
-      "loss": 3.6215,
+      "loss": 3.5932,
       "step": 12500
     },
     {
       "epoch": 0.832,
-      "grad_norm": 3.2972261905670166,
+      "grad_norm": 3.1690776348114014,
       "learning_rate": 3.6133333333333335e-05,
-      "loss": 3.5853,
+      "loss": 3.559,
       "step": 13000
     },
     {
       "epoch": 0.864,
-      "grad_norm": 3.358686685562134,
+      "grad_norm": 3.3345956802368164,
       "learning_rate": 3.56e-05,
-      "loss": 3.5631,
+      "loss": 3.5387,
       "step": 13500
     },
     {
       "epoch": 0.896,
-      "grad_norm": 3.1134519577026367,
+      "grad_norm": 3.0441858768463135,
       "learning_rate": 3.506666666666667e-05,
-      "loss": 3.5408,
+      "loss": 3.5169,
       "step": 14000
     },
     {
       "epoch": 0.928,
-      "grad_norm": 3.2438101768493652,
+      "grad_norm": 3.24686598777771,
       "learning_rate": 3.453333333333334e-05,
-      "loss": 3.5045,
+      "loss": 3.4847,
       "step": 14500
     },
     {
       "epoch": 0.96,
-      "grad_norm": 3.3038313388824463,
+      "grad_norm": 3.2995717525482178,
       "learning_rate": 3.4000000000000007e-05,
-      "loss": 3.4869,
+      "loss": 3.4684,
       "step": 15000
     },
     {
       "epoch": 0.992,
-      "grad_norm": 3.4895401000976562,
+      "grad_norm": 3.398087501525879,
       "learning_rate": 3.346666666666667e-05,
-      "loss": 3.4762,
+      "loss": 3.4568,
       "step": 15500
     },
     {
       "epoch": 1.024,
-      "grad_norm": 3.14111590385437,
+      "grad_norm": 3.1483492851257324,
       "learning_rate": 3.293333333333333e-05,
-      "loss": 3.445,
+      "loss": 3.429,
       "step": 16000
     },
     {
       "epoch": 1.056,
-      "grad_norm": 3.072317361831665,
+      "grad_norm": 3.1493332386016846,
       "learning_rate": 3.24e-05,
-      "loss": 3.4168,
+      "loss": 3.4016,
       "step": 16500
     },
     {
       "epoch": 1.088,
-      "grad_norm": 3.383622407913208,
+      "grad_norm": 3.290648937225342,
       "learning_rate": 3.1866666666666664e-05,
-      "loss": 3.4061,
+      "loss": 3.3935,
       "step": 17000
     },
     {
       "epoch": 1.12,
-      "grad_norm": 3.311823606491089,
+      "grad_norm": 3.2807164192199707,
       "learning_rate": 3.1333333333333334e-05,
-      "loss": 3.3891,
+      "loss": 3.3751,
       "step": 17500
     },
     {
       "epoch": 1.152,
-      "grad_norm": 3.3376855850219727,
+      "grad_norm": 3.3905978202819824,
       "learning_rate": 3.08e-05,
-      "loss": 3.3763,
+      "loss": 3.3667,
       "step": 18000
     },
     {
       "epoch": 1.184,
-      "grad_norm": 3.310765266418457,
+      "grad_norm": 3.1638753414154053,
       "learning_rate": 3.0266666666666666e-05,
-      "loss": 3.3565,
+      "loss": 3.3469,
       "step": 18500
     },
     {
       "epoch": 1.216,
-      "grad_norm": 3.721555709838867,
+      "grad_norm": 3.80301570892334,
       "learning_rate": 2.9733333333333336e-05,
-      "loss": 3.346,
+      "loss": 3.3369,
       "step": 19000
     },
     {
       "epoch": 1.248,
-      "grad_norm": 3.3782217502593994,
+      "grad_norm": 3.5537655353546143,
       "learning_rate": 2.9199999999999998e-05,
-      "loss": 3.3234,
+      "loss": 3.3132,
       "step": 19500
     },
     {
       "epoch": 1.28,
-      "grad_norm": 4.35649299621582,
+      "grad_norm": 3.344750165939331,
       "learning_rate": 2.8666666666666668e-05,
-      "loss": 3.3122,
+      "loss": 3.3075,
       "step": 20000
     },
     {
       "epoch": 1.312,
-      "grad_norm": 3.3569066524505615,
+      "grad_norm": 3.435274839401245,
       "learning_rate": 2.8133333333333334e-05,
-      "loss": 3.3093,
+      "loss": 3.3028,
       "step": 20500
     },
     {
       "epoch": 1.3439999999999999,
-      "grad_norm": 3.2398617267608643,
+      "grad_norm": 3.3021931648254395,
       "learning_rate": 2.7600000000000003e-05,
-      "loss": 3.2831,
+      "loss": 3.2771,
       "step": 21000
     },
     {
       "epoch": 1.376,
-      "grad_norm": 3.3986246585845947,
+      "grad_norm": 3.332207679748535,
       "learning_rate": 2.706666666666667e-05,
-      "loss": 3.2806,
+      "loss": 3.2732,
       "step": 21500
     },
     {
       "epoch": 1.408,
-      "grad_norm": 3.310861349105835,
+      "grad_norm": 3.382160186767578,
       "learning_rate": 2.6533333333333332e-05,
-      "loss": 3.2686,
+      "loss": 3.2632,
       "step": 22000
     },
     {
       "epoch": 1.44,
-      "grad_norm": 3.4433434009552,
+      "grad_norm": 3.399611711502075,
       "learning_rate": 2.6000000000000002e-05,
-      "loss": 3.2535,
+      "loss": 3.2474,
       "step": 22500
     },
     {
       "epoch": 1.472,
-      "grad_norm": 3.2779598236083984,
+      "grad_norm": 3.369029998779297,
       "learning_rate": 2.5466666666666668e-05,
-      "loss": 3.2505,
+      "loss": 3.2456,
       "step": 23000
     },
     {
       "epoch": 1.504,
-      "grad_norm": 3.1767616271972656,
+      "grad_norm": 3.1563122272491455,
       "learning_rate": 2.4933333333333334e-05,
-      "loss": 3.2289,
+      "loss": 3.2264,
       "step": 23500
     },
     {
       "epoch": 1.536,
-      "grad_norm": 3.530750274658203,
+      "grad_norm": 3.5122745037078857,
       "learning_rate": 2.44e-05,
-      "loss": 3.2241,
+      "loss": 3.219,
       "step": 24000
     },
     {
       "epoch": 1.568,
-      "grad_norm": 3.5728418827056885,
+      "grad_norm": 3.5229005813598633,
       "learning_rate": 2.3866666666666666e-05,
-      "loss": 3.2142,
+      "loss": 3.2115,
       "step": 24500
     },
     {
       "epoch": 1.6,
-      "grad_norm": 3.2718331813812256,
+      "grad_norm": 3.275137186050415,
       "learning_rate": 2.3333333333333336e-05,
-      "loss": 3.2034,
+      "loss": 3.1998,
       "step": 25000
     },
     {
       "epoch": 1.6320000000000001,
-      "grad_norm": 3.2709009647369385,
+      "grad_norm": 3.5656373500823975,
       "learning_rate": 2.2800000000000002e-05,
-      "loss": 3.1924,
+      "loss": 3.1895,
       "step": 25500
     },
     {
       "epoch": 1.6640000000000001,
-      "grad_norm": 3.323166608810425,
+      "grad_norm": 3.359915256500244,
       "learning_rate": 2.2266666666666668e-05,
-      "loss": 3.1797,
+      "loss": 3.1788,
       "step": 26000
     },
     {
       "epoch": 1.696,
-      "grad_norm": 3.4490272998809814,
+      "grad_norm": 3.496894121170044,
       "learning_rate": 2.1733333333333334e-05,
-      "loss": 3.1763,
+      "loss": 3.1757,
       "step": 26500
     },
     {
       "epoch": 1.728,
-      "grad_norm": 3.6509385108947754,
+      "grad_norm": 3.5603113174438477,
       "learning_rate": 2.12e-05,
-      "loss": 3.1778,
+      "loss": 3.1764,
       "step": 27000
     },
     {
       "epoch": 1.76,
-      "grad_norm": 3.481295108795166,
+      "grad_norm": 3.5351145267486572,
       "learning_rate": 2.0666666666666666e-05,
-      "loss": 3.1622,
+      "loss": 3.1619,
       "step": 27500
     },
     {
       "epoch": 1.792,
-      "grad_norm": 3.4964404106140137,
+      "grad_norm": 3.6555192470550537,
       "learning_rate": 2.0133333333333336e-05,
-      "loss": 3.1628,
+      "loss": 3.1618,
       "step": 28000
     },
     {
       "epoch": 1.8239999999999998,
-      "grad_norm": 3.550084352493286,
+      "grad_norm": 3.527869939804077,
       "learning_rate": 1.9600000000000002e-05,
-      "loss": 3.1446,
+      "loss": 3.1434,
       "step": 28500
     },
     {
       "epoch": 1.8559999999999999,
-      "grad_norm": 3.4653377532958984,
+      "grad_norm": 3.3731658458709717,
       "learning_rate": 1.9066666666666668e-05,
-      "loss": 3.1462,
+      "loss": 3.1443,
       "step": 29000
     },
     {
       "epoch": 1.888,
-      "grad_norm": 3.488093376159668,
+      "grad_norm": 3.5230460166931152,
       "learning_rate": 1.8533333333333334e-05,
-      "loss": 3.1319,
+      "loss": 3.1307,
       "step": 29500
     },
     {
       "epoch": 1.92,
-      "grad_norm": 3.3135769367218018,
+      "grad_norm": 3.2880985736846924,
       "learning_rate": 1.8e-05,
-      "loss": 3.135,
+      "loss": 3.1354,
       "step": 30000
     },
     {
       "epoch": 1.952,
-      "grad_norm": 3.5230553150177,
+      "grad_norm": 4.2027387619018555,
       "learning_rate": 1.7466666666666667e-05,
-      "loss": 3.1337,
+      "loss": 3.1349,
       "step": 30500
     },
     {
       "epoch": 1.984,
-      "grad_norm": 3.5588736534118652,
+      "grad_norm": 3.8834004402160645,
       "learning_rate": 1.6933333333333333e-05,
-      "loss": 3.1299,
+      "loss": 3.1287,
       "step": 31000
     },
     {
       "epoch": 2.016,
-      "grad_norm": 3.321937084197998,
+      "grad_norm": 3.5137991905212402,
       "learning_rate": 1.6400000000000002e-05,
-      "loss": 3.1097,
+      "loss": 3.107,
       "step": 31500
     },
     {
       "epoch": 2.048,
-      "grad_norm": 3.56290340423584,
+      "grad_norm": 5.010372161865234,
       "learning_rate": 1.586666666666667e-05,
-      "loss": 3.112,
+      "loss": 3.111,
       "step": 32000
     },
     {
       "epoch": 2.08,
-      "grad_norm": 3.408571720123291,
+      "grad_norm": 3.338728189468384,
       "learning_rate": 1.5333333333333334e-05,
-      "loss": 3.1055,
+      "loss": 3.1047,
       "step": 32500
     },
     {
       "epoch": 2.112,
-      "grad_norm": 3.9053122997283936,
+      "grad_norm": 3.71789813041687,
       "learning_rate": 1.48e-05,
-      "loss": 3.1013,
+      "loss": 3.1014,
       "step": 33000
     },
     {
       "epoch": 2.144,
-      "grad_norm": 3.7654247283935547,
+      "grad_norm": 3.7948594093322754,
       "learning_rate": 1.4266666666666667e-05,
-      "loss": 3.0976,
+      "loss": 3.097,
       "step": 33500
     },
     {
       "epoch": 2.176,
-      "grad_norm": 3.6201889514923096,
+      "grad_norm": 3.650618314743042,
       "learning_rate": 1.3733333333333335e-05,
-      "loss": 3.097,
+      "loss": 3.0967,
       "step": 34000
     },
     {
       "epoch": 2.208,
-      "grad_norm": 3.573209524154663,
+      "grad_norm": 3.4917681217193604,
       "learning_rate": 1.32e-05,
-      "loss": 3.1063,
+      "loss": 3.1064,
       "step": 34500
     },
     {
       "epoch": 2.24,
-      "grad_norm": 3.4236044883728027,
+      "grad_norm": 3.5514914989471436,
       "learning_rate": 1.2666666666666668e-05,
-      "loss": 3.0929,
+      "loss": 3.094,
       "step": 35000
     },
     {
       "epoch": 2.2720000000000002,
-      "grad_norm": 3.5758888721466064,
+      "grad_norm": 3.4345743656158447,
       "learning_rate": 1.2133333333333335e-05,
-      "loss": 3.0818,
+      "loss": 3.082,
       "step": 35500
     },
     {
       "epoch": 2.304,
-      "grad_norm": 3.563825845718384,
+      "grad_norm": 3.6274352073669434,
       "learning_rate": 1.16e-05,
-      "loss": 3.0857,
+      "loss": 3.0861,
       "step": 36000
     },
     {
       "epoch": 2.336,
-      "grad_norm": 3.4262824058532715,
+      "grad_norm": 3.6665496826171875,
       "learning_rate": 1.1066666666666667e-05,
-      "loss": 3.0721,
+      "loss": 3.0728,
       "step": 36500
     },
     {
       "epoch": 2.368,
-      "grad_norm": 3.4638020992279053,
+      "grad_norm": 3.657233953475952,
       "learning_rate": 1.0533333333333335e-05,
-      "loss": 3.0723,
+      "loss": 3.0727,
       "step": 37000
     },
     {
       "epoch": 2.4,
-      "grad_norm": 3.6780405044555664,
+      "grad_norm": 3.726337194442749,
       "learning_rate": 1e-05,
-      "loss": 3.07,
+      "loss": 3.0717,
       "step": 37500
     },
     {
       "epoch": 2.432,
-      "grad_norm": 3.6551175117492676,
+      "grad_norm": 3.7187366485595703,
       "learning_rate": 9.466666666666667e-06,
-      "loss": 3.0587,
+      "loss": 3.059,
       "step": 38000
     },
     {
       "epoch": 2.464,
-      "grad_norm": 3.607362747192383,
+      "grad_norm": 3.5927035808563232,
       "learning_rate": 8.933333333333333e-06,
-      "loss": 3.0675,
+      "loss": 3.0698,
       "step": 38500
     },
     {
       "epoch": 2.496,
-      "grad_norm": 3.5600104331970215,
+      "grad_norm": 3.5619006156921387,
       "learning_rate": 8.400000000000001e-06,
-      "loss": 3.0631,
+      "loss": 3.0625,
       "step": 39000
     },
     {
       "epoch": 2.528,
-      "grad_norm": 3.9264795780181885,
+      "grad_norm": 3.679213762283325,
       "learning_rate": 7.866666666666667e-06,
-      "loss": 3.0551,
+      "loss": 3.0565,
       "step": 39500
     },
     {
       "epoch": 2.56,
-      "grad_norm": 3.7913153171539307,
+      "grad_norm": 3.4748876094818115,
       "learning_rate": 7.333333333333334e-06,
       "loss": 3.0556,
       "step": 40000
     },
     {
       "epoch": 2.592,
-      "grad_norm": 3.5976860523223877,
+      "grad_norm": 3.483736515045166,
       "learning_rate": 6.800000000000001e-06,
-      "loss": 3.0407,
+      "loss": 3.0418,
       "step": 40500
     },
     {
       "epoch": 2.624,
-      "grad_norm": 3.384963035583496,
+      "grad_norm": 3.4663684368133545,
       "learning_rate": 6.266666666666666e-06,
-      "loss": 3.0453,
+      "loss": 3.0461,
       "step": 41000
     },
     {
       "epoch": 2.656,
-      "grad_norm": 3.6251819133758545,
+      "grad_norm": 3.416421413421631,
       "learning_rate": 5.733333333333333e-06,
-      "loss": 3.0485,
+      "loss": 3.0508,
       "step": 41500
     },
     {
       "epoch": 2.6879999999999997,
-      "grad_norm": 3.544095754623413,
+      "grad_norm": 3.5485949516296387,
       "learning_rate": 5.2e-06,
-      "loss": 3.0448,
+      "loss": 3.0447,
       "step": 42000
     },
     {
       "epoch": 2.7199999999999998,
-      "grad_norm": 3.513115882873535,
+      "grad_norm": 3.5995631217956543,
       "learning_rate": 4.666666666666667e-06,
-      "loss": 3.0477,
+      "loss": 3.0495,
       "step": 42500
     },
     {
       "epoch": 2.752,
-      "grad_norm": 3.5182816982269287,
+      "grad_norm": 3.676875591278076,
       "learning_rate": 4.133333333333333e-06,
-      "loss": 3.0303,
+      "loss": 3.0302,
       "step": 43000
     },
     {
       "epoch": 2.784,
-      "grad_norm": 3.7786800861358643,
+      "grad_norm": 3.650153160095215,
       "learning_rate": 3.6e-06,
       "loss": 3.0397,
       "step": 43500
     },
     {
       "epoch": 2.816,
-      "grad_norm": 3.6905770301818848,
+      "grad_norm": 3.6755716800689697,
       "learning_rate": 3.066666666666667e-06,
-      "loss": 3.0501,
+      "loss": 3.0504,
       "step": 44000
     },
     {
       "epoch": 2.848,
-      "grad_norm": 3.663628578186035,
+      "grad_norm": 3.6604747772216797,
       "learning_rate": 2.5333333333333334e-06,
-      "loss": 3.0525,
+      "loss": 3.0521,
       "step": 44500
     },
     {
       "epoch": 2.88,
-      "grad_norm": 3.723430871963501,
+      "grad_norm": 4.217142105102539,
       "learning_rate": 2.0000000000000003e-06,
-      "loss": 3.0353,
+      "loss": 3.0355,
       "step": 45000
     },
     {
       "epoch": 2.912,
-      "grad_norm": 3.6573081016540527,
+      "grad_norm": 3.6961653232574463,
       "learning_rate": 1.4666666666666667e-06,
-      "loss": 3.0442,
+      "loss": 3.0454,
       "step": 45500
     },
     {
       "epoch": 2.944,
-      "grad_norm": 4.172336101531982,
+      "grad_norm": 3.8596930503845215,
       "learning_rate": 9.333333333333334e-07,
-      "loss": 3.0454,
+      "loss": 3.0464,
       "step": 46000
     },
     {
       "epoch": 2.976,
-      "grad_norm": 3.7377212047576904,
+      "grad_norm": 3.8002963066101074,
       "learning_rate": 4.0000000000000003e-07,
-      "loss": 3.0469,
+      "loss": 3.0494,
       "step": 46500
     }
   ],

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:28be81758d786e53fa9d2c578564fc7ddc2b1377c8cc18dcda6f053214f10834
-size 5176
+oid sha256:365b0df6fa46689acb9348f6e307444d5eb3c995dadbc79368b21dcfcc70a321
+size 5240