Training in progress, step 100, checkpoint

last-checkpoint/config.json

@@ -7,11 +7,11 @@
   "bos_token_id": 2,
   "dtype": "bfloat16",
   "eos_token_id": 3,
-  "head_dim": 80,
+  "head_dim": 128,
   "hidden_act": "silu",
-  "hidden_size": 1280,
+  "hidden_size": 1536,
   "initializer_range": 0.02,
-  "intermediate_size": 5120,
+  "intermediate_size": 4608,
   "layer_types": [
     "full_attention",
     "full_attention",
@@ -48,14 +48,22 @@
     "full_attention",
     "full_attention",
     "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
     "full_attention"
   ],
   "max_position_embeddings": 8192,
-  "max_window_layers": 36,
-  "model_name": "random_model",
+  "max_window_layers": 44,
+  "model_name": "random_model2",
   "model_type": "qwen3",
-  "num_attention_heads": 40,
-  "num_hidden_layers": 36,
+  "num_attention_heads": 16,
+  "num_hidden_layers": 44,
   "num_key_value_heads": 8,
   "pad_token_id": 0,
   "rms_norm_eps": 1e-08,

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a31fb4e150696fbd81b685c233a7efae3a01549a539473980c0eb76b0c56dc3d
-size 2254879384
+oid sha256:4115292e473c854202107d5b5b763cf274fbd6445b61fcab4fdc825f3d91392e
+size 2856967144

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ad2e709ecb45e0af969d49e7f7cbb19ea0dd7bac74b71ef4d5d82503da0757b6
-size 1130472782
+oid sha256:10176e71c2e527c5378a53bb02722941ab104018709b0413788f5ac248d4efcb
+size 1431977238

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7c33e7e06e97e0f115a37527cc80045ed4276cbd1c8253ac1179212ed437b981
-size 8953
+oid sha256:225eac795dad119415278af3e8355e1ef3365025b06e35151a74d8fc845a7e57
+size 10617

last-checkpoint/trainer_state.json

@@ -11,702 +11,702 @@
   "log_history": [
     {
       "epoch": 0.001,
-      "grad_norm": 6.191351890563965,
+      "grad_norm": 5.846541404724121,
       "learning_rate": 0.0,
-      "loss": 11.086275100708008,
+      "loss": 11.106311798095703,
       "step": 1
     },
     {
       "epoch": 0.002,
-      "grad_norm": 6.183396339416504,
+      "grad_norm": 5.834257125854492,
       "learning_rate": 0.00011999999999999999,
-      "loss": 11.084210395812988,
+      "loss": 11.106182098388672,
       "step": 2
     },
     {
       "epoch": 0.003,
-      "grad_norm": 40.25852584838867,
+      "grad_norm": 60.36576461791992,
       "learning_rate": 0.00023999999999999998,
-      "loss": 11.06926441192627,
+      "loss": 11.599030494689941,
       "step": 3
     },
     {
       "epoch": 0.004,
-      "grad_norm": 3.6167638301849365,
+      "grad_norm": 3.9030606746673584,
       "learning_rate": 0.00035999999999999997,
-      "loss": 12.345776557922363,
+      "loss": 12.58960247039795,
       "step": 4
     },
     {
       "epoch": 0.005,
-      "grad_norm": 37.04316711425781,
+      "grad_norm": 34.81679916381836,
       "learning_rate": 0.00047999999999999996,
-      "loss": 12.08117961883545,
+      "loss": 11.170536994934082,
       "step": 5
     },
     {
       "epoch": 0.006,
-      "grad_norm": 2.5832877159118652,
+      "grad_norm": 2.888207197189331,
       "learning_rate": 0.0006,
-      "loss": 10.664973258972168,
+      "loss": 11.16025447845459,
       "step": 6
     },
     {
       "epoch": 0.007,
-      "grad_norm": 2.420443296432495,
+      "grad_norm": 2.596513509750366,
       "learning_rate": 0.0007199999999999999,
-      "loss": 9.976428031921387,
+      "loss": 10.094969749450684,
       "step": 7
     },
     {
       "epoch": 0.008,
-      "grad_norm": 2.208857774734497,
+      "grad_norm": 2.5529680252075195,
       "learning_rate": 0.0008399999999999999,
-      "loss": 9.32384204864502,
+      "loss": 9.325643539428711,
       "step": 8
     },
     {
       "epoch": 0.009,
-      "grad_norm": 1.4080778360366821,
+      "grad_norm": 2.799875259399414,
       "learning_rate": 0.0009599999999999999,
-      "loss": 8.844566345214844,
+      "loss": 8.808692932128906,
       "step": 9
     },
     {
       "epoch": 0.01,
-      "grad_norm": 0.932022750377655,
+      "grad_norm": 1.7801514863967896,
       "learning_rate": 0.00108,
-      "loss": 8.574441909790039,
+      "loss": 8.572772026062012,
       "step": 10
     },
     {
       "epoch": 0.011,
-      "grad_norm": 0.905206024646759,
+      "grad_norm": 1.6746641397476196,
       "learning_rate": 0.0012,
-      "loss": 8.570406913757324,
+      "loss": 8.614738464355469,
       "step": 11
     },
     {
       "epoch": 0.012,
-      "grad_norm": 1.9950780868530273,
+      "grad_norm": 1.4035816192626953,
       "learning_rate": 0.0012,
-      "loss": 8.654431343078613,
+      "loss": 8.762595176696777,
       "step": 12
     },
     {
       "epoch": 0.013,
-      "grad_norm": 1.4423072338104248,
+      "grad_norm": 5.130497455596924,
       "learning_rate": 0.0012,
-      "loss": 8.775788307189941,
+      "loss": 8.738160133361816,
       "step": 13
     },
     {
       "epoch": 0.014,
-      "grad_norm": 1.3934134244918823,
+      "grad_norm": 2.1428215503692627,
       "learning_rate": 0.0012,
-      "loss": 8.70737361907959,
+      "loss": 8.727628707885742,
       "step": 14
     },
     {
       "epoch": 0.015,
-      "grad_norm": 0.6730704307556152,
+      "grad_norm": 2.2844486236572266,
       "learning_rate": 0.0012,
-      "loss": 8.58652114868164,
+      "loss": 8.695308685302734,
       "step": 15
     },
     {
       "epoch": 0.016,
-      "grad_norm": 1.3850765228271484,
+      "grad_norm": 1.5479906797409058,
       "learning_rate": 0.0012,
-      "loss": 8.599067687988281,
+      "loss": 8.661114692687988,
       "step": 16
     },
     {
       "epoch": 0.017,
-      "grad_norm": 1.1575509309768677,
+      "grad_norm": 0.7431973218917847,
       "learning_rate": 0.0012,
-      "loss": 8.577943801879883,
+      "loss": 8.624454498291016,
       "step": 17
     },
     {
       "epoch": 0.018,
-      "grad_norm": 1.094803810119629,
+      "grad_norm": 1.2121706008911133,
       "learning_rate": 0.0012,
-      "loss": 8.583452224731445,
+      "loss": 8.594417572021484,
       "step": 18
     },
     {
       "epoch": 0.019,
-      "grad_norm": 0.9581138491630554,
+      "grad_norm": 0.8720210194587708,
       "learning_rate": 0.0012,
-      "loss": 8.573091506958008,
+      "loss": 8.565889358520508,
       "step": 19
     },
     {
       "epoch": 0.02,
-      "grad_norm": 0.8642209768295288,
+      "grad_norm": 0.7236080765724182,
       "learning_rate": 0.0012,
-      "loss": 8.558369636535645,
+      "loss": 8.575213432312012,
       "step": 20
     },
     {
       "epoch": 0.021,
-      "grad_norm": 0.5513532757759094,
+      "grad_norm": 1.5400056838989258,
       "learning_rate": 0.0012,
-      "loss": 8.547396659851074,
+      "loss": 8.587174415588379,
       "step": 21
     },
     {
       "epoch": 0.022,
-      "grad_norm": 0.5867564678192139,
+      "grad_norm": 0.48110371828079224,
       "learning_rate": 0.0012,
-      "loss": 8.557206153869629,
+      "loss": 8.618706703186035,
       "step": 22
     },
     {
       "epoch": 0.023,
-      "grad_norm": 0.5513554215431213,
+      "grad_norm": 0.9107239842414856,
       "learning_rate": 0.0012,
-      "loss": 8.519376754760742,
+      "loss": 8.568038940429688,
       "step": 23
     },
     {
       "epoch": 0.024,
-      "grad_norm": 0.5908657908439636,
+      "grad_norm": 0.3714894652366638,
       "learning_rate": 0.0012,
-      "loss": 8.571252822875977,
+      "loss": 8.594157218933105,
       "step": 24
     },
     {
       "epoch": 0.025,
-      "grad_norm": 0.5163703560829163,
+      "grad_norm": 0.8364576101303101,
       "learning_rate": 0.0012,
-      "loss": 8.550788879394531,
+      "loss": 8.578184127807617,
       "step": 25
     },
     {
       "epoch": 0.026,
-      "grad_norm": 0.6450480818748474,
+      "grad_norm": 0.4300350546836853,
       "learning_rate": 0.0012,
-      "loss": 8.553627014160156,
+      "loss": 8.577058792114258,
       "step": 26
     },
     {
       "epoch": 0.027,
-      "grad_norm": 0.4144176244735718,
+      "grad_norm": 0.707401692867279,
       "learning_rate": 0.0012,
-      "loss": 8.526437759399414,
+      "loss": 8.55170726776123,
       "step": 27
     },
     {
       "epoch": 0.028,
-      "grad_norm": 0.5550076961517334,
+      "grad_norm": 0.4554755389690399,
       "learning_rate": 0.0012,
-      "loss": 8.547487258911133,
+      "loss": 8.578386306762695,
       "step": 28
     },
     {
       "epoch": 0.029,
-      "grad_norm": 0.3792600631713867,
+      "grad_norm": 0.9380117654800415,
       "learning_rate": 0.0012,
-      "loss": 8.534337043762207,
+      "loss": 8.570149421691895,
       "step": 29
     },
     {
       "epoch": 0.03,
-      "grad_norm": 0.38315337896347046,
+      "grad_norm": 0.3442000448703766,
       "learning_rate": 0.0012,
-      "loss": 8.537225723266602,
+      "loss": 8.570940017700195,
       "step": 30
     },
     {
       "epoch": 0.031,
-      "grad_norm": 0.44758984446525574,
+      "grad_norm": 0.5590385794639587,
       "learning_rate": 0.0012,
-      "loss": 8.539337158203125,
+      "loss": 8.574797630310059,
       "step": 31
     },
     {
       "epoch": 0.032,
-      "grad_norm": 0.42012736201286316,
+      "grad_norm": 0.30500972270965576,
       "learning_rate": 0.0012,
-      "loss": 8.50920581817627,
+      "loss": 8.541150093078613,
       "step": 32
     },
     {
       "epoch": 0.033,
-      "grad_norm": 0.4103575348854065,
+      "grad_norm": 0.5824410319328308,
       "learning_rate": 0.0012,
-      "loss": 8.508569717407227,
+      "loss": 8.540495872497559,
       "step": 33
     },
     {
       "epoch": 0.034,
-      "grad_norm": 0.2795290946960449,
+      "grad_norm": 0.34406864643096924,
       "learning_rate": 0.0012,
-      "loss": 8.499092102050781,
+      "loss": 8.531632423400879,
       "step": 34
     },
     {
       "epoch": 0.035,
-      "grad_norm": 0.47300985455513,
+      "grad_norm": 0.45740342140197754,
       "learning_rate": 0.0012,
-      "loss": 8.508503913879395,
+      "loss": 8.53554630279541,
       "step": 35
     },
     {
       "epoch": 0.036,
-      "grad_norm": 0.27382028102874756,
+      "grad_norm": 0.2679101526737213,
       "learning_rate": 0.0012,
-      "loss": 8.498414039611816,
+      "loss": 8.523307800292969,
       "step": 36
     },
     {
       "epoch": 0.037,
-      "grad_norm": 0.37145861983299255,
+      "grad_norm": 0.39295539259910583,
       "learning_rate": 0.0012,
-      "loss": 8.516667366027832,
+      "loss": 8.545315742492676,
       "step": 37
     },
     {
       "epoch": 0.038,
-      "grad_norm": 0.26397329568862915,
+      "grad_norm": 0.29337337613105774,
       "learning_rate": 0.0012,
-      "loss": 8.515829086303711,
+      "loss": 8.546914100646973,
       "step": 38
     },
     {
       "epoch": 0.039,
-      "grad_norm": 0.2639530599117279,
+      "grad_norm": 0.4711088538169861,
       "learning_rate": 0.0012,
-      "loss": 8.509860038757324,
+      "loss": 8.54129695892334,
       "step": 39
     },
     {
       "epoch": 0.04,
-      "grad_norm": 0.22002728283405304,
+      "grad_norm": 0.2972702085971832,
       "learning_rate": 0.0012,
-      "loss": 8.498579978942871,
+      "loss": 8.523733139038086,
       "step": 40
     },
     {
       "epoch": 0.041,
-      "grad_norm": 0.2718033194541931,
+      "grad_norm": 0.5003634691238403,
       "learning_rate": 0.0012,
-      "loss": 8.4937105178833,
+      "loss": 8.516210556030273,
       "step": 41
     },
     {
       "epoch": 0.042,
-      "grad_norm": 0.2288224995136261,
+      "grad_norm": 0.28751179575920105,
       "learning_rate": 0.0012,
-      "loss": 8.510730743408203,
+      "loss": 8.531023025512695,
       "step": 42
     },
     {
       "epoch": 0.043,
-      "grad_norm": 0.2424091100692749,
+      "grad_norm": 0.44586503505706787,
       "learning_rate": 0.0012,
-      "loss": 8.484707832336426,
+      "loss": 8.501187324523926,
       "step": 43
     },
     {
       "epoch": 0.044,
-      "grad_norm": 0.2990792393684387,
+      "grad_norm": 0.46735402941703796,
       "learning_rate": 0.0012,
-      "loss": 8.510951042175293,
+      "loss": 8.52404499053955,
       "step": 44
     },
     {
       "epoch": 0.045,
-      "grad_norm": 0.2716998755931854,
+      "grad_norm": 0.23489262163639069,
       "learning_rate": 0.0012,
-      "loss": 8.487112998962402,
+      "loss": 8.489968299865723,
       "step": 45
     },
     {
       "epoch": 0.046,
-      "grad_norm": 0.31668758392333984,
+      "grad_norm": 0.4332050681114197,
       "learning_rate": 0.0012,
-      "loss": 8.497540473937988,
+      "loss": 8.493889808654785,
       "step": 46
     },
     {
       "epoch": 0.047,
-      "grad_norm": 0.2717650830745697,
+      "grad_norm": 1.0877212285995483,
       "learning_rate": 0.0012,
-      "loss": 8.497724533081055,
+      "loss": 8.496402740478516,
       "step": 47
     },
     {
       "epoch": 0.048,
-      "grad_norm": 0.2814731299877167,
+      "grad_norm": 1.1888394355773926,
       "learning_rate": 0.0012,
-      "loss": 8.501847267150879,
+      "loss": 8.49997615814209,
       "step": 48
     },
     {
       "epoch": 0.049,
-      "grad_norm": 0.31272783875465393,
+      "grad_norm": 0.699862003326416,
       "learning_rate": 0.0012,
-      "loss": 8.50948715209961,
+      "loss": 8.48403263092041,
       "step": 49
     },
     {
       "epoch": 0.05,
-      "grad_norm": 0.24348217248916626,
+      "grad_norm": 2.3642783164978027,
       "learning_rate": 0.0012,
-      "loss": 8.495022773742676,
+      "loss": 8.480637550354004,
       "step": 50
     },
     {
       "epoch": 0.051,
-      "grad_norm": 0.3011173605918884,
+      "grad_norm": 0.5330905318260193,
       "learning_rate": 0.0012,
-      "loss": 8.496807098388672,
+      "loss": 8.465481758117676,
       "step": 51
     },
     {
       "epoch": 0.052,
-      "grad_norm": 0.29326388239860535,
+      "grad_norm": 1.400657057762146,
       "learning_rate": 0.0012,
-      "loss": 8.492173194885254,
+      "loss": 8.481971740722656,
       "step": 52
     },
     {
       "epoch": 0.053,
-      "grad_norm": 0.29722103476524353,
+      "grad_norm": 0.39073649048805237,
       "learning_rate": 0.0012,
-      "loss": 8.480097770690918,
+      "loss": 8.406377792358398,
       "step": 53
     },
     {
       "epoch": 0.054,
-      "grad_norm": 0.21530401706695557,
+      "grad_norm": 2.112400531768799,
       "learning_rate": 0.0012,
-      "loss": 8.49035930633545,
+      "loss": 8.428634643554688,
       "step": 54
     },
     {
       "epoch": 0.055,
-      "grad_norm": 0.2602374851703644,
+      "grad_norm": 0.8586761355400085,
       "learning_rate": 0.0012,
-      "loss": 8.483044624328613,
+      "loss": 8.388739585876465,
       "step": 55
     },
     {
       "epoch": 0.056,
-      "grad_norm": 0.26086798310279846,
+      "grad_norm": 1.381065011024475,
       "learning_rate": 0.0012,
-      "loss": 8.479232788085938,
+      "loss": 8.386281967163086,
       "step": 56
     },
     {
       "epoch": 0.057,
-      "grad_norm": 0.3874048590660095,
+      "grad_norm": 1.4827899932861328,
       "learning_rate": 0.0012,
-      "loss": 8.478414535522461,
+      "loss": 8.36591911315918,
       "step": 57
     },
     {
       "epoch": 0.058,
-      "grad_norm": 0.7225645780563354,
+      "grad_norm": 0.5299570560455322,
       "learning_rate": 0.0012,
-      "loss": 8.495964050292969,
+      "loss": 8.3040771484375,
       "step": 58
     },
     {
       "epoch": 0.059,
-      "grad_norm": 1.4137455224990845,
+      "grad_norm": 2.805574417114258,
       "learning_rate": 0.0012,
-      "loss": 8.577896118164062,
+      "loss": 8.333112716674805,
       "step": 59
     },
     {
       "epoch": 0.06,
-      "grad_norm": 0.5455415844917297,
+      "grad_norm": 1.8315175771713257,
       "learning_rate": 0.0012,
-      "loss": 8.526337623596191,
+      "loss": 8.282299041748047,
       "step": 60
     },
     {
       "epoch": 0.061,
-      "grad_norm": 2.8609142303466797,
+      "grad_norm": 1.1118725538253784,
       "learning_rate": 0.0012,
-      "loss": 8.514455795288086,
+      "loss": 8.234102249145508,
       "step": 61
     },
     {
       "epoch": 0.062,
-      "grad_norm": 1.440839171409607,
+      "grad_norm": 1.0642567873001099,
       "learning_rate": 0.0012,
-      "loss": 8.634124755859375,
+      "loss": 8.242690086364746,
       "step": 62
     },
     {
       "epoch": 0.063,
-      "grad_norm": 1.2359699010849,
+      "grad_norm": 1.470217227935791,
       "learning_rate": 0.0012,
-      "loss": 8.568485260009766,
+      "loss": 8.215409278869629,
       "step": 63
     },
     {
       "epoch": 0.064,
-      "grad_norm": 1.180612564086914,
+      "grad_norm": 0.8547876477241516,
       "learning_rate": 0.0012,
-      "loss": 8.51024341583252,
+      "loss": 8.196215629577637,
       "step": 64
     },
     {
       "epoch": 0.065,
-      "grad_norm": 1.0028116703033447,
+      "grad_norm": 1.2849491834640503,
       "learning_rate": 0.0012,
-      "loss": 8.499656677246094,
+      "loss": 8.176141738891602,
       "step": 65
     },
     {
       "epoch": 0.066,
-      "grad_norm": 2.423152446746826,
+      "grad_norm": 1.0881035327911377,
       "learning_rate": 0.0012,
-      "loss": 8.472902297973633,
+      "loss": 8.161955833435059,
       "step": 66
     },
     {
       "epoch": 0.067,
-      "grad_norm": 1.8138713836669922,
+      "grad_norm": 0.9094507098197937,
       "learning_rate": 0.0012,
-      "loss": 8.483245849609375,
+      "loss": 8.14596176147461,
       "step": 67
     },
     {
       "epoch": 0.068,
-      "grad_norm": 0.959982693195343,
+      "grad_norm": 0.5833266377449036,
       "learning_rate": 0.0012,
-      "loss": 8.483409881591797,
+      "loss": 8.109943389892578,
       "step": 68
     },
     {
       "epoch": 0.069,
-      "grad_norm": 0.8686690330505371,
+      "grad_norm": 0.6731480956077576,
       "learning_rate": 0.0012,
-      "loss": 8.440892219543457,
+      "loss": 8.148503303527832,
       "step": 69
     },
     {
       "epoch": 0.07,
-      "grad_norm": 0.7423587441444397,
+      "grad_norm": 0.9029632210731506,
       "learning_rate": 0.0012,
-      "loss": 8.409979820251465,
+      "loss": 8.14487361907959,
       "step": 70
     },
     {
       "epoch": 0.071,
-      "grad_norm": 0.88447505235672,
+      "grad_norm": 1.3046966791152954,
       "learning_rate": 0.0012,
-      "loss": 8.390122413635254,
+      "loss": 8.236167907714844,
       "step": 71
     },
     {
       "epoch": 0.072,
-      "grad_norm": 0.982702910900116,
+      "grad_norm": 1.8148207664489746,
       "learning_rate": 0.0012,
-      "loss": 8.37291431427002,
+      "loss": 8.141400337219238,
       "step": 72
     },
     {
       "epoch": 0.073,
-      "grad_norm": 0.6741560101509094,
+      "grad_norm": 1.0058391094207764,
       "learning_rate": 0.0012,
-      "loss": 8.374544143676758,
+      "loss": 8.162654876708984,
       "step": 73
     },
     {
       "epoch": 0.074,
-      "grad_norm": 0.9037299156188965,
+      "grad_norm": 1.0639183521270752,
       "learning_rate": 0.0012,
-      "loss": 8.357539176940918,
+      "loss": 8.147970199584961,
       "step": 74
     },
     {
       "epoch": 0.075,
-      "grad_norm": 0.5122299790382385,
+      "grad_norm": 1.2224985361099243,
       "learning_rate": 0.0012,
-      "loss": 8.336565017700195,
+      "loss": 8.111481666564941,
       "step": 75
     },
     {
       "epoch": 0.076,
-      "grad_norm": 1.5479395389556885,
+      "grad_norm": 0.537449836730957,
       "learning_rate": 0.0012,
-      "loss": 8.325927734375,
+      "loss": 8.093454360961914,
       "step": 76
     },
     {
       "epoch": 0.077,
-      "grad_norm": 1.3528615236282349,
+      "grad_norm": 1.4961029291152954,
       "learning_rate": 0.0012,
-      "loss": 8.327680587768555,
+      "loss": 8.102242469787598,
       "step": 77
     },
     {
       "epoch": 0.078,
-      "grad_norm": 0.7083619832992554,
+      "grad_norm": 0.7618138790130615,
       "learning_rate": 0.0012,
-      "loss": 8.325244903564453,
+      "loss": 8.114128112792969,
       "step": 78
     },
     {
       "epoch": 0.079,
-      "grad_norm": 0.9852667450904846,
+      "grad_norm": 0.7566311359405518,
       "learning_rate": 0.0012,
-      "loss": 8.305098533630371,
+      "loss": 8.08398723602295,
       "step": 79
     },
     {
       "epoch": 0.08,
-      "grad_norm": 0.5205386281013489,
+      "grad_norm": 0.6958624720573425,
       "learning_rate": 0.0012,
-      "loss": 8.274360656738281,
+      "loss": 8.074750900268555,
       "step": 80
     },
     {
       "epoch": 0.081,
-      "grad_norm": 0.946526825428009,
+      "grad_norm": 0.6900841593742371,
       "learning_rate": 0.0012,
-      "loss": 8.282258033752441,
+      "loss": 8.09200382232666,
       "step": 81
     },
     {
       "epoch": 0.082,
-      "grad_norm": 0.6880609393119812,
+      "grad_norm": 0.6930720806121826,
       "learning_rate": 0.0012,
-      "loss": 8.254171371459961,
+      "loss": 8.054723739624023,
       "step": 82
     },
     {
       "epoch": 0.083,
-      "grad_norm": 0.5522202253341675,
+      "grad_norm": 0.7483615875244141,
       "learning_rate": 0.0012,
-      "loss": 8.278489112854004,
+      "loss": 8.086467742919922,
       "step": 83
     },
     {
       "epoch": 0.084,
-      "grad_norm": 0.6842210292816162,
+      "grad_norm": 0.8259903788566589,
       "learning_rate": 0.0012,
-      "loss": 8.220597267150879,
+      "loss": 8.040852546691895,
       "step": 84
     },
     {
       "epoch": 0.085,
-      "grad_norm": 0.29896456003189087,
+      "grad_norm": 0.6626695394515991,
       "learning_rate": 0.0012,
-      "loss": 8.204936981201172,
+      "loss": 8.026881217956543,
       "step": 85
     },
     {
       "epoch": 0.086,
-      "grad_norm": 0.748399555683136,
+      "grad_norm": 0.48757404088974,
       "learning_rate": 0.0012,
-      "loss": 8.220271110534668,
+      "loss": 8.038444519042969,
       "step": 86
     },
     {
       "epoch": 0.087,
-      "grad_norm": 0.28956496715545654,
+      "grad_norm": 0.7604665160179138,
       "learning_rate": 0.0012,
-      "loss": 8.216758728027344,
+      "loss": 8.045461654663086,
       "step": 87
     },
     {
       "epoch": 0.088,
-      "grad_norm": 0.5921151638031006,
+      "grad_norm": 0.48548105359077454,
       "learning_rate": 0.0012,
-      "loss": 8.215239524841309,
+      "loss": 8.03872013092041,
       "step": 88
     },
     {
       "epoch": 0.089,
-      "grad_norm": 0.451276570558548,
+      "grad_norm": 0.9413471817970276,
       "learning_rate": 0.0012,
-      "loss": 8.17009162902832,
+      "loss": 8.008406639099121,
       "step": 89
     },
     {
       "epoch": 0.09,
-      "grad_norm": 0.5010297894477844,
+      "grad_norm": 0.3605736494064331,
       "learning_rate": 0.0012,
-      "loss": 8.161727905273438,
+      "loss": 8.004989624023438,
       "step": 90
     },
     {
       "epoch": 0.091,
-      "grad_norm": 0.30494871735572815,
+      "grad_norm": 0.7028122544288635,
       "learning_rate": 0.0012,
-      "loss": 8.166382789611816,
+      "loss": 8.022098541259766,
       "step": 91
     },
     {
       "epoch": 0.092,
-      "grad_norm": 0.43188294768333435,
+      "grad_norm": 0.3218722641468048,
       "learning_rate": 0.0012,
-      "loss": 8.167814254760742,
+      "loss": 8.010289192199707,
       "step": 92
     },
     {
       "epoch": 0.093,
-      "grad_norm": 0.28245413303375244,
+      "grad_norm": 0.7810201048851013,
       "learning_rate": 0.0012,
-      "loss": 8.160371780395508,
+      "loss": 8.008184432983398,
       "step": 93
     },
     {
       "epoch": 0.094,
-      "grad_norm": 0.4027661383152008,
+      "grad_norm": 0.7134589552879333,
       "learning_rate": 0.0012,
-      "loss": 8.109189987182617,
+      "loss": 7.970849990844727,
       "step": 94
     },
     {
       "epoch": 0.095,
-      "grad_norm": 0.2549852132797241,
+      "grad_norm": 0.4336091876029968,
       "learning_rate": 0.0012,
-      "loss": 8.114521026611328,
+      "loss": 7.978301525115967,
       "step": 95
     },
     {
       "epoch": 0.096,
-      "grad_norm": 0.3145453929901123,
+      "grad_norm": 0.8183775544166565,
       "learning_rate": 0.0012,
-      "loss": 8.10285472869873,
+      "loss": 7.978800296783447,
       "step": 96
     },
     {
       "epoch": 0.097,
-      "grad_norm": 0.37485775351524353,
+      "grad_norm": 0.5193440914154053,
       "learning_rate": 0.0012,
-      "loss": 8.09223747253418,
+      "loss": 7.968331336975098,
       "step": 97
     },
     {
       "epoch": 0.098,
-      "grad_norm": 0.36243894696235657,
+      "grad_norm": 0.46330469846725464,
       "learning_rate": 0.0012,
-      "loss": 8.064014434814453,
+      "loss": 7.943108081817627,
       "step": 98
     },
     {
       "epoch": 0.099,
-      "grad_norm": 0.2771179676055908,
+      "grad_norm": 0.6331789493560791,
       "learning_rate": 0.0012,
-      "loss": 8.062220573425293,
+      "loss": 7.956433296203613,
       "step": 99
     },
     {
       "epoch": 0.1,
-      "grad_norm": 0.2642522156238556,
+      "grad_norm": 0.5161998271942139,
       "learning_rate": 0.0012,
-      "loss": 8.07100772857666,
+      "loss": 7.983428955078125,
       "step": 100
     }
   ],
@@ -727,7 +727,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.469577768971264e+17,
+  "total_flos": 2.9353509544157184e+17,
   "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 sha256:1d58e58b86ebc8ee74fda679712e3a52e109c5c26268157eabc8b5c0cd94155a
+oid sha256:dd9108a715e52b3714ac9b3d76525c6e8beadca24fb404442fc72e2ce144ff05
 size 5777