Upload task output 1bb7f5eb-6f15-4cc7-904c-cbb98a510983

config.json

@@ -130,6 +130,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.51.3",
-  "use_cache": true,
+  "use_cache": false,
   "vocab_size": 32064
 }

generation_config.json

@@ -1,5 +1,7 @@
 {
-  "temperature": null,
-  "top_p": null,
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 32000,
+  "pad_token_id": 32000,
   "transformers_version": "4.51.3"
 }

loss.txt

@@ -1 +1 @@
-419,no_eval
+500,9.067032814025879

model-00001-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4e2c8692400c1dce77feead9d8b9a3df70ca02d680dbcfed4ad9072f35897f10
+oid sha256:c06e4d9ef0079babfe5495ac87b368c1ac69fa74896dfe69d6f5c2dd65edd003
 size 4991370968

trainer_state.json

@@ -2,597 +2,724 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.17656974294142436,
+  "epoch": 0.42122999157540014,
   "eval_steps": 500,
-  "global_step": 419,
+  "global_step": 500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.002107037505267594,
+      "epoch": 0.004212299915754001,
       "grad_norm": Infinity,
       "learning_rate": 1.0425224359183675e-05,
-      "loss": 9.2528,
+      "loss": 9.1367,
       "step": 5
     },
     {
-      "epoch": 0.004214075010535188,
+      "epoch": 0.008424599831508003,
       "grad_norm": Infinity,
       "learning_rate": 2.3456754808163266e-05,
-      "loss": 9.0876,
+      "loss": 9.1854,
       "step": 10
     },
     {
-      "epoch": 0.006321112515802781,
+      "epoch": 0.012636899747262006,
       "grad_norm": Infinity,
       "learning_rate": 3.6488285257142865e-05,
-      "loss": 9.071,
+      "loss": 9.2157,
       "step": 15
     },
     {
-      "epoch": 0.008428150021070375,
+      "epoch": 0.016849199663016005,
       "grad_norm": Infinity,
       "learning_rate": 4.951981570612245e-05,
-      "loss": 9.0988,
+      "loss": 9.1202,
       "step": 20
     },
     {
-      "epoch": 0.01053518752633797,
+      "epoch": 0.02106149957877001,
       "grad_norm": Infinity,
       "learning_rate": 6.255134615510205e-05,
-      "loss": 9.1175,
+      "loss": 9.0366,
       "step": 25
     },
     {
-      "epoch": 0.012642225031605562,
+      "epoch": 0.02527379949452401,
       "grad_norm": Infinity,
       "learning_rate": 7.558287660408165e-05,
-      "loss": 9.1526,
+      "loss": 8.9155,
       "step": 30
     },
     {
-      "epoch": 0.014749262536873156,
+      "epoch": 0.02948609941027801,
       "grad_norm": Infinity,
       "learning_rate": 8.861440705306124e-05,
-      "loss": 9.009,
+      "loss": 9.0951,
       "step": 35
     },
     {
-      "epoch": 0.01685630004214075,
+      "epoch": 0.03369839932603201,
       "grad_norm": Infinity,
-      "learning_rate": 9.122065932117784e-05,
-      "loss": 9.2034,
+      "learning_rate": 9.122049589805829e-05,
+      "loss": 9.2247,
       "step": 40
     },
     {
-      "epoch": 0.018963337547408345,
+      "epoch": 0.037910699241786014,
       "grad_norm": Infinity,
-      "learning_rate": 9.122044067089586e-05,
-      "loss": 9.0025,
+      "learning_rate": 9.121961334579207e-05,
+      "loss": 9.1174,
       "step": 45
     },
     {
-      "epoch": 0.02107037505267594,
+      "epoch": 0.04212299915754002,
       "grad_norm": Infinity,
-      "learning_rate": 9.122005382923049e-05,
-      "loss": 9.0069,
+      "learning_rate": 9.121805192576025e-05,
+      "loss": 9.1988,
       "step": 50
     },
     {
-      "epoch": 0.023177412557943533,
+      "epoch": 0.04633529907329402,
       "grad_norm": Infinity,
-      "learning_rate": 9.121949879808376e-05,
-      "loss": 8.9929,
+      "learning_rate": 9.121581166895088e-05,
+      "loss": 9.1803,
       "step": 55
     },
     {
-      "epoch": 0.025284450063211124,
+      "epoch": 0.05054759898904802,
       "grad_norm": Infinity,
-      "learning_rate": 9.121877558018465e-05,
-      "loss": 8.8893,
+      "learning_rate": 9.121289261982425e-05,
+      "loss": 9.0852,
       "step": 60
     },
     {
-      "epoch": 0.027391487568478718,
+      "epoch": 0.05475989890480202,
       "grad_norm": Infinity,
-      "learning_rate": 9.121788417908908e-05,
-      "loss": 9.0498,
+      "learning_rate": 9.120929483631201e-05,
+      "loss": 9.161,
       "step": 65
     },
     {
-      "epoch": 0.029498525073746312,
+      "epoch": 0.05897219882055602,
       "grad_norm": Infinity,
-      "learning_rate": 9.121682459917987e-05,
-      "loss": 9.1409,
+      "learning_rate": 9.120501838981599e-05,
+      "loss": 9.0158,
       "step": 70
     },
     {
-      "epoch": 0.0316055625790139,
+      "epoch": 0.06318449873631002,
       "grad_norm": Infinity,
-      "learning_rate": 9.121559684566679e-05,
-      "loss": 9.3385,
+      "learning_rate": 9.120006336520685e-05,
+      "loss": 9.0559,
       "step": 75
     },
     {
-      "epoch": 0.0337126000842815,
+      "epoch": 0.06739679865206402,
       "grad_norm": Infinity,
-      "learning_rate": 9.121420092458643e-05,
-      "loss": 9.0298,
+      "learning_rate": 9.11944298608223e-05,
+      "loss": 9.0891,
       "step": 80
     },
     {
-      "epoch": 0.03581963758954909,
+      "epoch": 0.07160909856781802,
       "grad_norm": Infinity,
-      "learning_rate": 9.121263684280228e-05,
-      "loss": 9.0127,
+      "learning_rate": 9.118811798846527e-05,
+      "loss": 9.1733,
       "step": 85
     },
     {
-      "epoch": 0.03792667509481669,
+      "epoch": 0.07582139848357203,
       "grad_norm": Infinity,
-      "learning_rate": 9.121090460800458e-05,
-      "loss": 9.0721,
+      "learning_rate": 9.118112787340156e-05,
+      "loss": 9.2267,
       "step": 90
     },
     {
-      "epoch": 0.04003371260008428,
+      "epoch": 0.08003369839932603,
       "grad_norm": Infinity,
-      "learning_rate": 9.120900422871043e-05,
-      "loss": 9.1312,
+      "learning_rate": 9.117345965435748e-05,
+      "loss": 9.0607,
       "step": 95
     },
     {
-      "epoch": 0.04214075010535188,
+      "epoch": 0.08424599831508003,
       "grad_norm": Infinity,
-      "learning_rate": 9.120693571426357e-05,
-      "loss": 9.114,
+      "learning_rate": 9.116511348351699e-05,
+      "loss": 9.0453,
       "step": 100
     },
     {
-      "epoch": 0.04424778761061947,
+      "epoch": 0.08845829823083404,
       "grad_norm": Infinity,
-      "learning_rate": 9.12046990748345e-05,
-      "loss": 9.1711,
+      "learning_rate": 9.115608952651875e-05,
+      "loss": 9.0472,
       "step": 105
     },
     {
-      "epoch": 0.046354825115887066,
+      "epoch": 0.09267059814658804,
       "grad_norm": Infinity,
-      "learning_rate": 9.120229432142029e-05,
-      "loss": 9.1641,
+      "learning_rate": 9.11463879624528e-05,
+      "loss": 9.1655,
       "step": 110
     },
     {
-      "epoch": 0.04846186262115466,
+      "epoch": 0.09688289806234204,
       "grad_norm": Infinity,
-      "learning_rate": 9.119972146584466e-05,
-      "loss": 9.2252,
+      "learning_rate": 9.113600898385701e-05,
+      "loss": 8.9799,
       "step": 115
     },
     {
-      "epoch": 0.05056890012642225,
+      "epoch": 0.10109519797809605,
       "grad_norm": Infinity,
-      "learning_rate": 9.119698052075782e-05,
-      "loss": 9.0076,
+      "learning_rate": 9.112495279671328e-05,
+      "loss": 9.2456,
       "step": 120
     },
     {
-      "epoch": 0.052675937631689845,
+      "epoch": 0.10530749789385004,
       "grad_norm": Infinity,
-      "learning_rate": 9.119407149963643e-05,
-      "loss": 9.0771,
+      "learning_rate": 9.11132196204434e-05,
+      "loss": 9.0901,
       "step": 125
     },
     {
-      "epoch": 0.054782975136957436,
+      "epoch": 0.10951979780960404,
       "grad_norm": Infinity,
-      "learning_rate": 9.11909944167836e-05,
-      "loss": 9.262,
+      "learning_rate": 9.11008096879048e-05,
+      "loss": 9.0431,
       "step": 130
     },
     {
-      "epoch": 0.056890012642225034,
+      "epoch": 0.11373209772535804,
       "grad_norm": Infinity,
-      "learning_rate": 9.118774928732871e-05,
-      "loss": 9.0539,
+      "learning_rate": 9.108772324538577e-05,
+      "loss": 9.153,
       "step": 135
     },
     {
-      "epoch": 0.058997050147492625,
+      "epoch": 0.11794439764111204,
       "grad_norm": Infinity,
-      "learning_rate": 9.118433612722738e-05,
-      "loss": 9.0142,
+      "learning_rate": 9.107396055260072e-05,
+      "loss": 9.2556,
       "step": 140
     },
     {
-      "epoch": 0.06110408765276022,
+      "epoch": 0.12215669755686605,
       "grad_norm": Infinity,
-      "learning_rate": 9.118075495326146e-05,
-      "loss": 9.1751,
+      "learning_rate": 9.105952188268497e-05,
+      "loss": 8.9465,
       "step": 145
     },
     {
-      "epoch": 0.0632111251580278,
+      "epoch": 0.12636899747262004,
       "grad_norm": Infinity,
-      "learning_rate": 9.117700578303887e-05,
-      "loss": 8.805,
+      "learning_rate": 9.104440752218928e-05,
+      "loss": 9.0371,
       "step": 150
     },
     {
-      "epoch": 0.0653181626632954,
+      "epoch": 0.13058129738837404,
       "grad_norm": Infinity,
-      "learning_rate": 9.117308863499353e-05,
-      "loss": 9.0702,
+      "learning_rate": 9.102861777107424e-05,
+      "loss": 9.2312,
       "step": 155
     },
     {
-      "epoch": 0.067425200168563,
+      "epoch": 0.13479359730412804,
       "grad_norm": Infinity,
-      "learning_rate": 9.116900352838523e-05,
-      "loss": 9.073,
+      "learning_rate": 9.101215294270431e-05,
+      "loss": 9.2269,
       "step": 160
     },
     {
-      "epoch": 0.0695322376738306,
+      "epoch": 0.13900589721988205,
       "grad_norm": Infinity,
-      "learning_rate": 9.116475048329966e-05,
-      "loss": 9.2132,
+      "learning_rate": 9.099501336384152e-05,
+      "loss": 9.0931,
       "step": 165
     },
     {
-      "epoch": 0.07163927517909818,
+      "epoch": 0.14321819713563605,
       "grad_norm": Infinity,
-      "learning_rate": 9.116032952064818e-05,
-      "loss": 9.1914,
+      "learning_rate": 9.097719937463912e-05,
+      "loss": 9.072,
       "step": 170
     },
     {
-      "epoch": 0.07374631268436578,
+      "epoch": 0.14743049705139005,
       "grad_norm": Infinity,
-      "learning_rate": 9.115574066216778e-05,
-      "loss": 9.0857,
+      "learning_rate": 9.095871132863466e-05,
+      "loss": 8.983,
       "step": 175
     },
     {
-      "epoch": 0.07585335018963338,
+      "epoch": 0.15164279696714406,
       "grad_norm": Infinity,
-      "learning_rate": 9.115098393042094e-05,
-      "loss": 9.3902,
+      "learning_rate": 9.093954959274312e-05,
+      "loss": 9.1125,
       "step": 180
     },
     {
-      "epoch": 0.07796038769490098,
+      "epoch": 0.15585509688289806,
       "grad_norm": Infinity,
-      "learning_rate": 9.114605934879554e-05,
-      "loss": 9.1069,
+      "learning_rate": 9.091971454724959e-05,
+      "loss": 8.8615,
       "step": 185
     },
     {
-      "epoch": 0.08006742520016856,
+      "epoch": 0.16006739679865206,
       "grad_norm": Infinity,
-      "learning_rate": 9.11409669415048e-05,
-      "loss": 9.0304,
+      "learning_rate": 9.089920658580165e-05,
+      "loss": 9.1443,
       "step": 190
     },
     {
-      "epoch": 0.08217446270543616,
+      "epoch": 0.16427969671440606,
       "grad_norm": Infinity,
-      "learning_rate": 9.113570673358704e-05,
-      "loss": 8.8525,
+      "learning_rate": 9.087802611540166e-05,
+      "loss": 9.2218,
       "step": 195
     },
     {
-      "epoch": 0.08428150021070376,
+      "epoch": 0.16849199663016007,
       "grad_norm": Infinity,
-      "learning_rate": 9.113027875090565e-05,
-      "loss": 9.3074,
+      "learning_rate": 9.085617355639865e-05,
+      "loss": 9.1446,
       "step": 200
     },
     {
-      "epoch": 0.08638853771597134,
+      "epoch": 0.17270429654591407,
       "grad_norm": Infinity,
-      "learning_rate": 9.112468302014893e-05,
-      "loss": 9.1235,
+      "learning_rate": 9.083364934247988e-05,
+      "loss": 9.2042,
       "step": 205
     },
     {
-      "epoch": 0.08849557522123894,
+      "epoch": 0.17691659646166807,
       "grad_norm": Infinity,
-      "learning_rate": 9.111891956882998e-05,
-      "loss": 9.0219,
+      "learning_rate": 9.081045392066242e-05,
+      "loss": 9.2255,
       "step": 210
     },
     {
-      "epoch": 0.09060261272650653,
+      "epoch": 0.18112889637742208,
       "grad_norm": Infinity,
-      "learning_rate": 9.11129884252865e-05,
-      "loss": 9.1624,
+      "learning_rate": 9.07865877512841e-05,
+      "loss": 8.9658,
       "step": 215
     },
     {
-      "epoch": 0.09270965023177413,
+      "epoch": 0.18534119629317608,
       "grad_norm": Infinity,
-      "learning_rate": 9.110688961868076e-05,
-      "loss": 9.0277,
+      "learning_rate": 9.076205130799451e-05,
+      "loss": 9.0488,
       "step": 220
     },
     {
-      "epoch": 0.09481668773704172,
+      "epoch": 0.18955349620893008,
       "grad_norm": Infinity,
-      "learning_rate": 9.110062317899935e-05,
-      "loss": 9.0036,
+      "learning_rate": 9.073684507774549e-05,
+      "loss": 8.9994,
       "step": 225
     },
     {
-      "epoch": 0.09692372524230931,
+      "epoch": 0.1937657961246841,
       "grad_norm": Infinity,
-      "learning_rate": 9.109418913705311e-05,
-      "loss": 9.0376,
+      "learning_rate": 9.071096956078153e-05,
+      "loss": 9.2227,
       "step": 230
     },
     {
-      "epoch": 0.09903076274757691,
+      "epoch": 0.1979780960404381,
       "grad_norm": Infinity,
-      "learning_rate": 9.108758752447692e-05,
-      "loss": 9.1026,
+      "learning_rate": 9.068442527062987e-05,
+      "loss": 9.0947,
       "step": 235
     },
     {
-      "epoch": 0.1011378002528445,
+      "epoch": 0.2021903959561921,
       "grad_norm": Infinity,
-      "learning_rate": 9.108081837372961e-05,
-      "loss": 9.2828,
+      "learning_rate": 9.065721273409019e-05,
+      "loss": 8.9357,
       "step": 240
     },
     {
-      "epoch": 0.10324483775811209,
+      "epoch": 0.2064026958719461,
       "grad_norm": Infinity,
-      "learning_rate": 9.107388171809369e-05,
-      "loss": 9.1233,
+      "learning_rate": 9.062933249122428e-05,
+      "loss": 9.1496,
       "step": 245
     },
     {
-      "epoch": 0.10535187526337969,
+      "epoch": 0.21061499578770007,
       "grad_norm": Infinity,
-      "learning_rate": 9.106677759167533e-05,
-      "loss": 9.2087,
+      "learning_rate": 9.060078509534528e-05,
+      "loss": 9.198,
       "step": 250
     },
     {
-      "epoch": 0.10745891276864729,
+      "epoch": 0.21482729570345407,
       "grad_norm": Infinity,
-      "learning_rate": 9.10595060294041e-05,
-      "loss": 9.0822,
+      "learning_rate": 9.057157111300668e-05,
+      "loss": 9.2222,
       "step": 255
     },
     {
-      "epoch": 0.10956595027391487,
+      "epoch": 0.21903959561920808,
       "grad_norm": Infinity,
-      "learning_rate": 9.105206706703279e-05,
-      "loss": 9.0266,
+      "learning_rate": 9.054169112399107e-05,
+      "loss": 9.104,
       "step": 260
     },
     {
-      "epoch": 0.11167298777918247,
+      "epoch": 0.22325189553496208,
       "grad_norm": Infinity,
-      "learning_rate": 9.104446074113729e-05,
-      "loss": 9.0158,
+      "learning_rate": 9.051114572129868e-05,
+      "loss": 9.2462,
       "step": 265
     },
     {
-      "epoch": 0.11378002528445007,
+      "epoch": 0.22746419545071608,
       "grad_norm": Infinity,
-      "learning_rate": 9.10366870891164e-05,
-      "loss": 9.1653,
+      "learning_rate": 9.047993551113556e-05,
+      "loss": 9.2007,
       "step": 270
     },
     {
-      "epoch": 0.11588706278971765,
+      "epoch": 0.2316764953664701,
       "grad_norm": Infinity,
-      "learning_rate": 9.102874614919157e-05,
-      "loss": 9.1888,
+      "learning_rate": 9.044806111290159e-05,
+      "loss": 9.0567,
       "step": 275
     },
     {
-      "epoch": 0.11799410029498525,
+      "epoch": 0.2358887952822241,
       "grad_norm": Infinity,
-      "learning_rate": 9.102063796040684e-05,
-      "loss": 9.1944,
+      "learning_rate": 9.041552315917816e-05,
+      "loss": 9.3158,
       "step": 280
     },
     {
-      "epoch": 0.12010113780025285,
+      "epoch": 0.2401010951979781,
       "grad_norm": Infinity,
-      "learning_rate": 9.101236256262852e-05,
-      "loss": 8.8196,
+      "learning_rate": 9.038232229571564e-05,
+      "loss": 9.127,
       "step": 285
     },
     {
-      "epoch": 0.12220817530552044,
+      "epoch": 0.2443133951137321,
       "grad_norm": Infinity,
-      "learning_rate": 9.10039199965451e-05,
-      "loss": 9.1051,
+      "learning_rate": 9.034845918142056e-05,
+      "loss": 9.139,
       "step": 290
     },
     {
-      "epoch": 0.12431521281078803,
+      "epoch": 0.2485256950294861,
       "grad_norm": Infinity,
-      "learning_rate": 9.099531030366696e-05,
-      "loss": 9.0194,
+      "learning_rate": 9.031393448834246e-05,
+      "loss": 8.9646,
       "step": 295
     },
     {
-      "epoch": 0.1264222503160556,
+      "epoch": 0.2527379949452401,
       "grad_norm": Infinity,
-      "learning_rate": 9.098653352632625e-05,
-      "loss": 9.0343,
+      "learning_rate": 9.027874890166069e-05,
+      "loss": 9.1291,
       "step": 300
     },
     {
-      "epoch": 0.12852928782132322,
+      "epoch": 0.2569502948609941,
       "grad_norm": Infinity,
-      "learning_rate": 9.097758970767663e-05,
-      "loss": 9.0571,
+      "learning_rate": 9.024290311967066e-05,
+      "loss": 9.1037,
       "step": 305
     },
     {
-      "epoch": 0.1306363253265908,
+      "epoch": 0.2611625947767481,
       "grad_norm": Infinity,
-      "learning_rate": 9.096847889169301e-05,
-      "loss": 9.1159,
+      "learning_rate": 9.020639785377019e-05,
+      "loss": 9.0707,
       "step": 310
     },
     {
-      "epoch": 0.13274336283185842,
+      "epoch": 0.2653748946925021,
       "grad_norm": Infinity,
-      "learning_rate": 9.09592011231715e-05,
-      "loss": 9.3278,
+      "learning_rate": 9.01692338284451e-05,
+      "loss": 9.2199,
       "step": 315
     },
     {
-      "epoch": 0.134850400337126,
+      "epoch": 0.2695871946082561,
       "grad_norm": Infinity,
-      "learning_rate": 9.0949756447729e-05,
-      "loss": 9.049,
+      "learning_rate": 9.013141178125513e-05,
+      "loss": 9.2428,
       "step": 320
     },
     {
-      "epoch": 0.1369574378423936,
+      "epoch": 0.2737994945240101,
       "grad_norm": Infinity,
-      "learning_rate": 9.094014491180309e-05,
-      "loss": 9.053,
+      "learning_rate": 9.009293246281905e-05,
+      "loss": 9.163,
       "step": 325
     },
     {
-      "epoch": 0.1390644753476612,
+      "epoch": 0.2780117944397641,
       "grad_norm": Infinity,
-      "learning_rate": 9.093036656265176e-05,
-      "loss": 9.1193,
+      "learning_rate": 9.005379663679996e-05,
+      "loss": 9.22,
       "step": 330
     },
     {
-      "epoch": 0.14117151285292878,
+      "epoch": 0.2822240943555181,
       "grad_norm": Infinity,
-      "learning_rate": 9.092042144835316e-05,
-      "loss": 9.0794,
+      "learning_rate": 9.001400507989004e-05,
+      "loss": 9.0213,
       "step": 335
     },
     {
-      "epoch": 0.14327855035819637,
+      "epoch": 0.2864363942712721,
       "grad_norm": Infinity,
-      "learning_rate": 9.091030961780546e-05,
-      "loss": 9.0032,
+      "learning_rate": 8.99735585817951e-05,
+      "loss": 9.1937,
       "step": 340
     },
     {
-      "epoch": 0.14538558786346398,
+      "epoch": 0.2906486941870261,
       "grad_norm": Infinity,
-      "learning_rate": 9.090003112072649e-05,
-      "loss": 9.0861,
+      "learning_rate": 8.993245794521902e-05,
+      "loss": 9.1865,
       "step": 345
     },
     {
-      "epoch": 0.14749262536873156,
+      "epoch": 0.2948609941027801,
       "grad_norm": Infinity,
-      "learning_rate": 9.088958600765355e-05,
-      "loss": 8.9942,
+      "learning_rate": 8.98907039858477e-05,
+      "loss": 9.2424,
       "step": 350
     },
     {
-      "epoch": 0.14959966287399915,
+      "epoch": 0.2990732940185341,
       "grad_norm": Infinity,
-      "learning_rate": 9.087897432994316e-05,
-      "loss": 8.968,
+      "learning_rate": 8.984829753233298e-05,
+      "loss": 9.1797,
       "step": 355
     },
     {
-      "epoch": 0.15170670037926676,
+      "epoch": 0.3032855939342881,
       "grad_norm": Infinity,
-      "learning_rate": 9.086819613977083e-05,
-      "loss": 9.1462,
+      "learning_rate": 8.980523942627609e-05,
+      "loss": 9.127,
       "step": 360
     },
     {
-      "epoch": 0.15381373788453434,
+      "epoch": 0.3074978938500421,
       "grad_norm": Infinity,
-      "learning_rate": 9.085725149013074e-05,
-      "loss": 8.8035,
+      "learning_rate": 8.976153052221104e-05,
+      "loss": 9.0876,
       "step": 365
     },
     {
-      "epoch": 0.15592077538980195,
+      "epoch": 0.3117101937657961,
       "grad_norm": Infinity,
-      "learning_rate": 9.08461404348355e-05,
-      "loss": 9.0213,
+      "learning_rate": 8.971717168758756e-05,
+      "loss": 9.1596,
       "step": 370
     },
     {
-      "epoch": 0.15802781289506954,
+      "epoch": 0.3159224936815501,
       "grad_norm": Infinity,
-      "learning_rate": 9.083486302851602e-05,
-      "loss": 9.1359,
+      "learning_rate": 8.967216380275405e-05,
+      "loss": 9.048,
       "step": 375
     },
     {
-      "epoch": 0.16013485040033712,
+      "epoch": 0.3201347935973041,
       "grad_norm": Infinity,
-      "learning_rate": 9.082341932662096e-05,
-      "loss": 9.1997,
+      "learning_rate": 8.962650776093989e-05,
+      "loss": 9.0332,
       "step": 380
     },
     {
-      "epoch": 0.16224188790560473,
+      "epoch": 0.3243470935130581,
       "grad_norm": Infinity,
-      "learning_rate": 9.081180938541676e-05,
-      "loss": 8.9983,
+      "learning_rate": 8.958020446823789e-05,
+      "loss": 9.1551,
       "step": 385
     },
     {
-      "epoch": 0.16434892541087232,
+      "epoch": 0.32855939342881213,
       "grad_norm": Infinity,
-      "learning_rate": 9.080003326198714e-05,
-      "loss": 9.2442,
+      "learning_rate": 8.953325484358625e-05,
+      "loss": 8.9415,
       "step": 390
     },
     {
-      "epoch": 0.1664559629161399,
+      "epoch": 0.33277169334456613,
       "grad_norm": Infinity,
-      "learning_rate": 9.078809101423295e-05,
-      "loss": 9.283,
+      "learning_rate": 8.948565981875027e-05,
+      "loss": 9.0024,
       "step": 395
     },
     {
-      "epoch": 0.1685630004214075,
+      "epoch": 0.33698399326032014,
       "grad_norm": Infinity,
-      "learning_rate": 9.07759827008718e-05,
-      "loss": 9.0873,
+      "learning_rate": 8.943742033830394e-05,
+      "loss": 9.1814,
       "step": 400
     },
     {
-      "epoch": 0.1706700379266751,
+      "epoch": 0.34119629317607414,
       "grad_norm": Infinity,
-      "learning_rate": 9.076370838143787e-05,
-      "loss": 9.1848,
+      "learning_rate": 8.938853735961113e-05,
+      "loss": 9.2051,
       "step": 405
     },
     {
-      "epoch": 0.17277707543194268,
+      "epoch": 0.34540859309182814,
       "grad_norm": Infinity,
-      "learning_rate": 9.07512681162815e-05,
-      "loss": 9.1581,
+      "learning_rate": 8.933901185280665e-05,
+      "loss": 9.1032,
       "step": 410
     },
     {
-      "epoch": 0.1748841129372103,
+      "epoch": 0.34962089300758215,
       "grad_norm": Infinity,
-      "learning_rate": 9.073866196656897e-05,
-      "loss": 9.3075,
+      "learning_rate": 8.928884480077696e-05,
+      "loss": 9.0838,
       "step": 415
+    },
+    {
+      "epoch": 0.35383319292333615,
+      "grad_norm": Infinity,
+      "learning_rate": 8.923803719914063e-05,
+      "loss": 9.0075,
+      "step": 420
+    },
+    {
+      "epoch": 0.35804549283909015,
+      "grad_norm": Infinity,
+      "learning_rate": 8.91865900562287e-05,
+      "loss": 8.9951,
+      "step": 425
+    },
+    {
+      "epoch": 0.36225779275484415,
+      "grad_norm": Infinity,
+      "learning_rate": 8.913450439306448e-05,
+      "loss": 9.1082,
+      "step": 430
+    },
+    {
+      "epoch": 0.36647009267059816,
+      "grad_norm": Infinity,
+      "learning_rate": 8.908178124334348e-05,
+      "loss": 9.2004,
+      "step": 435
+    },
+    {
+      "epoch": 0.37068239258635216,
+      "grad_norm": Infinity,
+      "learning_rate": 8.902842165341278e-05,
+      "loss": 8.9807,
+      "step": 440
+    },
+    {
+      "epoch": 0.37489469250210616,
+      "grad_norm": Infinity,
+      "learning_rate": 8.89744266822503e-05,
+      "loss": 9.043,
+      "step": 445
+    },
+    {
+      "epoch": 0.37910699241786017,
+      "grad_norm": Infinity,
+      "learning_rate": 8.891979740144376e-05,
+      "loss": 9.1759,
+      "step": 450
+    },
+    {
+      "epoch": 0.38331929233361417,
+      "grad_norm": Infinity,
+      "learning_rate": 8.886453489516945e-05,
+      "loss": 9.0236,
+      "step": 455
+    },
+    {
+      "epoch": 0.3875315922493682,
+      "grad_norm": Infinity,
+      "learning_rate": 8.880864026017068e-05,
+      "loss": 9.2918,
+      "step": 460
+    },
+    {
+      "epoch": 0.3917438921651222,
+      "grad_norm": Infinity,
+      "learning_rate": 8.875211460573607e-05,
+      "loss": 9.0641,
+      "step": 465
+    },
+    {
+      "epoch": 0.3959561920808762,
+      "grad_norm": Infinity,
+      "learning_rate": 8.869495905367742e-05,
+      "loss": 9.048,
+      "step": 470
+    },
+    {
+      "epoch": 0.4001684919966302,
+      "grad_norm": Infinity,
+      "learning_rate": 8.863717473830758e-05,
+      "loss": 9.0192,
+      "step": 475
+    },
+    {
+      "epoch": 0.4043807919123842,
+      "grad_norm": Infinity,
+      "learning_rate": 8.857876280641784e-05,
+      "loss": 9.0275,
+      "step": 480
+    },
+    {
+      "epoch": 0.4085930918281382,
+      "grad_norm": Infinity,
+      "learning_rate": 8.851972441725522e-05,
+      "loss": 8.9821,
+      "step": 485
+    },
+    {
+      "epoch": 0.4128053917438922,
+      "grad_norm": Infinity,
+      "learning_rate": 8.846006074249951e-05,
+      "loss": 8.9907,
+      "step": 490
+    },
+    {
+      "epoch": 0.41701769165964614,
+      "grad_norm": Infinity,
+      "learning_rate": 8.839977296623983e-05,
+      "loss": 9.0855,
+      "step": 495
+    },
+    {
+      "epoch": 0.42122999157540014,
+      "grad_norm": Infinity,
+      "learning_rate": 8.833886228495139e-05,
+      "loss": 9.2186,
+      "step": 500
+    },
+    {
+      "epoch": 0.42122999157540014,
+      "eval_loss": 9.067032814025879,
+      "eval_runtime": 2.0409,
+      "eval_samples_per_second": 26.949,
+      "eval_steps_per_second": 13.719,
+      "step": 500
     }
   ],
   "logging_steps": 5,
-  "max_steps": 7119,
+  "max_steps": 3561,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
   "save_steps": 500,
@@ -608,7 +735,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 9.199842228845937e+17,
+  "total_flos": 2.1956663091134464e+18,
   "train_batch_size": 48,
   "trial_name": null,
   "trial_params": null

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5d61848344d261f7ee72ed40e543705d4c9b47355806f2308c7db674796a9a95
+oid sha256:d048f01f6853330251af6fe32add0aa6a3f4efcb452d1313fc0b4287ea95d91e
 size 5624