AhmedZaky1 commited on
Commit
c46cc1c
·
verified ·
1 Parent(s): 3577f70

Training in progress, step 21500, checkpoint

Browse files
last-checkpoint/adapter_config.json CHANGED
@@ -29,12 +29,12 @@
29
  "rank_pattern": {},
30
  "revision": null,
31
  "target_modules": [
32
- "up_proj",
33
  "down_proj",
34
- "k_proj",
35
- "v_proj",
36
  "o_proj",
37
  "gate_proj",
 
 
 
38
  "q_proj"
39
  ],
40
  "target_parameters": null,
 
29
  "rank_pattern": {},
30
  "revision": null,
31
  "target_modules": [
 
32
  "down_proj",
 
 
33
  "o_proj",
34
  "gate_proj",
35
+ "v_proj",
36
+ "up_proj",
37
+ "k_proj",
38
  "q_proj"
39
  ],
40
  "target_parameters": null,
last-checkpoint/adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9bc73624e965dba3332b4388cc496cbb8aad1907f94a8719548c95cf4eb17852
3
  size 295488936
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:69ad14b251ea8c975c5d62c4711c05a374bb07e36a6992565f356138763fa2c2
3
  size 295488936
last-checkpoint/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b00f227f2a749f1f8548a96c419b40d3d8a2bf63270369e77147ddd68f16c6e3
3
  size 591209083
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5ec34cb33ad9fff8f8a52d0ad649519df64e481d60cbcf0bea75f53c621d30e
3
  size 591209083
last-checkpoint/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2ff17261e7b5e5b975d26b5586b146fae1f6df93e96d395e1220068624854c3c
3
  size 14645
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:771529242bdb07b1cb665e7e57920e3e30f761a3d29ac1dacd1d5590d33d1212
3
  size 14645
last-checkpoint/scaler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:05bed5f269e502f1dd5aaddf0027a51378cefd049d2e03956ce54ec7ac8f9d02
3
  size 1383
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:24fea3ad573eb47743e5ae20df84584202336b8d095b8a57b3c387cc05e44c5c
3
  size 1383
last-checkpoint/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:77ec69abd8c24b53e58f32cf52a02c1aa256d4078db48086ac9619fa1df22bec
3
  size 1465
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0b370f3929273b07b346764620dd20c69bf8d76aaf108df2420db3fad8e4501c
3
  size 1465
last-checkpoint/trainer_state.json CHANGED
@@ -2,9 +2,9 @@
2
  "best_global_step": null,
3
  "best_metric": null,
4
  "best_model_checkpoint": null,
5
- "epoch": 2.2191896640553788,
6
  "eval_steps": 500,
7
- "global_step": 21000,
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
@@ -15044,6 +15044,364 @@
15044
  "eval_samples_per_second": 2.702,
15045
  "eval_steps_per_second": 2.702,
15046
  "step": 21000
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
15047
  }
15048
  ],
15049
  "logging_steps": 10,
@@ -15063,7 +15421,7 @@
15063
  "attributes": {}
15064
  }
15065
  },
15066
- "total_flos": 1.1339819789302579e+18,
15067
  "train_batch_size": 1,
15068
  "trial_name": null,
15069
  "trial_params": null
 
2
  "best_global_step": null,
3
  "best_metric": null,
4
  "best_model_checkpoint": null,
5
+ "epoch": 2.27203191672061,
6
  "eval_steps": 500,
7
+ "global_step": 21500,
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
 
15044
  "eval_samples_per_second": 2.702,
15045
  "eval_steps_per_second": 2.702,
15046
  "step": 21000
15047
+ },
15048
+ {
15049
+ "epoch": 2.220246509108683,
15050
+ "grad_norm": 1.6857826709747314,
15051
+ "learning_rate": 7.894517793812078e-06,
15052
+ "loss": 0.8496136665344238,
15053
+ "step": 21010
15054
+ },
15055
+ {
15056
+ "epoch": 2.221303354161988,
15057
+ "grad_norm": 1.4029793739318848,
15058
+ "learning_rate": 7.874338209672507e-06,
15059
+ "loss": 0.8891298294067382,
15060
+ "step": 21020
15061
+ },
15062
+ {
15063
+ "epoch": 2.2223601992152924,
15064
+ "grad_norm": 1.215582251548767,
15065
+ "learning_rate": 7.854179627453389e-06,
15066
+ "loss": 0.8832895278930664,
15067
+ "step": 21030
15068
+ },
15069
+ {
15070
+ "epoch": 2.223417044268597,
15071
+ "grad_norm": 1.3796038627624512,
15072
+ "learning_rate": 7.834042071876063e-06,
15073
+ "loss": 0.8879108428955078,
15074
+ "step": 21040
15075
+ },
15076
+ {
15077
+ "epoch": 2.2244738893219016,
15078
+ "grad_norm": 1.357569694519043,
15079
+ "learning_rate": 7.813925567636051e-06,
15080
+ "loss": 0.8945812225341797,
15081
+ "step": 21050
15082
+ },
15083
+ {
15084
+ "epoch": 2.2255307343752064,
15085
+ "grad_norm": 1.4801578521728516,
15086
+ "learning_rate": 7.793830139403072e-06,
15087
+ "loss": 0.9161118507385254,
15088
+ "step": 21060
15089
+ },
15090
+ {
15091
+ "epoch": 2.226587579428511,
15092
+ "grad_norm": 1.2665520906448364,
15093
+ "learning_rate": 7.773755811820985e-06,
15094
+ "loss": 0.8399243354797363,
15095
+ "step": 21070
15096
+ },
15097
+ {
15098
+ "epoch": 2.2276444244818157,
15099
+ "grad_norm": 1.1792436838150024,
15100
+ "learning_rate": 7.753702609507801e-06,
15101
+ "loss": 0.8259225845336914,
15102
+ "step": 21080
15103
+ },
15104
+ {
15105
+ "epoch": 2.2287012695351205,
15106
+ "grad_norm": 1.274572730064392,
15107
+ "learning_rate": 7.7336705570556e-06,
15108
+ "loss": 0.8148788452148438,
15109
+ "step": 21090
15110
+ },
15111
+ {
15112
+ "epoch": 2.229758114588425,
15113
+ "grad_norm": 1.6488457918167114,
15114
+ "learning_rate": 7.713659679030527e-06,
15115
+ "loss": 0.8844100952148437,
15116
+ "step": 21100
15117
+ },
15118
+ {
15119
+ "epoch": 2.2308149596417297,
15120
+ "grad_norm": 1.4412405490875244,
15121
+ "learning_rate": 7.693669999972763e-06,
15122
+ "loss": 0.8640574455261231,
15123
+ "step": 21110
15124
+ },
15125
+ {
15126
+ "epoch": 2.231871804695034,
15127
+ "grad_norm": 1.5173754692077637,
15128
+ "learning_rate": 7.673701544396508e-06,
15129
+ "loss": 0.8501100540161133,
15130
+ "step": 21120
15131
+ },
15132
+ {
15133
+ "epoch": 2.232928649748339,
15134
+ "grad_norm": 1.2773516178131104,
15135
+ "learning_rate": 7.653754336789909e-06,
15136
+ "loss": 0.900639533996582,
15137
+ "step": 21130
15138
+ },
15139
+ {
15140
+ "epoch": 2.2339854948016433,
15141
+ "grad_norm": 1.5317981243133545,
15142
+ "learning_rate": 7.633828401615071e-06,
15143
+ "loss": 0.9276094436645508,
15144
+ "step": 21140
15145
+ },
15146
+ {
15147
+ "epoch": 2.235042339854948,
15148
+ "grad_norm": 1.3817973136901855,
15149
+ "learning_rate": 7.6139237633079956e-06,
15150
+ "loss": 0.800080394744873,
15151
+ "step": 21150
15152
+ },
15153
+ {
15154
+ "epoch": 2.2360991849082525,
15155
+ "grad_norm": 1.4337248802185059,
15156
+ "learning_rate": 7.594040446278597e-06,
15157
+ "loss": 0.961793041229248,
15158
+ "step": 21160
15159
+ },
15160
+ {
15161
+ "epoch": 2.2371560299615574,
15162
+ "grad_norm": 1.4675263166427612,
15163
+ "learning_rate": 7.5741784749106145e-06,
15164
+ "loss": 0.8298232078552246,
15165
+ "step": 21170
15166
+ },
15167
+ {
15168
+ "epoch": 2.2382128750148618,
15169
+ "grad_norm": 1.1299012899398804,
15170
+ "learning_rate": 7.554337873561626e-06,
15171
+ "loss": 0.9074227333068847,
15172
+ "step": 21180
15173
+ },
15174
+ {
15175
+ "epoch": 2.2392697200681666,
15176
+ "grad_norm": 1.122652292251587,
15177
+ "learning_rate": 7.5345186665629925e-06,
15178
+ "loss": 0.9688516616821289,
15179
+ "step": 21190
15180
+ },
15181
+ {
15182
+ "epoch": 2.240326565121471,
15183
+ "grad_norm": 1.3946301937103271,
15184
+ "learning_rate": 7.514720878219833e-06,
15185
+ "loss": 0.8214277267456055,
15186
+ "step": 21200
15187
+ },
15188
+ {
15189
+ "epoch": 2.241383410174776,
15190
+ "grad_norm": 2.1314175128936768,
15191
+ "learning_rate": 7.494944532811038e-06,
15192
+ "loss": 0.9431739807128906,
15193
+ "step": 21210
15194
+ },
15195
+ {
15196
+ "epoch": 2.24244025522808,
15197
+ "grad_norm": 1.3471664190292358,
15198
+ "learning_rate": 7.475189654589138e-06,
15199
+ "loss": 0.7802850246429444,
15200
+ "step": 21220
15201
+ },
15202
+ {
15203
+ "epoch": 2.243497100281385,
15204
+ "grad_norm": 1.7846940755844116,
15205
+ "learning_rate": 7.455456267780395e-06,
15206
+ "loss": 0.8822522163391113,
15207
+ "step": 21230
15208
+ },
15209
+ {
15210
+ "epoch": 2.2445539453346894,
15211
+ "grad_norm": 1.3426530361175537,
15212
+ "learning_rate": 7.435744396584679e-06,
15213
+ "loss": 0.8295673370361328,
15214
+ "step": 21240
15215
+ },
15216
+ {
15217
+ "epoch": 2.2456107903879943,
15218
+ "grad_norm": 1.3698786497116089,
15219
+ "learning_rate": 7.416054065175513e-06,
15220
+ "loss": 0.815796184539795,
15221
+ "step": 21250
15222
+ },
15223
+ {
15224
+ "epoch": 2.2466676354412987,
15225
+ "grad_norm": 1.4198206663131714,
15226
+ "learning_rate": 7.396385297699943e-06,
15227
+ "loss": 0.9136198043823243,
15228
+ "step": 21260
15229
+ },
15230
+ {
15231
+ "epoch": 2.2477244804946035,
15232
+ "grad_norm": 0.9479147791862488,
15233
+ "learning_rate": 7.376738118278634e-06,
15234
+ "loss": 0.9125507354736329,
15235
+ "step": 21270
15236
+ },
15237
+ {
15238
+ "epoch": 2.248781325547908,
15239
+ "grad_norm": 1.5470222234725952,
15240
+ "learning_rate": 7.357112551005735e-06,
15241
+ "loss": 0.8383374214172363,
15242
+ "step": 21280
15243
+ },
15244
+ {
15245
+ "epoch": 2.2498381706012127,
15246
+ "grad_norm": 1.3305402994155884,
15247
+ "learning_rate": 7.33750861994893e-06,
15248
+ "loss": 0.8419397354125977,
15249
+ "step": 21290
15250
+ },
15251
+ {
15252
+ "epoch": 2.2508950156545176,
15253
+ "grad_norm": 1.2185657024383545,
15254
+ "learning_rate": 7.3179263491493095e-06,
15255
+ "loss": 0.8711481094360352,
15256
+ "step": 21300
15257
+ },
15258
+ {
15259
+ "epoch": 2.251951860707822,
15260
+ "grad_norm": 1.182152271270752,
15261
+ "learning_rate": 7.298365762621464e-06,
15262
+ "loss": 0.8621840476989746,
15263
+ "step": 21310
15264
+ },
15265
+ {
15266
+ "epoch": 2.2530087057611268,
15267
+ "grad_norm": 1.4320274591445923,
15268
+ "learning_rate": 7.278826884353351e-06,
15269
+ "loss": 0.8451634407043457,
15270
+ "step": 21320
15271
+ },
15272
+ {
15273
+ "epoch": 2.254065550814431,
15274
+ "grad_norm": 1.0900665521621704,
15275
+ "learning_rate": 7.2593097383063384e-06,
15276
+ "loss": 0.8664793968200684,
15277
+ "step": 21330
15278
+ },
15279
+ {
15280
+ "epoch": 2.255122395867736,
15281
+ "grad_norm": 0.867480456829071,
15282
+ "learning_rate": 7.2398143484150975e-06,
15283
+ "loss": 0.8150432586669922,
15284
+ "step": 21340
15285
+ },
15286
+ {
15287
+ "epoch": 2.2561792409210404,
15288
+ "grad_norm": 1.2458523511886597,
15289
+ "learning_rate": 7.220340738587669e-06,
15290
+ "loss": 0.9331179618835449,
15291
+ "step": 21350
15292
+ },
15293
+ {
15294
+ "epoch": 2.257236085974345,
15295
+ "grad_norm": 1.2626800537109375,
15296
+ "learning_rate": 7.200888932705344e-06,
15297
+ "loss": 0.8389029502868652,
15298
+ "step": 21360
15299
+ },
15300
+ {
15301
+ "epoch": 2.2582929310276496,
15302
+ "grad_norm": 1.0448240041732788,
15303
+ "learning_rate": 7.181458954622714e-06,
15304
+ "loss": 0.934419059753418,
15305
+ "step": 21370
15306
+ },
15307
+ {
15308
+ "epoch": 2.2593497760809544,
15309
+ "grad_norm": 1.4713551998138428,
15310
+ "learning_rate": 7.162050828167549e-06,
15311
+ "loss": 0.871403980255127,
15312
+ "step": 21380
15313
+ },
15314
+ {
15315
+ "epoch": 2.260406621134259,
15316
+ "grad_norm": 1.7052724361419678,
15317
+ "learning_rate": 7.142664577140876e-06,
15318
+ "loss": 0.9119677543640137,
15319
+ "step": 21390
15320
+ },
15321
+ {
15322
+ "epoch": 2.2614634661875637,
15323
+ "grad_norm": 1.777775764465332,
15324
+ "learning_rate": 7.123300225316859e-06,
15325
+ "loss": 0.9340780258178711,
15326
+ "step": 21400
15327
+ },
15328
+ {
15329
+ "epoch": 2.262520311240868,
15330
+ "grad_norm": 1.014190673828125,
15331
+ "learning_rate": 7.103957796442834e-06,
15332
+ "loss": 0.7185611248016357,
15333
+ "step": 21410
15334
+ },
15335
+ {
15336
+ "epoch": 2.263577156294173,
15337
+ "grad_norm": 1.2191739082336426,
15338
+ "learning_rate": 7.084637314239212e-06,
15339
+ "loss": 0.9621832847595215,
15340
+ "step": 21420
15341
+ },
15342
+ {
15343
+ "epoch": 2.2646340013474773,
15344
+ "grad_norm": 1.7386350631713867,
15345
+ "learning_rate": 7.065338802399538e-06,
15346
+ "loss": 0.8671950340270996,
15347
+ "step": 21430
15348
+ },
15349
+ {
15350
+ "epoch": 2.265690846400782,
15351
+ "grad_norm": 0.9992101192474365,
15352
+ "learning_rate": 7.0460622845903686e-06,
15353
+ "loss": 0.811929988861084,
15354
+ "step": 21440
15355
+ },
15356
+ {
15357
+ "epoch": 2.2667476914540865,
15358
+ "grad_norm": 1.4000431299209595,
15359
+ "learning_rate": 7.0268077844513335e-06,
15360
+ "loss": 0.8797445297241211,
15361
+ "step": 21450
15362
+ },
15363
+ {
15364
+ "epoch": 2.2678045365073913,
15365
+ "grad_norm": 1.0444756746292114,
15366
+ "learning_rate": 7.007575325595012e-06,
15367
+ "loss": 0.840949535369873,
15368
+ "step": 21460
15369
+ },
15370
+ {
15371
+ "epoch": 2.268861381560696,
15372
+ "grad_norm": 1.2659364938735962,
15373
+ "learning_rate": 6.988364931607e-06,
15374
+ "loss": 0.878659439086914,
15375
+ "step": 21470
15376
+ },
15377
+ {
15378
+ "epoch": 2.2699182266140006,
15379
+ "grad_norm": 1.2796216011047363,
15380
+ "learning_rate": 6.969176626045792e-06,
15381
+ "loss": 0.8297262191772461,
15382
+ "step": 21480
15383
+ },
15384
+ {
15385
+ "epoch": 2.270975071667305,
15386
+ "grad_norm": 0.9651225805282593,
15387
+ "learning_rate": 6.950010432442844e-06,
15388
+ "loss": 0.7489469051361084,
15389
+ "step": 21490
15390
+ },
15391
+ {
15392
+ "epoch": 2.27203191672061,
15393
+ "grad_norm": 1.1623140573501587,
15394
+ "learning_rate": 6.930866374302436e-06,
15395
+ "loss": 0.8541905403137207,
15396
+ "step": 21500
15397
+ },
15398
+ {
15399
+ "epoch": 2.27203191672061,
15400
+ "eval_loss": 0.4870993494987488,
15401
+ "eval_runtime": 3511.8057,
15402
+ "eval_samples_per_second": 2.536,
15403
+ "eval_steps_per_second": 2.536,
15404
+ "step": 21500
15405
  }
15406
  ],
15407
  "logging_steps": 10,
 
15421
  "attributes": {}
15422
  }
15423
  },
15424
+ "total_flos": 1.161339301405778e+18,
15425
  "train_batch_size": 1,
15426
  "trial_name": null,
15427
  "trial_params": null