End of training

README.md

@@ -16,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # qwen3-1.7b-dabstep-reasoning-30-fixed-reasoning-sharegpt-sft
 
-This model is a fine-tuned version of [Qwen/Qwen3-1.7B](https://huggingface.co/Qwen/Qwen3-1.7B) on an unknown dataset.
+This model is a fine-tuned version of [Qwen/Qwen3-1.7B](https://huggingface.co/Qwen/Qwen3-1.7B) on the eval-ds-dabstep-reasoning-30-fixed-reasoning-sharegpt dataset.
 
 ## Model description
 

all_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 5.0,
     "total_flos": 213154529280.0,
-    "train_loss": 0.7240357622504234,
-    "train_runtime": 218.9144,
-    "train_samples_per_second": 0.64,
-    "train_steps_per_second": 0.32
+    "train_loss": 0.7237587083663259,
+    "train_runtime": 212.7239,
+    "train_samples_per_second": 0.658,
+    "train_steps_per_second": 0.329
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 5.0,
     "total_flos": 213154529280.0,
-    "train_loss": 0.7240357622504234,
-    "train_runtime": 218.9144,
-    "train_samples_per_second": 0.64,
-    "train_steps_per_second": 0.32
+    "train_loss": 0.7237587083663259,
+    "train_runtime": 212.7239,
+    "train_samples_per_second": 0.658,
+    "train_steps_per_second": 0.329
 }

trainer_state.json

@@ -11,502 +11,502 @@
   "log_history": [
     {
       "epoch": 0.07142857142857142,
-      "grad_norm": 24.021249098223716,
+      "grad_norm": 24.023632843584195,
       "learning_rate": 0.0,
       "loss": 1.9336,
       "step": 1
     },
     {
       "epoch": 0.14285714285714285,
-      "grad_norm": 25.283596707859346,
+      "grad_norm": 25.27216550339971,
       "learning_rate": 1e-05,
       "loss": 1.4638,
       "step": 2
     },
     {
       "epoch": 0.21428571428571427,
-      "grad_norm": 13.753980587901806,
+      "grad_norm": 13.768008441403769,
       "learning_rate": 9.994818368233639e-06,
-      "loss": 1.3625,
+      "loss": 1.3641,
       "step": 3
     },
     {
       "epoch": 0.2857142857142857,
-      "grad_norm": 9.145524344533044,
+      "grad_norm": 9.119394933221514,
       "learning_rate": 9.979284212657658e-06,
-      "loss": 1.6794,
+      "loss": 1.6788,
       "step": 4
     },
     {
       "epoch": 0.35714285714285715,
-      "grad_norm": 8.35400132880214,
+      "grad_norm": 8.29180049726504,
       "learning_rate": 9.953429730181653e-06,
-      "loss": 1.4699,
+      "loss": 1.471,
       "step": 5
     },
     {
       "epoch": 0.42857142857142855,
-      "grad_norm": 6.7101769586899795,
+      "grad_norm": 6.6755630111422075,
       "learning_rate": 9.917308508168712e-06,
-      "loss": 1.5353,
+      "loss": 1.535,
       "step": 6
     },
     {
       "epoch": 0.5,
-      "grad_norm": 5.988822692253111,
+      "grad_norm": 5.927824954855512,
       "learning_rate": 9.870995413367397e-06,
-      "loss": 1.0191,
+      "loss": 1.0182,
       "step": 7
     },
     {
       "epoch": 0.5714285714285714,
-      "grad_norm": 5.309562103895213,
+      "grad_norm": 5.290712655595048,
       "learning_rate": 9.814586436738998e-06,
-      "loss": 1.2792,
+      "loss": 1.2797,
       "step": 8
     },
     {
       "epoch": 0.6428571428571429,
-      "grad_norm": 4.420439423024025,
+      "grad_norm": 4.450390639155862,
       "learning_rate": 9.748198494501598e-06,
-      "loss": 1.1122,
+      "loss": 1.1114,
       "step": 9
     },
     {
       "epoch": 0.7142857142857143,
-      "grad_norm": 4.691101929329496,
+      "grad_norm": 5.552257885415831,
       "learning_rate": 9.671969185803357e-06,
-      "loss": 1.1426,
+      "loss": 1.1423,
       "step": 10
     },
     {
       "epoch": 0.7857142857142857,
-      "grad_norm": 5.321128291286354,
+      "grad_norm": 5.335273326587211,
       "learning_rate": 9.586056507527266e-06,
-      "loss": 1.1301,
+      "loss": 1.1309,
       "step": 11
     },
     {
       "epoch": 0.8571428571428571,
-      "grad_norm": 5.041563930361967,
+      "grad_norm": 5.055921586049105,
       "learning_rate": 9.490638526818482e-06,
-      "loss": 1.324,
+      "loss": 1.3257,
       "step": 12
     },
     {
       "epoch": 0.9285714285714286,
-      "grad_norm": 4.670789707567545,
+      "grad_norm": 4.759961808264356,
       "learning_rate": 9.385913012012972e-06,
-      "loss": 1.1864,
+      "loss": 1.1885,
       "step": 13
     },
     {
       "epoch": 1.0,
-      "grad_norm": 5.0546136864842435,
+      "grad_norm": 4.871340975018161,
       "learning_rate": 9.272097022732444e-06,
       "loss": 1.2715,
       "step": 14
     },
     {
       "epoch": 1.0714285714285714,
-      "grad_norm": 4.47229822194355,
+      "grad_norm": 4.436955356609461,
       "learning_rate": 9.149426459995127e-06,
-      "loss": 1.2285,
+      "loss": 1.2302,
       "step": 15
     },
     {
       "epoch": 1.1428571428571428,
-      "grad_norm": 4.345645421364438,
+      "grad_norm": 4.28052498090576,
       "learning_rate": 9.018155577274891e-06,
-      "loss": 1.0186,
+      "loss": 1.02,
       "step": 16
     },
     {
       "epoch": 1.2142857142857142,
-      "grad_norm": 5.112930979741301,
+      "grad_norm": 5.067070847952673,
       "learning_rate": 8.8785564535221e-06,
-      "loss": 1.0241,
+      "loss": 1.0235,
       "step": 17
     },
     {
       "epoch": 1.2857142857142856,
-      "grad_norm": 4.1178820420377695,
+      "grad_norm": 4.087626184275003,
       "learning_rate": 8.730918429238429e-06,
-      "loss": 0.9642,
+      "loss": 0.9651,
       "step": 18
     },
     {
       "epoch": 1.3571428571428572,
-      "grad_norm": 4.337610285913757,
+      "grad_norm": 4.301072973315502,
       "learning_rate": 8.575547506774498e-06,
-      "loss": 0.9862,
+      "loss": 0.995,
       "step": 19
     },
     {
       "epoch": 1.4285714285714286,
-      "grad_norm": 4.059264208551376,
+      "grad_norm": 4.063688761012534,
       "learning_rate": 8.412765716093273e-06,
-      "loss": 0.9026,
+      "loss": 0.9102,
       "step": 20
     },
     {
       "epoch": 1.5,
-      "grad_norm": 4.641502887388555,
+      "grad_norm": 4.646743739276943,
       "learning_rate": 8.24291044731378e-06,
-      "loss": 0.8315,
+      "loss": 0.8329,
       "step": 21
     },
     {
       "epoch": 1.5714285714285714,
-      "grad_norm": 3.8478823458692903,
+      "grad_norm": 3.845579762800239,
       "learning_rate": 8.066333751418582e-06,
-      "loss": 0.7903,
+      "loss": 0.795,
       "step": 22
     },
     {
       "epoch": 1.6428571428571428,
-      "grad_norm": 4.163156027393506,
+      "grad_norm": 4.130688938229113,
       "learning_rate": 7.883401610574338e-06,
-      "loss": 0.8801,
+      "loss": 0.8774,
       "step": 23
     },
     {
       "epoch": 1.7142857142857144,
-      "grad_norm": 4.580115671438542,
+      "grad_norm": 4.444662774507303,
       "learning_rate": 7.69449317957788e-06,
-      "loss": 0.5991,
+      "loss": 0.5987,
       "step": 24
     },
     {
       "epoch": 1.7857142857142856,
-      "grad_norm": 5.19407596275134,
+      "grad_norm": 5.247198746213424,
       "learning_rate": 7.500000000000001e-06,
-      "loss": 0.6789,
+      "loss": 0.678,
       "step": 25
     },
     {
       "epoch": 1.8571428571428572,
-      "grad_norm": 4.665522623386005,
+      "grad_norm": 4.633300703755712,
       "learning_rate": 7.300325188655762e-06,
-      "loss": 0.8996,
+      "loss": 0.8991,
       "step": 26
     },
     {
       "epoch": 1.9285714285714286,
-      "grad_norm": 4.489581664483281,
+      "grad_norm": 4.341489445734741,
       "learning_rate": 7.095882602083321e-06,
-      "loss": 0.8013,
+      "loss": 0.802,
       "step": 27
     },
     {
       "epoch": 2.0,
-      "grad_norm": 3.972150256755104,
+      "grad_norm": 3.890840356813419,
       "learning_rate": 6.887095978763072e-06,
-      "loss": 0.7484,
+      "loss": 0.7479,
       "step": 28
     },
     {
       "epoch": 2.0714285714285716,
-      "grad_norm": 4.002751253836439,
+      "grad_norm": 3.953138214805454,
       "learning_rate": 6.674398060854931e-06,
-      "loss": 0.7164,
+      "loss": 0.7145,
       "step": 29
     },
     {
       "epoch": 2.142857142857143,
-      "grad_norm": 4.599302497344593,
+      "grad_norm": 4.246093007674121,
       "learning_rate": 6.458229697274125e-06,
-      "loss": 0.6523,
+      "loss": 0.6478,
       "step": 30
     },
     {
       "epoch": 2.2142857142857144,
-      "grad_norm": 4.459186000164494,
+      "grad_norm": 4.235637857831103,
       "learning_rate": 6.2390389299645e-06,
-      "loss": 0.7149,
+      "loss": 0.7132,
       "step": 31
     },
     {
       "epoch": 2.2857142857142856,
-      "grad_norm": 3.7742151881381254,
+      "grad_norm": 3.795047115132234,
       "learning_rate": 6.0172800652631706e-06,
       "loss": 0.5269,
       "step": 32
     },
     {
       "epoch": 2.357142857142857,
-      "grad_norm": 3.8044314857438533,
+      "grad_norm": 3.708500518296535,
       "learning_rate": 5.793412732281258e-06,
-      "loss": 0.6025,
+      "loss": 0.6023,
       "step": 33
     },
     {
       "epoch": 2.4285714285714284,
-      "grad_norm": 3.4365610343142423,
+      "grad_norm": 3.3778342286451735,
       "learning_rate": 5.567900930252375e-06,
-      "loss": 0.4838,
+      "loss": 0.4804,
       "step": 34
     },
     {
       "epoch": 2.5,
-      "grad_norm": 3.2175580299376634,
+      "grad_norm": 3.2763514955628454,
       "learning_rate": 5.341212066823356e-06,
-      "loss": 0.3622,
+      "loss": 0.3588,
       "step": 35
     },
     {
       "epoch": 2.571428571428571,
-      "grad_norm": 3.7968714193101976,
+      "grad_norm": 3.809622648646623,
       "learning_rate": 5.113815989280528e-06,
-      "loss": 0.5398,
+      "loss": 0.5365,
       "step": 36
     },
     {
       "epoch": 2.642857142857143,
-      "grad_norm": 4.874899187806918,
+      "grad_norm": 4.905824079180291,
       "learning_rate": 4.886184010719472e-06,
-      "loss": 0.7112,
+      "loss": 0.7041,
       "step": 37
     },
     {
       "epoch": 2.7142857142857144,
-      "grad_norm": 4.270728932769242,
+      "grad_norm": 4.331352426673357,
       "learning_rate": 4.6587879331766465e-06,
-      "loss": 0.6873,
+      "loss": 0.6861,
       "step": 38
     },
     {
       "epoch": 2.7857142857142856,
-      "grad_norm": 4.147689478796016,
+      "grad_norm": 4.268031298589448,
       "learning_rate": 4.432099069747625e-06,
-      "loss": 0.5413,
+      "loss": 0.538,
       "step": 39
     },
     {
       "epoch": 2.857142857142857,
-      "grad_norm": 4.501855975090803,
+      "grad_norm": 4.506172331148789,
       "learning_rate": 4.206587267718743e-06,
-      "loss": 0.6436,
+      "loss": 0.6443,
       "step": 40
     },
     {
       "epoch": 2.928571428571429,
-      "grad_norm": 3.873126903923081,
+      "grad_norm": 3.9399153380545116,
       "learning_rate": 3.982719934736832e-06,
-      "loss": 0.4471,
+      "loss": 0.461,
       "step": 41
     },
     {
       "epoch": 3.0,
-      "grad_norm": 4.522514310496286,
+      "grad_norm": 4.553905357043457,
       "learning_rate": 3.7609610700355014e-06,
-      "loss": 0.6648,
+      "loss": 0.6665,
       "step": 42
     },
     {
       "epoch": 3.0714285714285716,
-      "grad_norm": 3.52190753165053,
+      "grad_norm": 3.4967459109023253,
       "learning_rate": 3.5417703027258752e-06,
-      "loss": 0.4666,
+      "loss": 0.4645,
       "step": 43
     },
     {
       "epoch": 3.142857142857143,
-      "grad_norm": 3.481080222258227,
+      "grad_norm": 3.460264221490188,
       "learning_rate": 3.3256019391450696e-06,
-      "loss": 0.3418,
+      "loss": 0.3431,
       "step": 44
     },
     {
       "epoch": 3.2142857142857144,
-      "grad_norm": 3.763998470727338,
+      "grad_norm": 3.8276195456268285,
       "learning_rate": 3.1129040212369286e-06,
-      "loss": 0.5798,
+      "loss": 0.5799,
       "step": 45
     },
     {
       "epoch": 3.2857142857142856,
-      "grad_norm": 4.273847204368766,
+      "grad_norm": 4.281258985773756,
       "learning_rate": 2.9041173979166813e-06,
-      "loss": 0.51,
+      "loss": 0.5118,
       "step": 46
     },
     {
       "epoch": 3.357142857142857,
-      "grad_norm": 4.114567072348475,
+      "grad_norm": 4.006912491344759,
       "learning_rate": 2.6996748113442397e-06,
-      "loss": 0.4798,
+      "loss": 0.4746,
       "step": 47
     },
     {
       "epoch": 3.4285714285714284,
-      "grad_norm": 3.9147546587215465,
+      "grad_norm": 3.7962255242435337,
       "learning_rate": 2.5000000000000015e-06,
-      "loss": 0.446,
+      "loss": 0.4415,
       "step": 48
     },
     {
       "epoch": 3.5,
-      "grad_norm": 3.909860425388475,
+      "grad_norm": 3.93529721260128,
       "learning_rate": 2.3055068204221226e-06,
-      "loss": 0.372,
+      "loss": 0.3722,
       "step": 49
     },
     {
       "epoch": 3.571428571428571,
-      "grad_norm": 4.276313784444758,
+      "grad_norm": 4.201031566755138,
       "learning_rate": 2.1165983894256647e-06,
-      "loss": 0.5271,
+      "loss": 0.5201,
       "step": 50
     },
     {
       "epoch": 3.642857142857143,
-      "grad_norm": 3.920809797593642,
+      "grad_norm": 3.8906161720286927,
       "learning_rate": 1.933666248581418e-06,
-      "loss": 0.3067,
+      "loss": 0.3054,
       "step": 51
     },
     {
       "epoch": 3.7142857142857144,
-      "grad_norm": 4.019497689567704,
+      "grad_norm": 4.02310526884226,
       "learning_rate": 1.7570895526862202e-06,
-      "loss": 0.4295,
+      "loss": 0.4316,
       "step": 52
     },
     {
       "epoch": 3.7857142857142856,
-      "grad_norm": 3.4710315291966953,
+      "grad_norm": 3.3963029194169048,
       "learning_rate": 1.5872342839067305e-06,
-      "loss": 0.3069,
+      "loss": 0.3126,
       "step": 53
     },
     {
       "epoch": 3.857142857142857,
-      "grad_norm": 4.442931325188177,
+      "grad_norm": 4.3730733787425144,
       "learning_rate": 1.4244524932255026e-06,
-      "loss": 0.5745,
+      "loss": 0.5713,
       "step": 54
     },
     {
       "epoch": 3.928571428571429,
-      "grad_norm": 3.283331198334984,
+      "grad_norm": 3.353702364092209,
       "learning_rate": 1.2690815707615727e-06,
-      "loss": 0.2519,
+      "loss": 0.2574,
       "step": 55
     },
     {
       "epoch": 4.0,
-      "grad_norm": 4.600185580765784,
+      "grad_norm": 4.659106009637382,
       "learning_rate": 1.1214435464779006e-06,
-      "loss": 0.4206,
+      "loss": 0.4133,
       "step": 56
     },
     {
       "epoch": 4.071428571428571,
-      "grad_norm": 3.878198233905602,
+      "grad_norm": 3.868760155163949,
       "learning_rate": 9.81844422725109e-07,
-      "loss": 0.3339,
+      "loss": 0.334,
       "step": 57
     },
     {
       "epoch": 4.142857142857143,
-      "grad_norm": 3.558735485741733,
+      "grad_norm": 3.51726363976578,
       "learning_rate": 8.505735400048748e-07,
-      "loss": 0.4342,
+      "loss": 0.4336,
       "step": 58
     },
     {
       "epoch": 4.214285714285714,
-      "grad_norm": 3.4908015568361868,
+      "grad_norm": 3.4559508001260046,
       "learning_rate": 7.279029772675572e-07,
-      "loss": 0.3092,
+      "loss": 0.3091,
       "step": 59
     },
     {
       "epoch": 4.285714285714286,
-      "grad_norm": 3.442763565484024,
+      "grad_norm": 3.4232071286039503,
       "learning_rate": 6.140869879870287e-07,
-      "loss": 0.2904,
+      "loss": 0.2855,
       "step": 60
     },
     {
       "epoch": 4.357142857142857,
-      "grad_norm": 3.8067504051875685,
+      "grad_norm": 3.8024404087360923,
       "learning_rate": 5.0936147318152e-07,
-      "loss": 0.3512,
+      "loss": 0.352,
       "step": 61
     },
     {
       "epoch": 4.428571428571429,
-      "grad_norm": 4.105552776050043,
+      "grad_norm": 4.019294133630708,
       "learning_rate": 4.139434924727359e-07,
-      "loss": 0.4841,
+      "loss": 0.4831,
       "step": 62
     },
     {
       "epoch": 4.5,
-      "grad_norm": 3.388208677413764,
+      "grad_norm": 3.430280210521517,
       "learning_rate": 3.2803081419664483e-07,
-      "loss": 0.2912,
+      "loss": 0.2978,
       "step": 63
     },
     {
       "epoch": 4.571428571428571,
-      "grad_norm": 3.0263602419413873,
+      "grad_norm": 3.015729308739645,
       "learning_rate": 2.518015054984041e-07,
-      "loss": 0.2377,
+      "loss": 0.2341,
       "step": 64
     },
     {
       "epoch": 4.642857142857143,
-      "grad_norm": 4.005448888450202,
+      "grad_norm": 3.9855580227342284,
       "learning_rate": 1.8541356326100436e-07,
-      "loss": 0.4252,
+      "loss": 0.421,
       "step": 65
     },
     {
       "epoch": 4.714285714285714,
-      "grad_norm": 4.207717806182417,
+      "grad_norm": 4.166405494748373,
       "learning_rate": 1.2900458663260506e-07,
-      "loss": 0.4554,
+      "loss": 0.4509,
       "step": 66
     },
     {
       "epoch": 4.785714285714286,
-      "grad_norm": 4.53862291490365,
+      "grad_norm": 4.4514501143038325,
       "learning_rate": 8.269149183128988e-08,
-      "loss": 0.3974,
+      "loss": 0.3925,
       "step": 67
     },
     {
       "epoch": 4.857142857142857,
-      "grad_norm": 4.117090425838732,
+      "grad_norm": 4.231455128161484,
       "learning_rate": 4.657026981834623e-08,
-      "loss": 0.4093,
+      "loss": 0.408,
       "step": 68
     },
     {
       "epoch": 4.928571428571429,
-      "grad_norm": 4.169067110774366,
+      "grad_norm": 4.162171350169795,
       "learning_rate": 2.0715787342343586e-08,
-      "loss": 0.3711,
+      "loss": 0.3737,
       "step": 69
     },
     {
       "epoch": 5.0,
-      "grad_norm": 3.5840621133595802,
+      "grad_norm": 3.562544021343475,
       "learning_rate": 5.181631766362216e-09,
-      "loss": 0.3221,
+      "loss": 0.3182,
       "step": 70
     },
     {
       "epoch": 5.0,
       "step": 70,
       "total_flos": 213154529280.0,
-      "train_loss": 0.7240357622504234,
-      "train_runtime": 218.9144,
-      "train_samples_per_second": 0.64,
-      "train_steps_per_second": 0.32
+      "train_loss": 0.7237587083663259,
+      "train_runtime": 212.7239,
+      "train_samples_per_second": 0.658,
+      "train_steps_per_second": 0.329
     }
   ],
   "logging_steps": 1,