Training in progress, step 2000, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8760cf4fd71ffe55ac2b4376681bd1ece98f92647b8a0d7884a6bdb84448e4ac
+oid sha256:c2a193933a701dc0f71711b06b789f4ce61a8231b50c23d78cd67e1473cab528
 size 529337896

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:80a5647b5f60dde4cc65574d3b50f39a981867a2f21becb1ebd54b5014687f04
+oid sha256:e42073b4332f17525a5be43285c6bf9938b3269e377a9d0cc0a79fe7bbb1698e
 size 871247243

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9bb1461e40c7e08171c25a0bc2bf64964397ea8cc712bb88016a59954d83db2d
+oid sha256:14cd4a8cf54ce87e9e7818ae50060fa501ea05e8685c2025ab631bfde31c87cf
 size 14645

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eb2a89a85e1c8ac7ee9b5678d857a4d99090faa37a24489f44839e0801dfe16a
+oid sha256:b716fb1dcbc77ddc6a0ae7c4c98d559e36f1986fff006ef6982bd75247833235
 size 1465

last-checkpoint/trainer_state.json

@@ -2,486 +2,170 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.12,
+  "epoch": 0.04,
   "eval_steps": 1000,
-  "global_step": 6000,
+  "global_step": 2000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.002,
-      "grad_norm": 1.158125638961792,
+      "grad_norm": 0.8063191175460815,
       "learning_rate": 7.920000000000001e-05,
-      "loss": 8.2536,
+      "loss": 8.2537,
       "step": 100
     },
     {
       "epoch": 0.004,
-      "grad_norm": 0.6474116444587708,
+      "grad_norm": 0.6161019206047058,
       "learning_rate": 0.00015920000000000002,
-      "loss": 6.577,
+      "loss": 6.5731,
       "step": 200
     },
     {
       "epoch": 0.006,
-      "grad_norm": 0.8440109491348267,
+      "grad_norm": 0.8347169160842896,
       "learning_rate": 0.00023920000000000001,
-      "loss": 6.1636,
+      "loss": 6.1639,
       "step": 300
     },
     {
       "epoch": 0.008,
-      "grad_norm": 0.712805986404419,
+      "grad_norm": 0.9150793552398682,
       "learning_rate": 0.0003192,
-      "loss": 5.8686,
+      "loss": 5.8682,
       "step": 400
     },
     {
       "epoch": 0.01,
-      "grad_norm": 0.6756730675697327,
+      "grad_norm": 0.6698479652404785,
       "learning_rate": 0.0003992,
-      "loss": 5.5639,
+      "loss": 5.5708,
       "step": 500
     },
     {
       "epoch": 0.012,
-      "grad_norm": 0.8336403965950012,
+      "grad_norm": 0.5323518514633179,
       "learning_rate": 0.00047920000000000005,
-      "loss": 5.3098,
+      "loss": 5.3157,
       "step": 600
     },
     {
       "epoch": 0.014,
-      "grad_norm": 0.6417033672332764,
+      "grad_norm": 0.7472386956214905,
       "learning_rate": 0.0005592,
-      "loss": 5.0845,
+      "loss": 5.0907,
       "step": 700
     },
     {
       "epoch": 0.016,
-      "grad_norm": 0.5867908596992493,
+      "grad_norm": 0.6033297181129456,
       "learning_rate": 0.0006392,
-      "loss": 4.8767,
+      "loss": 4.8805,
       "step": 800
     },
     {
       "epoch": 0.018,
-      "grad_norm": 0.5321984887123108,
+      "grad_norm": 0.497113436460495,
       "learning_rate": 0.0007191999999999999,
-      "loss": 4.7211,
+      "loss": 4.7207,
       "step": 900
     },
     {
       "epoch": 0.02,
-      "grad_norm": 0.5662059187889099,
+      "grad_norm": 0.47698545455932617,
       "learning_rate": 0.0007992,
-      "loss": 4.6059,
+      "loss": 4.6068,
       "step": 1000
     },
     {
       "epoch": 0.02,
-      "eval_accuracy": 0.2709784735812133,
-      "eval_loss": 4.504483222961426,
-      "eval_runtime": 11.2003,
-      "eval_samples_per_second": 89.283,
-      "eval_steps_per_second": 0.714,
+      "eval_accuracy": 0.2703913894324853,
+      "eval_loss": 4.5053935050964355,
+      "eval_runtime": 13.0992,
+      "eval_samples_per_second": 76.341,
+      "eval_steps_per_second": 0.611,
       "step": 1000
     },
     {
       "epoch": 0.022,
-      "grad_norm": 0.45525625348091125,
+      "grad_norm": 0.44341349601745605,
       "learning_rate": 0.0008792,
-      "loss": 4.4901,
+      "loss": 4.4904,
       "step": 1100
     },
     {
       "epoch": 0.024,
-      "grad_norm": 0.4701859652996063,
+      "grad_norm": 0.4383724331855774,
       "learning_rate": 0.0009592000000000001,
-      "loss": 4.4344,
+      "loss": 4.439,
       "step": 1200
     },
     {
       "epoch": 0.026,
-      "grad_norm": 0.40884163975715637,
+      "grad_norm": 0.3894674777984619,
       "learning_rate": 0.0010391999999999999,
-      "loss": 4.3707,
+      "loss": 4.3735,
       "step": 1300
     },
     {
       "epoch": 0.028,
-      "grad_norm": 0.3846769332885742,
+      "grad_norm": 0.41903820633888245,
       "learning_rate": 0.0011192,
-      "loss": 4.336,
+      "loss": 4.3406,
       "step": 1400
     },
     {
       "epoch": 0.03,
-      "grad_norm": 0.3585458993911743,
+      "grad_norm": 0.36601418256759644,
       "learning_rate": 0.0011992,
-      "loss": 4.2832,
+      "loss": 4.2879,
       "step": 1500
     },
     {
       "epoch": 0.032,
-      "grad_norm": 0.4558027982711792,
+      "grad_norm": 0.4282030165195465,
       "learning_rate": 0.0012791999999999999,
-      "loss": 4.2599,
+      "loss": 4.2625,
       "step": 1600
     },
     {
       "epoch": 0.034,
-      "grad_norm": 0.33011195063591003,
+      "grad_norm": 0.3431214690208435,
       "learning_rate": 0.0013592,
-      "loss": 4.2287,
+      "loss": 4.2316,
       "step": 1700
     },
     {
       "epoch": 0.036,
-      "grad_norm": 0.32163774967193604,
+      "grad_norm": 0.3303460478782654,
       "learning_rate": 0.0014392,
-      "loss": 4.1932,
+      "loss": 4.1976,
       "step": 1800
     },
     {
       "epoch": 0.038,
-      "grad_norm": 0.30342814326286316,
+      "grad_norm": 0.3030431866645813,
       "learning_rate": 0.0015192,
-      "loss": 4.1447,
+      "loss": 4.1494,
       "step": 1900
     },
     {
       "epoch": 0.04,
-      "grad_norm": 0.32037997245788574,
+      "grad_norm": 0.3076256513595581,
       "learning_rate": 0.0015992,
-      "loss": 4.149,
+      "loss": 4.1688,
       "step": 2000
     },
     {
       "epoch": 0.04,
-      "eval_accuracy": 0.30734833659491195,
-      "eval_loss": 4.087185859680176,
-      "eval_runtime": 11.8275,
-      "eval_samples_per_second": 84.549,
-      "eval_steps_per_second": 0.676,
+      "eval_accuracy": 0.30658512720156555,
+      "eval_loss": 4.094621658325195,
+      "eval_runtime": 12.4723,
+      "eval_samples_per_second": 80.178,
+      "eval_steps_per_second": 0.641,
       "step": 2000
-    },
-    {
-      "epoch": 0.042,
-      "grad_norm": 0.3013302683830261,
-      "learning_rate": 0.0016792,
-      "loss": 4.1191,
-      "step": 2100
-    },
-    {
-      "epoch": 0.044,
-      "grad_norm": 0.25836440920829773,
-      "learning_rate": 0.0017592,
-      "loss": 4.075,
-      "step": 2200
-    },
-    {
-      "epoch": 0.046,
-      "grad_norm": 0.2489597499370575,
-      "learning_rate": 0.0018392,
-      "loss": 4.0824,
-      "step": 2300
-    },
-    {
-      "epoch": 0.048,
-      "grad_norm": 0.24732087552547455,
-      "learning_rate": 0.0019192,
-      "loss": 4.0533,
-      "step": 2400
-    },
-    {
-      "epoch": 0.05,
-      "grad_norm": 0.2429002970457077,
-      "learning_rate": 0.0019992,
-      "loss": 4.0472,
-      "step": 2500
-    },
-    {
-      "epoch": 0.052,
-      "grad_norm": 0.21534161269664764,
-      "learning_rate": 0.001999978563623903,
-      "loss": 4.0063,
-      "step": 2600
-    },
-    {
-      "epoch": 0.054,
-      "grad_norm": 0.2244805246591568,
-      "learning_rate": 0.0019999133871331223,
-      "loss": 4.0074,
-      "step": 2700
-    },
-    {
-      "epoch": 0.056,
-      "grad_norm": 0.1999787837266922,
-      "learning_rate": 0.0019998044711915177,
-      "loss": 3.9708,
-      "step": 2800
-    },
-    {
-      "epoch": 0.058,
-      "grad_norm": 0.19969643652439117,
-      "learning_rate": 0.0019996518205634257,
-      "loss": 3.9525,
-      "step": 2900
-    },
-    {
-      "epoch": 0.06,
-      "grad_norm": 0.21098633110523224,
-      "learning_rate": 0.0019994554419262797,
-      "loss": 3.9491,
-      "step": 3000
-    },
-    {
-      "epoch": 0.06,
-      "eval_accuracy": 0.324146771037182,
-      "eval_loss": 3.9009768962860107,
-      "eval_runtime": 18.6503,
-      "eval_samples_per_second": 53.618,
-      "eval_steps_per_second": 0.429,
-      "step": 3000
-    },
-    {
-      "epoch": 0.062,
-      "grad_norm": 0.1954544633626938,
-      "learning_rate": 0.001999215343870317,
-      "loss": 3.942,
-      "step": 3100
-    },
-    {
-      "epoch": 0.064,
-      "grad_norm": 0.19358158111572266,
-      "learning_rate": 0.0019989315368982054,
-      "loss": 3.9159,
-      "step": 3200
-    },
-    {
-      "epoch": 0.066,
-      "grad_norm": 0.18568383157253265,
-      "learning_rate": 0.0019986040334245797,
-      "loss": 3.8959,
-      "step": 3300
-    },
-    {
-      "epoch": 0.068,
-      "grad_norm": 0.1798350065946579,
-      "learning_rate": 0.001998232847775504,
-      "loss": 3.9035,
-      "step": 3400
-    },
-    {
-      "epoch": 0.07,
-      "grad_norm": 0.18326053023338318,
-      "learning_rate": 0.0019978179961878404,
-      "loss": 3.8771,
-      "step": 3500
-    },
-    {
-      "epoch": 0.072,
-      "grad_norm": 0.1763591170310974,
-      "learning_rate": 0.001997359496808541,
-      "loss": 3.8714,
-      "step": 3600
-    },
-    {
-      "epoch": 0.074,
-      "grad_norm": 0.16502562165260315,
-      "learning_rate": 0.001996857369693855,
-      "loss": 3.8516,
-      "step": 3700
-    },
-    {
-      "epoch": 0.076,
-      "grad_norm": 0.18450422585010529,
-      "learning_rate": 0.0019963116368084486,
-      "loss": 3.8573,
-      "step": 3800
-    },
-    {
-      "epoch": 0.078,
-      "grad_norm": 0.1750190556049347,
-      "learning_rate": 0.001995722322024446,
-      "loss": 3.8403,
-      "step": 3900
-    },
-    {
-      "epoch": 0.08,
-      "grad_norm": 0.1842149794101715,
-      "learning_rate": 0.001995089451120385,
-      "loss": 3.8158,
-      "step": 4000
-    },
-    {
-      "epoch": 0.08,
-      "eval_accuracy": 0.3340352250489237,
-      "eval_loss": 3.7926738262176514,
-      "eval_runtime": 12.3279,
-      "eval_samples_per_second": 81.117,
-      "eval_steps_per_second": 0.649,
-      "step": 4000
-    },
-    {
-      "epoch": 0.082,
-      "grad_norm": 0.1560104936361313,
-      "learning_rate": 0.0019944130517800893,
-      "loss": 3.8258,
-      "step": 4100
-    },
-    {
-      "epoch": 0.084,
-      "grad_norm": 0.15086719393730164,
-      "learning_rate": 0.001993693153591457,
-      "loss": 3.8246,
-      "step": 4200
-    },
-    {
-      "epoch": 0.086,
-      "grad_norm": 0.1565793752670288,
-      "learning_rate": 0.0019929297880451674,
-      "loss": 3.8036,
-      "step": 4300
-    },
-    {
-      "epoch": 0.088,
-      "grad_norm": 0.16023847460746765,
-      "learning_rate": 0.0019921229885333023,
-      "loss": 3.7936,
-      "step": 4400
-    },
-    {
-      "epoch": 0.09,
-      "grad_norm": 0.16122952103614807,
-      "learning_rate": 0.001991272790347886,
-      "loss": 3.8071,
-      "step": 4500
-    },
-    {
-      "epoch": 0.092,
-      "grad_norm": 0.1704457700252533,
-      "learning_rate": 0.0019903792306793415,
-      "loss": 3.7761,
-      "step": 4600
-    },
-    {
-      "epoch": 0.094,
-      "grad_norm": 0.14904244244098663,
-      "learning_rate": 0.001989442348614863,
-      "loss": 3.7839,
-      "step": 4700
-    },
-    {
-      "epoch": 0.096,
-      "grad_norm": 0.15502023696899414,
-      "learning_rate": 0.0019884621851367075,
-      "loss": 3.7759,
-      "step": 4800
-    },
-    {
-      "epoch": 0.098,
-      "grad_norm": 0.14749225974082947,
-      "learning_rate": 0.001987438783120401,
-      "loss": 3.7718,
-      "step": 4900
-    },
-    {
-      "epoch": 0.1,
-      "grad_norm": 0.15654946863651276,
-      "learning_rate": 0.001986372187332862,
-      "loss": 3.7804,
-      "step": 5000
-    },
-    {
-      "epoch": 0.1,
-      "eval_accuracy": 0.3406477495107632,
-      "eval_loss": 3.7244012355804443,
-      "eval_runtime": 12.2821,
-      "eval_samples_per_second": 81.419,
-      "eval_steps_per_second": 0.651,
-      "step": 5000
-    },
-    {
-      "epoch": 0.102,
-      "grad_norm": 0.15231221914291382,
-      "learning_rate": 0.0019852624444304467,
-      "loss": 3.7449,
-      "step": 5100
-    },
-    {
-      "epoch": 0.104,
-      "grad_norm": 0.15843184292316437,
-      "learning_rate": 0.0019841096029569044,
-      "loss": 3.7559,
-      "step": 5200
-    },
-    {
-      "epoch": 0.106,
-      "grad_norm": 0.14371678233146667,
-      "learning_rate": 0.0019829137133412556,
-      "loss": 3.7536,
-      "step": 5300
-    },
-    {
-      "epoch": 0.108,
-      "grad_norm": 0.1501024067401886,
-      "learning_rate": 0.001981674827895587,
-      "loss": 3.7346,
-      "step": 5400
-    },
-    {
-      "epoch": 0.11,
-      "grad_norm": 0.1276807337999344,
-      "learning_rate": 0.00198039300081276,
-      "loss": 3.7387,
-      "step": 5500
-    },
-    {
-      "epoch": 0.112,
-      "grad_norm": 0.13706138730049133,
-      "learning_rate": 0.0019790682881640448,
-      "loss": 3.738,
-      "step": 5600
-    },
-    {
-      "epoch": 0.114,
-      "grad_norm": 0.14035683870315552,
-      "learning_rate": 0.001977700747896664,
-      "loss": 3.7224,
-      "step": 5700
-    },
-    {
-      "epoch": 0.116,
-      "grad_norm": 0.1608739197254181,
-      "learning_rate": 0.001976290439831259,
-      "loss": 3.7252,
-      "step": 5800
-    },
-    {
-      "epoch": 0.118,
-      "grad_norm": 0.130873903632164,
-      "learning_rate": 0.0019748374256592736,
-      "loss": 3.7179,
-      "step": 5900
-    },
-    {
-      "epoch": 0.12,
-      "grad_norm": 0.13922768831253052,
-      "learning_rate": 0.0019733417689402543,
-      "loss": 3.712,
-      "step": 6000
-    },
-    {
-      "epoch": 0.12,
-      "eval_accuracy": 0.3451976516634051,
-      "eval_loss": 3.6720173358917236,
-      "eval_runtime": 12.3084,
-      "eval_samples_per_second": 81.245,
-      "eval_steps_per_second": 0.65,
-      "step": 6000
     }
   ],
   "logging_steps": 100,

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cecd7e947ce75b613d9846ddffb484a23218fd659b5b05c087445bc754630b37
+oid sha256:ab7799c96f82a5f3608933c037c72f58f0cf65bc6f2172c20a2b57dff04d327b
 size 5905