End of training

README.md

@@ -17,9 +17,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [youdiniplays/tl-ceb-model-v2](https://huggingface.co/youdiniplays/tl-ceb-model-v2) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3933
-- Bleu: 4.5385
-- Gen Len: 18.024
+- Loss: 0.3606
+- Bleu: 3.942
+- Gen Len: 18.31
 
 ## Model description
 
@@ -51,11 +51,11 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step  | Validation Loss | Bleu   | Gen Len |
 |:-------------:|:-----:|:-----:|:---------------:|:------:|:-------:|
-| 0.5165        | 1.0   | 6516  | 0.4074          | 4.3267 | 18.008  |
-| 0.4819        | 2.0   | 13032 | 0.4139          | 4.3977 | 18.051  |
-| 0.4448        | 3.0   | 19548 | 0.4067          | 4.4144 | 18.041  |
-| 0.3992        | 4.0   | 26064 | 0.3954          | 4.4996 | 18.016  |
-| 0.3604        | 5.0   | 32580 | 0.3933          | 4.5385 | 18.024  |
+| 0.4837        | 1.0   | 6516  | 0.3805          | 3.8228 | 18.313  |
+| 0.4479        | 2.0   | 13032 | 0.3810          | 3.7662 | 18.331  |
+| 0.4036        | 3.0   | 19548 | 0.3755          | 3.8306 | 18.343  |
+| 0.3572        | 4.0   | 26064 | 0.3673          | 3.8996 | 18.321  |
+| 0.3183        | 5.0   | 32580 | 0.3606          | 3.942  | 18.31   |
 
 
 ### Framework versions

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:12a56e4f696daf653f7e92b1bf3b257dfe9d43458e87219ae7e2d4a450ef9f0a
+oid sha256:4f41bb533aa86e301a1be3b65f96c1170612ae823deb447e90a35a957837ed0d
 size 242041896

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:58cc9abef2cbdea5218d8fa701a60440ccf98ca546e5f5c51a8311045ac25026
+oid sha256:5199d65e10b84399e00ecb642c52de2797b550f6e71e4e1f4b276272553ca552
 size 484163514

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8dab94ed4b71ebfd79df176ffa20e8a3d588127f93d1c43937f8527fd3a9c67d
+oid sha256:d1cd63170bb513ed1093722cd85dcb9976e6bd105066dd8d5d629488d4e94f5b
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:36e6f55bcc217a43fef0a794bb548420d34ff2f529f3e814b7aa83539a026216
+oid sha256:53e446aca6ce45dd19f0dd3bb36af3837e8ac7287ccd924c8a3f9b7a954a486f
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 4.987722529158994,
+  "epoch": 3.14610190300798,
   "eval_steps": 500,
-  "global_step": 32500,
+  "global_step": 20500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -11,439 +11,285 @@
     {
       "epoch": 0.08,
       "learning_rate": 0.0009846531614487416,
-      "loss": 0.4248,
+      "loss": 0.3358,
       "step": 500
     },
     {
       "epoch": 0.15,
       "learning_rate": 0.0009693063228974831,
-      "loss": 0.4521,
+      "loss": 0.3723,
       "step": 1000
     },
     {
       "epoch": 0.23,
       "learning_rate": 0.0009539594843462247,
-      "loss": 0.4654,
+      "loss": 0.3901,
       "step": 1500
     },
     {
       "epoch": 0.31,
       "learning_rate": 0.0009386126457949663,
-      "loss": 0.4686,
+      "loss": 0.4022,
       "step": 2000
     },
     {
       "epoch": 0.38,
       "learning_rate": 0.0009232658072437078,
-      "loss": 0.4705,
+      "loss": 0.4109,
       "step": 2500
     },
     {
       "epoch": 0.46,
       "learning_rate": 0.0009079189686924494,
-      "loss": 0.4757,
+      "loss": 0.4132,
       "step": 3000
     },
     {
       "epoch": 0.54,
       "learning_rate": 0.000892572130141191,
-      "loss": 0.478,
+      "loss": 0.4197,
       "step": 3500
     },
     {
       "epoch": 0.61,
       "learning_rate": 0.0008772252915899325,
-      "loss": 0.4802,
+      "loss": 0.4227,
       "step": 4000
     },
     {
       "epoch": 0.69,
       "learning_rate": 0.0008618784530386741,
-      "loss": 0.4848,
+      "loss": 0.424,
       "step": 4500
     },
     {
       "epoch": 0.77,
       "learning_rate": 0.0008465316144874156,
-      "loss": 0.4835,
+      "loss": 0.4228,
       "step": 5000
     },
     {
       "epoch": 0.84,
       "learning_rate": 0.0008311847759361571,
-      "loss": 0.4845,
+      "loss": 0.4246,
       "step": 5500
     },
     {
       "epoch": 0.92,
       "learning_rate": 0.0008158379373848988,
-      "loss": 0.485,
+      "loss": 0.4268,
       "step": 6000
     },
     {
       "epoch": 1.0,
-      "learning_rate": 0.0008005524861878454,
-      "loss": 0.4837,
+      "learning_rate": 0.0008005217925107428,
+      "loss": 0.4316,
       "step": 6500
     },
     {
       "epoch": 1.0,
-      "eval_bleu": 3.8228,
-      "eval_gen_len": 18.313,
-      "eval_loss": 0.3804944157600403,
-      "eval_runtime": 25.2075,
-      "eval_samples_per_second": 39.671,
-      "eval_steps_per_second": 2.499,
+      "eval_bleu": 4.2677,
+      "eval_gen_len": 18.024,
+      "eval_loss": 0.45959702134132385,
+      "eval_runtime": 23.3818,
+      "eval_samples_per_second": 42.768,
+      "eval_steps_per_second": 2.694,
       "step": 6516
     },
     {
       "epoch": 1.07,
-      "learning_rate": 0.0007852056476365869,
-      "loss": 0.4371,
+      "learning_rate": 0.0007851749539594844,
+      "loss": 0.3897,
       "step": 7000
     },
     {
       "epoch": 1.15,
-      "learning_rate": 0.0007698588090853285,
-      "loss": 0.4371,
+      "learning_rate": 0.0007698281154082259,
+      "loss": 0.3901,
       "step": 7500
     },
     {
       "epoch": 1.23,
-      "learning_rate": 0.0007545119705340699,
-      "loss": 0.4439,
+      "learning_rate": 0.0007544812768569675,
+      "loss": 0.3965,
       "step": 8000
     },
     {
       "epoch": 1.3,
       "learning_rate": 0.0007391651319828116,
-      "loss": 0.4449,
+      "loss": 0.394,
       "step": 8500
     },
     {
       "epoch": 1.38,
       "learning_rate": 0.0007238182934315531,
-      "loss": 0.4484,
+      "loss": 0.4003,
       "step": 9000
     },
     {
       "epoch": 1.46,
       "learning_rate": 0.0007084714548802946,
-      "loss": 0.4471,
+      "loss": 0.403,
       "step": 9500
     },
     {
       "epoch": 1.53,
       "learning_rate": 0.0006931246163290362,
-      "loss": 0.4562,
+      "loss": 0.4067,
       "step": 10000
     },
     {
       "epoch": 1.61,
       "learning_rate": 0.0006778084714548804,
-      "loss": 0.4496,
+      "loss": 0.4067,
       "step": 10500
     },
     {
       "epoch": 1.69,
       "learning_rate": 0.0006624616329036218,
-      "loss": 0.4485,
+      "loss": 0.4047,
       "step": 11000
     },
     {
       "epoch": 1.76,
       "learning_rate": 0.0006471147943523635,
-      "loss": 0.4481,
+      "loss": 0.4032,
       "step": 11500
     },
     {
       "epoch": 1.84,
       "learning_rate": 0.000631767955801105,
-      "loss": 0.4479,
+      "loss": 0.4077,
       "step": 12000
     },
     {
       "epoch": 1.92,
-      "learning_rate": 0.000616451810926949,
-      "loss": 0.4472,
+      "learning_rate": 0.0006164825046040515,
+      "loss": 0.4062,
       "step": 12500
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.0006011049723756906,
-      "loss": 0.4479,
+      "learning_rate": 0.0006011356660527931,
+      "loss": 0.4074,
       "step": 13000
     },
     {
       "epoch": 2.0,
-      "eval_bleu": 3.7662,
-      "eval_gen_len": 18.331,
-      "eval_loss": 0.3810326159000397,
-      "eval_runtime": 21.6354,
-      "eval_samples_per_second": 46.22,
-      "eval_steps_per_second": 2.912,
+      "eval_bleu": 4.2659,
+      "eval_gen_len": 18.019,
+      "eval_loss": 0.45938295125961304,
+      "eval_runtime": 24.1279,
+      "eval_samples_per_second": 41.446,
+      "eval_steps_per_second": 2.611,
       "step": 13032
     },
     {
       "epoch": 2.07,
-      "learning_rate": 0.0005857581338244322,
-      "loss": 0.4063,
+      "learning_rate": 0.0005857888275015347,
+      "loss": 0.3667,
       "step": 13500
     },
     {
       "epoch": 2.15,
-      "learning_rate": 0.0005704112952731737,
-      "loss": 0.4007,
+      "learning_rate": 0.0005704419889502763,
+      "loss": 0.3674,
       "step": 14000
     },
     {
       "epoch": 2.23,
       "learning_rate": 0.0005550951503990178,
-      "loss": 0.4059,
+      "loss": 0.3749,
       "step": 14500
     },
     {
       "epoch": 2.3,
       "learning_rate": 0.0005397483118477594,
-      "loss": 0.4023,
+      "loss": 0.3707,
       "step": 15000
     },
     {
       "epoch": 2.38,
       "learning_rate": 0.0005244014732965009,
-      "loss": 0.4061,
+      "loss": 0.3676,
       "step": 15500
     },
     {
       "epoch": 2.46,
       "learning_rate": 0.0005090546347452425,
-      "loss": 0.4046,
+      "loss": 0.3761,
       "step": 16000
     },
     {
       "epoch": 2.53,
       "learning_rate": 0.0004937384898710866,
-      "loss": 0.4082,
+      "loss": 0.3795,
       "step": 16500
     },
     {
       "epoch": 2.61,
       "learning_rate": 0.00047839165131982814,
-      "loss": 0.4122,
+      "loss": 0.377,
       "step": 17000
     },
     {
       "epoch": 2.69,
       "learning_rate": 0.0004630448127685697,
-      "loss": 0.4076,
+      "loss": 0.3804,
       "step": 17500
     },
     {
       "epoch": 2.76,
       "learning_rate": 0.00044769797421731125,
-      "loss": 0.4055,
+      "loss": 0.3824,
       "step": 18000
     },
     {
       "epoch": 2.84,
       "learning_rate": 0.00043238182934315537,
-      "loss": 0.4003,
+      "loss": 0.3724,
       "step": 18500
     },
     {
       "epoch": 2.92,
       "learning_rate": 0.00041703499079189687,
-      "loss": 0.4045,
+      "loss": 0.3795,
       "step": 19000
     },
     {
       "epoch": 2.99,
       "learning_rate": 0.0004016881522406384,
-      "loss": 0.4036,
+      "loss": 0.3833,
       "step": 19500
     },
     {
       "epoch": 3.0,
-      "eval_bleu": 3.8306,
-      "eval_gen_len": 18.343,
-      "eval_loss": 0.3754543364048004,
-      "eval_runtime": 21.8401,
-      "eval_samples_per_second": 45.787,
-      "eval_steps_per_second": 2.885,
+      "eval_bleu": 4.3571,
+      "eval_gen_len": 18.015,
+      "eval_loss": 0.4461449086666107,
+      "eval_runtime": 23.8381,
+      "eval_samples_per_second": 41.95,
+      "eval_steps_per_second": 2.643,
       "step": 19548
     },
     {
       "epoch": 3.07,
       "learning_rate": 0.00038634131368938,
-      "loss": 0.3611,
+      "loss": 0.3423,
       "step": 20000
     },
     {
       "epoch": 3.15,
-      "learning_rate": 0.0003710251688152241,
-      "loss": 0.3608,
+      "learning_rate": 0.0003710558624923266,
+      "loss": 0.3375,
       "step": 20500
-    },
-    {
-      "epoch": 3.22,
-      "learning_rate": 0.0003556783302639656,
-      "loss": 0.3676,
-      "step": 21000
-    },
-    {
-      "epoch": 3.3,
-      "learning_rate": 0.00034033149171270715,
-      "loss": 0.3643,
-      "step": 21500
-    },
-    {
-      "epoch": 3.38,
-      "learning_rate": 0.00032498465316144876,
-      "loss": 0.3591,
-      "step": 22000
-    },
-    {
-      "epoch": 3.45,
-      "learning_rate": 0.0003096378146101903,
-      "loss": 0.3651,
-      "step": 22500
-    },
-    {
-      "epoch": 3.53,
-      "learning_rate": 0.0002943216697360344,
-      "loss": 0.3669,
-      "step": 23000
-    },
-    {
-      "epoch": 3.61,
-      "learning_rate": 0.00027897483118477593,
-      "loss": 0.3609,
-      "step": 23500
-    },
-    {
-      "epoch": 3.68,
-      "learning_rate": 0.0002636279926335175,
-      "loss": 0.3626,
-      "step": 24000
-    },
-    {
-      "epoch": 3.76,
-      "learning_rate": 0.0002482811540822591,
-      "loss": 0.3625,
-      "step": 24500
-    },
-    {
-      "epoch": 3.84,
-      "learning_rate": 0.00023296500920810313,
-      "loss": 0.362,
-      "step": 25000
-    },
-    {
-      "epoch": 3.91,
-      "learning_rate": 0.00021761817065684468,
-      "loss": 0.3554,
-      "step": 25500
-    },
-    {
-      "epoch": 3.99,
-      "learning_rate": 0.00020227133210558627,
-      "loss": 0.3572,
-      "step": 26000
-    },
-    {
-      "epoch": 4.0,
-      "eval_bleu": 3.8996,
-      "eval_gen_len": 18.321,
-      "eval_loss": 0.3672715425491333,
-      "eval_runtime": 22.3643,
-      "eval_samples_per_second": 44.714,
-      "eval_steps_per_second": 2.817,
-      "step": 26064
-    },
-    {
-      "epoch": 4.07,
-      "learning_rate": 0.00018692449355432782,
-      "loss": 0.3273,
-      "step": 26500
-    },
-    {
-      "epoch": 4.14,
-      "learning_rate": 0.00017160834868017188,
-      "loss": 0.3241,
-      "step": 27000
-    },
-    {
-      "epoch": 4.22,
-      "learning_rate": 0.00015626151012891346,
-      "loss": 0.3266,
-      "step": 27500
-    },
-    {
-      "epoch": 4.3,
-      "learning_rate": 0.000140914671577655,
-      "loss": 0.3238,
-      "step": 28000
-    },
-    {
-      "epoch": 4.37,
-      "learning_rate": 0.00012556783302639657,
-      "loss": 0.3256,
-      "step": 28500
-    },
-    {
-      "epoch": 4.45,
-      "learning_rate": 0.00011025168815224063,
-      "loss": 0.3258,
-      "step": 29000
-    },
-    {
-      "epoch": 4.53,
-      "learning_rate": 9.49048496009822e-05,
-      "loss": 0.3269,
-      "step": 29500
-    },
-    {
-      "epoch": 4.6,
-      "learning_rate": 7.955801104972376e-05,
-      "loss": 0.3228,
-      "step": 30000
-    },
-    {
-      "epoch": 4.68,
-      "learning_rate": 6.421117249846533e-05,
-      "loss": 0.3179,
-      "step": 30500
-    },
-    {
-      "epoch": 4.76,
-      "learning_rate": 4.8895027624309394e-05,
-      "loss": 0.3187,
-      "step": 31000
-    },
-    {
-      "epoch": 4.83,
-      "learning_rate": 3.3548189073050956e-05,
-      "loss": 0.3178,
-      "step": 31500
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 1.820135052179251e-05,
-      "loss": 0.3183,
-      "step": 32000
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 2.85451197053407e-06,
-      "loss": 0.3183,
-      "step": 32500
     }
   ],
   "logging_steps": 500,
   "max_steps": 32580,
   "num_train_epochs": 5,
   "save_steps": 500,
-  "total_flos": 1.7229657373016064e+16,
+  "total_flos": 1.0876549321457664e+16,
   "trial_name": null,
   "trial_params": null
 }

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3b58968f338c312448777ee1ccf6dbc18e49e360004c218ec824e458e768bc68
+oid sha256:0d9e5fd144c9e568216db4505e71aa6bc722764749fdf639a5e84c071e512165
 size 4792

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:12a56e4f696daf653f7e92b1bf3b257dfe9d43458e87219ae7e2d4a450ef9f0a
+oid sha256:fb2b0a9b9f4a155763c87b51105045e65887afa5bfbc46b443dd01f5651826fc
 size 242041896

runs/Jan21_14-30-28_54c898f0920f/events.out.tfevents.1705847429.54c898f0920f.211.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8421fec6ebe0b1140a36339f157ceabd74d54d6603509b8b874f4a0075277118
-size 17066
+oid sha256:a8fb12633f17104e51c7e80afdea5d9c62195fbc60ccb09102029559af398b02
+size 17803