1870

.gitattributes

@@ -17,3 +17,5 @@ global_step1530/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt filter=lfs diff=l
 global_step1530/mp_rank_00_model_states.pt filter=lfs diff=lfs merge=lfs -text
 global_step1700/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt filter=lfs diff=lfs merge=lfs -text
 global_step1700/mp_rank_00_model_states.pt filter=lfs diff=lfs merge=lfs -text
+global_step1870/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt filter=lfs diff=lfs merge=lfs -text
+global_step1870/mp_rank_00_model_states.pt filter=lfs diff=lfs merge=lfs -text

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:493127a18f562097ef7c38d16886f04e034a9f418dc607b528b5051ed93d5a84
+oid sha256:93145fb67a6755f3358da8d36be3e02948c85fb7102b8b9e3b481f8674852c53
 size 335606536

{global_step1700 → global_step1870}/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:092d8bb7df9ac2ac664a2cf69679528ebff360207f68e8d7a545374ac1a29aac
+oid sha256:49c974b0eae7df4d70fa3958f5269a525f7333e726267a4b9a59af45154dccc2
 size 2013320965

{global_step1700 → global_step1870}/mp_rank_00_model_states.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:437c9e6fe407eb8bc2ff3462e0f9fcd63f91983a42fc7c9ee6677474afa21f33
+oid sha256:9844d738eb2fa46e0ae4955e3c82502b358271721cc524822f0ac9da6d4c9639
 size 335901529

latest

@@ -1 +1 @@
-global_step1700
+global_step1870

trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 1700,
-  "best_metric": 0.3144903779029846,
-  "best_model_checkpoint": "/home/longtail/data/outputs/exaone_7b_lora_rtlcoder_ratio1.0/checkpoint-1700",
-  "epoch": 1.2811381194648577,
+  "best_global_step": 1870,
+  "best_metric": 0.31162187457084656,
+  "best_model_checkpoint": "/home/longtail/data/outputs/exaone_7b_lora_rtlcoder_ratio1.0/checkpoint-1870",
+  "epoch": 1.409270774448841,
   "eval_steps": 170,
-  "global_step": 1700,
+  "global_step": 1870,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -1818,6 +1818,187 @@
       "eval_samples_per_second": 5.971,
       "eval_steps_per_second": 1.494,
       "step": 1700
+    },
+    {
+      "entropy": 0.2976318359375,
+      "epoch": 1.2886753344639157,
+      "grad_norm": 0.3675684630870819,
+      "learning_rate": 1.740689807638311e-05,
+      "loss": 0.3084413051605225,
+      "mean_token_accuracy": 0.9068222478032112,
+      "num_tokens": 21606886.0,
+      "step": 1710
+    },
+    {
+      "entropy": 0.26845703125,
+      "epoch": 1.2962125494629735,
+      "grad_norm": 0.48080354928970337,
+      "learning_rate": 1.7374010091410126e-05,
+      "loss": 0.27617788314819336,
+      "mean_token_accuracy": 0.9144353941082954,
+      "num_tokens": 21728204.0,
+      "step": 1720
+    },
+    {
+      "entropy": 0.3060546875,
+      "epoch": 1.3037497644620313,
+      "grad_norm": 0.4360472559928894,
+      "learning_rate": 1.7340946352319795e-05,
+      "loss": 0.3208837270736694,
+      "mean_token_accuracy": 0.9051508828997612,
+      "num_tokens": 21847166.0,
+      "step": 1730
+    },
+    {
+      "entropy": 0.275732421875,
+      "epoch": 1.311286979461089,
+      "grad_norm": 0.42607492208480835,
+      "learning_rate": 1.730770764716206e-05,
+      "loss": 0.2842782735824585,
+      "mean_token_accuracy": 0.912279462814331,
+      "num_tokens": 21972827.0,
+      "step": 1740
+    },
+    {
+      "entropy": 0.298291015625,
+      "epoch": 1.3188241944601469,
+      "grad_norm": 0.3897174298763275,
+      "learning_rate": 1.7274294768157065e-05,
+      "loss": 0.30409531593322753,
+      "mean_token_accuracy": 0.907126384973526,
+      "num_tokens": 22098702.0,
+      "step": 1750
+    },
+    {
+      "entropy": 0.301708984375,
+      "epoch": 1.3263614094592049,
+      "grad_norm": 0.3547886312007904,
+      "learning_rate": 1.7240708511676253e-05,
+      "loss": 0.30962510108947755,
+      "mean_token_accuracy": 0.9043201595544815,
+      "num_tokens": 22224280.0,
+      "step": 1760
+    },
+    {
+      "entropy": 0.2868896484375,
+      "epoch": 1.3338986244582627,
+      "grad_norm": 0.41709667444229126,
+      "learning_rate": 1.7206949678223388e-05,
+      "loss": 0.29194619655609133,
+      "mean_token_accuracy": 0.9089675724506379,
+      "num_tokens": 22350124.0,
+      "step": 1770
+    },
+    {
+      "entropy": 0.31044921875,
+      "epoch": 1.3414358394573205,
+      "grad_norm": 0.41593441367149353,
+      "learning_rate": 1.7173019072415488e-05,
+      "loss": 0.3232313871383667,
+      "mean_token_accuracy": 0.9014707505702972,
+      "num_tokens": 22466149.0,
+      "step": 1780
+    },
+    {
+      "entropy": 0.302587890625,
+      "epoch": 1.3489730544563785,
+      "grad_norm": 0.3737364113330841,
+      "learning_rate": 1.7138917502963627e-05,
+      "loss": 0.3045801639556885,
+      "mean_token_accuracy": 0.9061708480119706,
+      "num_tokens": 22589692.0,
+      "step": 1790
+    },
+    {
+      "entropy": 0.288525390625,
+      "epoch": 1.3565102694554363,
+      "grad_norm": 0.36714568734169006,
+      "learning_rate": 1.710464578265369e-05,
+      "loss": 0.2963777780532837,
+      "mean_token_accuracy": 0.9083131685853004,
+      "num_tokens": 22722538.0,
+      "step": 1800
+    },
+    {
+      "entropy": 0.273828125,
+      "epoch": 1.364047484454494,
+      "grad_norm": 0.35630372166633606,
+      "learning_rate": 1.7070204728326964e-05,
+      "loss": 0.27682127952575686,
+      "mean_token_accuracy": 0.9124655604362488,
+      "num_tokens": 22857508.0,
+      "step": 1810
+    },
+    {
+      "entropy": 0.2949951171875,
+      "epoch": 1.3715846994535519,
+      "grad_norm": 0.353568434715271,
+      "learning_rate": 1.7035595160860694e-05,
+      "loss": 0.3098414897918701,
+      "mean_token_accuracy": 0.9068989664316177,
+      "num_tokens": 22982244.0,
+      "step": 1820
+    },
+    {
+      "entropy": 0.30048828125,
+      "epoch": 1.3791219144526097,
+      "grad_norm": 0.33436575531959534,
+      "learning_rate": 1.7000817905148523e-05,
+      "loss": 0.3092353820800781,
+      "mean_token_accuracy": 0.9077757328748703,
+      "num_tokens": 23115582.0,
+      "step": 1830
+    },
+    {
+      "entropy": 0.293017578125,
+      "epoch": 1.3866591294516675,
+      "grad_norm": 0.3831446170806885,
+      "learning_rate": 1.6965873790080806e-05,
+      "loss": 0.2991267442703247,
+      "mean_token_accuracy": 0.9091545164585113,
+      "num_tokens": 23244458.0,
+      "step": 1840
+    },
+    {
+      "entropy": 0.2932373046875,
+      "epoch": 1.3941963444507255,
+      "grad_norm": 0.4001041352748871,
+      "learning_rate": 1.693076364852487e-05,
+      "loss": 0.3073925018310547,
+      "mean_token_accuracy": 0.9061064407229423,
+      "num_tokens": 23370896.0,
+      "step": 1850
+    },
+    {
+      "entropy": 0.2966796875,
+      "epoch": 1.4017335594497833,
+      "grad_norm": 0.36026492714881897,
+      "learning_rate": 1.6895488317305174e-05,
+      "loss": 0.3013612747192383,
+      "mean_token_accuracy": 0.9059083923697472,
+      "num_tokens": 23490463.0,
+      "step": 1860
+    },
+    {
+      "entropy": 0.28447265625,
+      "epoch": 1.409270774448841,
+      "grad_norm": 0.4457832872867584,
+      "learning_rate": 1.6860048637183336e-05,
+      "loss": 0.2879622936248779,
+      "mean_token_accuracy": 0.910649086534977,
+      "num_tokens": 23619624.0,
+      "step": 1870
+    },
+    {
+      "epoch": 1.409270774448841,
+      "eval_entropy": 0.2866534497364458,
+      "eval_loss": 0.31162187457084656,
+      "eval_mean_token_accuracy": 0.9055151030959854,
+      "eval_num_tokens": 23619624.0,
+      "eval_runtime": 444.4582,
+      "eval_samples_per_second": 5.969,
+      "eval_steps_per_second": 1.494,
+      "step": 1870
     }
   ],
   "logging_steps": 10,
@@ -1837,7 +2018,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.6260785315176776e+18,
+  "total_flos": 1.787706633963438e+18,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null