1530

.gitattributes

@@ -13,3 +13,5 @@ global_step1190/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt filter=lfs diff=l
 global_step1190/mp_rank_00_model_states.pt filter=lfs diff=lfs merge=lfs -text
 global_step1360/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt filter=lfs diff=lfs merge=lfs -text
 global_step1360/mp_rank_00_model_states.pt filter=lfs diff=lfs merge=lfs -text
+global_step1530/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt filter=lfs diff=lfs merge=lfs -text
+global_step1530/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:bd69e81817705d82e8afdf92f2fd4d1f80619039f9c66fcabe60abdb04b73da5
+oid sha256:f94e24087b7510bd5d44a550a951fb167d26aa832c2b96c28bf68c6769038d8c
 size 335606536

{global_step1360 → global_step1530}/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:b4c5b5e2f6a09944f9c8bc6488602eee5f996ce0355110d4ee2065a778062b15
+oid sha256:08562ef8d1e27aebfdb2e4aada7b7bfba33260f497d06f6998192605adaadb93
 size 2013320965

{global_step1360 → global_step1530}/mp_rank_00_model_states.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0a1eb4d71011b3a6773b58c7f7f3be471a3d173632543bd240c8c786c49b8c97
+oid sha256:e02272081e1f4c7013026c2d3c8f9fabedcf43bd7ae766f7d8face91998fe3b3
 size 335901529

latest

@@ -1 +1 @@
-global_step1360
+global_step1530

trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 1360,
-  "best_metric": 0.32009029388427734,
-  "best_model_checkpoint": "/home/longtail/data/outputs/exaone_7b_lora_rtlcoder_ratio1.0/checkpoint-1360",
-  "epoch": 1.0248728094968909,
+  "best_global_step": 1530,
+  "best_metric": 0.31701526045799255,
+  "best_model_checkpoint": "/home/longtail/data/outputs/exaone_7b_lora_rtlcoder_ratio1.0/checkpoint-1530",
+  "epoch": 1.1530054644808743,
   "eval_steps": 170,
-  "global_step": 1360,
+  "global_step": 1530,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -1456,6 +1456,187 @@
       "eval_samples_per_second": 5.963,
       "eval_steps_per_second": 1.492,
       "step": 1360
+    },
+    {
+      "entropy": 0.30123291015625,
+      "epoch": 1.0324100244959487,
+      "grad_norm": 0.40995725989341736,
+      "learning_rate": 1.841517375569583e-05,
+      "loss": 0.30909392833709715,
+      "mean_token_accuracy": 0.9055786192417145,
+      "num_tokens": 17352049.0,
+      "step": 1370
+    },
+    {
+      "entropy": 0.3074951171875,
+      "epoch": 1.0399472394950067,
+      "grad_norm": 0.36840590834617615,
+      "learning_rate": 1.8388699329326237e-05,
+      "loss": 0.31580018997192383,
+      "mean_token_accuracy": 0.9033578932285309,
+      "num_tokens": 17479720.0,
+      "step": 1380
+    },
+    {
+      "entropy": 0.2889404296875,
+      "epoch": 1.0474844544940645,
+      "grad_norm": 0.39798107743263245,
+      "learning_rate": 1.8362024964466185e-05,
+      "loss": 0.30143420696258544,
+      "mean_token_accuracy": 0.9072484865784645,
+      "num_tokens": 17602320.0,
+      "step": 1390
+    },
+    {
+      "entropy": 0.2903564453125,
+      "epoch": 1.0550216694931223,
+      "grad_norm": 0.36563146114349365,
+      "learning_rate": 1.8335151296879576e-05,
+      "loss": 0.2945702075958252,
+      "mean_token_accuracy": 0.9097679927945137,
+      "num_tokens": 17732035.0,
+      "step": 1400
+    },
+    {
+      "entropy": 0.2938720703125,
+      "epoch": 1.06255888449218,
+      "grad_norm": 0.3827485740184784,
+      "learning_rate": 1.8308078967080547e-05,
+      "loss": 0.3062156915664673,
+      "mean_token_accuracy": 0.9074965313076973,
+      "num_tokens": 17858692.0,
+      "step": 1410
+    },
+    {
+      "entropy": 0.293212890625,
+      "epoch": 1.070096099491238,
+      "grad_norm": 0.3732523024082184,
+      "learning_rate": 1.8280808620318207e-05,
+      "loss": 0.29337191581726074,
+      "mean_token_accuracy": 0.908596222102642,
+      "num_tokens": 17980887.0,
+      "step": 1420
+    },
+    {
+      "entropy": 0.3029052734375,
+      "epoch": 1.0776333144902959,
+      "grad_norm": 0.34282541275024414,
+      "learning_rate": 1.8253340906561257e-05,
+      "loss": 0.31957981586456297,
+      "mean_token_accuracy": 0.9054985359311104,
+      "num_tokens": 18118746.0,
+      "step": 1430
+    },
+    {
+      "entropy": 0.306103515625,
+      "epoch": 1.0851705294893537,
+      "grad_norm": 0.4450572729110718,
+      "learning_rate": 1.8225676480482484e-05,
+      "loss": 0.3173269033432007,
+      "mean_token_accuracy": 0.905036324262619,
+      "num_tokens": 18242535.0,
+      "step": 1440
+    },
+    {
+      "entropy": 0.3273193359375,
+      "epoch": 1.0927077444884115,
+      "grad_norm": 0.37454572319984436,
+      "learning_rate": 1.819781600144318e-05,
+      "loss": 0.339494514465332,
+      "mean_token_accuracy": 0.8988345369696618,
+      "num_tokens": 18369940.0,
+      "step": 1450
+    },
+    {
+      "entropy": 0.2994873046875,
+      "epoch": 1.1002449594874695,
+      "grad_norm": 0.39925429224967957,
+      "learning_rate": 1.8169760133477404e-05,
+      "loss": 0.3085808277130127,
+      "mean_token_accuracy": 0.9062377795577049,
+      "num_tokens": 18510860.0,
+      "step": 1460
+    },
+    {
+      "entropy": 0.3177734375,
+      "epoch": 1.1077821744865273,
+      "grad_norm": 0.3473636209964752,
+      "learning_rate": 1.814150954527618e-05,
+      "loss": 0.33002748489379885,
+      "mean_token_accuracy": 0.9014916539192199,
+      "num_tokens": 18637891.0,
+      "step": 1470
+    },
+    {
+      "entropy": 0.2923583984375,
+      "epoch": 1.115319389485585,
+      "grad_norm": 0.41963300108909607,
+      "learning_rate": 1.8113064910171538e-05,
+      "loss": 0.30164639949798583,
+      "mean_token_accuracy": 0.9075041651725769,
+      "num_tokens": 18762929.0,
+      "step": 1480
+    },
+    {
+      "entropy": 0.304736328125,
+      "epoch": 1.1228566044846429,
+      "grad_norm": 0.4134460985660553,
+      "learning_rate": 1.8084426906120472e-05,
+      "loss": 0.312406063079834,
+      "mean_token_accuracy": 0.9049404874444008,
+      "num_tokens": 18885165.0,
+      "step": 1490
+    },
+    {
+      "entropy": 0.2976318359375,
+      "epoch": 1.1303938194837007,
+      "grad_norm": 0.44191214442253113,
+      "learning_rate": 1.8055596215688788e-05,
+      "loss": 0.3146297693252563,
+      "mean_token_accuracy": 0.9052109360694885,
+      "num_tokens": 19003706.0,
+      "step": 1500
+    },
+    {
+      "entropy": 0.27841796875,
+      "epoch": 1.1379310344827587,
+      "grad_norm": 0.43751654028892517,
+      "learning_rate": 1.802657352603483e-05,
+      "loss": 0.28258237838745115,
+      "mean_token_accuracy": 0.9103972479701042,
+      "num_tokens": 19129867.0,
+      "step": 1510
+    },
+    {
+      "entropy": 0.3037841796875,
+      "epoch": 1.1454682494818165,
+      "grad_norm": 0.4770371913909912,
+      "learning_rate": 1.7997359528893098e-05,
+      "loss": 0.31419939994812013,
+      "mean_token_accuracy": 0.9043475434184074,
+      "num_tokens": 19245782.0,
+      "step": 1520
+    },
+    {
+      "entropy": 0.3114501953125,
+      "epoch": 1.1530054644808743,
+      "grad_norm": 0.40218523144721985,
+      "learning_rate": 1.7967954920557768e-05,
+      "loss": 0.3275810956954956,
+      "mean_token_accuracy": 0.9032333821058274,
+      "num_tokens": 19375196.0,
+      "step": 1530
+    },
+    {
+      "epoch": 1.1530054644808743,
+      "eval_entropy": 0.30540668533509036,
+      "eval_loss": 0.31701526045799255,
+      "eval_mean_token_accuracy": 0.9038175200841513,
+      "eval_num_tokens": 19375196.0,
+      "eval_runtime": 444.8306,
+      "eval_samples_per_second": 5.964,
+      "eval_steps_per_second": 1.493,
+      "step": 1530
     }
   ],
   "logging_steps": 10,
@@ -1475,7 +1656,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.3052692489515827e+18,
+  "total_flos": 1.4667250088552694e+18,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null