Upload folder using huggingface_hub

.gitattributes

@@ -56,3 +56,4 @@ checkpoints-v4.6/checkpoint-13312/eval_state.json filter=lfs diff=lfs merge=lfs
 checkpoints-v5.11-b/checkpoint-9216/eval_state.json filter=lfs diff=lfs merge=lfs -text
 checkpoints-v5.11-c/checkpoint-4096/eval_state.json filter=lfs diff=lfs merge=lfs -text
 checkpoints-v5.11-c/checkpoint-10240/eval_state.json filter=lfs diff=lfs merge=lfs -text
+checkpoints-v5.12/checkpoint-11264/eval_state.json filter=lfs diff=lfs merge=lfs -text

checkpoints-v5.12/checkpoint-11264/ema.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49447921d53c4db4a7732879b8ed6ef41aaa5f5dcd5936b849b445ed88234ab4
+size 55150648

checkpoints-v5.12/checkpoint-11264/eval_state.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:faddd8d911a462f35f87226f011181597a657945a211544fe4a9ce4aee931234
+size 60242873

checkpoints-v5.12/checkpoint-11264/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:395e2cbb78da104b0bc2dede3c66700b354d73a083dda41b441074e862954551
+size 55150680

checkpoints-v5.12/checkpoint-11264/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:00ef4f91f0a8c658831caf28f1a43bf89c2e049d126ecfaaca58e15fd1f99860
+size 77725643

checkpoints-v5.12/checkpoint-11264/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50310c5f8d737208cf5a23571aa56c041c1d6c581482c803aaf1740270054c7b
+size 14645

checkpoints-v5.12/checkpoint-11264/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ada79ea6c7ac33c12c71bea590d5ea2140ee809fed092f4670f19668dcbcc9d1
+size 1383

checkpoints-v5.12/checkpoint-11264/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21c700f8c7da85aa9d4c7cb7c4f2fe1f5cc1460165fd5b9ff6c072c94729a07e
+size 1465

checkpoints-v5.12/checkpoint-11264/trainer_state.json

@@ -0,0 +1,452 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.4854440062921542,
+  "eval_steps": 1024,
+  "global_step": 11264,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.04413127329928675,
+      "grad_norm": 1.2638587951660156,
+      "learning_rate": 9.990234375e-05,
+      "loss": 10.142199516296387,
+      "step": 1024
+    },
+    {
+      "epoch": 0.04413127329928675,
+      "eval_bleu": 0.058932490136422516,
+      "eval_ce_clean_loss": 3.8288235781289366,
+      "eval_ce_pred_loss": 5.581323942904279,
+      "eval_flow_consistency_loss": 0.19790819223755712,
+      "eval_flow_mse_loss": 0.9903420607672572,
+      "eval_loss": 7.053502456974119,
+      "flow/cos_sim": 0.6025428896519676,
+      "flow/improvement_ratio": 0.9951722414762989,
+      "flow/mag_ratio_mean": 0.5899381368144997,
+      "flow/mag_ratio_std": 0.07305362846043065,
+      "step": 1024
+    },
+    {
+      "epoch": 0.04413127329928675,
+      "eval_bleu": 0.058932490136422516,
+      "eval_ce_clean_loss": 3.8288235781289366,
+      "eval_ce_pred_loss": 5.581323942904279,
+      "eval_flow_consistency_loss": 0.19790819223755712,
+      "eval_flow_mse_loss": 0.9903420607672572,
+      "eval_loss": 7.053502456974119,
+      "eval_runtime": 210.5321,
+      "eval_samples_per_second": 142.496,
+      "eval_steps_per_second": 2.228,
+      "flow/cos_sim": 0.6025428896519676,
+      "flow/improvement_ratio": 0.9951722414762989,
+      "flow/mag_ratio_mean": 0.5899381368144997,
+      "flow/mag_ratio_std": 0.07305362846043065,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0882625465985735,
+      "grad_norm": 2.830021619796753,
+      "learning_rate": 9.9476028157316e-05,
+      "loss": 5.8488240242004395,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0882625465985735,
+      "eval_bleu": 0.25233143434843397,
+      "eval_ce_clean_loss": 1.226465684010276,
+      "eval_ce_pred_loss": 3.66637077921235,
+      "eval_flow_consistency_loss": 0.19513546225867037,
+      "eval_flow_mse_loss": 1.1100789928741293,
+      "eval_loss": 4.996664083842784,
+      "flow/cos_sim": 0.5829344555767361,
+      "flow/improvement_ratio": 0.9943469047292209,
+      "flow/mag_ratio_mean": 0.5073343900475167,
+      "flow/mag_ratio_std": 0.06693376618217049,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0882625465985735,
+      "eval_bleu": 0.25233143434843397,
+      "eval_ce_clean_loss": 1.226465684010276,
+      "eval_ce_pred_loss": 3.66637077921235,
+      "eval_flow_consistency_loss": 0.19513546225867037,
+      "eval_flow_mse_loss": 1.1100789928741293,
+      "eval_loss": 4.996664083842784,
+      "eval_runtime": 218.0358,
+      "eval_samples_per_second": 137.592,
+      "eval_steps_per_second": 2.151,
+      "flow/cos_sim": 0.5829344555767361,
+      "flow/improvement_ratio": 0.9943469047292209,
+      "flow/mag_ratio_mean": 0.5073343900475167,
+      "flow/mag_ratio_std": 0.06693376618217049,
+      "step": 2048
+    },
+    {
+      "epoch": 0.13239381989786023,
+      "grad_norm": 0.8616206645965576,
+      "learning_rate": 9.791307026072513e-05,
+      "loss": 4.643683433532715,
+      "step": 3072
+    },
+    {
+      "epoch": 0.13239381989786023,
+      "eval_bleu": 0.35994713222566843,
+      "eval_ce_clean_loss": 0.49520395553188284,
+      "eval_ce_pred_loss": 2.976453432396277,
+      "eval_flow_consistency_loss": 0.20647781791844602,
+      "eval_flow_mse_loss": 1.1147571771637972,
+      "eval_loss": 4.24396990713026,
+      "flow/cos_sim": 0.5977321162915179,
+      "flow/improvement_ratio": 0.9930047064956063,
+      "flow/mag_ratio_mean": 0.5432682163171423,
+      "flow/mag_ratio_std": 0.0777663038547105,
+      "step": 3072
+    },
+    {
+      "epoch": 0.13239381989786023,
+      "eval_bleu": 0.35994713222566843,
+      "eval_ce_clean_loss": 0.49520395553188284,
+      "eval_ce_pred_loss": 2.976453432396277,
+      "eval_flow_consistency_loss": 0.20647781791844602,
+      "eval_flow_mse_loss": 1.1147571771637972,
+      "eval_loss": 4.24396990713026,
+      "eval_runtime": 209.3799,
+      "eval_samples_per_second": 143.28,
+      "eval_steps_per_second": 2.24,
+      "flow/cos_sim": 0.5977321162915179,
+      "flow/improvement_ratio": 0.9930047064956063,
+      "flow/mag_ratio_mean": 0.5432682163171423,
+      "flow/mag_ratio_std": 0.0777663038547105,
+      "step": 3072
+    },
+    {
+      "epoch": 0.176525093197147,
+      "grad_norm": 1.9151488542556763,
+      "learning_rate": 9.53439476074686e-05,
+      "loss": 4.1653642654418945,
+      "step": 4096
+    },
+    {
+      "epoch": 0.176525093197147,
+      "eval_bleu": 0.40944445209476404,
+      "eval_ce_clean_loss": 0.25719291623086055,
+      "eval_ce_pred_loss": 2.6760470996787555,
+      "eval_flow_consistency_loss": 0.22640585670593197,
+      "eval_flow_mse_loss": 1.137088912382309,
+      "eval_loss": 3.952058234448626,
+      "flow/cos_sim": 0.6167135178915727,
+      "flow/improvement_ratio": 0.9935523163535194,
+      "flow/mag_ratio_mean": 0.5722723617228364,
+      "flow/mag_ratio_std": 0.10071539680268973,
+      "step": 4096
+    },
+    {
+      "epoch": 0.176525093197147,
+      "eval_bleu": 0.40944445209476404,
+      "eval_ce_clean_loss": 0.25719291623086055,
+      "eval_ce_pred_loss": 2.6760470996787555,
+      "eval_flow_consistency_loss": 0.22640585670593197,
+      "eval_flow_mse_loss": 1.137088912382309,
+      "eval_loss": 3.952058234448626,
+      "eval_runtime": 210.8013,
+      "eval_samples_per_second": 142.314,
+      "eval_steps_per_second": 2.225,
+      "flow/cos_sim": 0.6167135178915727,
+      "flow/improvement_ratio": 0.9935523163535194,
+      "flow/mag_ratio_mean": 0.5722723617228364,
+      "flow/mag_ratio_std": 0.10071539680268973,
+      "step": 4096
+    },
+    {
+      "epoch": 0.22065636649643372,
+      "grad_norm": 1.6371338367462158,
+      "learning_rate": 9.182261125213742e-05,
+      "loss": 3.9546115398406982,
+      "step": 5120
+    },
+    {
+      "epoch": 0.22065636649643372,
+      "eval_bleu": 0.44116786132852503,
+      "eval_ce_clean_loss": 0.15586312782408587,
+      "eval_ce_pred_loss": 2.4723648534400633,
+      "eval_flow_consistency_loss": 0.2307965545766135,
+      "eval_flow_mse_loss": 1.1533130791141535,
+      "eval_loss": 3.756662518230837,
+      "flow/cos_sim": 0.6392486124658889,
+      "flow/improvement_ratio": 0.9932308851528778,
+      "flow/mag_ratio_mean": 0.5966055151750284,
+      "flow/mag_ratio_std": 0.12516595324727772,
+      "step": 5120
+    },
+    {
+      "epoch": 0.22065636649643372,
+      "eval_bleu": 0.44116786132852503,
+      "eval_ce_clean_loss": 0.15586312782408587,
+      "eval_ce_pred_loss": 2.4723648534400633,
+      "eval_flow_consistency_loss": 0.2307965545766135,
+      "eval_flow_mse_loss": 1.1533130791141535,
+      "eval_loss": 3.756662518230837,
+      "eval_runtime": 211.3834,
+      "eval_samples_per_second": 141.922,
+      "eval_steps_per_second": 2.219,
+      "flow/cos_sim": 0.6392486124658889,
+      "flow/improvement_ratio": 0.9932308851528778,
+      "flow/mag_ratio_mean": 0.5966055151750284,
+      "flow/mag_ratio_std": 0.12516595324727772,
+      "step": 5120
+    },
+    {
+      "epoch": 0.26478763979572045,
+      "grad_norm": 1.4187636375427246,
+      "learning_rate": 8.742770483354739e-05,
+      "loss": 3.803468704223633,
+      "step": 6144
+    },
+    {
+      "epoch": 0.26478763979572045,
+      "eval_bleu": 0.4623410022491807,
+      "eval_ce_clean_loss": 0.10419444447514345,
+      "eval_ce_pred_loss": 2.367216001441484,
+      "eval_flow_consistency_loss": 0.23397522535659612,
+      "eval_flow_mse_loss": 1.1800828803576895,
+      "eval_loss": 3.6747059415398375,
+      "flow/cos_sim": 0.655801643059452,
+      "flow/improvement_ratio": 0.993431042887763,
+      "flow/mag_ratio_mean": 0.6149344101135157,
+      "flow/mag_ratio_std": 0.14765548493180955,
+      "step": 6144
+    },
+    {
+      "epoch": 0.26478763979572045,
+      "eval_bleu": 0.4623410022491807,
+      "eval_ce_clean_loss": 0.10419444447514345,
+      "eval_ce_pred_loss": 2.367216001441484,
+      "eval_flow_consistency_loss": 0.23397522535659612,
+      "eval_flow_mse_loss": 1.1800828803576895,
+      "eval_loss": 3.6747059415398375,
+      "eval_runtime": 213.1063,
+      "eval_samples_per_second": 140.775,
+      "eval_steps_per_second": 2.201,
+      "flow/cos_sim": 0.655801643059452,
+      "flow/improvement_ratio": 0.993431042887763,
+      "flow/mag_ratio_mean": 0.6149344101135157,
+      "flow/mag_ratio_std": 0.14765548493180955,
+      "step": 6144
+    },
+    {
+      "epoch": 0.30891891309500724,
+      "grad_norm": 0.977324366569519,
+      "learning_rate": 8.22483558761947e-05,
+      "loss": 3.6876096725463867,
+      "step": 7168
+    },
+    {
+      "epoch": 0.30891891309500724,
+      "eval_bleu": 0.48154751490891107,
+      "eval_ce_clean_loss": 0.07341260211204668,
+      "eval_ce_pred_loss": 2.2317311003772433,
+      "eval_flow_consistency_loss": 0.23049222644585282,
+      "eval_flow_mse_loss": 1.1846269220431476,
+      "eval_loss": 3.53894539021734,
+      "flow/cos_sim": 0.6778708653155167,
+      "flow/improvement_ratio": 0.9941845817098232,
+      "flow/mag_ratio_mean": 0.6407407661999213,
+      "flow/mag_ratio_std": 0.15953632788871652,
+      "step": 7168
+    },
+    {
+      "epoch": 0.30891891309500724,
+      "eval_bleu": 0.48154751490891107,
+      "eval_ce_clean_loss": 0.07341260211204668,
+      "eval_ce_pred_loss": 2.2317311003772433,
+      "eval_flow_consistency_loss": 0.23049222644585282,
+      "eval_flow_mse_loss": 1.1846269220431476,
+      "eval_loss": 3.53894539021734,
+      "eval_runtime": 211.8258,
+      "eval_samples_per_second": 141.626,
+      "eval_steps_per_second": 2.214,
+      "flow/cos_sim": 0.6778708653155167,
+      "flow/improvement_ratio": 0.9941845817098232,
+      "flow/mag_ratio_mean": 0.6407407661999213,
+      "flow/mag_ratio_std": 0.15953632788871652,
+      "step": 7168
+    },
+    {
+      "epoch": 0.353050186394294,
+      "grad_norm": 0.8968266844749451,
+      "learning_rate": 7.638710244802891e-05,
+      "loss": 3.5899417400360107,
+      "step": 8192
+    },
+    {
+      "epoch": 0.353050186394294,
+      "eval_bleu": 0.49368891647433405,
+      "eval_ce_clean_loss": 0.05484104749839951,
+      "eval_ce_pred_loss": 2.1758488762353276,
+      "eval_flow_consistency_loss": 0.2250279579589616,
+      "eval_flow_mse_loss": 1.2083756652976405,
+      "eval_loss": 3.5022226300066723,
+      "flow/cos_sim": 0.6907487210434383,
+      "flow/improvement_ratio": 0.9930832450832131,
+      "flow/mag_ratio_mean": 0.652735139006999,
+      "flow/mag_ratio_std": 0.16740097594794942,
+      "step": 8192
+    },
+    {
+      "epoch": 0.353050186394294,
+      "eval_bleu": 0.49368891647433405,
+      "eval_ce_clean_loss": 0.05484104749839951,
+      "eval_ce_pred_loss": 2.1758488762353276,
+      "eval_flow_consistency_loss": 0.2250279579589616,
+      "eval_flow_mse_loss": 1.2083756652976405,
+      "eval_loss": 3.5022226300066723,
+      "eval_runtime": 210.2696,
+      "eval_samples_per_second": 142.674,
+      "eval_steps_per_second": 2.23,
+      "flow/cos_sim": 0.6907487210434383,
+      "flow/improvement_ratio": 0.9930832450832131,
+      "flow/mag_ratio_mean": 0.652735139006999,
+      "flow/mag_ratio_std": 0.16740097594794942,
+      "step": 8192
+    },
+    {
+      "epoch": 0.3971814596935807,
+      "grad_norm": 1.9928840398788452,
+      "learning_rate": 6.997172522088177e-05,
+      "loss": 3.5251195430755615,
+      "step": 9216
+    },
+    {
+      "epoch": 0.3971814596935807,
+      "eval_bleu": 0.5060081279142332,
+      "eval_ce_clean_loss": 0.043055545256685604,
+      "eval_ce_pred_loss": 2.106957233282549,
+      "eval_flow_consistency_loss": 0.22306220640124544,
+      "eval_flow_mse_loss": 1.2162765316617514,
+      "eval_loss": 3.4390704296291004,
+      "flow/cos_sim": 0.7040034491878583,
+      "flow/improvement_ratio": 0.9930018180214775,
+      "flow/mag_ratio_mean": 0.6716202778348537,
+      "flow/mag_ratio_std": 0.17294156087486984,
+      "step": 9216
+    },
+    {
+      "epoch": 0.3971814596935807,
+      "eval_bleu": 0.5060081279142332,
+      "eval_ce_clean_loss": 0.043055545256685604,
+      "eval_ce_pred_loss": 2.106957233282549,
+      "eval_flow_consistency_loss": 0.22306220640124544,
+      "eval_flow_mse_loss": 1.2162765316617514,
+      "eval_loss": 3.4390704296291004,
+      "eval_runtime": 209.3455,
+      "eval_samples_per_second": 143.304,
+      "eval_steps_per_second": 2.24,
+      "flow/cos_sim": 0.7040034491878583,
+      "flow/improvement_ratio": 0.9930018180214775,
+      "flow/mag_ratio_mean": 0.6716202778348537,
+      "flow/mag_ratio_std": 0.17294156087486984,
+      "step": 9216
+    },
+    {
+      "epoch": 0.44131273299286744,
+      "grad_norm": 1.3912804126739502,
+      "learning_rate": 6.314377890922702e-05,
+      "loss": 3.487394332885742,
+      "step": 10240
+    },
+    {
+      "epoch": 0.44131273299286744,
+      "eval_bleu": 0.5124308712326037,
+      "eval_ce_clean_loss": 0.034350075578289244,
+      "eval_ce_pred_loss": 2.0600304321439538,
+      "eval_flow_consistency_loss": 0.21802372556886693,
+      "eval_flow_mse_loss": 1.2173504947599318,
+      "eval_loss": 3.3898277974077886,
+      "flow/cos_sim": 0.7159046165978731,
+      "flow/improvement_ratio": 0.9939875282458405,
+      "flow/mag_ratio_mean": 0.6801346318045659,
+      "flow/mag_ratio_std": 0.17366625556051096,
+      "step": 10240
+    },
+    {
+      "epoch": 0.44131273299286744,
+      "eval_bleu": 0.5124308712326037,
+      "eval_ce_clean_loss": 0.034350075578289244,
+      "eval_ce_pred_loss": 2.0600304321439538,
+      "eval_flow_consistency_loss": 0.21802372556886693,
+      "eval_flow_mse_loss": 1.2173504947599318,
+      "eval_loss": 3.3898277974077886,
+      "eval_runtime": 209.8004,
+      "eval_samples_per_second": 142.993,
+      "eval_steps_per_second": 2.235,
+      "flow/cos_sim": 0.7159046165978731,
+      "flow/improvement_ratio": 0.9939875282458405,
+      "flow/mag_ratio_mean": 0.6801346318045659,
+      "flow/mag_ratio_std": 0.17366625556051096,
+      "step": 10240
+    },
+    {
+      "epoch": 0.4854440062921542,
+      "grad_norm": 1.2457709312438965,
+      "learning_rate": 5.603332356428589e-05,
+      "loss": 3.454292058944702,
+      "step": 11264
+    },
+    {
+      "epoch": 0.4854440062921542,
+      "eval_bleu": 0.5169710884245263,
+      "eval_ce_clean_loss": 0.0282955012286205,
+      "eval_ce_pred_loss": 2.015394574289383,
+      "eval_flow_consistency_loss": 0.2058143128972572,
+      "eval_flow_mse_loss": 1.222560783693277,
+      "eval_loss": 3.343692081060999,
+      "flow/cos_sim": 0.7249747821008727,
+      "flow/improvement_ratio": 0.9951503481437911,
+      "flow/mag_ratio_mean": 0.692842490891658,
+      "flow/mag_ratio_std": 0.17570793422173336,
+      "step": 11264
+    },
+    {
+      "epoch": 0.4854440062921542,
+      "eval_bleu": 0.5169710884245263,
+      "eval_ce_clean_loss": 0.0282955012286205,
+      "eval_ce_pred_loss": 2.015394574289383,
+      "eval_flow_consistency_loss": 0.2058143128972572,
+      "eval_flow_mse_loss": 1.222560783693277,
+      "eval_loss": 3.343692081060999,
+      "eval_runtime": 208.4095,
+      "eval_samples_per_second": 143.947,
+      "eval_steps_per_second": 2.25,
+      "flow/cos_sim": 0.7249747821008727,
+      "flow/improvement_ratio": 0.9951503481437911,
+      "flow/mag_ratio_mean": 0.692842490891658,
+      "flow/mag_ratio_std": 0.17570793422173336,
+      "step": 11264
+    }
+  ],
+  "logging_steps": 1024,
+  "max_steps": 23204,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1024,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoints-v5.12/checkpoint-11264/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8469bbc81a2ba0be2f5b44007faafd15c75615abe30f4f4e56171816d31caa5b
+size 5137