Upload folder using huggingface_hub

.gitattributes

@@ -49,3 +49,4 @@ checkpoints-v5.8/checkpoint-10240/eval_state.json filter=lfs diff=lfs merge=lfs
 checkpoints-v4.5/checkpoint-7168/eval_state.json filter=lfs diff=lfs merge=lfs -text
 checkpoints-v4.5/checkpoint-15360/eval_state.json filter=lfs diff=lfs merge=lfs -text
 checkpoints-v5.9/checkpoint-11264/eval_state.json filter=lfs diff=lfs merge=lfs -text
+checkpoints-v5.10/checkpoint-7168/eval_state.json filter=lfs diff=lfs merge=lfs -text

checkpoints-v5.10/checkpoint-7168/ema.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:859b524c6f08cb8e4de0cdeac3d1078754a838adbe2bb2b52253d00d572d8949
+size 55150648

checkpoints-v5.10/checkpoint-7168/eval_state.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6172bae3b0d393a89c588da917a75b35b245d09483a6f2e1d40162c71c7a1b08
+size 58881754

checkpoints-v5.10/checkpoint-7168/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d73ce65202c4a4403f56015696a3586e66b41a2df367b58ac4eab6399e1b7796
+size 55150680

checkpoints-v5.10/checkpoint-7168/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0191582a7b4e76bacb808a998eca18354df820c14390b9ea35ba5b5a749a826b
+size 77724619

checkpoints-v5.10/checkpoint-7168/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:179c71fcead7f0f1a71ca3f32ded1de0ad5b7d83affaf2bafaa286d4b4ab393f
+size 14645

checkpoints-v5.10/checkpoint-7168/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d595dc04756955b718dbe40e12e3b42e9a74ec09bbdeec39a22714665de3cd13
+size 1383

checkpoints-v5.10/checkpoint-7168/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59ae6fdfd0cae3f8c7c6f88ba50d2b929ba08e8f73177205bb5416fc134c7f7b
+size 1465

checkpoints-v5.10/checkpoint-7168/trainer_state.json

@@ -0,0 +1,286 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.30891891309500724,
+  "eval_steps": 1024,
+  "global_step": 7168,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.04413127329928675,
+      "grad_norm": 0.7938550114631653,
+      "learning_rate": 9.990234375e-05,
+      "loss": 12.336601257324219,
+      "step": 1024
+    },
+    {
+      "epoch": 0.04413127329928675,
+      "eval_bleu": 0.16407143517604253,
+      "eval_ce_clean_loss": 1.724545789679993,
+      "eval_ce_pred_loss": 5.058502770690267,
+      "eval_flow_mse_loss": 0.929569506823127,
+      "eval_loss": 6.195067207442164,
+      "flow/cos_sim": 0.055166260543853235,
+      "flow/improvement_ratio": 0.596057767425773,
+      "flow/mag_ratio_mean": 0.4713213621680416,
+      "flow/mag_ratio_std": 0.4101032533371118,
+      "step": 1024
+    },
+    {
+      "epoch": 0.04413127329928675,
+      "eval_bleu": 0.16407143517604253,
+      "eval_ce_clean_loss": 1.724545789679993,
+      "eval_ce_pred_loss": 5.058502770690267,
+      "eval_flow_mse_loss": 0.929569506823127,
+      "eval_loss": 6.195067207442164,
+      "eval_runtime": 212.5631,
+      "eval_samples_per_second": 141.135,
+      "eval_steps_per_second": 2.206,
+      "flow/cos_sim": 0.055166260543853235,
+      "flow/improvement_ratio": 0.596057767425773,
+      "flow/mag_ratio_mean": 0.4713213621680416,
+      "flow/mag_ratio_std": 0.4101032533371118,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0882625465985735,
+      "grad_norm": 0.450653612613678,
+      "learning_rate": 9.9476028157316e-05,
+      "loss": 4.612443923950195,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0882625465985735,
+      "eval_bleu": 0.3321661463154056,
+      "eval_ce_clean_loss": 0.2987517699885216,
+      "eval_ce_pred_loss": 3.625808963389285,
+      "eval_flow_mse_loss": 1.015768651387839,
+      "eval_loss": 3.8525866531868225,
+      "flow/cos_sim": 0.1616891078921015,
+      "flow/improvement_ratio": 0.7423238582702588,
+      "flow/mag_ratio_mean": 0.6468003783017587,
+      "flow/mag_ratio_std": 0.5354506893834071,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0882625465985735,
+      "eval_bleu": 0.3321661463154056,
+      "eval_ce_clean_loss": 0.2987517699885216,
+      "eval_ce_pred_loss": 3.625808963389285,
+      "eval_flow_mse_loss": 1.015768651387839,
+      "eval_loss": 3.8525866531868225,
+      "eval_runtime": 208.1773,
+      "eval_samples_per_second": 144.108,
+      "eval_steps_per_second": 2.253,
+      "flow/cos_sim": 0.1616891078921015,
+      "flow/improvement_ratio": 0.7423238582702588,
+      "flow/mag_ratio_mean": 0.6468003783017587,
+      "flow/mag_ratio_std": 0.5354506893834071,
+      "step": 2048
+    },
+    {
+      "epoch": 0.13239381989786023,
+      "grad_norm": 0.37855076789855957,
+      "learning_rate": 9.791307026072513e-05,
+      "loss": 3.617213010787964,
+      "step": 3072
+    },
+    {
+      "epoch": 0.13239381989786023,
+      "eval_bleu": 0.3836053070830755,
+      "eval_ce_clean_loss": 0.12278842976860908,
+      "eval_ce_pred_loss": 3.1099086214484437,
+      "eval_flow_mse_loss": 1.0722369169121357,
+      "eval_loss": 3.3719613491090885,
+      "flow/cos_sim": 0.22163581698815196,
+      "flow/improvement_ratio": 0.8109787233602772,
+      "flow/mag_ratio_mean": 0.6704668670829171,
+      "flow/mag_ratio_std": 0.4992133626805694,
+      "step": 3072
+    },
+    {
+      "epoch": 0.13239381989786023,
+      "eval_bleu": 0.3836053070830755,
+      "eval_ce_clean_loss": 0.12278842976860908,
+      "eval_ce_pred_loss": 3.1099086214484437,
+      "eval_flow_mse_loss": 1.0722369169121357,
+      "eval_loss": 3.3719613491090885,
+      "eval_runtime": 209.1152,
+      "eval_samples_per_second": 143.462,
+      "eval_steps_per_second": 2.243,
+      "flow/cos_sim": 0.22163581698815196,
+      "flow/improvement_ratio": 0.8109787233602772,
+      "flow/mag_ratio_mean": 0.6704668670829171,
+      "flow/mag_ratio_std": 0.4992133626805694,
+      "step": 3072
+    },
+    {
+      "epoch": 0.176525093197147,
+      "grad_norm": 0.4389702081680298,
+      "learning_rate": 9.53439476074686e-05,
+      "loss": 3.2931454181671143,
+      "step": 4096
+    },
+    {
+      "epoch": 0.176525093197147,
+      "eval_bleu": 0.40933125434293716,
+      "eval_ce_clean_loss": 0.06711880822997612,
+      "eval_ce_pred_loss": 2.8558058159183592,
+      "eval_flow_mse_loss": 1.081392692604553,
+      "eval_loss": 3.147575543125047,
+      "flow/cos_sim": 0.25172679529769587,
+      "flow/improvement_ratio": 0.8514864339248966,
+      "flow/mag_ratio_mean": 0.6699917584594125,
+      "flow/mag_ratio_std": 0.44682612055654464,
+      "step": 4096
+    },
+    {
+      "epoch": 0.176525093197147,
+      "eval_bleu": 0.40933125434293716,
+      "eval_ce_clean_loss": 0.06711880822997612,
+      "eval_ce_pred_loss": 2.8558058159183592,
+      "eval_flow_mse_loss": 1.081392692604553,
+      "eval_loss": 3.147575543125047,
+      "eval_runtime": 210.459,
+      "eval_samples_per_second": 142.546,
+      "eval_steps_per_second": 2.228,
+      "flow/cos_sim": 0.25172679529769587,
+      "flow/improvement_ratio": 0.8514864339248966,
+      "flow/mag_ratio_mean": 0.6699917584594125,
+      "flow/mag_ratio_std": 0.44682612055654464,
+      "step": 4096
+    },
+    {
+      "epoch": 0.22065636649643372,
+      "grad_norm": 0.44552379846572876,
+      "learning_rate": 9.182261125213742e-05,
+      "loss": 3.127476692199707,
+      "step": 5120
+    },
+    {
+      "epoch": 0.22065636649643372,
+      "eval_bleu": 0.42390360625279244,
+      "eval_ce_clean_loss": 0.04179192618377554,
+      "eval_ce_pred_loss": 2.713604238495898,
+      "eval_flow_mse_loss": 1.0599637749606867,
+      "eval_loss": 3.00127863375617,
+      "flow/cos_sim": 0.25444072618413327,
+      "flow/improvement_ratio": 0.8746775842424649,
+      "flow/mag_ratio_mean": 0.671740317395501,
+      "flow/mag_ratio_std": 0.40047251164659,
+      "step": 5120
+    },
+    {
+      "epoch": 0.22065636649643372,
+      "eval_bleu": 0.42390360625279244,
+      "eval_ce_clean_loss": 0.04179192618377554,
+      "eval_ce_pred_loss": 2.713604238495898,
+      "eval_flow_mse_loss": 1.0599637749606867,
+      "eval_loss": 3.00127863375617,
+      "eval_runtime": 211.2198,
+      "eval_samples_per_second": 142.032,
+      "eval_steps_per_second": 2.22,
+      "flow/cos_sim": 0.25444072618413327,
+      "flow/improvement_ratio": 0.8746775842424649,
+      "flow/mag_ratio_mean": 0.671740317395501,
+      "flow/mag_ratio_std": 0.40047251164659,
+      "step": 5120
+    },
+    {
+      "epoch": 0.26478763979572045,
+      "grad_norm": 0.4731499254703522,
+      "learning_rate": 8.742300854391668e-05,
+      "loss": 3.012479782104492,
+      "step": 6144
+    },
+    {
+      "epoch": 0.26478763979572045,
+      "eval_bleu": 0.4346563578061454,
+      "eval_ce_clean_loss": 0.028318230050808586,
+      "eval_ce_pred_loss": 2.6327081789085858,
+      "eval_flow_mse_loss": 1.056245713854141,
+      "eval_loss": 2.927459636985112,
+      "flow/cos_sim": 0.2519478304808074,
+      "flow/improvement_ratio": 0.8869821624969368,
+      "flow/mag_ratio_mean": 0.6784741002867725,
+      "flow/mag_ratio_std": 0.37580890851869764,
+      "step": 6144
+    },
+    {
+      "epoch": 0.26478763979572045,
+      "eval_bleu": 0.4346563578061454,
+      "eval_ce_clean_loss": 0.028318230050808586,
+      "eval_ce_pred_loss": 2.6327081789085858,
+      "eval_flow_mse_loss": 1.056245713854141,
+      "eval_loss": 2.927459636985112,
+      "eval_runtime": 211.7504,
+      "eval_samples_per_second": 141.676,
+      "eval_steps_per_second": 2.215,
+      "flow/cos_sim": 0.2519478304808074,
+      "flow/improvement_ratio": 0.8869821624969368,
+      "flow/mag_ratio_mean": 0.6784741002867725,
+      "flow/mag_ratio_std": 0.37580890851869764,
+      "step": 6144
+    },
+    {
+      "epoch": 0.30891891309500724,
+      "grad_norm": 0.6240633726119995,
+      "learning_rate": 8.223753024725232e-05,
+      "loss": 2.9197511672973633,
+      "step": 7168
+    },
+    {
+      "epoch": 0.30891891309500724,
+      "eval_bleu": 0.44790457353944046,
+      "eval_ce_clean_loss": 0.020565879328656934,
+      "eval_ce_pred_loss": 2.529554044768246,
+      "eval_flow_mse_loss": 1.036346342009522,
+      "eval_loss": 2.8276000312650638,
+      "flow/cos_sim": 0.24807281547517918,
+      "flow/improvement_ratio": 0.8986854375298343,
+      "flow/mag_ratio_mean": 0.6740282399059613,
+      "flow/mag_ratio_std": 0.3316722640287139,
+      "step": 7168
+    },
+    {
+      "epoch": 0.30891891309500724,
+      "eval_bleu": 0.44790457353944046,
+      "eval_ce_clean_loss": 0.020565879328656934,
+      "eval_ce_pred_loss": 2.529554044768246,
+      "eval_flow_mse_loss": 1.036346342009522,
+      "eval_loss": 2.8276000312650638,
+      "eval_runtime": 211.5611,
+      "eval_samples_per_second": 141.803,
+      "eval_steps_per_second": 2.217,
+      "flow/cos_sim": 0.24807281547517918,
+      "flow/improvement_ratio": 0.8986854375298343,
+      "flow/mag_ratio_mean": 0.6740282399059613,
+      "flow/mag_ratio_std": 0.3316722640287139,
+      "step": 7168
+    }
+  ],
+  "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-v5.10/checkpoint-7168/training_args.bin

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