Upload folder using huggingface_hub

.gitattributes

@@ -37,3 +37,4 @@ checkpoints/checkpoint-4096/eval_state.json filter=lfs diff=lfs merge=lfs -text
 checkpoints-v2/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs -text
 checkpoints-v2.1/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs -text
 checkpoints-v2.2/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs -text
+checkpoints-v2.3/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs -text

checkpoints-v2.3/checkpoint-12288/eval_state.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c4f134b68139a8d9692400b30a578d25ce628b68d1ccb63cdfe9f0629ffdd3e
+size 44107313

checkpoints-v2.3/checkpoint-12288/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:200359ef2048194a9355317b1695f656f9d71cb2c1c615c78444d90674bca430
+size 37402680

checkpoints-v2.3/checkpoint-12288/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:936ee6981444997ba9bbef30a4a344c16273ec6e70a0245baf75df9c7250fdd9
+size 512267

checkpoints-v2.3/checkpoint-12288/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f36e6c28b77555b6da6de84681647b558ac8ebc553a1b458e45112e416a213c
+size 14645

checkpoints-v2.3/checkpoint-12288/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd87ecaac101c5fe8b1b0f15753a689841cfa96e9f0b15e5cdbaf87b7829ac89
+size 1383

checkpoints-v2.3/checkpoint-12288/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:220b4c233a19883d884d3d8ad9c7134f99ecd7f2b9282b526b3c0fd015c1a12c
+size 1465

checkpoints-v2.3/checkpoint-12288/trainer_state.json

@@ -0,0 +1,838 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.5675488430095608,
+  "eval_steps": 1024,
+  "global_step": 12288,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.011823934229365849,
+      "grad_norm": 1.1381809711456299,
+      "learning_rate": 1.9615384615384617e-05,
+      "loss": 10.3904,
+      "step": 256
+    },
+    {
+      "epoch": 0.023647868458731697,
+      "grad_norm": 1.2398728132247925,
+      "learning_rate": 3.930769230769231e-05,
+      "loss": 7.9162,
+      "step": 512
+    },
+    {
+      "epoch": 0.03547180268809755,
+      "grad_norm": 1.1707866191864014,
+      "learning_rate": 4.999617095521894e-05,
+      "loss": 5.6793,
+      "step": 768
+    },
+    {
+      "epoch": 0.047295736917463395,
+      "grad_norm": 0.7125558853149414,
+      "learning_rate": 4.9961092368776736e-05,
+      "loss": 3.8181,
+      "step": 1024
+    },
+    {
+      "epoch": 0.047295736917463395,
+      "eval_acr_loss": 0.9955622624588883,
+      "eval_across_var": 0.0022213522599452947,
+      "eval_bleu": 0.5783933701616512,
+      "eval_ce_loss": 2.3272575776870936,
+      "eval_cos_loss": 0.9208345009039526,
+      "eval_cov": 0.0631596116714826,
+      "eval_cov_loss": 0.006449785340815511,
+      "eval_global_kurtosis": 3.5549826104891356,
+      "eval_global_mean": -0.0015338476934389437,
+      "eval_global_var": 0.2108087409032534,
+      "eval_loss": 2.798912787001971,
+      "eval_mse_loss": 1.9040727988225685,
+      "eval_per_var": 0.2048030487478596,
+      "eval_within_var": 0.2086275832633994,
+      "step": 1024
+    },
+    {
+      "epoch": 0.047295736917463395,
+      "eval_acr_loss": 0.9955622624588883,
+      "eval_across_var": 0.0022213522599452947,
+      "eval_bleu": 0.5783933701616512,
+      "eval_ce_loss": 2.3272575776870936,
+      "eval_cos_loss": 0.9208345009039526,
+      "eval_cov": 0.0631596116714826,
+      "eval_cov_loss": 0.006449785340815511,
+      "eval_global_kurtosis": 3.5549826104891356,
+      "eval_global_mean": -0.0015338476934389437,
+      "eval_global_var": 0.2108087409032534,
+      "eval_loss": 2.798912787001971,
+      "eval_mse_loss": 1.9040727988225685,
+      "eval_per_var": 0.2048030487478596,
+      "eval_runtime": 160.1579,
+      "eval_samples_per_second": 174.784,
+      "eval_steps_per_second": 2.735,
+      "eval_within_var": 0.2086275832633994,
+      "step": 1024
+    },
+    {
+      "epoch": 0.05911967114682925,
+      "grad_norm": 0.39322802424430847,
+      "learning_rate": 4.988941132556799e-05,
+      "loss": 2.7155,
+      "step": 1280
+    },
+    {
+      "epoch": 0.0709436053761951,
+      "grad_norm": 0.3898029327392578,
+      "learning_rate": 4.9781232937269974e-05,
+      "loss": 2.1064,
+      "step": 1536
+    },
+    {
+      "epoch": 0.08276753960556095,
+      "grad_norm": 0.2939172387123108,
+      "learning_rate": 4.963671583455164e-05,
+      "loss": 1.7128,
+      "step": 1792
+    },
+    {
+      "epoch": 0.09459147383492679,
+      "grad_norm": 0.27661314606666565,
+      "learning_rate": 4.945607193446079e-05,
+      "loss": 1.4405,
+      "step": 2048
+    },
+    {
+      "epoch": 0.09459147383492679,
+      "eval_acr_loss": 0.9952529187071814,
+      "eval_across_var": 0.0023763778212280372,
+      "eval_bleu": 0.8110985901711817,
+      "eval_ce_loss": 0.6872511814990544,
+      "eval_cos_loss": 0.7538281038199386,
+      "eval_cov": 0.06513830733625856,
+      "eval_cov_loss": 0.006921744725739235,
+      "eval_global_kurtosis": 3.977810041000854,
+      "eval_global_mean": -0.0010168063858328345,
+      "eval_global_var": 0.2243758249500571,
+      "eval_loss": 1.0947843177677834,
+      "eval_mse_loss": 1.6233446619826364,
+      "eval_per_var": 0.21794232930222604,
+      "eval_within_var": 0.22208662303887547,
+      "step": 2048
+    },
+    {
+      "epoch": 0.09459147383492679,
+      "eval_acr_loss": 0.9952529187071814,
+      "eval_across_var": 0.0023763778212280372,
+      "eval_bleu": 0.8110985901711817,
+      "eval_ce_loss": 0.6872511814990544,
+      "eval_cos_loss": 0.7538281038199386,
+      "eval_cov": 0.06513830733625856,
+      "eval_cov_loss": 0.006921744725739235,
+      "eval_global_kurtosis": 3.977810041000854,
+      "eval_global_mean": -0.0010168063858328345,
+      "eval_global_var": 0.2243758249500571,
+      "eval_loss": 1.0947843177677834,
+      "eval_mse_loss": 1.6233446619826364,
+      "eval_per_var": 0.21794232930222604,
+      "eval_runtime": 155.3139,
+      "eval_samples_per_second": 180.235,
+      "eval_steps_per_second": 2.82,
+      "eval_within_var": 0.22208662303887547,
+      "step": 2048
+    },
+    {
+      "epoch": 0.10641540806429264,
+      "grad_norm": 0.19573667645454407,
+      "learning_rate": 4.923956612967301e-05,
+      "loss": 1.2426,
+      "step": 2304
+    },
+    {
+      "epoch": 0.1182393422936585,
+      "grad_norm": 0.19026412069797516,
+      "learning_rate": 4.898751590005826e-05,
+      "loss": 1.0857,
+      "step": 2560
+    },
+    {
+      "epoch": 0.13006327652302435,
+      "grad_norm": 0.17090244591236115,
+      "learning_rate": 4.870029084713462e-05,
+      "loss": 0.9634,
+      "step": 2816
+    },
+    {
+      "epoch": 0.1418872107523902,
+      "grad_norm": 0.15054035186767578,
+      "learning_rate": 4.837831215209188e-05,
+      "loss": 0.866,
+      "step": 3072
+    },
+    {
+      "epoch": 0.1418872107523902,
+      "eval_acr_loss": 0.9946045773486568,
+      "eval_across_var": 0.002701379077032753,
+      "eval_bleu": 0.8988671192714057,
+      "eval_ce_loss": 0.32117312840402945,
+      "eval_cos_loss": 0.5974249733637457,
+      "eval_cov": 0.06391607136486872,
+      "eval_cov_loss": 0.006677505914249445,
+      "eval_global_kurtosis": 4.353016280692462,
+      "eval_global_mean": 0.0002520815150378502,
+      "eval_global_var": 0.2451829605450913,
+      "eval_loss": 0.6672432675753555,
+      "eval_mse_loss": 1.3342886499073952,
+      "eval_per_var": 0.23811486649186644,
+      "eval_within_var": 0.2426148920094586,
+      "step": 3072
+    },
+    {
+      "epoch": 0.1418872107523902,
+      "eval_acr_loss": 0.9946045773486568,
+      "eval_across_var": 0.002701379077032753,
+      "eval_bleu": 0.8988671192714057,
+      "eval_ce_loss": 0.32117312840402945,
+      "eval_cos_loss": 0.5974249733637457,
+      "eval_cov": 0.06391607136486872,
+      "eval_cov_loss": 0.006677505914249445,
+      "eval_global_kurtosis": 4.353016280692462,
+      "eval_global_mean": 0.0002520815150378502,
+      "eval_global_var": 0.2451829605450913,
+      "eval_loss": 0.6672432675753555,
+      "eval_mse_loss": 1.3342886499073952,
+      "eval_per_var": 0.23811486649186644,
+      "eval_runtime": 155.4012,
+      "eval_samples_per_second": 180.134,
+      "eval_steps_per_second": 2.819,
+      "eval_within_var": 0.2426148920094586,
+      "step": 3072
+    },
+    {
+      "epoch": 0.15371114498175603,
+      "grad_norm": 0.1389162540435791,
+      "learning_rate": 4.802205195817963e-05,
+      "loss": 0.7877,
+      "step": 3328
+    },
+    {
+      "epoch": 0.1655350792111219,
+      "grad_norm": 0.13399961590766907,
+      "learning_rate": 4.763203267836576e-05,
+      "loss": 0.7208,
+      "step": 3584
+    },
+    {
+      "epoch": 0.17735901344048774,
+      "grad_norm": 0.11527423560619354,
+      "learning_rate": 4.720882622928019e-05,
+      "loss": 0.6654,
+      "step": 3840
+    },
+    {
+      "epoch": 0.18918294766985358,
+      "grad_norm": 0.11336533725261688,
+      "learning_rate": 4.675305319256765e-05,
+      "loss": 0.6197,
+      "step": 4096
+    },
+    {
+      "epoch": 0.18918294766985358,
+      "eval_acr_loss": 0.9938907430052213,
+      "eval_across_var": 0.0030593323472410984,
+      "eval_bleu": 0.9366627287855721,
+      "eval_ce_loss": 0.1859381996333327,
+      "eval_cos_loss": 0.4787035355829213,
+      "eval_cov": 0.0628127006635274,
+      "eval_cov_loss": 0.006437324723186286,
+      "eval_global_kurtosis": 4.670551087758312,
+      "eval_global_mean": -0.0007838325141227408,
+      "eval_global_var": 0.2651386696454053,
+      "eval_loss": 0.48502123158544164,
+      "eval_mse_loss": 1.1087831166236912,
+      "eval_per_var": 0.257416468232734,
+      "eval_within_var": 0.2622626631363342,
+      "step": 4096
+    },
+    {
+      "epoch": 0.18918294766985358,
+      "eval_acr_loss": 0.9938907430052213,
+      "eval_across_var": 0.0030593323472410984,
+      "eval_bleu": 0.9366627287855721,
+      "eval_ce_loss": 0.1859381996333327,
+      "eval_cos_loss": 0.4787035355829213,
+      "eval_cov": 0.0628127006635274,
+      "eval_cov_loss": 0.006437324723186286,
+      "eval_global_kurtosis": 4.670551087758312,
+      "eval_global_mean": -0.0007838325141227408,
+      "eval_global_var": 0.2651386696454053,
+      "eval_loss": 0.48502123158544164,
+      "eval_mse_loss": 1.1087831166236912,
+      "eval_per_var": 0.257416468232734,
+      "eval_runtime": 156.5211,
+      "eval_samples_per_second": 178.845,
+      "eval_steps_per_second": 2.798,
+      "eval_within_var": 0.2622626631363342,
+      "step": 4096
+    },
+    {
+      "epoch": 0.20100688189921945,
+      "grad_norm": 0.11432790011167526,
+      "learning_rate": 4.6265381904878854e-05,
+      "loss": 0.5778,
+      "step": 4352
+    },
+    {
+      "epoch": 0.2128308161285853,
+      "grad_norm": 0.1015966460108757,
+      "learning_rate": 4.57465274778347e-05,
+      "loss": 0.5464,
+      "step": 4608
+    },
+    {
+      "epoch": 0.22465475035795113,
+      "grad_norm": 0.0999189242720604,
+      "learning_rate": 4.519725074940068e-05,
+      "loss": 0.5131,
+      "step": 4864
+    },
+    {
+      "epoch": 0.236478684587317,
+      "grad_norm": 0.09335104376077652,
+      "learning_rate": 4.461835716820895e-05,
+      "loss": 0.4865,
+      "step": 5120
+    },
+    {
+      "epoch": 0.236478684587317,
+      "eval_acr_loss": 0.993116364903646,
+      "eval_across_var": 0.0034477898343912745,
+      "eval_bleu": 0.9567762367252449,
+      "eval_ce_loss": 0.12156322401136024,
+      "eval_cos_loss": 0.39367829685069655,
+      "eval_cov": 0.06181878912938784,
+      "eval_cov_loss": 0.006211797137995629,
+      "eval_global_kurtosis": 4.93804465472426,
+      "eval_global_mean": -0.00010290959654333384,
+      "eval_global_var": 0.2838530518692922,
+      "eval_loss": 0.3869279193687657,
+      "eval_mse_loss": 0.946567648213748,
+      "eval_per_var": 0.2755494836258562,
+      "eval_within_var": 0.28062804267831043,
+      "step": 5120
+    },
+    {
+      "epoch": 0.236478684587317,
+      "eval_acr_loss": 0.993116364903646,
+      "eval_across_var": 0.0034477898343912745,
+      "eval_bleu": 0.9567762367252449,
+      "eval_ce_loss": 0.12156322401136024,
+      "eval_cos_loss": 0.39367829685069655,
+      "eval_cov": 0.06181878912938784,
+      "eval_cov_loss": 0.006211797137995629,
+      "eval_global_kurtosis": 4.93804465472426,
+      "eval_global_mean": -0.00010290959654333384,
+      "eval_global_var": 0.2838530518692922,
+      "eval_loss": 0.3869279193687657,
+      "eval_mse_loss": 0.946567648213748,
+      "eval_per_var": 0.2755494836258562,
+      "eval_runtime": 155.284,
+      "eval_samples_per_second": 180.27,
+      "eval_steps_per_second": 2.821,
+      "eval_within_var": 0.28062804267831043,
+      "step": 5120
+    },
+    {
+      "epoch": 0.24830261881668284,
+      "grad_norm": 0.0941869243979454,
+      "learning_rate": 4.401069561246422e-05,
+      "loss": 0.4632,
+      "step": 5376
+    },
+    {
+      "epoch": 0.2601265530460487,
+      "grad_norm": 0.09946911782026291,
+      "learning_rate": 4.337515714516545e-05,
+      "loss": 0.4419,
+      "step": 5632
+    },
+    {
+      "epoch": 0.27195048727541454,
+      "grad_norm": 0.1010931134223938,
+      "learning_rate": 4.2712673707468434e-05,
+      "loss": 0.4267,
+      "step": 5888
+    },
+    {
+      "epoch": 0.2837744215047804,
+      "grad_norm": 0.08404899388551712,
+      "learning_rate": 4.202421675210565e-05,
+      "loss": 0.4103,
+      "step": 6144
+    },
+    {
+      "epoch": 0.2837744215047804,
+      "eval_acr_loss": 0.9921849604066648,
+      "eval_across_var": 0.003915222487135973,
+      "eval_bleu": 0.968595516462779,
+      "eval_ce_loss": 0.08580248714445933,
+      "eval_cos_loss": 0.3347833108956411,
+      "eval_cov": 0.06079366326876427,
+      "eval_cov_loss": 0.006003422991130246,
+      "eval_global_kurtosis": 5.130746393987577,
+      "eval_global_mean": -5.609749659011353e-05,
+      "eval_global_var": 0.3024344945062785,
+      "eval_loss": 0.3278743654625601,
+      "eval_mse_loss": 0.836360767143502,
+      "eval_per_var": 0.2935495594320776,
+      "eval_within_var": 0.2987868603506045,
+      "step": 6144
+    },
+    {
+      "epoch": 0.2837744215047804,
+      "eval_acr_loss": 0.9921849604066648,
+      "eval_across_var": 0.003915222487135973,
+      "eval_bleu": 0.968595516462779,
+      "eval_ce_loss": 0.08580248714445933,
+      "eval_cos_loss": 0.3347833108956411,
+      "eval_cov": 0.06079366326876427,
+      "eval_cov_loss": 0.006003422991130246,
+      "eval_global_kurtosis": 5.130746393987577,
+      "eval_global_mean": -5.609749659011353e-05,
+      "eval_global_var": 0.3024344945062785,
+      "eval_loss": 0.3278743654625601,
+      "eval_mse_loss": 0.836360767143502,
+      "eval_per_var": 0.2935495594320776,
+      "eval_runtime": 154.5518,
+      "eval_samples_per_second": 181.124,
+      "eval_steps_per_second": 2.834,
+      "eval_within_var": 0.2987868603506045,
+      "step": 6144
+    },
+    {
+      "epoch": 0.2955983557341462,
+      "grad_norm": 0.08095081150531769,
+      "learning_rate": 4.131079581886694e-05,
+      "loss": 0.393,
+      "step": 6400
+    },
+    {
+      "epoch": 0.30742228996351206,
+      "grad_norm": 0.08812420070171356,
+      "learning_rate": 4.057345705423016e-05,
+      "loss": 0.3806,
+      "step": 6656
+    },
+    {
+      "epoch": 0.3192462241928779,
+      "grad_norm": 0.08378447592258453,
+      "learning_rate": 3.981328167731251e-05,
+      "loss": 0.3703,
+      "step": 6912
+    },
+    {
+      "epoch": 0.3310701584222438,
+      "grad_norm": 0.0990639477968216,
+      "learning_rate": 3.9031384394391954e-05,
+      "loss": 0.3564,
+      "step": 7168
+    },
+    {
+      "epoch": 0.3310701584222438,
+      "eval_acr_loss": 0.9909341404699299,
+      "eval_across_var": 0.004543303069233316,
+      "eval_bleu": 0.9765803459841506,
+      "eval_ce_loss": 0.06368093232551938,
+      "eval_cos_loss": 0.2937984631894386,
+      "eval_cov": 0.059937150511023114,
+      "eval_cov_loss": 0.005822429295255033,
+      "eval_global_kurtosis": 5.257898571284394,
+      "eval_global_mean": -0.0003388445126955912,
+      "eval_global_var": 0.3208498323344749,
+      "eval_loss": 0.28962595475046604,
+      "eval_mse_loss": 0.762596405532262,
+      "eval_per_var": 0.311440385095605,
+      "eval_within_var": 0.31661272981123295,
+      "step": 7168
+    },
+    {
+      "epoch": 0.3310701584222438,
+      "eval_acr_loss": 0.9909341404699299,
+      "eval_across_var": 0.004543303069233316,
+      "eval_bleu": 0.9765803459841506,
+      "eval_ce_loss": 0.06368093232551938,
+      "eval_cos_loss": 0.2937984631894386,
+      "eval_cov": 0.059937150511023114,
+      "eval_cov_loss": 0.005822429295255033,
+      "eval_global_kurtosis": 5.257898571284394,
+      "eval_global_mean": -0.0003388445126955912,
+      "eval_global_var": 0.3208498323344749,
+      "eval_loss": 0.28962595475046604,
+      "eval_mse_loss": 0.762596405532262,
+      "eval_per_var": 0.311440385095605,
+      "eval_runtime": 152.728,
+      "eval_samples_per_second": 183.287,
+      "eval_steps_per_second": 2.868,
+      "eval_within_var": 0.31661272981123295,
+      "step": 7168
+    },
+    {
+      "epoch": 0.34289409265160964,
+      "grad_norm": 0.07939422130584717,
+      "learning_rate": 3.822891176432382e-05,
+      "loss": 0.3491,
+      "step": 7424
+    },
+    {
+      "epoch": 0.3547180268809755,
+      "grad_norm": 0.0864938348531723,
+      "learning_rate": 3.7407040517249335e-05,
+      "loss": 0.3399,
+      "step": 7680
+    },
+    {
+      "epoch": 0.3665419611103413,
+      "grad_norm": 0.08456117659807205,
+      "learning_rate": 3.6566975829061614e-05,
+      "loss": 0.3307,
+      "step": 7936
+    },
+    {
+      "epoch": 0.37836589533970716,
+      "grad_norm": 0.06939388811588287,
+      "learning_rate": 3.5709949554159355e-05,
+      "loss": 0.3222,
+      "step": 8192
+    },
+    {
+      "epoch": 0.37836589533970716,
+      "eval_acr_loss": 0.9887979859358644,
+      "eval_across_var": 0.005616884891232943,
+      "eval_bleu": 0.9814701772398943,
+      "eval_ce_loss": 0.04945951889877178,
+      "eval_cos_loss": 0.2659321248395258,
+      "eval_cov": 0.05915525950253282,
+      "eval_cov_loss": 0.005664959631766699,
+      "eval_global_kurtosis": 5.291959479519221,
+      "eval_global_mean": 0.00022150909519631025,
+      "eval_global_var": 0.3408019183433219,
+      "eval_loss": 0.26449544942134046,
+      "eval_mse_loss": 0.7161998211248825,
+      "eval_per_var": 0.3309253710045662,
+      "eval_within_var": 0.3355388471252842,
+      "step": 8192
+    },
+    {
+      "epoch": 0.37836589533970716,
+      "eval_acr_loss": 0.9887979859358644,
+      "eval_across_var": 0.005616884891232943,
+      "eval_bleu": 0.9814701772398943,
+      "eval_ce_loss": 0.04945951889877178,
+      "eval_cos_loss": 0.2659321248395258,
+      "eval_cov": 0.05915525950253282,
+      "eval_cov_loss": 0.005664959631766699,
+      "eval_global_kurtosis": 5.291959479519221,
+      "eval_global_mean": 0.00022150909519631025,
+      "eval_global_var": 0.3408019183433219,
+      "eval_loss": 0.26449544942134046,
+      "eval_mse_loss": 0.7161998211248825,
+      "eval_per_var": 0.3309253710045662,
+      "eval_runtime": 152.3388,
+      "eval_samples_per_second": 183.755,
+      "eval_steps_per_second": 2.875,
+      "eval_within_var": 0.3355388471252842,
+      "step": 8192
+    },
+    {
+      "epoch": 0.390189829569073,
+      "grad_norm": 0.07356765121221542,
+      "learning_rate": 3.483721841907964e-05,
+      "loss": 0.3166,
+      "step": 8448
+    },
+    {
+      "epoch": 0.4020137637984389,
+      "grad_norm": 0.10812926292419434,
+      "learning_rate": 3.395006217965885e-05,
+      "loss": 0.3106,
+      "step": 8704
+    },
+    {
+      "epoch": 0.41383769802780473,
+      "grad_norm": 0.08505494147539139,
+      "learning_rate": 3.3049781744423665e-05,
+      "loss": 0.3032,
+      "step": 8960
+    },
+    {
+      "epoch": 0.4256616322571706,
+      "grad_norm": 0.07096228003501892,
+      "learning_rate": 3.213769726696439e-05,
+      "loss": 0.2986,
+      "step": 9216
+    },
+    {
+      "epoch": 0.4256616322571706,
+      "eval_acr_loss": 0.9829728993923152,
+      "eval_across_var": 0.008550605183832993,
+      "eval_bleu": 0.9851257395247456,
+      "eval_ce_loss": 0.0397455885402484,
+      "eval_cos_loss": 0.24592996208362927,
+      "eval_cov": 0.05898399875588613,
+      "eval_cov_loss": 0.005632478560929157,
+      "eval_global_kurtosis": 5.217721328343431,
+      "eval_global_mean": 0.0015772416439230584,
+      "eval_global_var": 0.3661223111087329,
+      "eval_loss": 0.24666939634982854,
+      "eval_mse_loss": 0.6856855016592975,
+      "eval_per_var": 0.355949227668379,
+      "eval_within_var": 0.3579468384180983,
+      "step": 9216
+    },
+    {
+      "epoch": 0.4256616322571706,
+      "eval_acr_loss": 0.9829728993923152,
+      "eval_across_var": 0.008550605183832993,
+      "eval_bleu": 0.9851257395247456,
+      "eval_ce_loss": 0.0397455885402484,
+      "eval_cos_loss": 0.24592996208362927,
+      "eval_cov": 0.05898399875588613,
+      "eval_cov_loss": 0.005632478560929157,
+      "eval_global_kurtosis": 5.217721328343431,
+      "eval_global_mean": 0.0015772416439230584,
+      "eval_global_var": 0.3661223111087329,
+      "eval_loss": 0.24666939634982854,
+      "eval_mse_loss": 0.6856855016592975,
+      "eval_per_var": 0.355949227668379,
+      "eval_runtime": 150.6898,
+      "eval_samples_per_second": 185.766,
+      "eval_steps_per_second": 2.907,
+      "eval_within_var": 0.3579468384180983,
+      "step": 9216
+    },
+    {
+      "epoch": 0.4374855664865364,
+      "grad_norm": 0.07909992337226868,
+      "learning_rate": 3.121514621008757e-05,
+      "loss": 0.294,
+      "step": 9472
+    },
+    {
+      "epoch": 0.44930950071590225,
+      "grad_norm": 0.10224120318889618,
+      "learning_rate": 3.0283481384586697e-05,
+      "loss": 0.2906,
+      "step": 9728
+    },
+    {
+      "epoch": 0.4611334349452681,
+      "grad_norm": 0.07880751043558121,
+      "learning_rate": 2.9344068965507027e-05,
+      "loss": 0.2855,
+      "step": 9984
+    },
+    {
+      "epoch": 0.472957369174634,
+      "grad_norm": 0.09268064051866531,
+      "learning_rate": 2.839828648881323e-05,
+      "loss": 0.2825,
+      "step": 10240
+    },
+    {
+      "epoch": 0.472957369174634,
+      "eval_acr_loss": 0.9428148437036227,
+      "eval_across_var": 0.029026934383734722,
+      "eval_bleu": 0.9874667625866159,
+      "eval_ce_loss": 0.03316931340409673,
+      "eval_cos_loss": 0.23319579296868687,
+      "eval_cov": 0.06389001297624144,
+      "eval_cov_loss": 0.0071894081860151325,
+      "eval_global_kurtosis": 4.941469875100541,
+      "eval_global_mean": 0.0006307871102198074,
+      "eval_global_var": 0.4374793762485731,
+      "eval_loss": 0.23165297341537258,
+      "eval_mse_loss": 0.6704898216680849,
+      "eval_per_var": 0.4275816477597032,
+      "eval_within_var": 0.40892007232528843,
+      "step": 10240
+    },
+    {
+      "epoch": 0.472957369174634,
+      "eval_acr_loss": 0.9428148437036227,
+      "eval_across_var": 0.029026934383734722,
+      "eval_bleu": 0.9874667625866159,
+      "eval_ce_loss": 0.03316931340409673,
+      "eval_cos_loss": 0.23319579296868687,
+      "eval_cov": 0.06389001297624144,
+      "eval_cov_loss": 0.0071894081860151325,
+      "eval_global_kurtosis": 4.941469875100541,
+      "eval_global_mean": 0.0006307871102198074,
+      "eval_global_var": 0.4374793762485731,
+      "eval_loss": 0.23165297341537258,
+      "eval_mse_loss": 0.6704898216680849,
+      "eval_per_var": 0.4275816477597032,
+      "eval_runtime": 151.8849,
+      "eval_samples_per_second": 184.304,
+      "eval_steps_per_second": 2.884,
+      "eval_within_var": 0.40892007232528843,
+      "step": 10240
+    },
+    {
+      "epoch": 0.48478130340399983,
+      "grad_norm": 0.12065292149782181,
+      "learning_rate": 2.7447520831397623e-05,
+      "loss": 0.2767,
+      "step": 10496
+    },
+    {
+      "epoch": 0.49660523763336567,
+      "grad_norm": 0.12913434207439423,
+      "learning_rate": 2.6493166177391138e-05,
+      "loss": 0.2652,
+      "step": 10752
+    },
+    {
+      "epoch": 0.5084291718627315,
+      "grad_norm": 0.14671213924884796,
+      "learning_rate": 2.5536621973758952e-05,
+      "loss": 0.2329,
+      "step": 11008
+    },
+    {
+      "epoch": 0.5202531060920974,
+      "grad_norm": 0.13414837419986725,
+      "learning_rate": 2.4579290878178904e-05,
+      "loss": 0.2016,
+      "step": 11264
+    },
+    {
+      "epoch": 0.5202531060920974,
+      "eval_acr_loss": 0.02229548650542855,
+      "eval_across_var": 0.9048832782871647,
+      "eval_bleu": 0.987608544196938,
+      "eval_ce_loss": 0.031427863691869666,
+      "eval_cos_loss": 0.2381418139490907,
+      "eval_cov": 0.05971070729434218,
+      "eval_cov_loss": 0.007483123698522715,
+      "eval_global_kurtosis": 23.39030278872137,
+      "eval_global_mean": -0.007102423879109561,
+      "eval_global_var": 1.6357934681792237,
+      "eval_loss": 0.14085906721889702,
+      "eval_mse_loss": 0.7002158846757184,
+      "eval_per_var": 1.6518420911815068,
+      "eval_within_var": 0.7377716272933298,
+      "step": 11264
+    },
+    {
+      "epoch": 0.5202531060920974,
+      "eval_acr_loss": 0.02229548650542855,
+      "eval_across_var": 0.9048832782871647,
+      "eval_bleu": 0.987608544196938,
+      "eval_ce_loss": 0.031427863691869666,
+      "eval_cos_loss": 0.2381418139490907,
+      "eval_cov": 0.05971070729434218,
+      "eval_cov_loss": 0.007483123698522715,
+      "eval_global_kurtosis": 23.39030278872137,
+      "eval_global_mean": -0.007102423879109561,
+      "eval_global_var": 1.6357934681792237,
+      "eval_loss": 0.14085906721889702,
+      "eval_mse_loss": 0.7002158846757184,
+      "eval_per_var": 1.6518420911815068,
+      "eval_runtime": 152.4278,
+      "eval_samples_per_second": 183.648,
+      "eval_steps_per_second": 2.873,
+      "eval_within_var": 0.7377716272933298,
+      "step": 11264
+    },
+    {
+      "epoch": 0.5320770403214632,
+      "grad_norm": 0.1328643560409546,
+      "learning_rate": 2.362257670221181e-05,
+      "loss": 0.1901,
+      "step": 11520
+    },
+    {
+      "epoch": 0.5439009745508291,
+      "grad_norm": 0.1011626198887825,
+      "learning_rate": 2.2667882352779608e-05,
+      "loss": 0.1844,
+      "step": 11776
+    },
+    {
+      "epoch": 0.5557249087801949,
+      "grad_norm": 0.11414045095443726,
+      "learning_rate": 2.1720315230424133e-05,
+      "loss": 0.18,
+      "step": 12032
+    },
+    {
+      "epoch": 0.5675488430095608,
+      "grad_norm": 0.1001119315624237,
+      "learning_rate": 2.0773833841855016e-05,
+      "loss": 0.1774,
+      "step": 12288
+    },
+    {
+      "epoch": 0.5675488430095608,
+      "eval_acr_loss": 0.016273215680705114,
+      "eval_across_var": 0.9544410968207877,
+      "eval_bleu": 0.9895506585988,
+      "eval_ce_loss": 0.026713626858986678,
+      "eval_cos_loss": 0.22355560778074612,
+      "eval_cov": 0.055647828263234875,
+      "eval_cov_loss": 0.006082957281973468,
+      "eval_global_kurtosis": 32.18490342353577,
+      "eval_global_mean": -0.013440034569126286,
+      "eval_global_var": 1.8210059039668949,
+      "eval_loss": 0.1298266947609649,
+      "eval_mse_loss": 0.6762152809530633,
+      "eval_per_var": 1.8439484339326484,
+      "eval_within_var": 0.8750175644545795,
+      "step": 12288
+    },
+    {
+      "epoch": 0.5675488430095608,
+      "eval_acr_loss": 0.016273215680705114,
+      "eval_across_var": 0.9544410968207877,
+      "eval_bleu": 0.9895506585988,
+      "eval_ce_loss": 0.026713626858986678,
+      "eval_cos_loss": 0.22355560778074612,
+      "eval_cov": 0.055647828263234875,
+      "eval_cov_loss": 0.006082957281973468,
+      "eval_global_kurtosis": 32.18490342353577,
+      "eval_global_mean": -0.013440034569126286,
+      "eval_global_var": 1.8210059039668949,
+      "eval_loss": 0.1298266947609649,
+      "eval_mse_loss": 0.6762152809530633,
+      "eval_per_var": 1.8439484339326484,
+      "eval_runtime": 150.5474,
+      "eval_samples_per_second": 185.941,
+      "eval_steps_per_second": 2.909,
+      "eval_within_var": 0.8750175644545795,
+      "step": 12288
+    }
+  ],
+  "logging_steps": 256,
+  "max_steps": 21651,
+  "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-v2.3/checkpoint-12288/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:267efbdef838a34be981fa3b5630732fcfd9373e6d02bbe6840e02256e0daf01
+size 5777