Upload folder using huggingface_hub

.gitattributes

@@ -112,3 +112,4 @@ checkpoints-v2.0a-discrete-condition/checkpoint-20480/eval_state.json filter=lfs
 checkpoints-v2.0a-discrete-condition/checkpoint-21651/eval_state.json filter=lfs diff=lfs merge=lfs -text
 checkpoints-v2.0b-discrete-condition/checkpoint-8192/eval_state.json filter=lfs diff=lfs merge=lfs -text
 checkpoints-v2.0b-discrete-condition/checkpoint-15360/eval_state.json filter=lfs diff=lfs merge=lfs -text
+checkpoints-v2.0b-discrete-condition/checkpoint-21651/eval_state.json filter=lfs diff=lfs merge=lfs -text

checkpoints-v2.0b-discrete-condition/checkpoint-21651/eval_state.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6f652497e0a0b5c88ce02c6e1b8bb46c09294f43ba9e638d1f196be9e24d0f1
+size 44280903

checkpoints-v2.0b-discrete-condition/checkpoint-21651/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:620a031be9cdc3099f0b3c553afee563f357ef56b6441aacc2ea98e2b12fd943
+size 15210032

checkpoints-v2.0b-discrete-condition/checkpoint-21651/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:374f3740ea9b5f894d2eb7c887eafdbcf62bd4ab5a996ac6c299dee0d4c24aa0
+size 13675339

checkpoints-v2.0b-discrete-condition/checkpoint-21651/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:031698b2d035b0fc7504393c1960f3c00763dfdfabd9aa2a33d33daf359affa5
+size 14645

checkpoints-v2.0b-discrete-condition/checkpoint-21651/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c747e9fbe7804620d7663267b8a9d0d19e98f5e4b1c1e2d74ba2130e92fa5ce
+size 1383

checkpoints-v2.0b-discrete-condition/checkpoint-21651/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:70a493bf8fd6932c4ac60675b60fe483d4684704d3cab9dc59633db8961d91a7
+size 1465

checkpoints-v2.0b-discrete-condition/checkpoint-21651/trainer_state.json

@@ -0,0 +1,1216 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 1024,
+  "global_step": 21651,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.011823934229365849,
+      "grad_norm": 0.2690356373786926,
+      "learning_rate": 0.000498046875,
+      "loss": 2.567225456237793,
+      "step": 256
+    },
+    {
+      "epoch": 0.023647868458731697,
+      "grad_norm": 0.7925965785980225,
+      "learning_rate": 0.000998046875,
+      "loss": 1.9134947061538696,
+      "step": 512
+    },
+    {
+      "epoch": 0.03547180268809755,
+      "grad_norm": 1.418247103691101,
+      "learning_rate": 0.000999640996023194,
+      "loss": 1.630811333656311,
+      "step": 768
+    },
+    {
+      "epoch": 0.047295736917463395,
+      "grad_norm": 0.8778807520866394,
+      "learning_rate": 0.0009985588674043958,
+      "loss": 1.502601146697998,
+      "step": 1024
+    },
+    {
+      "epoch": 0.047295736917463395,
+      "eval_bleu": 0.0009829440445654548,
+      "eval_cos0_loss": 0.3603647586689692,
+      "eval_cos_loss": 0.3824953237488934,
+      "eval_loss": 1.4436643986397137,
+      "eval_mse0_loss": 1.4848588544484143,
+      "eval_mse_loss": 1.4004915827485525,
+      "eval_op_cos_loss": 0.3837161726875392,
+      "eval_op_mse_loss": 1.4044482999196335,
+      "step": 1024
+    },
+    {
+      "epoch": 0.047295736917463395,
+      "eval_bleu": 0.0009829440445654548,
+      "eval_cos0_loss": 0.3603647586689692,
+      "eval_cos_loss": 0.3824953237488934,
+      "eval_loss": 1.4436643986397137,
+      "eval_mse0_loss": 1.4848588544484143,
+      "eval_mse_loss": 1.4004915827485525,
+      "eval_op_cos_loss": 0.3837161726875392,
+      "eval_op_mse_loss": 1.4044482999196335,
+      "eval_runtime": 624.5636,
+      "eval_samples_per_second": 44.82,
+      "eval_steps_per_second": 0.701,
+      "step": 1024
+    },
+    {
+      "epoch": 0.05911967114682925,
+      "grad_norm": 1.361502766609192,
+      "learning_rate": 0.0009967551747861387,
+      "loss": 1.3899359703063965,
+      "step": 1280
+    },
+    {
+      "epoch": 0.0709436053761951,
+      "grad_norm": 1.701208472251892,
+      "learning_rate": 0.000994232528651847,
+      "loss": 1.308840036392212,
+      "step": 1536
+    },
+    {
+      "epoch": 0.08276753960556095,
+      "grad_norm": 1.2284104824066162,
+      "learning_rate": 0.0009909945800260092,
+      "loss": 1.2330195903778076,
+      "step": 1792
+    },
+    {
+      "epoch": 0.09459147383492679,
+      "grad_norm": 2.280683755874634,
+      "learning_rate": 0.0009870460151900522,
+      "loss": 1.163641095161438,
+      "step": 2048
+    },
+    {
+      "epoch": 0.09459147383492679,
+      "eval_bleu": 0.008117690959454648,
+      "eval_cos0_loss": 0.27556806632645053,
+      "eval_cos_loss": 0.33376140804051263,
+      "eval_loss": 1.1314381491648007,
+      "eval_mse0_loss": 1.0731379844282316,
+      "eval_mse_loss": 1.1875404857065035,
+      "eval_op_cos_loss": 0.33499729789827515,
+      "eval_op_mse_loss": 1.1919361415519019,
+      "step": 2048
+    },
+    {
+      "epoch": 0.09459147383492679,
+      "eval_bleu": 0.008117690959454648,
+      "eval_cos0_loss": 0.27556806632645053,
+      "eval_cos_loss": 0.33376140804051263,
+      "eval_loss": 1.1314381491648007,
+      "eval_mse0_loss": 1.0731379844282316,
+      "eval_mse_loss": 1.1875404857065035,
+      "eval_op_cos_loss": 0.33499729789827515,
+      "eval_op_mse_loss": 1.1919361415519019,
+      "eval_runtime": 617.7053,
+      "eval_samples_per_second": 45.318,
+      "eval_steps_per_second": 0.709,
+      "step": 2048
+    },
+    {
+      "epoch": 0.10641540806429264,
+      "grad_norm": 1.2745862007141113,
+      "learning_rate": 0.0009823925488998885,
+      "loss": 1.109260082244873,
+      "step": 2304
+    },
+    {
+      "epoch": 0.1182393422936585,
+      "grad_norm": 1.4149446487426758,
+      "learning_rate": 0.0009770409161149525,
+      "loss": 1.0698601007461548,
+      "step": 2560
+    },
+    {
+      "epoch": 0.13006327652302435,
+      "grad_norm": 1.6209443807601929,
+      "learning_rate": 0.0009709988622506973,
+      "loss": 1.0396567583084106,
+      "step": 2816
+    },
+    {
+      "epoch": 0.1418872107523902,
+      "grad_norm": 1.2111774682998657,
+      "learning_rate": 0.000964275131968659,
+      "loss": 1.0121068954467773,
+      "step": 3072
+    },
+    {
+      "epoch": 0.1418872107523902,
+      "eval_bleu": 0.033563816480858176,
+      "eval_cos0_loss": 0.241996217672139,
+      "eval_cos_loss": 0.30742351520279226,
+      "eval_loss": 1.0054655975947098,
+      "eval_mse0_loss": 0.9265603099239471,
+      "eval_mse_loss": 1.0819748758453211,
+      "eval_op_cos_loss": 0.3086226877148293,
+      "eval_op_mse_loss": 1.0867668872010219,
+      "step": 3072
+    },
+    {
+      "epoch": 0.1418872107523902,
+      "eval_bleu": 0.033563816480858176,
+      "eval_cos0_loss": 0.241996217672139,
+      "eval_cos_loss": 0.30742351520279226,
+      "eval_loss": 1.0054655975947098,
+      "eval_mse0_loss": 0.9265603099239471,
+      "eval_mse_loss": 1.0819748758453211,
+      "eval_op_cos_loss": 0.3086226877148293,
+      "eval_op_mse_loss": 1.0867668872010219,
+      "eval_runtime": 620.9592,
+      "eval_samples_per_second": 45.08,
+      "eval_steps_per_second": 0.705,
+      "step": 3072
+    },
+    {
+      "epoch": 0.15371114498175603,
+      "grad_norm": 1.7128350734710693,
+      "learning_rate": 0.0009568794565203123,
+      "loss": 0.9933720827102661,
+      "step": 3328
+    },
+    {
+      "epoch": 0.1655350792111219,
+      "grad_norm": 2.8741519451141357,
+      "learning_rate": 0.0009488225396630347,
+      "loss": 0.967034101486206,
+      "step": 3584
+    },
+    {
+      "epoch": 0.17735901344048774,
+      "grad_norm": 2.0762879848480225,
+      "learning_rate": 0.0009401160421685646,
+      "loss": 0.9411672949790955,
+      "step": 3840
+    },
+    {
+      "epoch": 0.18918294766985358,
+      "grad_norm": 1.4471893310546875,
+      "learning_rate": 0.0009307725649463714,
+      "loss": 0.9228276014328003,
+      "step": 4096
+    },
+    {
+      "epoch": 0.18918294766985358,
+      "eval_bleu": 0.052020585783013364,
+      "eval_cos0_loss": 0.21095720705784618,
+      "eval_cos_loss": 0.28916995336203816,
+      "eval_loss": 0.908736090137534,
+      "eval_mse0_loss": 0.8035404923027509,
+      "eval_mse_loss": 1.011488009261214,
+      "eval_op_cos_loss": 0.29041639147283826,
+      "eval_op_mse_loss": 1.0163753615122408,
+      "step": 4096
+    },
+    {
+      "epoch": 0.18918294766985358,
+      "eval_bleu": 0.052020585783013364,
+      "eval_cos0_loss": 0.21095720705784618,
+      "eval_cos_loss": 0.28916995336203816,
+      "eval_loss": 0.908736090137534,
+      "eval_mse0_loss": 0.8035404923027509,
+      "eval_mse_loss": 1.011488009261214,
+      "eval_op_cos_loss": 0.29041639147283826,
+      "eval_op_mse_loss": 1.0163753615122408,
+      "eval_runtime": 621.6523,
+      "eval_samples_per_second": 45.03,
+      "eval_steps_per_second": 0.705,
+      "step": 4096
+    },
+    {
+      "epoch": 0.20100688189921945,
+      "grad_norm": 2.4271538257598877,
+      "learning_rate": 0.0009208056308063659,
+      "loss": 0.9052898287773132,
+      "step": 4352
+    },
+    {
+      "epoch": 0.2128308161285853,
+      "grad_norm": 3.4062352180480957,
+      "learning_rate": 0.0009102296648873445,
+      "loss": 0.886395275592804,
+      "step": 4608
+    },
+    {
+      "epoch": 0.22465475035795113,
+      "grad_norm": 3.1074397563934326,
+      "learning_rate": 0.0008990599737794927,
+      "loss": 0.8743802905082703,
+      "step": 4864
+    },
+    {
+      "epoch": 0.236478684587317,
+      "grad_norm": 3.4324328899383545,
+      "learning_rate": 0.0008873127233711644,
+      "loss": 0.8594736456871033,
+      "step": 5120
+    },
+    {
+      "epoch": 0.236478684587317,
+      "eval_bleu": 0.07926506450551825,
+      "eval_cos0_loss": 0.20024654759119634,
+      "eval_cos_loss": 0.2791197347926767,
+      "eval_loss": 0.8700207179267657,
+      "eval_mse0_loss": 0.7612610093262642,
+      "eval_mse_loss": 0.9762748075674658,
+      "eval_op_cos_loss": 0.28037585025508655,
+      "eval_op_mse_loss": 0.9812860469839889,
+      "step": 5120
+    },
+    {
+      "epoch": 0.236478684587317,
+      "eval_bleu": 0.07926506450551825,
+      "eval_cos0_loss": 0.20024654759119634,
+      "eval_cos_loss": 0.2791197347926767,
+      "eval_loss": 0.8700207179267657,
+      "eval_mse0_loss": 0.7612610093262642,
+      "eval_mse_loss": 0.9762748075674658,
+      "eval_op_cos_loss": 0.28037585025508655,
+      "eval_op_mse_loss": 0.9812860469839889,
+      "eval_runtime": 621.0342,
+      "eval_samples_per_second": 45.075,
+      "eval_steps_per_second": 0.705,
+      "step": 5120
+    },
+    {
+      "epoch": 0.24830261881668284,
+      "grad_norm": 2.9458811283111572,
+      "learning_rate": 0.0008750049154520011,
+      "loss": 0.8498503565788269,
+      "step": 5376
+    },
+    {
+      "epoch": 0.2601265530460487,
+      "grad_norm": 3.388106107711792,
+      "learning_rate": 0.0008621543631062487,
+      "loss": 0.84010249376297,
+      "step": 5632
+    },
+    {
+      "epoch": 0.27195048727541454,
+      "grad_norm": 1.273484468460083,
+      "learning_rate": 0.0008487796649318904,
+      "loss": 0.8293308019638062,
+      "step": 5888
+    },
+    {
+      "epoch": 0.2837744215047804,
+      "grad_norm": 2.749922037124634,
+      "learning_rate": 0.0008349001781229053,
+      "loss": 0.8250018954277039,
+      "step": 6144
+    },
+    {
+      "epoch": 0.2837744215047804,
+      "eval_bleu": 0.1064852589300246,
+      "eval_cos0_loss": 0.18931846756096846,
+      "eval_cos_loss": 0.26743964381413915,
+      "eval_loss": 0.8231828152861225,
+      "eval_mse0_loss": 0.7119000969139953,
+      "eval_mse_loss": 0.9318836630751554,
+      "eval_op_cos_loss": 0.26869444518464886,
+      "eval_op_mse_loss": 0.9370474062825991,
+      "step": 6144
+    },
+    {
+      "epoch": 0.2837744215047804,
+      "eval_bleu": 0.1064852589300246,
+      "eval_cos0_loss": 0.18931846756096846,
+      "eval_cos_loss": 0.26743964381413915,
+      "eval_loss": 0.8231828152861225,
+      "eval_mse0_loss": 0.7119000969139953,
+      "eval_mse_loss": 0.9318836630751554,
+      "eval_op_cos_loss": 0.26869444518464886,
+      "eval_op_mse_loss": 0.9370474062825991,
+      "eval_runtime": 623.4472,
+      "eval_samples_per_second": 44.9,
+      "eval_steps_per_second": 0.703,
+      "step": 6144
+    },
+    {
+      "epoch": 0.2955983557341462,
+      "grad_norm": 1.9564300775527954,
+      "learning_rate": 0.0008205359904536107,
+      "loss": 0.8172543048858643,
+      "step": 6400
+    },
+    {
+      "epoch": 0.30742228996351206,
+      "grad_norm": 2.69132661819458,
+      "learning_rate": 0.0008057078912056363,
+      "loss": 0.8127719163894653,
+      "step": 6656
+    },
+    {
+      "epoch": 0.3192462241928779,
+      "grad_norm": 1.2459280490875244,
+      "learning_rate": 0.0007904373410796086,
+      "loss": 0.8043445348739624,
+      "step": 6912
+    },
+    {
+      "epoch": 0.3310701584222438,
+      "grad_norm": 1.679030179977417,
+      "learning_rate": 0.0007747464411350876,
+      "loss": 0.7973858714103699,
+      "step": 7168
+    },
+    {
+      "epoch": 0.3310701584222438,
+      "eval_bleu": 0.11764582579657883,
+      "eval_cos0_loss": 0.18387672284694567,
+      "eval_cos_loss": 0.25816574287877236,
+      "eval_loss": 0.7918145169406177,
+      "eval_mse0_loss": 0.6911438315698545,
+      "eval_mse_loss": 0.8898210989557989,
+      "eval_op_cos_loss": 0.25947807440066445,
+      "eval_op_mse_loss": 0.8951493017205364,
+      "step": 7168
+    },
+    {
+      "epoch": 0.3310701584222438,
+      "eval_bleu": 0.11764582579657883,
+      "eval_cos0_loss": 0.18387672284694567,
+      "eval_cos_loss": 0.25816574287877236,
+      "eval_loss": 0.7918145169406177,
+      "eval_mse0_loss": 0.6911438315698545,
+      "eval_mse_loss": 0.8898210989557989,
+      "eval_op_cos_loss": 0.25947807440066445,
+      "eval_op_mse_loss": 0.8951493017205364,
+      "eval_runtime": 626.051,
+      "eval_samples_per_second": 44.714,
+      "eval_steps_per_second": 0.7,
+      "step": 7168
+    },
+    {
+      "epoch": 0.34289409265160964,
+      "grad_norm": 2.451791763305664,
+      "learning_rate": 0.000758657900803716,
+      "loss": 0.7910720109939575,
+      "step": 7424
+    },
+    {
+      "epoch": 0.3547180268809755,
+      "grad_norm": 2.0209128856658936,
+      "learning_rate": 0.000742195005021869,
+      "loss": 0.7846375703811646,
+      "step": 7680
+    },
+    {
+      "epoch": 0.3665419611103413,
+      "grad_norm": 2.341242551803589,
+      "learning_rate": 0.0007253815805303786,
+      "loss": 0.7798209190368652,
+      "step": 7936
+    },
+    {
+      "epoch": 0.37836589533970716,
+      "grad_norm": 4.772252082824707,
+      "learning_rate": 0.0007082419613901028,
+      "loss": 0.7740265130996704,
+      "step": 8192
+    },
+    {
+      "epoch": 0.37836589533970716,
+      "eval_bleu": 0.11939785091250239,
+      "eval_cos0_loss": 0.18091393731636543,
+      "eval_cos_loss": 0.254198662732562,
+      "eval_loss": 0.7796027199046253,
+      "eval_mse0_loss": 0.680223688143029,
+      "eval_mse_loss": 0.8762989998135937,
+      "eval_op_cos_loss": 0.25553356053214094,
+      "eval_op_mse_loss": 0.881664497122917,
+      "step": 8192
+    },
+    {
+      "epoch": 0.37836589533970716,
+      "eval_bleu": 0.11939785091250239,
+      "eval_cos0_loss": 0.18091393731636543,
+      "eval_cos_loss": 0.254198662732562,
+      "eval_loss": 0.7796027199046253,
+      "eval_mse0_loss": 0.680223688143029,
+      "eval_mse_loss": 0.8762989998135937,
+      "eval_op_cos_loss": 0.25553356053214094,
+      "eval_op_mse_loss": 0.881664497122917,
+      "eval_runtime": 624.2884,
+      "eval_samples_per_second": 44.84,
+      "eval_steps_per_second": 0.702,
+      "step": 8192
+    },
+    {
+      "epoch": 0.390189829569073,
+      "grad_norm": 2.841437339782715,
+      "learning_rate": 0.0006908009537632514,
+      "loss": 0.7711653113365173,
+      "step": 8448
+    },
+    {
+      "epoch": 0.4020137637984389,
+      "grad_norm": 0.9838259816169739,
+      "learning_rate": 0.0006730838000114403,
+      "loss": 0.7693042755126953,
+      "step": 8704
+    },
+    {
+      "epoch": 0.41383769802780473,
+      "grad_norm": 0.9974409937858582,
+      "learning_rate": 0.0006551161421624341,
+      "loss": 0.7617323994636536,
+      "step": 8960
+    },
+    {
+      "epoch": 0.4256616322571706,
+      "grad_norm": 1.763875126838684,
+      "learning_rate": 0.0006369239847984517,
+      "loss": 0.7581734657287598,
+      "step": 9216
+    },
+    {
+      "epoch": 0.4256616322571706,
+      "eval_bleu": 0.13561947866145418,
+      "eval_cos0_loss": 0.17714670307287886,
+      "eval_cos_loss": 0.24925356851593,
+      "eval_loss": 0.7624340608511886,
+      "eval_mse0_loss": 0.664543849284246,
+      "eval_mse_loss": 0.8575849496338466,
+      "eval_op_cos_loss": 0.2506045452467927,
+      "eval_op_mse_loss": 0.8630635990127581,
+      "step": 9216
+    },
+    {
+      "epoch": 0.4256616322571706,
+      "eval_bleu": 0.13561947866145418,
+      "eval_cos0_loss": 0.17714670307287886,
+      "eval_cos_loss": 0.24925356851593,
+      "eval_loss": 0.7624340608511886,
+      "eval_mse0_loss": 0.664543849284246,
+      "eval_mse_loss": 0.8575849496338466,
+      "eval_op_cos_loss": 0.2506045452467927,
+      "eval_op_mse_loss": 0.8630635990127581,
+      "eval_runtime": 641.0279,
+      "eval_samples_per_second": 43.669,
+      "eval_steps_per_second": 0.683,
+      "step": 9216
+    },
+    {
+      "epoch": 0.4374855664865364,
+      "grad_norm": 1.7327455282211304,
+      "learning_rate": 0.0006185336574197479,
+      "loss": 0.7516188621520996,
+      "step": 9472
+    },
+    {
+      "epoch": 0.44930950071590225,
+      "grad_norm": 1.642829418182373,
+      "learning_rate": 0.0005999717763379407,
+      "loss": 0.7503851056098938,
+      "step": 9728
+    },
+    {
+      "epoch": 0.4611334349452681,
+      "grad_norm": 1.244384765625,
+      "learning_rate": 0.0005812652061542363,
+      "loss": 0.7449800968170166,
+      "step": 9984
+    },
+    {
+      "epoch": 0.472957369174634,
+      "grad_norm": 0.8096818923950195,
+      "learning_rate": 0.0005624410208783071,
+      "loss": 0.7398213148117065,
+      "step": 10240
+    },
+    {
+      "epoch": 0.472957369174634,
+      "eval_bleu": 0.14802557872775604,
+      "eval_cos0_loss": 0.17145923164473276,
+      "eval_cos_loss": 0.24269084288785447,
+      "eval_loss": 0.7362960164677607,
+      "eval_mse0_loss": 0.6412733903486435,
+      "eval_mse_loss": 0.8284871681640137,
+      "eval_op_cos_loss": 0.24404011319761407,
+      "eval_op_mse_loss": 0.8341501227252559,
+      "step": 10240
+    },
+    {
+      "epoch": 0.472957369174634,
+      "eval_bleu": 0.14802557872775604,
+      "eval_cos0_loss": 0.17145923164473276,
+      "eval_cos_loss": 0.24269084288785447,
+      "eval_loss": 0.7362960164677607,
+      "eval_mse0_loss": 0.6412733903486435,
+      "eval_mse_loss": 0.8284871681640137,
+      "eval_op_cos_loss": 0.24404011319761407,
+      "eval_op_mse_loss": 0.8341501227252559,
+      "eval_runtime": 636.3818,
+      "eval_samples_per_second": 43.988,
+      "eval_steps_per_second": 0.688,
+      "step": 10240
+    },
+    {
+      "epoch": 0.48478130340399983,
+      "grad_norm": 1.4516924619674683,
+      "learning_rate": 0.0005435264647440881,
+      "loss": 0.7356193661689758,
+      "step": 10496
+    },
+    {
+      "epoch": 0.49660523763336567,
+      "grad_norm": 1.1671243906021118,
+      "learning_rate": 0.000524548912779213,
+      "loss": 0.731113612651825,
+      "step": 10752
+    },
+    {
+      "epoch": 0.5084291718627315,
+      "grad_norm": 0.8795217871665955,
+      "learning_rate": 0.0005055358311851499,
+      "loss": 0.7303516268730164,
+      "step": 11008
+    },
+    {
+      "epoch": 0.5202531060920974,
+      "grad_norm": 1.0264520645141602,
+      "learning_rate": 0.0004865147375853812,
+      "loss": 0.7245017290115356,
+      "step": 11264
+    },
+    {
+      "epoch": 0.5202531060920974,
+      "eval_bleu": 0.15385697418347846,
+      "eval_cos0_loss": 0.1683639393753657,
+      "eval_cos_loss": 0.23805547740361463,
+      "eval_loss": 0.7221763148971888,
+      "eval_mse0_loss": 0.6295320896525362,
+      "eval_mse_loss": 0.8119698241693244,
+      "eval_op_cos_loss": 0.23943006539861905,
+      "eval_op_mse_loss": 0.817671249174092,
+      "step": 11264
+    },
+    {
+      "epoch": 0.5202531060920974,
+      "eval_bleu": 0.15385697418347846,
+      "eval_cos0_loss": 0.1683639393753657,
+      "eval_cos_loss": 0.23805547740361463,
+      "eval_loss": 0.7221763148971888,
+      "eval_mse0_loss": 0.6295320896525362,
+      "eval_mse_loss": 0.8119698241693244,
+      "eval_op_cos_loss": 0.23943006539861905,
+      "eval_op_mse_loss": 0.817671249174092,
+      "eval_runtime": 635.0392,
+      "eval_samples_per_second": 44.081,
+      "eval_steps_per_second": 0.69,
+      "step": 11264
+    },
+    {
+      "epoch": 0.5320770403214632,
+      "grad_norm": 1.1498006582260132,
+      "learning_rate": 0.0004675131611991607,
+      "loss": 0.7198587656021118,
+      "step": 11520
+    },
+    {
+      "epoch": 0.5439009745508291,
+      "grad_norm": 0.7187283635139465,
+      "learning_rate": 0.0004485586029984899,
+      "loss": 0.7185528874397278,
+      "step": 11776
+    },
+    {
+      "epoch": 0.5557249087801949,
+      "grad_norm": 3.2578821182250977,
+      "learning_rate": 0.00042967849590597266,
+      "loss": 0.7147178649902344,
+      "step": 12032
+    },
+    {
+      "epoch": 0.5675488430095608,
+      "grad_norm": 0.9559022784233093,
+      "learning_rate": 0.0004109001650911621,
+      "loss": 0.7136829495429993,
+      "step": 12288
+    },
+    {
+      "epoch": 0.5675488430095608,
+      "eval_bleu": 0.15709127331241565,
+      "eval_cos0_loss": 0.16574035799258377,
+      "eval_cos_loss": 0.2339096010819962,
+      "eval_loss": 0.7094371295682916,
+      "eval_mse0_loss": 0.6193709444237626,
+      "eval_mse_loss": 0.7965604800339703,
+      "eval_op_cos_loss": 0.23531686000900182,
+      "eval_op_mse_loss": 0.802446152113344,
+      "step": 12288
+    },
+    {
+      "epoch": 0.5675488430095608,
+      "eval_bleu": 0.15709127331241565,
+      "eval_cos0_loss": 0.16574035799258377,
+      "eval_cos_loss": 0.2339096010819962,
+      "eval_loss": 0.7094371295682916,
+      "eval_mse0_loss": 0.6193709444237626,
+      "eval_mse_loss": 0.7965604800339703,
+      "eval_op_cos_loss": 0.23531686000900182,
+      "eval_op_mse_loss": 0.802446152113344,
+      "eval_runtime": 634.9136,
+      "eval_samples_per_second": 44.089,
+      "eval_steps_per_second": 0.69,
+      "step": 12288
+    },
+    {
+      "epoch": 0.5793727772389267,
+      "grad_norm": 0.8688677549362183,
+      "learning_rate": 0.0003922507884228551,
+      "loss": 0.7095566987991333,
+      "step": 12544
+    },
+    {
+      "epoch": 0.5911967114682924,
+      "grad_norm": 1.2830818891525269,
+      "learning_rate": 0.00037375735713457723,
+      "loss": 0.7053347826004028,
+      "step": 12800
+    },
+    {
+      "epoch": 0.6030206456976583,
+      "grad_norm": 1.9164239168167114,
+      "learning_rate": 0.00035544663676018276,
+      "loss": 0.704186737537384,
+      "step": 13056
+    },
+    {
+      "epoch": 0.6148445799270241,
+      "grad_norm": 1.7377350330352783,
+      "learning_rate": 0.00033734512839611255,
+      "loss": 0.7017666101455688,
+      "step": 13312
+    },
+    {
+      "epoch": 0.6148445799270241,
+      "eval_bleu": 0.16868913199894975,
+      "eval_cos0_loss": 0.16362416999389048,
+      "eval_cos_loss": 0.23251915617620564,
+      "eval_loss": 0.701696240602563,
+      "eval_mse0_loss": 0.6112470790131451,
+      "eval_mse_loss": 0.7891615165150874,
+      "eval_op_cos_loss": 0.23392131468748936,
+      "eval_op_mse_loss": 0.7951292909987985,
+      "step": 13312
+    },
+    {
+      "epoch": 0.6148445799270241,
+      "eval_bleu": 0.16868913199894975,
+      "eval_cos0_loss": 0.16362416999389048,
+      "eval_cos_loss": 0.23251915617620564,
+      "eval_loss": 0.701696240602563,
+      "eval_mse0_loss": 0.6112470790131451,
+      "eval_mse_loss": 0.7891615165150874,
+      "eval_op_cos_loss": 0.23392131468748936,
+      "eval_op_mse_loss": 0.7951292909987985,
+      "eval_runtime": 636.8773,
+      "eval_samples_per_second": 43.954,
+      "eval_steps_per_second": 0.688,
+      "step": 13312
+    },
+    {
+      "epoch": 0.62666851415639,
+      "grad_norm": 4.23655366897583,
+      "learning_rate": 0.0003194790303463687,
+      "loss": 0.6976134777069092,
+      "step": 13568
+    },
+    {
+      "epoch": 0.6384924483857558,
+      "grad_norm": 0.7764115333557129,
+      "learning_rate": 0.00030187420020572406,
+      "loss": 0.6969444155693054,
+      "step": 13824
+    },
+    {
+      "epoch": 0.6503163826151217,
+      "grad_norm": 1.1612268686294556,
+      "learning_rate": 0.00028455611743603626,
+      "loss": 0.6931793093681335,
+      "step": 14080
+    },
+    {
+      "epoch": 0.6621403168444876,
+      "grad_norm": 0.970486044883728,
+      "learning_rate": 0.0002675498464898373,
+      "loss": 0.6915045976638794,
+      "step": 14336
+    },
+    {
+      "epoch": 0.6621403168444876,
+      "eval_bleu": 0.17463722283289504,
+      "eval_cos0_loss": 0.16116628955760504,
+      "eval_cos_loss": 0.2291939453175079,
+      "eval_loss": 0.6914327684874948,
+      "eval_mse0_loss": 0.6017627612640869,
+      "eval_mse_loss": 0.7781209033918163,
+      "eval_op_cos_loss": 0.23059950115746014,
+      "eval_op_mse_loss": 0.7840846474856547,
+      "step": 14336
+    },
+    {
+      "epoch": 0.6621403168444876,
+      "eval_bleu": 0.17463722283289504,
+      "eval_cos0_loss": 0.16116628955760504,
+      "eval_cos_loss": 0.2291939453175079,
+      "eval_loss": 0.6914327684874948,
+      "eval_mse0_loss": 0.6017627612640869,
+      "eval_mse_loss": 0.7781209033918163,
+      "eval_op_cos_loss": 0.23059950115746014,
+      "eval_op_mse_loss": 0.7840846474856547,
+      "eval_runtime": 635.4879,
+      "eval_samples_per_second": 44.05,
+      "eval_steps_per_second": 0.689,
+      "step": 14336
+    },
+    {
+      "epoch": 0.6739642510738534,
+      "grad_norm": 0.9553900957107544,
+      "learning_rate": 0.0002508800005345623,
+      "loss": 0.6907013654708862,
+      "step": 14592
+    },
+    {
+      "epoch": 0.6857881853032193,
+      "grad_norm": 0.900937020778656,
+      "learning_rate": 0.00023457070582992562,
+      "loss": 0.6890263557434082,
+      "step": 14848
+    },
+    {
+      "epoch": 0.6976121195325851,
+      "grad_norm": 1.1593620777130127,
+      "learning_rate": 0.00021864556680999692,
+      "loss": 0.6859737038612366,
+      "step": 15104
+    },
+    {
+      "epoch": 0.709436053761951,
+      "grad_norm": 1.2470667362213135,
+      "learning_rate": 0.0002031276319205152,
+      "loss": 0.6826637983322144,
+      "step": 15360
+    },
+    {
+      "epoch": 0.709436053761951,
+      "eval_bleu": 0.17806020853790655,
+      "eval_cos0_loss": 0.15932175193883513,
+      "eval_cos_loss": 0.22693043862167558,
+      "eval_loss": 0.6834767210429118,
+      "eval_mse0_loss": 0.5944504796369979,
+      "eval_mse_loss": 0.7694434859709108,
+      "eval_op_cos_loss": 0.22835160798678114,
+      "eval_op_mse_loss": 0.775562437838071,
+      "step": 15360
+    },
+    {
+      "epoch": 0.709436053761951,
+      "eval_bleu": 0.17806020853790655,
+      "eval_cos0_loss": 0.15932175193883513,
+      "eval_cos_loss": 0.22693043862167558,
+      "eval_loss": 0.6834767210429118,
+      "eval_mse0_loss": 0.5944504796369979,
+      "eval_mse_loss": 0.7694434859709108,
+      "eval_op_cos_loss": 0.22835160798678114,
+      "eval_op_mse_loss": 0.775562437838071,
+      "eval_runtime": 636.3715,
+      "eval_samples_per_second": 43.988,
+      "eval_steps_per_second": 0.688,
+      "step": 15360
+    },
+    {
+      "epoch": 0.7212599879913169,
+      "grad_norm": 0.9710851907730103,
+      "learning_rate": 0.00018803936026088542,
+      "loss": 0.6816228032112122,
+      "step": 15616
+    },
+    {
+      "epoch": 0.7330839222206826,
+      "grad_norm": 0.7932974696159363,
+      "learning_rate": 0.00017340258907913464,
+      "loss": 0.6800879836082458,
+      "step": 15872
+    },
+    {
+      "epoch": 0.7449078564500485,
+      "grad_norm": 0.9313244819641113,
+      "learning_rate": 0.0001592385021668743,
+      "loss": 0.677623450756073,
+      "step": 16128
+    },
+    {
+      "epoch": 0.7567317906794143,
+      "grad_norm": 0.7504497766494751,
+      "learning_rate": 0.0001455675992000087,
+      "loss": 0.6776302456855774,
+      "step": 16384
+    },
+    {
+      "epoch": 0.7567317906794143,
+      "eval_bleu": 0.1840349169111624,
+      "eval_cos0_loss": 0.1576695951833028,
+      "eval_cos_loss": 0.22468648969991023,
+      "eval_loss": 0.6754413573709253,
+      "eval_mse0_loss": 0.587967689587101,
+      "eval_mse_loss": 0.7598868886085406,
+      "eval_op_cos_loss": 0.22610502622034995,
+      "eval_op_mse_loss": 0.7659431604762056,
+      "step": 16384
+    },
+    {
+      "epoch": 0.7567317906794143,
+      "eval_bleu": 0.1840349169111624,
+      "eval_cos0_loss": 0.1576695951833028,
+      "eval_cos_loss": 0.22468648969991023,
+      "eval_loss": 0.6754413573709253,
+      "eval_mse0_loss": 0.587967689587101,
+      "eval_mse_loss": 0.7598868886085406,
+      "eval_op_cos_loss": 0.22610502622034995,
+      "eval_op_mse_loss": 0.7659431604762056,
+      "eval_runtime": 592.1361,
+      "eval_samples_per_second": 47.275,
+      "eval_steps_per_second": 0.74,
+      "step": 16384
+    },
+    {
+      "epoch": 0.7685557249087802,
+      "grad_norm": 0.8778237104415894,
+      "learning_rate": 0.000132409666069565,
+      "loss": 0.6726434230804443,
+      "step": 16640
+    },
+    {
+      "epoch": 0.780379659138146,
+      "grad_norm": 1.0196456909179688,
+      "learning_rate": 0.0001197837462455823,
+      "loss": 0.6718765497207642,
+      "step": 16896
+    },
+    {
+      "epoch": 0.7922035933675119,
+      "grad_norm": 0.7209889888763428,
+      "learning_rate": 0.00010770811321550749,
+      "loss": 0.671399712562561,
+      "step": 17152
+    },
+    {
+      "epoch": 0.8040275275968778,
+      "grad_norm": 0.7410993576049805,
+      "learning_rate": 9.620024403698591e-05,
+      "loss": 0.671210765838623,
+      "step": 17408
+    },
+    {
+      "epoch": 0.8040275275968778,
+      "eval_bleu": 0.1886717959007858,
+      "eval_cos0_loss": 0.1565227022726242,
+      "eval_cos_loss": 0.22315489463338026,
+      "eval_loss": 0.6706775726248685,
+      "eval_mse0_loss": 0.5836055650286478,
+      "eval_mse_loss": 0.7547007320134063,
+      "eval_op_cos_loss": 0.22457487874379442,
+      "eval_op_mse_loss": 0.7607984341442857,
+      "step": 17408
+    },
+    {
+      "epoch": 0.8040275275968778,
+      "eval_bleu": 0.1886717959007858,
+      "eval_cos0_loss": 0.1565227022726242,
+      "eval_cos_loss": 0.22315489463338026,
+      "eval_loss": 0.6706775726248685,
+      "eval_mse0_loss": 0.5836055650286478,
+      "eval_mse_loss": 0.7547007320134063,
+      "eval_op_cos_loss": 0.22457487874379442,
+      "eval_op_mse_loss": 0.7607984341442857,
+      "eval_runtime": 588.0523,
+      "eval_samples_per_second": 47.603,
+      "eval_steps_per_second": 0.745,
+      "step": 17408
+    },
+    {
+      "epoch": 0.8158514618262436,
+      "grad_norm": 0.8522108197212219,
+      "learning_rate": 8.527679404332429e-05,
+      "loss": 0.6695510149002075,
+      "step": 17664
+    },
+    {
+      "epoch": 0.8276753960556095,
+      "grad_norm": 0.7196120619773865,
+      "learning_rate": 7.495357273823544e-05,
+      "loss": 0.6697996854782104,
+      "step": 17920
+    },
+    {
+      "epoch": 0.8394993302849753,
+      "grad_norm": 0.7057328224182129,
+      "learning_rate": 6.524552091475183e-05,
+      "loss": 0.6694165468215942,
+      "step": 18176
+    },
+    {
+      "epoch": 0.8513232645143411,
+      "grad_norm": 0.6040672063827515,
+      "learning_rate": 5.6166689031422024e-05,
+      "loss": 0.6665940284729004,
+      "step": 18432
+    },
+    {
+      "epoch": 0.8513232645143411,
+      "eval_bleu": 0.19121834366216395,
+      "eval_cos0_loss": 0.15590576688040336,
+      "eval_cos_loss": 0.22263375186620782,
+      "eval_loss": 0.667947496317293,
+      "eval_mse0_loss": 0.5812215901673112,
+      "eval_mse_loss": 0.7516118535712429,
+      "eval_op_cos_loss": 0.22406025043769515,
+      "eval_op_mse_loss": 0.7577349534045615,
+      "step": 18432
+    },
+    {
+      "epoch": 0.8513232645143411,
+      "eval_bleu": 0.19121834366216395,
+      "eval_cos0_loss": 0.15590576688040336,
+      "eval_cos_loss": 0.22263375186620782,
+      "eval_loss": 0.667947496317293,
+      "eval_mse0_loss": 0.5812215901673112,
+      "eval_mse_loss": 0.7516118535712429,
+      "eval_op_cos_loss": 0.22406025043769515,
+      "eval_op_mse_loss": 0.7577349534045615,
+      "eval_runtime": 587.816,
+      "eval_samples_per_second": 47.622,
+      "eval_steps_per_second": 0.745,
+      "step": 18432
+    },
+    {
+      "epoch": 0.8631471987437069,
+      "grad_norm": 0.6761173009872437,
+      "learning_rate": 4.773021687709067e-05,
+      "loss": 0.6668224334716797,
+      "step": 18688
+    },
+    {
+      "epoch": 0.8749711329730728,
+      "grad_norm": 0.5916471481323242,
+      "learning_rate": 3.994831455368719e-05,
+      "loss": 0.6652066111564636,
+      "step": 18944
+    },
+    {
+      "epoch": 0.8867950672024387,
+      "grad_norm": 0.5401838421821594,
+      "learning_rate": 3.283224480455282e-05,
+      "loss": 0.6671025156974792,
+      "step": 19200
+    },
+    {
+      "epoch": 0.8986190014318045,
+      "grad_norm": 0.6440165638923645,
+      "learning_rate": 2.639230671387627e-05,
+      "loss": 0.6639235019683838,
+      "step": 19456
+    },
+    {
+      "epoch": 0.8986190014318045,
+      "eval_bleu": 0.1909364136393179,
+      "eval_cos0_loss": 0.1555118853370893,
+      "eval_cos_loss": 0.2215327832456593,
+      "eval_loss": 0.6657468673845405,
+      "eval_mse0_loss": 0.5797921034026908,
+      "eval_mse_loss": 0.7486564839539462,
+      "eval_op_cos_loss": 0.222952307631436,
+      "eval_op_mse_loss": 0.7547467840860967,
+      "step": 19456
+    },
+    {
+      "epoch": 0.8986190014318045,
+      "eval_bleu": 0.1909364136393179,
+      "eval_cos0_loss": 0.1555118853370893,
+      "eval_cos_loss": 0.2215327832456593,
+      "eval_loss": 0.6657468673845405,
+      "eval_mse0_loss": 0.5797921034026908,
+      "eval_mse_loss": 0.7486564839539462,
+      "eval_op_cos_loss": 0.222952307631436,
+      "eval_op_mse_loss": 0.7547467840860967,
+      "eval_runtime": 588.6982,
+      "eval_samples_per_second": 47.551,
+      "eval_steps_per_second": 0.744,
+      "step": 19456
+    },
+    {
+      "epoch": 0.9104429356611704,
+      "grad_norm": 0.6151320338249207,
+      "learning_rate": 2.063782080083576e-05,
+      "loss": 0.6646709442138672,
+      "step": 19712
+    },
+    {
+      "epoch": 0.9222668698905362,
+      "grad_norm": 0.5962440371513367,
+      "learning_rate": 1.557711553001523e-05,
+      "loss": 0.6645055413246155,
+      "step": 19968
+    },
+    {
+      "epoch": 0.9340908041199021,
+      "grad_norm": 0.5067774653434753,
+      "learning_rate": 1.1217515257622269e-05,
+      "loss": 0.6640968322753906,
+      "step": 20224
+    },
+    {
+      "epoch": 0.945914738349268,
+      "grad_norm": 0.5350014567375183,
+      "learning_rate": 7.565329630950746e-06,
+      "loss": 0.6651480197906494,
+      "step": 20480
+    },
+    {
+      "epoch": 0.945914738349268,
+      "eval_bleu": 0.19344947454126088,
+      "eval_cos0_loss": 0.15533433133336508,
+      "eval_cos_loss": 0.22171324658203342,
+      "eval_loss": 0.6652725612464017,
+      "eval_mse0_loss": 0.5790987470650781,
+      "eval_mse_loss": 0.7483866028317577,
+      "eval_op_cos_loss": 0.22313952146599828,
+      "eval_op_mse_loss": 0.7545061458463538,
+      "step": 20480
+    },
+    {
+      "epoch": 0.945914738349268,
+      "eval_bleu": 0.19344947454126088,
+      "eval_cos0_loss": 0.15533433133336508,
+      "eval_cos_loss": 0.22171324658203342,
+      "eval_loss": 0.6652725612464017,
+      "eval_mse0_loss": 0.5790987470650781,
+      "eval_mse_loss": 0.7483866028317577,
+      "eval_op_cos_loss": 0.22313952146599828,
+      "eval_op_mse_loss": 0.7545061458463538,
+      "eval_runtime": 588.7494,
+      "eval_samples_per_second": 47.547,
+      "eval_steps_per_second": 0.744,
+      "step": 20480
+    },
+    {
+      "epoch": 0.9577386725786338,
+      "grad_norm": 0.4614306390285492,
+      "learning_rate": 4.62584445643166e-06,
+      "loss": 0.6627260446548462,
+      "step": 20736
+    },
+    {
+      "epoch": 0.9695626068079997,
+      "grad_norm": 0.37359824776649475,
+      "learning_rate": 2.40331404948807e-06,
+      "loss": 0.6657180786132812,
+      "step": 20992
+    },
+    {
+      "epoch": 0.9813865410373654,
+      "grad_norm": 0.4106631875038147,
+      "learning_rate": 9.009550772663965e-07,
+      "loss": 0.6650002002716064,
+      "step": 21248
+    },
+    {
+      "epoch": 0.9932104752667313,
+      "grad_norm": 0.3390243947505951,
+      "learning_rate": 1.2094190315575791e-07,
+      "loss": 0.6650212407112122,
+      "step": 21504
+    },
+    {
+      "epoch": 0.9932104752667313,
+      "eval_bleu": 0.19311959623153893,
+      "eval_cos0_loss": 0.15524692571462562,
+      "eval_cos_loss": 0.22154003402141675,
+      "eval_loss": 0.6644643922102506,
+      "eval_mse0_loss": 0.5787450458905469,
+      "eval_mse_loss": 0.7471250993748234,
+      "eval_op_cos_loss": 0.2229611462313835,
+      "eval_op_mse_loss": 0.7532423829923481,
+      "step": 21504
+    },
+    {
+      "epoch": 0.9932104752667313,
+      "eval_bleu": 0.19311959623153893,
+      "eval_cos0_loss": 0.15524692571462562,
+      "eval_cos_loss": 0.22154003402141675,
+      "eval_loss": 0.6644643922102506,
+      "eval_mse0_loss": 0.5787450458905469,
+      "eval_mse_loss": 0.7471250993748234,
+      "eval_op_cos_loss": 0.2229611462313835,
+      "eval_op_mse_loss": 0.7532423829923481,
+      "eval_runtime": 591.4692,
+      "eval_samples_per_second": 47.328,
+      "eval_steps_per_second": 0.741,
+      "step": 21504
+    },
+    {
+      "epoch": 1.0,
+      "eval_bleu": 0.1930297042658782,
+      "eval_cos0_loss": 0.15528089971575018,
+      "eval_cos_loss": 0.22123827870305815,
+      "eval_loss": 0.6641868916276383,
+      "eval_mse0_loss": 0.5787552864584204,
+      "eval_mse_loss": 0.7465460198382808,
+      "eval_op_cos_loss": 0.2226699615178043,
+      "eval_op_mse_loss": 0.7526909737554315,
+      "step": 21651
+    },
+    {
+      "epoch": 1.0,
+      "eval_bleu": 0.1930297042658782,
+      "eval_cos0_loss": 0.15528089971575018,
+      "eval_cos_loss": 0.22123827870305815,
+      "eval_loss": 0.6641868916276383,
+      "eval_mse0_loss": 0.5787552864584204,
+      "eval_mse_loss": 0.7465460198382808,
+      "eval_op_cos_loss": 0.2226699615178043,
+      "eval_op_mse_loss": 0.7526909737554315,
+      "eval_runtime": 591.8162,
+      "eval_samples_per_second": 47.3,
+      "eval_steps_per_second": 0.74,
+      "step": 21651
+    }
+  ],
+  "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": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoints-v2.0b-discrete-condition/checkpoint-21651/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f63a8a711fb280c6320e95476c532b75c862f31d6a218a0d05079e32411a5ae1
+size 5201