Upload folder using huggingface_hub
Browse files- .gitattributes +1 -0
- checkpoints-d1.1/checkpoint-11264/ema.safetensors +3 -0
- checkpoints-d1.1/checkpoint-11264/eval_state.json +3 -0
- checkpoints-d1.1/checkpoint-11264/model.safetensors +3 -0
- checkpoints-d1.1/checkpoint-11264/optimizer.pt +3 -0
- checkpoints-d1.1/checkpoint-11264/rng_state.pth +3 -0
- checkpoints-d1.1/checkpoint-11264/scaler.pt +3 -0
- checkpoints-d1.1/checkpoint-11264/scheduler.pt +3 -0
- checkpoints-d1.1/checkpoint-11264/trainer_state.json +727 -0
- checkpoints-d1.1/checkpoint-11264/training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -74,3 +74,4 @@ checkpoints-v5.15/checkpoint-11264/eval_state.json filter=lfs diff=lfs merge=lfs
|
|
| 74 |
checkpoints-v4.7/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 75 |
checkpoints-v4.7/checkpoint-21504/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 76 |
checkpoints-d1.0/checkpoint-10240/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 74 |
checkpoints-v4.7/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 75 |
checkpoints-v4.7/checkpoint-21504/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 76 |
checkpoints-d1.0/checkpoint-10240/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 77 |
+
checkpoints-d1.1/checkpoint-11264/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
checkpoints-d1.1/checkpoint-11264/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:23e63d448e361121d7872532d98f47e7e52a6fd6ce2a62032928dc596aa1a095
|
| 3 |
+
size 746712
|
checkpoints-d1.1/checkpoint-11264/eval_state.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1277638d7496d99a91d95fa2f44fe42737fdbaba1aab4182cf93080471ea57b8
|
| 3 |
+
size 57292057
|
checkpoints-d1.1/checkpoint-11264/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:94d0ceb939d87027a12516dec601816c2d8c459b94f6a24d7fcf0309372977b4
|
| 3 |
+
size 24002016
|
checkpoints-d1.1/checkpoint-11264/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bd82bb557889ae3238f2b338685b12aeeeb1c01e3e6a66696e34878d50e2114e
|
| 3 |
+
size 1569995
|
checkpoints-d1.1/checkpoint-11264/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b7cfc61462969f5ba350924264a7a5144a5ddf7261e3539e2bdf80a4f26cc8ee
|
| 3 |
+
size 14645
|
checkpoints-d1.1/checkpoint-11264/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:04050294348d6e04878694718ea868fb71c948a4cc1b6bcc5bf74a159d253de6
|
| 3 |
+
size 1383
|
checkpoints-d1.1/checkpoint-11264/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ccf022a1be483726c969dc5e04a5e0174513bbea4e2812e09e97b7093b3e4e2e
|
| 3 |
+
size 1465
|
checkpoints-d1.1/checkpoint-11264/trainer_state.json
ADDED
|
@@ -0,0 +1,727 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.4854440062921542,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 11264,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011032818324821687,
|
| 14 |
+
"grad_norm": 0.10309942811727524,
|
| 15 |
+
"learning_rate": 0.000498046875,
|
| 16 |
+
"loss": 1.9074174165725708,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.022065636649643373,
|
| 21 |
+
"grad_norm": 0.2910110056400299,
|
| 22 |
+
"learning_rate": 0.000998046875,
|
| 23 |
+
"loss": 1.5273144245147705,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03309845497446506,
|
| 28 |
+
"grad_norm": 0.3859289586544037,
|
| 29 |
+
"learning_rate": 0.000999688448778502,
|
| 30 |
+
"loss": 1.3800736665725708,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.04413127329928675,
|
| 35 |
+
"grad_norm": 0.5722110867500305,
|
| 36 |
+
"learning_rate": 0.0009987492950653055,
|
| 37 |
+
"loss": 1.342606544494629,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.04413127329928675,
|
| 42 |
+
"eval_bleu": 0.9366650964401493,
|
| 43 |
+
"eval_cos_loss": 0.4710617309440174,
|
| 44 |
+
"eval_dec_loss": 0.11786629736169314,
|
| 45 |
+
"eval_loss": 1.3323029561845987,
|
| 46 |
+
"eval_mse2_loss": 0.1665979178824913,
|
| 47 |
+
"eval_mse_loss": 1.3323029561845987,
|
| 48 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 49 |
+
"eval_var_loss": 0.01723895594080501,
|
| 50 |
+
"flow/cos_sim": 0.5289382661329404,
|
| 51 |
+
"flow/improvement_ratio": 0.8936813888010948,
|
| 52 |
+
"flow/mag_ratio_mean": 0.5435932263382462,
|
| 53 |
+
"flow/mag_ratio_std": 0.2489985737210906,
|
| 54 |
+
"step": 1024
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.04413127329928675,
|
| 58 |
+
"eval_bleu": 0.9366650964401493,
|
| 59 |
+
"eval_cos_loss": 0.4710617309440174,
|
| 60 |
+
"eval_dec_loss": 0.11786629736169314,
|
| 61 |
+
"eval_loss": 1.3323029561845987,
|
| 62 |
+
"eval_mse2_loss": 0.1665979178824913,
|
| 63 |
+
"eval_mse_loss": 1.3323029561845987,
|
| 64 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 65 |
+
"eval_runtime": 157.3375,
|
| 66 |
+
"eval_samples_per_second": 190.673,
|
| 67 |
+
"eval_steps_per_second": 2.981,
|
| 68 |
+
"eval_var_loss": 0.01723895594080501,
|
| 69 |
+
"flow/cos_sim": 0.5289382661329404,
|
| 70 |
+
"flow/improvement_ratio": 0.8936813888010948,
|
| 71 |
+
"flow/mag_ratio_mean": 0.5435932263382462,
|
| 72 |
+
"flow/mag_ratio_std": 0.2489985737210906,
|
| 73 |
+
"step": 1024
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.05516409162410843,
|
| 77 |
+
"grad_norm": 0.6506242752075195,
|
| 78 |
+
"learning_rate": 0.0009971837136430763,
|
| 79 |
+
"loss": 1.3261979818344116,
|
| 80 |
+
"step": 1280
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.06619690994893011,
|
| 84 |
+
"grad_norm": 0.6324401497840881,
|
| 85 |
+
"learning_rate": 0.0009949936708776692,
|
| 86 |
+
"loss": 1.3123514652252197,
|
| 87 |
+
"step": 1536
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.07722972827375181,
|
| 91 |
+
"grad_norm": 1.1031574010849,
|
| 92 |
+
"learning_rate": 0.0009921819174566252,
|
| 93 |
+
"loss": 1.3050185441970825,
|
| 94 |
+
"step": 1792
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.0882625465985735,
|
| 98 |
+
"grad_norm": 0.762417733669281,
|
| 99 |
+
"learning_rate": 0.000988751984934317,
|
| 100 |
+
"loss": 1.3001574277877808,
|
| 101 |
+
"step": 2048
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.0882625465985735,
|
| 105 |
+
"eval_bleu": 0.938925796606621,
|
| 106 |
+
"eval_cos_loss": 0.4579503086330032,
|
| 107 |
+
"eval_dec_loss": 0.10506504188690867,
|
| 108 |
+
"eval_loss": 1.2970875999820766,
|
| 109 |
+
"eval_mse2_loss": 0.15707123614768229,
|
| 110 |
+
"eval_mse_loss": 1.2970875999820766,
|
| 111 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 112 |
+
"eval_var_loss": 0.01723895594080501,
|
| 113 |
+
"flow/cos_sim": 0.5420496905409197,
|
| 114 |
+
"flow/improvement_ratio": 0.8918823948038667,
|
| 115 |
+
"flow/mag_ratio_mean": 0.5503126610316702,
|
| 116 |
+
"flow/mag_ratio_std": 0.25175602854823254,
|
| 117 |
+
"step": 2048
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.0882625465985735,
|
| 121 |
+
"eval_bleu": 0.938925796606621,
|
| 122 |
+
"eval_cos_loss": 0.4579503086330032,
|
| 123 |
+
"eval_dec_loss": 0.10506504188690867,
|
| 124 |
+
"eval_loss": 1.2970875999820766,
|
| 125 |
+
"eval_mse2_loss": 0.15707123614768229,
|
| 126 |
+
"eval_mse_loss": 1.2970875999820766,
|
| 127 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 128 |
+
"eval_runtime": 151.9416,
|
| 129 |
+
"eval_samples_per_second": 197.444,
|
| 130 |
+
"eval_steps_per_second": 3.087,
|
| 131 |
+
"eval_var_loss": 0.01723895594080501,
|
| 132 |
+
"flow/cos_sim": 0.5420496905409197,
|
| 133 |
+
"flow/improvement_ratio": 0.8918823948038667,
|
| 134 |
+
"flow/mag_ratio_mean": 0.5503126610316702,
|
| 135 |
+
"flow/mag_ratio_std": 0.25175602854823254,
|
| 136 |
+
"step": 2048
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.09929536492339518,
|
| 140 |
+
"grad_norm": 0.39165085554122925,
|
| 141 |
+
"learning_rate": 0.0009847081812963268,
|
| 142 |
+
"loss": 1.2909460067749023,
|
| 143 |
+
"step": 2304
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.11032818324821686,
|
| 147 |
+
"grad_norm": 0.6050369739532471,
|
| 148 |
+
"learning_rate": 0.0009800555855486275,
|
| 149 |
+
"loss": 1.291382908821106,
|
| 150 |
+
"step": 2560
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.12136100157303854,
|
| 154 |
+
"grad_norm": 0.6340572237968445,
|
| 155 |
+
"learning_rate": 0.0009748000413383664,
|
| 156 |
+
"loss": 1.2860350608825684,
|
| 157 |
+
"step": 2816
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.13239381989786023,
|
| 161 |
+
"grad_norm": 0.8046131134033203,
|
| 162 |
+
"learning_rate": 0.0009689481496142604,
|
| 163 |
+
"loss": 1.2806360721588135,
|
| 164 |
+
"step": 3072
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.13239381989786023,
|
| 168 |
+
"eval_bleu": 0.9365596012238808,
|
| 169 |
+
"eval_cos_loss": 0.4510079253075728,
|
| 170 |
+
"eval_dec_loss": 0.1170106883853801,
|
| 171 |
+
"eval_loss": 1.2785198518208094,
|
| 172 |
+
"eval_mse2_loss": 0.15482012001372603,
|
| 173 |
+
"eval_mse_loss": 1.2785198518208094,
|
| 174 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 175 |
+
"eval_var_loss": 0.01723895594080501,
|
| 176 |
+
"flow/cos_sim": 0.5489920710703966,
|
| 177 |
+
"flow/improvement_ratio": 0.895310169598187,
|
| 178 |
+
"flow/mag_ratio_mean": 0.5600611698398712,
|
| 179 |
+
"flow/mag_ratio_std": 0.2589119763326035,
|
| 180 |
+
"step": 3072
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 0.13239381989786023,
|
| 184 |
+
"eval_bleu": 0.9365596012238808,
|
| 185 |
+
"eval_cos_loss": 0.4510079253075728,
|
| 186 |
+
"eval_dec_loss": 0.1170106883853801,
|
| 187 |
+
"eval_loss": 1.2785198518208094,
|
| 188 |
+
"eval_mse2_loss": 0.15482012001372603,
|
| 189 |
+
"eval_mse_loss": 1.2785198518208094,
|
| 190 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 191 |
+
"eval_runtime": 150.2303,
|
| 192 |
+
"eval_samples_per_second": 199.693,
|
| 193 |
+
"eval_steps_per_second": 3.122,
|
| 194 |
+
"eval_var_loss": 0.01723895594080501,
|
| 195 |
+
"flow/cos_sim": 0.5489920710703966,
|
| 196 |
+
"flow/improvement_ratio": 0.895310169598187,
|
| 197 |
+
"flow/mag_ratio_mean": 0.5600611698398712,
|
| 198 |
+
"flow/mag_ratio_std": 0.2589119763326035,
|
| 199 |
+
"step": 3072
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.14342663822268192,
|
| 203 |
+
"grad_norm": 0.7344346046447754,
|
| 204 |
+
"learning_rate": 0.0009625072603358231,
|
| 205 |
+
"loss": 1.277908444404602,
|
| 206 |
+
"step": 3328
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.15445945654750362,
|
| 210 |
+
"grad_norm": 0.7456739544868469,
|
| 211 |
+
"learning_rate": 0.0009554854632418371,
|
| 212 |
+
"loss": 1.274967074394226,
|
| 213 |
+
"step": 3584
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.1654922748723253,
|
| 217 |
+
"grad_norm": 0.528167724609375,
|
| 218 |
+
"learning_rate": 0.000947891577689663,
|
| 219 |
+
"loss": 1.2722811698913574,
|
| 220 |
+
"step": 3840
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.176525093197147,
|
| 224 |
+
"grad_norm": 0.7374073266983032,
|
| 225 |
+
"learning_rate": 0.0009397351415781539,
|
| 226 |
+
"loss": 1.2716022729873657,
|
| 227 |
+
"step": 4096
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.176525093197147,
|
| 231 |
+
"eval_bleu": 0.9383145863088955,
|
| 232 |
+
"eval_cos_loss": 0.44795799712890755,
|
| 233 |
+
"eval_dec_loss": 0.11301154795406597,
|
| 234 |
+
"eval_loss": 1.2707049117159488,
|
| 235 |
+
"eval_mse2_loss": 0.15204078735890927,
|
| 236 |
+
"eval_mse_loss": 1.2707049117159488,
|
| 237 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 238 |
+
"eval_var_loss": 0.01723895594080501,
|
| 239 |
+
"flow/cos_sim": 0.552042003760714,
|
| 240 |
+
"flow/improvement_ratio": 0.8948889724227157,
|
| 241 |
+
"flow/mag_ratio_mean": 0.5576132778674047,
|
| 242 |
+
"flow/mag_ratio_std": 0.25525683488672984,
|
| 243 |
+
"step": 4096
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.176525093197147,
|
| 247 |
+
"eval_bleu": 0.9383145863088955,
|
| 248 |
+
"eval_cos_loss": 0.44795799712890755,
|
| 249 |
+
"eval_dec_loss": 0.11301154795406597,
|
| 250 |
+
"eval_loss": 1.2707049117159488,
|
| 251 |
+
"eval_mse2_loss": 0.15204078735890927,
|
| 252 |
+
"eval_mse_loss": 1.2707049117159488,
|
| 253 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 254 |
+
"eval_runtime": 149.5476,
|
| 255 |
+
"eval_samples_per_second": 200.605,
|
| 256 |
+
"eval_steps_per_second": 3.136,
|
| 257 |
+
"eval_var_loss": 0.01723895594080501,
|
| 258 |
+
"flow/cos_sim": 0.552042003760714,
|
| 259 |
+
"flow/improvement_ratio": 0.8948889724227157,
|
| 260 |
+
"flow/mag_ratio_mean": 0.5576132778674047,
|
| 261 |
+
"flow/mag_ratio_std": 0.25525683488672984,
|
| 262 |
+
"step": 4096
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.18755791152196866,
|
| 266 |
+
"grad_norm": 1.123129963874817,
|
| 267 |
+
"learning_rate": 0.000931026399368079,
|
| 268 |
+
"loss": 1.2691912651062012,
|
| 269 |
+
"step": 4352
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.19859072984679035,
|
| 273 |
+
"grad_norm": 0.49173882603645325,
|
| 274 |
+
"learning_rate": 0.0009217762892151117,
|
| 275 |
+
"loss": 1.26752769947052,
|
| 276 |
+
"step": 4608
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.20962354817161205,
|
| 280 |
+
"grad_norm": 0.5665431618690491,
|
| 281 |
+
"learning_rate": 0.0009119964292315354,
|
| 282 |
+
"loss": 1.2669333219528198,
|
| 283 |
+
"step": 4864
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.22065636649643372,
|
| 287 |
+
"grad_norm": 0.4946308732032776,
|
| 288 |
+
"learning_rate": 0.0009016991028939279,
|
| 289 |
+
"loss": 1.2646225690841675,
|
| 290 |
+
"step": 5120
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.22065636649643372,
|
| 294 |
+
"eval_bleu": 0.9396675860722136,
|
| 295 |
+
"eval_cos_loss": 0.44516199083724767,
|
| 296 |
+
"eval_dec_loss": 0.10893038547893705,
|
| 297 |
+
"eval_loss": 1.264682760879175,
|
| 298 |
+
"eval_mse2_loss": 0.1498125367073108,
|
| 299 |
+
"eval_mse_loss": 1.264682760879175,
|
| 300 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 301 |
+
"eval_var_loss": 0.01723895594080501,
|
| 302 |
+
"flow/cos_sim": 0.5548380070657872,
|
| 303 |
+
"flow/improvement_ratio": 0.8946977740665997,
|
| 304 |
+
"flow/mag_ratio_mean": 0.5694006043456511,
|
| 305 |
+
"flow/mag_ratio_std": 0.2655116878211625,
|
| 306 |
+
"step": 5120
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.22065636649643372,
|
| 310 |
+
"eval_bleu": 0.9396675860722136,
|
| 311 |
+
"eval_cos_loss": 0.44516199083724767,
|
| 312 |
+
"eval_dec_loss": 0.10893038547893705,
|
| 313 |
+
"eval_loss": 1.264682760879175,
|
| 314 |
+
"eval_mse2_loss": 0.1498125367073108,
|
| 315 |
+
"eval_mse_loss": 1.264682760879175,
|
| 316 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 317 |
+
"eval_runtime": 151.8799,
|
| 318 |
+
"eval_samples_per_second": 197.524,
|
| 319 |
+
"eval_steps_per_second": 3.088,
|
| 320 |
+
"eval_var_loss": 0.01723895594080501,
|
| 321 |
+
"flow/cos_sim": 0.5548380070657872,
|
| 322 |
+
"flow/improvement_ratio": 0.8946977740665997,
|
| 323 |
+
"flow/mag_ratio_mean": 0.5694006043456511,
|
| 324 |
+
"flow/mag_ratio_std": 0.2655116878211625,
|
| 325 |
+
"step": 5120
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.23168918482125542,
|
| 329 |
+
"grad_norm": 0.5147830843925476,
|
| 330 |
+
"learning_rate": 0.0008908972436151494,
|
| 331 |
+
"loss": 1.261371374130249,
|
| 332 |
+
"step": 5376
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.2427220031460771,
|
| 336 |
+
"grad_norm": 0.7221893668174744,
|
| 337 |
+
"learning_rate": 0.0008796044185000127,
|
| 338 |
+
"loss": 1.259010672569275,
|
| 339 |
+
"step": 5632
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.2537548214708988,
|
| 343 |
+
"grad_norm": 0.6270182132720947,
|
| 344 |
+
"learning_rate": 0.0008678348113050368,
|
| 345 |
+
"loss": 1.2565613985061646,
|
| 346 |
+
"step": 5888
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.26478763979572045,
|
| 350 |
+
"grad_norm": 0.3954711854457855,
|
| 351 |
+
"learning_rate": 0.0008556032046236897,
|
| 352 |
+
"loss": 1.258548378944397,
|
| 353 |
+
"step": 6144
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.26478763979572045,
|
| 357 |
+
"eval_bleu": 0.9381239377332383,
|
| 358 |
+
"eval_cos_loss": 0.4434889930524806,
|
| 359 |
+
"eval_dec_loss": 0.11391587999226378,
|
| 360 |
+
"eval_loss": 1.2588644528439812,
|
| 361 |
+
"eval_mse2_loss": 0.15056055846181252,
|
| 362 |
+
"eval_mse_loss": 1.2588644528439812,
|
| 363 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 364 |
+
"eval_var_loss": 0.01723895594080501,
|
| 365 |
+
"flow/cos_sim": 0.5565110067568863,
|
| 366 |
+
"flow/improvement_ratio": 0.8946461940625074,
|
| 367 |
+
"flow/mag_ratio_mean": 0.5628604918146438,
|
| 368 |
+
"flow/mag_ratio_std": 0.2606462057528974,
|
| 369 |
+
"step": 6144
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.26478763979572045,
|
| 373 |
+
"eval_bleu": 0.9381239377332383,
|
| 374 |
+
"eval_cos_loss": 0.4434889930524806,
|
| 375 |
+
"eval_dec_loss": 0.11391587999226378,
|
| 376 |
+
"eval_loss": 1.2588644528439812,
|
| 377 |
+
"eval_mse2_loss": 0.15056055846181252,
|
| 378 |
+
"eval_mse_loss": 1.2588644528439812,
|
| 379 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 380 |
+
"eval_runtime": 153.8457,
|
| 381 |
+
"eval_samples_per_second": 195.001,
|
| 382 |
+
"eval_steps_per_second": 3.049,
|
| 383 |
+
"eval_var_loss": 0.01723895594080501,
|
| 384 |
+
"flow/cos_sim": 0.5565110067568863,
|
| 385 |
+
"flow/improvement_ratio": 0.8946461940625074,
|
| 386 |
+
"flow/mag_ratio_mean": 0.5628604918146438,
|
| 387 |
+
"flow/mag_ratio_std": 0.2606462057528974,
|
| 388 |
+
"step": 6144
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.2758204581205422,
|
| 392 |
+
"grad_norm": 0.8126729130744934,
|
| 393 |
+
"learning_rate": 0.000842924961319492,
|
| 394 |
+
"loss": 1.2565950155258179,
|
| 395 |
+
"step": 6400
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.28685327644536385,
|
| 399 |
+
"grad_norm": 0.84797203540802,
|
| 400 |
+
"learning_rate": 0.0008298160052303045,
|
| 401 |
+
"loss": 1.2548315525054932,
|
| 402 |
+
"step": 6656
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.2978860947701855,
|
| 406 |
+
"grad_norm": 0.561568021774292,
|
| 407 |
+
"learning_rate": 0.0008162928011680314,
|
| 408 |
+
"loss": 1.2526129484176636,
|
| 409 |
+
"step": 6912
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.30891891309500724,
|
| 413 |
+
"grad_norm": 0.45474377274513245,
|
| 414 |
+
"learning_rate": 0.000802372334238864,
|
| 415 |
+
"loss": 1.2513761520385742,
|
| 416 |
+
"step": 7168
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.30891891309500724,
|
| 420 |
+
"eval_bleu": 0.9385536520845816,
|
| 421 |
+
"eval_cos_loss": 0.4402598062557961,
|
| 422 |
+
"eval_dec_loss": 0.11249503215500858,
|
| 423 |
+
"eval_loss": 1.2510530173397267,
|
| 424 |
+
"eval_mse2_loss": 0.1480516226116274,
|
| 425 |
+
"eval_mse_loss": 1.2510530173397267,
|
| 426 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 427 |
+
"eval_var_loss": 0.01723895594080501,
|
| 428 |
+
"flow/cos_sim": 0.5597401952692694,
|
| 429 |
+
"flow/improvement_ratio": 0.895444744939743,
|
| 430 |
+
"flow/mag_ratio_mean": 0.5710282248220464,
|
| 431 |
+
"flow/mag_ratio_std": 0.26387540328858505,
|
| 432 |
+
"step": 7168
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.30891891309500724,
|
| 436 |
+
"eval_bleu": 0.9385536520845816,
|
| 437 |
+
"eval_cos_loss": 0.4402598062557961,
|
| 438 |
+
"eval_dec_loss": 0.11249503215500858,
|
| 439 |
+
"eval_loss": 1.2510530173397267,
|
| 440 |
+
"eval_mse2_loss": 0.1480516226116274,
|
| 441 |
+
"eval_mse_loss": 1.2510530173397267,
|
| 442 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 443 |
+
"eval_runtime": 152.7181,
|
| 444 |
+
"eval_samples_per_second": 196.44,
|
| 445 |
+
"eval_steps_per_second": 3.071,
|
| 446 |
+
"eval_var_loss": 0.01723895594080501,
|
| 447 |
+
"flow/cos_sim": 0.5597401952692694,
|
| 448 |
+
"flow/improvement_ratio": 0.895444744939743,
|
| 449 |
+
"flow/mag_ratio_mean": 0.5710282248220464,
|
| 450 |
+
"flow/mag_ratio_std": 0.26387540328858505,
|
| 451 |
+
"step": 7168
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.3199517314198289,
|
| 455 |
+
"grad_norm": 1.3543585538864136,
|
| 456 |
+
"learning_rate": 0.0007880720885100349,
|
| 457 |
+
"loss": 1.2521653175354004,
|
| 458 |
+
"step": 7424
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.3309845497446506,
|
| 462 |
+
"grad_norm": 0.4370076358318329,
|
| 463 |
+
"learning_rate": 0.0007734100250498788,
|
| 464 |
+
"loss": 1.249273419380188,
|
| 465 |
+
"step": 7680
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.3420173680694723,
|
| 469 |
+
"grad_norm": 1.0196475982666016,
|
| 470 |
+
"learning_rate": 0.000758404559368781,
|
| 471 |
+
"loss": 1.2500712871551514,
|
| 472 |
+
"step": 7936
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.353050186394294,
|
| 476 |
+
"grad_norm": 0.733001708984375,
|
| 477 |
+
"learning_rate": 0.0007430745382893488,
|
| 478 |
+
"loss": 1.245364785194397,
|
| 479 |
+
"step": 8192
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.353050186394294,
|
| 483 |
+
"eval_bleu": 0.9376793187397806,
|
| 484 |
+
"eval_cos_loss": 0.4385024095013706,
|
| 485 |
+
"eval_dec_loss": 0.11364057421017049,
|
| 486 |
+
"eval_loss": 1.2459661925016945,
|
| 487 |
+
"eval_mse2_loss": 0.148339767350571,
|
| 488 |
+
"eval_mse_loss": 1.2459661925016945,
|
| 489 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 490 |
+
"eval_var_loss": 0.01723895594080501,
|
| 491 |
+
"flow/cos_sim": 0.5614975882745755,
|
| 492 |
+
"flow/improvement_ratio": 0.8961417695352518,
|
| 493 |
+
"flow/mag_ratio_mean": 0.5688313084370547,
|
| 494 |
+
"flow/mag_ratio_std": 0.26494109700483554,
|
| 495 |
+
"step": 8192
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.353050186394294,
|
| 499 |
+
"eval_bleu": 0.9376793187397806,
|
| 500 |
+
"eval_cos_loss": 0.4385024095013706,
|
| 501 |
+
"eval_dec_loss": 0.11364057421017049,
|
| 502 |
+
"eval_loss": 1.2459661925016945,
|
| 503 |
+
"eval_mse2_loss": 0.148339767350571,
|
| 504 |
+
"eval_mse_loss": 1.2459661925016945,
|
| 505 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 506 |
+
"eval_runtime": 152.8054,
|
| 507 |
+
"eval_samples_per_second": 196.328,
|
| 508 |
+
"eval_steps_per_second": 3.069,
|
| 509 |
+
"eval_var_loss": 0.01723895594080501,
|
| 510 |
+
"flow/cos_sim": 0.5614975882745755,
|
| 511 |
+
"flow/improvement_ratio": 0.8961417695352518,
|
| 512 |
+
"flow/mag_ratio_mean": 0.5688313084370547,
|
| 513 |
+
"flow/mag_ratio_std": 0.26494109700483554,
|
| 514 |
+
"step": 8192
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.36408300471911564,
|
| 518 |
+
"grad_norm": 0.676328718662262,
|
| 519 |
+
"learning_rate": 0.0007274392162748551,
|
| 520 |
+
"loss": 1.2448910474777222,
|
| 521 |
+
"step": 8448
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.3751158230439373,
|
| 525 |
+
"grad_norm": 0.6379961967468262,
|
| 526 |
+
"learning_rate": 0.000711518231245687,
|
| 527 |
+
"loss": 1.2442706823349,
|
| 528 |
+
"step": 8704
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.38614864136875904,
|
| 532 |
+
"grad_norm": 0.5386805534362793,
|
| 533 |
+
"learning_rate": 0.0006953315799141723,
|
| 534 |
+
"loss": 1.2446835041046143,
|
| 535 |
+
"step": 8960
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.3971814596935807,
|
| 539 |
+
"grad_norm": 0.8263258934020996,
|
| 540 |
+
"learning_rate": 0.0006788995926687669,
|
| 541 |
+
"loss": 1.2411766052246094,
|
| 542 |
+
"step": 9216
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.3971814596935807,
|
| 546 |
+
"eval_bleu": 0.9372486918854673,
|
| 547 |
+
"eval_cos_loss": 0.43675092898452206,
|
| 548 |
+
"eval_dec_loss": 0.11516488874867273,
|
| 549 |
+
"eval_loss": 1.241364901762273,
|
| 550 |
+
"eval_mse2_loss": 0.1478570194196091,
|
| 551 |
+
"eval_mse_loss": 1.241364901762273,
|
| 552 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 553 |
+
"eval_var_loss": 0.01723895594080501,
|
| 554 |
+
"flow/cos_sim": 0.5632490722863659,
|
| 555 |
+
"flow/improvement_ratio": 0.8974738620491679,
|
| 556 |
+
"flow/mag_ratio_mean": 0.5655419154207844,
|
| 557 |
+
"flow/mag_ratio_std": 0.2603240320041998,
|
| 558 |
+
"step": 9216
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 0.3971814596935807,
|
| 562 |
+
"eval_bleu": 0.9372486918854673,
|
| 563 |
+
"eval_cos_loss": 0.43675092898452206,
|
| 564 |
+
"eval_dec_loss": 0.11516488874867273,
|
| 565 |
+
"eval_loss": 1.241364901762273,
|
| 566 |
+
"eval_mse2_loss": 0.1478570194196091,
|
| 567 |
+
"eval_mse_loss": 1.241364901762273,
|
| 568 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 569 |
+
"eval_runtime": 152.8433,
|
| 570 |
+
"eval_samples_per_second": 196.28,
|
| 571 |
+
"eval_steps_per_second": 3.069,
|
| 572 |
+
"eval_var_loss": 0.01723895594080501,
|
| 573 |
+
"flow/cos_sim": 0.5632490722863659,
|
| 574 |
+
"flow/improvement_ratio": 0.8974738620491679,
|
| 575 |
+
"flow/mag_ratio_mean": 0.5655419154207844,
|
| 576 |
+
"flow/mag_ratio_std": 0.2603240320041998,
|
| 577 |
+
"step": 9216
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.4082142780184024,
|
| 581 |
+
"grad_norm": 0.7855456471443176,
|
| 582 |
+
"learning_rate": 0.0006622429080391422,
|
| 583 |
+
"loss": 1.2460049390792847,
|
| 584 |
+
"step": 9472
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.4192470963432241,
|
| 588 |
+
"grad_norm": 0.4608207941055298,
|
| 589 |
+
"learning_rate": 0.0006453824467742515,
|
| 590 |
+
"loss": 1.2414920330047607,
|
| 591 |
+
"step": 9728
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.43027991466804577,
|
| 595 |
+
"grad_norm": 0.5247617959976196,
|
| 596 |
+
"learning_rate": 0.0006283393855659275,
|
| 597 |
+
"loss": 1.2424880266189575,
|
| 598 |
+
"step": 9984
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.44131273299286744,
|
| 602 |
+
"grad_norm": 0.8765453100204468,
|
| 603 |
+
"learning_rate": 0.0006111351304510173,
|
| 604 |
+
"loss": 1.237776517868042,
|
| 605 |
+
"step": 10240
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.44131273299286744,
|
| 609 |
+
"eval_bleu": 0.937646836000478,
|
| 610 |
+
"eval_cos_loss": 0.4353823194752878,
|
| 611 |
+
"eval_dec_loss": 0.11402556833737632,
|
| 612 |
+
"eval_loss": 1.2377641976260936,
|
| 613 |
+
"eval_mse2_loss": 0.1474350707204357,
|
| 614 |
+
"eval_mse_loss": 1.2377641976260936,
|
| 615 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 616 |
+
"eval_var_loss": 0.01723895594080501,
|
| 617 |
+
"flow/cos_sim": 0.564617679063191,
|
| 618 |
+
"flow/improvement_ratio": 0.899760089830549,
|
| 619 |
+
"flow/mag_ratio_mean": 0.5730336795229394,
|
| 620 |
+
"flow/mag_ratio_std": 0.26344449729172154,
|
| 621 |
+
"step": 10240
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.44131273299286744,
|
| 625 |
+
"eval_bleu": 0.937646836000478,
|
| 626 |
+
"eval_cos_loss": 0.4353823194752878,
|
| 627 |
+
"eval_dec_loss": 0.11402556833737632,
|
| 628 |
+
"eval_loss": 1.2377641976260936,
|
| 629 |
+
"eval_mse2_loss": 0.1474350707204357,
|
| 630 |
+
"eval_mse_loss": 1.2377641976260936,
|
| 631 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 632 |
+
"eval_runtime": 151.9737,
|
| 633 |
+
"eval_samples_per_second": 197.403,
|
| 634 |
+
"eval_steps_per_second": 3.086,
|
| 635 |
+
"eval_var_loss": 0.01723895594080501,
|
| 636 |
+
"flow/cos_sim": 0.564617679063191,
|
| 637 |
+
"flow/improvement_ratio": 0.899760089830549,
|
| 638 |
+
"flow/mag_ratio_mean": 0.5730336795229394,
|
| 639 |
+
"flow/mag_ratio_std": 0.26344449729172154,
|
| 640 |
+
"step": 10240
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.45234555131768917,
|
| 644 |
+
"grad_norm": 0.6895334124565125,
|
| 645 |
+
"learning_rate": 0.0005937912899254605,
|
| 646 |
+
"loss": 1.2384426593780518,
|
| 647 |
+
"step": 10496
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.46337836964251083,
|
| 651 |
+
"grad_norm": 0.6421330571174622,
|
| 652 |
+
"learning_rate": 0.0005763296478040787,
|
| 653 |
+
"loss": 1.240878939628601,
|
| 654 |
+
"step": 10752
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.4744111879673325,
|
| 658 |
+
"grad_norm": 0.7770284414291382,
|
| 659 |
+
"learning_rate": 0.0005587721358601663,
|
| 660 |
+
"loss": 1.2393468618392944,
|
| 661 |
+
"step": 11008
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.4854440062921542,
|
| 665 |
+
"grad_norm": 1.0520166158676147,
|
| 666 |
+
"learning_rate": 0.0005411408062792448,
|
| 667 |
+
"loss": 1.237922191619873,
|
| 668 |
+
"step": 11264
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.4854440062921542,
|
| 672 |
+
"eval_bleu": 0.93652744913201,
|
| 673 |
+
"eval_cos_loss": 0.4366011674851497,
|
| 674 |
+
"eval_dec_loss": 0.11468809016390459,
|
| 675 |
+
"eval_loss": 1.2409222840246108,
|
| 676 |
+
"eval_mse2_loss": 0.14564816977804912,
|
| 677 |
+
"eval_mse_loss": 1.2409222840246108,
|
| 678 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 679 |
+
"eval_var_loss": 0.01723895594080501,
|
| 680 |
+
"flow/cos_sim": 0.5633988297824413,
|
| 681 |
+
"flow/improvement_ratio": 0.897065937773251,
|
| 682 |
+
"flow/mag_ratio_mean": 0.5639294942558956,
|
| 683 |
+
"flow/mag_ratio_std": 0.25510865748564066,
|
| 684 |
+
"step": 11264
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.4854440062921542,
|
| 688 |
+
"eval_bleu": 0.93652744913201,
|
| 689 |
+
"eval_cos_loss": 0.4366011674851497,
|
| 690 |
+
"eval_dec_loss": 0.11468809016390459,
|
| 691 |
+
"eval_loss": 1.2409222840246108,
|
| 692 |
+
"eval_mse2_loss": 0.14564816977804912,
|
| 693 |
+
"eval_mse_loss": 1.2409222840246108,
|
| 694 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 695 |
+
"eval_runtime": 152.4483,
|
| 696 |
+
"eval_samples_per_second": 196.788,
|
| 697 |
+
"eval_steps_per_second": 3.076,
|
| 698 |
+
"eval_var_loss": 0.01723895594080501,
|
| 699 |
+
"flow/cos_sim": 0.5633988297824413,
|
| 700 |
+
"flow/improvement_ratio": 0.897065937773251,
|
| 701 |
+
"flow/mag_ratio_mean": 0.5639294942558956,
|
| 702 |
+
"flow/mag_ratio_std": 0.25510865748564066,
|
| 703 |
+
"step": 11264
|
| 704 |
+
}
|
| 705 |
+
],
|
| 706 |
+
"logging_steps": 256,
|
| 707 |
+
"max_steps": 23204,
|
| 708 |
+
"num_input_tokens_seen": 0,
|
| 709 |
+
"num_train_epochs": 1,
|
| 710 |
+
"save_steps": 1024,
|
| 711 |
+
"stateful_callbacks": {
|
| 712 |
+
"TrainerControl": {
|
| 713 |
+
"args": {
|
| 714 |
+
"should_epoch_stop": false,
|
| 715 |
+
"should_evaluate": false,
|
| 716 |
+
"should_log": false,
|
| 717 |
+
"should_save": true,
|
| 718 |
+
"should_training_stop": false
|
| 719 |
+
},
|
| 720 |
+
"attributes": {}
|
| 721 |
+
}
|
| 722 |
+
},
|
| 723 |
+
"total_flos": 0.0,
|
| 724 |
+
"train_batch_size": 64,
|
| 725 |
+
"trial_name": null,
|
| 726 |
+
"trial_params": null
|
| 727 |
+
}
|
checkpoints-d1.1/checkpoint-11264/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:61a16bb839f687414b8e48611327c4b9cfddeefe38c031ca70808f9a97c476b7
|
| 3 |
+
size 5137
|