Upload folder using huggingface_hub
Browse files- checkpoints-v1/checkpoint-14336/ema.safetensors +3 -0
- checkpoints-v1/checkpoint-14336/model.safetensors +3 -0
- checkpoints-v1/checkpoint-14336/optimizer.pt +3 -0
- checkpoints-v1/checkpoint-14336/rng_state.pth +3 -0
- checkpoints-v1/checkpoint-14336/scaler.pt +3 -0
- checkpoints-v1/checkpoint-14336/scheduler.pt +3 -0
- checkpoints-v1/checkpoint-14336/trainer_state.json +748 -0
- checkpoints-v1/checkpoint-14336/training_args.bin +3 -0
checkpoints-v1/checkpoint-14336/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:42037039f322c66e4db015fad3ecc3ae3e427264853ff9d0066aa6a7ee231f01
|
| 3 |
+
size 571816
|
checkpoints-v1/checkpoint-14336/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4151eb5921dad621a3441c3b6a39b6d8702f93fae0d4f8d5ba00a31d216219d1
|
| 3 |
+
size 19265392
|
checkpoints-v1/checkpoint-14336/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:88946300e31ba750dde4ed985f5abfb8cb17c978e5106d43da1339fdf32d9a2f
|
| 3 |
+
size 1212939
|
checkpoints-v1/checkpoint-14336/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0fb1ced3c1a73919bac75ddd04d7fab6d5228194743b711df8de676564c549fd
|
| 3 |
+
size 14645
|
checkpoints-v1/checkpoint-14336/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bd345a16eac44d3b6973d1b101c7e88ac305f5d055942cd287e3b7f78937bd29
|
| 3 |
+
size 1383
|
checkpoints-v1/checkpoint-14336/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d083abe2a0e28e3d2a41416c38d70182db4da2363a1af9e95380e20ec2d4d1e
|
| 3 |
+
size 1465
|
checkpoints-v1/checkpoint-14336/trainer_state.json
ADDED
|
@@ -0,0 +1,748 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.6621403168444876,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 14336,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011823934229365849,
|
| 14 |
+
"grad_norm": 0.17313924431800842,
|
| 15 |
+
"learning_rate": 0.000498046875,
|
| 16 |
+
"loss": 1.7064213752746582,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.15720121562480927,
|
| 22 |
+
"learning_rate": 0.000998046875,
|
| 23 |
+
"loss": 1.1101791858673096,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.145066499710083,
|
| 29 |
+
"learning_rate": 0.000999640996023194,
|
| 30 |
+
"loss": 1.0319945812225342,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.1564043015241623,
|
| 36 |
+
"learning_rate": 0.0009985588674043958,
|
| 37 |
+
"loss": 0.9714428186416626,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_loss": 0.9119398776254698,
|
| 43 |
+
"eval_mse_loss": 0.9119398776254698,
|
| 44 |
+
"flow/cos_sim": 0.705863389794685,
|
| 45 |
+
"flow/improvement_ratio": 0.4745521226974383,
|
| 46 |
+
"flow/mag_ratio_mean": 0.6946878124317623,
|
| 47 |
+
"flow/mag_ratio_std": 0.17089739207127322,
|
| 48 |
+
"step": 1024
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.047295736917463395,
|
| 52 |
+
"eval_loss": 0.9119398776254698,
|
| 53 |
+
"eval_mse_loss": 0.9119398776254698,
|
| 54 |
+
"eval_runtime": 100.3557,
|
| 55 |
+
"eval_samples_per_second": 278.938,
|
| 56 |
+
"eval_steps_per_second": 4.364,
|
| 57 |
+
"flow/cos_sim": 0.705863389794685,
|
| 58 |
+
"flow/improvement_ratio": 0.4745521226974383,
|
| 59 |
+
"flow/mag_ratio_mean": 0.6946878124317623,
|
| 60 |
+
"flow/mag_ratio_std": 0.17089739207127322,
|
| 61 |
+
"step": 1024
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.05911967114682925,
|
| 65 |
+
"grad_norm": 0.16647565364837646,
|
| 66 |
+
"learning_rate": 0.0009967551747861387,
|
| 67 |
+
"loss": 0.929740846157074,
|
| 68 |
+
"step": 1280
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.0709436053761951,
|
| 72 |
+
"grad_norm": 0.18282100558280945,
|
| 73 |
+
"learning_rate": 0.000994232528651847,
|
| 74 |
+
"loss": 0.895330548286438,
|
| 75 |
+
"step": 1536
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.08276753960556095,
|
| 79 |
+
"grad_norm": 0.17141221463680267,
|
| 80 |
+
"learning_rate": 0.0009909945800260092,
|
| 81 |
+
"loss": 0.8683519959449768,
|
| 82 |
+
"step": 1792
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.09459147383492679,
|
| 86 |
+
"grad_norm": 0.19230426847934723,
|
| 87 |
+
"learning_rate": 0.0009870460151900522,
|
| 88 |
+
"loss": 0.8509060144424438,
|
| 89 |
+
"step": 2048
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.09459147383492679,
|
| 93 |
+
"eval_loss": 0.8054492728078746,
|
| 94 |
+
"eval_mse_loss": 0.8054492728078746,
|
| 95 |
+
"flow/cos_sim": 0.7493123941497716,
|
| 96 |
+
"flow/improvement_ratio": 0.476775190669652,
|
| 97 |
+
"flow/mag_ratio_mean": 0.7334817982699773,
|
| 98 |
+
"flow/mag_ratio_std": 0.1754700492846367,
|
| 99 |
+
"step": 2048
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.09459147383492679,
|
| 103 |
+
"eval_loss": 0.8054492728078746,
|
| 104 |
+
"eval_mse_loss": 0.8054492728078746,
|
| 105 |
+
"eval_runtime": 101.0321,
|
| 106 |
+
"eval_samples_per_second": 277.07,
|
| 107 |
+
"eval_steps_per_second": 4.335,
|
| 108 |
+
"flow/cos_sim": 0.7493123941497716,
|
| 109 |
+
"flow/improvement_ratio": 0.476775190669652,
|
| 110 |
+
"flow/mag_ratio_mean": 0.7334817982699773,
|
| 111 |
+
"flow/mag_ratio_std": 0.1754700492846367,
|
| 112 |
+
"step": 2048
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.10641540806429264,
|
| 116 |
+
"grad_norm": 0.1906282752752304,
|
| 117 |
+
"learning_rate": 0.0009823925488998885,
|
| 118 |
+
"loss": 0.8359999060630798,
|
| 119 |
+
"step": 2304
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.1182393422936585,
|
| 123 |
+
"grad_norm": 0.24338403344154358,
|
| 124 |
+
"learning_rate": 0.0009770409161149525,
|
| 125 |
+
"loss": 0.8268019556999207,
|
| 126 |
+
"step": 2560
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 0.13006327652302435,
|
| 130 |
+
"grad_norm": 0.2402328997850418,
|
| 131 |
+
"learning_rate": 0.0009709988622506973,
|
| 132 |
+
"loss": 0.8164141178131104,
|
| 133 |
+
"step": 2816
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 0.1418872107523902,
|
| 137 |
+
"grad_norm": 0.19165532290935516,
|
| 138 |
+
"learning_rate": 0.000964275131968659,
|
| 139 |
+
"loss": 0.8093737363815308,
|
| 140 |
+
"step": 3072
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.1418872107523902,
|
| 144 |
+
"eval_loss": 0.7729098539646357,
|
| 145 |
+
"eval_mse_loss": 0.7729098539646357,
|
| 146 |
+
"flow/cos_sim": 0.7615711595097633,
|
| 147 |
+
"flow/improvement_ratio": 0.48455443985113816,
|
| 148 |
+
"flow/mag_ratio_mean": 0.7450809510059009,
|
| 149 |
+
"flow/mag_ratio_std": 0.17553651608424645,
|
| 150 |
+
"step": 3072
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.1418872107523902,
|
| 154 |
+
"eval_loss": 0.7729098539646357,
|
| 155 |
+
"eval_mse_loss": 0.7729098539646357,
|
| 156 |
+
"eval_runtime": 100.8602,
|
| 157 |
+
"eval_samples_per_second": 277.543,
|
| 158 |
+
"eval_steps_per_second": 4.343,
|
| 159 |
+
"flow/cos_sim": 0.7615711595097633,
|
| 160 |
+
"flow/improvement_ratio": 0.48455443985113816,
|
| 161 |
+
"flow/mag_ratio_mean": 0.7450809510059009,
|
| 162 |
+
"flow/mag_ratio_std": 0.17553651608424645,
|
| 163 |
+
"step": 3072
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.15371114498175603,
|
| 167 |
+
"grad_norm": 0.23683102428913116,
|
| 168 |
+
"learning_rate": 0.0009568794565203123,
|
| 169 |
+
"loss": 0.8039254546165466,
|
| 170 |
+
"step": 3328
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.1655350792111219,
|
| 174 |
+
"grad_norm": 0.24094223976135254,
|
| 175 |
+
"learning_rate": 0.0009488225396630347,
|
| 176 |
+
"loss": 0.8009832501411438,
|
| 177 |
+
"step": 3584
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.17735901344048774,
|
| 181 |
+
"grad_norm": 0.22388198971748352,
|
| 182 |
+
"learning_rate": 0.0009401160421685646,
|
| 183 |
+
"loss": 0.7924661636352539,
|
| 184 |
+
"step": 3840
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.18918294766985358,
|
| 188 |
+
"grad_norm": 0.22257934510707855,
|
| 189 |
+
"learning_rate": 0.0009307725649463714,
|
| 190 |
+
"loss": 0.7945993542671204,
|
| 191 |
+
"step": 4096
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.18918294766985358,
|
| 195 |
+
"eval_loss": 0.7515348514465436,
|
| 196 |
+
"eval_mse_loss": 0.7515348514465436,
|
| 197 |
+
"flow/cos_sim": 0.7693686923479925,
|
| 198 |
+
"flow/improvement_ratio": 0.4754656074934354,
|
| 199 |
+
"flow/mag_ratio_mean": 0.7561575877884207,
|
| 200 |
+
"flow/mag_ratio_std": 0.1765889780570383,
|
| 201 |
+
"step": 4096
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.18918294766985358,
|
| 205 |
+
"eval_loss": 0.7515348514465436,
|
| 206 |
+
"eval_mse_loss": 0.7515348514465436,
|
| 207 |
+
"eval_runtime": 100.3871,
|
| 208 |
+
"eval_samples_per_second": 278.851,
|
| 209 |
+
"eval_steps_per_second": 4.363,
|
| 210 |
+
"flow/cos_sim": 0.7693686923479925,
|
| 211 |
+
"flow/improvement_ratio": 0.4754656074934354,
|
| 212 |
+
"flow/mag_ratio_mean": 0.7561575877884207,
|
| 213 |
+
"flow/mag_ratio_std": 0.1765889780570383,
|
| 214 |
+
"step": 4096
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.20100688189921945,
|
| 218 |
+
"grad_norm": 0.22517219185829163,
|
| 219 |
+
"learning_rate": 0.0009208056308063659,
|
| 220 |
+
"loss": 0.7901711463928223,
|
| 221 |
+
"step": 4352
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.2128308161285853,
|
| 225 |
+
"grad_norm": 0.2154047191143036,
|
| 226 |
+
"learning_rate": 0.0009102296648873445,
|
| 227 |
+
"loss": 0.7860417366027832,
|
| 228 |
+
"step": 4608
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 0.22465475035795113,
|
| 232 |
+
"grad_norm": 0.2263442873954773,
|
| 233 |
+
"learning_rate": 0.0008990599737794927,
|
| 234 |
+
"loss": 0.785137414932251,
|
| 235 |
+
"step": 4864
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.236478684587317,
|
| 239 |
+
"grad_norm": 0.2566796839237213,
|
| 240 |
+
"learning_rate": 0.0008873127233711644,
|
| 241 |
+
"loss": 0.7781056761741638,
|
| 242 |
+
"step": 5120
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.236478684587317,
|
| 246 |
+
"eval_loss": 0.7397306067486332,
|
| 247 |
+
"eval_mse_loss": 0.7397306067486332,
|
| 248 |
+
"flow/cos_sim": 0.7743443068304018,
|
| 249 |
+
"flow/improvement_ratio": 0.47565921052405824,
|
| 250 |
+
"flow/mag_ratio_mean": 0.7624385751545701,
|
| 251 |
+
"flow/mag_ratio_std": 0.17572799042615717,
|
| 252 |
+
"step": 5120
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.236478684587317,
|
| 256 |
+
"eval_loss": 0.7397306067486332,
|
| 257 |
+
"eval_mse_loss": 0.7397306067486332,
|
| 258 |
+
"eval_runtime": 100.3693,
|
| 259 |
+
"eval_samples_per_second": 278.9,
|
| 260 |
+
"eval_steps_per_second": 4.364,
|
| 261 |
+
"flow/cos_sim": 0.7743443068304018,
|
| 262 |
+
"flow/improvement_ratio": 0.47565921052405824,
|
| 263 |
+
"flow/mag_ratio_mean": 0.7624385751545701,
|
| 264 |
+
"flow/mag_ratio_std": 0.17572799042615717,
|
| 265 |
+
"step": 5120
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.24830261881668284,
|
| 269 |
+
"grad_norm": 0.2051643282175064,
|
| 270 |
+
"learning_rate": 0.0008750049154520011,
|
| 271 |
+
"loss": 0.7773663997650146,
|
| 272 |
+
"step": 5376
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.2601265530460487,
|
| 276 |
+
"grad_norm": 0.23271359503269196,
|
| 277 |
+
"learning_rate": 0.0008621543631062487,
|
| 278 |
+
"loss": 0.7789486050605774,
|
| 279 |
+
"step": 5632
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.27195048727541454,
|
| 283 |
+
"grad_norm": 0.22768262028694153,
|
| 284 |
+
"learning_rate": 0.0008487796649318904,
|
| 285 |
+
"loss": 0.7753859758377075,
|
| 286 |
+
"step": 5888
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 0.2837744215047804,
|
| 290 |
+
"grad_norm": 0.23808346688747406,
|
| 291 |
+
"learning_rate": 0.0008349001781229053,
|
| 292 |
+
"loss": 0.7760079503059387,
|
| 293 |
+
"step": 6144
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 0.2837744215047804,
|
| 297 |
+
"eval_loss": 0.7382390670580407,
|
| 298 |
+
"eval_mse_loss": 0.7382390670580407,
|
| 299 |
+
"flow/cos_sim": 0.7747776206225565,
|
| 300 |
+
"flow/improvement_ratio": 0.48073648889315185,
|
| 301 |
+
"flow/mag_ratio_mean": 0.7609562668354,
|
| 302 |
+
"flow/mag_ratio_std": 0.17754725645802336,
|
| 303 |
+
"step": 6144
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.2837744215047804,
|
| 307 |
+
"eval_loss": 0.7382390670580407,
|
| 308 |
+
"eval_mse_loss": 0.7382390670580407,
|
| 309 |
+
"eval_runtime": 101.1596,
|
| 310 |
+
"eval_samples_per_second": 276.721,
|
| 311 |
+
"eval_steps_per_second": 4.33,
|
| 312 |
+
"flow/cos_sim": 0.7747776206225565,
|
| 313 |
+
"flow/improvement_ratio": 0.48073648889315185,
|
| 314 |
+
"flow/mag_ratio_mean": 0.7609562668354,
|
| 315 |
+
"flow/mag_ratio_std": 0.17754725645802336,
|
| 316 |
+
"step": 6144
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 0.2955983557341462,
|
| 320 |
+
"grad_norm": 0.22078828513622284,
|
| 321 |
+
"learning_rate": 0.0008205359904536107,
|
| 322 |
+
"loss": 0.7707664966583252,
|
| 323 |
+
"step": 6400
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 0.30742228996351206,
|
| 327 |
+
"grad_norm": 0.21839560568332672,
|
| 328 |
+
"learning_rate": 0.0008057078912056363,
|
| 329 |
+
"loss": 0.7725305557250977,
|
| 330 |
+
"step": 6656
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 0.3192462241928779,
|
| 334 |
+
"grad_norm": 0.23544862866401672,
|
| 335 |
+
"learning_rate": 0.0007904373410796086,
|
| 336 |
+
"loss": 0.7722088694572449,
|
| 337 |
+
"step": 6912
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 0.3310701584222438,
|
| 341 |
+
"grad_norm": 0.23949392139911652,
|
| 342 |
+
"learning_rate": 0.0007747464411350876,
|
| 343 |
+
"loss": 0.7707281708717346,
|
| 344 |
+
"step": 7168
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.3310701584222438,
|
| 348 |
+
"eval_loss": 0.7320967807889529,
|
| 349 |
+
"eval_mse_loss": 0.7320967807889529,
|
| 350 |
+
"flow/cos_sim": 0.7774923427464211,
|
| 351 |
+
"flow/improvement_ratio": 0.48102973714538905,
|
| 352 |
+
"flow/mag_ratio_mean": 0.7606384374507486,
|
| 353 |
+
"flow/mag_ratio_std": 0.1749520302839475,
|
| 354 |
+
"step": 7168
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.3310701584222438,
|
| 358 |
+
"eval_loss": 0.7320967807889529,
|
| 359 |
+
"eval_mse_loss": 0.7320967807889529,
|
| 360 |
+
"eval_runtime": 101.0362,
|
| 361 |
+
"eval_samples_per_second": 277.059,
|
| 362 |
+
"eval_steps_per_second": 4.335,
|
| 363 |
+
"flow/cos_sim": 0.7774923427464211,
|
| 364 |
+
"flow/improvement_ratio": 0.48102973714538905,
|
| 365 |
+
"flow/mag_ratio_mean": 0.7606384374507486,
|
| 366 |
+
"flow/mag_ratio_std": 0.1749520302839475,
|
| 367 |
+
"step": 7168
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.34289409265160964,
|
| 371 |
+
"grad_norm": 0.21281319856643677,
|
| 372 |
+
"learning_rate": 0.000758657900803716,
|
| 373 |
+
"loss": 0.7705408930778503,
|
| 374 |
+
"step": 7424
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.3547180268809755,
|
| 378 |
+
"grad_norm": 0.2182183712720871,
|
| 379 |
+
"learning_rate": 0.000742195005021869,
|
| 380 |
+
"loss": 0.7691267728805542,
|
| 381 |
+
"step": 7680
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.3665419611103413,
|
| 385 |
+
"grad_norm": 0.20540344715118408,
|
| 386 |
+
"learning_rate": 0.0007253815805303786,
|
| 387 |
+
"loss": 0.7687782049179077,
|
| 388 |
+
"step": 7936
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.37836589533970716,
|
| 392 |
+
"grad_norm": 0.25242024660110474,
|
| 393 |
+
"learning_rate": 0.000708309515673374,
|
| 394 |
+
"loss": 0.7675922513008118,
|
| 395 |
+
"step": 8192
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.37836589533970716,
|
| 399 |
+
"eval_loss": 0.72715279745729,
|
| 400 |
+
"eval_mse_loss": 0.72715279745729,
|
| 401 |
+
"flow/cos_sim": 0.7793520910826992,
|
| 402 |
+
"flow/improvement_ratio": 0.47947503408612724,
|
| 403 |
+
"flow/mag_ratio_mean": 0.7625162698090349,
|
| 404 |
+
"flow/mag_ratio_std": 0.17530882001331408,
|
| 405 |
+
"step": 8192
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.37836589533970716,
|
| 409 |
+
"eval_loss": 0.72715279745729,
|
| 410 |
+
"eval_mse_loss": 0.72715279745729,
|
| 411 |
+
"eval_runtime": 100.656,
|
| 412 |
+
"eval_samples_per_second": 278.106,
|
| 413 |
+
"eval_steps_per_second": 4.351,
|
| 414 |
+
"flow/cos_sim": 0.7793520910826992,
|
| 415 |
+
"flow/improvement_ratio": 0.47947503408612724,
|
| 416 |
+
"flow/mag_ratio_mean": 0.7625162698090349,
|
| 417 |
+
"flow/mag_ratio_std": 0.17530882001331408,
|
| 418 |
+
"step": 8192
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 0.390189829569073,
|
| 422 |
+
"grad_norm": 0.18503619730472565,
|
| 423 |
+
"learning_rate": 0.0006908696365085842,
|
| 424 |
+
"loss": 0.7652720808982849,
|
| 425 |
+
"step": 8448
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.4020137637984389,
|
| 429 |
+
"grad_norm": 0.20311063528060913,
|
| 430 |
+
"learning_rate": 0.0006731535118143318,
|
| 431 |
+
"loss": 0.7671474814414978,
|
| 432 |
+
"step": 8704
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.41383769802780473,
|
| 436 |
+
"grad_norm": 0.17226819694042206,
|
| 437 |
+
"learning_rate": 0.0006551867821290267,
|
| 438 |
+
"loss": 0.7648245096206665,
|
| 439 |
+
"step": 8960
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 0.4256616322571706,
|
| 443 |
+
"grad_norm": 0.20406724512577057,
|
| 444 |
+
"learning_rate": 0.0006369954506915572,
|
| 445 |
+
"loss": 0.7627850770950317,
|
| 446 |
+
"step": 9216
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.4256616322571706,
|
| 450 |
+
"eval_loss": 0.7245442253541728,
|
| 451 |
+
"eval_mse_loss": 0.7245442253541728,
|
| 452 |
+
"flow/cos_sim": 0.7804642662610093,
|
| 453 |
+
"flow/improvement_ratio": 0.478297841671395,
|
| 454 |
+
"flow/mag_ratio_mean": 0.7637837749637969,
|
| 455 |
+
"flow/mag_ratio_std": 0.17519798711554646,
|
| 456 |
+
"step": 9216
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 0.4256616322571706,
|
| 460 |
+
"eval_loss": 0.7245442253541728,
|
| 461 |
+
"eval_mse_loss": 0.7245442253541728,
|
| 462 |
+
"eval_runtime": 100.5176,
|
| 463 |
+
"eval_samples_per_second": 278.488,
|
| 464 |
+
"eval_steps_per_second": 4.357,
|
| 465 |
+
"flow/cos_sim": 0.7804642662610093,
|
| 466 |
+
"flow/improvement_ratio": 0.478297841671395,
|
| 467 |
+
"flow/mag_ratio_mean": 0.7637837749637969,
|
| 468 |
+
"flow/mag_ratio_std": 0.17519798711554646,
|
| 469 |
+
"step": 9216
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 0.4374855664865364,
|
| 473 |
+
"grad_norm": 0.19549715518951416,
|
| 474 |
+
"learning_rate": 0.0006186058458068149,
|
| 475 |
+
"loss": 0.7646156549453735,
|
| 476 |
+
"step": 9472
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.44930950071590225,
|
| 480 |
+
"grad_norm": 0.22935132682323456,
|
| 481 |
+
"learning_rate": 0.0006000445827407526,
|
| 482 |
+
"loss": 0.7637522220611572,
|
| 483 |
+
"step": 9728
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 0.4611334349452681,
|
| 487 |
+
"grad_norm": 0.1892293095588684,
|
| 488 |
+
"learning_rate": 0.0005813385252001232,
|
| 489 |
+
"loss": 0.7614347338676453,
|
| 490 |
+
"step": 9984
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 0.472957369174634,
|
| 494 |
+
"grad_norm": 0.19596756994724274,
|
| 495 |
+
"learning_rate": 0.00056251474645265,
|
| 496 |
+
"loss": 0.7602492570877075,
|
| 497 |
+
"step": 10240
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.472957369174634,
|
| 501 |
+
"eval_loss": 0.7225635307564583,
|
| 502 |
+
"eval_mse_loss": 0.7225635307564583,
|
| 503 |
+
"flow/cos_sim": 0.781412697137763,
|
| 504 |
+
"flow/improvement_ratio": 0.48148406279958,
|
| 505 |
+
"flow/mag_ratio_mean": 0.7642885947336345,
|
| 506 |
+
"flow/mag_ratio_std": 0.17388889067657462,
|
| 507 |
+
"step": 10240
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.472957369174634,
|
| 511 |
+
"eval_loss": 0.7225635307564583,
|
| 512 |
+
"eval_mse_loss": 0.7225635307564583,
|
| 513 |
+
"eval_runtime": 100.4132,
|
| 514 |
+
"eval_samples_per_second": 278.778,
|
| 515 |
+
"eval_steps_per_second": 4.362,
|
| 516 |
+
"flow/cos_sim": 0.781412697137763,
|
| 517 |
+
"flow/improvement_ratio": 0.48148406279958,
|
| 518 |
+
"flow/mag_ratio_mean": 0.7642885947336345,
|
| 519 |
+
"flow/mag_ratio_std": 0.17388889067657462,
|
| 520 |
+
"step": 10240
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.48478130340399983,
|
| 524 |
+
"grad_norm": 0.2039279341697693,
|
| 525 |
+
"learning_rate": 0.0005436004901439003,
|
| 526 |
+
"loss": 0.7591882944107056,
|
| 527 |
+
"step": 10496
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.49660523763336567,
|
| 531 |
+
"grad_norm": 0.18449348211288452,
|
| 532 |
+
"learning_rate": 0.0005246973484120841,
|
| 533 |
+
"loss": 0.7581324577331543,
|
| 534 |
+
"step": 10752
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.5084291718627315,
|
| 538 |
+
"grad_norm": 0.18734613060951233,
|
| 539 |
+
"learning_rate": 0.0005056844377834413,
|
| 540 |
+
"loss": 0.7585782408714294,
|
| 541 |
+
"step": 11008
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.5202531060920974,
|
| 545 |
+
"grad_norm": 0.19812439382076263,
|
| 546 |
+
"learning_rate": 0.0004866633000708374,
|
| 547 |
+
"loss": 0.758400559425354,
|
| 548 |
+
"step": 11264
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.5202531060920974,
|
| 552 |
+
"eval_loss": 0.7193547954025878,
|
| 553 |
+
"eval_mse_loss": 0.7193547954025878,
|
| 554 |
+
"flow/cos_sim": 0.7829214582160183,
|
| 555 |
+
"flow/improvement_ratio": 0.4840309496747849,
|
| 556 |
+
"flow/mag_ratio_mean": 0.7646983984398515,
|
| 557 |
+
"flow/mag_ratio_std": 0.17315611838614015,
|
| 558 |
+
"step": 11264
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 0.5202531060920974,
|
| 562 |
+
"eval_loss": 0.7193547954025878,
|
| 563 |
+
"eval_mse_loss": 0.7193547954025878,
|
| 564 |
+
"eval_runtime": 100.2674,
|
| 565 |
+
"eval_samples_per_second": 279.183,
|
| 566 |
+
"eval_steps_per_second": 4.368,
|
| 567 |
+
"flow/cos_sim": 0.7829214582160183,
|
| 568 |
+
"flow/improvement_ratio": 0.4840309496747849,
|
| 569 |
+
"flow/mag_ratio_mean": 0.7646983984398515,
|
| 570 |
+
"flow/mag_ratio_std": 0.17315611838614015,
|
| 571 |
+
"step": 11264
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 0.5320770403214632,
|
| 575 |
+
"grad_norm": 0.20949678122997284,
|
| 576 |
+
"learning_rate": 0.00046766146455737116,
|
| 577 |
+
"loss": 0.756182074546814,
|
| 578 |
+
"step": 11520
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.5439009745508291,
|
| 582 |
+
"grad_norm": 0.20575636625289917,
|
| 583 |
+
"learning_rate": 0.00044870643259007823,
|
| 584 |
+
"loss": 0.7595986723899841,
|
| 585 |
+
"step": 11776
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.5557249087801949,
|
| 589 |
+
"grad_norm": 0.18551619350910187,
|
| 590 |
+
"learning_rate": 0.000429825637777245,
|
| 591 |
+
"loss": 0.7588385939598083,
|
| 592 |
+
"step": 12032
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 0.5675488430095608,
|
| 596 |
+
"grad_norm": 0.21281953155994415,
|
| 597 |
+
"learning_rate": 0.00041104640628376166,
|
| 598 |
+
"loss": 0.755641758441925,
|
| 599 |
+
"step": 12288
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.5675488430095608,
|
| 603 |
+
"eval_loss": 0.7185944062934074,
|
| 604 |
+
"eval_mse_loss": 0.7185944062934074,
|
| 605 |
+
"flow/cos_sim": 0.7829709806943048,
|
| 606 |
+
"flow/improvement_ratio": 0.4833299268598426,
|
| 607 |
+
"flow/mag_ratio_mean": 0.7666111965429837,
|
| 608 |
+
"flow/mag_ratio_std": 0.17369252120114897,
|
| 609 |
+
"step": 12288
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.5675488430095608,
|
| 613 |
+
"eval_loss": 0.7185944062934074,
|
| 614 |
+
"eval_mse_loss": 0.7185944062934074,
|
| 615 |
+
"eval_runtime": 100.2742,
|
| 616 |
+
"eval_samples_per_second": 279.165,
|
| 617 |
+
"eval_steps_per_second": 4.368,
|
| 618 |
+
"flow/cos_sim": 0.7829709806943048,
|
| 619 |
+
"flow/improvement_ratio": 0.4833299268598426,
|
| 620 |
+
"flow/mag_ratio_mean": 0.7666111965429837,
|
| 621 |
+
"flow/mag_ratio_std": 0.17369252120114897,
|
| 622 |
+
"step": 12288
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 0.5793727772389267,
|
| 626 |
+
"grad_norm": 0.1611846536397934,
|
| 627 |
+
"learning_rate": 0.00039239591728197724,
|
| 628 |
+
"loss": 0.7571735382080078,
|
| 629 |
+
"step": 12544
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 0.5911967114682924,
|
| 633 |
+
"grad_norm": 0.17743456363677979,
|
| 634 |
+
"learning_rate": 0.0003739011636152962,
|
| 635 |
+
"loss": 0.7581431269645691,
|
| 636 |
+
"step": 12800
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 0.6030206456976583,
|
| 640 |
+
"grad_norm": 0.17940299212932587,
|
| 641 |
+
"learning_rate": 0.0003556600555022182,
|
| 642 |
+
"loss": 0.7566176652908325,
|
| 643 |
+
"step": 13056
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 0.6148445799270241,
|
| 647 |
+
"grad_norm": 0.1930851936340332,
|
| 648 |
+
"learning_rate": 0.00033755594309981104,
|
| 649 |
+
"loss": 0.7547513246536255,
|
| 650 |
+
"step": 13312
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.6148445799270241,
|
| 654 |
+
"eval_loss": 0.7150395362888842,
|
| 655 |
+
"eval_mse_loss": 0.7150395362888842,
|
| 656 |
+
"flow/cos_sim": 0.7844073471685523,
|
| 657 |
+
"flow/improvement_ratio": 0.47775867540542394,
|
| 658 |
+
"flow/mag_ratio_mean": 0.7674100999146292,
|
| 659 |
+
"flow/mag_ratio_std": 0.1724472795039007,
|
| 660 |
+
"step": 13312
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.6148445799270241,
|
| 664 |
+
"eval_loss": 0.7150395362888842,
|
| 665 |
+
"eval_mse_loss": 0.7150395362888842,
|
| 666 |
+
"eval_runtime": 100.7945,
|
| 667 |
+
"eval_samples_per_second": 277.724,
|
| 668 |
+
"eval_steps_per_second": 4.345,
|
| 669 |
+
"flow/cos_sim": 0.7844073471685523,
|
| 670 |
+
"flow/improvement_ratio": 0.47775867540542394,
|
| 671 |
+
"flow/mag_ratio_mean": 0.7674100999146292,
|
| 672 |
+
"flow/mag_ratio_std": 0.1724472795039007,
|
| 673 |
+
"step": 13312
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 0.62666851415639,
|
| 677 |
+
"grad_norm": 0.17439058423042297,
|
| 678 |
+
"learning_rate": 0.000319686935899715,
|
| 679 |
+
"loss": 0.7525946497917175,
|
| 680 |
+
"step": 13568
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.6384924483857558,
|
| 684 |
+
"grad_norm": 0.16240930557250977,
|
| 685 |
+
"learning_rate": 0.0003020788957071146,
|
| 686 |
+
"loss": 0.7547707557678223,
|
| 687 |
+
"step": 13824
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 0.6503163826151217,
|
| 691 |
+
"grad_norm": 0.16282112896442413,
|
| 692 |
+
"learning_rate": 0.0002847573066297746,
|
| 693 |
+
"loss": 0.7539600729942322,
|
| 694 |
+
"step": 14080
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 0.6621403168444876,
|
| 698 |
+
"grad_norm": 0.19493776559829712,
|
| 699 |
+
"learning_rate": 0.0002677472381949038,
|
| 700 |
+
"loss": 0.753170371055603,
|
| 701 |
+
"step": 14336
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 0.6621403168444876,
|
| 705 |
+
"eval_loss": 0.7136020788319035,
|
| 706 |
+
"eval_mse_loss": 0.7136020788319035,
|
| 707 |
+
"flow/cos_sim": 0.7848518109757062,
|
| 708 |
+
"flow/improvement_ratio": 0.47987128467592477,
|
| 709 |
+
"flow/mag_ratio_mean": 0.7672100954404161,
|
| 710 |
+
"flow/mag_ratio_std": 0.17320666945280005,
|
| 711 |
+
"step": 14336
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.6621403168444876,
|
| 715 |
+
"eval_loss": 0.7136020788319035,
|
| 716 |
+
"eval_mse_loss": 0.7136020788319035,
|
| 717 |
+
"eval_runtime": 100.4201,
|
| 718 |
+
"eval_samples_per_second": 278.759,
|
| 719 |
+
"eval_steps_per_second": 4.362,
|
| 720 |
+
"flow/cos_sim": 0.7848518109757062,
|
| 721 |
+
"flow/improvement_ratio": 0.47987128467592477,
|
| 722 |
+
"flow/mag_ratio_mean": 0.7672100954404161,
|
| 723 |
+
"flow/mag_ratio_std": 0.17320666945280005,
|
| 724 |
+
"step": 14336
|
| 725 |
+
}
|
| 726 |
+
],
|
| 727 |
+
"logging_steps": 256,
|
| 728 |
+
"max_steps": 21651,
|
| 729 |
+
"num_input_tokens_seen": 0,
|
| 730 |
+
"num_train_epochs": 1,
|
| 731 |
+
"save_steps": 1024,
|
| 732 |
+
"stateful_callbacks": {
|
| 733 |
+
"TrainerControl": {
|
| 734 |
+
"args": {
|
| 735 |
+
"should_epoch_stop": false,
|
| 736 |
+
"should_evaluate": false,
|
| 737 |
+
"should_log": false,
|
| 738 |
+
"should_save": true,
|
| 739 |
+
"should_training_stop": false
|
| 740 |
+
},
|
| 741 |
+
"attributes": {}
|
| 742 |
+
}
|
| 743 |
+
},
|
| 744 |
+
"total_flos": 0.0,
|
| 745 |
+
"train_batch_size": 64,
|
| 746 |
+
"trial_name": null,
|
| 747 |
+
"trial_params": null
|
| 748 |
+
}
|
checkpoints-v1/checkpoint-14336/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:03a8c5803c6e5663005c9dc7a4c98c213c8267649ccec42581475db390cc8017
|
| 3 |
+
size 5137
|