Upload folder using huggingface_hub
Browse files- checkpoints-v2.8-h-2/checkpoint-21651/ema.safetensors +3 -0
- checkpoints-v2.8-h-2/checkpoint-21651/model.safetensors +3 -0
- checkpoints-v2.8-h-2/checkpoint-21651/optimizer.pt +3 -0
- checkpoints-v2.8-h-2/checkpoint-21651/rng_state.pth +3 -0
- checkpoints-v2.8-h-2/checkpoint-21651/scaler.pt +3 -0
- checkpoints-v2.8-h-2/checkpoint-21651/scheduler.pt +3 -0
- checkpoints-v2.8-h-2/checkpoint-21651/trainer_state.json +706 -0
- checkpoints-v2.8-h-2/checkpoint-21651/training_args.bin +3 -0
checkpoints-v2.8-h-2/checkpoint-21651/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f6cab88e892c6d3d398146ba3e47141b43f87d7a5854120892dae7c50a7d5a44
|
| 3 |
+
size 550088
|
checkpoints-v2.8-h-2/checkpoint-21651/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:71f293191018d1ae9cebc826c13788048803dbb6c59d84bd30c059b8f5e5a8e5
|
| 3 |
+
size 19318464
|
checkpoints-v2.8-h-2/checkpoint-21651/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2972d42888f6f806590a53719723fbe36f4b3bb3d0cff683e1abb63e5ed184c2
|
| 3 |
+
size 1175115
|
checkpoints-v2.8-h-2/checkpoint-21651/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:13e90563fac8d100230efe8ed83cb90f5dba5cae4ef0fd5b21a7a334409667e3
|
| 3 |
+
size 14645
|
checkpoints-v2.8-h-2/checkpoint-21651/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:292ad3fa2862942d57211b5dc215b794892749e4755a645c17374656a0c1b093
|
| 3 |
+
size 1383
|
checkpoints-v2.8-h-2/checkpoint-21651/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:832d6d1252c8f66c3851354151edb931a66032460323a9421e37c6a2a7b4fa65
|
| 3 |
+
size 1465
|
checkpoints-v2.8-h-2/checkpoint-21651/trainer_state.json
ADDED
|
@@ -0,0 +1,706 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 21651,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.047295736917463395,
|
| 14 |
+
"grad_norm": 0.13216529786586761,
|
| 15 |
+
"learning_rate": 0.0003330078125,
|
| 16 |
+
"loss": 2.2540314197540283,
|
| 17 |
+
"step": 1024
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.047295736917463395,
|
| 21 |
+
"eval_cos_loss": 0.5608168508364185,
|
| 22 |
+
"eval_loss": 1.840172099740538,
|
| 23 |
+
"eval_mse_loss": 1.5597636724171573,
|
| 24 |
+
"flow/cos_sim": 0.4391831588255216,
|
| 25 |
+
"flow/improvement_ratio": 0.9548978971564062,
|
| 26 |
+
"flow/mag_ratio_mean": 0.4252509521022779,
|
| 27 |
+
"flow/mag_ratio_std": 0.15304358966954767,
|
| 28 |
+
"step": 1024
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"epoch": 0.047295736917463395,
|
| 32 |
+
"eval_cos_loss": 0.5608168508364185,
|
| 33 |
+
"eval_loss": 1.840172099740538,
|
| 34 |
+
"eval_mse_loss": 1.5597636724171573,
|
| 35 |
+
"eval_runtime": 36.5199,
|
| 36 |
+
"eval_samples_per_second": 766.514,
|
| 37 |
+
"eval_steps_per_second": 11.993,
|
| 38 |
+
"flow/cos_sim": 0.4391831588255216,
|
| 39 |
+
"flow/improvement_ratio": 0.9548978971564062,
|
| 40 |
+
"flow/mag_ratio_mean": 0.4252509521022779,
|
| 41 |
+
"flow/mag_ratio_std": 0.15304358966954767,
|
| 42 |
+
"step": 1024
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 0.09459147383492679,
|
| 46 |
+
"grad_norm": 0.4216933250427246,
|
| 47 |
+
"learning_rate": 0.0006663411458333333,
|
| 48 |
+
"loss": 1.7353554964065552,
|
| 49 |
+
"step": 2048
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.09459147383492679,
|
| 53 |
+
"eval_cos_loss": 0.470765205482914,
|
| 54 |
+
"eval_loss": 1.593590806063996,
|
| 55 |
+
"eval_mse_loss": 1.3582082039689365,
|
| 56 |
+
"flow/cos_sim": 0.5292348125481714,
|
| 57 |
+
"flow/improvement_ratio": 0.9650679060585423,
|
| 58 |
+
"flow/mag_ratio_mean": 0.5083687867475971,
|
| 59 |
+
"flow/mag_ratio_std": 0.18558434835852008,
|
| 60 |
+
"step": 2048
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.09459147383492679,
|
| 64 |
+
"eval_cos_loss": 0.470765205482914,
|
| 65 |
+
"eval_loss": 1.593590806063996,
|
| 66 |
+
"eval_mse_loss": 1.3582082039689365,
|
| 67 |
+
"eval_runtime": 36.3508,
|
| 68 |
+
"eval_samples_per_second": 770.08,
|
| 69 |
+
"eval_steps_per_second": 12.049,
|
| 70 |
+
"flow/cos_sim": 0.5292348125481714,
|
| 71 |
+
"flow/improvement_ratio": 0.9650679060585423,
|
| 72 |
+
"flow/mag_ratio_mean": 0.5083687867475971,
|
| 73 |
+
"flow/mag_ratio_std": 0.18558434835852008,
|
| 74 |
+
"step": 2048
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.1418872107523902,
|
| 78 |
+
"grad_norm": 0.27868369221687317,
|
| 79 |
+
"learning_rate": 0.0009996744791666667,
|
| 80 |
+
"loss": 1.6089088916778564,
|
| 81 |
+
"step": 3072
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.1418872107523902,
|
| 85 |
+
"eval_cos_loss": 0.44682060161682025,
|
| 86 |
+
"eval_loss": 1.5164495962395517,
|
| 87 |
+
"eval_mse_loss": 1.293039296043518,
|
| 88 |
+
"flow/cos_sim": 0.5531794064121159,
|
| 89 |
+
"flow/improvement_ratio": 0.9667330909537398,
|
| 90 |
+
"flow/mag_ratio_mean": 0.5384583678691899,
|
| 91 |
+
"flow/mag_ratio_std": 0.2053203195265439,
|
| 92 |
+
"step": 3072
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.1418872107523902,
|
| 96 |
+
"eval_cos_loss": 0.44682060161682025,
|
| 97 |
+
"eval_loss": 1.5164495962395517,
|
| 98 |
+
"eval_mse_loss": 1.293039296043518,
|
| 99 |
+
"eval_runtime": 36.3439,
|
| 100 |
+
"eval_samples_per_second": 770.225,
|
| 101 |
+
"eval_steps_per_second": 12.052,
|
| 102 |
+
"flow/cos_sim": 0.5531794064121159,
|
| 103 |
+
"flow/improvement_ratio": 0.9667330909537398,
|
| 104 |
+
"flow/mag_ratio_mean": 0.5384583678691899,
|
| 105 |
+
"flow/mag_ratio_std": 0.2053203195265439,
|
| 106 |
+
"step": 3072
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.18918294766985358,
|
| 110 |
+
"grad_norm": 0.15250740945339203,
|
| 111 |
+
"learning_rate": 0.0009925378645256155,
|
| 112 |
+
"loss": 1.5571845769882202,
|
| 113 |
+
"step": 4096
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.18918294766985358,
|
| 117 |
+
"eval_cos_loss": 0.43560234120447344,
|
| 118 |
+
"eval_loss": 1.4811601312193152,
|
| 119 |
+
"eval_mse_loss": 1.2633589582356144,
|
| 120 |
+
"flow/cos_sim": 0.5643976608367816,
|
| 121 |
+
"flow/improvement_ratio": 0.967472982324966,
|
| 122 |
+
"flow/mag_ratio_mean": 0.5474448244865626,
|
| 123 |
+
"flow/mag_ratio_std": 0.21047595840746955,
|
| 124 |
+
"step": 4096
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.18918294766985358,
|
| 128 |
+
"eval_cos_loss": 0.43560234120447344,
|
| 129 |
+
"eval_loss": 1.4811601312193152,
|
| 130 |
+
"eval_mse_loss": 1.2633589582356144,
|
| 131 |
+
"eval_runtime": 37.7437,
|
| 132 |
+
"eval_samples_per_second": 741.661,
|
| 133 |
+
"eval_steps_per_second": 11.605,
|
| 134 |
+
"flow/cos_sim": 0.5643976608367816,
|
| 135 |
+
"flow/improvement_ratio": 0.967472982324966,
|
| 136 |
+
"flow/mag_ratio_mean": 0.5474448244865626,
|
| 137 |
+
"flow/mag_ratio_std": 0.21047595840746955,
|
| 138 |
+
"step": 4096
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.236478684587317,
|
| 142 |
+
"grad_norm": 0.13595673441886902,
|
| 143 |
+
"learning_rate": 0.0009703455149398919,
|
| 144 |
+
"loss": 1.531019687652588,
|
| 145 |
+
"step": 5120
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.236478684587317,
|
| 149 |
+
"eval_cos_loss": 0.42780890196697896,
|
| 150 |
+
"eval_loss": 1.4542832099683753,
|
| 151 |
+
"eval_mse_loss": 1.2403787578621956,
|
| 152 |
+
"flow/cos_sim": 0.572191093202051,
|
| 153 |
+
"flow/improvement_ratio": 0.9686366873516884,
|
| 154 |
+
"flow/mag_ratio_mean": 0.5504985138705877,
|
| 155 |
+
"flow/mag_ratio_std": 0.21518446338367245,
|
| 156 |
+
"step": 5120
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.236478684587317,
|
| 160 |
+
"eval_cos_loss": 0.42780890196697896,
|
| 161 |
+
"eval_loss": 1.4542832099683753,
|
| 162 |
+
"eval_mse_loss": 1.2403787578621956,
|
| 163 |
+
"eval_runtime": 36.5108,
|
| 164 |
+
"eval_samples_per_second": 766.704,
|
| 165 |
+
"eval_steps_per_second": 11.996,
|
| 166 |
+
"flow/cos_sim": 0.572191093202051,
|
| 167 |
+
"flow/improvement_ratio": 0.9686366873516884,
|
| 168 |
+
"flow/mag_ratio_mean": 0.5504985138705877,
|
| 169 |
+
"flow/mag_ratio_std": 0.21518446338367245,
|
| 170 |
+
"step": 5120
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.2837744215047804,
|
| 174 |
+
"grad_norm": 0.1523830145597458,
|
| 175 |
+
"learning_rate": 0.0009340866457980386,
|
| 176 |
+
"loss": 1.5127094984054565,
|
| 177 |
+
"step": 6144
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.2837744215047804,
|
| 181 |
+
"eval_cos_loss": 0.4244332621767096,
|
| 182 |
+
"eval_loss": 1.4425126316884882,
|
| 183 |
+
"eval_mse_loss": 1.2302960026754093,
|
| 184 |
+
"flow/cos_sim": 0.5755667399325871,
|
| 185 |
+
"flow/improvement_ratio": 0.9678758921688551,
|
| 186 |
+
"flow/mag_ratio_mean": 0.5517790397824762,
|
| 187 |
+
"flow/mag_ratio_std": 0.21870258596798056,
|
| 188 |
+
"step": 6144
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.2837744215047804,
|
| 192 |
+
"eval_cos_loss": 0.4244332621767096,
|
| 193 |
+
"eval_loss": 1.4425126316884882,
|
| 194 |
+
"eval_mse_loss": 1.2302960026754093,
|
| 195 |
+
"eval_runtime": 37.204,
|
| 196 |
+
"eval_samples_per_second": 752.419,
|
| 197 |
+
"eval_steps_per_second": 11.773,
|
| 198 |
+
"flow/cos_sim": 0.5755667399325871,
|
| 199 |
+
"flow/improvement_ratio": 0.9678758921688551,
|
| 200 |
+
"flow/mag_ratio_mean": 0.5517790397824762,
|
| 201 |
+
"flow/mag_ratio_std": 0.21870258596798056,
|
| 202 |
+
"step": 6144
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.3310701584222438,
|
| 206 |
+
"grad_norm": 0.3283827602863312,
|
| 207 |
+
"learning_rate": 0.0008848456431007006,
|
| 208 |
+
"loss": 1.4992624521255493,
|
| 209 |
+
"step": 7168
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.3310701584222438,
|
| 213 |
+
"eval_cos_loss": 0.42202579097388543,
|
| 214 |
+
"eval_loss": 1.4335319138553044,
|
| 215 |
+
"eval_mse_loss": 1.2225190172456715,
|
| 216 |
+
"flow/cos_sim": 0.5779742277376184,
|
| 217 |
+
"flow/improvement_ratio": 0.9681859147058774,
|
| 218 |
+
"flow/mag_ratio_mean": 0.556161379705281,
|
| 219 |
+
"flow/mag_ratio_std": 0.22328348849053797,
|
| 220 |
+
"step": 7168
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.3310701584222438,
|
| 224 |
+
"eval_cos_loss": 0.42202579097388543,
|
| 225 |
+
"eval_loss": 1.4335319138553044,
|
| 226 |
+
"eval_mse_loss": 1.2225190172456715,
|
| 227 |
+
"eval_runtime": 37.4238,
|
| 228 |
+
"eval_samples_per_second": 748.0,
|
| 229 |
+
"eval_steps_per_second": 11.704,
|
| 230 |
+
"flow/cos_sim": 0.5779742277376184,
|
| 231 |
+
"flow/improvement_ratio": 0.9681859147058774,
|
| 232 |
+
"flow/mag_ratio_mean": 0.556161379705281,
|
| 233 |
+
"flow/mag_ratio_std": 0.22328348849053797,
|
| 234 |
+
"step": 7168
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.37836589533970716,
|
| 238 |
+
"grad_norm": 0.132944718003273,
|
| 239 |
+
"learning_rate": 0.0008240951466528818,
|
| 240 |
+
"loss": 1.4886243343353271,
|
| 241 |
+
"step": 8192
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.37836589533970716,
|
| 245 |
+
"eval_cos_loss": 0.41840511710132094,
|
| 246 |
+
"eval_loss": 1.4239725503202987,
|
| 247 |
+
"eval_mse_loss": 1.21476999183768,
|
| 248 |
+
"flow/cos_sim": 0.581594914061838,
|
| 249 |
+
"flow/improvement_ratio": 0.9676709669093563,
|
| 250 |
+
"flow/mag_ratio_mean": 0.5514429012934366,
|
| 251 |
+
"flow/mag_ratio_std": 0.21821844441705643,
|
| 252 |
+
"step": 8192
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.37836589533970716,
|
| 256 |
+
"eval_cos_loss": 0.41840511710132094,
|
| 257 |
+
"eval_loss": 1.4239725503202987,
|
| 258 |
+
"eval_mse_loss": 1.21476999183768,
|
| 259 |
+
"eval_runtime": 37.4921,
|
| 260 |
+
"eval_samples_per_second": 746.637,
|
| 261 |
+
"eval_steps_per_second": 11.682,
|
| 262 |
+
"flow/cos_sim": 0.581594914061838,
|
| 263 |
+
"flow/improvement_ratio": 0.9676709669093563,
|
| 264 |
+
"flow/mag_ratio_mean": 0.5514429012934366,
|
| 265 |
+
"flow/mag_ratio_std": 0.21821844441705643,
|
| 266 |
+
"step": 8192
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.4256616322571706,
|
| 270 |
+
"grad_norm": 0.20388394594192505,
|
| 271 |
+
"learning_rate": 0.0007536520081501641,
|
| 272 |
+
"loss": 1.4808920621871948,
|
| 273 |
+
"step": 9216
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.4256616322571706,
|
| 277 |
+
"eval_cos_loss": 0.4158792528387618,
|
| 278 |
+
"eval_loss": 1.4131894620586203,
|
| 279 |
+
"eval_mse_loss": 1.20524983863308,
|
| 280 |
+
"flow/cos_sim": 0.584120759408768,
|
| 281 |
+
"flow/improvement_ratio": 0.9670752337261966,
|
| 282 |
+
"flow/mag_ratio_mean": 0.5632989504565932,
|
| 283 |
+
"flow/mag_ratio_std": 0.22649330343013485,
|
| 284 |
+
"step": 9216
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.4256616322571706,
|
| 288 |
+
"eval_cos_loss": 0.4158792528387618,
|
| 289 |
+
"eval_loss": 1.4131894620586203,
|
| 290 |
+
"eval_mse_loss": 1.20524983863308,
|
| 291 |
+
"eval_runtime": 37.6013,
|
| 292 |
+
"eval_samples_per_second": 744.468,
|
| 293 |
+
"eval_steps_per_second": 11.649,
|
| 294 |
+
"flow/cos_sim": 0.584120759408768,
|
| 295 |
+
"flow/improvement_ratio": 0.9670752337261966,
|
| 296 |
+
"flow/mag_ratio_mean": 0.5632989504565932,
|
| 297 |
+
"flow/mag_ratio_std": 0.22649330343013485,
|
| 298 |
+
"step": 9216
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 0.472957369174634,
|
| 302 |
+
"grad_norm": 0.16685892641544342,
|
| 303 |
+
"learning_rate": 0.0006756229549958484,
|
| 304 |
+
"loss": 1.4752637147903442,
|
| 305 |
+
"step": 10240
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.472957369174634,
|
| 309 |
+
"eval_cos_loss": 0.4138242877917747,
|
| 310 |
+
"eval_loss": 1.409086674043577,
|
| 311 |
+
"eval_mse_loss": 1.2021745326312165,
|
| 312 |
+
"flow/cos_sim": 0.5861757269733028,
|
| 313 |
+
"flow/improvement_ratio": 0.9699954621868047,
|
| 314 |
+
"flow/mag_ratio_mean": 0.5634166458153833,
|
| 315 |
+
"flow/mag_ratio_std": 0.22239429697598498,
|
| 316 |
+
"step": 10240
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 0.472957369174634,
|
| 320 |
+
"eval_cos_loss": 0.4138242877917747,
|
| 321 |
+
"eval_loss": 1.409086674043577,
|
| 322 |
+
"eval_mse_loss": 1.2021745326312165,
|
| 323 |
+
"eval_runtime": 37.5678,
|
| 324 |
+
"eval_samples_per_second": 745.133,
|
| 325 |
+
"eval_steps_per_second": 11.659,
|
| 326 |
+
"flow/cos_sim": 0.5861757269733028,
|
| 327 |
+
"flow/improvement_ratio": 0.9699954621868047,
|
| 328 |
+
"flow/mag_ratio_mean": 0.5634166458153833,
|
| 329 |
+
"flow/mag_ratio_std": 0.22239429697598498,
|
| 330 |
+
"step": 10240
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 0.5202531060920974,
|
| 334 |
+
"grad_norm": 0.19414471089839935,
|
| 335 |
+
"learning_rate": 0.0005923415848692534,
|
| 336 |
+
"loss": 1.4696978330612183,
|
| 337 |
+
"step": 11264
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 0.5202531060920974,
|
| 341 |
+
"eval_cos_loss": 0.4130789393295436,
|
| 342 |
+
"eval_loss": 1.4056184373489797,
|
| 343 |
+
"eval_mse_loss": 1.1990789692151491,
|
| 344 |
+
"flow/cos_sim": 0.5869210932624939,
|
| 345 |
+
"flow/improvement_ratio": 0.9684095639888555,
|
| 346 |
+
"flow/mag_ratio_mean": 0.5640594227128921,
|
| 347 |
+
"flow/mag_ratio_std": 0.22638205625967348,
|
| 348 |
+
"step": 11264
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 0.5202531060920974,
|
| 352 |
+
"eval_cos_loss": 0.4130789393295436,
|
| 353 |
+
"eval_loss": 1.4056184373489797,
|
| 354 |
+
"eval_mse_loss": 1.1990789692151491,
|
| 355 |
+
"eval_runtime": 37.4767,
|
| 356 |
+
"eval_samples_per_second": 746.944,
|
| 357 |
+
"eval_steps_per_second": 11.687,
|
| 358 |
+
"flow/cos_sim": 0.5869210932624939,
|
| 359 |
+
"flow/improvement_ratio": 0.9684095639888555,
|
| 360 |
+
"flow/mag_ratio_mean": 0.5640594227128921,
|
| 361 |
+
"flow/mag_ratio_std": 0.22638205625967348,
|
| 362 |
+
"step": 11264
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.5675488430095608,
|
| 366 |
+
"grad_norm": 0.21493041515350342,
|
| 367 |
+
"learning_rate": 0.000506383115407268,
|
| 368 |
+
"loss": 1.4641783237457275,
|
| 369 |
+
"step": 12288
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.5675488430095608,
|
| 373 |
+
"eval_cos_loss": 0.4106343852602728,
|
| 374 |
+
"eval_loss": 1.397022104426606,
|
| 375 |
+
"eval_mse_loss": 1.1917049117284277,
|
| 376 |
+
"flow/cos_sim": 0.5893656173253168,
|
| 377 |
+
"flow/improvement_ratio": 0.9678930984512312,
|
| 378 |
+
"flow/mag_ratio_mean": 0.565856612164136,
|
| 379 |
+
"flow/mag_ratio_std": 0.2252416678079187,
|
| 380 |
+
"step": 12288
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.5675488430095608,
|
| 384 |
+
"eval_cos_loss": 0.4106343852602728,
|
| 385 |
+
"eval_loss": 1.397022104426606,
|
| 386 |
+
"eval_mse_loss": 1.1917049117284277,
|
| 387 |
+
"eval_runtime": 37.5318,
|
| 388 |
+
"eval_samples_per_second": 745.848,
|
| 389 |
+
"eval_steps_per_second": 11.67,
|
| 390 |
+
"flow/cos_sim": 0.5893656173253168,
|
| 391 |
+
"flow/improvement_ratio": 0.9678930984512312,
|
| 392 |
+
"flow/mag_ratio_mean": 0.565856612164136,
|
| 393 |
+
"flow/mag_ratio_std": 0.2252416678079187,
|
| 394 |
+
"step": 12288
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.6148445799270241,
|
| 398 |
+
"grad_norm": 0.1553519070148468,
|
| 399 |
+
"learning_rate": 0.00042015065633581203,
|
| 400 |
+
"loss": 1.460407018661499,
|
| 401 |
+
"step": 13312
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.6148445799270241,
|
| 405 |
+
"eval_cos_loss": 0.41024401954047757,
|
| 406 |
+
"eval_loss": 1.3953286480141558,
|
| 407 |
+
"eval_mse_loss": 1.190206638209896,
|
| 408 |
+
"flow/cos_sim": 0.5897559979462732,
|
| 409 |
+
"flow/improvement_ratio": 0.9680638820885523,
|
| 410 |
+
"flow/mag_ratio_mean": 0.5661686862984748,
|
| 411 |
+
"flow/mag_ratio_std": 0.22554758708226627,
|
| 412 |
+
"step": 13312
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 0.6148445799270241,
|
| 416 |
+
"eval_cos_loss": 0.41024401954047757,
|
| 417 |
+
"eval_loss": 1.3953286480141558,
|
| 418 |
+
"eval_mse_loss": 1.190206638209896,
|
| 419 |
+
"eval_runtime": 37.5911,
|
| 420 |
+
"eval_samples_per_second": 744.671,
|
| 421 |
+
"eval_steps_per_second": 11.652,
|
| 422 |
+
"flow/cos_sim": 0.5897559979462732,
|
| 423 |
+
"flow/improvement_ratio": 0.9680638820885523,
|
| 424 |
+
"flow/mag_ratio_mean": 0.5661686862984748,
|
| 425 |
+
"flow/mag_ratio_std": 0.22554758708226627,
|
| 426 |
+
"step": 13312
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 0.6621403168444876,
|
| 430 |
+
"grad_norm": 0.17843684554100037,
|
| 431 |
+
"learning_rate": 0.0003363861238167604,
|
| 432 |
+
"loss": 1.456861138343811,
|
| 433 |
+
"step": 14336
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.6621403168444876,
|
| 437 |
+
"eval_cos_loss": 0.4097557538720571,
|
| 438 |
+
"eval_loss": 1.3941423579982426,
|
| 439 |
+
"eval_mse_loss": 1.1892644820147997,
|
| 440 |
+
"flow/cos_sim": 0.5902442677652455,
|
| 441 |
+
"flow/improvement_ratio": 0.9690267289882382,
|
| 442 |
+
"flow/mag_ratio_mean": 0.5675587929002771,
|
| 443 |
+
"flow/mag_ratio_std": 0.22511094835795223,
|
| 444 |
+
"step": 14336
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.6621403168444876,
|
| 448 |
+
"eval_cos_loss": 0.4097557538720571,
|
| 449 |
+
"eval_loss": 1.3941423579982426,
|
| 450 |
+
"eval_mse_loss": 1.1892644820147997,
|
| 451 |
+
"eval_runtime": 37.725,
|
| 452 |
+
"eval_samples_per_second": 742.027,
|
| 453 |
+
"eval_steps_per_second": 11.61,
|
| 454 |
+
"flow/cos_sim": 0.5902442677652455,
|
| 455 |
+
"flow/improvement_ratio": 0.9690267289882382,
|
| 456 |
+
"flow/mag_ratio_mean": 0.5675587929002771,
|
| 457 |
+
"flow/mag_ratio_std": 0.22511094835795223,
|
| 458 |
+
"step": 14336
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.709436053761951,
|
| 462 |
+
"grad_norm": 0.16084638237953186,
|
| 463 |
+
"learning_rate": 0.0002574312924773668,
|
| 464 |
+
"loss": 1.4563266038894653,
|
| 465 |
+
"step": 15360
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.709436053761951,
|
| 469 |
+
"eval_cos_loss": 0.4082320715602674,
|
| 470 |
+
"eval_loss": 1.3902559054496626,
|
| 471 |
+
"eval_mse_loss": 1.1861398674581694,
|
| 472 |
+
"flow/cos_sim": 0.5917679436130611,
|
| 473 |
+
"flow/improvement_ratio": 0.9698736070225772,
|
| 474 |
+
"flow/mag_ratio_mean": 0.5668952558955102,
|
| 475 |
+
"flow/mag_ratio_std": 0.22523612784196254,
|
| 476 |
+
"step": 15360
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.709436053761951,
|
| 480 |
+
"eval_cos_loss": 0.4082320715602674,
|
| 481 |
+
"eval_loss": 1.3902559054496626,
|
| 482 |
+
"eval_mse_loss": 1.1861398674581694,
|
| 483 |
+
"eval_runtime": 37.6782,
|
| 484 |
+
"eval_samples_per_second": 742.949,
|
| 485 |
+
"eval_steps_per_second": 11.625,
|
| 486 |
+
"flow/cos_sim": 0.5917679436130611,
|
| 487 |
+
"flow/improvement_ratio": 0.9698736070225772,
|
| 488 |
+
"flow/mag_ratio_mean": 0.5668952558955102,
|
| 489 |
+
"flow/mag_ratio_std": 0.22523612784196254,
|
| 490 |
+
"step": 15360
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 0.7567317906794143,
|
| 494 |
+
"grad_norm": 0.1939423829317093,
|
| 495 |
+
"learning_rate": 0.0001857309099627023,
|
| 496 |
+
"loss": 1.4537469148635864,
|
| 497 |
+
"step": 16384
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.7567317906794143,
|
| 501 |
+
"eval_cos_loss": 0.4083600237489291,
|
| 502 |
+
"eval_loss": 1.3896791839708476,
|
| 503 |
+
"eval_mse_loss": 1.18549917087163,
|
| 504 |
+
"flow/cos_sim": 0.5916399929893615,
|
| 505 |
+
"flow/improvement_ratio": 0.9679969251155853,
|
| 506 |
+
"flow/mag_ratio_mean": 0.5672158710488445,
|
| 507 |
+
"flow/mag_ratio_std": 0.2283102631841076,
|
| 508 |
+
"step": 16384
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.7567317906794143,
|
| 512 |
+
"eval_cos_loss": 0.4083600237489291,
|
| 513 |
+
"eval_loss": 1.3896791839708476,
|
| 514 |
+
"eval_mse_loss": 1.18549917087163,
|
| 515 |
+
"eval_runtime": 37.6784,
|
| 516 |
+
"eval_samples_per_second": 742.946,
|
| 517 |
+
"eval_steps_per_second": 11.625,
|
| 518 |
+
"flow/cos_sim": 0.5916399929893615,
|
| 519 |
+
"flow/improvement_ratio": 0.9679969251155853,
|
| 520 |
+
"flow/mag_ratio_mean": 0.5672158710488445,
|
| 521 |
+
"flow/mag_ratio_std": 0.2283102631841076,
|
| 522 |
+
"step": 16384
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.8040275275968778,
|
| 526 |
+
"grad_norm": 0.16456730663776398,
|
| 527 |
+
"learning_rate": 0.0001234293037953827,
|
| 528 |
+
"loss": 1.4506418704986572,
|
| 529 |
+
"step": 17408
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.8040275275968778,
|
| 533 |
+
"eval_cos_loss": 0.4077592275730551,
|
| 534 |
+
"eval_loss": 1.3878156499231242,
|
| 535 |
+
"eval_mse_loss": 1.1839360382458934,
|
| 536 |
+
"flow/cos_sim": 0.5922407781424588,
|
| 537 |
+
"flow/improvement_ratio": 0.9692025500345448,
|
| 538 |
+
"flow/mag_ratio_mean": 0.5673814232218756,
|
| 539 |
+
"flow/mag_ratio_std": 0.22620373494821053,
|
| 540 |
+
"step": 17408
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.8040275275968778,
|
| 544 |
+
"eval_cos_loss": 0.4077592275730551,
|
| 545 |
+
"eval_loss": 1.3878156499231242,
|
| 546 |
+
"eval_mse_loss": 1.1839360382458934,
|
| 547 |
+
"eval_runtime": 37.7046,
|
| 548 |
+
"eval_samples_per_second": 742.43,
|
| 549 |
+
"eval_steps_per_second": 11.617,
|
| 550 |
+
"flow/cos_sim": 0.5922407781424588,
|
| 551 |
+
"flow/improvement_ratio": 0.9692025500345448,
|
| 552 |
+
"flow/mag_ratio_mean": 0.5673814232218756,
|
| 553 |
+
"flow/mag_ratio_std": 0.22620373494821053,
|
| 554 |
+
"step": 17408
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.8513232645143411,
|
| 558 |
+
"grad_norm": 0.16574722528457642,
|
| 559 |
+
"learning_rate": 7.243353809647096e-05,
|
| 560 |
+
"loss": 1.450889229774475,
|
| 561 |
+
"step": 18432
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.8513232645143411,
|
| 565 |
+
"eval_cos_loss": 0.4075628997392306,
|
| 566 |
+
"eval_loss": 1.389668641025073,
|
| 567 |
+
"eval_mse_loss": 1.185887193026608,
|
| 568 |
+
"flow/cos_sim": 0.5924371161145162,
|
| 569 |
+
"flow/improvement_ratio": 0.9703709175597587,
|
| 570 |
+
"flow/mag_ratio_mean": 0.5677012568467283,
|
| 571 |
+
"flow/mag_ratio_std": 0.22380314015498445,
|
| 572 |
+
"step": 18432
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.8513232645143411,
|
| 576 |
+
"eval_cos_loss": 0.4075628997392306,
|
| 577 |
+
"eval_loss": 1.389668641025073,
|
| 578 |
+
"eval_mse_loss": 1.185887193026608,
|
| 579 |
+
"eval_runtime": 37.4946,
|
| 580 |
+
"eval_samples_per_second": 746.588,
|
| 581 |
+
"eval_steps_per_second": 11.682,
|
| 582 |
+
"flow/cos_sim": 0.5924371161145162,
|
| 583 |
+
"flow/improvement_ratio": 0.9703709175597587,
|
| 584 |
+
"flow/mag_ratio_mean": 0.5677012568467283,
|
| 585 |
+
"flow/mag_ratio_std": 0.22380314015498445,
|
| 586 |
+
"step": 18432
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 0.8986190014318045,
|
| 590 |
+
"grad_norm": 0.1395236700773239,
|
| 591 |
+
"learning_rate": 3.416928297501892e-05,
|
| 592 |
+
"loss": 1.4504221677780151,
|
| 593 |
+
"step": 19456
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.8986190014318045,
|
| 597 |
+
"eval_cos_loss": 0.4083901912518288,
|
| 598 |
+
"eval_loss": 1.390020596926615,
|
| 599 |
+
"eval_mse_loss": 1.185825499497592,
|
| 600 |
+
"flow/cos_sim": 0.5916098165729818,
|
| 601 |
+
"flow/improvement_ratio": 0.9690576137200883,
|
| 602 |
+
"flow/mag_ratio_mean": 0.5676344796402814,
|
| 603 |
+
"flow/mag_ratio_std": 0.22553705900363183,
|
| 604 |
+
"step": 19456
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.8986190014318045,
|
| 608 |
+
"eval_cos_loss": 0.4083901912518288,
|
| 609 |
+
"eval_loss": 1.390020596926615,
|
| 610 |
+
"eval_mse_loss": 1.185825499497592,
|
| 611 |
+
"eval_runtime": 37.3515,
|
| 612 |
+
"eval_samples_per_second": 749.448,
|
| 613 |
+
"eval_steps_per_second": 11.726,
|
| 614 |
+
"flow/cos_sim": 0.5916098165729818,
|
| 615 |
+
"flow/improvement_ratio": 0.9690576137200883,
|
| 616 |
+
"flow/mag_ratio_mean": 0.5676344796402814,
|
| 617 |
+
"flow/mag_ratio_std": 0.22553705900363183,
|
| 618 |
+
"step": 19456
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.945914738349268,
|
| 622 |
+
"grad_norm": 0.1532638967037201,
|
| 623 |
+
"learning_rate": 9.853219257125512e-06,
|
| 624 |
+
"loss": 1.4503555297851562,
|
| 625 |
+
"step": 20480
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.945914738349268,
|
| 629 |
+
"eval_cos_loss": 0.4074147270830799,
|
| 630 |
+
"eval_loss": 1.3868377687724214,
|
| 631 |
+
"eval_mse_loss": 1.1831304028153964,
|
| 632 |
+
"flow/cos_sim": 0.5925852951665992,
|
| 633 |
+
"flow/improvement_ratio": 0.9679995860955487,
|
| 634 |
+
"flow/mag_ratio_mean": 0.5671717142133408,
|
| 635 |
+
"flow/mag_ratio_std": 0.2263773354833529,
|
| 636 |
+
"step": 20480
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 0.945914738349268,
|
| 640 |
+
"eval_cos_loss": 0.4074147270830799,
|
| 641 |
+
"eval_loss": 1.3868377687724214,
|
| 642 |
+
"eval_mse_loss": 1.1831304028153964,
|
| 643 |
+
"eval_runtime": 37.367,
|
| 644 |
+
"eval_samples_per_second": 749.138,
|
| 645 |
+
"eval_steps_per_second": 11.722,
|
| 646 |
+
"flow/cos_sim": 0.5925852951665992,
|
| 647 |
+
"flow/improvement_ratio": 0.9679995860955487,
|
| 648 |
+
"flow/mag_ratio_mean": 0.5671717142133408,
|
| 649 |
+
"flow/mag_ratio_std": 0.2263773354833529,
|
| 650 |
+
"step": 20480
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.9932104752667313,
|
| 654 |
+
"grad_norm": 0.11826281249523163,
|
| 655 |
+
"learning_rate": 1.651423088821491e-07,
|
| 656 |
+
"loss": 1.450731635093689,
|
| 657 |
+
"step": 21504
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.9932104752667313,
|
| 661 |
+
"eval_cos_loss": 0.4083047036983107,
|
| 662 |
+
"eval_loss": 1.3886951202671278,
|
| 663 |
+
"eval_mse_loss": 1.184542767805596,
|
| 664 |
+
"flow/cos_sim": 0.5916953210689161,
|
| 665 |
+
"flow/improvement_ratio": 0.9686407382901945,
|
| 666 |
+
"flow/mag_ratio_mean": 0.5667071160116152,
|
| 667 |
+
"flow/mag_ratio_std": 0.22649903095340077,
|
| 668 |
+
"step": 21504
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.9932104752667313,
|
| 672 |
+
"eval_cos_loss": 0.4083047036983107,
|
| 673 |
+
"eval_loss": 1.3886951202671278,
|
| 674 |
+
"eval_mse_loss": 1.184542767805596,
|
| 675 |
+
"eval_runtime": 37.4955,
|
| 676 |
+
"eval_samples_per_second": 746.569,
|
| 677 |
+
"eval_steps_per_second": 11.681,
|
| 678 |
+
"flow/cos_sim": 0.5916953210689161,
|
| 679 |
+
"flow/improvement_ratio": 0.9686407382901945,
|
| 680 |
+
"flow/mag_ratio_mean": 0.5667071160116152,
|
| 681 |
+
"flow/mag_ratio_std": 0.22649903095340077,
|
| 682 |
+
"step": 21504
|
| 683 |
+
}
|
| 684 |
+
],
|
| 685 |
+
"logging_steps": 1024,
|
| 686 |
+
"max_steps": 21651,
|
| 687 |
+
"num_input_tokens_seen": 0,
|
| 688 |
+
"num_train_epochs": 1,
|
| 689 |
+
"save_steps": 1024,
|
| 690 |
+
"stateful_callbacks": {
|
| 691 |
+
"TrainerControl": {
|
| 692 |
+
"args": {
|
| 693 |
+
"should_epoch_stop": false,
|
| 694 |
+
"should_evaluate": false,
|
| 695 |
+
"should_log": false,
|
| 696 |
+
"should_save": true,
|
| 697 |
+
"should_training_stop": true
|
| 698 |
+
},
|
| 699 |
+
"attributes": {}
|
| 700 |
+
}
|
| 701 |
+
},
|
| 702 |
+
"total_flos": 0.0,
|
| 703 |
+
"train_batch_size": 64,
|
| 704 |
+
"trial_name": null,
|
| 705 |
+
"trial_params": null
|
| 706 |
+
}
|
checkpoints-v2.8-h-2/checkpoint-21651/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bd8764e705e3d3a01283602f7bbb7db8de6272269197e8d0fb5f615cd86459fe
|
| 3 |
+
size 5137
|