Upload folder using huggingface_hub
Browse files- checkpoints-v3.2/checkpoint-13312/ema.safetensors +3 -0
- checkpoints-v3.2/checkpoint-13312/model.safetensors +3 -0
- checkpoints-v3.2/checkpoint-13312/optimizer.pt +3 -0
- checkpoints-v3.2/checkpoint-13312/rng_state.pth +3 -0
- checkpoints-v3.2/checkpoint-13312/scaler.pt +3 -0
- checkpoints-v3.2/checkpoint-13312/scheduler.pt +3 -0
- checkpoints-v3.2/checkpoint-13312/trainer_state.json +723 -0
- checkpoints-v3.2/checkpoint-13312/training_args.bin +3 -0
checkpoints-v3.2/checkpoint-13312/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:60513d19bff81e6a5fcb79031f7ebae52a5769cb9754f010f5e9124df4a4da7e
|
| 3 |
+
size 550088
|
checkpoints-v3.2/checkpoint-13312/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e6280d0308e8d832f04993a87e716f6e009f7e62d78663b2fb024084c4147edb
|
| 3 |
+
size 19318680
|
checkpoints-v3.2/checkpoint-13312/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:121badc1a64dee6ae374365777046995b4dfe2da5aeafab1bb5b5cfde091737e
|
| 3 |
+
size 1176715
|
checkpoints-v3.2/checkpoint-13312/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eda5b5f997ef36bbf2c1f11b2ae3b11660b167472d3b79365952ee618890d73b
|
| 3 |
+
size 14645
|
checkpoints-v3.2/checkpoint-13312/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f52db02d1ef343c718b20949e8402af29ccf7d4ae00897235ad12dfc91f027cb
|
| 3 |
+
size 1383
|
checkpoints-v3.2/checkpoint-13312/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4f5fdf3e9e8835f981e5fcf3f2ce5d6a8ac2e3b6cfb6f151337a5e9701411146
|
| 3 |
+
size 1465
|
checkpoints-v3.2/checkpoint-13312/trainer_state.json
ADDED
|
@@ -0,0 +1,723 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.6148445799270241,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 13312,
|
| 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": 0.17750050127506256,
|
| 15 |
+
"learning_rate": 0.000498046875,
|
| 16 |
+
"loss": 1.7531359195709229,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.14886082708835602,
|
| 22 |
+
"learning_rate": 0.000998046875,
|
| 23 |
+
"loss": 1.120242953300476,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.11476853489875793,
|
| 29 |
+
"learning_rate": 0.000999640996023194,
|
| 30 |
+
"loss": 1.0460094213485718,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.26101624965667725,
|
| 36 |
+
"learning_rate": 0.0009985588674043958,
|
| 37 |
+
"loss": 1.0132286548614502,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_cos_loss": 0.3147891739459887,
|
| 43 |
+
"eval_loss": 0.9696788661283989,
|
| 44 |
+
"eval_mse_loss": 0.9696788661283989,
|
| 45 |
+
"flow/cos_sim": 0.6852108316334415,
|
| 46 |
+
"flow/improvement_ratio": 0.4729253878332164,
|
| 47 |
+
"flow/mag_ratio_mean": 0.6834825821663146,
|
| 48 |
+
"flow/mag_ratio_std": 0.18169553397589078,
|
| 49 |
+
"step": 1024
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.047295736917463395,
|
| 53 |
+
"eval_cos_loss": 0.3147891739459887,
|
| 54 |
+
"eval_loss": 0.9696788661283989,
|
| 55 |
+
"eval_mse_loss": 0.9696788661283989,
|
| 56 |
+
"eval_runtime": 37.5026,
|
| 57 |
+
"eval_samples_per_second": 746.427,
|
| 58 |
+
"eval_steps_per_second": 11.679,
|
| 59 |
+
"flow/cos_sim": 0.6852108316334415,
|
| 60 |
+
"flow/improvement_ratio": 0.4729253878332164,
|
| 61 |
+
"flow/mag_ratio_mean": 0.6834825821663146,
|
| 62 |
+
"flow/mag_ratio_std": 0.18169553397589078,
|
| 63 |
+
"step": 1024
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.05911967114682925,
|
| 67 |
+
"grad_norm": 0.32471492886543274,
|
| 68 |
+
"learning_rate": 0.0009967551747861387,
|
| 69 |
+
"loss": 0.9918397068977356,
|
| 70 |
+
"step": 1280
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.0709436053761951,
|
| 74 |
+
"grad_norm": 0.19580507278442383,
|
| 75 |
+
"learning_rate": 0.000994232528651847,
|
| 76 |
+
"loss": 0.9722790718078613,
|
| 77 |
+
"step": 1536
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.08276753960556095,
|
| 81 |
+
"grad_norm": 0.13594156503677368,
|
| 82 |
+
"learning_rate": 0.0009909945800260092,
|
| 83 |
+
"loss": 0.9547147750854492,
|
| 84 |
+
"step": 1792
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.09459147383492679,
|
| 88 |
+
"grad_norm": 0.49703535437583923,
|
| 89 |
+
"learning_rate": 0.0009870460151900522,
|
| 90 |
+
"loss": 0.9414160847663879,
|
| 91 |
+
"step": 2048
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.09459147383492679,
|
| 95 |
+
"eval_cos_loss": 0.2932011633430986,
|
| 96 |
+
"eval_loss": 0.8981472950547797,
|
| 97 |
+
"eval_mse_loss": 0.8981472950547797,
|
| 98 |
+
"flow/cos_sim": 0.7067988623767139,
|
| 99 |
+
"flow/improvement_ratio": 0.47438520363204556,
|
| 100 |
+
"flow/mag_ratio_mean": 0.6956081727868346,
|
| 101 |
+
"flow/mag_ratio_std": 0.19316728430115468,
|
| 102 |
+
"step": 2048
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.09459147383492679,
|
| 106 |
+
"eval_cos_loss": 0.2932011633430986,
|
| 107 |
+
"eval_loss": 0.8981472950547797,
|
| 108 |
+
"eval_mse_loss": 0.8981472950547797,
|
| 109 |
+
"eval_runtime": 37.712,
|
| 110 |
+
"eval_samples_per_second": 742.284,
|
| 111 |
+
"eval_steps_per_second": 11.614,
|
| 112 |
+
"flow/cos_sim": 0.7067988623767139,
|
| 113 |
+
"flow/improvement_ratio": 0.47438520363204556,
|
| 114 |
+
"flow/mag_ratio_mean": 0.6956081727868346,
|
| 115 |
+
"flow/mag_ratio_std": 0.19316728430115468,
|
| 116 |
+
"step": 2048
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.10641540806429264,
|
| 120 |
+
"grad_norm": 0.49009642004966736,
|
| 121 |
+
"learning_rate": 0.0009823925488998885,
|
| 122 |
+
"loss": 0.929905116558075,
|
| 123 |
+
"step": 2304
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.1182393422936585,
|
| 127 |
+
"grad_norm": 0.21833102405071259,
|
| 128 |
+
"learning_rate": 0.0009770409161149525,
|
| 129 |
+
"loss": 0.9224135875701904,
|
| 130 |
+
"step": 2560
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.13006327652302435,
|
| 134 |
+
"grad_norm": 0.3465236723423004,
|
| 135 |
+
"learning_rate": 0.0009709988622506973,
|
| 136 |
+
"loss": 0.9120264649391174,
|
| 137 |
+
"step": 2816
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.1418872107523902,
|
| 141 |
+
"grad_norm": 0.3057152032852173,
|
| 142 |
+
"learning_rate": 0.000964275131968659,
|
| 143 |
+
"loss": 0.9054349064826965,
|
| 144 |
+
"step": 3072
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.1418872107523902,
|
| 148 |
+
"eval_cos_loss": 0.28495094285571954,
|
| 149 |
+
"eval_loss": 0.8664921860988826,
|
| 150 |
+
"eval_mse_loss": 0.8664921860988826,
|
| 151 |
+
"flow/cos_sim": 0.7150490750733032,
|
| 152 |
+
"flow/improvement_ratio": 0.48076002510715293,
|
| 153 |
+
"flow/mag_ratio_mean": 0.7135201504785721,
|
| 154 |
+
"flow/mag_ratio_std": 0.19565439009911392,
|
| 155 |
+
"step": 3072
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.1418872107523902,
|
| 159 |
+
"eval_cos_loss": 0.28495094285571954,
|
| 160 |
+
"eval_loss": 0.8664921860988826,
|
| 161 |
+
"eval_mse_loss": 0.8664921860988826,
|
| 162 |
+
"eval_runtime": 38.3291,
|
| 163 |
+
"eval_samples_per_second": 730.333,
|
| 164 |
+
"eval_steps_per_second": 11.427,
|
| 165 |
+
"flow/cos_sim": 0.7150490750733032,
|
| 166 |
+
"flow/improvement_ratio": 0.48076002510715293,
|
| 167 |
+
"flow/mag_ratio_mean": 0.7135201504785721,
|
| 168 |
+
"flow/mag_ratio_std": 0.19565439009911392,
|
| 169 |
+
"step": 3072
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.15371114498175603,
|
| 173 |
+
"grad_norm": 0.599277138710022,
|
| 174 |
+
"learning_rate": 0.0009568794565203123,
|
| 175 |
+
"loss": 0.9003790020942688,
|
| 176 |
+
"step": 3328
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.1655350792111219,
|
| 180 |
+
"grad_norm": 0.6535385251045227,
|
| 181 |
+
"learning_rate": 0.0009488225396630347,
|
| 182 |
+
"loss": 0.897746205329895,
|
| 183 |
+
"step": 3584
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.17735901344048774,
|
| 187 |
+
"grad_norm": 0.4937800467014313,
|
| 188 |
+
"learning_rate": 0.0009401160421685646,
|
| 189 |
+
"loss": 0.8885043859481812,
|
| 190 |
+
"step": 3840
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.18918294766985358,
|
| 194 |
+
"grad_norm": 1.3498679399490356,
|
| 195 |
+
"learning_rate": 0.0009307725649463714,
|
| 196 |
+
"loss": 0.8906658887863159,
|
| 197 |
+
"step": 4096
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.18918294766985358,
|
| 201 |
+
"eval_cos_loss": 0.281152941626773,
|
| 202 |
+
"eval_loss": 0.8519824494237769,
|
| 203 |
+
"eval_mse_loss": 0.8519824494237769,
|
| 204 |
+
"flow/cos_sim": 0.718847086304399,
|
| 205 |
+
"flow/improvement_ratio": 0.47086378285601804,
|
| 206 |
+
"flow/mag_ratio_mean": 0.7225797281145505,
|
| 207 |
+
"flow/mag_ratio_std": 0.19704299981463447,
|
| 208 |
+
"step": 4096
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.18918294766985358,
|
| 212 |
+
"eval_cos_loss": 0.281152941626773,
|
| 213 |
+
"eval_loss": 0.8519824494237769,
|
| 214 |
+
"eval_mse_loss": 0.8519824494237769,
|
| 215 |
+
"eval_runtime": 37.6235,
|
| 216 |
+
"eval_samples_per_second": 744.029,
|
| 217 |
+
"eval_steps_per_second": 11.642,
|
| 218 |
+
"flow/cos_sim": 0.718847086304399,
|
| 219 |
+
"flow/improvement_ratio": 0.47086378285601804,
|
| 220 |
+
"flow/mag_ratio_mean": 0.7225797281145505,
|
| 221 |
+
"flow/mag_ratio_std": 0.19704299981463447,
|
| 222 |
+
"step": 4096
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.20100688189921945,
|
| 226 |
+
"grad_norm": 0.24068038165569305,
|
| 227 |
+
"learning_rate": 0.0009208056308063659,
|
| 228 |
+
"loss": 0.885456383228302,
|
| 229 |
+
"step": 4352
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.2128308161285853,
|
| 233 |
+
"grad_norm": 0.33502769470214844,
|
| 234 |
+
"learning_rate": 0.0009102296648873445,
|
| 235 |
+
"loss": 0.8807857632637024,
|
| 236 |
+
"step": 4608
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.22465475035795113,
|
| 240 |
+
"grad_norm": 0.176764577627182,
|
| 241 |
+
"learning_rate": 0.0008990599737794927,
|
| 242 |
+
"loss": 0.8787178993225098,
|
| 243 |
+
"step": 4864
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.236478684587317,
|
| 247 |
+
"grad_norm": 0.1856887936592102,
|
| 248 |
+
"learning_rate": 0.0008873127233711644,
|
| 249 |
+
"loss": 0.8718249797821045,
|
| 250 |
+
"step": 5120
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.236478684587317,
|
| 254 |
+
"eval_cos_loss": 0.2779194979022627,
|
| 255 |
+
"eval_loss": 0.8412302969253227,
|
| 256 |
+
"eval_mse_loss": 0.8412302969253227,
|
| 257 |
+
"flow/cos_sim": 0.7220805322742898,
|
| 258 |
+
"flow/improvement_ratio": 0.4709661888749632,
|
| 259 |
+
"flow/mag_ratio_mean": 0.7200435751105008,
|
| 260 |
+
"flow/mag_ratio_std": 0.19833092303036554,
|
| 261 |
+
"step": 5120
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.236478684587317,
|
| 265 |
+
"eval_cos_loss": 0.2779194979022627,
|
| 266 |
+
"eval_loss": 0.8412302969253227,
|
| 267 |
+
"eval_mse_loss": 0.8412302969253227,
|
| 268 |
+
"eval_runtime": 37.8514,
|
| 269 |
+
"eval_samples_per_second": 739.549,
|
| 270 |
+
"eval_steps_per_second": 11.572,
|
| 271 |
+
"flow/cos_sim": 0.7220805322742898,
|
| 272 |
+
"flow/improvement_ratio": 0.4709661888749632,
|
| 273 |
+
"flow/mag_ratio_mean": 0.7200435751105008,
|
| 274 |
+
"flow/mag_ratio_std": 0.19833092303036554,
|
| 275 |
+
"step": 5120
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.24830261881668284,
|
| 279 |
+
"grad_norm": 0.315166711807251,
|
| 280 |
+
"learning_rate": 0.0008750049154520011,
|
| 281 |
+
"loss": 0.8694944977760315,
|
| 282 |
+
"step": 5376
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.2601265530460487,
|
| 286 |
+
"grad_norm": 0.6202583909034729,
|
| 287 |
+
"learning_rate": 0.0008621543631062487,
|
| 288 |
+
"loss": 0.8698850870132446,
|
| 289 |
+
"step": 5632
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.27195048727541454,
|
| 293 |
+
"grad_norm": 0.154885396361351,
|
| 294 |
+
"learning_rate": 0.0008487796649318904,
|
| 295 |
+
"loss": 0.866125762462616,
|
| 296 |
+
"step": 5888
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.2837744215047804,
|
| 300 |
+
"grad_norm": 0.1593707948923111,
|
| 301 |
+
"learning_rate": 0.0008349001781229053,
|
| 302 |
+
"loss": 0.8656928539276123,
|
| 303 |
+
"step": 6144
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.2837744215047804,
|
| 307 |
+
"eval_cos_loss": 0.27761552666556344,
|
| 308 |
+
"eval_loss": 0.8367355261763482,
|
| 309 |
+
"eval_mse_loss": 0.8367355261763482,
|
| 310 |
+
"flow/cos_sim": 0.7223844902428318,
|
| 311 |
+
"flow/improvement_ratio": 0.47543454585282224,
|
| 312 |
+
"flow/mag_ratio_mean": 0.7271437070685435,
|
| 313 |
+
"flow/mag_ratio_std": 0.2028201749063518,
|
| 314 |
+
"step": 6144
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.2837744215047804,
|
| 318 |
+
"eval_cos_loss": 0.27761552666556344,
|
| 319 |
+
"eval_loss": 0.8367355261763482,
|
| 320 |
+
"eval_mse_loss": 0.8367355261763482,
|
| 321 |
+
"eval_runtime": 37.5912,
|
| 322 |
+
"eval_samples_per_second": 744.669,
|
| 323 |
+
"eval_steps_per_second": 11.652,
|
| 324 |
+
"flow/cos_sim": 0.7223844902428318,
|
| 325 |
+
"flow/improvement_ratio": 0.47543454585282224,
|
| 326 |
+
"flow/mag_ratio_mean": 0.7271437070685435,
|
| 327 |
+
"flow/mag_ratio_std": 0.2028201749063518,
|
| 328 |
+
"step": 6144
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.2955983557341462,
|
| 332 |
+
"grad_norm": 0.24749045073986053,
|
| 333 |
+
"learning_rate": 0.0008205359904536107,
|
| 334 |
+
"loss": 0.8599100112915039,
|
| 335 |
+
"step": 6400
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.30742228996351206,
|
| 339 |
+
"grad_norm": 0.4366367757320404,
|
| 340 |
+
"learning_rate": 0.0008057078912056363,
|
| 341 |
+
"loss": 0.8599902987480164,
|
| 342 |
+
"step": 6656
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.3192462241928779,
|
| 346 |
+
"grad_norm": 0.2574009597301483,
|
| 347 |
+
"learning_rate": 0.0007904373410796086,
|
| 348 |
+
"loss": 0.8590140342712402,
|
| 349 |
+
"step": 6912
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.3310701584222438,
|
| 353 |
+
"grad_norm": 0.24796369671821594,
|
| 354 |
+
"learning_rate": 0.0007747464411350876,
|
| 355 |
+
"loss": 0.8581823110580444,
|
| 356 |
+
"step": 7168
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.3310701584222438,
|
| 360 |
+
"eval_cos_loss": 0.27323219904752627,
|
| 361 |
+
"eval_loss": 0.8229639839091801,
|
| 362 |
+
"eval_mse_loss": 0.8229639839091801,
|
| 363 |
+
"flow/cos_sim": 0.7267678196299566,
|
| 364 |
+
"flow/improvement_ratio": 0.4771028495814702,
|
| 365 |
+
"flow/mag_ratio_mean": 0.7125569941246346,
|
| 366 |
+
"flow/mag_ratio_std": 0.20383604319944773,
|
| 367 |
+
"step": 7168
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.3310701584222438,
|
| 371 |
+
"eval_cos_loss": 0.27323219904752627,
|
| 372 |
+
"eval_loss": 0.8229639839091801,
|
| 373 |
+
"eval_mse_loss": 0.8229639839091801,
|
| 374 |
+
"eval_runtime": 37.8121,
|
| 375 |
+
"eval_samples_per_second": 740.319,
|
| 376 |
+
"eval_steps_per_second": 11.584,
|
| 377 |
+
"flow/cos_sim": 0.7267678196299566,
|
| 378 |
+
"flow/improvement_ratio": 0.4771028495814702,
|
| 379 |
+
"flow/mag_ratio_mean": 0.7125569941246346,
|
| 380 |
+
"flow/mag_ratio_std": 0.20383604319944773,
|
| 381 |
+
"step": 7168
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.34289409265160964,
|
| 385 |
+
"grad_norm": 0.7159921526908875,
|
| 386 |
+
"learning_rate": 0.000758657900803716,
|
| 387 |
+
"loss": 0.858252763748169,
|
| 388 |
+
"step": 7424
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.3547180268809755,
|
| 392 |
+
"grad_norm": 0.6132074594497681,
|
| 393 |
+
"learning_rate": 0.000742195005021869,
|
| 394 |
+
"loss": 0.8558468818664551,
|
| 395 |
+
"step": 7680
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.3665419611103413,
|
| 399 |
+
"grad_norm": 0.18163549900054932,
|
| 400 |
+
"learning_rate": 0.0007253815805303786,
|
| 401 |
+
"loss": 0.85396808385849,
|
| 402 |
+
"step": 7936
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.37836589533970716,
|
| 406 |
+
"grad_norm": 0.572221040725708,
|
| 407 |
+
"learning_rate": 0.0007082419613901028,
|
| 408 |
+
"loss": 0.8530430197715759,
|
| 409 |
+
"step": 8192
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.37836589533970716,
|
| 413 |
+
"eval_cos_loss": 0.2715048284427216,
|
| 414 |
+
"eval_loss": 0.8155939645691005,
|
| 415 |
+
"eval_mse_loss": 0.8155939645691005,
|
| 416 |
+
"flow/cos_sim": 0.7284952074153238,
|
| 417 |
+
"flow/improvement_ratio": 0.47482473080016707,
|
| 418 |
+
"flow/mag_ratio_mean": 0.724112270765653,
|
| 419 |
+
"flow/mag_ratio_std": 0.2020510737210104,
|
| 420 |
+
"step": 8192
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.37836589533970716,
|
| 424 |
+
"eval_cos_loss": 0.2715048284427216,
|
| 425 |
+
"eval_loss": 0.8155939645691005,
|
| 426 |
+
"eval_mse_loss": 0.8155939645691005,
|
| 427 |
+
"eval_runtime": 37.7077,
|
| 428 |
+
"eval_samples_per_second": 742.367,
|
| 429 |
+
"eval_steps_per_second": 11.616,
|
| 430 |
+
"flow/cos_sim": 0.7284952074153238,
|
| 431 |
+
"flow/improvement_ratio": 0.47482473080016707,
|
| 432 |
+
"flow/mag_ratio_mean": 0.724112270765653,
|
| 433 |
+
"flow/mag_ratio_std": 0.2020510737210104,
|
| 434 |
+
"step": 8192
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.390189829569073,
|
| 438 |
+
"grad_norm": 0.2026418000459671,
|
| 439 |
+
"learning_rate": 0.0006908009537632514,
|
| 440 |
+
"loss": 0.8493704199790955,
|
| 441 |
+
"step": 8448
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.4020137637984389,
|
| 445 |
+
"grad_norm": 0.5044918656349182,
|
| 446 |
+
"learning_rate": 0.0006730838000114403,
|
| 447 |
+
"loss": 0.8506228923797607,
|
| 448 |
+
"step": 8704
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.41383769802780473,
|
| 452 |
+
"grad_norm": 0.2458736002445221,
|
| 453 |
+
"learning_rate": 0.0006551161421624341,
|
| 454 |
+
"loss": 0.8482614159584045,
|
| 455 |
+
"step": 8960
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.4256616322571706,
|
| 459 |
+
"grad_norm": 0.7097583413124084,
|
| 460 |
+
"learning_rate": 0.0006369239847984517,
|
| 461 |
+
"loss": 0.846942126750946,
|
| 462 |
+
"step": 9216
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.4256616322571706,
|
| 466 |
+
"eval_cos_loss": 0.27180098626575516,
|
| 467 |
+
"eval_loss": 0.8161548636003172,
|
| 468 |
+
"eval_mse_loss": 0.8161548636003172,
|
| 469 |
+
"flow/cos_sim": 0.7281990317993512,
|
| 470 |
+
"flow/improvement_ratio": 0.4732581986003815,
|
| 471 |
+
"flow/mag_ratio_mean": 0.7231416398803937,
|
| 472 |
+
"flow/mag_ratio_std": 0.201789186425405,
|
| 473 |
+
"step": 9216
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.4256616322571706,
|
| 477 |
+
"eval_cos_loss": 0.27180098626575516,
|
| 478 |
+
"eval_loss": 0.8161548636003172,
|
| 479 |
+
"eval_mse_loss": 0.8161548636003172,
|
| 480 |
+
"eval_runtime": 37.5698,
|
| 481 |
+
"eval_samples_per_second": 745.094,
|
| 482 |
+
"eval_steps_per_second": 11.658,
|
| 483 |
+
"flow/cos_sim": 0.7281990317993512,
|
| 484 |
+
"flow/improvement_ratio": 0.4732581986003815,
|
| 485 |
+
"flow/mag_ratio_mean": 0.7231416398803937,
|
| 486 |
+
"flow/mag_ratio_std": 0.201789186425405,
|
| 487 |
+
"step": 9216
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.4374855664865364,
|
| 491 |
+
"grad_norm": 0.3167099952697754,
|
| 492 |
+
"learning_rate": 0.0006185336574197479,
|
| 493 |
+
"loss": 0.8480910062789917,
|
| 494 |
+
"step": 9472
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.44930950071590225,
|
| 498 |
+
"grad_norm": 0.4810751974582672,
|
| 499 |
+
"learning_rate": 0.0005999717763379407,
|
| 500 |
+
"loss": 0.8465522527694702,
|
| 501 |
+
"step": 9728
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.4611334349452681,
|
| 505 |
+
"grad_norm": 0.2259974628686905,
|
| 506 |
+
"learning_rate": 0.0005812652061542363,
|
| 507 |
+
"loss": 0.844083309173584,
|
| 508 |
+
"step": 9984
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.472957369174634,
|
| 512 |
+
"grad_norm": 0.6505718231201172,
|
| 513 |
+
"learning_rate": 0.0005624410208783071,
|
| 514 |
+
"loss": 0.8436377644538879,
|
| 515 |
+
"step": 10240
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.472957369174634,
|
| 519 |
+
"eval_cos_loss": 0.2699208431047936,
|
| 520 |
+
"eval_loss": 0.808695926922097,
|
| 521 |
+
"eval_mse_loss": 0.808695926922097,
|
| 522 |
+
"flow/cos_sim": 0.7300791967677199,
|
| 523 |
+
"flow/improvement_ratio": 0.47688411740951886,
|
| 524 |
+
"flow/mag_ratio_mean": 0.7248269360359401,
|
| 525 |
+
"flow/mag_ratio_std": 0.20361059002544238,
|
| 526 |
+
"step": 10240
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.472957369174634,
|
| 530 |
+
"eval_cos_loss": 0.2699208431047936,
|
| 531 |
+
"eval_loss": 0.808695926922097,
|
| 532 |
+
"eval_mse_loss": 0.808695926922097,
|
| 533 |
+
"eval_runtime": 37.5891,
|
| 534 |
+
"eval_samples_per_second": 744.71,
|
| 535 |
+
"eval_steps_per_second": 11.652,
|
| 536 |
+
"flow/cos_sim": 0.7300791967677199,
|
| 537 |
+
"flow/improvement_ratio": 0.47688411740951886,
|
| 538 |
+
"flow/mag_ratio_mean": 0.7248269360359401,
|
| 539 |
+
"flow/mag_ratio_std": 0.20361059002544238,
|
| 540 |
+
"step": 10240
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.48478130340399983,
|
| 544 |
+
"grad_norm": 0.29194557666778564,
|
| 545 |
+
"learning_rate": 0.0005435264647440881,
|
| 546 |
+
"loss": 0.8419358730316162,
|
| 547 |
+
"step": 10496
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.49660523763336567,
|
| 551 |
+
"grad_norm": 0.20549535751342773,
|
| 552 |
+
"learning_rate": 0.000524548912779213,
|
| 553 |
+
"loss": 0.8400572538375854,
|
| 554 |
+
"step": 10752
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.5084291718627315,
|
| 558 |
+
"grad_norm": 0.7953479290008545,
|
| 559 |
+
"learning_rate": 0.0005055358311851499,
|
| 560 |
+
"loss": 0.8401579260826111,
|
| 561 |
+
"step": 11008
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.5202531060920974,
|
| 565 |
+
"grad_norm": 0.14846356213092804,
|
| 566 |
+
"learning_rate": 0.0004865147375853812,
|
| 567 |
+
"loss": 0.840056836605072,
|
| 568 |
+
"step": 11264
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.5202531060920974,
|
| 572 |
+
"eval_cos_loss": 0.2704421652642559,
|
| 573 |
+
"eval_loss": 0.8102181156476339,
|
| 574 |
+
"eval_mse_loss": 0.8102181156476339,
|
| 575 |
+
"flow/cos_sim": 0.7295578481399849,
|
| 576 |
+
"flow/improvement_ratio": 0.4786316101273445,
|
| 577 |
+
"flow/mag_ratio_mean": 0.730471750780872,
|
| 578 |
+
"flow/mag_ratio_std": 0.2050654717368078,
|
| 579 |
+
"step": 11264
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.5202531060920974,
|
| 583 |
+
"eval_cos_loss": 0.2704421652642559,
|
| 584 |
+
"eval_loss": 0.8102181156476339,
|
| 585 |
+
"eval_mse_loss": 0.8102181156476339,
|
| 586 |
+
"eval_runtime": 37.702,
|
| 587 |
+
"eval_samples_per_second": 742.48,
|
| 588 |
+
"eval_steps_per_second": 11.617,
|
| 589 |
+
"flow/cos_sim": 0.7295578481399849,
|
| 590 |
+
"flow/improvement_ratio": 0.4786316101273445,
|
| 591 |
+
"flow/mag_ratio_mean": 0.730471750780872,
|
| 592 |
+
"flow/mag_ratio_std": 0.2050654717368078,
|
| 593 |
+
"step": 11264
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.5320770403214632,
|
| 597 |
+
"grad_norm": 0.49196043610572815,
|
| 598 |
+
"learning_rate": 0.0004675131611991607,
|
| 599 |
+
"loss": 0.8376814723014832,
|
| 600 |
+
"step": 11520
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.5439009745508291,
|
| 604 |
+
"grad_norm": 0.3101634383201599,
|
| 605 |
+
"learning_rate": 0.0004485586029984899,
|
| 606 |
+
"loss": 0.8409138321876526,
|
| 607 |
+
"step": 11776
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 0.5557249087801949,
|
| 611 |
+
"grad_norm": 0.6860947608947754,
|
| 612 |
+
"learning_rate": 0.00042967849590597266,
|
| 613 |
+
"loss": 0.8401282429695129,
|
| 614 |
+
"step": 12032
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.5675488430095608,
|
| 618 |
+
"grad_norm": 0.5976177453994751,
|
| 619 |
+
"learning_rate": 0.0004109001650911621,
|
| 620 |
+
"loss": 0.8376214504241943,
|
| 621 |
+
"step": 12288
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.5675488430095608,
|
| 625 |
+
"eval_cos_loss": 0.2699217140742633,
|
| 626 |
+
"eval_loss": 0.8073501400479443,
|
| 627 |
+
"eval_mse_loss": 0.8073501400479443,
|
| 628 |
+
"flow/cos_sim": 0.7300783122239047,
|
| 629 |
+
"flow/improvement_ratio": 0.4786103154154129,
|
| 630 |
+
"flow/mag_ratio_mean": 0.7241527468102164,
|
| 631 |
+
"flow/mag_ratio_std": 0.20363660513946455,
|
| 632 |
+
"step": 12288
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.5675488430095608,
|
| 636 |
+
"eval_cos_loss": 0.2699217140742633,
|
| 637 |
+
"eval_loss": 0.8073501400479443,
|
| 638 |
+
"eval_mse_loss": 0.8073501400479443,
|
| 639 |
+
"eval_runtime": 37.7433,
|
| 640 |
+
"eval_samples_per_second": 741.668,
|
| 641 |
+
"eval_steps_per_second": 11.605,
|
| 642 |
+
"flow/cos_sim": 0.7300783122239047,
|
| 643 |
+
"flow/improvement_ratio": 0.4786103154154129,
|
| 644 |
+
"flow/mag_ratio_mean": 0.7241527468102164,
|
| 645 |
+
"flow/mag_ratio_std": 0.20363660513946455,
|
| 646 |
+
"step": 12288
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.5793727772389267,
|
| 650 |
+
"grad_norm": 0.19840270280838013,
|
| 651 |
+
"learning_rate": 0.0003922507884228551,
|
| 652 |
+
"loss": 0.8377624154090881,
|
| 653 |
+
"step": 12544
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.5911967114682924,
|
| 657 |
+
"grad_norm": 0.6053916215896606,
|
| 658 |
+
"learning_rate": 0.00037375735713457723,
|
| 659 |
+
"loss": 0.838446855545044,
|
| 660 |
+
"step": 12800
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.6030206456976583,
|
| 664 |
+
"grad_norm": 0.1777360439300537,
|
| 665 |
+
"learning_rate": 0.00035544663676018276,
|
| 666 |
+
"loss": 0.8392548561096191,
|
| 667 |
+
"step": 13056
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.6148445799270241,
|
| 671 |
+
"grad_norm": 0.2625690698623657,
|
| 672 |
+
"learning_rate": 0.00033734512839611255,
|
| 673 |
+
"loss": 0.8352001309394836,
|
| 674 |
+
"step": 13312
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.6148445799270241,
|
| 678 |
+
"eval_cos_loss": 0.2678832187568216,
|
| 679 |
+
"eval_loss": 0.8018131276516065,
|
| 680 |
+
"eval_mse_loss": 0.8018131276516065,
|
| 681 |
+
"flow/cos_sim": 0.732116795565984,
|
| 682 |
+
"flow/improvement_ratio": 0.47306563096231524,
|
| 683 |
+
"flow/mag_ratio_mean": 0.730011395259535,
|
| 684 |
+
"flow/mag_ratio_std": 0.20525332758007528,
|
| 685 |
+
"step": 13312
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 0.6148445799270241,
|
| 689 |
+
"eval_cos_loss": 0.2678832187568216,
|
| 690 |
+
"eval_loss": 0.8018131276516065,
|
| 691 |
+
"eval_mse_loss": 0.8018131276516065,
|
| 692 |
+
"eval_runtime": 37.6397,
|
| 693 |
+
"eval_samples_per_second": 743.71,
|
| 694 |
+
"eval_steps_per_second": 11.637,
|
| 695 |
+
"flow/cos_sim": 0.732116795565984,
|
| 696 |
+
"flow/improvement_ratio": 0.47306563096231524,
|
| 697 |
+
"flow/mag_ratio_mean": 0.730011395259535,
|
| 698 |
+
"flow/mag_ratio_std": 0.20525332758007528,
|
| 699 |
+
"step": 13312
|
| 700 |
+
}
|
| 701 |
+
],
|
| 702 |
+
"logging_steps": 256,
|
| 703 |
+
"max_steps": 21651,
|
| 704 |
+
"num_input_tokens_seen": 0,
|
| 705 |
+
"num_train_epochs": 1,
|
| 706 |
+
"save_steps": 1024,
|
| 707 |
+
"stateful_callbacks": {
|
| 708 |
+
"TrainerControl": {
|
| 709 |
+
"args": {
|
| 710 |
+
"should_epoch_stop": false,
|
| 711 |
+
"should_evaluate": false,
|
| 712 |
+
"should_log": false,
|
| 713 |
+
"should_save": true,
|
| 714 |
+
"should_training_stop": false
|
| 715 |
+
},
|
| 716 |
+
"attributes": {}
|
| 717 |
+
}
|
| 718 |
+
},
|
| 719 |
+
"total_flos": 0.0,
|
| 720 |
+
"train_batch_size": 64,
|
| 721 |
+
"trial_name": null,
|
| 722 |
+
"trial_params": null
|
| 723 |
+
}
|
checkpoints-v3.2/checkpoint-13312/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ee9cb7ebeac20164f12c8a5bf77bbe01722a7871151494aba288a393bf7de081
|
| 3 |
+
size 5137
|