Upload folder using huggingface_hub
Browse files- checkpoints-v4.3/checkpoint-8192/ema.safetensors +3 -0
- checkpoints-v4.3/checkpoint-8192/model.safetensors +3 -0
- checkpoints-v4.3/checkpoint-8192/optimizer.pt +3 -0
- checkpoints-v4.3/checkpoint-8192/rng_state.pth +3 -0
- checkpoints-v4.3/checkpoint-8192/scaler.pt +3 -0
- checkpoints-v4.3/checkpoint-8192/scheduler.pt +3 -0
- checkpoints-v4.3/checkpoint-8192/trainer_state.json +474 -0
- checkpoints-v4.3/checkpoint-8192/training_args.bin +3 -0
checkpoints-v4.3/checkpoint-8192/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:24749aa3a3cb39e795293be24719e592e674380377948091fcd8faeaf6ae435f
|
| 3 |
+
size 4843064
|
checkpoints-v4.3/checkpoint-8192/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:904ff3b21299d1cd750424ef50340e22e3a7b5442d189c6a6d998aafa69b1feb
|
| 3 |
+
size 4843096
|
checkpoints-v4.3/checkpoint-8192/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:384e97f24eefb7ca909bca3a92f6722ffff360d6ee4274d75c1bf14fe9589bc0
|
| 3 |
+
size 9762827
|
checkpoints-v4.3/checkpoint-8192/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0f031696cc588e3f79e8591a51b5d738ec86d9b366bc4fc90e9a96d189566e95
|
| 3 |
+
size 14645
|
checkpoints-v4.3/checkpoint-8192/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:755d5b486bfa2379c7d3018b4fe45f6a08e6cb9b1a32e33a8616611f099da1b7
|
| 3 |
+
size 1383
|
checkpoints-v4.3/checkpoint-8192/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:633f4a51e3130420dcf105f63b87d79cc31db8e4efd560f5f6c0e0209526c2e9
|
| 3 |
+
size 1465
|
checkpoints-v4.3/checkpoint-8192/trainer_state.json
ADDED
|
@@ -0,0 +1,474 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.37836589533970716,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 8192,
|
| 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": 0.12324434518814087,
|
| 15 |
+
"learning_rate": 0.000498046875,
|
| 16 |
+
"loss": 2.881023645401001,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.15631070733070374,
|
| 22 |
+
"learning_rate": 0.000998046875,
|
| 23 |
+
"loss": 2.335212469100952,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.32382041215896606,
|
| 29 |
+
"learning_rate": 0.000999640996023194,
|
| 30 |
+
"loss": 1.905767560005188,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.3657301664352417,
|
| 36 |
+
"learning_rate": 0.0009985588674043958,
|
| 37 |
+
"loss": 1.565263271331787,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_cos_loss": 0.3709636884465065,
|
| 43 |
+
"eval_loss": 1.4077782859540966,
|
| 44 |
+
"eval_mse_loss": 0.915621253589517,
|
| 45 |
+
"eval_sim_loss": 0.492157032636747,
|
| 46 |
+
"flow/cos_sim": 0.6290363127782465,
|
| 47 |
+
"flow/improvement_ratio": 0.9817231620011264,
|
| 48 |
+
"flow/mag_ratio_mean": 0.6408119081906533,
|
| 49 |
+
"flow/mag_ratio_std": 0.1935653368087664,
|
| 50 |
+
"step": 1024
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.047295736917463395,
|
| 54 |
+
"eval_cos_loss": 0.3709636884465065,
|
| 55 |
+
"eval_loss": 1.4077782859540966,
|
| 56 |
+
"eval_mse_loss": 0.915621253589517,
|
| 57 |
+
"eval_runtime": 44.3003,
|
| 58 |
+
"eval_samples_per_second": 631.892,
|
| 59 |
+
"eval_sim_loss": 0.492157032636747,
|
| 60 |
+
"eval_steps_per_second": 9.887,
|
| 61 |
+
"flow/cos_sim": 0.6290363127782465,
|
| 62 |
+
"flow/improvement_ratio": 0.9817231620011264,
|
| 63 |
+
"flow/mag_ratio_mean": 0.6408119081906533,
|
| 64 |
+
"flow/mag_ratio_std": 0.1935653368087664,
|
| 65 |
+
"step": 1024
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.05911967114682925,
|
| 69 |
+
"grad_norm": 0.8176831007003784,
|
| 70 |
+
"learning_rate": 0.0009967551747861387,
|
| 71 |
+
"loss": 1.272852897644043,
|
| 72 |
+
"step": 1280
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.0709436053761951,
|
| 76 |
+
"grad_norm": 0.5458641052246094,
|
| 77 |
+
"learning_rate": 0.000994232528651847,
|
| 78 |
+
"loss": 1.0125640630722046,
|
| 79 |
+
"step": 1536
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.08276753960556095,
|
| 83 |
+
"grad_norm": 1.5209648609161377,
|
| 84 |
+
"learning_rate": 0.0009909945800260092,
|
| 85 |
+
"loss": 0.7900597453117371,
|
| 86 |
+
"step": 1792
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.09459147383492679,
|
| 90 |
+
"grad_norm": 0.5339614152908325,
|
| 91 |
+
"learning_rate": 0.0009870460151900522,
|
| 92 |
+
"loss": 0.5982019305229187,
|
| 93 |
+
"step": 2048
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.09459147383492679,
|
| 97 |
+
"eval_cos_loss": 0.166214491882825,
|
| 98 |
+
"eval_loss": 0.5164470963826463,
|
| 99 |
+
"eval_mse_loss": 0.39311050484169563,
|
| 100 |
+
"eval_sim_loss": 0.12333659069042772,
|
| 101 |
+
"flow/cos_sim": 0.8337855092738862,
|
| 102 |
+
"flow/improvement_ratio": 0.9917055516754656,
|
| 103 |
+
"flow/mag_ratio_mean": 0.8421911159062494,
|
| 104 |
+
"flow/mag_ratio_std": 0.158547153926059,
|
| 105 |
+
"step": 2048
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.09459147383492679,
|
| 109 |
+
"eval_cos_loss": 0.166214491882825,
|
| 110 |
+
"eval_loss": 0.5164470963826463,
|
| 111 |
+
"eval_mse_loss": 0.39311050484169563,
|
| 112 |
+
"eval_runtime": 44.6023,
|
| 113 |
+
"eval_samples_per_second": 627.614,
|
| 114 |
+
"eval_sim_loss": 0.12333659069042772,
|
| 115 |
+
"eval_steps_per_second": 9.82,
|
| 116 |
+
"flow/cos_sim": 0.8337855092738862,
|
| 117 |
+
"flow/improvement_ratio": 0.9917055516754656,
|
| 118 |
+
"flow/mag_ratio_mean": 0.8421911159062494,
|
| 119 |
+
"flow/mag_ratio_std": 0.158547153926059,
|
| 120 |
+
"step": 2048
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 0.10641540806429264,
|
| 124 |
+
"grad_norm": 0.6704728007316589,
|
| 125 |
+
"learning_rate": 0.0009823925488998885,
|
| 126 |
+
"loss": 0.449043333530426,
|
| 127 |
+
"step": 2304
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 0.1182393422936585,
|
| 131 |
+
"grad_norm": 0.7579811215400696,
|
| 132 |
+
"learning_rate": 0.0009770409161149525,
|
| 133 |
+
"loss": 0.3412686586380005,
|
| 134 |
+
"step": 2560
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.13006327652302435,
|
| 138 |
+
"grad_norm": 0.5465359091758728,
|
| 139 |
+
"learning_rate": 0.0009709988622506973,
|
| 140 |
+
"loss": 0.2806977331638336,
|
| 141 |
+
"step": 2816
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.1418872107523902,
|
| 145 |
+
"grad_norm": 0.4062241017818451,
|
| 146 |
+
"learning_rate": 0.000964275131968659,
|
| 147 |
+
"loss": 0.222763791680336,
|
| 148 |
+
"step": 3072
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.1418872107523902,
|
| 152 |
+
"eval_cos_loss": 0.0714544884469275,
|
| 153 |
+
"eval_loss": 0.20498143738671526,
|
| 154 |
+
"eval_mse_loss": 0.1675494800419568,
|
| 155 |
+
"eval_sim_loss": 0.037431957672209776,
|
| 156 |
+
"flow/cos_sim": 0.9285455110682744,
|
| 157 |
+
"flow/improvement_ratio": 0.9945728493063417,
|
| 158 |
+
"flow/mag_ratio_mean": 0.9371804536477616,
|
| 159 |
+
"flow/mag_ratio_std": 0.11087696671009609,
|
| 160 |
+
"step": 3072
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.1418872107523902,
|
| 164 |
+
"eval_cos_loss": 0.0714544884469275,
|
| 165 |
+
"eval_loss": 0.20498143738671526,
|
| 166 |
+
"eval_mse_loss": 0.1675494800419568,
|
| 167 |
+
"eval_runtime": 44.6285,
|
| 168 |
+
"eval_samples_per_second": 627.246,
|
| 169 |
+
"eval_sim_loss": 0.037431957672209776,
|
| 170 |
+
"eval_steps_per_second": 9.814,
|
| 171 |
+
"flow/cos_sim": 0.9285455110682744,
|
| 172 |
+
"flow/improvement_ratio": 0.9945728493063417,
|
| 173 |
+
"flow/mag_ratio_mean": 0.9371804536477616,
|
| 174 |
+
"flow/mag_ratio_std": 0.11087696671009609,
|
| 175 |
+
"step": 3072
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.15371114498175603,
|
| 179 |
+
"grad_norm": 0.6429200172424316,
|
| 180 |
+
"learning_rate": 0.0009568794565203123,
|
| 181 |
+
"loss": 0.19131025671958923,
|
| 182 |
+
"step": 3328
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.1655350792111219,
|
| 186 |
+
"grad_norm": 0.4758129119873047,
|
| 187 |
+
"learning_rate": 0.0009488225396630347,
|
| 188 |
+
"loss": 0.16897854208946228,
|
| 189 |
+
"step": 3584
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.17735901344048774,
|
| 193 |
+
"grad_norm": 0.2732395827770233,
|
| 194 |
+
"learning_rate": 0.0009401160421685646,
|
| 195 |
+
"loss": 0.15373976528644562,
|
| 196 |
+
"step": 3840
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.18918294766985358,
|
| 200 |
+
"grad_norm": 0.3236653208732605,
|
| 201 |
+
"learning_rate": 0.0009307725649463714,
|
| 202 |
+
"loss": 0.1414027065038681,
|
| 203 |
+
"step": 4096
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.18918294766985358,
|
| 207 |
+
"eval_cos_loss": 0.05107848004323162,
|
| 208 |
+
"eval_loss": 0.138613528890969,
|
| 209 |
+
"eval_mse_loss": 0.11963267840411021,
|
| 210 |
+
"eval_sim_loss": 0.01898085068885798,
|
| 211 |
+
"flow/cos_sim": 0.9489215186469631,
|
| 212 |
+
"flow/improvement_ratio": 0.9940151711032815,
|
| 213 |
+
"flow/mag_ratio_mean": 0.9492184031227408,
|
| 214 |
+
"flow/mag_ratio_std": 0.08317253486036438,
|
| 215 |
+
"step": 4096
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.18918294766985358,
|
| 219 |
+
"eval_cos_loss": 0.05107848004323162,
|
| 220 |
+
"eval_loss": 0.138613528890969,
|
| 221 |
+
"eval_mse_loss": 0.11963267840411021,
|
| 222 |
+
"eval_runtime": 44.2576,
|
| 223 |
+
"eval_samples_per_second": 632.501,
|
| 224 |
+
"eval_sim_loss": 0.01898085068885798,
|
| 225 |
+
"eval_steps_per_second": 9.897,
|
| 226 |
+
"flow/cos_sim": 0.9489215186469631,
|
| 227 |
+
"flow/improvement_ratio": 0.9940151711032815,
|
| 228 |
+
"flow/mag_ratio_mean": 0.9492184031227408,
|
| 229 |
+
"flow/mag_ratio_std": 0.08317253486036438,
|
| 230 |
+
"step": 4096
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.20100688189921945,
|
| 234 |
+
"grad_norm": 0.2813233435153961,
|
| 235 |
+
"learning_rate": 0.0009208056308063659,
|
| 236 |
+
"loss": 0.1309625208377838,
|
| 237 |
+
"step": 4352
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.2128308161285853,
|
| 241 |
+
"grad_norm": 1.0844590663909912,
|
| 242 |
+
"learning_rate": 0.0009102296648873445,
|
| 243 |
+
"loss": 0.124587781727314,
|
| 244 |
+
"step": 4608
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.22465475035795113,
|
| 248 |
+
"grad_norm": 0.21950708329677582,
|
| 249 |
+
"learning_rate": 0.0008990599737794927,
|
| 250 |
+
"loss": 0.11813892424106598,
|
| 251 |
+
"step": 4864
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.236478684587317,
|
| 255 |
+
"grad_norm": 0.2600741684436798,
|
| 256 |
+
"learning_rate": 0.0008873127233711644,
|
| 257 |
+
"loss": 0.11322201043367386,
|
| 258 |
+
"step": 5120
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.236478684587317,
|
| 262 |
+
"eval_cos_loss": 0.04255827353016971,
|
| 263 |
+
"eval_loss": 0.11155674004391448,
|
| 264 |
+
"eval_mse_loss": 0.09941426802375546,
|
| 265 |
+
"eval_sim_loss": 0.01214247206162202,
|
| 266 |
+
"flow/cos_sim": 0.9574417269121022,
|
| 267 |
+
"flow/improvement_ratio": 0.9936971177249194,
|
| 268 |
+
"flow/mag_ratio_mean": 0.962604379000729,
|
| 269 |
+
"flow/mag_ratio_std": 0.0713872539483385,
|
| 270 |
+
"step": 5120
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.236478684587317,
|
| 274 |
+
"eval_cos_loss": 0.04255827353016971,
|
| 275 |
+
"eval_loss": 0.11155674004391448,
|
| 276 |
+
"eval_mse_loss": 0.09941426802375546,
|
| 277 |
+
"eval_runtime": 44.4031,
|
| 278 |
+
"eval_samples_per_second": 630.429,
|
| 279 |
+
"eval_sim_loss": 0.01214247206162202,
|
| 280 |
+
"eval_steps_per_second": 9.864,
|
| 281 |
+
"flow/cos_sim": 0.9574417269121022,
|
| 282 |
+
"flow/improvement_ratio": 0.9936971177249194,
|
| 283 |
+
"flow/mag_ratio_mean": 0.962604379000729,
|
| 284 |
+
"flow/mag_ratio_std": 0.0713872539483385,
|
| 285 |
+
"step": 5120
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.24830261881668284,
|
| 289 |
+
"grad_norm": 0.2182897925376892,
|
| 290 |
+
"learning_rate": 0.0008750049154520011,
|
| 291 |
+
"loss": 0.10881217569112778,
|
| 292 |
+
"step": 5376
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.2601265530460487,
|
| 296 |
+
"grad_norm": 0.2122986763715744,
|
| 297 |
+
"learning_rate": 0.0008621543631062487,
|
| 298 |
+
"loss": 0.11977330595254898,
|
| 299 |
+
"step": 5632
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 0.27195048727541454,
|
| 303 |
+
"grad_norm": 0.18399468064308167,
|
| 304 |
+
"learning_rate": 0.0008487796649318904,
|
| 305 |
+
"loss": 0.10303137451410294,
|
| 306 |
+
"step": 5888
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.2837744215047804,
|
| 310 |
+
"grad_norm": 0.22411631047725677,
|
| 311 |
+
"learning_rate": 0.0008349001781229053,
|
| 312 |
+
"loss": 0.10049661993980408,
|
| 313 |
+
"step": 6144
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.2837744215047804,
|
| 317 |
+
"eval_cos_loss": 0.03894300645719108,
|
| 318 |
+
"eval_loss": 0.09966377367916172,
|
| 319 |
+
"eval_mse_loss": 0.09095577063829932,
|
| 320 |
+
"eval_sim_loss": 0.008708002842052682,
|
| 321 |
+
"flow/cos_sim": 0.961056991390986,
|
| 322 |
+
"flow/improvement_ratio": 0.9948118436554251,
|
| 323 |
+
"flow/mag_ratio_mean": 0.9735807504283783,
|
| 324 |
+
"flow/mag_ratio_std": 0.06451816036106517,
|
| 325 |
+
"step": 6144
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.2837744215047804,
|
| 329 |
+
"eval_cos_loss": 0.03894300645719108,
|
| 330 |
+
"eval_loss": 0.09966377367916172,
|
| 331 |
+
"eval_mse_loss": 0.09095577063829932,
|
| 332 |
+
"eval_runtime": 46.781,
|
| 333 |
+
"eval_samples_per_second": 598.384,
|
| 334 |
+
"eval_sim_loss": 0.008708002842052682,
|
| 335 |
+
"eval_steps_per_second": 9.363,
|
| 336 |
+
"flow/cos_sim": 0.961056991390986,
|
| 337 |
+
"flow/improvement_ratio": 0.9948118436554251,
|
| 338 |
+
"flow/mag_ratio_mean": 0.9735807504283783,
|
| 339 |
+
"flow/mag_ratio_std": 0.06451816036106517,
|
| 340 |
+
"step": 6144
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.2955983557341462,
|
| 344 |
+
"grad_norm": 0.16586197912693024,
|
| 345 |
+
"learning_rate": 0.0008205359904536107,
|
| 346 |
+
"loss": 0.13418957591056824,
|
| 347 |
+
"step": 6400
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 0.30742228996351206,
|
| 351 |
+
"grad_norm": 0.2780066728591919,
|
| 352 |
+
"learning_rate": 0.0008057078912056363,
|
| 353 |
+
"loss": 0.09691102057695389,
|
| 354 |
+
"step": 6656
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.3192462241928779,
|
| 358 |
+
"grad_norm": 0.1937379539012909,
|
| 359 |
+
"learning_rate": 0.0007904373410796086,
|
| 360 |
+
"loss": 0.09377767890691757,
|
| 361 |
+
"step": 6912
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.3310701584222438,
|
| 365 |
+
"grad_norm": 0.5644144415855408,
|
| 366 |
+
"learning_rate": 0.0007747464411350876,
|
| 367 |
+
"loss": 0.09154396504163742,
|
| 368 |
+
"step": 7168
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.3310701584222438,
|
| 372 |
+
"eval_cos_loss": 0.03662780873035187,
|
| 373 |
+
"eval_loss": 0.0921711324794924,
|
| 374 |
+
"eval_mse_loss": 0.08544895486676529,
|
| 375 |
+
"eval_sim_loss": 0.006722177483553926,
|
| 376 |
+
"flow/cos_sim": 0.9633721935150286,
|
| 377 |
+
"flow/improvement_ratio": 0.9947773780180439,
|
| 378 |
+
"flow/mag_ratio_mean": 0.9676769978651717,
|
| 379 |
+
"flow/mag_ratio_std": 0.05803668045664215,
|
| 380 |
+
"step": 7168
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.3310701584222438,
|
| 384 |
+
"eval_cos_loss": 0.03662780873035187,
|
| 385 |
+
"eval_loss": 0.0921711324794924,
|
| 386 |
+
"eval_mse_loss": 0.08544895486676529,
|
| 387 |
+
"eval_runtime": 45.5346,
|
| 388 |
+
"eval_samples_per_second": 614.763,
|
| 389 |
+
"eval_sim_loss": 0.006722177483553926,
|
| 390 |
+
"eval_steps_per_second": 9.619,
|
| 391 |
+
"flow/cos_sim": 0.9633721935150286,
|
| 392 |
+
"flow/improvement_ratio": 0.9947773780180439,
|
| 393 |
+
"flow/mag_ratio_mean": 0.9676769978651717,
|
| 394 |
+
"flow/mag_ratio_std": 0.05803668045664215,
|
| 395 |
+
"step": 7168
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.34289409265160964,
|
| 399 |
+
"grad_norm": 0.45590853691101074,
|
| 400 |
+
"learning_rate": 0.000758657900803716,
|
| 401 |
+
"loss": 0.09103628993034363,
|
| 402 |
+
"step": 7424
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.3547180268809755,
|
| 406 |
+
"grad_norm": 0.3166259825229645,
|
| 407 |
+
"learning_rate": 0.000742195005021869,
|
| 408 |
+
"loss": 0.09036355465650558,
|
| 409 |
+
"step": 7680
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.3665419611103413,
|
| 413 |
+
"grad_norm": 0.14437530934810638,
|
| 414 |
+
"learning_rate": 0.0007253815805303786,
|
| 415 |
+
"loss": 0.08894525468349457,
|
| 416 |
+
"step": 7936
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.37836589533970716,
|
| 420 |
+
"grad_norm": 0.21061654388904572,
|
| 421 |
+
"learning_rate": 0.0007082419613901028,
|
| 422 |
+
"loss": 0.08701565861701965,
|
| 423 |
+
"step": 8192
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.37836589533970716,
|
| 427 |
+
"eval_cos_loss": 0.035134583094995044,
|
| 428 |
+
"eval_loss": 0.08715778672599901,
|
| 429 |
+
"eval_mse_loss": 0.08172946989740411,
|
| 430 |
+
"eval_sim_loss": 0.005428316740353154,
|
| 431 |
+
"flow/cos_sim": 0.964865413022368,
|
| 432 |
+
"flow/improvement_ratio": 0.9948916134496802,
|
| 433 |
+
"flow/mag_ratio_mean": 0.9663348095874263,
|
| 434 |
+
"flow/mag_ratio_std": 0.05209753118013138,
|
| 435 |
+
"step": 8192
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.37836589533970716,
|
| 439 |
+
"eval_cos_loss": 0.035134583094995044,
|
| 440 |
+
"eval_loss": 0.08715778672599901,
|
| 441 |
+
"eval_mse_loss": 0.08172946989740411,
|
| 442 |
+
"eval_runtime": 45.6593,
|
| 443 |
+
"eval_samples_per_second": 613.085,
|
| 444 |
+
"eval_sim_loss": 0.005428316740353154,
|
| 445 |
+
"eval_steps_per_second": 9.593,
|
| 446 |
+
"flow/cos_sim": 0.964865413022368,
|
| 447 |
+
"flow/improvement_ratio": 0.9948916134496802,
|
| 448 |
+
"flow/mag_ratio_mean": 0.9663348095874263,
|
| 449 |
+
"flow/mag_ratio_std": 0.05209753118013138,
|
| 450 |
+
"step": 8192
|
| 451 |
+
}
|
| 452 |
+
],
|
| 453 |
+
"logging_steps": 256,
|
| 454 |
+
"max_steps": 21651,
|
| 455 |
+
"num_input_tokens_seen": 0,
|
| 456 |
+
"num_train_epochs": 1,
|
| 457 |
+
"save_steps": 1024,
|
| 458 |
+
"stateful_callbacks": {
|
| 459 |
+
"TrainerControl": {
|
| 460 |
+
"args": {
|
| 461 |
+
"should_epoch_stop": false,
|
| 462 |
+
"should_evaluate": false,
|
| 463 |
+
"should_log": false,
|
| 464 |
+
"should_save": true,
|
| 465 |
+
"should_training_stop": false
|
| 466 |
+
},
|
| 467 |
+
"attributes": {}
|
| 468 |
+
}
|
| 469 |
+
},
|
| 470 |
+
"total_flos": 0.0,
|
| 471 |
+
"train_batch_size": 64,
|
| 472 |
+
"trial_name": null,
|
| 473 |
+
"trial_params": null
|
| 474 |
+
}
|
checkpoints-v4.3/checkpoint-8192/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ee9cb7ebeac20164f12c8a5bf77bbe01722a7871151494aba288a393bf7de081
|
| 3 |
+
size 5137
|