Training in progress, step 20000, checkpoint
Browse files
checkpoint-20000/config.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "openai/clip-vit-large-patch14-336",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"CLIPModel"
|
| 5 |
+
],
|
| 6 |
+
"initializer_factor": 1.0,
|
| 7 |
+
"logit_scale_init_value": 2.6592,
|
| 8 |
+
"model_type": "clip",
|
| 9 |
+
"projection_dim": 768,
|
| 10 |
+
"text_config": {
|
| 11 |
+
"dropout": 0.0,
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"intermediate_size": 3072,
|
| 14 |
+
"model_type": "clip_text_model",
|
| 15 |
+
"num_attention_heads": 12,
|
| 16 |
+
"projection_dim": 768
|
| 17 |
+
},
|
| 18 |
+
"torch_dtype": "float32",
|
| 19 |
+
"transformers_version": "4.45.0.dev0",
|
| 20 |
+
"vision_config": {
|
| 21 |
+
"dropout": 0.0,
|
| 22 |
+
"hidden_size": 1024,
|
| 23 |
+
"image_size": 336,
|
| 24 |
+
"intermediate_size": 4096,
|
| 25 |
+
"model_type": "clip_vision_model",
|
| 26 |
+
"num_attention_heads": 16,
|
| 27 |
+
"num_hidden_layers": 24,
|
| 28 |
+
"patch_size": 14,
|
| 29 |
+
"projection_dim": 768
|
| 30 |
+
}
|
| 31 |
+
}
|
checkpoint-20000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ff2112bf65f4c96c9eeb9c1f6d9bfa80ad5f10f7a8c7c28fc1777fa6c1ec3b95
|
| 3 |
+
size 1711848436
|
checkpoint-20000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a500573c50606e0a7898aba5fcbaaf01e05c21f2591d898405157cf538cd9924
|
| 3 |
+
size 3424043887
|
checkpoint-20000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad4eaa1ac460f8adba5a5367eeb11edc34cf992ad45639aa9b8336a625442250
|
| 3 |
+
size 14503
|
checkpoint-20000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:07a3b3cccdb311bfff4cf38c4bbb669692b246a481728f3e579a7a14e8a3e478
|
| 3 |
+
size 623
|
checkpoint-20000/trainer_state.json
ADDED
|
@@ -0,0 +1,633 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.4830440580844879,
|
| 3 |
+
"best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e3l57-l/checkpoint-20000",
|
| 4 |
+
"epoch": 2.1331058020477816,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 20000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.05332764505119454,
|
| 13 |
+
"grad_norm": 414.62841796875,
|
| 14 |
+
"learning_rate": 4.911120591581342e-07,
|
| 15 |
+
"loss": 0.3812,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.05332764505119454,
|
| 20 |
+
"eval_loss": 1.1163015365600586,
|
| 21 |
+
"eval_runtime": 61.8962,
|
| 22 |
+
"eval_samples_per_second": 15.946,
|
| 23 |
+
"eval_steps_per_second": 2.003,
|
| 24 |
+
"step": 500
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.10665529010238908,
|
| 28 |
+
"grad_norm": 34.08418655395508,
|
| 29 |
+
"learning_rate": 4.822241183162685e-07,
|
| 30 |
+
"loss": 0.2683,
|
| 31 |
+
"step": 1000
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.10665529010238908,
|
| 35 |
+
"eval_loss": 0.9684447646141052,
|
| 36 |
+
"eval_runtime": 61.821,
|
| 37 |
+
"eval_samples_per_second": 15.965,
|
| 38 |
+
"eval_steps_per_second": 2.006,
|
| 39 |
+
"step": 1000
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.1599829351535836,
|
| 43 |
+
"grad_norm": 350.4574890136719,
|
| 44 |
+
"learning_rate": 4.733361774744027e-07,
|
| 45 |
+
"loss": 0.2119,
|
| 46 |
+
"step": 1500
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.1599829351535836,
|
| 50 |
+
"eval_loss": 0.9099885821342468,
|
| 51 |
+
"eval_runtime": 61.9606,
|
| 52 |
+
"eval_samples_per_second": 15.929,
|
| 53 |
+
"eval_steps_per_second": 2.001,
|
| 54 |
+
"step": 1500
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.21331058020477817,
|
| 58 |
+
"grad_norm": 0.01787523180246353,
|
| 59 |
+
"learning_rate": 4.6444823663253695e-07,
|
| 60 |
+
"loss": 0.1889,
|
| 61 |
+
"step": 2000
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.21331058020477817,
|
| 65 |
+
"eval_loss": 0.8620074987411499,
|
| 66 |
+
"eval_runtime": 63.2824,
|
| 67 |
+
"eval_samples_per_second": 15.597,
|
| 68 |
+
"eval_steps_per_second": 1.959,
|
| 69 |
+
"step": 2000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.2666382252559727,
|
| 73 |
+
"grad_norm": 118.66907501220703,
|
| 74 |
+
"learning_rate": 4.5556029579067116e-07,
|
| 75 |
+
"loss": 0.2071,
|
| 76 |
+
"step": 2500
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.2666382252559727,
|
| 80 |
+
"eval_loss": 0.791786253452301,
|
| 81 |
+
"eval_runtime": 63.7932,
|
| 82 |
+
"eval_samples_per_second": 15.472,
|
| 83 |
+
"eval_steps_per_second": 1.944,
|
| 84 |
+
"step": 2500
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.3199658703071672,
|
| 88 |
+
"grad_norm": 7.21087323185543e-09,
|
| 89 |
+
"learning_rate": 4.4667235494880547e-07,
|
| 90 |
+
"loss": 0.1588,
|
| 91 |
+
"step": 3000
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.3199658703071672,
|
| 95 |
+
"eval_loss": 0.7657254338264465,
|
| 96 |
+
"eval_runtime": 62.4738,
|
| 97 |
+
"eval_samples_per_second": 15.799,
|
| 98 |
+
"eval_steps_per_second": 1.985,
|
| 99 |
+
"step": 3000
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.37329351535836175,
|
| 103 |
+
"grad_norm": 0.0014836661284789443,
|
| 104 |
+
"learning_rate": 4.377844141069397e-07,
|
| 105 |
+
"loss": 0.1718,
|
| 106 |
+
"step": 3500
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.37329351535836175,
|
| 110 |
+
"eval_loss": 0.7610095739364624,
|
| 111 |
+
"eval_runtime": 63.9118,
|
| 112 |
+
"eval_samples_per_second": 15.443,
|
| 113 |
+
"eval_steps_per_second": 1.94,
|
| 114 |
+
"step": 3500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.42662116040955633,
|
| 118 |
+
"grad_norm": 3.2739710807800293,
|
| 119 |
+
"learning_rate": 4.2889647326507393e-07,
|
| 120 |
+
"loss": 0.1113,
|
| 121 |
+
"step": 4000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.42662116040955633,
|
| 125 |
+
"eval_loss": 0.7458393573760986,
|
| 126 |
+
"eval_runtime": 62.1803,
|
| 127 |
+
"eval_samples_per_second": 15.873,
|
| 128 |
+
"eval_steps_per_second": 1.994,
|
| 129 |
+
"step": 4000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.47994880546075086,
|
| 133 |
+
"grad_norm": 1.788787841796875,
|
| 134 |
+
"learning_rate": 4.2000853242320814e-07,
|
| 135 |
+
"loss": 0.1313,
|
| 136 |
+
"step": 4500
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.47994880546075086,
|
| 140 |
+
"eval_loss": 0.7168156504631042,
|
| 141 |
+
"eval_runtime": 62.1477,
|
| 142 |
+
"eval_samples_per_second": 15.882,
|
| 143 |
+
"eval_steps_per_second": 1.995,
|
| 144 |
+
"step": 4500
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.5332764505119454,
|
| 148 |
+
"grad_norm": 4.031916250823997e-05,
|
| 149 |
+
"learning_rate": 4.1112059158134245e-07,
|
| 150 |
+
"loss": 0.1649,
|
| 151 |
+
"step": 5000
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.5332764505119454,
|
| 155 |
+
"eval_loss": 0.7019046545028687,
|
| 156 |
+
"eval_runtime": 62.2092,
|
| 157 |
+
"eval_samples_per_second": 15.866,
|
| 158 |
+
"eval_steps_per_second": 1.993,
|
| 159 |
+
"step": 5000
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.58660409556314,
|
| 163 |
+
"grad_norm": 3.883213139488362e-06,
|
| 164 |
+
"learning_rate": 4.0223265073947665e-07,
|
| 165 |
+
"loss": 0.1245,
|
| 166 |
+
"step": 5500
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.58660409556314,
|
| 170 |
+
"eval_loss": 0.6812178492546082,
|
| 171 |
+
"eval_runtime": 62.492,
|
| 172 |
+
"eval_samples_per_second": 15.794,
|
| 173 |
+
"eval_steps_per_second": 1.984,
|
| 174 |
+
"step": 5500
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.6399317406143344,
|
| 178 |
+
"grad_norm": 80.17021942138672,
|
| 179 |
+
"learning_rate": 3.933447098976109e-07,
|
| 180 |
+
"loss": 0.1286,
|
| 181 |
+
"step": 6000
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 0.6399317406143344,
|
| 185 |
+
"eval_loss": 0.6501584649085999,
|
| 186 |
+
"eval_runtime": 62.5236,
|
| 187 |
+
"eval_samples_per_second": 15.786,
|
| 188 |
+
"eval_steps_per_second": 1.983,
|
| 189 |
+
"step": 6000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.693259385665529,
|
| 193 |
+
"grad_norm": 0.0001418297761119902,
|
| 194 |
+
"learning_rate": 3.8445676905574517e-07,
|
| 195 |
+
"loss": 0.1076,
|
| 196 |
+
"step": 6500
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.693259385665529,
|
| 200 |
+
"eval_loss": 0.6153982281684875,
|
| 201 |
+
"eval_runtime": 62.2927,
|
| 202 |
+
"eval_samples_per_second": 15.845,
|
| 203 |
+
"eval_steps_per_second": 1.991,
|
| 204 |
+
"step": 6500
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.7465870307167235,
|
| 208 |
+
"grad_norm": 5.804526495012396e-07,
|
| 209 |
+
"learning_rate": 3.755688282138794e-07,
|
| 210 |
+
"loss": 0.1477,
|
| 211 |
+
"step": 7000
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 0.7465870307167235,
|
| 215 |
+
"eval_loss": 0.6118016839027405,
|
| 216 |
+
"eval_runtime": 63.1588,
|
| 217 |
+
"eval_samples_per_second": 15.627,
|
| 218 |
+
"eval_steps_per_second": 1.963,
|
| 219 |
+
"step": 7000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.7999146757679181,
|
| 223 |
+
"grad_norm": 2.66489315032959,
|
| 224 |
+
"learning_rate": 3.6668088737201363e-07,
|
| 225 |
+
"loss": 0.1315,
|
| 226 |
+
"step": 7500
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.7999146757679181,
|
| 230 |
+
"eval_loss": 0.601563036441803,
|
| 231 |
+
"eval_runtime": 62.9163,
|
| 232 |
+
"eval_samples_per_second": 15.688,
|
| 233 |
+
"eval_steps_per_second": 1.971,
|
| 234 |
+
"step": 7500
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.8532423208191127,
|
| 238 |
+
"grad_norm": 851.2547607421875,
|
| 239 |
+
"learning_rate": 3.5779294653014783e-07,
|
| 240 |
+
"loss": 0.1413,
|
| 241 |
+
"step": 8000
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.8532423208191127,
|
| 245 |
+
"eval_loss": 0.584888756275177,
|
| 246 |
+
"eval_runtime": 62.6735,
|
| 247 |
+
"eval_samples_per_second": 15.748,
|
| 248 |
+
"eval_steps_per_second": 1.979,
|
| 249 |
+
"step": 8000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.9065699658703071,
|
| 253 |
+
"grad_norm": 6.041275628376752e-05,
|
| 254 |
+
"learning_rate": 3.4890500568828214e-07,
|
| 255 |
+
"loss": 0.124,
|
| 256 |
+
"step": 8500
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.9065699658703071,
|
| 260 |
+
"eval_loss": 0.5766209363937378,
|
| 261 |
+
"eval_runtime": 63.3493,
|
| 262 |
+
"eval_samples_per_second": 15.58,
|
| 263 |
+
"eval_steps_per_second": 1.957,
|
| 264 |
+
"step": 8500
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.9598976109215017,
|
| 268 |
+
"grad_norm": 3.0901598930358887,
|
| 269 |
+
"learning_rate": 3.4001706484641635e-07,
|
| 270 |
+
"loss": 0.1215,
|
| 271 |
+
"step": 9000
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.9598976109215017,
|
| 275 |
+
"eval_loss": 0.5558739304542542,
|
| 276 |
+
"eval_runtime": 63.0862,
|
| 277 |
+
"eval_samples_per_second": 15.645,
|
| 278 |
+
"eval_steps_per_second": 1.966,
|
| 279 |
+
"step": 9000
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.0132252559726962,
|
| 283 |
+
"grad_norm": 2509.93212890625,
|
| 284 |
+
"learning_rate": 3.311291240045506e-07,
|
| 285 |
+
"loss": 0.131,
|
| 286 |
+
"step": 9500
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 1.0132252559726962,
|
| 290 |
+
"eval_loss": 0.563326358795166,
|
| 291 |
+
"eval_runtime": 62.3607,
|
| 292 |
+
"eval_samples_per_second": 15.827,
|
| 293 |
+
"eval_steps_per_second": 1.988,
|
| 294 |
+
"step": 9500
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.0665529010238908,
|
| 298 |
+
"grad_norm": 32.05735778808594,
|
| 299 |
+
"learning_rate": 3.2224118316268486e-07,
|
| 300 |
+
"loss": 0.0348,
|
| 301 |
+
"step": 10000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.0665529010238908,
|
| 305 |
+
"eval_loss": 0.553093671798706,
|
| 306 |
+
"eval_runtime": 62.9895,
|
| 307 |
+
"eval_samples_per_second": 15.669,
|
| 308 |
+
"eval_steps_per_second": 1.969,
|
| 309 |
+
"step": 10000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 1.1198805460750854,
|
| 313 |
+
"grad_norm": 0.004324838053435087,
|
| 314 |
+
"learning_rate": 3.133532423208191e-07,
|
| 315 |
+
"loss": 0.0687,
|
| 316 |
+
"step": 10500
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 1.1198805460750854,
|
| 320 |
+
"eval_loss": 0.5494681000709534,
|
| 321 |
+
"eval_runtime": 62.3211,
|
| 322 |
+
"eval_samples_per_second": 15.837,
|
| 323 |
+
"eval_steps_per_second": 1.99,
|
| 324 |
+
"step": 10500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.17320819112628,
|
| 328 |
+
"grad_norm": 0.09391242265701294,
|
| 329 |
+
"learning_rate": 3.044653014789533e-07,
|
| 330 |
+
"loss": 0.0749,
|
| 331 |
+
"step": 11000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.17320819112628,
|
| 335 |
+
"eval_loss": 0.5473881363868713,
|
| 336 |
+
"eval_runtime": 62.3454,
|
| 337 |
+
"eval_samples_per_second": 15.831,
|
| 338 |
+
"eval_steps_per_second": 1.989,
|
| 339 |
+
"step": 11000
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.2265358361774745,
|
| 343 |
+
"grad_norm": 1.9322777752850406e-11,
|
| 344 |
+
"learning_rate": 2.955773606370876e-07,
|
| 345 |
+
"loss": 0.1083,
|
| 346 |
+
"step": 11500
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 1.2265358361774745,
|
| 350 |
+
"eval_loss": 0.541618287563324,
|
| 351 |
+
"eval_runtime": 63.1868,
|
| 352 |
+
"eval_samples_per_second": 15.62,
|
| 353 |
+
"eval_steps_per_second": 1.962,
|
| 354 |
+
"step": 11500
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.2798634812286689,
|
| 358 |
+
"grad_norm": 0.0002804399700835347,
|
| 359 |
+
"learning_rate": 2.8668941979522184e-07,
|
| 360 |
+
"loss": 0.0485,
|
| 361 |
+
"step": 12000
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.2798634812286689,
|
| 365 |
+
"eval_loss": 0.5412191152572632,
|
| 366 |
+
"eval_runtime": 62.9217,
|
| 367 |
+
"eval_samples_per_second": 15.686,
|
| 368 |
+
"eval_steps_per_second": 1.971,
|
| 369 |
+
"step": 12000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.3331911262798635,
|
| 373 |
+
"grad_norm": 8.90180089641035e-08,
|
| 374 |
+
"learning_rate": 2.778014789533561e-07,
|
| 375 |
+
"loss": 0.0449,
|
| 376 |
+
"step": 12500
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 1.3331911262798635,
|
| 380 |
+
"eval_loss": 0.5510581731796265,
|
| 381 |
+
"eval_runtime": 62.525,
|
| 382 |
+
"eval_samples_per_second": 15.786,
|
| 383 |
+
"eval_steps_per_second": 1.983,
|
| 384 |
+
"step": 12500
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.386518771331058,
|
| 388 |
+
"grad_norm": 1.0178096429905303e-12,
|
| 389 |
+
"learning_rate": 2.689135381114903e-07,
|
| 390 |
+
"loss": 0.0847,
|
| 391 |
+
"step": 13000
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 1.386518771331058,
|
| 395 |
+
"eval_loss": 0.5492247939109802,
|
| 396 |
+
"eval_runtime": 62.8314,
|
| 397 |
+
"eval_samples_per_second": 15.709,
|
| 398 |
+
"eval_steps_per_second": 1.974,
|
| 399 |
+
"step": 13000
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.4398464163822526,
|
| 403 |
+
"grad_norm": 0.06238893046975136,
|
| 404 |
+
"learning_rate": 2.600255972696246e-07,
|
| 405 |
+
"loss": 0.0702,
|
| 406 |
+
"step": 13500
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.4398464163822526,
|
| 410 |
+
"eval_loss": 0.5508909821510315,
|
| 411 |
+
"eval_runtime": 63.08,
|
| 412 |
+
"eval_samples_per_second": 15.647,
|
| 413 |
+
"eval_steps_per_second": 1.966,
|
| 414 |
+
"step": 13500
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 1.493174061433447,
|
| 418 |
+
"grad_norm": 3.558750449883519e-08,
|
| 419 |
+
"learning_rate": 2.511376564277588e-07,
|
| 420 |
+
"loss": 0.0487,
|
| 421 |
+
"step": 14000
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 1.493174061433447,
|
| 425 |
+
"eval_loss": 0.544665515422821,
|
| 426 |
+
"eval_runtime": 62.2772,
|
| 427 |
+
"eval_samples_per_second": 15.849,
|
| 428 |
+
"eval_steps_per_second": 1.991,
|
| 429 |
+
"step": 14000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.5465017064846416,
|
| 433 |
+
"grad_norm": 0.29742875695228577,
|
| 434 |
+
"learning_rate": 2.422497155858931e-07,
|
| 435 |
+
"loss": 0.072,
|
| 436 |
+
"step": 14500
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.5465017064846416,
|
| 440 |
+
"eval_loss": 0.5490172505378723,
|
| 441 |
+
"eval_runtime": 62.363,
|
| 442 |
+
"eval_samples_per_second": 15.827,
|
| 443 |
+
"eval_steps_per_second": 1.988,
|
| 444 |
+
"step": 14500
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.5998293515358362,
|
| 448 |
+
"grad_norm": 664.765380859375,
|
| 449 |
+
"learning_rate": 2.333617747440273e-07,
|
| 450 |
+
"loss": 0.0325,
|
| 451 |
+
"step": 15000
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.5998293515358362,
|
| 455 |
+
"eval_loss": 0.5443011522293091,
|
| 456 |
+
"eval_runtime": 62.1023,
|
| 457 |
+
"eval_samples_per_second": 15.893,
|
| 458 |
+
"eval_steps_per_second": 1.997,
|
| 459 |
+
"step": 15000
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.6531569965870307,
|
| 463 |
+
"grad_norm": 0.0024932425003498793,
|
| 464 |
+
"learning_rate": 2.2447383390216154e-07,
|
| 465 |
+
"loss": 0.0894,
|
| 466 |
+
"step": 15500
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.6531569965870307,
|
| 470 |
+
"eval_loss": 0.5264189839363098,
|
| 471 |
+
"eval_runtime": 62.2333,
|
| 472 |
+
"eval_samples_per_second": 15.86,
|
| 473 |
+
"eval_steps_per_second": 1.993,
|
| 474 |
+
"step": 15500
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 1.7064846416382253,
|
| 478 |
+
"grad_norm": 1.1819661855697632,
|
| 479 |
+
"learning_rate": 2.155858930602958e-07,
|
| 480 |
+
"loss": 0.0503,
|
| 481 |
+
"step": 16000
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 1.7064846416382253,
|
| 485 |
+
"eval_loss": 0.52089524269104,
|
| 486 |
+
"eval_runtime": 62.2763,
|
| 487 |
+
"eval_samples_per_second": 15.849,
|
| 488 |
+
"eval_steps_per_second": 1.991,
|
| 489 |
+
"step": 16000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 1.75981228668942,
|
| 493 |
+
"grad_norm": 9.822515023572009e-11,
|
| 494 |
+
"learning_rate": 2.0669795221843002e-07,
|
| 495 |
+
"loss": 0.0568,
|
| 496 |
+
"step": 16500
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 1.75981228668942,
|
| 500 |
+
"eval_loss": 0.5083423852920532,
|
| 501 |
+
"eval_runtime": 63.6327,
|
| 502 |
+
"eval_samples_per_second": 15.511,
|
| 503 |
+
"eval_steps_per_second": 1.949,
|
| 504 |
+
"step": 16500
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 1.8131399317406145,
|
| 508 |
+
"grad_norm": 9.34051013617987e-13,
|
| 509 |
+
"learning_rate": 1.9781001137656428e-07,
|
| 510 |
+
"loss": 0.0589,
|
| 511 |
+
"step": 17000
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 1.8131399317406145,
|
| 515 |
+
"eval_loss": 0.5093010067939758,
|
| 516 |
+
"eval_runtime": 62.0049,
|
| 517 |
+
"eval_samples_per_second": 15.918,
|
| 518 |
+
"eval_steps_per_second": 2.0,
|
| 519 |
+
"step": 17000
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 1.8664675767918089,
|
| 523 |
+
"grad_norm": 0.00014494755305349827,
|
| 524 |
+
"learning_rate": 1.889220705346985e-07,
|
| 525 |
+
"loss": 0.0892,
|
| 526 |
+
"step": 17500
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 1.8664675767918089,
|
| 530 |
+
"eval_loss": 0.4983443319797516,
|
| 531 |
+
"eval_runtime": 62.0795,
|
| 532 |
+
"eval_samples_per_second": 15.899,
|
| 533 |
+
"eval_steps_per_second": 1.997,
|
| 534 |
+
"step": 17500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.9197952218430034,
|
| 538 |
+
"grad_norm": 1.963637918933614e-09,
|
| 539 |
+
"learning_rate": 1.8003412969283277e-07,
|
| 540 |
+
"loss": 0.0584,
|
| 541 |
+
"step": 18000
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.9197952218430034,
|
| 545 |
+
"eval_loss": 0.48863324522972107,
|
| 546 |
+
"eval_runtime": 63.6145,
|
| 547 |
+
"eval_samples_per_second": 15.515,
|
| 548 |
+
"eval_steps_per_second": 1.949,
|
| 549 |
+
"step": 18000
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.9731228668941978,
|
| 553 |
+
"grad_norm": 386.5431823730469,
|
| 554 |
+
"learning_rate": 1.7114618885096697e-07,
|
| 555 |
+
"loss": 0.063,
|
| 556 |
+
"step": 18500
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.9731228668941978,
|
| 560 |
+
"eval_loss": 0.4945463240146637,
|
| 561 |
+
"eval_runtime": 62.2077,
|
| 562 |
+
"eval_samples_per_second": 15.866,
|
| 563 |
+
"eval_steps_per_second": 1.993,
|
| 564 |
+
"step": 18500
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 2.0264505119453924,
|
| 568 |
+
"grad_norm": 0.10051131248474121,
|
| 569 |
+
"learning_rate": 1.6225824800910123e-07,
|
| 570 |
+
"loss": 0.0493,
|
| 571 |
+
"step": 19000
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 2.0264505119453924,
|
| 575 |
+
"eval_loss": 0.4956463575363159,
|
| 576 |
+
"eval_runtime": 62.1433,
|
| 577 |
+
"eval_samples_per_second": 15.883,
|
| 578 |
+
"eval_steps_per_second": 1.995,
|
| 579 |
+
"step": 19000
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 2.079778156996587,
|
| 583 |
+
"grad_norm": 1.1283660569461063e-05,
|
| 584 |
+
"learning_rate": 1.533703071672355e-07,
|
| 585 |
+
"loss": 0.0246,
|
| 586 |
+
"step": 19500
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 2.079778156996587,
|
| 590 |
+
"eval_loss": 0.48714280128479004,
|
| 591 |
+
"eval_runtime": 63.7786,
|
| 592 |
+
"eval_samples_per_second": 15.475,
|
| 593 |
+
"eval_steps_per_second": 1.944,
|
| 594 |
+
"step": 19500
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 2.1331058020477816,
|
| 598 |
+
"grad_norm": 2.3514085114584304e-06,
|
| 599 |
+
"learning_rate": 1.4448236632536972e-07,
|
| 600 |
+
"loss": 0.0385,
|
| 601 |
+
"step": 20000
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 2.1331058020477816,
|
| 605 |
+
"eval_loss": 0.4830440580844879,
|
| 606 |
+
"eval_runtime": 62.2754,
|
| 607 |
+
"eval_samples_per_second": 15.849,
|
| 608 |
+
"eval_steps_per_second": 1.991,
|
| 609 |
+
"step": 20000
|
| 610 |
+
}
|
| 611 |
+
],
|
| 612 |
+
"logging_steps": 500,
|
| 613 |
+
"max_steps": 28128,
|
| 614 |
+
"num_input_tokens_seen": 0,
|
| 615 |
+
"num_train_epochs": 3,
|
| 616 |
+
"save_steps": 500,
|
| 617 |
+
"stateful_callbacks": {
|
| 618 |
+
"TrainerControl": {
|
| 619 |
+
"args": {
|
| 620 |
+
"should_epoch_stop": false,
|
| 621 |
+
"should_evaluate": false,
|
| 622 |
+
"should_log": false,
|
| 623 |
+
"should_save": true,
|
| 624 |
+
"should_training_stop": false
|
| 625 |
+
},
|
| 626 |
+
"attributes": {}
|
| 627 |
+
}
|
| 628 |
+
},
|
| 629 |
+
"total_flos": 7195886541902520.0,
|
| 630 |
+
"train_batch_size": 2,
|
| 631 |
+
"trial_name": null,
|
| 632 |
+
"trial_params": null
|
| 633 |
+
}
|
checkpoint-20000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bd16a1abd112ae60a2700f590eb3bd73e96acd9a0a37b30bbfa3f8ee2dc23bc1
|
| 3 |
+
size 4847
|