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 sha256:73e0b58fe79a147ffe82fabfe7651ebfccc723170efda8338f8272212671787a
|
| 3 |
+
size 1711848436
|
checkpoint-20000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1368a789d412c606b85038c15f9aeab964538bde12ca85ef03bf5e8e4ce47536
|
| 3 |
+
size 3424043887
|
checkpoint-20000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0112b25beedbe5c58a79d56a364c54bf302957f0afcdbd9d18fb998c348fc8c2
|
| 3 |
+
size 14503
|
checkpoint-20000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0138aee610d696a759720c35e948ed2ee290cc1fb6d6c5b579e7b783997f3e04
|
| 3 |
+
size 623
|
checkpoint-20000/trainer_state.json
ADDED
|
@@ -0,0 +1,633 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.7751660943031311,
|
| 3 |
+
"best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l58-l/checkpoint-20000",
|
| 4 |
+
"epoch": 3.6845983787767134,
|
| 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": 0.09209799226376865,
|
| 13 |
+
"grad_norm": 0.0014697719598188996,
|
| 14 |
+
"learning_rate": 4.884877509670289e-08,
|
| 15 |
+
"loss": 0.4588,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.09209799226376865,
|
| 20 |
+
"eval_loss": 1.4194883108139038,
|
| 21 |
+
"eval_runtime": 78.7703,
|
| 22 |
+
"eval_samples_per_second": 15.31,
|
| 23 |
+
"eval_steps_per_second": 1.917,
|
| 24 |
+
"step": 500
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.1841959845275373,
|
| 28 |
+
"grad_norm": 301.76043701171875,
|
| 29 |
+
"learning_rate": 4.769755019340578e-08,
|
| 30 |
+
"loss": 0.4255,
|
| 31 |
+
"step": 1000
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.1841959845275373,
|
| 35 |
+
"eval_loss": 1.3416553735733032,
|
| 36 |
+
"eval_runtime": 77.0051,
|
| 37 |
+
"eval_samples_per_second": 15.661,
|
| 38 |
+
"eval_steps_per_second": 1.961,
|
| 39 |
+
"step": 1000
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.27629397679130596,
|
| 43 |
+
"grad_norm": 35.28853225708008,
|
| 44 |
+
"learning_rate": 4.654632529010867e-08,
|
| 45 |
+
"loss": 0.3724,
|
| 46 |
+
"step": 1500
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.27629397679130596,
|
| 50 |
+
"eval_loss": 1.2872973680496216,
|
| 51 |
+
"eval_runtime": 77.5439,
|
| 52 |
+
"eval_samples_per_second": 15.552,
|
| 53 |
+
"eval_steps_per_second": 1.947,
|
| 54 |
+
"step": 1500
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.3683919690550746,
|
| 58 |
+
"grad_norm": 338.9349060058594,
|
| 59 |
+
"learning_rate": 4.539510038681156e-08,
|
| 60 |
+
"loss": 0.3251,
|
| 61 |
+
"step": 2000
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.3683919690550746,
|
| 65 |
+
"eval_loss": 1.234910249710083,
|
| 66 |
+
"eval_runtime": 78.7715,
|
| 67 |
+
"eval_samples_per_second": 15.31,
|
| 68 |
+
"eval_steps_per_second": 1.917,
|
| 69 |
+
"step": 2000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.46048996131884323,
|
| 73 |
+
"grad_norm": 325.85101318359375,
|
| 74 |
+
"learning_rate": 4.4243875483514457e-08,
|
| 75 |
+
"loss": 0.3308,
|
| 76 |
+
"step": 2500
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.46048996131884323,
|
| 80 |
+
"eval_loss": 1.1944907903671265,
|
| 81 |
+
"eval_runtime": 79.0397,
|
| 82 |
+
"eval_samples_per_second": 15.258,
|
| 83 |
+
"eval_steps_per_second": 1.91,
|
| 84 |
+
"step": 2500
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.5525879535826119,
|
| 88 |
+
"grad_norm": 100.85855102539062,
|
| 89 |
+
"learning_rate": 4.309265058021735e-08,
|
| 90 |
+
"loss": 0.3017,
|
| 91 |
+
"step": 3000
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.5525879535826119,
|
| 95 |
+
"eval_loss": 1.1593009233474731,
|
| 96 |
+
"eval_runtime": 77.6354,
|
| 97 |
+
"eval_samples_per_second": 15.534,
|
| 98 |
+
"eval_steps_per_second": 1.945,
|
| 99 |
+
"step": 3000
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.6446859458463805,
|
| 103 |
+
"grad_norm": 8.865598678588867,
|
| 104 |
+
"learning_rate": 4.194142567692024e-08,
|
| 105 |
+
"loss": 0.2962,
|
| 106 |
+
"step": 3500
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.6446859458463805,
|
| 110 |
+
"eval_loss": 1.1259427070617676,
|
| 111 |
+
"eval_runtime": 78.2715,
|
| 112 |
+
"eval_samples_per_second": 15.408,
|
| 113 |
+
"eval_steps_per_second": 1.929,
|
| 114 |
+
"step": 3500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.7367839381101492,
|
| 118 |
+
"grad_norm": 4.393447399139404,
|
| 119 |
+
"learning_rate": 4.079020077362314e-08,
|
| 120 |
+
"loss": 0.2919,
|
| 121 |
+
"step": 4000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.7367839381101492,
|
| 125 |
+
"eval_loss": 1.0954023599624634,
|
| 126 |
+
"eval_runtime": 78.5366,
|
| 127 |
+
"eval_samples_per_second": 15.356,
|
| 128 |
+
"eval_steps_per_second": 1.923,
|
| 129 |
+
"step": 4000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.8288819303739179,
|
| 133 |
+
"grad_norm": 1.6036031246185303,
|
| 134 |
+
"learning_rate": 3.9638975870326026e-08,
|
| 135 |
+
"loss": 0.307,
|
| 136 |
+
"step": 4500
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.8288819303739179,
|
| 140 |
+
"eval_loss": 1.0728861093521118,
|
| 141 |
+
"eval_runtime": 78.4032,
|
| 142 |
+
"eval_samples_per_second": 15.382,
|
| 143 |
+
"eval_steps_per_second": 1.926,
|
| 144 |
+
"step": 4500
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.9209799226376865,
|
| 148 |
+
"grad_norm": 11.990758895874023,
|
| 149 |
+
"learning_rate": 3.8487750967028915e-08,
|
| 150 |
+
"loss": 0.2764,
|
| 151 |
+
"step": 5000
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.9209799226376865,
|
| 155 |
+
"eval_loss": 1.0523593425750732,
|
| 156 |
+
"eval_runtime": 78.3726,
|
| 157 |
+
"eval_samples_per_second": 15.388,
|
| 158 |
+
"eval_steps_per_second": 1.927,
|
| 159 |
+
"step": 5000
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 1.013077914901455,
|
| 163 |
+
"grad_norm": 43.865562438964844,
|
| 164 |
+
"learning_rate": 3.733652606373181e-08,
|
| 165 |
+
"loss": 0.2456,
|
| 166 |
+
"step": 5500
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 1.013077914901455,
|
| 170 |
+
"eval_loss": 1.0375442504882812,
|
| 171 |
+
"eval_runtime": 78.2473,
|
| 172 |
+
"eval_samples_per_second": 15.413,
|
| 173 |
+
"eval_steps_per_second": 1.93,
|
| 174 |
+
"step": 5500
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 1.1051759071652238,
|
| 178 |
+
"grad_norm": 105.91036987304688,
|
| 179 |
+
"learning_rate": 3.61853011604347e-08,
|
| 180 |
+
"loss": 0.2642,
|
| 181 |
+
"step": 6000
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 1.1051759071652238,
|
| 185 |
+
"eval_loss": 1.0233356952667236,
|
| 186 |
+
"eval_runtime": 78.3135,
|
| 187 |
+
"eval_samples_per_second": 15.4,
|
| 188 |
+
"eval_steps_per_second": 1.928,
|
| 189 |
+
"step": 6000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.1972738994289924,
|
| 193 |
+
"grad_norm": 221.4896240234375,
|
| 194 |
+
"learning_rate": 3.503407625713759e-08,
|
| 195 |
+
"loss": 0.2066,
|
| 196 |
+
"step": 6500
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.1972738994289924,
|
| 200 |
+
"eval_loss": 1.0104238986968994,
|
| 201 |
+
"eval_runtime": 78.6485,
|
| 202 |
+
"eval_samples_per_second": 15.334,
|
| 203 |
+
"eval_steps_per_second": 1.92,
|
| 204 |
+
"step": 6500
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 1.289371891692761,
|
| 208 |
+
"grad_norm": 270.53253173828125,
|
| 209 |
+
"learning_rate": 3.3882851353840485e-08,
|
| 210 |
+
"loss": 0.2376,
|
| 211 |
+
"step": 7000
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 1.289371891692761,
|
| 215 |
+
"eval_loss": 0.998353898525238,
|
| 216 |
+
"eval_runtime": 78.5502,
|
| 217 |
+
"eval_samples_per_second": 15.353,
|
| 218 |
+
"eval_steps_per_second": 1.922,
|
| 219 |
+
"step": 7000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.3814698839565298,
|
| 223 |
+
"grad_norm": 402.9339904785156,
|
| 224 |
+
"learning_rate": 3.2731626450543374e-08,
|
| 225 |
+
"loss": 0.1931,
|
| 226 |
+
"step": 7500
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.3814698839565298,
|
| 230 |
+
"eval_loss": 0.9887453317642212,
|
| 231 |
+
"eval_runtime": 78.4833,
|
| 232 |
+
"eval_samples_per_second": 15.366,
|
| 233 |
+
"eval_steps_per_second": 1.924,
|
| 234 |
+
"step": 7500
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.4735678762202984,
|
| 238 |
+
"grad_norm": 0.006355441175401211,
|
| 239 |
+
"learning_rate": 3.158040154724626e-08,
|
| 240 |
+
"loss": 0.2163,
|
| 241 |
+
"step": 8000
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.4735678762202984,
|
| 245 |
+
"eval_loss": 0.9767189621925354,
|
| 246 |
+
"eval_runtime": 78.5499,
|
| 247 |
+
"eval_samples_per_second": 15.353,
|
| 248 |
+
"eval_steps_per_second": 1.922,
|
| 249 |
+
"step": 8000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.565665868484067,
|
| 253 |
+
"grad_norm": 0.12982100248336792,
|
| 254 |
+
"learning_rate": 3.042917664394916e-08,
|
| 255 |
+
"loss": 0.1903,
|
| 256 |
+
"step": 8500
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.565665868484067,
|
| 260 |
+
"eval_loss": 0.9664921760559082,
|
| 261 |
+
"eval_runtime": 78.6832,
|
| 262 |
+
"eval_samples_per_second": 15.327,
|
| 263 |
+
"eval_steps_per_second": 1.919,
|
| 264 |
+
"step": 8500
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 1.6577638607478358,
|
| 268 |
+
"grad_norm": 173.31788635253906,
|
| 269 |
+
"learning_rate": 2.927795174065205e-08,
|
| 270 |
+
"loss": 0.2069,
|
| 271 |
+
"step": 9000
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 1.6577638607478358,
|
| 275 |
+
"eval_loss": 0.9571623206138611,
|
| 276 |
+
"eval_runtime": 78.544,
|
| 277 |
+
"eval_samples_per_second": 15.354,
|
| 278 |
+
"eval_steps_per_second": 1.922,
|
| 279 |
+
"step": 9000
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.7498618530116044,
|
| 283 |
+
"grad_norm": 358.7403259277344,
|
| 284 |
+
"learning_rate": 2.8126726837354947e-08,
|
| 285 |
+
"loss": 0.2093,
|
| 286 |
+
"step": 9500
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 1.7498618530116044,
|
| 290 |
+
"eval_loss": 0.9496769905090332,
|
| 291 |
+
"eval_runtime": 78.3124,
|
| 292 |
+
"eval_samples_per_second": 15.4,
|
| 293 |
+
"eval_steps_per_second": 1.928,
|
| 294 |
+
"step": 9500
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.841959845275373,
|
| 298 |
+
"grad_norm": 435.9449768066406,
|
| 299 |
+
"learning_rate": 2.6975501934057836e-08,
|
| 300 |
+
"loss": 0.2523,
|
| 301 |
+
"step": 10000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.841959845275373,
|
| 305 |
+
"eval_loss": 0.9420022368431091,
|
| 306 |
+
"eval_runtime": 78.3702,
|
| 307 |
+
"eval_samples_per_second": 15.388,
|
| 308 |
+
"eval_steps_per_second": 1.927,
|
| 309 |
+
"step": 10000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 1.9340578375391417,
|
| 313 |
+
"grad_norm": 23.49118423461914,
|
| 314 |
+
"learning_rate": 2.582427703076073e-08,
|
| 315 |
+
"loss": 0.2127,
|
| 316 |
+
"step": 10500
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 1.9340578375391417,
|
| 320 |
+
"eval_loss": 0.9328628182411194,
|
| 321 |
+
"eval_runtime": 76.6904,
|
| 322 |
+
"eval_samples_per_second": 15.726,
|
| 323 |
+
"eval_steps_per_second": 1.969,
|
| 324 |
+
"step": 10500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 2.02615582980291,
|
| 328 |
+
"grad_norm": 0.4029064476490021,
|
| 329 |
+
"learning_rate": 2.467305212746362e-08,
|
| 330 |
+
"loss": 0.1968,
|
| 331 |
+
"step": 11000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 2.02615582980291,
|
| 335 |
+
"eval_loss": 0.926975429058075,
|
| 336 |
+
"eval_runtime": 76.7161,
|
| 337 |
+
"eval_samples_per_second": 15.72,
|
| 338 |
+
"eval_steps_per_second": 1.968,
|
| 339 |
+
"step": 11000
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.118253822066679,
|
| 343 |
+
"grad_norm": 274.0943603515625,
|
| 344 |
+
"learning_rate": 2.3521827224166513e-08,
|
| 345 |
+
"loss": 0.1879,
|
| 346 |
+
"step": 11500
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 2.118253822066679,
|
| 350 |
+
"eval_loss": 0.9231382012367249,
|
| 351 |
+
"eval_runtime": 78.6224,
|
| 352 |
+
"eval_samples_per_second": 15.339,
|
| 353 |
+
"eval_steps_per_second": 1.921,
|
| 354 |
+
"step": 11500
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.2103518143304477,
|
| 358 |
+
"grad_norm": 0.000765918695833534,
|
| 359 |
+
"learning_rate": 2.2370602320869402e-08,
|
| 360 |
+
"loss": 0.1981,
|
| 361 |
+
"step": 12000
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.2103518143304477,
|
| 365 |
+
"eval_loss": 0.9183884263038635,
|
| 366 |
+
"eval_runtime": 76.7707,
|
| 367 |
+
"eval_samples_per_second": 15.709,
|
| 368 |
+
"eval_steps_per_second": 1.967,
|
| 369 |
+
"step": 12000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 2.302449806594216,
|
| 373 |
+
"grad_norm": 0.6250036358833313,
|
| 374 |
+
"learning_rate": 2.1219377417572295e-08,
|
| 375 |
+
"loss": 0.1964,
|
| 376 |
+
"step": 12500
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 2.302449806594216,
|
| 380 |
+
"eval_loss": 0.9134928584098816,
|
| 381 |
+
"eval_runtime": 78.5128,
|
| 382 |
+
"eval_samples_per_second": 15.361,
|
| 383 |
+
"eval_steps_per_second": 1.923,
|
| 384 |
+
"step": 12500
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 2.394547798857985,
|
| 388 |
+
"grad_norm": 158.6990203857422,
|
| 389 |
+
"learning_rate": 2.0068152514275187e-08,
|
| 390 |
+
"loss": 0.1697,
|
| 391 |
+
"step": 13000
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 2.394547798857985,
|
| 395 |
+
"eval_loss": 0.910048246383667,
|
| 396 |
+
"eval_runtime": 76.7476,
|
| 397 |
+
"eval_samples_per_second": 15.714,
|
| 398 |
+
"eval_steps_per_second": 1.967,
|
| 399 |
+
"step": 13000
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 2.4866457911217537,
|
| 403 |
+
"grad_norm": 1.288072943687439,
|
| 404 |
+
"learning_rate": 1.891692761097808e-08,
|
| 405 |
+
"loss": 0.2015,
|
| 406 |
+
"step": 13500
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 2.4866457911217537,
|
| 410 |
+
"eval_loss": 0.9051916599273682,
|
| 411 |
+
"eval_runtime": 78.3909,
|
| 412 |
+
"eval_samples_per_second": 15.384,
|
| 413 |
+
"eval_steps_per_second": 1.926,
|
| 414 |
+
"step": 13500
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 2.578743783385522,
|
| 418 |
+
"grad_norm": 5.835799311171286e-05,
|
| 419 |
+
"learning_rate": 1.7765702707680972e-08,
|
| 420 |
+
"loss": 0.1827,
|
| 421 |
+
"step": 14000
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 2.578743783385522,
|
| 425 |
+
"eval_loss": 0.9026296734809875,
|
| 426 |
+
"eval_runtime": 76.6591,
|
| 427 |
+
"eval_samples_per_second": 15.732,
|
| 428 |
+
"eval_steps_per_second": 1.97,
|
| 429 |
+
"step": 14000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 2.670841775649291,
|
| 433 |
+
"grad_norm": 30.217784881591797,
|
| 434 |
+
"learning_rate": 1.6614477804383865e-08,
|
| 435 |
+
"loss": 0.1435,
|
| 436 |
+
"step": 14500
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 2.670841775649291,
|
| 440 |
+
"eval_loss": 0.899847686290741,
|
| 441 |
+
"eval_runtime": 76.742,
|
| 442 |
+
"eval_samples_per_second": 15.715,
|
| 443 |
+
"eval_steps_per_second": 1.968,
|
| 444 |
+
"step": 14500
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 2.7629397679130596,
|
| 448 |
+
"grad_norm": 202.30496215820312,
|
| 449 |
+
"learning_rate": 1.5463252901086757e-08,
|
| 450 |
+
"loss": 0.1541,
|
| 451 |
+
"step": 15000
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 2.7629397679130596,
|
| 455 |
+
"eval_loss": 0.896297812461853,
|
| 456 |
+
"eval_runtime": 78.5539,
|
| 457 |
+
"eval_samples_per_second": 15.353,
|
| 458 |
+
"eval_steps_per_second": 1.922,
|
| 459 |
+
"step": 15000
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 2.8550377601768284,
|
| 463 |
+
"grad_norm": 241.57069396972656,
|
| 464 |
+
"learning_rate": 1.4312027997789646e-08,
|
| 465 |
+
"loss": 0.1716,
|
| 466 |
+
"step": 15500
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 2.8550377601768284,
|
| 470 |
+
"eval_loss": 0.8935310244560242,
|
| 471 |
+
"eval_runtime": 76.9195,
|
| 472 |
+
"eval_samples_per_second": 15.679,
|
| 473 |
+
"eval_steps_per_second": 1.963,
|
| 474 |
+
"step": 15500
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 2.947135752440597,
|
| 478 |
+
"grad_norm": 0.25739413499832153,
|
| 479 |
+
"learning_rate": 1.316080309449254e-08,
|
| 480 |
+
"loss": 0.2056,
|
| 481 |
+
"step": 16000
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 2.947135752440597,
|
| 485 |
+
"eval_loss": 0.890542209148407,
|
| 486 |
+
"eval_runtime": 76.7268,
|
| 487 |
+
"eval_samples_per_second": 15.718,
|
| 488 |
+
"eval_steps_per_second": 1.968,
|
| 489 |
+
"step": 16000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 3.0392337447043656,
|
| 493 |
+
"grad_norm": 86.5535659790039,
|
| 494 |
+
"learning_rate": 1.2009578191195431e-08,
|
| 495 |
+
"loss": 0.1843,
|
| 496 |
+
"step": 16500
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 3.0392337447043656,
|
| 500 |
+
"eval_loss": 0.8875145316123962,
|
| 501 |
+
"eval_runtime": 76.4556,
|
| 502 |
+
"eval_samples_per_second": 15.774,
|
| 503 |
+
"eval_steps_per_second": 1.975,
|
| 504 |
+
"step": 16500
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 3.131331736968134,
|
| 508 |
+
"grad_norm": 0.05445564165711403,
|
| 509 |
+
"learning_rate": 1.0858353287898323e-08,
|
| 510 |
+
"loss": 0.1611,
|
| 511 |
+
"step": 17000
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 3.131331736968134,
|
| 515 |
+
"eval_loss": 0.8858152627944946,
|
| 516 |
+
"eval_runtime": 77.4814,
|
| 517 |
+
"eval_samples_per_second": 15.565,
|
| 518 |
+
"eval_steps_per_second": 1.949,
|
| 519 |
+
"step": 17000
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 3.2240235814296243,
|
| 523 |
+
"grad_norm": 44.673702239990234,
|
| 524 |
+
"learning_rate": 9.699705232129698e-09,
|
| 525 |
+
"loss": 0.1568,
|
| 526 |
+
"step": 17500
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 3.2240235814296243,
|
| 530 |
+
"eval_loss": 0.7821062207221985,
|
| 531 |
+
"eval_runtime": 75.6502,
|
| 532 |
+
"eval_samples_per_second": 15.955,
|
| 533 |
+
"eval_steps_per_second": 1.996,
|
| 534 |
+
"step": 17500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 3.316138540899042,
|
| 538 |
+
"grad_norm": 0.5617901086807251,
|
| 539 |
+
"learning_rate": 8.548268238761974e-09,
|
| 540 |
+
"loss": 0.1395,
|
| 541 |
+
"step": 18000
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 3.316138540899042,
|
| 545 |
+
"eval_loss": 0.7793659567832947,
|
| 546 |
+
"eval_runtime": 76.0283,
|
| 547 |
+
"eval_samples_per_second": 15.876,
|
| 548 |
+
"eval_steps_per_second": 1.986,
|
| 549 |
+
"step": 18000
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 3.4082535003684598,
|
| 553 |
+
"grad_norm": 6.4552321434021,
|
| 554 |
+
"learning_rate": 7.396831245394252e-09,
|
| 555 |
+
"loss": 0.1804,
|
| 556 |
+
"step": 18500
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 3.4082535003684598,
|
| 560 |
+
"eval_loss": 0.7778191566467285,
|
| 561 |
+
"eval_runtime": 76.0536,
|
| 562 |
+
"eval_samples_per_second": 15.87,
|
| 563 |
+
"eval_steps_per_second": 1.985,
|
| 564 |
+
"step": 18500
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 3.500368459837878,
|
| 568 |
+
"grad_norm": 0.31470850110054016,
|
| 569 |
+
"learning_rate": 6.2453942520265294e-09,
|
| 570 |
+
"loss": 0.1728,
|
| 571 |
+
"step": 19000
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 3.500368459837878,
|
| 575 |
+
"eval_loss": 0.7768829464912415,
|
| 576 |
+
"eval_runtime": 76.116,
|
| 577 |
+
"eval_samples_per_second": 15.857,
|
| 578 |
+
"eval_steps_per_second": 1.984,
|
| 579 |
+
"step": 19000
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 3.5924834193072956,
|
| 583 |
+
"grad_norm": 0.029128307476639748,
|
| 584 |
+
"learning_rate": 5.093957258658806e-09,
|
| 585 |
+
"loss": 0.179,
|
| 586 |
+
"step": 19500
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 3.5924834193072956,
|
| 590 |
+
"eval_loss": 0.7758385539054871,
|
| 591 |
+
"eval_runtime": 76.1889,
|
| 592 |
+
"eval_samples_per_second": 15.842,
|
| 593 |
+
"eval_steps_per_second": 1.982,
|
| 594 |
+
"step": 19500
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 3.6845983787767134,
|
| 598 |
+
"grad_norm": 0.0002703067730180919,
|
| 599 |
+
"learning_rate": 3.942520265291083e-09,
|
| 600 |
+
"loss": 0.179,
|
| 601 |
+
"step": 20000
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 3.6845983787767134,
|
| 605 |
+
"eval_loss": 0.7751660943031311,
|
| 606 |
+
"eval_runtime": 76.1566,
|
| 607 |
+
"eval_samples_per_second": 15.849,
|
| 608 |
+
"eval_steps_per_second": 1.983,
|
| 609 |
+
"step": 20000
|
| 610 |
+
}
|
| 611 |
+
],
|
| 612 |
+
"logging_steps": 500,
|
| 613 |
+
"max_steps": 21712,
|
| 614 |
+
"num_input_tokens_seen": 0,
|
| 615 |
+
"num_train_epochs": 4,
|
| 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": 7195346904375630.0,
|
| 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:af19727b1180ae73a7b4839d8539085faed51bc7e6cd0858188fac5c6ef07636
|
| 3 |
+
size 4847
|