Training in progress, step 15500, checkpoint
Browse files
checkpoint-15500/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-15500/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:88627be1fb9ff402575072fe529b6d40100f02c480764dedc20c678a5feed043
|
| 3 |
+
size 1711848436
|
checkpoint-15500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c979c0cab23610e6245c51e47267a7068c15133ff05b0b264657f33d2d159816
|
| 3 |
+
size 3424043887
|
checkpoint-15500/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c5557b4d8fcc20e6c5c0ce89835be50d8684913e3edc6ba4c0e67d6c8e9cce93
|
| 3 |
+
size 14503
|
checkpoint-15500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6a7f92dbc187e70e7ebab38af01497c58b8ab0cb80967344f17451ae31dbdd89
|
| 3 |
+
size 623
|
checkpoint-15500/trainer_state.json
ADDED
|
@@ -0,0 +1,498 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9261995553970337,
|
| 3 |
+
"best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e3l58-l/checkpoint-15500",
|
| 4 |
+
"epoch": 2.855563743551953,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 15500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.09211495946941783,
|
| 13 |
+
"grad_norm": 19.808820724487305,
|
| 14 |
+
"learning_rate": 4.84647506755097e-08,
|
| 15 |
+
"loss": 0.3758,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.09211495946941783,
|
| 20 |
+
"eval_loss": 1.418845295906067,
|
| 21 |
+
"eval_runtime": 75.2377,
|
| 22 |
+
"eval_samples_per_second": 16.042,
|
| 23 |
+
"eval_steps_per_second": 2.007,
|
| 24 |
+
"step": 500
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.18422991893883567,
|
| 28 |
+
"grad_norm": 76.37991333007812,
|
| 29 |
+
"learning_rate": 4.69295013510194e-08,
|
| 30 |
+
"loss": 0.4108,
|
| 31 |
+
"step": 1000
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.18422991893883567,
|
| 35 |
+
"eval_loss": 1.3513683080673218,
|
| 36 |
+
"eval_runtime": 74.9686,
|
| 37 |
+
"eval_samples_per_second": 16.1,
|
| 38 |
+
"eval_steps_per_second": 2.014,
|
| 39 |
+
"step": 1000
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.2763448784082535,
|
| 43 |
+
"grad_norm": 0.5795987844467163,
|
| 44 |
+
"learning_rate": 4.539425202652911e-08,
|
| 45 |
+
"loss": 0.4335,
|
| 46 |
+
"step": 1500
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.2763448784082535,
|
| 50 |
+
"eval_loss": 1.2895005941390991,
|
| 51 |
+
"eval_runtime": 76.713,
|
| 52 |
+
"eval_samples_per_second": 15.734,
|
| 53 |
+
"eval_steps_per_second": 1.968,
|
| 54 |
+
"step": 1500
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.36845983787767134,
|
| 58 |
+
"grad_norm": 0.1463615745306015,
|
| 59 |
+
"learning_rate": 4.385900270203881e-08,
|
| 60 |
+
"loss": 0.3436,
|
| 61 |
+
"step": 2000
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.36845983787767134,
|
| 65 |
+
"eval_loss": 1.2414860725402832,
|
| 66 |
+
"eval_runtime": 76.9749,
|
| 67 |
+
"eval_samples_per_second": 15.68,
|
| 68 |
+
"eval_steps_per_second": 1.962,
|
| 69 |
+
"step": 2000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.46057479734708917,
|
| 73 |
+
"grad_norm": 323.8576354980469,
|
| 74 |
+
"learning_rate": 4.232375337754851e-08,
|
| 75 |
+
"loss": 0.366,
|
| 76 |
+
"step": 2500
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.46057479734708917,
|
| 80 |
+
"eval_loss": 1.1949334144592285,
|
| 81 |
+
"eval_runtime": 77.0381,
|
| 82 |
+
"eval_samples_per_second": 15.668,
|
| 83 |
+
"eval_steps_per_second": 1.96,
|
| 84 |
+
"step": 2500
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.552689756816507,
|
| 88 |
+
"grad_norm": 339.28533935546875,
|
| 89 |
+
"learning_rate": 4.078850405305821e-08,
|
| 90 |
+
"loss": 0.3191,
|
| 91 |
+
"step": 3000
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.552689756816507,
|
| 95 |
+
"eval_loss": 1.1560280323028564,
|
| 96 |
+
"eval_runtime": 77.0026,
|
| 97 |
+
"eval_samples_per_second": 15.675,
|
| 98 |
+
"eval_steps_per_second": 1.961,
|
| 99 |
+
"step": 3000
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.6448047162859248,
|
| 103 |
+
"grad_norm": 60.27859878540039,
|
| 104 |
+
"learning_rate": 3.925325472856792e-08,
|
| 105 |
+
"loss": 0.2779,
|
| 106 |
+
"step": 3500
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.6448047162859248,
|
| 110 |
+
"eval_loss": 1.1233855485916138,
|
| 111 |
+
"eval_runtime": 76.9006,
|
| 112 |
+
"eval_samples_per_second": 15.696,
|
| 113 |
+
"eval_steps_per_second": 1.964,
|
| 114 |
+
"step": 3500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.7369196757553427,
|
| 118 |
+
"grad_norm": 18.268455505371094,
|
| 119 |
+
"learning_rate": 3.7718005404077616e-08,
|
| 120 |
+
"loss": 0.2942,
|
| 121 |
+
"step": 4000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.7369196757553427,
|
| 125 |
+
"eval_loss": 1.0993207693099976,
|
| 126 |
+
"eval_runtime": 77.2834,
|
| 127 |
+
"eval_samples_per_second": 15.618,
|
| 128 |
+
"eval_steps_per_second": 1.954,
|
| 129 |
+
"step": 4000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.8290346352247605,
|
| 133 |
+
"grad_norm": 505.22723388671875,
|
| 134 |
+
"learning_rate": 3.6182756079587324e-08,
|
| 135 |
+
"loss": 0.2615,
|
| 136 |
+
"step": 4500
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.8290346352247605,
|
| 140 |
+
"eval_loss": 1.080470085144043,
|
| 141 |
+
"eval_runtime": 77.1214,
|
| 142 |
+
"eval_samples_per_second": 15.651,
|
| 143 |
+
"eval_steps_per_second": 1.958,
|
| 144 |
+
"step": 4500
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.9211495946941783,
|
| 148 |
+
"grad_norm": 1.2971173524856567,
|
| 149 |
+
"learning_rate": 3.4647506755097025e-08,
|
| 150 |
+
"loss": 0.2715,
|
| 151 |
+
"step": 5000
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.9211495946941783,
|
| 155 |
+
"eval_loss": 1.0614728927612305,
|
| 156 |
+
"eval_runtime": 77.2982,
|
| 157 |
+
"eval_samples_per_second": 15.615,
|
| 158 |
+
"eval_steps_per_second": 1.953,
|
| 159 |
+
"step": 5000
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 1.013264554163596,
|
| 163 |
+
"grad_norm": 0.17611829936504364,
|
| 164 |
+
"learning_rate": 3.3112257430606727e-08,
|
| 165 |
+
"loss": 0.2509,
|
| 166 |
+
"step": 5500
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 1.013264554163596,
|
| 170 |
+
"eval_loss": 1.0461839437484741,
|
| 171 |
+
"eval_runtime": 75.3494,
|
| 172 |
+
"eval_samples_per_second": 16.019,
|
| 173 |
+
"eval_steps_per_second": 2.004,
|
| 174 |
+
"step": 5500
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 1.105379513633014,
|
| 178 |
+
"grad_norm": 228.2562713623047,
|
| 179 |
+
"learning_rate": 3.1577008106116435e-08,
|
| 180 |
+
"loss": 0.2653,
|
| 181 |
+
"step": 6000
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 1.105379513633014,
|
| 185 |
+
"eval_loss": 1.0305675268173218,
|
| 186 |
+
"eval_runtime": 77.0542,
|
| 187 |
+
"eval_samples_per_second": 15.664,
|
| 188 |
+
"eval_steps_per_second": 1.96,
|
| 189 |
+
"step": 6000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.1974944731024317,
|
| 193 |
+
"grad_norm": 432.4784240722656,
|
| 194 |
+
"learning_rate": 3.0041758781626136e-08,
|
| 195 |
+
"loss": 0.2199,
|
| 196 |
+
"step": 6500
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.1974944731024317,
|
| 200 |
+
"eval_loss": 1.0182703733444214,
|
| 201 |
+
"eval_runtime": 76.296,
|
| 202 |
+
"eval_samples_per_second": 15.82,
|
| 203 |
+
"eval_steps_per_second": 1.979,
|
| 204 |
+
"step": 6500
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 1.2896094325718497,
|
| 208 |
+
"grad_norm": 0.6385967135429382,
|
| 209 |
+
"learning_rate": 2.8506509457135834e-08,
|
| 210 |
+
"loss": 0.1844,
|
| 211 |
+
"step": 7000
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 1.2896094325718497,
|
| 215 |
+
"eval_loss": 1.0078601837158203,
|
| 216 |
+
"eval_runtime": 76.5441,
|
| 217 |
+
"eval_samples_per_second": 15.769,
|
| 218 |
+
"eval_steps_per_second": 1.973,
|
| 219 |
+
"step": 7000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.3817243920412676,
|
| 223 |
+
"grad_norm": 57.47743606567383,
|
| 224 |
+
"learning_rate": 2.697126013264554e-08,
|
| 225 |
+
"loss": 0.2143,
|
| 226 |
+
"step": 7500
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.3817243920412676,
|
| 230 |
+
"eval_loss": 0.9969784617424011,
|
| 231 |
+
"eval_runtime": 75.6186,
|
| 232 |
+
"eval_samples_per_second": 15.962,
|
| 233 |
+
"eval_steps_per_second": 1.997,
|
| 234 |
+
"step": 7500
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.4738393515106853,
|
| 238 |
+
"grad_norm": 0.11704438179731369,
|
| 239 |
+
"learning_rate": 2.5436010808155244e-08,
|
| 240 |
+
"loss": 0.2011,
|
| 241 |
+
"step": 8000
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.4738393515106853,
|
| 245 |
+
"eval_loss": 0.9886976480484009,
|
| 246 |
+
"eval_runtime": 75.238,
|
| 247 |
+
"eval_samples_per_second": 16.042,
|
| 248 |
+
"eval_steps_per_second": 2.007,
|
| 249 |
+
"step": 8000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.565954310980103,
|
| 253 |
+
"grad_norm": 39.71356964111328,
|
| 254 |
+
"learning_rate": 2.3900761483664948e-08,
|
| 255 |
+
"loss": 0.1995,
|
| 256 |
+
"step": 8500
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.565954310980103,
|
| 260 |
+
"eval_loss": 0.980699360370636,
|
| 261 |
+
"eval_runtime": 75.7533,
|
| 262 |
+
"eval_samples_per_second": 15.933,
|
| 263 |
+
"eval_steps_per_second": 1.993,
|
| 264 |
+
"step": 8500
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 1.658069270449521,
|
| 268 |
+
"grad_norm": 0.09355564415454865,
|
| 269 |
+
"learning_rate": 2.236551215917465e-08,
|
| 270 |
+
"loss": 0.2493,
|
| 271 |
+
"step": 9000
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 1.658069270449521,
|
| 275 |
+
"eval_loss": 0.9733120203018188,
|
| 276 |
+
"eval_runtime": 75.1219,
|
| 277 |
+
"eval_samples_per_second": 16.067,
|
| 278 |
+
"eval_steps_per_second": 2.01,
|
| 279 |
+
"step": 9000
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.750184229918939,
|
| 283 |
+
"grad_norm": 0.12566468119621277,
|
| 284 |
+
"learning_rate": 2.083026283468435e-08,
|
| 285 |
+
"loss": 0.2192,
|
| 286 |
+
"step": 9500
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 1.750184229918939,
|
| 290 |
+
"eval_loss": 0.966866672039032,
|
| 291 |
+
"eval_runtime": 76.5812,
|
| 292 |
+
"eval_samples_per_second": 15.761,
|
| 293 |
+
"eval_steps_per_second": 1.972,
|
| 294 |
+
"step": 9500
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.8422991893883567,
|
| 298 |
+
"grad_norm": 1.9423623085021973,
|
| 299 |
+
"learning_rate": 1.9295013510194056e-08,
|
| 300 |
+
"loss": 0.1882,
|
| 301 |
+
"step": 10000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.8422991893883567,
|
| 305 |
+
"eval_loss": 0.9616385698318481,
|
| 306 |
+
"eval_runtime": 75.5454,
|
| 307 |
+
"eval_samples_per_second": 15.977,
|
| 308 |
+
"eval_steps_per_second": 1.999,
|
| 309 |
+
"step": 10000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 1.9344141488577744,
|
| 313 |
+
"grad_norm": 0.12048076093196869,
|
| 314 |
+
"learning_rate": 1.7759764185703757e-08,
|
| 315 |
+
"loss": 0.2193,
|
| 316 |
+
"step": 10500
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 1.9344141488577744,
|
| 320 |
+
"eval_loss": 0.9550592303276062,
|
| 321 |
+
"eval_runtime": 75.1134,
|
| 322 |
+
"eval_samples_per_second": 16.069,
|
| 323 |
+
"eval_steps_per_second": 2.01,
|
| 324 |
+
"step": 10500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 2.026529108327192,
|
| 328 |
+
"grad_norm": 0.19933176040649414,
|
| 329 |
+
"learning_rate": 1.622451486121346e-08,
|
| 330 |
+
"loss": 0.2148,
|
| 331 |
+
"step": 11000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 2.026529108327192,
|
| 335 |
+
"eval_loss": 0.9494845271110535,
|
| 336 |
+
"eval_runtime": 76.2803,
|
| 337 |
+
"eval_samples_per_second": 15.823,
|
| 338 |
+
"eval_steps_per_second": 1.98,
|
| 339 |
+
"step": 11000
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.1186440677966103,
|
| 343 |
+
"grad_norm": 13.599303245544434,
|
| 344 |
+
"learning_rate": 1.4689265536723163e-08,
|
| 345 |
+
"loss": 0.1975,
|
| 346 |
+
"step": 11500
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 2.1186440677966103,
|
| 350 |
+
"eval_loss": 0.9449298977851868,
|
| 351 |
+
"eval_runtime": 75.6422,
|
| 352 |
+
"eval_samples_per_second": 15.957,
|
| 353 |
+
"eval_steps_per_second": 1.996,
|
| 354 |
+
"step": 11500
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.210759027266028,
|
| 358 |
+
"grad_norm": 0.8174373507499695,
|
| 359 |
+
"learning_rate": 1.3154016212232866e-08,
|
| 360 |
+
"loss": 0.1791,
|
| 361 |
+
"step": 12000
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.210759027266028,
|
| 365 |
+
"eval_loss": 0.940927267074585,
|
| 366 |
+
"eval_runtime": 75.2379,
|
| 367 |
+
"eval_samples_per_second": 16.042,
|
| 368 |
+
"eval_steps_per_second": 2.007,
|
| 369 |
+
"step": 12000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 2.3028739867354457,
|
| 373 |
+
"grad_norm": 705.062744140625,
|
| 374 |
+
"learning_rate": 1.161876688774257e-08,
|
| 375 |
+
"loss": 0.2057,
|
| 376 |
+
"step": 12500
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 2.3028739867354457,
|
| 380 |
+
"eval_loss": 0.9382032155990601,
|
| 381 |
+
"eval_runtime": 76.1075,
|
| 382 |
+
"eval_samples_per_second": 15.859,
|
| 383 |
+
"eval_steps_per_second": 1.984,
|
| 384 |
+
"step": 12500
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 2.3949889462048635,
|
| 388 |
+
"grad_norm": 0.7270774841308594,
|
| 389 |
+
"learning_rate": 1.0083517563252273e-08,
|
| 390 |
+
"loss": 0.2037,
|
| 391 |
+
"step": 13000
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 2.3949889462048635,
|
| 395 |
+
"eval_loss": 0.9352426528930664,
|
| 396 |
+
"eval_runtime": 75.5821,
|
| 397 |
+
"eval_samples_per_second": 15.969,
|
| 398 |
+
"eval_steps_per_second": 1.998,
|
| 399 |
+
"step": 13000
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 2.4871039056742816,
|
| 403 |
+
"grad_norm": 2.034397602081299,
|
| 404 |
+
"learning_rate": 8.548268238761974e-09,
|
| 405 |
+
"loss": 0.2011,
|
| 406 |
+
"step": 13500
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 2.4871039056742816,
|
| 410 |
+
"eval_loss": 0.9317955374717712,
|
| 411 |
+
"eval_runtime": 75.3697,
|
| 412 |
+
"eval_samples_per_second": 16.014,
|
| 413 |
+
"eval_steps_per_second": 2.003,
|
| 414 |
+
"step": 13500
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 2.5792188651436994,
|
| 418 |
+
"grad_norm": 299.313232421875,
|
| 419 |
+
"learning_rate": 7.013018914271678e-09,
|
| 420 |
+
"loss": 0.1551,
|
| 421 |
+
"step": 14000
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 2.5792188651436994,
|
| 425 |
+
"eval_loss": 0.9295068979263306,
|
| 426 |
+
"eval_runtime": 76.109,
|
| 427 |
+
"eval_samples_per_second": 15.859,
|
| 428 |
+
"eval_steps_per_second": 1.984,
|
| 429 |
+
"step": 14000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 2.671333824613117,
|
| 433 |
+
"grad_norm": 0.04151364788413048,
|
| 434 |
+
"learning_rate": 5.47776958978138e-09,
|
| 435 |
+
"loss": 0.1565,
|
| 436 |
+
"step": 14500
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 2.671333824613117,
|
| 440 |
+
"eval_loss": 0.9278931617736816,
|
| 441 |
+
"eval_runtime": 75.8114,
|
| 442 |
+
"eval_samples_per_second": 15.921,
|
| 443 |
+
"eval_steps_per_second": 1.992,
|
| 444 |
+
"step": 14500
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 2.7634487840825352,
|
| 448 |
+
"grad_norm": 4.239564895629883,
|
| 449 |
+
"learning_rate": 3.942520265291083e-09,
|
| 450 |
+
"loss": 0.1963,
|
| 451 |
+
"step": 15000
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 2.7634487840825352,
|
| 455 |
+
"eval_loss": 0.9268263578414917,
|
| 456 |
+
"eval_runtime": 75.5159,
|
| 457 |
+
"eval_samples_per_second": 15.983,
|
| 458 |
+
"eval_steps_per_second": 2.0,
|
| 459 |
+
"step": 15000
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 2.855563743551953,
|
| 463 |
+
"grad_norm": 0.001815026975236833,
|
| 464 |
+
"learning_rate": 2.407270940800786e-09,
|
| 465 |
+
"loss": 0.1823,
|
| 466 |
+
"step": 15500
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 2.855563743551953,
|
| 470 |
+
"eval_loss": 0.9261995553970337,
|
| 471 |
+
"eval_runtime": 76.2027,
|
| 472 |
+
"eval_samples_per_second": 15.839,
|
| 473 |
+
"eval_steps_per_second": 1.982,
|
| 474 |
+
"step": 15500
|
| 475 |
+
}
|
| 476 |
+
],
|
| 477 |
+
"logging_steps": 500,
|
| 478 |
+
"max_steps": 16284,
|
| 479 |
+
"num_input_tokens_seen": 0,
|
| 480 |
+
"num_train_epochs": 3,
|
| 481 |
+
"save_steps": 500,
|
| 482 |
+
"stateful_callbacks": {
|
| 483 |
+
"TrainerControl": {
|
| 484 |
+
"args": {
|
| 485 |
+
"should_epoch_stop": false,
|
| 486 |
+
"should_evaluate": false,
|
| 487 |
+
"should_log": false,
|
| 488 |
+
"should_save": true,
|
| 489 |
+
"should_training_stop": false
|
| 490 |
+
},
|
| 491 |
+
"attributes": {}
|
| 492 |
+
}
|
| 493 |
+
},
|
| 494 |
+
"total_flos": 5576973961232520.0,
|
| 495 |
+
"train_batch_size": 2,
|
| 496 |
+
"trial_name": null,
|
| 497 |
+
"trial_params": null
|
| 498 |
+
}
|
checkpoint-15500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cbcece3d46fd39b0ba80eb3e495fdf46c9f891af5b4bddce9de77ad5a467b625
|
| 3 |
+
size 4847
|