Upload folder using huggingface_hub
Browse files- checkpoint-14616/config.json +42 -0
- checkpoint-14616/model.safetensors +3 -0
- checkpoint-14616/optimizer.pt +3 -0
- checkpoint-14616/preprocessor_config.json +36 -0
- checkpoint-14616/rng_state.pth +3 -0
- checkpoint-14616/scheduler.pt +3 -0
- checkpoint-14616/trainer_state.json +350 -0
- checkpoint-14616/training_args.bin +3 -0
- checkpoint-17748/config.json +42 -0
- checkpoint-17748/model.safetensors +3 -0
- checkpoint-17748/optimizer.pt +3 -0
- checkpoint-17748/preprocessor_config.json +36 -0
- checkpoint-17748/rng_state.pth +3 -0
- checkpoint-17748/scheduler.pt +3 -0
- checkpoint-17748/trainer_state.json +419 -0
- checkpoint-17748/training_args.bin +3 -0
checkpoint-14616/config.json
ADDED
|
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "google/vit-base-patch16-224",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ViTForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.0,
|
| 7 |
+
"encoder_stride": 16,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.0,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "disguist",
|
| 13 |
+
"1": "anger",
|
| 14 |
+
"2": "fear",
|
| 15 |
+
"3": "happy",
|
| 16 |
+
"4": "surprise",
|
| 17 |
+
"5": "neutral",
|
| 18 |
+
"6": "sad"
|
| 19 |
+
},
|
| 20 |
+
"image_size": 224,
|
| 21 |
+
"initializer_range": 0.02,
|
| 22 |
+
"intermediate_size": 3072,
|
| 23 |
+
"label2id": {
|
| 24 |
+
"anger": 1,
|
| 25 |
+
"disguist": 0,
|
| 26 |
+
"fear": 2,
|
| 27 |
+
"happy": 3,
|
| 28 |
+
"neutral": 5,
|
| 29 |
+
"sad": 6,
|
| 30 |
+
"surprise": 4
|
| 31 |
+
},
|
| 32 |
+
"layer_norm_eps": 1e-12,
|
| 33 |
+
"model_type": "vit",
|
| 34 |
+
"num_attention_heads": 12,
|
| 35 |
+
"num_channels": 3,
|
| 36 |
+
"num_hidden_layers": 12,
|
| 37 |
+
"patch_size": 16,
|
| 38 |
+
"problem_type": "single_label_classification",
|
| 39 |
+
"qkv_bias": true,
|
| 40 |
+
"torch_dtype": "float32",
|
| 41 |
+
"transformers_version": "4.39.3"
|
| 42 |
+
}
|
checkpoint-14616/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:73c9bc11b6a6c8ad4b674efad682676b150e033e243aa817584f336564d11cb4
|
| 3 |
+
size 343239356
|
checkpoint-14616/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8eeed73e1d0ebfd568fbd8582d74e127a4523da017d9988ff7c2a659d5a75bfd
|
| 3 |
+
size 686599610
|
checkpoint-14616/preprocessor_config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_valid_processor_keys": [
|
| 3 |
+
"images",
|
| 4 |
+
"do_resize",
|
| 5 |
+
"size",
|
| 6 |
+
"resample",
|
| 7 |
+
"do_rescale",
|
| 8 |
+
"rescale_factor",
|
| 9 |
+
"do_normalize",
|
| 10 |
+
"image_mean",
|
| 11 |
+
"image_std",
|
| 12 |
+
"return_tensors",
|
| 13 |
+
"data_format",
|
| 14 |
+
"input_data_format"
|
| 15 |
+
],
|
| 16 |
+
"do_normalize": true,
|
| 17 |
+
"do_rescale": true,
|
| 18 |
+
"do_resize": true,
|
| 19 |
+
"image_mean": [
|
| 20 |
+
0.5,
|
| 21 |
+
0.5,
|
| 22 |
+
0.5
|
| 23 |
+
],
|
| 24 |
+
"image_processor_type": "ViTImageProcessor",
|
| 25 |
+
"image_std": [
|
| 26 |
+
0.5,
|
| 27 |
+
0.5,
|
| 28 |
+
0.5
|
| 29 |
+
],
|
| 30 |
+
"resample": 2,
|
| 31 |
+
"rescale_factor": 0.00392156862745098,
|
| 32 |
+
"size": {
|
| 33 |
+
"height": 224,
|
| 34 |
+
"width": 224
|
| 35 |
+
}
|
| 36 |
+
}
|
checkpoint-14616/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b7d6513867adac690b95294e66a85a01cd3c4cc2c44aed15e66414ecec3c165b
|
| 3 |
+
size 14244
|
checkpoint-14616/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4e02491a9810330626ef5f61900ae5a59e7685369be8fc8362fbbc9cc56877fb
|
| 3 |
+
size 1064
|
checkpoint-14616/trainer_state.json
ADDED
|
@@ -0,0 +1,350 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.5935187339782715,
|
| 3 |
+
"best_model_checkpoint": "facial_emotions_image_detection_rafdb_google_vit/checkpoint-14616",
|
| 4 |
+
"epoch": 14.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 14616,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.48,
|
| 13 |
+
"grad_norm": 4.908149719238281,
|
| 14 |
+
"learning_rate": 2.9351896303408547e-06,
|
| 15 |
+
"loss": 1.6796,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.96,
|
| 20 |
+
"grad_norm": 6.714761257171631,
|
| 21 |
+
"learning_rate": 2.8631781084973597e-06,
|
| 22 |
+
"loss": 1.1697,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 1.0,
|
| 27 |
+
"eval_accuracy": 0.6166883963494133,
|
| 28 |
+
"eval_loss": 1.1752560138702393,
|
| 29 |
+
"eval_runtime": 43.7506,
|
| 30 |
+
"eval_samples_per_second": 70.125,
|
| 31 |
+
"eval_steps_per_second": 4.389,
|
| 32 |
+
"step": 1044
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 1.44,
|
| 36 |
+
"grad_norm": 5.284860134124756,
|
| 37 |
+
"learning_rate": 2.7911665866538647e-06,
|
| 38 |
+
"loss": 0.9317,
|
| 39 |
+
"step": 1500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 1.92,
|
| 43 |
+
"grad_norm": 6.737391471862793,
|
| 44 |
+
"learning_rate": 2.7191550648103698e-06,
|
| 45 |
+
"loss": 0.7959,
|
| 46 |
+
"step": 2000
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 2.0,
|
| 50 |
+
"eval_accuracy": 0.6988265971316818,
|
| 51 |
+
"eval_loss": 0.9186690449714661,
|
| 52 |
+
"eval_runtime": 27.1747,
|
| 53 |
+
"eval_samples_per_second": 112.899,
|
| 54 |
+
"eval_steps_per_second": 7.065,
|
| 55 |
+
"step": 2088
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 2.39,
|
| 59 |
+
"grad_norm": 7.392563819885254,
|
| 60 |
+
"learning_rate": 2.6471435429668748e-06,
|
| 61 |
+
"loss": 0.6865,
|
| 62 |
+
"step": 2500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 2.87,
|
| 66 |
+
"grad_norm": 6.488108158111572,
|
| 67 |
+
"learning_rate": 2.57513202112338e-06,
|
| 68 |
+
"loss": 0.6157,
|
| 69 |
+
"step": 3000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 3.0,
|
| 73 |
+
"eval_accuracy": 0.7314211212516297,
|
| 74 |
+
"eval_loss": 0.798936128616333,
|
| 75 |
+
"eval_runtime": 26.6756,
|
| 76 |
+
"eval_samples_per_second": 115.012,
|
| 77 |
+
"eval_steps_per_second": 7.198,
|
| 78 |
+
"step": 3132
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 3.35,
|
| 82 |
+
"grad_norm": 6.2023773193359375,
|
| 83 |
+
"learning_rate": 2.503120499279885e-06,
|
| 84 |
+
"loss": 0.5444,
|
| 85 |
+
"step": 3500
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 3.83,
|
| 89 |
+
"grad_norm": 7.885190010070801,
|
| 90 |
+
"learning_rate": 2.43110897743639e-06,
|
| 91 |
+
"loss": 0.4788,
|
| 92 |
+
"step": 4000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 4.0,
|
| 96 |
+
"eval_accuracy": 0.7496740547588006,
|
| 97 |
+
"eval_loss": 0.7347991466522217,
|
| 98 |
+
"eval_runtime": 26.7532,
|
| 99 |
+
"eval_samples_per_second": 114.678,
|
| 100 |
+
"eval_steps_per_second": 7.177,
|
| 101 |
+
"step": 4176
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 4.31,
|
| 105 |
+
"grad_norm": 6.879584312438965,
|
| 106 |
+
"learning_rate": 2.359097455592895e-06,
|
| 107 |
+
"loss": 0.4213,
|
| 108 |
+
"step": 4500
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 4.79,
|
| 112 |
+
"grad_norm": 5.687563896179199,
|
| 113 |
+
"learning_rate": 2.2870859337494e-06,
|
| 114 |
+
"loss": 0.3859,
|
| 115 |
+
"step": 5000
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 5.0,
|
| 119 |
+
"eval_accuracy": 0.7601043024771839,
|
| 120 |
+
"eval_loss": 0.6939254999160767,
|
| 121 |
+
"eval_runtime": 34.3214,
|
| 122 |
+
"eval_samples_per_second": 89.39,
|
| 123 |
+
"eval_steps_per_second": 5.594,
|
| 124 |
+
"step": 5220
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 5.27,
|
| 128 |
+
"grad_norm": 8.583109855651855,
|
| 129 |
+
"learning_rate": 2.215074411905905e-06,
|
| 130 |
+
"loss": 0.3529,
|
| 131 |
+
"step": 5500
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 5.75,
|
| 135 |
+
"grad_norm": 8.925189971923828,
|
| 136 |
+
"learning_rate": 2.14306289006241e-06,
|
| 137 |
+
"loss": 0.319,
|
| 138 |
+
"step": 6000
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 6.0,
|
| 142 |
+
"eval_accuracy": 0.7728161668839635,
|
| 143 |
+
"eval_loss": 0.6545543074607849,
|
| 144 |
+
"eval_runtime": 26.3981,
|
| 145 |
+
"eval_samples_per_second": 116.22,
|
| 146 |
+
"eval_steps_per_second": 7.273,
|
| 147 |
+
"step": 6264
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 6.23,
|
| 151 |
+
"grad_norm": 6.763670921325684,
|
| 152 |
+
"learning_rate": 2.071051368218915e-06,
|
| 153 |
+
"loss": 0.2914,
|
| 154 |
+
"step": 6500
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 6.7,
|
| 158 |
+
"grad_norm": 5.260948657989502,
|
| 159 |
+
"learning_rate": 1.9990398463754204e-06,
|
| 160 |
+
"loss": 0.2635,
|
| 161 |
+
"step": 7000
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 7.0,
|
| 165 |
+
"eval_accuracy": 0.7692307692307693,
|
| 166 |
+
"eval_loss": 0.6396156549453735,
|
| 167 |
+
"eval_runtime": 29.4081,
|
| 168 |
+
"eval_samples_per_second": 104.325,
|
| 169 |
+
"eval_steps_per_second": 6.529,
|
| 170 |
+
"step": 7308
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 7.18,
|
| 174 |
+
"grad_norm": 5.159913539886475,
|
| 175 |
+
"learning_rate": 1.9270283245319254e-06,
|
| 176 |
+
"loss": 0.2463,
|
| 177 |
+
"step": 7500
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 7.66,
|
| 181 |
+
"grad_norm": 6.131282806396484,
|
| 182 |
+
"learning_rate": 1.8550168026884303e-06,
|
| 183 |
+
"loss": 0.2254,
|
| 184 |
+
"step": 8000
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 8.0,
|
| 188 |
+
"eval_accuracy": 0.7829204693611473,
|
| 189 |
+
"eval_loss": 0.6089679598808289,
|
| 190 |
+
"eval_runtime": 26.3772,
|
| 191 |
+
"eval_samples_per_second": 116.313,
|
| 192 |
+
"eval_steps_per_second": 7.279,
|
| 193 |
+
"step": 8352
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 8.14,
|
| 197 |
+
"grad_norm": 9.69652271270752,
|
| 198 |
+
"learning_rate": 1.7830052808449353e-06,
|
| 199 |
+
"loss": 0.206,
|
| 200 |
+
"step": 8500
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 8.62,
|
| 204 |
+
"grad_norm": 19.338401794433594,
|
| 205 |
+
"learning_rate": 1.71099375900144e-06,
|
| 206 |
+
"loss": 0.189,
|
| 207 |
+
"step": 9000
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 9.0,
|
| 211 |
+
"eval_accuracy": 0.7842242503259452,
|
| 212 |
+
"eval_loss": 0.6003961563110352,
|
| 213 |
+
"eval_runtime": 26.7103,
|
| 214 |
+
"eval_samples_per_second": 114.862,
|
| 215 |
+
"eval_steps_per_second": 7.188,
|
| 216 |
+
"step": 9396
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 9.1,
|
| 220 |
+
"grad_norm": 7.628934383392334,
|
| 221 |
+
"learning_rate": 1.6389822371579451e-06,
|
| 222 |
+
"loss": 0.1867,
|
| 223 |
+
"step": 9500
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 9.58,
|
| 227 |
+
"grad_norm": 7.943005084991455,
|
| 228 |
+
"learning_rate": 1.5669707153144506e-06,
|
| 229 |
+
"loss": 0.1675,
|
| 230 |
+
"step": 10000
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 10.0,
|
| 234 |
+
"eval_accuracy": 0.7832464146023468,
|
| 235 |
+
"eval_loss": 0.6076902747154236,
|
| 236 |
+
"eval_runtime": 28.2446,
|
| 237 |
+
"eval_samples_per_second": 108.623,
|
| 238 |
+
"eval_steps_per_second": 6.798,
|
| 239 |
+
"step": 10440
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 10.06,
|
| 243 |
+
"grad_norm": 5.741328239440918,
|
| 244 |
+
"learning_rate": 1.4949591934709554e-06,
|
| 245 |
+
"loss": 0.154,
|
| 246 |
+
"step": 10500
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 10.54,
|
| 250 |
+
"grad_norm": 12.926642417907715,
|
| 251 |
+
"learning_rate": 1.4229476716274604e-06,
|
| 252 |
+
"loss": 0.1515,
|
| 253 |
+
"step": 11000
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 11.0,
|
| 257 |
+
"eval_accuracy": 0.7845501955671447,
|
| 258 |
+
"eval_loss": 0.5965496897697449,
|
| 259 |
+
"eval_runtime": 26.4509,
|
| 260 |
+
"eval_samples_per_second": 115.989,
|
| 261 |
+
"eval_steps_per_second": 7.259,
|
| 262 |
+
"step": 11484
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 11.02,
|
| 266 |
+
"grad_norm": 7.206885814666748,
|
| 267 |
+
"learning_rate": 1.3509361497839654e-06,
|
| 268 |
+
"loss": 0.1346,
|
| 269 |
+
"step": 11500
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 11.49,
|
| 273 |
+
"grad_norm": 5.869232177734375,
|
| 274 |
+
"learning_rate": 1.2789246279404704e-06,
|
| 275 |
+
"loss": 0.1292,
|
| 276 |
+
"step": 12000
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 11.97,
|
| 280 |
+
"grad_norm": 5.3356828689575195,
|
| 281 |
+
"learning_rate": 1.2069131060969755e-06,
|
| 282 |
+
"loss": 0.1212,
|
| 283 |
+
"step": 12500
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 12.0,
|
| 287 |
+
"eval_accuracy": 0.7868318122555411,
|
| 288 |
+
"eval_loss": 0.5986641645431519,
|
| 289 |
+
"eval_runtime": 28.927,
|
| 290 |
+
"eval_samples_per_second": 106.06,
|
| 291 |
+
"eval_steps_per_second": 6.637,
|
| 292 |
+
"step": 12528
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 12.45,
|
| 296 |
+
"grad_norm": 4.9229350090026855,
|
| 297 |
+
"learning_rate": 1.1349015842534805e-06,
|
| 298 |
+
"loss": 0.1127,
|
| 299 |
+
"step": 13000
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 12.93,
|
| 303 |
+
"grad_norm": 2.7797281742095947,
|
| 304 |
+
"learning_rate": 1.0628900624099857e-06,
|
| 305 |
+
"loss": 0.1147,
|
| 306 |
+
"step": 13500
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 13.0,
|
| 310 |
+
"eval_accuracy": 0.7891134289439374,
|
| 311 |
+
"eval_loss": 0.6010680794715881,
|
| 312 |
+
"eval_runtime": 26.5314,
|
| 313 |
+
"eval_samples_per_second": 115.637,
|
| 314 |
+
"eval_steps_per_second": 7.237,
|
| 315 |
+
"step": 13572
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 13.41,
|
| 319 |
+
"grad_norm": 6.798295497894287,
|
| 320 |
+
"learning_rate": 9.908785405664905e-07,
|
| 321 |
+
"loss": 0.1059,
|
| 322 |
+
"step": 14000
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 13.89,
|
| 326 |
+
"grad_norm": 7.220839500427246,
|
| 327 |
+
"learning_rate": 9.188670187229958e-07,
|
| 328 |
+
"loss": 0.0989,
|
| 329 |
+
"step": 14500
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 14.0,
|
| 333 |
+
"eval_accuracy": 0.795632333767927,
|
| 334 |
+
"eval_loss": 0.5935187339782715,
|
| 335 |
+
"eval_runtime": 26.5413,
|
| 336 |
+
"eval_samples_per_second": 115.593,
|
| 337 |
+
"eval_steps_per_second": 7.234,
|
| 338 |
+
"step": 14616
|
| 339 |
+
}
|
| 340 |
+
],
|
| 341 |
+
"logging_steps": 500,
|
| 342 |
+
"max_steps": 20880,
|
| 343 |
+
"num_input_tokens_seen": 0,
|
| 344 |
+
"num_train_epochs": 20,
|
| 345 |
+
"save_steps": 500,
|
| 346 |
+
"total_flos": 3.6241217915288666e+19,
|
| 347 |
+
"train_batch_size": 32,
|
| 348 |
+
"trial_name": null,
|
| 349 |
+
"trial_params": null
|
| 350 |
+
}
|
checkpoint-14616/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:61935e964cb014773997bd6603b4e4c4a0a91af98ad1de8925b113ef61252125
|
| 3 |
+
size 4920
|
checkpoint-17748/config.json
ADDED
|
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "google/vit-base-patch16-224",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ViTForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.0,
|
| 7 |
+
"encoder_stride": 16,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.0,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "disguist",
|
| 13 |
+
"1": "anger",
|
| 14 |
+
"2": "fear",
|
| 15 |
+
"3": "happy",
|
| 16 |
+
"4": "surprise",
|
| 17 |
+
"5": "neutral",
|
| 18 |
+
"6": "sad"
|
| 19 |
+
},
|
| 20 |
+
"image_size": 224,
|
| 21 |
+
"initializer_range": 0.02,
|
| 22 |
+
"intermediate_size": 3072,
|
| 23 |
+
"label2id": {
|
| 24 |
+
"anger": 1,
|
| 25 |
+
"disguist": 0,
|
| 26 |
+
"fear": 2,
|
| 27 |
+
"happy": 3,
|
| 28 |
+
"neutral": 5,
|
| 29 |
+
"sad": 6,
|
| 30 |
+
"surprise": 4
|
| 31 |
+
},
|
| 32 |
+
"layer_norm_eps": 1e-12,
|
| 33 |
+
"model_type": "vit",
|
| 34 |
+
"num_attention_heads": 12,
|
| 35 |
+
"num_channels": 3,
|
| 36 |
+
"num_hidden_layers": 12,
|
| 37 |
+
"patch_size": 16,
|
| 38 |
+
"problem_type": "single_label_classification",
|
| 39 |
+
"qkv_bias": true,
|
| 40 |
+
"torch_dtype": "float32",
|
| 41 |
+
"transformers_version": "4.39.3"
|
| 42 |
+
}
|
checkpoint-17748/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5c24f766742d6293755a3cbb94446ae1043b7ca2a1a63fb01258b0aa80584689
|
| 3 |
+
size 343239356
|
checkpoint-17748/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bf56edad8d1de532c865621ada2fb5f32d3b501a9884d3bb823082344249c70d
|
| 3 |
+
size 686599610
|
checkpoint-17748/preprocessor_config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_valid_processor_keys": [
|
| 3 |
+
"images",
|
| 4 |
+
"do_resize",
|
| 5 |
+
"size",
|
| 6 |
+
"resample",
|
| 7 |
+
"do_rescale",
|
| 8 |
+
"rescale_factor",
|
| 9 |
+
"do_normalize",
|
| 10 |
+
"image_mean",
|
| 11 |
+
"image_std",
|
| 12 |
+
"return_tensors",
|
| 13 |
+
"data_format",
|
| 14 |
+
"input_data_format"
|
| 15 |
+
],
|
| 16 |
+
"do_normalize": true,
|
| 17 |
+
"do_rescale": true,
|
| 18 |
+
"do_resize": true,
|
| 19 |
+
"image_mean": [
|
| 20 |
+
0.5,
|
| 21 |
+
0.5,
|
| 22 |
+
0.5
|
| 23 |
+
],
|
| 24 |
+
"image_processor_type": "ViTImageProcessor",
|
| 25 |
+
"image_std": [
|
| 26 |
+
0.5,
|
| 27 |
+
0.5,
|
| 28 |
+
0.5
|
| 29 |
+
],
|
| 30 |
+
"resample": 2,
|
| 31 |
+
"rescale_factor": 0.00392156862745098,
|
| 32 |
+
"size": {
|
| 33 |
+
"height": 224,
|
| 34 |
+
"width": 224
|
| 35 |
+
}
|
| 36 |
+
}
|
checkpoint-17748/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cf287b8b3094e74830aac58c786d3a1ac1c694f32db1ad4cb77047b0ecd80fe3
|
| 3 |
+
size 14244
|
checkpoint-17748/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e7c85352a3ece7a837d6e30238104b0d0afe0a39d7fc4fc1f471824f396d1d1e
|
| 3 |
+
size 1064
|
checkpoint-17748/trainer_state.json
ADDED
|
@@ -0,0 +1,419 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.5935187339782715,
|
| 3 |
+
"best_model_checkpoint": "facial_emotions_image_detection_rafdb_google_vit/checkpoint-14616",
|
| 4 |
+
"epoch": 17.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 17748,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.48,
|
| 13 |
+
"grad_norm": 4.908149719238281,
|
| 14 |
+
"learning_rate": 2.9351896303408547e-06,
|
| 15 |
+
"loss": 1.6796,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.96,
|
| 20 |
+
"grad_norm": 6.714761257171631,
|
| 21 |
+
"learning_rate": 2.8631781084973597e-06,
|
| 22 |
+
"loss": 1.1697,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 1.0,
|
| 27 |
+
"eval_accuracy": 0.6166883963494133,
|
| 28 |
+
"eval_loss": 1.1752560138702393,
|
| 29 |
+
"eval_runtime": 43.7506,
|
| 30 |
+
"eval_samples_per_second": 70.125,
|
| 31 |
+
"eval_steps_per_second": 4.389,
|
| 32 |
+
"step": 1044
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 1.44,
|
| 36 |
+
"grad_norm": 5.284860134124756,
|
| 37 |
+
"learning_rate": 2.7911665866538647e-06,
|
| 38 |
+
"loss": 0.9317,
|
| 39 |
+
"step": 1500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 1.92,
|
| 43 |
+
"grad_norm": 6.737391471862793,
|
| 44 |
+
"learning_rate": 2.7191550648103698e-06,
|
| 45 |
+
"loss": 0.7959,
|
| 46 |
+
"step": 2000
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 2.0,
|
| 50 |
+
"eval_accuracy": 0.6988265971316818,
|
| 51 |
+
"eval_loss": 0.9186690449714661,
|
| 52 |
+
"eval_runtime": 27.1747,
|
| 53 |
+
"eval_samples_per_second": 112.899,
|
| 54 |
+
"eval_steps_per_second": 7.065,
|
| 55 |
+
"step": 2088
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 2.39,
|
| 59 |
+
"grad_norm": 7.392563819885254,
|
| 60 |
+
"learning_rate": 2.6471435429668748e-06,
|
| 61 |
+
"loss": 0.6865,
|
| 62 |
+
"step": 2500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 2.87,
|
| 66 |
+
"grad_norm": 6.488108158111572,
|
| 67 |
+
"learning_rate": 2.57513202112338e-06,
|
| 68 |
+
"loss": 0.6157,
|
| 69 |
+
"step": 3000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 3.0,
|
| 73 |
+
"eval_accuracy": 0.7314211212516297,
|
| 74 |
+
"eval_loss": 0.798936128616333,
|
| 75 |
+
"eval_runtime": 26.6756,
|
| 76 |
+
"eval_samples_per_second": 115.012,
|
| 77 |
+
"eval_steps_per_second": 7.198,
|
| 78 |
+
"step": 3132
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 3.35,
|
| 82 |
+
"grad_norm": 6.2023773193359375,
|
| 83 |
+
"learning_rate": 2.503120499279885e-06,
|
| 84 |
+
"loss": 0.5444,
|
| 85 |
+
"step": 3500
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 3.83,
|
| 89 |
+
"grad_norm": 7.885190010070801,
|
| 90 |
+
"learning_rate": 2.43110897743639e-06,
|
| 91 |
+
"loss": 0.4788,
|
| 92 |
+
"step": 4000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 4.0,
|
| 96 |
+
"eval_accuracy": 0.7496740547588006,
|
| 97 |
+
"eval_loss": 0.7347991466522217,
|
| 98 |
+
"eval_runtime": 26.7532,
|
| 99 |
+
"eval_samples_per_second": 114.678,
|
| 100 |
+
"eval_steps_per_second": 7.177,
|
| 101 |
+
"step": 4176
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 4.31,
|
| 105 |
+
"grad_norm": 6.879584312438965,
|
| 106 |
+
"learning_rate": 2.359097455592895e-06,
|
| 107 |
+
"loss": 0.4213,
|
| 108 |
+
"step": 4500
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 4.79,
|
| 112 |
+
"grad_norm": 5.687563896179199,
|
| 113 |
+
"learning_rate": 2.2870859337494e-06,
|
| 114 |
+
"loss": 0.3859,
|
| 115 |
+
"step": 5000
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 5.0,
|
| 119 |
+
"eval_accuracy": 0.7601043024771839,
|
| 120 |
+
"eval_loss": 0.6939254999160767,
|
| 121 |
+
"eval_runtime": 34.3214,
|
| 122 |
+
"eval_samples_per_second": 89.39,
|
| 123 |
+
"eval_steps_per_second": 5.594,
|
| 124 |
+
"step": 5220
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 5.27,
|
| 128 |
+
"grad_norm": 8.583109855651855,
|
| 129 |
+
"learning_rate": 2.215074411905905e-06,
|
| 130 |
+
"loss": 0.3529,
|
| 131 |
+
"step": 5500
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 5.75,
|
| 135 |
+
"grad_norm": 8.925189971923828,
|
| 136 |
+
"learning_rate": 2.14306289006241e-06,
|
| 137 |
+
"loss": 0.319,
|
| 138 |
+
"step": 6000
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 6.0,
|
| 142 |
+
"eval_accuracy": 0.7728161668839635,
|
| 143 |
+
"eval_loss": 0.6545543074607849,
|
| 144 |
+
"eval_runtime": 26.3981,
|
| 145 |
+
"eval_samples_per_second": 116.22,
|
| 146 |
+
"eval_steps_per_second": 7.273,
|
| 147 |
+
"step": 6264
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 6.23,
|
| 151 |
+
"grad_norm": 6.763670921325684,
|
| 152 |
+
"learning_rate": 2.071051368218915e-06,
|
| 153 |
+
"loss": 0.2914,
|
| 154 |
+
"step": 6500
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 6.7,
|
| 158 |
+
"grad_norm": 5.260948657989502,
|
| 159 |
+
"learning_rate": 1.9990398463754204e-06,
|
| 160 |
+
"loss": 0.2635,
|
| 161 |
+
"step": 7000
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 7.0,
|
| 165 |
+
"eval_accuracy": 0.7692307692307693,
|
| 166 |
+
"eval_loss": 0.6396156549453735,
|
| 167 |
+
"eval_runtime": 29.4081,
|
| 168 |
+
"eval_samples_per_second": 104.325,
|
| 169 |
+
"eval_steps_per_second": 6.529,
|
| 170 |
+
"step": 7308
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 7.18,
|
| 174 |
+
"grad_norm": 5.159913539886475,
|
| 175 |
+
"learning_rate": 1.9270283245319254e-06,
|
| 176 |
+
"loss": 0.2463,
|
| 177 |
+
"step": 7500
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 7.66,
|
| 181 |
+
"grad_norm": 6.131282806396484,
|
| 182 |
+
"learning_rate": 1.8550168026884303e-06,
|
| 183 |
+
"loss": 0.2254,
|
| 184 |
+
"step": 8000
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 8.0,
|
| 188 |
+
"eval_accuracy": 0.7829204693611473,
|
| 189 |
+
"eval_loss": 0.6089679598808289,
|
| 190 |
+
"eval_runtime": 26.3772,
|
| 191 |
+
"eval_samples_per_second": 116.313,
|
| 192 |
+
"eval_steps_per_second": 7.279,
|
| 193 |
+
"step": 8352
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 8.14,
|
| 197 |
+
"grad_norm": 9.69652271270752,
|
| 198 |
+
"learning_rate": 1.7830052808449353e-06,
|
| 199 |
+
"loss": 0.206,
|
| 200 |
+
"step": 8500
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 8.62,
|
| 204 |
+
"grad_norm": 19.338401794433594,
|
| 205 |
+
"learning_rate": 1.71099375900144e-06,
|
| 206 |
+
"loss": 0.189,
|
| 207 |
+
"step": 9000
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 9.0,
|
| 211 |
+
"eval_accuracy": 0.7842242503259452,
|
| 212 |
+
"eval_loss": 0.6003961563110352,
|
| 213 |
+
"eval_runtime": 26.7103,
|
| 214 |
+
"eval_samples_per_second": 114.862,
|
| 215 |
+
"eval_steps_per_second": 7.188,
|
| 216 |
+
"step": 9396
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 9.1,
|
| 220 |
+
"grad_norm": 7.628934383392334,
|
| 221 |
+
"learning_rate": 1.6389822371579451e-06,
|
| 222 |
+
"loss": 0.1867,
|
| 223 |
+
"step": 9500
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 9.58,
|
| 227 |
+
"grad_norm": 7.943005084991455,
|
| 228 |
+
"learning_rate": 1.5669707153144506e-06,
|
| 229 |
+
"loss": 0.1675,
|
| 230 |
+
"step": 10000
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 10.0,
|
| 234 |
+
"eval_accuracy": 0.7832464146023468,
|
| 235 |
+
"eval_loss": 0.6076902747154236,
|
| 236 |
+
"eval_runtime": 28.2446,
|
| 237 |
+
"eval_samples_per_second": 108.623,
|
| 238 |
+
"eval_steps_per_second": 6.798,
|
| 239 |
+
"step": 10440
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 10.06,
|
| 243 |
+
"grad_norm": 5.741328239440918,
|
| 244 |
+
"learning_rate": 1.4949591934709554e-06,
|
| 245 |
+
"loss": 0.154,
|
| 246 |
+
"step": 10500
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 10.54,
|
| 250 |
+
"grad_norm": 12.926642417907715,
|
| 251 |
+
"learning_rate": 1.4229476716274604e-06,
|
| 252 |
+
"loss": 0.1515,
|
| 253 |
+
"step": 11000
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 11.0,
|
| 257 |
+
"eval_accuracy": 0.7845501955671447,
|
| 258 |
+
"eval_loss": 0.5965496897697449,
|
| 259 |
+
"eval_runtime": 26.4509,
|
| 260 |
+
"eval_samples_per_second": 115.989,
|
| 261 |
+
"eval_steps_per_second": 7.259,
|
| 262 |
+
"step": 11484
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 11.02,
|
| 266 |
+
"grad_norm": 7.206885814666748,
|
| 267 |
+
"learning_rate": 1.3509361497839654e-06,
|
| 268 |
+
"loss": 0.1346,
|
| 269 |
+
"step": 11500
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 11.49,
|
| 273 |
+
"grad_norm": 5.869232177734375,
|
| 274 |
+
"learning_rate": 1.2789246279404704e-06,
|
| 275 |
+
"loss": 0.1292,
|
| 276 |
+
"step": 12000
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 11.97,
|
| 280 |
+
"grad_norm": 5.3356828689575195,
|
| 281 |
+
"learning_rate": 1.2069131060969755e-06,
|
| 282 |
+
"loss": 0.1212,
|
| 283 |
+
"step": 12500
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 12.0,
|
| 287 |
+
"eval_accuracy": 0.7868318122555411,
|
| 288 |
+
"eval_loss": 0.5986641645431519,
|
| 289 |
+
"eval_runtime": 28.927,
|
| 290 |
+
"eval_samples_per_second": 106.06,
|
| 291 |
+
"eval_steps_per_second": 6.637,
|
| 292 |
+
"step": 12528
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 12.45,
|
| 296 |
+
"grad_norm": 4.9229350090026855,
|
| 297 |
+
"learning_rate": 1.1349015842534805e-06,
|
| 298 |
+
"loss": 0.1127,
|
| 299 |
+
"step": 13000
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 12.93,
|
| 303 |
+
"grad_norm": 2.7797281742095947,
|
| 304 |
+
"learning_rate": 1.0628900624099857e-06,
|
| 305 |
+
"loss": 0.1147,
|
| 306 |
+
"step": 13500
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 13.0,
|
| 310 |
+
"eval_accuracy": 0.7891134289439374,
|
| 311 |
+
"eval_loss": 0.6010680794715881,
|
| 312 |
+
"eval_runtime": 26.5314,
|
| 313 |
+
"eval_samples_per_second": 115.637,
|
| 314 |
+
"eval_steps_per_second": 7.237,
|
| 315 |
+
"step": 13572
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 13.41,
|
| 319 |
+
"grad_norm": 6.798295497894287,
|
| 320 |
+
"learning_rate": 9.908785405664905e-07,
|
| 321 |
+
"loss": 0.1059,
|
| 322 |
+
"step": 14000
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 13.89,
|
| 326 |
+
"grad_norm": 7.220839500427246,
|
| 327 |
+
"learning_rate": 9.188670187229958e-07,
|
| 328 |
+
"loss": 0.0989,
|
| 329 |
+
"step": 14500
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 14.0,
|
| 333 |
+
"eval_accuracy": 0.795632333767927,
|
| 334 |
+
"eval_loss": 0.5935187339782715,
|
| 335 |
+
"eval_runtime": 26.5413,
|
| 336 |
+
"eval_samples_per_second": 115.593,
|
| 337 |
+
"eval_steps_per_second": 7.234,
|
| 338 |
+
"step": 14616
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 14.37,
|
| 342 |
+
"grad_norm": 7.945505619049072,
|
| 343 |
+
"learning_rate": 8.468554968795008e-07,
|
| 344 |
+
"loss": 0.0972,
|
| 345 |
+
"step": 15000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 14.85,
|
| 349 |
+
"grad_norm": 2.5897045135498047,
|
| 350 |
+
"learning_rate": 7.748439750360057e-07,
|
| 351 |
+
"loss": 0.0915,
|
| 352 |
+
"step": 15500
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 15.0,
|
| 356 |
+
"eval_accuracy": 0.7940026075619296,
|
| 357 |
+
"eval_loss": 0.5987181663513184,
|
| 358 |
+
"eval_runtime": 26.6784,
|
| 359 |
+
"eval_samples_per_second": 114.999,
|
| 360 |
+
"eval_steps_per_second": 7.197,
|
| 361 |
+
"step": 15660
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 15.33,
|
| 365 |
+
"grad_norm": 5.564732074737549,
|
| 366 |
+
"learning_rate": 7.028324531925108e-07,
|
| 367 |
+
"loss": 0.0898,
|
| 368 |
+
"step": 16000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 15.8,
|
| 372 |
+
"grad_norm": 2.869757890701294,
|
| 373 |
+
"learning_rate": 6.308209313490159e-07,
|
| 374 |
+
"loss": 0.0887,
|
| 375 |
+
"step": 16500
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 16.0,
|
| 379 |
+
"eval_accuracy": 0.7966101694915254,
|
| 380 |
+
"eval_loss": 0.6017541885375977,
|
| 381 |
+
"eval_runtime": 29.5882,
|
| 382 |
+
"eval_samples_per_second": 103.69,
|
| 383 |
+
"eval_steps_per_second": 6.489,
|
| 384 |
+
"step": 16704
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 16.28,
|
| 388 |
+
"grad_norm": 10.52423095703125,
|
| 389 |
+
"learning_rate": 5.588094095055209e-07,
|
| 390 |
+
"loss": 0.0812,
|
| 391 |
+
"step": 17000
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 16.76,
|
| 395 |
+
"grad_norm": 2.270158052444458,
|
| 396 |
+
"learning_rate": 4.867978876620259e-07,
|
| 397 |
+
"loss": 0.0857,
|
| 398 |
+
"step": 17500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 17.0,
|
| 402 |
+
"eval_accuracy": 0.7985658409387223,
|
| 403 |
+
"eval_loss": 0.6071637272834778,
|
| 404 |
+
"eval_runtime": 32.54,
|
| 405 |
+
"eval_samples_per_second": 94.284,
|
| 406 |
+
"eval_steps_per_second": 5.9,
|
| 407 |
+
"step": 17748
|
| 408 |
+
}
|
| 409 |
+
],
|
| 410 |
+
"logging_steps": 500,
|
| 411 |
+
"max_steps": 20880,
|
| 412 |
+
"num_input_tokens_seen": 0,
|
| 413 |
+
"num_train_epochs": 20,
|
| 414 |
+
"save_steps": 500,
|
| 415 |
+
"total_flos": 4.400719318285052e+19,
|
| 416 |
+
"train_batch_size": 32,
|
| 417 |
+
"trial_name": null,
|
| 418 |
+
"trial_params": null
|
| 419 |
+
}
|
checkpoint-17748/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:61935e964cb014773997bd6603b4e4c4a0a91af98ad1de8925b113ef61252125
|
| 3 |
+
size 4920
|