json files from train
Browse files- all_results.json +12 -0
- config.json +127 -0
- eval_results.json +8 -0
- preprocessor_config.json +9 -0
- train_results.json +7 -0
- trainer_state.json +409 -0
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.9587628865979382,
|
| 4 |
+
"eval_loss": 0.15296605229377747,
|
| 5 |
+
"eval_runtime": 37.6058,
|
| 6 |
+
"eval_samples_per_second": 2.579,
|
| 7 |
+
"eval_steps_per_second": 0.186,
|
| 8 |
+
"train_loss": 0.30126741142833935,
|
| 9 |
+
"train_runtime": 1869.8807,
|
| 10 |
+
"train_samples_per_second": 5.722,
|
| 11 |
+
"train_steps_per_second": 0.182
|
| 12 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,127 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/wav2vec2-base",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"adapter_attn_dim": null,
|
| 5 |
+
"adapter_kernel_size": 3,
|
| 6 |
+
"adapter_stride": 2,
|
| 7 |
+
"add_adapter": false,
|
| 8 |
+
"apply_spec_augment": true,
|
| 9 |
+
"architectures": [
|
| 10 |
+
"Wav2Vec2ForSequenceClassification"
|
| 11 |
+
],
|
| 12 |
+
"attention_dropout": 0.1,
|
| 13 |
+
"bos_token_id": 1,
|
| 14 |
+
"classifier_proj_size": 256,
|
| 15 |
+
"codevector_dim": 256,
|
| 16 |
+
"contrastive_logits_temperature": 0.1,
|
| 17 |
+
"conv_bias": false,
|
| 18 |
+
"conv_dim": [
|
| 19 |
+
512,
|
| 20 |
+
512,
|
| 21 |
+
512,
|
| 22 |
+
512,
|
| 23 |
+
512,
|
| 24 |
+
512,
|
| 25 |
+
512
|
| 26 |
+
],
|
| 27 |
+
"conv_kernel": [
|
| 28 |
+
10,
|
| 29 |
+
3,
|
| 30 |
+
3,
|
| 31 |
+
3,
|
| 32 |
+
3,
|
| 33 |
+
2,
|
| 34 |
+
2
|
| 35 |
+
],
|
| 36 |
+
"conv_stride": [
|
| 37 |
+
5,
|
| 38 |
+
2,
|
| 39 |
+
2,
|
| 40 |
+
2,
|
| 41 |
+
2,
|
| 42 |
+
2,
|
| 43 |
+
2
|
| 44 |
+
],
|
| 45 |
+
"ctc_loss_reduction": "sum",
|
| 46 |
+
"ctc_zero_infinity": false,
|
| 47 |
+
"diversity_loss_weight": 0.1,
|
| 48 |
+
"do_stable_layer_norm": false,
|
| 49 |
+
"eos_token_id": 2,
|
| 50 |
+
"feat_extract_activation": "gelu",
|
| 51 |
+
"feat_extract_norm": "group",
|
| 52 |
+
"feat_proj_dropout": 0.1,
|
| 53 |
+
"feat_quantizer_dropout": 0.0,
|
| 54 |
+
"final_dropout": 0.0,
|
| 55 |
+
"finetuning_task": "audio-classification",
|
| 56 |
+
"freeze_feat_extract_train": true,
|
| 57 |
+
"hidden_act": "gelu",
|
| 58 |
+
"hidden_dropout": 0.1,
|
| 59 |
+
"hidden_size": 768,
|
| 60 |
+
"id2label": {
|
| 61 |
+
"0": "I+have+one+now",
|
| 62 |
+
"1": "I+only+have+one"
|
| 63 |
+
},
|
| 64 |
+
"initializer_range": 0.02,
|
| 65 |
+
"intermediate_size": 3072,
|
| 66 |
+
"label2id": {
|
| 67 |
+
"I+have+one+now": "0",
|
| 68 |
+
"I+only+have+one": "1"
|
| 69 |
+
},
|
| 70 |
+
"layer_norm_eps": 1e-05,
|
| 71 |
+
"layerdrop": 0.0,
|
| 72 |
+
"mask_channel_length": 10,
|
| 73 |
+
"mask_channel_min_space": 1,
|
| 74 |
+
"mask_channel_other": 0.0,
|
| 75 |
+
"mask_channel_prob": 0.0,
|
| 76 |
+
"mask_channel_selection": "static",
|
| 77 |
+
"mask_feature_length": 10,
|
| 78 |
+
"mask_feature_min_masks": 0,
|
| 79 |
+
"mask_feature_prob": 0.0,
|
| 80 |
+
"mask_time_length": 10,
|
| 81 |
+
"mask_time_min_masks": 2,
|
| 82 |
+
"mask_time_min_space": 1,
|
| 83 |
+
"mask_time_other": 0.0,
|
| 84 |
+
"mask_time_prob": 0.05,
|
| 85 |
+
"mask_time_selection": "static",
|
| 86 |
+
"model_type": "wav2vec2",
|
| 87 |
+
"no_mask_channel_overlap": false,
|
| 88 |
+
"no_mask_time_overlap": false,
|
| 89 |
+
"num_adapter_layers": 3,
|
| 90 |
+
"num_attention_heads": 12,
|
| 91 |
+
"num_codevector_groups": 2,
|
| 92 |
+
"num_codevectors_per_group": 320,
|
| 93 |
+
"num_conv_pos_embedding_groups": 16,
|
| 94 |
+
"num_conv_pos_embeddings": 128,
|
| 95 |
+
"num_feat_extract_layers": 7,
|
| 96 |
+
"num_hidden_layers": 12,
|
| 97 |
+
"num_negatives": 100,
|
| 98 |
+
"output_hidden_size": 768,
|
| 99 |
+
"pad_token_id": 0,
|
| 100 |
+
"proj_codevector_dim": 256,
|
| 101 |
+
"tdnn_dilation": [
|
| 102 |
+
1,
|
| 103 |
+
2,
|
| 104 |
+
3,
|
| 105 |
+
1,
|
| 106 |
+
1
|
| 107 |
+
],
|
| 108 |
+
"tdnn_dim": [
|
| 109 |
+
512,
|
| 110 |
+
512,
|
| 111 |
+
512,
|
| 112 |
+
512,
|
| 113 |
+
1500
|
| 114 |
+
],
|
| 115 |
+
"tdnn_kernel": [
|
| 116 |
+
5,
|
| 117 |
+
3,
|
| 118 |
+
3,
|
| 119 |
+
1,
|
| 120 |
+
1
|
| 121 |
+
],
|
| 122 |
+
"torch_dtype": "float32",
|
| 123 |
+
"transformers_version": "4.31.0.dev0",
|
| 124 |
+
"use_weighted_layer_sum": false,
|
| 125 |
+
"vocab_size": 32,
|
| 126 |
+
"xvector_output_dim": 512
|
| 127 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.9587628865979382,
|
| 4 |
+
"eval_loss": 0.15296605229377747,
|
| 5 |
+
"eval_runtime": 37.6058,
|
| 6 |
+
"eval_samples_per_second": 2.579,
|
| 7 |
+
"eval_steps_per_second": 0.186
|
| 8 |
+
}
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_normalize": true,
|
| 3 |
+
"feature_extractor_type": "Wav2Vec2FeatureExtractor",
|
| 4 |
+
"feature_size": 1,
|
| 5 |
+
"padding_side": "right",
|
| 6 |
+
"padding_value": 0.0,
|
| 7 |
+
"return_attention_mask": false,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"train_loss": 0.30126741142833935,
|
| 4 |
+
"train_runtime": 1869.8807,
|
| 5 |
+
"train_samples_per_second": 5.722,
|
| 6 |
+
"train_steps_per_second": 0.182
|
| 7 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,409 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9587628865979382,
|
| 3 |
+
"best_model_checkpoint": "wav2vec2-base_have_one/checkpoint-187",
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
+
"global_step": 340,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.59,
|
| 12 |
+
"learning_rate": 8.823529411764707e-06,
|
| 13 |
+
"loss": 0.6902,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.0,
|
| 18 |
+
"eval_accuracy": 0.4845360824742268,
|
| 19 |
+
"eval_loss": 0.6957051157951355,
|
| 20 |
+
"eval_runtime": 37.1182,
|
| 21 |
+
"eval_samples_per_second": 2.613,
|
| 22 |
+
"eval_steps_per_second": 0.189,
|
| 23 |
+
"step": 17
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 1.18,
|
| 27 |
+
"learning_rate": 1.7647058823529414e-05,
|
| 28 |
+
"loss": 0.6902,
|
| 29 |
+
"step": 20
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 1.76,
|
| 33 |
+
"learning_rate": 2.647058823529412e-05,
|
| 34 |
+
"loss": 0.68,
|
| 35 |
+
"step": 30
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"epoch": 2.0,
|
| 39 |
+
"eval_accuracy": 0.6907216494845361,
|
| 40 |
+
"eval_loss": 0.6642839908599854,
|
| 41 |
+
"eval_runtime": 38.6941,
|
| 42 |
+
"eval_samples_per_second": 2.507,
|
| 43 |
+
"eval_steps_per_second": 0.181,
|
| 44 |
+
"step": 34
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 2.35,
|
| 48 |
+
"learning_rate": 2.9411764705882354e-05,
|
| 49 |
+
"loss": 0.6778,
|
| 50 |
+
"step": 40
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 2.94,
|
| 54 |
+
"learning_rate": 2.8431372549019608e-05,
|
| 55 |
+
"loss": 0.6298,
|
| 56 |
+
"step": 50
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 3.0,
|
| 60 |
+
"eval_accuracy": 0.5360824742268041,
|
| 61 |
+
"eval_loss": 0.6741653680801392,
|
| 62 |
+
"eval_runtime": 38.2882,
|
| 63 |
+
"eval_samples_per_second": 2.533,
|
| 64 |
+
"eval_steps_per_second": 0.183,
|
| 65 |
+
"step": 51
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 3.53,
|
| 69 |
+
"learning_rate": 2.745098039215686e-05,
|
| 70 |
+
"loss": 0.5278,
|
| 71 |
+
"step": 60
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 4.0,
|
| 75 |
+
"eval_accuracy": 0.6804123711340206,
|
| 76 |
+
"eval_loss": 0.6006929278373718,
|
| 77 |
+
"eval_runtime": 38.4017,
|
| 78 |
+
"eval_samples_per_second": 2.526,
|
| 79 |
+
"eval_steps_per_second": 0.182,
|
| 80 |
+
"step": 68
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 4.12,
|
| 84 |
+
"learning_rate": 2.656862745098039e-05,
|
| 85 |
+
"loss": 0.5259,
|
| 86 |
+
"step": 70
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 4.71,
|
| 90 |
+
"learning_rate": 2.5588235294117648e-05,
|
| 91 |
+
"loss": 0.4355,
|
| 92 |
+
"step": 80
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 5.0,
|
| 96 |
+
"eval_accuracy": 0.7422680412371134,
|
| 97 |
+
"eval_loss": 0.517099916934967,
|
| 98 |
+
"eval_runtime": 35.8028,
|
| 99 |
+
"eval_samples_per_second": 2.709,
|
| 100 |
+
"eval_steps_per_second": 0.196,
|
| 101 |
+
"step": 85
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 5.29,
|
| 105 |
+
"learning_rate": 2.4705882352941174e-05,
|
| 106 |
+
"loss": 0.3802,
|
| 107 |
+
"step": 90
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 5.88,
|
| 111 |
+
"learning_rate": 2.372549019607843e-05,
|
| 112 |
+
"loss": 0.4019,
|
| 113 |
+
"step": 100
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 6.0,
|
| 117 |
+
"eval_accuracy": 0.865979381443299,
|
| 118 |
+
"eval_loss": 0.3196536898612976,
|
| 119 |
+
"eval_runtime": 38.286,
|
| 120 |
+
"eval_samples_per_second": 2.534,
|
| 121 |
+
"eval_steps_per_second": 0.183,
|
| 122 |
+
"step": 102
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 6.47,
|
| 126 |
+
"learning_rate": 2.2745098039215685e-05,
|
| 127 |
+
"loss": 0.3744,
|
| 128 |
+
"step": 110
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 7.0,
|
| 132 |
+
"eval_accuracy": 0.865979381443299,
|
| 133 |
+
"eval_loss": 0.3025638461112976,
|
| 134 |
+
"eval_runtime": 38.4396,
|
| 135 |
+
"eval_samples_per_second": 2.523,
|
| 136 |
+
"eval_steps_per_second": 0.182,
|
| 137 |
+
"step": 119
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 7.06,
|
| 141 |
+
"learning_rate": 2.1764705882352943e-05,
|
| 142 |
+
"loss": 0.3023,
|
| 143 |
+
"step": 120
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 7.65,
|
| 147 |
+
"learning_rate": 2.0784313725490197e-05,
|
| 148 |
+
"loss": 0.2918,
|
| 149 |
+
"step": 130
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 8.0,
|
| 153 |
+
"eval_accuracy": 0.8969072164948454,
|
| 154 |
+
"eval_loss": 0.2628060579299927,
|
| 155 |
+
"eval_runtime": 38.2434,
|
| 156 |
+
"eval_samples_per_second": 2.536,
|
| 157 |
+
"eval_steps_per_second": 0.183,
|
| 158 |
+
"step": 136
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 8.24,
|
| 162 |
+
"learning_rate": 1.980392156862745e-05,
|
| 163 |
+
"loss": 0.293,
|
| 164 |
+
"step": 140
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 8.82,
|
| 168 |
+
"learning_rate": 1.8823529411764708e-05,
|
| 169 |
+
"loss": 0.2183,
|
| 170 |
+
"step": 150
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 9.0,
|
| 174 |
+
"eval_accuracy": 0.9072164948453608,
|
| 175 |
+
"eval_loss": 0.24250493943691254,
|
| 176 |
+
"eval_runtime": 38.2877,
|
| 177 |
+
"eval_samples_per_second": 2.533,
|
| 178 |
+
"eval_steps_per_second": 0.183,
|
| 179 |
+
"step": 153
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 9.41,
|
| 183 |
+
"learning_rate": 1.784313725490196e-05,
|
| 184 |
+
"loss": 0.282,
|
| 185 |
+
"step": 160
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 10.0,
|
| 189 |
+
"learning_rate": 1.6862745098039215e-05,
|
| 190 |
+
"loss": 0.279,
|
| 191 |
+
"step": 170
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 10.0,
|
| 195 |
+
"eval_accuracy": 0.8556701030927835,
|
| 196 |
+
"eval_loss": 0.3291594386100769,
|
| 197 |
+
"eval_runtime": 38.2272,
|
| 198 |
+
"eval_samples_per_second": 2.537,
|
| 199 |
+
"eval_steps_per_second": 0.183,
|
| 200 |
+
"step": 170
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 10.59,
|
| 204 |
+
"learning_rate": 1.5882352941176473e-05,
|
| 205 |
+
"loss": 0.241,
|
| 206 |
+
"step": 180
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 11.0,
|
| 210 |
+
"eval_accuracy": 0.9587628865979382,
|
| 211 |
+
"eval_loss": 0.15296605229377747,
|
| 212 |
+
"eval_runtime": 38.4242,
|
| 213 |
+
"eval_samples_per_second": 2.524,
|
| 214 |
+
"eval_steps_per_second": 0.182,
|
| 215 |
+
"step": 187
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 11.18,
|
| 219 |
+
"learning_rate": 1.4901960784313726e-05,
|
| 220 |
+
"loss": 0.1683,
|
| 221 |
+
"step": 190
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 11.76,
|
| 225 |
+
"learning_rate": 1.3921568627450982e-05,
|
| 226 |
+
"loss": 0.1902,
|
| 227 |
+
"step": 200
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 12.0,
|
| 231 |
+
"eval_accuracy": 0.9381443298969072,
|
| 232 |
+
"eval_loss": 0.16121618449687958,
|
| 233 |
+
"eval_runtime": 38.3684,
|
| 234 |
+
"eval_samples_per_second": 2.528,
|
| 235 |
+
"eval_steps_per_second": 0.182,
|
| 236 |
+
"step": 204
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 12.35,
|
| 240 |
+
"learning_rate": 1.2941176470588236e-05,
|
| 241 |
+
"loss": 0.1848,
|
| 242 |
+
"step": 210
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 12.94,
|
| 246 |
+
"learning_rate": 1.1960784313725491e-05,
|
| 247 |
+
"loss": 0.2008,
|
| 248 |
+
"step": 220
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 13.0,
|
| 252 |
+
"eval_accuracy": 0.9278350515463918,
|
| 253 |
+
"eval_loss": 0.19607779383659363,
|
| 254 |
+
"eval_runtime": 38.2401,
|
| 255 |
+
"eval_samples_per_second": 2.537,
|
| 256 |
+
"eval_steps_per_second": 0.183,
|
| 257 |
+
"step": 221
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 13.53,
|
| 261 |
+
"learning_rate": 1.0980392156862745e-05,
|
| 262 |
+
"loss": 0.1522,
|
| 263 |
+
"step": 230
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 14.0,
|
| 267 |
+
"eval_accuracy": 0.9278350515463918,
|
| 268 |
+
"eval_loss": 0.19184765219688416,
|
| 269 |
+
"eval_runtime": 38.2777,
|
| 270 |
+
"eval_samples_per_second": 2.534,
|
| 271 |
+
"eval_steps_per_second": 0.183,
|
| 272 |
+
"step": 238
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 14.12,
|
| 276 |
+
"learning_rate": 9.999999999999999e-06,
|
| 277 |
+
"loss": 0.1117,
|
| 278 |
+
"step": 240
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 14.71,
|
| 282 |
+
"learning_rate": 9.019607843137255e-06,
|
| 283 |
+
"loss": 0.1758,
|
| 284 |
+
"step": 250
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 15.0,
|
| 288 |
+
"eval_accuracy": 0.9278350515463918,
|
| 289 |
+
"eval_loss": 0.19222897291183472,
|
| 290 |
+
"eval_runtime": 38.3475,
|
| 291 |
+
"eval_samples_per_second": 2.53,
|
| 292 |
+
"eval_steps_per_second": 0.183,
|
| 293 |
+
"step": 255
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 15.29,
|
| 297 |
+
"learning_rate": 8.03921568627451e-06,
|
| 298 |
+
"loss": 0.1421,
|
| 299 |
+
"step": 260
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 15.88,
|
| 303 |
+
"learning_rate": 7.058823529411765e-06,
|
| 304 |
+
"loss": 0.1571,
|
| 305 |
+
"step": 270
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 16.0,
|
| 309 |
+
"eval_accuracy": 0.9484536082474226,
|
| 310 |
+
"eval_loss": 0.18052217364311218,
|
| 311 |
+
"eval_runtime": 38.3194,
|
| 312 |
+
"eval_samples_per_second": 2.531,
|
| 313 |
+
"eval_steps_per_second": 0.183,
|
| 314 |
+
"step": 272
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 16.47,
|
| 318 |
+
"learning_rate": 6.07843137254902e-06,
|
| 319 |
+
"loss": 0.1148,
|
| 320 |
+
"step": 280
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 17.0,
|
| 324 |
+
"eval_accuracy": 0.9484536082474226,
|
| 325 |
+
"eval_loss": 0.15847636759281158,
|
| 326 |
+
"eval_runtime": 38.2687,
|
| 327 |
+
"eval_samples_per_second": 2.535,
|
| 328 |
+
"eval_steps_per_second": 0.183,
|
| 329 |
+
"step": 289
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 17.06,
|
| 333 |
+
"learning_rate": 5.098039215686275e-06,
|
| 334 |
+
"loss": 0.1421,
|
| 335 |
+
"step": 290
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 17.65,
|
| 339 |
+
"learning_rate": 4.11764705882353e-06,
|
| 340 |
+
"loss": 0.1343,
|
| 341 |
+
"step": 300
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 18.0,
|
| 345 |
+
"eval_accuracy": 0.9278350515463918,
|
| 346 |
+
"eval_loss": 0.157073974609375,
|
| 347 |
+
"eval_runtime": 38.2381,
|
| 348 |
+
"eval_samples_per_second": 2.537,
|
| 349 |
+
"eval_steps_per_second": 0.183,
|
| 350 |
+
"step": 306
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 18.24,
|
| 354 |
+
"learning_rate": 3.1372549019607846e-06,
|
| 355 |
+
"loss": 0.1074,
|
| 356 |
+
"step": 310
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 18.82,
|
| 360 |
+
"learning_rate": 2.1568627450980393e-06,
|
| 361 |
+
"loss": 0.134,
|
| 362 |
+
"step": 320
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 19.0,
|
| 366 |
+
"eval_accuracy": 0.9381443298969072,
|
| 367 |
+
"eval_loss": 0.15740494430065155,
|
| 368 |
+
"eval_runtime": 38.1402,
|
| 369 |
+
"eval_samples_per_second": 2.543,
|
| 370 |
+
"eval_steps_per_second": 0.184,
|
| 371 |
+
"step": 323
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 19.41,
|
| 375 |
+
"learning_rate": 1.1764705882352942e-06,
|
| 376 |
+
"loss": 0.1036,
|
| 377 |
+
"step": 330
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 20.0,
|
| 381 |
+
"learning_rate": 1.9607843137254904e-07,
|
| 382 |
+
"loss": 0.1029,
|
| 383 |
+
"step": 340
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 20.0,
|
| 387 |
+
"eval_accuracy": 0.9381443298969072,
|
| 388 |
+
"eval_loss": 0.16617804765701294,
|
| 389 |
+
"eval_runtime": 38.1011,
|
| 390 |
+
"eval_samples_per_second": 2.546,
|
| 391 |
+
"eval_steps_per_second": 0.184,
|
| 392 |
+
"step": 340
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 20.0,
|
| 396 |
+
"step": 340,
|
| 397 |
+
"total_flos": 9.284582669750386e+16,
|
| 398 |
+
"train_loss": 0.30126741142833935,
|
| 399 |
+
"train_runtime": 1869.8807,
|
| 400 |
+
"train_samples_per_second": 5.722,
|
| 401 |
+
"train_steps_per_second": 0.182
|
| 402 |
+
}
|
| 403 |
+
],
|
| 404 |
+
"max_steps": 340,
|
| 405 |
+
"num_train_epochs": 20,
|
| 406 |
+
"total_flos": 9.284582669750386e+16,
|
| 407 |
+
"trial_name": null,
|
| 408 |
+
"trial_params": null
|
| 409 |
+
}
|