Commit
•
f5d2c47
1
Parent(s):
fef3774
Training in progress, epoch 1
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- model.safetensors +1 -1
- run-13/checkpoint-144/config.json +80 -0
- run-13/checkpoint-144/model.safetensors +3 -0
- run-13/checkpoint-144/optimizer.pt +3 -0
- run-13/checkpoint-144/preprocessor_config.json +9 -0
- run-13/checkpoint-144/rng_state.pth +3 -0
- run-13/checkpoint-144/scheduler.pt +3 -0
- run-13/checkpoint-144/trainer_state.json +247 -0
- run-13/checkpoint-144/training_args.bin +3 -0
- run-13/checkpoint-192/config.json +80 -0
- run-13/checkpoint-192/model.safetensors +3 -0
- run-13/checkpoint-192/optimizer.pt +3 -0
- run-13/checkpoint-192/preprocessor_config.json +9 -0
- run-13/checkpoint-192/rng_state.pth +3 -0
- run-13/checkpoint-192/scheduler.pt +3 -0
- run-13/checkpoint-192/trainer_state.json +326 -0
- run-13/checkpoint-192/training_args.bin +3 -0
- run-13/checkpoint-96/config.json +80 -0
- run-13/checkpoint-96/model.safetensors +3 -0
- run-13/checkpoint-96/optimizer.pt +3 -0
- run-13/checkpoint-96/preprocessor_config.json +9 -0
- run-13/checkpoint-96/rng_state.pth +3 -0
- run-13/checkpoint-96/scheduler.pt +3 -0
- run-13/checkpoint-96/trainer_state.json +175 -0
- run-13/checkpoint-96/training_args.bin +3 -0
- run-14/checkpoint-144/config.json +80 -0
- run-14/checkpoint-144/model.safetensors +3 -0
- run-14/checkpoint-144/optimizer.pt +3 -0
- run-14/checkpoint-144/preprocessor_config.json +9 -0
- run-14/checkpoint-144/rng_state.pth +3 -0
- run-14/checkpoint-144/scheduler.pt +3 -0
- run-14/checkpoint-144/trainer_state.json +247 -0
- run-14/checkpoint-144/training_args.bin +3 -0
- run-14/checkpoint-48/config.json +80 -0
- run-14/checkpoint-48/model.safetensors +3 -0
- run-14/checkpoint-48/optimizer.pt +3 -0
- run-14/checkpoint-48/preprocessor_config.json +9 -0
- run-14/checkpoint-48/rng_state.pth +3 -0
- run-14/checkpoint-48/scheduler.pt +3 -0
- run-14/checkpoint-48/trainer_state.json +96 -0
- run-14/checkpoint-48/training_args.bin +3 -0
- run-14/checkpoint-96/config.json +80 -0
- run-14/checkpoint-96/model.safetensors +3 -0
- run-14/checkpoint-96/optimizer.pt +3 -0
- run-14/checkpoint-96/preprocessor_config.json +9 -0
- run-14/checkpoint-96/rng_state.pth +3 -0
- run-14/checkpoint-96/scheduler.pt +3 -0
- run-14/checkpoint-96/trainer_state.json +175 -0
- run-14/checkpoint-96/training_args.bin +3 -0
- run-15/checkpoint-144/config.json +80 -0
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 94763496
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dfa9c57c1a422417816b316606c2d878a3bcb3f7c6d7607c8f81010ace74cee3
|
| 3 |
size 94763496
|
run-13/checkpoint-144/config.json
ADDED
|
@@ -0,0 +1,80 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ntu-spml/distilhubert",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"HubertForSequenceClassification"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 1,
|
| 10 |
+
"classifier_proj_size": 256,
|
| 11 |
+
"conv_bias": false,
|
| 12 |
+
"conv_dim": [
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512
|
| 20 |
+
],
|
| 21 |
+
"conv_kernel": [
|
| 22 |
+
10,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
3,
|
| 27 |
+
2,
|
| 28 |
+
2
|
| 29 |
+
],
|
| 30 |
+
"conv_stride": [
|
| 31 |
+
5,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2,
|
| 37 |
+
2
|
| 38 |
+
],
|
| 39 |
+
"ctc_loss_reduction": "sum",
|
| 40 |
+
"ctc_zero_infinity": false,
|
| 41 |
+
"do_stable_layer_norm": false,
|
| 42 |
+
"eos_token_id": 2,
|
| 43 |
+
"feat_extract_activation": "gelu",
|
| 44 |
+
"feat_extract_norm": "group",
|
| 45 |
+
"feat_proj_dropout": 0.0,
|
| 46 |
+
"feat_proj_layer_norm": false,
|
| 47 |
+
"final_dropout": 0.0,
|
| 48 |
+
"hidden_act": "gelu",
|
| 49 |
+
"hidden_dropout": 0.1,
|
| 50 |
+
"hidden_size": 768,
|
| 51 |
+
"id2label": {
|
| 52 |
+
"0": "NOT_WORD",
|
| 53 |
+
"1": "WORD"
|
| 54 |
+
},
|
| 55 |
+
"initializer_range": 0.02,
|
| 56 |
+
"intermediate_size": 3072,
|
| 57 |
+
"label2id": {
|
| 58 |
+
"NOT_WORD": "0",
|
| 59 |
+
"WORD": "1"
|
| 60 |
+
},
|
| 61 |
+
"layer_norm_eps": 1e-05,
|
| 62 |
+
"layerdrop": 0.0,
|
| 63 |
+
"mask_feature_length": 10,
|
| 64 |
+
"mask_feature_min_masks": 0,
|
| 65 |
+
"mask_feature_prob": 0.0,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.05,
|
| 69 |
+
"model_type": "hubert",
|
| 70 |
+
"num_attention_heads": 12,
|
| 71 |
+
"num_conv_pos_embedding_groups": 16,
|
| 72 |
+
"num_conv_pos_embeddings": 128,
|
| 73 |
+
"num_feat_extract_layers": 7,
|
| 74 |
+
"num_hidden_layers": 2,
|
| 75 |
+
"pad_token_id": 0,
|
| 76 |
+
"torch_dtype": "float32",
|
| 77 |
+
"transformers_version": "4.38.1",
|
| 78 |
+
"use_weighted_layer_sum": false,
|
| 79 |
+
"vocab_size": 32
|
| 80 |
+
}
|
run-13/checkpoint-144/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a3876c6c1df2b682d9bf903a731bf1f13fc85ae1801e9e6f5c4bca95546bf699
|
| 3 |
+
size 94763496
|
run-13/checkpoint-144/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bd7e97044eeee28229692efe53fb1816326bd164c3a15e2c78dfa1fe22d19388
|
| 3 |
+
size 189552570
|
run-13/checkpoint-144/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,
|
| 7 |
+
"return_attention_mask": true,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
run-13/checkpoint-144/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:87b8bdd7b4355fd23f0b8256efb0158e4240e11263e992a13d50944c37692500
|
| 3 |
+
size 14244
|
run-13/checkpoint-144/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0d9cdbc15afa9058e3779fdb5b5182ac793f6da43b2bd229a164f2a22e5982ae
|
| 3 |
+
size 1064
|
run-13/checkpoint-144/trainer_state.json
ADDED
|
@@ -0,0 +1,247 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.7326732673267327,
|
| 3 |
+
"best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-13/checkpoint-48",
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 144,
|
| 7 |
+
"is_hyper_param_search": true,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.1,
|
| 13 |
+
"grad_norm": 1.2512309551239014,
|
| 14 |
+
"learning_rate": 1.034188035217011e-05,
|
| 15 |
+
"loss": 0.7013,
|
| 16 |
+
"step": 5
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.21,
|
| 20 |
+
"grad_norm": 1.0938308238983154,
|
| 21 |
+
"learning_rate": 2.068376070434022e-05,
|
| 22 |
+
"loss": 0.6709,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.31,
|
| 27 |
+
"grad_norm": 1.029103398323059,
|
| 28 |
+
"learning_rate": 3.102564105651033e-05,
|
| 29 |
+
"loss": 0.676,
|
| 30 |
+
"step": 15
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.42,
|
| 34 |
+
"grad_norm": 0.7623715400695801,
|
| 35 |
+
"learning_rate": 4.136752140868044e-05,
|
| 36 |
+
"loss": 0.6371,
|
| 37 |
+
"step": 20
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.52,
|
| 41 |
+
"grad_norm": 0.6755096912384033,
|
| 42 |
+
"learning_rate": 5.170940176085055e-05,
|
| 43 |
+
"loss": 0.6148,
|
| 44 |
+
"step": 25
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.62,
|
| 48 |
+
"grad_norm": 0.48364195227622986,
|
| 49 |
+
"learning_rate": 6.205128211302065e-05,
|
| 50 |
+
"loss": 0.609,
|
| 51 |
+
"step": 30
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.73,
|
| 55 |
+
"grad_norm": 3.724107503890991,
|
| 56 |
+
"learning_rate": 7.239316246519077e-05,
|
| 57 |
+
"loss": 0.5257,
|
| 58 |
+
"step": 35
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.83,
|
| 62 |
+
"grad_norm": 1.4992948770523071,
|
| 63 |
+
"learning_rate": 8.273504281736088e-05,
|
| 64 |
+
"loss": 0.7212,
|
| 65 |
+
"step": 40
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.94,
|
| 69 |
+
"grad_norm": 4.220996856689453,
|
| 70 |
+
"learning_rate": 9.307692316953099e-05,
|
| 71 |
+
"loss": 0.6503,
|
| 72 |
+
"step": 45
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.0,
|
| 76 |
+
"eval_f1": 0.7326732673267327,
|
| 77 |
+
"eval_loss": 0.7372970581054688,
|
| 78 |
+
"eval_runtime": 1.384,
|
| 79 |
+
"eval_samples_per_second": 46.242,
|
| 80 |
+
"eval_steps_per_second": 5.78,
|
| 81 |
+
"step": 48
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.04,
|
| 85 |
+
"grad_norm": 4.6990180015563965,
|
| 86 |
+
"learning_rate": 9.882241225406994e-05,
|
| 87 |
+
"loss": 0.6895,
|
| 88 |
+
"step": 50
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.15,
|
| 92 |
+
"grad_norm": 1.5386626720428467,
|
| 93 |
+
"learning_rate": 9.767331443716214e-05,
|
| 94 |
+
"loss": 0.6802,
|
| 95 |
+
"step": 55
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.25,
|
| 99 |
+
"grad_norm": 1.7925231456756592,
|
| 100 |
+
"learning_rate": 9.652421662025435e-05,
|
| 101 |
+
"loss": 0.4231,
|
| 102 |
+
"step": 60
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.35,
|
| 106 |
+
"grad_norm": 1.0101344585418701,
|
| 107 |
+
"learning_rate": 9.560493836672811e-05,
|
| 108 |
+
"loss": 0.5768,
|
| 109 |
+
"step": 65
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.46,
|
| 113 |
+
"grad_norm": 1.2472072839736938,
|
| 114 |
+
"learning_rate": 9.445584054982033e-05,
|
| 115 |
+
"loss": 0.5095,
|
| 116 |
+
"step": 70
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.56,
|
| 120 |
+
"grad_norm": 3.8222692012786865,
|
| 121 |
+
"learning_rate": 9.330674273291253e-05,
|
| 122 |
+
"loss": 0.8618,
|
| 123 |
+
"step": 75
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.67,
|
| 127 |
+
"grad_norm": 2.367830276489258,
|
| 128 |
+
"learning_rate": 9.215764491600475e-05,
|
| 129 |
+
"loss": 0.6321,
|
| 130 |
+
"step": 80
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.77,
|
| 134 |
+
"grad_norm": 1.2242908477783203,
|
| 135 |
+
"learning_rate": 9.100854709909696e-05,
|
| 136 |
+
"loss": 0.6086,
|
| 137 |
+
"step": 85
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.88,
|
| 141 |
+
"grad_norm": 1.1424415111541748,
|
| 142 |
+
"learning_rate": 8.985944928218916e-05,
|
| 143 |
+
"loss": 0.6172,
|
| 144 |
+
"step": 90
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.98,
|
| 148 |
+
"grad_norm": 1.9207895994186401,
|
| 149 |
+
"learning_rate": 8.871035146528138e-05,
|
| 150 |
+
"loss": 0.6318,
|
| 151 |
+
"step": 95
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.0,
|
| 155 |
+
"eval_f1": 0.7326732673267327,
|
| 156 |
+
"eval_loss": 0.6706695556640625,
|
| 157 |
+
"eval_runtime": 1.3744,
|
| 158 |
+
"eval_samples_per_second": 46.567,
|
| 159 |
+
"eval_steps_per_second": 5.821,
|
| 160 |
+
"step": 96
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.08,
|
| 164 |
+
"grad_norm": 1.3165736198425293,
|
| 165 |
+
"learning_rate": 8.756125364837358e-05,
|
| 166 |
+
"loss": 0.6359,
|
| 167 |
+
"step": 100
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 2.19,
|
| 171 |
+
"grad_norm": 1.2420493364334106,
|
| 172 |
+
"learning_rate": 8.64121558314658e-05,
|
| 173 |
+
"loss": 0.5142,
|
| 174 |
+
"step": 105
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 2.29,
|
| 178 |
+
"grad_norm": 1.2290922403335571,
|
| 179 |
+
"learning_rate": 8.526305801455801e-05,
|
| 180 |
+
"loss": 0.643,
|
| 181 |
+
"step": 110
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 2.4,
|
| 185 |
+
"grad_norm": 1.1532477140426636,
|
| 186 |
+
"learning_rate": 8.434377976103179e-05,
|
| 187 |
+
"loss": 0.6774,
|
| 188 |
+
"step": 115
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.5,
|
| 192 |
+
"grad_norm": 6.360584259033203,
|
| 193 |
+
"learning_rate": 8.319468194412399e-05,
|
| 194 |
+
"loss": 0.5839,
|
| 195 |
+
"step": 120
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 2.6,
|
| 199 |
+
"grad_norm": 7.5169677734375,
|
| 200 |
+
"learning_rate": 8.204558412721619e-05,
|
| 201 |
+
"loss": 0.5138,
|
| 202 |
+
"step": 125
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 2.71,
|
| 206 |
+
"grad_norm": 14.515143394470215,
|
| 207 |
+
"learning_rate": 8.089648631030841e-05,
|
| 208 |
+
"loss": 0.6207,
|
| 209 |
+
"step": 130
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 2.81,
|
| 213 |
+
"grad_norm": 2.682206153869629,
|
| 214 |
+
"learning_rate": 7.974738849340061e-05,
|
| 215 |
+
"loss": 0.4247,
|
| 216 |
+
"step": 135
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 2.92,
|
| 220 |
+
"grad_norm": 5.204345226287842,
|
| 221 |
+
"learning_rate": 7.88281102398744e-05,
|
| 222 |
+
"loss": 0.3698,
|
| 223 |
+
"step": 140
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 3.0,
|
| 227 |
+
"eval_f1": 0.7326732673267327,
|
| 228 |
+
"eval_loss": 0.9202766418457031,
|
| 229 |
+
"eval_runtime": 1.4182,
|
| 230 |
+
"eval_samples_per_second": 45.129,
|
| 231 |
+
"eval_steps_per_second": 5.641,
|
| 232 |
+
"step": 144
|
| 233 |
+
}
|
| 234 |
+
],
|
| 235 |
+
"logging_steps": 5,
|
| 236 |
+
"max_steps": 480,
|
| 237 |
+
"num_input_tokens_seen": 0,
|
| 238 |
+
"num_train_epochs": 10,
|
| 239 |
+
"save_steps": 500,
|
| 240 |
+
"total_flos": 2121874430755872.0,
|
| 241 |
+
"train_batch_size": 4,
|
| 242 |
+
"trial_name": null,
|
| 243 |
+
"trial_params": {
|
| 244 |
+
"learning_rate": 9.928205138083305e-05,
|
| 245 |
+
"per_device_train_batch_size": 4
|
| 246 |
+
}
|
| 247 |
+
}
|
run-13/checkpoint-144/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6181d77c6ec1f535fee0fbb436cf5e7d6cbe4ec68fcd3368e6e1056eece160ff
|
| 3 |
+
size 4920
|
run-13/checkpoint-192/config.json
ADDED
|
@@ -0,0 +1,80 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ntu-spml/distilhubert",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"HubertForSequenceClassification"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 1,
|
| 10 |
+
"classifier_proj_size": 256,
|
| 11 |
+
"conv_bias": false,
|
| 12 |
+
"conv_dim": [
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512
|
| 20 |
+
],
|
| 21 |
+
"conv_kernel": [
|
| 22 |
+
10,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
3,
|
| 27 |
+
2,
|
| 28 |
+
2
|
| 29 |
+
],
|
| 30 |
+
"conv_stride": [
|
| 31 |
+
5,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2,
|
| 37 |
+
2
|
| 38 |
+
],
|
| 39 |
+
"ctc_loss_reduction": "sum",
|
| 40 |
+
"ctc_zero_infinity": false,
|
| 41 |
+
"do_stable_layer_norm": false,
|
| 42 |
+
"eos_token_id": 2,
|
| 43 |
+
"feat_extract_activation": "gelu",
|
| 44 |
+
"feat_extract_norm": "group",
|
| 45 |
+
"feat_proj_dropout": 0.0,
|
| 46 |
+
"feat_proj_layer_norm": false,
|
| 47 |
+
"final_dropout": 0.0,
|
| 48 |
+
"hidden_act": "gelu",
|
| 49 |
+
"hidden_dropout": 0.1,
|
| 50 |
+
"hidden_size": 768,
|
| 51 |
+
"id2label": {
|
| 52 |
+
"0": "NOT_WORD",
|
| 53 |
+
"1": "WORD"
|
| 54 |
+
},
|
| 55 |
+
"initializer_range": 0.02,
|
| 56 |
+
"intermediate_size": 3072,
|
| 57 |
+
"label2id": {
|
| 58 |
+
"NOT_WORD": "0",
|
| 59 |
+
"WORD": "1"
|
| 60 |
+
},
|
| 61 |
+
"layer_norm_eps": 1e-05,
|
| 62 |
+
"layerdrop": 0.0,
|
| 63 |
+
"mask_feature_length": 10,
|
| 64 |
+
"mask_feature_min_masks": 0,
|
| 65 |
+
"mask_feature_prob": 0.0,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.05,
|
| 69 |
+
"model_type": "hubert",
|
| 70 |
+
"num_attention_heads": 12,
|
| 71 |
+
"num_conv_pos_embedding_groups": 16,
|
| 72 |
+
"num_conv_pos_embeddings": 128,
|
| 73 |
+
"num_feat_extract_layers": 7,
|
| 74 |
+
"num_hidden_layers": 2,
|
| 75 |
+
"pad_token_id": 0,
|
| 76 |
+
"torch_dtype": "float32",
|
| 77 |
+
"transformers_version": "4.38.1",
|
| 78 |
+
"use_weighted_layer_sum": false,
|
| 79 |
+
"vocab_size": 32
|
| 80 |
+
}
|
run-13/checkpoint-192/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3262e5adad1b2fffb615ffc1fb15d11f7d2eb9a344b05ff26e79afd8d8379e71
|
| 3 |
+
size 94763496
|
run-13/checkpoint-192/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:df96c8c3ac0bd44f961e18982f85baa470a90ac9079a906e6858e0085e01ae40
|
| 3 |
+
size 189552570
|
run-13/checkpoint-192/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,
|
| 7 |
+
"return_attention_mask": true,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
run-13/checkpoint-192/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d5aef020ca2df517540ac9ff4e195e1c41a7b85939e93195d118078f119bc949
|
| 3 |
+
size 14244
|
run-13/checkpoint-192/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3b32d127c3fc208eaf13706655c4c8adef63953c167a18e059a63ae70d423407
|
| 3 |
+
size 1064
|
run-13/checkpoint-192/trainer_state.json
ADDED
|
@@ -0,0 +1,326 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.7474747474747475,
|
| 3 |
+
"best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-13/checkpoint-192",
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 192,
|
| 7 |
+
"is_hyper_param_search": true,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.1,
|
| 13 |
+
"grad_norm": 1.2512309551239014,
|
| 14 |
+
"learning_rate": 1.034188035217011e-05,
|
| 15 |
+
"loss": 0.7013,
|
| 16 |
+
"step": 5
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.21,
|
| 20 |
+
"grad_norm": 1.0938308238983154,
|
| 21 |
+
"learning_rate": 2.068376070434022e-05,
|
| 22 |
+
"loss": 0.6709,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.31,
|
| 27 |
+
"grad_norm": 1.029103398323059,
|
| 28 |
+
"learning_rate": 3.102564105651033e-05,
|
| 29 |
+
"loss": 0.676,
|
| 30 |
+
"step": 15
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.42,
|
| 34 |
+
"grad_norm": 0.7623715400695801,
|
| 35 |
+
"learning_rate": 4.136752140868044e-05,
|
| 36 |
+
"loss": 0.6371,
|
| 37 |
+
"step": 20
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.52,
|
| 41 |
+
"grad_norm": 0.6755096912384033,
|
| 42 |
+
"learning_rate": 5.170940176085055e-05,
|
| 43 |
+
"loss": 0.6148,
|
| 44 |
+
"step": 25
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.62,
|
| 48 |
+
"grad_norm": 0.48364195227622986,
|
| 49 |
+
"learning_rate": 6.205128211302065e-05,
|
| 50 |
+
"loss": 0.609,
|
| 51 |
+
"step": 30
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.73,
|
| 55 |
+
"grad_norm": 3.724107503890991,
|
| 56 |
+
"learning_rate": 7.239316246519077e-05,
|
| 57 |
+
"loss": 0.5257,
|
| 58 |
+
"step": 35
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.83,
|
| 62 |
+
"grad_norm": 1.4992948770523071,
|
| 63 |
+
"learning_rate": 8.273504281736088e-05,
|
| 64 |
+
"loss": 0.7212,
|
| 65 |
+
"step": 40
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.94,
|
| 69 |
+
"grad_norm": 4.220996856689453,
|
| 70 |
+
"learning_rate": 9.307692316953099e-05,
|
| 71 |
+
"loss": 0.6503,
|
| 72 |
+
"step": 45
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.0,
|
| 76 |
+
"eval_f1": 0.7326732673267327,
|
| 77 |
+
"eval_loss": 0.7372970581054688,
|
| 78 |
+
"eval_runtime": 1.384,
|
| 79 |
+
"eval_samples_per_second": 46.242,
|
| 80 |
+
"eval_steps_per_second": 5.78,
|
| 81 |
+
"step": 48
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.04,
|
| 85 |
+
"grad_norm": 4.6990180015563965,
|
| 86 |
+
"learning_rate": 9.882241225406994e-05,
|
| 87 |
+
"loss": 0.6895,
|
| 88 |
+
"step": 50
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.15,
|
| 92 |
+
"grad_norm": 1.5386626720428467,
|
| 93 |
+
"learning_rate": 9.767331443716214e-05,
|
| 94 |
+
"loss": 0.6802,
|
| 95 |
+
"step": 55
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.25,
|
| 99 |
+
"grad_norm": 1.7925231456756592,
|
| 100 |
+
"learning_rate": 9.652421662025435e-05,
|
| 101 |
+
"loss": 0.4231,
|
| 102 |
+
"step": 60
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.35,
|
| 106 |
+
"grad_norm": 1.0101344585418701,
|
| 107 |
+
"learning_rate": 9.560493836672811e-05,
|
| 108 |
+
"loss": 0.5768,
|
| 109 |
+
"step": 65
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.46,
|
| 113 |
+
"grad_norm": 1.2472072839736938,
|
| 114 |
+
"learning_rate": 9.445584054982033e-05,
|
| 115 |
+
"loss": 0.5095,
|
| 116 |
+
"step": 70
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.56,
|
| 120 |
+
"grad_norm": 3.8222692012786865,
|
| 121 |
+
"learning_rate": 9.330674273291253e-05,
|
| 122 |
+
"loss": 0.8618,
|
| 123 |
+
"step": 75
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.67,
|
| 127 |
+
"grad_norm": 2.367830276489258,
|
| 128 |
+
"learning_rate": 9.215764491600475e-05,
|
| 129 |
+
"loss": 0.6321,
|
| 130 |
+
"step": 80
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.77,
|
| 134 |
+
"grad_norm": 1.2242908477783203,
|
| 135 |
+
"learning_rate": 9.100854709909696e-05,
|
| 136 |
+
"loss": 0.6086,
|
| 137 |
+
"step": 85
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.88,
|
| 141 |
+
"grad_norm": 1.1424415111541748,
|
| 142 |
+
"learning_rate": 8.985944928218916e-05,
|
| 143 |
+
"loss": 0.6172,
|
| 144 |
+
"step": 90
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.98,
|
| 148 |
+
"grad_norm": 1.9207895994186401,
|
| 149 |
+
"learning_rate": 8.871035146528138e-05,
|
| 150 |
+
"loss": 0.6318,
|
| 151 |
+
"step": 95
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.0,
|
| 155 |
+
"eval_f1": 0.7326732673267327,
|
| 156 |
+
"eval_loss": 0.6706695556640625,
|
| 157 |
+
"eval_runtime": 1.3744,
|
| 158 |
+
"eval_samples_per_second": 46.567,
|
| 159 |
+
"eval_steps_per_second": 5.821,
|
| 160 |
+
"step": 96
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.08,
|
| 164 |
+
"grad_norm": 1.3165736198425293,
|
| 165 |
+
"learning_rate": 8.756125364837358e-05,
|
| 166 |
+
"loss": 0.6359,
|
| 167 |
+
"step": 100
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 2.19,
|
| 171 |
+
"grad_norm": 1.2420493364334106,
|
| 172 |
+
"learning_rate": 8.64121558314658e-05,
|
| 173 |
+
"loss": 0.5142,
|
| 174 |
+
"step": 105
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 2.29,
|
| 178 |
+
"grad_norm": 1.2290922403335571,
|
| 179 |
+
"learning_rate": 8.526305801455801e-05,
|
| 180 |
+
"loss": 0.643,
|
| 181 |
+
"step": 110
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 2.4,
|
| 185 |
+
"grad_norm": 1.1532477140426636,
|
| 186 |
+
"learning_rate": 8.434377976103179e-05,
|
| 187 |
+
"loss": 0.6774,
|
| 188 |
+
"step": 115
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.5,
|
| 192 |
+
"grad_norm": 6.360584259033203,
|
| 193 |
+
"learning_rate": 8.319468194412399e-05,
|
| 194 |
+
"loss": 0.5839,
|
| 195 |
+
"step": 120
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 2.6,
|
| 199 |
+
"grad_norm": 7.5169677734375,
|
| 200 |
+
"learning_rate": 8.204558412721619e-05,
|
| 201 |
+
"loss": 0.5138,
|
| 202 |
+
"step": 125
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 2.71,
|
| 206 |
+
"grad_norm": 14.515143394470215,
|
| 207 |
+
"learning_rate": 8.089648631030841e-05,
|
| 208 |
+
"loss": 0.6207,
|
| 209 |
+
"step": 130
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 2.81,
|
| 213 |
+
"grad_norm": 2.682206153869629,
|
| 214 |
+
"learning_rate": 7.974738849340061e-05,
|
| 215 |
+
"loss": 0.4247,
|
| 216 |
+
"step": 135
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 2.92,
|
| 220 |
+
"grad_norm": 5.204345226287842,
|
| 221 |
+
"learning_rate": 7.88281102398744e-05,
|
| 222 |
+
"loss": 0.3698,
|
| 223 |
+
"step": 140
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 3.0,
|
| 227 |
+
"eval_f1": 0.7326732673267327,
|
| 228 |
+
"eval_loss": 0.9202766418457031,
|
| 229 |
+
"eval_runtime": 1.4182,
|
| 230 |
+
"eval_samples_per_second": 45.129,
|
| 231 |
+
"eval_steps_per_second": 5.641,
|
| 232 |
+
"step": 144
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.02,
|
| 236 |
+
"grad_norm": 6.481989860534668,
|
| 237 |
+
"learning_rate": 7.76790124229666e-05,
|
| 238 |
+
"loss": 0.6196,
|
| 239 |
+
"step": 145
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.12,
|
| 243 |
+
"grad_norm": Infinity,
|
| 244 |
+
"learning_rate": 7.675973416944037e-05,
|
| 245 |
+
"loss": 0.4192,
|
| 246 |
+
"step": 150
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.23,
|
| 250 |
+
"grad_norm": 2.9529616832733154,
|
| 251 |
+
"learning_rate": 7.561063635253257e-05,
|
| 252 |
+
"loss": 0.2996,
|
| 253 |
+
"step": 155
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 3.33,
|
| 257 |
+
"grad_norm": 13.315085411071777,
|
| 258 |
+
"learning_rate": 7.469135809900635e-05,
|
| 259 |
+
"loss": 0.6014,
|
| 260 |
+
"step": 160
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 3.44,
|
| 264 |
+
"grad_norm": 3.830242156982422,
|
| 265 |
+
"learning_rate": 7.354226028209855e-05,
|
| 266 |
+
"loss": 0.2117,
|
| 267 |
+
"step": 165
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 3.54,
|
| 271 |
+
"grad_norm": 26.10947608947754,
|
| 272 |
+
"learning_rate": 7.239316246519077e-05,
|
| 273 |
+
"loss": 0.2918,
|
| 274 |
+
"step": 170
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 3.65,
|
| 278 |
+
"grad_norm": 55.73325729370117,
|
| 279 |
+
"learning_rate": 7.124406464828297e-05,
|
| 280 |
+
"loss": 0.775,
|
| 281 |
+
"step": 175
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 3.75,
|
| 285 |
+
"grad_norm": 1.9401394128799438,
|
| 286 |
+
"learning_rate": 7.009496683137518e-05,
|
| 287 |
+
"loss": 0.4163,
|
| 288 |
+
"step": 180
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 3.85,
|
| 292 |
+
"grad_norm": 0.818503201007843,
|
| 293 |
+
"learning_rate": 6.89458690144674e-05,
|
| 294 |
+
"loss": 0.3559,
|
| 295 |
+
"step": 185
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 3.96,
|
| 299 |
+
"grad_norm": 6.375278472900391,
|
| 300 |
+
"learning_rate": 6.77967711975596e-05,
|
| 301 |
+
"loss": 0.5548,
|
| 302 |
+
"step": 190
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 4.0,
|
| 306 |
+
"eval_f1": 0.7474747474747475,
|
| 307 |
+
"eval_loss": 1.170898675918579,
|
| 308 |
+
"eval_runtime": 1.3939,
|
| 309 |
+
"eval_samples_per_second": 45.913,
|
| 310 |
+
"eval_steps_per_second": 5.739,
|
| 311 |
+
"step": 192
|
| 312 |
+
}
|
| 313 |
+
],
|
| 314 |
+
"logging_steps": 5,
|
| 315 |
+
"max_steps": 480,
|
| 316 |
+
"num_input_tokens_seen": 0,
|
| 317 |
+
"num_train_epochs": 10,
|
| 318 |
+
"save_steps": 500,
|
| 319 |
+
"total_flos": 2891755054954176.0,
|
| 320 |
+
"train_batch_size": 4,
|
| 321 |
+
"trial_name": null,
|
| 322 |
+
"trial_params": {
|
| 323 |
+
"learning_rate": 9.928205138083305e-05,
|
| 324 |
+
"per_device_train_batch_size": 4
|
| 325 |
+
}
|
| 326 |
+
}
|
run-13/checkpoint-192/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6181d77c6ec1f535fee0fbb436cf5e7d6cbe4ec68fcd3368e6e1056eece160ff
|
| 3 |
+
size 4920
|
run-13/checkpoint-96/config.json
ADDED
|
@@ -0,0 +1,80 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ntu-spml/distilhubert",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"HubertForSequenceClassification"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 1,
|
| 10 |
+
"classifier_proj_size": 256,
|
| 11 |
+
"conv_bias": false,
|
| 12 |
+
"conv_dim": [
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512
|
| 20 |
+
],
|
| 21 |
+
"conv_kernel": [
|
| 22 |
+
10,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
3,
|
| 27 |
+
2,
|
| 28 |
+
2
|
| 29 |
+
],
|
| 30 |
+
"conv_stride": [
|
| 31 |
+
5,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2,
|
| 37 |
+
2
|
| 38 |
+
],
|
| 39 |
+
"ctc_loss_reduction": "sum",
|
| 40 |
+
"ctc_zero_infinity": false,
|
| 41 |
+
"do_stable_layer_norm": false,
|
| 42 |
+
"eos_token_id": 2,
|
| 43 |
+
"feat_extract_activation": "gelu",
|
| 44 |
+
"feat_extract_norm": "group",
|
| 45 |
+
"feat_proj_dropout": 0.0,
|
| 46 |
+
"feat_proj_layer_norm": false,
|
| 47 |
+
"final_dropout": 0.0,
|
| 48 |
+
"hidden_act": "gelu",
|
| 49 |
+
"hidden_dropout": 0.1,
|
| 50 |
+
"hidden_size": 768,
|
| 51 |
+
"id2label": {
|
| 52 |
+
"0": "NOT_WORD",
|
| 53 |
+
"1": "WORD"
|
| 54 |
+
},
|
| 55 |
+
"initializer_range": 0.02,
|
| 56 |
+
"intermediate_size": 3072,
|
| 57 |
+
"label2id": {
|
| 58 |
+
"NOT_WORD": "0",
|
| 59 |
+
"WORD": "1"
|
| 60 |
+
},
|
| 61 |
+
"layer_norm_eps": 1e-05,
|
| 62 |
+
"layerdrop": 0.0,
|
| 63 |
+
"mask_feature_length": 10,
|
| 64 |
+
"mask_feature_min_masks": 0,
|
| 65 |
+
"mask_feature_prob": 0.0,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.05,
|
| 69 |
+
"model_type": "hubert",
|
| 70 |
+
"num_attention_heads": 12,
|
| 71 |
+
"num_conv_pos_embedding_groups": 16,
|
| 72 |
+
"num_conv_pos_embeddings": 128,
|
| 73 |
+
"num_feat_extract_layers": 7,
|
| 74 |
+
"num_hidden_layers": 2,
|
| 75 |
+
"pad_token_id": 0,
|
| 76 |
+
"torch_dtype": "float32",
|
| 77 |
+
"transformers_version": "4.38.1",
|
| 78 |
+
"use_weighted_layer_sum": false,
|
| 79 |
+
"vocab_size": 32
|
| 80 |
+
}
|
run-13/checkpoint-96/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c12b8d2b75a071bf5e832285aa494f6c0b5a6337c618f408c9b1b0619da5aaca
|
| 3 |
+
size 94763496
|
run-13/checkpoint-96/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:27693d8a54d2c8c8ccf19cc1c96d45d895fda3a5dd6ee169ec6503d7194cbd67
|
| 3 |
+
size 189552570
|
run-13/checkpoint-96/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,
|
| 7 |
+
"return_attention_mask": true,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
run-13/checkpoint-96/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:39dbf03bf644af79257aec95c925042cb81a469bfcc7a839a95d68f1d0425513
|
| 3 |
+
size 14244
|
run-13/checkpoint-96/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5c63af2dd8f0e1bb01f0d7e66947ad84d12f8c6c65a2423368262dc1e83a7ab1
|
| 3 |
+
size 1064
|
run-13/checkpoint-96/trainer_state.json
ADDED
|
@@ -0,0 +1,175 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.7326732673267327,
|
| 3 |
+
"best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-13/checkpoint-48",
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 96,
|
| 7 |
+
"is_hyper_param_search": true,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.1,
|
| 13 |
+
"grad_norm": 1.2512309551239014,
|
| 14 |
+
"learning_rate": 1.034188035217011e-05,
|
| 15 |
+
"loss": 0.7013,
|
| 16 |
+
"step": 5
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.21,
|
| 20 |
+
"grad_norm": 1.0938308238983154,
|
| 21 |
+
"learning_rate": 2.068376070434022e-05,
|
| 22 |
+
"loss": 0.6709,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.31,
|
| 27 |
+
"grad_norm": 1.029103398323059,
|
| 28 |
+
"learning_rate": 3.102564105651033e-05,
|
| 29 |
+
"loss": 0.676,
|
| 30 |
+
"step": 15
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.42,
|
| 34 |
+
"grad_norm": 0.7623715400695801,
|
| 35 |
+
"learning_rate": 4.136752140868044e-05,
|
| 36 |
+
"loss": 0.6371,
|
| 37 |
+
"step": 20
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.52,
|
| 41 |
+
"grad_norm": 0.6755096912384033,
|
| 42 |
+
"learning_rate": 5.170940176085055e-05,
|
| 43 |
+
"loss": 0.6148,
|
| 44 |
+
"step": 25
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.62,
|
| 48 |
+
"grad_norm": 0.48364195227622986,
|
| 49 |
+
"learning_rate": 6.205128211302065e-05,
|
| 50 |
+
"loss": 0.609,
|
| 51 |
+
"step": 30
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.73,
|
| 55 |
+
"grad_norm": 3.724107503890991,
|
| 56 |
+
"learning_rate": 7.239316246519077e-05,
|
| 57 |
+
"loss": 0.5257,
|
| 58 |
+
"step": 35
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.83,
|
| 62 |
+
"grad_norm": 1.4992948770523071,
|
| 63 |
+
"learning_rate": 8.273504281736088e-05,
|
| 64 |
+
"loss": 0.7212,
|
| 65 |
+
"step": 40
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.94,
|
| 69 |
+
"grad_norm": 4.220996856689453,
|
| 70 |
+
"learning_rate": 9.307692316953099e-05,
|
| 71 |
+
"loss": 0.6503,
|
| 72 |
+
"step": 45
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.0,
|
| 76 |
+
"eval_f1": 0.7326732673267327,
|
| 77 |
+
"eval_loss": 0.7372970581054688,
|
| 78 |
+
"eval_runtime": 1.384,
|
| 79 |
+
"eval_samples_per_second": 46.242,
|
| 80 |
+
"eval_steps_per_second": 5.78,
|
| 81 |
+
"step": 48
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.04,
|
| 85 |
+
"grad_norm": 4.6990180015563965,
|
| 86 |
+
"learning_rate": 9.882241225406994e-05,
|
| 87 |
+
"loss": 0.6895,
|
| 88 |
+
"step": 50
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.15,
|
| 92 |
+
"grad_norm": 1.5386626720428467,
|
| 93 |
+
"learning_rate": 9.767331443716214e-05,
|
| 94 |
+
"loss": 0.6802,
|
| 95 |
+
"step": 55
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.25,
|
| 99 |
+
"grad_norm": 1.7925231456756592,
|
| 100 |
+
"learning_rate": 9.652421662025435e-05,
|
| 101 |
+
"loss": 0.4231,
|
| 102 |
+
"step": 60
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.35,
|
| 106 |
+
"grad_norm": 1.0101344585418701,
|
| 107 |
+
"learning_rate": 9.560493836672811e-05,
|
| 108 |
+
"loss": 0.5768,
|
| 109 |
+
"step": 65
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.46,
|
| 113 |
+
"grad_norm": 1.2472072839736938,
|
| 114 |
+
"learning_rate": 9.445584054982033e-05,
|
| 115 |
+
"loss": 0.5095,
|
| 116 |
+
"step": 70
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.56,
|
| 120 |
+
"grad_norm": 3.8222692012786865,
|
| 121 |
+
"learning_rate": 9.330674273291253e-05,
|
| 122 |
+
"loss": 0.8618,
|
| 123 |
+
"step": 75
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.67,
|
| 127 |
+
"grad_norm": 2.367830276489258,
|
| 128 |
+
"learning_rate": 9.215764491600475e-05,
|
| 129 |
+
"loss": 0.6321,
|
| 130 |
+
"step": 80
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.77,
|
| 134 |
+
"grad_norm": 1.2242908477783203,
|
| 135 |
+
"learning_rate": 9.100854709909696e-05,
|
| 136 |
+
"loss": 0.6086,
|
| 137 |
+
"step": 85
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.88,
|
| 141 |
+
"grad_norm": 1.1424415111541748,
|
| 142 |
+
"learning_rate": 8.985944928218916e-05,
|
| 143 |
+
"loss": 0.6172,
|
| 144 |
+
"step": 90
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.98,
|
| 148 |
+
"grad_norm": 1.9207895994186401,
|
| 149 |
+
"learning_rate": 8.871035146528138e-05,
|
| 150 |
+
"loss": 0.6318,
|
| 151 |
+
"step": 95
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.0,
|
| 155 |
+
"eval_f1": 0.7326732673267327,
|
| 156 |
+
"eval_loss": 0.6706695556640625,
|
| 157 |
+
"eval_runtime": 1.3744,
|
| 158 |
+
"eval_samples_per_second": 46.567,
|
| 159 |
+
"eval_steps_per_second": 5.821,
|
| 160 |
+
"step": 96
|
| 161 |
+
}
|
| 162 |
+
],
|
| 163 |
+
"logging_steps": 5,
|
| 164 |
+
"max_steps": 480,
|
| 165 |
+
"num_input_tokens_seen": 0,
|
| 166 |
+
"num_train_epochs": 10,
|
| 167 |
+
"save_steps": 500,
|
| 168 |
+
"total_flos": 1442567462539200.0,
|
| 169 |
+
"train_batch_size": 4,
|
| 170 |
+
"trial_name": null,
|
| 171 |
+
"trial_params": {
|
| 172 |
+
"learning_rate": 9.928205138083305e-05,
|
| 173 |
+
"per_device_train_batch_size": 4
|
| 174 |
+
}
|
| 175 |
+
}
|
run-13/checkpoint-96/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6181d77c6ec1f535fee0fbb436cf5e7d6cbe4ec68fcd3368e6e1056eece160ff
|
| 3 |
+
size 4920
|
run-14/checkpoint-144/config.json
ADDED
|
@@ -0,0 +1,80 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ntu-spml/distilhubert",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"HubertForSequenceClassification"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 1,
|
| 10 |
+
"classifier_proj_size": 256,
|
| 11 |
+
"conv_bias": false,
|
| 12 |
+
"conv_dim": [
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512
|
| 20 |
+
],
|
| 21 |
+
"conv_kernel": [
|
| 22 |
+
10,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
3,
|
| 27 |
+
2,
|
| 28 |
+
2
|
| 29 |
+
],
|
| 30 |
+
"conv_stride": [
|
| 31 |
+
5,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2,
|
| 37 |
+
2
|
| 38 |
+
],
|
| 39 |
+
"ctc_loss_reduction": "sum",
|
| 40 |
+
"ctc_zero_infinity": false,
|
| 41 |
+
"do_stable_layer_norm": false,
|
| 42 |
+
"eos_token_id": 2,
|
| 43 |
+
"feat_extract_activation": "gelu",
|
| 44 |
+
"feat_extract_norm": "group",
|
| 45 |
+
"feat_proj_dropout": 0.0,
|
| 46 |
+
"feat_proj_layer_norm": false,
|
| 47 |
+
"final_dropout": 0.0,
|
| 48 |
+
"hidden_act": "gelu",
|
| 49 |
+
"hidden_dropout": 0.1,
|
| 50 |
+
"hidden_size": 768,
|
| 51 |
+
"id2label": {
|
| 52 |
+
"0": "NOT_WORD",
|
| 53 |
+
"1": "WORD"
|
| 54 |
+
},
|
| 55 |
+
"initializer_range": 0.02,
|
| 56 |
+
"intermediate_size": 3072,
|
| 57 |
+
"label2id": {
|
| 58 |
+
"NOT_WORD": "0",
|
| 59 |
+
"WORD": "1"
|
| 60 |
+
},
|
| 61 |
+
"layer_norm_eps": 1e-05,
|
| 62 |
+
"layerdrop": 0.0,
|
| 63 |
+
"mask_feature_length": 10,
|
| 64 |
+
"mask_feature_min_masks": 0,
|
| 65 |
+
"mask_feature_prob": 0.0,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.05,
|
| 69 |
+
"model_type": "hubert",
|
| 70 |
+
"num_attention_heads": 12,
|
| 71 |
+
"num_conv_pos_embedding_groups": 16,
|
| 72 |
+
"num_conv_pos_embeddings": 128,
|
| 73 |
+
"num_feat_extract_layers": 7,
|
| 74 |
+
"num_hidden_layers": 2,
|
| 75 |
+
"pad_token_id": 0,
|
| 76 |
+
"torch_dtype": "float32",
|
| 77 |
+
"transformers_version": "4.38.1",
|
| 78 |
+
"use_weighted_layer_sum": false,
|
| 79 |
+
"vocab_size": 32
|
| 80 |
+
}
|
run-14/checkpoint-144/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dd4f9758ea480b5414cd4643e6fb5ae119b64487091753c4af492052263ff061
|
| 3 |
+
size 94763496
|
run-14/checkpoint-144/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c82c386655c5e3c53096812ec800237e6815adf2ae47b1765069888f4bc9b7a1
|
| 3 |
+
size 189552570
|
run-14/checkpoint-144/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,
|
| 7 |
+
"return_attention_mask": true,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
run-14/checkpoint-144/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:87b8bdd7b4355fd23f0b8256efb0158e4240e11263e992a13d50944c37692500
|
| 3 |
+
size 14244
|
run-14/checkpoint-144/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:51de9a2f2a1ea0a358d768e1268974f6355f99f9552b45c19d9f7b2227d52121
|
| 3 |
+
size 1064
|
run-14/checkpoint-144/trainer_state.json
ADDED
|
@@ -0,0 +1,247 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.7346938775510203,
|
| 3 |
+
"best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-14/checkpoint-144",
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 144,
|
| 7 |
+
"is_hyper_param_search": true,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.1,
|
| 13 |
+
"grad_norm": 1.263960838317871,
|
| 14 |
+
"learning_rate": 4.556569949810922e-06,
|
| 15 |
+
"loss": 0.7021,
|
| 16 |
+
"step": 5
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.21,
|
| 20 |
+
"grad_norm": 1.1750551462173462,
|
| 21 |
+
"learning_rate": 9.113139899621844e-06,
|
| 22 |
+
"loss": 0.6857,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.31,
|
| 27 |
+
"grad_norm": 1.3693294525146484,
|
| 28 |
+
"learning_rate": 1.3669709849432766e-05,
|
| 29 |
+
"loss": 0.6849,
|
| 30 |
+
"step": 15
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.42,
|
| 34 |
+
"grad_norm": 0.9440478086471558,
|
| 35 |
+
"learning_rate": 1.822627979924369e-05,
|
| 36 |
+
"loss": 0.6652,
|
| 37 |
+
"step": 20
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.52,
|
| 41 |
+
"grad_norm": 0.9551452398300171,
|
| 42 |
+
"learning_rate": 2.2782849749054612e-05,
|
| 43 |
+
"loss": 0.64,
|
| 44 |
+
"step": 25
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.62,
|
| 48 |
+
"grad_norm": 0.777751088142395,
|
| 49 |
+
"learning_rate": 2.7339419698865533e-05,
|
| 50 |
+
"loss": 0.6269,
|
| 51 |
+
"step": 30
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.73,
|
| 55 |
+
"grad_norm": 2.4816713333129883,
|
| 56 |
+
"learning_rate": 3.1895989648676456e-05,
|
| 57 |
+
"loss": 0.5586,
|
| 58 |
+
"step": 35
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.83,
|
| 62 |
+
"grad_norm": 1.696999430656433,
|
| 63 |
+
"learning_rate": 3.645255959848738e-05,
|
| 64 |
+
"loss": 0.6586,
|
| 65 |
+
"step": 40
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.94,
|
| 69 |
+
"grad_norm": 3.2336232662200928,
|
| 70 |
+
"learning_rate": 4.10091295482983e-05,
|
| 71 |
+
"loss": 0.6259,
|
| 72 |
+
"step": 45
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.0,
|
| 76 |
+
"eval_f1": 0.7326732673267327,
|
| 77 |
+
"eval_loss": 0.7427406311035156,
|
| 78 |
+
"eval_runtime": 1.4051,
|
| 79 |
+
"eval_samples_per_second": 45.548,
|
| 80 |
+
"eval_steps_per_second": 5.694,
|
| 81 |
+
"step": 48
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.04,
|
| 85 |
+
"grad_norm": 4.544165134429932,
|
| 86 |
+
"learning_rate": 4.354055729819326e-05,
|
| 87 |
+
"loss": 0.7,
|
| 88 |
+
"step": 50
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.15,
|
| 92 |
+
"grad_norm": 0.8700355887413025,
|
| 93 |
+
"learning_rate": 4.3034271748214265e-05,
|
| 94 |
+
"loss": 0.6972,
|
| 95 |
+
"step": 55
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.25,
|
| 99 |
+
"grad_norm": 1.7657314538955688,
|
| 100 |
+
"learning_rate": 4.252798619823527e-05,
|
| 101 |
+
"loss": 0.4085,
|
| 102 |
+
"step": 60
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.35,
|
| 106 |
+
"grad_norm": 0.6718289852142334,
|
| 107 |
+
"learning_rate": 4.202170064825628e-05,
|
| 108 |
+
"loss": 0.5618,
|
| 109 |
+
"step": 65
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.46,
|
| 113 |
+
"grad_norm": 1.5080432891845703,
|
| 114 |
+
"learning_rate": 4.151541509827729e-05,
|
| 115 |
+
"loss": 0.4822,
|
| 116 |
+
"step": 70
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.56,
|
| 120 |
+
"grad_norm": Infinity,
|
| 121 |
+
"learning_rate": 4.11103866582941e-05,
|
| 122 |
+
"loss": 0.7679,
|
| 123 |
+
"step": 75
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.67,
|
| 127 |
+
"grad_norm": 1.8398224115371704,
|
| 128 |
+
"learning_rate": 4.0604101108315106e-05,
|
| 129 |
+
"loss": 0.6183,
|
| 130 |
+
"step": 80
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.77,
|
| 134 |
+
"grad_norm": 1.7079250812530518,
|
| 135 |
+
"learning_rate": 4.0097815558336114e-05,
|
| 136 |
+
"loss": 0.5954,
|
| 137 |
+
"step": 85
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.88,
|
| 141 |
+
"grad_norm": 1.8899829387664795,
|
| 142 |
+
"learning_rate": 3.959153000835712e-05,
|
| 143 |
+
"loss": 0.596,
|
| 144 |
+
"step": 90
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.98,
|
| 148 |
+
"grad_norm": 3.904428243637085,
|
| 149 |
+
"learning_rate": 3.908524445837813e-05,
|
| 150 |
+
"loss": 0.6094,
|
| 151 |
+
"step": 95
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.0,
|
| 155 |
+
"eval_f1": 0.7326732673267327,
|
| 156 |
+
"eval_loss": 0.673431396484375,
|
| 157 |
+
"eval_runtime": 1.3605,
|
| 158 |
+
"eval_samples_per_second": 47.041,
|
| 159 |
+
"eval_steps_per_second": 5.88,
|
| 160 |
+
"step": 96
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.08,
|
| 164 |
+
"grad_norm": 1.0948293209075928,
|
| 165 |
+
"learning_rate": 3.857895890839914e-05,
|
| 166 |
+
"loss": 0.5749,
|
| 167 |
+
"step": 100
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 2.19,
|
| 171 |
+
"grad_norm": 1.2664576768875122,
|
| 172 |
+
"learning_rate": 3.807267335842015e-05,
|
| 173 |
+
"loss": 0.4808,
|
| 174 |
+
"step": 105
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 2.29,
|
| 178 |
+
"grad_norm": 1.7244511842727661,
|
| 179 |
+
"learning_rate": 3.7566387808441155e-05,
|
| 180 |
+
"loss": 0.5641,
|
| 181 |
+
"step": 110
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 2.4,
|
| 185 |
+
"grad_norm": 1.1561750173568726,
|
| 186 |
+
"learning_rate": 3.706010225846216e-05,
|
| 187 |
+
"loss": 0.5395,
|
| 188 |
+
"step": 115
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.5,
|
| 192 |
+
"grad_norm": Infinity,
|
| 193 |
+
"learning_rate": 3.665507381847897e-05,
|
| 194 |
+
"loss": 0.5224,
|
| 195 |
+
"step": 120
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 2.6,
|
| 199 |
+
"grad_norm": 6.8102827072143555,
|
| 200 |
+
"learning_rate": 3.614878826849998e-05,
|
| 201 |
+
"loss": 0.4427,
|
| 202 |
+
"step": 125
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 2.71,
|
| 206 |
+
"grad_norm": Infinity,
|
| 207 |
+
"learning_rate": 3.574375982851679e-05,
|
| 208 |
+
"loss": 0.6088,
|
| 209 |
+
"step": 130
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 2.81,
|
| 213 |
+
"grad_norm": 3.278684139251709,
|
| 214 |
+
"learning_rate": 3.52374742785378e-05,
|
| 215 |
+
"loss": 0.3563,
|
| 216 |
+
"step": 135
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 2.92,
|
| 220 |
+
"grad_norm": 4.377391815185547,
|
| 221 |
+
"learning_rate": 3.4731188728558806e-05,
|
| 222 |
+
"loss": 0.3177,
|
| 223 |
+
"step": 140
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 3.0,
|
| 227 |
+
"eval_f1": 0.7346938775510203,
|
| 228 |
+
"eval_loss": 0.804656982421875,
|
| 229 |
+
"eval_runtime": 1.3676,
|
| 230 |
+
"eval_samples_per_second": 46.796,
|
| 231 |
+
"eval_steps_per_second": 5.85,
|
| 232 |
+
"step": 144
|
| 233 |
+
}
|
| 234 |
+
],
|
| 235 |
+
"logging_steps": 5,
|
| 236 |
+
"max_steps": 480,
|
| 237 |
+
"num_input_tokens_seen": 0,
|
| 238 |
+
"num_train_epochs": 10,
|
| 239 |
+
"save_steps": 500,
|
| 240 |
+
"total_flos": 2121874430755872.0,
|
| 241 |
+
"train_batch_size": 4,
|
| 242 |
+
"trial_name": null,
|
| 243 |
+
"trial_params": {
|
| 244 |
+
"learning_rate": 4.374307151818485e-05,
|
| 245 |
+
"per_device_train_batch_size": 4
|
| 246 |
+
}
|
| 247 |
+
}
|
run-14/checkpoint-144/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:08323d3f888db4d96cf1275214c0bc1494a12c503c7492df7675d3c2976246ad
|
| 3 |
+
size 4920
|
run-14/checkpoint-48/config.json
ADDED
|
@@ -0,0 +1,80 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ntu-spml/distilhubert",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"HubertForSequenceClassification"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 1,
|
| 10 |
+
"classifier_proj_size": 256,
|
| 11 |
+
"conv_bias": false,
|
| 12 |
+
"conv_dim": [
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512
|
| 20 |
+
],
|
| 21 |
+
"conv_kernel": [
|
| 22 |
+
10,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
3,
|
| 27 |
+
2,
|
| 28 |
+
2
|
| 29 |
+
],
|
| 30 |
+
"conv_stride": [
|
| 31 |
+
5,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2,
|
| 37 |
+
2
|
| 38 |
+
],
|
| 39 |
+
"ctc_loss_reduction": "sum",
|
| 40 |
+
"ctc_zero_infinity": false,
|
| 41 |
+
"do_stable_layer_norm": false,
|
| 42 |
+
"eos_token_id": 2,
|
| 43 |
+
"feat_extract_activation": "gelu",
|
| 44 |
+
"feat_extract_norm": "group",
|
| 45 |
+
"feat_proj_dropout": 0.0,
|
| 46 |
+
"feat_proj_layer_norm": false,
|
| 47 |
+
"final_dropout": 0.0,
|
| 48 |
+
"hidden_act": "gelu",
|
| 49 |
+
"hidden_dropout": 0.1,
|
| 50 |
+
"hidden_size": 768,
|
| 51 |
+
"id2label": {
|
| 52 |
+
"0": "NOT_WORD",
|
| 53 |
+
"1": "WORD"
|
| 54 |
+
},
|
| 55 |
+
"initializer_range": 0.02,
|
| 56 |
+
"intermediate_size": 3072,
|
| 57 |
+
"label2id": {
|
| 58 |
+
"NOT_WORD": "0",
|
| 59 |
+
"WORD": "1"
|
| 60 |
+
},
|
| 61 |
+
"layer_norm_eps": 1e-05,
|
| 62 |
+
"layerdrop": 0.0,
|
| 63 |
+
"mask_feature_length": 10,
|
| 64 |
+
"mask_feature_min_masks": 0,
|
| 65 |
+
"mask_feature_prob": 0.0,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.05,
|
| 69 |
+
"model_type": "hubert",
|
| 70 |
+
"num_attention_heads": 12,
|
| 71 |
+
"num_conv_pos_embedding_groups": 16,
|
| 72 |
+
"num_conv_pos_embeddings": 128,
|
| 73 |
+
"num_feat_extract_layers": 7,
|
| 74 |
+
"num_hidden_layers": 2,
|
| 75 |
+
"pad_token_id": 0,
|
| 76 |
+
"torch_dtype": "float32",
|
| 77 |
+
"transformers_version": "4.38.1",
|
| 78 |
+
"use_weighted_layer_sum": false,
|
| 79 |
+
"vocab_size": 32
|
| 80 |
+
}
|
run-14/checkpoint-48/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5cc860643e0442d477e3db2d795cea596ec406f49262f0b2921abc98e53a278a
|
| 3 |
+
size 94763496
|
run-14/checkpoint-48/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:41f0c7d3378515129e4de3df8bdd50f14018dea647741bfa2a5a7b0160fb69c8
|
| 3 |
+
size 189552570
|
run-14/checkpoint-48/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,
|
| 7 |
+
"return_attention_mask": true,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
run-14/checkpoint-48/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b7251f0e64bf9e5675ed89b468a7ff74c1c3fd6457742f84db0e5e361db11f13
|
| 3 |
+
size 14244
|
run-14/checkpoint-48/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:850ccdd16609d418223240119cbf87863f10dc2294422d17358d1418f070270d
|
| 3 |
+
size 1064
|
run-14/checkpoint-48/trainer_state.json
ADDED
|
@@ -0,0 +1,96 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.7326732673267327,
|
| 3 |
+
"best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-14/checkpoint-48",
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 48,
|
| 7 |
+
"is_hyper_param_search": true,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.1,
|
| 13 |
+
"grad_norm": 1.263960838317871,
|
| 14 |
+
"learning_rate": 4.556569949810922e-06,
|
| 15 |
+
"loss": 0.7021,
|
| 16 |
+
"step": 5
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.21,
|
| 20 |
+
"grad_norm": 1.1750551462173462,
|
| 21 |
+
"learning_rate": 9.113139899621844e-06,
|
| 22 |
+
"loss": 0.6857,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.31,
|
| 27 |
+
"grad_norm": 1.3693294525146484,
|
| 28 |
+
"learning_rate": 1.3669709849432766e-05,
|
| 29 |
+
"loss": 0.6849,
|
| 30 |
+
"step": 15
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.42,
|
| 34 |
+
"grad_norm": 0.9440478086471558,
|
| 35 |
+
"learning_rate": 1.822627979924369e-05,
|
| 36 |
+
"loss": 0.6652,
|
| 37 |
+
"step": 20
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.52,
|
| 41 |
+
"grad_norm": 0.9551452398300171,
|
| 42 |
+
"learning_rate": 2.2782849749054612e-05,
|
| 43 |
+
"loss": 0.64,
|
| 44 |
+
"step": 25
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.62,
|
| 48 |
+
"grad_norm": 0.777751088142395,
|
| 49 |
+
"learning_rate": 2.7339419698865533e-05,
|
| 50 |
+
"loss": 0.6269,
|
| 51 |
+
"step": 30
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.73,
|
| 55 |
+
"grad_norm": 2.4816713333129883,
|
| 56 |
+
"learning_rate": 3.1895989648676456e-05,
|
| 57 |
+
"loss": 0.5586,
|
| 58 |
+
"step": 35
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.83,
|
| 62 |
+
"grad_norm": 1.696999430656433,
|
| 63 |
+
"learning_rate": 3.645255959848738e-05,
|
| 64 |
+
"loss": 0.6586,
|
| 65 |
+
"step": 40
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.94,
|
| 69 |
+
"grad_norm": 3.2336232662200928,
|
| 70 |
+
"learning_rate": 4.10091295482983e-05,
|
| 71 |
+
"loss": 0.6259,
|
| 72 |
+
"step": 45
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.0,
|
| 76 |
+
"eval_f1": 0.7326732673267327,
|
| 77 |
+
"eval_loss": 0.7427406311035156,
|
| 78 |
+
"eval_runtime": 1.4051,
|
| 79 |
+
"eval_samples_per_second": 45.548,
|
| 80 |
+
"eval_steps_per_second": 5.694,
|
| 81 |
+
"step": 48
|
| 82 |
+
}
|
| 83 |
+
],
|
| 84 |
+
"logging_steps": 5,
|
| 85 |
+
"max_steps": 480,
|
| 86 |
+
"num_input_tokens_seen": 0,
|
| 87 |
+
"num_train_epochs": 10,
|
| 88 |
+
"save_steps": 500,
|
| 89 |
+
"total_flos": 670686130935120.0,
|
| 90 |
+
"train_batch_size": 4,
|
| 91 |
+
"trial_name": null,
|
| 92 |
+
"trial_params": {
|
| 93 |
+
"learning_rate": 4.374307151818485e-05,
|
| 94 |
+
"per_device_train_batch_size": 4
|
| 95 |
+
}
|
| 96 |
+
}
|
run-14/checkpoint-48/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:08323d3f888db4d96cf1275214c0bc1494a12c503c7492df7675d3c2976246ad
|
| 3 |
+
size 4920
|
run-14/checkpoint-96/config.json
ADDED
|
@@ -0,0 +1,80 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ntu-spml/distilhubert",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"HubertForSequenceClassification"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 1,
|
| 10 |
+
"classifier_proj_size": 256,
|
| 11 |
+
"conv_bias": false,
|
| 12 |
+
"conv_dim": [
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512
|
| 20 |
+
],
|
| 21 |
+
"conv_kernel": [
|
| 22 |
+
10,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
3,
|
| 27 |
+
2,
|
| 28 |
+
2
|
| 29 |
+
],
|
| 30 |
+
"conv_stride": [
|
| 31 |
+
5,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2,
|
| 37 |
+
2
|
| 38 |
+
],
|
| 39 |
+
"ctc_loss_reduction": "sum",
|
| 40 |
+
"ctc_zero_infinity": false,
|
| 41 |
+
"do_stable_layer_norm": false,
|
| 42 |
+
"eos_token_id": 2,
|
| 43 |
+
"feat_extract_activation": "gelu",
|
| 44 |
+
"feat_extract_norm": "group",
|
| 45 |
+
"feat_proj_dropout": 0.0,
|
| 46 |
+
"feat_proj_layer_norm": false,
|
| 47 |
+
"final_dropout": 0.0,
|
| 48 |
+
"hidden_act": "gelu",
|
| 49 |
+
"hidden_dropout": 0.1,
|
| 50 |
+
"hidden_size": 768,
|
| 51 |
+
"id2label": {
|
| 52 |
+
"0": "NOT_WORD",
|
| 53 |
+
"1": "WORD"
|
| 54 |
+
},
|
| 55 |
+
"initializer_range": 0.02,
|
| 56 |
+
"intermediate_size": 3072,
|
| 57 |
+
"label2id": {
|
| 58 |
+
"NOT_WORD": "0",
|
| 59 |
+
"WORD": "1"
|
| 60 |
+
},
|
| 61 |
+
"layer_norm_eps": 1e-05,
|
| 62 |
+
"layerdrop": 0.0,
|
| 63 |
+
"mask_feature_length": 10,
|
| 64 |
+
"mask_feature_min_masks": 0,
|
| 65 |
+
"mask_feature_prob": 0.0,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.05,
|
| 69 |
+
"model_type": "hubert",
|
| 70 |
+
"num_attention_heads": 12,
|
| 71 |
+
"num_conv_pos_embedding_groups": 16,
|
| 72 |
+
"num_conv_pos_embeddings": 128,
|
| 73 |
+
"num_feat_extract_layers": 7,
|
| 74 |
+
"num_hidden_layers": 2,
|
| 75 |
+
"pad_token_id": 0,
|
| 76 |
+
"torch_dtype": "float32",
|
| 77 |
+
"transformers_version": "4.38.1",
|
| 78 |
+
"use_weighted_layer_sum": false,
|
| 79 |
+
"vocab_size": 32
|
| 80 |
+
}
|
run-14/checkpoint-96/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:970da1c4501d894071e1745ddd25b3433221a0df24b92676742e39b527c82fba
|
| 3 |
+
size 94763496
|
run-14/checkpoint-96/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:630d07de2c2d75949124972c6fcff497d75f5edc0096dc94b40290bcd0d6109c
|
| 3 |
+
size 189552570
|
run-14/checkpoint-96/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,
|
| 7 |
+
"return_attention_mask": true,
|
| 8 |
+
"sampling_rate": 16000
|
| 9 |
+
}
|
run-14/checkpoint-96/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:39dbf03bf644af79257aec95c925042cb81a469bfcc7a839a95d68f1d0425513
|
| 3 |
+
size 14244
|
run-14/checkpoint-96/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5a0a0fb57ed8a0d88d453ecc2fea8f8423b6a7ed1fd472dbd8806bc9412c8db3
|
| 3 |
+
size 1064
|
run-14/checkpoint-96/trainer_state.json
ADDED
|
@@ -0,0 +1,175 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.7326732673267327,
|
| 3 |
+
"best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-14/checkpoint-48",
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 96,
|
| 7 |
+
"is_hyper_param_search": true,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.1,
|
| 13 |
+
"grad_norm": 1.263960838317871,
|
| 14 |
+
"learning_rate": 4.556569949810922e-06,
|
| 15 |
+
"loss": 0.7021,
|
| 16 |
+
"step": 5
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.21,
|
| 20 |
+
"grad_norm": 1.1750551462173462,
|
| 21 |
+
"learning_rate": 9.113139899621844e-06,
|
| 22 |
+
"loss": 0.6857,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.31,
|
| 27 |
+
"grad_norm": 1.3693294525146484,
|
| 28 |
+
"learning_rate": 1.3669709849432766e-05,
|
| 29 |
+
"loss": 0.6849,
|
| 30 |
+
"step": 15
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.42,
|
| 34 |
+
"grad_norm": 0.9440478086471558,
|
| 35 |
+
"learning_rate": 1.822627979924369e-05,
|
| 36 |
+
"loss": 0.6652,
|
| 37 |
+
"step": 20
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.52,
|
| 41 |
+
"grad_norm": 0.9551452398300171,
|
| 42 |
+
"learning_rate": 2.2782849749054612e-05,
|
| 43 |
+
"loss": 0.64,
|
| 44 |
+
"step": 25
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.62,
|
| 48 |
+
"grad_norm": 0.777751088142395,
|
| 49 |
+
"learning_rate": 2.7339419698865533e-05,
|
| 50 |
+
"loss": 0.6269,
|
| 51 |
+
"step": 30
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.73,
|
| 55 |
+
"grad_norm": 2.4816713333129883,
|
| 56 |
+
"learning_rate": 3.1895989648676456e-05,
|
| 57 |
+
"loss": 0.5586,
|
| 58 |
+
"step": 35
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.83,
|
| 62 |
+
"grad_norm": 1.696999430656433,
|
| 63 |
+
"learning_rate": 3.645255959848738e-05,
|
| 64 |
+
"loss": 0.6586,
|
| 65 |
+
"step": 40
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.94,
|
| 69 |
+
"grad_norm": 3.2336232662200928,
|
| 70 |
+
"learning_rate": 4.10091295482983e-05,
|
| 71 |
+
"loss": 0.6259,
|
| 72 |
+
"step": 45
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.0,
|
| 76 |
+
"eval_f1": 0.7326732673267327,
|
| 77 |
+
"eval_loss": 0.7427406311035156,
|
| 78 |
+
"eval_runtime": 1.4051,
|
| 79 |
+
"eval_samples_per_second": 45.548,
|
| 80 |
+
"eval_steps_per_second": 5.694,
|
| 81 |
+
"step": 48
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.04,
|
| 85 |
+
"grad_norm": 4.544165134429932,
|
| 86 |
+
"learning_rate": 4.354055729819326e-05,
|
| 87 |
+
"loss": 0.7,
|
| 88 |
+
"step": 50
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.15,
|
| 92 |
+
"grad_norm": 0.8700355887413025,
|
| 93 |
+
"learning_rate": 4.3034271748214265e-05,
|
| 94 |
+
"loss": 0.6972,
|
| 95 |
+
"step": 55
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.25,
|
| 99 |
+
"grad_norm": 1.7657314538955688,
|
| 100 |
+
"learning_rate": 4.252798619823527e-05,
|
| 101 |
+
"loss": 0.4085,
|
| 102 |
+
"step": 60
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.35,
|
| 106 |
+
"grad_norm": 0.6718289852142334,
|
| 107 |
+
"learning_rate": 4.202170064825628e-05,
|
| 108 |
+
"loss": 0.5618,
|
| 109 |
+
"step": 65
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.46,
|
| 113 |
+
"grad_norm": 1.5080432891845703,
|
| 114 |
+
"learning_rate": 4.151541509827729e-05,
|
| 115 |
+
"loss": 0.4822,
|
| 116 |
+
"step": 70
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.56,
|
| 120 |
+
"grad_norm": Infinity,
|
| 121 |
+
"learning_rate": 4.11103866582941e-05,
|
| 122 |
+
"loss": 0.7679,
|
| 123 |
+
"step": 75
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.67,
|
| 127 |
+
"grad_norm": 1.8398224115371704,
|
| 128 |
+
"learning_rate": 4.0604101108315106e-05,
|
| 129 |
+
"loss": 0.6183,
|
| 130 |
+
"step": 80
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.77,
|
| 134 |
+
"grad_norm": 1.7079250812530518,
|
| 135 |
+
"learning_rate": 4.0097815558336114e-05,
|
| 136 |
+
"loss": 0.5954,
|
| 137 |
+
"step": 85
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.88,
|
| 141 |
+
"grad_norm": 1.8899829387664795,
|
| 142 |
+
"learning_rate": 3.959153000835712e-05,
|
| 143 |
+
"loss": 0.596,
|
| 144 |
+
"step": 90
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.98,
|
| 148 |
+
"grad_norm": 3.904428243637085,
|
| 149 |
+
"learning_rate": 3.908524445837813e-05,
|
| 150 |
+
"loss": 0.6094,
|
| 151 |
+
"step": 95
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.0,
|
| 155 |
+
"eval_f1": 0.7326732673267327,
|
| 156 |
+
"eval_loss": 0.673431396484375,
|
| 157 |
+
"eval_runtime": 1.3605,
|
| 158 |
+
"eval_samples_per_second": 47.041,
|
| 159 |
+
"eval_steps_per_second": 5.88,
|
| 160 |
+
"step": 96
|
| 161 |
+
}
|
| 162 |
+
],
|
| 163 |
+
"logging_steps": 5,
|
| 164 |
+
"max_steps": 480,
|
| 165 |
+
"num_input_tokens_seen": 0,
|
| 166 |
+
"num_train_epochs": 10,
|
| 167 |
+
"save_steps": 500,
|
| 168 |
+
"total_flos": 1442567462539200.0,
|
| 169 |
+
"train_batch_size": 4,
|
| 170 |
+
"trial_name": null,
|
| 171 |
+
"trial_params": {
|
| 172 |
+
"learning_rate": 4.374307151818485e-05,
|
| 173 |
+
"per_device_train_batch_size": 4
|
| 174 |
+
}
|
| 175 |
+
}
|
run-14/checkpoint-96/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:08323d3f888db4d96cf1275214c0bc1494a12c503c7492df7675d3c2976246ad
|
| 3 |
+
size 4920
|
run-15/checkpoint-144/config.json
ADDED
|
@@ -0,0 +1,80 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "ntu-spml/distilhubert",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"HubertForSequenceClassification"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.1,
|
| 9 |
+
"bos_token_id": 1,
|
| 10 |
+
"classifier_proj_size": 256,
|
| 11 |
+
"conv_bias": false,
|
| 12 |
+
"conv_dim": [
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512
|
| 20 |
+
],
|
| 21 |
+
"conv_kernel": [
|
| 22 |
+
10,
|
| 23 |
+
3,
|
| 24 |
+
3,
|
| 25 |
+
3,
|
| 26 |
+
3,
|
| 27 |
+
2,
|
| 28 |
+
2
|
| 29 |
+
],
|
| 30 |
+
"conv_stride": [
|
| 31 |
+
5,
|
| 32 |
+
2,
|
| 33 |
+
2,
|
| 34 |
+
2,
|
| 35 |
+
2,
|
| 36 |
+
2,
|
| 37 |
+
2
|
| 38 |
+
],
|
| 39 |
+
"ctc_loss_reduction": "sum",
|
| 40 |
+
"ctc_zero_infinity": false,
|
| 41 |
+
"do_stable_layer_norm": false,
|
| 42 |
+
"eos_token_id": 2,
|
| 43 |
+
"feat_extract_activation": "gelu",
|
| 44 |
+
"feat_extract_norm": "group",
|
| 45 |
+
"feat_proj_dropout": 0.0,
|
| 46 |
+
"feat_proj_layer_norm": false,
|
| 47 |
+
"final_dropout": 0.0,
|
| 48 |
+
"hidden_act": "gelu",
|
| 49 |
+
"hidden_dropout": 0.1,
|
| 50 |
+
"hidden_size": 768,
|
| 51 |
+
"id2label": {
|
| 52 |
+
"0": "NOT_WORD",
|
| 53 |
+
"1": "WORD"
|
| 54 |
+
},
|
| 55 |
+
"initializer_range": 0.02,
|
| 56 |
+
"intermediate_size": 3072,
|
| 57 |
+
"label2id": {
|
| 58 |
+
"NOT_WORD": "0",
|
| 59 |
+
"WORD": "1"
|
| 60 |
+
},
|
| 61 |
+
"layer_norm_eps": 1e-05,
|
| 62 |
+
"layerdrop": 0.0,
|
| 63 |
+
"mask_feature_length": 10,
|
| 64 |
+
"mask_feature_min_masks": 0,
|
| 65 |
+
"mask_feature_prob": 0.0,
|
| 66 |
+
"mask_time_length": 10,
|
| 67 |
+
"mask_time_min_masks": 2,
|
| 68 |
+
"mask_time_prob": 0.05,
|
| 69 |
+
"model_type": "hubert",
|
| 70 |
+
"num_attention_heads": 12,
|
| 71 |
+
"num_conv_pos_embedding_groups": 16,
|
| 72 |
+
"num_conv_pos_embeddings": 128,
|
| 73 |
+
"num_feat_extract_layers": 7,
|
| 74 |
+
"num_hidden_layers": 2,
|
| 75 |
+
"pad_token_id": 0,
|
| 76 |
+
"torch_dtype": "float32",
|
| 77 |
+
"transformers_version": "4.38.1",
|
| 78 |
+
"use_weighted_layer_sum": false,
|
| 79 |
+
"vocab_size": 32
|
| 80 |
+
}
|