Retrained binary detector with corrected focal loss alpha weighting
Browse files- checkpoint-200/config.json +54 -0
- checkpoint-200/model.safetensors +3 -0
- checkpoint-200/optimizer.pt +3 -0
- checkpoint-200/rng_state.pth +3 -0
- checkpoint-200/scaler.pt +3 -0
- checkpoint-200/scheduler.pt +3 -0
- checkpoint-200/trainer_state.json +139 -0
- checkpoint-200/training_args.bin +3 -0
- checkpoint-400/config.json +54 -0
- checkpoint-400/model.safetensors +3 -0
- checkpoint-400/optimizer.pt +3 -0
- checkpoint-400/rng_state.pth +3 -0
- checkpoint-400/scaler.pt +3 -0
- checkpoint-400/scheduler.pt +3 -0
- checkpoint-400/trainer_state.json +235 -0
- checkpoint-400/training_args.bin +3 -0
- checkpoint-600/config.json +54 -0
- checkpoint-600/model.safetensors +3 -0
- checkpoint-600/optimizer.pt +3 -0
- checkpoint-600/rng_state.pth +3 -0
- checkpoint-600/scaler.pt +3 -0
- checkpoint-600/scheduler.pt +3 -0
- checkpoint-600/trainer_state.json +331 -0
- checkpoint-600/training_args.bin +3 -0
- model.safetensors +3 -0
- test_results.json +8 -8
- training_args.bin +2 -2
checkpoint-200/config.json
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"ModernBertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 50281,
|
| 8 |
+
"classifier_activation": "gelu",
|
| 9 |
+
"classifier_bias": false,
|
| 10 |
+
"classifier_dropout": 0.0,
|
| 11 |
+
"classifier_pooling": "mean",
|
| 12 |
+
"cls_token_id": 50281,
|
| 13 |
+
"decoder_bias": true,
|
| 14 |
+
"deterministic_flash_attn": false,
|
| 15 |
+
"dtype": "float32",
|
| 16 |
+
"embedding_dropout": 0.0,
|
| 17 |
+
"eos_token_id": 50282,
|
| 18 |
+
"global_attn_every_n_layers": 3,
|
| 19 |
+
"global_rope_theta": 160000.0,
|
| 20 |
+
"gradient_checkpointing": false,
|
| 21 |
+
"hidden_activation": "gelu",
|
| 22 |
+
"hidden_size": 768,
|
| 23 |
+
"id2label": {
|
| 24 |
+
"0": "no_propaganda",
|
| 25 |
+
"1": "has_propaganda"
|
| 26 |
+
},
|
| 27 |
+
"initializer_cutoff_factor": 2.0,
|
| 28 |
+
"initializer_range": 0.02,
|
| 29 |
+
"intermediate_size": 1152,
|
| 30 |
+
"label2id": {
|
| 31 |
+
"has_propaganda": 1,
|
| 32 |
+
"no_propaganda": 0
|
| 33 |
+
},
|
| 34 |
+
"layer_norm_eps": 1e-05,
|
| 35 |
+
"local_attention": 128,
|
| 36 |
+
"local_rope_theta": 10000.0,
|
| 37 |
+
"max_position_embeddings": 8192,
|
| 38 |
+
"mlp_bias": false,
|
| 39 |
+
"mlp_dropout": 0.0,
|
| 40 |
+
"model_type": "modernbert",
|
| 41 |
+
"norm_bias": false,
|
| 42 |
+
"norm_eps": 1e-05,
|
| 43 |
+
"num_attention_heads": 12,
|
| 44 |
+
"num_hidden_layers": 22,
|
| 45 |
+
"pad_token_id": 50283,
|
| 46 |
+
"position_embedding_type": "absolute",
|
| 47 |
+
"problem_type": "single_label_classification",
|
| 48 |
+
"repad_logits_with_grad": false,
|
| 49 |
+
"sep_token_id": 50282,
|
| 50 |
+
"sparse_pred_ignore_index": -100,
|
| 51 |
+
"sparse_prediction": false,
|
| 52 |
+
"transformers_version": "4.57.3",
|
| 53 |
+
"vocab_size": 50368
|
| 54 |
+
}
|
checkpoint-200/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b2aae000f6b136f5ba37b84e00f9f4b3849f2c3ebbb9f15316a1658b785ec20c
|
| 3 |
+
size 598439784
|
checkpoint-200/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2de83188c578a2771dba49cb1868183db64f2ccfcf87a8cfabbaa5259771db3c
|
| 3 |
+
size 1196969803
|
checkpoint-200/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4a9f217e852f439efa6bd32fde98d6867f11aa6ea13ddc021ba10af6a0b0934
|
| 3 |
+
size 14645
|
checkpoint-200/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:17a0d72ed9080212eb9b54c712205e4caf5fd660e6ac2976551fc7e5c10f31bb
|
| 3 |
+
size 1383
|
checkpoint-200/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8fc52dc9b611603860209b456c0a27c6d5387c43752f103e88a490557670867
|
| 3 |
+
size 1465
|
checkpoint-200/trainer_state.json
ADDED
|
@@ -0,0 +1,139 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 200,
|
| 3 |
+
"best_metric": 0.9974785678265254,
|
| 4 |
+
"best_model_checkpoint": "nci-binary-detector/checkpoint-200",
|
| 5 |
+
"epoch": 0.32679738562091504,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
+
"global_step": 200,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.032679738562091505,
|
| 14 |
+
"grad_norm": 2.849823236465454,
|
| 15 |
+
"learning_rate": 1.2418300653594772e-06,
|
| 16 |
+
"loss": 0.1789,
|
| 17 |
+
"step": 20
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.06535947712418301,
|
| 21 |
+
"grad_norm": 1.1499906778335571,
|
| 22 |
+
"learning_rate": 2.549019607843137e-06,
|
| 23 |
+
"loss": 0.1045,
|
| 24 |
+
"step": 40
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.09803921568627451,
|
| 28 |
+
"grad_norm": 1.5816751718521118,
|
| 29 |
+
"learning_rate": 3.856209150326798e-06,
|
| 30 |
+
"loss": 0.0389,
|
| 31 |
+
"step": 60
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.13071895424836602,
|
| 35 |
+
"grad_norm": 0.3664398789405823,
|
| 36 |
+
"learning_rate": 5.163398692810458e-06,
|
| 37 |
+
"loss": 0.0164,
|
| 38 |
+
"step": 80
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.16339869281045752,
|
| 42 |
+
"grad_norm": 0.15663352608680725,
|
| 43 |
+
"learning_rate": 6.470588235294119e-06,
|
| 44 |
+
"loss": 0.0123,
|
| 45 |
+
"step": 100
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.16339869281045752,
|
| 49 |
+
"eval_accuracy": 0.9837869137232195,
|
| 50 |
+
"eval_f1": 0.9860279441117764,
|
| 51 |
+
"eval_loss": 0.005597150418907404,
|
| 52 |
+
"eval_precision": 0.9743589743589743,
|
| 53 |
+
"eval_recall": 0.997979797979798,
|
| 54 |
+
"eval_roc_auc": 0.9985759905705631,
|
| 55 |
+
"eval_runtime": 11.5844,
|
| 56 |
+
"eval_samples_per_second": 149.08,
|
| 57 |
+
"eval_steps_per_second": 4.661,
|
| 58 |
+
"step": 100
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.19607843137254902,
|
| 62 |
+
"grad_norm": 0.10866066068410873,
|
| 63 |
+
"learning_rate": 7.77777777777778e-06,
|
| 64 |
+
"loss": 0.0073,
|
| 65 |
+
"step": 120
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.22875816993464052,
|
| 69 |
+
"grad_norm": 0.06542709469795227,
|
| 70 |
+
"learning_rate": 9.084967320261438e-06,
|
| 71 |
+
"loss": 0.0019,
|
| 72 |
+
"step": 140
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.26143790849673204,
|
| 76 |
+
"grad_norm": 0.15572959184646606,
|
| 77 |
+
"learning_rate": 1.03921568627451e-05,
|
| 78 |
+
"loss": 0.0014,
|
| 79 |
+
"step": 160
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.29411764705882354,
|
| 83 |
+
"grad_norm": 0.009457970038056374,
|
| 84 |
+
"learning_rate": 1.169934640522876e-05,
|
| 85 |
+
"loss": 0.0066,
|
| 86 |
+
"step": 180
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.32679738562091504,
|
| 90 |
+
"grad_norm": 0.016104549169540405,
|
| 91 |
+
"learning_rate": 1.3006535947712419e-05,
|
| 92 |
+
"loss": 0.0007,
|
| 93 |
+
"step": 200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.32679738562091504,
|
| 97 |
+
"eval_accuracy": 0.9971048060220035,
|
| 98 |
+
"eval_f1": 0.9974785678265254,
|
| 99 |
+
"eval_loss": 0.0025405457708984613,
|
| 100 |
+
"eval_precision": 0.9959718026183283,
|
| 101 |
+
"eval_recall": 0.998989898989899,
|
| 102 |
+
"eval_roc_auc": 0.998836396529748,
|
| 103 |
+
"eval_runtime": 8.7693,
|
| 104 |
+
"eval_samples_per_second": 196.936,
|
| 105 |
+
"eval_steps_per_second": 6.158,
|
| 106 |
+
"step": 200
|
| 107 |
+
}
|
| 108 |
+
],
|
| 109 |
+
"logging_steps": 20,
|
| 110 |
+
"max_steps": 3060,
|
| 111 |
+
"num_input_tokens_seen": 0,
|
| 112 |
+
"num_train_epochs": 5,
|
| 113 |
+
"save_steps": 200,
|
| 114 |
+
"stateful_callbacks": {
|
| 115 |
+
"EarlyStoppingCallback": {
|
| 116 |
+
"args": {
|
| 117 |
+
"early_stopping_patience": 3,
|
| 118 |
+
"early_stopping_threshold": 0.0
|
| 119 |
+
},
|
| 120 |
+
"attributes": {
|
| 121 |
+
"early_stopping_patience_counter": 0
|
| 122 |
+
}
|
| 123 |
+
},
|
| 124 |
+
"TrainerControl": {
|
| 125 |
+
"args": {
|
| 126 |
+
"should_epoch_stop": false,
|
| 127 |
+
"should_evaluate": false,
|
| 128 |
+
"should_log": false,
|
| 129 |
+
"should_save": true,
|
| 130 |
+
"should_training_stop": false
|
| 131 |
+
},
|
| 132 |
+
"attributes": {}
|
| 133 |
+
}
|
| 134 |
+
},
|
| 135 |
+
"total_flos": 2180850214502400.0,
|
| 136 |
+
"train_batch_size": 16,
|
| 137 |
+
"trial_name": null,
|
| 138 |
+
"trial_params": null
|
| 139 |
+
}
|
checkpoint-200/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c1be75a7bbc1e87d42052a69bee01f6eaf1e6586d629ebf02cacfdf4eb951e52
|
| 3 |
+
size 5841
|
checkpoint-400/config.json
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"ModernBertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 50281,
|
| 8 |
+
"classifier_activation": "gelu",
|
| 9 |
+
"classifier_bias": false,
|
| 10 |
+
"classifier_dropout": 0.0,
|
| 11 |
+
"classifier_pooling": "mean",
|
| 12 |
+
"cls_token_id": 50281,
|
| 13 |
+
"decoder_bias": true,
|
| 14 |
+
"deterministic_flash_attn": false,
|
| 15 |
+
"dtype": "float32",
|
| 16 |
+
"embedding_dropout": 0.0,
|
| 17 |
+
"eos_token_id": 50282,
|
| 18 |
+
"global_attn_every_n_layers": 3,
|
| 19 |
+
"global_rope_theta": 160000.0,
|
| 20 |
+
"gradient_checkpointing": false,
|
| 21 |
+
"hidden_activation": "gelu",
|
| 22 |
+
"hidden_size": 768,
|
| 23 |
+
"id2label": {
|
| 24 |
+
"0": "no_propaganda",
|
| 25 |
+
"1": "has_propaganda"
|
| 26 |
+
},
|
| 27 |
+
"initializer_cutoff_factor": 2.0,
|
| 28 |
+
"initializer_range": 0.02,
|
| 29 |
+
"intermediate_size": 1152,
|
| 30 |
+
"label2id": {
|
| 31 |
+
"has_propaganda": 1,
|
| 32 |
+
"no_propaganda": 0
|
| 33 |
+
},
|
| 34 |
+
"layer_norm_eps": 1e-05,
|
| 35 |
+
"local_attention": 128,
|
| 36 |
+
"local_rope_theta": 10000.0,
|
| 37 |
+
"max_position_embeddings": 8192,
|
| 38 |
+
"mlp_bias": false,
|
| 39 |
+
"mlp_dropout": 0.0,
|
| 40 |
+
"model_type": "modernbert",
|
| 41 |
+
"norm_bias": false,
|
| 42 |
+
"norm_eps": 1e-05,
|
| 43 |
+
"num_attention_heads": 12,
|
| 44 |
+
"num_hidden_layers": 22,
|
| 45 |
+
"pad_token_id": 50283,
|
| 46 |
+
"position_embedding_type": "absolute",
|
| 47 |
+
"problem_type": "single_label_classification",
|
| 48 |
+
"repad_logits_with_grad": false,
|
| 49 |
+
"sep_token_id": 50282,
|
| 50 |
+
"sparse_pred_ignore_index": -100,
|
| 51 |
+
"sparse_prediction": false,
|
| 52 |
+
"transformers_version": "4.57.3",
|
| 53 |
+
"vocab_size": 50368
|
| 54 |
+
}
|
checkpoint-400/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:606ea7da887dcedca22a8e326e47f5e00602fd243a036853988d022eb7e406c3
|
| 3 |
+
size 598439784
|
checkpoint-400/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:be1ab97a46359294ec6b5ffff2c8cad6f8fdc87bbf4fdc243126dddaa578a945
|
| 3 |
+
size 1196969803
|
checkpoint-400/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:718a0f3db00824213036a2c0441849791319b7d9cf189065873bb26a7020738e
|
| 3 |
+
size 14645
|
checkpoint-400/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:00a368419dfa117317e118dc1cc99dd02f7f47d2c85fd92b9bc06654abbfb3ca
|
| 3 |
+
size 1383
|
checkpoint-400/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7934a8cae2282d33f468182cc96f87a5e2a1baf0ca87868dd26a10b5ddb95cf6
|
| 3 |
+
size 1465
|
checkpoint-400/trainer_state.json
ADDED
|
@@ -0,0 +1,235 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 300,
|
| 3 |
+
"best_metric": 0.997979797979798,
|
| 4 |
+
"best_model_checkpoint": "nci-binary-detector/checkpoint-200",
|
| 5 |
+
"epoch": 0.6535947712418301,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
+
"global_step": 400,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.032679738562091505,
|
| 14 |
+
"grad_norm": 2.849823236465454,
|
| 15 |
+
"learning_rate": 1.2418300653594772e-06,
|
| 16 |
+
"loss": 0.1789,
|
| 17 |
+
"step": 20
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.06535947712418301,
|
| 21 |
+
"grad_norm": 1.1499906778335571,
|
| 22 |
+
"learning_rate": 2.549019607843137e-06,
|
| 23 |
+
"loss": 0.1045,
|
| 24 |
+
"step": 40
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.09803921568627451,
|
| 28 |
+
"grad_norm": 1.5816751718521118,
|
| 29 |
+
"learning_rate": 3.856209150326798e-06,
|
| 30 |
+
"loss": 0.0389,
|
| 31 |
+
"step": 60
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.13071895424836602,
|
| 35 |
+
"grad_norm": 0.3664398789405823,
|
| 36 |
+
"learning_rate": 5.163398692810458e-06,
|
| 37 |
+
"loss": 0.0164,
|
| 38 |
+
"step": 80
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.16339869281045752,
|
| 42 |
+
"grad_norm": 0.15663352608680725,
|
| 43 |
+
"learning_rate": 6.470588235294119e-06,
|
| 44 |
+
"loss": 0.0123,
|
| 45 |
+
"step": 100
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.16339869281045752,
|
| 49 |
+
"eval_accuracy": 0.9837869137232195,
|
| 50 |
+
"eval_f1": 0.9860279441117764,
|
| 51 |
+
"eval_loss": 0.005597150418907404,
|
| 52 |
+
"eval_precision": 0.9743589743589743,
|
| 53 |
+
"eval_recall": 0.997979797979798,
|
| 54 |
+
"eval_roc_auc": 0.9985759905705631,
|
| 55 |
+
"eval_runtime": 11.5844,
|
| 56 |
+
"eval_samples_per_second": 149.08,
|
| 57 |
+
"eval_steps_per_second": 4.661,
|
| 58 |
+
"step": 100
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.19607843137254902,
|
| 62 |
+
"grad_norm": 0.10866066068410873,
|
| 63 |
+
"learning_rate": 7.77777777777778e-06,
|
| 64 |
+
"loss": 0.0073,
|
| 65 |
+
"step": 120
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.22875816993464052,
|
| 69 |
+
"grad_norm": 0.06542709469795227,
|
| 70 |
+
"learning_rate": 9.084967320261438e-06,
|
| 71 |
+
"loss": 0.0019,
|
| 72 |
+
"step": 140
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.26143790849673204,
|
| 76 |
+
"grad_norm": 0.15572959184646606,
|
| 77 |
+
"learning_rate": 1.03921568627451e-05,
|
| 78 |
+
"loss": 0.0014,
|
| 79 |
+
"step": 160
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.29411764705882354,
|
| 83 |
+
"grad_norm": 0.009457970038056374,
|
| 84 |
+
"learning_rate": 1.169934640522876e-05,
|
| 85 |
+
"loss": 0.0066,
|
| 86 |
+
"step": 180
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.32679738562091504,
|
| 90 |
+
"grad_norm": 0.016104549169540405,
|
| 91 |
+
"learning_rate": 1.3006535947712419e-05,
|
| 92 |
+
"loss": 0.0007,
|
| 93 |
+
"step": 200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.32679738562091504,
|
| 97 |
+
"eval_accuracy": 0.9971048060220035,
|
| 98 |
+
"eval_f1": 0.9974785678265254,
|
| 99 |
+
"eval_loss": 0.0025405457708984613,
|
| 100 |
+
"eval_precision": 0.9959718026183283,
|
| 101 |
+
"eval_recall": 0.998989898989899,
|
| 102 |
+
"eval_roc_auc": 0.998836396529748,
|
| 103 |
+
"eval_runtime": 8.7693,
|
| 104 |
+
"eval_samples_per_second": 196.936,
|
| 105 |
+
"eval_steps_per_second": 6.158,
|
| 106 |
+
"step": 200
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.35947712418300654,
|
| 110 |
+
"grad_norm": 0.10450628399848938,
|
| 111 |
+
"learning_rate": 1.431372549019608e-05,
|
| 112 |
+
"loss": 0.0166,
|
| 113 |
+
"step": 220
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.39215686274509803,
|
| 117 |
+
"grad_norm": 0.11410112679004669,
|
| 118 |
+
"learning_rate": 1.562091503267974e-05,
|
| 119 |
+
"loss": 0.001,
|
| 120 |
+
"step": 240
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 0.42483660130718953,
|
| 124 |
+
"grad_norm": 0.056244976818561554,
|
| 125 |
+
"learning_rate": 1.69281045751634e-05,
|
| 126 |
+
"loss": 0.0044,
|
| 127 |
+
"step": 260
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 0.45751633986928103,
|
| 131 |
+
"grad_norm": 0.044668033719062805,
|
| 132 |
+
"learning_rate": 1.823529411764706e-05,
|
| 133 |
+
"loss": 0.0019,
|
| 134 |
+
"step": 280
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.49019607843137253,
|
| 138 |
+
"grad_norm": 0.004644064232707024,
|
| 139 |
+
"learning_rate": 1.954248366013072e-05,
|
| 140 |
+
"loss": 0.0004,
|
| 141 |
+
"step": 300
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.49019607843137253,
|
| 145 |
+
"eval_accuracy": 0.9976838448176027,
|
| 146 |
+
"eval_f1": 0.997979797979798,
|
| 147 |
+
"eval_loss": 0.002409732900559902,
|
| 148 |
+
"eval_precision": 0.997979797979798,
|
| 149 |
+
"eval_recall": 0.997979797979798,
|
| 150 |
+
"eval_roc_auc": 0.9996368022148212,
|
| 151 |
+
"eval_runtime": 8.7756,
|
| 152 |
+
"eval_samples_per_second": 196.796,
|
| 153 |
+
"eval_steps_per_second": 6.153,
|
| 154 |
+
"step": 300
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.5228758169934641,
|
| 158 |
+
"grad_norm": 0.31552189588546753,
|
| 159 |
+
"learning_rate": 1.990559186637618e-05,
|
| 160 |
+
"loss": 0.0033,
|
| 161 |
+
"step": 320
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.5555555555555556,
|
| 165 |
+
"grad_norm": 0.0014637637650594115,
|
| 166 |
+
"learning_rate": 1.9760348583877997e-05,
|
| 167 |
+
"loss": 0.0002,
|
| 168 |
+
"step": 340
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 0.5882352941176471,
|
| 172 |
+
"grad_norm": 2.498281955718994,
|
| 173 |
+
"learning_rate": 1.9615105301379813e-05,
|
| 174 |
+
"loss": 0.0008,
|
| 175 |
+
"step": 360
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.6209150326797386,
|
| 179 |
+
"grad_norm": 1.514452014816925e-05,
|
| 180 |
+
"learning_rate": 1.946986201888163e-05,
|
| 181 |
+
"loss": 0.0001,
|
| 182 |
+
"step": 380
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.6535947712418301,
|
| 186 |
+
"grad_norm": 0.016483725979924202,
|
| 187 |
+
"learning_rate": 1.9324618736383445e-05,
|
| 188 |
+
"loss": 0.0046,
|
| 189 |
+
"step": 400
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.6535947712418301,
|
| 193 |
+
"eval_accuracy": 0.9936305732484076,
|
| 194 |
+
"eval_f1": 0.994475138121547,
|
| 195 |
+
"eval_loss": 0.0015705447876825929,
|
| 196 |
+
"eval_precision": 0.989010989010989,
|
| 197 |
+
"eval_recall": 1.0,
|
| 198 |
+
"eval_roc_auc": 0.9999013198470458,
|
| 199 |
+
"eval_runtime": 8.774,
|
| 200 |
+
"eval_samples_per_second": 196.831,
|
| 201 |
+
"eval_steps_per_second": 6.155,
|
| 202 |
+
"step": 400
|
| 203 |
+
}
|
| 204 |
+
],
|
| 205 |
+
"logging_steps": 20,
|
| 206 |
+
"max_steps": 3060,
|
| 207 |
+
"num_input_tokens_seen": 0,
|
| 208 |
+
"num_train_epochs": 5,
|
| 209 |
+
"save_steps": 200,
|
| 210 |
+
"stateful_callbacks": {
|
| 211 |
+
"EarlyStoppingCallback": {
|
| 212 |
+
"args": {
|
| 213 |
+
"early_stopping_patience": 3,
|
| 214 |
+
"early_stopping_threshold": 0.0
|
| 215 |
+
},
|
| 216 |
+
"attributes": {
|
| 217 |
+
"early_stopping_patience_counter": 1
|
| 218 |
+
}
|
| 219 |
+
},
|
| 220 |
+
"TrainerControl": {
|
| 221 |
+
"args": {
|
| 222 |
+
"should_epoch_stop": false,
|
| 223 |
+
"should_evaluate": false,
|
| 224 |
+
"should_log": false,
|
| 225 |
+
"should_save": true,
|
| 226 |
+
"should_training_stop": false
|
| 227 |
+
},
|
| 228 |
+
"attributes": {}
|
| 229 |
+
}
|
| 230 |
+
},
|
| 231 |
+
"total_flos": 4361700429004800.0,
|
| 232 |
+
"train_batch_size": 16,
|
| 233 |
+
"trial_name": null,
|
| 234 |
+
"trial_params": null
|
| 235 |
+
}
|
checkpoint-400/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c1be75a7bbc1e87d42052a69bee01f6eaf1e6586d629ebf02cacfdf4eb951e52
|
| 3 |
+
size 5841
|
checkpoint-600/config.json
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"ModernBertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 50281,
|
| 8 |
+
"classifier_activation": "gelu",
|
| 9 |
+
"classifier_bias": false,
|
| 10 |
+
"classifier_dropout": 0.0,
|
| 11 |
+
"classifier_pooling": "mean",
|
| 12 |
+
"cls_token_id": 50281,
|
| 13 |
+
"decoder_bias": true,
|
| 14 |
+
"deterministic_flash_attn": false,
|
| 15 |
+
"dtype": "float32",
|
| 16 |
+
"embedding_dropout": 0.0,
|
| 17 |
+
"eos_token_id": 50282,
|
| 18 |
+
"global_attn_every_n_layers": 3,
|
| 19 |
+
"global_rope_theta": 160000.0,
|
| 20 |
+
"gradient_checkpointing": false,
|
| 21 |
+
"hidden_activation": "gelu",
|
| 22 |
+
"hidden_size": 768,
|
| 23 |
+
"id2label": {
|
| 24 |
+
"0": "no_propaganda",
|
| 25 |
+
"1": "has_propaganda"
|
| 26 |
+
},
|
| 27 |
+
"initializer_cutoff_factor": 2.0,
|
| 28 |
+
"initializer_range": 0.02,
|
| 29 |
+
"intermediate_size": 1152,
|
| 30 |
+
"label2id": {
|
| 31 |
+
"has_propaganda": 1,
|
| 32 |
+
"no_propaganda": 0
|
| 33 |
+
},
|
| 34 |
+
"layer_norm_eps": 1e-05,
|
| 35 |
+
"local_attention": 128,
|
| 36 |
+
"local_rope_theta": 10000.0,
|
| 37 |
+
"max_position_embeddings": 8192,
|
| 38 |
+
"mlp_bias": false,
|
| 39 |
+
"mlp_dropout": 0.0,
|
| 40 |
+
"model_type": "modernbert",
|
| 41 |
+
"norm_bias": false,
|
| 42 |
+
"norm_eps": 1e-05,
|
| 43 |
+
"num_attention_heads": 12,
|
| 44 |
+
"num_hidden_layers": 22,
|
| 45 |
+
"pad_token_id": 50283,
|
| 46 |
+
"position_embedding_type": "absolute",
|
| 47 |
+
"problem_type": "single_label_classification",
|
| 48 |
+
"repad_logits_with_grad": false,
|
| 49 |
+
"sep_token_id": 50282,
|
| 50 |
+
"sparse_pred_ignore_index": -100,
|
| 51 |
+
"sparse_prediction": false,
|
| 52 |
+
"transformers_version": "4.57.3",
|
| 53 |
+
"vocab_size": 50368
|
| 54 |
+
}
|
checkpoint-600/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f133a493cb63a16ccbeb38b71d01caf73ac23fc445fb95601f4f5dbc988cd9da
|
| 3 |
+
size 598439784
|
checkpoint-600/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:08e19cbc488fe04403301c8fdd20a93d2aca52c250300e6def624c9a87c2f965
|
| 3 |
+
size 1196969803
|
checkpoint-600/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f196323d7423b60f8e4ceb7dbf8715ee326c0d068e5ff164f13c63b279b9f1a0
|
| 3 |
+
size 14645
|
checkpoint-600/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:78a1d1a5c8df0ff87f3d9a923fa98aa9b5f51a25e6518dd0f5b32d02e9a66dfd
|
| 3 |
+
size 1383
|
checkpoint-600/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7d9ed1fd0d4f00acf322bcd65b9233aaeae538bd83fe7c42e86772af6f795983
|
| 3 |
+
size 1465
|
checkpoint-600/trainer_state.json
ADDED
|
@@ -0,0 +1,331 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 300,
|
| 3 |
+
"best_metric": 0.997979797979798,
|
| 4 |
+
"best_model_checkpoint": "nci-binary-detector/checkpoint-200",
|
| 5 |
+
"epoch": 0.9803921568627451,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
+
"global_step": 600,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.032679738562091505,
|
| 14 |
+
"grad_norm": 2.849823236465454,
|
| 15 |
+
"learning_rate": 1.2418300653594772e-06,
|
| 16 |
+
"loss": 0.1789,
|
| 17 |
+
"step": 20
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.06535947712418301,
|
| 21 |
+
"grad_norm": 1.1499906778335571,
|
| 22 |
+
"learning_rate": 2.549019607843137e-06,
|
| 23 |
+
"loss": 0.1045,
|
| 24 |
+
"step": 40
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.09803921568627451,
|
| 28 |
+
"grad_norm": 1.5816751718521118,
|
| 29 |
+
"learning_rate": 3.856209150326798e-06,
|
| 30 |
+
"loss": 0.0389,
|
| 31 |
+
"step": 60
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.13071895424836602,
|
| 35 |
+
"grad_norm": 0.3664398789405823,
|
| 36 |
+
"learning_rate": 5.163398692810458e-06,
|
| 37 |
+
"loss": 0.0164,
|
| 38 |
+
"step": 80
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.16339869281045752,
|
| 42 |
+
"grad_norm": 0.15663352608680725,
|
| 43 |
+
"learning_rate": 6.470588235294119e-06,
|
| 44 |
+
"loss": 0.0123,
|
| 45 |
+
"step": 100
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.16339869281045752,
|
| 49 |
+
"eval_accuracy": 0.9837869137232195,
|
| 50 |
+
"eval_f1": 0.9860279441117764,
|
| 51 |
+
"eval_loss": 0.005597150418907404,
|
| 52 |
+
"eval_precision": 0.9743589743589743,
|
| 53 |
+
"eval_recall": 0.997979797979798,
|
| 54 |
+
"eval_roc_auc": 0.9985759905705631,
|
| 55 |
+
"eval_runtime": 11.5844,
|
| 56 |
+
"eval_samples_per_second": 149.08,
|
| 57 |
+
"eval_steps_per_second": 4.661,
|
| 58 |
+
"step": 100
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.19607843137254902,
|
| 62 |
+
"grad_norm": 0.10866066068410873,
|
| 63 |
+
"learning_rate": 7.77777777777778e-06,
|
| 64 |
+
"loss": 0.0073,
|
| 65 |
+
"step": 120
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.22875816993464052,
|
| 69 |
+
"grad_norm": 0.06542709469795227,
|
| 70 |
+
"learning_rate": 9.084967320261438e-06,
|
| 71 |
+
"loss": 0.0019,
|
| 72 |
+
"step": 140
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.26143790849673204,
|
| 76 |
+
"grad_norm": 0.15572959184646606,
|
| 77 |
+
"learning_rate": 1.03921568627451e-05,
|
| 78 |
+
"loss": 0.0014,
|
| 79 |
+
"step": 160
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.29411764705882354,
|
| 83 |
+
"grad_norm": 0.009457970038056374,
|
| 84 |
+
"learning_rate": 1.169934640522876e-05,
|
| 85 |
+
"loss": 0.0066,
|
| 86 |
+
"step": 180
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.32679738562091504,
|
| 90 |
+
"grad_norm": 0.016104549169540405,
|
| 91 |
+
"learning_rate": 1.3006535947712419e-05,
|
| 92 |
+
"loss": 0.0007,
|
| 93 |
+
"step": 200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.32679738562091504,
|
| 97 |
+
"eval_accuracy": 0.9971048060220035,
|
| 98 |
+
"eval_f1": 0.9974785678265254,
|
| 99 |
+
"eval_loss": 0.0025405457708984613,
|
| 100 |
+
"eval_precision": 0.9959718026183283,
|
| 101 |
+
"eval_recall": 0.998989898989899,
|
| 102 |
+
"eval_roc_auc": 0.998836396529748,
|
| 103 |
+
"eval_runtime": 8.7693,
|
| 104 |
+
"eval_samples_per_second": 196.936,
|
| 105 |
+
"eval_steps_per_second": 6.158,
|
| 106 |
+
"step": 200
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.35947712418300654,
|
| 110 |
+
"grad_norm": 0.10450628399848938,
|
| 111 |
+
"learning_rate": 1.431372549019608e-05,
|
| 112 |
+
"loss": 0.0166,
|
| 113 |
+
"step": 220
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.39215686274509803,
|
| 117 |
+
"grad_norm": 0.11410112679004669,
|
| 118 |
+
"learning_rate": 1.562091503267974e-05,
|
| 119 |
+
"loss": 0.001,
|
| 120 |
+
"step": 240
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 0.42483660130718953,
|
| 124 |
+
"grad_norm": 0.056244976818561554,
|
| 125 |
+
"learning_rate": 1.69281045751634e-05,
|
| 126 |
+
"loss": 0.0044,
|
| 127 |
+
"step": 260
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 0.45751633986928103,
|
| 131 |
+
"grad_norm": 0.044668033719062805,
|
| 132 |
+
"learning_rate": 1.823529411764706e-05,
|
| 133 |
+
"loss": 0.0019,
|
| 134 |
+
"step": 280
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.49019607843137253,
|
| 138 |
+
"grad_norm": 0.004644064232707024,
|
| 139 |
+
"learning_rate": 1.954248366013072e-05,
|
| 140 |
+
"loss": 0.0004,
|
| 141 |
+
"step": 300
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.49019607843137253,
|
| 145 |
+
"eval_accuracy": 0.9976838448176027,
|
| 146 |
+
"eval_f1": 0.997979797979798,
|
| 147 |
+
"eval_loss": 0.002409732900559902,
|
| 148 |
+
"eval_precision": 0.997979797979798,
|
| 149 |
+
"eval_recall": 0.997979797979798,
|
| 150 |
+
"eval_roc_auc": 0.9996368022148212,
|
| 151 |
+
"eval_runtime": 8.7756,
|
| 152 |
+
"eval_samples_per_second": 196.796,
|
| 153 |
+
"eval_steps_per_second": 6.153,
|
| 154 |
+
"step": 300
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.5228758169934641,
|
| 158 |
+
"grad_norm": 0.31552189588546753,
|
| 159 |
+
"learning_rate": 1.990559186637618e-05,
|
| 160 |
+
"loss": 0.0033,
|
| 161 |
+
"step": 320
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.5555555555555556,
|
| 165 |
+
"grad_norm": 0.0014637637650594115,
|
| 166 |
+
"learning_rate": 1.9760348583877997e-05,
|
| 167 |
+
"loss": 0.0002,
|
| 168 |
+
"step": 340
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 0.5882352941176471,
|
| 172 |
+
"grad_norm": 2.498281955718994,
|
| 173 |
+
"learning_rate": 1.9615105301379813e-05,
|
| 174 |
+
"loss": 0.0008,
|
| 175 |
+
"step": 360
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.6209150326797386,
|
| 179 |
+
"grad_norm": 1.514452014816925e-05,
|
| 180 |
+
"learning_rate": 1.946986201888163e-05,
|
| 181 |
+
"loss": 0.0001,
|
| 182 |
+
"step": 380
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.6535947712418301,
|
| 186 |
+
"grad_norm": 0.016483725979924202,
|
| 187 |
+
"learning_rate": 1.9324618736383445e-05,
|
| 188 |
+
"loss": 0.0046,
|
| 189 |
+
"step": 400
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.6535947712418301,
|
| 193 |
+
"eval_accuracy": 0.9936305732484076,
|
| 194 |
+
"eval_f1": 0.994475138121547,
|
| 195 |
+
"eval_loss": 0.0015705447876825929,
|
| 196 |
+
"eval_precision": 0.989010989010989,
|
| 197 |
+
"eval_recall": 1.0,
|
| 198 |
+
"eval_roc_auc": 0.9999013198470458,
|
| 199 |
+
"eval_runtime": 8.774,
|
| 200 |
+
"eval_samples_per_second": 196.831,
|
| 201 |
+
"eval_steps_per_second": 6.155,
|
| 202 |
+
"step": 400
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.6862745098039216,
|
| 206 |
+
"grad_norm": 0.006017300765961409,
|
| 207 |
+
"learning_rate": 1.9179375453885257e-05,
|
| 208 |
+
"loss": 0.0006,
|
| 209 |
+
"step": 420
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.7189542483660131,
|
| 213 |
+
"grad_norm": 0.0029072267934679985,
|
| 214 |
+
"learning_rate": 1.9034132171387073e-05,
|
| 215 |
+
"loss": 0.0008,
|
| 216 |
+
"step": 440
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 0.7516339869281046,
|
| 220 |
+
"grad_norm": 0.011737746186554432,
|
| 221 |
+
"learning_rate": 1.888888888888889e-05,
|
| 222 |
+
"loss": 0.0001,
|
| 223 |
+
"step": 460
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 0.7843137254901961,
|
| 227 |
+
"grad_norm": 0.043070435523986816,
|
| 228 |
+
"learning_rate": 1.8743645606390705e-05,
|
| 229 |
+
"loss": 0.0054,
|
| 230 |
+
"step": 480
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.8169934640522876,
|
| 234 |
+
"grad_norm": 0.0014130477793514729,
|
| 235 |
+
"learning_rate": 1.859840232389252e-05,
|
| 236 |
+
"loss": 0.0002,
|
| 237 |
+
"step": 500
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.8169934640522876,
|
| 241 |
+
"eval_accuracy": 0.9976838448176027,
|
| 242 |
+
"eval_f1": 0.997979797979798,
|
| 243 |
+
"eval_loss": 0.0020458686631172895,
|
| 244 |
+
"eval_precision": 0.997979797979798,
|
| 245 |
+
"eval_recall": 0.997979797979798,
|
| 246 |
+
"eval_roc_auc": 0.9999479188081631,
|
| 247 |
+
"eval_runtime": 8.7772,
|
| 248 |
+
"eval_samples_per_second": 196.761,
|
| 249 |
+
"eval_steps_per_second": 6.152,
|
| 250 |
+
"step": 500
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.8496732026143791,
|
| 254 |
+
"grad_norm": 0.1288619190454483,
|
| 255 |
+
"learning_rate": 1.8453159041394337e-05,
|
| 256 |
+
"loss": 0.0,
|
| 257 |
+
"step": 520
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.8823529411764706,
|
| 261 |
+
"grad_norm": 0.001577255199663341,
|
| 262 |
+
"learning_rate": 1.8307915758896153e-05,
|
| 263 |
+
"loss": 0.0036,
|
| 264 |
+
"step": 540
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.9150326797385621,
|
| 268 |
+
"grad_norm": 0.0329805426299572,
|
| 269 |
+
"learning_rate": 1.816267247639797e-05,
|
| 270 |
+
"loss": 0.0034,
|
| 271 |
+
"step": 560
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.9477124183006536,
|
| 275 |
+
"grad_norm": 0.0058128549717366695,
|
| 276 |
+
"learning_rate": 1.801742919389978e-05,
|
| 277 |
+
"loss": 0.0035,
|
| 278 |
+
"step": 580
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.9803921568627451,
|
| 282 |
+
"grad_norm": 0.00015560630708932877,
|
| 283 |
+
"learning_rate": 1.7872185911401597e-05,
|
| 284 |
+
"loss": 0.0008,
|
| 285 |
+
"step": 600
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.9803921568627451,
|
| 289 |
+
"eval_accuracy": 0.9971048060220035,
|
| 290 |
+
"eval_f1": 0.9974785678265254,
|
| 291 |
+
"eval_loss": 0.0012797409435734153,
|
| 292 |
+
"eval_precision": 0.9959718026183283,
|
| 293 |
+
"eval_recall": 0.998989898989899,
|
| 294 |
+
"eval_roc_auc": 0.9999218782122447,
|
| 295 |
+
"eval_runtime": 8.7835,
|
| 296 |
+
"eval_samples_per_second": 196.619,
|
| 297 |
+
"eval_steps_per_second": 6.148,
|
| 298 |
+
"step": 600
|
| 299 |
+
}
|
| 300 |
+
],
|
| 301 |
+
"logging_steps": 20,
|
| 302 |
+
"max_steps": 3060,
|
| 303 |
+
"num_input_tokens_seen": 0,
|
| 304 |
+
"num_train_epochs": 5,
|
| 305 |
+
"save_steps": 200,
|
| 306 |
+
"stateful_callbacks": {
|
| 307 |
+
"EarlyStoppingCallback": {
|
| 308 |
+
"args": {
|
| 309 |
+
"early_stopping_patience": 3,
|
| 310 |
+
"early_stopping_threshold": 0.0
|
| 311 |
+
},
|
| 312 |
+
"attributes": {
|
| 313 |
+
"early_stopping_patience_counter": 3
|
| 314 |
+
}
|
| 315 |
+
},
|
| 316 |
+
"TrainerControl": {
|
| 317 |
+
"args": {
|
| 318 |
+
"should_epoch_stop": false,
|
| 319 |
+
"should_evaluate": false,
|
| 320 |
+
"should_log": false,
|
| 321 |
+
"should_save": true,
|
| 322 |
+
"should_training_stop": true
|
| 323 |
+
},
|
| 324 |
+
"attributes": {}
|
| 325 |
+
}
|
| 326 |
+
},
|
| 327 |
+
"total_flos": 6542550643507200.0,
|
| 328 |
+
"train_batch_size": 16,
|
| 329 |
+
"trial_name": null,
|
| 330 |
+
"trial_params": null
|
| 331 |
+
}
|
checkpoint-600/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c1be75a7bbc1e87d42052a69bee01f6eaf1e6586d629ebf02cacfdf4eb951e52
|
| 3 |
+
size 5841
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b2aae000f6b136f5ba37b84e00f9f4b3849f2c3ebbb9f15316a1658b785ec20c
|
| 3 |
+
size 598439784
|
test_results.json
CHANGED
|
@@ -1,12 +1,12 @@
|
|
| 1 |
{
|
| 2 |
-
"eval_loss": 0.
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_f1": 0.
|
| 5 |
-
"eval_precision": 0.
|
| 6 |
"eval_recall": 1.0,
|
| 7 |
-
"eval_roc_auc": 0.
|
| 8 |
-
"eval_runtime": 10.
|
| 9 |
-
"eval_samples_per_second": 169.
|
| 10 |
-
"eval_steps_per_second": 5.
|
| 11 |
"epoch": 0.9803921568627451
|
| 12 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"eval_loss": 0.0017990026390179992,
|
| 3 |
+
"eval_accuracy": 0.9965297860034702,
|
| 4 |
+
"eval_f1": 0.9969543147208122,
|
| 5 |
+
"eval_precision": 0.9939271255060729,
|
| 6 |
"eval_recall": 1.0,
|
| 7 |
+
"eval_roc_auc": 0.9992372749654421,
|
| 8 |
+
"eval_runtime": 10.1974,
|
| 9 |
+
"eval_samples_per_second": 169.553,
|
| 10 |
+
"eval_steps_per_second": 5.394,
|
| 11 |
"epoch": 0.9803921568627451
|
| 12 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c1be75a7bbc1e87d42052a69bee01f6eaf1e6586d629ebf02cacfdf4eb951e52
|
| 3 |
+
size 5841
|