Training in progress, epoch 2, checkpoint
Browse files
checkpoint-1110/config.json
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/bart-large-mnli",
|
| 3 |
+
"_num_labels": 3,
|
| 4 |
+
"activation_dropout": 0.0,
|
| 5 |
+
"activation_function": "gelu",
|
| 6 |
+
"add_final_layer_norm": false,
|
| 7 |
+
"architectures": [
|
| 8 |
+
"BartForSequenceClassification"
|
| 9 |
+
],
|
| 10 |
+
"attention_dropout": 0.0,
|
| 11 |
+
"bos_token_id": 0,
|
| 12 |
+
"classif_dropout": 0.0,
|
| 13 |
+
"classifier_dropout": 0.0,
|
| 14 |
+
"d_model": 1024,
|
| 15 |
+
"decoder_attention_heads": 16,
|
| 16 |
+
"decoder_ffn_dim": 4096,
|
| 17 |
+
"decoder_layerdrop": 0.0,
|
| 18 |
+
"decoder_layers": 12,
|
| 19 |
+
"decoder_start_token_id": 2,
|
| 20 |
+
"dropout": 0.1,
|
| 21 |
+
"encoder_attention_heads": 16,
|
| 22 |
+
"encoder_ffn_dim": 4096,
|
| 23 |
+
"encoder_layerdrop": 0.0,
|
| 24 |
+
"encoder_layers": 12,
|
| 25 |
+
"eos_token_id": 2,
|
| 26 |
+
"forced_eos_token_id": 2,
|
| 27 |
+
"gradient_checkpointing": false,
|
| 28 |
+
"id2label": {
|
| 29 |
+
"0": "LABEL_0",
|
| 30 |
+
"1": "LABEL_1"
|
| 31 |
+
},
|
| 32 |
+
"init_std": 0.02,
|
| 33 |
+
"is_encoder_decoder": true,
|
| 34 |
+
"label2id": {
|
| 35 |
+
"LABEL_0": 0,
|
| 36 |
+
"LABEL_1": 1
|
| 37 |
+
},
|
| 38 |
+
"max_position_embeddings": 1024,
|
| 39 |
+
"model_type": "bart",
|
| 40 |
+
"normalize_before": false,
|
| 41 |
+
"num_hidden_layers": 12,
|
| 42 |
+
"output_past": false,
|
| 43 |
+
"pad_token_id": 1,
|
| 44 |
+
"scale_embedding": false,
|
| 45 |
+
"torch_dtype": "float32",
|
| 46 |
+
"transformers_version": "4.47.0",
|
| 47 |
+
"use_cache": true,
|
| 48 |
+
"vocab_size": 50265
|
| 49 |
+
}
|
checkpoint-1110/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d5161d3f28e6c123f9d2da6382a171942340cf80f568d275d32ce21b53e0f4a1
|
| 3 |
+
size 1629432864
|
checkpoint-1110/trainer_state.json
ADDED
|
@@ -0,0 +1,832 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 693,
|
| 6 |
+
"global_step": 1110,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0,
|
| 13 |
+
"eval_f1": 0.14772727272727273,
|
| 14 |
+
"eval_loss": 0.8850772380828857,
|
| 15 |
+
"eval_precision": 0.08080808080808081,
|
| 16 |
+
"eval_recall": 0.859504132231405,
|
| 17 |
+
"eval_runtime": 99.3343,
|
| 18 |
+
"eval_samples_per_second": 47.043,
|
| 19 |
+
"eval_steps_per_second": 5.889,
|
| 20 |
+
"step": 0
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.018026137899954935,
|
| 24 |
+
"grad_norm": 6.000438690185547,
|
| 25 |
+
"learning_rate": 4.981949458483755e-05,
|
| 26 |
+
"loss": 0.3931,
|
| 27 |
+
"step": 10
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.03605227579990987,
|
| 31 |
+
"grad_norm": 47.497467041015625,
|
| 32 |
+
"learning_rate": 4.963898916967509e-05,
|
| 33 |
+
"loss": 0.2607,
|
| 34 |
+
"step": 20
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 0.054078413699864804,
|
| 38 |
+
"grad_norm": 13.464141845703125,
|
| 39 |
+
"learning_rate": 4.945848375451264e-05,
|
| 40 |
+
"loss": 0.1813,
|
| 41 |
+
"step": 30
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.07210455159981974,
|
| 45 |
+
"grad_norm": 15.57222843170166,
|
| 46 |
+
"learning_rate": 4.927797833935018e-05,
|
| 47 |
+
"loss": 0.2618,
|
| 48 |
+
"step": 40
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.09013068949977468,
|
| 52 |
+
"grad_norm": 11.892234802246094,
|
| 53 |
+
"learning_rate": 4.909747292418773e-05,
|
| 54 |
+
"loss": 0.1559,
|
| 55 |
+
"step": 50
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 0.10815682739972961,
|
| 59 |
+
"grad_norm": 6.340595722198486,
|
| 60 |
+
"learning_rate": 4.891696750902527e-05,
|
| 61 |
+
"loss": 0.1648,
|
| 62 |
+
"step": 60
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.12618296529968454,
|
| 66 |
+
"grad_norm": 10.196683883666992,
|
| 67 |
+
"learning_rate": 4.873646209386282e-05,
|
| 68 |
+
"loss": 0.1472,
|
| 69 |
+
"step": 70
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.14420910319963948,
|
| 73 |
+
"grad_norm": 11.511515617370605,
|
| 74 |
+
"learning_rate": 4.855595667870036e-05,
|
| 75 |
+
"loss": 0.2108,
|
| 76 |
+
"step": 80
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.16223524109959442,
|
| 80 |
+
"grad_norm": 8.681073188781738,
|
| 81 |
+
"learning_rate": 4.837545126353791e-05,
|
| 82 |
+
"loss": 0.1561,
|
| 83 |
+
"step": 90
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.18026137899954936,
|
| 87 |
+
"grad_norm": 41.062801361083984,
|
| 88 |
+
"learning_rate": 4.819494584837546e-05,
|
| 89 |
+
"loss": 0.1563,
|
| 90 |
+
"step": 100
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.19828751689950427,
|
| 94 |
+
"grad_norm": 19.93891716003418,
|
| 95 |
+
"learning_rate": 4.8014440433213e-05,
|
| 96 |
+
"loss": 0.1881,
|
| 97 |
+
"step": 110
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.21631365479945922,
|
| 101 |
+
"grad_norm": 21.90877914428711,
|
| 102 |
+
"learning_rate": 4.783393501805055e-05,
|
| 103 |
+
"loss": 0.1262,
|
| 104 |
+
"step": 120
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.23433979269941416,
|
| 108 |
+
"grad_norm": 14.123740196228027,
|
| 109 |
+
"learning_rate": 4.765342960288809e-05,
|
| 110 |
+
"loss": 0.1535,
|
| 111 |
+
"step": 130
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.25236593059936907,
|
| 115 |
+
"grad_norm": 30.671157836914062,
|
| 116 |
+
"learning_rate": 4.747292418772563e-05,
|
| 117 |
+
"loss": 0.1458,
|
| 118 |
+
"step": 140
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 0.270392068499324,
|
| 122 |
+
"grad_norm": 7.560745716094971,
|
| 123 |
+
"learning_rate": 4.7292418772563177e-05,
|
| 124 |
+
"loss": 0.1797,
|
| 125 |
+
"step": 150
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.28841820639927895,
|
| 129 |
+
"grad_norm": 12.448927879333496,
|
| 130 |
+
"learning_rate": 4.711191335740072e-05,
|
| 131 |
+
"loss": 0.1898,
|
| 132 |
+
"step": 160
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 0.3064443442992339,
|
| 136 |
+
"grad_norm": 20.040008544921875,
|
| 137 |
+
"learning_rate": 4.693140794223827e-05,
|
| 138 |
+
"loss": 0.1698,
|
| 139 |
+
"step": 170
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 0.32447048219918884,
|
| 143 |
+
"grad_norm": 7.592446804046631,
|
| 144 |
+
"learning_rate": 4.675090252707581e-05,
|
| 145 |
+
"loss": 0.156,
|
| 146 |
+
"step": 180
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.3424966200991438,
|
| 150 |
+
"grad_norm": 18.830904006958008,
|
| 151 |
+
"learning_rate": 4.657039711191336e-05,
|
| 152 |
+
"loss": 0.1536,
|
| 153 |
+
"step": 190
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.3605227579990987,
|
| 157 |
+
"grad_norm": 3.684938430786133,
|
| 158 |
+
"learning_rate": 4.63898916967509e-05,
|
| 159 |
+
"loss": 0.1152,
|
| 160 |
+
"step": 200
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.3785488958990536,
|
| 164 |
+
"grad_norm": 13.293309211730957,
|
| 165 |
+
"learning_rate": 4.620938628158845e-05,
|
| 166 |
+
"loss": 0.2116,
|
| 167 |
+
"step": 210
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 0.39657503379900855,
|
| 171 |
+
"grad_norm": 14.630786895751953,
|
| 172 |
+
"learning_rate": 4.602888086642599e-05,
|
| 173 |
+
"loss": 0.1549,
|
| 174 |
+
"step": 220
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.4146011716989635,
|
| 178 |
+
"grad_norm": 20.37642478942871,
|
| 179 |
+
"learning_rate": 4.584837545126354e-05,
|
| 180 |
+
"loss": 0.1679,
|
| 181 |
+
"step": 230
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 0.43262730959891843,
|
| 185 |
+
"grad_norm": 13.183340072631836,
|
| 186 |
+
"learning_rate": 4.566787003610109e-05,
|
| 187 |
+
"loss": 0.1581,
|
| 188 |
+
"step": 240
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.45065344749887337,
|
| 192 |
+
"grad_norm": 21.777263641357422,
|
| 193 |
+
"learning_rate": 4.548736462093863e-05,
|
| 194 |
+
"loss": 0.1618,
|
| 195 |
+
"step": 250
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.4686795853988283,
|
| 199 |
+
"grad_norm": 20.683944702148438,
|
| 200 |
+
"learning_rate": 4.530685920577618e-05,
|
| 201 |
+
"loss": 0.1606,
|
| 202 |
+
"step": 260
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.48670572329878325,
|
| 206 |
+
"grad_norm": 30.506752014160156,
|
| 207 |
+
"learning_rate": 4.5126353790613716e-05,
|
| 208 |
+
"loss": 0.1831,
|
| 209 |
+
"step": 270
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.5047318611987381,
|
| 213 |
+
"grad_norm": 22.953815460205078,
|
| 214 |
+
"learning_rate": 4.494584837545127e-05,
|
| 215 |
+
"loss": 0.1565,
|
| 216 |
+
"step": 280
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 0.5227579990986931,
|
| 220 |
+
"grad_norm": 6.177365303039551,
|
| 221 |
+
"learning_rate": 4.4765342960288806e-05,
|
| 222 |
+
"loss": 0.1287,
|
| 223 |
+
"step": 290
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 0.540784136998648,
|
| 227 |
+
"grad_norm": 8.373098373413086,
|
| 228 |
+
"learning_rate": 4.458483754512636e-05,
|
| 229 |
+
"loss": 0.1553,
|
| 230 |
+
"step": 300
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.558810274898603,
|
| 234 |
+
"grad_norm": 10.07934284210205,
|
| 235 |
+
"learning_rate": 4.44043321299639e-05,
|
| 236 |
+
"loss": 0.1868,
|
| 237 |
+
"step": 310
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.5768364127985579,
|
| 241 |
+
"grad_norm": 18.686702728271484,
|
| 242 |
+
"learning_rate": 4.422382671480145e-05,
|
| 243 |
+
"loss": 0.1576,
|
| 244 |
+
"step": 320
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.5948625506985128,
|
| 248 |
+
"grad_norm": 10.679413795471191,
|
| 249 |
+
"learning_rate": 4.404332129963899e-05,
|
| 250 |
+
"loss": 0.1409,
|
| 251 |
+
"step": 330
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.6128886885984678,
|
| 255 |
+
"grad_norm": 11.381563186645508,
|
| 256 |
+
"learning_rate": 4.386281588447654e-05,
|
| 257 |
+
"loss": 0.2043,
|
| 258 |
+
"step": 340
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.6309148264984227,
|
| 262 |
+
"grad_norm": 40.68292236328125,
|
| 263 |
+
"learning_rate": 4.368231046931408e-05,
|
| 264 |
+
"loss": 0.1288,
|
| 265 |
+
"step": 350
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.6489409643983777,
|
| 269 |
+
"grad_norm": 17.46027946472168,
|
| 270 |
+
"learning_rate": 4.350180505415163e-05,
|
| 271 |
+
"loss": 0.1537,
|
| 272 |
+
"step": 360
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.6669671022983326,
|
| 276 |
+
"grad_norm": 13.656320571899414,
|
| 277 |
+
"learning_rate": 4.332129963898917e-05,
|
| 278 |
+
"loss": 0.1201,
|
| 279 |
+
"step": 370
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.6849932401982876,
|
| 283 |
+
"grad_norm": 20.412939071655273,
|
| 284 |
+
"learning_rate": 4.314079422382672e-05,
|
| 285 |
+
"loss": 0.1321,
|
| 286 |
+
"step": 380
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 0.7030193780982424,
|
| 290 |
+
"grad_norm": 9.957433700561523,
|
| 291 |
+
"learning_rate": 4.296028880866426e-05,
|
| 292 |
+
"loss": 0.141,
|
| 293 |
+
"step": 390
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 0.7210455159981974,
|
| 297 |
+
"grad_norm": 9.682558059692383,
|
| 298 |
+
"learning_rate": 4.277978339350181e-05,
|
| 299 |
+
"loss": 0.1607,
|
| 300 |
+
"step": 400
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 0.7390716538981523,
|
| 304 |
+
"grad_norm": 9.733131408691406,
|
| 305 |
+
"learning_rate": 4.259927797833935e-05,
|
| 306 |
+
"loss": 0.1037,
|
| 307 |
+
"step": 410
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.7570977917981072,
|
| 311 |
+
"grad_norm": 15.76230525970459,
|
| 312 |
+
"learning_rate": 4.24187725631769e-05,
|
| 313 |
+
"loss": 0.2101,
|
| 314 |
+
"step": 420
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.7751239296980622,
|
| 318 |
+
"grad_norm": 22.166545867919922,
|
| 319 |
+
"learning_rate": 4.223826714801444e-05,
|
| 320 |
+
"loss": 0.1459,
|
| 321 |
+
"step": 430
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.7931500675980171,
|
| 325 |
+
"grad_norm": 17.87278175354004,
|
| 326 |
+
"learning_rate": 4.205776173285199e-05,
|
| 327 |
+
"loss": 0.147,
|
| 328 |
+
"step": 440
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.8111762054979721,
|
| 332 |
+
"grad_norm": 5.201169490814209,
|
| 333 |
+
"learning_rate": 4.187725631768953e-05,
|
| 334 |
+
"loss": 0.1247,
|
| 335 |
+
"step": 450
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.829202343397927,
|
| 339 |
+
"grad_norm": 8.68251895904541,
|
| 340 |
+
"learning_rate": 4.169675090252708e-05,
|
| 341 |
+
"loss": 0.1172,
|
| 342 |
+
"step": 460
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.847228481297882,
|
| 346 |
+
"grad_norm": 14.086528778076172,
|
| 347 |
+
"learning_rate": 4.151624548736462e-05,
|
| 348 |
+
"loss": 0.111,
|
| 349 |
+
"step": 470
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.8652546191978369,
|
| 353 |
+
"grad_norm": 5.421584606170654,
|
| 354 |
+
"learning_rate": 4.1335740072202167e-05,
|
| 355 |
+
"loss": 0.1264,
|
| 356 |
+
"step": 480
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.8832807570977917,
|
| 360 |
+
"grad_norm": 7.700821399688721,
|
| 361 |
+
"learning_rate": 4.115523465703972e-05,
|
| 362 |
+
"loss": 0.1278,
|
| 363 |
+
"step": 490
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.9013068949977467,
|
| 367 |
+
"grad_norm": 9.820999145507812,
|
| 368 |
+
"learning_rate": 4.0974729241877256e-05,
|
| 369 |
+
"loss": 0.1294,
|
| 370 |
+
"step": 500
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.9193330328977016,
|
| 374 |
+
"grad_norm": 15.016349792480469,
|
| 375 |
+
"learning_rate": 4.079422382671481e-05,
|
| 376 |
+
"loss": 0.1607,
|
| 377 |
+
"step": 510
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.9373591707976566,
|
| 381 |
+
"grad_norm": 10.43041706085205,
|
| 382 |
+
"learning_rate": 4.0613718411552346e-05,
|
| 383 |
+
"loss": 0.1091,
|
| 384 |
+
"step": 520
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.9553853086976115,
|
| 388 |
+
"grad_norm": 26.99540901184082,
|
| 389 |
+
"learning_rate": 4.043321299638989e-05,
|
| 390 |
+
"loss": 0.1442,
|
| 391 |
+
"step": 530
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 0.9734114465975665,
|
| 395 |
+
"grad_norm": 8.318906784057617,
|
| 396 |
+
"learning_rate": 4.0252707581227436e-05,
|
| 397 |
+
"loss": 0.1233,
|
| 398 |
+
"step": 540
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.9914375844975214,
|
| 402 |
+
"grad_norm": 8.85923957824707,
|
| 403 |
+
"learning_rate": 4.007220216606498e-05,
|
| 404 |
+
"loss": 0.1466,
|
| 405 |
+
"step": 550
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 1.0090130689499774,
|
| 409 |
+
"grad_norm": 4.69356107711792,
|
| 410 |
+
"learning_rate": 3.989169675090253e-05,
|
| 411 |
+
"loss": 0.1433,
|
| 412 |
+
"step": 560
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 1.0270392068499323,
|
| 416 |
+
"grad_norm": 17.66107177734375,
|
| 417 |
+
"learning_rate": 3.971119133574007e-05,
|
| 418 |
+
"loss": 0.0944,
|
| 419 |
+
"step": 570
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 1.0450653447498874,
|
| 423 |
+
"grad_norm": 4.721229553222656,
|
| 424 |
+
"learning_rate": 3.953068592057762e-05,
|
| 425 |
+
"loss": 0.1372,
|
| 426 |
+
"step": 580
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 1.0630914826498423,
|
| 430 |
+
"grad_norm": 6.2178955078125,
|
| 431 |
+
"learning_rate": 3.935018050541516e-05,
|
| 432 |
+
"loss": 0.1412,
|
| 433 |
+
"step": 590
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 1.0811176205497972,
|
| 437 |
+
"grad_norm": 6.32495641708374,
|
| 438 |
+
"learning_rate": 3.916967509025271e-05,
|
| 439 |
+
"loss": 0.0892,
|
| 440 |
+
"step": 600
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.099143758449752,
|
| 444 |
+
"grad_norm": 12.221674919128418,
|
| 445 |
+
"learning_rate": 3.898916967509025e-05,
|
| 446 |
+
"loss": 0.1333,
|
| 447 |
+
"step": 610
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 1.117169896349707,
|
| 451 |
+
"grad_norm": 1.7088432312011719,
|
| 452 |
+
"learning_rate": 3.88086642599278e-05,
|
| 453 |
+
"loss": 0.1077,
|
| 454 |
+
"step": 620
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 1.135196034249662,
|
| 458 |
+
"grad_norm": 17.136337280273438,
|
| 459 |
+
"learning_rate": 3.862815884476535e-05,
|
| 460 |
+
"loss": 0.1381,
|
| 461 |
+
"step": 630
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 1.153222172149617,
|
| 465 |
+
"grad_norm": 5.424814224243164,
|
| 466 |
+
"learning_rate": 3.844765342960289e-05,
|
| 467 |
+
"loss": 0.1228,
|
| 468 |
+
"step": 640
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 1.1712483100495719,
|
| 472 |
+
"grad_norm": 18.775466918945312,
|
| 473 |
+
"learning_rate": 3.826714801444044e-05,
|
| 474 |
+
"loss": 0.099,
|
| 475 |
+
"step": 650
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 1.1892744479495267,
|
| 479 |
+
"grad_norm": 8.731025695800781,
|
| 480 |
+
"learning_rate": 3.8086642599277976e-05,
|
| 481 |
+
"loss": 0.1324,
|
| 482 |
+
"step": 660
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 1.2073005858494819,
|
| 486 |
+
"grad_norm": 9.53122615814209,
|
| 487 |
+
"learning_rate": 3.790613718411553e-05,
|
| 488 |
+
"loss": 0.1085,
|
| 489 |
+
"step": 670
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 1.2253267237494367,
|
| 493 |
+
"grad_norm": 17.20035743713379,
|
| 494 |
+
"learning_rate": 3.7725631768953066e-05,
|
| 495 |
+
"loss": 0.0774,
|
| 496 |
+
"step": 680
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 1.2433528616493916,
|
| 500 |
+
"grad_norm": 18.292253494262695,
|
| 501 |
+
"learning_rate": 3.754512635379062e-05,
|
| 502 |
+
"loss": 0.0965,
|
| 503 |
+
"step": 690
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 1.2487607030193781,
|
| 507 |
+
"eval_f1": 0.7416879795396419,
|
| 508 |
+
"eval_loss": 0.10558413714170456,
|
| 509 |
+
"eval_precision": 0.6921241050119332,
|
| 510 |
+
"eval_recall": 0.7988980716253443,
|
| 511 |
+
"eval_runtime": 115.001,
|
| 512 |
+
"eval_samples_per_second": 40.634,
|
| 513 |
+
"eval_steps_per_second": 5.087,
|
| 514 |
+
"step": 693
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.2613789995493465,
|
| 518 |
+
"grad_norm": 6.922005653381348,
|
| 519 |
+
"learning_rate": 3.7364620938628155e-05,
|
| 520 |
+
"loss": 0.0906,
|
| 521 |
+
"step": 700
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.2794051374493014,
|
| 525 |
+
"grad_norm": 33.214107513427734,
|
| 526 |
+
"learning_rate": 3.718411552346571e-05,
|
| 527 |
+
"loss": 0.072,
|
| 528 |
+
"step": 710
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.2974312753492563,
|
| 532 |
+
"grad_norm": 9.25637149810791,
|
| 533 |
+
"learning_rate": 3.700361010830325e-05,
|
| 534 |
+
"loss": 0.1597,
|
| 535 |
+
"step": 720
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.3154574132492114,
|
| 539 |
+
"grad_norm": 10.336318016052246,
|
| 540 |
+
"learning_rate": 3.68231046931408e-05,
|
| 541 |
+
"loss": 0.1063,
|
| 542 |
+
"step": 730
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.3334835511491663,
|
| 546 |
+
"grad_norm": 11.632000923156738,
|
| 547 |
+
"learning_rate": 3.664259927797834e-05,
|
| 548 |
+
"loss": 0.1175,
|
| 549 |
+
"step": 740
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.3515096890491212,
|
| 553 |
+
"grad_norm": 12.909192085266113,
|
| 554 |
+
"learning_rate": 3.646209386281589e-05,
|
| 555 |
+
"loss": 0.1285,
|
| 556 |
+
"step": 750
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.3695358269490763,
|
| 560 |
+
"grad_norm": 7.545075416564941,
|
| 561 |
+
"learning_rate": 3.628158844765343e-05,
|
| 562 |
+
"loss": 0.0864,
|
| 563 |
+
"step": 760
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.3875619648490312,
|
| 567 |
+
"grad_norm": 25.248464584350586,
|
| 568 |
+
"learning_rate": 3.610108303249098e-05,
|
| 569 |
+
"loss": 0.0769,
|
| 570 |
+
"step": 770
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.405588102748986,
|
| 574 |
+
"grad_norm": 10.5347318649292,
|
| 575 |
+
"learning_rate": 3.592057761732852e-05,
|
| 576 |
+
"loss": 0.088,
|
| 577 |
+
"step": 780
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.423614240648941,
|
| 581 |
+
"grad_norm": 13.101734161376953,
|
| 582 |
+
"learning_rate": 3.574007220216607e-05,
|
| 583 |
+
"loss": 0.1084,
|
| 584 |
+
"step": 790
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.4416403785488958,
|
| 588 |
+
"grad_norm": 12.188843727111816,
|
| 589 |
+
"learning_rate": 3.555956678700361e-05,
|
| 590 |
+
"loss": 0.1101,
|
| 591 |
+
"step": 800
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.4596665164488507,
|
| 595 |
+
"grad_norm": 11.552268981933594,
|
| 596 |
+
"learning_rate": 3.537906137184116e-05,
|
| 597 |
+
"loss": 0.0785,
|
| 598 |
+
"step": 810
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.4776926543488058,
|
| 602 |
+
"grad_norm": 10.269721031188965,
|
| 603 |
+
"learning_rate": 3.51985559566787e-05,
|
| 604 |
+
"loss": 0.1161,
|
| 605 |
+
"step": 820
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.4957187922487607,
|
| 609 |
+
"grad_norm": 12.09333610534668,
|
| 610 |
+
"learning_rate": 3.5018050541516247e-05,
|
| 611 |
+
"loss": 0.1113,
|
| 612 |
+
"step": 830
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.5137449301487156,
|
| 616 |
+
"grad_norm": 5.4690141677856445,
|
| 617 |
+
"learning_rate": 3.483754512635379e-05,
|
| 618 |
+
"loss": 0.1046,
|
| 619 |
+
"step": 840
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.5317710680486707,
|
| 623 |
+
"grad_norm": 6.296072959899902,
|
| 624 |
+
"learning_rate": 3.4657039711191336e-05,
|
| 625 |
+
"loss": 0.1202,
|
| 626 |
+
"step": 850
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.5497972059486256,
|
| 630 |
+
"grad_norm": 8.854303359985352,
|
| 631 |
+
"learning_rate": 3.447653429602888e-05,
|
| 632 |
+
"loss": 0.0929,
|
| 633 |
+
"step": 860
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.5678233438485805,
|
| 637 |
+
"grad_norm": 23.987024307250977,
|
| 638 |
+
"learning_rate": 3.4296028880866426e-05,
|
| 639 |
+
"loss": 0.1326,
|
| 640 |
+
"step": 870
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.5858494817485354,
|
| 644 |
+
"grad_norm": 7.229393005371094,
|
| 645 |
+
"learning_rate": 3.411552346570397e-05,
|
| 646 |
+
"loss": 0.1169,
|
| 647 |
+
"step": 880
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.6038756196484902,
|
| 651 |
+
"grad_norm": 5.463610649108887,
|
| 652 |
+
"learning_rate": 3.3935018050541516e-05,
|
| 653 |
+
"loss": 0.0809,
|
| 654 |
+
"step": 890
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.6219017575484451,
|
| 658 |
+
"grad_norm": 5.9659624099731445,
|
| 659 |
+
"learning_rate": 3.375451263537907e-05,
|
| 660 |
+
"loss": 0.1265,
|
| 661 |
+
"step": 900
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.6399278954484002,
|
| 665 |
+
"grad_norm": 17.582469940185547,
|
| 666 |
+
"learning_rate": 3.3574007220216606e-05,
|
| 667 |
+
"loss": 0.0846,
|
| 668 |
+
"step": 910
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.6579540333483551,
|
| 672 |
+
"grad_norm": 13.633155822753906,
|
| 673 |
+
"learning_rate": 3.339350180505416e-05,
|
| 674 |
+
"loss": 0.0999,
|
| 675 |
+
"step": 920
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.67598017124831,
|
| 679 |
+
"grad_norm": 14.496593475341797,
|
| 680 |
+
"learning_rate": 3.3212996389891696e-05,
|
| 681 |
+
"loss": 0.0903,
|
| 682 |
+
"step": 930
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.694006309148265,
|
| 686 |
+
"grad_norm": 7.594356536865234,
|
| 687 |
+
"learning_rate": 3.303249097472924e-05,
|
| 688 |
+
"loss": 0.1021,
|
| 689 |
+
"step": 940
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.71203244704822,
|
| 693 |
+
"grad_norm": 20.628849029541016,
|
| 694 |
+
"learning_rate": 3.2851985559566786e-05,
|
| 695 |
+
"loss": 0.1131,
|
| 696 |
+
"step": 950
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.7300585849481749,
|
| 700 |
+
"grad_norm": 8.25106430053711,
|
| 701 |
+
"learning_rate": 3.267148014440433e-05,
|
| 702 |
+
"loss": 0.0872,
|
| 703 |
+
"step": 960
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.7480847228481298,
|
| 707 |
+
"grad_norm": 17.08070182800293,
|
| 708 |
+
"learning_rate": 3.249097472924188e-05,
|
| 709 |
+
"loss": 0.0663,
|
| 710 |
+
"step": 970
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.7661108607480847,
|
| 714 |
+
"grad_norm": 38.33162307739258,
|
| 715 |
+
"learning_rate": 3.231046931407942e-05,
|
| 716 |
+
"loss": 0.142,
|
| 717 |
+
"step": 980
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.7841369986480395,
|
| 721 |
+
"grad_norm": 17.89957618713379,
|
| 722 |
+
"learning_rate": 3.212996389891697e-05,
|
| 723 |
+
"loss": 0.0942,
|
| 724 |
+
"step": 990
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.8021631365479944,
|
| 728 |
+
"grad_norm": 9.691226959228516,
|
| 729 |
+
"learning_rate": 3.194945848375451e-05,
|
| 730 |
+
"loss": 0.0902,
|
| 731 |
+
"step": 1000
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.8201892744479495,
|
| 735 |
+
"grad_norm": 9.764283180236816,
|
| 736 |
+
"learning_rate": 3.176895306859206e-05,
|
| 737 |
+
"loss": 0.1108,
|
| 738 |
+
"step": 1010
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.8382154123479044,
|
| 742 |
+
"grad_norm": 13.829035758972168,
|
| 743 |
+
"learning_rate": 3.15884476534296e-05,
|
| 744 |
+
"loss": 0.0881,
|
| 745 |
+
"step": 1020
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.8562415502478595,
|
| 749 |
+
"grad_norm": 8.372889518737793,
|
| 750 |
+
"learning_rate": 3.140794223826715e-05,
|
| 751 |
+
"loss": 0.1272,
|
| 752 |
+
"step": 1030
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.8742676881478144,
|
| 756 |
+
"grad_norm": 12.576041221618652,
|
| 757 |
+
"learning_rate": 3.12274368231047e-05,
|
| 758 |
+
"loss": 0.1076,
|
| 759 |
+
"step": 1040
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.8922938260477693,
|
| 763 |
+
"grad_norm": 12.995355606079102,
|
| 764 |
+
"learning_rate": 3.104693140794224e-05,
|
| 765 |
+
"loss": 0.0932,
|
| 766 |
+
"step": 1050
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.9103199639477242,
|
| 770 |
+
"grad_norm": 6.626877784729004,
|
| 771 |
+
"learning_rate": 3.086642599277979e-05,
|
| 772 |
+
"loss": 0.105,
|
| 773 |
+
"step": 1060
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.928346101847679,
|
| 777 |
+
"grad_norm": 11.209733009338379,
|
| 778 |
+
"learning_rate": 3.0685920577617325e-05,
|
| 779 |
+
"loss": 0.1331,
|
| 780 |
+
"step": 1070
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.946372239747634,
|
| 784 |
+
"grad_norm": 7.832246780395508,
|
| 785 |
+
"learning_rate": 3.0505415162454877e-05,
|
| 786 |
+
"loss": 0.087,
|
| 787 |
+
"step": 1080
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.9643983776475888,
|
| 791 |
+
"grad_norm": 9.489190101623535,
|
| 792 |
+
"learning_rate": 3.032490974729242e-05,
|
| 793 |
+
"loss": 0.0904,
|
| 794 |
+
"step": 1090
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.982424515547544,
|
| 798 |
+
"grad_norm": 14.940125465393066,
|
| 799 |
+
"learning_rate": 3.0144404332129967e-05,
|
| 800 |
+
"loss": 0.0959,
|
| 801 |
+
"step": 1100
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 2.0,
|
| 805 |
+
"grad_norm": 8.826193809509277,
|
| 806 |
+
"learning_rate": 2.996389891696751e-05,
|
| 807 |
+
"loss": 0.1276,
|
| 808 |
+
"step": 1110
|
| 809 |
+
}
|
| 810 |
+
],
|
| 811 |
+
"logging_steps": 10,
|
| 812 |
+
"max_steps": 2770,
|
| 813 |
+
"num_input_tokens_seen": 0,
|
| 814 |
+
"num_train_epochs": 5,
|
| 815 |
+
"save_steps": 500,
|
| 816 |
+
"stateful_callbacks": {
|
| 817 |
+
"TrainerControl": {
|
| 818 |
+
"args": {
|
| 819 |
+
"should_epoch_stop": false,
|
| 820 |
+
"should_evaluate": false,
|
| 821 |
+
"should_log": false,
|
| 822 |
+
"should_save": true,
|
| 823 |
+
"should_training_stop": false
|
| 824 |
+
},
|
| 825 |
+
"attributes": {}
|
| 826 |
+
}
|
| 827 |
+
},
|
| 828 |
+
"total_flos": 3.858514256176742e+16,
|
| 829 |
+
"train_batch_size": 8,
|
| 830 |
+
"trial_name": null,
|
| 831 |
+
"trial_params": null
|
| 832 |
+
}
|
checkpoint-1110/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a1cb6c2831bf6310695ea4ef6539addf4fc4092fd2b8fd2fd7364adece0d315f
|
| 3 |
+
size 5368
|