Upload checkpoint and eval results at step 2400
Browse files- checkpoint-2400/config.json +37 -0
- checkpoint-2400/eval_results.json +12 -0
- checkpoint-2400/eval_step_002400.json +16 -0
- checkpoint-2400/model.safetensors +3 -0
- checkpoint-2400/optimizer.pt +3 -0
- checkpoint-2400/rng_state.pth +3 -0
- checkpoint-2400/scaler.pt +3 -0
- checkpoint-2400/scheduler.pt +3 -0
- checkpoint-2400/tokenizer.json +0 -0
- checkpoint-2400/tokenizer_config.json +17 -0
- checkpoint-2400/trainer_state.json +794 -0
- checkpoint-2400/training_args.bin +3 -0
checkpoint-2400/config.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_cross_attention": false,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"RobertaForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": 0,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"dtype": "float32",
|
| 10 |
+
"eos_token_id": 2,
|
| 11 |
+
"hidden_act": "gelu",
|
| 12 |
+
"hidden_dropout_prob": 0.1,
|
| 13 |
+
"hidden_size": 768,
|
| 14 |
+
"id2label": {
|
| 15 |
+
"0": "benign",
|
| 16 |
+
"1": "malicious"
|
| 17 |
+
},
|
| 18 |
+
"initializer_range": 0.02,
|
| 19 |
+
"intermediate_size": 3072,
|
| 20 |
+
"is_decoder": false,
|
| 21 |
+
"label2id": {
|
| 22 |
+
"benign": 0,
|
| 23 |
+
"malicious": 1
|
| 24 |
+
},
|
| 25 |
+
"layer_norm_eps": 1e-05,
|
| 26 |
+
"max_position_embeddings": 514,
|
| 27 |
+
"model_type": "roberta",
|
| 28 |
+
"num_attention_heads": 12,
|
| 29 |
+
"num_hidden_layers": 12,
|
| 30 |
+
"output_past": true,
|
| 31 |
+
"pad_token_id": 1,
|
| 32 |
+
"tie_word_embeddings": true,
|
| 33 |
+
"transformers_version": "5.12.1",
|
| 34 |
+
"type_vocab_size": 1,
|
| 35 |
+
"use_cache": false,
|
| 36 |
+
"vocab_size": 50265
|
| 37 |
+
}
|
checkpoint-2400/eval_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"step": 2400,
|
| 3 |
+
"epoch": 1.941747572815534,
|
| 4 |
+
"time": "2026-06-26 20:55:19",
|
| 5 |
+
"metrics": {
|
| 6 |
+
"eval_loss": 0.14186660945415497,
|
| 7 |
+
"eval_accuracy": 0.9631126831733199,
|
| 8 |
+
"eval_precision": 0.9644012944983819,
|
| 9 |
+
"eval_recall": 0.9880636604774535,
|
| 10 |
+
"eval_f1": 0.9760890926957091
|
| 11 |
+
}
|
| 12 |
+
}
|
checkpoint-2400/eval_step_002400.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"step": 2400,
|
| 3 |
+
"epoch": 1.941747572815534,
|
| 4 |
+
"time": "2026-06-26 20:55:19",
|
| 5 |
+
"metrics": {
|
| 6 |
+
"eval_loss": 0.14186660945415497,
|
| 7 |
+
"eval_accuracy": 0.9631126831733199,
|
| 8 |
+
"eval_precision": 0.9644012944983819,
|
| 9 |
+
"eval_recall": 0.9880636604774535,
|
| 10 |
+
"eval_f1": 0.9760890926957091,
|
| 11 |
+
"eval_runtime": 34.6039,
|
| 12 |
+
"eval_samples_per_second": 57.19,
|
| 13 |
+
"eval_steps_per_second": 3.583,
|
| 14 |
+
"epoch": 1.941747572815534
|
| 15 |
+
}
|
| 16 |
+
}
|
checkpoint-2400/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:485fd536bb3ddaaa764ebcc88395217a83549faf406ecb8653a8f92d93659d07
|
| 3 |
+
size 498612824
|
checkpoint-2400/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b6e50568f53dc95802006242fe8a3572e250793077702bb08e7e18d0bd39f266
|
| 3 |
+
size 997348747
|
checkpoint-2400/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8037dbacdfe9b7b77db2dd97ca0b5a4952ecf6f0774f70bcc551051358fa3d97
|
| 3 |
+
size 14645
|
checkpoint-2400/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7da7c5085795b13d2bf0030671cbddb9f62ae43221bf1424a3830d4cf8c19012
|
| 3 |
+
size 1383
|
checkpoint-2400/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bd8a30c72a1259c9d694db421b0ef3449b4c169404caca82507ff76fe60a7e0e
|
| 3 |
+
size 1465
|
checkpoint-2400/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-2400/tokenizer_config.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "<s>",
|
| 5 |
+
"cls_token": "<s>",
|
| 6 |
+
"eos_token": "</s>",
|
| 7 |
+
"errors": "replace",
|
| 8 |
+
"is_local": false,
|
| 9 |
+
"local_files_only": false,
|
| 10 |
+
"mask_token": "<mask>",
|
| 11 |
+
"model_max_length": 1000000000,
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"tokenizer_class": "RobertaTokenizer",
|
| 15 |
+
"trim_offsets": true,
|
| 16 |
+
"unk_token": "<unk>"
|
| 17 |
+
}
|
checkpoint-2400/trainer_state.json
ADDED
|
@@ -0,0 +1,794 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1600,
|
| 3 |
+
"best_metric": 0.9785549323655559,
|
| 4 |
+
"best_model_checkpoint": "/kaggle/working/outputs/codebert_rd_keystrokes/checkpoint-1600",
|
| 5 |
+
"epoch": 1.941747572815534,
|
| 6 |
+
"eval_steps": 400,
|
| 7 |
+
"global_step": 2400,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0008090614886731392,
|
| 14 |
+
"grad_norm": 4.001261234283447,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 0.748644232749939,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.020226537216828478,
|
| 21 |
+
"grad_norm": 2.0581886768341064,
|
| 22 |
+
"learning_rate": 9.716599190283402e-07,
|
| 23 |
+
"loss": 0.7263915538787842,
|
| 24 |
+
"step": 25
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.040453074433656956,
|
| 28 |
+
"grad_norm": 3.1348977088928223,
|
| 29 |
+
"learning_rate": 1.9838056680161946e-06,
|
| 30 |
+
"loss": 0.7004936218261719,
|
| 31 |
+
"step": 50
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.06067961165048544,
|
| 35 |
+
"grad_norm": 7.536289215087891,
|
| 36 |
+
"learning_rate": 2.995951417004049e-06,
|
| 37 |
+
"loss": 0.6983621978759765,
|
| 38 |
+
"step": 75
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.08090614886731391,
|
| 42 |
+
"grad_norm": 5.387710094451904,
|
| 43 |
+
"learning_rate": 4.008097165991903e-06,
|
| 44 |
+
"loss": 0.6408506774902344,
|
| 45 |
+
"step": 100
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.1011326860841424,
|
| 49 |
+
"grad_norm": 4.4866790771484375,
|
| 50 |
+
"learning_rate": 5.020242914979757e-06,
|
| 51 |
+
"loss": 0.5292000961303711,
|
| 52 |
+
"step": 125
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.12135922330097088,
|
| 56 |
+
"grad_norm": 5.9393534660339355,
|
| 57 |
+
"learning_rate": 6.0323886639676124e-06,
|
| 58 |
+
"loss": 0.4315451049804688,
|
| 59 |
+
"step": 150
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.14158576051779936,
|
| 63 |
+
"grad_norm": 4.814444065093994,
|
| 64 |
+
"learning_rate": 7.044534412955466e-06,
|
| 65 |
+
"loss": 0.3674881362915039,
|
| 66 |
+
"step": 175
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.16181229773462782,
|
| 70 |
+
"grad_norm": 2.463817834854126,
|
| 71 |
+
"learning_rate": 8.056680161943322e-06,
|
| 72 |
+
"loss": 0.29549776077270506,
|
| 73 |
+
"step": 200
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.1820388349514563,
|
| 77 |
+
"grad_norm": 1.6270427703857422,
|
| 78 |
+
"learning_rate": 9.068825910931175e-06,
|
| 79 |
+
"loss": 0.19513965606689454,
|
| 80 |
+
"step": 225
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.2022653721682848,
|
| 84 |
+
"grad_norm": 15.206025123596191,
|
| 85 |
+
"learning_rate": 1.008097165991903e-05,
|
| 86 |
+
"loss": 0.258267822265625,
|
| 87 |
+
"step": 250
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.22249190938511326,
|
| 91 |
+
"grad_norm": 5.6893134117126465,
|
| 92 |
+
"learning_rate": 1.1093117408906884e-05,
|
| 93 |
+
"loss": 0.3252030563354492,
|
| 94 |
+
"step": 275
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.24271844660194175,
|
| 98 |
+
"grad_norm": 15.90676498413086,
|
| 99 |
+
"learning_rate": 1.2105263157894737e-05,
|
| 100 |
+
"loss": 0.2558871841430664,
|
| 101 |
+
"step": 300
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.26294498381877024,
|
| 105 |
+
"grad_norm": 1.5257281064987183,
|
| 106 |
+
"learning_rate": 1.3117408906882592e-05,
|
| 107 |
+
"loss": 0.20428178787231446,
|
| 108 |
+
"step": 325
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.28317152103559873,
|
| 112 |
+
"grad_norm": 20.172571182250977,
|
| 113 |
+
"learning_rate": 1.4129554655870446e-05,
|
| 114 |
+
"loss": 0.2579817008972168,
|
| 115 |
+
"step": 350
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.30339805825242716,
|
| 119 |
+
"grad_norm": 0.42674902081489563,
|
| 120 |
+
"learning_rate": 1.5141700404858302e-05,
|
| 121 |
+
"loss": 0.3013817977905273,
|
| 122 |
+
"step": 375
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.32362459546925565,
|
| 126 |
+
"grad_norm": 0.6443495750427246,
|
| 127 |
+
"learning_rate": 1.6153846153846154e-05,
|
| 128 |
+
"loss": 0.22500619888305665,
|
| 129 |
+
"step": 400
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.32362459546925565,
|
| 133 |
+
"eval_accuracy": 0.9560384032339565,
|
| 134 |
+
"eval_f1": 0.9716889033517735,
|
| 135 |
+
"eval_loss": 0.16199301183223724,
|
| 136 |
+
"eval_precision": 0.9539936102236422,
|
| 137 |
+
"eval_recall": 0.9900530503978779,
|
| 138 |
+
"eval_runtime": 34.745,
|
| 139 |
+
"eval_samples_per_second": 56.958,
|
| 140 |
+
"eval_steps_per_second": 3.569,
|
| 141 |
+
"step": 400
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.34385113268608414,
|
| 145 |
+
"grad_norm": 12.952281951904297,
|
| 146 |
+
"learning_rate": 1.716599190283401e-05,
|
| 147 |
+
"loss": 0.23191883087158202,
|
| 148 |
+
"step": 425
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.3640776699029126,
|
| 152 |
+
"grad_norm": 19.496694564819336,
|
| 153 |
+
"learning_rate": 1.8178137651821864e-05,
|
| 154 |
+
"loss": 0.26200836181640624,
|
| 155 |
+
"step": 450
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.3843042071197411,
|
| 159 |
+
"grad_norm": 2.9963057041168213,
|
| 160 |
+
"learning_rate": 1.9190283400809718e-05,
|
| 161 |
+
"loss": 0.164515438079834,
|
| 162 |
+
"step": 475
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 0.4045307443365696,
|
| 166 |
+
"grad_norm": 0.2180398553609848,
|
| 167 |
+
"learning_rate": 1.999993769982229e-05,
|
| 168 |
+
"loss": 0.2513391304016113,
|
| 169 |
+
"step": 500
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.42475728155339804,
|
| 173 |
+
"grad_norm": 15.512799263000488,
|
| 174 |
+
"learning_rate": 1.9997757275108847e-05,
|
| 175 |
+
"loss": 0.30474105834960935,
|
| 176 |
+
"step": 525
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.4449838187702265,
|
| 180 |
+
"grad_norm": 23.762832641601562,
|
| 181 |
+
"learning_rate": 1.9992462617728763e-05,
|
| 182 |
+
"loss": 0.214493408203125,
|
| 183 |
+
"step": 550
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.465210355987055,
|
| 187 |
+
"grad_norm": 0.5767886638641357,
|
| 188 |
+
"learning_rate": 1.9984055376931414e-05,
|
| 189 |
+
"loss": 0.23740652084350586,
|
| 190 |
+
"step": 575
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.4854368932038835,
|
| 194 |
+
"grad_norm": 3.6927132606506348,
|
| 195 |
+
"learning_rate": 1.9972538171514525e-05,
|
| 196 |
+
"loss": 0.2517130470275879,
|
| 197 |
+
"step": 600
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.5056634304207119,
|
| 201 |
+
"grad_norm": 2.146721363067627,
|
| 202 |
+
"learning_rate": 1.9957914589008405e-05,
|
| 203 |
+
"loss": 0.21869991302490235,
|
| 204 |
+
"step": 625
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.5258899676375405,
|
| 208 |
+
"grad_norm": 9.424388885498047,
|
| 209 |
+
"learning_rate": 1.9940189184558487e-05,
|
| 210 |
+
"loss": 0.2582887077331543,
|
| 211 |
+
"step": 650
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 0.5461165048543689,
|
| 215 |
+
"grad_norm": 4.679764747619629,
|
| 216 |
+
"learning_rate": 1.9919367479506413e-05,
|
| 217 |
+
"loss": 0.2324064254760742,
|
| 218 |
+
"step": 675
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.5663430420711975,
|
| 222 |
+
"grad_norm": 3.0119147300720215,
|
| 223 |
+
"learning_rate": 1.989545595967018e-05,
|
| 224 |
+
"loss": 0.23698102951049804,
|
| 225 |
+
"step": 700
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.5865695792880259,
|
| 229 |
+
"grad_norm": 10.184627532958984,
|
| 230 |
+
"learning_rate": 1.986846207332384e-05,
|
| 231 |
+
"loss": 0.23535886764526368,
|
| 232 |
+
"step": 725
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.6067961165048543,
|
| 236 |
+
"grad_norm": 51.39314651489258,
|
| 237 |
+
"learning_rate": 1.9838394228877443e-05,
|
| 238 |
+
"loss": 0.13176690101623534,
|
| 239 |
+
"step": 750
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.6270226537216829,
|
| 243 |
+
"grad_norm": 18.441652297973633,
|
| 244 |
+
"learning_rate": 1.980526179225785e-05,
|
| 245 |
+
"loss": 0.2888069725036621,
|
| 246 |
+
"step": 775
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.6472491909385113,
|
| 250 |
+
"grad_norm": 0.9691627621650696,
|
| 251 |
+
"learning_rate": 1.97690750839913e-05,
|
| 252 |
+
"loss": 0.29139081954956053,
|
| 253 |
+
"step": 800
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 0.6472491909385113,
|
| 257 |
+
"eval_accuracy": 0.965133906013138,
|
| 258 |
+
"eval_f1": 0.9773547751887102,
|
| 259 |
+
"eval_loss": 0.14249353110790253,
|
| 260 |
+
"eval_precision": 0.967511371020143,
|
| 261 |
+
"eval_recall": 0.9874005305039788,
|
| 262 |
+
"eval_runtime": 34.6417,
|
| 263 |
+
"eval_samples_per_second": 57.128,
|
| 264 |
+
"eval_steps_per_second": 3.58,
|
| 265 |
+
"step": 800
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.6674757281553398,
|
| 269 |
+
"grad_norm": 7.6904096603393555,
|
| 270 |
+
"learning_rate": 1.9729845375988694e-05,
|
| 271 |
+
"loss": 0.1711592674255371,
|
| 272 |
+
"step": 825
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.6877022653721683,
|
| 276 |
+
"grad_norm": 6.999588489532471,
|
| 277 |
+
"learning_rate": 1.968758488803439e-05,
|
| 278 |
+
"loss": 0.2283180809020996,
|
| 279 |
+
"step": 850
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.7079288025889967,
|
| 283 |
+
"grad_norm": 6.149271488189697,
|
| 284 |
+
"learning_rate": 1.9642306783979902e-05,
|
| 285 |
+
"loss": 0.26335405349731444,
|
| 286 |
+
"step": 875
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 0.7281553398058253,
|
| 290 |
+
"grad_norm": 9.907973289489746,
|
| 291 |
+
"learning_rate": 1.959402516764339e-05,
|
| 292 |
+
"loss": 0.1903523635864258,
|
| 293 |
+
"step": 900
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 0.7483818770226537,
|
| 297 |
+
"grad_norm": 12.104832649230957,
|
| 298 |
+
"learning_rate": 1.954275507841646e-05,
|
| 299 |
+
"loss": 0.1740720748901367,
|
| 300 |
+
"step": 925
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 0.7686084142394822,
|
| 304 |
+
"grad_norm": 2.8893074989318848,
|
| 305 |
+
"learning_rate": 1.948851248657947e-05,
|
| 306 |
+
"loss": 0.3097032928466797,
|
| 307 |
+
"step": 950
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.7888349514563107,
|
| 311 |
+
"grad_norm": 0.07049937546253204,
|
| 312 |
+
"learning_rate": 1.9431314288326925e-05,
|
| 313 |
+
"loss": 0.14465970993041993,
|
| 314 |
+
"step": 975
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.8090614886731392,
|
| 318 |
+
"grad_norm": 18.34374237060547,
|
| 319 |
+
"learning_rate": 1.9371178300504394e-05,
|
| 320 |
+
"loss": 0.22320323944091797,
|
| 321 |
+
"step": 1000
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.8292880258899676,
|
| 325 |
+
"grad_norm": 5.898121356964111,
|
| 326 |
+
"learning_rate": 1.930812325505871e-05,
|
| 327 |
+
"loss": 0.09155179023742675,
|
| 328 |
+
"step": 1025
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.8495145631067961,
|
| 332 |
+
"grad_norm": 9.933682441711426,
|
| 333 |
+
"learning_rate": 1.9242168793203088e-05,
|
| 334 |
+
"loss": 0.28490724563598635,
|
| 335 |
+
"step": 1050
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.8697411003236246,
|
| 339 |
+
"grad_norm": 0.8073019981384277,
|
| 340 |
+
"learning_rate": 1.9173335459299025e-05,
|
| 341 |
+
"loss": 0.14135926246643066,
|
| 342 |
+
"step": 1075
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.889967637540453,
|
| 346 |
+
"grad_norm": 23.247236251831055,
|
| 347 |
+
"learning_rate": 1.9101644694456873e-05,
|
| 348 |
+
"loss": 0.2752204895019531,
|
| 349 |
+
"step": 1100
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.9101941747572816,
|
| 353 |
+
"grad_norm": 10.088881492614746,
|
| 354 |
+
"learning_rate": 1.902711882985708e-05,
|
| 355 |
+
"loss": 0.14879122734069825,
|
| 356 |
+
"step": 1125
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.93042071197411,
|
| 360 |
+
"grad_norm": 0.5620437860488892,
|
| 361 |
+
"learning_rate": 1.894978107979417e-05,
|
| 362 |
+
"loss": 0.23470134735107423,
|
| 363 |
+
"step": 1150
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.9506472491909385,
|
| 367 |
+
"grad_norm": 3.2695255279541016,
|
| 368 |
+
"learning_rate": 1.886965553444568e-05,
|
| 369 |
+
"loss": 0.19982004165649414,
|
| 370 |
+
"step": 1175
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.970873786407767,
|
| 374 |
+
"grad_norm": 19.842267990112305,
|
| 375 |
+
"learning_rate": 1.8786767152368197e-05,
|
| 376 |
+
"loss": 0.22285139083862304,
|
| 377 |
+
"step": 1200
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.970873786407767,
|
| 381 |
+
"eval_accuracy": 0.9641232945932289,
|
| 382 |
+
"eval_f1": 0.9767441860465116,
|
| 383 |
+
"eval_loss": 0.14106644690036774,
|
| 384 |
+
"eval_precision": 0.9650485436893204,
|
| 385 |
+
"eval_recall": 0.9887267904509284,
|
| 386 |
+
"eval_runtime": 34.6101,
|
| 387 |
+
"eval_samples_per_second": 57.18,
|
| 388 |
+
"eval_steps_per_second": 3.583,
|
| 389 |
+
"step": 1200
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.9911003236245954,
|
| 393 |
+
"grad_norm": 17.05675506591797,
|
| 394 |
+
"learning_rate": 1.8701141752722966e-05,
|
| 395 |
+
"loss": 0.13252583503723145,
|
| 396 |
+
"step": 1225
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.0113268608414239,
|
| 400 |
+
"grad_norm": 14.713388442993164,
|
| 401 |
+
"learning_rate": 1.8612806007233382e-05,
|
| 402 |
+
"loss": 0.13512847900390626,
|
| 403 |
+
"step": 1250
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.0315533980582525,
|
| 407 |
+
"grad_norm": 16.86681365966797,
|
| 408 |
+
"learning_rate": 1.8521787431876954e-05,
|
| 409 |
+
"loss": 0.15763619422912598,
|
| 410 |
+
"step": 1275
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.051779935275081,
|
| 414 |
+
"grad_norm": 0.6122269630432129,
|
| 415 |
+
"learning_rate": 1.8428114378314237e-05,
|
| 416 |
+
"loss": 0.18742298126220702,
|
| 417 |
+
"step": 1300
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.0720064724919094,
|
| 421 |
+
"grad_norm": 6.803472995758057,
|
| 422 |
+
"learning_rate": 1.8331816025057508e-05,
|
| 423 |
+
"loss": 0.16315383911132814,
|
| 424 |
+
"step": 1325
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.0922330097087378,
|
| 428 |
+
"grad_norm": 3.42342209815979,
|
| 429 |
+
"learning_rate": 1.823292236838186e-05,
|
| 430 |
+
"loss": 0.1462923812866211,
|
| 431 |
+
"step": 1350
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.1124595469255663,
|
| 435 |
+
"grad_norm": 6.770485877990723,
|
| 436 |
+
"learning_rate": 1.813146421298154e-05,
|
| 437 |
+
"loss": 0.2096657371520996,
|
| 438 |
+
"step": 1375
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.132686084142395,
|
| 442 |
+
"grad_norm": 6.892830848693848,
|
| 443 |
+
"learning_rate": 1.802747316237454e-05,
|
| 444 |
+
"loss": 0.1903732681274414,
|
| 445 |
+
"step": 1400
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.1529126213592233,
|
| 449 |
+
"grad_norm": 1.7113957405090332,
|
| 450 |
+
"learning_rate": 1.792098160905829e-05,
|
| 451 |
+
"loss": 0.22764249801635741,
|
| 452 |
+
"step": 1425
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.1731391585760518,
|
| 456 |
+
"grad_norm": 0.2954396903514862,
|
| 457 |
+
"learning_rate": 1.7812022724419624e-05,
|
| 458 |
+
"loss": 0.08850884437561035,
|
| 459 |
+
"step": 1450
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.1933656957928802,
|
| 463 |
+
"grad_norm": 1.2150897979736328,
|
| 464 |
+
"learning_rate": 1.7700630448402125e-05,
|
| 465 |
+
"loss": 0.15286758422851562,
|
| 466 |
+
"step": 1475
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.2135922330097086,
|
| 470 |
+
"grad_norm": 2.9238908290863037,
|
| 471 |
+
"learning_rate": 1.758683947893406e-05,
|
| 472 |
+
"loss": 0.14677520751953124,
|
| 473 |
+
"step": 1500
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.2338187702265373,
|
| 477 |
+
"grad_norm": 9.288032531738281,
|
| 478 |
+
"learning_rate": 1.747068526112022e-05,
|
| 479 |
+
"loss": 0.13224196434020996,
|
| 480 |
+
"step": 1525
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.2540453074433657,
|
| 484 |
+
"grad_norm": 21.540422439575195,
|
| 485 |
+
"learning_rate": 1.735220397620101e-05,
|
| 486 |
+
"loss": 0.08127058982849121,
|
| 487 |
+
"step": 1550
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.2742718446601942,
|
| 491 |
+
"grad_norm": 0.07893255352973938,
|
| 492 |
+
"learning_rate": 1.7231432530282246e-05,
|
| 493 |
+
"loss": 0.10571711540222167,
|
| 494 |
+
"step": 1575
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.2944983818770226,
|
| 498 |
+
"grad_norm": 0.5910708904266357,
|
| 499 |
+
"learning_rate": 1.7108408542839146e-05,
|
| 500 |
+
"loss": 0.1872186279296875,
|
| 501 |
+
"step": 1600
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.2944983818770226,
|
| 505 |
+
"eval_accuracy": 0.9671551288529561,
|
| 506 |
+
"eval_f1": 0.9785549323655559,
|
| 507 |
+
"eval_loss": 0.1455843597650528,
|
| 508 |
+
"eval_precision": 0.9737360472751149,
|
| 509 |
+
"eval_recall": 0.98342175066313,
|
| 510 |
+
"eval_runtime": 34.8064,
|
| 511 |
+
"eval_samples_per_second": 56.857,
|
| 512 |
+
"eval_steps_per_second": 3.563,
|
| 513 |
+
"step": 1600
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.3147249190938513,
|
| 517 |
+
"grad_norm": 0.48737797141075134,
|
| 518 |
+
"learning_rate": 1.698317033499813e-05,
|
| 519 |
+
"loss": 0.1566315174102783,
|
| 520 |
+
"step": 1625
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.3349514563106797,
|
| 524 |
+
"grad_norm": 17.18292808532715,
|
| 525 |
+
"learning_rate": 1.6855756917600057e-05,
|
| 526 |
+
"loss": 0.15158571243286134,
|
| 527 |
+
"step": 1650
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.3551779935275081,
|
| 531 |
+
"grad_norm": 10.572613716125488,
|
| 532 |
+
"learning_rate": 1.6726207979048604e-05,
|
| 533 |
+
"loss": 0.16150398254394532,
|
| 534 |
+
"step": 1675
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.3754045307443366,
|
| 538 |
+
"grad_norm": 0.1977110058069229,
|
| 539 |
+
"learning_rate": 1.6594563872947613e-05,
|
| 540 |
+
"loss": 0.09849663734436036,
|
| 541 |
+
"step": 1700
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.395631067961165,
|
| 545 |
+
"grad_norm": 2.6522152423858643,
|
| 546 |
+
"learning_rate": 1.6460865605531214e-05,
|
| 547 |
+
"loss": 0.17041929244995116,
|
| 548 |
+
"step": 1725
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.4158576051779934,
|
| 552 |
+
"grad_norm": 0.18808063864707947,
|
| 553 |
+
"learning_rate": 1.6325154822890663e-05,
|
| 554 |
+
"loss": 0.1783625602722168,
|
| 555 |
+
"step": 1750
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.436084142394822,
|
| 559 |
+
"grad_norm": 33.89837646484375,
|
| 560 |
+
"learning_rate": 1.618747379800188e-05,
|
| 561 |
+
"loss": 0.15096524238586426,
|
| 562 |
+
"step": 1775
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.4563106796116505,
|
| 566 |
+
"grad_norm": 10.100881576538086,
|
| 567 |
+
"learning_rate": 1.6047865417557695e-05,
|
| 568 |
+
"loss": 0.2288828659057617,
|
| 569 |
+
"step": 1800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.476537216828479,
|
| 573 |
+
"grad_norm": 7.704436302185059,
|
| 574 |
+
"learning_rate": 1.5906373168608952e-05,
|
| 575 |
+
"loss": 0.22033729553222656,
|
| 576 |
+
"step": 1825
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.4967637540453074,
|
| 580 |
+
"grad_norm": 9.382397651672363,
|
| 581 |
+
"learning_rate": 1.576304112501858e-05,
|
| 582 |
+
"loss": 0.15806819915771483,
|
| 583 |
+
"step": 1850
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.516990291262136,
|
| 587 |
+
"grad_norm": 6.415167808532715,
|
| 588 |
+
"learning_rate": 1.5617913933732892e-05,
|
| 589 |
+
"loss": 0.16990087509155274,
|
| 590 |
+
"step": 1875
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.5372168284789645,
|
| 594 |
+
"grad_norm": 8.820036888122559,
|
| 595 |
+
"learning_rate": 1.5471036800874375e-05,
|
| 596 |
+
"loss": 0.1730282974243164,
|
| 597 |
+
"step": 1900
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.557443365695793,
|
| 601 |
+
"grad_norm": 28.52211570739746,
|
| 602 |
+
"learning_rate": 1.5322455477660274e-05,
|
| 603 |
+
"loss": 0.15791168212890624,
|
| 604 |
+
"step": 1925
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.5776699029126213,
|
| 608 |
+
"grad_norm": 15.133056640625,
|
| 609 |
+
"learning_rate": 1.5172216246151426e-05,
|
| 610 |
+
"loss": 0.24158744812011718,
|
| 611 |
+
"step": 1950
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.5978964401294498,
|
| 615 |
+
"grad_norm": 13.767041206359863,
|
| 616 |
+
"learning_rate": 1.50203659048357e-05,
|
| 617 |
+
"loss": 0.07609220504760743,
|
| 618 |
+
"step": 1975
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.6181229773462782,
|
| 622 |
+
"grad_norm": 2.377875328063965,
|
| 623 |
+
"learning_rate": 1.4866951754050605e-05,
|
| 624 |
+
"loss": 0.2856024742126465,
|
| 625 |
+
"step": 2000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.6181229773462782,
|
| 629 |
+
"eval_accuracy": 0.9671551288529561,
|
| 630 |
+
"eval_f1": 0.978369384359401,
|
| 631 |
+
"eval_loss": 0.12881171703338623,
|
| 632 |
+
"eval_precision": 0.9819639278557114,
|
| 633 |
+
"eval_recall": 0.9748010610079576,
|
| 634 |
+
"eval_runtime": 34.7861,
|
| 635 |
+
"eval_samples_per_second": 56.891,
|
| 636 |
+
"eval_steps_per_second": 3.565,
|
| 637 |
+
"step": 2000
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 1.6383495145631068,
|
| 641 |
+
"grad_norm": 1.8338792324066162,
|
| 642 |
+
"learning_rate": 1.4712021581249534e-05,
|
| 643 |
+
"loss": 0.18048593521118164,
|
| 644 |
+
"step": 2025
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 1.6585760517799353,
|
| 648 |
+
"grad_norm": 0.2766644060611725,
|
| 649 |
+
"learning_rate": 1.4555623646116325e-05,
|
| 650 |
+
"loss": 0.1664619827270508,
|
| 651 |
+
"step": 2050
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 1.6788025889967637,
|
| 655 |
+
"grad_norm": 2.5218565464019775,
|
| 656 |
+
"learning_rate": 1.4397806665532693e-05,
|
| 657 |
+
"loss": 0.16300722122192382,
|
| 658 |
+
"step": 2075
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 1.6990291262135924,
|
| 662 |
+
"grad_norm": 0.9285559058189392,
|
| 663 |
+
"learning_rate": 1.4238619798403266e-05,
|
| 664 |
+
"loss": 0.14837112426757812,
|
| 665 |
+
"step": 2100
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 1.7192556634304208,
|
| 669 |
+
"grad_norm": 10.178064346313477,
|
| 670 |
+
"learning_rate": 1.4078112630342891e-05,
|
| 671 |
+
"loss": 0.15740616798400878,
|
| 672 |
+
"step": 2125
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 1.7394822006472492,
|
| 676 |
+
"grad_norm": 19.79058074951172,
|
| 677 |
+
"learning_rate": 1.3916335158231101e-05,
|
| 678 |
+
"loss": 0.22019144058227538,
|
| 679 |
+
"step": 2150
|
| 680 |
+
},
|
| 681 |
+
{
|
| 682 |
+
"epoch": 1.7597087378640777,
|
| 683 |
+
"grad_norm": 8.479653358459473,
|
| 684 |
+
"learning_rate": 1.3753337774638397e-05,
|
| 685 |
+
"loss": 0.10831342697143555,
|
| 686 |
+
"step": 2175
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 1.779935275080906,
|
| 690 |
+
"grad_norm": 0.27186402678489685,
|
| 691 |
+
"learning_rate": 1.3589171252129329e-05,
|
| 692 |
+
"loss": 0.14887014389038086,
|
| 693 |
+
"step": 2200
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 1.8001618122977345,
|
| 697 |
+
"grad_norm": 5.499264240264893,
|
| 698 |
+
"learning_rate": 1.3423886727447176e-05,
|
| 699 |
+
"loss": 0.19739471435546874,
|
| 700 |
+
"step": 2225
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 1.820388349514563,
|
| 704 |
+
"grad_norm": 2.787201166152954,
|
| 705 |
+
"learning_rate": 1.325753568558521e-05,
|
| 706 |
+
"loss": 0.19788772583007813,
|
| 707 |
+
"step": 2250
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 1.8406148867313916,
|
| 711 |
+
"grad_norm": 3.942160129547119,
|
| 712 |
+
"learning_rate": 1.3090169943749475e-05,
|
| 713 |
+
"loss": 0.11929478645324706,
|
| 714 |
+
"step": 2275
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 1.86084142394822,
|
| 718 |
+
"grad_norm": 0.19959710538387299,
|
| 719 |
+
"learning_rate": 1.2921841635218082e-05,
|
| 720 |
+
"loss": 0.09196714401245117,
|
| 721 |
+
"step": 2300
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 1.8810679611650487,
|
| 725 |
+
"grad_norm": 0.05811980366706848,
|
| 726 |
+
"learning_rate": 1.275260319310205e-05,
|
| 727 |
+
"loss": 0.2154999542236328,
|
| 728 |
+
"step": 2325
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 1.9012944983818771,
|
| 732 |
+
"grad_norm": 1.1418527364730835,
|
| 733 |
+
"learning_rate": 1.2582507334012755e-05,
|
| 734 |
+
"loss": 0.1920686912536621,
|
| 735 |
+
"step": 2350
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 1.9215210355987056,
|
| 739 |
+
"grad_norm": 3.8587708473205566,
|
| 740 |
+
"learning_rate": 1.2411607041641062e-05,
|
| 741 |
+
"loss": 0.15039829254150391,
|
| 742 |
+
"step": 2375
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 1.941747572815534,
|
| 746 |
+
"grad_norm": 0.10201486200094223,
|
| 747 |
+
"learning_rate": 1.2239955550253277e-05,
|
| 748 |
+
"loss": 0.1844302558898926,
|
| 749 |
+
"step": 2400
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 1.941747572815534,
|
| 753 |
+
"eval_accuracy": 0.9631126831733199,
|
| 754 |
+
"eval_f1": 0.9760890926957091,
|
| 755 |
+
"eval_loss": 0.14186660945415497,
|
| 756 |
+
"eval_precision": 0.9644012944983819,
|
| 757 |
+
"eval_recall": 0.9880636604774535,
|
| 758 |
+
"eval_runtime": 34.6039,
|
| 759 |
+
"eval_samples_per_second": 57.19,
|
| 760 |
+
"eval_steps_per_second": 3.583,
|
| 761 |
+
"step": 2400
|
| 762 |
+
}
|
| 763 |
+
],
|
| 764 |
+
"logging_steps": 25,
|
| 765 |
+
"max_steps": 4944,
|
| 766 |
+
"num_input_tokens_seen": 0,
|
| 767 |
+
"num_train_epochs": 4,
|
| 768 |
+
"save_steps": 400,
|
| 769 |
+
"stateful_callbacks": {
|
| 770 |
+
"EarlyStoppingCallback": {
|
| 771 |
+
"args": {
|
| 772 |
+
"early_stopping_patience": 2,
|
| 773 |
+
"early_stopping_threshold": 0.0
|
| 774 |
+
},
|
| 775 |
+
"attributes": {
|
| 776 |
+
"early_stopping_patience_counter": 2
|
| 777 |
+
}
|
| 778 |
+
},
|
| 779 |
+
"TrainerControl": {
|
| 780 |
+
"args": {
|
| 781 |
+
"should_epoch_stop": false,
|
| 782 |
+
"should_evaluate": false,
|
| 783 |
+
"should_log": false,
|
| 784 |
+
"should_save": true,
|
| 785 |
+
"should_training_stop": true
|
| 786 |
+
},
|
| 787 |
+
"attributes": {}
|
| 788 |
+
}
|
| 789 |
+
},
|
| 790 |
+
"total_flos": 1.010030719315968e+16,
|
| 791 |
+
"train_batch_size": 16,
|
| 792 |
+
"trial_name": null,
|
| 793 |
+
"trial_params": null
|
| 794 |
+
}
|
checkpoint-2400/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ab81eb8f627d496428c93c1d48c52e0479518be7570628b71c7a023704823dee
|
| 3 |
+
size 5201
|