Upload 8 files
Browse files- config.json +42 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- trainer_state.json +947 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"directionality": "bidi",
|
| 8 |
+
"gradient_checkpointing": false,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 1024,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "LABEL_0",
|
| 14 |
+
"1": "LABEL_1",
|
| 15 |
+
"2": "LABEL_2"
|
| 16 |
+
},
|
| 17 |
+
"initializer_range": 0.02,
|
| 18 |
+
"intermediate_size": 4096,
|
| 19 |
+
"label2id": {
|
| 20 |
+
"LABEL_0": 0,
|
| 21 |
+
"LABEL_1": 1,
|
| 22 |
+
"LABEL_2": 2
|
| 23 |
+
},
|
| 24 |
+
"layer_norm_eps": 1e-12,
|
| 25 |
+
"max_position_embeddings": 512,
|
| 26 |
+
"model_type": "bert",
|
| 27 |
+
"num_attention_heads": 16,
|
| 28 |
+
"num_hidden_layers": 24,
|
| 29 |
+
"pad_token_id": 0,
|
| 30 |
+
"pooler_fc_size": 768,
|
| 31 |
+
"pooler_num_attention_heads": 12,
|
| 32 |
+
"pooler_num_fc_layers": 3,
|
| 33 |
+
"pooler_size_per_head": 128,
|
| 34 |
+
"pooler_type": "first_token_transform",
|
| 35 |
+
"position_embedding_type": "absolute",
|
| 36 |
+
"problem_type": "single_label_classification",
|
| 37 |
+
"torch_dtype": "float32",
|
| 38 |
+
"transformers_version": "4.52.4",
|
| 39 |
+
"type_vocab_size": 2,
|
| 40 |
+
"use_cache": true,
|
| 41 |
+
"vocab_size": 120138
|
| 42 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c7de74fe3938d6788996e2a24689e2c34b93a223e2fe509fe70b4f87429ad4a4
|
| 3 |
+
size 1707694236
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1e2e3c583dbbde9300a4e68e041344db56ac35fef9461798e5fc7805aefc1c32
|
| 3 |
+
size 3415628798
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7f13812f18c42fb4b59edc7c75a0004448ba5eaab5e720aaa64756c19dfcc2af
|
| 3 |
+
size 14645
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f9ae40af54fc54c68dc89dd5c42f0ffe04508dc31726ad9b231173a5fe90ff10
|
| 3 |
+
size 1383
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:122a8d2a0d89629e5d418dcda26c7a4da744789bc4a4fefdfd1281b76e9942a5
|
| 3 |
+
size 1465
|
trainer_state.json
ADDED
|
@@ -0,0 +1,947 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 25817,
|
| 3 |
+
"best_metric": 0.7780532248284215,
|
| 4 |
+
"best_model_checkpoint": "./sbert_large_sentiment/checkpoint-25817",
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 25817,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.007746833481814309,
|
| 14 |
+
"grad_norm": 7.575475215911865,
|
| 15 |
+
"learning_rate": 2.5690679060160085e-07,
|
| 16 |
+
"loss": 1.0936,
|
| 17 |
+
"step": 200
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.015493666963628618,
|
| 21 |
+
"grad_norm": 5.179758548736572,
|
| 22 |
+
"learning_rate": 5.151045701006972e-07,
|
| 23 |
+
"loss": 1.0065,
|
| 24 |
+
"step": 400
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.023240500445442926,
|
| 28 |
+
"grad_norm": 5.142735958099365,
|
| 29 |
+
"learning_rate": 7.733023495997935e-07,
|
| 30 |
+
"loss": 0.9436,
|
| 31 |
+
"step": 600
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.030987333927257235,
|
| 35 |
+
"grad_norm": 6.526517868041992,
|
| 36 |
+
"learning_rate": 1.0315001290988898e-06,
|
| 37 |
+
"loss": 0.8957,
|
| 38 |
+
"step": 800
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.03873416740907154,
|
| 42 |
+
"grad_norm": 8.164371490478516,
|
| 43 |
+
"learning_rate": 1.2896979085979861e-06,
|
| 44 |
+
"loss": 0.7752,
|
| 45 |
+
"step": 1000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.04648100089088585,
|
| 49 |
+
"grad_norm": 9.465094566345215,
|
| 50 |
+
"learning_rate": 1.5478956880970825e-06,
|
| 51 |
+
"loss": 0.7205,
|
| 52 |
+
"step": 1200
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.05422783437270016,
|
| 56 |
+
"grad_norm": 7.948523998260498,
|
| 57 |
+
"learning_rate": 1.806093467596179e-06,
|
| 58 |
+
"loss": 0.6522,
|
| 59 |
+
"step": 1400
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.06197466785451447,
|
| 63 |
+
"grad_norm": 7.701857089996338,
|
| 64 |
+
"learning_rate": 2.064291247095275e-06,
|
| 65 |
+
"loss": 0.6041,
|
| 66 |
+
"step": 1600
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.06972150133632878,
|
| 70 |
+
"grad_norm": 10.935823440551758,
|
| 71 |
+
"learning_rate": 2.3224890265943715e-06,
|
| 72 |
+
"loss": 0.5922,
|
| 73 |
+
"step": 1800
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.07746833481814308,
|
| 77 |
+
"grad_norm": 6.648813247680664,
|
| 78 |
+
"learning_rate": 2.5806868060934675e-06,
|
| 79 |
+
"loss": 0.6016,
|
| 80 |
+
"step": 2000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.08521516829995739,
|
| 84 |
+
"grad_norm": 9.691781997680664,
|
| 85 |
+
"learning_rate": 2.838884585592564e-06,
|
| 86 |
+
"loss": 0.5578,
|
| 87 |
+
"step": 2200
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.0929620017817717,
|
| 91 |
+
"grad_norm": 4.378147125244141,
|
| 92 |
+
"learning_rate": 3.0970823650916602e-06,
|
| 93 |
+
"loss": 0.5731,
|
| 94 |
+
"step": 2400
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.10070883526358601,
|
| 98 |
+
"grad_norm": 10.45535945892334,
|
| 99 |
+
"learning_rate": 3.3552801445907566e-06,
|
| 100 |
+
"loss": 0.5728,
|
| 101 |
+
"step": 2600
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.10845566874540032,
|
| 105 |
+
"grad_norm": 9.362472534179688,
|
| 106 |
+
"learning_rate": 3.613477924089853e-06,
|
| 107 |
+
"loss": 0.5498,
|
| 108 |
+
"step": 2800
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.11620250222721462,
|
| 112 |
+
"grad_norm": 15.068540573120117,
|
| 113 |
+
"learning_rate": 3.87167570358895e-06,
|
| 114 |
+
"loss": 0.5497,
|
| 115 |
+
"step": 3000
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.12394933570902894,
|
| 119 |
+
"grad_norm": 6.796774387359619,
|
| 120 |
+
"learning_rate": 4.129873483088046e-06,
|
| 121 |
+
"loss": 0.5392,
|
| 122 |
+
"step": 3200
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.13169616919084323,
|
| 126 |
+
"grad_norm": 10.11823844909668,
|
| 127 |
+
"learning_rate": 4.388071262587142e-06,
|
| 128 |
+
"loss": 0.5499,
|
| 129 |
+
"step": 3400
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.13944300267265755,
|
| 133 |
+
"grad_norm": 9.95432186126709,
|
| 134 |
+
"learning_rate": 4.646269042086238e-06,
|
| 135 |
+
"loss": 0.5475,
|
| 136 |
+
"step": 3600
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.14718983615447187,
|
| 140 |
+
"grad_norm": 22.89798355102539,
|
| 141 |
+
"learning_rate": 4.9044668215853345e-06,
|
| 142 |
+
"loss": 0.5188,
|
| 143 |
+
"step": 3800
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.15493666963628616,
|
| 147 |
+
"grad_norm": 3.961637496948242,
|
| 148 |
+
"learning_rate": 5.162664601084431e-06,
|
| 149 |
+
"loss": 0.5502,
|
| 150 |
+
"step": 4000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.16268350311810048,
|
| 154 |
+
"grad_norm": 3.6492996215820312,
|
| 155 |
+
"learning_rate": 5.420862380583527e-06,
|
| 156 |
+
"loss": 0.5352,
|
| 157 |
+
"step": 4200
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.17043033659991477,
|
| 161 |
+
"grad_norm": 5.891486644744873,
|
| 162 |
+
"learning_rate": 5.679060160082624e-06,
|
| 163 |
+
"loss": 0.5126,
|
| 164 |
+
"step": 4400
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.1781771700817291,
|
| 168 |
+
"grad_norm": 10.750714302062988,
|
| 169 |
+
"learning_rate": 5.93725793958172e-06,
|
| 170 |
+
"loss": 0.5269,
|
| 171 |
+
"step": 4600
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.1859240035635434,
|
| 175 |
+
"grad_norm": 7.711459636688232,
|
| 176 |
+
"learning_rate": 6.1954557190808165e-06,
|
| 177 |
+
"loss": 0.5504,
|
| 178 |
+
"step": 4800
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.1936708370453577,
|
| 182 |
+
"grad_norm": 11.42162036895752,
|
| 183 |
+
"learning_rate": 6.453653498579913e-06,
|
| 184 |
+
"loss": 0.4935,
|
| 185 |
+
"step": 5000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.20141767052717202,
|
| 189 |
+
"grad_norm": 8.477349281311035,
|
| 190 |
+
"learning_rate": 6.711851278079009e-06,
|
| 191 |
+
"loss": 0.5252,
|
| 192 |
+
"step": 5200
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.20916450400898634,
|
| 196 |
+
"grad_norm": 8.771768569946289,
|
| 197 |
+
"learning_rate": 6.970049057578106e-06,
|
| 198 |
+
"loss": 0.4995,
|
| 199 |
+
"step": 5400
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.21691133749080063,
|
| 203 |
+
"grad_norm": 12.874269485473633,
|
| 204 |
+
"learning_rate": 7.228246837077202e-06,
|
| 205 |
+
"loss": 0.5089,
|
| 206 |
+
"step": 5600
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.22465817097261495,
|
| 210 |
+
"grad_norm": 10.64578628540039,
|
| 211 |
+
"learning_rate": 7.4864446165762984e-06,
|
| 212 |
+
"loss": 0.5227,
|
| 213 |
+
"step": 5800
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.23240500445442924,
|
| 217 |
+
"grad_norm": 5.0879950523376465,
|
| 218 |
+
"learning_rate": 7.744642396075395e-06,
|
| 219 |
+
"loss": 0.4903,
|
| 220 |
+
"step": 6000
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.24015183793624356,
|
| 224 |
+
"grad_norm": 7.7417402267456055,
|
| 225 |
+
"learning_rate": 8.00284017557449e-06,
|
| 226 |
+
"loss": 0.5321,
|
| 227 |
+
"step": 6200
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.24789867141805788,
|
| 231 |
+
"grad_norm": 16.59149932861328,
|
| 232 |
+
"learning_rate": 8.261037955073588e-06,
|
| 233 |
+
"loss": 0.5125,
|
| 234 |
+
"step": 6400
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.2556455048998722,
|
| 238 |
+
"grad_norm": 10.474128723144531,
|
| 239 |
+
"learning_rate": 8.519235734572683e-06,
|
| 240 |
+
"loss": 0.5244,
|
| 241 |
+
"step": 6600
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.26339233838168646,
|
| 245 |
+
"grad_norm": 7.808370113372803,
|
| 246 |
+
"learning_rate": 8.777433514071779e-06,
|
| 247 |
+
"loss": 0.5206,
|
| 248 |
+
"step": 6800
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.2711391718635008,
|
| 252 |
+
"grad_norm": 14.291008949279785,
|
| 253 |
+
"learning_rate": 9.035631293570876e-06,
|
| 254 |
+
"loss": 0.5635,
|
| 255 |
+
"step": 7000
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.2788860053453151,
|
| 259 |
+
"grad_norm": 4.823723793029785,
|
| 260 |
+
"learning_rate": 9.293829073069971e-06,
|
| 261 |
+
"loss": 0.5186,
|
| 262 |
+
"step": 7200
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.2866328388271294,
|
| 266 |
+
"grad_norm": 8.087597846984863,
|
| 267 |
+
"learning_rate": 9.552026852569069e-06,
|
| 268 |
+
"loss": 0.5106,
|
| 269 |
+
"step": 7400
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.29437967230894374,
|
| 273 |
+
"grad_norm": 19.820837020874023,
|
| 274 |
+
"learning_rate": 9.810224632068164e-06,
|
| 275 |
+
"loss": 0.5069,
|
| 276 |
+
"step": 7600
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.302126505790758,
|
| 280 |
+
"grad_norm": 7.311481475830078,
|
| 281 |
+
"learning_rate": 9.992396528226095e-06,
|
| 282 |
+
"loss": 0.5305,
|
| 283 |
+
"step": 7800
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.3098733392725723,
|
| 287 |
+
"grad_norm": 8.929943084716797,
|
| 288 |
+
"learning_rate": 9.963704181909475e-06,
|
| 289 |
+
"loss": 0.5463,
|
| 290 |
+
"step": 8000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.31762017275438664,
|
| 294 |
+
"grad_norm": 9.268566131591797,
|
| 295 |
+
"learning_rate": 9.935011835592857e-06,
|
| 296 |
+
"loss": 0.5328,
|
| 297 |
+
"step": 8200
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.32536700623620096,
|
| 301 |
+
"grad_norm": 9.394050598144531,
|
| 302 |
+
"learning_rate": 9.906319489276236e-06,
|
| 303 |
+
"loss": 0.5097,
|
| 304 |
+
"step": 8400
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.3331138397180153,
|
| 308 |
+
"grad_norm": 6.343189716339111,
|
| 309 |
+
"learning_rate": 9.877627142959616e-06,
|
| 310 |
+
"loss": 0.5063,
|
| 311 |
+
"step": 8600
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.34086067319982954,
|
| 315 |
+
"grad_norm": 8.755598068237305,
|
| 316 |
+
"learning_rate": 9.848934796642996e-06,
|
| 317 |
+
"loss": 0.5202,
|
| 318 |
+
"step": 8800
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.34860750668164386,
|
| 322 |
+
"grad_norm": 13.985782623291016,
|
| 323 |
+
"learning_rate": 9.820242450326375e-06,
|
| 324 |
+
"loss": 0.5019,
|
| 325 |
+
"step": 9000
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.3563543401634582,
|
| 329 |
+
"grad_norm": 7.063718318939209,
|
| 330 |
+
"learning_rate": 9.791550104009757e-06,
|
| 331 |
+
"loss": 0.4844,
|
| 332 |
+
"step": 9200
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.3641011736452725,
|
| 336 |
+
"grad_norm": 15.77436351776123,
|
| 337 |
+
"learning_rate": 9.762857757693137e-06,
|
| 338 |
+
"loss": 0.5007,
|
| 339 |
+
"step": 9400
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.3718480071270868,
|
| 343 |
+
"grad_norm": 8.492551803588867,
|
| 344 |
+
"learning_rate": 9.734165411376516e-06,
|
| 345 |
+
"loss": 0.4676,
|
| 346 |
+
"step": 9600
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.3795948406089011,
|
| 350 |
+
"grad_norm": 13.73719596862793,
|
| 351 |
+
"learning_rate": 9.705473065059896e-06,
|
| 352 |
+
"loss": 0.5103,
|
| 353 |
+
"step": 9800
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.3873416740907154,
|
| 357 |
+
"grad_norm": 3.9511256217956543,
|
| 358 |
+
"learning_rate": 9.676780718743276e-06,
|
| 359 |
+
"loss": 0.4936,
|
| 360 |
+
"step": 10000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.3950885075725297,
|
| 364 |
+
"grad_norm": 7.649303436279297,
|
| 365 |
+
"learning_rate": 9.648088372426655e-06,
|
| 366 |
+
"loss": 0.4997,
|
| 367 |
+
"step": 10200
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.40283534105434404,
|
| 371 |
+
"grad_norm": 12.8671236038208,
|
| 372 |
+
"learning_rate": 9.619396026110037e-06,
|
| 373 |
+
"loss": 0.5093,
|
| 374 |
+
"step": 10400
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.41058217453615836,
|
| 378 |
+
"grad_norm": 9.342642784118652,
|
| 379 |
+
"learning_rate": 9.590703679793416e-06,
|
| 380 |
+
"loss": 0.4908,
|
| 381 |
+
"step": 10600
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.4183290080179727,
|
| 385 |
+
"grad_norm": 10.612739562988281,
|
| 386 |
+
"learning_rate": 9.562011333476796e-06,
|
| 387 |
+
"loss": 0.5141,
|
| 388 |
+
"step": 10800
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.42607584149978694,
|
| 392 |
+
"grad_norm": 5.249755382537842,
|
| 393 |
+
"learning_rate": 9.533318987160176e-06,
|
| 394 |
+
"loss": 0.4738,
|
| 395 |
+
"step": 11000
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.43382267498160126,
|
| 399 |
+
"grad_norm": 15.617774963378906,
|
| 400 |
+
"learning_rate": 9.504626640843556e-06,
|
| 401 |
+
"loss": 0.4782,
|
| 402 |
+
"step": 11200
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.4415695084634156,
|
| 406 |
+
"grad_norm": 20.442564010620117,
|
| 407 |
+
"learning_rate": 9.475934294526937e-06,
|
| 408 |
+
"loss": 0.4772,
|
| 409 |
+
"step": 11400
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.4493163419452299,
|
| 413 |
+
"grad_norm": 5.345826148986816,
|
| 414 |
+
"learning_rate": 9.447241948210317e-06,
|
| 415 |
+
"loss": 0.4764,
|
| 416 |
+
"step": 11600
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.4570631754270442,
|
| 420 |
+
"grad_norm": 21.79194450378418,
|
| 421 |
+
"learning_rate": 9.418549601893696e-06,
|
| 422 |
+
"loss": 0.4892,
|
| 423 |
+
"step": 11800
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.4648100089088585,
|
| 427 |
+
"grad_norm": 15.640652656555176,
|
| 428 |
+
"learning_rate": 9.389857255577074e-06,
|
| 429 |
+
"loss": 0.5199,
|
| 430 |
+
"step": 12000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.4725568423906728,
|
| 434 |
+
"grad_norm": 6.19819974899292,
|
| 435 |
+
"learning_rate": 9.361164909260456e-06,
|
| 436 |
+
"loss": 0.4474,
|
| 437 |
+
"step": 12200
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.4803036758724871,
|
| 441 |
+
"grad_norm": 7.241565704345703,
|
| 442 |
+
"learning_rate": 9.332472562943835e-06,
|
| 443 |
+
"loss": 0.4672,
|
| 444 |
+
"step": 12400
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.48805050935430144,
|
| 448 |
+
"grad_norm": 9.49963092803955,
|
| 449 |
+
"learning_rate": 9.303780216627215e-06,
|
| 450 |
+
"loss": 0.4701,
|
| 451 |
+
"step": 12600
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.49579734283611576,
|
| 455 |
+
"grad_norm": 16.092571258544922,
|
| 456 |
+
"learning_rate": 9.275087870310595e-06,
|
| 457 |
+
"loss": 0.4628,
|
| 458 |
+
"step": 12800
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.50354417631793,
|
| 462 |
+
"grad_norm": 10.93955135345459,
|
| 463 |
+
"learning_rate": 9.246395523993975e-06,
|
| 464 |
+
"loss": 0.474,
|
| 465 |
+
"step": 13000
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.5112910097997444,
|
| 469 |
+
"grad_norm": 9.083719253540039,
|
| 470 |
+
"learning_rate": 9.217703177677354e-06,
|
| 471 |
+
"loss": 0.4771,
|
| 472 |
+
"step": 13200
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.5190378432815587,
|
| 476 |
+
"grad_norm": 9.085740089416504,
|
| 477 |
+
"learning_rate": 9.189010831360736e-06,
|
| 478 |
+
"loss": 0.4803,
|
| 479 |
+
"step": 13400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.5267846767633729,
|
| 483 |
+
"grad_norm": 17.334964752197266,
|
| 484 |
+
"learning_rate": 9.160318485044115e-06,
|
| 485 |
+
"loss": 0.478,
|
| 486 |
+
"step": 13600
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.5345315102451873,
|
| 490 |
+
"grad_norm": 4.976545810699463,
|
| 491 |
+
"learning_rate": 9.131626138727495e-06,
|
| 492 |
+
"loss": 0.5116,
|
| 493 |
+
"step": 13800
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.5422783437270016,
|
| 497 |
+
"grad_norm": 8.329999923706055,
|
| 498 |
+
"learning_rate": 9.102933792410875e-06,
|
| 499 |
+
"loss": 0.4831,
|
| 500 |
+
"step": 14000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.5500251772088159,
|
| 504 |
+
"grad_norm": 10.104872703552246,
|
| 505 |
+
"learning_rate": 9.074241446094255e-06,
|
| 506 |
+
"loss": 0.4905,
|
| 507 |
+
"step": 14200
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.5577720106906302,
|
| 511 |
+
"grad_norm": 8.581650733947754,
|
| 512 |
+
"learning_rate": 9.045549099777636e-06,
|
| 513 |
+
"loss": 0.4656,
|
| 514 |
+
"step": 14400
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.5655188441724445,
|
| 518 |
+
"grad_norm": 8.831694602966309,
|
| 519 |
+
"learning_rate": 9.016856753461016e-06,
|
| 520 |
+
"loss": 0.455,
|
| 521 |
+
"step": 14600
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.5732656776542588,
|
| 525 |
+
"grad_norm": 12.80400276184082,
|
| 526 |
+
"learning_rate": 8.988164407144395e-06,
|
| 527 |
+
"loss": 0.4556,
|
| 528 |
+
"step": 14800
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.5810125111360731,
|
| 532 |
+
"grad_norm": 10.323575019836426,
|
| 533 |
+
"learning_rate": 8.959472060827775e-06,
|
| 534 |
+
"loss": 0.4952,
|
| 535 |
+
"step": 15000
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.5887593446178875,
|
| 539 |
+
"grad_norm": 8.978657722473145,
|
| 540 |
+
"learning_rate": 8.930779714511155e-06,
|
| 541 |
+
"loss": 0.4774,
|
| 542 |
+
"step": 15200
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.5965061780997017,
|
| 546 |
+
"grad_norm": 9.755346298217773,
|
| 547 |
+
"learning_rate": 8.902087368194534e-06,
|
| 548 |
+
"loss": 0.4607,
|
| 549 |
+
"step": 15400
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.604253011581516,
|
| 553 |
+
"grad_norm": 4.952932357788086,
|
| 554 |
+
"learning_rate": 8.873395021877916e-06,
|
| 555 |
+
"loss": 0.4931,
|
| 556 |
+
"step": 15600
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.6119998450633304,
|
| 560 |
+
"grad_norm": 14.65958309173584,
|
| 561 |
+
"learning_rate": 8.844702675561296e-06,
|
| 562 |
+
"loss": 0.4723,
|
| 563 |
+
"step": 15800
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.6197466785451446,
|
| 567 |
+
"grad_norm": 12.291789054870605,
|
| 568 |
+
"learning_rate": 8.816010329244675e-06,
|
| 569 |
+
"loss": 0.4916,
|
| 570 |
+
"step": 16000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.627493512026959,
|
| 574 |
+
"grad_norm": 6.109698295593262,
|
| 575 |
+
"learning_rate": 8.787317982928055e-06,
|
| 576 |
+
"loss": 0.4749,
|
| 577 |
+
"step": 16200
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.6352403455087733,
|
| 581 |
+
"grad_norm": 9.7838134765625,
|
| 582 |
+
"learning_rate": 8.758625636611435e-06,
|
| 583 |
+
"loss": 0.4621,
|
| 584 |
+
"step": 16400
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.6429871789905875,
|
| 588 |
+
"grad_norm": 6.760039806365967,
|
| 589 |
+
"learning_rate": 8.729933290294814e-06,
|
| 590 |
+
"loss": 0.4806,
|
| 591 |
+
"step": 16600
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.6507340124724019,
|
| 595 |
+
"grad_norm": 10.97524642944336,
|
| 596 |
+
"learning_rate": 8.701240943978194e-06,
|
| 597 |
+
"loss": 0.4591,
|
| 598 |
+
"step": 16800
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.6584808459542162,
|
| 602 |
+
"grad_norm": 9.859977722167969,
|
| 603 |
+
"learning_rate": 8.672548597661574e-06,
|
| 604 |
+
"loss": 0.4742,
|
| 605 |
+
"step": 17000
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.6662276794360306,
|
| 609 |
+
"grad_norm": 11.13088321685791,
|
| 610 |
+
"learning_rate": 8.643856251344953e-06,
|
| 611 |
+
"loss": 0.4593,
|
| 612 |
+
"step": 17200
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.6739745129178448,
|
| 616 |
+
"grad_norm": 8.611045837402344,
|
| 617 |
+
"learning_rate": 8.615163905028335e-06,
|
| 618 |
+
"loss": 0.4711,
|
| 619 |
+
"step": 17400
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.6817213463996591,
|
| 623 |
+
"grad_norm": 9.801880836486816,
|
| 624 |
+
"learning_rate": 8.586471558711715e-06,
|
| 625 |
+
"loss": 0.4639,
|
| 626 |
+
"step": 17600
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.6894681798814735,
|
| 630 |
+
"grad_norm": 6.184256553649902,
|
| 631 |
+
"learning_rate": 8.557779212395094e-06,
|
| 632 |
+
"loss": 0.4921,
|
| 633 |
+
"step": 17800
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.6972150133632877,
|
| 637 |
+
"grad_norm": 9.72716999053955,
|
| 638 |
+
"learning_rate": 8.529086866078474e-06,
|
| 639 |
+
"loss": 0.4829,
|
| 640 |
+
"step": 18000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.7049618468451021,
|
| 644 |
+
"grad_norm": 7.580606937408447,
|
| 645 |
+
"learning_rate": 8.500394519761854e-06,
|
| 646 |
+
"loss": 0.4638,
|
| 647 |
+
"step": 18200
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.7127086803269164,
|
| 651 |
+
"grad_norm": 12.59068775177002,
|
| 652 |
+
"learning_rate": 8.471702173445235e-06,
|
| 653 |
+
"loss": 0.4818,
|
| 654 |
+
"step": 18400
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.7204555138087306,
|
| 658 |
+
"grad_norm": 5.276298522949219,
|
| 659 |
+
"learning_rate": 8.443009827128615e-06,
|
| 660 |
+
"loss": 0.4736,
|
| 661 |
+
"step": 18600
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.728202347290545,
|
| 665 |
+
"grad_norm": 17.732887268066406,
|
| 666 |
+
"learning_rate": 8.414317480811995e-06,
|
| 667 |
+
"loss": 0.4515,
|
| 668 |
+
"step": 18800
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.7359491807723593,
|
| 672 |
+
"grad_norm": 6.198581695556641,
|
| 673 |
+
"learning_rate": 8.385625134495374e-06,
|
| 674 |
+
"loss": 0.4766,
|
| 675 |
+
"step": 19000
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.7436960142541736,
|
| 679 |
+
"grad_norm": 5.835597038269043,
|
| 680 |
+
"learning_rate": 8.356932788178754e-06,
|
| 681 |
+
"loss": 0.4506,
|
| 682 |
+
"step": 19200
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.7514428477359879,
|
| 686 |
+
"grad_norm": 10.405342102050781,
|
| 687 |
+
"learning_rate": 8.328240441862134e-06,
|
| 688 |
+
"loss": 0.4703,
|
| 689 |
+
"step": 19400
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.7591896812178022,
|
| 693 |
+
"grad_norm": 10.531950950622559,
|
| 694 |
+
"learning_rate": 8.299548095545515e-06,
|
| 695 |
+
"loss": 0.471,
|
| 696 |
+
"step": 19600
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.7669365146996165,
|
| 700 |
+
"grad_norm": 11.877711296081543,
|
| 701 |
+
"learning_rate": 8.270855749228895e-06,
|
| 702 |
+
"loss": 0.4678,
|
| 703 |
+
"step": 19800
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.7746833481814308,
|
| 707 |
+
"grad_norm": 7.640944957733154,
|
| 708 |
+
"learning_rate": 8.242163402912274e-06,
|
| 709 |
+
"loss": 0.4662,
|
| 710 |
+
"step": 20000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.7824301816632452,
|
| 714 |
+
"grad_norm": 17.27083969116211,
|
| 715 |
+
"learning_rate": 8.213471056595652e-06,
|
| 716 |
+
"loss": 0.4871,
|
| 717 |
+
"step": 20200
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.7901770151450594,
|
| 721 |
+
"grad_norm": 3.0645809173583984,
|
| 722 |
+
"learning_rate": 8.184778710279034e-06,
|
| 723 |
+
"loss": 0.4574,
|
| 724 |
+
"step": 20400
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.7979238486268737,
|
| 728 |
+
"grad_norm": 4.7715935707092285,
|
| 729 |
+
"learning_rate": 8.156086363962414e-06,
|
| 730 |
+
"loss": 0.4433,
|
| 731 |
+
"step": 20600
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.8056706821086881,
|
| 735 |
+
"grad_norm": 4.5974202156066895,
|
| 736 |
+
"learning_rate": 8.127394017645793e-06,
|
| 737 |
+
"loss": 0.4439,
|
| 738 |
+
"step": 20800
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.8134175155905023,
|
| 742 |
+
"grad_norm": 13.890436172485352,
|
| 743 |
+
"learning_rate": 8.098701671329173e-06,
|
| 744 |
+
"loss": 0.4354,
|
| 745 |
+
"step": 21000
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.8211643490723167,
|
| 749 |
+
"grad_norm": 27.995084762573242,
|
| 750 |
+
"learning_rate": 8.070009325012553e-06,
|
| 751 |
+
"loss": 0.4115,
|
| 752 |
+
"step": 21200
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.828911182554131,
|
| 756 |
+
"grad_norm": 2.1986517906188965,
|
| 757 |
+
"learning_rate": 8.041316978695934e-06,
|
| 758 |
+
"loss": 0.4243,
|
| 759 |
+
"step": 21400
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.8366580160359454,
|
| 763 |
+
"grad_norm": 17.675518035888672,
|
| 764 |
+
"learning_rate": 8.012624632379314e-06,
|
| 765 |
+
"loss": 0.495,
|
| 766 |
+
"step": 21600
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.8444048495177596,
|
| 770 |
+
"grad_norm": 7.1354146003723145,
|
| 771 |
+
"learning_rate": 7.983932286062693e-06,
|
| 772 |
+
"loss": 0.4688,
|
| 773 |
+
"step": 21800
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.8521516829995739,
|
| 777 |
+
"grad_norm": 5.76608943939209,
|
| 778 |
+
"learning_rate": 7.955239939746073e-06,
|
| 779 |
+
"loss": 0.4364,
|
| 780 |
+
"step": 22000
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.8598985164813883,
|
| 784 |
+
"grad_norm": 15.63326358795166,
|
| 785 |
+
"learning_rate": 7.926547593429453e-06,
|
| 786 |
+
"loss": 0.4496,
|
| 787 |
+
"step": 22200
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.8676453499632025,
|
| 791 |
+
"grad_norm": 10.311728477478027,
|
| 792 |
+
"learning_rate": 7.897855247112833e-06,
|
| 793 |
+
"loss": 0.4649,
|
| 794 |
+
"step": 22400
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.8753921834450169,
|
| 798 |
+
"grad_norm": 5.432675838470459,
|
| 799 |
+
"learning_rate": 7.869162900796214e-06,
|
| 800 |
+
"loss": 0.4307,
|
| 801 |
+
"step": 22600
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.8831390169268312,
|
| 805 |
+
"grad_norm": 6.74233341217041,
|
| 806 |
+
"learning_rate": 7.840470554479594e-06,
|
| 807 |
+
"loss": 0.4629,
|
| 808 |
+
"step": 22800
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.8908858504086454,
|
| 812 |
+
"grad_norm": 9.045225143432617,
|
| 813 |
+
"learning_rate": 7.811778208162973e-06,
|
| 814 |
+
"loss": 0.4742,
|
| 815 |
+
"step": 23000
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.8986326838904598,
|
| 819 |
+
"grad_norm": 16.038272857666016,
|
| 820 |
+
"learning_rate": 7.783085861846353e-06,
|
| 821 |
+
"loss": 0.4335,
|
| 822 |
+
"step": 23200
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.9063795173722741,
|
| 826 |
+
"grad_norm": 11.987189292907715,
|
| 827 |
+
"learning_rate": 7.754393515529733e-06,
|
| 828 |
+
"loss": 0.4534,
|
| 829 |
+
"step": 23400
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.9141263508540884,
|
| 833 |
+
"grad_norm": 2.1488113403320312,
|
| 834 |
+
"learning_rate": 7.725701169213114e-06,
|
| 835 |
+
"loss": 0.4566,
|
| 836 |
+
"step": 23600
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.9218731843359027,
|
| 840 |
+
"grad_norm": 0.6424712538719177,
|
| 841 |
+
"learning_rate": 7.697008822896494e-06,
|
| 842 |
+
"loss": 0.4397,
|
| 843 |
+
"step": 23800
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.929620017817717,
|
| 847 |
+
"grad_norm": 9.486328125,
|
| 848 |
+
"learning_rate": 7.668316476579874e-06,
|
| 849 |
+
"loss": 0.4564,
|
| 850 |
+
"step": 24000
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.9373668512995313,
|
| 854 |
+
"grad_norm": 9.135612487792969,
|
| 855 |
+
"learning_rate": 7.639624130263253e-06,
|
| 856 |
+
"loss": 0.4608,
|
| 857 |
+
"step": 24200
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.9451136847813456,
|
| 861 |
+
"grad_norm": 9.029398918151855,
|
| 862 |
+
"learning_rate": 7.610931783946632e-06,
|
| 863 |
+
"loss": 0.4596,
|
| 864 |
+
"step": 24400
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.95286051826316,
|
| 868 |
+
"grad_norm": 5.138970851898193,
|
| 869 |
+
"learning_rate": 7.582239437630013e-06,
|
| 870 |
+
"loss": 0.4445,
|
| 871 |
+
"step": 24600
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.9606073517449742,
|
| 875 |
+
"grad_norm": 8.201314926147461,
|
| 876 |
+
"learning_rate": 7.5535470913133925e-06,
|
| 877 |
+
"loss": 0.4514,
|
| 878 |
+
"step": 24800
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.9683541852267885,
|
| 882 |
+
"grad_norm": 5.1406145095825195,
|
| 883 |
+
"learning_rate": 7.524854744996772e-06,
|
| 884 |
+
"loss": 0.4546,
|
| 885 |
+
"step": 25000
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.9761010187086029,
|
| 889 |
+
"grad_norm": 7.5569353103637695,
|
| 890 |
+
"learning_rate": 7.496162398680153e-06,
|
| 891 |
+
"loss": 0.4465,
|
| 892 |
+
"step": 25200
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.9838478521904171,
|
| 896 |
+
"grad_norm": 5.765118598937988,
|
| 897 |
+
"learning_rate": 7.4674700523635324e-06,
|
| 898 |
+
"loss": 0.4355,
|
| 899 |
+
"step": 25400
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.9915946856722315,
|
| 903 |
+
"grad_norm": 7.945804595947266,
|
| 904 |
+
"learning_rate": 7.438777706046912e-06,
|
| 905 |
+
"loss": 0.4694,
|
| 906 |
+
"step": 25600
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.9993415191540458,
|
| 910 |
+
"grad_norm": 10.627592086791992,
|
| 911 |
+
"learning_rate": 7.410085359730293e-06,
|
| 912 |
+
"loss": 0.5065,
|
| 913 |
+
"step": 25800
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.0,
|
| 917 |
+
"eval_accuracy": 0.8090112858948102,
|
| 918 |
+
"eval_loss": 0.4515499174594879,
|
| 919 |
+
"eval_macro_f1": 0.7780532248284215,
|
| 920 |
+
"eval_runtime": 57.5502,
|
| 921 |
+
"eval_samples_per_second": 398.765,
|
| 922 |
+
"eval_steps_per_second": 49.852,
|
| 923 |
+
"step": 25817
|
| 924 |
+
}
|
| 925 |
+
],
|
| 926 |
+
"logging_steps": 200,
|
| 927 |
+
"max_steps": 77451,
|
| 928 |
+
"num_input_tokens_seen": 0,
|
| 929 |
+
"num_train_epochs": 3,
|
| 930 |
+
"save_steps": 500,
|
| 931 |
+
"stateful_callbacks": {
|
| 932 |
+
"TrainerControl": {
|
| 933 |
+
"args": {
|
| 934 |
+
"should_epoch_stop": false,
|
| 935 |
+
"should_evaluate": false,
|
| 936 |
+
"should_log": false,
|
| 937 |
+
"should_save": true,
|
| 938 |
+
"should_training_stop": false
|
| 939 |
+
},
|
| 940 |
+
"attributes": {}
|
| 941 |
+
}
|
| 942 |
+
},
|
| 943 |
+
"total_flos": 4.811880764833306e+16,
|
| 944 |
+
"train_batch_size": 8,
|
| 945 |
+
"trial_name": null,
|
| 946 |
+
"trial_params": null
|
| 947 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:47753f608345e6ebeec215278254e6a2cfb468d428be3e7ae66a34e212b3c15a
|
| 3 |
+
size 5713
|