Training in progress, epoch 4, checkpoint
Browse files
checkpoint-2770/config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "bert-base-uncased",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"gradient_checkpointing": false,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 3072,
|
| 14 |
+
"layer_norm_eps": 1e-12,
|
| 15 |
+
"max_position_embeddings": 512,
|
| 16 |
+
"model_type": "bert",
|
| 17 |
+
"num_attention_heads": 12,
|
| 18 |
+
"num_hidden_layers": 12,
|
| 19 |
+
"pad_token_id": 0,
|
| 20 |
+
"position_embedding_type": "absolute",
|
| 21 |
+
"torch_dtype": "float32",
|
| 22 |
+
"transformers_version": "4.47.0",
|
| 23 |
+
"type_vocab_size": 2,
|
| 24 |
+
"use_cache": true,
|
| 25 |
+
"vocab_size": 30522
|
| 26 |
+
}
|
checkpoint-2770/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:53c5fe79bbf518d643867451985234b558ec4eaa25d9a66001e599901ad5c8a4
|
| 3 |
+
size 437958648
|
checkpoint-2770/trainer_state.json
ADDED
|
@@ -0,0 +1,2027 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.991437584497522,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2770,
|
| 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": 0.6562770562770562,
|
| 14 |
+
"eval_loss": 0.7476533651351929,
|
| 15 |
+
"eval_precision": 0.48976954555244456,
|
| 16 |
+
"eval_recall": 0.9943156974202011,
|
| 17 |
+
"eval_runtime": 16.4688,
|
| 18 |
+
"eval_samples_per_second": 283.749,
|
| 19 |
+
"eval_steps_per_second": 35.522,
|
| 20 |
+
"step": 0
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.018026137899954935,
|
| 24 |
+
"grad_norm": 14.20566463470459,
|
| 25 |
+
"learning_rate": 4.981949458483755e-05,
|
| 26 |
+
"loss": 0.6876,
|
| 27 |
+
"step": 10
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.03605227579990987,
|
| 31 |
+
"grad_norm": 27.79998016357422,
|
| 32 |
+
"learning_rate": 4.963898916967509e-05,
|
| 33 |
+
"loss": 0.5988,
|
| 34 |
+
"step": 20
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 0.054078413699864804,
|
| 38 |
+
"grad_norm": 39.63682556152344,
|
| 39 |
+
"learning_rate": 4.945848375451264e-05,
|
| 40 |
+
"loss": 0.5677,
|
| 41 |
+
"step": 30
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.07210455159981974,
|
| 45 |
+
"grad_norm": 38.49867630004883,
|
| 46 |
+
"learning_rate": 4.927797833935018e-05,
|
| 47 |
+
"loss": 0.5922,
|
| 48 |
+
"step": 40
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.09013068949977468,
|
| 52 |
+
"grad_norm": 52.28614044189453,
|
| 53 |
+
"learning_rate": 4.909747292418773e-05,
|
| 54 |
+
"loss": 0.5506,
|
| 55 |
+
"step": 50
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 0.10815682739972961,
|
| 59 |
+
"grad_norm": 29.180490493774414,
|
| 60 |
+
"learning_rate": 4.891696750902527e-05,
|
| 61 |
+
"loss": 0.5173,
|
| 62 |
+
"step": 60
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.12618296529968454,
|
| 66 |
+
"grad_norm": 33.55237579345703,
|
| 67 |
+
"learning_rate": 4.873646209386282e-05,
|
| 68 |
+
"loss": 0.4931,
|
| 69 |
+
"step": 70
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.14420910319963948,
|
| 73 |
+
"grad_norm": 32.063228607177734,
|
| 74 |
+
"learning_rate": 4.855595667870036e-05,
|
| 75 |
+
"loss": 0.483,
|
| 76 |
+
"step": 80
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.16223524109959442,
|
| 80 |
+
"grad_norm": 24.234926223754883,
|
| 81 |
+
"learning_rate": 4.837545126353791e-05,
|
| 82 |
+
"loss": 0.47,
|
| 83 |
+
"step": 90
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.18026137899954936,
|
| 87 |
+
"grad_norm": 33.66775894165039,
|
| 88 |
+
"learning_rate": 4.819494584837546e-05,
|
| 89 |
+
"loss": 0.4564,
|
| 90 |
+
"step": 100
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.19828751689950427,
|
| 94 |
+
"grad_norm": 53.322574615478516,
|
| 95 |
+
"learning_rate": 4.8014440433213e-05,
|
| 96 |
+
"loss": 0.5083,
|
| 97 |
+
"step": 110
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.21631365479945922,
|
| 101 |
+
"grad_norm": 29.443470001220703,
|
| 102 |
+
"learning_rate": 4.783393501805055e-05,
|
| 103 |
+
"loss": 0.463,
|
| 104 |
+
"step": 120
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.23433979269941416,
|
| 108 |
+
"grad_norm": 33.13877487182617,
|
| 109 |
+
"learning_rate": 4.765342960288809e-05,
|
| 110 |
+
"loss": 0.4626,
|
| 111 |
+
"step": 130
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.25236593059936907,
|
| 115 |
+
"grad_norm": 25.509536743164062,
|
| 116 |
+
"learning_rate": 4.747292418772563e-05,
|
| 117 |
+
"loss": 0.5113,
|
| 118 |
+
"step": 140
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 0.270392068499324,
|
| 122 |
+
"grad_norm": 21.015031814575195,
|
| 123 |
+
"learning_rate": 4.7292418772563177e-05,
|
| 124 |
+
"loss": 0.4611,
|
| 125 |
+
"step": 150
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.28841820639927895,
|
| 129 |
+
"grad_norm": 28.079387664794922,
|
| 130 |
+
"learning_rate": 4.711191335740072e-05,
|
| 131 |
+
"loss": 0.4783,
|
| 132 |
+
"step": 160
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 0.3064443442992339,
|
| 136 |
+
"grad_norm": 19.978599548339844,
|
| 137 |
+
"learning_rate": 4.693140794223827e-05,
|
| 138 |
+
"loss": 0.5026,
|
| 139 |
+
"step": 170
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 0.32447048219918884,
|
| 143 |
+
"grad_norm": 29.316265106201172,
|
| 144 |
+
"learning_rate": 4.675090252707581e-05,
|
| 145 |
+
"loss": 0.4906,
|
| 146 |
+
"step": 180
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.3424966200991438,
|
| 150 |
+
"grad_norm": 18.79282569885254,
|
| 151 |
+
"learning_rate": 4.657039711191336e-05,
|
| 152 |
+
"loss": 0.5016,
|
| 153 |
+
"step": 190
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.3605227579990987,
|
| 157 |
+
"grad_norm": 18.66801643371582,
|
| 158 |
+
"learning_rate": 4.63898916967509e-05,
|
| 159 |
+
"loss": 0.4502,
|
| 160 |
+
"step": 200
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.3785488958990536,
|
| 164 |
+
"grad_norm": 27.929332733154297,
|
| 165 |
+
"learning_rate": 4.620938628158845e-05,
|
| 166 |
+
"loss": 0.4302,
|
| 167 |
+
"step": 210
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 0.39657503379900855,
|
| 171 |
+
"grad_norm": 20.541879653930664,
|
| 172 |
+
"learning_rate": 4.602888086642599e-05,
|
| 173 |
+
"loss": 0.4496,
|
| 174 |
+
"step": 220
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.4146011716989635,
|
| 178 |
+
"grad_norm": 19.034873962402344,
|
| 179 |
+
"learning_rate": 4.584837545126354e-05,
|
| 180 |
+
"loss": 0.4541,
|
| 181 |
+
"step": 230
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 0.43262730959891843,
|
| 185 |
+
"grad_norm": 32.077945709228516,
|
| 186 |
+
"learning_rate": 4.566787003610109e-05,
|
| 187 |
+
"loss": 0.4562,
|
| 188 |
+
"step": 240
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.45065344749887337,
|
| 192 |
+
"grad_norm": 25.475997924804688,
|
| 193 |
+
"learning_rate": 4.548736462093863e-05,
|
| 194 |
+
"loss": 0.4456,
|
| 195 |
+
"step": 250
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.4686795853988283,
|
| 199 |
+
"grad_norm": 17.132720947265625,
|
| 200 |
+
"learning_rate": 4.530685920577618e-05,
|
| 201 |
+
"loss": 0.4629,
|
| 202 |
+
"step": 260
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.48670572329878325,
|
| 206 |
+
"grad_norm": 22.313941955566406,
|
| 207 |
+
"learning_rate": 4.5126353790613716e-05,
|
| 208 |
+
"loss": 0.4444,
|
| 209 |
+
"step": 270
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.5047318611987381,
|
| 213 |
+
"grad_norm": 22.373477935791016,
|
| 214 |
+
"learning_rate": 4.494584837545127e-05,
|
| 215 |
+
"loss": 0.4404,
|
| 216 |
+
"step": 280
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 0.5227579990986931,
|
| 220 |
+
"grad_norm": 16.72430992126465,
|
| 221 |
+
"learning_rate": 4.4765342960288806e-05,
|
| 222 |
+
"loss": 0.4611,
|
| 223 |
+
"step": 290
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 0.540784136998648,
|
| 227 |
+
"grad_norm": 27.037927627563477,
|
| 228 |
+
"learning_rate": 4.458483754512636e-05,
|
| 229 |
+
"loss": 0.4272,
|
| 230 |
+
"step": 300
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.558810274898603,
|
| 234 |
+
"grad_norm": 31.637985229492188,
|
| 235 |
+
"learning_rate": 4.44043321299639e-05,
|
| 236 |
+
"loss": 0.448,
|
| 237 |
+
"step": 310
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.5768364127985579,
|
| 241 |
+
"grad_norm": 35.010459899902344,
|
| 242 |
+
"learning_rate": 4.422382671480145e-05,
|
| 243 |
+
"loss": 0.4215,
|
| 244 |
+
"step": 320
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.5948625506985128,
|
| 248 |
+
"grad_norm": 26.25156021118164,
|
| 249 |
+
"learning_rate": 4.404332129963899e-05,
|
| 250 |
+
"loss": 0.4396,
|
| 251 |
+
"step": 330
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.6128886885984678,
|
| 255 |
+
"grad_norm": 19.4619083404541,
|
| 256 |
+
"learning_rate": 4.386281588447654e-05,
|
| 257 |
+
"loss": 0.4378,
|
| 258 |
+
"step": 340
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.6309148264984227,
|
| 262 |
+
"grad_norm": 19.16460609436035,
|
| 263 |
+
"learning_rate": 4.368231046931408e-05,
|
| 264 |
+
"loss": 0.4366,
|
| 265 |
+
"step": 350
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.6489409643983777,
|
| 269 |
+
"grad_norm": 30.745384216308594,
|
| 270 |
+
"learning_rate": 4.350180505415163e-05,
|
| 271 |
+
"loss": 0.4062,
|
| 272 |
+
"step": 360
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.6669671022983326,
|
| 276 |
+
"grad_norm": 16.873245239257812,
|
| 277 |
+
"learning_rate": 4.332129963898917e-05,
|
| 278 |
+
"loss": 0.439,
|
| 279 |
+
"step": 370
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.6849932401982876,
|
| 283 |
+
"grad_norm": 25.771486282348633,
|
| 284 |
+
"learning_rate": 4.314079422382672e-05,
|
| 285 |
+
"loss": 0.4053,
|
| 286 |
+
"step": 380
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 0.7030193780982424,
|
| 290 |
+
"grad_norm": 14.48513412475586,
|
| 291 |
+
"learning_rate": 4.296028880866426e-05,
|
| 292 |
+
"loss": 0.4164,
|
| 293 |
+
"step": 390
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 0.7210455159981974,
|
| 297 |
+
"grad_norm": 34.993465423583984,
|
| 298 |
+
"learning_rate": 4.277978339350181e-05,
|
| 299 |
+
"loss": 0.4438,
|
| 300 |
+
"step": 400
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 0.7390716538981523,
|
| 304 |
+
"grad_norm": 22.576608657836914,
|
| 305 |
+
"learning_rate": 4.259927797833935e-05,
|
| 306 |
+
"loss": 0.4518,
|
| 307 |
+
"step": 410
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.7570977917981072,
|
| 311 |
+
"grad_norm": 19.088132858276367,
|
| 312 |
+
"learning_rate": 4.24187725631769e-05,
|
| 313 |
+
"loss": 0.4495,
|
| 314 |
+
"step": 420
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.7751239296980622,
|
| 318 |
+
"grad_norm": 35.2998046875,
|
| 319 |
+
"learning_rate": 4.223826714801444e-05,
|
| 320 |
+
"loss": 0.3897,
|
| 321 |
+
"step": 430
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.7931500675980171,
|
| 325 |
+
"grad_norm": 20.395572662353516,
|
| 326 |
+
"learning_rate": 4.205776173285199e-05,
|
| 327 |
+
"loss": 0.4818,
|
| 328 |
+
"step": 440
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.8111762054979721,
|
| 332 |
+
"grad_norm": 21.174386978149414,
|
| 333 |
+
"learning_rate": 4.187725631768953e-05,
|
| 334 |
+
"loss": 0.4103,
|
| 335 |
+
"step": 450
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.829202343397927,
|
| 339 |
+
"grad_norm": 32.97265625,
|
| 340 |
+
"learning_rate": 4.169675090252708e-05,
|
| 341 |
+
"loss": 0.4036,
|
| 342 |
+
"step": 460
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.847228481297882,
|
| 346 |
+
"grad_norm": 23.78690528869629,
|
| 347 |
+
"learning_rate": 4.151624548736462e-05,
|
| 348 |
+
"loss": 0.4079,
|
| 349 |
+
"step": 470
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.8652546191978369,
|
| 353 |
+
"grad_norm": 17.087247848510742,
|
| 354 |
+
"learning_rate": 4.1335740072202167e-05,
|
| 355 |
+
"loss": 0.4371,
|
| 356 |
+
"step": 480
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.8832807570977917,
|
| 360 |
+
"grad_norm": 13.810160636901855,
|
| 361 |
+
"learning_rate": 4.115523465703972e-05,
|
| 362 |
+
"loss": 0.4249,
|
| 363 |
+
"step": 490
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.9013068949977467,
|
| 367 |
+
"grad_norm": 20.197996139526367,
|
| 368 |
+
"learning_rate": 4.0974729241877256e-05,
|
| 369 |
+
"loss": 0.3991,
|
| 370 |
+
"step": 500
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.9193330328977016,
|
| 374 |
+
"grad_norm": 33.1153564453125,
|
| 375 |
+
"learning_rate": 4.079422382671481e-05,
|
| 376 |
+
"loss": 0.3994,
|
| 377 |
+
"step": 510
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.9373591707976566,
|
| 381 |
+
"grad_norm": 18.20081901550293,
|
| 382 |
+
"learning_rate": 4.0613718411552346e-05,
|
| 383 |
+
"loss": 0.4176,
|
| 384 |
+
"step": 520
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.9553853086976115,
|
| 388 |
+
"grad_norm": 20.09193992614746,
|
| 389 |
+
"learning_rate": 4.043321299638989e-05,
|
| 390 |
+
"loss": 0.4297,
|
| 391 |
+
"step": 530
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 0.9734114465975665,
|
| 395 |
+
"grad_norm": 15.939692497253418,
|
| 396 |
+
"learning_rate": 4.0252707581227436e-05,
|
| 397 |
+
"loss": 0.4309,
|
| 398 |
+
"step": 540
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.9914375844975214,
|
| 402 |
+
"grad_norm": 24.811601638793945,
|
| 403 |
+
"learning_rate": 4.007220216606498e-05,
|
| 404 |
+
"loss": 0.3908,
|
| 405 |
+
"step": 550
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 1.0,
|
| 409 |
+
"eval_f1": 0.8367177786987153,
|
| 410 |
+
"eval_loss": 0.37827184796333313,
|
| 411 |
+
"eval_precision": 0.7951949586451359,
|
| 412 |
+
"eval_recall": 0.8828159160472234,
|
| 413 |
+
"eval_runtime": 15.9582,
|
| 414 |
+
"eval_samples_per_second": 292.827,
|
| 415 |
+
"eval_steps_per_second": 36.658,
|
| 416 |
+
"step": 555
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.0090130689499774,
|
| 420 |
+
"grad_norm": 16.960424423217773,
|
| 421 |
+
"learning_rate": 3.989169675090253e-05,
|
| 422 |
+
"loss": 0.3719,
|
| 423 |
+
"step": 560
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.0270392068499323,
|
| 427 |
+
"grad_norm": 15.816173553466797,
|
| 428 |
+
"learning_rate": 3.971119133574007e-05,
|
| 429 |
+
"loss": 0.3238,
|
| 430 |
+
"step": 570
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.0450653447498874,
|
| 434 |
+
"grad_norm": 19.375043869018555,
|
| 435 |
+
"learning_rate": 3.953068592057762e-05,
|
| 436 |
+
"loss": 0.3516,
|
| 437 |
+
"step": 580
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.0630914826498423,
|
| 441 |
+
"grad_norm": 36.740943908691406,
|
| 442 |
+
"learning_rate": 3.935018050541516e-05,
|
| 443 |
+
"loss": 0.3329,
|
| 444 |
+
"step": 590
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.0811176205497972,
|
| 448 |
+
"grad_norm": 23.23931884765625,
|
| 449 |
+
"learning_rate": 3.916967509025271e-05,
|
| 450 |
+
"loss": 0.345,
|
| 451 |
+
"step": 600
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.099143758449752,
|
| 455 |
+
"grad_norm": 23.009553909301758,
|
| 456 |
+
"learning_rate": 3.898916967509025e-05,
|
| 457 |
+
"loss": 0.3054,
|
| 458 |
+
"step": 610
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.117169896349707,
|
| 462 |
+
"grad_norm": 18.865903854370117,
|
| 463 |
+
"learning_rate": 3.88086642599278e-05,
|
| 464 |
+
"loss": 0.3236,
|
| 465 |
+
"step": 620
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.135196034249662,
|
| 469 |
+
"grad_norm": 21.043642044067383,
|
| 470 |
+
"learning_rate": 3.862815884476535e-05,
|
| 471 |
+
"loss": 0.3092,
|
| 472 |
+
"step": 630
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.153222172149617,
|
| 476 |
+
"grad_norm": 22.50505256652832,
|
| 477 |
+
"learning_rate": 3.844765342960289e-05,
|
| 478 |
+
"loss": 0.3274,
|
| 479 |
+
"step": 640
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.1712483100495719,
|
| 483 |
+
"grad_norm": 20.4163818359375,
|
| 484 |
+
"learning_rate": 3.826714801444044e-05,
|
| 485 |
+
"loss": 0.2459,
|
| 486 |
+
"step": 650
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.1892744479495267,
|
| 490 |
+
"grad_norm": 21.210535049438477,
|
| 491 |
+
"learning_rate": 3.8086642599277976e-05,
|
| 492 |
+
"loss": 0.3365,
|
| 493 |
+
"step": 660
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.2073005858494819,
|
| 497 |
+
"grad_norm": 21.084487915039062,
|
| 498 |
+
"learning_rate": 3.790613718411553e-05,
|
| 499 |
+
"loss": 0.3242,
|
| 500 |
+
"step": 670
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.2253267237494367,
|
| 504 |
+
"grad_norm": 22.721454620361328,
|
| 505 |
+
"learning_rate": 3.7725631768953066e-05,
|
| 506 |
+
"loss": 0.2844,
|
| 507 |
+
"step": 680
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.2433528616493916,
|
| 511 |
+
"grad_norm": 49.28988265991211,
|
| 512 |
+
"learning_rate": 3.754512635379062e-05,
|
| 513 |
+
"loss": 0.3247,
|
| 514 |
+
"step": 690
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.2613789995493465,
|
| 518 |
+
"grad_norm": 48.35593795776367,
|
| 519 |
+
"learning_rate": 3.7364620938628155e-05,
|
| 520 |
+
"loss": 0.3217,
|
| 521 |
+
"step": 700
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.2794051374493014,
|
| 525 |
+
"grad_norm": 30.997228622436523,
|
| 526 |
+
"learning_rate": 3.718411552346571e-05,
|
| 527 |
+
"loss": 0.2854,
|
| 528 |
+
"step": 710
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.2974312753492563,
|
| 532 |
+
"grad_norm": 35.57414627075195,
|
| 533 |
+
"learning_rate": 3.700361010830325e-05,
|
| 534 |
+
"loss": 0.3315,
|
| 535 |
+
"step": 720
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.3154574132492114,
|
| 539 |
+
"grad_norm": 27.17759895324707,
|
| 540 |
+
"learning_rate": 3.68231046931408e-05,
|
| 541 |
+
"loss": 0.2965,
|
| 542 |
+
"step": 730
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.3334835511491663,
|
| 546 |
+
"grad_norm": 31.296627044677734,
|
| 547 |
+
"learning_rate": 3.664259927797834e-05,
|
| 548 |
+
"loss": 0.3396,
|
| 549 |
+
"step": 740
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.3515096890491212,
|
| 553 |
+
"grad_norm": 18.639175415039062,
|
| 554 |
+
"learning_rate": 3.646209386281589e-05,
|
| 555 |
+
"loss": 0.3178,
|
| 556 |
+
"step": 750
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.3695358269490763,
|
| 560 |
+
"grad_norm": 31.972665786743164,
|
| 561 |
+
"learning_rate": 3.628158844765343e-05,
|
| 562 |
+
"loss": 0.2927,
|
| 563 |
+
"step": 760
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.3875619648490312,
|
| 567 |
+
"grad_norm": 14.907872200012207,
|
| 568 |
+
"learning_rate": 3.610108303249098e-05,
|
| 569 |
+
"loss": 0.3391,
|
| 570 |
+
"step": 770
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.405588102748986,
|
| 574 |
+
"grad_norm": 33.58858871459961,
|
| 575 |
+
"learning_rate": 3.592057761732852e-05,
|
| 576 |
+
"loss": 0.3111,
|
| 577 |
+
"step": 780
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.423614240648941,
|
| 581 |
+
"grad_norm": 17.019819259643555,
|
| 582 |
+
"learning_rate": 3.574007220216607e-05,
|
| 583 |
+
"loss": 0.3401,
|
| 584 |
+
"step": 790
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.4416403785488958,
|
| 588 |
+
"grad_norm": 27.324296951293945,
|
| 589 |
+
"learning_rate": 3.555956678700361e-05,
|
| 590 |
+
"loss": 0.2917,
|
| 591 |
+
"step": 800
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.4596665164488507,
|
| 595 |
+
"grad_norm": 30.170120239257812,
|
| 596 |
+
"learning_rate": 3.537906137184116e-05,
|
| 597 |
+
"loss": 0.2871,
|
| 598 |
+
"step": 810
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.4776926543488058,
|
| 602 |
+
"grad_norm": 44.26614761352539,
|
| 603 |
+
"learning_rate": 3.51985559566787e-05,
|
| 604 |
+
"loss": 0.3307,
|
| 605 |
+
"step": 820
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.4957187922487607,
|
| 609 |
+
"grad_norm": 28.978620529174805,
|
| 610 |
+
"learning_rate": 3.5018050541516247e-05,
|
| 611 |
+
"loss": 0.3207,
|
| 612 |
+
"step": 830
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.5137449301487156,
|
| 616 |
+
"grad_norm": 22.961040496826172,
|
| 617 |
+
"learning_rate": 3.483754512635379e-05,
|
| 618 |
+
"loss": 0.3075,
|
| 619 |
+
"step": 840
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.5317710680486707,
|
| 623 |
+
"grad_norm": 22.511327743530273,
|
| 624 |
+
"learning_rate": 3.4657039711191336e-05,
|
| 625 |
+
"loss": 0.3035,
|
| 626 |
+
"step": 850
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.5497972059486256,
|
| 630 |
+
"grad_norm": 33.944190979003906,
|
| 631 |
+
"learning_rate": 3.447653429602888e-05,
|
| 632 |
+
"loss": 0.3243,
|
| 633 |
+
"step": 860
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.5678233438485805,
|
| 637 |
+
"grad_norm": 24.72734832763672,
|
| 638 |
+
"learning_rate": 3.4296028880866426e-05,
|
| 639 |
+
"loss": 0.3403,
|
| 640 |
+
"step": 870
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.5858494817485354,
|
| 644 |
+
"grad_norm": 31.211387634277344,
|
| 645 |
+
"learning_rate": 3.411552346570397e-05,
|
| 646 |
+
"loss": 0.3142,
|
| 647 |
+
"step": 880
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.6038756196484902,
|
| 651 |
+
"grad_norm": 23.974918365478516,
|
| 652 |
+
"learning_rate": 3.3935018050541516e-05,
|
| 653 |
+
"loss": 0.3015,
|
| 654 |
+
"step": 890
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.6219017575484451,
|
| 658 |
+
"grad_norm": 30.705829620361328,
|
| 659 |
+
"learning_rate": 3.375451263537907e-05,
|
| 660 |
+
"loss": 0.3529,
|
| 661 |
+
"step": 900
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.6399278954484002,
|
| 665 |
+
"grad_norm": 20.697528839111328,
|
| 666 |
+
"learning_rate": 3.3574007220216606e-05,
|
| 667 |
+
"loss": 0.302,
|
| 668 |
+
"step": 910
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.6579540333483551,
|
| 672 |
+
"grad_norm": 28.308101654052734,
|
| 673 |
+
"learning_rate": 3.339350180505416e-05,
|
| 674 |
+
"loss": 0.2735,
|
| 675 |
+
"step": 920
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.67598017124831,
|
| 679 |
+
"grad_norm": 22.28900909423828,
|
| 680 |
+
"learning_rate": 3.3212996389891696e-05,
|
| 681 |
+
"loss": 0.3194,
|
| 682 |
+
"step": 930
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.694006309148265,
|
| 686 |
+
"grad_norm": 29.04107666015625,
|
| 687 |
+
"learning_rate": 3.303249097472924e-05,
|
| 688 |
+
"loss": 0.3265,
|
| 689 |
+
"step": 940
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.71203244704822,
|
| 693 |
+
"grad_norm": 40.08032989501953,
|
| 694 |
+
"learning_rate": 3.2851985559566786e-05,
|
| 695 |
+
"loss": 0.3331,
|
| 696 |
+
"step": 950
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.7300585849481749,
|
| 700 |
+
"grad_norm": 29.40334129333496,
|
| 701 |
+
"learning_rate": 3.267148014440433e-05,
|
| 702 |
+
"loss": 0.2664,
|
| 703 |
+
"step": 960
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.7480847228481298,
|
| 707 |
+
"grad_norm": 33.51088333129883,
|
| 708 |
+
"learning_rate": 3.249097472924188e-05,
|
| 709 |
+
"loss": 0.3031,
|
| 710 |
+
"step": 970
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.7661108607480847,
|
| 714 |
+
"grad_norm": 27.003469467163086,
|
| 715 |
+
"learning_rate": 3.231046931407942e-05,
|
| 716 |
+
"loss": 0.3195,
|
| 717 |
+
"step": 980
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.7841369986480395,
|
| 721 |
+
"grad_norm": 26.585309982299805,
|
| 722 |
+
"learning_rate": 3.212996389891697e-05,
|
| 723 |
+
"loss": 0.3385,
|
| 724 |
+
"step": 990
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.8021631365479944,
|
| 728 |
+
"grad_norm": 36.866294860839844,
|
| 729 |
+
"learning_rate": 3.194945848375451e-05,
|
| 730 |
+
"loss": 0.3029,
|
| 731 |
+
"step": 1000
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.8201892744479495,
|
| 735 |
+
"grad_norm": 44.910362243652344,
|
| 736 |
+
"learning_rate": 3.176895306859206e-05,
|
| 737 |
+
"loss": 0.3239,
|
| 738 |
+
"step": 1010
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.8382154123479044,
|
| 742 |
+
"grad_norm": 20.13945198059082,
|
| 743 |
+
"learning_rate": 3.15884476534296e-05,
|
| 744 |
+
"loss": 0.3071,
|
| 745 |
+
"step": 1020
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.8562415502478595,
|
| 749 |
+
"grad_norm": 42.44215774536133,
|
| 750 |
+
"learning_rate": 3.140794223826715e-05,
|
| 751 |
+
"loss": 0.3246,
|
| 752 |
+
"step": 1030
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.8742676881478144,
|
| 756 |
+
"grad_norm": 31.791410446166992,
|
| 757 |
+
"learning_rate": 3.12274368231047e-05,
|
| 758 |
+
"loss": 0.3273,
|
| 759 |
+
"step": 1040
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.8922938260477693,
|
| 763 |
+
"grad_norm": 21.244722366333008,
|
| 764 |
+
"learning_rate": 3.104693140794224e-05,
|
| 765 |
+
"loss": 0.3093,
|
| 766 |
+
"step": 1050
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.9103199639477242,
|
| 770 |
+
"grad_norm": 27.103349685668945,
|
| 771 |
+
"learning_rate": 3.086642599277979e-05,
|
| 772 |
+
"loss": 0.2694,
|
| 773 |
+
"step": 1060
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.928346101847679,
|
| 777 |
+
"grad_norm": 35.73145294189453,
|
| 778 |
+
"learning_rate": 3.0685920577617325e-05,
|
| 779 |
+
"loss": 0.2934,
|
| 780 |
+
"step": 1070
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.946372239747634,
|
| 784 |
+
"grad_norm": 19.97008514404297,
|
| 785 |
+
"learning_rate": 3.0505415162454877e-05,
|
| 786 |
+
"loss": 0.3493,
|
| 787 |
+
"step": 1080
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.9643983776475888,
|
| 791 |
+
"grad_norm": 23.084754943847656,
|
| 792 |
+
"learning_rate": 3.032490974729242e-05,
|
| 793 |
+
"loss": 0.3243,
|
| 794 |
+
"step": 1090
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.982424515547544,
|
| 798 |
+
"grad_norm": 29.515155792236328,
|
| 799 |
+
"learning_rate": 3.0144404332129967e-05,
|
| 800 |
+
"loss": 0.3235,
|
| 801 |
+
"step": 1100
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 2.0,
|
| 805 |
+
"grad_norm": 24.17915916442871,
|
| 806 |
+
"learning_rate": 2.996389891696751e-05,
|
| 807 |
+
"loss": 0.2751,
|
| 808 |
+
"step": 1110
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 2.0,
|
| 812 |
+
"eval_f1": 0.8428896708755951,
|
| 813 |
+
"eval_loss": 0.3940623104572296,
|
| 814 |
+
"eval_precision": 0.800314465408805,
|
| 815 |
+
"eval_recall": 0.890249234805422,
|
| 816 |
+
"eval_runtime": 12.3989,
|
| 817 |
+
"eval_samples_per_second": 376.889,
|
| 818 |
+
"eval_steps_per_second": 47.182,
|
| 819 |
+
"step": 1110
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 2.018026137899955,
|
| 823 |
+
"grad_norm": 18.65660858154297,
|
| 824 |
+
"learning_rate": 2.9783393501805057e-05,
|
| 825 |
+
"loss": 0.1912,
|
| 826 |
+
"step": 1120
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 2.0360522757999098,
|
| 830 |
+
"grad_norm": 15.253448486328125,
|
| 831 |
+
"learning_rate": 2.9602888086642598e-05,
|
| 832 |
+
"loss": 0.1425,
|
| 833 |
+
"step": 1130
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 2.0540784136998647,
|
| 837 |
+
"grad_norm": 58.709293365478516,
|
| 838 |
+
"learning_rate": 2.9422382671480147e-05,
|
| 839 |
+
"loss": 0.1945,
|
| 840 |
+
"step": 1140
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 2.0721045515998195,
|
| 844 |
+
"grad_norm": 45.032073974609375,
|
| 845 |
+
"learning_rate": 2.924187725631769e-05,
|
| 846 |
+
"loss": 0.1477,
|
| 847 |
+
"step": 1150
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 2.090130689499775,
|
| 851 |
+
"grad_norm": 26.54606819152832,
|
| 852 |
+
"learning_rate": 2.906137184115524e-05,
|
| 853 |
+
"loss": 0.1731,
|
| 854 |
+
"step": 1160
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 2.1081568273997298,
|
| 858 |
+
"grad_norm": 24.949045181274414,
|
| 859 |
+
"learning_rate": 2.888086642599278e-05,
|
| 860 |
+
"loss": 0.1504,
|
| 861 |
+
"step": 1170
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 2.1261829652996846,
|
| 865 |
+
"grad_norm": 71.09517669677734,
|
| 866 |
+
"learning_rate": 2.870036101083033e-05,
|
| 867 |
+
"loss": 0.149,
|
| 868 |
+
"step": 1180
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 2.1442091031996395,
|
| 872 |
+
"grad_norm": 43.31296157836914,
|
| 873 |
+
"learning_rate": 2.851985559566787e-05,
|
| 874 |
+
"loss": 0.1826,
|
| 875 |
+
"step": 1190
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 2.1622352410995944,
|
| 879 |
+
"grad_norm": 23.145519256591797,
|
| 880 |
+
"learning_rate": 2.8339350180505413e-05,
|
| 881 |
+
"loss": 0.1716,
|
| 882 |
+
"step": 1200
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 2.1802613789995493,
|
| 886 |
+
"grad_norm": 41.57735061645508,
|
| 887 |
+
"learning_rate": 2.815884476534296e-05,
|
| 888 |
+
"loss": 0.1784,
|
| 889 |
+
"step": 1210
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 2.198287516899504,
|
| 893 |
+
"grad_norm": 51.243560791015625,
|
| 894 |
+
"learning_rate": 2.7978339350180506e-05,
|
| 895 |
+
"loss": 0.161,
|
| 896 |
+
"step": 1220
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 2.216313654799459,
|
| 900 |
+
"grad_norm": 38.89030838012695,
|
| 901 |
+
"learning_rate": 2.779783393501805e-05,
|
| 902 |
+
"loss": 0.1688,
|
| 903 |
+
"step": 1230
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 2.234339792699414,
|
| 907 |
+
"grad_norm": 43.14015197753906,
|
| 908 |
+
"learning_rate": 2.7617328519855596e-05,
|
| 909 |
+
"loss": 0.1916,
|
| 910 |
+
"step": 1240
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 2.2523659305993693,
|
| 914 |
+
"grad_norm": 35.20539474487305,
|
| 915 |
+
"learning_rate": 2.7436823104693144e-05,
|
| 916 |
+
"loss": 0.1652,
|
| 917 |
+
"step": 1250
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 2.270392068499324,
|
| 921 |
+
"grad_norm": 39.17034912109375,
|
| 922 |
+
"learning_rate": 2.7256317689530686e-05,
|
| 923 |
+
"loss": 0.1888,
|
| 924 |
+
"step": 1260
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 2.288418206399279,
|
| 928 |
+
"grad_norm": 26.350337982177734,
|
| 929 |
+
"learning_rate": 2.7075812274368234e-05,
|
| 930 |
+
"loss": 0.1585,
|
| 931 |
+
"step": 1270
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 2.306444344299234,
|
| 935 |
+
"grad_norm": 33.05204772949219,
|
| 936 |
+
"learning_rate": 2.6895306859205776e-05,
|
| 937 |
+
"loss": 0.1449,
|
| 938 |
+
"step": 1280
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 2.324470482199189,
|
| 942 |
+
"grad_norm": 53.03055191040039,
|
| 943 |
+
"learning_rate": 2.6714801444043324e-05,
|
| 944 |
+
"loss": 0.1457,
|
| 945 |
+
"step": 1290
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 2.3424966200991437,
|
| 949 |
+
"grad_norm": 13.711820602416992,
|
| 950 |
+
"learning_rate": 2.6534296028880866e-05,
|
| 951 |
+
"loss": 0.1451,
|
| 952 |
+
"step": 1300
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 2.3605227579990986,
|
| 956 |
+
"grad_norm": 60.76357650756836,
|
| 957 |
+
"learning_rate": 2.6353790613718414e-05,
|
| 958 |
+
"loss": 0.1803,
|
| 959 |
+
"step": 1310
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 2.3785488958990535,
|
| 963 |
+
"grad_norm": 47.53473663330078,
|
| 964 |
+
"learning_rate": 2.617328519855596e-05,
|
| 965 |
+
"loss": 0.1405,
|
| 966 |
+
"step": 1320
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 2.3965750337990084,
|
| 970 |
+
"grad_norm": 46.01548385620117,
|
| 971 |
+
"learning_rate": 2.59927797833935e-05,
|
| 972 |
+
"loss": 0.1701,
|
| 973 |
+
"step": 1330
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 2.4146011716989637,
|
| 977 |
+
"grad_norm": 36.47391128540039,
|
| 978 |
+
"learning_rate": 2.581227436823105e-05,
|
| 979 |
+
"loss": 0.1362,
|
| 980 |
+
"step": 1340
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 2.4326273095989186,
|
| 984 |
+
"grad_norm": 69.90142059326172,
|
| 985 |
+
"learning_rate": 2.563176895306859e-05,
|
| 986 |
+
"loss": 0.143,
|
| 987 |
+
"step": 1350
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 2.4506534474988735,
|
| 991 |
+
"grad_norm": 52.88728713989258,
|
| 992 |
+
"learning_rate": 2.545126353790614e-05,
|
| 993 |
+
"loss": 0.1776,
|
| 994 |
+
"step": 1360
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 2.4686795853988284,
|
| 998 |
+
"grad_norm": 26.982271194458008,
|
| 999 |
+
"learning_rate": 2.527075812274368e-05,
|
| 1000 |
+
"loss": 0.1922,
|
| 1001 |
+
"step": 1370
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 2.4867057232987833,
|
| 1005 |
+
"grad_norm": 41.308414459228516,
|
| 1006 |
+
"learning_rate": 2.509025270758123e-05,
|
| 1007 |
+
"loss": 0.1699,
|
| 1008 |
+
"step": 1380
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 2.504731861198738,
|
| 1012 |
+
"grad_norm": 68.30561828613281,
|
| 1013 |
+
"learning_rate": 2.4909747292418774e-05,
|
| 1014 |
+
"loss": 0.1503,
|
| 1015 |
+
"step": 1390
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 2.522757999098693,
|
| 1019 |
+
"grad_norm": 32.563026428222656,
|
| 1020 |
+
"learning_rate": 2.472924187725632e-05,
|
| 1021 |
+
"loss": 0.2153,
|
| 1022 |
+
"step": 1400
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 2.540784136998648,
|
| 1026 |
+
"grad_norm": 24.238431930541992,
|
| 1027 |
+
"learning_rate": 2.4548736462093864e-05,
|
| 1028 |
+
"loss": 0.1935,
|
| 1029 |
+
"step": 1410
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 2.558810274898603,
|
| 1033 |
+
"grad_norm": 52.790916442871094,
|
| 1034 |
+
"learning_rate": 2.436823104693141e-05,
|
| 1035 |
+
"loss": 0.21,
|
| 1036 |
+
"step": 1420
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 2.576836412798558,
|
| 1040 |
+
"grad_norm": 39.62615966796875,
|
| 1041 |
+
"learning_rate": 2.4187725631768953e-05,
|
| 1042 |
+
"loss": 0.1756,
|
| 1043 |
+
"step": 1430
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 2.5948625506985126,
|
| 1047 |
+
"grad_norm": 28.220199584960938,
|
| 1048 |
+
"learning_rate": 2.40072202166065e-05,
|
| 1049 |
+
"loss": 0.1704,
|
| 1050 |
+
"step": 1440
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 2.612888688598468,
|
| 1054 |
+
"grad_norm": 36.99165344238281,
|
| 1055 |
+
"learning_rate": 2.3826714801444043e-05,
|
| 1056 |
+
"loss": 0.1632,
|
| 1057 |
+
"step": 1450
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 2.630914826498423,
|
| 1061 |
+
"grad_norm": 79.26138305664062,
|
| 1062 |
+
"learning_rate": 2.3646209386281588e-05,
|
| 1063 |
+
"loss": 0.1562,
|
| 1064 |
+
"step": 1460
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 2.6489409643983777,
|
| 1068 |
+
"grad_norm": 47.28179931640625,
|
| 1069 |
+
"learning_rate": 2.3465703971119137e-05,
|
| 1070 |
+
"loss": 0.2061,
|
| 1071 |
+
"step": 1470
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 2.6669671022983326,
|
| 1075 |
+
"grad_norm": 40.676551818847656,
|
| 1076 |
+
"learning_rate": 2.328519855595668e-05,
|
| 1077 |
+
"loss": 0.1955,
|
| 1078 |
+
"step": 1480
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 2.6849932401982874,
|
| 1082 |
+
"grad_norm": 29.851699829101562,
|
| 1083 |
+
"learning_rate": 2.3104693140794227e-05,
|
| 1084 |
+
"loss": 0.1447,
|
| 1085 |
+
"step": 1490
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 2.7030193780982423,
|
| 1089 |
+
"grad_norm": 46.22013473510742,
|
| 1090 |
+
"learning_rate": 2.292418772563177e-05,
|
| 1091 |
+
"loss": 0.1635,
|
| 1092 |
+
"step": 1500
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 2.721045515998197,
|
| 1096 |
+
"grad_norm": 35.77067565917969,
|
| 1097 |
+
"learning_rate": 2.2743682310469316e-05,
|
| 1098 |
+
"loss": 0.1611,
|
| 1099 |
+
"step": 1510
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 2.7390716538981525,
|
| 1103 |
+
"grad_norm": 44.474876403808594,
|
| 1104 |
+
"learning_rate": 2.2563176895306858e-05,
|
| 1105 |
+
"loss": 0.159,
|
| 1106 |
+
"step": 1520
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 2.757097791798107,
|
| 1110 |
+
"grad_norm": 49.898887634277344,
|
| 1111 |
+
"learning_rate": 2.2382671480144403e-05,
|
| 1112 |
+
"loss": 0.1716,
|
| 1113 |
+
"step": 1530
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 2.7751239296980623,
|
| 1117 |
+
"grad_norm": 54.783653259277344,
|
| 1118 |
+
"learning_rate": 2.220216606498195e-05,
|
| 1119 |
+
"loss": 0.1921,
|
| 1120 |
+
"step": 1540
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 2.793150067598017,
|
| 1124 |
+
"grad_norm": 14.532942771911621,
|
| 1125 |
+
"learning_rate": 2.2021660649819496e-05,
|
| 1126 |
+
"loss": 0.1436,
|
| 1127 |
+
"step": 1550
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 2.811176205497972,
|
| 1131 |
+
"grad_norm": 47.2420768737793,
|
| 1132 |
+
"learning_rate": 2.184115523465704e-05,
|
| 1133 |
+
"loss": 0.2074,
|
| 1134 |
+
"step": 1560
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 2.829202343397927,
|
| 1138 |
+
"grad_norm": 39.21223449707031,
|
| 1139 |
+
"learning_rate": 2.1660649819494586e-05,
|
| 1140 |
+
"loss": 0.1871,
|
| 1141 |
+
"step": 1570
|
| 1142 |
+
},
|
| 1143 |
+
{
|
| 1144 |
+
"epoch": 2.847228481297882,
|
| 1145 |
+
"grad_norm": 43.29368591308594,
|
| 1146 |
+
"learning_rate": 2.148014440433213e-05,
|
| 1147 |
+
"loss": 0.1474,
|
| 1148 |
+
"step": 1580
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 2.8652546191978367,
|
| 1152 |
+
"grad_norm": 30.276391983032227,
|
| 1153 |
+
"learning_rate": 2.1299638989169676e-05,
|
| 1154 |
+
"loss": 0.2214,
|
| 1155 |
+
"step": 1590
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 2.8832807570977916,
|
| 1159 |
+
"grad_norm": 38.97634506225586,
|
| 1160 |
+
"learning_rate": 2.111913357400722e-05,
|
| 1161 |
+
"loss": 0.1692,
|
| 1162 |
+
"step": 1600
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 2.901306894997747,
|
| 1166 |
+
"grad_norm": 46.147945404052734,
|
| 1167 |
+
"learning_rate": 2.0938628158844766e-05,
|
| 1168 |
+
"loss": 0.1587,
|
| 1169 |
+
"step": 1610
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 2.9193330328977014,
|
| 1173 |
+
"grad_norm": 57.70730209350586,
|
| 1174 |
+
"learning_rate": 2.075812274368231e-05,
|
| 1175 |
+
"loss": 0.1642,
|
| 1176 |
+
"step": 1620
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 2.9373591707976567,
|
| 1180 |
+
"grad_norm": 48.976478576660156,
|
| 1181 |
+
"learning_rate": 2.057761732851986e-05,
|
| 1182 |
+
"loss": 0.1536,
|
| 1183 |
+
"step": 1630
|
| 1184 |
+
},
|
| 1185 |
+
{
|
| 1186 |
+
"epoch": 2.9553853086976116,
|
| 1187 |
+
"grad_norm": 49.945587158203125,
|
| 1188 |
+
"learning_rate": 2.0397111913357404e-05,
|
| 1189 |
+
"loss": 0.1612,
|
| 1190 |
+
"step": 1640
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 2.9734114465975665,
|
| 1194 |
+
"grad_norm": 74.91495513916016,
|
| 1195 |
+
"learning_rate": 2.0216606498194946e-05,
|
| 1196 |
+
"loss": 0.1557,
|
| 1197 |
+
"step": 1650
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 2.9914375844975214,
|
| 1201 |
+
"grad_norm": 30.24640464782715,
|
| 1202 |
+
"learning_rate": 2.003610108303249e-05,
|
| 1203 |
+
"loss": 0.1631,
|
| 1204 |
+
"step": 1660
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 3.0,
|
| 1208 |
+
"eval_f1": 0.8290598290598291,
|
| 1209 |
+
"eval_loss": 0.5290467739105225,
|
| 1210 |
+
"eval_precision": 0.8106978687839532,
|
| 1211 |
+
"eval_recall": 0.8482728465238304,
|
| 1212 |
+
"eval_runtime": 11.4074,
|
| 1213 |
+
"eval_samples_per_second": 409.646,
|
| 1214 |
+
"eval_steps_per_second": 51.282,
|
| 1215 |
+
"step": 1665
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 3.0090130689499777,
|
| 1219 |
+
"grad_norm": 7.1513214111328125,
|
| 1220 |
+
"learning_rate": 1.9855595667870036e-05,
|
| 1221 |
+
"loss": 0.1199,
|
| 1222 |
+
"step": 1670
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 3.0270392068499326,
|
| 1226 |
+
"grad_norm": 36.67481994628906,
|
| 1227 |
+
"learning_rate": 1.967509025270758e-05,
|
| 1228 |
+
"loss": 0.0607,
|
| 1229 |
+
"step": 1680
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 3.0450653447498874,
|
| 1233 |
+
"grad_norm": 41.55238723754883,
|
| 1234 |
+
"learning_rate": 1.9494584837545125e-05,
|
| 1235 |
+
"loss": 0.0821,
|
| 1236 |
+
"step": 1690
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 3.0630914826498423,
|
| 1240 |
+
"grad_norm": 37.779537200927734,
|
| 1241 |
+
"learning_rate": 1.9314079422382674e-05,
|
| 1242 |
+
"loss": 0.0624,
|
| 1243 |
+
"step": 1700
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 3.081117620549797,
|
| 1247 |
+
"grad_norm": 13.272784233093262,
|
| 1248 |
+
"learning_rate": 1.913357400722022e-05,
|
| 1249 |
+
"loss": 0.0745,
|
| 1250 |
+
"step": 1710
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 3.099143758449752,
|
| 1254 |
+
"grad_norm": 19.611099243164062,
|
| 1255 |
+
"learning_rate": 1.8953068592057764e-05,
|
| 1256 |
+
"loss": 0.0953,
|
| 1257 |
+
"step": 1720
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 3.117169896349707,
|
| 1261 |
+
"grad_norm": 26.23482894897461,
|
| 1262 |
+
"learning_rate": 1.877256317689531e-05,
|
| 1263 |
+
"loss": 0.0576,
|
| 1264 |
+
"step": 1730
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 3.135196034249662,
|
| 1268 |
+
"grad_norm": 31.549488067626953,
|
| 1269 |
+
"learning_rate": 1.8592057761732854e-05,
|
| 1270 |
+
"loss": 0.1034,
|
| 1271 |
+
"step": 1740
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 3.1532221721496168,
|
| 1275 |
+
"grad_norm": 44.37907409667969,
|
| 1276 |
+
"learning_rate": 1.84115523465704e-05,
|
| 1277 |
+
"loss": 0.0807,
|
| 1278 |
+
"step": 1750
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 3.171248310049572,
|
| 1282 |
+
"grad_norm": 29.255582809448242,
|
| 1283 |
+
"learning_rate": 1.8231046931407943e-05,
|
| 1284 |
+
"loss": 0.08,
|
| 1285 |
+
"step": 1760
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 3.189274447949527,
|
| 1289 |
+
"grad_norm": 11.359423637390137,
|
| 1290 |
+
"learning_rate": 1.805054151624549e-05,
|
| 1291 |
+
"loss": 0.073,
|
| 1292 |
+
"step": 1770
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 3.207300585849482,
|
| 1296 |
+
"grad_norm": 51.08259201049805,
|
| 1297 |
+
"learning_rate": 1.7870036101083033e-05,
|
| 1298 |
+
"loss": 0.0869,
|
| 1299 |
+
"step": 1780
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 3.2253267237494367,
|
| 1303 |
+
"grad_norm": 28.016101837158203,
|
| 1304 |
+
"learning_rate": 1.768953068592058e-05,
|
| 1305 |
+
"loss": 0.1044,
|
| 1306 |
+
"step": 1790
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 3.2433528616493916,
|
| 1310 |
+
"grad_norm": 81.0438003540039,
|
| 1311 |
+
"learning_rate": 1.7509025270758123e-05,
|
| 1312 |
+
"loss": 0.0818,
|
| 1313 |
+
"step": 1800
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 3.2613789995493465,
|
| 1317 |
+
"grad_norm": 25.233104705810547,
|
| 1318 |
+
"learning_rate": 1.7328519855595668e-05,
|
| 1319 |
+
"loss": 0.0808,
|
| 1320 |
+
"step": 1810
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 3.2794051374493014,
|
| 1324 |
+
"grad_norm": 127.22054290771484,
|
| 1325 |
+
"learning_rate": 1.7148014440433213e-05,
|
| 1326 |
+
"loss": 0.0847,
|
| 1327 |
+
"step": 1820
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 3.2974312753492563,
|
| 1331 |
+
"grad_norm": 37.59592819213867,
|
| 1332 |
+
"learning_rate": 1.6967509025270758e-05,
|
| 1333 |
+
"loss": 0.1276,
|
| 1334 |
+
"step": 1830
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 3.315457413249211,
|
| 1338 |
+
"grad_norm": 33.60833740234375,
|
| 1339 |
+
"learning_rate": 1.6787003610108303e-05,
|
| 1340 |
+
"loss": 0.0718,
|
| 1341 |
+
"step": 1840
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 3.3334835511491665,
|
| 1345 |
+
"grad_norm": 47.96291732788086,
|
| 1346 |
+
"learning_rate": 1.6606498194945848e-05,
|
| 1347 |
+
"loss": 0.0769,
|
| 1348 |
+
"step": 1850
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 3.3515096890491214,
|
| 1352 |
+
"grad_norm": 1.6931202411651611,
|
| 1353 |
+
"learning_rate": 1.6425992779783393e-05,
|
| 1354 |
+
"loss": 0.0694,
|
| 1355 |
+
"step": 1860
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 3.3695358269490763,
|
| 1359 |
+
"grad_norm": 65.2172622680664,
|
| 1360 |
+
"learning_rate": 1.624548736462094e-05,
|
| 1361 |
+
"loss": 0.1009,
|
| 1362 |
+
"step": 1870
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 3.387561964849031,
|
| 1366 |
+
"grad_norm": 11.240936279296875,
|
| 1367 |
+
"learning_rate": 1.6064981949458486e-05,
|
| 1368 |
+
"loss": 0.0749,
|
| 1369 |
+
"step": 1880
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 3.405588102748986,
|
| 1373 |
+
"grad_norm": 24.050607681274414,
|
| 1374 |
+
"learning_rate": 1.588447653429603e-05,
|
| 1375 |
+
"loss": 0.0979,
|
| 1376 |
+
"step": 1890
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 3.423614240648941,
|
| 1380 |
+
"grad_norm": 92.54638671875,
|
| 1381 |
+
"learning_rate": 1.5703971119133576e-05,
|
| 1382 |
+
"loss": 0.0968,
|
| 1383 |
+
"step": 1900
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 3.441640378548896,
|
| 1387 |
+
"grad_norm": 10.564998626708984,
|
| 1388 |
+
"learning_rate": 1.552346570397112e-05,
|
| 1389 |
+
"loss": 0.0642,
|
| 1390 |
+
"step": 1910
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 3.4596665164488507,
|
| 1394 |
+
"grad_norm": 63.50326919555664,
|
| 1395 |
+
"learning_rate": 1.5342960288808663e-05,
|
| 1396 |
+
"loss": 0.0941,
|
| 1397 |
+
"step": 1920
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 3.4776926543488056,
|
| 1401 |
+
"grad_norm": 74.58378601074219,
|
| 1402 |
+
"learning_rate": 1.516245487364621e-05,
|
| 1403 |
+
"loss": 0.1205,
|
| 1404 |
+
"step": 1930
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 3.495718792248761,
|
| 1408 |
+
"grad_norm": 9.73009967803955,
|
| 1409 |
+
"learning_rate": 1.4981949458483754e-05,
|
| 1410 |
+
"loss": 0.0826,
|
| 1411 |
+
"step": 1940
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 3.5137449301487154,
|
| 1415 |
+
"grad_norm": 14.504748344421387,
|
| 1416 |
+
"learning_rate": 1.4801444043321299e-05,
|
| 1417 |
+
"loss": 0.0861,
|
| 1418 |
+
"step": 1950
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 3.5317710680486707,
|
| 1422 |
+
"grad_norm": 3.4811439514160156,
|
| 1423 |
+
"learning_rate": 1.4620938628158846e-05,
|
| 1424 |
+
"loss": 0.0658,
|
| 1425 |
+
"step": 1960
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 3.5497972059486256,
|
| 1429 |
+
"grad_norm": 60.301517486572266,
|
| 1430 |
+
"learning_rate": 1.444043321299639e-05,
|
| 1431 |
+
"loss": 0.074,
|
| 1432 |
+
"step": 1970
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 3.5678233438485805,
|
| 1436 |
+
"grad_norm": 58.39499282836914,
|
| 1437 |
+
"learning_rate": 1.4259927797833936e-05,
|
| 1438 |
+
"loss": 0.1037,
|
| 1439 |
+
"step": 1980
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 3.5858494817485354,
|
| 1443 |
+
"grad_norm": 28.403823852539062,
|
| 1444 |
+
"learning_rate": 1.407942238267148e-05,
|
| 1445 |
+
"loss": 0.0807,
|
| 1446 |
+
"step": 1990
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 3.6038756196484902,
|
| 1450 |
+
"grad_norm": 47.89693832397461,
|
| 1451 |
+
"learning_rate": 1.3898916967509026e-05,
|
| 1452 |
+
"loss": 0.0818,
|
| 1453 |
+
"step": 2000
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 3.621901757548445,
|
| 1457 |
+
"grad_norm": 43.21398162841797,
|
| 1458 |
+
"learning_rate": 1.3718411552346572e-05,
|
| 1459 |
+
"loss": 0.099,
|
| 1460 |
+
"step": 2010
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 3.6399278954484,
|
| 1464 |
+
"grad_norm": 84.97303771972656,
|
| 1465 |
+
"learning_rate": 1.3537906137184117e-05,
|
| 1466 |
+
"loss": 0.0916,
|
| 1467 |
+
"step": 2020
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 3.6579540333483553,
|
| 1471 |
+
"grad_norm": 62.820228576660156,
|
| 1472 |
+
"learning_rate": 1.3357400722021662e-05,
|
| 1473 |
+
"loss": 0.1033,
|
| 1474 |
+
"step": 2030
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 3.67598017124831,
|
| 1478 |
+
"grad_norm": 41.46628952026367,
|
| 1479 |
+
"learning_rate": 1.3176895306859207e-05,
|
| 1480 |
+
"loss": 0.0775,
|
| 1481 |
+
"step": 2040
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 3.694006309148265,
|
| 1485 |
+
"grad_norm": 9.337126731872559,
|
| 1486 |
+
"learning_rate": 1.299638989169675e-05,
|
| 1487 |
+
"loss": 0.083,
|
| 1488 |
+
"step": 2050
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 3.71203244704822,
|
| 1492 |
+
"grad_norm": 86.4469985961914,
|
| 1493 |
+
"learning_rate": 1.2815884476534295e-05,
|
| 1494 |
+
"loss": 0.0702,
|
| 1495 |
+
"step": 2060
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 3.730058584948175,
|
| 1499 |
+
"grad_norm": 44.70920944213867,
|
| 1500 |
+
"learning_rate": 1.263537906137184e-05,
|
| 1501 |
+
"loss": 0.0908,
|
| 1502 |
+
"step": 2070
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 3.7480847228481298,
|
| 1506 |
+
"grad_norm": 83.01412963867188,
|
| 1507 |
+
"learning_rate": 1.2454873646209387e-05,
|
| 1508 |
+
"loss": 0.0831,
|
| 1509 |
+
"step": 2080
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 3.7661108607480847,
|
| 1513 |
+
"grad_norm": 28.753332138061523,
|
| 1514 |
+
"learning_rate": 1.2274368231046932e-05,
|
| 1515 |
+
"loss": 0.1054,
|
| 1516 |
+
"step": 2090
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 3.7841369986480395,
|
| 1520 |
+
"grad_norm": 39.22829055786133,
|
| 1521 |
+
"learning_rate": 1.2093862815884477e-05,
|
| 1522 |
+
"loss": 0.0976,
|
| 1523 |
+
"step": 2100
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 3.8021631365479944,
|
| 1527 |
+
"grad_norm": 41.21415328979492,
|
| 1528 |
+
"learning_rate": 1.1913357400722022e-05,
|
| 1529 |
+
"loss": 0.0995,
|
| 1530 |
+
"step": 2110
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 3.8201892744479498,
|
| 1534 |
+
"grad_norm": 25.80561065673828,
|
| 1535 |
+
"learning_rate": 1.1732851985559568e-05,
|
| 1536 |
+
"loss": 0.1185,
|
| 1537 |
+
"step": 2120
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 3.838215412347904,
|
| 1541 |
+
"grad_norm": 23.873735427856445,
|
| 1542 |
+
"learning_rate": 1.1552346570397113e-05,
|
| 1543 |
+
"loss": 0.0618,
|
| 1544 |
+
"step": 2130
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 3.8562415502478595,
|
| 1548 |
+
"grad_norm": 20.476320266723633,
|
| 1549 |
+
"learning_rate": 1.1371841155234658e-05,
|
| 1550 |
+
"loss": 0.1164,
|
| 1551 |
+
"step": 2140
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 3.8742676881478144,
|
| 1555 |
+
"grad_norm": 75.58688354492188,
|
| 1556 |
+
"learning_rate": 1.1191335740072201e-05,
|
| 1557 |
+
"loss": 0.1166,
|
| 1558 |
+
"step": 2150
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 3.8922938260477693,
|
| 1562 |
+
"grad_norm": 49.5202751159668,
|
| 1563 |
+
"learning_rate": 1.1010830324909748e-05,
|
| 1564 |
+
"loss": 0.0982,
|
| 1565 |
+
"step": 2160
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 3.910319963947724,
|
| 1569 |
+
"grad_norm": 37.534568786621094,
|
| 1570 |
+
"learning_rate": 1.0830324909747293e-05,
|
| 1571 |
+
"loss": 0.0652,
|
| 1572 |
+
"step": 2170
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 3.928346101847679,
|
| 1576 |
+
"grad_norm": 70.99764251708984,
|
| 1577 |
+
"learning_rate": 1.0649819494584838e-05,
|
| 1578 |
+
"loss": 0.0876,
|
| 1579 |
+
"step": 2180
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 3.946372239747634,
|
| 1583 |
+
"grad_norm": 35.94860076904297,
|
| 1584 |
+
"learning_rate": 1.0469314079422383e-05,
|
| 1585 |
+
"loss": 0.0642,
|
| 1586 |
+
"step": 2190
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 3.964398377647589,
|
| 1590 |
+
"grad_norm": 6.9634785652160645,
|
| 1591 |
+
"learning_rate": 1.028880866425993e-05,
|
| 1592 |
+
"loss": 0.1256,
|
| 1593 |
+
"step": 2200
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 3.982424515547544,
|
| 1597 |
+
"grad_norm": 33.263797760009766,
|
| 1598 |
+
"learning_rate": 1.0108303249097473e-05,
|
| 1599 |
+
"loss": 0.0646,
|
| 1600 |
+
"step": 2210
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 4.0,
|
| 1604 |
+
"grad_norm": 24.34364128112793,
|
| 1605 |
+
"learning_rate": 9.927797833935018e-06,
|
| 1606 |
+
"loss": 0.0782,
|
| 1607 |
+
"step": 2220
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 4.0,
|
| 1611 |
+
"eval_f1": 0.827906976744186,
|
| 1612 |
+
"eval_loss": 0.6959211826324463,
|
| 1613 |
+
"eval_precision": 0.8014735980352026,
|
| 1614 |
+
"eval_recall": 0.8561434193266287,
|
| 1615 |
+
"eval_runtime": 10.9179,
|
| 1616 |
+
"eval_samples_per_second": 428.013,
|
| 1617 |
+
"eval_steps_per_second": 53.582,
|
| 1618 |
+
"step": 2220
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 4.018026137899955,
|
| 1622 |
+
"grad_norm": 26.636980056762695,
|
| 1623 |
+
"learning_rate": 9.747292418772563e-06,
|
| 1624 |
+
"loss": 0.0584,
|
| 1625 |
+
"step": 2230
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 4.03605227579991,
|
| 1629 |
+
"grad_norm": 52.13910675048828,
|
| 1630 |
+
"learning_rate": 9.56678700361011e-06,
|
| 1631 |
+
"loss": 0.0696,
|
| 1632 |
+
"step": 2240
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 4.054078413699865,
|
| 1636 |
+
"grad_norm": 19.250686645507812,
|
| 1637 |
+
"learning_rate": 9.386281588447654e-06,
|
| 1638 |
+
"loss": 0.0624,
|
| 1639 |
+
"step": 2250
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 4.0721045515998195,
|
| 1643 |
+
"grad_norm": 17.650712966918945,
|
| 1644 |
+
"learning_rate": 9.2057761732852e-06,
|
| 1645 |
+
"loss": 0.0488,
|
| 1646 |
+
"step": 2260
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 4.090130689499775,
|
| 1650 |
+
"grad_norm": 44.8015251159668,
|
| 1651 |
+
"learning_rate": 9.025270758122744e-06,
|
| 1652 |
+
"loss": 0.042,
|
| 1653 |
+
"step": 2270
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 4.108156827399729,
|
| 1657 |
+
"grad_norm": 16.66750144958496,
|
| 1658 |
+
"learning_rate": 8.84476534296029e-06,
|
| 1659 |
+
"loss": 0.0636,
|
| 1660 |
+
"step": 2280
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 4.126182965299685,
|
| 1664 |
+
"grad_norm": 8.76605224609375,
|
| 1665 |
+
"learning_rate": 8.664259927797834e-06,
|
| 1666 |
+
"loss": 0.0494,
|
| 1667 |
+
"step": 2290
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 4.144209103199639,
|
| 1671 |
+
"grad_norm": 54.90278625488281,
|
| 1672 |
+
"learning_rate": 8.483754512635379e-06,
|
| 1673 |
+
"loss": 0.0393,
|
| 1674 |
+
"step": 2300
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 4.162235241099594,
|
| 1678 |
+
"grad_norm": 49.901302337646484,
|
| 1679 |
+
"learning_rate": 8.303249097472924e-06,
|
| 1680 |
+
"loss": 0.0504,
|
| 1681 |
+
"step": 2310
|
| 1682 |
+
},
|
| 1683 |
+
{
|
| 1684 |
+
"epoch": 4.18026137899955,
|
| 1685 |
+
"grad_norm": 2.451421022415161,
|
| 1686 |
+
"learning_rate": 8.12274368231047e-06,
|
| 1687 |
+
"loss": 0.0422,
|
| 1688 |
+
"step": 2320
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 4.198287516899504,
|
| 1692 |
+
"grad_norm": 2.6284492015838623,
|
| 1693 |
+
"learning_rate": 7.942238267148016e-06,
|
| 1694 |
+
"loss": 0.0657,
|
| 1695 |
+
"step": 2330
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 4.2163136547994595,
|
| 1699 |
+
"grad_norm": 30.951091766357422,
|
| 1700 |
+
"learning_rate": 7.76173285198556e-06,
|
| 1701 |
+
"loss": 0.0391,
|
| 1702 |
+
"step": 2340
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"epoch": 4.234339792699414,
|
| 1706 |
+
"grad_norm": 83.39810180664062,
|
| 1707 |
+
"learning_rate": 7.581227436823105e-06,
|
| 1708 |
+
"loss": 0.0588,
|
| 1709 |
+
"step": 2350
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 4.252365930599369,
|
| 1713 |
+
"grad_norm": 23.396623611450195,
|
| 1714 |
+
"learning_rate": 7.4007220216606496e-06,
|
| 1715 |
+
"loss": 0.0269,
|
| 1716 |
+
"step": 2360
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 4.270392068499324,
|
| 1720 |
+
"grad_norm": 45.39535140991211,
|
| 1721 |
+
"learning_rate": 7.220216606498195e-06,
|
| 1722 |
+
"loss": 0.015,
|
| 1723 |
+
"step": 2370
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"epoch": 4.288418206399279,
|
| 1727 |
+
"grad_norm": 17.25420379638672,
|
| 1728 |
+
"learning_rate": 7.03971119133574e-06,
|
| 1729 |
+
"loss": 0.0458,
|
| 1730 |
+
"step": 2380
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 4.3064443442992335,
|
| 1734 |
+
"grad_norm": 72.16468811035156,
|
| 1735 |
+
"learning_rate": 6.859205776173286e-06,
|
| 1736 |
+
"loss": 0.089,
|
| 1737 |
+
"step": 2390
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 4.324470482199189,
|
| 1741 |
+
"grad_norm": 3.1122958660125732,
|
| 1742 |
+
"learning_rate": 6.678700361010831e-06,
|
| 1743 |
+
"loss": 0.0721,
|
| 1744 |
+
"step": 2400
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 4.342496620099144,
|
| 1748 |
+
"grad_norm": 22.552589416503906,
|
| 1749 |
+
"learning_rate": 6.498194945848375e-06,
|
| 1750 |
+
"loss": 0.0669,
|
| 1751 |
+
"step": 2410
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 4.360522757999099,
|
| 1755 |
+
"grad_norm": 15.435254096984863,
|
| 1756 |
+
"learning_rate": 6.31768953068592e-06,
|
| 1757 |
+
"loss": 0.0452,
|
| 1758 |
+
"step": 2420
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 4.378548895899054,
|
| 1762 |
+
"grad_norm": 30.996776580810547,
|
| 1763 |
+
"learning_rate": 6.137184115523466e-06,
|
| 1764 |
+
"loss": 0.0718,
|
| 1765 |
+
"step": 2430
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 4.396575033799008,
|
| 1769 |
+
"grad_norm": 44.87395477294922,
|
| 1770 |
+
"learning_rate": 5.956678700361011e-06,
|
| 1771 |
+
"loss": 0.0627,
|
| 1772 |
+
"step": 2440
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 4.414601171698964,
|
| 1776 |
+
"grad_norm": 64.19063568115234,
|
| 1777 |
+
"learning_rate": 5.776173285198557e-06,
|
| 1778 |
+
"loss": 0.0631,
|
| 1779 |
+
"step": 2450
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 4.432627309598918,
|
| 1783 |
+
"grad_norm": 60.25037384033203,
|
| 1784 |
+
"learning_rate": 5.595667870036101e-06,
|
| 1785 |
+
"loss": 0.0458,
|
| 1786 |
+
"step": 2460
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 4.4506534474988735,
|
| 1790 |
+
"grad_norm": 48.32729721069336,
|
| 1791 |
+
"learning_rate": 5.4151624548736465e-06,
|
| 1792 |
+
"loss": 0.0414,
|
| 1793 |
+
"step": 2470
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 4.468679585398828,
|
| 1797 |
+
"grad_norm": 121.8750228881836,
|
| 1798 |
+
"learning_rate": 5.2346570397111915e-06,
|
| 1799 |
+
"loss": 0.0358,
|
| 1800 |
+
"step": 2480
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 4.486705723298783,
|
| 1804 |
+
"grad_norm": 107.8392562866211,
|
| 1805 |
+
"learning_rate": 5.054151624548736e-06,
|
| 1806 |
+
"loss": 0.0455,
|
| 1807 |
+
"step": 2490
|
| 1808 |
+
},
|
| 1809 |
+
{
|
| 1810 |
+
"epoch": 4.504731861198739,
|
| 1811 |
+
"grad_norm": 307.8359375,
|
| 1812 |
+
"learning_rate": 4.873646209386281e-06,
|
| 1813 |
+
"loss": 0.0715,
|
| 1814 |
+
"step": 2500
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 4.522757999098693,
|
| 1818 |
+
"grad_norm": 2.502087116241455,
|
| 1819 |
+
"learning_rate": 4.693140794223827e-06,
|
| 1820 |
+
"loss": 0.0453,
|
| 1821 |
+
"step": 2510
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 4.540784136998648,
|
| 1825 |
+
"grad_norm": 22.081605911254883,
|
| 1826 |
+
"learning_rate": 4.512635379061372e-06,
|
| 1827 |
+
"loss": 0.0372,
|
| 1828 |
+
"step": 2520
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 4.558810274898603,
|
| 1832 |
+
"grad_norm": 90.6872329711914,
|
| 1833 |
+
"learning_rate": 4.332129963898917e-06,
|
| 1834 |
+
"loss": 0.0516,
|
| 1835 |
+
"step": 2530
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 4.576836412798558,
|
| 1839 |
+
"grad_norm": 82.89095306396484,
|
| 1840 |
+
"learning_rate": 4.151624548736462e-06,
|
| 1841 |
+
"loss": 0.0686,
|
| 1842 |
+
"step": 2540
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 4.594862550698513,
|
| 1846 |
+
"grad_norm": 7.579428672790527,
|
| 1847 |
+
"learning_rate": 3.971119133574008e-06,
|
| 1848 |
+
"loss": 0.0462,
|
| 1849 |
+
"step": 2550
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 4.612888688598468,
|
| 1853 |
+
"grad_norm": 21.60133171081543,
|
| 1854 |
+
"learning_rate": 3.7906137184115523e-06,
|
| 1855 |
+
"loss": 0.0783,
|
| 1856 |
+
"step": 2560
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 4.630914826498422,
|
| 1860 |
+
"grad_norm": 137.67620849609375,
|
| 1861 |
+
"learning_rate": 3.6101083032490977e-06,
|
| 1862 |
+
"loss": 0.0397,
|
| 1863 |
+
"step": 2570
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 4.648940964398378,
|
| 1867 |
+
"grad_norm": 122.53430938720703,
|
| 1868 |
+
"learning_rate": 3.429602888086643e-06,
|
| 1869 |
+
"loss": 0.0622,
|
| 1870 |
+
"step": 2580
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 4.666967102298333,
|
| 1874 |
+
"grad_norm": 15.056447982788086,
|
| 1875 |
+
"learning_rate": 3.2490974729241876e-06,
|
| 1876 |
+
"loss": 0.0295,
|
| 1877 |
+
"step": 2590
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 4.6849932401982874,
|
| 1881 |
+
"grad_norm": 2.0812671184539795,
|
| 1882 |
+
"learning_rate": 3.068592057761733e-06,
|
| 1883 |
+
"loss": 0.049,
|
| 1884 |
+
"step": 2600
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 4.703019378098243,
|
| 1888 |
+
"grad_norm": 46.54938507080078,
|
| 1889 |
+
"learning_rate": 2.8880866425992783e-06,
|
| 1890 |
+
"loss": 0.0473,
|
| 1891 |
+
"step": 2610
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 4.721045515998197,
|
| 1895 |
+
"grad_norm": 32.64075469970703,
|
| 1896 |
+
"learning_rate": 2.7075812274368233e-06,
|
| 1897 |
+
"loss": 0.0706,
|
| 1898 |
+
"step": 2620
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 4.7390716538981525,
|
| 1902 |
+
"grad_norm": 0.8347495794296265,
|
| 1903 |
+
"learning_rate": 2.527075812274368e-06,
|
| 1904 |
+
"loss": 0.0278,
|
| 1905 |
+
"step": 2630
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 4.757097791798107,
|
| 1909 |
+
"grad_norm": 80.3661117553711,
|
| 1910 |
+
"learning_rate": 2.3465703971119136e-06,
|
| 1911 |
+
"loss": 0.0757,
|
| 1912 |
+
"step": 2640
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 4.775123929698062,
|
| 1916 |
+
"grad_norm": 80.65737915039062,
|
| 1917 |
+
"learning_rate": 2.1660649819494585e-06,
|
| 1918 |
+
"loss": 0.0584,
|
| 1919 |
+
"step": 2650
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 4.793150067598017,
|
| 1923 |
+
"grad_norm": 31.22911834716797,
|
| 1924 |
+
"learning_rate": 1.985559566787004e-06,
|
| 1925 |
+
"loss": 0.0448,
|
| 1926 |
+
"step": 2660
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 4.811176205497972,
|
| 1930 |
+
"grad_norm": 50.2165641784668,
|
| 1931 |
+
"learning_rate": 1.8050541516245488e-06,
|
| 1932 |
+
"loss": 0.0218,
|
| 1933 |
+
"step": 2670
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 4.829202343397927,
|
| 1937 |
+
"grad_norm": 31.585453033447266,
|
| 1938 |
+
"learning_rate": 1.6245487364620938e-06,
|
| 1939 |
+
"loss": 0.0152,
|
| 1940 |
+
"step": 2680
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 4.847228481297882,
|
| 1944 |
+
"grad_norm": 46.6591796875,
|
| 1945 |
+
"learning_rate": 1.4440433212996392e-06,
|
| 1946 |
+
"loss": 0.0495,
|
| 1947 |
+
"step": 2690
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 4.865254619197837,
|
| 1951 |
+
"grad_norm": 50.74000549316406,
|
| 1952 |
+
"learning_rate": 1.263537906137184e-06,
|
| 1953 |
+
"loss": 0.0386,
|
| 1954 |
+
"step": 2700
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 4.883280757097792,
|
| 1958 |
+
"grad_norm": 1.2886587381362915,
|
| 1959 |
+
"learning_rate": 1.0830324909747293e-06,
|
| 1960 |
+
"loss": 0.0479,
|
| 1961 |
+
"step": 2710
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 4.901306894997747,
|
| 1965 |
+
"grad_norm": 27.440536499023438,
|
| 1966 |
+
"learning_rate": 9.025270758122744e-07,
|
| 1967 |
+
"loss": 0.0641,
|
| 1968 |
+
"step": 2720
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 4.919333032897701,
|
| 1972 |
+
"grad_norm": 84.98222351074219,
|
| 1973 |
+
"learning_rate": 7.220216606498196e-07,
|
| 1974 |
+
"loss": 0.0861,
|
| 1975 |
+
"step": 2730
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 4.937359170797657,
|
| 1979 |
+
"grad_norm": 55.485774993896484,
|
| 1980 |
+
"learning_rate": 5.415162454873646e-07,
|
| 1981 |
+
"loss": 0.0191,
|
| 1982 |
+
"step": 2740
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 4.955385308697611,
|
| 1986 |
+
"grad_norm": 15.337183952331543,
|
| 1987 |
+
"learning_rate": 3.610108303249098e-07,
|
| 1988 |
+
"loss": 0.0328,
|
| 1989 |
+
"step": 2750
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 4.9734114465975665,
|
| 1993 |
+
"grad_norm": 72.6976089477539,
|
| 1994 |
+
"learning_rate": 1.805054151624549e-07,
|
| 1995 |
+
"loss": 0.0364,
|
| 1996 |
+
"step": 2760
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 4.991437584497522,
|
| 2000 |
+
"grad_norm": 0.4997097849845886,
|
| 2001 |
+
"learning_rate": 0.0,
|
| 2002 |
+
"loss": 0.0301,
|
| 2003 |
+
"step": 2770
|
| 2004 |
+
}
|
| 2005 |
+
],
|
| 2006 |
+
"logging_steps": 10,
|
| 2007 |
+
"max_steps": 2770,
|
| 2008 |
+
"num_input_tokens_seen": 0,
|
| 2009 |
+
"num_train_epochs": 5,
|
| 2010 |
+
"save_steps": 500,
|
| 2011 |
+
"stateful_callbacks": {
|
| 2012 |
+
"TrainerControl": {
|
| 2013 |
+
"args": {
|
| 2014 |
+
"should_epoch_stop": false,
|
| 2015 |
+
"should_evaluate": false,
|
| 2016 |
+
"should_log": false,
|
| 2017 |
+
"should_save": true,
|
| 2018 |
+
"should_training_stop": true
|
| 2019 |
+
},
|
| 2020 |
+
"attributes": {}
|
| 2021 |
+
}
|
| 2022 |
+
},
|
| 2023 |
+
"total_flos": 2.331374439333888e+16,
|
| 2024 |
+
"train_batch_size": 8,
|
| 2025 |
+
"trial_name": null,
|
| 2026 |
+
"trial_params": null
|
| 2027 |
+
}
|
checkpoint-2770/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:92aaf17ba6b51dd0179acd28630a512a049c53024ba01863d7232c815d6018dd
|
| 3 |
+
size 5432
|