Upload 4 files
Browse files- config.json +64 -0
- model.safetensors +3 -0
- trainer_state.json +2253 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "answerdotai/ModernBERT-base",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ModernBertForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": false,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"bos_token_id": 50281,
|
| 9 |
+
"classifier_activation": "gelu",
|
| 10 |
+
"classifier_bias": false,
|
| 11 |
+
"classifier_dropout": 0.0,
|
| 12 |
+
"classifier_pooling": "mean",
|
| 13 |
+
"cls_token_id": 50281,
|
| 14 |
+
"decoder_bias": true,
|
| 15 |
+
"deterministic_flash_attn": false,
|
| 16 |
+
"embedding_dropout": 0.0,
|
| 17 |
+
"eos_token_id": 50282,
|
| 18 |
+
"global_attn_every_n_layers": 3,
|
| 19 |
+
"global_rope_theta": 160000.0,
|
| 20 |
+
"gradient_checkpointing": false,
|
| 21 |
+
"hidden_activation": "gelu",
|
| 22 |
+
"hidden_size": 768,
|
| 23 |
+
"id2label": {
|
| 24 |
+
"0": "class_0",
|
| 25 |
+
"1": "class_1",
|
| 26 |
+
"2": "class_2",
|
| 27 |
+
"3": "class_3",
|
| 28 |
+
"4": "class_4",
|
| 29 |
+
"5": "class_5"
|
| 30 |
+
},
|
| 31 |
+
"initializer_cutoff_factor": 2.0,
|
| 32 |
+
"initializer_range": 0.02,
|
| 33 |
+
"intermediate_size": 1152,
|
| 34 |
+
"label2id": {
|
| 35 |
+
"class_0": "0",
|
| 36 |
+
"class_1": "1",
|
| 37 |
+
"class_2": "2",
|
| 38 |
+
"class_3": "3",
|
| 39 |
+
"class_4": "4",
|
| 40 |
+
"class_5": "5"
|
| 41 |
+
},
|
| 42 |
+
"layer_norm_eps": 1e-05,
|
| 43 |
+
"local_attention": 128,
|
| 44 |
+
"local_rope_theta": 10000.0,
|
| 45 |
+
"max_position_embeddings": 8192,
|
| 46 |
+
"mlp_bias": false,
|
| 47 |
+
"mlp_dropout": 0.0,
|
| 48 |
+
"model_type": "modernbert",
|
| 49 |
+
"norm_bias": false,
|
| 50 |
+
"norm_eps": 1e-05,
|
| 51 |
+
"num_attention_heads": 12,
|
| 52 |
+
"num_hidden_layers": 22,
|
| 53 |
+
"pad_token_id": 50283,
|
| 54 |
+
"position_embedding_type": "absolute",
|
| 55 |
+
"problem_type": "single_label_classification",
|
| 56 |
+
"reference_compile": false,
|
| 57 |
+
"repad_logits_with_grad": false,
|
| 58 |
+
"sep_token_id": 50282,
|
| 59 |
+
"sparse_pred_ignore_index": -100,
|
| 60 |
+
"sparse_prediction": false,
|
| 61 |
+
"torch_dtype": "float32",
|
| 62 |
+
"transformers_version": "4.49.0",
|
| 63 |
+
"vocab_size": 50368
|
| 64 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:30e26eae53294f869a6fc8ec17755be44d5146c76bb28992a4def088376ede79
|
| 3 |
+
size 598452088
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2253 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.6869426704202687,
|
| 3 |
+
"best_model_checkpoint": "/userstorage/modernbert-llm-grader/checkpoint-31216",
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 31216,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.012813941568426449,
|
| 13 |
+
"grad_norm": 5.321617603302002,
|
| 14 |
+
"learning_rate": 4.987186058431574e-05,
|
| 15 |
+
"loss": 1.4033,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.025627883136852898,
|
| 20 |
+
"grad_norm": 3.621730089187622,
|
| 21 |
+
"learning_rate": 4.974372116863147e-05,
|
| 22 |
+
"loss": 1.3035,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.03844182470527934,
|
| 27 |
+
"grad_norm": 7.95962381362915,
|
| 28 |
+
"learning_rate": 4.961558175294721e-05,
|
| 29 |
+
"loss": 1.2506,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.051255766273705795,
|
| 34 |
+
"grad_norm": 3.631398916244507,
|
| 35 |
+
"learning_rate": 4.9487442337262944e-05,
|
| 36 |
+
"loss": 1.2354,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.06406970784213224,
|
| 41 |
+
"grad_norm": 2.6680333614349365,
|
| 42 |
+
"learning_rate": 4.935930292157868e-05,
|
| 43 |
+
"loss": 1.2397,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.07688364941055868,
|
| 48 |
+
"grad_norm": 6.042360305786133,
|
| 49 |
+
"learning_rate": 4.9231163505894415e-05,
|
| 50 |
+
"loss": 1.1811,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.08969759097898514,
|
| 55 |
+
"grad_norm": 6.7501959800720215,
|
| 56 |
+
"learning_rate": 4.9103024090210154e-05,
|
| 57 |
+
"loss": 1.1947,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.10251153254741159,
|
| 62 |
+
"grad_norm": 3.4089736938476562,
|
| 63 |
+
"learning_rate": 4.8974884674525886e-05,
|
| 64 |
+
"loss": 1.1812,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.11532547411583803,
|
| 69 |
+
"grad_norm": 5.0775604248046875,
|
| 70 |
+
"learning_rate": 4.884674525884162e-05,
|
| 71 |
+
"loss": 1.1752,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.12813941568426448,
|
| 76 |
+
"grad_norm": 4.529630184173584,
|
| 77 |
+
"learning_rate": 4.8718605843157357e-05,
|
| 78 |
+
"loss": 1.1867,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.14095335725269093,
|
| 83 |
+
"grad_norm": 4.961220741271973,
|
| 84 |
+
"learning_rate": 4.859046642747309e-05,
|
| 85 |
+
"loss": 1.2129,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.15376729882111737,
|
| 90 |
+
"grad_norm": 4.113813400268555,
|
| 91 |
+
"learning_rate": 4.846232701178883e-05,
|
| 92 |
+
"loss": 1.1293,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.16658124038954383,
|
| 97 |
+
"grad_norm": 7.25917387008667,
|
| 98 |
+
"learning_rate": 4.8334187596104566e-05,
|
| 99 |
+
"loss": 1.1008,
|
| 100 |
+
"step": 1300
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.17939518195797027,
|
| 104 |
+
"grad_norm": 5.579372882843018,
|
| 105 |
+
"learning_rate": 4.82060481804203e-05,
|
| 106 |
+
"loss": 1.1327,
|
| 107 |
+
"step": 1400
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.1922091235263967,
|
| 111 |
+
"grad_norm": 9.794898986816406,
|
| 112 |
+
"learning_rate": 4.807790876473604e-05,
|
| 113 |
+
"loss": 1.1355,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.20502306509482318,
|
| 118 |
+
"grad_norm": 9.875951766967773,
|
| 119 |
+
"learning_rate": 4.794976934905177e-05,
|
| 120 |
+
"loss": 1.0057,
|
| 121 |
+
"step": 1600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.21783700666324962,
|
| 125 |
+
"grad_norm": 7.271333694458008,
|
| 126 |
+
"learning_rate": 4.782162993336751e-05,
|
| 127 |
+
"loss": 1.1101,
|
| 128 |
+
"step": 1700
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.23065094823167606,
|
| 132 |
+
"grad_norm": 6.730026721954346,
|
| 133 |
+
"learning_rate": 4.769349051768324e-05,
|
| 134 |
+
"loss": 1.0762,
|
| 135 |
+
"step": 1800
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.2434648898001025,
|
| 139 |
+
"grad_norm": 5.596224784851074,
|
| 140 |
+
"learning_rate": 4.756535110199898e-05,
|
| 141 |
+
"loss": 1.0413,
|
| 142 |
+
"step": 1900
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.25627883136852897,
|
| 146 |
+
"grad_norm": 4.591865539550781,
|
| 147 |
+
"learning_rate": 4.743721168631472e-05,
|
| 148 |
+
"loss": 1.0593,
|
| 149 |
+
"step": 2000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.2690927729369554,
|
| 153 |
+
"grad_norm": 6.357232570648193,
|
| 154 |
+
"learning_rate": 4.730907227063045e-05,
|
| 155 |
+
"loss": 1.0434,
|
| 156 |
+
"step": 2100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.28190671450538185,
|
| 160 |
+
"grad_norm": 5.185873508453369,
|
| 161 |
+
"learning_rate": 4.718093285494619e-05,
|
| 162 |
+
"loss": 1.021,
|
| 163 |
+
"step": 2200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.2947206560738083,
|
| 167 |
+
"grad_norm": 8.19482135772705,
|
| 168 |
+
"learning_rate": 4.705279343926192e-05,
|
| 169 |
+
"loss": 1.1102,
|
| 170 |
+
"step": 2300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.30753459764223473,
|
| 174 |
+
"grad_norm": 6.1499176025390625,
|
| 175 |
+
"learning_rate": 4.692465402357765e-05,
|
| 176 |
+
"loss": 1.0115,
|
| 177 |
+
"step": 2400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.3203485392106612,
|
| 181 |
+
"grad_norm": 11.092570304870605,
|
| 182 |
+
"learning_rate": 4.679651460789339e-05,
|
| 183 |
+
"loss": 0.9576,
|
| 184 |
+
"step": 2500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.33316248077908767,
|
| 188 |
+
"grad_norm": 5.10243034362793,
|
| 189 |
+
"learning_rate": 4.666837519220912e-05,
|
| 190 |
+
"loss": 1.0674,
|
| 191 |
+
"step": 2600
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.3459764223475141,
|
| 195 |
+
"grad_norm": 4.633431434631348,
|
| 196 |
+
"learning_rate": 4.654023577652486e-05,
|
| 197 |
+
"loss": 1.0004,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.35879036391594055,
|
| 202 |
+
"grad_norm": 5.507874488830566,
|
| 203 |
+
"learning_rate": 4.6412096360840594e-05,
|
| 204 |
+
"loss": 1.0439,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.371604305484367,
|
| 209 |
+
"grad_norm": 5.591798305511475,
|
| 210 |
+
"learning_rate": 4.628395694515633e-05,
|
| 211 |
+
"loss": 1.0509,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.3844182470527934,
|
| 216 |
+
"grad_norm": 4.341959476470947,
|
| 217 |
+
"learning_rate": 4.6155817529472065e-05,
|
| 218 |
+
"loss": 0.9864,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.3972321886212199,
|
| 223 |
+
"grad_norm": 3.6542482376098633,
|
| 224 |
+
"learning_rate": 4.6027678113787804e-05,
|
| 225 |
+
"loss": 0.9897,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.41004613018964636,
|
| 230 |
+
"grad_norm": 6.769758701324463,
|
| 231 |
+
"learning_rate": 4.589953869810354e-05,
|
| 232 |
+
"loss": 1.0528,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.4228600717580728,
|
| 237 |
+
"grad_norm": 5.762277603149414,
|
| 238 |
+
"learning_rate": 4.5771399282419274e-05,
|
| 239 |
+
"loss": 1.0036,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.43567401332649924,
|
| 244 |
+
"grad_norm": 7.389179229736328,
|
| 245 |
+
"learning_rate": 4.564325986673501e-05,
|
| 246 |
+
"loss": 1.0304,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.44848795489492566,
|
| 251 |
+
"grad_norm": 3.9039294719696045,
|
| 252 |
+
"learning_rate": 4.5515120451050745e-05,
|
| 253 |
+
"loss": 0.9962,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.4613018964633521,
|
| 258 |
+
"grad_norm": 3.0561447143554688,
|
| 259 |
+
"learning_rate": 4.5386981035366484e-05,
|
| 260 |
+
"loss": 0.9777,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.4741158380317786,
|
| 265 |
+
"grad_norm": 6.340303897857666,
|
| 266 |
+
"learning_rate": 4.5258841619682216e-05,
|
| 267 |
+
"loss": 0.9603,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.486929779600205,
|
| 272 |
+
"grad_norm": 9.058144569396973,
|
| 273 |
+
"learning_rate": 4.5130702203997955e-05,
|
| 274 |
+
"loss": 0.9658,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.49974372116863147,
|
| 279 |
+
"grad_norm": 8.219672203063965,
|
| 280 |
+
"learning_rate": 4.500256278831369e-05,
|
| 281 |
+
"loss": 0.9856,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.5125576627370579,
|
| 286 |
+
"grad_norm": 3.6466543674468994,
|
| 287 |
+
"learning_rate": 4.487442337262942e-05,
|
| 288 |
+
"loss": 0.9734,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.5253716043054844,
|
| 293 |
+
"grad_norm": 7.289781093597412,
|
| 294 |
+
"learning_rate": 4.474628395694516e-05,
|
| 295 |
+
"loss": 0.9045,
|
| 296 |
+
"step": 4100
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.5381855458739108,
|
| 300 |
+
"grad_norm": 6.18227481842041,
|
| 301 |
+
"learning_rate": 4.461814454126089e-05,
|
| 302 |
+
"loss": 0.9679,
|
| 303 |
+
"step": 4200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.5509994874423373,
|
| 307 |
+
"grad_norm": 3.994476318359375,
|
| 308 |
+
"learning_rate": 4.449000512557663e-05,
|
| 309 |
+
"loss": 0.8958,
|
| 310 |
+
"step": 4300
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.5638134290107637,
|
| 314 |
+
"grad_norm": 3.913896322250366,
|
| 315 |
+
"learning_rate": 4.436186570989236e-05,
|
| 316 |
+
"loss": 0.9453,
|
| 317 |
+
"step": 4400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.5766273705791901,
|
| 321 |
+
"grad_norm": 4.39192008972168,
|
| 322 |
+
"learning_rate": 4.42337262942081e-05,
|
| 323 |
+
"loss": 0.9132,
|
| 324 |
+
"step": 4500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.5894413121476166,
|
| 328 |
+
"grad_norm": 5.574671745300293,
|
| 329 |
+
"learning_rate": 4.410558687852384e-05,
|
| 330 |
+
"loss": 0.9069,
|
| 331 |
+
"step": 4600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.602255253716043,
|
| 335 |
+
"grad_norm": 4.218778610229492,
|
| 336 |
+
"learning_rate": 4.397744746283957e-05,
|
| 337 |
+
"loss": 0.9631,
|
| 338 |
+
"step": 4700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.6150691952844695,
|
| 342 |
+
"grad_norm": 7.804980754852295,
|
| 343 |
+
"learning_rate": 4.384930804715531e-05,
|
| 344 |
+
"loss": 0.9121,
|
| 345 |
+
"step": 4800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.627883136852896,
|
| 349 |
+
"grad_norm": 7.064172744750977,
|
| 350 |
+
"learning_rate": 4.372116863147104e-05,
|
| 351 |
+
"loss": 0.9387,
|
| 352 |
+
"step": 4900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.6406970784213224,
|
| 356 |
+
"grad_norm": 5.293111324310303,
|
| 357 |
+
"learning_rate": 4.359302921578678e-05,
|
| 358 |
+
"loss": 0.9264,
|
| 359 |
+
"step": 5000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.6535110199897488,
|
| 363 |
+
"grad_norm": 7.019448757171631,
|
| 364 |
+
"learning_rate": 4.346488980010251e-05,
|
| 365 |
+
"loss": 0.9452,
|
| 366 |
+
"step": 5100
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.6663249615581753,
|
| 370 |
+
"grad_norm": 6.714709758758545,
|
| 371 |
+
"learning_rate": 4.333675038441825e-05,
|
| 372 |
+
"loss": 0.8648,
|
| 373 |
+
"step": 5200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.6791389031266017,
|
| 377 |
+
"grad_norm": 8.232748031616211,
|
| 378 |
+
"learning_rate": 4.320861096873399e-05,
|
| 379 |
+
"loss": 0.904,
|
| 380 |
+
"step": 5300
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.6919528446950282,
|
| 384 |
+
"grad_norm": 9.853933334350586,
|
| 385 |
+
"learning_rate": 4.308047155304972e-05,
|
| 386 |
+
"loss": 0.8895,
|
| 387 |
+
"step": 5400
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.7047667862634547,
|
| 391 |
+
"grad_norm": 6.8710455894470215,
|
| 392 |
+
"learning_rate": 4.2952332137365454e-05,
|
| 393 |
+
"loss": 0.86,
|
| 394 |
+
"step": 5500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.7175807278318811,
|
| 398 |
+
"grad_norm": 6.45287561416626,
|
| 399 |
+
"learning_rate": 4.2824192721681186e-05,
|
| 400 |
+
"loss": 0.8718,
|
| 401 |
+
"step": 5600
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.7303946694003075,
|
| 405 |
+
"grad_norm": 5.772899627685547,
|
| 406 |
+
"learning_rate": 4.2696053305996924e-05,
|
| 407 |
+
"loss": 0.8477,
|
| 408 |
+
"step": 5700
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.743208610968734,
|
| 412 |
+
"grad_norm": 6.193540573120117,
|
| 413 |
+
"learning_rate": 4.256791389031266e-05,
|
| 414 |
+
"loss": 0.9184,
|
| 415 |
+
"step": 5800
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.7560225525371604,
|
| 419 |
+
"grad_norm": 2.5397393703460693,
|
| 420 |
+
"learning_rate": 4.2439774474628395e-05,
|
| 421 |
+
"loss": 0.9537,
|
| 422 |
+
"step": 5900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.7688364941055869,
|
| 426 |
+
"grad_norm": 8.280569076538086,
|
| 427 |
+
"learning_rate": 4.2311635058944134e-05,
|
| 428 |
+
"loss": 0.8988,
|
| 429 |
+
"step": 6000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.7816504356740134,
|
| 433 |
+
"grad_norm": 10.563502311706543,
|
| 434 |
+
"learning_rate": 4.2183495643259866e-05,
|
| 435 |
+
"loss": 0.8518,
|
| 436 |
+
"step": 6100
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.7944643772424398,
|
| 440 |
+
"grad_norm": 3.090008497238159,
|
| 441 |
+
"learning_rate": 4.2055356227575605e-05,
|
| 442 |
+
"loss": 0.8731,
|
| 443 |
+
"step": 6200
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.8072783188108662,
|
| 447 |
+
"grad_norm": 4.051167011260986,
|
| 448 |
+
"learning_rate": 4.192721681189134e-05,
|
| 449 |
+
"loss": 0.8713,
|
| 450 |
+
"step": 6300
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.8200922603792927,
|
| 454 |
+
"grad_norm": 7.207763671875,
|
| 455 |
+
"learning_rate": 4.1799077396207076e-05,
|
| 456 |
+
"loss": 0.8781,
|
| 457 |
+
"step": 6400
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.8329062019477191,
|
| 461 |
+
"grad_norm": 6.396823883056641,
|
| 462 |
+
"learning_rate": 4.1670937980522815e-05,
|
| 463 |
+
"loss": 0.8231,
|
| 464 |
+
"step": 6500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.8457201435161456,
|
| 468 |
+
"grad_norm": 6.260582447052002,
|
| 469 |
+
"learning_rate": 4.1542798564838547e-05,
|
| 470 |
+
"loss": 0.8658,
|
| 471 |
+
"step": 6600
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.858534085084572,
|
| 475 |
+
"grad_norm": 8.35356616973877,
|
| 476 |
+
"learning_rate": 4.1414659149154285e-05,
|
| 477 |
+
"loss": 0.8637,
|
| 478 |
+
"step": 6700
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.8713480266529985,
|
| 482 |
+
"grad_norm": 7.236725330352783,
|
| 483 |
+
"learning_rate": 4.128651973347002e-05,
|
| 484 |
+
"loss": 0.8525,
|
| 485 |
+
"step": 6800
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.8841619682214249,
|
| 489 |
+
"grad_norm": 14.001522064208984,
|
| 490 |
+
"learning_rate": 4.1158380317785756e-05,
|
| 491 |
+
"loss": 0.8628,
|
| 492 |
+
"step": 6900
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.8969759097898513,
|
| 496 |
+
"grad_norm": 4.257541179656982,
|
| 497 |
+
"learning_rate": 4.103024090210149e-05,
|
| 498 |
+
"loss": 0.8443,
|
| 499 |
+
"step": 7000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.9097898513582778,
|
| 503 |
+
"grad_norm": 5.065970420837402,
|
| 504 |
+
"learning_rate": 4.090210148641722e-05,
|
| 505 |
+
"loss": 0.8329,
|
| 506 |
+
"step": 7100
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.9226037929267042,
|
| 510 |
+
"grad_norm": 6.647068977355957,
|
| 511 |
+
"learning_rate": 4.077396207073296e-05,
|
| 512 |
+
"loss": 0.8585,
|
| 513 |
+
"step": 7200
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.9354177344951307,
|
| 517 |
+
"grad_norm": 8.440242767333984,
|
| 518 |
+
"learning_rate": 4.064582265504869e-05,
|
| 519 |
+
"loss": 0.8749,
|
| 520 |
+
"step": 7300
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.9482316760635572,
|
| 524 |
+
"grad_norm": 7.684078216552734,
|
| 525 |
+
"learning_rate": 4.051768323936443e-05,
|
| 526 |
+
"loss": 0.7771,
|
| 527 |
+
"step": 7400
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.9610456176319836,
|
| 531 |
+
"grad_norm": 6.4709577560424805,
|
| 532 |
+
"learning_rate": 4.038954382368016e-05,
|
| 533 |
+
"loss": 0.8597,
|
| 534 |
+
"step": 7500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.97385955920041,
|
| 538 |
+
"grad_norm": 4.3970489501953125,
|
| 539 |
+
"learning_rate": 4.02614044079959e-05,
|
| 540 |
+
"loss": 0.7852,
|
| 541 |
+
"step": 7600
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.9866735007688365,
|
| 545 |
+
"grad_norm": 9.167794227600098,
|
| 546 |
+
"learning_rate": 4.013326499231164e-05,
|
| 547 |
+
"loss": 0.8563,
|
| 548 |
+
"step": 7700
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.9994874423372629,
|
| 552 |
+
"grad_norm": 6.251096248626709,
|
| 553 |
+
"learning_rate": 4.000512557662737e-05,
|
| 554 |
+
"loss": 0.8243,
|
| 555 |
+
"step": 7800
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.0,
|
| 559 |
+
"eval_f1": 0.640692076906927,
|
| 560 |
+
"eval_loss": 0.8794865608215332,
|
| 561 |
+
"eval_runtime": 744.6214,
|
| 562 |
+
"eval_samples_per_second": 10.48,
|
| 563 |
+
"eval_steps_per_second": 2.62,
|
| 564 |
+
"step": 7804
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.0123013839056894,
|
| 568 |
+
"grad_norm": 5.928829669952393,
|
| 569 |
+
"learning_rate": 3.987698616094311e-05,
|
| 570 |
+
"loss": 0.7046,
|
| 571 |
+
"step": 7900
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.0251153254741159,
|
| 575 |
+
"grad_norm": 2.885106086730957,
|
| 576 |
+
"learning_rate": 3.974884674525884e-05,
|
| 577 |
+
"loss": 0.7663,
|
| 578 |
+
"step": 8000
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.0379292670425422,
|
| 582 |
+
"grad_norm": 5.951350212097168,
|
| 583 |
+
"learning_rate": 3.962070732957458e-05,
|
| 584 |
+
"loss": 0.7374,
|
| 585 |
+
"step": 8100
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.0507432086109687,
|
| 589 |
+
"grad_norm": 2.5160486698150635,
|
| 590 |
+
"learning_rate": 3.949256791389031e-05,
|
| 591 |
+
"loss": 0.7126,
|
| 592 |
+
"step": 8200
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.0635571501793952,
|
| 596 |
+
"grad_norm": 6.847401142120361,
|
| 597 |
+
"learning_rate": 3.936442849820605e-05,
|
| 598 |
+
"loss": 0.6785,
|
| 599 |
+
"step": 8300
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.0763710917478215,
|
| 603 |
+
"grad_norm": 4.729136943817139,
|
| 604 |
+
"learning_rate": 3.923628908252179e-05,
|
| 605 |
+
"loss": 0.7085,
|
| 606 |
+
"step": 8400
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.089185033316248,
|
| 610 |
+
"grad_norm": 5.535890102386475,
|
| 611 |
+
"learning_rate": 3.910814966683752e-05,
|
| 612 |
+
"loss": 0.7548,
|
| 613 |
+
"step": 8500
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.1019989748846746,
|
| 617 |
+
"grad_norm": 6.188892364501953,
|
| 618 |
+
"learning_rate": 3.8980010251153255e-05,
|
| 619 |
+
"loss": 0.7193,
|
| 620 |
+
"step": 8600
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.1148129164531009,
|
| 624 |
+
"grad_norm": 5.806282997131348,
|
| 625 |
+
"learning_rate": 3.885187083546899e-05,
|
| 626 |
+
"loss": 0.7143,
|
| 627 |
+
"step": 8700
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.1276268580215274,
|
| 631 |
+
"grad_norm": 10.726571083068848,
|
| 632 |
+
"learning_rate": 3.8723731419784726e-05,
|
| 633 |
+
"loss": 0.6892,
|
| 634 |
+
"step": 8800
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.140440799589954,
|
| 638 |
+
"grad_norm": 7.0307512283325195,
|
| 639 |
+
"learning_rate": 3.8595592004100465e-05,
|
| 640 |
+
"loss": 0.7264,
|
| 641 |
+
"step": 8900
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.1532547411583802,
|
| 645 |
+
"grad_norm": 20.715412139892578,
|
| 646 |
+
"learning_rate": 3.8467452588416197e-05,
|
| 647 |
+
"loss": 0.6987,
|
| 648 |
+
"step": 9000
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.1660686827268067,
|
| 652 |
+
"grad_norm": 6.620629787445068,
|
| 653 |
+
"learning_rate": 3.8339313172731935e-05,
|
| 654 |
+
"loss": 0.7041,
|
| 655 |
+
"step": 9100
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.1788826242952333,
|
| 659 |
+
"grad_norm": 5.27125883102417,
|
| 660 |
+
"learning_rate": 3.821117375704767e-05,
|
| 661 |
+
"loss": 0.67,
|
| 662 |
+
"step": 9200
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.1916965658636596,
|
| 666 |
+
"grad_norm": 6.010765552520752,
|
| 667 |
+
"learning_rate": 3.8083034341363406e-05,
|
| 668 |
+
"loss": 0.6737,
|
| 669 |
+
"step": 9300
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 1.204510507432086,
|
| 673 |
+
"grad_norm": 14.393863677978516,
|
| 674 |
+
"learning_rate": 3.795489492567914e-05,
|
| 675 |
+
"loss": 0.7097,
|
| 676 |
+
"step": 9400
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.2173244490005126,
|
| 680 |
+
"grad_norm": 6.37823486328125,
|
| 681 |
+
"learning_rate": 3.782675550999488e-05,
|
| 682 |
+
"loss": 0.7157,
|
| 683 |
+
"step": 9500
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.230138390568939,
|
| 687 |
+
"grad_norm": 11.626152992248535,
|
| 688 |
+
"learning_rate": 3.7698616094310616e-05,
|
| 689 |
+
"loss": 0.7066,
|
| 690 |
+
"step": 9600
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.2429523321373654,
|
| 694 |
+
"grad_norm": 5.520190238952637,
|
| 695 |
+
"learning_rate": 3.757047667862635e-05,
|
| 696 |
+
"loss": 0.7303,
|
| 697 |
+
"step": 9700
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.255766273705792,
|
| 701 |
+
"grad_norm": 9.865089416503906,
|
| 702 |
+
"learning_rate": 3.744233726294209e-05,
|
| 703 |
+
"loss": 0.7559,
|
| 704 |
+
"step": 9800
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.2685802152742183,
|
| 708 |
+
"grad_norm": 7.075952529907227,
|
| 709 |
+
"learning_rate": 3.731419784725782e-05,
|
| 710 |
+
"loss": 0.6941,
|
| 711 |
+
"step": 9900
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.2813941568426448,
|
| 715 |
+
"grad_norm": 3.4892656803131104,
|
| 716 |
+
"learning_rate": 3.718605843157356e-05,
|
| 717 |
+
"loss": 0.7164,
|
| 718 |
+
"step": 10000
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.2942080984110713,
|
| 722 |
+
"grad_norm": 9.843413352966309,
|
| 723 |
+
"learning_rate": 3.705791901588929e-05,
|
| 724 |
+
"loss": 0.695,
|
| 725 |
+
"step": 10100
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.3070220399794976,
|
| 729 |
+
"grad_norm": 12.128110885620117,
|
| 730 |
+
"learning_rate": 3.692977960020502e-05,
|
| 731 |
+
"loss": 0.6563,
|
| 732 |
+
"step": 10200
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.3198359815479241,
|
| 736 |
+
"grad_norm": 11.26876163482666,
|
| 737 |
+
"learning_rate": 3.680164018452076e-05,
|
| 738 |
+
"loss": 0.6803,
|
| 739 |
+
"step": 10300
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.3326499231163507,
|
| 743 |
+
"grad_norm": 12.95758056640625,
|
| 744 |
+
"learning_rate": 3.667350076883649e-05,
|
| 745 |
+
"loss": 0.6864,
|
| 746 |
+
"step": 10400
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.345463864684777,
|
| 750 |
+
"grad_norm": 4.91602897644043,
|
| 751 |
+
"learning_rate": 3.654536135315223e-05,
|
| 752 |
+
"loss": 0.7184,
|
| 753 |
+
"step": 10500
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.3582778062532035,
|
| 757 |
+
"grad_norm": 4.799069881439209,
|
| 758 |
+
"learning_rate": 3.641722193746796e-05,
|
| 759 |
+
"loss": 0.7558,
|
| 760 |
+
"step": 10600
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.37109174782163,
|
| 764 |
+
"grad_norm": 64.9485855102539,
|
| 765 |
+
"learning_rate": 3.62890825217837e-05,
|
| 766 |
+
"loss": 0.7292,
|
| 767 |
+
"step": 10700
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.3839056893900563,
|
| 771 |
+
"grad_norm": 6.147428512573242,
|
| 772 |
+
"learning_rate": 3.616094310609944e-05,
|
| 773 |
+
"loss": 0.6623,
|
| 774 |
+
"step": 10800
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.3967196309584828,
|
| 778 |
+
"grad_norm": 7.638481140136719,
|
| 779 |
+
"learning_rate": 3.603280369041517e-05,
|
| 780 |
+
"loss": 0.6981,
|
| 781 |
+
"step": 10900
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.4095335725269091,
|
| 785 |
+
"grad_norm": 4.798500061035156,
|
| 786 |
+
"learning_rate": 3.590466427473091e-05,
|
| 787 |
+
"loss": 0.7569,
|
| 788 |
+
"step": 11000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.4223475140953357,
|
| 792 |
+
"grad_norm": 4.413691520690918,
|
| 793 |
+
"learning_rate": 3.5776524859046644e-05,
|
| 794 |
+
"loss": 0.6391,
|
| 795 |
+
"step": 11100
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.4351614556637622,
|
| 799 |
+
"grad_norm": 6.2526421546936035,
|
| 800 |
+
"learning_rate": 3.564838544336238e-05,
|
| 801 |
+
"loss": 0.7045,
|
| 802 |
+
"step": 11200
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.4479753972321885,
|
| 806 |
+
"grad_norm": 6.3732805252075195,
|
| 807 |
+
"learning_rate": 3.5520246027678114e-05,
|
| 808 |
+
"loss": 0.6916,
|
| 809 |
+
"step": 11300
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.460789338800615,
|
| 813 |
+
"grad_norm": 25.24698829650879,
|
| 814 |
+
"learning_rate": 3.539210661199385e-05,
|
| 815 |
+
"loss": 0.7652,
|
| 816 |
+
"step": 11400
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.4736032803690415,
|
| 820 |
+
"grad_norm": 4.716599941253662,
|
| 821 |
+
"learning_rate": 3.5263967196309585e-05,
|
| 822 |
+
"loss": 0.7199,
|
| 823 |
+
"step": 11500
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.4864172219374678,
|
| 827 |
+
"grad_norm": 13.750917434692383,
|
| 828 |
+
"learning_rate": 3.5135827780625324e-05,
|
| 829 |
+
"loss": 0.7032,
|
| 830 |
+
"step": 11600
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.4992311635058944,
|
| 834 |
+
"grad_norm": 3.6678273677825928,
|
| 835 |
+
"learning_rate": 3.500768836494106e-05,
|
| 836 |
+
"loss": 0.6821,
|
| 837 |
+
"step": 11700
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 1.5120451050743209,
|
| 841 |
+
"grad_norm": 7.891080856323242,
|
| 842 |
+
"learning_rate": 3.487954894925679e-05,
|
| 843 |
+
"loss": 0.7301,
|
| 844 |
+
"step": 11800
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 1.5248590466427472,
|
| 848 |
+
"grad_norm": 3.25317645072937,
|
| 849 |
+
"learning_rate": 3.475140953357253e-05,
|
| 850 |
+
"loss": 0.6665,
|
| 851 |
+
"step": 11900
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 1.5376729882111737,
|
| 855 |
+
"grad_norm": 12.75395679473877,
|
| 856 |
+
"learning_rate": 3.462327011788826e-05,
|
| 857 |
+
"loss": 0.733,
|
| 858 |
+
"step": 12000
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 1.5504869297796002,
|
| 862 |
+
"grad_norm": 10.9820556640625,
|
| 863 |
+
"learning_rate": 3.4495130702204e-05,
|
| 864 |
+
"loss": 0.7064,
|
| 865 |
+
"step": 12100
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 1.5633008713480265,
|
| 869 |
+
"grad_norm": 6.558383941650391,
|
| 870 |
+
"learning_rate": 3.4366991286519737e-05,
|
| 871 |
+
"loss": 0.7105,
|
| 872 |
+
"step": 12200
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.576114812916453,
|
| 876 |
+
"grad_norm": 8.5501070022583,
|
| 877 |
+
"learning_rate": 3.423885187083547e-05,
|
| 878 |
+
"loss": 0.706,
|
| 879 |
+
"step": 12300
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.5889287544848796,
|
| 883 |
+
"grad_norm": 5.319694995880127,
|
| 884 |
+
"learning_rate": 3.411071245515121e-05,
|
| 885 |
+
"loss": 0.7239,
|
| 886 |
+
"step": 12400
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.6017426960533059,
|
| 890 |
+
"grad_norm": 5.92519474029541,
|
| 891 |
+
"learning_rate": 3.398257303946694e-05,
|
| 892 |
+
"loss": 0.7043,
|
| 893 |
+
"step": 12500
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 1.6145566376217324,
|
| 897 |
+
"grad_norm": 8.853275299072266,
|
| 898 |
+
"learning_rate": 3.385443362378268e-05,
|
| 899 |
+
"loss": 0.6831,
|
| 900 |
+
"step": 12600
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.627370579190159,
|
| 904 |
+
"grad_norm": 9.30588150024414,
|
| 905 |
+
"learning_rate": 3.372629420809841e-05,
|
| 906 |
+
"loss": 0.6756,
|
| 907 |
+
"step": 12700
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.6401845207585852,
|
| 911 |
+
"grad_norm": 5.903197288513184,
|
| 912 |
+
"learning_rate": 3.359815479241415e-05,
|
| 913 |
+
"loss": 0.725,
|
| 914 |
+
"step": 12800
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.6529984623270118,
|
| 918 |
+
"grad_norm": 5.500326156616211,
|
| 919 |
+
"learning_rate": 3.347001537672989e-05,
|
| 920 |
+
"loss": 0.6801,
|
| 921 |
+
"step": 12900
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.6658124038954383,
|
| 925 |
+
"grad_norm": 7.896096229553223,
|
| 926 |
+
"learning_rate": 3.334187596104562e-05,
|
| 927 |
+
"loss": 0.6975,
|
| 928 |
+
"step": 13000
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.6786263454638646,
|
| 932 |
+
"grad_norm": 6.674001216888428,
|
| 933 |
+
"learning_rate": 3.321373654536136e-05,
|
| 934 |
+
"loss": 0.6681,
|
| 935 |
+
"step": 13100
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.691440287032291,
|
| 939 |
+
"grad_norm": 21.74435806274414,
|
| 940 |
+
"learning_rate": 3.308559712967709e-05,
|
| 941 |
+
"loss": 0.7045,
|
| 942 |
+
"step": 13200
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.7042542286007176,
|
| 946 |
+
"grad_norm": 6.329532146453857,
|
| 947 |
+
"learning_rate": 3.295745771399282e-05,
|
| 948 |
+
"loss": 0.6885,
|
| 949 |
+
"step": 13300
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.717068170169144,
|
| 953 |
+
"grad_norm": 24.047470092773438,
|
| 954 |
+
"learning_rate": 3.282931829830856e-05,
|
| 955 |
+
"loss": 0.7003,
|
| 956 |
+
"step": 13400
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.7298821117375704,
|
| 960 |
+
"grad_norm": 7.407759666442871,
|
| 961 |
+
"learning_rate": 3.2701178882624294e-05,
|
| 962 |
+
"loss": 0.6856,
|
| 963 |
+
"step": 13500
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.742696053305997,
|
| 967 |
+
"grad_norm": 5.755215167999268,
|
| 968 |
+
"learning_rate": 3.257303946694003e-05,
|
| 969 |
+
"loss": 0.7005,
|
| 970 |
+
"step": 13600
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.7555099948744233,
|
| 974 |
+
"grad_norm": 11.444562911987305,
|
| 975 |
+
"learning_rate": 3.2444900051255764e-05,
|
| 976 |
+
"loss": 0.7136,
|
| 977 |
+
"step": 13700
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 1.7683239364428498,
|
| 981 |
+
"grad_norm": 8.267853736877441,
|
| 982 |
+
"learning_rate": 3.23167606355715e-05,
|
| 983 |
+
"loss": 0.7029,
|
| 984 |
+
"step": 13800
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.7811378780112763,
|
| 988 |
+
"grad_norm": 6.73785924911499,
|
| 989 |
+
"learning_rate": 3.2188621219887235e-05,
|
| 990 |
+
"loss": 0.6572,
|
| 991 |
+
"step": 13900
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 1.7939518195797026,
|
| 995 |
+
"grad_norm": 5.369395732879639,
|
| 996 |
+
"learning_rate": 3.2060481804202974e-05,
|
| 997 |
+
"loss": 0.6617,
|
| 998 |
+
"step": 14000
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.8067657611481291,
|
| 1002 |
+
"grad_norm": 2.288243293762207,
|
| 1003 |
+
"learning_rate": 3.193234238851871e-05,
|
| 1004 |
+
"loss": 0.6688,
|
| 1005 |
+
"step": 14100
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.8195797027165557,
|
| 1009 |
+
"grad_norm": 14.942804336547852,
|
| 1010 |
+
"learning_rate": 3.1804202972834445e-05,
|
| 1011 |
+
"loss": 0.6792,
|
| 1012 |
+
"step": 14200
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.832393644284982,
|
| 1016 |
+
"grad_norm": 8.988631248474121,
|
| 1017 |
+
"learning_rate": 3.1676063557150184e-05,
|
| 1018 |
+
"loss": 0.6518,
|
| 1019 |
+
"step": 14300
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 1.8452075858534085,
|
| 1023 |
+
"grad_norm": 7.9590630531311035,
|
| 1024 |
+
"learning_rate": 3.1547924141465916e-05,
|
| 1025 |
+
"loss": 0.6503,
|
| 1026 |
+
"step": 14400
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 1.858021527421835,
|
| 1030 |
+
"grad_norm": 9.33973503112793,
|
| 1031 |
+
"learning_rate": 3.1419784725781655e-05,
|
| 1032 |
+
"loss": 0.6647,
|
| 1033 |
+
"step": 14500
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 1.8708354689902613,
|
| 1037 |
+
"grad_norm": 9.39842700958252,
|
| 1038 |
+
"learning_rate": 3.1291645310097387e-05,
|
| 1039 |
+
"loss": 0.6515,
|
| 1040 |
+
"step": 14600
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.8836494105586878,
|
| 1044 |
+
"grad_norm": 10.142439842224121,
|
| 1045 |
+
"learning_rate": 3.1163505894413125e-05,
|
| 1046 |
+
"loss": 0.6794,
|
| 1047 |
+
"step": 14700
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 1.8964633521271144,
|
| 1051 |
+
"grad_norm": 11.658042907714844,
|
| 1052 |
+
"learning_rate": 3.1035366478728864e-05,
|
| 1053 |
+
"loss": 0.6931,
|
| 1054 |
+
"step": 14800
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 1.9092772936955407,
|
| 1058 |
+
"grad_norm": 8.672663688659668,
|
| 1059 |
+
"learning_rate": 3.090722706304459e-05,
|
| 1060 |
+
"loss": 0.6377,
|
| 1061 |
+
"step": 14900
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.9220912352639672,
|
| 1065 |
+
"grad_norm": 6.620725631713867,
|
| 1066 |
+
"learning_rate": 3.077908764736033e-05,
|
| 1067 |
+
"loss": 0.7044,
|
| 1068 |
+
"step": 15000
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 1.9349051768323937,
|
| 1072 |
+
"grad_norm": 8.3103609085083,
|
| 1073 |
+
"learning_rate": 3.065094823167606e-05,
|
| 1074 |
+
"loss": 0.641,
|
| 1075 |
+
"step": 15100
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 1.94771911840082,
|
| 1079 |
+
"grad_norm": 8.163315773010254,
|
| 1080 |
+
"learning_rate": 3.05228088159918e-05,
|
| 1081 |
+
"loss": 0.7094,
|
| 1082 |
+
"step": 15200
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 1.9605330599692465,
|
| 1086 |
+
"grad_norm": 3.6365621089935303,
|
| 1087 |
+
"learning_rate": 3.0394669400307534e-05,
|
| 1088 |
+
"loss": 0.7022,
|
| 1089 |
+
"step": 15300
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 1.973347001537673,
|
| 1093 |
+
"grad_norm": 4.264801502227783,
|
| 1094 |
+
"learning_rate": 3.026652998462327e-05,
|
| 1095 |
+
"loss": 0.6833,
|
| 1096 |
+
"step": 15400
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 1.9861609431060994,
|
| 1100 |
+
"grad_norm": 6.547428131103516,
|
| 1101 |
+
"learning_rate": 3.0138390568939005e-05,
|
| 1102 |
+
"loss": 0.6126,
|
| 1103 |
+
"step": 15500
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 1.9989748846745259,
|
| 1107 |
+
"grad_norm": 6.155936241149902,
|
| 1108 |
+
"learning_rate": 3.0010251153254744e-05,
|
| 1109 |
+
"loss": 0.6851,
|
| 1110 |
+
"step": 15600
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 2.0,
|
| 1114 |
+
"eval_f1": 0.6772664805551888,
|
| 1115 |
+
"eval_loss": 0.781230092048645,
|
| 1116 |
+
"eval_runtime": 778.3436,
|
| 1117 |
+
"eval_samples_per_second": 10.026,
|
| 1118 |
+
"eval_steps_per_second": 2.507,
|
| 1119 |
+
"step": 15608
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 2.0117888262429524,
|
| 1123 |
+
"grad_norm": 8.777030944824219,
|
| 1124 |
+
"learning_rate": 2.988211173757048e-05,
|
| 1125 |
+
"loss": 0.4707,
|
| 1126 |
+
"step": 15700
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 2.0246027678113787,
|
| 1130 |
+
"grad_norm": 4.798321723937988,
|
| 1131 |
+
"learning_rate": 2.9753972321886215e-05,
|
| 1132 |
+
"loss": 0.4366,
|
| 1133 |
+
"step": 15800
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 2.037416709379805,
|
| 1137 |
+
"grad_norm": 2.5244762897491455,
|
| 1138 |
+
"learning_rate": 2.962583290620195e-05,
|
| 1139 |
+
"loss": 0.504,
|
| 1140 |
+
"step": 15900
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 2.0502306509482318,
|
| 1144 |
+
"grad_norm": 15.636524200439453,
|
| 1145 |
+
"learning_rate": 2.9497693490517686e-05,
|
| 1146 |
+
"loss": 0.4234,
|
| 1147 |
+
"step": 16000
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 2.063044592516658,
|
| 1151 |
+
"grad_norm": 8.811060905456543,
|
| 1152 |
+
"learning_rate": 2.936955407483342e-05,
|
| 1153 |
+
"loss": 0.3911,
|
| 1154 |
+
"step": 16100
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 2.0758585340850844,
|
| 1158 |
+
"grad_norm": 4.1310930252075195,
|
| 1159 |
+
"learning_rate": 2.9241414659149157e-05,
|
| 1160 |
+
"loss": 0.4538,
|
| 1161 |
+
"step": 16200
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 2.088672475653511,
|
| 1165 |
+
"grad_norm": 9.516937255859375,
|
| 1166 |
+
"learning_rate": 2.9113275243464892e-05,
|
| 1167 |
+
"loss": 0.4461,
|
| 1168 |
+
"step": 16300
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 2.1014864172219374,
|
| 1172 |
+
"grad_norm": 4.6523756980896,
|
| 1173 |
+
"learning_rate": 2.8985135827780624e-05,
|
| 1174 |
+
"loss": 0.4808,
|
| 1175 |
+
"step": 16400
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 2.1143003587903637,
|
| 1179 |
+
"grad_norm": 4.160647392272949,
|
| 1180 |
+
"learning_rate": 2.885699641209636e-05,
|
| 1181 |
+
"loss": 0.4879,
|
| 1182 |
+
"step": 16500
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.1271143003587905,
|
| 1186 |
+
"grad_norm": 11.32701587677002,
|
| 1187 |
+
"learning_rate": 2.8728856996412095e-05,
|
| 1188 |
+
"loss": 0.4544,
|
| 1189 |
+
"step": 16600
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 2.1399282419272168,
|
| 1193 |
+
"grad_norm": 4.703444004058838,
|
| 1194 |
+
"learning_rate": 2.860071758072783e-05,
|
| 1195 |
+
"loss": 0.466,
|
| 1196 |
+
"step": 16700
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.152742183495643,
|
| 1200 |
+
"grad_norm": 8.985660552978516,
|
| 1201 |
+
"learning_rate": 2.847257816504357e-05,
|
| 1202 |
+
"loss": 0.4734,
|
| 1203 |
+
"step": 16800
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 2.16555612506407,
|
| 1207 |
+
"grad_norm": 12.306890487670898,
|
| 1208 |
+
"learning_rate": 2.8344438749359304e-05,
|
| 1209 |
+
"loss": 0.4287,
|
| 1210 |
+
"step": 16900
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 2.178370066632496,
|
| 1214 |
+
"grad_norm": 5.025609016418457,
|
| 1215 |
+
"learning_rate": 2.821629933367504e-05,
|
| 1216 |
+
"loss": 0.4657,
|
| 1217 |
+
"step": 17000
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 2.1911840082009224,
|
| 1221 |
+
"grad_norm": 31.554025650024414,
|
| 1222 |
+
"learning_rate": 2.8088159917990775e-05,
|
| 1223 |
+
"loss": 0.4378,
|
| 1224 |
+
"step": 17100
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 2.203997949769349,
|
| 1228 |
+
"grad_norm": 9.015434265136719,
|
| 1229 |
+
"learning_rate": 2.796002050230651e-05,
|
| 1230 |
+
"loss": 0.4538,
|
| 1231 |
+
"step": 17200
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 2.2168118913377755,
|
| 1235 |
+
"grad_norm": 15.61099624633789,
|
| 1236 |
+
"learning_rate": 2.7831881086622246e-05,
|
| 1237 |
+
"loss": 0.4134,
|
| 1238 |
+
"step": 17300
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 2.2296258329062018,
|
| 1242 |
+
"grad_norm": 10.191957473754883,
|
| 1243 |
+
"learning_rate": 2.770374167093798e-05,
|
| 1244 |
+
"loss": 0.5188,
|
| 1245 |
+
"step": 17400
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 2.2424397744746285,
|
| 1249 |
+
"grad_norm": 2.2506730556488037,
|
| 1250 |
+
"learning_rate": 2.7575602255253717e-05,
|
| 1251 |
+
"loss": 0.4028,
|
| 1252 |
+
"step": 17500
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 2.255253716043055,
|
| 1256 |
+
"grad_norm": 23.088764190673828,
|
| 1257 |
+
"learning_rate": 2.7447462839569456e-05,
|
| 1258 |
+
"loss": 0.4814,
|
| 1259 |
+
"step": 17600
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 2.268067657611481,
|
| 1263 |
+
"grad_norm": 4.473659515380859,
|
| 1264 |
+
"learning_rate": 2.731932342388519e-05,
|
| 1265 |
+
"loss": 0.4822,
|
| 1266 |
+
"step": 17700
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 2.280881599179908,
|
| 1270 |
+
"grad_norm": 2.1489970684051514,
|
| 1271 |
+
"learning_rate": 2.7191184008200927e-05,
|
| 1272 |
+
"loss": 0.4934,
|
| 1273 |
+
"step": 17800
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 2.293695540748334,
|
| 1277 |
+
"grad_norm": 1.4255170822143555,
|
| 1278 |
+
"learning_rate": 2.7063044592516662e-05,
|
| 1279 |
+
"loss": 0.4314,
|
| 1280 |
+
"step": 17900
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 2.3065094823167605,
|
| 1284 |
+
"grad_norm": 4.612204074859619,
|
| 1285 |
+
"learning_rate": 2.693490517683239e-05,
|
| 1286 |
+
"loss": 0.4322,
|
| 1287 |
+
"step": 18000
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 2.319323423885187,
|
| 1291 |
+
"grad_norm": 3.1022679805755615,
|
| 1292 |
+
"learning_rate": 2.680676576114813e-05,
|
| 1293 |
+
"loss": 0.424,
|
| 1294 |
+
"step": 18100
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 2.3321373654536135,
|
| 1298 |
+
"grad_norm": 3.745171070098877,
|
| 1299 |
+
"learning_rate": 2.6678626345463865e-05,
|
| 1300 |
+
"loss": 0.4269,
|
| 1301 |
+
"step": 18200
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 2.34495130702204,
|
| 1305 |
+
"grad_norm": 4.0442328453063965,
|
| 1306 |
+
"learning_rate": 2.65504869297796e-05,
|
| 1307 |
+
"loss": 0.4698,
|
| 1308 |
+
"step": 18300
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 2.3577652485904665,
|
| 1312 |
+
"grad_norm": 21.303607940673828,
|
| 1313 |
+
"learning_rate": 2.6422347514095336e-05,
|
| 1314 |
+
"loss": 0.4909,
|
| 1315 |
+
"step": 18400
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 2.370579190158893,
|
| 1319 |
+
"grad_norm": 9.175422668457031,
|
| 1320 |
+
"learning_rate": 2.629420809841107e-05,
|
| 1321 |
+
"loss": 0.4598,
|
| 1322 |
+
"step": 18500
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 2.383393131727319,
|
| 1326 |
+
"grad_norm": 5.787283420562744,
|
| 1327 |
+
"learning_rate": 2.6166068682726807e-05,
|
| 1328 |
+
"loss": 0.4409,
|
| 1329 |
+
"step": 18600
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 2.396207073295746,
|
| 1333 |
+
"grad_norm": 7.338250637054443,
|
| 1334 |
+
"learning_rate": 2.6037929267042542e-05,
|
| 1335 |
+
"loss": 0.4157,
|
| 1336 |
+
"step": 18700
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 2.409021014864172,
|
| 1340 |
+
"grad_norm": 13.879666328430176,
|
| 1341 |
+
"learning_rate": 2.590978985135828e-05,
|
| 1342 |
+
"loss": 0.4584,
|
| 1343 |
+
"step": 18800
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 2.4218349564325985,
|
| 1347 |
+
"grad_norm": 9.484577178955078,
|
| 1348 |
+
"learning_rate": 2.5781650435674016e-05,
|
| 1349 |
+
"loss": 0.4914,
|
| 1350 |
+
"step": 18900
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 2.4346488980010252,
|
| 1354 |
+
"grad_norm": 10.865300178527832,
|
| 1355 |
+
"learning_rate": 2.565351101998975e-05,
|
| 1356 |
+
"loss": 0.4259,
|
| 1357 |
+
"step": 19000
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 2.4474628395694515,
|
| 1361 |
+
"grad_norm": 16.69988441467285,
|
| 1362 |
+
"learning_rate": 2.5525371604305487e-05,
|
| 1363 |
+
"loss": 0.4563,
|
| 1364 |
+
"step": 19100
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 2.460276781137878,
|
| 1368 |
+
"grad_norm": 19.711631774902344,
|
| 1369 |
+
"learning_rate": 2.5397232188621222e-05,
|
| 1370 |
+
"loss": 0.4159,
|
| 1371 |
+
"step": 19200
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 2.4730907227063046,
|
| 1375 |
+
"grad_norm": 13.3755521774292,
|
| 1376 |
+
"learning_rate": 2.5269092772936958e-05,
|
| 1377 |
+
"loss": 0.537,
|
| 1378 |
+
"step": 19300
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 2.485904664274731,
|
| 1382 |
+
"grad_norm": 6.953076362609863,
|
| 1383 |
+
"learning_rate": 2.5140953357252693e-05,
|
| 1384 |
+
"loss": 0.4288,
|
| 1385 |
+
"step": 19400
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 2.498718605843157,
|
| 1389 |
+
"grad_norm": 47.91322708129883,
|
| 1390 |
+
"learning_rate": 2.5012813941568432e-05,
|
| 1391 |
+
"loss": 0.5049,
|
| 1392 |
+
"step": 19500
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 2.511532547411584,
|
| 1396 |
+
"grad_norm": 1.6553832292556763,
|
| 1397 |
+
"learning_rate": 2.4884674525884164e-05,
|
| 1398 |
+
"loss": 0.4779,
|
| 1399 |
+
"step": 19600
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 2.5243464889800102,
|
| 1403 |
+
"grad_norm": 12.199808120727539,
|
| 1404 |
+
"learning_rate": 2.47565351101999e-05,
|
| 1405 |
+
"loss": 0.4246,
|
| 1406 |
+
"step": 19700
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 2.5371604305484365,
|
| 1410 |
+
"grad_norm": 11.326825141906738,
|
| 1411 |
+
"learning_rate": 2.4628395694515635e-05,
|
| 1412 |
+
"loss": 0.4482,
|
| 1413 |
+
"step": 19800
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 2.5499743721168633,
|
| 1417 |
+
"grad_norm": 9.247246742248535,
|
| 1418 |
+
"learning_rate": 2.450025627883137e-05,
|
| 1419 |
+
"loss": 0.4656,
|
| 1420 |
+
"step": 19900
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 2.5627883136852896,
|
| 1424 |
+
"grad_norm": 1.773540735244751,
|
| 1425 |
+
"learning_rate": 2.4372116863147106e-05,
|
| 1426 |
+
"loss": 0.4776,
|
| 1427 |
+
"step": 20000
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 2.575602255253716,
|
| 1431 |
+
"grad_norm": 7.454749584197998,
|
| 1432 |
+
"learning_rate": 2.424397744746284e-05,
|
| 1433 |
+
"loss": 0.4161,
|
| 1434 |
+
"step": 20100
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 2.5884161968221426,
|
| 1438 |
+
"grad_norm": 19.77891731262207,
|
| 1439 |
+
"learning_rate": 2.4115838031778577e-05,
|
| 1440 |
+
"loss": 0.4609,
|
| 1441 |
+
"step": 20200
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 2.601230138390569,
|
| 1445 |
+
"grad_norm": 12.208200454711914,
|
| 1446 |
+
"learning_rate": 2.3987698616094312e-05,
|
| 1447 |
+
"loss": 0.453,
|
| 1448 |
+
"step": 20300
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 2.6140440799589952,
|
| 1452 |
+
"grad_norm": 11.438812255859375,
|
| 1453 |
+
"learning_rate": 2.3859559200410047e-05,
|
| 1454 |
+
"loss": 0.4439,
|
| 1455 |
+
"step": 20400
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 2.626858021527422,
|
| 1459 |
+
"grad_norm": 1.6863147020339966,
|
| 1460 |
+
"learning_rate": 2.3731419784725783e-05,
|
| 1461 |
+
"loss": 0.3987,
|
| 1462 |
+
"step": 20500
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 2.6396719630958483,
|
| 1466 |
+
"grad_norm": 1.3637946844100952,
|
| 1467 |
+
"learning_rate": 2.3603280369041518e-05,
|
| 1468 |
+
"loss": 0.4523,
|
| 1469 |
+
"step": 20600
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 2.6524859046642746,
|
| 1473 |
+
"grad_norm": 21.555208206176758,
|
| 1474 |
+
"learning_rate": 2.3475140953357254e-05,
|
| 1475 |
+
"loss": 0.4624,
|
| 1476 |
+
"step": 20700
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 2.6652998462327013,
|
| 1480 |
+
"grad_norm": 8.768684387207031,
|
| 1481 |
+
"learning_rate": 2.334700153767299e-05,
|
| 1482 |
+
"loss": 0.4585,
|
| 1483 |
+
"step": 20800
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 2.6781137878011276,
|
| 1487 |
+
"grad_norm": 3.2959704399108887,
|
| 1488 |
+
"learning_rate": 2.3218862121988724e-05,
|
| 1489 |
+
"loss": 0.4579,
|
| 1490 |
+
"step": 20900
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 2.690927729369554,
|
| 1494 |
+
"grad_norm": 16.97565269470215,
|
| 1495 |
+
"learning_rate": 2.309072270630446e-05,
|
| 1496 |
+
"loss": 0.4132,
|
| 1497 |
+
"step": 21000
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 2.7037416709379807,
|
| 1501 |
+
"grad_norm": 14.613641738891602,
|
| 1502 |
+
"learning_rate": 2.2962583290620195e-05,
|
| 1503 |
+
"loss": 0.4297,
|
| 1504 |
+
"step": 21100
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 2.716555612506407,
|
| 1508 |
+
"grad_norm": 28.61090087890625,
|
| 1509 |
+
"learning_rate": 2.283444387493593e-05,
|
| 1510 |
+
"loss": 0.4479,
|
| 1511 |
+
"step": 21200
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 2.7293695540748333,
|
| 1515 |
+
"grad_norm": 9.84257984161377,
|
| 1516 |
+
"learning_rate": 2.2706304459251666e-05,
|
| 1517 |
+
"loss": 0.4428,
|
| 1518 |
+
"step": 21300
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 2.74218349564326,
|
| 1522 |
+
"grad_norm": 8.199345588684082,
|
| 1523 |
+
"learning_rate": 2.2578165043567405e-05,
|
| 1524 |
+
"loss": 0.3999,
|
| 1525 |
+
"step": 21400
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 2.7549974372116863,
|
| 1529 |
+
"grad_norm": 15.411248207092285,
|
| 1530 |
+
"learning_rate": 2.2450025627883137e-05,
|
| 1531 |
+
"loss": 0.4423,
|
| 1532 |
+
"step": 21500
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 2.7678113787801126,
|
| 1536 |
+
"grad_norm": 7.122200012207031,
|
| 1537 |
+
"learning_rate": 2.2321886212198872e-05,
|
| 1538 |
+
"loss": 0.4675,
|
| 1539 |
+
"step": 21600
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 2.7806253203485394,
|
| 1543 |
+
"grad_norm": 11.358266830444336,
|
| 1544 |
+
"learning_rate": 2.2193746796514608e-05,
|
| 1545 |
+
"loss": 0.4885,
|
| 1546 |
+
"step": 21700
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 2.7934392619169657,
|
| 1550 |
+
"grad_norm": 9.456644058227539,
|
| 1551 |
+
"learning_rate": 2.2065607380830343e-05,
|
| 1552 |
+
"loss": 0.4973,
|
| 1553 |
+
"step": 21800
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 2.806253203485392,
|
| 1557 |
+
"grad_norm": 28.7235164642334,
|
| 1558 |
+
"learning_rate": 2.193746796514608e-05,
|
| 1559 |
+
"loss": 0.429,
|
| 1560 |
+
"step": 21900
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 2.8190671450538183,
|
| 1564 |
+
"grad_norm": 14.859136581420898,
|
| 1565 |
+
"learning_rate": 2.1809328549461817e-05,
|
| 1566 |
+
"loss": 0.4867,
|
| 1567 |
+
"step": 22000
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 2.831881086622245,
|
| 1571 |
+
"grad_norm": 3.089897394180298,
|
| 1572 |
+
"learning_rate": 2.1681189133777553e-05,
|
| 1573 |
+
"loss": 0.4249,
|
| 1574 |
+
"step": 22100
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 2.8446950281906713,
|
| 1578 |
+
"grad_norm": 14.606719970703125,
|
| 1579 |
+
"learning_rate": 2.1553049718093288e-05,
|
| 1580 |
+
"loss": 0.4429,
|
| 1581 |
+
"step": 22200
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 2.857508969759098,
|
| 1585 |
+
"grad_norm": 7.761451244354248,
|
| 1586 |
+
"learning_rate": 2.142491030240902e-05,
|
| 1587 |
+
"loss": 0.4639,
|
| 1588 |
+
"step": 22300
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 2.8703229113275244,
|
| 1592 |
+
"grad_norm": 6.9101362228393555,
|
| 1593 |
+
"learning_rate": 2.1296770886724756e-05,
|
| 1594 |
+
"loss": 0.4606,
|
| 1595 |
+
"step": 22400
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 2.8831368528959507,
|
| 1599 |
+
"grad_norm": 6.754969120025635,
|
| 1600 |
+
"learning_rate": 2.116863147104049e-05,
|
| 1601 |
+
"loss": 0.4784,
|
| 1602 |
+
"step": 22500
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 2.895950794464377,
|
| 1606 |
+
"grad_norm": 20.884119033813477,
|
| 1607 |
+
"learning_rate": 2.104049205535623e-05,
|
| 1608 |
+
"loss": 0.4625,
|
| 1609 |
+
"step": 22600
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 2.9087647360328037,
|
| 1613 |
+
"grad_norm": 18.428529739379883,
|
| 1614 |
+
"learning_rate": 2.0912352639671965e-05,
|
| 1615 |
+
"loss": 0.4121,
|
| 1616 |
+
"step": 22700
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 2.92157867760123,
|
| 1620 |
+
"grad_norm": 9.211915969848633,
|
| 1621 |
+
"learning_rate": 2.07842132239877e-05,
|
| 1622 |
+
"loss": 0.457,
|
| 1623 |
+
"step": 22800
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 2.9343926191696568,
|
| 1627 |
+
"grad_norm": 5.744906425476074,
|
| 1628 |
+
"learning_rate": 2.0656073808303436e-05,
|
| 1629 |
+
"loss": 0.4169,
|
| 1630 |
+
"step": 22900
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 2.947206560738083,
|
| 1634 |
+
"grad_norm": 10.679366111755371,
|
| 1635 |
+
"learning_rate": 2.052793439261917e-05,
|
| 1636 |
+
"loss": 0.4719,
|
| 1637 |
+
"step": 23000
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 2.9600205023065094,
|
| 1641 |
+
"grad_norm": 8.72630500793457,
|
| 1642 |
+
"learning_rate": 2.0399794976934904e-05,
|
| 1643 |
+
"loss": 0.4743,
|
| 1644 |
+
"step": 23100
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 2.9728344438749357,
|
| 1648 |
+
"grad_norm": 5.53284215927124,
|
| 1649 |
+
"learning_rate": 2.0271655561250642e-05,
|
| 1650 |
+
"loss": 0.4592,
|
| 1651 |
+
"step": 23200
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 2.9856483854433624,
|
| 1655 |
+
"grad_norm": 10.75283432006836,
|
| 1656 |
+
"learning_rate": 2.0143516145566378e-05,
|
| 1657 |
+
"loss": 0.3971,
|
| 1658 |
+
"step": 23300
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 2.9984623270117887,
|
| 1662 |
+
"grad_norm": 10.634764671325684,
|
| 1663 |
+
"learning_rate": 2.0015376729882113e-05,
|
| 1664 |
+
"loss": 0.4295,
|
| 1665 |
+
"step": 23400
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 3.0,
|
| 1669 |
+
"eval_f1": 0.6853715205850849,
|
| 1670 |
+
"eval_loss": 1.0191140174865723,
|
| 1671 |
+
"eval_runtime": 837.6905,
|
| 1672 |
+
"eval_samples_per_second": 9.316,
|
| 1673 |
+
"eval_steps_per_second": 2.329,
|
| 1674 |
+
"step": 23412
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 3.0112762685802155,
|
| 1678 |
+
"grad_norm": 3.43902587890625,
|
| 1679 |
+
"learning_rate": 1.988723731419785e-05,
|
| 1680 |
+
"loss": 0.2448,
|
| 1681 |
+
"step": 23500
|
| 1682 |
+
},
|
| 1683 |
+
{
|
| 1684 |
+
"epoch": 3.0240902101486418,
|
| 1685 |
+
"grad_norm": 0.5649552941322327,
|
| 1686 |
+
"learning_rate": 1.9759097898513584e-05,
|
| 1687 |
+
"loss": 0.1908,
|
| 1688 |
+
"step": 23600
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 3.036904151717068,
|
| 1692 |
+
"grad_norm": 1.3035610914230347,
|
| 1693 |
+
"learning_rate": 1.963095848282932e-05,
|
| 1694 |
+
"loss": 0.275,
|
| 1695 |
+
"step": 23700
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 3.049718093285495,
|
| 1699 |
+
"grad_norm": 27.42232322692871,
|
| 1700 |
+
"learning_rate": 1.9502819067145055e-05,
|
| 1701 |
+
"loss": 0.2727,
|
| 1702 |
+
"step": 23800
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"epoch": 3.062532034853921,
|
| 1706 |
+
"grad_norm": 1.675907015800476,
|
| 1707 |
+
"learning_rate": 1.937467965146079e-05,
|
| 1708 |
+
"loss": 0.2916,
|
| 1709 |
+
"step": 23900
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 3.0753459764223474,
|
| 1713 |
+
"grad_norm": 9.602179527282715,
|
| 1714 |
+
"learning_rate": 1.9246540235776526e-05,
|
| 1715 |
+
"loss": 0.2645,
|
| 1716 |
+
"step": 24000
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 3.088159917990774,
|
| 1720 |
+
"grad_norm": 16.757831573486328,
|
| 1721 |
+
"learning_rate": 1.911840082009226e-05,
|
| 1722 |
+
"loss": 0.2476,
|
| 1723 |
+
"step": 24100
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"epoch": 3.1009738595592005,
|
| 1727 |
+
"grad_norm": 5.842043876647949,
|
| 1728 |
+
"learning_rate": 1.8990261404407997e-05,
|
| 1729 |
+
"loss": 0.2829,
|
| 1730 |
+
"step": 24200
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 3.1137878011276268,
|
| 1734 |
+
"grad_norm": 0.593449592590332,
|
| 1735 |
+
"learning_rate": 1.8862121988723732e-05,
|
| 1736 |
+
"loss": 0.289,
|
| 1737 |
+
"step": 24300
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 3.1266017426960535,
|
| 1741 |
+
"grad_norm": 5.712982177734375,
|
| 1742 |
+
"learning_rate": 1.8733982573039467e-05,
|
| 1743 |
+
"loss": 0.2355,
|
| 1744 |
+
"step": 24400
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 3.13941568426448,
|
| 1748 |
+
"grad_norm": 0.3152589201927185,
|
| 1749 |
+
"learning_rate": 1.8605843157355203e-05,
|
| 1750 |
+
"loss": 0.2491,
|
| 1751 |
+
"step": 24500
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 3.152229625832906,
|
| 1755 |
+
"grad_norm": 19.951833724975586,
|
| 1756 |
+
"learning_rate": 1.8477703741670938e-05,
|
| 1757 |
+
"loss": 0.271,
|
| 1758 |
+
"step": 24600
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 3.165043567401333,
|
| 1762 |
+
"grad_norm": 5.257028579711914,
|
| 1763 |
+
"learning_rate": 1.8349564325986674e-05,
|
| 1764 |
+
"loss": 0.277,
|
| 1765 |
+
"step": 24700
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 3.177857508969759,
|
| 1769 |
+
"grad_norm": 3.6717381477355957,
|
| 1770 |
+
"learning_rate": 1.822142491030241e-05,
|
| 1771 |
+
"loss": 0.2736,
|
| 1772 |
+
"step": 24800
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 3.1906714505381855,
|
| 1776 |
+
"grad_norm": 38.49631881713867,
|
| 1777 |
+
"learning_rate": 1.8093285494618144e-05,
|
| 1778 |
+
"loss": 0.2789,
|
| 1779 |
+
"step": 24900
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 3.2034853921066118,
|
| 1783 |
+
"grad_norm": 5.944704055786133,
|
| 1784 |
+
"learning_rate": 1.796514607893388e-05,
|
| 1785 |
+
"loss": 0.3111,
|
| 1786 |
+
"step": 25000
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 3.2162993336750385,
|
| 1790 |
+
"grad_norm": 3.278078079223633,
|
| 1791 |
+
"learning_rate": 1.7837006663249615e-05,
|
| 1792 |
+
"loss": 0.287,
|
| 1793 |
+
"step": 25100
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 3.229113275243465,
|
| 1797 |
+
"grad_norm": 13.320869445800781,
|
| 1798 |
+
"learning_rate": 1.7708867247565354e-05,
|
| 1799 |
+
"loss": 0.2708,
|
| 1800 |
+
"step": 25200
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 3.2419272168118916,
|
| 1804 |
+
"grad_norm": 9.01321029663086,
|
| 1805 |
+
"learning_rate": 1.758072783188109e-05,
|
| 1806 |
+
"loss": 0.2891,
|
| 1807 |
+
"step": 25300
|
| 1808 |
+
},
|
| 1809 |
+
{
|
| 1810 |
+
"epoch": 3.254741158380318,
|
| 1811 |
+
"grad_norm": 14.35201644897461,
|
| 1812 |
+
"learning_rate": 1.745258841619682e-05,
|
| 1813 |
+
"loss": 0.1523,
|
| 1814 |
+
"step": 25400
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 3.267555099948744,
|
| 1818 |
+
"grad_norm": 5.268370628356934,
|
| 1819 |
+
"learning_rate": 1.7324449000512557e-05,
|
| 1820 |
+
"loss": 0.3608,
|
| 1821 |
+
"step": 25500
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 3.2803690415171705,
|
| 1825 |
+
"grad_norm": 3.338168144226074,
|
| 1826 |
+
"learning_rate": 1.7196309584828292e-05,
|
| 1827 |
+
"loss": 0.2829,
|
| 1828 |
+
"step": 25600
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 3.293182983085597,
|
| 1832 |
+
"grad_norm": 12.441572189331055,
|
| 1833 |
+
"learning_rate": 1.7068170169144028e-05,
|
| 1834 |
+
"loss": 0.2563,
|
| 1835 |
+
"step": 25700
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 3.3059969246540235,
|
| 1839 |
+
"grad_norm": 2.870978832244873,
|
| 1840 |
+
"learning_rate": 1.6940030753459767e-05,
|
| 1841 |
+
"loss": 0.2957,
|
| 1842 |
+
"step": 25800
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 3.3188108662224503,
|
| 1846 |
+
"grad_norm": 10.626642227172852,
|
| 1847 |
+
"learning_rate": 1.6811891337775502e-05,
|
| 1848 |
+
"loss": 0.3493,
|
| 1849 |
+
"step": 25900
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 3.3316248077908766,
|
| 1853 |
+
"grad_norm": 1.1796225309371948,
|
| 1854 |
+
"learning_rate": 1.6683751922091237e-05,
|
| 1855 |
+
"loss": 0.293,
|
| 1856 |
+
"step": 26000
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 3.344438749359303,
|
| 1860 |
+
"grad_norm": 46.64753341674805,
|
| 1861 |
+
"learning_rate": 1.6555612506406973e-05,
|
| 1862 |
+
"loss": 0.2739,
|
| 1863 |
+
"step": 26100
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 3.357252690927729,
|
| 1867 |
+
"grad_norm": 17.778207778930664,
|
| 1868 |
+
"learning_rate": 1.6427473090722705e-05,
|
| 1869 |
+
"loss": 0.2897,
|
| 1870 |
+
"step": 26200
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 3.370066632496156,
|
| 1874 |
+
"grad_norm": 1.6698403358459473,
|
| 1875 |
+
"learning_rate": 1.629933367503844e-05,
|
| 1876 |
+
"loss": 0.2661,
|
| 1877 |
+
"step": 26300
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 3.382880574064582,
|
| 1881 |
+
"grad_norm": 0.18206116557121277,
|
| 1882 |
+
"learning_rate": 1.617119425935418e-05,
|
| 1883 |
+
"loss": 0.2847,
|
| 1884 |
+
"step": 26400
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 3.395694515633009,
|
| 1888 |
+
"grad_norm": 6.839690208435059,
|
| 1889 |
+
"learning_rate": 1.6043054843669915e-05,
|
| 1890 |
+
"loss": 0.3044,
|
| 1891 |
+
"step": 26500
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 3.4085084572014352,
|
| 1895 |
+
"grad_norm": 0.6313930749893188,
|
| 1896 |
+
"learning_rate": 1.591491542798565e-05,
|
| 1897 |
+
"loss": 0.2623,
|
| 1898 |
+
"step": 26600
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 3.4213223987698616,
|
| 1902 |
+
"grad_norm": 70.23905181884766,
|
| 1903 |
+
"learning_rate": 1.5786776012301385e-05,
|
| 1904 |
+
"loss": 0.2573,
|
| 1905 |
+
"step": 26700
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 3.434136340338288,
|
| 1909 |
+
"grad_norm": 16.72913360595703,
|
| 1910 |
+
"learning_rate": 1.565863659661712e-05,
|
| 1911 |
+
"loss": 0.2626,
|
| 1912 |
+
"step": 26800
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 3.4469502819067146,
|
| 1916 |
+
"grad_norm": 43.662845611572266,
|
| 1917 |
+
"learning_rate": 1.5530497180932856e-05,
|
| 1918 |
+
"loss": 0.2679,
|
| 1919 |
+
"step": 26900
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 3.459764223475141,
|
| 1923 |
+
"grad_norm": 20.96466064453125,
|
| 1924 |
+
"learning_rate": 1.540235776524859e-05,
|
| 1925 |
+
"loss": 0.3082,
|
| 1926 |
+
"step": 27000
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 3.4725781650435676,
|
| 1930 |
+
"grad_norm": 45.02407455444336,
|
| 1931 |
+
"learning_rate": 1.5274218349564327e-05,
|
| 1932 |
+
"loss": 0.2492,
|
| 1933 |
+
"step": 27100
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 3.485392106611994,
|
| 1937 |
+
"grad_norm": 14.404077529907227,
|
| 1938 |
+
"learning_rate": 1.5146078933880062e-05,
|
| 1939 |
+
"loss": 0.2704,
|
| 1940 |
+
"step": 27200
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 3.4982060481804202,
|
| 1944 |
+
"grad_norm": 19.40283966064453,
|
| 1945 |
+
"learning_rate": 1.5017939518195798e-05,
|
| 1946 |
+
"loss": 0.3089,
|
| 1947 |
+
"step": 27300
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 3.5110199897488465,
|
| 1951 |
+
"grad_norm": 13.016902923583984,
|
| 1952 |
+
"learning_rate": 1.4889800102511533e-05,
|
| 1953 |
+
"loss": 0.2953,
|
| 1954 |
+
"step": 27400
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 3.5238339313172733,
|
| 1958 |
+
"grad_norm": 6.934922695159912,
|
| 1959 |
+
"learning_rate": 1.4761660686827269e-05,
|
| 1960 |
+
"loss": 0.2132,
|
| 1961 |
+
"step": 27500
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 3.5366478728856996,
|
| 1965 |
+
"grad_norm": 49.58895492553711,
|
| 1966 |
+
"learning_rate": 1.4633521271143006e-05,
|
| 1967 |
+
"loss": 0.271,
|
| 1968 |
+
"step": 27600
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 3.5494618144541263,
|
| 1972 |
+
"grad_norm": 4.814508438110352,
|
| 1973 |
+
"learning_rate": 1.4505381855458741e-05,
|
| 1974 |
+
"loss": 0.3195,
|
| 1975 |
+
"step": 27700
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 3.5622757560225526,
|
| 1979 |
+
"grad_norm": 28.65342903137207,
|
| 1980 |
+
"learning_rate": 1.4377242439774475e-05,
|
| 1981 |
+
"loss": 0.2869,
|
| 1982 |
+
"step": 27800
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 3.575089697590979,
|
| 1986 |
+
"grad_norm": 5.931487083435059,
|
| 1987 |
+
"learning_rate": 1.424910302409021e-05,
|
| 1988 |
+
"loss": 0.2982,
|
| 1989 |
+
"step": 27900
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 3.5879036391594052,
|
| 1993 |
+
"grad_norm": 0.22432470321655273,
|
| 1994 |
+
"learning_rate": 1.4120963608405946e-05,
|
| 1995 |
+
"loss": 0.3167,
|
| 1996 |
+
"step": 28000
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 3.600717580727832,
|
| 2000 |
+
"grad_norm": 27.89299964904785,
|
| 2001 |
+
"learning_rate": 1.3992824192721681e-05,
|
| 2002 |
+
"loss": 0.2831,
|
| 2003 |
+
"step": 28100
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 3.6135315222962583,
|
| 2007 |
+
"grad_norm": 6.232203006744385,
|
| 2008 |
+
"learning_rate": 1.3864684777037418e-05,
|
| 2009 |
+
"loss": 0.2328,
|
| 2010 |
+
"step": 28200
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 3.626345463864685,
|
| 2014 |
+
"grad_norm": 0.3798358738422394,
|
| 2015 |
+
"learning_rate": 1.3736545361353154e-05,
|
| 2016 |
+
"loss": 0.2565,
|
| 2017 |
+
"step": 28300
|
| 2018 |
+
},
|
| 2019 |
+
{
|
| 2020 |
+
"epoch": 3.6391594054331113,
|
| 2021 |
+
"grad_norm": 2.3177566528320312,
|
| 2022 |
+
"learning_rate": 1.3608405945668889e-05,
|
| 2023 |
+
"loss": 0.2822,
|
| 2024 |
+
"step": 28400
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 3.6519733470015376,
|
| 2028 |
+
"grad_norm": 0.9287611246109009,
|
| 2029 |
+
"learning_rate": 1.3480266529984623e-05,
|
| 2030 |
+
"loss": 0.2206,
|
| 2031 |
+
"step": 28500
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 3.664787288569964,
|
| 2035 |
+
"grad_norm": 19.89398765563965,
|
| 2036 |
+
"learning_rate": 1.3352127114300358e-05,
|
| 2037 |
+
"loss": 0.2934,
|
| 2038 |
+
"step": 28600
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 3.6776012301383907,
|
| 2042 |
+
"grad_norm": 14.735712051391602,
|
| 2043 |
+
"learning_rate": 1.3223987698616094e-05,
|
| 2044 |
+
"loss": 0.2667,
|
| 2045 |
+
"step": 28700
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 3.690415171706817,
|
| 2049 |
+
"grad_norm": 2.782954454421997,
|
| 2050 |
+
"learning_rate": 1.309584828293183e-05,
|
| 2051 |
+
"loss": 0.2565,
|
| 2052 |
+
"step": 28800
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 3.7032291132752437,
|
| 2056 |
+
"grad_norm": 20.082395553588867,
|
| 2057 |
+
"learning_rate": 1.2967708867247566e-05,
|
| 2058 |
+
"loss": 0.3069,
|
| 2059 |
+
"step": 28900
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 3.71604305484367,
|
| 2063 |
+
"grad_norm": 1.8632967472076416,
|
| 2064 |
+
"learning_rate": 1.2839569451563302e-05,
|
| 2065 |
+
"loss": 0.2484,
|
| 2066 |
+
"step": 29000
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 3.7288569964120963,
|
| 2070 |
+
"grad_norm": 6.2880330085754395,
|
| 2071 |
+
"learning_rate": 1.2711430035879037e-05,
|
| 2072 |
+
"loss": 0.2769,
|
| 2073 |
+
"step": 29100
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 3.7416709379805226,
|
| 2077 |
+
"grad_norm": 18.328922271728516,
|
| 2078 |
+
"learning_rate": 1.2583290620194774e-05,
|
| 2079 |
+
"loss": 0.284,
|
| 2080 |
+
"step": 29200
|
| 2081 |
+
},
|
| 2082 |
+
{
|
| 2083 |
+
"epoch": 3.7544848795489494,
|
| 2084 |
+
"grad_norm": 0.2658964991569519,
|
| 2085 |
+
"learning_rate": 1.2455151204510508e-05,
|
| 2086 |
+
"loss": 0.2725,
|
| 2087 |
+
"step": 29300
|
| 2088 |
+
},
|
| 2089 |
+
{
|
| 2090 |
+
"epoch": 3.7672988211173757,
|
| 2091 |
+
"grad_norm": 7.819123268127441,
|
| 2092 |
+
"learning_rate": 1.2327011788826243e-05,
|
| 2093 |
+
"loss": 0.2513,
|
| 2094 |
+
"step": 29400
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 3.7801127626858024,
|
| 2098 |
+
"grad_norm": 4.6279144287109375,
|
| 2099 |
+
"learning_rate": 1.2198872373141979e-05,
|
| 2100 |
+
"loss": 0.2573,
|
| 2101 |
+
"step": 29500
|
| 2102 |
+
},
|
| 2103 |
+
{
|
| 2104 |
+
"epoch": 3.7929267042542287,
|
| 2105 |
+
"grad_norm": 24.996662139892578,
|
| 2106 |
+
"learning_rate": 1.2070732957457714e-05,
|
| 2107 |
+
"loss": 0.2621,
|
| 2108 |
+
"step": 29600
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 3.805740645822655,
|
| 2112 |
+
"grad_norm": 20.87746810913086,
|
| 2113 |
+
"learning_rate": 1.194259354177345e-05,
|
| 2114 |
+
"loss": 0.2499,
|
| 2115 |
+
"step": 29700
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 3.8185545873910813,
|
| 2119 |
+
"grad_norm": 1.5061414241790771,
|
| 2120 |
+
"learning_rate": 1.1814454126089187e-05,
|
| 2121 |
+
"loss": 0.265,
|
| 2122 |
+
"step": 29800
|
| 2123 |
+
},
|
| 2124 |
+
{
|
| 2125 |
+
"epoch": 3.831368528959508,
|
| 2126 |
+
"grad_norm": 2.7230064868927,
|
| 2127 |
+
"learning_rate": 1.168631471040492e-05,
|
| 2128 |
+
"loss": 0.2469,
|
| 2129 |
+
"step": 29900
|
| 2130 |
+
},
|
| 2131 |
+
{
|
| 2132 |
+
"epoch": 3.8441824705279344,
|
| 2133 |
+
"grad_norm": 0.6768075823783875,
|
| 2134 |
+
"learning_rate": 1.1558175294720656e-05,
|
| 2135 |
+
"loss": 0.2686,
|
| 2136 |
+
"step": 30000
|
| 2137 |
+
},
|
| 2138 |
+
{
|
| 2139 |
+
"epoch": 3.8569964120963607,
|
| 2140 |
+
"grad_norm": 0.08343211561441422,
|
| 2141 |
+
"learning_rate": 1.1430035879036393e-05,
|
| 2142 |
+
"loss": 0.2565,
|
| 2143 |
+
"step": 30100
|
| 2144 |
+
},
|
| 2145 |
+
{
|
| 2146 |
+
"epoch": 3.8698103536647874,
|
| 2147 |
+
"grad_norm": 25.58348274230957,
|
| 2148 |
+
"learning_rate": 1.1301896463352128e-05,
|
| 2149 |
+
"loss": 0.2967,
|
| 2150 |
+
"step": 30200
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 3.8826242952332137,
|
| 2154 |
+
"grad_norm": 1.0459709167480469,
|
| 2155 |
+
"learning_rate": 1.1173757047667862e-05,
|
| 2156 |
+
"loss": 0.3028,
|
| 2157 |
+
"step": 30300
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 3.89543823680164,
|
| 2161 |
+
"grad_norm": 0.33878639340400696,
|
| 2162 |
+
"learning_rate": 1.1045617631983599e-05,
|
| 2163 |
+
"loss": 0.2243,
|
| 2164 |
+
"step": 30400
|
| 2165 |
+
},
|
| 2166 |
+
{
|
| 2167 |
+
"epoch": 3.9082521783700668,
|
| 2168 |
+
"grad_norm": 2.021047592163086,
|
| 2169 |
+
"learning_rate": 1.0917478216299335e-05,
|
| 2170 |
+
"loss": 0.3656,
|
| 2171 |
+
"step": 30500
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 3.921066119938493,
|
| 2175 |
+
"grad_norm": 1.6855653524398804,
|
| 2176 |
+
"learning_rate": 1.078933880061507e-05,
|
| 2177 |
+
"loss": 0.2323,
|
| 2178 |
+
"step": 30600
|
| 2179 |
+
},
|
| 2180 |
+
{
|
| 2181 |
+
"epoch": 3.9338800615069194,
|
| 2182 |
+
"grad_norm": 21.66104507446289,
|
| 2183 |
+
"learning_rate": 1.0661199384930805e-05,
|
| 2184 |
+
"loss": 0.2205,
|
| 2185 |
+
"step": 30700
|
| 2186 |
+
},
|
| 2187 |
+
{
|
| 2188 |
+
"epoch": 3.946694003075346,
|
| 2189 |
+
"grad_norm": 2.4428458213806152,
|
| 2190 |
+
"learning_rate": 1.053305996924654e-05,
|
| 2191 |
+
"loss": 0.2436,
|
| 2192 |
+
"step": 30800
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 3.9595079446437724,
|
| 2196 |
+
"grad_norm": 39.37623596191406,
|
| 2197 |
+
"learning_rate": 1.0404920553562276e-05,
|
| 2198 |
+
"loss": 0.2831,
|
| 2199 |
+
"step": 30900
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 3.9723218862121987,
|
| 2203 |
+
"grad_norm": 44.4313850402832,
|
| 2204 |
+
"learning_rate": 1.0276781137878012e-05,
|
| 2205 |
+
"loss": 0.2522,
|
| 2206 |
+
"step": 31000
|
| 2207 |
+
},
|
| 2208 |
+
{
|
| 2209 |
+
"epoch": 3.9851358277806255,
|
| 2210 |
+
"grad_norm": 2.6004929542541504,
|
| 2211 |
+
"learning_rate": 1.0148641722193747e-05,
|
| 2212 |
+
"loss": 0.3209,
|
| 2213 |
+
"step": 31100
|
| 2214 |
+
},
|
| 2215 |
+
{
|
| 2216 |
+
"epoch": 3.9979497693490518,
|
| 2217 |
+
"grad_norm": 2.536029815673828,
|
| 2218 |
+
"learning_rate": 1.0020502306509482e-05,
|
| 2219 |
+
"loss": 0.2807,
|
| 2220 |
+
"step": 31200
|
| 2221 |
+
},
|
| 2222 |
+
{
|
| 2223 |
+
"epoch": 4.0,
|
| 2224 |
+
"eval_f1": 0.6869426704202687,
|
| 2225 |
+
"eval_loss": 2.03011155128479,
|
| 2226 |
+
"eval_runtime": 825.2142,
|
| 2227 |
+
"eval_samples_per_second": 9.457,
|
| 2228 |
+
"eval_steps_per_second": 2.364,
|
| 2229 |
+
"step": 31216
|
| 2230 |
+
}
|
| 2231 |
+
],
|
| 2232 |
+
"logging_steps": 100,
|
| 2233 |
+
"max_steps": 39020,
|
| 2234 |
+
"num_input_tokens_seen": 0,
|
| 2235 |
+
"num_train_epochs": 5,
|
| 2236 |
+
"save_steps": 500,
|
| 2237 |
+
"stateful_callbacks": {
|
| 2238 |
+
"TrainerControl": {
|
| 2239 |
+
"args": {
|
| 2240 |
+
"should_epoch_stop": false,
|
| 2241 |
+
"should_evaluate": false,
|
| 2242 |
+
"should_log": false,
|
| 2243 |
+
"should_save": true,
|
| 2244 |
+
"should_training_stop": false
|
| 2245 |
+
},
|
| 2246 |
+
"attributes": {}
|
| 2247 |
+
}
|
| 2248 |
+
},
|
| 2249 |
+
"total_flos": 5.105948110057636e+17,
|
| 2250 |
+
"train_batch_size": 8,
|
| 2251 |
+
"trial_name": null,
|
| 2252 |
+
"trial_params": null
|
| 2253 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6ee0657691bc4bc86642e2a1842bd363a7f0f18eafb897662025781f2323eea7
|
| 3 |
+
size 5368
|