Training in progress, epoch 3, checkpoint
Browse files
checkpoint-1665/config.json
ADDED
|
@@ -0,0 +1,61 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "google-t5/t5-large",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"T5ForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"classifier_dropout": 0.0,
|
| 7 |
+
"d_ff": 4096,
|
| 8 |
+
"d_kv": 64,
|
| 9 |
+
"d_model": 1024,
|
| 10 |
+
"decoder_start_token_id": 0,
|
| 11 |
+
"dense_act_fn": "relu",
|
| 12 |
+
"dropout_rate": 0.1,
|
| 13 |
+
"eos_token_id": 1,
|
| 14 |
+
"feed_forward_proj": "relu",
|
| 15 |
+
"initializer_factor": 1.0,
|
| 16 |
+
"is_encoder_decoder": true,
|
| 17 |
+
"is_gated_act": false,
|
| 18 |
+
"layer_norm_epsilon": 1e-06,
|
| 19 |
+
"model_type": "t5",
|
| 20 |
+
"n_positions": 512,
|
| 21 |
+
"num_decoder_layers": 24,
|
| 22 |
+
"num_heads": 16,
|
| 23 |
+
"num_layers": 24,
|
| 24 |
+
"output_past": true,
|
| 25 |
+
"pad_token_id": 0,
|
| 26 |
+
"relative_attention_max_distance": 128,
|
| 27 |
+
"relative_attention_num_buckets": 32,
|
| 28 |
+
"task_specific_params": {
|
| 29 |
+
"summarization": {
|
| 30 |
+
"early_stopping": true,
|
| 31 |
+
"length_penalty": 2.0,
|
| 32 |
+
"max_length": 200,
|
| 33 |
+
"min_length": 30,
|
| 34 |
+
"no_repeat_ngram_size": 3,
|
| 35 |
+
"num_beams": 4,
|
| 36 |
+
"prefix": "summarize: "
|
| 37 |
+
},
|
| 38 |
+
"translation_en_to_de": {
|
| 39 |
+
"early_stopping": true,
|
| 40 |
+
"max_length": 300,
|
| 41 |
+
"num_beams": 4,
|
| 42 |
+
"prefix": "translate English to German: "
|
| 43 |
+
},
|
| 44 |
+
"translation_en_to_fr": {
|
| 45 |
+
"early_stopping": true,
|
| 46 |
+
"max_length": 300,
|
| 47 |
+
"num_beams": 4,
|
| 48 |
+
"prefix": "translate English to French: "
|
| 49 |
+
},
|
| 50 |
+
"translation_en_to_ro": {
|
| 51 |
+
"early_stopping": true,
|
| 52 |
+
"max_length": 300,
|
| 53 |
+
"num_beams": 4,
|
| 54 |
+
"prefix": "translate English to Romanian: "
|
| 55 |
+
}
|
| 56 |
+
},
|
| 57 |
+
"torch_dtype": "float32",
|
| 58 |
+
"transformers_version": "4.47.0",
|
| 59 |
+
"use_cache": true,
|
| 60 |
+
"vocab_size": 32128
|
| 61 |
+
}
|
checkpoint-1665/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f5e92f39aeb083320afb6b98c8598c50c96081197c14b8e1a32b266f5ec8d5e4
|
| 3 |
+
size 2954947648
|
checkpoint-1665/trainer_state.json
ADDED
|
@@ -0,0 +1,1255 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1665,
|
| 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": 0.35766423357664234,
|
| 14 |
+
"eval_fn": 1724,
|
| 15 |
+
"eval_fp": 740,
|
| 16 |
+
"eval_loss": 0.6976209878921509,
|
| 17 |
+
"eval_precision": 0.48106591865357645,
|
| 18 |
+
"eval_recall": 0.2846473029045643,
|
| 19 |
+
"eval_runtime": 68.4179,
|
| 20 |
+
"eval_samples_per_second": 68.301,
|
| 21 |
+
"eval_steps_per_second": 4.283,
|
| 22 |
+
"eval_tn": 1523,
|
| 23 |
+
"eval_tp": 686,
|
| 24 |
+
"step": 0
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.018026137899954935,
|
| 28 |
+
"grad_norm": 9.306479454040527,
|
| 29 |
+
"learning_rate": 4.981949458483755e-05,
|
| 30 |
+
"loss": 0.6855,
|
| 31 |
+
"step": 10
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.03605227579990987,
|
| 35 |
+
"grad_norm": 3.271188497543335,
|
| 36 |
+
"learning_rate": 4.963898916967509e-05,
|
| 37 |
+
"loss": 0.667,
|
| 38 |
+
"step": 20
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.054078413699864804,
|
| 42 |
+
"grad_norm": 7.420863628387451,
|
| 43 |
+
"learning_rate": 4.945848375451264e-05,
|
| 44 |
+
"loss": 0.6415,
|
| 45 |
+
"step": 30
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.07210455159981974,
|
| 49 |
+
"grad_norm": 3.4994685649871826,
|
| 50 |
+
"learning_rate": 4.927797833935018e-05,
|
| 51 |
+
"loss": 0.6188,
|
| 52 |
+
"step": 40
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.09013068949977468,
|
| 56 |
+
"grad_norm": 12.319598197937012,
|
| 57 |
+
"learning_rate": 4.909747292418773e-05,
|
| 58 |
+
"loss": 0.6083,
|
| 59 |
+
"step": 50
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.10815682739972961,
|
| 63 |
+
"grad_norm": 5.15464448928833,
|
| 64 |
+
"learning_rate": 4.891696750902527e-05,
|
| 65 |
+
"loss": 0.5685,
|
| 66 |
+
"step": 60
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.12618296529968454,
|
| 70 |
+
"grad_norm": 8.32410717010498,
|
| 71 |
+
"learning_rate": 4.873646209386282e-05,
|
| 72 |
+
"loss": 0.5153,
|
| 73 |
+
"step": 70
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.14420910319963948,
|
| 77 |
+
"grad_norm": 26.5758113861084,
|
| 78 |
+
"learning_rate": 4.855595667870036e-05,
|
| 79 |
+
"loss": 0.518,
|
| 80 |
+
"step": 80
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.16223524109959442,
|
| 84 |
+
"grad_norm": 10.276503562927246,
|
| 85 |
+
"learning_rate": 4.837545126353791e-05,
|
| 86 |
+
"loss": 0.4902,
|
| 87 |
+
"step": 90
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.18026137899954936,
|
| 91 |
+
"grad_norm": 8.947182655334473,
|
| 92 |
+
"learning_rate": 4.819494584837546e-05,
|
| 93 |
+
"loss": 0.4889,
|
| 94 |
+
"step": 100
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.19828751689950427,
|
| 98 |
+
"grad_norm": 8.54738998413086,
|
| 99 |
+
"learning_rate": 4.8014440433213e-05,
|
| 100 |
+
"loss": 0.4671,
|
| 101 |
+
"step": 110
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.21631365479945922,
|
| 105 |
+
"grad_norm": 10.866387367248535,
|
| 106 |
+
"learning_rate": 4.783393501805055e-05,
|
| 107 |
+
"loss": 0.4543,
|
| 108 |
+
"step": 120
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.23433979269941416,
|
| 112 |
+
"grad_norm": 38.420108795166016,
|
| 113 |
+
"learning_rate": 4.765342960288809e-05,
|
| 114 |
+
"loss": 0.4354,
|
| 115 |
+
"step": 130
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.25236593059936907,
|
| 119 |
+
"grad_norm": 13.113662719726562,
|
| 120 |
+
"learning_rate": 4.747292418772563e-05,
|
| 121 |
+
"loss": 0.5018,
|
| 122 |
+
"step": 140
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.270392068499324,
|
| 126 |
+
"grad_norm": 36.51860809326172,
|
| 127 |
+
"learning_rate": 4.7292418772563177e-05,
|
| 128 |
+
"loss": 0.4681,
|
| 129 |
+
"step": 150
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.28841820639927895,
|
| 133 |
+
"grad_norm": 40.15293502807617,
|
| 134 |
+
"learning_rate": 4.711191335740072e-05,
|
| 135 |
+
"loss": 0.4798,
|
| 136 |
+
"step": 160
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.3064443442992339,
|
| 140 |
+
"grad_norm": 12.45967960357666,
|
| 141 |
+
"learning_rate": 4.693140794223827e-05,
|
| 142 |
+
"loss": 0.4849,
|
| 143 |
+
"step": 170
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.32447048219918884,
|
| 147 |
+
"grad_norm": 10.393397331237793,
|
| 148 |
+
"learning_rate": 4.675090252707581e-05,
|
| 149 |
+
"loss": 0.4496,
|
| 150 |
+
"step": 180
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.3424966200991438,
|
| 154 |
+
"grad_norm": 8.426735877990723,
|
| 155 |
+
"learning_rate": 4.657039711191336e-05,
|
| 156 |
+
"loss": 0.4799,
|
| 157 |
+
"step": 190
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.3605227579990987,
|
| 161 |
+
"grad_norm": 10.40811824798584,
|
| 162 |
+
"learning_rate": 4.63898916967509e-05,
|
| 163 |
+
"loss": 0.45,
|
| 164 |
+
"step": 200
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.3785488958990536,
|
| 168 |
+
"grad_norm": 10.157293319702148,
|
| 169 |
+
"learning_rate": 4.620938628158845e-05,
|
| 170 |
+
"loss": 0.4516,
|
| 171 |
+
"step": 210
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.39657503379900855,
|
| 175 |
+
"grad_norm": 9.582480430603027,
|
| 176 |
+
"learning_rate": 4.602888086642599e-05,
|
| 177 |
+
"loss": 0.4491,
|
| 178 |
+
"step": 220
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.4146011716989635,
|
| 182 |
+
"grad_norm": 17.83095359802246,
|
| 183 |
+
"learning_rate": 4.584837545126354e-05,
|
| 184 |
+
"loss": 0.4575,
|
| 185 |
+
"step": 230
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.43262730959891843,
|
| 189 |
+
"grad_norm": 10.945122718811035,
|
| 190 |
+
"learning_rate": 4.566787003610109e-05,
|
| 191 |
+
"loss": 0.4324,
|
| 192 |
+
"step": 240
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.45065344749887337,
|
| 196 |
+
"grad_norm": 5.155959606170654,
|
| 197 |
+
"learning_rate": 4.548736462093863e-05,
|
| 198 |
+
"loss": 0.421,
|
| 199 |
+
"step": 250
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.4686795853988283,
|
| 203 |
+
"grad_norm": 12.092293739318848,
|
| 204 |
+
"learning_rate": 4.530685920577618e-05,
|
| 205 |
+
"loss": 0.4406,
|
| 206 |
+
"step": 260
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.48670572329878325,
|
| 210 |
+
"grad_norm": 12.724702835083008,
|
| 211 |
+
"learning_rate": 4.5126353790613716e-05,
|
| 212 |
+
"loss": 0.3947,
|
| 213 |
+
"step": 270
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.5047318611987381,
|
| 217 |
+
"grad_norm": 26.710527420043945,
|
| 218 |
+
"learning_rate": 4.494584837545127e-05,
|
| 219 |
+
"loss": 0.4068,
|
| 220 |
+
"step": 280
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.5227579990986931,
|
| 224 |
+
"grad_norm": 8.311524391174316,
|
| 225 |
+
"learning_rate": 4.4765342960288806e-05,
|
| 226 |
+
"loss": 0.4483,
|
| 227 |
+
"step": 290
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.540784136998648,
|
| 231 |
+
"grad_norm": 12.505125045776367,
|
| 232 |
+
"learning_rate": 4.458483754512636e-05,
|
| 233 |
+
"loss": 0.3945,
|
| 234 |
+
"step": 300
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.558810274898603,
|
| 238 |
+
"grad_norm": 10.924736976623535,
|
| 239 |
+
"learning_rate": 4.44043321299639e-05,
|
| 240 |
+
"loss": 0.4386,
|
| 241 |
+
"step": 310
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.5768364127985579,
|
| 245 |
+
"grad_norm": 9.005159378051758,
|
| 246 |
+
"learning_rate": 4.422382671480145e-05,
|
| 247 |
+
"loss": 0.4157,
|
| 248 |
+
"step": 320
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.5948625506985128,
|
| 252 |
+
"grad_norm": 13.740028381347656,
|
| 253 |
+
"learning_rate": 4.404332129963899e-05,
|
| 254 |
+
"loss": 0.4392,
|
| 255 |
+
"step": 330
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.6128886885984678,
|
| 259 |
+
"grad_norm": 58.44659423828125,
|
| 260 |
+
"learning_rate": 4.386281588447654e-05,
|
| 261 |
+
"loss": 0.4345,
|
| 262 |
+
"step": 340
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.6309148264984227,
|
| 266 |
+
"grad_norm": 8.443741798400879,
|
| 267 |
+
"learning_rate": 4.368231046931408e-05,
|
| 268 |
+
"loss": 0.4439,
|
| 269 |
+
"step": 350
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.6489409643983777,
|
| 273 |
+
"grad_norm": 7.2340521812438965,
|
| 274 |
+
"learning_rate": 4.350180505415163e-05,
|
| 275 |
+
"loss": 0.4026,
|
| 276 |
+
"step": 360
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.6669671022983326,
|
| 280 |
+
"grad_norm": 7.140610218048096,
|
| 281 |
+
"learning_rate": 4.332129963898917e-05,
|
| 282 |
+
"loss": 0.4284,
|
| 283 |
+
"step": 370
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.6849932401982876,
|
| 287 |
+
"grad_norm": 26.290523529052734,
|
| 288 |
+
"learning_rate": 4.314079422382672e-05,
|
| 289 |
+
"loss": 0.3861,
|
| 290 |
+
"step": 380
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.7030193780982424,
|
| 294 |
+
"grad_norm": 18.140789031982422,
|
| 295 |
+
"learning_rate": 4.296028880866426e-05,
|
| 296 |
+
"loss": 0.4209,
|
| 297 |
+
"step": 390
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.7210455159981974,
|
| 301 |
+
"grad_norm": 9.561443328857422,
|
| 302 |
+
"learning_rate": 4.277978339350181e-05,
|
| 303 |
+
"loss": 0.4432,
|
| 304 |
+
"step": 400
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.7390716538981523,
|
| 308 |
+
"grad_norm": 11.68012809753418,
|
| 309 |
+
"learning_rate": 4.259927797833935e-05,
|
| 310 |
+
"loss": 0.4329,
|
| 311 |
+
"step": 410
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.7570977917981072,
|
| 315 |
+
"grad_norm": 13.150470733642578,
|
| 316 |
+
"learning_rate": 4.24187725631769e-05,
|
| 317 |
+
"loss": 0.4157,
|
| 318 |
+
"step": 420
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.7751239296980622,
|
| 322 |
+
"grad_norm": 11.43278980255127,
|
| 323 |
+
"learning_rate": 4.223826714801444e-05,
|
| 324 |
+
"loss": 0.4084,
|
| 325 |
+
"step": 430
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.7931500675980171,
|
| 329 |
+
"grad_norm": 6.537344455718994,
|
| 330 |
+
"learning_rate": 4.205776173285199e-05,
|
| 331 |
+
"loss": 0.4095,
|
| 332 |
+
"step": 440
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.8111762054979721,
|
| 336 |
+
"grad_norm": 8.772361755371094,
|
| 337 |
+
"learning_rate": 4.187725631768953e-05,
|
| 338 |
+
"loss": 0.3763,
|
| 339 |
+
"step": 450
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.829202343397927,
|
| 343 |
+
"grad_norm": 7.250331878662109,
|
| 344 |
+
"learning_rate": 4.169675090252708e-05,
|
| 345 |
+
"loss": 0.4262,
|
| 346 |
+
"step": 460
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.847228481297882,
|
| 350 |
+
"grad_norm": 8.936566352844238,
|
| 351 |
+
"learning_rate": 4.151624548736462e-05,
|
| 352 |
+
"loss": 0.3947,
|
| 353 |
+
"step": 470
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.8652546191978369,
|
| 357 |
+
"grad_norm": 8.440159797668457,
|
| 358 |
+
"learning_rate": 4.1335740072202167e-05,
|
| 359 |
+
"loss": 0.4056,
|
| 360 |
+
"step": 480
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.8832807570977917,
|
| 364 |
+
"grad_norm": 9.320460319519043,
|
| 365 |
+
"learning_rate": 4.115523465703972e-05,
|
| 366 |
+
"loss": 0.4019,
|
| 367 |
+
"step": 490
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.9013068949977467,
|
| 371 |
+
"grad_norm": 6.329681873321533,
|
| 372 |
+
"learning_rate": 4.0974729241877256e-05,
|
| 373 |
+
"loss": 0.4092,
|
| 374 |
+
"step": 500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.9193330328977016,
|
| 378 |
+
"grad_norm": 16.446720123291016,
|
| 379 |
+
"learning_rate": 4.079422382671481e-05,
|
| 380 |
+
"loss": 0.4132,
|
| 381 |
+
"step": 510
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.9373591707976566,
|
| 385 |
+
"grad_norm": 7.657299041748047,
|
| 386 |
+
"learning_rate": 4.0613718411552346e-05,
|
| 387 |
+
"loss": 0.4091,
|
| 388 |
+
"step": 520
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.9553853086976115,
|
| 392 |
+
"grad_norm": 8.183615684509277,
|
| 393 |
+
"learning_rate": 4.043321299638989e-05,
|
| 394 |
+
"loss": 0.3985,
|
| 395 |
+
"step": 530
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.9734114465975665,
|
| 399 |
+
"grad_norm": 7.5413384437561035,
|
| 400 |
+
"learning_rate": 4.0252707581227436e-05,
|
| 401 |
+
"loss": 0.4404,
|
| 402 |
+
"step": 540
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.9914375844975214,
|
| 406 |
+
"grad_norm": 10.21881103515625,
|
| 407 |
+
"learning_rate": 4.007220216606498e-05,
|
| 408 |
+
"loss": 0.421,
|
| 409 |
+
"step": 550
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.0,
|
| 413 |
+
"eval_f1": 0.8386043664558254,
|
| 414 |
+
"eval_fn": 355,
|
| 415 |
+
"eval_fp": 436,
|
| 416 |
+
"eval_loss": 0.3710826635360718,
|
| 417 |
+
"eval_precision": 0.8249698916097953,
|
| 418 |
+
"eval_recall": 0.8526970954356846,
|
| 419 |
+
"eval_runtime": 78.3531,
|
| 420 |
+
"eval_samples_per_second": 59.64,
|
| 421 |
+
"eval_steps_per_second": 3.739,
|
| 422 |
+
"eval_tn": 1827,
|
| 423 |
+
"eval_tp": 2055,
|
| 424 |
+
"step": 555
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.0090130689499774,
|
| 428 |
+
"grad_norm": 7.165682792663574,
|
| 429 |
+
"learning_rate": 3.989169675090253e-05,
|
| 430 |
+
"loss": 0.3664,
|
| 431 |
+
"step": 560
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.0270392068499323,
|
| 435 |
+
"grad_norm": 7.220385551452637,
|
| 436 |
+
"learning_rate": 3.971119133574007e-05,
|
| 437 |
+
"loss": 0.3672,
|
| 438 |
+
"step": 570
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.0450653447498874,
|
| 442 |
+
"grad_norm": 51.91734313964844,
|
| 443 |
+
"learning_rate": 3.953068592057762e-05,
|
| 444 |
+
"loss": 0.3674,
|
| 445 |
+
"step": 580
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.0630914826498423,
|
| 449 |
+
"grad_norm": 9.907829284667969,
|
| 450 |
+
"learning_rate": 3.935018050541516e-05,
|
| 451 |
+
"loss": 0.3315,
|
| 452 |
+
"step": 590
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.0811176205497972,
|
| 456 |
+
"grad_norm": 13.948915481567383,
|
| 457 |
+
"learning_rate": 3.916967509025271e-05,
|
| 458 |
+
"loss": 0.4092,
|
| 459 |
+
"step": 600
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.099143758449752,
|
| 463 |
+
"grad_norm": 8.972203254699707,
|
| 464 |
+
"learning_rate": 3.898916967509025e-05,
|
| 465 |
+
"loss": 0.3407,
|
| 466 |
+
"step": 610
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.117169896349707,
|
| 470 |
+
"grad_norm": 6.633499622344971,
|
| 471 |
+
"learning_rate": 3.88086642599278e-05,
|
| 472 |
+
"loss": 0.3509,
|
| 473 |
+
"step": 620
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.135196034249662,
|
| 477 |
+
"grad_norm": 8.704663276672363,
|
| 478 |
+
"learning_rate": 3.862815884476535e-05,
|
| 479 |
+
"loss": 0.3468,
|
| 480 |
+
"step": 630
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.153222172149617,
|
| 484 |
+
"grad_norm": 16.486459732055664,
|
| 485 |
+
"learning_rate": 3.844765342960289e-05,
|
| 486 |
+
"loss": 0.3504,
|
| 487 |
+
"step": 640
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.1712483100495719,
|
| 491 |
+
"grad_norm": 36.337745666503906,
|
| 492 |
+
"learning_rate": 3.826714801444044e-05,
|
| 493 |
+
"loss": 0.3634,
|
| 494 |
+
"step": 650
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.1892744479495267,
|
| 498 |
+
"grad_norm": 6.744643211364746,
|
| 499 |
+
"learning_rate": 3.8086642599277976e-05,
|
| 500 |
+
"loss": 0.3729,
|
| 501 |
+
"step": 660
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.2073005858494819,
|
| 505 |
+
"grad_norm": 5.880044460296631,
|
| 506 |
+
"learning_rate": 3.790613718411553e-05,
|
| 507 |
+
"loss": 0.3685,
|
| 508 |
+
"step": 670
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.2253267237494367,
|
| 512 |
+
"grad_norm": 10.722400665283203,
|
| 513 |
+
"learning_rate": 3.7725631768953066e-05,
|
| 514 |
+
"loss": 0.3134,
|
| 515 |
+
"step": 680
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.2433528616493916,
|
| 519 |
+
"grad_norm": 20.80872917175293,
|
| 520 |
+
"learning_rate": 3.754512635379062e-05,
|
| 521 |
+
"loss": 0.3843,
|
| 522 |
+
"step": 690
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 1.2613789995493465,
|
| 526 |
+
"grad_norm": 7.892972469329834,
|
| 527 |
+
"learning_rate": 3.7364620938628155e-05,
|
| 528 |
+
"loss": 0.3312,
|
| 529 |
+
"step": 700
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.2794051374493014,
|
| 533 |
+
"grad_norm": 8.907954216003418,
|
| 534 |
+
"learning_rate": 3.718411552346571e-05,
|
| 535 |
+
"loss": 0.3645,
|
| 536 |
+
"step": 710
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.2974312753492563,
|
| 540 |
+
"grad_norm": 8.709657669067383,
|
| 541 |
+
"learning_rate": 3.700361010830325e-05,
|
| 542 |
+
"loss": 0.3575,
|
| 543 |
+
"step": 720
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.3154574132492114,
|
| 547 |
+
"grad_norm": 12.098897933959961,
|
| 548 |
+
"learning_rate": 3.68231046931408e-05,
|
| 549 |
+
"loss": 0.3509,
|
| 550 |
+
"step": 730
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.3334835511491663,
|
| 554 |
+
"grad_norm": 6.648235321044922,
|
| 555 |
+
"learning_rate": 3.664259927797834e-05,
|
| 556 |
+
"loss": 0.3539,
|
| 557 |
+
"step": 740
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.3515096890491212,
|
| 561 |
+
"grad_norm": 7.914617538452148,
|
| 562 |
+
"learning_rate": 3.646209386281589e-05,
|
| 563 |
+
"loss": 0.3765,
|
| 564 |
+
"step": 750
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.3695358269490763,
|
| 568 |
+
"grad_norm": 7.850524425506592,
|
| 569 |
+
"learning_rate": 3.628158844765343e-05,
|
| 570 |
+
"loss": 0.3653,
|
| 571 |
+
"step": 760
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.3875619648490312,
|
| 575 |
+
"grad_norm": 13.03383731842041,
|
| 576 |
+
"learning_rate": 3.610108303249098e-05,
|
| 577 |
+
"loss": 0.3917,
|
| 578 |
+
"step": 770
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.405588102748986,
|
| 582 |
+
"grad_norm": 6.959196090698242,
|
| 583 |
+
"learning_rate": 3.592057761732852e-05,
|
| 584 |
+
"loss": 0.3633,
|
| 585 |
+
"step": 780
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.423614240648941,
|
| 589 |
+
"grad_norm": 6.3111252784729,
|
| 590 |
+
"learning_rate": 3.574007220216607e-05,
|
| 591 |
+
"loss": 0.3652,
|
| 592 |
+
"step": 790
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.4416403785488958,
|
| 596 |
+
"grad_norm": 40.7597541809082,
|
| 597 |
+
"learning_rate": 3.555956678700361e-05,
|
| 598 |
+
"loss": 0.3568,
|
| 599 |
+
"step": 800
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.4596665164488507,
|
| 603 |
+
"grad_norm": 7.647876262664795,
|
| 604 |
+
"learning_rate": 3.537906137184116e-05,
|
| 605 |
+
"loss": 0.329,
|
| 606 |
+
"step": 810
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.4776926543488058,
|
| 610 |
+
"grad_norm": 23.181804656982422,
|
| 611 |
+
"learning_rate": 3.51985559566787e-05,
|
| 612 |
+
"loss": 0.3216,
|
| 613 |
+
"step": 820
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.4957187922487607,
|
| 617 |
+
"grad_norm": 26.149192810058594,
|
| 618 |
+
"learning_rate": 3.5018050541516247e-05,
|
| 619 |
+
"loss": 0.3415,
|
| 620 |
+
"step": 830
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.5137449301487156,
|
| 624 |
+
"grad_norm": 11.924449920654297,
|
| 625 |
+
"learning_rate": 3.483754512635379e-05,
|
| 626 |
+
"loss": 0.3539,
|
| 627 |
+
"step": 840
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.5317710680486707,
|
| 631 |
+
"grad_norm": 8.165074348449707,
|
| 632 |
+
"learning_rate": 3.4657039711191336e-05,
|
| 633 |
+
"loss": 0.3407,
|
| 634 |
+
"step": 850
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.5497972059486256,
|
| 638 |
+
"grad_norm": 6.361219882965088,
|
| 639 |
+
"learning_rate": 3.447653429602888e-05,
|
| 640 |
+
"loss": 0.3412,
|
| 641 |
+
"step": 860
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.5678233438485805,
|
| 645 |
+
"grad_norm": 8.405501365661621,
|
| 646 |
+
"learning_rate": 3.4296028880866426e-05,
|
| 647 |
+
"loss": 0.3115,
|
| 648 |
+
"step": 870
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.5858494817485354,
|
| 652 |
+
"grad_norm": 8.599916458129883,
|
| 653 |
+
"learning_rate": 3.411552346570397e-05,
|
| 654 |
+
"loss": 0.339,
|
| 655 |
+
"step": 880
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.6038756196484902,
|
| 659 |
+
"grad_norm": 6.046282768249512,
|
| 660 |
+
"learning_rate": 3.3935018050541516e-05,
|
| 661 |
+
"loss": 0.3503,
|
| 662 |
+
"step": 890
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.6219017575484451,
|
| 666 |
+
"grad_norm": 19.992610931396484,
|
| 667 |
+
"learning_rate": 3.375451263537907e-05,
|
| 668 |
+
"loss": 0.3483,
|
| 669 |
+
"step": 900
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 1.6399278954484002,
|
| 673 |
+
"grad_norm": 10.666455268859863,
|
| 674 |
+
"learning_rate": 3.3574007220216606e-05,
|
| 675 |
+
"loss": 0.3637,
|
| 676 |
+
"step": 910
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.6579540333483551,
|
| 680 |
+
"grad_norm": 19.68842124938965,
|
| 681 |
+
"learning_rate": 3.339350180505416e-05,
|
| 682 |
+
"loss": 0.3554,
|
| 683 |
+
"step": 920
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.67598017124831,
|
| 687 |
+
"grad_norm": 15.999527931213379,
|
| 688 |
+
"learning_rate": 3.3212996389891696e-05,
|
| 689 |
+
"loss": 0.345,
|
| 690 |
+
"step": 930
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.694006309148265,
|
| 694 |
+
"grad_norm": 8.567536354064941,
|
| 695 |
+
"learning_rate": 3.303249097472924e-05,
|
| 696 |
+
"loss": 0.3611,
|
| 697 |
+
"step": 940
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.71203244704822,
|
| 701 |
+
"grad_norm": 13.945199012756348,
|
| 702 |
+
"learning_rate": 3.2851985559566786e-05,
|
| 703 |
+
"loss": 0.3527,
|
| 704 |
+
"step": 950
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.7300585849481749,
|
| 708 |
+
"grad_norm": 12.041361808776855,
|
| 709 |
+
"learning_rate": 3.267148014440433e-05,
|
| 710 |
+
"loss": 0.3543,
|
| 711 |
+
"step": 960
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.7480847228481298,
|
| 715 |
+
"grad_norm": 8.660093307495117,
|
| 716 |
+
"learning_rate": 3.249097472924188e-05,
|
| 717 |
+
"loss": 0.3211,
|
| 718 |
+
"step": 970
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.7661108607480847,
|
| 722 |
+
"grad_norm": 19.359636306762695,
|
| 723 |
+
"learning_rate": 3.231046931407942e-05,
|
| 724 |
+
"loss": 0.354,
|
| 725 |
+
"step": 980
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.7841369986480395,
|
| 729 |
+
"grad_norm": 14.446808815002441,
|
| 730 |
+
"learning_rate": 3.212996389891697e-05,
|
| 731 |
+
"loss": 0.4002,
|
| 732 |
+
"step": 990
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.8021631365479944,
|
| 736 |
+
"grad_norm": 8.190558433532715,
|
| 737 |
+
"learning_rate": 3.194945848375451e-05,
|
| 738 |
+
"loss": 0.3757,
|
| 739 |
+
"step": 1000
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.8201892744479495,
|
| 743 |
+
"grad_norm": 7.9925055503845215,
|
| 744 |
+
"learning_rate": 3.176895306859206e-05,
|
| 745 |
+
"loss": 0.3397,
|
| 746 |
+
"step": 1010
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.8382154123479044,
|
| 750 |
+
"grad_norm": 18.883665084838867,
|
| 751 |
+
"learning_rate": 3.15884476534296e-05,
|
| 752 |
+
"loss": 0.3367,
|
| 753 |
+
"step": 1020
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.8562415502478595,
|
| 757 |
+
"grad_norm": 8.38142204284668,
|
| 758 |
+
"learning_rate": 3.140794223826715e-05,
|
| 759 |
+
"loss": 0.363,
|
| 760 |
+
"step": 1030
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.8742676881478144,
|
| 764 |
+
"grad_norm": 11.258013725280762,
|
| 765 |
+
"learning_rate": 3.12274368231047e-05,
|
| 766 |
+
"loss": 0.3584,
|
| 767 |
+
"step": 1040
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.8922938260477693,
|
| 771 |
+
"grad_norm": 9.164728164672852,
|
| 772 |
+
"learning_rate": 3.104693140794224e-05,
|
| 773 |
+
"loss": 0.313,
|
| 774 |
+
"step": 1050
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.9103199639477242,
|
| 778 |
+
"grad_norm": 7.581974983215332,
|
| 779 |
+
"learning_rate": 3.086642599277979e-05,
|
| 780 |
+
"loss": 0.3294,
|
| 781 |
+
"step": 1060
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.928346101847679,
|
| 785 |
+
"grad_norm": 45.30561828613281,
|
| 786 |
+
"learning_rate": 3.0685920577617325e-05,
|
| 787 |
+
"loss": 0.318,
|
| 788 |
+
"step": 1070
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.946372239747634,
|
| 792 |
+
"grad_norm": 5.765126705169678,
|
| 793 |
+
"learning_rate": 3.0505415162454877e-05,
|
| 794 |
+
"loss": 0.3459,
|
| 795 |
+
"step": 1080
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.9643983776475888,
|
| 799 |
+
"grad_norm": 8.358948707580566,
|
| 800 |
+
"learning_rate": 3.032490974729242e-05,
|
| 801 |
+
"loss": 0.3022,
|
| 802 |
+
"step": 1090
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.982424515547544,
|
| 806 |
+
"grad_norm": 9.03178596496582,
|
| 807 |
+
"learning_rate": 3.0144404332129967e-05,
|
| 808 |
+
"loss": 0.3374,
|
| 809 |
+
"step": 1100
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 2.0,
|
| 813 |
+
"grad_norm": 9.664074897766113,
|
| 814 |
+
"learning_rate": 2.996389891696751e-05,
|
| 815 |
+
"loss": 0.3068,
|
| 816 |
+
"step": 1110
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 2.0,
|
| 820 |
+
"eval_f1": 0.8450475997569374,
|
| 821 |
+
"eval_fn": 324,
|
| 822 |
+
"eval_fp": 441,
|
| 823 |
+
"eval_loss": 0.3717796802520752,
|
| 824 |
+
"eval_precision": 0.8254847645429363,
|
| 825 |
+
"eval_recall": 0.8655601659751038,
|
| 826 |
+
"eval_runtime": 72.1654,
|
| 827 |
+
"eval_samples_per_second": 64.754,
|
| 828 |
+
"eval_steps_per_second": 4.06,
|
| 829 |
+
"eval_tn": 1822,
|
| 830 |
+
"eval_tp": 2086,
|
| 831 |
+
"step": 1110
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 2.018026137899955,
|
| 835 |
+
"grad_norm": 7.0051350593566895,
|
| 836 |
+
"learning_rate": 2.9783393501805057e-05,
|
| 837 |
+
"loss": 0.2963,
|
| 838 |
+
"step": 1120
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 2.0360522757999098,
|
| 842 |
+
"grad_norm": 33.976837158203125,
|
| 843 |
+
"learning_rate": 2.9602888086642598e-05,
|
| 844 |
+
"loss": 0.2376,
|
| 845 |
+
"step": 1130
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 2.0540784136998647,
|
| 849 |
+
"grad_norm": 8.37226390838623,
|
| 850 |
+
"learning_rate": 2.9422382671480147e-05,
|
| 851 |
+
"loss": 0.2855,
|
| 852 |
+
"step": 1140
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 2.0721045515998195,
|
| 856 |
+
"grad_norm": 28.015409469604492,
|
| 857 |
+
"learning_rate": 2.924187725631769e-05,
|
| 858 |
+
"loss": 0.2624,
|
| 859 |
+
"step": 1150
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 2.090130689499775,
|
| 863 |
+
"grad_norm": 81.69696807861328,
|
| 864 |
+
"learning_rate": 2.906137184115524e-05,
|
| 865 |
+
"loss": 0.3098,
|
| 866 |
+
"step": 1160
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 2.1081568273997298,
|
| 870 |
+
"grad_norm": 12.006569862365723,
|
| 871 |
+
"learning_rate": 2.888086642599278e-05,
|
| 872 |
+
"loss": 0.2975,
|
| 873 |
+
"step": 1170
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 2.1261829652996846,
|
| 877 |
+
"grad_norm": 12.094599723815918,
|
| 878 |
+
"learning_rate": 2.870036101083033e-05,
|
| 879 |
+
"loss": 0.2865,
|
| 880 |
+
"step": 1180
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 2.1442091031996395,
|
| 884 |
+
"grad_norm": 10.78286361694336,
|
| 885 |
+
"learning_rate": 2.851985559566787e-05,
|
| 886 |
+
"loss": 0.2778,
|
| 887 |
+
"step": 1190
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 2.1622352410995944,
|
| 891 |
+
"grad_norm": 7.721689701080322,
|
| 892 |
+
"learning_rate": 2.8339350180505413e-05,
|
| 893 |
+
"loss": 0.2603,
|
| 894 |
+
"step": 1200
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 2.1802613789995493,
|
| 898 |
+
"grad_norm": 8.49467945098877,
|
| 899 |
+
"learning_rate": 2.815884476534296e-05,
|
| 900 |
+
"loss": 0.2722,
|
| 901 |
+
"step": 1210
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 2.198287516899504,
|
| 905 |
+
"grad_norm": 10.548264503479004,
|
| 906 |
+
"learning_rate": 2.7978339350180506e-05,
|
| 907 |
+
"loss": 0.2625,
|
| 908 |
+
"step": 1220
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 2.216313654799459,
|
| 912 |
+
"grad_norm": 13.317939758300781,
|
| 913 |
+
"learning_rate": 2.779783393501805e-05,
|
| 914 |
+
"loss": 0.3161,
|
| 915 |
+
"step": 1230
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 2.234339792699414,
|
| 919 |
+
"grad_norm": 8.207959175109863,
|
| 920 |
+
"learning_rate": 2.7617328519855596e-05,
|
| 921 |
+
"loss": 0.2829,
|
| 922 |
+
"step": 1240
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 2.2523659305993693,
|
| 926 |
+
"grad_norm": 10.530780792236328,
|
| 927 |
+
"learning_rate": 2.7436823104693144e-05,
|
| 928 |
+
"loss": 0.2604,
|
| 929 |
+
"step": 1250
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 2.270392068499324,
|
| 933 |
+
"grad_norm": 12.912242889404297,
|
| 934 |
+
"learning_rate": 2.7256317689530686e-05,
|
| 935 |
+
"loss": 0.2925,
|
| 936 |
+
"step": 1260
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 2.288418206399279,
|
| 940 |
+
"grad_norm": 9.53242015838623,
|
| 941 |
+
"learning_rate": 2.7075812274368234e-05,
|
| 942 |
+
"loss": 0.2566,
|
| 943 |
+
"step": 1270
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 2.306444344299234,
|
| 947 |
+
"grad_norm": 8.965003967285156,
|
| 948 |
+
"learning_rate": 2.6895306859205776e-05,
|
| 949 |
+
"loss": 0.2734,
|
| 950 |
+
"step": 1280
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 2.324470482199189,
|
| 954 |
+
"grad_norm": 11.137091636657715,
|
| 955 |
+
"learning_rate": 2.6714801444043324e-05,
|
| 956 |
+
"loss": 0.2906,
|
| 957 |
+
"step": 1290
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 2.3424966200991437,
|
| 961 |
+
"grad_norm": 11.334556579589844,
|
| 962 |
+
"learning_rate": 2.6534296028880866e-05,
|
| 963 |
+
"loss": 0.2951,
|
| 964 |
+
"step": 1300
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 2.3605227579990986,
|
| 968 |
+
"grad_norm": 9.768026351928711,
|
| 969 |
+
"learning_rate": 2.6353790613718414e-05,
|
| 970 |
+
"loss": 0.2779,
|
| 971 |
+
"step": 1310
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 2.3785488958990535,
|
| 975 |
+
"grad_norm": 10.248305320739746,
|
| 976 |
+
"learning_rate": 2.617328519855596e-05,
|
| 977 |
+
"loss": 0.2836,
|
| 978 |
+
"step": 1320
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 2.3965750337990084,
|
| 982 |
+
"grad_norm": 10.527518272399902,
|
| 983 |
+
"learning_rate": 2.59927797833935e-05,
|
| 984 |
+
"loss": 0.2864,
|
| 985 |
+
"step": 1330
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 2.4146011716989637,
|
| 989 |
+
"grad_norm": 7.96708869934082,
|
| 990 |
+
"learning_rate": 2.581227436823105e-05,
|
| 991 |
+
"loss": 0.2618,
|
| 992 |
+
"step": 1340
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 2.4326273095989186,
|
| 996 |
+
"grad_norm": 10.488567352294922,
|
| 997 |
+
"learning_rate": 2.563176895306859e-05,
|
| 998 |
+
"loss": 0.2543,
|
| 999 |
+
"step": 1350
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 2.4506534474988735,
|
| 1003 |
+
"grad_norm": 47.32187271118164,
|
| 1004 |
+
"learning_rate": 2.545126353790614e-05,
|
| 1005 |
+
"loss": 0.2538,
|
| 1006 |
+
"step": 1360
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 2.4686795853988284,
|
| 1010 |
+
"grad_norm": 12.458937644958496,
|
| 1011 |
+
"learning_rate": 2.527075812274368e-05,
|
| 1012 |
+
"loss": 0.2416,
|
| 1013 |
+
"step": 1370
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 2.4867057232987833,
|
| 1017 |
+
"grad_norm": 19.352401733398438,
|
| 1018 |
+
"learning_rate": 2.509025270758123e-05,
|
| 1019 |
+
"loss": 0.2431,
|
| 1020 |
+
"step": 1380
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 2.504731861198738,
|
| 1024 |
+
"grad_norm": 11.721567153930664,
|
| 1025 |
+
"learning_rate": 2.4909747292418774e-05,
|
| 1026 |
+
"loss": 0.3046,
|
| 1027 |
+
"step": 1390
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 2.522757999098693,
|
| 1031 |
+
"grad_norm": 8.958959579467773,
|
| 1032 |
+
"learning_rate": 2.472924187725632e-05,
|
| 1033 |
+
"loss": 0.3058,
|
| 1034 |
+
"step": 1400
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 2.540784136998648,
|
| 1038 |
+
"grad_norm": 11.200770378112793,
|
| 1039 |
+
"learning_rate": 2.4548736462093864e-05,
|
| 1040 |
+
"loss": 0.2726,
|
| 1041 |
+
"step": 1410
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 2.558810274898603,
|
| 1045 |
+
"grad_norm": 20.568017959594727,
|
| 1046 |
+
"learning_rate": 2.436823104693141e-05,
|
| 1047 |
+
"loss": 0.2762,
|
| 1048 |
+
"step": 1420
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 2.576836412798558,
|
| 1052 |
+
"grad_norm": 12.618547439575195,
|
| 1053 |
+
"learning_rate": 2.4187725631768953e-05,
|
| 1054 |
+
"loss": 0.2784,
|
| 1055 |
+
"step": 1430
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 2.5948625506985126,
|
| 1059 |
+
"grad_norm": 10.008194923400879,
|
| 1060 |
+
"learning_rate": 2.40072202166065e-05,
|
| 1061 |
+
"loss": 0.294,
|
| 1062 |
+
"step": 1440
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 2.612888688598468,
|
| 1066 |
+
"grad_norm": 9.454615592956543,
|
| 1067 |
+
"learning_rate": 2.3826714801444043e-05,
|
| 1068 |
+
"loss": 0.3102,
|
| 1069 |
+
"step": 1450
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 2.630914826498423,
|
| 1073 |
+
"grad_norm": 15.94392204284668,
|
| 1074 |
+
"learning_rate": 2.3646209386281588e-05,
|
| 1075 |
+
"loss": 0.2378,
|
| 1076 |
+
"step": 1460
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 2.6489409643983777,
|
| 1080 |
+
"grad_norm": 10.787693977355957,
|
| 1081 |
+
"learning_rate": 2.3465703971119137e-05,
|
| 1082 |
+
"loss": 0.2617,
|
| 1083 |
+
"step": 1470
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 2.6669671022983326,
|
| 1087 |
+
"grad_norm": 17.901201248168945,
|
| 1088 |
+
"learning_rate": 2.328519855595668e-05,
|
| 1089 |
+
"loss": 0.3069,
|
| 1090 |
+
"step": 1480
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 2.6849932401982874,
|
| 1094 |
+
"grad_norm": 38.81023406982422,
|
| 1095 |
+
"learning_rate": 2.3104693140794227e-05,
|
| 1096 |
+
"loss": 0.2764,
|
| 1097 |
+
"step": 1490
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 2.7030193780982423,
|
| 1101 |
+
"grad_norm": 14.45818042755127,
|
| 1102 |
+
"learning_rate": 2.292418772563177e-05,
|
| 1103 |
+
"loss": 0.291,
|
| 1104 |
+
"step": 1500
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 2.721045515998197,
|
| 1108 |
+
"grad_norm": 7.874146461486816,
|
| 1109 |
+
"learning_rate": 2.2743682310469316e-05,
|
| 1110 |
+
"loss": 0.2743,
|
| 1111 |
+
"step": 1510
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 2.7390716538981525,
|
| 1115 |
+
"grad_norm": 9.052064895629883,
|
| 1116 |
+
"learning_rate": 2.2563176895306858e-05,
|
| 1117 |
+
"loss": 0.3207,
|
| 1118 |
+
"step": 1520
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 2.757097791798107,
|
| 1122 |
+
"grad_norm": 9.65465259552002,
|
| 1123 |
+
"learning_rate": 2.2382671480144403e-05,
|
| 1124 |
+
"loss": 0.2747,
|
| 1125 |
+
"step": 1530
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 2.7751239296980623,
|
| 1129 |
+
"grad_norm": 8.10863208770752,
|
| 1130 |
+
"learning_rate": 2.220216606498195e-05,
|
| 1131 |
+
"loss": 0.2883,
|
| 1132 |
+
"step": 1540
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 2.793150067598017,
|
| 1136 |
+
"grad_norm": 9.309361457824707,
|
| 1137 |
+
"learning_rate": 2.2021660649819496e-05,
|
| 1138 |
+
"loss": 0.3039,
|
| 1139 |
+
"step": 1550
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 2.811176205497972,
|
| 1143 |
+
"grad_norm": 54.50463104248047,
|
| 1144 |
+
"learning_rate": 2.184115523465704e-05,
|
| 1145 |
+
"loss": 0.2912,
|
| 1146 |
+
"step": 1560
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 2.829202343397927,
|
| 1150 |
+
"grad_norm": 11.151042938232422,
|
| 1151 |
+
"learning_rate": 2.1660649819494586e-05,
|
| 1152 |
+
"loss": 0.3161,
|
| 1153 |
+
"step": 1570
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 2.847228481297882,
|
| 1157 |
+
"grad_norm": 11.726816177368164,
|
| 1158 |
+
"learning_rate": 2.148014440433213e-05,
|
| 1159 |
+
"loss": 0.2485,
|
| 1160 |
+
"step": 1580
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 2.8652546191978367,
|
| 1164 |
+
"grad_norm": 9.904781341552734,
|
| 1165 |
+
"learning_rate": 2.1299638989169676e-05,
|
| 1166 |
+
"loss": 0.3208,
|
| 1167 |
+
"step": 1590
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 2.8832807570977916,
|
| 1171 |
+
"grad_norm": 13.534920692443848,
|
| 1172 |
+
"learning_rate": 2.111913357400722e-05,
|
| 1173 |
+
"loss": 0.3127,
|
| 1174 |
+
"step": 1600
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 2.901306894997747,
|
| 1178 |
+
"grad_norm": 8.898674964904785,
|
| 1179 |
+
"learning_rate": 2.0938628158844766e-05,
|
| 1180 |
+
"loss": 0.2989,
|
| 1181 |
+
"step": 1610
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 2.9193330328977014,
|
| 1185 |
+
"grad_norm": 14.875420570373535,
|
| 1186 |
+
"learning_rate": 2.075812274368231e-05,
|
| 1187 |
+
"loss": 0.3275,
|
| 1188 |
+
"step": 1620
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 2.9373591707976567,
|
| 1192 |
+
"grad_norm": 7.563508033752441,
|
| 1193 |
+
"learning_rate": 2.057761732851986e-05,
|
| 1194 |
+
"loss": 0.2747,
|
| 1195 |
+
"step": 1630
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 2.9553853086976116,
|
| 1199 |
+
"grad_norm": 15.946221351623535,
|
| 1200 |
+
"learning_rate": 2.0397111913357404e-05,
|
| 1201 |
+
"loss": 0.2661,
|
| 1202 |
+
"step": 1640
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 2.9734114465975665,
|
| 1206 |
+
"grad_norm": 14.538371086120605,
|
| 1207 |
+
"learning_rate": 2.0216606498194946e-05,
|
| 1208 |
+
"loss": 0.2767,
|
| 1209 |
+
"step": 1650
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 2.9914375844975214,
|
| 1213 |
+
"grad_norm": 14.545721054077148,
|
| 1214 |
+
"learning_rate": 2.003610108303249e-05,
|
| 1215 |
+
"loss": 0.2721,
|
| 1216 |
+
"step": 1660
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 3.0,
|
| 1220 |
+
"eval_f1": 0.8490452261306533,
|
| 1221 |
+
"eval_fn": 298,
|
| 1222 |
+
"eval_fp": 453,
|
| 1223 |
+
"eval_loss": 0.39444267749786377,
|
| 1224 |
+
"eval_precision": 0.8233918128654971,
|
| 1225 |
+
"eval_recall": 0.8763485477178423,
|
| 1226 |
+
"eval_runtime": 72.7061,
|
| 1227 |
+
"eval_samples_per_second": 64.272,
|
| 1228 |
+
"eval_steps_per_second": 4.03,
|
| 1229 |
+
"eval_tn": 1810,
|
| 1230 |
+
"eval_tp": 2112,
|
| 1231 |
+
"step": 1665
|
| 1232 |
+
}
|
| 1233 |
+
],
|
| 1234 |
+
"logging_steps": 10,
|
| 1235 |
+
"max_steps": 2770,
|
| 1236 |
+
"num_input_tokens_seen": 0,
|
| 1237 |
+
"num_train_epochs": 5,
|
| 1238 |
+
"save_steps": 500,
|
| 1239 |
+
"stateful_callbacks": {
|
| 1240 |
+
"TrainerControl": {
|
| 1241 |
+
"args": {
|
| 1242 |
+
"should_epoch_stop": false,
|
| 1243 |
+
"should_evaluate": false,
|
| 1244 |
+
"should_log": false,
|
| 1245 |
+
"should_save": true,
|
| 1246 |
+
"should_training_stop": false
|
| 1247 |
+
},
|
| 1248 |
+
"attributes": {}
|
| 1249 |
+
}
|
| 1250 |
+
},
|
| 1251 |
+
"total_flos": 1.154738112933888e+17,
|
| 1252 |
+
"train_batch_size": 16,
|
| 1253 |
+
"trial_name": null,
|
| 1254 |
+
"trial_params": null
|
| 1255 |
+
}
|
checkpoint-1665/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1e02c48bb13e93b20c0e596d2f6d047e5f779b15b95c9b353feb183f1e6b75bf
|
| 3 |
+
size 5368
|