Training in progress, epoch 4, checkpoint
Browse files- checkpoint-1568/config.json +71 -0
- checkpoint-1568/model.safetensors +3 -0
- checkpoint-1568/optimizer.pt +3 -0
- checkpoint-1568/preprocessor_config.json +14 -0
- checkpoint-1568/rng_state.pth +3 -0
- checkpoint-1568/scheduler.pt +3 -0
- checkpoint-1568/trainer_state.json +2260 -0
- checkpoint-1568/training_args.bin +3 -0
checkpoint-1568/config.json
ADDED
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "vinai/PhoWhisper-small",
|
3 |
+
"activation_dropout": 0.0,
|
4 |
+
"activation_function": "gelu",
|
5 |
+
"apply_spec_augment": false,
|
6 |
+
"architectures": [
|
7 |
+
"ViSpeechClassification"
|
8 |
+
],
|
9 |
+
"attention_dropout": 0.0,
|
10 |
+
"begin_suppress_tokens": [
|
11 |
+
220,
|
12 |
+
50257
|
13 |
+
],
|
14 |
+
"bos_token_id": 50257,
|
15 |
+
"classifier_proj_size": 256,
|
16 |
+
"d_model": 768,
|
17 |
+
"decoder_attention_heads": 12,
|
18 |
+
"decoder_ffn_dim": 3072,
|
19 |
+
"decoder_layerdrop": 0.0,
|
20 |
+
"decoder_layers": 12,
|
21 |
+
"decoder_start_token_id": 50258,
|
22 |
+
"dropout": 0.0,
|
23 |
+
"encoder_attention_heads": 12,
|
24 |
+
"encoder_ffn_dim": 3072,
|
25 |
+
"encoder_layerdrop": 0.0,
|
26 |
+
"encoder_layers": 12,
|
27 |
+
"eos_token_id": 50257,
|
28 |
+
"forced_decoder_ids": null,
|
29 |
+
"id2label": {
|
30 |
+
"0": "male, northern dialect",
|
31 |
+
"1": "male, central dialect",
|
32 |
+
"2": "male, highland central dialect",
|
33 |
+
"3": "male, southern dialect",
|
34 |
+
"4": "female, northern dialect",
|
35 |
+
"5": "female, central dialect",
|
36 |
+
"6": "female, highland central dialect",
|
37 |
+
"7": "female, southern dialect"
|
38 |
+
},
|
39 |
+
"init_std": 0.02,
|
40 |
+
"is_encoder_decoder": true,
|
41 |
+
"label2id": {
|
42 |
+
"female, central dialect": "5",
|
43 |
+
"female, highland central dialect": "6",
|
44 |
+
"female, northern dialect": "4",
|
45 |
+
"female, southern dialect": "7",
|
46 |
+
"male, central dialect": "1",
|
47 |
+
"male, highland central dialect": "2",
|
48 |
+
"male, northern dialect": "0",
|
49 |
+
"male, southern dialect": "3"
|
50 |
+
},
|
51 |
+
"mask_feature_length": 10,
|
52 |
+
"mask_feature_min_masks": 0,
|
53 |
+
"mask_feature_prob": 0.0,
|
54 |
+
"mask_time_length": 10,
|
55 |
+
"mask_time_min_masks": 2,
|
56 |
+
"mask_time_prob": 0.05,
|
57 |
+
"max_length": 448,
|
58 |
+
"max_source_positions": 1500,
|
59 |
+
"max_target_positions": 448,
|
60 |
+
"median_filter_width": 7,
|
61 |
+
"model_type": "whisper",
|
62 |
+
"num_hidden_layers": 12,
|
63 |
+
"num_mel_bins": 80,
|
64 |
+
"pad_token_id": 50257,
|
65 |
+
"scale_embedding": false,
|
66 |
+
"torch_dtype": "float32",
|
67 |
+
"transformers_version": "4.41.2",
|
68 |
+
"use_cache": true,
|
69 |
+
"use_weighted_layer_sum": false,
|
70 |
+
"vocab_size": 51865
|
71 |
+
}
|
checkpoint-1568/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:df0e7e47b9048b91ea6dcea3047e633e108449de3b8a4098742a849ebf6bc975
|
3 |
+
size 358419776
|
checkpoint-1568/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:25d54d9acf083c7438b0a5a8161b5518202f136866ddcdf03bd8c2467eace0a1
|
3 |
+
size 11572486
|
checkpoint-1568/preprocessor_config.json
ADDED
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"chunk_length": 30,
|
3 |
+
"feature_extractor_type": "WhisperFeatureExtractor",
|
4 |
+
"feature_size": 80,
|
5 |
+
"hop_length": 160,
|
6 |
+
"n_fft": 400,
|
7 |
+
"n_samples": 480000,
|
8 |
+
"nb_max_frames": 3000,
|
9 |
+
"padding_side": "right",
|
10 |
+
"padding_value": 0.0,
|
11 |
+
"processor_class": "WhisperProcessor",
|
12 |
+
"return_attention_mask": false,
|
13 |
+
"sampling_rate": 16000
|
14 |
+
}
|
checkpoint-1568/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4f94d472e7e5b550ef785d21f64b5205c52737e1172cd92d797859792253643e
|
3 |
+
size 14244
|
checkpoint-1568/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8a4a076a31a50168b375ee3e820c527106d601aeea9deb49819180bc244bd66d
|
3 |
+
size 1064
|
checkpoint-1568/trainer_state.json
ADDED
@@ -0,0 +1,2260 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.4682603455202893,
|
3 |
+
"best_model_checkpoint": "PhoWhisper-small-vispeech-classifier-v3/checkpoint-1568",
|
4 |
+
"epoch": 4.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 1568,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.012755102040816327,
|
13 |
+
"grad_norm": 48589.94921875,
|
14 |
+
"learning_rate": 6.377551020408164e-08,
|
15 |
+
"loss": 2.0884,
|
16 |
+
"step": 5
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.025510204081632654,
|
20 |
+
"grad_norm": 50532.125,
|
21 |
+
"learning_rate": 1.2755102040816328e-07,
|
22 |
+
"loss": 2.0887,
|
23 |
+
"step": 10
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.03826530612244898,
|
27 |
+
"grad_norm": 50246.4296875,
|
28 |
+
"learning_rate": 1.913265306122449e-07,
|
29 |
+
"loss": 2.0885,
|
30 |
+
"step": 15
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.05102040816326531,
|
34 |
+
"grad_norm": 57370.44140625,
|
35 |
+
"learning_rate": 2.5510204081632656e-07,
|
36 |
+
"loss": 2.0908,
|
37 |
+
"step": 20
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.06377551020408163,
|
41 |
+
"grad_norm": 40133.53125,
|
42 |
+
"learning_rate": 3.188775510204082e-07,
|
43 |
+
"loss": 2.0889,
|
44 |
+
"step": 25
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.07653061224489796,
|
48 |
+
"grad_norm": 58377.3671875,
|
49 |
+
"learning_rate": 3.826530612244898e-07,
|
50 |
+
"loss": 2.0874,
|
51 |
+
"step": 30
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.08928571428571429,
|
55 |
+
"grad_norm": 50888.7421875,
|
56 |
+
"learning_rate": 4.4642857142857147e-07,
|
57 |
+
"loss": 2.088,
|
58 |
+
"step": 35
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.10204081632653061,
|
62 |
+
"grad_norm": 44107.75390625,
|
63 |
+
"learning_rate": 5.102040816326531e-07,
|
64 |
+
"loss": 2.0872,
|
65 |
+
"step": 40
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.11479591836734694,
|
69 |
+
"grad_norm": 62939.18359375,
|
70 |
+
"learning_rate": 5.739795918367347e-07,
|
71 |
+
"loss": 2.0873,
|
72 |
+
"step": 45
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.12755102040816327,
|
76 |
+
"grad_norm": 42077.53515625,
|
77 |
+
"learning_rate": 6.377551020408164e-07,
|
78 |
+
"loss": 2.0852,
|
79 |
+
"step": 50
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.14030612244897958,
|
83 |
+
"grad_norm": 50233.00390625,
|
84 |
+
"learning_rate": 7.015306122448979e-07,
|
85 |
+
"loss": 2.0861,
|
86 |
+
"step": 55
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.15306122448979592,
|
90 |
+
"grad_norm": 60839.4765625,
|
91 |
+
"learning_rate": 7.653061224489796e-07,
|
92 |
+
"loss": 2.0856,
|
93 |
+
"step": 60
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.16581632653061223,
|
97 |
+
"grad_norm": 57499.3515625,
|
98 |
+
"learning_rate": 8.290816326530612e-07,
|
99 |
+
"loss": 2.0838,
|
100 |
+
"step": 65
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.17857142857142858,
|
104 |
+
"grad_norm": 65030.765625,
|
105 |
+
"learning_rate": 8.928571428571429e-07,
|
106 |
+
"loss": 2.083,
|
107 |
+
"step": 70
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.1913265306122449,
|
111 |
+
"grad_norm": 53192.35546875,
|
112 |
+
"learning_rate": 9.566326530612244e-07,
|
113 |
+
"loss": 2.0841,
|
114 |
+
"step": 75
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.20408163265306123,
|
118 |
+
"grad_norm": 52991.80078125,
|
119 |
+
"learning_rate": 1.0204081632653063e-06,
|
120 |
+
"loss": 2.0817,
|
121 |
+
"step": 80
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.21683673469387754,
|
125 |
+
"grad_norm": 48656.234375,
|
126 |
+
"learning_rate": 1.0841836734693879e-06,
|
127 |
+
"loss": 2.0813,
|
128 |
+
"step": 85
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.22959183673469388,
|
132 |
+
"grad_norm": 63459.2421875,
|
133 |
+
"learning_rate": 1.1479591836734695e-06,
|
134 |
+
"loss": 2.081,
|
135 |
+
"step": 90
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.2423469387755102,
|
139 |
+
"grad_norm": 47800.3515625,
|
140 |
+
"learning_rate": 1.211734693877551e-06,
|
141 |
+
"loss": 2.0775,
|
142 |
+
"step": 95
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.25510204081632654,
|
146 |
+
"grad_norm": 64353.37890625,
|
147 |
+
"learning_rate": 1.2755102040816329e-06,
|
148 |
+
"loss": 2.0772,
|
149 |
+
"step": 100
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.26785714285714285,
|
153 |
+
"grad_norm": 59652.29296875,
|
154 |
+
"learning_rate": 1.3392857142857143e-06,
|
155 |
+
"loss": 2.0779,
|
156 |
+
"step": 105
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.28061224489795916,
|
160 |
+
"grad_norm": 44752.69140625,
|
161 |
+
"learning_rate": 1.4030612244897959e-06,
|
162 |
+
"loss": 2.0764,
|
163 |
+
"step": 110
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.29336734693877553,
|
167 |
+
"grad_norm": 49883.0,
|
168 |
+
"learning_rate": 1.4668367346938777e-06,
|
169 |
+
"loss": 2.0763,
|
170 |
+
"step": 115
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.30612244897959184,
|
174 |
+
"grad_norm": 65950.28125,
|
175 |
+
"learning_rate": 1.5306122448979593e-06,
|
176 |
+
"loss": 2.0739,
|
177 |
+
"step": 120
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.31887755102040816,
|
181 |
+
"grad_norm": 61354.83203125,
|
182 |
+
"learning_rate": 1.5943877551020409e-06,
|
183 |
+
"loss": 2.0706,
|
184 |
+
"step": 125
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.33163265306122447,
|
188 |
+
"grad_norm": 53169.91796875,
|
189 |
+
"learning_rate": 1.6581632653061225e-06,
|
190 |
+
"loss": 2.0709,
|
191 |
+
"step": 130
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.34438775510204084,
|
195 |
+
"grad_norm": 55883.31640625,
|
196 |
+
"learning_rate": 1.7219387755102043e-06,
|
197 |
+
"loss": 2.0695,
|
198 |
+
"step": 135
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.35714285714285715,
|
202 |
+
"grad_norm": 38134.765625,
|
203 |
+
"learning_rate": 1.7857142857142859e-06,
|
204 |
+
"loss": 2.0675,
|
205 |
+
"step": 140
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.36989795918367346,
|
209 |
+
"grad_norm": 53920.7265625,
|
210 |
+
"learning_rate": 1.8494897959183675e-06,
|
211 |
+
"loss": 2.0659,
|
212 |
+
"step": 145
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.3826530612244898,
|
216 |
+
"grad_norm": 68777.28125,
|
217 |
+
"learning_rate": 1.913265306122449e-06,
|
218 |
+
"loss": 2.0638,
|
219 |
+
"step": 150
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.39540816326530615,
|
223 |
+
"grad_norm": 46247.8984375,
|
224 |
+
"learning_rate": 1.977040816326531e-06,
|
225 |
+
"loss": 2.0656,
|
226 |
+
"step": 155
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.40816326530612246,
|
230 |
+
"grad_norm": 49636.04296875,
|
231 |
+
"learning_rate": 2.0408163265306125e-06,
|
232 |
+
"loss": 2.0621,
|
233 |
+
"step": 160
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.42091836734693877,
|
237 |
+
"grad_norm": 51880.3515625,
|
238 |
+
"learning_rate": 2.104591836734694e-06,
|
239 |
+
"loss": 2.057,
|
240 |
+
"step": 165
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.4336734693877551,
|
244 |
+
"grad_norm": 52808.734375,
|
245 |
+
"learning_rate": 2.1683673469387757e-06,
|
246 |
+
"loss": 2.0585,
|
247 |
+
"step": 170
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.44642857142857145,
|
251 |
+
"grad_norm": 46398.125,
|
252 |
+
"learning_rate": 2.2321428571428573e-06,
|
253 |
+
"loss": 2.0505,
|
254 |
+
"step": 175
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.45918367346938777,
|
258 |
+
"grad_norm": 59347.2578125,
|
259 |
+
"learning_rate": 2.295918367346939e-06,
|
260 |
+
"loss": 2.0487,
|
261 |
+
"step": 180
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.4719387755102041,
|
265 |
+
"grad_norm": 58343.73828125,
|
266 |
+
"learning_rate": 2.3596938775510205e-06,
|
267 |
+
"loss": 2.0563,
|
268 |
+
"step": 185
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 0.4846938775510204,
|
272 |
+
"grad_norm": 43836.42578125,
|
273 |
+
"learning_rate": 2.423469387755102e-06,
|
274 |
+
"loss": 2.0486,
|
275 |
+
"step": 190
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 0.49744897959183676,
|
279 |
+
"grad_norm": 69661.5859375,
|
280 |
+
"learning_rate": 2.487244897959184e-06,
|
281 |
+
"loss": 2.0443,
|
282 |
+
"step": 195
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.5102040816326531,
|
286 |
+
"grad_norm": 53888.24609375,
|
287 |
+
"learning_rate": 2.5510204081632657e-06,
|
288 |
+
"loss": 2.0475,
|
289 |
+
"step": 200
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.5229591836734694,
|
293 |
+
"grad_norm": 50369.1640625,
|
294 |
+
"learning_rate": 2.6147959183673473e-06,
|
295 |
+
"loss": 2.035,
|
296 |
+
"step": 205
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 0.5357142857142857,
|
300 |
+
"grad_norm": 62733.9609375,
|
301 |
+
"learning_rate": 2.6785714285714285e-06,
|
302 |
+
"loss": 2.0354,
|
303 |
+
"step": 210
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.548469387755102,
|
307 |
+
"grad_norm": 48340.67578125,
|
308 |
+
"learning_rate": 2.74234693877551e-06,
|
309 |
+
"loss": 2.0373,
|
310 |
+
"step": 215
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 0.5612244897959183,
|
314 |
+
"grad_norm": 50353.734375,
|
315 |
+
"learning_rate": 2.8061224489795917e-06,
|
316 |
+
"loss": 2.0389,
|
317 |
+
"step": 220
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 0.5739795918367347,
|
321 |
+
"grad_norm": 62351.765625,
|
322 |
+
"learning_rate": 2.869897959183674e-06,
|
323 |
+
"loss": 2.0263,
|
324 |
+
"step": 225
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 0.5867346938775511,
|
328 |
+
"grad_norm": 73855.609375,
|
329 |
+
"learning_rate": 2.9336734693877553e-06,
|
330 |
+
"loss": 2.0269,
|
331 |
+
"step": 230
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 0.5994897959183674,
|
335 |
+
"grad_norm": 84417.859375,
|
336 |
+
"learning_rate": 2.997448979591837e-06,
|
337 |
+
"loss": 2.0311,
|
338 |
+
"step": 235
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 0.6122448979591837,
|
342 |
+
"grad_norm": 57347.55078125,
|
343 |
+
"learning_rate": 3.0612244897959185e-06,
|
344 |
+
"loss": 2.0239,
|
345 |
+
"step": 240
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 0.625,
|
349 |
+
"grad_norm": 65611.1875,
|
350 |
+
"learning_rate": 3.125e-06,
|
351 |
+
"loss": 2.0064,
|
352 |
+
"step": 245
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 0.6377551020408163,
|
356 |
+
"grad_norm": 70810.75,
|
357 |
+
"learning_rate": 3.1887755102040818e-06,
|
358 |
+
"loss": 2.0224,
|
359 |
+
"step": 250
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 0.6505102040816326,
|
363 |
+
"grad_norm": 46479.8828125,
|
364 |
+
"learning_rate": 3.2525510204081634e-06,
|
365 |
+
"loss": 2.0057,
|
366 |
+
"step": 255
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 0.6632653061224489,
|
370 |
+
"grad_norm": 34124.89453125,
|
371 |
+
"learning_rate": 3.316326530612245e-06,
|
372 |
+
"loss": 2.0135,
|
373 |
+
"step": 260
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 0.6760204081632653,
|
377 |
+
"grad_norm": 53196.58203125,
|
378 |
+
"learning_rate": 3.3801020408163266e-06,
|
379 |
+
"loss": 2.0087,
|
380 |
+
"step": 265
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 0.6887755102040817,
|
384 |
+
"grad_norm": 64588.515625,
|
385 |
+
"learning_rate": 3.4438775510204086e-06,
|
386 |
+
"loss": 2.0006,
|
387 |
+
"step": 270
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 0.701530612244898,
|
391 |
+
"grad_norm": 73780.546875,
|
392 |
+
"learning_rate": 3.50765306122449e-06,
|
393 |
+
"loss": 1.9965,
|
394 |
+
"step": 275
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 0.7142857142857143,
|
398 |
+
"grad_norm": 63101.05859375,
|
399 |
+
"learning_rate": 3.5714285714285718e-06,
|
400 |
+
"loss": 2.0005,
|
401 |
+
"step": 280
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 0.7270408163265306,
|
405 |
+
"grad_norm": 78616.109375,
|
406 |
+
"learning_rate": 3.6352040816326534e-06,
|
407 |
+
"loss": 1.9904,
|
408 |
+
"step": 285
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 0.7397959183673469,
|
412 |
+
"grad_norm": 79120.6484375,
|
413 |
+
"learning_rate": 3.698979591836735e-06,
|
414 |
+
"loss": 1.9891,
|
415 |
+
"step": 290
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 0.7525510204081632,
|
419 |
+
"grad_norm": 92711.09375,
|
420 |
+
"learning_rate": 3.7627551020408166e-06,
|
421 |
+
"loss": 1.9611,
|
422 |
+
"step": 295
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 0.7653061224489796,
|
426 |
+
"grad_norm": 77935.4765625,
|
427 |
+
"learning_rate": 3.826530612244898e-06,
|
428 |
+
"loss": 1.9746,
|
429 |
+
"step": 300
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 0.7780612244897959,
|
433 |
+
"grad_norm": 85825.375,
|
434 |
+
"learning_rate": 3.89030612244898e-06,
|
435 |
+
"loss": 1.9753,
|
436 |
+
"step": 305
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 0.7908163265306123,
|
440 |
+
"grad_norm": 65178.671875,
|
441 |
+
"learning_rate": 3.954081632653062e-06,
|
442 |
+
"loss": 1.9534,
|
443 |
+
"step": 310
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 0.8035714285714286,
|
447 |
+
"grad_norm": 71973.671875,
|
448 |
+
"learning_rate": 4.017857142857143e-06,
|
449 |
+
"loss": 1.9474,
|
450 |
+
"step": 315
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 0.8163265306122449,
|
454 |
+
"grad_norm": 88983.125,
|
455 |
+
"learning_rate": 4.081632653061225e-06,
|
456 |
+
"loss": 1.9492,
|
457 |
+
"step": 320
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 0.8290816326530612,
|
461 |
+
"grad_norm": 76345.84375,
|
462 |
+
"learning_rate": 4.145408163265306e-06,
|
463 |
+
"loss": 1.9513,
|
464 |
+
"step": 325
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 0.8418367346938775,
|
468 |
+
"grad_norm": 64801.640625,
|
469 |
+
"learning_rate": 4.209183673469388e-06,
|
470 |
+
"loss": 1.9583,
|
471 |
+
"step": 330
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 0.8545918367346939,
|
475 |
+
"grad_norm": 72903.8203125,
|
476 |
+
"learning_rate": 4.272959183673469e-06,
|
477 |
+
"loss": 1.9579,
|
478 |
+
"step": 335
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 0.8673469387755102,
|
482 |
+
"grad_norm": 65284.796875,
|
483 |
+
"learning_rate": 4.336734693877551e-06,
|
484 |
+
"loss": 1.9272,
|
485 |
+
"step": 340
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 0.8801020408163265,
|
489 |
+
"grad_norm": 65237.38671875,
|
490 |
+
"learning_rate": 4.400510204081633e-06,
|
491 |
+
"loss": 1.9065,
|
492 |
+
"step": 345
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 0.8928571428571429,
|
496 |
+
"grad_norm": 79606.9296875,
|
497 |
+
"learning_rate": 4.464285714285715e-06,
|
498 |
+
"loss": 1.9013,
|
499 |
+
"step": 350
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 0.9056122448979592,
|
503 |
+
"grad_norm": 76137.6328125,
|
504 |
+
"learning_rate": 4.528061224489797e-06,
|
505 |
+
"loss": 1.8892,
|
506 |
+
"step": 355
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 0.9183673469387755,
|
510 |
+
"grad_norm": 40717.60546875,
|
511 |
+
"learning_rate": 4.591836734693878e-06,
|
512 |
+
"loss": 1.9072,
|
513 |
+
"step": 360
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 0.9311224489795918,
|
517 |
+
"grad_norm": 110416.125,
|
518 |
+
"learning_rate": 4.65561224489796e-06,
|
519 |
+
"loss": 1.9024,
|
520 |
+
"step": 365
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 0.9438775510204082,
|
524 |
+
"grad_norm": 99101.6875,
|
525 |
+
"learning_rate": 4.719387755102041e-06,
|
526 |
+
"loss": 1.8979,
|
527 |
+
"step": 370
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 0.9566326530612245,
|
531 |
+
"grad_norm": 86707.7109375,
|
532 |
+
"learning_rate": 4.783163265306123e-06,
|
533 |
+
"loss": 1.878,
|
534 |
+
"step": 375
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 0.9693877551020408,
|
538 |
+
"grad_norm": 85074.75,
|
539 |
+
"learning_rate": 4.846938775510204e-06,
|
540 |
+
"loss": 1.871,
|
541 |
+
"step": 380
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 0.9821428571428571,
|
545 |
+
"grad_norm": 47668.7109375,
|
546 |
+
"learning_rate": 4.910714285714286e-06,
|
547 |
+
"loss": 1.8762,
|
548 |
+
"step": 385
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 0.9948979591836735,
|
552 |
+
"grad_norm": 133846.6875,
|
553 |
+
"learning_rate": 4.974489795918368e-06,
|
554 |
+
"loss": 1.8344,
|
555 |
+
"step": 390
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 1.0,
|
559 |
+
"eval_accuracy": 0.0857774206508638,
|
560 |
+
"eval_loss": 1.9574071168899536,
|
561 |
+
"eval_runtime": 313.0247,
|
562 |
+
"eval_samples_per_second": 15.903,
|
563 |
+
"eval_steps_per_second": 0.53,
|
564 |
+
"step": 392
|
565 |
+
},
|
566 |
+
{
|
567 |
+
"epoch": 1.0076530612244898,
|
568 |
+
"grad_norm": 95410.25,
|
569 |
+
"learning_rate": 5.0382653061224495e-06,
|
570 |
+
"loss": 1.8635,
|
571 |
+
"step": 395
|
572 |
+
},
|
573 |
+
{
|
574 |
+
"epoch": 1.0204081632653061,
|
575 |
+
"grad_norm": 86022.1484375,
|
576 |
+
"learning_rate": 5.1020408163265315e-06,
|
577 |
+
"loss": 1.8147,
|
578 |
+
"step": 400
|
579 |
+
},
|
580 |
+
{
|
581 |
+
"epoch": 1.0331632653061225,
|
582 |
+
"grad_norm": 154860.171875,
|
583 |
+
"learning_rate": 5.165816326530613e-06,
|
584 |
+
"loss": 1.8016,
|
585 |
+
"step": 405
|
586 |
+
},
|
587 |
+
{
|
588 |
+
"epoch": 1.0459183673469388,
|
589 |
+
"grad_norm": 105390.4609375,
|
590 |
+
"learning_rate": 5.229591836734695e-06,
|
591 |
+
"loss": 1.8458,
|
592 |
+
"step": 410
|
593 |
+
},
|
594 |
+
{
|
595 |
+
"epoch": 1.058673469387755,
|
596 |
+
"grad_norm": 94511.6171875,
|
597 |
+
"learning_rate": 5.293367346938776e-06,
|
598 |
+
"loss": 1.7888,
|
599 |
+
"step": 415
|
600 |
+
},
|
601 |
+
{
|
602 |
+
"epoch": 1.0714285714285714,
|
603 |
+
"grad_norm": 93031.234375,
|
604 |
+
"learning_rate": 5.357142857142857e-06,
|
605 |
+
"loss": 1.7809,
|
606 |
+
"step": 420
|
607 |
+
},
|
608 |
+
{
|
609 |
+
"epoch": 1.0841836734693877,
|
610 |
+
"grad_norm": 71911.4296875,
|
611 |
+
"learning_rate": 5.420918367346939e-06,
|
612 |
+
"loss": 1.8523,
|
613 |
+
"step": 425
|
614 |
+
},
|
615 |
+
{
|
616 |
+
"epoch": 1.096938775510204,
|
617 |
+
"grad_norm": 59985.96875,
|
618 |
+
"learning_rate": 5.48469387755102e-06,
|
619 |
+
"loss": 1.7427,
|
620 |
+
"step": 430
|
621 |
+
},
|
622 |
+
{
|
623 |
+
"epoch": 1.1096938775510203,
|
624 |
+
"grad_norm": 86531.875,
|
625 |
+
"learning_rate": 5.548469387755102e-06,
|
626 |
+
"loss": 1.7646,
|
627 |
+
"step": 435
|
628 |
+
},
|
629 |
+
{
|
630 |
+
"epoch": 1.1224489795918366,
|
631 |
+
"grad_norm": 90977.859375,
|
632 |
+
"learning_rate": 5.6122448979591834e-06,
|
633 |
+
"loss": 1.7529,
|
634 |
+
"step": 440
|
635 |
+
},
|
636 |
+
{
|
637 |
+
"epoch": 1.135204081632653,
|
638 |
+
"grad_norm": 70691.375,
|
639 |
+
"learning_rate": 5.6760204081632655e-06,
|
640 |
+
"loss": 1.7754,
|
641 |
+
"step": 445
|
642 |
+
},
|
643 |
+
{
|
644 |
+
"epoch": 1.1479591836734695,
|
645 |
+
"grad_norm": 137077.8125,
|
646 |
+
"learning_rate": 5.739795918367348e-06,
|
647 |
+
"loss": 1.73,
|
648 |
+
"step": 450
|
649 |
+
},
|
650 |
+
{
|
651 |
+
"epoch": 1.1607142857142858,
|
652 |
+
"grad_norm": 133976.8125,
|
653 |
+
"learning_rate": 5.8035714285714295e-06,
|
654 |
+
"loss": 1.7906,
|
655 |
+
"step": 455
|
656 |
+
},
|
657 |
+
{
|
658 |
+
"epoch": 1.1734693877551021,
|
659 |
+
"grad_norm": 84754.7421875,
|
660 |
+
"learning_rate": 5.867346938775511e-06,
|
661 |
+
"loss": 1.7722,
|
662 |
+
"step": 460
|
663 |
+
},
|
664 |
+
{
|
665 |
+
"epoch": 1.1862244897959184,
|
666 |
+
"grad_norm": 67924.859375,
|
667 |
+
"learning_rate": 5.931122448979593e-06,
|
668 |
+
"loss": 1.8274,
|
669 |
+
"step": 465
|
670 |
+
},
|
671 |
+
{
|
672 |
+
"epoch": 1.1989795918367347,
|
673 |
+
"grad_norm": 110309.015625,
|
674 |
+
"learning_rate": 5.994897959183674e-06,
|
675 |
+
"loss": 1.7558,
|
676 |
+
"step": 470
|
677 |
+
},
|
678 |
+
{
|
679 |
+
"epoch": 1.211734693877551,
|
680 |
+
"grad_norm": 108489.703125,
|
681 |
+
"learning_rate": 6.058673469387756e-06,
|
682 |
+
"loss": 1.7387,
|
683 |
+
"step": 475
|
684 |
+
},
|
685 |
+
{
|
686 |
+
"epoch": 1.2244897959183674,
|
687 |
+
"grad_norm": 99533.59375,
|
688 |
+
"learning_rate": 6.122448979591837e-06,
|
689 |
+
"loss": 1.7367,
|
690 |
+
"step": 480
|
691 |
+
},
|
692 |
+
{
|
693 |
+
"epoch": 1.2372448979591837,
|
694 |
+
"grad_norm": 85590.296875,
|
695 |
+
"learning_rate": 6.186224489795919e-06,
|
696 |
+
"loss": 1.7304,
|
697 |
+
"step": 485
|
698 |
+
},
|
699 |
+
{
|
700 |
+
"epoch": 1.25,
|
701 |
+
"grad_norm": 242811.53125,
|
702 |
+
"learning_rate": 6.25e-06,
|
703 |
+
"loss": 1.6764,
|
704 |
+
"step": 490
|
705 |
+
},
|
706 |
+
{
|
707 |
+
"epoch": 1.2627551020408163,
|
708 |
+
"grad_norm": 88326.6171875,
|
709 |
+
"learning_rate": 6.313775510204082e-06,
|
710 |
+
"loss": 1.7684,
|
711 |
+
"step": 495
|
712 |
+
},
|
713 |
+
{
|
714 |
+
"epoch": 1.2755102040816326,
|
715 |
+
"grad_norm": 64742.67578125,
|
716 |
+
"learning_rate": 6.3775510204081635e-06,
|
717 |
+
"loss": 1.788,
|
718 |
+
"step": 500
|
719 |
+
},
|
720 |
+
{
|
721 |
+
"epoch": 1.288265306122449,
|
722 |
+
"grad_norm": 130764.1015625,
|
723 |
+
"learning_rate": 6.4413265306122455e-06,
|
724 |
+
"loss": 1.6374,
|
725 |
+
"step": 505
|
726 |
+
},
|
727 |
+
{
|
728 |
+
"epoch": 1.3010204081632653,
|
729 |
+
"grad_norm": 105262.7578125,
|
730 |
+
"learning_rate": 6.505102040816327e-06,
|
731 |
+
"loss": 1.7669,
|
732 |
+
"step": 510
|
733 |
+
},
|
734 |
+
{
|
735 |
+
"epoch": 1.3137755102040816,
|
736 |
+
"grad_norm": 132783.3125,
|
737 |
+
"learning_rate": 6.568877551020409e-06,
|
738 |
+
"loss": 1.7428,
|
739 |
+
"step": 515
|
740 |
+
},
|
741 |
+
{
|
742 |
+
"epoch": 1.3265306122448979,
|
743 |
+
"grad_norm": 109851.921875,
|
744 |
+
"learning_rate": 6.63265306122449e-06,
|
745 |
+
"loss": 1.7023,
|
746 |
+
"step": 520
|
747 |
+
},
|
748 |
+
{
|
749 |
+
"epoch": 1.3392857142857144,
|
750 |
+
"grad_norm": 65197.0859375,
|
751 |
+
"learning_rate": 6.696428571428571e-06,
|
752 |
+
"loss": 1.7407,
|
753 |
+
"step": 525
|
754 |
+
},
|
755 |
+
{
|
756 |
+
"epoch": 1.3520408163265305,
|
757 |
+
"grad_norm": 158283.9375,
|
758 |
+
"learning_rate": 6.760204081632653e-06,
|
759 |
+
"loss": 1.7629,
|
760 |
+
"step": 530
|
761 |
+
},
|
762 |
+
{
|
763 |
+
"epoch": 1.364795918367347,
|
764 |
+
"grad_norm": 121252.4140625,
|
765 |
+
"learning_rate": 6.823979591836736e-06,
|
766 |
+
"loss": 1.7242,
|
767 |
+
"step": 535
|
768 |
+
},
|
769 |
+
{
|
770 |
+
"epoch": 1.3775510204081631,
|
771 |
+
"grad_norm": 110898.328125,
|
772 |
+
"learning_rate": 6.887755102040817e-06,
|
773 |
+
"loss": 1.6922,
|
774 |
+
"step": 540
|
775 |
+
},
|
776 |
+
{
|
777 |
+
"epoch": 1.3903061224489797,
|
778 |
+
"grad_norm": 127766.8046875,
|
779 |
+
"learning_rate": 6.951530612244899e-06,
|
780 |
+
"loss": 1.7274,
|
781 |
+
"step": 545
|
782 |
+
},
|
783 |
+
{
|
784 |
+
"epoch": 1.403061224489796,
|
785 |
+
"grad_norm": 138821.59375,
|
786 |
+
"learning_rate": 7.01530612244898e-06,
|
787 |
+
"loss": 1.6859,
|
788 |
+
"step": 550
|
789 |
+
},
|
790 |
+
{
|
791 |
+
"epoch": 1.4158163265306123,
|
792 |
+
"grad_norm": 154254.75,
|
793 |
+
"learning_rate": 7.079081632653062e-06,
|
794 |
+
"loss": 1.6988,
|
795 |
+
"step": 555
|
796 |
+
},
|
797 |
+
{
|
798 |
+
"epoch": 1.4285714285714286,
|
799 |
+
"grad_norm": 122805.65625,
|
800 |
+
"learning_rate": 7.1428571428571436e-06,
|
801 |
+
"loss": 1.7003,
|
802 |
+
"step": 560
|
803 |
+
},
|
804 |
+
{
|
805 |
+
"epoch": 1.441326530612245,
|
806 |
+
"grad_norm": 35300.796875,
|
807 |
+
"learning_rate": 7.206632653061226e-06,
|
808 |
+
"loss": 1.7224,
|
809 |
+
"step": 565
|
810 |
+
},
|
811 |
+
{
|
812 |
+
"epoch": 1.4540816326530612,
|
813 |
+
"grad_norm": 125603.1015625,
|
814 |
+
"learning_rate": 7.270408163265307e-06,
|
815 |
+
"loss": 1.8201,
|
816 |
+
"step": 570
|
817 |
+
},
|
818 |
+
{
|
819 |
+
"epoch": 1.4668367346938775,
|
820 |
+
"grad_norm": 167185.671875,
|
821 |
+
"learning_rate": 7.334183673469388e-06,
|
822 |
+
"loss": 1.7959,
|
823 |
+
"step": 575
|
824 |
+
},
|
825 |
+
{
|
826 |
+
"epoch": 1.4795918367346939,
|
827 |
+
"grad_norm": 59943.359375,
|
828 |
+
"learning_rate": 7.39795918367347e-06,
|
829 |
+
"loss": 1.6133,
|
830 |
+
"step": 580
|
831 |
+
},
|
832 |
+
{
|
833 |
+
"epoch": 1.4923469387755102,
|
834 |
+
"grad_norm": 197770.234375,
|
835 |
+
"learning_rate": 7.461734693877551e-06,
|
836 |
+
"loss": 1.709,
|
837 |
+
"step": 585
|
838 |
+
},
|
839 |
+
{
|
840 |
+
"epoch": 1.5051020408163265,
|
841 |
+
"grad_norm": 182674.5,
|
842 |
+
"learning_rate": 7.525510204081633e-06,
|
843 |
+
"loss": 1.688,
|
844 |
+
"step": 590
|
845 |
+
},
|
846 |
+
{
|
847 |
+
"epoch": 1.5178571428571428,
|
848 |
+
"grad_norm": 90262.0390625,
|
849 |
+
"learning_rate": 7.589285714285714e-06,
|
850 |
+
"loss": 1.7639,
|
851 |
+
"step": 595
|
852 |
+
},
|
853 |
+
{
|
854 |
+
"epoch": 1.5306122448979593,
|
855 |
+
"grad_norm": 119569.0859375,
|
856 |
+
"learning_rate": 7.653061224489796e-06,
|
857 |
+
"loss": 1.7587,
|
858 |
+
"step": 600
|
859 |
+
},
|
860 |
+
{
|
861 |
+
"epoch": 1.5433673469387754,
|
862 |
+
"grad_norm": 134583.546875,
|
863 |
+
"learning_rate": 7.716836734693878e-06,
|
864 |
+
"loss": 1.7174,
|
865 |
+
"step": 605
|
866 |
+
},
|
867 |
+
{
|
868 |
+
"epoch": 1.556122448979592,
|
869 |
+
"grad_norm": 123799.515625,
|
870 |
+
"learning_rate": 7.78061224489796e-06,
|
871 |
+
"loss": 1.668,
|
872 |
+
"step": 610
|
873 |
+
},
|
874 |
+
{
|
875 |
+
"epoch": 1.568877551020408,
|
876 |
+
"grad_norm": 124765.65625,
|
877 |
+
"learning_rate": 7.844387755102042e-06,
|
878 |
+
"loss": 1.691,
|
879 |
+
"step": 615
|
880 |
+
},
|
881 |
+
{
|
882 |
+
"epoch": 1.5816326530612246,
|
883 |
+
"grad_norm": 180459.28125,
|
884 |
+
"learning_rate": 7.908163265306124e-06,
|
885 |
+
"loss": 1.6133,
|
886 |
+
"step": 620
|
887 |
+
},
|
888 |
+
{
|
889 |
+
"epoch": 1.5943877551020407,
|
890 |
+
"grad_norm": 115367.0859375,
|
891 |
+
"learning_rate": 7.971938775510205e-06,
|
892 |
+
"loss": 1.7529,
|
893 |
+
"step": 625
|
894 |
+
},
|
895 |
+
{
|
896 |
+
"epoch": 1.6071428571428572,
|
897 |
+
"grad_norm": 66548.40625,
|
898 |
+
"learning_rate": 8.035714285714286e-06,
|
899 |
+
"loss": 1.7157,
|
900 |
+
"step": 630
|
901 |
+
},
|
902 |
+
{
|
903 |
+
"epoch": 1.6198979591836735,
|
904 |
+
"grad_norm": 152392.640625,
|
905 |
+
"learning_rate": 8.099489795918369e-06,
|
906 |
+
"loss": 1.7417,
|
907 |
+
"step": 635
|
908 |
+
},
|
909 |
+
{
|
910 |
+
"epoch": 1.6326530612244898,
|
911 |
+
"grad_norm": 131193.015625,
|
912 |
+
"learning_rate": 8.16326530612245e-06,
|
913 |
+
"loss": 1.7704,
|
914 |
+
"step": 640
|
915 |
+
},
|
916 |
+
{
|
917 |
+
"epoch": 1.6454081632653061,
|
918 |
+
"grad_norm": 154705.71875,
|
919 |
+
"learning_rate": 8.227040816326531e-06,
|
920 |
+
"loss": 1.6573,
|
921 |
+
"step": 645
|
922 |
+
},
|
923 |
+
{
|
924 |
+
"epoch": 1.6581632653061225,
|
925 |
+
"grad_norm": 113254.484375,
|
926 |
+
"learning_rate": 8.290816326530612e-06,
|
927 |
+
"loss": 1.6394,
|
928 |
+
"step": 650
|
929 |
+
},
|
930 |
+
{
|
931 |
+
"epoch": 1.6709183673469388,
|
932 |
+
"grad_norm": 163821.484375,
|
933 |
+
"learning_rate": 8.354591836734695e-06,
|
934 |
+
"loss": 1.6884,
|
935 |
+
"step": 655
|
936 |
+
},
|
937 |
+
{
|
938 |
+
"epoch": 1.683673469387755,
|
939 |
+
"grad_norm": 141981.828125,
|
940 |
+
"learning_rate": 8.418367346938776e-06,
|
941 |
+
"loss": 1.5955,
|
942 |
+
"step": 660
|
943 |
+
},
|
944 |
+
{
|
945 |
+
"epoch": 1.6964285714285714,
|
946 |
+
"grad_norm": 158287.6875,
|
947 |
+
"learning_rate": 8.482142857142858e-06,
|
948 |
+
"loss": 1.6885,
|
949 |
+
"step": 665
|
950 |
+
},
|
951 |
+
{
|
952 |
+
"epoch": 1.7091836734693877,
|
953 |
+
"grad_norm": 166876.4375,
|
954 |
+
"learning_rate": 8.545918367346939e-06,
|
955 |
+
"loss": 1.7253,
|
956 |
+
"step": 670
|
957 |
+
},
|
958 |
+
{
|
959 |
+
"epoch": 1.7219387755102042,
|
960 |
+
"grad_norm": 111882.171875,
|
961 |
+
"learning_rate": 8.609693877551022e-06,
|
962 |
+
"loss": 1.7057,
|
963 |
+
"step": 675
|
964 |
+
},
|
965 |
+
{
|
966 |
+
"epoch": 1.7346938775510203,
|
967 |
+
"grad_norm": 237840.828125,
|
968 |
+
"learning_rate": 8.673469387755103e-06,
|
969 |
+
"loss": 1.7132,
|
970 |
+
"step": 680
|
971 |
+
},
|
972 |
+
{
|
973 |
+
"epoch": 1.7474489795918369,
|
974 |
+
"grad_norm": 83344.59375,
|
975 |
+
"learning_rate": 8.737244897959184e-06,
|
976 |
+
"loss": 1.733,
|
977 |
+
"step": 685
|
978 |
+
},
|
979 |
+
{
|
980 |
+
"epoch": 1.760204081632653,
|
981 |
+
"grad_norm": 30312.462890625,
|
982 |
+
"learning_rate": 8.801020408163265e-06,
|
983 |
+
"loss": 1.7926,
|
984 |
+
"step": 690
|
985 |
+
},
|
986 |
+
{
|
987 |
+
"epoch": 1.7729591836734695,
|
988 |
+
"grad_norm": 89868.9609375,
|
989 |
+
"learning_rate": 8.864795918367348e-06,
|
990 |
+
"loss": 1.6389,
|
991 |
+
"step": 695
|
992 |
+
},
|
993 |
+
{
|
994 |
+
"epoch": 1.7857142857142856,
|
995 |
+
"grad_norm": 182723.09375,
|
996 |
+
"learning_rate": 8.92857142857143e-06,
|
997 |
+
"loss": 1.7533,
|
998 |
+
"step": 700
|
999 |
+
},
|
1000 |
+
{
|
1001 |
+
"epoch": 1.7984693877551021,
|
1002 |
+
"grad_norm": 104328.2421875,
|
1003 |
+
"learning_rate": 8.992346938775512e-06,
|
1004 |
+
"loss": 1.7424,
|
1005 |
+
"step": 705
|
1006 |
+
},
|
1007 |
+
{
|
1008 |
+
"epoch": 1.8112244897959182,
|
1009 |
+
"grad_norm": 110120.609375,
|
1010 |
+
"learning_rate": 9.056122448979593e-06,
|
1011 |
+
"loss": 1.6961,
|
1012 |
+
"step": 710
|
1013 |
+
},
|
1014 |
+
{
|
1015 |
+
"epoch": 1.8239795918367347,
|
1016 |
+
"grad_norm": 114349.7578125,
|
1017 |
+
"learning_rate": 9.119897959183674e-06,
|
1018 |
+
"loss": 1.7753,
|
1019 |
+
"step": 715
|
1020 |
+
},
|
1021 |
+
{
|
1022 |
+
"epoch": 1.836734693877551,
|
1023 |
+
"grad_norm": 103603.8828125,
|
1024 |
+
"learning_rate": 9.183673469387756e-06,
|
1025 |
+
"loss": 1.777,
|
1026 |
+
"step": 720
|
1027 |
+
},
|
1028 |
+
{
|
1029 |
+
"epoch": 1.8494897959183674,
|
1030 |
+
"grad_norm": 173858.828125,
|
1031 |
+
"learning_rate": 9.247448979591837e-06,
|
1032 |
+
"loss": 1.7218,
|
1033 |
+
"step": 725
|
1034 |
+
},
|
1035 |
+
{
|
1036 |
+
"epoch": 1.8622448979591837,
|
1037 |
+
"grad_norm": 255108.96875,
|
1038 |
+
"learning_rate": 9.31122448979592e-06,
|
1039 |
+
"loss": 1.7319,
|
1040 |
+
"step": 730
|
1041 |
+
},
|
1042 |
+
{
|
1043 |
+
"epoch": 1.875,
|
1044 |
+
"grad_norm": 38653.734375,
|
1045 |
+
"learning_rate": 9.375000000000001e-06,
|
1046 |
+
"loss": 1.6574,
|
1047 |
+
"step": 735
|
1048 |
+
},
|
1049 |
+
{
|
1050 |
+
"epoch": 1.8877551020408163,
|
1051 |
+
"grad_norm": 141534.234375,
|
1052 |
+
"learning_rate": 9.438775510204082e-06,
|
1053 |
+
"loss": 1.7595,
|
1054 |
+
"step": 740
|
1055 |
+
},
|
1056 |
+
{
|
1057 |
+
"epoch": 1.9005102040816326,
|
1058 |
+
"grad_norm": 182591.890625,
|
1059 |
+
"learning_rate": 9.502551020408163e-06,
|
1060 |
+
"loss": 1.6796,
|
1061 |
+
"step": 745
|
1062 |
+
},
|
1063 |
+
{
|
1064 |
+
"epoch": 1.913265306122449,
|
1065 |
+
"grad_norm": 75052.21875,
|
1066 |
+
"learning_rate": 9.566326530612246e-06,
|
1067 |
+
"loss": 1.754,
|
1068 |
+
"step": 750
|
1069 |
+
},
|
1070 |
+
{
|
1071 |
+
"epoch": 1.9260204081632653,
|
1072 |
+
"grad_norm": 108983.921875,
|
1073 |
+
"learning_rate": 9.630102040816327e-06,
|
1074 |
+
"loss": 1.6907,
|
1075 |
+
"step": 755
|
1076 |
+
},
|
1077 |
+
{
|
1078 |
+
"epoch": 1.9387755102040818,
|
1079 |
+
"grad_norm": 110641.828125,
|
1080 |
+
"learning_rate": 9.693877551020408e-06,
|
1081 |
+
"loss": 1.7041,
|
1082 |
+
"step": 760
|
1083 |
+
},
|
1084 |
+
{
|
1085 |
+
"epoch": 1.9515306122448979,
|
1086 |
+
"grad_norm": 119189.8203125,
|
1087 |
+
"learning_rate": 9.75765306122449e-06,
|
1088 |
+
"loss": 1.6054,
|
1089 |
+
"step": 765
|
1090 |
+
},
|
1091 |
+
{
|
1092 |
+
"epoch": 1.9642857142857144,
|
1093 |
+
"grad_norm": 62365.9453125,
|
1094 |
+
"learning_rate": 9.821428571428573e-06,
|
1095 |
+
"loss": 1.6862,
|
1096 |
+
"step": 770
|
1097 |
+
},
|
1098 |
+
{
|
1099 |
+
"epoch": 1.9770408163265305,
|
1100 |
+
"grad_norm": 103043.6171875,
|
1101 |
+
"learning_rate": 9.885204081632654e-06,
|
1102 |
+
"loss": 1.6596,
|
1103 |
+
"step": 775
|
1104 |
+
},
|
1105 |
+
{
|
1106 |
+
"epoch": 1.989795918367347,
|
1107 |
+
"grad_norm": 117368.546875,
|
1108 |
+
"learning_rate": 9.948979591836737e-06,
|
1109 |
+
"loss": 1.6737,
|
1110 |
+
"step": 780
|
1111 |
+
},
|
1112 |
+
{
|
1113 |
+
"epoch": 2.0,
|
1114 |
+
"eval_accuracy": 0.0857774206508638,
|
1115 |
+
"eval_loss": 1.8835231065750122,
|
1116 |
+
"eval_runtime": 308.4538,
|
1117 |
+
"eval_samples_per_second": 16.139,
|
1118 |
+
"eval_steps_per_second": 0.538,
|
1119 |
+
"step": 784
|
1120 |
+
},
|
1121 |
+
{
|
1122 |
+
"epoch": 2.002551020408163,
|
1123 |
+
"grad_norm": 219828.25,
|
1124 |
+
"learning_rate": 9.996811224489797e-06,
|
1125 |
+
"loss": 1.604,
|
1126 |
+
"step": 785
|
1127 |
+
},
|
1128 |
+
{
|
1129 |
+
"epoch": 2.0153061224489797,
|
1130 |
+
"grad_norm": 132893.65625,
|
1131 |
+
"learning_rate": 9.980867346938775e-06,
|
1132 |
+
"loss": 1.7278,
|
1133 |
+
"step": 790
|
1134 |
+
},
|
1135 |
+
{
|
1136 |
+
"epoch": 2.0280612244897958,
|
1137 |
+
"grad_norm": 103370.2109375,
|
1138 |
+
"learning_rate": 9.964923469387756e-06,
|
1139 |
+
"loss": 1.6631,
|
1140 |
+
"step": 795
|
1141 |
+
},
|
1142 |
+
{
|
1143 |
+
"epoch": 2.0408163265306123,
|
1144 |
+
"grad_norm": 201988.0625,
|
1145 |
+
"learning_rate": 9.948979591836737e-06,
|
1146 |
+
"loss": 1.7244,
|
1147 |
+
"step": 800
|
1148 |
+
},
|
1149 |
+
{
|
1150 |
+
"epoch": 2.0535714285714284,
|
1151 |
+
"grad_norm": 67003.359375,
|
1152 |
+
"learning_rate": 9.933035714285715e-06,
|
1153 |
+
"loss": 1.7057,
|
1154 |
+
"step": 805
|
1155 |
+
},
|
1156 |
+
{
|
1157 |
+
"epoch": 2.066326530612245,
|
1158 |
+
"grad_norm": 306698.625,
|
1159 |
+
"learning_rate": 9.917091836734694e-06,
|
1160 |
+
"loss": 1.7915,
|
1161 |
+
"step": 810
|
1162 |
+
},
|
1163 |
+
{
|
1164 |
+
"epoch": 2.079081632653061,
|
1165 |
+
"grad_norm": 168723.765625,
|
1166 |
+
"learning_rate": 9.901147959183675e-06,
|
1167 |
+
"loss": 1.7387,
|
1168 |
+
"step": 815
|
1169 |
+
},
|
1170 |
+
{
|
1171 |
+
"epoch": 2.0918367346938775,
|
1172 |
+
"grad_norm": 159844.828125,
|
1173 |
+
"learning_rate": 9.885204081632654e-06,
|
1174 |
+
"loss": 1.716,
|
1175 |
+
"step": 820
|
1176 |
+
},
|
1177 |
+
{
|
1178 |
+
"epoch": 2.104591836734694,
|
1179 |
+
"grad_norm": 131381.34375,
|
1180 |
+
"learning_rate": 9.869260204081633e-06,
|
1181 |
+
"loss": 1.7793,
|
1182 |
+
"step": 825
|
1183 |
+
},
|
1184 |
+
{
|
1185 |
+
"epoch": 2.11734693877551,
|
1186 |
+
"grad_norm": 134116.8125,
|
1187 |
+
"learning_rate": 9.853316326530613e-06,
|
1188 |
+
"loss": 1.6806,
|
1189 |
+
"step": 830
|
1190 |
+
},
|
1191 |
+
{
|
1192 |
+
"epoch": 2.1301020408163267,
|
1193 |
+
"grad_norm": 222184.515625,
|
1194 |
+
"learning_rate": 9.837372448979594e-06,
|
1195 |
+
"loss": 1.6068,
|
1196 |
+
"step": 835
|
1197 |
+
},
|
1198 |
+
{
|
1199 |
+
"epoch": 2.142857142857143,
|
1200 |
+
"grad_norm": 151780.703125,
|
1201 |
+
"learning_rate": 9.821428571428573e-06,
|
1202 |
+
"loss": 1.7369,
|
1203 |
+
"step": 840
|
1204 |
+
},
|
1205 |
+
{
|
1206 |
+
"epoch": 2.1556122448979593,
|
1207 |
+
"grad_norm": 150746.140625,
|
1208 |
+
"learning_rate": 9.805484693877551e-06,
|
1209 |
+
"loss": 1.6551,
|
1210 |
+
"step": 845
|
1211 |
+
},
|
1212 |
+
{
|
1213 |
+
"epoch": 2.1683673469387754,
|
1214 |
+
"grad_norm": 136112.78125,
|
1215 |
+
"learning_rate": 9.789540816326532e-06,
|
1216 |
+
"loss": 1.7609,
|
1217 |
+
"step": 850
|
1218 |
+
},
|
1219 |
+
{
|
1220 |
+
"epoch": 2.181122448979592,
|
1221 |
+
"grad_norm": 101995.4140625,
|
1222 |
+
"learning_rate": 9.77359693877551e-06,
|
1223 |
+
"loss": 1.7171,
|
1224 |
+
"step": 855
|
1225 |
+
},
|
1226 |
+
{
|
1227 |
+
"epoch": 2.193877551020408,
|
1228 |
+
"grad_norm": 46676.8046875,
|
1229 |
+
"learning_rate": 9.75765306122449e-06,
|
1230 |
+
"loss": 1.5907,
|
1231 |
+
"step": 860
|
1232 |
+
},
|
1233 |
+
{
|
1234 |
+
"epoch": 2.2066326530612246,
|
1235 |
+
"grad_norm": 143790.296875,
|
1236 |
+
"learning_rate": 9.74170918367347e-06,
|
1237 |
+
"loss": 1.6003,
|
1238 |
+
"step": 865
|
1239 |
+
},
|
1240 |
+
{
|
1241 |
+
"epoch": 2.2193877551020407,
|
1242 |
+
"grad_norm": 152429.078125,
|
1243 |
+
"learning_rate": 9.72576530612245e-06,
|
1244 |
+
"loss": 1.5825,
|
1245 |
+
"step": 870
|
1246 |
+
},
|
1247 |
+
{
|
1248 |
+
"epoch": 2.232142857142857,
|
1249 |
+
"grad_norm": 232178.03125,
|
1250 |
+
"learning_rate": 9.70982142857143e-06,
|
1251 |
+
"loss": 1.6564,
|
1252 |
+
"step": 875
|
1253 |
+
},
|
1254 |
+
{
|
1255 |
+
"epoch": 2.2448979591836733,
|
1256 |
+
"grad_norm": 87462.2109375,
|
1257 |
+
"learning_rate": 9.693877551020408e-06,
|
1258 |
+
"loss": 1.6744,
|
1259 |
+
"step": 880
|
1260 |
+
},
|
1261 |
+
{
|
1262 |
+
"epoch": 2.25765306122449,
|
1263 |
+
"grad_norm": 144245.515625,
|
1264 |
+
"learning_rate": 9.677933673469389e-06,
|
1265 |
+
"loss": 1.6895,
|
1266 |
+
"step": 885
|
1267 |
+
},
|
1268 |
+
{
|
1269 |
+
"epoch": 2.270408163265306,
|
1270 |
+
"grad_norm": 110020.5546875,
|
1271 |
+
"learning_rate": 9.661989795918368e-06,
|
1272 |
+
"loss": 1.7014,
|
1273 |
+
"step": 890
|
1274 |
+
},
|
1275 |
+
{
|
1276 |
+
"epoch": 2.2831632653061225,
|
1277 |
+
"grad_norm": 192982.6875,
|
1278 |
+
"learning_rate": 9.646045918367348e-06,
|
1279 |
+
"loss": 1.6613,
|
1280 |
+
"step": 895
|
1281 |
+
},
|
1282 |
+
{
|
1283 |
+
"epoch": 2.295918367346939,
|
1284 |
+
"grad_norm": 103864.7734375,
|
1285 |
+
"learning_rate": 9.630102040816327e-06,
|
1286 |
+
"loss": 1.6505,
|
1287 |
+
"step": 900
|
1288 |
+
},
|
1289 |
+
{
|
1290 |
+
"epoch": 2.308673469387755,
|
1291 |
+
"grad_norm": 171332.34375,
|
1292 |
+
"learning_rate": 9.614158163265306e-06,
|
1293 |
+
"loss": 1.764,
|
1294 |
+
"step": 905
|
1295 |
+
},
|
1296 |
+
{
|
1297 |
+
"epoch": 2.3214285714285716,
|
1298 |
+
"grad_norm": 151310.984375,
|
1299 |
+
"learning_rate": 9.598214285714287e-06,
|
1300 |
+
"loss": 1.7049,
|
1301 |
+
"step": 910
|
1302 |
+
},
|
1303 |
+
{
|
1304 |
+
"epoch": 2.3341836734693877,
|
1305 |
+
"grad_norm": 50600.53125,
|
1306 |
+
"learning_rate": 9.582270408163266e-06,
|
1307 |
+
"loss": 1.7203,
|
1308 |
+
"step": 915
|
1309 |
+
},
|
1310 |
+
{
|
1311 |
+
"epoch": 2.3469387755102042,
|
1312 |
+
"grad_norm": 219012.625,
|
1313 |
+
"learning_rate": 9.566326530612246e-06,
|
1314 |
+
"loss": 1.7004,
|
1315 |
+
"step": 920
|
1316 |
+
},
|
1317 |
+
{
|
1318 |
+
"epoch": 2.3596938775510203,
|
1319 |
+
"grad_norm": 120568.4765625,
|
1320 |
+
"learning_rate": 9.550382653061225e-06,
|
1321 |
+
"loss": 1.6287,
|
1322 |
+
"step": 925
|
1323 |
+
},
|
1324 |
+
{
|
1325 |
+
"epoch": 2.372448979591837,
|
1326 |
+
"grad_norm": 103431.390625,
|
1327 |
+
"learning_rate": 9.534438775510206e-06,
|
1328 |
+
"loss": 1.6989,
|
1329 |
+
"step": 930
|
1330 |
+
},
|
1331 |
+
{
|
1332 |
+
"epoch": 2.385204081632653,
|
1333 |
+
"grad_norm": 75834.46875,
|
1334 |
+
"learning_rate": 9.518494897959184e-06,
|
1335 |
+
"loss": 1.666,
|
1336 |
+
"step": 935
|
1337 |
+
},
|
1338 |
+
{
|
1339 |
+
"epoch": 2.3979591836734695,
|
1340 |
+
"grad_norm": 106555.3828125,
|
1341 |
+
"learning_rate": 9.502551020408163e-06,
|
1342 |
+
"loss": 1.6855,
|
1343 |
+
"step": 940
|
1344 |
+
},
|
1345 |
+
{
|
1346 |
+
"epoch": 2.4107142857142856,
|
1347 |
+
"grad_norm": 172204.640625,
|
1348 |
+
"learning_rate": 9.486607142857144e-06,
|
1349 |
+
"loss": 1.5667,
|
1350 |
+
"step": 945
|
1351 |
+
},
|
1352 |
+
{
|
1353 |
+
"epoch": 2.423469387755102,
|
1354 |
+
"grad_norm": 234761.484375,
|
1355 |
+
"learning_rate": 9.470663265306124e-06,
|
1356 |
+
"loss": 1.5812,
|
1357 |
+
"step": 950
|
1358 |
+
},
|
1359 |
+
{
|
1360 |
+
"epoch": 2.436224489795918,
|
1361 |
+
"grad_norm": 184700.15625,
|
1362 |
+
"learning_rate": 9.454719387755103e-06,
|
1363 |
+
"loss": 1.5293,
|
1364 |
+
"step": 955
|
1365 |
+
},
|
1366 |
+
{
|
1367 |
+
"epoch": 2.4489795918367347,
|
1368 |
+
"grad_norm": 189192.515625,
|
1369 |
+
"learning_rate": 9.438775510204082e-06,
|
1370 |
+
"loss": 1.5867,
|
1371 |
+
"step": 960
|
1372 |
+
},
|
1373 |
+
{
|
1374 |
+
"epoch": 2.461734693877551,
|
1375 |
+
"grad_norm": 123937.0078125,
|
1376 |
+
"learning_rate": 9.422831632653063e-06,
|
1377 |
+
"loss": 1.692,
|
1378 |
+
"step": 965
|
1379 |
+
},
|
1380 |
+
{
|
1381 |
+
"epoch": 2.4744897959183674,
|
1382 |
+
"grad_norm": 215251.9375,
|
1383 |
+
"learning_rate": 9.406887755102041e-06,
|
1384 |
+
"loss": 1.6901,
|
1385 |
+
"step": 970
|
1386 |
+
},
|
1387 |
+
{
|
1388 |
+
"epoch": 2.487244897959184,
|
1389 |
+
"grad_norm": 176675.5625,
|
1390 |
+
"learning_rate": 9.39094387755102e-06,
|
1391 |
+
"loss": 1.59,
|
1392 |
+
"step": 975
|
1393 |
+
},
|
1394 |
+
{
|
1395 |
+
"epoch": 2.5,
|
1396 |
+
"grad_norm": 134754.96875,
|
1397 |
+
"learning_rate": 9.375000000000001e-06,
|
1398 |
+
"loss": 1.5592,
|
1399 |
+
"step": 980
|
1400 |
+
},
|
1401 |
+
{
|
1402 |
+
"epoch": 2.512755102040816,
|
1403 |
+
"grad_norm": 103404.9921875,
|
1404 |
+
"learning_rate": 9.359056122448981e-06,
|
1405 |
+
"loss": 1.6001,
|
1406 |
+
"step": 985
|
1407 |
+
},
|
1408 |
+
{
|
1409 |
+
"epoch": 2.5255102040816326,
|
1410 |
+
"grad_norm": 147972.421875,
|
1411 |
+
"learning_rate": 9.343112244897959e-06,
|
1412 |
+
"loss": 1.6028,
|
1413 |
+
"step": 990
|
1414 |
+
},
|
1415 |
+
{
|
1416 |
+
"epoch": 2.538265306122449,
|
1417 |
+
"grad_norm": 152136.828125,
|
1418 |
+
"learning_rate": 9.327168367346939e-06,
|
1419 |
+
"loss": 1.6255,
|
1420 |
+
"step": 995
|
1421 |
+
},
|
1422 |
+
{
|
1423 |
+
"epoch": 2.5510204081632653,
|
1424 |
+
"grad_norm": 118834.9609375,
|
1425 |
+
"learning_rate": 9.31122448979592e-06,
|
1426 |
+
"loss": 1.6274,
|
1427 |
+
"step": 1000
|
1428 |
+
},
|
1429 |
+
{
|
1430 |
+
"epoch": 2.563775510204082,
|
1431 |
+
"grad_norm": 127266.28125,
|
1432 |
+
"learning_rate": 9.295280612244899e-06,
|
1433 |
+
"loss": 1.6929,
|
1434 |
+
"step": 1005
|
1435 |
+
},
|
1436 |
+
{
|
1437 |
+
"epoch": 2.576530612244898,
|
1438 |
+
"grad_norm": 131886.953125,
|
1439 |
+
"learning_rate": 9.279336734693877e-06,
|
1440 |
+
"loss": 1.5158,
|
1441 |
+
"step": 1010
|
1442 |
+
},
|
1443 |
+
{
|
1444 |
+
"epoch": 2.5892857142857144,
|
1445 |
+
"grad_norm": 221068.84375,
|
1446 |
+
"learning_rate": 9.263392857142858e-06,
|
1447 |
+
"loss": 1.6322,
|
1448 |
+
"step": 1015
|
1449 |
+
},
|
1450 |
+
{
|
1451 |
+
"epoch": 2.6020408163265305,
|
1452 |
+
"grad_norm": 260819.6875,
|
1453 |
+
"learning_rate": 9.247448979591837e-06,
|
1454 |
+
"loss": 1.6103,
|
1455 |
+
"step": 1020
|
1456 |
+
},
|
1457 |
+
{
|
1458 |
+
"epoch": 2.614795918367347,
|
1459 |
+
"grad_norm": 106676.7421875,
|
1460 |
+
"learning_rate": 9.231505102040817e-06,
|
1461 |
+
"loss": 1.5057,
|
1462 |
+
"step": 1025
|
1463 |
+
},
|
1464 |
+
{
|
1465 |
+
"epoch": 2.627551020408163,
|
1466 |
+
"grad_norm": 217982.078125,
|
1467 |
+
"learning_rate": 9.215561224489796e-06,
|
1468 |
+
"loss": 1.6204,
|
1469 |
+
"step": 1030
|
1470 |
+
},
|
1471 |
+
{
|
1472 |
+
"epoch": 2.6403061224489797,
|
1473 |
+
"grad_norm": 110120.5390625,
|
1474 |
+
"learning_rate": 9.199617346938777e-06,
|
1475 |
+
"loss": 1.5733,
|
1476 |
+
"step": 1035
|
1477 |
+
},
|
1478 |
+
{
|
1479 |
+
"epoch": 2.6530612244897958,
|
1480 |
+
"grad_norm": 80831.828125,
|
1481 |
+
"learning_rate": 9.183673469387756e-06,
|
1482 |
+
"loss": 1.569,
|
1483 |
+
"step": 1040
|
1484 |
+
},
|
1485 |
+
{
|
1486 |
+
"epoch": 2.6658163265306123,
|
1487 |
+
"grad_norm": 181651.515625,
|
1488 |
+
"learning_rate": 9.167729591836736e-06,
|
1489 |
+
"loss": 1.6483,
|
1490 |
+
"step": 1045
|
1491 |
+
},
|
1492 |
+
{
|
1493 |
+
"epoch": 2.678571428571429,
|
1494 |
+
"grad_norm": 272886.125,
|
1495 |
+
"learning_rate": 9.151785714285715e-06,
|
1496 |
+
"loss": 1.4335,
|
1497 |
+
"step": 1050
|
1498 |
+
},
|
1499 |
+
{
|
1500 |
+
"epoch": 2.691326530612245,
|
1501 |
+
"grad_norm": 108633.09375,
|
1502 |
+
"learning_rate": 9.135841836734694e-06,
|
1503 |
+
"loss": 1.5432,
|
1504 |
+
"step": 1055
|
1505 |
+
},
|
1506 |
+
{
|
1507 |
+
"epoch": 2.704081632653061,
|
1508 |
+
"grad_norm": 165070.234375,
|
1509 |
+
"learning_rate": 9.119897959183674e-06,
|
1510 |
+
"loss": 1.6833,
|
1511 |
+
"step": 1060
|
1512 |
+
},
|
1513 |
+
{
|
1514 |
+
"epoch": 2.7168367346938775,
|
1515 |
+
"grad_norm": 165150.328125,
|
1516 |
+
"learning_rate": 9.103954081632653e-06,
|
1517 |
+
"loss": 1.6997,
|
1518 |
+
"step": 1065
|
1519 |
+
},
|
1520 |
+
{
|
1521 |
+
"epoch": 2.729591836734694,
|
1522 |
+
"grad_norm": 97369.4765625,
|
1523 |
+
"learning_rate": 9.088010204081634e-06,
|
1524 |
+
"loss": 1.638,
|
1525 |
+
"step": 1070
|
1526 |
+
},
|
1527 |
+
{
|
1528 |
+
"epoch": 2.74234693877551,
|
1529 |
+
"grad_norm": 175838.71875,
|
1530 |
+
"learning_rate": 9.072066326530613e-06,
|
1531 |
+
"loss": 1.7033,
|
1532 |
+
"step": 1075
|
1533 |
+
},
|
1534 |
+
{
|
1535 |
+
"epoch": 2.7551020408163263,
|
1536 |
+
"grad_norm": 162969.171875,
|
1537 |
+
"learning_rate": 9.056122448979593e-06,
|
1538 |
+
"loss": 1.5911,
|
1539 |
+
"step": 1080
|
1540 |
+
},
|
1541 |
+
{
|
1542 |
+
"epoch": 2.767857142857143,
|
1543 |
+
"grad_norm": 189243.03125,
|
1544 |
+
"learning_rate": 9.040178571428572e-06,
|
1545 |
+
"loss": 1.4564,
|
1546 |
+
"step": 1085
|
1547 |
+
},
|
1548 |
+
{
|
1549 |
+
"epoch": 2.7806122448979593,
|
1550 |
+
"grad_norm": 191947.484375,
|
1551 |
+
"learning_rate": 9.024234693877551e-06,
|
1552 |
+
"loss": 1.7732,
|
1553 |
+
"step": 1090
|
1554 |
+
},
|
1555 |
+
{
|
1556 |
+
"epoch": 2.7933673469387754,
|
1557 |
+
"grad_norm": 143765.234375,
|
1558 |
+
"learning_rate": 9.008290816326532e-06,
|
1559 |
+
"loss": 1.5865,
|
1560 |
+
"step": 1095
|
1561 |
+
},
|
1562 |
+
{
|
1563 |
+
"epoch": 2.806122448979592,
|
1564 |
+
"grad_norm": 193076.421875,
|
1565 |
+
"learning_rate": 8.992346938775512e-06,
|
1566 |
+
"loss": 1.3999,
|
1567 |
+
"step": 1100
|
1568 |
+
},
|
1569 |
+
{
|
1570 |
+
"epoch": 2.818877551020408,
|
1571 |
+
"grad_norm": 116660.359375,
|
1572 |
+
"learning_rate": 8.97640306122449e-06,
|
1573 |
+
"loss": 1.5108,
|
1574 |
+
"step": 1105
|
1575 |
+
},
|
1576 |
+
{
|
1577 |
+
"epoch": 2.8316326530612246,
|
1578 |
+
"grad_norm": 130521.6484375,
|
1579 |
+
"learning_rate": 8.96045918367347e-06,
|
1580 |
+
"loss": 1.5463,
|
1581 |
+
"step": 1110
|
1582 |
+
},
|
1583 |
+
{
|
1584 |
+
"epoch": 2.8443877551020407,
|
1585 |
+
"grad_norm": 122570.4609375,
|
1586 |
+
"learning_rate": 8.94451530612245e-06,
|
1587 |
+
"loss": 1.5438,
|
1588 |
+
"step": 1115
|
1589 |
+
},
|
1590 |
+
{
|
1591 |
+
"epoch": 2.857142857142857,
|
1592 |
+
"grad_norm": 162270.4375,
|
1593 |
+
"learning_rate": 8.92857142857143e-06,
|
1594 |
+
"loss": 1.4326,
|
1595 |
+
"step": 1120
|
1596 |
+
},
|
1597 |
+
{
|
1598 |
+
"epoch": 2.8698979591836737,
|
1599 |
+
"grad_norm": 98064.515625,
|
1600 |
+
"learning_rate": 8.912627551020408e-06,
|
1601 |
+
"loss": 1.5871,
|
1602 |
+
"step": 1125
|
1603 |
+
},
|
1604 |
+
{
|
1605 |
+
"epoch": 2.88265306122449,
|
1606 |
+
"grad_norm": 211316.21875,
|
1607 |
+
"learning_rate": 8.896683673469389e-06,
|
1608 |
+
"loss": 1.4914,
|
1609 |
+
"step": 1130
|
1610 |
+
},
|
1611 |
+
{
|
1612 |
+
"epoch": 2.895408163265306,
|
1613 |
+
"grad_norm": 156801.890625,
|
1614 |
+
"learning_rate": 8.880739795918368e-06,
|
1615 |
+
"loss": 1.6379,
|
1616 |
+
"step": 1135
|
1617 |
+
},
|
1618 |
+
{
|
1619 |
+
"epoch": 2.9081632653061225,
|
1620 |
+
"grad_norm": 142554.421875,
|
1621 |
+
"learning_rate": 8.864795918367348e-06,
|
1622 |
+
"loss": 1.604,
|
1623 |
+
"step": 1140
|
1624 |
+
},
|
1625 |
+
{
|
1626 |
+
"epoch": 2.920918367346939,
|
1627 |
+
"grad_norm": 173248.734375,
|
1628 |
+
"learning_rate": 8.848852040816327e-06,
|
1629 |
+
"loss": 1.6187,
|
1630 |
+
"step": 1145
|
1631 |
+
},
|
1632 |
+
{
|
1633 |
+
"epoch": 2.933673469387755,
|
1634 |
+
"grad_norm": 108233.78125,
|
1635 |
+
"learning_rate": 8.832908163265307e-06,
|
1636 |
+
"loss": 1.5321,
|
1637 |
+
"step": 1150
|
1638 |
+
},
|
1639 |
+
{
|
1640 |
+
"epoch": 2.946428571428571,
|
1641 |
+
"grad_norm": 190112.34375,
|
1642 |
+
"learning_rate": 8.816964285714286e-06,
|
1643 |
+
"loss": 1.584,
|
1644 |
+
"step": 1155
|
1645 |
+
},
|
1646 |
+
{
|
1647 |
+
"epoch": 2.9591836734693877,
|
1648 |
+
"grad_norm": 244386.828125,
|
1649 |
+
"learning_rate": 8.801020408163265e-06,
|
1650 |
+
"loss": 1.5631,
|
1651 |
+
"step": 1160
|
1652 |
+
},
|
1653 |
+
{
|
1654 |
+
"epoch": 2.9719387755102042,
|
1655 |
+
"grad_norm": 169686.921875,
|
1656 |
+
"learning_rate": 8.785076530612246e-06,
|
1657 |
+
"loss": 1.5537,
|
1658 |
+
"step": 1165
|
1659 |
+
},
|
1660 |
+
{
|
1661 |
+
"epoch": 2.9846938775510203,
|
1662 |
+
"grad_norm": 162361.109375,
|
1663 |
+
"learning_rate": 8.769132653061225e-06,
|
1664 |
+
"loss": 1.4138,
|
1665 |
+
"step": 1170
|
1666 |
+
},
|
1667 |
+
{
|
1668 |
+
"epoch": 2.997448979591837,
|
1669 |
+
"grad_norm": 203224.546875,
|
1670 |
+
"learning_rate": 8.753188775510205e-06,
|
1671 |
+
"loss": 1.4787,
|
1672 |
+
"step": 1175
|
1673 |
+
},
|
1674 |
+
{
|
1675 |
+
"epoch": 3.0,
|
1676 |
+
"eval_accuracy": 0.35918039373242266,
|
1677 |
+
"eval_loss": 1.7329986095428467,
|
1678 |
+
"eval_runtime": 304.5388,
|
1679 |
+
"eval_samples_per_second": 16.346,
|
1680 |
+
"eval_steps_per_second": 0.545,
|
1681 |
+
"step": 1176
|
1682 |
+
},
|
1683 |
+
{
|
1684 |
+
"epoch": 3.010204081632653,
|
1685 |
+
"grad_norm": 157683.71875,
|
1686 |
+
"learning_rate": 8.737244897959184e-06,
|
1687 |
+
"loss": 1.4765,
|
1688 |
+
"step": 1180
|
1689 |
+
},
|
1690 |
+
{
|
1691 |
+
"epoch": 3.0229591836734695,
|
1692 |
+
"grad_norm": 133687.40625,
|
1693 |
+
"learning_rate": 8.721301020408165e-06,
|
1694 |
+
"loss": 1.4614,
|
1695 |
+
"step": 1185
|
1696 |
+
},
|
1697 |
+
{
|
1698 |
+
"epoch": 3.0357142857142856,
|
1699 |
+
"grad_norm": 156001.25,
|
1700 |
+
"learning_rate": 8.705357142857143e-06,
|
1701 |
+
"loss": 1.5644,
|
1702 |
+
"step": 1190
|
1703 |
+
},
|
1704 |
+
{
|
1705 |
+
"epoch": 3.048469387755102,
|
1706 |
+
"grad_norm": 134224.234375,
|
1707 |
+
"learning_rate": 8.689413265306124e-06,
|
1708 |
+
"loss": 1.5812,
|
1709 |
+
"step": 1195
|
1710 |
+
},
|
1711 |
+
{
|
1712 |
+
"epoch": 3.061224489795918,
|
1713 |
+
"grad_norm": 165535.9375,
|
1714 |
+
"learning_rate": 8.673469387755103e-06,
|
1715 |
+
"loss": 1.4321,
|
1716 |
+
"step": 1200
|
1717 |
+
},
|
1718 |
+
{
|
1719 |
+
"epoch": 3.0739795918367347,
|
1720 |
+
"grad_norm": 256160.484375,
|
1721 |
+
"learning_rate": 8.657525510204082e-06,
|
1722 |
+
"loss": 1.5011,
|
1723 |
+
"step": 1205
|
1724 |
+
},
|
1725 |
+
{
|
1726 |
+
"epoch": 3.086734693877551,
|
1727 |
+
"grad_norm": 175838.65625,
|
1728 |
+
"learning_rate": 8.641581632653062e-06,
|
1729 |
+
"loss": 1.4771,
|
1730 |
+
"step": 1210
|
1731 |
+
},
|
1732 |
+
{
|
1733 |
+
"epoch": 3.0994897959183674,
|
1734 |
+
"grad_norm": 105858.28125,
|
1735 |
+
"learning_rate": 8.625637755102043e-06,
|
1736 |
+
"loss": 1.4688,
|
1737 |
+
"step": 1215
|
1738 |
+
},
|
1739 |
+
{
|
1740 |
+
"epoch": 3.1122448979591835,
|
1741 |
+
"grad_norm": 221998.046875,
|
1742 |
+
"learning_rate": 8.609693877551022e-06,
|
1743 |
+
"loss": 1.5992,
|
1744 |
+
"step": 1220
|
1745 |
+
},
|
1746 |
+
{
|
1747 |
+
"epoch": 3.125,
|
1748 |
+
"grad_norm": 169362.953125,
|
1749 |
+
"learning_rate": 8.59375e-06,
|
1750 |
+
"loss": 1.4513,
|
1751 |
+
"step": 1225
|
1752 |
+
},
|
1753 |
+
{
|
1754 |
+
"epoch": 3.137755102040816,
|
1755 |
+
"grad_norm": 135083.890625,
|
1756 |
+
"learning_rate": 8.577806122448981e-06,
|
1757 |
+
"loss": 1.4633,
|
1758 |
+
"step": 1230
|
1759 |
+
},
|
1760 |
+
{
|
1761 |
+
"epoch": 3.1505102040816326,
|
1762 |
+
"grad_norm": 131832.9375,
|
1763 |
+
"learning_rate": 8.56186224489796e-06,
|
1764 |
+
"loss": 1.4095,
|
1765 |
+
"step": 1235
|
1766 |
+
},
|
1767 |
+
{
|
1768 |
+
"epoch": 3.163265306122449,
|
1769 |
+
"grad_norm": 178172.671875,
|
1770 |
+
"learning_rate": 8.545918367346939e-06,
|
1771 |
+
"loss": 1.5183,
|
1772 |
+
"step": 1240
|
1773 |
+
},
|
1774 |
+
{
|
1775 |
+
"epoch": 3.1760204081632653,
|
1776 |
+
"grad_norm": 190825.671875,
|
1777 |
+
"learning_rate": 8.52997448979592e-06,
|
1778 |
+
"loss": 1.5573,
|
1779 |
+
"step": 1245
|
1780 |
+
},
|
1781 |
+
{
|
1782 |
+
"epoch": 3.188775510204082,
|
1783 |
+
"grad_norm": 147478.296875,
|
1784 |
+
"learning_rate": 8.514030612244898e-06,
|
1785 |
+
"loss": 1.5018,
|
1786 |
+
"step": 1250
|
1787 |
+
},
|
1788 |
+
{
|
1789 |
+
"epoch": 3.201530612244898,
|
1790 |
+
"grad_norm": 320122.65625,
|
1791 |
+
"learning_rate": 8.498086734693877e-06,
|
1792 |
+
"loss": 1.4697,
|
1793 |
+
"step": 1255
|
1794 |
+
},
|
1795 |
+
{
|
1796 |
+
"epoch": 3.2142857142857144,
|
1797 |
+
"grad_norm": 118118.5234375,
|
1798 |
+
"learning_rate": 8.482142857142858e-06,
|
1799 |
+
"loss": 1.5815,
|
1800 |
+
"step": 1260
|
1801 |
+
},
|
1802 |
+
{
|
1803 |
+
"epoch": 3.2270408163265305,
|
1804 |
+
"grad_norm": 133998.90625,
|
1805 |
+
"learning_rate": 8.466198979591838e-06,
|
1806 |
+
"loss": 1.5255,
|
1807 |
+
"step": 1265
|
1808 |
+
},
|
1809 |
+
{
|
1810 |
+
"epoch": 3.239795918367347,
|
1811 |
+
"grad_norm": 162490.890625,
|
1812 |
+
"learning_rate": 8.450255102040817e-06,
|
1813 |
+
"loss": 1.4765,
|
1814 |
+
"step": 1270
|
1815 |
+
},
|
1816 |
+
{
|
1817 |
+
"epoch": 3.252551020408163,
|
1818 |
+
"grad_norm": 232293.328125,
|
1819 |
+
"learning_rate": 8.434311224489796e-06,
|
1820 |
+
"loss": 1.5323,
|
1821 |
+
"step": 1275
|
1822 |
+
},
|
1823 |
+
{
|
1824 |
+
"epoch": 3.2653061224489797,
|
1825 |
+
"grad_norm": 291055.28125,
|
1826 |
+
"learning_rate": 8.418367346938776e-06,
|
1827 |
+
"loss": 1.4856,
|
1828 |
+
"step": 1280
|
1829 |
+
},
|
1830 |
+
{
|
1831 |
+
"epoch": 3.2780612244897958,
|
1832 |
+
"grad_norm": 228705.5,
|
1833 |
+
"learning_rate": 8.402423469387755e-06,
|
1834 |
+
"loss": 1.3867,
|
1835 |
+
"step": 1285
|
1836 |
+
},
|
1837 |
+
{
|
1838 |
+
"epoch": 3.2908163265306123,
|
1839 |
+
"grad_norm": 154756.125,
|
1840 |
+
"learning_rate": 8.386479591836736e-06,
|
1841 |
+
"loss": 1.4489,
|
1842 |
+
"step": 1290
|
1843 |
+
},
|
1844 |
+
{
|
1845 |
+
"epoch": 3.3035714285714284,
|
1846 |
+
"grad_norm": 151405.671875,
|
1847 |
+
"learning_rate": 8.370535714285715e-06,
|
1848 |
+
"loss": 1.5342,
|
1849 |
+
"step": 1295
|
1850 |
+
},
|
1851 |
+
{
|
1852 |
+
"epoch": 3.316326530612245,
|
1853 |
+
"grad_norm": 136867.859375,
|
1854 |
+
"learning_rate": 8.354591836734695e-06,
|
1855 |
+
"loss": 1.4447,
|
1856 |
+
"step": 1300
|
1857 |
+
},
|
1858 |
+
{
|
1859 |
+
"epoch": 3.329081632653061,
|
1860 |
+
"grad_norm": 218093.40625,
|
1861 |
+
"learning_rate": 8.338647959183674e-06,
|
1862 |
+
"loss": 1.4035,
|
1863 |
+
"step": 1305
|
1864 |
+
},
|
1865 |
+
{
|
1866 |
+
"epoch": 3.3418367346938775,
|
1867 |
+
"grad_norm": 159188.140625,
|
1868 |
+
"learning_rate": 8.322704081632653e-06,
|
1869 |
+
"loss": 1.4324,
|
1870 |
+
"step": 1310
|
1871 |
+
},
|
1872 |
+
{
|
1873 |
+
"epoch": 3.354591836734694,
|
1874 |
+
"grad_norm": 247810.203125,
|
1875 |
+
"learning_rate": 8.306760204081634e-06,
|
1876 |
+
"loss": 1.478,
|
1877 |
+
"step": 1315
|
1878 |
+
},
|
1879 |
+
{
|
1880 |
+
"epoch": 3.36734693877551,
|
1881 |
+
"grad_norm": 121079.1015625,
|
1882 |
+
"learning_rate": 8.290816326530612e-06,
|
1883 |
+
"loss": 1.436,
|
1884 |
+
"step": 1320
|
1885 |
+
},
|
1886 |
+
{
|
1887 |
+
"epoch": 3.3801020408163267,
|
1888 |
+
"grad_norm": 191956.90625,
|
1889 |
+
"learning_rate": 8.274872448979593e-06,
|
1890 |
+
"loss": 1.3561,
|
1891 |
+
"step": 1325
|
1892 |
+
},
|
1893 |
+
{
|
1894 |
+
"epoch": 3.392857142857143,
|
1895 |
+
"grad_norm": 145911.25,
|
1896 |
+
"learning_rate": 8.258928571428572e-06,
|
1897 |
+
"loss": 1.5296,
|
1898 |
+
"step": 1330
|
1899 |
+
},
|
1900 |
+
{
|
1901 |
+
"epoch": 3.4056122448979593,
|
1902 |
+
"grad_norm": 236938.0625,
|
1903 |
+
"learning_rate": 8.242984693877552e-06,
|
1904 |
+
"loss": 1.3109,
|
1905 |
+
"step": 1335
|
1906 |
+
},
|
1907 |
+
{
|
1908 |
+
"epoch": 3.4183673469387754,
|
1909 |
+
"grad_norm": 91069.84375,
|
1910 |
+
"learning_rate": 8.227040816326531e-06,
|
1911 |
+
"loss": 1.4443,
|
1912 |
+
"step": 1340
|
1913 |
+
},
|
1914 |
+
{
|
1915 |
+
"epoch": 3.431122448979592,
|
1916 |
+
"grad_norm": 197523.359375,
|
1917 |
+
"learning_rate": 8.211096938775512e-06,
|
1918 |
+
"loss": 1.467,
|
1919 |
+
"step": 1345
|
1920 |
+
},
|
1921 |
+
{
|
1922 |
+
"epoch": 3.443877551020408,
|
1923 |
+
"grad_norm": 154592.609375,
|
1924 |
+
"learning_rate": 8.19515306122449e-06,
|
1925 |
+
"loss": 1.4467,
|
1926 |
+
"step": 1350
|
1927 |
+
},
|
1928 |
+
{
|
1929 |
+
"epoch": 3.4566326530612246,
|
1930 |
+
"grad_norm": 180109.609375,
|
1931 |
+
"learning_rate": 8.17920918367347e-06,
|
1932 |
+
"loss": 1.4724,
|
1933 |
+
"step": 1355
|
1934 |
+
},
|
1935 |
+
{
|
1936 |
+
"epoch": 3.4693877551020407,
|
1937 |
+
"grad_norm": 206644.03125,
|
1938 |
+
"learning_rate": 8.16326530612245e-06,
|
1939 |
+
"loss": 1.3233,
|
1940 |
+
"step": 1360
|
1941 |
+
},
|
1942 |
+
{
|
1943 |
+
"epoch": 3.482142857142857,
|
1944 |
+
"grad_norm": 128785.90625,
|
1945 |
+
"learning_rate": 8.147321428571429e-06,
|
1946 |
+
"loss": 1.4069,
|
1947 |
+
"step": 1365
|
1948 |
+
},
|
1949 |
+
{
|
1950 |
+
"epoch": 3.4948979591836733,
|
1951 |
+
"grad_norm": 227628.8125,
|
1952 |
+
"learning_rate": 8.131377551020408e-06,
|
1953 |
+
"loss": 1.4277,
|
1954 |
+
"step": 1370
|
1955 |
+
},
|
1956 |
+
{
|
1957 |
+
"epoch": 3.50765306122449,
|
1958 |
+
"grad_norm": 146029.109375,
|
1959 |
+
"learning_rate": 8.115433673469388e-06,
|
1960 |
+
"loss": 1.4764,
|
1961 |
+
"step": 1375
|
1962 |
+
},
|
1963 |
+
{
|
1964 |
+
"epoch": 3.520408163265306,
|
1965 |
+
"grad_norm": 99046.8984375,
|
1966 |
+
"learning_rate": 8.099489795918369e-06,
|
1967 |
+
"loss": 1.3589,
|
1968 |
+
"step": 1380
|
1969 |
+
},
|
1970 |
+
{
|
1971 |
+
"epoch": 3.5331632653061225,
|
1972 |
+
"grad_norm": 187931.296875,
|
1973 |
+
"learning_rate": 8.083545918367348e-06,
|
1974 |
+
"loss": 1.5433,
|
1975 |
+
"step": 1385
|
1976 |
+
},
|
1977 |
+
{
|
1978 |
+
"epoch": 3.545918367346939,
|
1979 |
+
"grad_norm": 297121.15625,
|
1980 |
+
"learning_rate": 8.067602040816327e-06,
|
1981 |
+
"loss": 1.3575,
|
1982 |
+
"step": 1390
|
1983 |
+
},
|
1984 |
+
{
|
1985 |
+
"epoch": 3.558673469387755,
|
1986 |
+
"grad_norm": 221463.234375,
|
1987 |
+
"learning_rate": 8.051658163265307e-06,
|
1988 |
+
"loss": 1.4635,
|
1989 |
+
"step": 1395
|
1990 |
+
},
|
1991 |
+
{
|
1992 |
+
"epoch": 3.571428571428571,
|
1993 |
+
"grad_norm": 130613.59375,
|
1994 |
+
"learning_rate": 8.035714285714286e-06,
|
1995 |
+
"loss": 1.4821,
|
1996 |
+
"step": 1400
|
1997 |
+
},
|
1998 |
+
{
|
1999 |
+
"epoch": 3.5841836734693877,
|
2000 |
+
"grad_norm": 243868.609375,
|
2001 |
+
"learning_rate": 8.019770408163265e-06,
|
2002 |
+
"loss": 1.4667,
|
2003 |
+
"step": 1405
|
2004 |
+
},
|
2005 |
+
{
|
2006 |
+
"epoch": 3.5969387755102042,
|
2007 |
+
"grad_norm": 185059.859375,
|
2008 |
+
"learning_rate": 8.003826530612245e-06,
|
2009 |
+
"loss": 1.4263,
|
2010 |
+
"step": 1410
|
2011 |
+
},
|
2012 |
+
{
|
2013 |
+
"epoch": 3.6096938775510203,
|
2014 |
+
"grad_norm": 124080.9609375,
|
2015 |
+
"learning_rate": 7.987882653061226e-06,
|
2016 |
+
"loss": 1.3818,
|
2017 |
+
"step": 1415
|
2018 |
+
},
|
2019 |
+
{
|
2020 |
+
"epoch": 3.622448979591837,
|
2021 |
+
"grad_norm": 130482.6484375,
|
2022 |
+
"learning_rate": 7.971938775510205e-06,
|
2023 |
+
"loss": 1.4006,
|
2024 |
+
"step": 1420
|
2025 |
+
},
|
2026 |
+
{
|
2027 |
+
"epoch": 3.635204081632653,
|
2028 |
+
"grad_norm": 150170.875,
|
2029 |
+
"learning_rate": 7.955994897959184e-06,
|
2030 |
+
"loss": 1.3012,
|
2031 |
+
"step": 1425
|
2032 |
+
},
|
2033 |
+
{
|
2034 |
+
"epoch": 3.6479591836734695,
|
2035 |
+
"grad_norm": 88223.3984375,
|
2036 |
+
"learning_rate": 7.940051020408164e-06,
|
2037 |
+
"loss": 1.5183,
|
2038 |
+
"step": 1430
|
2039 |
+
},
|
2040 |
+
{
|
2041 |
+
"epoch": 3.6607142857142856,
|
2042 |
+
"grad_norm": 128506.078125,
|
2043 |
+
"learning_rate": 7.924107142857143e-06,
|
2044 |
+
"loss": 1.3076,
|
2045 |
+
"step": 1435
|
2046 |
+
},
|
2047 |
+
{
|
2048 |
+
"epoch": 3.673469387755102,
|
2049 |
+
"grad_norm": 152920.75,
|
2050 |
+
"learning_rate": 7.908163265306124e-06,
|
2051 |
+
"loss": 1.3972,
|
2052 |
+
"step": 1440
|
2053 |
+
},
|
2054 |
+
{
|
2055 |
+
"epoch": 3.686224489795918,
|
2056 |
+
"grad_norm": 167115.84375,
|
2057 |
+
"learning_rate": 7.892219387755102e-06,
|
2058 |
+
"loss": 1.3973,
|
2059 |
+
"step": 1445
|
2060 |
+
},
|
2061 |
+
{
|
2062 |
+
"epoch": 3.6989795918367347,
|
2063 |
+
"grad_norm": 150445.15625,
|
2064 |
+
"learning_rate": 7.876275510204083e-06,
|
2065 |
+
"loss": 1.5096,
|
2066 |
+
"step": 1450
|
2067 |
+
},
|
2068 |
+
{
|
2069 |
+
"epoch": 3.711734693877551,
|
2070 |
+
"grad_norm": 125257.3359375,
|
2071 |
+
"learning_rate": 7.860331632653062e-06,
|
2072 |
+
"loss": 1.4464,
|
2073 |
+
"step": 1455
|
2074 |
+
},
|
2075 |
+
{
|
2076 |
+
"epoch": 3.7244897959183674,
|
2077 |
+
"grad_norm": 174646.5625,
|
2078 |
+
"learning_rate": 7.844387755102042e-06,
|
2079 |
+
"loss": 1.4094,
|
2080 |
+
"step": 1460
|
2081 |
+
},
|
2082 |
+
{
|
2083 |
+
"epoch": 3.737244897959184,
|
2084 |
+
"grad_norm": 128753.625,
|
2085 |
+
"learning_rate": 7.828443877551021e-06,
|
2086 |
+
"loss": 1.3139,
|
2087 |
+
"step": 1465
|
2088 |
+
},
|
2089 |
+
{
|
2090 |
+
"epoch": 3.75,
|
2091 |
+
"grad_norm": 76255.046875,
|
2092 |
+
"learning_rate": 7.8125e-06,
|
2093 |
+
"loss": 1.3626,
|
2094 |
+
"step": 1470
|
2095 |
+
},
|
2096 |
+
{
|
2097 |
+
"epoch": 3.762755102040816,
|
2098 |
+
"grad_norm": 239767.5,
|
2099 |
+
"learning_rate": 7.79655612244898e-06,
|
2100 |
+
"loss": 1.278,
|
2101 |
+
"step": 1475
|
2102 |
+
},
|
2103 |
+
{
|
2104 |
+
"epoch": 3.7755102040816326,
|
2105 |
+
"grad_norm": 192308.640625,
|
2106 |
+
"learning_rate": 7.78061224489796e-06,
|
2107 |
+
"loss": 1.5477,
|
2108 |
+
"step": 1480
|
2109 |
+
},
|
2110 |
+
{
|
2111 |
+
"epoch": 3.788265306122449,
|
2112 |
+
"grad_norm": 249644.578125,
|
2113 |
+
"learning_rate": 7.764668367346938e-06,
|
2114 |
+
"loss": 1.2447,
|
2115 |
+
"step": 1485
|
2116 |
+
},
|
2117 |
+
{
|
2118 |
+
"epoch": 3.8010204081632653,
|
2119 |
+
"grad_norm": 94861.9765625,
|
2120 |
+
"learning_rate": 7.748724489795919e-06,
|
2121 |
+
"loss": 1.3764,
|
2122 |
+
"step": 1490
|
2123 |
+
},
|
2124 |
+
{
|
2125 |
+
"epoch": 3.813775510204082,
|
2126 |
+
"grad_norm": 223123.515625,
|
2127 |
+
"learning_rate": 7.7327806122449e-06,
|
2128 |
+
"loss": 1.512,
|
2129 |
+
"step": 1495
|
2130 |
+
},
|
2131 |
+
{
|
2132 |
+
"epoch": 3.826530612244898,
|
2133 |
+
"grad_norm": 140383.140625,
|
2134 |
+
"learning_rate": 7.716836734693878e-06,
|
2135 |
+
"loss": 1.4049,
|
2136 |
+
"step": 1500
|
2137 |
+
},
|
2138 |
+
{
|
2139 |
+
"epoch": 3.8392857142857144,
|
2140 |
+
"grad_norm": 145005.65625,
|
2141 |
+
"learning_rate": 7.700892857142857e-06,
|
2142 |
+
"loss": 1.1713,
|
2143 |
+
"step": 1505
|
2144 |
+
},
|
2145 |
+
{
|
2146 |
+
"epoch": 3.8520408163265305,
|
2147 |
+
"grad_norm": 193081.1875,
|
2148 |
+
"learning_rate": 7.684948979591838e-06,
|
2149 |
+
"loss": 1.2455,
|
2150 |
+
"step": 1510
|
2151 |
+
},
|
2152 |
+
{
|
2153 |
+
"epoch": 3.864795918367347,
|
2154 |
+
"grad_norm": 123069.03125,
|
2155 |
+
"learning_rate": 7.669005102040817e-06,
|
2156 |
+
"loss": 1.3668,
|
2157 |
+
"step": 1515
|
2158 |
+
},
|
2159 |
+
{
|
2160 |
+
"epoch": 3.877551020408163,
|
2161 |
+
"grad_norm": 105425.453125,
|
2162 |
+
"learning_rate": 7.653061224489796e-06,
|
2163 |
+
"loss": 1.2862,
|
2164 |
+
"step": 1520
|
2165 |
+
},
|
2166 |
+
{
|
2167 |
+
"epoch": 3.8903061224489797,
|
2168 |
+
"grad_norm": 200306.828125,
|
2169 |
+
"learning_rate": 7.637117346938776e-06,
|
2170 |
+
"loss": 1.4318,
|
2171 |
+
"step": 1525
|
2172 |
+
},
|
2173 |
+
{
|
2174 |
+
"epoch": 3.9030612244897958,
|
2175 |
+
"grad_norm": 162285.671875,
|
2176 |
+
"learning_rate": 7.621173469387756e-06,
|
2177 |
+
"loss": 1.2454,
|
2178 |
+
"step": 1530
|
2179 |
+
},
|
2180 |
+
{
|
2181 |
+
"epoch": 3.9158163265306123,
|
2182 |
+
"grad_norm": 197908.765625,
|
2183 |
+
"learning_rate": 7.6052295918367355e-06,
|
2184 |
+
"loss": 1.2989,
|
2185 |
+
"step": 1535
|
2186 |
+
},
|
2187 |
+
{
|
2188 |
+
"epoch": 3.928571428571429,
|
2189 |
+
"grad_norm": 201493.890625,
|
2190 |
+
"learning_rate": 7.589285714285714e-06,
|
2191 |
+
"loss": 1.3827,
|
2192 |
+
"step": 1540
|
2193 |
+
},
|
2194 |
+
{
|
2195 |
+
"epoch": 3.941326530612245,
|
2196 |
+
"grad_norm": 219259.453125,
|
2197 |
+
"learning_rate": 7.573341836734694e-06,
|
2198 |
+
"loss": 1.335,
|
2199 |
+
"step": 1545
|
2200 |
+
},
|
2201 |
+
{
|
2202 |
+
"epoch": 3.954081632653061,
|
2203 |
+
"grad_norm": 198004.703125,
|
2204 |
+
"learning_rate": 7.557397959183675e-06,
|
2205 |
+
"loss": 1.445,
|
2206 |
+
"step": 1550
|
2207 |
+
},
|
2208 |
+
{
|
2209 |
+
"epoch": 3.9668367346938775,
|
2210 |
+
"grad_norm": 135796.671875,
|
2211 |
+
"learning_rate": 7.5414540816326535e-06,
|
2212 |
+
"loss": 1.3096,
|
2213 |
+
"step": 1555
|
2214 |
+
},
|
2215 |
+
{
|
2216 |
+
"epoch": 3.979591836734694,
|
2217 |
+
"grad_norm": 201893.0625,
|
2218 |
+
"learning_rate": 7.525510204081633e-06,
|
2219 |
+
"loss": 1.3282,
|
2220 |
+
"step": 1560
|
2221 |
+
},
|
2222 |
+
{
|
2223 |
+
"epoch": 3.99234693877551,
|
2224 |
+
"grad_norm": 97372.8671875,
|
2225 |
+
"learning_rate": 7.509566326530613e-06,
|
2226 |
+
"loss": 1.344,
|
2227 |
+
"step": 1565
|
2228 |
+
},
|
2229 |
+
{
|
2230 |
+
"epoch": 4.0,
|
2231 |
+
"eval_accuracy": 0.4682603455202893,
|
2232 |
+
"eval_loss": 1.5575779676437378,
|
2233 |
+
"eval_runtime": 304.4215,
|
2234 |
+
"eval_samples_per_second": 16.352,
|
2235 |
+
"eval_steps_per_second": 0.545,
|
2236 |
+
"step": 1568
|
2237 |
+
}
|
2238 |
+
],
|
2239 |
+
"logging_steps": 5,
|
2240 |
+
"max_steps": 3920,
|
2241 |
+
"num_input_tokens_seen": 0,
|
2242 |
+
"num_train_epochs": 10,
|
2243 |
+
"save_steps": 500,
|
2244 |
+
"stateful_callbacks": {
|
2245 |
+
"TrainerControl": {
|
2246 |
+
"args": {
|
2247 |
+
"should_epoch_stop": false,
|
2248 |
+
"should_evaluate": false,
|
2249 |
+
"should_log": false,
|
2250 |
+
"should_save": true,
|
2251 |
+
"should_training_stop": false
|
2252 |
+
},
|
2253 |
+
"attributes": {}
|
2254 |
+
}
|
2255 |
+
},
|
2256 |
+
"total_flos": 5.98716277346304e+18,
|
2257 |
+
"train_batch_size": 30,
|
2258 |
+
"trial_name": null,
|
2259 |
+
"trial_params": null
|
2260 |
+
}
|
checkpoint-1568/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:62d4afa24b2e5344ef9e2ace3afc7237855a36ba40c6701cd39cbf7db81b4cdc
|
3 |
+
size 5176
|