Training in progress, epoch 3, checkpoint
Browse files- checkpoint-1470/config.json +71 -0
- checkpoint-1470/model.safetensors +3 -0
- checkpoint-1470/optimizer.pt +3 -0
- checkpoint-1470/preprocessor_config.json +14 -0
- checkpoint-1470/rng_state.pth +3 -0
- checkpoint-1470/scheduler.pt +3 -0
- checkpoint-1470/trainer_state.json +2118 -0
- checkpoint-1470/training_args.bin +3 -0
checkpoint-1470/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-1470/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9b272c51a9f9a792ad6657e61300dc37a029dd0c1258b712ab47702c940d909b
|
3 |
+
size 358419776
|
checkpoint-1470/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cfbcaff6fcd6cf0f97245422927cf37855c0145863be3fd6317b413f1b43f912
|
3 |
+
size 707744314
|
checkpoint-1470/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-1470/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e63fabf8c6dfc894df2a38a4ee2b78b4d0e7da0bfa170def35bbea93d577d18d
|
3 |
+
size 14244
|
checkpoint-1470/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c25be42a9c5e0038995bf37a5b878f85a50342967123ef552d357c4c3f785025
|
3 |
+
size 1064
|
checkpoint-1470/trainer_state.json
ADDED
@@ -0,0 +1,2118 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.8829644533869886,
|
3 |
+
"best_model_checkpoint": "PhoWhisper-small-vispeech-classifier-v3/checkpoint-980",
|
4 |
+
"epoch": 3.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 1470,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.01020408163265306,
|
13 |
+
"grad_norm": 58959.5,
|
14 |
+
"learning_rate": 1.7006802721088437e-08,
|
15 |
+
"loss": 2.0854,
|
16 |
+
"step": 5
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.02040816326530612,
|
20 |
+
"grad_norm": 54177.4375,
|
21 |
+
"learning_rate": 3.4013605442176873e-08,
|
22 |
+
"loss": 2.0829,
|
23 |
+
"step": 10
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.030612244897959183,
|
27 |
+
"grad_norm": 47591.8828125,
|
28 |
+
"learning_rate": 5.1020408163265303e-08,
|
29 |
+
"loss": 2.0833,
|
30 |
+
"step": 15
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.04081632653061224,
|
34 |
+
"grad_norm": 54382.19921875,
|
35 |
+
"learning_rate": 6.802721088435375e-08,
|
36 |
+
"loss": 2.0864,
|
37 |
+
"step": 20
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.05102040816326531,
|
41 |
+
"grad_norm": 66346.9609375,
|
42 |
+
"learning_rate": 8.50340136054422e-08,
|
43 |
+
"loss": 2.0852,
|
44 |
+
"step": 25
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.061224489795918366,
|
48 |
+
"grad_norm": 57170.58203125,
|
49 |
+
"learning_rate": 1.0204081632653061e-07,
|
50 |
+
"loss": 2.0855,
|
51 |
+
"step": 30
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.07142857142857142,
|
55 |
+
"grad_norm": 54366.90625,
|
56 |
+
"learning_rate": 1.1904761904761906e-07,
|
57 |
+
"loss": 2.0841,
|
58 |
+
"step": 35
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.08163265306122448,
|
62 |
+
"grad_norm": 62029.41796875,
|
63 |
+
"learning_rate": 1.360544217687075e-07,
|
64 |
+
"loss": 2.082,
|
65 |
+
"step": 40
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.09183673469387756,
|
69 |
+
"grad_norm": 58306.72265625,
|
70 |
+
"learning_rate": 1.5306122448979592e-07,
|
71 |
+
"loss": 2.0836,
|
72 |
+
"step": 45
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.10204081632653061,
|
76 |
+
"grad_norm": 53709.69921875,
|
77 |
+
"learning_rate": 1.700680272108844e-07,
|
78 |
+
"loss": 2.0841,
|
79 |
+
"step": 50
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.11224489795918367,
|
83 |
+
"grad_norm": 74088.328125,
|
84 |
+
"learning_rate": 1.8707482993197282e-07,
|
85 |
+
"loss": 2.0825,
|
86 |
+
"step": 55
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.12244897959183673,
|
90 |
+
"grad_norm": 57089.6796875,
|
91 |
+
"learning_rate": 2.0408163265306121e-07,
|
92 |
+
"loss": 2.0799,
|
93 |
+
"step": 60
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.1326530612244898,
|
97 |
+
"grad_norm": 60232.51953125,
|
98 |
+
"learning_rate": 2.210884353741497e-07,
|
99 |
+
"loss": 2.0791,
|
100 |
+
"step": 65
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.14285714285714285,
|
104 |
+
"grad_norm": 56240.4609375,
|
105 |
+
"learning_rate": 2.3809523809523811e-07,
|
106 |
+
"loss": 2.0828,
|
107 |
+
"step": 70
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.15306122448979592,
|
111 |
+
"grad_norm": 74859.0859375,
|
112 |
+
"learning_rate": 2.5510204081632656e-07,
|
113 |
+
"loss": 2.0817,
|
114 |
+
"step": 75
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.16326530612244897,
|
118 |
+
"grad_norm": 57771.18359375,
|
119 |
+
"learning_rate": 2.72108843537415e-07,
|
120 |
+
"loss": 2.0796,
|
121 |
+
"step": 80
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.17346938775510204,
|
125 |
+
"grad_norm": 64205.13671875,
|
126 |
+
"learning_rate": 2.891156462585034e-07,
|
127 |
+
"loss": 2.0785,
|
128 |
+
"step": 85
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.1836734693877551,
|
132 |
+
"grad_norm": 70020.09375,
|
133 |
+
"learning_rate": 3.0612244897959183e-07,
|
134 |
+
"loss": 2.0784,
|
135 |
+
"step": 90
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.19387755102040816,
|
139 |
+
"grad_norm": 55830.03125,
|
140 |
+
"learning_rate": 3.231292517006803e-07,
|
141 |
+
"loss": 2.0786,
|
142 |
+
"step": 95
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.20408163265306123,
|
146 |
+
"grad_norm": 62736.1328125,
|
147 |
+
"learning_rate": 3.401360544217688e-07,
|
148 |
+
"loss": 2.0761,
|
149 |
+
"step": 100
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.21428571428571427,
|
153 |
+
"grad_norm": 48753.46875,
|
154 |
+
"learning_rate": 3.5714285714285716e-07,
|
155 |
+
"loss": 2.0777,
|
156 |
+
"step": 105
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.22448979591836735,
|
160 |
+
"grad_norm": 70695.0078125,
|
161 |
+
"learning_rate": 3.7414965986394563e-07,
|
162 |
+
"loss": 2.0776,
|
163 |
+
"step": 110
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.23469387755102042,
|
167 |
+
"grad_norm": 55496.5078125,
|
168 |
+
"learning_rate": 3.9115646258503406e-07,
|
169 |
+
"loss": 2.0755,
|
170 |
+
"step": 115
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.24489795918367346,
|
174 |
+
"grad_norm": 82219.4375,
|
175 |
+
"learning_rate": 4.0816326530612243e-07,
|
176 |
+
"loss": 2.0698,
|
177 |
+
"step": 120
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.25510204081632654,
|
181 |
+
"grad_norm": 69695.9375,
|
182 |
+
"learning_rate": 4.251700680272109e-07,
|
183 |
+
"loss": 2.0722,
|
184 |
+
"step": 125
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.2653061224489796,
|
188 |
+
"grad_norm": 73932.9453125,
|
189 |
+
"learning_rate": 4.421768707482994e-07,
|
190 |
+
"loss": 2.0718,
|
191 |
+
"step": 130
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.2755102040816326,
|
195 |
+
"grad_norm": 67854.5859375,
|
196 |
+
"learning_rate": 4.591836734693878e-07,
|
197 |
+
"loss": 2.07,
|
198 |
+
"step": 135
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.2857142857142857,
|
202 |
+
"grad_norm": 53210.015625,
|
203 |
+
"learning_rate": 4.7619047619047623e-07,
|
204 |
+
"loss": 2.0712,
|
205 |
+
"step": 140
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.29591836734693877,
|
209 |
+
"grad_norm": 62421.94140625,
|
210 |
+
"learning_rate": 4.931972789115647e-07,
|
211 |
+
"loss": 2.0736,
|
212 |
+
"step": 145
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.30612244897959184,
|
216 |
+
"grad_norm": 77439.4453125,
|
217 |
+
"learning_rate": 5.102040816326531e-07,
|
218 |
+
"loss": 2.0666,
|
219 |
+
"step": 150
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.3163265306122449,
|
223 |
+
"grad_norm": 56657.57421875,
|
224 |
+
"learning_rate": 5.272108843537416e-07,
|
225 |
+
"loss": 2.0646,
|
226 |
+
"step": 155
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.32653061224489793,
|
230 |
+
"grad_norm": 74708.640625,
|
231 |
+
"learning_rate": 5.4421768707483e-07,
|
232 |
+
"loss": 2.0625,
|
233 |
+
"step": 160
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.336734693877551,
|
237 |
+
"grad_norm": 48720.30078125,
|
238 |
+
"learning_rate": 5.612244897959184e-07,
|
239 |
+
"loss": 2.0658,
|
240 |
+
"step": 165
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.3469387755102041,
|
244 |
+
"grad_norm": 51835.54296875,
|
245 |
+
"learning_rate": 5.782312925170068e-07,
|
246 |
+
"loss": 2.0615,
|
247 |
+
"step": 170
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.35714285714285715,
|
251 |
+
"grad_norm": 45173.23046875,
|
252 |
+
"learning_rate": 5.952380952380953e-07,
|
253 |
+
"loss": 2.0602,
|
254 |
+
"step": 175
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.3673469387755102,
|
258 |
+
"grad_norm": 57442.7265625,
|
259 |
+
"learning_rate": 6.122448979591837e-07,
|
260 |
+
"loss": 2.0592,
|
261 |
+
"step": 180
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.37755102040816324,
|
265 |
+
"grad_norm": 77342.5,
|
266 |
+
"learning_rate": 6.292517006802721e-07,
|
267 |
+
"loss": 2.0581,
|
268 |
+
"step": 185
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 0.3877551020408163,
|
272 |
+
"grad_norm": 75636.84375,
|
273 |
+
"learning_rate": 6.462585034013606e-07,
|
274 |
+
"loss": 2.0556,
|
275 |
+
"step": 190
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 0.3979591836734694,
|
279 |
+
"grad_norm": 58702.5703125,
|
280 |
+
"learning_rate": 6.632653061224491e-07,
|
281 |
+
"loss": 2.0579,
|
282 |
+
"step": 195
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.40816326530612246,
|
286 |
+
"grad_norm": 66054.546875,
|
287 |
+
"learning_rate": 6.802721088435376e-07,
|
288 |
+
"loss": 2.0559,
|
289 |
+
"step": 200
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.41836734693877553,
|
293 |
+
"grad_norm": 62451.17578125,
|
294 |
+
"learning_rate": 6.972789115646258e-07,
|
295 |
+
"loss": 2.0471,
|
296 |
+
"step": 205
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 0.42857142857142855,
|
300 |
+
"grad_norm": 53168.203125,
|
301 |
+
"learning_rate": 7.142857142857143e-07,
|
302 |
+
"loss": 2.0459,
|
303 |
+
"step": 210
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.4387755102040816,
|
307 |
+
"grad_norm": 67945.484375,
|
308 |
+
"learning_rate": 7.312925170068028e-07,
|
309 |
+
"loss": 2.0444,
|
310 |
+
"step": 215
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 0.4489795918367347,
|
314 |
+
"grad_norm": 51722.9140625,
|
315 |
+
"learning_rate": 7.482993197278913e-07,
|
316 |
+
"loss": 2.0402,
|
317 |
+
"step": 220
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 0.45918367346938777,
|
321 |
+
"grad_norm": 74699.6796875,
|
322 |
+
"learning_rate": 7.653061224489796e-07,
|
323 |
+
"loss": 2.0316,
|
324 |
+
"step": 225
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 0.46938775510204084,
|
328 |
+
"grad_norm": 41176.36328125,
|
329 |
+
"learning_rate": 7.823129251700681e-07,
|
330 |
+
"loss": 2.0442,
|
331 |
+
"step": 230
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 0.47959183673469385,
|
335 |
+
"grad_norm": 56116.875,
|
336 |
+
"learning_rate": 7.993197278911566e-07,
|
337 |
+
"loss": 2.0403,
|
338 |
+
"step": 235
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 0.4897959183673469,
|
342 |
+
"grad_norm": 69240.3359375,
|
343 |
+
"learning_rate": 8.163265306122449e-07,
|
344 |
+
"loss": 2.037,
|
345 |
+
"step": 240
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 0.5,
|
349 |
+
"grad_norm": 67816.90625,
|
350 |
+
"learning_rate": 8.333333333333333e-07,
|
351 |
+
"loss": 2.0273,
|
352 |
+
"step": 245
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 0.5102040816326531,
|
356 |
+
"grad_norm": 64964.8515625,
|
357 |
+
"learning_rate": 8.503401360544218e-07,
|
358 |
+
"loss": 2.0358,
|
359 |
+
"step": 250
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 0.5204081632653061,
|
363 |
+
"grad_norm": 75715.4765625,
|
364 |
+
"learning_rate": 8.673469387755103e-07,
|
365 |
+
"loss": 2.0108,
|
366 |
+
"step": 255
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 0.5306122448979592,
|
370 |
+
"grad_norm": 66728.0390625,
|
371 |
+
"learning_rate": 8.843537414965988e-07,
|
372 |
+
"loss": 2.0219,
|
373 |
+
"step": 260
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 0.5408163265306123,
|
377 |
+
"grad_norm": 75688.15625,
|
378 |
+
"learning_rate": 9.013605442176871e-07,
|
379 |
+
"loss": 2.0165,
|
380 |
+
"step": 265
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 0.5510204081632653,
|
384 |
+
"grad_norm": 74948.90625,
|
385 |
+
"learning_rate": 9.183673469387756e-07,
|
386 |
+
"loss": 2.0228,
|
387 |
+
"step": 270
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 0.5612244897959183,
|
391 |
+
"grad_norm": 58538.5,
|
392 |
+
"learning_rate": 9.353741496598641e-07,
|
393 |
+
"loss": 2.0197,
|
394 |
+
"step": 275
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 0.5714285714285714,
|
398 |
+
"grad_norm": 58524.6796875,
|
399 |
+
"learning_rate": 9.523809523809525e-07,
|
400 |
+
"loss": 1.9988,
|
401 |
+
"step": 280
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 0.5816326530612245,
|
405 |
+
"grad_norm": 66490.1171875,
|
406 |
+
"learning_rate": 9.69387755102041e-07,
|
407 |
+
"loss": 1.9995,
|
408 |
+
"step": 285
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 0.5918367346938775,
|
412 |
+
"grad_norm": 76682.40625,
|
413 |
+
"learning_rate": 9.863945578231293e-07,
|
414 |
+
"loss": 2.0109,
|
415 |
+
"step": 290
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 0.6020408163265306,
|
419 |
+
"grad_norm": 76082.6953125,
|
420 |
+
"learning_rate": 1.0034013605442177e-06,
|
421 |
+
"loss": 1.9987,
|
422 |
+
"step": 295
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 0.6122448979591837,
|
426 |
+
"grad_norm": 75684.1796875,
|
427 |
+
"learning_rate": 1.0204081632653063e-06,
|
428 |
+
"loss": 1.9903,
|
429 |
+
"step": 300
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 0.6224489795918368,
|
433 |
+
"grad_norm": 108901.2890625,
|
434 |
+
"learning_rate": 1.0374149659863946e-06,
|
435 |
+
"loss": 1.9677,
|
436 |
+
"step": 305
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 0.6326530612244898,
|
440 |
+
"grad_norm": 85066.4375,
|
441 |
+
"learning_rate": 1.0544217687074832e-06,
|
442 |
+
"loss": 1.9941,
|
443 |
+
"step": 310
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 0.6428571428571429,
|
447 |
+
"grad_norm": 98574.0234375,
|
448 |
+
"learning_rate": 1.0714285714285714e-06,
|
449 |
+
"loss": 1.9778,
|
450 |
+
"step": 315
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 0.6530612244897959,
|
454 |
+
"grad_norm": 79184.9375,
|
455 |
+
"learning_rate": 1.08843537414966e-06,
|
456 |
+
"loss": 1.9723,
|
457 |
+
"step": 320
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 0.6632653061224489,
|
461 |
+
"grad_norm": 62203.5625,
|
462 |
+
"learning_rate": 1.1054421768707483e-06,
|
463 |
+
"loss": 1.9821,
|
464 |
+
"step": 325
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 0.673469387755102,
|
468 |
+
"grad_norm": 69015.2734375,
|
469 |
+
"learning_rate": 1.122448979591837e-06,
|
470 |
+
"loss": 1.9708,
|
471 |
+
"step": 330
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 0.6836734693877551,
|
475 |
+
"grad_norm": 83259.71875,
|
476 |
+
"learning_rate": 1.1394557823129253e-06,
|
477 |
+
"loss": 1.9581,
|
478 |
+
"step": 335
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 0.6938775510204082,
|
482 |
+
"grad_norm": 93754.0546875,
|
483 |
+
"learning_rate": 1.1564625850340136e-06,
|
484 |
+
"loss": 1.9614,
|
485 |
+
"step": 340
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 0.7040816326530612,
|
489 |
+
"grad_norm": 97343.171875,
|
490 |
+
"learning_rate": 1.1734693877551022e-06,
|
491 |
+
"loss": 1.9372,
|
492 |
+
"step": 345
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 0.7142857142857143,
|
496 |
+
"grad_norm": 86450.984375,
|
497 |
+
"learning_rate": 1.1904761904761906e-06,
|
498 |
+
"loss": 1.9635,
|
499 |
+
"step": 350
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 0.7244897959183674,
|
503 |
+
"grad_norm": 109065.9140625,
|
504 |
+
"learning_rate": 1.207482993197279e-06,
|
505 |
+
"loss": 1.9432,
|
506 |
+
"step": 355
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 0.7346938775510204,
|
510 |
+
"grad_norm": 96424.5625,
|
511 |
+
"learning_rate": 1.2244897959183673e-06,
|
512 |
+
"loss": 1.9276,
|
513 |
+
"step": 360
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 0.7448979591836735,
|
517 |
+
"grad_norm": 91580.3671875,
|
518 |
+
"learning_rate": 1.241496598639456e-06,
|
519 |
+
"loss": 1.9075,
|
520 |
+
"step": 365
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 0.7551020408163265,
|
524 |
+
"grad_norm": 74389.96875,
|
525 |
+
"learning_rate": 1.2585034013605443e-06,
|
526 |
+
"loss": 1.9058,
|
527 |
+
"step": 370
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 0.7653061224489796,
|
531 |
+
"grad_norm": 114560.078125,
|
532 |
+
"learning_rate": 1.2755102040816329e-06,
|
533 |
+
"loss": 1.9247,
|
534 |
+
"step": 375
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 0.7755102040816326,
|
538 |
+
"grad_norm": 104689.0078125,
|
539 |
+
"learning_rate": 1.2925170068027212e-06,
|
540 |
+
"loss": 1.9261,
|
541 |
+
"step": 380
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 0.7857142857142857,
|
545 |
+
"grad_norm": 115092.71875,
|
546 |
+
"learning_rate": 1.3095238095238096e-06,
|
547 |
+
"loss": 1.8644,
|
548 |
+
"step": 385
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 0.7959183673469388,
|
552 |
+
"grad_norm": 77653.03125,
|
553 |
+
"learning_rate": 1.3265306122448982e-06,
|
554 |
+
"loss": 1.8945,
|
555 |
+
"step": 390
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 0.8061224489795918,
|
559 |
+
"grad_norm": 97794.46875,
|
560 |
+
"learning_rate": 1.3435374149659866e-06,
|
561 |
+
"loss": 1.8651,
|
562 |
+
"step": 395
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 0.8163265306122449,
|
566 |
+
"grad_norm": 123392.59375,
|
567 |
+
"learning_rate": 1.3605442176870751e-06,
|
568 |
+
"loss": 1.881,
|
569 |
+
"step": 400
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 0.826530612244898,
|
573 |
+
"grad_norm": 78845.140625,
|
574 |
+
"learning_rate": 1.3775510204081633e-06,
|
575 |
+
"loss": 1.8774,
|
576 |
+
"step": 405
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 0.8367346938775511,
|
580 |
+
"grad_norm": 106911.1328125,
|
581 |
+
"learning_rate": 1.3945578231292517e-06,
|
582 |
+
"loss": 1.8784,
|
583 |
+
"step": 410
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 0.8469387755102041,
|
587 |
+
"grad_norm": 97048.953125,
|
588 |
+
"learning_rate": 1.4115646258503403e-06,
|
589 |
+
"loss": 1.8848,
|
590 |
+
"step": 415
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 0.8571428571428571,
|
594 |
+
"grad_norm": 112101.5859375,
|
595 |
+
"learning_rate": 1.4285714285714286e-06,
|
596 |
+
"loss": 1.8932,
|
597 |
+
"step": 420
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 0.8673469387755102,
|
601 |
+
"grad_norm": 90785.6640625,
|
602 |
+
"learning_rate": 1.445578231292517e-06,
|
603 |
+
"loss": 1.8413,
|
604 |
+
"step": 425
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 0.8775510204081632,
|
608 |
+
"grad_norm": 150235.328125,
|
609 |
+
"learning_rate": 1.4625850340136056e-06,
|
610 |
+
"loss": 1.7915,
|
611 |
+
"step": 430
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 0.8877551020408163,
|
615 |
+
"grad_norm": 92524.75,
|
616 |
+
"learning_rate": 1.479591836734694e-06,
|
617 |
+
"loss": 1.8316,
|
618 |
+
"step": 435
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 0.8979591836734694,
|
622 |
+
"grad_norm": 123792.203125,
|
623 |
+
"learning_rate": 1.4965986394557825e-06,
|
624 |
+
"loss": 1.7894,
|
625 |
+
"step": 440
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 0.9081632653061225,
|
629 |
+
"grad_norm": 105375.3671875,
|
630 |
+
"learning_rate": 1.513605442176871e-06,
|
631 |
+
"loss": 1.7826,
|
632 |
+
"step": 445
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 0.9183673469387755,
|
636 |
+
"grad_norm": 83843.8828125,
|
637 |
+
"learning_rate": 1.5306122448979593e-06,
|
638 |
+
"loss": 1.7905,
|
639 |
+
"step": 450
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 0.9285714285714286,
|
643 |
+
"grad_norm": 142571.53125,
|
644 |
+
"learning_rate": 1.5476190476190479e-06,
|
645 |
+
"loss": 1.8053,
|
646 |
+
"step": 455
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 0.9387755102040817,
|
650 |
+
"grad_norm": 119875.6328125,
|
651 |
+
"learning_rate": 1.5646258503401362e-06,
|
652 |
+
"loss": 1.8039,
|
653 |
+
"step": 460
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 0.9489795918367347,
|
657 |
+
"grad_norm": 112758.0390625,
|
658 |
+
"learning_rate": 1.5816326530612248e-06,
|
659 |
+
"loss": 1.7292,
|
660 |
+
"step": 465
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 0.9591836734693877,
|
664 |
+
"grad_norm": 125497.7890625,
|
665 |
+
"learning_rate": 1.5986394557823132e-06,
|
666 |
+
"loss": 1.7456,
|
667 |
+
"step": 470
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 0.9693877551020408,
|
671 |
+
"grad_norm": 110933.9609375,
|
672 |
+
"learning_rate": 1.6156462585034016e-06,
|
673 |
+
"loss": 1.7572,
|
674 |
+
"step": 475
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 0.9795918367346939,
|
678 |
+
"grad_norm": 149434.625,
|
679 |
+
"learning_rate": 1.6326530612244897e-06,
|
680 |
+
"loss": 1.7261,
|
681 |
+
"step": 480
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 0.9897959183673469,
|
685 |
+
"grad_norm": 149440.328125,
|
686 |
+
"learning_rate": 1.6496598639455783e-06,
|
687 |
+
"loss": 1.75,
|
688 |
+
"step": 485
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 1.0,
|
692 |
+
"grad_norm": 189548.890625,
|
693 |
+
"learning_rate": 1.6666666666666667e-06,
|
694 |
+
"loss": 1.7295,
|
695 |
+
"step": 490
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 1.0,
|
699 |
+
"eval_accuracy": 0.4966465459423206,
|
700 |
+
"eval_loss": 1.6658564805984497,
|
701 |
+
"eval_runtime": 190.4552,
|
702 |
+
"eval_samples_per_second": 15.657,
|
703 |
+
"eval_steps_per_second": 0.656,
|
704 |
+
"step": 490
|
705 |
+
},
|
706 |
+
{
|
707 |
+
"epoch": 1.010204081632653,
|
708 |
+
"grad_norm": 151543.5,
|
709 |
+
"learning_rate": 1.6836734693877552e-06,
|
710 |
+
"loss": 1.7048,
|
711 |
+
"step": 495
|
712 |
+
},
|
713 |
+
{
|
714 |
+
"epoch": 1.0204081632653061,
|
715 |
+
"grad_norm": 140440.59375,
|
716 |
+
"learning_rate": 1.7006802721088436e-06,
|
717 |
+
"loss": 1.6821,
|
718 |
+
"step": 500
|
719 |
+
},
|
720 |
+
{
|
721 |
+
"epoch": 1.030612244897959,
|
722 |
+
"grad_norm": 141126.046875,
|
723 |
+
"learning_rate": 1.717687074829932e-06,
|
724 |
+
"loss": 1.6729,
|
725 |
+
"step": 505
|
726 |
+
},
|
727 |
+
{
|
728 |
+
"epoch": 1.0408163265306123,
|
729 |
+
"grad_norm": 148147.0625,
|
730 |
+
"learning_rate": 1.7346938775510206e-06,
|
731 |
+
"loss": 1.6891,
|
732 |
+
"step": 510
|
733 |
+
},
|
734 |
+
{
|
735 |
+
"epoch": 1.0510204081632653,
|
736 |
+
"grad_norm": 152547.671875,
|
737 |
+
"learning_rate": 1.751700680272109e-06,
|
738 |
+
"loss": 1.6697,
|
739 |
+
"step": 515
|
740 |
+
},
|
741 |
+
{
|
742 |
+
"epoch": 1.0612244897959184,
|
743 |
+
"grad_norm": 172365.578125,
|
744 |
+
"learning_rate": 1.7687074829931975e-06,
|
745 |
+
"loss": 1.6462,
|
746 |
+
"step": 520
|
747 |
+
},
|
748 |
+
{
|
749 |
+
"epoch": 1.0714285714285714,
|
750 |
+
"grad_norm": 165208.671875,
|
751 |
+
"learning_rate": 1.7857142857142859e-06,
|
752 |
+
"loss": 1.6297,
|
753 |
+
"step": 525
|
754 |
+
},
|
755 |
+
{
|
756 |
+
"epoch": 1.0816326530612246,
|
757 |
+
"grad_norm": 111541.8046875,
|
758 |
+
"learning_rate": 1.8027210884353743e-06,
|
759 |
+
"loss": 1.6863,
|
760 |
+
"step": 530
|
761 |
+
},
|
762 |
+
{
|
763 |
+
"epoch": 1.0918367346938775,
|
764 |
+
"grad_norm": 176715.46875,
|
765 |
+
"learning_rate": 1.8197278911564628e-06,
|
766 |
+
"loss": 1.5767,
|
767 |
+
"step": 535
|
768 |
+
},
|
769 |
+
{
|
770 |
+
"epoch": 1.1020408163265305,
|
771 |
+
"grad_norm": 127892.1953125,
|
772 |
+
"learning_rate": 1.8367346938775512e-06,
|
773 |
+
"loss": 1.6373,
|
774 |
+
"step": 540
|
775 |
+
},
|
776 |
+
{
|
777 |
+
"epoch": 1.1122448979591837,
|
778 |
+
"grad_norm": 178854.4375,
|
779 |
+
"learning_rate": 1.8537414965986398e-06,
|
780 |
+
"loss": 1.5393,
|
781 |
+
"step": 545
|
782 |
+
},
|
783 |
+
{
|
784 |
+
"epoch": 1.1224489795918366,
|
785 |
+
"grad_norm": 159664.796875,
|
786 |
+
"learning_rate": 1.8707482993197282e-06,
|
787 |
+
"loss": 1.605,
|
788 |
+
"step": 550
|
789 |
+
},
|
790 |
+
{
|
791 |
+
"epoch": 1.1326530612244898,
|
792 |
+
"grad_norm": 136787.71875,
|
793 |
+
"learning_rate": 1.8877551020408163e-06,
|
794 |
+
"loss": 1.601,
|
795 |
+
"step": 555
|
796 |
+
},
|
797 |
+
{
|
798 |
+
"epoch": 1.1428571428571428,
|
799 |
+
"grad_norm": 149823.046875,
|
800 |
+
"learning_rate": 1.904761904761905e-06,
|
801 |
+
"loss": 1.5489,
|
802 |
+
"step": 560
|
803 |
+
},
|
804 |
+
{
|
805 |
+
"epoch": 1.153061224489796,
|
806 |
+
"grad_norm": 162967.0,
|
807 |
+
"learning_rate": 1.921768707482993e-06,
|
808 |
+
"loss": 1.5755,
|
809 |
+
"step": 565
|
810 |
+
},
|
811 |
+
{
|
812 |
+
"epoch": 1.163265306122449,
|
813 |
+
"grad_norm": 130929.3203125,
|
814 |
+
"learning_rate": 1.938775510204082e-06,
|
815 |
+
"loss": 1.6032,
|
816 |
+
"step": 570
|
817 |
+
},
|
818 |
+
{
|
819 |
+
"epoch": 1.1734693877551021,
|
820 |
+
"grad_norm": 178277.828125,
|
821 |
+
"learning_rate": 1.9557823129251702e-06,
|
822 |
+
"loss": 1.5356,
|
823 |
+
"step": 575
|
824 |
+
},
|
825 |
+
{
|
826 |
+
"epoch": 1.183673469387755,
|
827 |
+
"grad_norm": 147472.265625,
|
828 |
+
"learning_rate": 1.9727891156462586e-06,
|
829 |
+
"loss": 1.6313,
|
830 |
+
"step": 580
|
831 |
+
},
|
832 |
+
{
|
833 |
+
"epoch": 1.193877551020408,
|
834 |
+
"grad_norm": 224884.78125,
|
835 |
+
"learning_rate": 1.989795918367347e-06,
|
836 |
+
"loss": 1.5901,
|
837 |
+
"step": 585
|
838 |
+
},
|
839 |
+
{
|
840 |
+
"epoch": 1.2040816326530612,
|
841 |
+
"grad_norm": 161460.28125,
|
842 |
+
"learning_rate": 2.0068027210884353e-06,
|
843 |
+
"loss": 1.5122,
|
844 |
+
"step": 590
|
845 |
+
},
|
846 |
+
{
|
847 |
+
"epoch": 1.2142857142857142,
|
848 |
+
"grad_norm": 175408.140625,
|
849 |
+
"learning_rate": 2.023809523809524e-06,
|
850 |
+
"loss": 1.5233,
|
851 |
+
"step": 595
|
852 |
+
},
|
853 |
+
{
|
854 |
+
"epoch": 1.2244897959183674,
|
855 |
+
"grad_norm": 201887.96875,
|
856 |
+
"learning_rate": 2.0408163265306125e-06,
|
857 |
+
"loss": 1.5105,
|
858 |
+
"step": 600
|
859 |
+
},
|
860 |
+
{
|
861 |
+
"epoch": 1.2346938775510203,
|
862 |
+
"grad_norm": 161774.4375,
|
863 |
+
"learning_rate": 2.057823129251701e-06,
|
864 |
+
"loss": 1.5056,
|
865 |
+
"step": 605
|
866 |
+
},
|
867 |
+
{
|
868 |
+
"epoch": 1.2448979591836735,
|
869 |
+
"grad_norm": 150156.140625,
|
870 |
+
"learning_rate": 2.0748299319727892e-06,
|
871 |
+
"loss": 1.5062,
|
872 |
+
"step": 610
|
873 |
+
},
|
874 |
+
{
|
875 |
+
"epoch": 1.2551020408163265,
|
876 |
+
"grad_norm": 176416.453125,
|
877 |
+
"learning_rate": 2.0918367346938776e-06,
|
878 |
+
"loss": 1.4338,
|
879 |
+
"step": 615
|
880 |
+
},
|
881 |
+
{
|
882 |
+
"epoch": 1.2653061224489797,
|
883 |
+
"grad_norm": 177887.078125,
|
884 |
+
"learning_rate": 2.1088435374149664e-06,
|
885 |
+
"loss": 1.52,
|
886 |
+
"step": 620
|
887 |
+
},
|
888 |
+
{
|
889 |
+
"epoch": 1.2755102040816326,
|
890 |
+
"grad_norm": 288881.9375,
|
891 |
+
"learning_rate": 2.1258503401360544e-06,
|
892 |
+
"loss": 1.5005,
|
893 |
+
"step": 625
|
894 |
+
},
|
895 |
+
{
|
896 |
+
"epoch": 1.2857142857142856,
|
897 |
+
"grad_norm": 199033.71875,
|
898 |
+
"learning_rate": 2.1428571428571427e-06,
|
899 |
+
"loss": 1.4222,
|
900 |
+
"step": 630
|
901 |
+
},
|
902 |
+
{
|
903 |
+
"epoch": 1.2959183673469388,
|
904 |
+
"grad_norm": 195794.4375,
|
905 |
+
"learning_rate": 2.1598639455782315e-06,
|
906 |
+
"loss": 1.5098,
|
907 |
+
"step": 635
|
908 |
+
},
|
909 |
+
{
|
910 |
+
"epoch": 1.306122448979592,
|
911 |
+
"grad_norm": 157052.59375,
|
912 |
+
"learning_rate": 2.17687074829932e-06,
|
913 |
+
"loss": 1.4446,
|
914 |
+
"step": 640
|
915 |
+
},
|
916 |
+
{
|
917 |
+
"epoch": 1.316326530612245,
|
918 |
+
"grad_norm": 148277.3125,
|
919 |
+
"learning_rate": 2.1938775510204083e-06,
|
920 |
+
"loss": 1.4534,
|
921 |
+
"step": 645
|
922 |
+
},
|
923 |
+
{
|
924 |
+
"epoch": 1.3265306122448979,
|
925 |
+
"grad_norm": 138956.078125,
|
926 |
+
"learning_rate": 2.2108843537414966e-06,
|
927 |
+
"loss": 1.4171,
|
928 |
+
"step": 650
|
929 |
+
},
|
930 |
+
{
|
931 |
+
"epoch": 1.336734693877551,
|
932 |
+
"grad_norm": 182219.75,
|
933 |
+
"learning_rate": 2.227891156462585e-06,
|
934 |
+
"loss": 1.4169,
|
935 |
+
"step": 655
|
936 |
+
},
|
937 |
+
{
|
938 |
+
"epoch": 1.346938775510204,
|
939 |
+
"grad_norm": 196649.171875,
|
940 |
+
"learning_rate": 2.244897959183674e-06,
|
941 |
+
"loss": 1.3707,
|
942 |
+
"step": 660
|
943 |
+
},
|
944 |
+
{
|
945 |
+
"epoch": 1.3571428571428572,
|
946 |
+
"grad_norm": 195908.453125,
|
947 |
+
"learning_rate": 2.261904761904762e-06,
|
948 |
+
"loss": 1.391,
|
949 |
+
"step": 665
|
950 |
+
},
|
951 |
+
{
|
952 |
+
"epoch": 1.3673469387755102,
|
953 |
+
"grad_norm": 245094.96875,
|
954 |
+
"learning_rate": 2.2789115646258505e-06,
|
955 |
+
"loss": 1.3471,
|
956 |
+
"step": 670
|
957 |
+
},
|
958 |
+
{
|
959 |
+
"epoch": 1.3775510204081631,
|
960 |
+
"grad_norm": 173560.40625,
|
961 |
+
"learning_rate": 2.295918367346939e-06,
|
962 |
+
"loss": 1.3596,
|
963 |
+
"step": 675
|
964 |
+
},
|
965 |
+
{
|
966 |
+
"epoch": 1.3877551020408163,
|
967 |
+
"grad_norm": 453190.40625,
|
968 |
+
"learning_rate": 2.3129251700680273e-06,
|
969 |
+
"loss": 1.3595,
|
970 |
+
"step": 680
|
971 |
+
},
|
972 |
+
{
|
973 |
+
"epoch": 1.3979591836734695,
|
974 |
+
"grad_norm": 256905.390625,
|
975 |
+
"learning_rate": 2.329931972789116e-06,
|
976 |
+
"loss": 1.3721,
|
977 |
+
"step": 685
|
978 |
+
},
|
979 |
+
{
|
980 |
+
"epoch": 1.4081632653061225,
|
981 |
+
"grad_norm": 172486.375,
|
982 |
+
"learning_rate": 2.3469387755102044e-06,
|
983 |
+
"loss": 1.3023,
|
984 |
+
"step": 690
|
985 |
+
},
|
986 |
+
{
|
987 |
+
"epoch": 1.4183673469387754,
|
988 |
+
"grad_norm": 399484.28125,
|
989 |
+
"learning_rate": 2.363945578231293e-06,
|
990 |
+
"loss": 1.3021,
|
991 |
+
"step": 695
|
992 |
+
},
|
993 |
+
{
|
994 |
+
"epoch": 1.4285714285714286,
|
995 |
+
"grad_norm": 187642.390625,
|
996 |
+
"learning_rate": 2.380952380952381e-06,
|
997 |
+
"loss": 1.2988,
|
998 |
+
"step": 700
|
999 |
+
},
|
1000 |
+
{
|
1001 |
+
"epoch": 1.4387755102040816,
|
1002 |
+
"grad_norm": 166884.390625,
|
1003 |
+
"learning_rate": 2.3979591836734696e-06,
|
1004 |
+
"loss": 1.3776,
|
1005 |
+
"step": 705
|
1006 |
+
},
|
1007 |
+
{
|
1008 |
+
"epoch": 1.4489795918367347,
|
1009 |
+
"grad_norm": 127044.1484375,
|
1010 |
+
"learning_rate": 2.414965986394558e-06,
|
1011 |
+
"loss": 1.3648,
|
1012 |
+
"step": 710
|
1013 |
+
},
|
1014 |
+
{
|
1015 |
+
"epoch": 1.4591836734693877,
|
1016 |
+
"grad_norm": 139377.5625,
|
1017 |
+
"learning_rate": 2.4319727891156463e-06,
|
1018 |
+
"loss": 1.3505,
|
1019 |
+
"step": 715
|
1020 |
+
},
|
1021 |
+
{
|
1022 |
+
"epoch": 1.469387755102041,
|
1023 |
+
"grad_norm": 286594.53125,
|
1024 |
+
"learning_rate": 2.4489795918367347e-06,
|
1025 |
+
"loss": 1.3059,
|
1026 |
+
"step": 720
|
1027 |
+
},
|
1028 |
+
{
|
1029 |
+
"epoch": 1.4795918367346939,
|
1030 |
+
"grad_norm": 220758.40625,
|
1031 |
+
"learning_rate": 2.4659863945578235e-06,
|
1032 |
+
"loss": 1.2052,
|
1033 |
+
"step": 725
|
1034 |
+
},
|
1035 |
+
{
|
1036 |
+
"epoch": 1.489795918367347,
|
1037 |
+
"grad_norm": 159674.75,
|
1038 |
+
"learning_rate": 2.482993197278912e-06,
|
1039 |
+
"loss": 1.2049,
|
1040 |
+
"step": 730
|
1041 |
+
},
|
1042 |
+
{
|
1043 |
+
"epoch": 1.5,
|
1044 |
+
"grad_norm": 188546.5,
|
1045 |
+
"learning_rate": 2.5e-06,
|
1046 |
+
"loss": 1.2727,
|
1047 |
+
"step": 735
|
1048 |
+
},
|
1049 |
+
{
|
1050 |
+
"epoch": 1.510204081632653,
|
1051 |
+
"grad_norm": 339846.125,
|
1052 |
+
"learning_rate": 2.5170068027210886e-06,
|
1053 |
+
"loss": 1.2233,
|
1054 |
+
"step": 740
|
1055 |
+
},
|
1056 |
+
{
|
1057 |
+
"epoch": 1.5204081632653061,
|
1058 |
+
"grad_norm": 143014.203125,
|
1059 |
+
"learning_rate": 2.534013605442177e-06,
|
1060 |
+
"loss": 1.2485,
|
1061 |
+
"step": 745
|
1062 |
+
},
|
1063 |
+
{
|
1064 |
+
"epoch": 1.5306122448979593,
|
1065 |
+
"grad_norm": 347331.875,
|
1066 |
+
"learning_rate": 2.5510204081632657e-06,
|
1067 |
+
"loss": 1.2635,
|
1068 |
+
"step": 750
|
1069 |
+
},
|
1070 |
+
{
|
1071 |
+
"epoch": 1.5408163265306123,
|
1072 |
+
"grad_norm": 201180.96875,
|
1073 |
+
"learning_rate": 2.568027210884354e-06,
|
1074 |
+
"loss": 1.317,
|
1075 |
+
"step": 755
|
1076 |
+
},
|
1077 |
+
{
|
1078 |
+
"epoch": 1.5510204081632653,
|
1079 |
+
"grad_norm": 264098.125,
|
1080 |
+
"learning_rate": 2.5850340136054425e-06,
|
1081 |
+
"loss": 1.2379,
|
1082 |
+
"step": 760
|
1083 |
+
},
|
1084 |
+
{
|
1085 |
+
"epoch": 1.5612244897959182,
|
1086 |
+
"grad_norm": 198832.515625,
|
1087 |
+
"learning_rate": 2.602040816326531e-06,
|
1088 |
+
"loss": 1.2452,
|
1089 |
+
"step": 765
|
1090 |
+
},
|
1091 |
+
{
|
1092 |
+
"epoch": 1.5714285714285714,
|
1093 |
+
"grad_norm": 393651.0625,
|
1094 |
+
"learning_rate": 2.6190476190476192e-06,
|
1095 |
+
"loss": 1.1995,
|
1096 |
+
"step": 770
|
1097 |
+
},
|
1098 |
+
{
|
1099 |
+
"epoch": 1.5816326530612246,
|
1100 |
+
"grad_norm": 219208.8125,
|
1101 |
+
"learning_rate": 2.636054421768708e-06,
|
1102 |
+
"loss": 1.1063,
|
1103 |
+
"step": 775
|
1104 |
+
},
|
1105 |
+
{
|
1106 |
+
"epoch": 1.5918367346938775,
|
1107 |
+
"grad_norm": 398729.21875,
|
1108 |
+
"learning_rate": 2.6530612244897964e-06,
|
1109 |
+
"loss": 1.2484,
|
1110 |
+
"step": 780
|
1111 |
+
},
|
1112 |
+
{
|
1113 |
+
"epoch": 1.6020408163265305,
|
1114 |
+
"grad_norm": 232996.40625,
|
1115 |
+
"learning_rate": 2.6700680272108848e-06,
|
1116 |
+
"loss": 1.272,
|
1117 |
+
"step": 785
|
1118 |
+
},
|
1119 |
+
{
|
1120 |
+
"epoch": 1.6122448979591837,
|
1121 |
+
"grad_norm": 283414.875,
|
1122 |
+
"learning_rate": 2.687074829931973e-06,
|
1123 |
+
"loss": 1.1957,
|
1124 |
+
"step": 790
|
1125 |
+
},
|
1126 |
+
{
|
1127 |
+
"epoch": 1.6224489795918369,
|
1128 |
+
"grad_norm": 193658.25,
|
1129 |
+
"learning_rate": 2.7040816326530615e-06,
|
1130 |
+
"loss": 1.2704,
|
1131 |
+
"step": 795
|
1132 |
+
},
|
1133 |
+
{
|
1134 |
+
"epoch": 1.6326530612244898,
|
1135 |
+
"grad_norm": 196565.8125,
|
1136 |
+
"learning_rate": 2.7210884353741503e-06,
|
1137 |
+
"loss": 1.219,
|
1138 |
+
"step": 800
|
1139 |
+
},
|
1140 |
+
{
|
1141 |
+
"epoch": 1.6428571428571428,
|
1142 |
+
"grad_norm": 211995.5,
|
1143 |
+
"learning_rate": 2.7380952380952387e-06,
|
1144 |
+
"loss": 1.1321,
|
1145 |
+
"step": 805
|
1146 |
+
},
|
1147 |
+
{
|
1148 |
+
"epoch": 1.6530612244897958,
|
1149 |
+
"grad_norm": 206511.265625,
|
1150 |
+
"learning_rate": 2.7551020408163266e-06,
|
1151 |
+
"loss": 1.156,
|
1152 |
+
"step": 810
|
1153 |
+
},
|
1154 |
+
{
|
1155 |
+
"epoch": 1.663265306122449,
|
1156 |
+
"grad_norm": 269477.4375,
|
1157 |
+
"learning_rate": 2.772108843537415e-06,
|
1158 |
+
"loss": 1.0278,
|
1159 |
+
"step": 815
|
1160 |
+
},
|
1161 |
+
{
|
1162 |
+
"epoch": 1.6734693877551021,
|
1163 |
+
"grad_norm": 305353.09375,
|
1164 |
+
"learning_rate": 2.7891156462585034e-06,
|
1165 |
+
"loss": 1.1282,
|
1166 |
+
"step": 820
|
1167 |
+
},
|
1168 |
+
{
|
1169 |
+
"epoch": 1.683673469387755,
|
1170 |
+
"grad_norm": 405249.5625,
|
1171 |
+
"learning_rate": 2.8061224489795917e-06,
|
1172 |
+
"loss": 1.1018,
|
1173 |
+
"step": 825
|
1174 |
+
},
|
1175 |
+
{
|
1176 |
+
"epoch": 1.693877551020408,
|
1177 |
+
"grad_norm": 215439.359375,
|
1178 |
+
"learning_rate": 2.8231292517006805e-06,
|
1179 |
+
"loss": 1.0834,
|
1180 |
+
"step": 830
|
1181 |
+
},
|
1182 |
+
{
|
1183 |
+
"epoch": 1.7040816326530612,
|
1184 |
+
"grad_norm": 293742.0,
|
1185 |
+
"learning_rate": 2.840136054421769e-06,
|
1186 |
+
"loss": 1.1826,
|
1187 |
+
"step": 835
|
1188 |
+
},
|
1189 |
+
{
|
1190 |
+
"epoch": 1.7142857142857144,
|
1191 |
+
"grad_norm": 333310.65625,
|
1192 |
+
"learning_rate": 2.8571428571428573e-06,
|
1193 |
+
"loss": 1.1268,
|
1194 |
+
"step": 840
|
1195 |
+
},
|
1196 |
+
{
|
1197 |
+
"epoch": 1.7244897959183674,
|
1198 |
+
"grad_norm": 223134.421875,
|
1199 |
+
"learning_rate": 2.8741496598639456e-06,
|
1200 |
+
"loss": 1.1159,
|
1201 |
+
"step": 845
|
1202 |
+
},
|
1203 |
+
{
|
1204 |
+
"epoch": 1.7346938775510203,
|
1205 |
+
"grad_norm": 405755.96875,
|
1206 |
+
"learning_rate": 2.891156462585034e-06,
|
1207 |
+
"loss": 1.0942,
|
1208 |
+
"step": 850
|
1209 |
+
},
|
1210 |
+
{
|
1211 |
+
"epoch": 1.7448979591836735,
|
1212 |
+
"grad_norm": 285503.9375,
|
1213 |
+
"learning_rate": 2.908163265306123e-06,
|
1214 |
+
"loss": 1.1104,
|
1215 |
+
"step": 855
|
1216 |
+
},
|
1217 |
+
{
|
1218 |
+
"epoch": 1.7551020408163265,
|
1219 |
+
"grad_norm": 234126.296875,
|
1220 |
+
"learning_rate": 2.925170068027211e-06,
|
1221 |
+
"loss": 1.2139,
|
1222 |
+
"step": 860
|
1223 |
+
},
|
1224 |
+
{
|
1225 |
+
"epoch": 1.7653061224489797,
|
1226 |
+
"grad_norm": 554950.4375,
|
1227 |
+
"learning_rate": 2.9421768707482995e-06,
|
1228 |
+
"loss": 1.1507,
|
1229 |
+
"step": 865
|
1230 |
+
},
|
1231 |
+
{
|
1232 |
+
"epoch": 1.7755102040816326,
|
1233 |
+
"grad_norm": 168897.890625,
|
1234 |
+
"learning_rate": 2.959183673469388e-06,
|
1235 |
+
"loss": 1.0302,
|
1236 |
+
"step": 870
|
1237 |
+
},
|
1238 |
+
{
|
1239 |
+
"epoch": 1.7857142857142856,
|
1240 |
+
"grad_norm": 174446.953125,
|
1241 |
+
"learning_rate": 2.9761904761904763e-06,
|
1242 |
+
"loss": 1.0847,
|
1243 |
+
"step": 875
|
1244 |
+
},
|
1245 |
+
{
|
1246 |
+
"epoch": 1.7959183673469388,
|
1247 |
+
"grad_norm": 913355.6875,
|
1248 |
+
"learning_rate": 2.993197278911565e-06,
|
1249 |
+
"loss": 1.0703,
|
1250 |
+
"step": 880
|
1251 |
+
},
|
1252 |
+
{
|
1253 |
+
"epoch": 1.806122448979592,
|
1254 |
+
"grad_norm": 329008.75,
|
1255 |
+
"learning_rate": 3.0102040816326534e-06,
|
1256 |
+
"loss": 1.0336,
|
1257 |
+
"step": 885
|
1258 |
+
},
|
1259 |
+
{
|
1260 |
+
"epoch": 1.816326530612245,
|
1261 |
+
"grad_norm": 186949.75,
|
1262 |
+
"learning_rate": 3.027210884353742e-06,
|
1263 |
+
"loss": 1.1538,
|
1264 |
+
"step": 890
|
1265 |
+
},
|
1266 |
+
{
|
1267 |
+
"epoch": 1.8265306122448979,
|
1268 |
+
"grad_norm": 227551.765625,
|
1269 |
+
"learning_rate": 3.04421768707483e-06,
|
1270 |
+
"loss": 1.1304,
|
1271 |
+
"step": 895
|
1272 |
+
},
|
1273 |
+
{
|
1274 |
+
"epoch": 1.836734693877551,
|
1275 |
+
"grad_norm": 152557.625,
|
1276 |
+
"learning_rate": 3.0612244897959185e-06,
|
1277 |
+
"loss": 1.125,
|
1278 |
+
"step": 900
|
1279 |
+
},
|
1280 |
+
{
|
1281 |
+
"epoch": 1.8469387755102042,
|
1282 |
+
"grad_norm": 329511.65625,
|
1283 |
+
"learning_rate": 3.0782312925170073e-06,
|
1284 |
+
"loss": 1.1481,
|
1285 |
+
"step": 905
|
1286 |
+
},
|
1287 |
+
{
|
1288 |
+
"epoch": 1.8571428571428572,
|
1289 |
+
"grad_norm": 872083.75,
|
1290 |
+
"learning_rate": 3.0952380952380957e-06,
|
1291 |
+
"loss": 1.0087,
|
1292 |
+
"step": 910
|
1293 |
+
},
|
1294 |
+
{
|
1295 |
+
"epoch": 1.8673469387755102,
|
1296 |
+
"grad_norm": 440686.78125,
|
1297 |
+
"learning_rate": 3.112244897959184e-06,
|
1298 |
+
"loss": 1.1177,
|
1299 |
+
"step": 915
|
1300 |
+
},
|
1301 |
+
{
|
1302 |
+
"epoch": 1.8775510204081631,
|
1303 |
+
"grad_norm": 181735.3125,
|
1304 |
+
"learning_rate": 3.1292517006802725e-06,
|
1305 |
+
"loss": 0.9354,
|
1306 |
+
"step": 920
|
1307 |
+
},
|
1308 |
+
{
|
1309 |
+
"epoch": 1.8877551020408163,
|
1310 |
+
"grad_norm": 337831.65625,
|
1311 |
+
"learning_rate": 3.146258503401361e-06,
|
1312 |
+
"loss": 1.1114,
|
1313 |
+
"step": 925
|
1314 |
+
},
|
1315 |
+
{
|
1316 |
+
"epoch": 1.8979591836734695,
|
1317 |
+
"grad_norm": 350815.09375,
|
1318 |
+
"learning_rate": 3.1632653061224496e-06,
|
1319 |
+
"loss": 0.9113,
|
1320 |
+
"step": 930
|
1321 |
+
},
|
1322 |
+
{
|
1323 |
+
"epoch": 1.9081632653061225,
|
1324 |
+
"grad_norm": 341441.9375,
|
1325 |
+
"learning_rate": 3.180272108843538e-06,
|
1326 |
+
"loss": 1.1561,
|
1327 |
+
"step": 935
|
1328 |
+
},
|
1329 |
+
{
|
1330 |
+
"epoch": 1.9183673469387754,
|
1331 |
+
"grad_norm": 268450.65625,
|
1332 |
+
"learning_rate": 3.1972789115646264e-06,
|
1333 |
+
"loss": 1.0589,
|
1334 |
+
"step": 940
|
1335 |
+
},
|
1336 |
+
{
|
1337 |
+
"epoch": 1.9285714285714286,
|
1338 |
+
"grad_norm": 258875.890625,
|
1339 |
+
"learning_rate": 3.2142857142857147e-06,
|
1340 |
+
"loss": 1.0471,
|
1341 |
+
"step": 945
|
1342 |
+
},
|
1343 |
+
{
|
1344 |
+
"epoch": 1.9387755102040818,
|
1345 |
+
"grad_norm": 206252.34375,
|
1346 |
+
"learning_rate": 3.231292517006803e-06,
|
1347 |
+
"loss": 0.9862,
|
1348 |
+
"step": 950
|
1349 |
+
},
|
1350 |
+
{
|
1351 |
+
"epoch": 1.9489795918367347,
|
1352 |
+
"grad_norm": 193751.328125,
|
1353 |
+
"learning_rate": 3.248299319727892e-06,
|
1354 |
+
"loss": 0.9427,
|
1355 |
+
"step": 955
|
1356 |
+
},
|
1357 |
+
{
|
1358 |
+
"epoch": 1.9591836734693877,
|
1359 |
+
"grad_norm": 307227.28125,
|
1360 |
+
"learning_rate": 3.2653061224489794e-06,
|
1361 |
+
"loss": 0.9851,
|
1362 |
+
"step": 960
|
1363 |
+
},
|
1364 |
+
{
|
1365 |
+
"epoch": 1.9693877551020407,
|
1366 |
+
"grad_norm": 203521.03125,
|
1367 |
+
"learning_rate": 3.2823129251700682e-06,
|
1368 |
+
"loss": 0.9827,
|
1369 |
+
"step": 965
|
1370 |
+
},
|
1371 |
+
{
|
1372 |
+
"epoch": 1.9795918367346939,
|
1373 |
+
"grad_norm": 254505.734375,
|
1374 |
+
"learning_rate": 3.2993197278911566e-06,
|
1375 |
+
"loss": 0.8342,
|
1376 |
+
"step": 970
|
1377 |
+
},
|
1378 |
+
{
|
1379 |
+
"epoch": 1.989795918367347,
|
1380 |
+
"grad_norm": 185893.46875,
|
1381 |
+
"learning_rate": 3.316326530612245e-06,
|
1382 |
+
"loss": 0.977,
|
1383 |
+
"step": 975
|
1384 |
+
},
|
1385 |
+
{
|
1386 |
+
"epoch": 2.0,
|
1387 |
+
"grad_norm": 302582.59375,
|
1388 |
+
"learning_rate": 3.3333333333333333e-06,
|
1389 |
+
"loss": 0.8945,
|
1390 |
+
"step": 980
|
1391 |
+
},
|
1392 |
+
{
|
1393 |
+
"epoch": 2.0,
|
1394 |
+
"eval_accuracy": 0.8829644533869886,
|
1395 |
+
"eval_loss": 0.7550393342971802,
|
1396 |
+
"eval_runtime": 181.962,
|
1397 |
+
"eval_samples_per_second": 16.388,
|
1398 |
+
"eval_steps_per_second": 0.687,
|
1399 |
+
"step": 980
|
1400 |
+
},
|
1401 |
+
{
|
1402 |
+
"epoch": 2.010204081632653,
|
1403 |
+
"grad_norm": 373234.625,
|
1404 |
+
"learning_rate": 3.3503401360544217e-06,
|
1405 |
+
"loss": 0.8913,
|
1406 |
+
"step": 985
|
1407 |
+
},
|
1408 |
+
{
|
1409 |
+
"epoch": 2.020408163265306,
|
1410 |
+
"grad_norm": 224935.203125,
|
1411 |
+
"learning_rate": 3.3673469387755105e-06,
|
1412 |
+
"loss": 0.8764,
|
1413 |
+
"step": 990
|
1414 |
+
},
|
1415 |
+
{
|
1416 |
+
"epoch": 2.0306122448979593,
|
1417 |
+
"grad_norm": 329704.375,
|
1418 |
+
"learning_rate": 3.384353741496599e-06,
|
1419 |
+
"loss": 0.9428,
|
1420 |
+
"step": 995
|
1421 |
+
},
|
1422 |
+
{
|
1423 |
+
"epoch": 2.0408163265306123,
|
1424 |
+
"grad_norm": 251769.0625,
|
1425 |
+
"learning_rate": 3.4013605442176872e-06,
|
1426 |
+
"loss": 0.9975,
|
1427 |
+
"step": 1000
|
1428 |
+
},
|
1429 |
+
{
|
1430 |
+
"epoch": 2.0510204081632653,
|
1431 |
+
"grad_norm": 547890.875,
|
1432 |
+
"learning_rate": 3.4183673469387756e-06,
|
1433 |
+
"loss": 0.9587,
|
1434 |
+
"step": 1005
|
1435 |
+
},
|
1436 |
+
{
|
1437 |
+
"epoch": 2.061224489795918,
|
1438 |
+
"grad_norm": 286754.53125,
|
1439 |
+
"learning_rate": 3.435374149659864e-06,
|
1440 |
+
"loss": 1.075,
|
1441 |
+
"step": 1010
|
1442 |
+
},
|
1443 |
+
{
|
1444 |
+
"epoch": 2.0714285714285716,
|
1445 |
+
"grad_norm": 784546.6875,
|
1446 |
+
"learning_rate": 3.4523809523809528e-06,
|
1447 |
+
"loss": 0.879,
|
1448 |
+
"step": 1015
|
1449 |
+
},
|
1450 |
+
{
|
1451 |
+
"epoch": 2.0816326530612246,
|
1452 |
+
"grad_norm": 434627.28125,
|
1453 |
+
"learning_rate": 3.469387755102041e-06,
|
1454 |
+
"loss": 0.9471,
|
1455 |
+
"step": 1020
|
1456 |
+
},
|
1457 |
+
{
|
1458 |
+
"epoch": 2.0918367346938775,
|
1459 |
+
"grad_norm": 550331.0,
|
1460 |
+
"learning_rate": 3.4863945578231295e-06,
|
1461 |
+
"loss": 0.8507,
|
1462 |
+
"step": 1025
|
1463 |
+
},
|
1464 |
+
{
|
1465 |
+
"epoch": 2.1020408163265305,
|
1466 |
+
"grad_norm": 227700.0,
|
1467 |
+
"learning_rate": 3.503401360544218e-06,
|
1468 |
+
"loss": 0.9994,
|
1469 |
+
"step": 1030
|
1470 |
+
},
|
1471 |
+
{
|
1472 |
+
"epoch": 2.1122448979591835,
|
1473 |
+
"grad_norm": 153363.25,
|
1474 |
+
"learning_rate": 3.5204081632653062e-06,
|
1475 |
+
"loss": 0.8697,
|
1476 |
+
"step": 1035
|
1477 |
+
},
|
1478 |
+
{
|
1479 |
+
"epoch": 2.122448979591837,
|
1480 |
+
"grad_norm": 649990.125,
|
1481 |
+
"learning_rate": 3.537414965986395e-06,
|
1482 |
+
"loss": 0.9213,
|
1483 |
+
"step": 1040
|
1484 |
+
},
|
1485 |
+
{
|
1486 |
+
"epoch": 2.13265306122449,
|
1487 |
+
"grad_norm": 255040.046875,
|
1488 |
+
"learning_rate": 3.5544217687074834e-06,
|
1489 |
+
"loss": 0.7944,
|
1490 |
+
"step": 1045
|
1491 |
+
},
|
1492 |
+
{
|
1493 |
+
"epoch": 2.142857142857143,
|
1494 |
+
"grad_norm": 730054.6875,
|
1495 |
+
"learning_rate": 3.5714285714285718e-06,
|
1496 |
+
"loss": 0.9368,
|
1497 |
+
"step": 1050
|
1498 |
+
},
|
1499 |
+
{
|
1500 |
+
"epoch": 2.1530612244897958,
|
1501 |
+
"grad_norm": 333649.6875,
|
1502 |
+
"learning_rate": 3.58843537414966e-06,
|
1503 |
+
"loss": 0.8134,
|
1504 |
+
"step": 1055
|
1505 |
+
},
|
1506 |
+
{
|
1507 |
+
"epoch": 2.163265306122449,
|
1508 |
+
"grad_norm": 555207.0,
|
1509 |
+
"learning_rate": 3.6054421768707485e-06,
|
1510 |
+
"loss": 0.9628,
|
1511 |
+
"step": 1060
|
1512 |
+
},
|
1513 |
+
{
|
1514 |
+
"epoch": 2.173469387755102,
|
1515 |
+
"grad_norm": 195841.140625,
|
1516 |
+
"learning_rate": 3.6224489795918373e-06,
|
1517 |
+
"loss": 0.9173,
|
1518 |
+
"step": 1065
|
1519 |
+
},
|
1520 |
+
{
|
1521 |
+
"epoch": 2.183673469387755,
|
1522 |
+
"grad_norm": 280471.59375,
|
1523 |
+
"learning_rate": 3.6394557823129257e-06,
|
1524 |
+
"loss": 0.7997,
|
1525 |
+
"step": 1070
|
1526 |
+
},
|
1527 |
+
{
|
1528 |
+
"epoch": 2.193877551020408,
|
1529 |
+
"grad_norm": 273992.375,
|
1530 |
+
"learning_rate": 3.656462585034014e-06,
|
1531 |
+
"loss": 0.8739,
|
1532 |
+
"step": 1075
|
1533 |
+
},
|
1534 |
+
{
|
1535 |
+
"epoch": 2.204081632653061,
|
1536 |
+
"grad_norm": 289144.0,
|
1537 |
+
"learning_rate": 3.6734693877551024e-06,
|
1538 |
+
"loss": 0.8514,
|
1539 |
+
"step": 1080
|
1540 |
+
},
|
1541 |
+
{
|
1542 |
+
"epoch": 2.2142857142857144,
|
1543 |
+
"grad_norm": 520327.90625,
|
1544 |
+
"learning_rate": 3.690476190476191e-06,
|
1545 |
+
"loss": 0.7891,
|
1546 |
+
"step": 1085
|
1547 |
+
},
|
1548 |
+
{
|
1549 |
+
"epoch": 2.2244897959183674,
|
1550 |
+
"grad_norm": 180651.515625,
|
1551 |
+
"learning_rate": 3.7074829931972796e-06,
|
1552 |
+
"loss": 0.6701,
|
1553 |
+
"step": 1090
|
1554 |
+
},
|
1555 |
+
{
|
1556 |
+
"epoch": 2.2346938775510203,
|
1557 |
+
"grad_norm": 600967.4375,
|
1558 |
+
"learning_rate": 3.724489795918368e-06,
|
1559 |
+
"loss": 0.8048,
|
1560 |
+
"step": 1095
|
1561 |
+
},
|
1562 |
+
{
|
1563 |
+
"epoch": 2.2448979591836733,
|
1564 |
+
"grad_norm": 583708.1875,
|
1565 |
+
"learning_rate": 3.7414965986394563e-06,
|
1566 |
+
"loss": 0.8849,
|
1567 |
+
"step": 1100
|
1568 |
+
},
|
1569 |
+
{
|
1570 |
+
"epoch": 2.2551020408163267,
|
1571 |
+
"grad_norm": 359820.625,
|
1572 |
+
"learning_rate": 3.7585034013605443e-06,
|
1573 |
+
"loss": 0.9233,
|
1574 |
+
"step": 1105
|
1575 |
+
},
|
1576 |
+
{
|
1577 |
+
"epoch": 2.2653061224489797,
|
1578 |
+
"grad_norm": 485576.6875,
|
1579 |
+
"learning_rate": 3.7755102040816327e-06,
|
1580 |
+
"loss": 0.7481,
|
1581 |
+
"step": 1110
|
1582 |
+
},
|
1583 |
+
{
|
1584 |
+
"epoch": 2.2755102040816326,
|
1585 |
+
"grad_norm": 630407.5,
|
1586 |
+
"learning_rate": 3.792517006802721e-06,
|
1587 |
+
"loss": 0.9346,
|
1588 |
+
"step": 1115
|
1589 |
+
},
|
1590 |
+
{
|
1591 |
+
"epoch": 2.2857142857142856,
|
1592 |
+
"grad_norm": 223983.765625,
|
1593 |
+
"learning_rate": 3.80952380952381e-06,
|
1594 |
+
"loss": 0.8671,
|
1595 |
+
"step": 1120
|
1596 |
+
},
|
1597 |
+
{
|
1598 |
+
"epoch": 2.295918367346939,
|
1599 |
+
"grad_norm": 276714.1875,
|
1600 |
+
"learning_rate": 3.826530612244898e-06,
|
1601 |
+
"loss": 0.9081,
|
1602 |
+
"step": 1125
|
1603 |
+
},
|
1604 |
+
{
|
1605 |
+
"epoch": 2.306122448979592,
|
1606 |
+
"grad_norm": 609418.25,
|
1607 |
+
"learning_rate": 3.843537414965986e-06,
|
1608 |
+
"loss": 0.8401,
|
1609 |
+
"step": 1130
|
1610 |
+
},
|
1611 |
+
{
|
1612 |
+
"epoch": 2.316326530612245,
|
1613 |
+
"grad_norm": 336448.625,
|
1614 |
+
"learning_rate": 3.860544217687075e-06,
|
1615 |
+
"loss": 0.8894,
|
1616 |
+
"step": 1135
|
1617 |
+
},
|
1618 |
+
{
|
1619 |
+
"epoch": 2.326530612244898,
|
1620 |
+
"grad_norm": 264061.8125,
|
1621 |
+
"learning_rate": 3.877551020408164e-06,
|
1622 |
+
"loss": 0.8953,
|
1623 |
+
"step": 1140
|
1624 |
+
},
|
1625 |
+
{
|
1626 |
+
"epoch": 2.336734693877551,
|
1627 |
+
"grad_norm": 395307.84375,
|
1628 |
+
"learning_rate": 3.894557823129252e-06,
|
1629 |
+
"loss": 0.8055,
|
1630 |
+
"step": 1145
|
1631 |
+
},
|
1632 |
+
{
|
1633 |
+
"epoch": 2.3469387755102042,
|
1634 |
+
"grad_norm": 252475.984375,
|
1635 |
+
"learning_rate": 3.9115646258503405e-06,
|
1636 |
+
"loss": 0.9308,
|
1637 |
+
"step": 1150
|
1638 |
+
},
|
1639 |
+
{
|
1640 |
+
"epoch": 2.357142857142857,
|
1641 |
+
"grad_norm": 303853.15625,
|
1642 |
+
"learning_rate": 3.928571428571429e-06,
|
1643 |
+
"loss": 0.7525,
|
1644 |
+
"step": 1155
|
1645 |
+
},
|
1646 |
+
{
|
1647 |
+
"epoch": 2.36734693877551,
|
1648 |
+
"grad_norm": 534944.0625,
|
1649 |
+
"learning_rate": 3.945578231292517e-06,
|
1650 |
+
"loss": 0.8447,
|
1651 |
+
"step": 1160
|
1652 |
+
},
|
1653 |
+
{
|
1654 |
+
"epoch": 2.377551020408163,
|
1655 |
+
"grad_norm": 866857.6875,
|
1656 |
+
"learning_rate": 3.9625850340136056e-06,
|
1657 |
+
"loss": 0.8686,
|
1658 |
+
"step": 1165
|
1659 |
+
},
|
1660 |
+
{
|
1661 |
+
"epoch": 2.387755102040816,
|
1662 |
+
"grad_norm": 890087.75,
|
1663 |
+
"learning_rate": 3.979591836734694e-06,
|
1664 |
+
"loss": 0.8068,
|
1665 |
+
"step": 1170
|
1666 |
+
},
|
1667 |
+
{
|
1668 |
+
"epoch": 2.3979591836734695,
|
1669 |
+
"grad_norm": 421775.9375,
|
1670 |
+
"learning_rate": 3.996598639455782e-06,
|
1671 |
+
"loss": 0.7673,
|
1672 |
+
"step": 1175
|
1673 |
+
},
|
1674 |
+
{
|
1675 |
+
"epoch": 2.4081632653061225,
|
1676 |
+
"grad_norm": 217559.484375,
|
1677 |
+
"learning_rate": 4.013605442176871e-06,
|
1678 |
+
"loss": 0.6329,
|
1679 |
+
"step": 1180
|
1680 |
+
},
|
1681 |
+
{
|
1682 |
+
"epoch": 2.4183673469387754,
|
1683 |
+
"grad_norm": 452332.21875,
|
1684 |
+
"learning_rate": 4.03061224489796e-06,
|
1685 |
+
"loss": 0.6945,
|
1686 |
+
"step": 1185
|
1687 |
+
},
|
1688 |
+
{
|
1689 |
+
"epoch": 2.4285714285714284,
|
1690 |
+
"grad_norm": 150821.421875,
|
1691 |
+
"learning_rate": 4.047619047619048e-06,
|
1692 |
+
"loss": 0.7392,
|
1693 |
+
"step": 1190
|
1694 |
+
},
|
1695 |
+
{
|
1696 |
+
"epoch": 2.438775510204082,
|
1697 |
+
"grad_norm": 474443.5,
|
1698 |
+
"learning_rate": 4.064625850340137e-06,
|
1699 |
+
"loss": 0.6836,
|
1700 |
+
"step": 1195
|
1701 |
+
},
|
1702 |
+
{
|
1703 |
+
"epoch": 2.4489795918367347,
|
1704 |
+
"grad_norm": 673342.75,
|
1705 |
+
"learning_rate": 4.081632653061225e-06,
|
1706 |
+
"loss": 0.8028,
|
1707 |
+
"step": 1200
|
1708 |
+
},
|
1709 |
+
{
|
1710 |
+
"epoch": 2.4591836734693877,
|
1711 |
+
"grad_norm": 404749.40625,
|
1712 |
+
"learning_rate": 4.098639455782313e-06,
|
1713 |
+
"loss": 0.7749,
|
1714 |
+
"step": 1205
|
1715 |
+
},
|
1716 |
+
{
|
1717 |
+
"epoch": 2.4693877551020407,
|
1718 |
+
"grad_norm": 462878.5,
|
1719 |
+
"learning_rate": 4.115646258503402e-06,
|
1720 |
+
"loss": 0.8205,
|
1721 |
+
"step": 1210
|
1722 |
+
},
|
1723 |
+
{
|
1724 |
+
"epoch": 2.479591836734694,
|
1725 |
+
"grad_norm": 1012966.9375,
|
1726 |
+
"learning_rate": 4.13265306122449e-06,
|
1727 |
+
"loss": 0.9547,
|
1728 |
+
"step": 1215
|
1729 |
+
},
|
1730 |
+
{
|
1731 |
+
"epoch": 2.489795918367347,
|
1732 |
+
"grad_norm": 288577.46875,
|
1733 |
+
"learning_rate": 4.1496598639455785e-06,
|
1734 |
+
"loss": 0.859,
|
1735 |
+
"step": 1220
|
1736 |
+
},
|
1737 |
+
{
|
1738 |
+
"epoch": 2.5,
|
1739 |
+
"grad_norm": 563161.625,
|
1740 |
+
"learning_rate": 4.166666666666667e-06,
|
1741 |
+
"loss": 0.7339,
|
1742 |
+
"step": 1225
|
1743 |
+
},
|
1744 |
+
{
|
1745 |
+
"epoch": 2.510204081632653,
|
1746 |
+
"grad_norm": 346156.8125,
|
1747 |
+
"learning_rate": 4.183673469387755e-06,
|
1748 |
+
"loss": 0.7978,
|
1749 |
+
"step": 1230
|
1750 |
+
},
|
1751 |
+
{
|
1752 |
+
"epoch": 2.520408163265306,
|
1753 |
+
"grad_norm": 333493.28125,
|
1754 |
+
"learning_rate": 4.2006802721088445e-06,
|
1755 |
+
"loss": 0.6839,
|
1756 |
+
"step": 1235
|
1757 |
+
},
|
1758 |
+
{
|
1759 |
+
"epoch": 2.5306122448979593,
|
1760 |
+
"grad_norm": 339418.625,
|
1761 |
+
"learning_rate": 4.217687074829933e-06,
|
1762 |
+
"loss": 0.7972,
|
1763 |
+
"step": 1240
|
1764 |
+
},
|
1765 |
+
{
|
1766 |
+
"epoch": 2.5408163265306123,
|
1767 |
+
"grad_norm": 927960.0,
|
1768 |
+
"learning_rate": 4.234693877551021e-06,
|
1769 |
+
"loss": 0.8328,
|
1770 |
+
"step": 1245
|
1771 |
+
},
|
1772 |
+
{
|
1773 |
+
"epoch": 2.5510204081632653,
|
1774 |
+
"grad_norm": 279461.03125,
|
1775 |
+
"learning_rate": 4.251700680272109e-06,
|
1776 |
+
"loss": 0.8894,
|
1777 |
+
"step": 1250
|
1778 |
+
},
|
1779 |
+
{
|
1780 |
+
"epoch": 2.561224489795918,
|
1781 |
+
"grad_norm": 798350.625,
|
1782 |
+
"learning_rate": 4.268707482993197e-06,
|
1783 |
+
"loss": 0.9658,
|
1784 |
+
"step": 1255
|
1785 |
+
},
|
1786 |
+
{
|
1787 |
+
"epoch": 2.571428571428571,
|
1788 |
+
"grad_norm": 483177.21875,
|
1789 |
+
"learning_rate": 4.2857142857142855e-06,
|
1790 |
+
"loss": 0.7366,
|
1791 |
+
"step": 1260
|
1792 |
+
},
|
1793 |
+
{
|
1794 |
+
"epoch": 2.5816326530612246,
|
1795 |
+
"grad_norm": 835873.375,
|
1796 |
+
"learning_rate": 4.302721088435375e-06,
|
1797 |
+
"loss": 0.9837,
|
1798 |
+
"step": 1265
|
1799 |
+
},
|
1800 |
+
{
|
1801 |
+
"epoch": 2.5918367346938775,
|
1802 |
+
"grad_norm": 440312.0625,
|
1803 |
+
"learning_rate": 4.319727891156463e-06,
|
1804 |
+
"loss": 0.8315,
|
1805 |
+
"step": 1270
|
1806 |
+
},
|
1807 |
+
{
|
1808 |
+
"epoch": 2.6020408163265305,
|
1809 |
+
"grad_norm": 584700.75,
|
1810 |
+
"learning_rate": 4.336734693877551e-06,
|
1811 |
+
"loss": 0.7716,
|
1812 |
+
"step": 1275
|
1813 |
+
},
|
1814 |
+
{
|
1815 |
+
"epoch": 2.612244897959184,
|
1816 |
+
"grad_norm": 417963.6875,
|
1817 |
+
"learning_rate": 4.35374149659864e-06,
|
1818 |
+
"loss": 0.6108,
|
1819 |
+
"step": 1280
|
1820 |
+
},
|
1821 |
+
{
|
1822 |
+
"epoch": 2.622448979591837,
|
1823 |
+
"grad_norm": 510078.15625,
|
1824 |
+
"learning_rate": 4.370748299319728e-06,
|
1825 |
+
"loss": 0.7456,
|
1826 |
+
"step": 1285
|
1827 |
+
},
|
1828 |
+
{
|
1829 |
+
"epoch": 2.63265306122449,
|
1830 |
+
"grad_norm": 430475.28125,
|
1831 |
+
"learning_rate": 4.3877551020408165e-06,
|
1832 |
+
"loss": 0.9455,
|
1833 |
+
"step": 1290
|
1834 |
+
},
|
1835 |
+
{
|
1836 |
+
"epoch": 2.642857142857143,
|
1837 |
+
"grad_norm": 622921.375,
|
1838 |
+
"learning_rate": 4.404761904761905e-06,
|
1839 |
+
"loss": 0.8095,
|
1840 |
+
"step": 1295
|
1841 |
+
},
|
1842 |
+
{
|
1843 |
+
"epoch": 2.6530612244897958,
|
1844 |
+
"grad_norm": 389947.09375,
|
1845 |
+
"learning_rate": 4.421768707482993e-06,
|
1846 |
+
"loss": 0.9038,
|
1847 |
+
"step": 1300
|
1848 |
+
},
|
1849 |
+
{
|
1850 |
+
"epoch": 2.663265306122449,
|
1851 |
+
"grad_norm": 331616.6875,
|
1852 |
+
"learning_rate": 4.438775510204082e-06,
|
1853 |
+
"loss": 0.831,
|
1854 |
+
"step": 1305
|
1855 |
+
},
|
1856 |
+
{
|
1857 |
+
"epoch": 2.673469387755102,
|
1858 |
+
"grad_norm": 861532.0625,
|
1859 |
+
"learning_rate": 4.45578231292517e-06,
|
1860 |
+
"loss": 0.6686,
|
1861 |
+
"step": 1310
|
1862 |
+
},
|
1863 |
+
{
|
1864 |
+
"epoch": 2.683673469387755,
|
1865 |
+
"grad_norm": 742906.1875,
|
1866 |
+
"learning_rate": 4.472789115646259e-06,
|
1867 |
+
"loss": 0.6233,
|
1868 |
+
"step": 1315
|
1869 |
+
},
|
1870 |
+
{
|
1871 |
+
"epoch": 2.693877551020408,
|
1872 |
+
"grad_norm": 563013.25,
|
1873 |
+
"learning_rate": 4.489795918367348e-06,
|
1874 |
+
"loss": 0.7734,
|
1875 |
+
"step": 1320
|
1876 |
+
},
|
1877 |
+
{
|
1878 |
+
"epoch": 2.704081632653061,
|
1879 |
+
"grad_norm": 488701.28125,
|
1880 |
+
"learning_rate": 4.506802721088436e-06,
|
1881 |
+
"loss": 0.8453,
|
1882 |
+
"step": 1325
|
1883 |
+
},
|
1884 |
+
{
|
1885 |
+
"epoch": 2.7142857142857144,
|
1886 |
+
"grad_norm": 566946.875,
|
1887 |
+
"learning_rate": 4.523809523809524e-06,
|
1888 |
+
"loss": 0.8815,
|
1889 |
+
"step": 1330
|
1890 |
+
},
|
1891 |
+
{
|
1892 |
+
"epoch": 2.7244897959183674,
|
1893 |
+
"grad_norm": 327734.34375,
|
1894 |
+
"learning_rate": 4.540816326530613e-06,
|
1895 |
+
"loss": 0.861,
|
1896 |
+
"step": 1335
|
1897 |
+
},
|
1898 |
+
{
|
1899 |
+
"epoch": 2.7346938775510203,
|
1900 |
+
"grad_norm": 297249.40625,
|
1901 |
+
"learning_rate": 4.557823129251701e-06,
|
1902 |
+
"loss": 0.7455,
|
1903 |
+
"step": 1340
|
1904 |
+
},
|
1905 |
+
{
|
1906 |
+
"epoch": 2.7448979591836737,
|
1907 |
+
"grad_norm": 953682.0,
|
1908 |
+
"learning_rate": 4.5748299319727895e-06,
|
1909 |
+
"loss": 0.8965,
|
1910 |
+
"step": 1345
|
1911 |
+
},
|
1912 |
+
{
|
1913 |
+
"epoch": 2.7551020408163263,
|
1914 |
+
"grad_norm": 430745.6875,
|
1915 |
+
"learning_rate": 4.591836734693878e-06,
|
1916 |
+
"loss": 0.7581,
|
1917 |
+
"step": 1350
|
1918 |
+
},
|
1919 |
+
{
|
1920 |
+
"epoch": 2.7653061224489797,
|
1921 |
+
"grad_norm": 475997.59375,
|
1922 |
+
"learning_rate": 4.608843537414966e-06,
|
1923 |
+
"loss": 0.7165,
|
1924 |
+
"step": 1355
|
1925 |
+
},
|
1926 |
+
{
|
1927 |
+
"epoch": 2.7755102040816326,
|
1928 |
+
"grad_norm": 331845.90625,
|
1929 |
+
"learning_rate": 4.6258503401360546e-06,
|
1930 |
+
"loss": 0.8438,
|
1931 |
+
"step": 1360
|
1932 |
+
},
|
1933 |
+
{
|
1934 |
+
"epoch": 2.7857142857142856,
|
1935 |
+
"grad_norm": 1252833.625,
|
1936 |
+
"learning_rate": 4.642857142857144e-06,
|
1937 |
+
"loss": 0.9688,
|
1938 |
+
"step": 1365
|
1939 |
+
},
|
1940 |
+
{
|
1941 |
+
"epoch": 2.795918367346939,
|
1942 |
+
"grad_norm": 416551.65625,
|
1943 |
+
"learning_rate": 4.659863945578232e-06,
|
1944 |
+
"loss": 0.7577,
|
1945 |
+
"step": 1370
|
1946 |
+
},
|
1947 |
+
{
|
1948 |
+
"epoch": 2.806122448979592,
|
1949 |
+
"grad_norm": 568027.8125,
|
1950 |
+
"learning_rate": 4.6768707482993205e-06,
|
1951 |
+
"loss": 0.6014,
|
1952 |
+
"step": 1375
|
1953 |
+
},
|
1954 |
+
{
|
1955 |
+
"epoch": 2.816326530612245,
|
1956 |
+
"grad_norm": 281311.375,
|
1957 |
+
"learning_rate": 4.693877551020409e-06,
|
1958 |
+
"loss": 0.7129,
|
1959 |
+
"step": 1380
|
1960 |
+
},
|
1961 |
+
{
|
1962 |
+
"epoch": 2.826530612244898,
|
1963 |
+
"grad_norm": 376276.53125,
|
1964 |
+
"learning_rate": 4.710884353741497e-06,
|
1965 |
+
"loss": 0.6655,
|
1966 |
+
"step": 1385
|
1967 |
+
},
|
1968 |
+
{
|
1969 |
+
"epoch": 2.836734693877551,
|
1970 |
+
"grad_norm": 655210.0625,
|
1971 |
+
"learning_rate": 4.727891156462586e-06,
|
1972 |
+
"loss": 0.7257,
|
1973 |
+
"step": 1390
|
1974 |
+
},
|
1975 |
+
{
|
1976 |
+
"epoch": 2.8469387755102042,
|
1977 |
+
"grad_norm": 250385.671875,
|
1978 |
+
"learning_rate": 4.744897959183674e-06,
|
1979 |
+
"loss": 0.662,
|
1980 |
+
"step": 1395
|
1981 |
+
},
|
1982 |
+
{
|
1983 |
+
"epoch": 2.857142857142857,
|
1984 |
+
"grad_norm": 481249.96875,
|
1985 |
+
"learning_rate": 4.761904761904762e-06,
|
1986 |
+
"loss": 0.6335,
|
1987 |
+
"step": 1400
|
1988 |
+
},
|
1989 |
+
{
|
1990 |
+
"epoch": 2.86734693877551,
|
1991 |
+
"grad_norm": 742787.8125,
|
1992 |
+
"learning_rate": 4.778911564625851e-06,
|
1993 |
+
"loss": 0.8095,
|
1994 |
+
"step": 1405
|
1995 |
+
},
|
1996 |
+
{
|
1997 |
+
"epoch": 2.877551020408163,
|
1998 |
+
"grad_norm": 914541.125,
|
1999 |
+
"learning_rate": 4.795918367346939e-06,
|
2000 |
+
"loss": 0.7427,
|
2001 |
+
"step": 1410
|
2002 |
+
},
|
2003 |
+
{
|
2004 |
+
"epoch": 2.887755102040816,
|
2005 |
+
"grad_norm": 296453.9375,
|
2006 |
+
"learning_rate": 4.8129251700680275e-06,
|
2007 |
+
"loss": 0.6538,
|
2008 |
+
"step": 1415
|
2009 |
+
},
|
2010 |
+
{
|
2011 |
+
"epoch": 2.8979591836734695,
|
2012 |
+
"grad_norm": 804697.8125,
|
2013 |
+
"learning_rate": 4.829931972789116e-06,
|
2014 |
+
"loss": 1.0127,
|
2015 |
+
"step": 1420
|
2016 |
+
},
|
2017 |
+
{
|
2018 |
+
"epoch": 2.9081632653061225,
|
2019 |
+
"grad_norm": 274531.59375,
|
2020 |
+
"learning_rate": 4.846938775510204e-06,
|
2021 |
+
"loss": 0.8091,
|
2022 |
+
"step": 1425
|
2023 |
+
},
|
2024 |
+
{
|
2025 |
+
"epoch": 2.9183673469387754,
|
2026 |
+
"grad_norm": 337409.71875,
|
2027 |
+
"learning_rate": 4.863945578231293e-06,
|
2028 |
+
"loss": 0.8296,
|
2029 |
+
"step": 1430
|
2030 |
+
},
|
2031 |
+
{
|
2032 |
+
"epoch": 2.928571428571429,
|
2033 |
+
"grad_norm": 714898.125,
|
2034 |
+
"learning_rate": 4.880952380952381e-06,
|
2035 |
+
"loss": 0.7965,
|
2036 |
+
"step": 1435
|
2037 |
+
},
|
2038 |
+
{
|
2039 |
+
"epoch": 2.938775510204082,
|
2040 |
+
"grad_norm": 901902.9375,
|
2041 |
+
"learning_rate": 4.897959183673469e-06,
|
2042 |
+
"loss": 0.7622,
|
2043 |
+
"step": 1440
|
2044 |
+
},
|
2045 |
+
{
|
2046 |
+
"epoch": 2.9489795918367347,
|
2047 |
+
"grad_norm": 645474.3125,
|
2048 |
+
"learning_rate": 4.914965986394558e-06,
|
2049 |
+
"loss": 0.7526,
|
2050 |
+
"step": 1445
|
2051 |
+
},
|
2052 |
+
{
|
2053 |
+
"epoch": 2.9591836734693877,
|
2054 |
+
"grad_norm": 641594.5,
|
2055 |
+
"learning_rate": 4.931972789115647e-06,
|
2056 |
+
"loss": 0.764,
|
2057 |
+
"step": 1450
|
2058 |
+
},
|
2059 |
+
{
|
2060 |
+
"epoch": 2.9693877551020407,
|
2061 |
+
"grad_norm": 1178937.25,
|
2062 |
+
"learning_rate": 4.948979591836735e-06,
|
2063 |
+
"loss": 0.7085,
|
2064 |
+
"step": 1455
|
2065 |
+
},
|
2066 |
+
{
|
2067 |
+
"epoch": 2.979591836734694,
|
2068 |
+
"grad_norm": 601102.75,
|
2069 |
+
"learning_rate": 4.965986394557824e-06,
|
2070 |
+
"loss": 0.7176,
|
2071 |
+
"step": 1460
|
2072 |
+
},
|
2073 |
+
{
|
2074 |
+
"epoch": 2.989795918367347,
|
2075 |
+
"grad_norm": 662777.9375,
|
2076 |
+
"learning_rate": 4.982993197278912e-06,
|
2077 |
+
"loss": 0.6701,
|
2078 |
+
"step": 1465
|
2079 |
+
},
|
2080 |
+
{
|
2081 |
+
"epoch": 3.0,
|
2082 |
+
"grad_norm": 331967.96875,
|
2083 |
+
"learning_rate": 5e-06,
|
2084 |
+
"loss": 0.5434,
|
2085 |
+
"step": 1470
|
2086 |
+
},
|
2087 |
+
{
|
2088 |
+
"epoch": 3.0,
|
2089 |
+
"eval_accuracy": 0.8792756539235412,
|
2090 |
+
"eval_loss": 0.5291165113449097,
|
2091 |
+
"eval_runtime": 180.5859,
|
2092 |
+
"eval_samples_per_second": 16.513,
|
2093 |
+
"eval_steps_per_second": 0.692,
|
2094 |
+
"step": 1470
|
2095 |
+
}
|
2096 |
+
],
|
2097 |
+
"logging_steps": 5,
|
2098 |
+
"max_steps": 7350,
|
2099 |
+
"num_input_tokens_seen": 0,
|
2100 |
+
"num_train_epochs": 15,
|
2101 |
+
"save_steps": 500,
|
2102 |
+
"stateful_callbacks": {
|
2103 |
+
"TrainerControl": {
|
2104 |
+
"args": {
|
2105 |
+
"should_epoch_stop": false,
|
2106 |
+
"should_evaluate": false,
|
2107 |
+
"should_log": false,
|
2108 |
+
"should_save": true,
|
2109 |
+
"should_training_stop": false
|
2110 |
+
},
|
2111 |
+
"attributes": {}
|
2112 |
+
}
|
2113 |
+
},
|
2114 |
+
"total_flos": 4.49037208009728e+18,
|
2115 |
+
"train_batch_size": 24,
|
2116 |
+
"trial_name": null,
|
2117 |
+
"trial_params": null
|
2118 |
+
}
|
checkpoint-1470/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3c8330f8f1eea2830a3d99e288103b577fce3ec104205e265b5fe5492787aa5a
|
3 |
+
size 5176
|