Upload folder using huggingface_hub
Browse files- config.json +143 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1825 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,143 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"adapter_attn_dim": null,
|
| 4 |
+
"adapter_kernel_size": 3,
|
| 5 |
+
"adapter_stride": 2,
|
| 6 |
+
"add_adapter": false,
|
| 7 |
+
"apply_spec_augment": true,
|
| 8 |
+
"architectures": [
|
| 9 |
+
"Wav2Vec2ForSequenceClassification"
|
| 10 |
+
],
|
| 11 |
+
"attention_dropout": 0.1,
|
| 12 |
+
"bos_token_id": 1,
|
| 13 |
+
"classifier_proj_size": 256,
|
| 14 |
+
"codevector_dim": 256,
|
| 15 |
+
"contrastive_logits_temperature": 0.1,
|
| 16 |
+
"conv_bias": false,
|
| 17 |
+
"conv_dim": [
|
| 18 |
+
512,
|
| 19 |
+
512,
|
| 20 |
+
512,
|
| 21 |
+
512,
|
| 22 |
+
512,
|
| 23 |
+
512,
|
| 24 |
+
512
|
| 25 |
+
],
|
| 26 |
+
"conv_kernel": [
|
| 27 |
+
10,
|
| 28 |
+
3,
|
| 29 |
+
3,
|
| 30 |
+
3,
|
| 31 |
+
3,
|
| 32 |
+
2,
|
| 33 |
+
2
|
| 34 |
+
],
|
| 35 |
+
"conv_stride": [
|
| 36 |
+
5,
|
| 37 |
+
2,
|
| 38 |
+
2,
|
| 39 |
+
2,
|
| 40 |
+
2,
|
| 41 |
+
2,
|
| 42 |
+
2
|
| 43 |
+
],
|
| 44 |
+
"ctc_loss_reduction": "sum",
|
| 45 |
+
"ctc_zero_infinity": false,
|
| 46 |
+
"diversity_loss_weight": 0.1,
|
| 47 |
+
"do_stable_layer_norm": false,
|
| 48 |
+
"eos_token_id": 2,
|
| 49 |
+
"feat_extract_activation": "gelu",
|
| 50 |
+
"feat_extract_norm": "group",
|
| 51 |
+
"feat_proj_dropout": 0.1,
|
| 52 |
+
"feat_quantizer_dropout": 0.0,
|
| 53 |
+
"final_dropout": 0.0,
|
| 54 |
+
"freeze_feat_extract_train": true,
|
| 55 |
+
"hidden_act": "gelu",
|
| 56 |
+
"hidden_dropout": 0.1,
|
| 57 |
+
"hidden_size": 768,
|
| 58 |
+
"id2label": {
|
| 59 |
+
"0": "LABEL_0",
|
| 60 |
+
"1": "LABEL_1",
|
| 61 |
+
"2": "LABEL_2",
|
| 62 |
+
"3": "LABEL_3",
|
| 63 |
+
"4": "LABEL_4",
|
| 64 |
+
"5": "LABEL_5",
|
| 65 |
+
"6": "LABEL_6",
|
| 66 |
+
"7": "LABEL_7",
|
| 67 |
+
"8": "LABEL_8",
|
| 68 |
+
"9": "LABEL_9",
|
| 69 |
+
"10": "LABEL_10"
|
| 70 |
+
},
|
| 71 |
+
"initializer_range": 0.02,
|
| 72 |
+
"intermediate_size": 3072,
|
| 73 |
+
"label2id": {
|
| 74 |
+
"LABEL_0": 0,
|
| 75 |
+
"LABEL_1": 1,
|
| 76 |
+
"LABEL_10": 10,
|
| 77 |
+
"LABEL_2": 2,
|
| 78 |
+
"LABEL_3": 3,
|
| 79 |
+
"LABEL_4": 4,
|
| 80 |
+
"LABEL_5": 5,
|
| 81 |
+
"LABEL_6": 6,
|
| 82 |
+
"LABEL_7": 7,
|
| 83 |
+
"LABEL_8": 8,
|
| 84 |
+
"LABEL_9": 9
|
| 85 |
+
},
|
| 86 |
+
"layer_norm_eps": 1e-05,
|
| 87 |
+
"layerdrop": 0.0,
|
| 88 |
+
"mask_channel_length": 10,
|
| 89 |
+
"mask_channel_min_space": 1,
|
| 90 |
+
"mask_channel_other": 0.0,
|
| 91 |
+
"mask_channel_prob": 0.0,
|
| 92 |
+
"mask_channel_selection": "static",
|
| 93 |
+
"mask_feature_length": 10,
|
| 94 |
+
"mask_feature_min_masks": 0,
|
| 95 |
+
"mask_feature_prob": 0.0,
|
| 96 |
+
"mask_time_length": 10,
|
| 97 |
+
"mask_time_min_masks": 2,
|
| 98 |
+
"mask_time_min_space": 1,
|
| 99 |
+
"mask_time_other": 0.0,
|
| 100 |
+
"mask_time_prob": 0.05,
|
| 101 |
+
"mask_time_selection": "static",
|
| 102 |
+
"model_type": "wav2vec2",
|
| 103 |
+
"no_mask_channel_overlap": false,
|
| 104 |
+
"no_mask_time_overlap": false,
|
| 105 |
+
"num_adapter_layers": 3,
|
| 106 |
+
"num_attention_heads": 12,
|
| 107 |
+
"num_codevector_groups": 2,
|
| 108 |
+
"num_codevectors_per_group": 320,
|
| 109 |
+
"num_conv_pos_embedding_groups": 16,
|
| 110 |
+
"num_conv_pos_embeddings": 128,
|
| 111 |
+
"num_feat_extract_layers": 7,
|
| 112 |
+
"num_hidden_layers": 12,
|
| 113 |
+
"num_negatives": 100,
|
| 114 |
+
"output_hidden_size": 768,
|
| 115 |
+
"pad_token_id": 0,
|
| 116 |
+
"proj_codevector_dim": 256,
|
| 117 |
+
"tdnn_dilation": [
|
| 118 |
+
1,
|
| 119 |
+
2,
|
| 120 |
+
3,
|
| 121 |
+
1,
|
| 122 |
+
1
|
| 123 |
+
],
|
| 124 |
+
"tdnn_dim": [
|
| 125 |
+
512,
|
| 126 |
+
512,
|
| 127 |
+
512,
|
| 128 |
+
512,
|
| 129 |
+
1500
|
| 130 |
+
],
|
| 131 |
+
"tdnn_kernel": [
|
| 132 |
+
5,
|
| 133 |
+
3,
|
| 134 |
+
3,
|
| 135 |
+
1,
|
| 136 |
+
1
|
| 137 |
+
],
|
| 138 |
+
"torch_dtype": "float32",
|
| 139 |
+
"transformers_version": "4.50.0",
|
| 140 |
+
"use_weighted_layer_sum": false,
|
| 141 |
+
"vocab_size": 32,
|
| 142 |
+
"xvector_output_dim": 512
|
| 143 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7b8c22d6220c0fe2b2de3f57d272ce4b5cb98bda61bc4ac7517c15881a305922
|
| 3 |
+
size 378311620
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7d86c5907323c06b1608134d45a800b9d612452f34092afbde0a84ca5d84afc8
|
| 3 |
+
size 4771843
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b09097c1953912c5807314e9f0a1a4766cc4f0aaf84c36d549545d82ce75f791
|
| 3 |
+
size 14308
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0617efbfbb7fc93b69f89ca8af67f24948e1b623a640583096a01c1e14d1a779
|
| 3 |
+
size 988
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:36e97e9eaaa508c9e09406ab76f0eef3b7e362bc390058bc654ab1cf816f9cfc
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1825 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 5049,
|
| 3 |
+
"best_metric": 0.36363636363636365,
|
| 4 |
+
"best_model_checkpoint": "./wav2vec2_accent_classification_exp2/checkpoint-5049",
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 5049,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011883541295306001,
|
| 14 |
+
"grad_norm": 5.2643561363220215,
|
| 15 |
+
"learning_rate": 3.7623762376237625e-06,
|
| 16 |
+
"loss": 1.8575,
|
| 17 |
+
"step": 20
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023767082590612002,
|
| 21 |
+
"grad_norm": 6.11646032333374,
|
| 22 |
+
"learning_rate": 7.722772277227722e-06,
|
| 23 |
+
"loss": 1.8142,
|
| 24 |
+
"step": 40
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.035650623885918005,
|
| 28 |
+
"grad_norm": 6.610026836395264,
|
| 29 |
+
"learning_rate": 1.1683168316831684e-05,
|
| 30 |
+
"loss": 1.9892,
|
| 31 |
+
"step": 60
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047534165181224004,
|
| 35 |
+
"grad_norm": 4.259169578552246,
|
| 36 |
+
"learning_rate": 1.5643564356435644e-05,
|
| 37 |
+
"loss": 2.0514,
|
| 38 |
+
"step": 80
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.059417706476530004,
|
| 42 |
+
"grad_norm": 6.260364055633545,
|
| 43 |
+
"learning_rate": 1.9603960396039607e-05,
|
| 44 |
+
"loss": 1.9523,
|
| 45 |
+
"step": 100
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.07130124777183601,
|
| 49 |
+
"grad_norm": 3.329735279083252,
|
| 50 |
+
"learning_rate": 2.3564356435643563e-05,
|
| 51 |
+
"loss": 1.6386,
|
| 52 |
+
"step": 120
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.08318478906714201,
|
| 56 |
+
"grad_norm": 4.184360027313232,
|
| 57 |
+
"learning_rate": 2.7524752475247527e-05,
|
| 58 |
+
"loss": 1.9444,
|
| 59 |
+
"step": 140
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.09506833036244801,
|
| 63 |
+
"grad_norm": 5.585748672485352,
|
| 64 |
+
"learning_rate": 3.1485148514851487e-05,
|
| 65 |
+
"loss": 1.8409,
|
| 66 |
+
"step": 160
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.10695187165775401,
|
| 70 |
+
"grad_norm": 3.903047561645508,
|
| 71 |
+
"learning_rate": 3.5445544554455446e-05,
|
| 72 |
+
"loss": 1.5714,
|
| 73 |
+
"step": 180
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.11883541295306001,
|
| 77 |
+
"grad_norm": 4.687364101409912,
|
| 78 |
+
"learning_rate": 3.9405940594059406e-05,
|
| 79 |
+
"loss": 1.6461,
|
| 80 |
+
"step": 200
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.13071895424836602,
|
| 84 |
+
"grad_norm": 5.168088436126709,
|
| 85 |
+
"learning_rate": 4.3366336633663366e-05,
|
| 86 |
+
"loss": 1.7803,
|
| 87 |
+
"step": 220
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.14260249554367202,
|
| 91 |
+
"grad_norm": 7.197409152984619,
|
| 92 |
+
"learning_rate": 4.7326732673267326e-05,
|
| 93 |
+
"loss": 1.6711,
|
| 94 |
+
"step": 240
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.15448603683897802,
|
| 98 |
+
"grad_norm": 4.677839279174805,
|
| 99 |
+
"learning_rate": 5.128712871287129e-05,
|
| 100 |
+
"loss": 1.8529,
|
| 101 |
+
"step": 260
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.16636957813428402,
|
| 105 |
+
"grad_norm": 5.7448649406433105,
|
| 106 |
+
"learning_rate": 5.524752475247525e-05,
|
| 107 |
+
"loss": 2.1005,
|
| 108 |
+
"step": 280
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.17825311942959002,
|
| 112 |
+
"grad_norm": 8.236382484436035,
|
| 113 |
+
"learning_rate": 5.9207920792079205e-05,
|
| 114 |
+
"loss": 2.3531,
|
| 115 |
+
"step": 300
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.19013666072489602,
|
| 119 |
+
"grad_norm": 4.243819713592529,
|
| 120 |
+
"learning_rate": 6.316831683168317e-05,
|
| 121 |
+
"loss": 1.8971,
|
| 122 |
+
"step": 320
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.20202020202020202,
|
| 126 |
+
"grad_norm": 5.6959381103515625,
|
| 127 |
+
"learning_rate": 6.712871287128713e-05,
|
| 128 |
+
"loss": 1.9127,
|
| 129 |
+
"step": 340
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.21390374331550802,
|
| 133 |
+
"grad_norm": 5.795928001403809,
|
| 134 |
+
"learning_rate": 7.108910891089109e-05,
|
| 135 |
+
"loss": 1.9464,
|
| 136 |
+
"step": 360
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.22578728461081402,
|
| 140 |
+
"grad_norm": 7.922089099884033,
|
| 141 |
+
"learning_rate": 7.504950495049505e-05,
|
| 142 |
+
"loss": 1.9045,
|
| 143 |
+
"step": 380
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.23767082590612001,
|
| 147 |
+
"grad_norm": 7.137391567230225,
|
| 148 |
+
"learning_rate": 7.900990099009902e-05,
|
| 149 |
+
"loss": 1.4941,
|
| 150 |
+
"step": 400
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.24955436720142601,
|
| 154 |
+
"grad_norm": 5.790843486785889,
|
| 155 |
+
"learning_rate": 8.297029702970297e-05,
|
| 156 |
+
"loss": 1.6947,
|
| 157 |
+
"step": 420
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.26143790849673204,
|
| 161 |
+
"grad_norm": 6.460991382598877,
|
| 162 |
+
"learning_rate": 8.693069306930694e-05,
|
| 163 |
+
"loss": 1.751,
|
| 164 |
+
"step": 440
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.27332144979203804,
|
| 168 |
+
"grad_norm": 6.993760585784912,
|
| 169 |
+
"learning_rate": 9.06930693069307e-05,
|
| 170 |
+
"loss": 2.2948,
|
| 171 |
+
"step": 460
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.28520499108734404,
|
| 175 |
+
"grad_norm": 6.212869644165039,
|
| 176 |
+
"learning_rate": 9.465346534653465e-05,
|
| 177 |
+
"loss": 1.8421,
|
| 178 |
+
"step": 480
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.29708853238265004,
|
| 182 |
+
"grad_norm": 6.37160587310791,
|
| 183 |
+
"learning_rate": 9.861386138613862e-05,
|
| 184 |
+
"loss": 1.8579,
|
| 185 |
+
"step": 500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.30897207367795604,
|
| 189 |
+
"grad_norm": 6.82785701751709,
|
| 190 |
+
"learning_rate": 9.971390845070423e-05,
|
| 191 |
+
"loss": 1.6312,
|
| 192 |
+
"step": 520
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.32085561497326204,
|
| 196 |
+
"grad_norm": 5.461722373962402,
|
| 197 |
+
"learning_rate": 9.929577464788733e-05,
|
| 198 |
+
"loss": 1.4322,
|
| 199 |
+
"step": 540
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.33273915626856804,
|
| 203 |
+
"grad_norm": 11.987320899963379,
|
| 204 |
+
"learning_rate": 9.88556338028169e-05,
|
| 205 |
+
"loss": 2.1953,
|
| 206 |
+
"step": 560
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.34462269756387404,
|
| 210 |
+
"grad_norm": 11.54291820526123,
|
| 211 |
+
"learning_rate": 9.841549295774648e-05,
|
| 212 |
+
"loss": 2.2474,
|
| 213 |
+
"step": 580
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.35650623885918004,
|
| 217 |
+
"grad_norm": 3.708719491958618,
|
| 218 |
+
"learning_rate": 9.797535211267606e-05,
|
| 219 |
+
"loss": 1.7945,
|
| 220 |
+
"step": 600
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.36838978015448604,
|
| 224 |
+
"grad_norm": 3.2041361331939697,
|
| 225 |
+
"learning_rate": 9.753521126760564e-05,
|
| 226 |
+
"loss": 1.8463,
|
| 227 |
+
"step": 620
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.38027332144979203,
|
| 231 |
+
"grad_norm": 11.988749504089355,
|
| 232 |
+
"learning_rate": 9.709507042253521e-05,
|
| 233 |
+
"loss": 1.8383,
|
| 234 |
+
"step": 640
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.39215686274509803,
|
| 238 |
+
"grad_norm": 2.1846208572387695,
|
| 239 |
+
"learning_rate": 9.66549295774648e-05,
|
| 240 |
+
"loss": 1.5908,
|
| 241 |
+
"step": 660
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.40404040404040403,
|
| 245 |
+
"grad_norm": 6.514430046081543,
|
| 246 |
+
"learning_rate": 9.621478873239436e-05,
|
| 247 |
+
"loss": 2.0864,
|
| 248 |
+
"step": 680
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.41592394533571003,
|
| 252 |
+
"grad_norm": 4.143547534942627,
|
| 253 |
+
"learning_rate": 9.577464788732394e-05,
|
| 254 |
+
"loss": 1.8506,
|
| 255 |
+
"step": 700
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.42780748663101603,
|
| 259 |
+
"grad_norm": 5.5958781242370605,
|
| 260 |
+
"learning_rate": 9.533450704225353e-05,
|
| 261 |
+
"loss": 2.488,
|
| 262 |
+
"step": 720
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.43969102792632203,
|
| 266 |
+
"grad_norm": 3.9092047214508057,
|
| 267 |
+
"learning_rate": 9.489436619718311e-05,
|
| 268 |
+
"loss": 1.7667,
|
| 269 |
+
"step": 740
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.45157456922162803,
|
| 273 |
+
"grad_norm": 12.434467315673828,
|
| 274 |
+
"learning_rate": 9.445422535211268e-05,
|
| 275 |
+
"loss": 2.0171,
|
| 276 |
+
"step": 760
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.46345811051693403,
|
| 280 |
+
"grad_norm": 5.3435468673706055,
|
| 281 |
+
"learning_rate": 9.401408450704226e-05,
|
| 282 |
+
"loss": 2.1507,
|
| 283 |
+
"step": 780
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.47534165181224003,
|
| 287 |
+
"grad_norm": 6.318358421325684,
|
| 288 |
+
"learning_rate": 9.357394366197183e-05,
|
| 289 |
+
"loss": 1.7404,
|
| 290 |
+
"step": 800
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.48722519310754603,
|
| 294 |
+
"grad_norm": 6.812596797943115,
|
| 295 |
+
"learning_rate": 9.313380281690142e-05,
|
| 296 |
+
"loss": 2.4855,
|
| 297 |
+
"step": 820
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.49910873440285203,
|
| 301 |
+
"grad_norm": 4.214794158935547,
|
| 302 |
+
"learning_rate": 9.269366197183099e-05,
|
| 303 |
+
"loss": 1.8801,
|
| 304 |
+
"step": 840
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.5109922756981581,
|
| 308 |
+
"grad_norm": 7.025059700012207,
|
| 309 |
+
"learning_rate": 9.225352112676057e-05,
|
| 310 |
+
"loss": 1.7031,
|
| 311 |
+
"step": 860
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.5228758169934641,
|
| 315 |
+
"grad_norm": 5.63538932800293,
|
| 316 |
+
"learning_rate": 9.181338028169014e-05,
|
| 317 |
+
"loss": 1.9399,
|
| 318 |
+
"step": 880
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.5347593582887701,
|
| 322 |
+
"grad_norm": 3.4773788452148438,
|
| 323 |
+
"learning_rate": 9.137323943661972e-05,
|
| 324 |
+
"loss": 1.7851,
|
| 325 |
+
"step": 900
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.5466428995840761,
|
| 329 |
+
"grad_norm": 6.2494025230407715,
|
| 330 |
+
"learning_rate": 9.09330985915493e-05,
|
| 331 |
+
"loss": 2.0177,
|
| 332 |
+
"step": 920
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.5585264408793821,
|
| 336 |
+
"grad_norm": 7.094064235687256,
|
| 337 |
+
"learning_rate": 9.049295774647887e-05,
|
| 338 |
+
"loss": 1.688,
|
| 339 |
+
"step": 940
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.5704099821746881,
|
| 343 |
+
"grad_norm": 9.590660095214844,
|
| 344 |
+
"learning_rate": 9.005281690140845e-05,
|
| 345 |
+
"loss": 1.9148,
|
| 346 |
+
"step": 960
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.5822935234699941,
|
| 350 |
+
"grad_norm": 5.741421222686768,
|
| 351 |
+
"learning_rate": 8.961267605633804e-05,
|
| 352 |
+
"loss": 1.8483,
|
| 353 |
+
"step": 980
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.5941770647653001,
|
| 357 |
+
"grad_norm": 2.007276773452759,
|
| 358 |
+
"learning_rate": 8.91725352112676e-05,
|
| 359 |
+
"loss": 1.9433,
|
| 360 |
+
"step": 1000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.6060606060606061,
|
| 364 |
+
"grad_norm": 7.266458511352539,
|
| 365 |
+
"learning_rate": 8.873239436619719e-05,
|
| 366 |
+
"loss": 2.2806,
|
| 367 |
+
"step": 1020
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.6179441473559121,
|
| 371 |
+
"grad_norm": 6.307204723358154,
|
| 372 |
+
"learning_rate": 8.829225352112677e-05,
|
| 373 |
+
"loss": 2.1339,
|
| 374 |
+
"step": 1040
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.6298276886512181,
|
| 378 |
+
"grad_norm": 4.725139617919922,
|
| 379 |
+
"learning_rate": 8.785211267605634e-05,
|
| 380 |
+
"loss": 1.8299,
|
| 381 |
+
"step": 1060
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.6417112299465241,
|
| 385 |
+
"grad_norm": 5.144646644592285,
|
| 386 |
+
"learning_rate": 8.741197183098592e-05,
|
| 387 |
+
"loss": 1.604,
|
| 388 |
+
"step": 1080
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.6535947712418301,
|
| 392 |
+
"grad_norm": 5.759939670562744,
|
| 393 |
+
"learning_rate": 8.697183098591549e-05,
|
| 394 |
+
"loss": 1.8779,
|
| 395 |
+
"step": 1100
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.6654783125371361,
|
| 399 |
+
"grad_norm": 6.637747287750244,
|
| 400 |
+
"learning_rate": 8.653169014084508e-05,
|
| 401 |
+
"loss": 1.7935,
|
| 402 |
+
"step": 1120
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.6773618538324421,
|
| 406 |
+
"grad_norm": 4.596299171447754,
|
| 407 |
+
"learning_rate": 8.609154929577465e-05,
|
| 408 |
+
"loss": 2.0089,
|
| 409 |
+
"step": 1140
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.6892453951277481,
|
| 413 |
+
"grad_norm": 5.862588882446289,
|
| 414 |
+
"learning_rate": 8.565140845070423e-05,
|
| 415 |
+
"loss": 1.8758,
|
| 416 |
+
"step": 1160
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.7011289364230541,
|
| 420 |
+
"grad_norm": 6.179210186004639,
|
| 421 |
+
"learning_rate": 8.52112676056338e-05,
|
| 422 |
+
"loss": 1.8561,
|
| 423 |
+
"step": 1180
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.7130124777183601,
|
| 427 |
+
"grad_norm": 5.406636714935303,
|
| 428 |
+
"learning_rate": 8.477112676056338e-05,
|
| 429 |
+
"loss": 2.1628,
|
| 430 |
+
"step": 1200
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.7248960190136661,
|
| 434 |
+
"grad_norm": 5.283446788787842,
|
| 435 |
+
"learning_rate": 8.433098591549296e-05,
|
| 436 |
+
"loss": 1.5859,
|
| 437 |
+
"step": 1220
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.7367795603089721,
|
| 441 |
+
"grad_norm": 4.130381107330322,
|
| 442 |
+
"learning_rate": 8.389084507042255e-05,
|
| 443 |
+
"loss": 1.7202,
|
| 444 |
+
"step": 1240
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.7486631016042781,
|
| 448 |
+
"grad_norm": 7.399511814117432,
|
| 449 |
+
"learning_rate": 8.345070422535211e-05,
|
| 450 |
+
"loss": 1.8998,
|
| 451 |
+
"step": 1260
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.7605466428995841,
|
| 455 |
+
"grad_norm": 9.544116020202637,
|
| 456 |
+
"learning_rate": 8.30105633802817e-05,
|
| 457 |
+
"loss": 2.1312,
|
| 458 |
+
"step": 1280
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.7724301841948901,
|
| 462 |
+
"grad_norm": 5.536484241485596,
|
| 463 |
+
"learning_rate": 8.257042253521126e-05,
|
| 464 |
+
"loss": 1.8044,
|
| 465 |
+
"step": 1300
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.7843137254901961,
|
| 469 |
+
"grad_norm": 1.6260414123535156,
|
| 470 |
+
"learning_rate": 8.213028169014086e-05,
|
| 471 |
+
"loss": 1.7572,
|
| 472 |
+
"step": 1320
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.7961972667855021,
|
| 476 |
+
"grad_norm": 4.77047872543335,
|
| 477 |
+
"learning_rate": 8.169014084507043e-05,
|
| 478 |
+
"loss": 1.7056,
|
| 479 |
+
"step": 1340
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.8080808080808081,
|
| 483 |
+
"grad_norm": 8.282206535339355,
|
| 484 |
+
"learning_rate": 8.125000000000001e-05,
|
| 485 |
+
"loss": 2.2542,
|
| 486 |
+
"step": 1360
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.8199643493761141,
|
| 490 |
+
"grad_norm": 4.410986423492432,
|
| 491 |
+
"learning_rate": 8.080985915492958e-05,
|
| 492 |
+
"loss": 1.6738,
|
| 493 |
+
"step": 1380
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.8318478906714201,
|
| 497 |
+
"grad_norm": 4.1801323890686035,
|
| 498 |
+
"learning_rate": 8.036971830985916e-05,
|
| 499 |
+
"loss": 1.9745,
|
| 500 |
+
"step": 1400
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.8437314319667261,
|
| 504 |
+
"grad_norm": 6.4103875160217285,
|
| 505 |
+
"learning_rate": 7.992957746478874e-05,
|
| 506 |
+
"loss": 1.8702,
|
| 507 |
+
"step": 1420
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.8556149732620321,
|
| 511 |
+
"grad_norm": 7.729459762573242,
|
| 512 |
+
"learning_rate": 7.948943661971831e-05,
|
| 513 |
+
"loss": 1.7677,
|
| 514 |
+
"step": 1440
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.8674985145573381,
|
| 518 |
+
"grad_norm": 4.993863105773926,
|
| 519 |
+
"learning_rate": 7.904929577464789e-05,
|
| 520 |
+
"loss": 1.8287,
|
| 521 |
+
"step": 1460
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.8793820558526441,
|
| 525 |
+
"grad_norm": 7.292968273162842,
|
| 526 |
+
"learning_rate": 7.860915492957746e-05,
|
| 527 |
+
"loss": 1.8698,
|
| 528 |
+
"step": 1480
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.8912655971479501,
|
| 532 |
+
"grad_norm": 8.821915626525879,
|
| 533 |
+
"learning_rate": 7.816901408450704e-05,
|
| 534 |
+
"loss": 1.8932,
|
| 535 |
+
"step": 1500
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.9031491384432561,
|
| 539 |
+
"grad_norm": 6.973383903503418,
|
| 540 |
+
"learning_rate": 7.772887323943662e-05,
|
| 541 |
+
"loss": 1.8421,
|
| 542 |
+
"step": 1520
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.9150326797385621,
|
| 546 |
+
"grad_norm": 4.604738235473633,
|
| 547 |
+
"learning_rate": 7.72887323943662e-05,
|
| 548 |
+
"loss": 1.7423,
|
| 549 |
+
"step": 1540
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.9269162210338681,
|
| 553 |
+
"grad_norm": 13.173360824584961,
|
| 554 |
+
"learning_rate": 7.684859154929577e-05,
|
| 555 |
+
"loss": 2.0015,
|
| 556 |
+
"step": 1560
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.9387997623291741,
|
| 560 |
+
"grad_norm": 8.873441696166992,
|
| 561 |
+
"learning_rate": 7.640845070422536e-05,
|
| 562 |
+
"loss": 2.0531,
|
| 563 |
+
"step": 1580
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.9506833036244801,
|
| 567 |
+
"grad_norm": 5.304136753082275,
|
| 568 |
+
"learning_rate": 7.596830985915492e-05,
|
| 569 |
+
"loss": 1.9365,
|
| 570 |
+
"step": 1600
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.9625668449197861,
|
| 574 |
+
"grad_norm": 6.267118453979492,
|
| 575 |
+
"learning_rate": 7.552816901408452e-05,
|
| 576 |
+
"loss": 1.7111,
|
| 577 |
+
"step": 1620
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.9744503862150921,
|
| 581 |
+
"grad_norm": 8.107028007507324,
|
| 582 |
+
"learning_rate": 7.508802816901409e-05,
|
| 583 |
+
"loss": 1.7125,
|
| 584 |
+
"step": 1640
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.9863339275103981,
|
| 588 |
+
"grad_norm": 4.288895130157471,
|
| 589 |
+
"learning_rate": 7.464788732394367e-05,
|
| 590 |
+
"loss": 1.9968,
|
| 591 |
+
"step": 1660
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.9982174688057041,
|
| 595 |
+
"grad_norm": 6.284876823425293,
|
| 596 |
+
"learning_rate": 7.420774647887324e-05,
|
| 597 |
+
"loss": 1.6101,
|
| 598 |
+
"step": 1680
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.0,
|
| 602 |
+
"eval_accuracy": 0.3333333333333333,
|
| 603 |
+
"eval_loss": 1.8843854665756226,
|
| 604 |
+
"eval_runtime": 105.159,
|
| 605 |
+
"eval_samples_per_second": 2.824,
|
| 606 |
+
"eval_steps_per_second": 2.824,
|
| 607 |
+
"step": 1683
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 1.0101010101010102,
|
| 611 |
+
"grad_norm": 8.582942962646484,
|
| 612 |
+
"learning_rate": 7.376760563380282e-05,
|
| 613 |
+
"loss": 1.7318,
|
| 614 |
+
"step": 1700
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.0219845513963162,
|
| 618 |
+
"grad_norm": 9.720425605773926,
|
| 619 |
+
"learning_rate": 7.33274647887324e-05,
|
| 620 |
+
"loss": 1.5491,
|
| 621 |
+
"step": 1720
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 1.0338680926916222,
|
| 625 |
+
"grad_norm": 7.132287502288818,
|
| 626 |
+
"learning_rate": 7.288732394366198e-05,
|
| 627 |
+
"loss": 1.8271,
|
| 628 |
+
"step": 1740
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 1.0457516339869282,
|
| 632 |
+
"grad_norm": 9.671852111816406,
|
| 633 |
+
"learning_rate": 7.244718309859155e-05,
|
| 634 |
+
"loss": 1.8371,
|
| 635 |
+
"step": 1760
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 1.0576351752822342,
|
| 639 |
+
"grad_norm": 6.084751605987549,
|
| 640 |
+
"learning_rate": 7.200704225352113e-05,
|
| 641 |
+
"loss": 1.8624,
|
| 642 |
+
"step": 1780
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 1.0695187165775402,
|
| 646 |
+
"grad_norm": 0.5046180486679077,
|
| 647 |
+
"learning_rate": 7.15669014084507e-05,
|
| 648 |
+
"loss": 1.5029,
|
| 649 |
+
"step": 1800
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 1.0814022578728462,
|
| 653 |
+
"grad_norm": 6.817294120788574,
|
| 654 |
+
"learning_rate": 7.112676056338028e-05,
|
| 655 |
+
"loss": 1.6382,
|
| 656 |
+
"step": 1820
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 1.0932857991681522,
|
| 660 |
+
"grad_norm": 5.308507919311523,
|
| 661 |
+
"learning_rate": 7.068661971830987e-05,
|
| 662 |
+
"loss": 1.6164,
|
| 663 |
+
"step": 1840
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 1.1051693404634582,
|
| 667 |
+
"grad_norm": 6.489559173583984,
|
| 668 |
+
"learning_rate": 7.024647887323943e-05,
|
| 669 |
+
"loss": 2.1488,
|
| 670 |
+
"step": 1860
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.1170528817587642,
|
| 674 |
+
"grad_norm": 6.588403701782227,
|
| 675 |
+
"learning_rate": 6.980633802816902e-05,
|
| 676 |
+
"loss": 1.5704,
|
| 677 |
+
"step": 1880
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 1.1289364230540702,
|
| 681 |
+
"grad_norm": 20.61360740661621,
|
| 682 |
+
"learning_rate": 6.93661971830986e-05,
|
| 683 |
+
"loss": 1.6548,
|
| 684 |
+
"step": 1900
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 1.1408199643493762,
|
| 688 |
+
"grad_norm": 7.221196174621582,
|
| 689 |
+
"learning_rate": 6.892605633802818e-05,
|
| 690 |
+
"loss": 2.0391,
|
| 691 |
+
"step": 1920
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 1.1527035056446822,
|
| 695 |
+
"grad_norm": 8.964258193969727,
|
| 696 |
+
"learning_rate": 6.848591549295775e-05,
|
| 697 |
+
"loss": 1.3504,
|
| 698 |
+
"step": 1940
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 1.1645870469399882,
|
| 702 |
+
"grad_norm": 9.5341796875,
|
| 703 |
+
"learning_rate": 6.804577464788733e-05,
|
| 704 |
+
"loss": 1.7378,
|
| 705 |
+
"step": 1960
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 1.1764705882352942,
|
| 709 |
+
"grad_norm": 5.866186141967773,
|
| 710 |
+
"learning_rate": 6.76056338028169e-05,
|
| 711 |
+
"loss": 1.6568,
|
| 712 |
+
"step": 1980
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 1.1883541295306002,
|
| 716 |
+
"grad_norm": 8.703054428100586,
|
| 717 |
+
"learning_rate": 6.716549295774648e-05,
|
| 718 |
+
"loss": 2.0636,
|
| 719 |
+
"step": 2000
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 1.2002376708259062,
|
| 723 |
+
"grad_norm": 4.982656002044678,
|
| 724 |
+
"learning_rate": 6.672535211267606e-05,
|
| 725 |
+
"loss": 1.7215,
|
| 726 |
+
"step": 2020
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 1.2121212121212122,
|
| 730 |
+
"grad_norm": 0.16455787420272827,
|
| 731 |
+
"learning_rate": 6.628521126760564e-05,
|
| 732 |
+
"loss": 1.4068,
|
| 733 |
+
"step": 2040
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 1.2240047534165182,
|
| 737 |
+
"grad_norm": 8.127079963684082,
|
| 738 |
+
"learning_rate": 6.584507042253521e-05,
|
| 739 |
+
"loss": 1.321,
|
| 740 |
+
"step": 2060
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 1.2358882947118242,
|
| 744 |
+
"grad_norm": 6.172501564025879,
|
| 745 |
+
"learning_rate": 6.54049295774648e-05,
|
| 746 |
+
"loss": 1.6227,
|
| 747 |
+
"step": 2080
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 1.2477718360071302,
|
| 751 |
+
"grad_norm": 14.086821556091309,
|
| 752 |
+
"learning_rate": 6.496478873239436e-05,
|
| 753 |
+
"loss": 1.964,
|
| 754 |
+
"step": 2100
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 1.2596553773024362,
|
| 758 |
+
"grad_norm": 8.586699485778809,
|
| 759 |
+
"learning_rate": 6.452464788732394e-05,
|
| 760 |
+
"loss": 1.8143,
|
| 761 |
+
"step": 2120
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 1.2715389185977422,
|
| 765 |
+
"grad_norm": 7.064262390136719,
|
| 766 |
+
"learning_rate": 6.408450704225353e-05,
|
| 767 |
+
"loss": 1.5955,
|
| 768 |
+
"step": 2140
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 1.2834224598930482,
|
| 772 |
+
"grad_norm": 7.700599193572998,
|
| 773 |
+
"learning_rate": 6.364436619718311e-05,
|
| 774 |
+
"loss": 1.7971,
|
| 775 |
+
"step": 2160
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 1.2953060011883542,
|
| 779 |
+
"grad_norm": 1.7084336280822754,
|
| 780 |
+
"learning_rate": 6.320422535211268e-05,
|
| 781 |
+
"loss": 1.682,
|
| 782 |
+
"step": 2180
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 1.3071895424836601,
|
| 786 |
+
"grad_norm": 6.063055992126465,
|
| 787 |
+
"learning_rate": 6.276408450704226e-05,
|
| 788 |
+
"loss": 1.8566,
|
| 789 |
+
"step": 2200
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 1.3190730837789661,
|
| 793 |
+
"grad_norm": 0.21841351687908173,
|
| 794 |
+
"learning_rate": 6.232394366197183e-05,
|
| 795 |
+
"loss": 1.3266,
|
| 796 |
+
"step": 2220
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 1.3309566250742721,
|
| 800 |
+
"grad_norm": 5.327434062957764,
|
| 801 |
+
"learning_rate": 6.188380281690142e-05,
|
| 802 |
+
"loss": 1.6329,
|
| 803 |
+
"step": 2240
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 1.3428401663695781,
|
| 807 |
+
"grad_norm": 8.940739631652832,
|
| 808 |
+
"learning_rate": 6.144366197183099e-05,
|
| 809 |
+
"loss": 2.1826,
|
| 810 |
+
"step": 2260
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 1.3547237076648841,
|
| 814 |
+
"grad_norm": 8.249178886413574,
|
| 815 |
+
"learning_rate": 6.1003521126760565e-05,
|
| 816 |
+
"loss": 1.8603,
|
| 817 |
+
"step": 2280
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 1.3666072489601901,
|
| 821 |
+
"grad_norm": 7.17087459564209,
|
| 822 |
+
"learning_rate": 6.056338028169014e-05,
|
| 823 |
+
"loss": 1.4936,
|
| 824 |
+
"step": 2300
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 1.3784907902554961,
|
| 828 |
+
"grad_norm": 9.176311492919922,
|
| 829 |
+
"learning_rate": 6.0123239436619715e-05,
|
| 830 |
+
"loss": 1.7654,
|
| 831 |
+
"step": 2320
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 1.3903743315508021,
|
| 835 |
+
"grad_norm": 7.676440238952637,
|
| 836 |
+
"learning_rate": 5.9683098591549303e-05,
|
| 837 |
+
"loss": 1.881,
|
| 838 |
+
"step": 2340
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 1.4022578728461081,
|
| 842 |
+
"grad_norm": 4.362059593200684,
|
| 843 |
+
"learning_rate": 5.924295774647888e-05,
|
| 844 |
+
"loss": 1.6177,
|
| 845 |
+
"step": 2360
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 1.4141414141414141,
|
| 849 |
+
"grad_norm": 9.416321754455566,
|
| 850 |
+
"learning_rate": 5.8802816901408454e-05,
|
| 851 |
+
"loss": 1.6711,
|
| 852 |
+
"step": 2380
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 1.4260249554367201,
|
| 856 |
+
"grad_norm": 8.262443542480469,
|
| 857 |
+
"learning_rate": 5.836267605633803e-05,
|
| 858 |
+
"loss": 1.7803,
|
| 859 |
+
"step": 2400
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 1.4379084967320261,
|
| 863 |
+
"grad_norm": 14.148792266845703,
|
| 864 |
+
"learning_rate": 5.7922535211267604e-05,
|
| 865 |
+
"loss": 1.814,
|
| 866 |
+
"step": 2420
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.4497920380273321,
|
| 870 |
+
"grad_norm": 5.738523483276367,
|
| 871 |
+
"learning_rate": 5.748239436619719e-05,
|
| 872 |
+
"loss": 1.7585,
|
| 873 |
+
"step": 2440
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 1.4616755793226381,
|
| 877 |
+
"grad_norm": 7.487630367279053,
|
| 878 |
+
"learning_rate": 5.704225352112677e-05,
|
| 879 |
+
"loss": 1.8126,
|
| 880 |
+
"step": 2460
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 1.4735591206179441,
|
| 884 |
+
"grad_norm": 4.155028820037842,
|
| 885 |
+
"learning_rate": 5.660211267605634e-05,
|
| 886 |
+
"loss": 2.0536,
|
| 887 |
+
"step": 2480
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 1.4854426619132501,
|
| 891 |
+
"grad_norm": 5.266646862030029,
|
| 892 |
+
"learning_rate": 5.616197183098592e-05,
|
| 893 |
+
"loss": 1.7383,
|
| 894 |
+
"step": 2500
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 1.4973262032085561,
|
| 898 |
+
"grad_norm": 7.279095649719238,
|
| 899 |
+
"learning_rate": 5.572183098591549e-05,
|
| 900 |
+
"loss": 1.2653,
|
| 901 |
+
"step": 2520
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 1.5092097445038621,
|
| 905 |
+
"grad_norm": 7.2541937828063965,
|
| 906 |
+
"learning_rate": 5.5281690140845074e-05,
|
| 907 |
+
"loss": 1.836,
|
| 908 |
+
"step": 2540
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 1.5210932857991681,
|
| 912 |
+
"grad_norm": 0.5907230377197266,
|
| 913 |
+
"learning_rate": 5.484154929577465e-05,
|
| 914 |
+
"loss": 1.4867,
|
| 915 |
+
"step": 2560
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 1.5329768270944741,
|
| 919 |
+
"grad_norm": 6.433846473693848,
|
| 920 |
+
"learning_rate": 5.4401408450704225e-05,
|
| 921 |
+
"loss": 1.8376,
|
| 922 |
+
"step": 2580
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 1.5448603683897801,
|
| 926 |
+
"grad_norm": 5.489759922027588,
|
| 927 |
+
"learning_rate": 5.39612676056338e-05,
|
| 928 |
+
"loss": 1.4454,
|
| 929 |
+
"step": 2600
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 1.5567439096850861,
|
| 933 |
+
"grad_norm": 10.458736419677734,
|
| 934 |
+
"learning_rate": 5.352112676056338e-05,
|
| 935 |
+
"loss": 1.5792,
|
| 936 |
+
"step": 2620
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 1.5686274509803921,
|
| 940 |
+
"grad_norm": 4.136570453643799,
|
| 941 |
+
"learning_rate": 5.308098591549296e-05,
|
| 942 |
+
"loss": 1.8496,
|
| 943 |
+
"step": 2640
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 1.5805109922756981,
|
| 947 |
+
"grad_norm": 6.3958420753479,
|
| 948 |
+
"learning_rate": 5.264084507042254e-05,
|
| 949 |
+
"loss": 1.5628,
|
| 950 |
+
"step": 2660
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 1.5923945335710041,
|
| 954 |
+
"grad_norm": 5.275241851806641,
|
| 955 |
+
"learning_rate": 5.2200704225352113e-05,
|
| 956 |
+
"loss": 1.8609,
|
| 957 |
+
"step": 2680
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 1.6042780748663101,
|
| 961 |
+
"grad_norm": 7.227801322937012,
|
| 962 |
+
"learning_rate": 5.176056338028169e-05,
|
| 963 |
+
"loss": 1.5597,
|
| 964 |
+
"step": 2700
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 1.6161616161616161,
|
| 968 |
+
"grad_norm": 9.648629188537598,
|
| 969 |
+
"learning_rate": 5.1320422535211264e-05,
|
| 970 |
+
"loss": 1.7436,
|
| 971 |
+
"step": 2720
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 1.6280451574569221,
|
| 975 |
+
"grad_norm": 4.721320629119873,
|
| 976 |
+
"learning_rate": 5.088028169014085e-05,
|
| 977 |
+
"loss": 1.5865,
|
| 978 |
+
"step": 2740
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 1.6399286987522281,
|
| 982 |
+
"grad_norm": 9.709477424621582,
|
| 983 |
+
"learning_rate": 5.044014084507043e-05,
|
| 984 |
+
"loss": 1.5168,
|
| 985 |
+
"step": 2760
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 1.6518122400475341,
|
| 989 |
+
"grad_norm": 4.350772380828857,
|
| 990 |
+
"learning_rate": 5e-05,
|
| 991 |
+
"loss": 1.6576,
|
| 992 |
+
"step": 2780
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 1.6636957813428401,
|
| 996 |
+
"grad_norm": 6.966721534729004,
|
| 997 |
+
"learning_rate": 4.955985915492958e-05,
|
| 998 |
+
"loss": 1.61,
|
| 999 |
+
"step": 2800
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 1.6755793226381461,
|
| 1003 |
+
"grad_norm": 10.847317695617676,
|
| 1004 |
+
"learning_rate": 4.914172535211268e-05,
|
| 1005 |
+
"loss": 1.8436,
|
| 1006 |
+
"step": 2820
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 1.6874628639334521,
|
| 1010 |
+
"grad_norm": 7.786866188049316,
|
| 1011 |
+
"learning_rate": 4.870158450704226e-05,
|
| 1012 |
+
"loss": 1.7339,
|
| 1013 |
+
"step": 2840
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 1.6993464052287581,
|
| 1017 |
+
"grad_norm": 0.1721613109111786,
|
| 1018 |
+
"learning_rate": 4.8261443661971836e-05,
|
| 1019 |
+
"loss": 1.9236,
|
| 1020 |
+
"step": 2860
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 1.7112299465240641,
|
| 1024 |
+
"grad_norm": 11.29725170135498,
|
| 1025 |
+
"learning_rate": 4.782130281690141e-05,
|
| 1026 |
+
"loss": 2.2586,
|
| 1027 |
+
"step": 2880
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 1.7231134878193701,
|
| 1031 |
+
"grad_norm": 8.11932373046875,
|
| 1032 |
+
"learning_rate": 4.7381161971830987e-05,
|
| 1033 |
+
"loss": 1.6086,
|
| 1034 |
+
"step": 2900
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 1.7349970291146761,
|
| 1038 |
+
"grad_norm": 12.539628028869629,
|
| 1039 |
+
"learning_rate": 4.694102112676056e-05,
|
| 1040 |
+
"loss": 1.8505,
|
| 1041 |
+
"step": 2920
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 1.7468805704099821,
|
| 1045 |
+
"grad_norm": 6.338120460510254,
|
| 1046 |
+
"learning_rate": 4.6500880281690144e-05,
|
| 1047 |
+
"loss": 2.1694,
|
| 1048 |
+
"step": 2940
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 1.7587641117052881,
|
| 1052 |
+
"grad_norm": 6.731840133666992,
|
| 1053 |
+
"learning_rate": 4.606073943661972e-05,
|
| 1054 |
+
"loss": 1.3678,
|
| 1055 |
+
"step": 2960
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 1.7706476530005941,
|
| 1059 |
+
"grad_norm": 6.22773551940918,
|
| 1060 |
+
"learning_rate": 4.5620598591549294e-05,
|
| 1061 |
+
"loss": 1.5555,
|
| 1062 |
+
"step": 2980
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 1.7825311942959001,
|
| 1066 |
+
"grad_norm": 9.717683792114258,
|
| 1067 |
+
"learning_rate": 4.5180457746478875e-05,
|
| 1068 |
+
"loss": 1.6688,
|
| 1069 |
+
"step": 3000
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 1.7944147355912063,
|
| 1073 |
+
"grad_norm": 6.945573806762695,
|
| 1074 |
+
"learning_rate": 4.474031690140845e-05,
|
| 1075 |
+
"loss": 1.5924,
|
| 1076 |
+
"step": 3020
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 1.8062982768865123,
|
| 1080 |
+
"grad_norm": 6.494745254516602,
|
| 1081 |
+
"learning_rate": 4.430017605633803e-05,
|
| 1082 |
+
"loss": 1.7125,
|
| 1083 |
+
"step": 3040
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 1.8181818181818183,
|
| 1087 |
+
"grad_norm": 10.390026092529297,
|
| 1088 |
+
"learning_rate": 4.386003521126761e-05,
|
| 1089 |
+
"loss": 1.6515,
|
| 1090 |
+
"step": 3060
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 1.8300653594771243,
|
| 1094 |
+
"grad_norm": 18.333261489868164,
|
| 1095 |
+
"learning_rate": 4.341989436619718e-05,
|
| 1096 |
+
"loss": 1.6783,
|
| 1097 |
+
"step": 3080
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 1.8419489007724303,
|
| 1101 |
+
"grad_norm": 10.250723838806152,
|
| 1102 |
+
"learning_rate": 4.2979753521126764e-05,
|
| 1103 |
+
"loss": 2.1914,
|
| 1104 |
+
"step": 3100
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 1.8538324420677363,
|
| 1108 |
+
"grad_norm": 4.86543607711792,
|
| 1109 |
+
"learning_rate": 4.253961267605634e-05,
|
| 1110 |
+
"loss": 1.4082,
|
| 1111 |
+
"step": 3120
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 1.8657159833630423,
|
| 1115 |
+
"grad_norm": 11.374370574951172,
|
| 1116 |
+
"learning_rate": 4.2099471830985915e-05,
|
| 1117 |
+
"loss": 1.8741,
|
| 1118 |
+
"step": 3140
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 1.8775995246583483,
|
| 1122 |
+
"grad_norm": 0.07825623452663422,
|
| 1123 |
+
"learning_rate": 4.1659330985915496e-05,
|
| 1124 |
+
"loss": 1.2668,
|
| 1125 |
+
"step": 3160
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 1.8894830659536543,
|
| 1129 |
+
"grad_norm": 10.809089660644531,
|
| 1130 |
+
"learning_rate": 4.121919014084507e-05,
|
| 1131 |
+
"loss": 2.0245,
|
| 1132 |
+
"step": 3180
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 1.9013666072489603,
|
| 1136 |
+
"grad_norm": 5.572234153747559,
|
| 1137 |
+
"learning_rate": 4.077904929577465e-05,
|
| 1138 |
+
"loss": 1.3338,
|
| 1139 |
+
"step": 3200
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 1.9132501485442663,
|
| 1143 |
+
"grad_norm": 6.5486159324646,
|
| 1144 |
+
"learning_rate": 4.033890845070423e-05,
|
| 1145 |
+
"loss": 1.4283,
|
| 1146 |
+
"step": 3220
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 1.9251336898395723,
|
| 1150 |
+
"grad_norm": 7.642528057098389,
|
| 1151 |
+
"learning_rate": 3.9898767605633803e-05,
|
| 1152 |
+
"loss": 1.8407,
|
| 1153 |
+
"step": 3240
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 1.9370172311348783,
|
| 1157 |
+
"grad_norm": 4.357378959655762,
|
| 1158 |
+
"learning_rate": 3.9458626760563385e-05,
|
| 1159 |
+
"loss": 1.5338,
|
| 1160 |
+
"step": 3260
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 1.9489007724301843,
|
| 1164 |
+
"grad_norm": 12.393529891967773,
|
| 1165 |
+
"learning_rate": 3.901848591549296e-05,
|
| 1166 |
+
"loss": 1.621,
|
| 1167 |
+
"step": 3280
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 1.9607843137254903,
|
| 1171 |
+
"grad_norm": 4.674510478973389,
|
| 1172 |
+
"learning_rate": 3.857834507042254e-05,
|
| 1173 |
+
"loss": 1.8173,
|
| 1174 |
+
"step": 3300
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 1.9726678550207963,
|
| 1178 |
+
"grad_norm": 6.138635635375977,
|
| 1179 |
+
"learning_rate": 3.813820422535212e-05,
|
| 1180 |
+
"loss": 1.3999,
|
| 1181 |
+
"step": 3320
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 1.9845513963161023,
|
| 1185 |
+
"grad_norm": 9.038865089416504,
|
| 1186 |
+
"learning_rate": 3.769806338028169e-05,
|
| 1187 |
+
"loss": 1.7756,
|
| 1188 |
+
"step": 3340
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 1.9964349376114083,
|
| 1192 |
+
"grad_norm": 6.41239595413208,
|
| 1193 |
+
"learning_rate": 3.725792253521127e-05,
|
| 1194 |
+
"loss": 1.6539,
|
| 1195 |
+
"step": 3360
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 2.0,
|
| 1199 |
+
"eval_accuracy": 0.32996632996632996,
|
| 1200 |
+
"eval_loss": 1.9400980472564697,
|
| 1201 |
+
"eval_runtime": 102.7215,
|
| 1202 |
+
"eval_samples_per_second": 2.891,
|
| 1203 |
+
"eval_steps_per_second": 2.891,
|
| 1204 |
+
"step": 3366
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 2.0083184789067143,
|
| 1208 |
+
"grad_norm": 10.546788215637207,
|
| 1209 |
+
"learning_rate": 3.681778169014084e-05,
|
| 1210 |
+
"loss": 1.7238,
|
| 1211 |
+
"step": 3380
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 2.0202020202020203,
|
| 1215 |
+
"grad_norm": 7.634305953979492,
|
| 1216 |
+
"learning_rate": 3.6377640845070424e-05,
|
| 1217 |
+
"loss": 1.3036,
|
| 1218 |
+
"step": 3400
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 2.0320855614973263,
|
| 1222 |
+
"grad_norm": 4.861294269561768,
|
| 1223 |
+
"learning_rate": 3.59375e-05,
|
| 1224 |
+
"loss": 1.6293,
|
| 1225 |
+
"step": 3420
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 2.0439691027926323,
|
| 1229 |
+
"grad_norm": 7.750176429748535,
|
| 1230 |
+
"learning_rate": 3.5497359154929574e-05,
|
| 1231 |
+
"loss": 1.5016,
|
| 1232 |
+
"step": 3440
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 2.0558526440879383,
|
| 1236 |
+
"grad_norm": 58.997310638427734,
|
| 1237 |
+
"learning_rate": 3.5057218309859156e-05,
|
| 1238 |
+
"loss": 1.5281,
|
| 1239 |
+
"step": 3460
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 2.0677361853832443,
|
| 1243 |
+
"grad_norm": 0.3851041793823242,
|
| 1244 |
+
"learning_rate": 3.461707746478873e-05,
|
| 1245 |
+
"loss": 1.4736,
|
| 1246 |
+
"step": 3480
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 2.0796197266785503,
|
| 1250 |
+
"grad_norm": 7.343233108520508,
|
| 1251 |
+
"learning_rate": 3.417693661971831e-05,
|
| 1252 |
+
"loss": 1.5908,
|
| 1253 |
+
"step": 3500
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 2.0915032679738563,
|
| 1257 |
+
"grad_norm": 26.710235595703125,
|
| 1258 |
+
"learning_rate": 3.373679577464789e-05,
|
| 1259 |
+
"loss": 1.4212,
|
| 1260 |
+
"step": 3520
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 2.1033868092691623,
|
| 1264 |
+
"grad_norm": 4.58949613571167,
|
| 1265 |
+
"learning_rate": 3.329665492957746e-05,
|
| 1266 |
+
"loss": 1.3747,
|
| 1267 |
+
"step": 3540
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 2.1152703505644683,
|
| 1271 |
+
"grad_norm": 4.62943696975708,
|
| 1272 |
+
"learning_rate": 3.2856514084507045e-05,
|
| 1273 |
+
"loss": 1.466,
|
| 1274 |
+
"step": 3560
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 2.1271538918597743,
|
| 1278 |
+
"grad_norm": 5.863485336303711,
|
| 1279 |
+
"learning_rate": 3.243838028169015e-05,
|
| 1280 |
+
"loss": 1.5392,
|
| 1281 |
+
"step": 3580
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 2.1390374331550803,
|
| 1285 |
+
"grad_norm": 10.572720527648926,
|
| 1286 |
+
"learning_rate": 3.199823943661972e-05,
|
| 1287 |
+
"loss": 1.3615,
|
| 1288 |
+
"step": 3600
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 2.1509209744503863,
|
| 1292 |
+
"grad_norm": 12.56803035736084,
|
| 1293 |
+
"learning_rate": 3.15580985915493e-05,
|
| 1294 |
+
"loss": 1.579,
|
| 1295 |
+
"step": 3620
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 2.1628045157456923,
|
| 1299 |
+
"grad_norm": 9.435176849365234,
|
| 1300 |
+
"learning_rate": 3.111795774647888e-05,
|
| 1301 |
+
"loss": 1.6307,
|
| 1302 |
+
"step": 3640
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 2.1746880570409983,
|
| 1306 |
+
"grad_norm": 12.345380783081055,
|
| 1307 |
+
"learning_rate": 3.0677816901408454e-05,
|
| 1308 |
+
"loss": 2.0893,
|
| 1309 |
+
"step": 3660
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"epoch": 2.1865715983363043,
|
| 1313 |
+
"grad_norm": 6.393479347229004,
|
| 1314 |
+
"learning_rate": 3.0237676056338033e-05,
|
| 1315 |
+
"loss": 1.6022,
|
| 1316 |
+
"step": 3680
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 2.1984551396316103,
|
| 1320 |
+
"grad_norm": 23.872873306274414,
|
| 1321 |
+
"learning_rate": 2.9797535211267608e-05,
|
| 1322 |
+
"loss": 1.8693,
|
| 1323 |
+
"step": 3700
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 2.2103386809269163,
|
| 1327 |
+
"grad_norm": 10.536294937133789,
|
| 1328 |
+
"learning_rate": 2.9357394366197183e-05,
|
| 1329 |
+
"loss": 1.4288,
|
| 1330 |
+
"step": 3720
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 2.2222222222222223,
|
| 1334 |
+
"grad_norm": 14.060868263244629,
|
| 1335 |
+
"learning_rate": 2.8917253521126765e-05,
|
| 1336 |
+
"loss": 1.6282,
|
| 1337 |
+
"step": 3740
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 2.2341057635175283,
|
| 1341 |
+
"grad_norm": 6.455112934112549,
|
| 1342 |
+
"learning_rate": 2.847711267605634e-05,
|
| 1343 |
+
"loss": 1.5679,
|
| 1344 |
+
"step": 3760
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 2.2459893048128343,
|
| 1348 |
+
"grad_norm": 9.203805923461914,
|
| 1349 |
+
"learning_rate": 2.8036971830985915e-05,
|
| 1350 |
+
"loss": 1.8668,
|
| 1351 |
+
"step": 3780
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 2.2578728461081403,
|
| 1355 |
+
"grad_norm": 6.185215950012207,
|
| 1356 |
+
"learning_rate": 2.7596830985915493e-05,
|
| 1357 |
+
"loss": 1.4878,
|
| 1358 |
+
"step": 3800
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 2.2697563874034463,
|
| 1362 |
+
"grad_norm": 5.0371246337890625,
|
| 1363 |
+
"learning_rate": 2.7156690140845072e-05,
|
| 1364 |
+
"loss": 1.4435,
|
| 1365 |
+
"step": 3820
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 2.2816399286987523,
|
| 1369 |
+
"grad_norm": 6.076733589172363,
|
| 1370 |
+
"learning_rate": 2.671654929577465e-05,
|
| 1371 |
+
"loss": 1.554,
|
| 1372 |
+
"step": 3840
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 2.2935234699940583,
|
| 1376 |
+
"grad_norm": 41.63713455200195,
|
| 1377 |
+
"learning_rate": 2.6276408450704225e-05,
|
| 1378 |
+
"loss": 1.3157,
|
| 1379 |
+
"step": 3860
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 2.3054070112893643,
|
| 1383 |
+
"grad_norm": 8.987900733947754,
|
| 1384 |
+
"learning_rate": 2.58362676056338e-05,
|
| 1385 |
+
"loss": 1.7254,
|
| 1386 |
+
"step": 3880
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 2.3172905525846703,
|
| 1390 |
+
"grad_norm": 3.579362392425537,
|
| 1391 |
+
"learning_rate": 2.5396126760563382e-05,
|
| 1392 |
+
"loss": 1.3395,
|
| 1393 |
+
"step": 3900
|
| 1394 |
+
},
|
| 1395 |
+
{
|
| 1396 |
+
"epoch": 2.3291740938799763,
|
| 1397 |
+
"grad_norm": 9.25098705291748,
|
| 1398 |
+
"learning_rate": 2.4955985915492957e-05,
|
| 1399 |
+
"loss": 1.7785,
|
| 1400 |
+
"step": 3920
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 2.3410576351752823,
|
| 1404 |
+
"grad_norm": 8.279118537902832,
|
| 1405 |
+
"learning_rate": 2.4515845070422536e-05,
|
| 1406 |
+
"loss": 1.2131,
|
| 1407 |
+
"step": 3940
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 2.3529411764705883,
|
| 1411 |
+
"grad_norm": 5.703310012817383,
|
| 1412 |
+
"learning_rate": 2.4075704225352114e-05,
|
| 1413 |
+
"loss": 1.7614,
|
| 1414 |
+
"step": 3960
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 2.3648247177658943,
|
| 1418 |
+
"grad_norm": 7.569683074951172,
|
| 1419 |
+
"learning_rate": 2.3635563380281693e-05,
|
| 1420 |
+
"loss": 1.7231,
|
| 1421 |
+
"step": 3980
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 2.3767082590612003,
|
| 1425 |
+
"grad_norm": 7.967580795288086,
|
| 1426 |
+
"learning_rate": 2.319542253521127e-05,
|
| 1427 |
+
"loss": 1.4492,
|
| 1428 |
+
"step": 4000
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 2.3885918003565063,
|
| 1432 |
+
"grad_norm": 8.638738632202148,
|
| 1433 |
+
"learning_rate": 2.2755281690140846e-05,
|
| 1434 |
+
"loss": 1.5889,
|
| 1435 |
+
"step": 4020
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 2.4004753416518123,
|
| 1439 |
+
"grad_norm": 8.532800674438477,
|
| 1440 |
+
"learning_rate": 2.2315140845070425e-05,
|
| 1441 |
+
"loss": 1.937,
|
| 1442 |
+
"step": 4040
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 2.4123588829471183,
|
| 1446 |
+
"grad_norm": 8.64955997467041,
|
| 1447 |
+
"learning_rate": 2.1875e-05,
|
| 1448 |
+
"loss": 1.548,
|
| 1449 |
+
"step": 4060
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 2.4242424242424243,
|
| 1453 |
+
"grad_norm": 7.558818340301514,
|
| 1454 |
+
"learning_rate": 2.1434859154929578e-05,
|
| 1455 |
+
"loss": 1.5885,
|
| 1456 |
+
"step": 4080
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 2.4361259655377303,
|
| 1460 |
+
"grad_norm": 7.279077529907227,
|
| 1461 |
+
"learning_rate": 2.0994718309859157e-05,
|
| 1462 |
+
"loss": 1.668,
|
| 1463 |
+
"step": 4100
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 2.4480095068330363,
|
| 1467 |
+
"grad_norm": 8.037296295166016,
|
| 1468 |
+
"learning_rate": 2.0554577464788732e-05,
|
| 1469 |
+
"loss": 1.5021,
|
| 1470 |
+
"step": 4120
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 2.4598930481283423,
|
| 1474 |
+
"grad_norm": 0.8108901977539062,
|
| 1475 |
+
"learning_rate": 2.011443661971831e-05,
|
| 1476 |
+
"loss": 1.4816,
|
| 1477 |
+
"step": 4140
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 2.4717765894236483,
|
| 1481 |
+
"grad_norm": 7.269354343414307,
|
| 1482 |
+
"learning_rate": 1.967429577464789e-05,
|
| 1483 |
+
"loss": 2.1402,
|
| 1484 |
+
"step": 4160
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 2.4836601307189543,
|
| 1488 |
+
"grad_norm": 4.942673683166504,
|
| 1489 |
+
"learning_rate": 1.9234154929577467e-05,
|
| 1490 |
+
"loss": 1.7061,
|
| 1491 |
+
"step": 4180
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 2.4955436720142603,
|
| 1495 |
+
"grad_norm": 6.884216785430908,
|
| 1496 |
+
"learning_rate": 1.8794014084507046e-05,
|
| 1497 |
+
"loss": 1.6234,
|
| 1498 |
+
"step": 4200
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 2.5074272133095663,
|
| 1502 |
+
"grad_norm": 6.9878010749816895,
|
| 1503 |
+
"learning_rate": 1.835387323943662e-05,
|
| 1504 |
+
"loss": 1.5431,
|
| 1505 |
+
"step": 4220
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 2.5193107546048723,
|
| 1509 |
+
"grad_norm": 14.848012924194336,
|
| 1510 |
+
"learning_rate": 1.79137323943662e-05,
|
| 1511 |
+
"loss": 1.5482,
|
| 1512 |
+
"step": 4240
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 2.5311942959001783,
|
| 1516 |
+
"grad_norm": 0.01329022366553545,
|
| 1517 |
+
"learning_rate": 1.7473591549295774e-05,
|
| 1518 |
+
"loss": 1.6875,
|
| 1519 |
+
"step": 4260
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 2.5430778371954843,
|
| 1523 |
+
"grad_norm": 8.419339179992676,
|
| 1524 |
+
"learning_rate": 1.7033450704225353e-05,
|
| 1525 |
+
"loss": 1.5234,
|
| 1526 |
+
"step": 4280
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 2.5549613784907903,
|
| 1530 |
+
"grad_norm": 0.9078832268714905,
|
| 1531 |
+
"learning_rate": 1.6593309859154928e-05,
|
| 1532 |
+
"loss": 1.1912,
|
| 1533 |
+
"step": 4300
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 2.5668449197860963,
|
| 1537 |
+
"grad_norm": 9.658162117004395,
|
| 1538 |
+
"learning_rate": 1.6153169014084506e-05,
|
| 1539 |
+
"loss": 1.5156,
|
| 1540 |
+
"step": 4320
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 2.5787284610814023,
|
| 1544 |
+
"grad_norm": 6.657388687133789,
|
| 1545 |
+
"learning_rate": 1.5713028169014085e-05,
|
| 1546 |
+
"loss": 1.5352,
|
| 1547 |
+
"step": 4340
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 2.5906120023767083,
|
| 1551 |
+
"grad_norm": 7.032742977142334,
|
| 1552 |
+
"learning_rate": 1.5272887323943663e-05,
|
| 1553 |
+
"loss": 1.8476,
|
| 1554 |
+
"step": 4360
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 2.6024955436720143,
|
| 1558 |
+
"grad_norm": 17.715106964111328,
|
| 1559 |
+
"learning_rate": 1.4832746478873242e-05,
|
| 1560 |
+
"loss": 1.2804,
|
| 1561 |
+
"step": 4380
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 2.6143790849673203,
|
| 1565 |
+
"grad_norm": 7.869295120239258,
|
| 1566 |
+
"learning_rate": 1.4392605633802817e-05,
|
| 1567 |
+
"loss": 1.7243,
|
| 1568 |
+
"step": 4400
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 2.6262626262626263,
|
| 1572 |
+
"grad_norm": 17.751676559448242,
|
| 1573 |
+
"learning_rate": 1.3952464788732395e-05,
|
| 1574 |
+
"loss": 1.3411,
|
| 1575 |
+
"step": 4420
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 2.6381461675579323,
|
| 1579 |
+
"grad_norm": 8.237560272216797,
|
| 1580 |
+
"learning_rate": 1.3512323943661972e-05,
|
| 1581 |
+
"loss": 1.4966,
|
| 1582 |
+
"step": 4440
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 2.6500297088532383,
|
| 1586 |
+
"grad_norm": 5.366451740264893,
|
| 1587 |
+
"learning_rate": 1.307218309859155e-05,
|
| 1588 |
+
"loss": 1.5866,
|
| 1589 |
+
"step": 4460
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 2.6619132501485443,
|
| 1593 |
+
"grad_norm": 21.4503173828125,
|
| 1594 |
+
"learning_rate": 1.2632042253521129e-05,
|
| 1595 |
+
"loss": 1.9733,
|
| 1596 |
+
"step": 4480
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 2.6737967914438503,
|
| 1600 |
+
"grad_norm": 8.402605056762695,
|
| 1601 |
+
"learning_rate": 1.2191901408450706e-05,
|
| 1602 |
+
"loss": 1.8483,
|
| 1603 |
+
"step": 4500
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 2.6856803327391563,
|
| 1607 |
+
"grad_norm": 5.882747173309326,
|
| 1608 |
+
"learning_rate": 1.1751760563380282e-05,
|
| 1609 |
+
"loss": 1.4537,
|
| 1610 |
+
"step": 4520
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 2.6975638740344623,
|
| 1614 |
+
"grad_norm": 9.415022850036621,
|
| 1615 |
+
"learning_rate": 1.1311619718309859e-05,
|
| 1616 |
+
"loss": 1.411,
|
| 1617 |
+
"step": 4540
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 2.7094474153297683,
|
| 1621 |
+
"grad_norm": 4.848978042602539,
|
| 1622 |
+
"learning_rate": 1.0871478873239436e-05,
|
| 1623 |
+
"loss": 1.2521,
|
| 1624 |
+
"step": 4560
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 2.7213309566250743,
|
| 1628 |
+
"grad_norm": 8.204416275024414,
|
| 1629 |
+
"learning_rate": 1.0431338028169014e-05,
|
| 1630 |
+
"loss": 1.4406,
|
| 1631 |
+
"step": 4580
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 2.7332144979203803,
|
| 1635 |
+
"grad_norm": 0.049533355981111526,
|
| 1636 |
+
"learning_rate": 9.991197183098593e-06,
|
| 1637 |
+
"loss": 1.477,
|
| 1638 |
+
"step": 4600
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 2.7450980392156863,
|
| 1642 |
+
"grad_norm": 7.250455856323242,
|
| 1643 |
+
"learning_rate": 9.55105633802817e-06,
|
| 1644 |
+
"loss": 1.1354,
|
| 1645 |
+
"step": 4620
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 2.7569815805109923,
|
| 1649 |
+
"grad_norm": 1.7252225875854492,
|
| 1650 |
+
"learning_rate": 9.110915492957748e-06,
|
| 1651 |
+
"loss": 1.1712,
|
| 1652 |
+
"step": 4640
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 2.7688651218062983,
|
| 1656 |
+
"grad_norm": 7.220792770385742,
|
| 1657 |
+
"learning_rate": 8.670774647887325e-06,
|
| 1658 |
+
"loss": 2.1642,
|
| 1659 |
+
"step": 4660
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 2.7807486631016043,
|
| 1663 |
+
"grad_norm": 8.489096641540527,
|
| 1664 |
+
"learning_rate": 8.230633802816901e-06,
|
| 1665 |
+
"loss": 1.4366,
|
| 1666 |
+
"step": 4680
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 2.7926322043969103,
|
| 1670 |
+
"grad_norm": 18.79722023010254,
|
| 1671 |
+
"learning_rate": 7.790492957746478e-06,
|
| 1672 |
+
"loss": 1.3067,
|
| 1673 |
+
"step": 4700
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 2.8045157456922163,
|
| 1677 |
+
"grad_norm": 18.766353607177734,
|
| 1678 |
+
"learning_rate": 7.350352112676057e-06,
|
| 1679 |
+
"loss": 1.9347,
|
| 1680 |
+
"step": 4720
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 2.8163992869875223,
|
| 1684 |
+
"grad_norm": 14.669636726379395,
|
| 1685 |
+
"learning_rate": 6.910211267605634e-06,
|
| 1686 |
+
"loss": 1.8854,
|
| 1687 |
+
"step": 4740
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 2.8282828282828283,
|
| 1691 |
+
"grad_norm": 7.950643062591553,
|
| 1692 |
+
"learning_rate": 6.470070422535211e-06,
|
| 1693 |
+
"loss": 1.564,
|
| 1694 |
+
"step": 4760
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 2.8401663695781343,
|
| 1698 |
+
"grad_norm": 8.613475799560547,
|
| 1699 |
+
"learning_rate": 6.0299295774647895e-06,
|
| 1700 |
+
"loss": 1.3353,
|
| 1701 |
+
"step": 4780
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 2.8520499108734403,
|
| 1705 |
+
"grad_norm": 5.613901138305664,
|
| 1706 |
+
"learning_rate": 5.589788732394366e-06,
|
| 1707 |
+
"loss": 1.159,
|
| 1708 |
+
"step": 4800
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 2.8639334521687463,
|
| 1712 |
+
"grad_norm": 12.659928321838379,
|
| 1713 |
+
"learning_rate": 5.149647887323944e-06,
|
| 1714 |
+
"loss": 1.6703,
|
| 1715 |
+
"step": 4820
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 2.8758169934640523,
|
| 1719 |
+
"grad_norm": 21.733898162841797,
|
| 1720 |
+
"learning_rate": 4.7095070422535215e-06,
|
| 1721 |
+
"loss": 1.5962,
|
| 1722 |
+
"step": 4840
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 2.8877005347593583,
|
| 1726 |
+
"grad_norm": 10.480060577392578,
|
| 1727 |
+
"learning_rate": 4.269366197183098e-06,
|
| 1728 |
+
"loss": 1.4349,
|
| 1729 |
+
"step": 4860
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"epoch": 2.8995840760546643,
|
| 1733 |
+
"grad_norm": 8.418682098388672,
|
| 1734 |
+
"learning_rate": 3.829225352112676e-06,
|
| 1735 |
+
"loss": 1.7244,
|
| 1736 |
+
"step": 4880
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 2.9114676173499703,
|
| 1740 |
+
"grad_norm": 6.427100658416748,
|
| 1741 |
+
"learning_rate": 3.389084507042254e-06,
|
| 1742 |
+
"loss": 1.7937,
|
| 1743 |
+
"step": 4900
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 2.9233511586452763,
|
| 1747 |
+
"grad_norm": 10.744721412658691,
|
| 1748 |
+
"learning_rate": 2.948943661971831e-06,
|
| 1749 |
+
"loss": 1.3575,
|
| 1750 |
+
"step": 4920
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 2.9352346999405823,
|
| 1754 |
+
"grad_norm": 9.011435508728027,
|
| 1755 |
+
"learning_rate": 2.5088028169014087e-06,
|
| 1756 |
+
"loss": 1.3527,
|
| 1757 |
+
"step": 4940
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 2.9471182412358883,
|
| 1761 |
+
"grad_norm": 5.02351713180542,
|
| 1762 |
+
"learning_rate": 2.068661971830986e-06,
|
| 1763 |
+
"loss": 1.9237,
|
| 1764 |
+
"step": 4960
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 2.9590017825311943,
|
| 1768 |
+
"grad_norm": 5.158333778381348,
|
| 1769 |
+
"learning_rate": 1.6285211267605633e-06,
|
| 1770 |
+
"loss": 1.727,
|
| 1771 |
+
"step": 4980
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"epoch": 2.9708853238265003,
|
| 1775 |
+
"grad_norm": 5.139293193817139,
|
| 1776 |
+
"learning_rate": 1.188380281690141e-06,
|
| 1777 |
+
"loss": 1.2477,
|
| 1778 |
+
"step": 5000
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 2.9827688651218063,
|
| 1782 |
+
"grad_norm": 8.442879676818848,
|
| 1783 |
+
"learning_rate": 7.482394366197183e-07,
|
| 1784 |
+
"loss": 1.5919,
|
| 1785 |
+
"step": 5020
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 2.9946524064171123,
|
| 1789 |
+
"grad_norm": 7.629014492034912,
|
| 1790 |
+
"learning_rate": 3.0809859154929577e-07,
|
| 1791 |
+
"loss": 1.6945,
|
| 1792 |
+
"step": 5040
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 3.0,
|
| 1796 |
+
"eval_accuracy": 0.36363636363636365,
|
| 1797 |
+
"eval_loss": 1.9549455642700195,
|
| 1798 |
+
"eval_runtime": 105.2419,
|
| 1799 |
+
"eval_samples_per_second": 2.822,
|
| 1800 |
+
"eval_steps_per_second": 2.822,
|
| 1801 |
+
"step": 5049
|
| 1802 |
+
}
|
| 1803 |
+
],
|
| 1804 |
+
"logging_steps": 20,
|
| 1805 |
+
"max_steps": 5049,
|
| 1806 |
+
"num_input_tokens_seen": 0,
|
| 1807 |
+
"num_train_epochs": 3,
|
| 1808 |
+
"save_steps": 500,
|
| 1809 |
+
"stateful_callbacks": {
|
| 1810 |
+
"TrainerControl": {
|
| 1811 |
+
"args": {
|
| 1812 |
+
"should_epoch_stop": false,
|
| 1813 |
+
"should_evaluate": false,
|
| 1814 |
+
"should_log": false,
|
| 1815 |
+
"should_save": true,
|
| 1816 |
+
"should_training_stop": true
|
| 1817 |
+
},
|
| 1818 |
+
"attributes": {}
|
| 1819 |
+
}
|
| 1820 |
+
},
|
| 1821 |
+
"total_flos": 1.2716768052351475e+18,
|
| 1822 |
+
"train_batch_size": 1,
|
| 1823 |
+
"trial_name": null,
|
| 1824 |
+
"trial_params": null
|
| 1825 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2c0e752b8f37590774467b277e14b9b30c1538a66a00e4b537633e9d5515a618
|
| 3 |
+
size 5368
|