Upload folder using huggingface_hub
Browse files- config.json +68 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- preprocessor_config.json +15 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1800 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,68 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"activation_function": "gelu",
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"WhisperForAudioClassification"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"begin_suppress_tokens": null,
|
| 10 |
+
"bos_token_id": 50257,
|
| 11 |
+
"classifier_proj_size": 256,
|
| 12 |
+
"d_model": 768,
|
| 13 |
+
"decoder_attention_heads": 12,
|
| 14 |
+
"decoder_ffn_dim": 3072,
|
| 15 |
+
"decoder_layerdrop": 0.0,
|
| 16 |
+
"decoder_layers": 12,
|
| 17 |
+
"decoder_start_token_id": 50258,
|
| 18 |
+
"dropout": 0.0,
|
| 19 |
+
"dtype": "float32",
|
| 20 |
+
"encoder_attention_heads": 12,
|
| 21 |
+
"encoder_ffn_dim": 3072,
|
| 22 |
+
"encoder_layerdrop": 0.0,
|
| 23 |
+
"encoder_layers": 12,
|
| 24 |
+
"eos_token_id": 50257,
|
| 25 |
+
"forced_decoder_ids": [
|
| 26 |
+
[
|
| 27 |
+
1,
|
| 28 |
+
50259
|
| 29 |
+
],
|
| 30 |
+
[
|
| 31 |
+
2,
|
| 32 |
+
50359
|
| 33 |
+
],
|
| 34 |
+
[
|
| 35 |
+
3,
|
| 36 |
+
50363
|
| 37 |
+
]
|
| 38 |
+
],
|
| 39 |
+
"id2label": {
|
| 40 |
+
"0": "music",
|
| 41 |
+
"1": "speech"
|
| 42 |
+
},
|
| 43 |
+
"init_std": 0.02,
|
| 44 |
+
"is_encoder_decoder": true,
|
| 45 |
+
"label2id": {
|
| 46 |
+
"music": "0",
|
| 47 |
+
"speech": "1"
|
| 48 |
+
},
|
| 49 |
+
"mask_feature_length": 10,
|
| 50 |
+
"mask_feature_min_masks": 0,
|
| 51 |
+
"mask_feature_prob": 0.0,
|
| 52 |
+
"mask_time_length": 10,
|
| 53 |
+
"mask_time_min_masks": 2,
|
| 54 |
+
"mask_time_prob": 0.05,
|
| 55 |
+
"max_length": null,
|
| 56 |
+
"max_source_positions": 1500,
|
| 57 |
+
"max_target_positions": 448,
|
| 58 |
+
"median_filter_width": 7,
|
| 59 |
+
"model_type": "whisper",
|
| 60 |
+
"num_hidden_layers": 12,
|
| 61 |
+
"num_mel_bins": 80,
|
| 62 |
+
"pad_token_id": 50257,
|
| 63 |
+
"scale_embedding": false,
|
| 64 |
+
"transformers_version": "4.57.1",
|
| 65 |
+
"use_cache": true,
|
| 66 |
+
"use_weighted_layer_sum": false,
|
| 67 |
+
"vocab_size": 51865
|
| 68 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6b0a8fc0e1659b6cb5f8472dc5de46f3e73b09a0b99f211efc0be66fee2c56e0
|
| 3 |
+
size 353426360
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f0b082012f29577066bd98d3b851e85dd285a2c987237be2edcfa4f4edb77c52
|
| 3 |
+
size 697758027
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"chunk_length": 30,
|
| 3 |
+
"dither": 0.0,
|
| 4 |
+
"feature_extractor_type": "WhisperFeatureExtractor",
|
| 5 |
+
"feature_size": 80,
|
| 6 |
+
"hop_length": 160,
|
| 7 |
+
"n_fft": 400,
|
| 8 |
+
"n_samples": 480000,
|
| 9 |
+
"nb_max_frames": 3000,
|
| 10 |
+
"padding_side": "right",
|
| 11 |
+
"padding_value": 0.0,
|
| 12 |
+
"processor_class": "WhisperProcessor",
|
| 13 |
+
"return_attention_mask": false,
|
| 14 |
+
"sampling_rate": 16000
|
| 15 |
+
}
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:18b31b0d8a3a6bc7166b7ad0f2db6e541330af7f57e9e45a277f50ba1642cbb9
|
| 3 |
+
size 14645
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8d2f73728d18f7191909788bfa8b788003f2a18ba13f3cec5eea3061771c720f
|
| 3 |
+
size 1383
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d7f28e91f3775a4ab88567cc5d4f7e058f36f09d1ea036b9269cc0f3cf759110
|
| 3 |
+
size 1465
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1800 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 2500,
|
| 3 |
+
"best_metric": 0.999875,
|
| 4 |
+
"best_model_checkpoint": "./output/checkpoint-2500",
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2500,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0004,
|
| 14 |
+
"grad_norm": 1.8289061784744263,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 0.7116,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.004,
|
| 21 |
+
"grad_norm": 1.8662855625152588,
|
| 22 |
+
"learning_rate": 1.44e-05,
|
| 23 |
+
"loss": 0.6501,
|
| 24 |
+
"step": 10
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.008,
|
| 28 |
+
"grad_norm": 0.9602453112602234,
|
| 29 |
+
"learning_rate": 3.04e-05,
|
| 30 |
+
"loss": 0.2983,
|
| 31 |
+
"step": 20
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.012,
|
| 35 |
+
"grad_norm": 0.017806751653552055,
|
| 36 |
+
"learning_rate": 4.64e-05,
|
| 37 |
+
"loss": 0.0069,
|
| 38 |
+
"step": 30
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.016,
|
| 42 |
+
"grad_norm": 0.005567730404436588,
|
| 43 |
+
"learning_rate": 6.24e-05,
|
| 44 |
+
"loss": 0.0056,
|
| 45 |
+
"step": 40
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.02,
|
| 49 |
+
"grad_norm": 0.0028145266696810722,
|
| 50 |
+
"learning_rate": 7.840000000000001e-05,
|
| 51 |
+
"loss": 0.0083,
|
| 52 |
+
"step": 50
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.024,
|
| 56 |
+
"grad_norm": 0.002064670203253627,
|
| 57 |
+
"learning_rate": 9.44e-05,
|
| 58 |
+
"loss": 0.0001,
|
| 59 |
+
"step": 60
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.028,
|
| 63 |
+
"grad_norm": 0.0016958912601694465,
|
| 64 |
+
"learning_rate": 0.00011040000000000001,
|
| 65 |
+
"loss": 0.0,
|
| 66 |
+
"step": 70
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.032,
|
| 70 |
+
"grad_norm": 0.0036686942912638187,
|
| 71 |
+
"learning_rate": 0.0001264,
|
| 72 |
+
"loss": 0.0447,
|
| 73 |
+
"step": 80
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.036,
|
| 77 |
+
"grad_norm": 8.377481460571289,
|
| 78 |
+
"learning_rate": 0.0001424,
|
| 79 |
+
"loss": 0.0066,
|
| 80 |
+
"step": 90
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.04,
|
| 84 |
+
"grad_norm": 8.64479923248291,
|
| 85 |
+
"learning_rate": 0.00015840000000000003,
|
| 86 |
+
"loss": 0.0222,
|
| 87 |
+
"step": 100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.044,
|
| 91 |
+
"grad_norm": 0.01056673564016819,
|
| 92 |
+
"learning_rate": 0.0001744,
|
| 93 |
+
"loss": 0.0128,
|
| 94 |
+
"step": 110
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.048,
|
| 98 |
+
"grad_norm": 0.007756778970360756,
|
| 99 |
+
"learning_rate": 0.0001904,
|
| 100 |
+
"loss": 0.0306,
|
| 101 |
+
"step": 120
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.052,
|
| 105 |
+
"grad_norm": 0.027605902403593063,
|
| 106 |
+
"learning_rate": 0.0002064,
|
| 107 |
+
"loss": 0.0185,
|
| 108 |
+
"step": 130
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.056,
|
| 112 |
+
"grad_norm": 0.0016576085472479463,
|
| 113 |
+
"learning_rate": 0.00022240000000000004,
|
| 114 |
+
"loss": 0.0001,
|
| 115 |
+
"step": 140
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.06,
|
| 119 |
+
"grad_norm": 16.3161678314209,
|
| 120 |
+
"learning_rate": 0.0002384,
|
| 121 |
+
"loss": 0.0472,
|
| 122 |
+
"step": 150
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.064,
|
| 126 |
+
"grad_norm": 0.385207861661911,
|
| 127 |
+
"learning_rate": 0.0002544,
|
| 128 |
+
"loss": 0.0789,
|
| 129 |
+
"step": 160
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.068,
|
| 133 |
+
"grad_norm": 1.33426833152771,
|
| 134 |
+
"learning_rate": 0.0002704,
|
| 135 |
+
"loss": 0.0577,
|
| 136 |
+
"step": 170
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.072,
|
| 140 |
+
"grad_norm": 0.7534772753715515,
|
| 141 |
+
"learning_rate": 0.0002864,
|
| 142 |
+
"loss": 0.0337,
|
| 143 |
+
"step": 180
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.076,
|
| 147 |
+
"grad_norm": 0.4073263108730316,
|
| 148 |
+
"learning_rate": 0.00030240000000000003,
|
| 149 |
+
"loss": 0.0354,
|
| 150 |
+
"step": 190
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.08,
|
| 154 |
+
"grad_norm": 0.5311892628669739,
|
| 155 |
+
"learning_rate": 0.00031840000000000004,
|
| 156 |
+
"loss": 0.0886,
|
| 157 |
+
"step": 200
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.084,
|
| 161 |
+
"grad_norm": 0.04398029297590256,
|
| 162 |
+
"learning_rate": 0.0003344,
|
| 163 |
+
"loss": 0.0649,
|
| 164 |
+
"step": 210
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.088,
|
| 168 |
+
"grad_norm": 0.02103251777589321,
|
| 169 |
+
"learning_rate": 0.0003504,
|
| 170 |
+
"loss": 0.0122,
|
| 171 |
+
"step": 220
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.092,
|
| 175 |
+
"grad_norm": 0.021729949861764908,
|
| 176 |
+
"learning_rate": 0.0003664,
|
| 177 |
+
"loss": 0.0007,
|
| 178 |
+
"step": 230
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.096,
|
| 182 |
+
"grad_norm": 10.518205642700195,
|
| 183 |
+
"learning_rate": 0.0003824,
|
| 184 |
+
"loss": 0.0542,
|
| 185 |
+
"step": 240
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.1,
|
| 189 |
+
"grad_norm": 0.2691439688205719,
|
| 190 |
+
"learning_rate": 0.00039840000000000003,
|
| 191 |
+
"loss": 0.1148,
|
| 192 |
+
"step": 250
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.104,
|
| 196 |
+
"grad_norm": 0.006621910724788904,
|
| 197 |
+
"learning_rate": 0.00039840000000000003,
|
| 198 |
+
"loss": 0.0033,
|
| 199 |
+
"step": 260
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.108,
|
| 203 |
+
"grad_norm": 0.006356918718665838,
|
| 204 |
+
"learning_rate": 0.00039662222222222224,
|
| 205 |
+
"loss": 0.0142,
|
| 206 |
+
"step": 270
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.112,
|
| 210 |
+
"grad_norm": 0.19780497252941132,
|
| 211 |
+
"learning_rate": 0.0003948444444444445,
|
| 212 |
+
"loss": 0.0532,
|
| 213 |
+
"step": 280
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.116,
|
| 217 |
+
"grad_norm": 0.022000627592206,
|
| 218 |
+
"learning_rate": 0.0003930666666666667,
|
| 219 |
+
"loss": 0.023,
|
| 220 |
+
"step": 290
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.12,
|
| 224 |
+
"grad_norm": 0.021432537585496902,
|
| 225 |
+
"learning_rate": 0.0003912888888888889,
|
| 226 |
+
"loss": 0.0132,
|
| 227 |
+
"step": 300
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.124,
|
| 231 |
+
"grad_norm": 0.45686644315719604,
|
| 232 |
+
"learning_rate": 0.00038951111111111113,
|
| 233 |
+
"loss": 0.0491,
|
| 234 |
+
"step": 310
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.128,
|
| 238 |
+
"grad_norm": 1.6421903371810913,
|
| 239 |
+
"learning_rate": 0.0003877333333333334,
|
| 240 |
+
"loss": 0.034,
|
| 241 |
+
"step": 320
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.132,
|
| 245 |
+
"grad_norm": 0.10952561348676682,
|
| 246 |
+
"learning_rate": 0.0003859555555555556,
|
| 247 |
+
"loss": 0.0945,
|
| 248 |
+
"step": 330
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.136,
|
| 252 |
+
"grad_norm": 0.05203436315059662,
|
| 253 |
+
"learning_rate": 0.0003841777777777778,
|
| 254 |
+
"loss": 0.01,
|
| 255 |
+
"step": 340
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.14,
|
| 259 |
+
"grad_norm": 0.06594817340373993,
|
| 260 |
+
"learning_rate": 0.0003824,
|
| 261 |
+
"loss": 0.0229,
|
| 262 |
+
"step": 350
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.144,
|
| 266 |
+
"grad_norm": 0.03412980958819389,
|
| 267 |
+
"learning_rate": 0.00038062222222222223,
|
| 268 |
+
"loss": 0.0086,
|
| 269 |
+
"step": 360
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.148,
|
| 273 |
+
"grad_norm": 0.022151971235871315,
|
| 274 |
+
"learning_rate": 0.00037884444444444444,
|
| 275 |
+
"loss": 0.0098,
|
| 276 |
+
"step": 370
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.152,
|
| 280 |
+
"grad_norm": 4.759343147277832,
|
| 281 |
+
"learning_rate": 0.00037706666666666665,
|
| 282 |
+
"loss": 0.0014,
|
| 283 |
+
"step": 380
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.156,
|
| 287 |
+
"grad_norm": 0.031497757881879807,
|
| 288 |
+
"learning_rate": 0.0003752888888888889,
|
| 289 |
+
"loss": 0.0455,
|
| 290 |
+
"step": 390
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.16,
|
| 294 |
+
"grad_norm": 0.11943238973617554,
|
| 295 |
+
"learning_rate": 0.0003735111111111111,
|
| 296 |
+
"loss": 0.017,
|
| 297 |
+
"step": 400
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.164,
|
| 301 |
+
"grad_norm": 0.020562870427966118,
|
| 302 |
+
"learning_rate": 0.00037173333333333333,
|
| 303 |
+
"loss": 0.0036,
|
| 304 |
+
"step": 410
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.168,
|
| 308 |
+
"grad_norm": 0.00521384971216321,
|
| 309 |
+
"learning_rate": 0.00036995555555555554,
|
| 310 |
+
"loss": 0.0004,
|
| 311 |
+
"step": 420
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.172,
|
| 315 |
+
"grad_norm": 2.1800899505615234,
|
| 316 |
+
"learning_rate": 0.00036817777777777775,
|
| 317 |
+
"loss": 0.0337,
|
| 318 |
+
"step": 430
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.176,
|
| 322 |
+
"grad_norm": 0.14933565258979797,
|
| 323 |
+
"learning_rate": 0.0003664,
|
| 324 |
+
"loss": 0.02,
|
| 325 |
+
"step": 440
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.18,
|
| 329 |
+
"grad_norm": 1.230485439300537,
|
| 330 |
+
"learning_rate": 0.0003646222222222222,
|
| 331 |
+
"loss": 0.0351,
|
| 332 |
+
"step": 450
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.184,
|
| 336 |
+
"grad_norm": 0.012220942415297031,
|
| 337 |
+
"learning_rate": 0.00036284444444444443,
|
| 338 |
+
"loss": 0.0026,
|
| 339 |
+
"step": 460
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.188,
|
| 343 |
+
"grad_norm": 0.00523759750649333,
|
| 344 |
+
"learning_rate": 0.00036106666666666664,
|
| 345 |
+
"loss": 0.0002,
|
| 346 |
+
"step": 470
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.192,
|
| 350 |
+
"grad_norm": 0.19926713407039642,
|
| 351 |
+
"learning_rate": 0.0003592888888888889,
|
| 352 |
+
"loss": 0.0263,
|
| 353 |
+
"step": 480
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.196,
|
| 357 |
+
"grad_norm": 0.0035884054377675056,
|
| 358 |
+
"learning_rate": 0.0003575111111111111,
|
| 359 |
+
"loss": 0.0267,
|
| 360 |
+
"step": 490
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.2,
|
| 364 |
+
"grad_norm": 0.30148079991340637,
|
| 365 |
+
"learning_rate": 0.0003557333333333333,
|
| 366 |
+
"loss": 0.0446,
|
| 367 |
+
"step": 500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.204,
|
| 371 |
+
"grad_norm": 0.046939630061388016,
|
| 372 |
+
"learning_rate": 0.00035395555555555553,
|
| 373 |
+
"loss": 0.0084,
|
| 374 |
+
"step": 510
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.208,
|
| 378 |
+
"grad_norm": 0.02836071327328682,
|
| 379 |
+
"learning_rate": 0.0003521777777777778,
|
| 380 |
+
"loss": 0.0161,
|
| 381 |
+
"step": 520
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.212,
|
| 385 |
+
"grad_norm": 0.06878186017274857,
|
| 386 |
+
"learning_rate": 0.0003504,
|
| 387 |
+
"loss": 0.0082,
|
| 388 |
+
"step": 530
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.216,
|
| 392 |
+
"grad_norm": 0.14635486900806427,
|
| 393 |
+
"learning_rate": 0.0003486222222222222,
|
| 394 |
+
"loss": 0.033,
|
| 395 |
+
"step": 540
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.22,
|
| 399 |
+
"grad_norm": 0.02621046081185341,
|
| 400 |
+
"learning_rate": 0.0003468444444444445,
|
| 401 |
+
"loss": 0.0234,
|
| 402 |
+
"step": 550
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.224,
|
| 406 |
+
"grad_norm": 0.020144036039710045,
|
| 407 |
+
"learning_rate": 0.0003450666666666667,
|
| 408 |
+
"loss": 0.005,
|
| 409 |
+
"step": 560
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.228,
|
| 413 |
+
"grad_norm": 0.06109397113323212,
|
| 414 |
+
"learning_rate": 0.0003432888888888889,
|
| 415 |
+
"loss": 0.0126,
|
| 416 |
+
"step": 570
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.232,
|
| 420 |
+
"grad_norm": 0.02218439429998398,
|
| 421 |
+
"learning_rate": 0.0003415111111111111,
|
| 422 |
+
"loss": 0.0011,
|
| 423 |
+
"step": 580
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.236,
|
| 427 |
+
"grad_norm": 0.0069372705183923244,
|
| 428 |
+
"learning_rate": 0.00033973333333333337,
|
| 429 |
+
"loss": 0.0003,
|
| 430 |
+
"step": 590
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.24,
|
| 434 |
+
"grad_norm": 0.004556322004646063,
|
| 435 |
+
"learning_rate": 0.0003379555555555556,
|
| 436 |
+
"loss": 0.0001,
|
| 437 |
+
"step": 600
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.244,
|
| 441 |
+
"grad_norm": 0.003253826405853033,
|
| 442 |
+
"learning_rate": 0.0003361777777777778,
|
| 443 |
+
"loss": 0.0001,
|
| 444 |
+
"step": 610
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.248,
|
| 448 |
+
"grad_norm": 0.0028496645390987396,
|
| 449 |
+
"learning_rate": 0.0003344,
|
| 450 |
+
"loss": 0.0001,
|
| 451 |
+
"step": 620
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.252,
|
| 455 |
+
"grad_norm": 0.0023998478427529335,
|
| 456 |
+
"learning_rate": 0.00033262222222222226,
|
| 457 |
+
"loss": 0.0001,
|
| 458 |
+
"step": 630
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.256,
|
| 462 |
+
"grad_norm": 0.0020449627190828323,
|
| 463 |
+
"learning_rate": 0.00033084444444444447,
|
| 464 |
+
"loss": 0.0,
|
| 465 |
+
"step": 640
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.26,
|
| 469 |
+
"grad_norm": 0.13762225210666656,
|
| 470 |
+
"learning_rate": 0.0003290666666666667,
|
| 471 |
+
"loss": 0.0216,
|
| 472 |
+
"step": 650
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.264,
|
| 476 |
+
"grad_norm": 9.54612922668457,
|
| 477 |
+
"learning_rate": 0.0003272888888888889,
|
| 478 |
+
"loss": 0.0634,
|
| 479 |
+
"step": 660
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.268,
|
| 483 |
+
"grad_norm": 2.0493884086608887,
|
| 484 |
+
"learning_rate": 0.00032551111111111115,
|
| 485 |
+
"loss": 0.116,
|
| 486 |
+
"step": 670
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.272,
|
| 490 |
+
"grad_norm": 0.08442521095275879,
|
| 491 |
+
"learning_rate": 0.00032373333333333336,
|
| 492 |
+
"loss": 0.0113,
|
| 493 |
+
"step": 680
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.276,
|
| 497 |
+
"grad_norm": 0.010931574739515781,
|
| 498 |
+
"learning_rate": 0.00032195555555555557,
|
| 499 |
+
"loss": 0.0089,
|
| 500 |
+
"step": 690
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.28,
|
| 504 |
+
"grad_norm": 0.08980958163738251,
|
| 505 |
+
"learning_rate": 0.0003201777777777778,
|
| 506 |
+
"loss": 0.0173,
|
| 507 |
+
"step": 700
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.284,
|
| 511 |
+
"grad_norm": 0.0746162161231041,
|
| 512 |
+
"learning_rate": 0.00031840000000000004,
|
| 513 |
+
"loss": 0.009,
|
| 514 |
+
"step": 710
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.288,
|
| 518 |
+
"grad_norm": 0.022072354331612587,
|
| 519 |
+
"learning_rate": 0.00031662222222222225,
|
| 520 |
+
"loss": 0.0016,
|
| 521 |
+
"step": 720
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.292,
|
| 525 |
+
"grad_norm": 0.007012135349214077,
|
| 526 |
+
"learning_rate": 0.00031484444444444446,
|
| 527 |
+
"loss": 0.0084,
|
| 528 |
+
"step": 730
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.296,
|
| 532 |
+
"grad_norm": 0.009781530126929283,
|
| 533 |
+
"learning_rate": 0.00031306666666666667,
|
| 534 |
+
"loss": 0.0291,
|
| 535 |
+
"step": 740
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.3,
|
| 539 |
+
"grad_norm": 0.03511832281947136,
|
| 540 |
+
"learning_rate": 0.00031128888888888893,
|
| 541 |
+
"loss": 0.0103,
|
| 542 |
+
"step": 750
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.304,
|
| 546 |
+
"grad_norm": 0.03628341481089592,
|
| 547 |
+
"learning_rate": 0.00030951111111111114,
|
| 548 |
+
"loss": 0.0013,
|
| 549 |
+
"step": 760
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.308,
|
| 553 |
+
"grad_norm": 0.008655968122184277,
|
| 554 |
+
"learning_rate": 0.00030773333333333335,
|
| 555 |
+
"loss": 0.0005,
|
| 556 |
+
"step": 770
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.312,
|
| 560 |
+
"grad_norm": 0.015080983750522137,
|
| 561 |
+
"learning_rate": 0.00030595555555555556,
|
| 562 |
+
"loss": 0.0209,
|
| 563 |
+
"step": 780
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.316,
|
| 567 |
+
"grad_norm": 0.025626640766859055,
|
| 568 |
+
"learning_rate": 0.0003041777777777778,
|
| 569 |
+
"loss": 0.0068,
|
| 570 |
+
"step": 790
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.32,
|
| 574 |
+
"grad_norm": 0.02122630551457405,
|
| 575 |
+
"learning_rate": 0.00030240000000000003,
|
| 576 |
+
"loss": 0.0018,
|
| 577 |
+
"step": 800
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.324,
|
| 581 |
+
"grad_norm": 0.019404616206884384,
|
| 582 |
+
"learning_rate": 0.00030062222222222224,
|
| 583 |
+
"loss": 0.0088,
|
| 584 |
+
"step": 810
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.328,
|
| 588 |
+
"grad_norm": 0.037797220051288605,
|
| 589 |
+
"learning_rate": 0.00029884444444444445,
|
| 590 |
+
"loss": 0.0139,
|
| 591 |
+
"step": 820
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.332,
|
| 595 |
+
"grad_norm": 1.059258222579956,
|
| 596 |
+
"learning_rate": 0.0002970666666666667,
|
| 597 |
+
"loss": 0.0051,
|
| 598 |
+
"step": 830
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.336,
|
| 602 |
+
"grad_norm": 2.2363479137420654,
|
| 603 |
+
"learning_rate": 0.0002952888888888889,
|
| 604 |
+
"loss": 0.0125,
|
| 605 |
+
"step": 840
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.34,
|
| 609 |
+
"grad_norm": 0.0257643461227417,
|
| 610 |
+
"learning_rate": 0.00029351111111111113,
|
| 611 |
+
"loss": 0.0121,
|
| 612 |
+
"step": 850
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.344,
|
| 616 |
+
"grad_norm": 0.023524092510342598,
|
| 617 |
+
"learning_rate": 0.00029173333333333334,
|
| 618 |
+
"loss": 0.0011,
|
| 619 |
+
"step": 860
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.348,
|
| 623 |
+
"grad_norm": 0.07415185868740082,
|
| 624 |
+
"learning_rate": 0.0002899555555555556,
|
| 625 |
+
"loss": 0.0008,
|
| 626 |
+
"step": 870
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.352,
|
| 630 |
+
"grad_norm": 0.003202820662409067,
|
| 631 |
+
"learning_rate": 0.0002881777777777778,
|
| 632 |
+
"loss": 0.0011,
|
| 633 |
+
"step": 880
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.356,
|
| 637 |
+
"grad_norm": 0.4437563121318817,
|
| 638 |
+
"learning_rate": 0.0002864,
|
| 639 |
+
"loss": 0.0088,
|
| 640 |
+
"step": 890
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.36,
|
| 644 |
+
"grad_norm": 0.00555839529260993,
|
| 645 |
+
"learning_rate": 0.00028462222222222223,
|
| 646 |
+
"loss": 0.0107,
|
| 647 |
+
"step": 900
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.364,
|
| 651 |
+
"grad_norm": 0.008060933090746403,
|
| 652 |
+
"learning_rate": 0.00028284444444444444,
|
| 653 |
+
"loss": 0.002,
|
| 654 |
+
"step": 910
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.368,
|
| 658 |
+
"grad_norm": 0.011799403466284275,
|
| 659 |
+
"learning_rate": 0.0002810666666666667,
|
| 660 |
+
"loss": 0.0248,
|
| 661 |
+
"step": 920
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.372,
|
| 665 |
+
"grad_norm": 0.042705778032541275,
|
| 666 |
+
"learning_rate": 0.0002792888888888889,
|
| 667 |
+
"loss": 0.0022,
|
| 668 |
+
"step": 930
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.376,
|
| 672 |
+
"grad_norm": 0.005020072218030691,
|
| 673 |
+
"learning_rate": 0.0002775111111111111,
|
| 674 |
+
"loss": 0.0072,
|
| 675 |
+
"step": 940
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.38,
|
| 679 |
+
"grad_norm": 0.002978922799229622,
|
| 680 |
+
"learning_rate": 0.00027573333333333333,
|
| 681 |
+
"loss": 0.0154,
|
| 682 |
+
"step": 950
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.384,
|
| 686 |
+
"grad_norm": 3.051888942718506,
|
| 687 |
+
"learning_rate": 0.00027395555555555554,
|
| 688 |
+
"loss": 0.0065,
|
| 689 |
+
"step": 960
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.388,
|
| 693 |
+
"grad_norm": 0.10347535461187363,
|
| 694 |
+
"learning_rate": 0.00027217777777777775,
|
| 695 |
+
"loss": 0.0033,
|
| 696 |
+
"step": 970
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.392,
|
| 700 |
+
"grad_norm": 0.008343281224370003,
|
| 701 |
+
"learning_rate": 0.0002704,
|
| 702 |
+
"loss": 0.0003,
|
| 703 |
+
"step": 980
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.396,
|
| 707 |
+
"grad_norm": 0.0071385628543794155,
|
| 708 |
+
"learning_rate": 0.0002686222222222222,
|
| 709 |
+
"loss": 0.0003,
|
| 710 |
+
"step": 990
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.4,
|
| 714 |
+
"grad_norm": 0.003130185417830944,
|
| 715 |
+
"learning_rate": 0.00026684444444444443,
|
| 716 |
+
"loss": 0.0062,
|
| 717 |
+
"step": 1000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.404,
|
| 721 |
+
"grad_norm": 0.001496330020017922,
|
| 722 |
+
"learning_rate": 0.00026506666666666664,
|
| 723 |
+
"loss": 0.0005,
|
| 724 |
+
"step": 1010
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.408,
|
| 728 |
+
"grad_norm": 0.8067265748977661,
|
| 729 |
+
"learning_rate": 0.0002632888888888889,
|
| 730 |
+
"loss": 0.0147,
|
| 731 |
+
"step": 1020
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.412,
|
| 735 |
+
"grad_norm": 0.0030374648049473763,
|
| 736 |
+
"learning_rate": 0.0002615111111111111,
|
| 737 |
+
"loss": 0.0,
|
| 738 |
+
"step": 1030
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.416,
|
| 742 |
+
"grad_norm": 0.005266970954835415,
|
| 743 |
+
"learning_rate": 0.0002597333333333333,
|
| 744 |
+
"loss": 0.0018,
|
| 745 |
+
"step": 1040
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.42,
|
| 749 |
+
"grad_norm": 0.015546767972409725,
|
| 750 |
+
"learning_rate": 0.00025795555555555553,
|
| 751 |
+
"loss": 0.0078,
|
| 752 |
+
"step": 1050
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.424,
|
| 756 |
+
"grad_norm": 0.009915663860738277,
|
| 757 |
+
"learning_rate": 0.0002561777777777778,
|
| 758 |
+
"loss": 0.0021,
|
| 759 |
+
"step": 1060
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.428,
|
| 763 |
+
"grad_norm": 0.027595413848757744,
|
| 764 |
+
"learning_rate": 0.0002544,
|
| 765 |
+
"loss": 0.0069,
|
| 766 |
+
"step": 1070
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.432,
|
| 770 |
+
"grad_norm": 0.033019836992025375,
|
| 771 |
+
"learning_rate": 0.0002526222222222222,
|
| 772 |
+
"loss": 0.0231,
|
| 773 |
+
"step": 1080
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.436,
|
| 777 |
+
"grad_norm": 0.06228971853852272,
|
| 778 |
+
"learning_rate": 0.0002508444444444444,
|
| 779 |
+
"loss": 0.0162,
|
| 780 |
+
"step": 1090
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.44,
|
| 784 |
+
"grad_norm": 0.02616875059902668,
|
| 785 |
+
"learning_rate": 0.0002490666666666667,
|
| 786 |
+
"loss": 0.0029,
|
| 787 |
+
"step": 1100
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.444,
|
| 791 |
+
"grad_norm": 0.01255160290747881,
|
| 792 |
+
"learning_rate": 0.0002472888888888889,
|
| 793 |
+
"loss": 0.0014,
|
| 794 |
+
"step": 1110
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.448,
|
| 798 |
+
"grad_norm": 0.007698772940784693,
|
| 799 |
+
"learning_rate": 0.0002455111111111111,
|
| 800 |
+
"loss": 0.0003,
|
| 801 |
+
"step": 1120
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.452,
|
| 805 |
+
"grad_norm": 0.00799358356744051,
|
| 806 |
+
"learning_rate": 0.0002437333333333333,
|
| 807 |
+
"loss": 0.0003,
|
| 808 |
+
"step": 1130
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.456,
|
| 812 |
+
"grad_norm": 0.09039674699306488,
|
| 813 |
+
"learning_rate": 0.00024195555555555558,
|
| 814 |
+
"loss": 0.0003,
|
| 815 |
+
"step": 1140
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.46,
|
| 819 |
+
"grad_norm": 0.0030508041381835938,
|
| 820 |
+
"learning_rate": 0.00024017777777777779,
|
| 821 |
+
"loss": 0.0004,
|
| 822 |
+
"step": 1150
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.464,
|
| 826 |
+
"grad_norm": 0.0018616029992699623,
|
| 827 |
+
"learning_rate": 0.0002384,
|
| 828 |
+
"loss": 0.0053,
|
| 829 |
+
"step": 1160
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.468,
|
| 833 |
+
"grad_norm": 0.0019433458801358938,
|
| 834 |
+
"learning_rate": 0.0002366222222222222,
|
| 835 |
+
"loss": 0.0,
|
| 836 |
+
"step": 1170
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.472,
|
| 840 |
+
"grad_norm": 0.0020802582148462534,
|
| 841 |
+
"learning_rate": 0.00023484444444444447,
|
| 842 |
+
"loss": 0.0018,
|
| 843 |
+
"step": 1180
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.476,
|
| 847 |
+
"grad_norm": 0.2650572657585144,
|
| 848 |
+
"learning_rate": 0.00023306666666666668,
|
| 849 |
+
"loss": 0.0005,
|
| 850 |
+
"step": 1190
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.48,
|
| 854 |
+
"grad_norm": 0.0014364586677402258,
|
| 855 |
+
"learning_rate": 0.00023128888888888889,
|
| 856 |
+
"loss": 0.0114,
|
| 857 |
+
"step": 1200
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.484,
|
| 861 |
+
"grad_norm": 0.0016622517723590136,
|
| 862 |
+
"learning_rate": 0.00022951111111111115,
|
| 863 |
+
"loss": 0.0032,
|
| 864 |
+
"step": 1210
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.488,
|
| 868 |
+
"grad_norm": 0.0018463089363649487,
|
| 869 |
+
"learning_rate": 0.00022773333333333336,
|
| 870 |
+
"loss": 0.0001,
|
| 871 |
+
"step": 1220
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.492,
|
| 875 |
+
"grad_norm": 0.0017726500518620014,
|
| 876 |
+
"learning_rate": 0.00022595555555555557,
|
| 877 |
+
"loss": 0.0,
|
| 878 |
+
"step": 1230
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.496,
|
| 882 |
+
"grad_norm": 0.0016171345487236977,
|
| 883 |
+
"learning_rate": 0.00022417777777777778,
|
| 884 |
+
"loss": 0.0,
|
| 885 |
+
"step": 1240
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.5,
|
| 889 |
+
"grad_norm": 0.001747308298945427,
|
| 890 |
+
"learning_rate": 0.00022240000000000004,
|
| 891 |
+
"loss": 0.0,
|
| 892 |
+
"step": 1250
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.504,
|
| 896 |
+
"grad_norm": 0.01567615382373333,
|
| 897 |
+
"learning_rate": 0.00022062222222222225,
|
| 898 |
+
"loss": 0.0,
|
| 899 |
+
"step": 1260
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.508,
|
| 903 |
+
"grad_norm": 0.03938116133213043,
|
| 904 |
+
"learning_rate": 0.00021884444444444446,
|
| 905 |
+
"loss": 0.0001,
|
| 906 |
+
"step": 1270
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.512,
|
| 910 |
+
"grad_norm": 0.001731475000269711,
|
| 911 |
+
"learning_rate": 0.00021706666666666667,
|
| 912 |
+
"loss": 0.0,
|
| 913 |
+
"step": 1280
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.516,
|
| 917 |
+
"grad_norm": 0.0011080361437052488,
|
| 918 |
+
"learning_rate": 0.00021528888888888893,
|
| 919 |
+
"loss": 0.0,
|
| 920 |
+
"step": 1290
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.52,
|
| 924 |
+
"grad_norm": 0.001253710244782269,
|
| 925 |
+
"learning_rate": 0.00021351111111111114,
|
| 926 |
+
"loss": 0.0,
|
| 927 |
+
"step": 1300
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.524,
|
| 931 |
+
"grad_norm": 0.0009362171404063702,
|
| 932 |
+
"learning_rate": 0.00021173333333333335,
|
| 933 |
+
"loss": 0.0002,
|
| 934 |
+
"step": 1310
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.528,
|
| 938 |
+
"grad_norm": 0.0010680261766538024,
|
| 939 |
+
"learning_rate": 0.00020995555555555556,
|
| 940 |
+
"loss": 0.0,
|
| 941 |
+
"step": 1320
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.532,
|
| 945 |
+
"grad_norm": 0.0009266381966881454,
|
| 946 |
+
"learning_rate": 0.0002081777777777778,
|
| 947 |
+
"loss": 0.0,
|
| 948 |
+
"step": 1330
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.536,
|
| 952 |
+
"grad_norm": 0.0008506786543875933,
|
| 953 |
+
"learning_rate": 0.0002064,
|
| 954 |
+
"loss": 0.0,
|
| 955 |
+
"step": 1340
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.54,
|
| 959 |
+
"grad_norm": 0.000809083750937134,
|
| 960 |
+
"learning_rate": 0.00020462222222222224,
|
| 961 |
+
"loss": 0.0,
|
| 962 |
+
"step": 1350
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.544,
|
| 966 |
+
"grad_norm": 0.0008981467690318823,
|
| 967 |
+
"learning_rate": 0.00020284444444444445,
|
| 968 |
+
"loss": 0.0,
|
| 969 |
+
"step": 1360
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.548,
|
| 973 |
+
"grad_norm": 0.006928270682692528,
|
| 974 |
+
"learning_rate": 0.00020106666666666668,
|
| 975 |
+
"loss": 0.0157,
|
| 976 |
+
"step": 1370
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.552,
|
| 980 |
+
"grad_norm": 0.019026894122362137,
|
| 981 |
+
"learning_rate": 0.0001992888888888889,
|
| 982 |
+
"loss": 0.0003,
|
| 983 |
+
"step": 1380
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.556,
|
| 987 |
+
"grad_norm": 0.018439382314682007,
|
| 988 |
+
"learning_rate": 0.0001975111111111111,
|
| 989 |
+
"loss": 0.0019,
|
| 990 |
+
"step": 1390
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.56,
|
| 994 |
+
"grad_norm": 0.00941744539886713,
|
| 995 |
+
"learning_rate": 0.00019573333333333334,
|
| 996 |
+
"loss": 0.0007,
|
| 997 |
+
"step": 1400
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.564,
|
| 1001 |
+
"grad_norm": 0.00815660785883665,
|
| 1002 |
+
"learning_rate": 0.00019395555555555555,
|
| 1003 |
+
"loss": 0.0002,
|
| 1004 |
+
"step": 1410
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.568,
|
| 1008 |
+
"grad_norm": 0.039481475949287415,
|
| 1009 |
+
"learning_rate": 0.00019217777777777778,
|
| 1010 |
+
"loss": 0.0096,
|
| 1011 |
+
"step": 1420
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.572,
|
| 1015 |
+
"grad_norm": 0.005729417782276869,
|
| 1016 |
+
"learning_rate": 0.0001904,
|
| 1017 |
+
"loss": 0.0112,
|
| 1018 |
+
"step": 1430
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.576,
|
| 1022 |
+
"grad_norm": 0.007418328896164894,
|
| 1023 |
+
"learning_rate": 0.00018862222222222223,
|
| 1024 |
+
"loss": 0.0077,
|
| 1025 |
+
"step": 1440
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.58,
|
| 1029 |
+
"grad_norm": 0.012174281291663647,
|
| 1030 |
+
"learning_rate": 0.00018684444444444444,
|
| 1031 |
+
"loss": 0.0101,
|
| 1032 |
+
"step": 1450
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.584,
|
| 1036 |
+
"grad_norm": 0.015086760744452477,
|
| 1037 |
+
"learning_rate": 0.00018506666666666667,
|
| 1038 |
+
"loss": 0.0003,
|
| 1039 |
+
"step": 1460
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.588,
|
| 1043 |
+
"grad_norm": 0.01587488315999508,
|
| 1044 |
+
"learning_rate": 0.00018328888888888888,
|
| 1045 |
+
"loss": 0.0004,
|
| 1046 |
+
"step": 1470
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.592,
|
| 1050 |
+
"grad_norm": 0.03278527408838272,
|
| 1051 |
+
"learning_rate": 0.00018151111111111112,
|
| 1052 |
+
"loss": 0.014,
|
| 1053 |
+
"step": 1480
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.596,
|
| 1057 |
+
"grad_norm": 0.10622735321521759,
|
| 1058 |
+
"learning_rate": 0.00017973333333333333,
|
| 1059 |
+
"loss": 0.0012,
|
| 1060 |
+
"step": 1490
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.6,
|
| 1064 |
+
"grad_norm": 0.014216107316315174,
|
| 1065 |
+
"learning_rate": 0.00017795555555555557,
|
| 1066 |
+
"loss": 0.0006,
|
| 1067 |
+
"step": 1500
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.604,
|
| 1071 |
+
"grad_norm": 0.008536259643733501,
|
| 1072 |
+
"learning_rate": 0.0001761777777777778,
|
| 1073 |
+
"loss": 0.0004,
|
| 1074 |
+
"step": 1510
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.608,
|
| 1078 |
+
"grad_norm": 0.005456692539155483,
|
| 1079 |
+
"learning_rate": 0.0001744,
|
| 1080 |
+
"loss": 0.003,
|
| 1081 |
+
"step": 1520
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.612,
|
| 1085 |
+
"grad_norm": 0.005484251771122217,
|
| 1086 |
+
"learning_rate": 0.00017262222222222225,
|
| 1087 |
+
"loss": 0.0002,
|
| 1088 |
+
"step": 1530
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.616,
|
| 1092 |
+
"grad_norm": 0.006969142239540815,
|
| 1093 |
+
"learning_rate": 0.00017084444444444446,
|
| 1094 |
+
"loss": 0.0152,
|
| 1095 |
+
"step": 1540
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.62,
|
| 1099 |
+
"grad_norm": 0.01335048582404852,
|
| 1100 |
+
"learning_rate": 0.0001690666666666667,
|
| 1101 |
+
"loss": 0.0003,
|
| 1102 |
+
"step": 1550
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.624,
|
| 1106 |
+
"grad_norm": 0.012201100587844849,
|
| 1107 |
+
"learning_rate": 0.0001672888888888889,
|
| 1108 |
+
"loss": 0.0026,
|
| 1109 |
+
"step": 1560
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.628,
|
| 1113 |
+
"grad_norm": 0.01430495735257864,
|
| 1114 |
+
"learning_rate": 0.00016551111111111114,
|
| 1115 |
+
"loss": 0.0006,
|
| 1116 |
+
"step": 1570
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.632,
|
| 1120 |
+
"grad_norm": 0.03964800387620926,
|
| 1121 |
+
"learning_rate": 0.00016373333333333335,
|
| 1122 |
+
"loss": 0.0052,
|
| 1123 |
+
"step": 1580
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.636,
|
| 1127 |
+
"grad_norm": 0.007560222875326872,
|
| 1128 |
+
"learning_rate": 0.00016195555555555558,
|
| 1129 |
+
"loss": 0.0002,
|
| 1130 |
+
"step": 1590
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.64,
|
| 1134 |
+
"grad_norm": 0.00928121991455555,
|
| 1135 |
+
"learning_rate": 0.0001601777777777778,
|
| 1136 |
+
"loss": 0.0002,
|
| 1137 |
+
"step": 1600
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.644,
|
| 1141 |
+
"grad_norm": 0.016387827694416046,
|
| 1142 |
+
"learning_rate": 0.00015840000000000003,
|
| 1143 |
+
"loss": 0.0002,
|
| 1144 |
+
"step": 1610
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.648,
|
| 1148 |
+
"grad_norm": 0.005653001833707094,
|
| 1149 |
+
"learning_rate": 0.00015662222222222224,
|
| 1150 |
+
"loss": 0.0002,
|
| 1151 |
+
"step": 1620
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.652,
|
| 1155 |
+
"grad_norm": 0.005010788328945637,
|
| 1156 |
+
"learning_rate": 0.00015484444444444445,
|
| 1157 |
+
"loss": 0.0026,
|
| 1158 |
+
"step": 1630
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.656,
|
| 1162 |
+
"grad_norm": 0.00448651984333992,
|
| 1163 |
+
"learning_rate": 0.00015306666666666666,
|
| 1164 |
+
"loss": 0.0001,
|
| 1165 |
+
"step": 1640
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.66,
|
| 1169 |
+
"grad_norm": 0.011779513210058212,
|
| 1170 |
+
"learning_rate": 0.0001512888888888889,
|
| 1171 |
+
"loss": 0.0001,
|
| 1172 |
+
"step": 1650
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.664,
|
| 1176 |
+
"grad_norm": 0.005999232176691294,
|
| 1177 |
+
"learning_rate": 0.0001495111111111111,
|
| 1178 |
+
"loss": 0.0026,
|
| 1179 |
+
"step": 1660
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.668,
|
| 1183 |
+
"grad_norm": 0.005314210895448923,
|
| 1184 |
+
"learning_rate": 0.00014773333333333334,
|
| 1185 |
+
"loss": 0.0001,
|
| 1186 |
+
"step": 1670
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.672,
|
| 1190 |
+
"grad_norm": 0.00803409144282341,
|
| 1191 |
+
"learning_rate": 0.00014595555555555555,
|
| 1192 |
+
"loss": 0.0008,
|
| 1193 |
+
"step": 1680
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.676,
|
| 1197 |
+
"grad_norm": 1.069173812866211,
|
| 1198 |
+
"learning_rate": 0.00014417777777777778,
|
| 1199 |
+
"loss": 0.0009,
|
| 1200 |
+
"step": 1690
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.68,
|
| 1204 |
+
"grad_norm": 0.013021737337112427,
|
| 1205 |
+
"learning_rate": 0.0001424,
|
| 1206 |
+
"loss": 0.0002,
|
| 1207 |
+
"step": 1700
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.684,
|
| 1211 |
+
"grad_norm": 0.0024978595320135355,
|
| 1212 |
+
"learning_rate": 0.00014062222222222223,
|
| 1213 |
+
"loss": 0.0089,
|
| 1214 |
+
"step": 1710
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.688,
|
| 1218 |
+
"grad_norm": 0.003698146902024746,
|
| 1219 |
+
"learning_rate": 0.00013884444444444444,
|
| 1220 |
+
"loss": 0.0138,
|
| 1221 |
+
"step": 1720
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.692,
|
| 1225 |
+
"grad_norm": 0.03162381052970886,
|
| 1226 |
+
"learning_rate": 0.00013706666666666667,
|
| 1227 |
+
"loss": 0.0047,
|
| 1228 |
+
"step": 1730
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.696,
|
| 1232 |
+
"grad_norm": 0.022371523082256317,
|
| 1233 |
+
"learning_rate": 0.00013528888888888888,
|
| 1234 |
+
"loss": 0.0021,
|
| 1235 |
+
"step": 1740
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.7,
|
| 1239 |
+
"grad_norm": 0.008335133083164692,
|
| 1240 |
+
"learning_rate": 0.00013351111111111112,
|
| 1241 |
+
"loss": 0.004,
|
| 1242 |
+
"step": 1750
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.704,
|
| 1246 |
+
"grad_norm": 0.01031657587736845,
|
| 1247 |
+
"learning_rate": 0.00013173333333333333,
|
| 1248 |
+
"loss": 0.0057,
|
| 1249 |
+
"step": 1760
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.708,
|
| 1253 |
+
"grad_norm": 0.008161673322319984,
|
| 1254 |
+
"learning_rate": 0.00012995555555555556,
|
| 1255 |
+
"loss": 0.0062,
|
| 1256 |
+
"step": 1770
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.712,
|
| 1260 |
+
"grad_norm": 0.007272409275174141,
|
| 1261 |
+
"learning_rate": 0.00012817777777777777,
|
| 1262 |
+
"loss": 0.0002,
|
| 1263 |
+
"step": 1780
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.716,
|
| 1267 |
+
"grad_norm": 0.006052941549569368,
|
| 1268 |
+
"learning_rate": 0.0001264,
|
| 1269 |
+
"loss": 0.0004,
|
| 1270 |
+
"step": 1790
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.72,
|
| 1274 |
+
"grad_norm": 0.00472453935071826,
|
| 1275 |
+
"learning_rate": 0.00012462222222222222,
|
| 1276 |
+
"loss": 0.0015,
|
| 1277 |
+
"step": 1800
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.724,
|
| 1281 |
+
"grad_norm": 0.004381476901471615,
|
| 1282 |
+
"learning_rate": 0.00012284444444444445,
|
| 1283 |
+
"loss": 0.0001,
|
| 1284 |
+
"step": 1810
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 0.728,
|
| 1288 |
+
"grad_norm": 0.004298892803490162,
|
| 1289 |
+
"learning_rate": 0.00012106666666666666,
|
| 1290 |
+
"loss": 0.0001,
|
| 1291 |
+
"step": 1820
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 0.732,
|
| 1295 |
+
"grad_norm": 0.003755762241780758,
|
| 1296 |
+
"learning_rate": 0.0001192888888888889,
|
| 1297 |
+
"loss": 0.0069,
|
| 1298 |
+
"step": 1830
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.736,
|
| 1302 |
+
"grad_norm": 0.002545682480558753,
|
| 1303 |
+
"learning_rate": 0.00011751111111111112,
|
| 1304 |
+
"loss": 0.0001,
|
| 1305 |
+
"step": 1840
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.74,
|
| 1309 |
+
"grad_norm": 0.0083426209166646,
|
| 1310 |
+
"learning_rate": 0.00011573333333333333,
|
| 1311 |
+
"loss": 0.0145,
|
| 1312 |
+
"step": 1850
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 0.744,
|
| 1316 |
+
"grad_norm": 0.03931976854801178,
|
| 1317 |
+
"learning_rate": 0.00011395555555555557,
|
| 1318 |
+
"loss": 0.0011,
|
| 1319 |
+
"step": 1860
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 0.748,
|
| 1323 |
+
"grad_norm": 0.012124504894018173,
|
| 1324 |
+
"learning_rate": 0.00011217777777777778,
|
| 1325 |
+
"loss": 0.0008,
|
| 1326 |
+
"step": 1870
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 0.752,
|
| 1330 |
+
"grad_norm": 0.00980482716113329,
|
| 1331 |
+
"learning_rate": 0.00011040000000000001,
|
| 1332 |
+
"loss": 0.0037,
|
| 1333 |
+
"step": 1880
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 0.756,
|
| 1337 |
+
"grad_norm": 2.7182815074920654,
|
| 1338 |
+
"learning_rate": 0.00010862222222222222,
|
| 1339 |
+
"loss": 0.0013,
|
| 1340 |
+
"step": 1890
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.76,
|
| 1344 |
+
"grad_norm": 0.00282659032382071,
|
| 1345 |
+
"learning_rate": 0.00010684444444444446,
|
| 1346 |
+
"loss": 0.0015,
|
| 1347 |
+
"step": 1900
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.764,
|
| 1351 |
+
"grad_norm": 0.0028170158620923758,
|
| 1352 |
+
"learning_rate": 0.00010506666666666667,
|
| 1353 |
+
"loss": 0.0002,
|
| 1354 |
+
"step": 1910
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 0.768,
|
| 1358 |
+
"grad_norm": 0.0013493013102561235,
|
| 1359 |
+
"learning_rate": 0.0001032888888888889,
|
| 1360 |
+
"loss": 0.0011,
|
| 1361 |
+
"step": 1920
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 0.772,
|
| 1365 |
+
"grad_norm": 0.0013236267259344459,
|
| 1366 |
+
"learning_rate": 0.00010151111111111111,
|
| 1367 |
+
"loss": 0.0113,
|
| 1368 |
+
"step": 1930
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 0.776,
|
| 1372 |
+
"grad_norm": 0.0017112682107836008,
|
| 1373 |
+
"learning_rate": 9.973333333333334e-05,
|
| 1374 |
+
"loss": 0.0,
|
| 1375 |
+
"step": 1940
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 0.78,
|
| 1379 |
+
"grad_norm": 0.0022746031172573566,
|
| 1380 |
+
"learning_rate": 9.795555555555556e-05,
|
| 1381 |
+
"loss": 0.0002,
|
| 1382 |
+
"step": 1950
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.784,
|
| 1386 |
+
"grad_norm": 0.0015560939209535718,
|
| 1387 |
+
"learning_rate": 9.617777777777778e-05,
|
| 1388 |
+
"loss": 0.0003,
|
| 1389 |
+
"step": 1960
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.788,
|
| 1393 |
+
"grad_norm": 0.001704063848592341,
|
| 1394 |
+
"learning_rate": 9.44e-05,
|
| 1395 |
+
"loss": 0.0001,
|
| 1396 |
+
"step": 1970
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 0.792,
|
| 1400 |
+
"grad_norm": 0.13681542873382568,
|
| 1401 |
+
"learning_rate": 9.262222222222223e-05,
|
| 1402 |
+
"loss": 0.0001,
|
| 1403 |
+
"step": 1980
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 0.796,
|
| 1407 |
+
"grad_norm": 0.0013939260970801115,
|
| 1408 |
+
"learning_rate": 9.084444444444445e-05,
|
| 1409 |
+
"loss": 0.0,
|
| 1410 |
+
"step": 1990
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 0.8,
|
| 1414 |
+
"grad_norm": 0.0016302280128002167,
|
| 1415 |
+
"learning_rate": 8.906666666666667e-05,
|
| 1416 |
+
"loss": 0.0,
|
| 1417 |
+
"step": 2000
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 0.804,
|
| 1421 |
+
"grad_norm": 0.0014311623526737094,
|
| 1422 |
+
"learning_rate": 8.72888888888889e-05,
|
| 1423 |
+
"loss": 0.0001,
|
| 1424 |
+
"step": 2010
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 0.808,
|
| 1428 |
+
"grad_norm": 0.0013674832880496979,
|
| 1429 |
+
"learning_rate": 8.551111111111112e-05,
|
| 1430 |
+
"loss": 0.0,
|
| 1431 |
+
"step": 2020
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 0.812,
|
| 1435 |
+
"grad_norm": 0.0013156217755749822,
|
| 1436 |
+
"learning_rate": 8.373333333333334e-05,
|
| 1437 |
+
"loss": 0.0,
|
| 1438 |
+
"step": 2030
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 0.816,
|
| 1442 |
+
"grad_norm": 0.7331569790840149,
|
| 1443 |
+
"learning_rate": 8.195555555555556e-05,
|
| 1444 |
+
"loss": 0.0131,
|
| 1445 |
+
"step": 2040
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 0.82,
|
| 1449 |
+
"grad_norm": 0.006600831635296345,
|
| 1450 |
+
"learning_rate": 8.017777777777779e-05,
|
| 1451 |
+
"loss": 0.0168,
|
| 1452 |
+
"step": 2050
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.824,
|
| 1456 |
+
"grad_norm": 0.012540962547063828,
|
| 1457 |
+
"learning_rate": 7.840000000000001e-05,
|
| 1458 |
+
"loss": 0.0005,
|
| 1459 |
+
"step": 2060
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 0.828,
|
| 1463 |
+
"grad_norm": 0.023732764646410942,
|
| 1464 |
+
"learning_rate": 7.662222222222223e-05,
|
| 1465 |
+
"loss": 0.0004,
|
| 1466 |
+
"step": 2070
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.832,
|
| 1470 |
+
"grad_norm": 0.01323721744120121,
|
| 1471 |
+
"learning_rate": 7.484444444444445e-05,
|
| 1472 |
+
"loss": 0.0009,
|
| 1473 |
+
"step": 2080
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.836,
|
| 1477 |
+
"grad_norm": 0.009665262885391712,
|
| 1478 |
+
"learning_rate": 7.306666666666668e-05,
|
| 1479 |
+
"loss": 0.0003,
|
| 1480 |
+
"step": 2090
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 0.84,
|
| 1484 |
+
"grad_norm": 0.008318779990077019,
|
| 1485 |
+
"learning_rate": 7.12888888888889e-05,
|
| 1486 |
+
"loss": 0.0002,
|
| 1487 |
+
"step": 2100
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 0.844,
|
| 1491 |
+
"grad_norm": 0.0076893046498298645,
|
| 1492 |
+
"learning_rate": 6.951111111111112e-05,
|
| 1493 |
+
"loss": 0.0002,
|
| 1494 |
+
"step": 2110
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 0.848,
|
| 1498 |
+
"grad_norm": 0.007633762434124947,
|
| 1499 |
+
"learning_rate": 6.773333333333333e-05,
|
| 1500 |
+
"loss": 0.0002,
|
| 1501 |
+
"step": 2120
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 0.852,
|
| 1505 |
+
"grad_norm": 0.0073342034593224525,
|
| 1506 |
+
"learning_rate": 6.595555555555555e-05,
|
| 1507 |
+
"loss": 0.0002,
|
| 1508 |
+
"step": 2130
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 0.856,
|
| 1512 |
+
"grad_norm": 0.005839989520609379,
|
| 1513 |
+
"learning_rate": 6.417777777777778e-05,
|
| 1514 |
+
"loss": 0.0001,
|
| 1515 |
+
"step": 2140
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 0.86,
|
| 1519 |
+
"grad_norm": 0.005278122611343861,
|
| 1520 |
+
"learning_rate": 6.24e-05,
|
| 1521 |
+
"loss": 0.0001,
|
| 1522 |
+
"step": 2150
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 0.864,
|
| 1526 |
+
"grad_norm": 0.006889206822961569,
|
| 1527 |
+
"learning_rate": 6.062222222222222e-05,
|
| 1528 |
+
"loss": 0.0001,
|
| 1529 |
+
"step": 2160
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 0.868,
|
| 1533 |
+
"grad_norm": 0.7518675923347473,
|
| 1534 |
+
"learning_rate": 5.8844444444444444e-05,
|
| 1535 |
+
"loss": 0.0011,
|
| 1536 |
+
"step": 2170
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 0.872,
|
| 1540 |
+
"grad_norm": 0.009561669081449509,
|
| 1541 |
+
"learning_rate": 5.706666666666667e-05,
|
| 1542 |
+
"loss": 0.0015,
|
| 1543 |
+
"step": 2180
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 0.876,
|
| 1547 |
+
"grad_norm": 0.003935567103326321,
|
| 1548 |
+
"learning_rate": 5.528888888888889e-05,
|
| 1549 |
+
"loss": 0.0013,
|
| 1550 |
+
"step": 2190
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 0.88,
|
| 1554 |
+
"grad_norm": 0.003710120217874646,
|
| 1555 |
+
"learning_rate": 5.351111111111111e-05,
|
| 1556 |
+
"loss": 0.0001,
|
| 1557 |
+
"step": 2200
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 0.884,
|
| 1561 |
+
"grad_norm": 0.002907069865614176,
|
| 1562 |
+
"learning_rate": 5.1733333333333335e-05,
|
| 1563 |
+
"loss": 0.0002,
|
| 1564 |
+
"step": 2210
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 0.888,
|
| 1568 |
+
"grad_norm": 0.0030114708933979273,
|
| 1569 |
+
"learning_rate": 4.995555555555556e-05,
|
| 1570 |
+
"loss": 0.0023,
|
| 1571 |
+
"step": 2220
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 0.892,
|
| 1575 |
+
"grad_norm": 0.0034878470469266176,
|
| 1576 |
+
"learning_rate": 4.817777777777778e-05,
|
| 1577 |
+
"loss": 0.0001,
|
| 1578 |
+
"step": 2230
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 0.896,
|
| 1582 |
+
"grad_norm": 0.003132345387712121,
|
| 1583 |
+
"learning_rate": 4.64e-05,
|
| 1584 |
+
"loss": 0.0004,
|
| 1585 |
+
"step": 2240
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 0.9,
|
| 1589 |
+
"grad_norm": 0.0030635548755526543,
|
| 1590 |
+
"learning_rate": 4.4622222222222226e-05,
|
| 1591 |
+
"loss": 0.0001,
|
| 1592 |
+
"step": 2250
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 0.904,
|
| 1596 |
+
"grad_norm": 0.0031918687745928764,
|
| 1597 |
+
"learning_rate": 4.284444444444445e-05,
|
| 1598 |
+
"loss": 0.0031,
|
| 1599 |
+
"step": 2260
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 0.908,
|
| 1603 |
+
"grad_norm": 0.006079728715121746,
|
| 1604 |
+
"learning_rate": 4.106666666666667e-05,
|
| 1605 |
+
"loss": 0.0001,
|
| 1606 |
+
"step": 2270
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 0.912,
|
| 1610 |
+
"grad_norm": 0.0049600861966609955,
|
| 1611 |
+
"learning_rate": 3.9288888888888894e-05,
|
| 1612 |
+
"loss": 0.0001,
|
| 1613 |
+
"step": 2280
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 0.916,
|
| 1617 |
+
"grad_norm": 0.006550566293299198,
|
| 1618 |
+
"learning_rate": 3.7511111111111116e-05,
|
| 1619 |
+
"loss": 0.0001,
|
| 1620 |
+
"step": 2290
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 0.92,
|
| 1624 |
+
"grad_norm": 0.0037377977278083563,
|
| 1625 |
+
"learning_rate": 3.573333333333333e-05,
|
| 1626 |
+
"loss": 0.0043,
|
| 1627 |
+
"step": 2300
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 0.924,
|
| 1631 |
+
"grad_norm": 0.0033655581064522266,
|
| 1632 |
+
"learning_rate": 3.3955555555555555e-05,
|
| 1633 |
+
"loss": 0.0001,
|
| 1634 |
+
"step": 2310
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.928,
|
| 1638 |
+
"grad_norm": 0.0035526896826922894,
|
| 1639 |
+
"learning_rate": 3.217777777777778e-05,
|
| 1640 |
+
"loss": 0.0001,
|
| 1641 |
+
"step": 2320
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 0.932,
|
| 1645 |
+
"grad_norm": 0.0034034913405776024,
|
| 1646 |
+
"learning_rate": 3.04e-05,
|
| 1647 |
+
"loss": 0.0001,
|
| 1648 |
+
"step": 2330
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 0.936,
|
| 1652 |
+
"grad_norm": 0.006957338657230139,
|
| 1653 |
+
"learning_rate": 2.8622222222222223e-05,
|
| 1654 |
+
"loss": 0.0156,
|
| 1655 |
+
"step": 2340
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.94,
|
| 1659 |
+
"grad_norm": 0.007531840819865465,
|
| 1660 |
+
"learning_rate": 2.6844444444444446e-05,
|
| 1661 |
+
"loss": 0.0001,
|
| 1662 |
+
"step": 2350
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 0.944,
|
| 1666 |
+
"grad_norm": 0.00489840330556035,
|
| 1667 |
+
"learning_rate": 2.5066666666666665e-05,
|
| 1668 |
+
"loss": 0.0001,
|
| 1669 |
+
"step": 2360
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 0.948,
|
| 1673 |
+
"grad_norm": 0.005002748221158981,
|
| 1674 |
+
"learning_rate": 2.328888888888889e-05,
|
| 1675 |
+
"loss": 0.0001,
|
| 1676 |
+
"step": 2370
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.952,
|
| 1680 |
+
"grad_norm": 0.0048461793921887875,
|
| 1681 |
+
"learning_rate": 2.1511111111111114e-05,
|
| 1682 |
+
"loss": 0.0001,
|
| 1683 |
+
"step": 2380
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.956,
|
| 1687 |
+
"grad_norm": 0.0058443606831133366,
|
| 1688 |
+
"learning_rate": 1.9733333333333333e-05,
|
| 1689 |
+
"loss": 0.0004,
|
| 1690 |
+
"step": 2390
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.96,
|
| 1694 |
+
"grad_norm": 0.0049603781662881374,
|
| 1695 |
+
"learning_rate": 1.7955555555555556e-05,
|
| 1696 |
+
"loss": 0.0002,
|
| 1697 |
+
"step": 2400
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.964,
|
| 1701 |
+
"grad_norm": 0.005183044355362654,
|
| 1702 |
+
"learning_rate": 1.617777777777778e-05,
|
| 1703 |
+
"loss": 0.0035,
|
| 1704 |
+
"step": 2410
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 0.968,
|
| 1708 |
+
"grad_norm": 0.0050833080895245075,
|
| 1709 |
+
"learning_rate": 1.44e-05,
|
| 1710 |
+
"loss": 0.0012,
|
| 1711 |
+
"step": 2420
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 0.972,
|
| 1715 |
+
"grad_norm": 0.004844759125262499,
|
| 1716 |
+
"learning_rate": 1.2622222222222224e-05,
|
| 1717 |
+
"loss": 0.0001,
|
| 1718 |
+
"step": 2430
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.976,
|
| 1722 |
+
"grad_norm": 0.004592160694301128,
|
| 1723 |
+
"learning_rate": 1.0844444444444445e-05,
|
| 1724 |
+
"loss": 0.0012,
|
| 1725 |
+
"step": 2440
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.98,
|
| 1729 |
+
"grad_norm": 0.004524698480963707,
|
| 1730 |
+
"learning_rate": 9.066666666666667e-06,
|
| 1731 |
+
"loss": 0.0001,
|
| 1732 |
+
"step": 2450
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 0.984,
|
| 1736 |
+
"grad_norm": 0.005091739818453789,
|
| 1737 |
+
"learning_rate": 7.288888888888889e-06,
|
| 1738 |
+
"loss": 0.0002,
|
| 1739 |
+
"step": 2460
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.988,
|
| 1743 |
+
"grad_norm": 0.004614758305251598,
|
| 1744 |
+
"learning_rate": 5.511111111111111e-06,
|
| 1745 |
+
"loss": 0.0001,
|
| 1746 |
+
"step": 2470
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 0.992,
|
| 1750 |
+
"grad_norm": 0.010195998474955559,
|
| 1751 |
+
"learning_rate": 3.7333333333333337e-06,
|
| 1752 |
+
"loss": 0.0003,
|
| 1753 |
+
"step": 2480
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.996,
|
| 1757 |
+
"grad_norm": 0.004264904651790857,
|
| 1758 |
+
"learning_rate": 1.9555555555555556e-06,
|
| 1759 |
+
"loss": 0.0002,
|
| 1760 |
+
"step": 2490
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 1.0,
|
| 1764 |
+
"grad_norm": 0.004516505636274815,
|
| 1765 |
+
"learning_rate": 1.777777777777778e-07,
|
| 1766 |
+
"loss": 0.0002,
|
| 1767 |
+
"step": 2500
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 1.0,
|
| 1771 |
+
"eval_accuracy": 0.999875,
|
| 1772 |
+
"eval_loss": 0.0004970150184817612,
|
| 1773 |
+
"eval_runtime": 2540.0992,
|
| 1774 |
+
"eval_samples_per_second": 15.747,
|
| 1775 |
+
"eval_steps_per_second": 0.984,
|
| 1776 |
+
"step": 2500
|
| 1777 |
+
}
|
| 1778 |
+
],
|
| 1779 |
+
"logging_steps": 10,
|
| 1780 |
+
"max_steps": 2500,
|
| 1781 |
+
"num_input_tokens_seen": 0,
|
| 1782 |
+
"num_train_epochs": 1,
|
| 1783 |
+
"save_steps": 500,
|
| 1784 |
+
"stateful_callbacks": {
|
| 1785 |
+
"TrainerControl": {
|
| 1786 |
+
"args": {
|
| 1787 |
+
"should_epoch_stop": false,
|
| 1788 |
+
"should_evaluate": false,
|
| 1789 |
+
"should_log": false,
|
| 1790 |
+
"should_save": true,
|
| 1791 |
+
"should_training_stop": true
|
| 1792 |
+
},
|
| 1793 |
+
"attributes": {}
|
| 1794 |
+
}
|
| 1795 |
+
},
|
| 1796 |
+
"total_flos": 2.0090762496e+19,
|
| 1797 |
+
"train_batch_size": 16,
|
| 1798 |
+
"trial_name": null,
|
| 1799 |
+
"trial_params": null
|
| 1800 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0a1d65cfcef105499fedbf288fbe105350008ba80e044b2c8b41ab81f1688332
|
| 3 |
+
size 5841
|