Training in progress, epoch 3, checkpoint
Browse files- checkpoint-4407/config.json +71 -0
- checkpoint-4407/model.safetensors +3 -0
- checkpoint-4407/optimizer.pt +3 -0
- checkpoint-4407/preprocessor_config.json +14 -0
- checkpoint-4407/rng_state.pth +3 -0
- checkpoint-4407/scheduler.pt +3 -0
- checkpoint-4407/trainer_state.json +2111 -0
- checkpoint-4407/training_args.bin +3 -0
checkpoint-4407/config.json
ADDED
|
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "vinai/PhoWhisper-medium",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"activation_function": "gelu",
|
| 5 |
+
"apply_spec_augment": false,
|
| 6 |
+
"architectures": [
|
| 7 |
+
"ViSpeechClassification"
|
| 8 |
+
],
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"begin_suppress_tokens": [
|
| 11 |
+
220,
|
| 12 |
+
50257
|
| 13 |
+
],
|
| 14 |
+
"bos_token_id": 50257,
|
| 15 |
+
"classifier_proj_size": 256,
|
| 16 |
+
"d_model": 1024,
|
| 17 |
+
"decoder_attention_heads": 16,
|
| 18 |
+
"decoder_ffn_dim": 4096,
|
| 19 |
+
"decoder_layerdrop": 0.0,
|
| 20 |
+
"decoder_layers": 24,
|
| 21 |
+
"decoder_start_token_id": 50258,
|
| 22 |
+
"dropout": 0.0,
|
| 23 |
+
"encoder_attention_heads": 16,
|
| 24 |
+
"encoder_ffn_dim": 4096,
|
| 25 |
+
"encoder_layerdrop": 0.0,
|
| 26 |
+
"encoder_layers": 24,
|
| 27 |
+
"eos_token_id": 50257,
|
| 28 |
+
"forced_decoder_ids": null,
|
| 29 |
+
"id2label": {
|
| 30 |
+
"0": "male, northern dialect",
|
| 31 |
+
"1": "male, central dialect",
|
| 32 |
+
"2": "male, highland central dialect",
|
| 33 |
+
"3": "male, southern dialect",
|
| 34 |
+
"4": "female, northern dialect",
|
| 35 |
+
"5": "female, central dialect",
|
| 36 |
+
"6": "female, highland central dialect",
|
| 37 |
+
"7": "female, southern dialect"
|
| 38 |
+
},
|
| 39 |
+
"init_std": 0.02,
|
| 40 |
+
"is_encoder_decoder": true,
|
| 41 |
+
"label2id": {
|
| 42 |
+
"female, central dialect": "5",
|
| 43 |
+
"female, highland central dialect": "6",
|
| 44 |
+
"female, northern dialect": "4",
|
| 45 |
+
"female, southern dialect": "7",
|
| 46 |
+
"male, central dialect": "1",
|
| 47 |
+
"male, highland central dialect": "2",
|
| 48 |
+
"male, northern dialect": "0",
|
| 49 |
+
"male, southern dialect": "3"
|
| 50 |
+
},
|
| 51 |
+
"mask_feature_length": 10,
|
| 52 |
+
"mask_feature_min_masks": 0,
|
| 53 |
+
"mask_feature_prob": 0.0,
|
| 54 |
+
"mask_time_length": 10,
|
| 55 |
+
"mask_time_min_masks": 2,
|
| 56 |
+
"mask_time_prob": 0.05,
|
| 57 |
+
"max_length": 448,
|
| 58 |
+
"max_source_positions": 1500,
|
| 59 |
+
"max_target_positions": 448,
|
| 60 |
+
"median_filter_width": 7,
|
| 61 |
+
"model_type": "whisper",
|
| 62 |
+
"num_hidden_layers": 24,
|
| 63 |
+
"num_mel_bins": 80,
|
| 64 |
+
"pad_token_id": 50257,
|
| 65 |
+
"scale_embedding": false,
|
| 66 |
+
"torch_dtype": "float32",
|
| 67 |
+
"transformers_version": "4.41.2",
|
| 68 |
+
"use_cache": true,
|
| 69 |
+
"use_weighted_layer_sum": false,
|
| 70 |
+
"vocab_size": 51865
|
| 71 |
+
}
|
checkpoint-4407/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bd8a0c81feeb8fa7dae7cc0c71954e889904a68d974ef9f89cd54212b71949e2
|
| 3 |
+
size 1235737744
|
checkpoint-4407/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c971bc0addc96266e5fd0a88d59dff8267122c24ff657e1b5322f452bd728fe8
|
| 3 |
+
size 2459418356
|
checkpoint-4407/preprocessor_config.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"chunk_length": 30,
|
| 3 |
+
"feature_extractor_type": "WhisperFeatureExtractor",
|
| 4 |
+
"feature_size": 80,
|
| 5 |
+
"hop_length": 160,
|
| 6 |
+
"n_fft": 400,
|
| 7 |
+
"n_samples": 480000,
|
| 8 |
+
"nb_max_frames": 3000,
|
| 9 |
+
"padding_side": "right",
|
| 10 |
+
"padding_value": 0.0,
|
| 11 |
+
"processor_class": "WhisperProcessor",
|
| 12 |
+
"return_attention_mask": false,
|
| 13 |
+
"sampling_rate": 16000
|
| 14 |
+
}
|
checkpoint-4407/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d4e120b7b55e946fc347b1c55d2feaafbbe8122a499390fb1dc5c82b4d32326
|
| 3 |
+
size 14244
|
checkpoint-4407/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d727d45f6b93582ee1790992a0922ba823ded942ec0e8fc8a2c0ed67b9f33a34
|
| 3 |
+
size 1064
|
checkpoint-4407/trainer_state.json
ADDED
|
@@ -0,0 +1,2111 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9131455399061033,
|
| 3 |
+
"best_model_checkpoint": "PhoWhisper-medium-vispeech-classifier-v1/checkpoint-4407",
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 4407,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.010211027910142955,
|
| 13 |
+
"grad_norm": 79976.4921875,
|
| 14 |
+
"learning_rate": 6.122448979591837e-08,
|
| 15 |
+
"loss": 2.0788,
|
| 16 |
+
"step": 15
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.02042205582028591,
|
| 20 |
+
"grad_norm": 49252.46484375,
|
| 21 |
+
"learning_rate": 1.2244897959183673e-07,
|
| 22 |
+
"loss": 2.0787,
|
| 23 |
+
"step": 30
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.03063308373042886,
|
| 27 |
+
"grad_norm": 66237.7890625,
|
| 28 |
+
"learning_rate": 1.836734693877551e-07,
|
| 29 |
+
"loss": 2.0734,
|
| 30 |
+
"step": 45
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.04084411164057182,
|
| 34 |
+
"grad_norm": 66530.140625,
|
| 35 |
+
"learning_rate": 2.4489795918367347e-07,
|
| 36 |
+
"loss": 2.0745,
|
| 37 |
+
"step": 60
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.05105513955071477,
|
| 41 |
+
"grad_norm": 75123.3515625,
|
| 42 |
+
"learning_rate": 3.0612244897959183e-07,
|
| 43 |
+
"loss": 2.0715,
|
| 44 |
+
"step": 75
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.06126616746085772,
|
| 48 |
+
"grad_norm": 91761.9921875,
|
| 49 |
+
"learning_rate": 3.673469387755102e-07,
|
| 50 |
+
"loss": 2.0702,
|
| 51 |
+
"step": 90
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.07147719537100068,
|
| 55 |
+
"grad_norm": 76544.7421875,
|
| 56 |
+
"learning_rate": 4.2857142857142857e-07,
|
| 57 |
+
"loss": 2.0659,
|
| 58 |
+
"step": 105
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.08168822328114364,
|
| 62 |
+
"grad_norm": 85211.5078125,
|
| 63 |
+
"learning_rate": 4.897959183673469e-07,
|
| 64 |
+
"loss": 2.0632,
|
| 65 |
+
"step": 120
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.09189925119128659,
|
| 69 |
+
"grad_norm": 86731.6171875,
|
| 70 |
+
"learning_rate": 5.510204081632653e-07,
|
| 71 |
+
"loss": 2.0579,
|
| 72 |
+
"step": 135
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.10211027910142954,
|
| 76 |
+
"grad_norm": 73202.421875,
|
| 77 |
+
"learning_rate": 6.122448979591837e-07,
|
| 78 |
+
"loss": 2.0466,
|
| 79 |
+
"step": 150
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.1123213070115725,
|
| 83 |
+
"grad_norm": 99795.828125,
|
| 84 |
+
"learning_rate": 6.73469387755102e-07,
|
| 85 |
+
"loss": 2.0266,
|
| 86 |
+
"step": 165
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.12253233492171545,
|
| 90 |
+
"grad_norm": 78665.6015625,
|
| 91 |
+
"learning_rate": 7.346938775510204e-07,
|
| 92 |
+
"loss": 2.0133,
|
| 93 |
+
"step": 180
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.13274336283185842,
|
| 97 |
+
"grad_norm": 109804.5546875,
|
| 98 |
+
"learning_rate": 7.959183673469389e-07,
|
| 99 |
+
"loss": 2.0033,
|
| 100 |
+
"step": 195
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.14295439074200136,
|
| 104 |
+
"grad_norm": 88988.0234375,
|
| 105 |
+
"learning_rate": 8.571428571428571e-07,
|
| 106 |
+
"loss": 1.9912,
|
| 107 |
+
"step": 210
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.1531654186521443,
|
| 111 |
+
"grad_norm": 163108.234375,
|
| 112 |
+
"learning_rate": 9.183673469387755e-07,
|
| 113 |
+
"loss": 1.9714,
|
| 114 |
+
"step": 225
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.16337644656228728,
|
| 118 |
+
"grad_norm": 147248.25,
|
| 119 |
+
"learning_rate": 9.795918367346939e-07,
|
| 120 |
+
"loss": 1.9368,
|
| 121 |
+
"step": 240
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.17358747447243023,
|
| 125 |
+
"grad_norm": 161750.640625,
|
| 126 |
+
"learning_rate": 1.0408163265306123e-06,
|
| 127 |
+
"loss": 1.8955,
|
| 128 |
+
"step": 255
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.18379850238257317,
|
| 132 |
+
"grad_norm": 189184.296875,
|
| 133 |
+
"learning_rate": 1.1020408163265306e-06,
|
| 134 |
+
"loss": 1.8849,
|
| 135 |
+
"step": 270
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.19400953029271614,
|
| 139 |
+
"grad_norm": 160164.859375,
|
| 140 |
+
"learning_rate": 1.163265306122449e-06,
|
| 141 |
+
"loss": 1.8496,
|
| 142 |
+
"step": 285
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.2042205582028591,
|
| 146 |
+
"grad_norm": 147315.296875,
|
| 147 |
+
"learning_rate": 1.2244897959183673e-06,
|
| 148 |
+
"loss": 1.8118,
|
| 149 |
+
"step": 300
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.21443158611300203,
|
| 153 |
+
"grad_norm": 164890.859375,
|
| 154 |
+
"learning_rate": 1.2857142857142856e-06,
|
| 155 |
+
"loss": 1.7829,
|
| 156 |
+
"step": 315
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.224642614023145,
|
| 160 |
+
"grad_norm": 195027.578125,
|
| 161 |
+
"learning_rate": 1.346938775510204e-06,
|
| 162 |
+
"loss": 1.7938,
|
| 163 |
+
"step": 330
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.23485364193328795,
|
| 167 |
+
"grad_norm": 206460.03125,
|
| 168 |
+
"learning_rate": 1.4081632653061225e-06,
|
| 169 |
+
"loss": 1.7258,
|
| 170 |
+
"step": 345
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.2450646698434309,
|
| 174 |
+
"grad_norm": 247498.296875,
|
| 175 |
+
"learning_rate": 1.4693877551020408e-06,
|
| 176 |
+
"loss": 1.6498,
|
| 177 |
+
"step": 360
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.25527569775357384,
|
| 181 |
+
"grad_norm": 282144.96875,
|
| 182 |
+
"learning_rate": 1.5306122448979593e-06,
|
| 183 |
+
"loss": 1.6515,
|
| 184 |
+
"step": 375
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.26548672566371684,
|
| 188 |
+
"grad_norm": 244598.515625,
|
| 189 |
+
"learning_rate": 1.5918367346938777e-06,
|
| 190 |
+
"loss": 1.6327,
|
| 191 |
+
"step": 390
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.2756977535738598,
|
| 195 |
+
"grad_norm": 169616.5,
|
| 196 |
+
"learning_rate": 1.6530612244897958e-06,
|
| 197 |
+
"loss": 1.576,
|
| 198 |
+
"step": 405
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.2859087814840027,
|
| 202 |
+
"grad_norm": 186143.078125,
|
| 203 |
+
"learning_rate": 1.7142857142857143e-06,
|
| 204 |
+
"loss": 1.6067,
|
| 205 |
+
"step": 420
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.29611980939414567,
|
| 209 |
+
"grad_norm": 239726.171875,
|
| 210 |
+
"learning_rate": 1.7755102040816327e-06,
|
| 211 |
+
"loss": 1.5676,
|
| 212 |
+
"step": 435
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.3063308373042886,
|
| 216 |
+
"grad_norm": 224399.15625,
|
| 217 |
+
"learning_rate": 1.836734693877551e-06,
|
| 218 |
+
"loss": 1.4658,
|
| 219 |
+
"step": 450
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.31654186521443156,
|
| 223 |
+
"grad_norm": 585805.375,
|
| 224 |
+
"learning_rate": 1.8979591836734695e-06,
|
| 225 |
+
"loss": 1.4288,
|
| 226 |
+
"step": 465
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.32675289312457456,
|
| 230 |
+
"grad_norm": 204082.328125,
|
| 231 |
+
"learning_rate": 1.9591836734693877e-06,
|
| 232 |
+
"loss": 1.4079,
|
| 233 |
+
"step": 480
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.3369639210347175,
|
| 237 |
+
"grad_norm": 253897.078125,
|
| 238 |
+
"learning_rate": 2.020408163265306e-06,
|
| 239 |
+
"loss": 1.4658,
|
| 240 |
+
"step": 495
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.34717494894486045,
|
| 244 |
+
"grad_norm": 322156.15625,
|
| 245 |
+
"learning_rate": 2.0816326530612247e-06,
|
| 246 |
+
"loss": 1.3193,
|
| 247 |
+
"step": 510
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.3573859768550034,
|
| 251 |
+
"grad_norm": 209002.03125,
|
| 252 |
+
"learning_rate": 2.142857142857143e-06,
|
| 253 |
+
"loss": 1.3258,
|
| 254 |
+
"step": 525
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.36759700476514634,
|
| 258 |
+
"grad_norm": 239622.390625,
|
| 259 |
+
"learning_rate": 2.204081632653061e-06,
|
| 260 |
+
"loss": 1.3396,
|
| 261 |
+
"step": 540
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.3778080326752893,
|
| 265 |
+
"grad_norm": 302531.125,
|
| 266 |
+
"learning_rate": 2.2653061224489797e-06,
|
| 267 |
+
"loss": 1.2724,
|
| 268 |
+
"step": 555
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.3880190605854323,
|
| 272 |
+
"grad_norm": 277807.9375,
|
| 273 |
+
"learning_rate": 2.326530612244898e-06,
|
| 274 |
+
"loss": 1.2506,
|
| 275 |
+
"step": 570
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.39823008849557523,
|
| 279 |
+
"grad_norm": 543051.25,
|
| 280 |
+
"learning_rate": 2.387755102040816e-06,
|
| 281 |
+
"loss": 1.3114,
|
| 282 |
+
"step": 585
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.4084411164057182,
|
| 286 |
+
"grad_norm": 278458.59375,
|
| 287 |
+
"learning_rate": 2.4489795918367347e-06,
|
| 288 |
+
"loss": 1.2505,
|
| 289 |
+
"step": 600
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.4186521443158611,
|
| 293 |
+
"grad_norm": 227854.078125,
|
| 294 |
+
"learning_rate": 2.510204081632653e-06,
|
| 295 |
+
"loss": 1.2032,
|
| 296 |
+
"step": 615
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.42886317222600406,
|
| 300 |
+
"grad_norm": 669112.0625,
|
| 301 |
+
"learning_rate": 2.571428571428571e-06,
|
| 302 |
+
"loss": 1.2144,
|
| 303 |
+
"step": 630
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.43907420013614706,
|
| 307 |
+
"grad_norm": 291158.375,
|
| 308 |
+
"learning_rate": 2.6326530612244897e-06,
|
| 309 |
+
"loss": 1.1663,
|
| 310 |
+
"step": 645
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.44928522804629,
|
| 314 |
+
"grad_norm": 322840.65625,
|
| 315 |
+
"learning_rate": 2.693877551020408e-06,
|
| 316 |
+
"loss": 1.1802,
|
| 317 |
+
"step": 660
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.45949625595643295,
|
| 321 |
+
"grad_norm": 399617.4375,
|
| 322 |
+
"learning_rate": 2.7551020408163266e-06,
|
| 323 |
+
"loss": 1.0181,
|
| 324 |
+
"step": 675
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.4697072838665759,
|
| 328 |
+
"grad_norm": 399545.5,
|
| 329 |
+
"learning_rate": 2.816326530612245e-06,
|
| 330 |
+
"loss": 1.2468,
|
| 331 |
+
"step": 690
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.47991831177671884,
|
| 335 |
+
"grad_norm": 1116027.875,
|
| 336 |
+
"learning_rate": 2.877551020408163e-06,
|
| 337 |
+
"loss": 1.2023,
|
| 338 |
+
"step": 705
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.4901293396868618,
|
| 342 |
+
"grad_norm": 300487.4375,
|
| 343 |
+
"learning_rate": 2.9387755102040816e-06,
|
| 344 |
+
"loss": 1.147,
|
| 345 |
+
"step": 720
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.5003403675970047,
|
| 349 |
+
"grad_norm": 413331.65625,
|
| 350 |
+
"learning_rate": 3e-06,
|
| 351 |
+
"loss": 0.9871,
|
| 352 |
+
"step": 735
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.5105513955071477,
|
| 356 |
+
"grad_norm": 715814.125,
|
| 357 |
+
"learning_rate": 2.993192133131619e-06,
|
| 358 |
+
"loss": 1.1643,
|
| 359 |
+
"step": 750
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.5207624234172906,
|
| 363 |
+
"grad_norm": 1614147.0,
|
| 364 |
+
"learning_rate": 2.9863842662632373e-06,
|
| 365 |
+
"loss": 0.9421,
|
| 366 |
+
"step": 765
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.5309734513274337,
|
| 370 |
+
"grad_norm": 639895.0625,
|
| 371 |
+
"learning_rate": 2.9795763993948565e-06,
|
| 372 |
+
"loss": 1.0573,
|
| 373 |
+
"step": 780
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.5411844792375766,
|
| 377 |
+
"grad_norm": 345273.84375,
|
| 378 |
+
"learning_rate": 2.9727685325264754e-06,
|
| 379 |
+
"loss": 1.0337,
|
| 380 |
+
"step": 795
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.5513955071477196,
|
| 384 |
+
"grad_norm": 611370.375,
|
| 385 |
+
"learning_rate": 2.9659606656580938e-06,
|
| 386 |
+
"loss": 1.1131,
|
| 387 |
+
"step": 810
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.5616065350578625,
|
| 391 |
+
"grad_norm": 239903.25,
|
| 392 |
+
"learning_rate": 2.9591527987897126e-06,
|
| 393 |
+
"loss": 0.9076,
|
| 394 |
+
"step": 825
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.5718175629680055,
|
| 398 |
+
"grad_norm": 257958.703125,
|
| 399 |
+
"learning_rate": 2.9523449319213314e-06,
|
| 400 |
+
"loss": 0.9152,
|
| 401 |
+
"step": 840
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.5820285908781484,
|
| 405 |
+
"grad_norm": 590921.375,
|
| 406 |
+
"learning_rate": 2.9455370650529502e-06,
|
| 407 |
+
"loss": 0.8795,
|
| 408 |
+
"step": 855
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.5922396187882913,
|
| 412 |
+
"grad_norm": 363986.1875,
|
| 413 |
+
"learning_rate": 2.938729198184569e-06,
|
| 414 |
+
"loss": 1.0641,
|
| 415 |
+
"step": 870
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.6024506466984343,
|
| 419 |
+
"grad_norm": 1878926.875,
|
| 420 |
+
"learning_rate": 2.9319213313161875e-06,
|
| 421 |
+
"loss": 0.9414,
|
| 422 |
+
"step": 885
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.6126616746085772,
|
| 426 |
+
"grad_norm": 2941328.0,
|
| 427 |
+
"learning_rate": 2.9251134644478063e-06,
|
| 428 |
+
"loss": 0.9541,
|
| 429 |
+
"step": 900
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.6228727025187202,
|
| 433 |
+
"grad_norm": 1327942.75,
|
| 434 |
+
"learning_rate": 2.918305597579425e-06,
|
| 435 |
+
"loss": 0.8319,
|
| 436 |
+
"step": 915
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.6330837304288631,
|
| 440 |
+
"grad_norm": 161807.125,
|
| 441 |
+
"learning_rate": 2.911497730711044e-06,
|
| 442 |
+
"loss": 0.9268,
|
| 443 |
+
"step": 930
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.6432947583390062,
|
| 447 |
+
"grad_norm": 899555.875,
|
| 448 |
+
"learning_rate": 2.9046898638426628e-06,
|
| 449 |
+
"loss": 0.8278,
|
| 450 |
+
"step": 945
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.6535057862491491,
|
| 454 |
+
"grad_norm": 268360.0625,
|
| 455 |
+
"learning_rate": 2.8978819969742816e-06,
|
| 456 |
+
"loss": 0.9021,
|
| 457 |
+
"step": 960
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.6637168141592921,
|
| 461 |
+
"grad_norm": 555703.0625,
|
| 462 |
+
"learning_rate": 2.8910741301059e-06,
|
| 463 |
+
"loss": 1.0146,
|
| 464 |
+
"step": 975
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.673927842069435,
|
| 468 |
+
"grad_norm": 222425.40625,
|
| 469 |
+
"learning_rate": 2.884266263237519e-06,
|
| 470 |
+
"loss": 1.0005,
|
| 471 |
+
"step": 990
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.684138869979578,
|
| 475 |
+
"grad_norm": 681254.4375,
|
| 476 |
+
"learning_rate": 2.877458396369138e-06,
|
| 477 |
+
"loss": 0.842,
|
| 478 |
+
"step": 1005
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.6943498978897209,
|
| 482 |
+
"grad_norm": 1348221.125,
|
| 483 |
+
"learning_rate": 2.8706505295007565e-06,
|
| 484 |
+
"loss": 0.9394,
|
| 485 |
+
"step": 1020
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.7045609257998638,
|
| 489 |
+
"grad_norm": 225728.046875,
|
| 490 |
+
"learning_rate": 2.8638426626323753e-06,
|
| 491 |
+
"loss": 0.8289,
|
| 492 |
+
"step": 1035
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.7147719537100068,
|
| 496 |
+
"grad_norm": 413089.65625,
|
| 497 |
+
"learning_rate": 2.857034795763994e-06,
|
| 498 |
+
"loss": 1.0175,
|
| 499 |
+
"step": 1050
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.7249829816201497,
|
| 503 |
+
"grad_norm": 279751.71875,
|
| 504 |
+
"learning_rate": 2.8502269288956125e-06,
|
| 505 |
+
"loss": 0.9563,
|
| 506 |
+
"step": 1065
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.7351940095302927,
|
| 510 |
+
"grad_norm": 422167.125,
|
| 511 |
+
"learning_rate": 2.8434190620272318e-06,
|
| 512 |
+
"loss": 0.8566,
|
| 513 |
+
"step": 1080
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.7454050374404356,
|
| 517 |
+
"grad_norm": 2015877.125,
|
| 518 |
+
"learning_rate": 2.83661119515885e-06,
|
| 519 |
+
"loss": 0.7012,
|
| 520 |
+
"step": 1095
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.7556160653505786,
|
| 524 |
+
"grad_norm": 502859.5,
|
| 525 |
+
"learning_rate": 2.829803328290469e-06,
|
| 526 |
+
"loss": 0.7563,
|
| 527 |
+
"step": 1110
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.7658270932607216,
|
| 531 |
+
"grad_norm": 537785.125,
|
| 532 |
+
"learning_rate": 2.822995461422088e-06,
|
| 533 |
+
"loss": 0.9854,
|
| 534 |
+
"step": 1125
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.7760381211708646,
|
| 538 |
+
"grad_norm": 390231.125,
|
| 539 |
+
"learning_rate": 2.8161875945537066e-06,
|
| 540 |
+
"loss": 1.0036,
|
| 541 |
+
"step": 1140
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.7862491490810075,
|
| 545 |
+
"grad_norm": 370246.0625,
|
| 546 |
+
"learning_rate": 2.8093797276853255e-06,
|
| 547 |
+
"loss": 0.8186,
|
| 548 |
+
"step": 1155
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.7964601769911505,
|
| 552 |
+
"grad_norm": 1279901.375,
|
| 553 |
+
"learning_rate": 2.8025718608169443e-06,
|
| 554 |
+
"loss": 0.9289,
|
| 555 |
+
"step": 1170
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.8066712049012934,
|
| 559 |
+
"grad_norm": 316319.25,
|
| 560 |
+
"learning_rate": 2.7957639939485627e-06,
|
| 561 |
+
"loss": 0.7542,
|
| 562 |
+
"step": 1185
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.8168822328114363,
|
| 566 |
+
"grad_norm": 577071.5,
|
| 567 |
+
"learning_rate": 2.7889561270801815e-06,
|
| 568 |
+
"loss": 0.8579,
|
| 569 |
+
"step": 1200
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.8270932607215793,
|
| 573 |
+
"grad_norm": 412667.71875,
|
| 574 |
+
"learning_rate": 2.7821482602118003e-06,
|
| 575 |
+
"loss": 0.8424,
|
| 576 |
+
"step": 1215
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.8373042886317222,
|
| 580 |
+
"grad_norm": 555197.125,
|
| 581 |
+
"learning_rate": 2.775340393343419e-06,
|
| 582 |
+
"loss": 0.8225,
|
| 583 |
+
"step": 1230
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.8475153165418652,
|
| 587 |
+
"grad_norm": 429715.0,
|
| 588 |
+
"learning_rate": 2.768532526475038e-06,
|
| 589 |
+
"loss": 1.1042,
|
| 590 |
+
"step": 1245
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.8577263444520081,
|
| 594 |
+
"grad_norm": 1407773.25,
|
| 595 |
+
"learning_rate": 2.761724659606657e-06,
|
| 596 |
+
"loss": 0.9892,
|
| 597 |
+
"step": 1260
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.8679373723621511,
|
| 601 |
+
"grad_norm": 328839.6875,
|
| 602 |
+
"learning_rate": 2.754916792738275e-06,
|
| 603 |
+
"loss": 0.7921,
|
| 604 |
+
"step": 1275
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.8781484002722941,
|
| 608 |
+
"grad_norm": 196251.03125,
|
| 609 |
+
"learning_rate": 2.748108925869894e-06,
|
| 610 |
+
"loss": 0.6631,
|
| 611 |
+
"step": 1290
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.8883594281824371,
|
| 615 |
+
"grad_norm": 761081.6875,
|
| 616 |
+
"learning_rate": 2.7413010590015133e-06,
|
| 617 |
+
"loss": 0.8396,
|
| 618 |
+
"step": 1305
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.89857045609258,
|
| 622 |
+
"grad_norm": 905756.375,
|
| 623 |
+
"learning_rate": 2.7344931921331317e-06,
|
| 624 |
+
"loss": 0.8238,
|
| 625 |
+
"step": 1320
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.908781484002723,
|
| 629 |
+
"grad_norm": 1221640.875,
|
| 630 |
+
"learning_rate": 2.7276853252647505e-06,
|
| 631 |
+
"loss": 0.7228,
|
| 632 |
+
"step": 1335
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.9189925119128659,
|
| 636 |
+
"grad_norm": 2633817.0,
|
| 637 |
+
"learning_rate": 2.7208774583963693e-06,
|
| 638 |
+
"loss": 0.8766,
|
| 639 |
+
"step": 1350
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.9292035398230089,
|
| 643 |
+
"grad_norm": 454206.53125,
|
| 644 |
+
"learning_rate": 2.7140695915279877e-06,
|
| 645 |
+
"loss": 0.922,
|
| 646 |
+
"step": 1365
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.9394145677331518,
|
| 650 |
+
"grad_norm": 715028.8125,
|
| 651 |
+
"learning_rate": 2.707261724659607e-06,
|
| 652 |
+
"loss": 0.8212,
|
| 653 |
+
"step": 1380
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.9496255956432947,
|
| 657 |
+
"grad_norm": 384052.40625,
|
| 658 |
+
"learning_rate": 2.7004538577912254e-06,
|
| 659 |
+
"loss": 0.6728,
|
| 660 |
+
"step": 1395
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.9598366235534377,
|
| 664 |
+
"grad_norm": 268327.0,
|
| 665 |
+
"learning_rate": 2.693645990922844e-06,
|
| 666 |
+
"loss": 0.7077,
|
| 667 |
+
"step": 1410
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.9700476514635806,
|
| 671 |
+
"grad_norm": 430912.0625,
|
| 672 |
+
"learning_rate": 2.686838124054463e-06,
|
| 673 |
+
"loss": 0.8267,
|
| 674 |
+
"step": 1425
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.9802586793737236,
|
| 678 |
+
"grad_norm": 448801.65625,
|
| 679 |
+
"learning_rate": 2.680030257186082e-06,
|
| 680 |
+
"loss": 0.6444,
|
| 681 |
+
"step": 1440
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.9904697072838666,
|
| 685 |
+
"grad_norm": 625466.3125,
|
| 686 |
+
"learning_rate": 2.6732223903177007e-06,
|
| 687 |
+
"loss": 0.8255,
|
| 688 |
+
"step": 1455
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.0,
|
| 692 |
+
"eval_accuracy": 0.8819584171696848,
|
| 693 |
+
"eval_loss": 0.5119155645370483,
|
| 694 |
+
"eval_runtime": 579.0751,
|
| 695 |
+
"eval_samples_per_second": 5.15,
|
| 696 |
+
"eval_steps_per_second": 0.644,
|
| 697 |
+
"step": 1469
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.0006807351940095,
|
| 701 |
+
"grad_norm": 166140.59375,
|
| 702 |
+
"learning_rate": 2.6664145234493195e-06,
|
| 703 |
+
"loss": 0.8523,
|
| 704 |
+
"step": 1470
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.0108917631041525,
|
| 708 |
+
"grad_norm": 1578961.375,
|
| 709 |
+
"learning_rate": 2.659606656580938e-06,
|
| 710 |
+
"loss": 0.8466,
|
| 711 |
+
"step": 1485
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.0211027910142954,
|
| 715 |
+
"grad_norm": 3046324.75,
|
| 716 |
+
"learning_rate": 2.6527987897125567e-06,
|
| 717 |
+
"loss": 0.7421,
|
| 718 |
+
"step": 1500
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.0313138189244384,
|
| 722 |
+
"grad_norm": 486359.28125,
|
| 723 |
+
"learning_rate": 2.6459909228441755e-06,
|
| 724 |
+
"loss": 0.586,
|
| 725 |
+
"step": 1515
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.0415248468345812,
|
| 729 |
+
"grad_norm": 476911.5,
|
| 730 |
+
"learning_rate": 2.6391830559757944e-06,
|
| 731 |
+
"loss": 0.7123,
|
| 732 |
+
"step": 1530
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.0517358747447243,
|
| 736 |
+
"grad_norm": 312268.75,
|
| 737 |
+
"learning_rate": 2.632375189107413e-06,
|
| 738 |
+
"loss": 0.7437,
|
| 739 |
+
"step": 1545
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.0619469026548674,
|
| 743 |
+
"grad_norm": 555973.375,
|
| 744 |
+
"learning_rate": 2.625567322239032e-06,
|
| 745 |
+
"loss": 0.7994,
|
| 746 |
+
"step": 1560
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.0721579305650102,
|
| 750 |
+
"grad_norm": 441313.59375,
|
| 751 |
+
"learning_rate": 2.6187594553706504e-06,
|
| 752 |
+
"loss": 0.7675,
|
| 753 |
+
"step": 1575
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.0823689584751532,
|
| 757 |
+
"grad_norm": 2035757.875,
|
| 758 |
+
"learning_rate": 2.6119515885022692e-06,
|
| 759 |
+
"loss": 0.7534,
|
| 760 |
+
"step": 1590
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.092579986385296,
|
| 764 |
+
"grad_norm": 453271.59375,
|
| 765 |
+
"learning_rate": 2.605143721633888e-06,
|
| 766 |
+
"loss": 0.645,
|
| 767 |
+
"step": 1605
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.1027910142954391,
|
| 771 |
+
"grad_norm": 319587.0625,
|
| 772 |
+
"learning_rate": 2.598335854765507e-06,
|
| 773 |
+
"loss": 0.765,
|
| 774 |
+
"step": 1620
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.113002042205582,
|
| 778 |
+
"grad_norm": 732482.8125,
|
| 779 |
+
"learning_rate": 2.5915279878971257e-06,
|
| 780 |
+
"loss": 0.6155,
|
| 781 |
+
"step": 1635
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.123213070115725,
|
| 785 |
+
"grad_norm": 396029.65625,
|
| 786 |
+
"learning_rate": 2.5847201210287445e-06,
|
| 787 |
+
"loss": 0.7213,
|
| 788 |
+
"step": 1650
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.1334240980258679,
|
| 792 |
+
"grad_norm": 326143.03125,
|
| 793 |
+
"learning_rate": 2.577912254160363e-06,
|
| 794 |
+
"loss": 0.7134,
|
| 795 |
+
"step": 1665
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.143635125936011,
|
| 799 |
+
"grad_norm": 1083100.875,
|
| 800 |
+
"learning_rate": 2.571104387291982e-06,
|
| 801 |
+
"loss": 0.7249,
|
| 802 |
+
"step": 1680
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.1538461538461537,
|
| 806 |
+
"grad_norm": 905634.25,
|
| 807 |
+
"learning_rate": 2.5642965204236006e-06,
|
| 808 |
+
"loss": 0.767,
|
| 809 |
+
"step": 1695
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.1640571817562968,
|
| 813 |
+
"grad_norm": 656056.9375,
|
| 814 |
+
"learning_rate": 2.5574886535552194e-06,
|
| 815 |
+
"loss": 0.8304,
|
| 816 |
+
"step": 1710
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.1742682096664399,
|
| 820 |
+
"grad_norm": 722782.5625,
|
| 821 |
+
"learning_rate": 2.5506807866868382e-06,
|
| 822 |
+
"loss": 0.6995,
|
| 823 |
+
"step": 1725
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.1844792375765827,
|
| 827 |
+
"grad_norm": 577313.5625,
|
| 828 |
+
"learning_rate": 2.5438729198184566e-06,
|
| 829 |
+
"loss": 0.9264,
|
| 830 |
+
"step": 1740
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.1946902654867257,
|
| 834 |
+
"grad_norm": 2527231.5,
|
| 835 |
+
"learning_rate": 2.537065052950076e-06,
|
| 836 |
+
"loss": 0.9283,
|
| 837 |
+
"step": 1755
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 1.2049012933968686,
|
| 841 |
+
"grad_norm": 1089097.5,
|
| 842 |
+
"learning_rate": 2.5302571860816947e-06,
|
| 843 |
+
"loss": 0.6552,
|
| 844 |
+
"step": 1770
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 1.2151123213070116,
|
| 848 |
+
"grad_norm": 747680.25,
|
| 849 |
+
"learning_rate": 2.523449319213313e-06,
|
| 850 |
+
"loss": 0.7181,
|
| 851 |
+
"step": 1785
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 1.2253233492171545,
|
| 855 |
+
"grad_norm": 143748.9375,
|
| 856 |
+
"learning_rate": 2.516641452344932e-06,
|
| 857 |
+
"loss": 0.5906,
|
| 858 |
+
"step": 1800
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 1.2355343771272975,
|
| 862 |
+
"grad_norm": 782987.3125,
|
| 863 |
+
"learning_rate": 2.5098335854765507e-06,
|
| 864 |
+
"loss": 0.7338,
|
| 865 |
+
"step": 1815
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 1.2457454050374404,
|
| 869 |
+
"grad_norm": 199057.203125,
|
| 870 |
+
"learning_rate": 2.5030257186081696e-06,
|
| 871 |
+
"loss": 0.6712,
|
| 872 |
+
"step": 1830
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.2559564329475834,
|
| 876 |
+
"grad_norm": 169465.65625,
|
| 877 |
+
"learning_rate": 2.4962178517397884e-06,
|
| 878 |
+
"loss": 0.4682,
|
| 879 |
+
"step": 1845
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.2661674608577265,
|
| 883 |
+
"grad_norm": 458018.5,
|
| 884 |
+
"learning_rate": 2.4894099848714072e-06,
|
| 885 |
+
"loss": 0.7659,
|
| 886 |
+
"step": 1860
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.2763784887678693,
|
| 890 |
+
"grad_norm": 373506.46875,
|
| 891 |
+
"learning_rate": 2.4826021180030256e-06,
|
| 892 |
+
"loss": 0.7384,
|
| 893 |
+
"step": 1875
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 1.2865895166780121,
|
| 897 |
+
"grad_norm": 140858.234375,
|
| 898 |
+
"learning_rate": 2.4757942511346444e-06,
|
| 899 |
+
"loss": 0.723,
|
| 900 |
+
"step": 1890
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.2968005445881552,
|
| 904 |
+
"grad_norm": 758790.1875,
|
| 905 |
+
"learning_rate": 2.4689863842662633e-06,
|
| 906 |
+
"loss": 0.9274,
|
| 907 |
+
"step": 1905
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.3070115724982982,
|
| 911 |
+
"grad_norm": 311553.5625,
|
| 912 |
+
"learning_rate": 2.462178517397882e-06,
|
| 913 |
+
"loss": 0.6113,
|
| 914 |
+
"step": 1920
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.317222600408441,
|
| 918 |
+
"grad_norm": 167712.234375,
|
| 919 |
+
"learning_rate": 2.455370650529501e-06,
|
| 920 |
+
"loss": 0.6976,
|
| 921 |
+
"step": 1935
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.3274336283185841,
|
| 925 |
+
"grad_norm": 500111.15625,
|
| 926 |
+
"learning_rate": 2.4485627836611193e-06,
|
| 927 |
+
"loss": 0.7833,
|
| 928 |
+
"step": 1950
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.337644656228727,
|
| 932 |
+
"grad_norm": 517175.0,
|
| 933 |
+
"learning_rate": 2.441754916792738e-06,
|
| 934 |
+
"loss": 0.6971,
|
| 935 |
+
"step": 1965
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.34785568413887,
|
| 939 |
+
"grad_norm": 659138.6875,
|
| 940 |
+
"learning_rate": 2.4349470499243574e-06,
|
| 941 |
+
"loss": 0.7602,
|
| 942 |
+
"step": 1980
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.3580667120490129,
|
| 946 |
+
"grad_norm": 301056.09375,
|
| 947 |
+
"learning_rate": 2.4281391830559758e-06,
|
| 948 |
+
"loss": 0.6703,
|
| 949 |
+
"step": 1995
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.368277739959156,
|
| 953 |
+
"grad_norm": 273580.0625,
|
| 954 |
+
"learning_rate": 2.4213313161875946e-06,
|
| 955 |
+
"loss": 0.6183,
|
| 956 |
+
"step": 2010
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.3784887678692987,
|
| 960 |
+
"grad_norm": 559704.875,
|
| 961 |
+
"learning_rate": 2.4145234493192134e-06,
|
| 962 |
+
"loss": 0.564,
|
| 963 |
+
"step": 2025
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.3886997957794418,
|
| 967 |
+
"grad_norm": 593091.6875,
|
| 968 |
+
"learning_rate": 2.407715582450832e-06,
|
| 969 |
+
"loss": 0.605,
|
| 970 |
+
"step": 2040
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.3989108236895849,
|
| 974 |
+
"grad_norm": 307976.4375,
|
| 975 |
+
"learning_rate": 2.400907715582451e-06,
|
| 976 |
+
"loss": 0.771,
|
| 977 |
+
"step": 2055
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 1.4091218515997277,
|
| 981 |
+
"grad_norm": 593720.625,
|
| 982 |
+
"learning_rate": 2.39409984871407e-06,
|
| 983 |
+
"loss": 0.7066,
|
| 984 |
+
"step": 2070
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.4193328795098705,
|
| 988 |
+
"grad_norm": 831806.0625,
|
| 989 |
+
"learning_rate": 2.3872919818456883e-06,
|
| 990 |
+
"loss": 0.5807,
|
| 991 |
+
"step": 2085
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 1.4295439074200136,
|
| 995 |
+
"grad_norm": 206191.21875,
|
| 996 |
+
"learning_rate": 2.380484114977307e-06,
|
| 997 |
+
"loss": 0.6574,
|
| 998 |
+
"step": 2100
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.4397549353301566,
|
| 1002 |
+
"grad_norm": 258367.828125,
|
| 1003 |
+
"learning_rate": 2.373676248108926e-06,
|
| 1004 |
+
"loss": 0.6792,
|
| 1005 |
+
"step": 2115
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.4499659632402995,
|
| 1009 |
+
"grad_norm": 1177121.25,
|
| 1010 |
+
"learning_rate": 2.3668683812405448e-06,
|
| 1011 |
+
"loss": 0.8608,
|
| 1012 |
+
"step": 2130
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.4601769911504425,
|
| 1016 |
+
"grad_norm": 368604.9375,
|
| 1017 |
+
"learning_rate": 2.3600605143721636e-06,
|
| 1018 |
+
"loss": 0.7224,
|
| 1019 |
+
"step": 2145
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 1.4703880190605854,
|
| 1023 |
+
"grad_norm": 570852.125,
|
| 1024 |
+
"learning_rate": 2.3532526475037824e-06,
|
| 1025 |
+
"loss": 0.6717,
|
| 1026 |
+
"step": 2160
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 1.4805990469707284,
|
| 1030 |
+
"grad_norm": 240292.90625,
|
| 1031 |
+
"learning_rate": 2.346444780635401e-06,
|
| 1032 |
+
"loss": 0.5415,
|
| 1033 |
+
"step": 2175
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 1.4908100748808715,
|
| 1037 |
+
"grad_norm": 1409754.875,
|
| 1038 |
+
"learning_rate": 2.3396369137670197e-06,
|
| 1039 |
+
"loss": 0.66,
|
| 1040 |
+
"step": 2190
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.5010211027910143,
|
| 1044 |
+
"grad_norm": 1670060.75,
|
| 1045 |
+
"learning_rate": 2.3328290468986385e-06,
|
| 1046 |
+
"loss": 0.686,
|
| 1047 |
+
"step": 2205
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 1.5112321307011571,
|
| 1051 |
+
"grad_norm": 530769.5,
|
| 1052 |
+
"learning_rate": 2.3260211800302573e-06,
|
| 1053 |
+
"loss": 0.612,
|
| 1054 |
+
"step": 2220
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 1.5214431586113002,
|
| 1058 |
+
"grad_norm": 1758232.75,
|
| 1059 |
+
"learning_rate": 2.319213313161876e-06,
|
| 1060 |
+
"loss": 0.6849,
|
| 1061 |
+
"step": 2235
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.5316541865214433,
|
| 1065 |
+
"grad_norm": 1104782.0,
|
| 1066 |
+
"learning_rate": 2.3124054462934945e-06,
|
| 1067 |
+
"loss": 0.7704,
|
| 1068 |
+
"step": 2250
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 1.541865214431586,
|
| 1072 |
+
"grad_norm": 826967.125,
|
| 1073 |
+
"learning_rate": 2.3055975794251133e-06,
|
| 1074 |
+
"loss": 0.5943,
|
| 1075 |
+
"step": 2265
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 1.552076242341729,
|
| 1079 |
+
"grad_norm": 1028021.3125,
|
| 1080 |
+
"learning_rate": 2.2987897125567326e-06,
|
| 1081 |
+
"loss": 0.6579,
|
| 1082 |
+
"step": 2280
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 1.562287270251872,
|
| 1086 |
+
"grad_norm": 250268.609375,
|
| 1087 |
+
"learning_rate": 2.291981845688351e-06,
|
| 1088 |
+
"loss": 0.671,
|
| 1089 |
+
"step": 2295
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 1.572498298162015,
|
| 1093 |
+
"grad_norm": 608503.5625,
|
| 1094 |
+
"learning_rate": 2.28517397881997e-06,
|
| 1095 |
+
"loss": 0.546,
|
| 1096 |
+
"step": 2310
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 1.582709326072158,
|
| 1100 |
+
"grad_norm": 2906189.5,
|
| 1101 |
+
"learning_rate": 2.2783661119515886e-06,
|
| 1102 |
+
"loss": 0.5798,
|
| 1103 |
+
"step": 2325
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 1.592920353982301,
|
| 1107 |
+
"grad_norm": 1186615.875,
|
| 1108 |
+
"learning_rate": 2.271558245083207e-06,
|
| 1109 |
+
"loss": 0.8234,
|
| 1110 |
+
"step": 2340
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 1.6031313818924438,
|
| 1114 |
+
"grad_norm": 189114.234375,
|
| 1115 |
+
"learning_rate": 2.2647503782148263e-06,
|
| 1116 |
+
"loss": 0.7017,
|
| 1117 |
+
"step": 2355
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 1.6133424098025868,
|
| 1121 |
+
"grad_norm": 1086454.375,
|
| 1122 |
+
"learning_rate": 2.257942511346445e-06,
|
| 1123 |
+
"loss": 0.6067,
|
| 1124 |
+
"step": 2370
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 1.6235534377127299,
|
| 1128 |
+
"grad_norm": 438766.8125,
|
| 1129 |
+
"learning_rate": 2.2511346444780635e-06,
|
| 1130 |
+
"loss": 0.7277,
|
| 1131 |
+
"step": 2385
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 1.6337644656228727,
|
| 1135 |
+
"grad_norm": 746603.625,
|
| 1136 |
+
"learning_rate": 2.2443267776096823e-06,
|
| 1137 |
+
"loss": 0.7049,
|
| 1138 |
+
"step": 2400
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 1.6439754935330155,
|
| 1142 |
+
"grad_norm": 1114921.375,
|
| 1143 |
+
"learning_rate": 2.237518910741301e-06,
|
| 1144 |
+
"loss": 0.6517,
|
| 1145 |
+
"step": 2415
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 1.6541865214431586,
|
| 1149 |
+
"grad_norm": 119118.2421875,
|
| 1150 |
+
"learning_rate": 2.23071104387292e-06,
|
| 1151 |
+
"loss": 0.6355,
|
| 1152 |
+
"step": 2430
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 1.6643975493533016,
|
| 1156 |
+
"grad_norm": 278499.84375,
|
| 1157 |
+
"learning_rate": 2.223903177004539e-06,
|
| 1158 |
+
"loss": 0.528,
|
| 1159 |
+
"step": 2445
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 1.6746085772634445,
|
| 1163 |
+
"grad_norm": 2842614.5,
|
| 1164 |
+
"learning_rate": 2.2170953101361572e-06,
|
| 1165 |
+
"loss": 0.818,
|
| 1166 |
+
"step": 2460
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 1.6848196051735873,
|
| 1170 |
+
"grad_norm": 681713.6875,
|
| 1171 |
+
"learning_rate": 2.210287443267776e-06,
|
| 1172 |
+
"loss": 0.5879,
|
| 1173 |
+
"step": 2475
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 1.6950306330837304,
|
| 1177 |
+
"grad_norm": 955702.4375,
|
| 1178 |
+
"learning_rate": 2.203479576399395e-06,
|
| 1179 |
+
"loss": 0.6868,
|
| 1180 |
+
"step": 2490
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 1.7052416609938734,
|
| 1184 |
+
"grad_norm": 293277.28125,
|
| 1185 |
+
"learning_rate": 2.1966717095310137e-06,
|
| 1186 |
+
"loss": 0.6262,
|
| 1187 |
+
"step": 2505
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 1.7154526889040165,
|
| 1191 |
+
"grad_norm": 1372940.875,
|
| 1192 |
+
"learning_rate": 2.1898638426626325e-06,
|
| 1193 |
+
"loss": 0.7799,
|
| 1194 |
+
"step": 2520
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 1.7256637168141593,
|
| 1198 |
+
"grad_norm": 277903.0625,
|
| 1199 |
+
"learning_rate": 2.1830559757942513e-06,
|
| 1200 |
+
"loss": 0.6204,
|
| 1201 |
+
"step": 2535
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 1.7358747447243021,
|
| 1205 |
+
"grad_norm": 959110.125,
|
| 1206 |
+
"learning_rate": 2.1762481089258697e-06,
|
| 1207 |
+
"loss": 0.7283,
|
| 1208 |
+
"step": 2550
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 1.7460857726344452,
|
| 1212 |
+
"grad_norm": 870804.125,
|
| 1213 |
+
"learning_rate": 2.1694402420574886e-06,
|
| 1214 |
+
"loss": 0.6073,
|
| 1215 |
+
"step": 2565
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 1.7562968005445883,
|
| 1219 |
+
"grad_norm": 723633.5,
|
| 1220 |
+
"learning_rate": 2.162632375189108e-06,
|
| 1221 |
+
"loss": 0.8315,
|
| 1222 |
+
"step": 2580
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 1.766507828454731,
|
| 1226 |
+
"grad_norm": 279254.0625,
|
| 1227 |
+
"learning_rate": 2.155824508320726e-06,
|
| 1228 |
+
"loss": 0.6998,
|
| 1229 |
+
"step": 2595
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 1.776718856364874,
|
| 1233 |
+
"grad_norm": 816422.0625,
|
| 1234 |
+
"learning_rate": 2.149016641452345e-06,
|
| 1235 |
+
"loss": 0.7695,
|
| 1236 |
+
"step": 2610
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 1.786929884275017,
|
| 1240 |
+
"grad_norm": 277112.28125,
|
| 1241 |
+
"learning_rate": 2.142208774583964e-06,
|
| 1242 |
+
"loss": 0.5903,
|
| 1243 |
+
"step": 2625
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 1.79714091218516,
|
| 1247 |
+
"grad_norm": 1158773.75,
|
| 1248 |
+
"learning_rate": 2.1354009077155823e-06,
|
| 1249 |
+
"loss": 0.5811,
|
| 1250 |
+
"step": 2640
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 1.807351940095303,
|
| 1254 |
+
"grad_norm": 3944267.75,
|
| 1255 |
+
"learning_rate": 2.1285930408472015e-06,
|
| 1256 |
+
"loss": 0.5903,
|
| 1257 |
+
"step": 2655
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 1.817562968005446,
|
| 1261 |
+
"grad_norm": 774781.9375,
|
| 1262 |
+
"learning_rate": 2.12178517397882e-06,
|
| 1263 |
+
"loss": 0.7741,
|
| 1264 |
+
"step": 2670
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 1.8277739959155888,
|
| 1268 |
+
"grad_norm": 744016.25,
|
| 1269 |
+
"learning_rate": 2.1149773071104387e-06,
|
| 1270 |
+
"loss": 0.7171,
|
| 1271 |
+
"step": 2685
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 1.8379850238257318,
|
| 1275 |
+
"grad_norm": 5070120.5,
|
| 1276 |
+
"learning_rate": 2.1081694402420575e-06,
|
| 1277 |
+
"loss": 0.5936,
|
| 1278 |
+
"step": 2700
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 1.8481960517358749,
|
| 1282 |
+
"grad_norm": 249426.0,
|
| 1283 |
+
"learning_rate": 2.1013615733736764e-06,
|
| 1284 |
+
"loss": 0.7609,
|
| 1285 |
+
"step": 2715
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 1.8584070796460177,
|
| 1289 |
+
"grad_norm": 403828.75,
|
| 1290 |
+
"learning_rate": 2.094553706505295e-06,
|
| 1291 |
+
"loss": 0.6597,
|
| 1292 |
+
"step": 2730
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 1.8686181075561605,
|
| 1296 |
+
"grad_norm": 829779.0625,
|
| 1297 |
+
"learning_rate": 2.087745839636914e-06,
|
| 1298 |
+
"loss": 0.6788,
|
| 1299 |
+
"step": 2745
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 1.8788291354663036,
|
| 1303 |
+
"grad_norm": 939866.5625,
|
| 1304 |
+
"learning_rate": 2.0809379727685324e-06,
|
| 1305 |
+
"loss": 0.5589,
|
| 1306 |
+
"step": 2760
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 1.8890401633764466,
|
| 1310 |
+
"grad_norm": 1203267.0,
|
| 1311 |
+
"learning_rate": 2.0741301059001512e-06,
|
| 1312 |
+
"loss": 0.7759,
|
| 1313 |
+
"step": 2775
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 1.8992511912865895,
|
| 1317 |
+
"grad_norm": 1656486.5,
|
| 1318 |
+
"learning_rate": 2.06732223903177e-06,
|
| 1319 |
+
"loss": 0.4373,
|
| 1320 |
+
"step": 2790
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 1.9094622191967323,
|
| 1324 |
+
"grad_norm": 1833536.875,
|
| 1325 |
+
"learning_rate": 2.060514372163389e-06,
|
| 1326 |
+
"loss": 1.0234,
|
| 1327 |
+
"step": 2805
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 1.9196732471068754,
|
| 1331 |
+
"grad_norm": 429027.65625,
|
| 1332 |
+
"learning_rate": 2.0537065052950077e-06,
|
| 1333 |
+
"loss": 0.5515,
|
| 1334 |
+
"step": 2820
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 1.9298842750170184,
|
| 1338 |
+
"grad_norm": 2480347.0,
|
| 1339 |
+
"learning_rate": 2.0468986384266265e-06,
|
| 1340 |
+
"loss": 0.679,
|
| 1341 |
+
"step": 2835
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 1.9400953029271615,
|
| 1345 |
+
"grad_norm": 682525.625,
|
| 1346 |
+
"learning_rate": 2.040090771558245e-06,
|
| 1347 |
+
"loss": 0.5025,
|
| 1348 |
+
"step": 2850
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 1.9503063308373043,
|
| 1352 |
+
"grad_norm": 289022.15625,
|
| 1353 |
+
"learning_rate": 2.0332829046898638e-06,
|
| 1354 |
+
"loss": 0.5361,
|
| 1355 |
+
"step": 2865
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 1.9605173587474471,
|
| 1359 |
+
"grad_norm": 633536.75,
|
| 1360 |
+
"learning_rate": 2.0264750378214826e-06,
|
| 1361 |
+
"loss": 0.6829,
|
| 1362 |
+
"step": 2880
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 1.9707283866575902,
|
| 1366 |
+
"grad_norm": 551145.0,
|
| 1367 |
+
"learning_rate": 2.0196671709531014e-06,
|
| 1368 |
+
"loss": 0.591,
|
| 1369 |
+
"step": 2895
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 1.9809394145677333,
|
| 1373 |
+
"grad_norm": 951443.5,
|
| 1374 |
+
"learning_rate": 2.0128593040847202e-06,
|
| 1375 |
+
"loss": 0.5114,
|
| 1376 |
+
"step": 2910
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 1.991150442477876,
|
| 1380 |
+
"grad_norm": 1422993.0,
|
| 1381 |
+
"learning_rate": 2.006051437216339e-06,
|
| 1382 |
+
"loss": 0.6774,
|
| 1383 |
+
"step": 2925
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 2.0,
|
| 1387 |
+
"eval_accuracy": 0.9124748490945674,
|
| 1388 |
+
"eval_loss": 0.3880421221256256,
|
| 1389 |
+
"eval_runtime": 580.6935,
|
| 1390 |
+
"eval_samples_per_second": 5.135,
|
| 1391 |
+
"eval_steps_per_second": 0.642,
|
| 1392 |
+
"step": 2938
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 2.001361470388019,
|
| 1396 |
+
"grad_norm": 149413.0625,
|
| 1397 |
+
"learning_rate": 1.9992435703479575e-06,
|
| 1398 |
+
"loss": 0.5516,
|
| 1399 |
+
"step": 2940
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 2.011572498298162,
|
| 1403 |
+
"grad_norm": 92282.625,
|
| 1404 |
+
"learning_rate": 1.9924357034795767e-06,
|
| 1405 |
+
"loss": 0.5141,
|
| 1406 |
+
"step": 2955
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 2.021783526208305,
|
| 1410 |
+
"grad_norm": 771094.375,
|
| 1411 |
+
"learning_rate": 1.985627836611195e-06,
|
| 1412 |
+
"loss": 0.5162,
|
| 1413 |
+
"step": 2970
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 2.031994554118448,
|
| 1417 |
+
"grad_norm": 1059304.125,
|
| 1418 |
+
"learning_rate": 1.978819969742814e-06,
|
| 1419 |
+
"loss": 0.5794,
|
| 1420 |
+
"step": 2985
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 2.0422055820285907,
|
| 1424 |
+
"grad_norm": 588466.6875,
|
| 1425 |
+
"learning_rate": 1.9720121028744328e-06,
|
| 1426 |
+
"loss": 0.6778,
|
| 1427 |
+
"step": 3000
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 2.0524166099387338,
|
| 1431 |
+
"grad_norm": 571242.3125,
|
| 1432 |
+
"learning_rate": 1.965204236006051e-06,
|
| 1433 |
+
"loss": 0.6637,
|
| 1434 |
+
"step": 3015
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 2.062627637848877,
|
| 1438 |
+
"grad_norm": 1182853.5,
|
| 1439 |
+
"learning_rate": 1.9583963691376704e-06,
|
| 1440 |
+
"loss": 0.4922,
|
| 1441 |
+
"step": 3030
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 2.07283866575902,
|
| 1445 |
+
"grad_norm": 669133.8125,
|
| 1446 |
+
"learning_rate": 1.9515885022692892e-06,
|
| 1447 |
+
"loss": 0.4657,
|
| 1448 |
+
"step": 3045
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 2.0830496936691625,
|
| 1452 |
+
"grad_norm": 323752.25,
|
| 1453 |
+
"learning_rate": 1.9447806354009076e-06,
|
| 1454 |
+
"loss": 0.596,
|
| 1455 |
+
"step": 3060
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 2.0932607215793055,
|
| 1459 |
+
"grad_norm": 1483114.375,
|
| 1460 |
+
"learning_rate": 1.9379727685325265e-06,
|
| 1461 |
+
"loss": 0.5028,
|
| 1462 |
+
"step": 3075
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 2.1034717494894486,
|
| 1466 |
+
"grad_norm": 214087.359375,
|
| 1467 |
+
"learning_rate": 1.9311649016641453e-06,
|
| 1468 |
+
"loss": 0.763,
|
| 1469 |
+
"step": 3090
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 2.1136827773995917,
|
| 1473 |
+
"grad_norm": 136251.953125,
|
| 1474 |
+
"learning_rate": 1.924357034795764e-06,
|
| 1475 |
+
"loss": 0.4622,
|
| 1476 |
+
"step": 3105
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 2.1238938053097347,
|
| 1480 |
+
"grad_norm": 132665.53125,
|
| 1481 |
+
"learning_rate": 1.917549167927383e-06,
|
| 1482 |
+
"loss": 0.5564,
|
| 1483 |
+
"step": 3120
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 2.1341048332198773,
|
| 1487 |
+
"grad_norm": 741566.0,
|
| 1488 |
+
"learning_rate": 1.9107413010590018e-06,
|
| 1489 |
+
"loss": 0.5569,
|
| 1490 |
+
"step": 3135
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 2.1443158611300204,
|
| 1494 |
+
"grad_norm": 612487.625,
|
| 1495 |
+
"learning_rate": 1.9039334341906204e-06,
|
| 1496 |
+
"loss": 0.576,
|
| 1497 |
+
"step": 3150
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 2.1545268890401634,
|
| 1501 |
+
"grad_norm": 953638.5625,
|
| 1502 |
+
"learning_rate": 1.897125567322239e-06,
|
| 1503 |
+
"loss": 0.3974,
|
| 1504 |
+
"step": 3165
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 2.1647379169503065,
|
| 1508 |
+
"grad_norm": 663514.3125,
|
| 1509 |
+
"learning_rate": 1.890317700453858e-06,
|
| 1510 |
+
"loss": 0.471,
|
| 1511 |
+
"step": 3180
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 2.174948944860449,
|
| 1515 |
+
"grad_norm": 197147.5,
|
| 1516 |
+
"learning_rate": 1.8835098335854766e-06,
|
| 1517 |
+
"loss": 0.6322,
|
| 1518 |
+
"step": 3195
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 2.185159972770592,
|
| 1522 |
+
"grad_norm": 61634.29296875,
|
| 1523 |
+
"learning_rate": 1.8767019667170954e-06,
|
| 1524 |
+
"loss": 0.5017,
|
| 1525 |
+
"step": 3210
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 2.195371000680735,
|
| 1529 |
+
"grad_norm": 309183.3125,
|
| 1530 |
+
"learning_rate": 1.869894099848714e-06,
|
| 1531 |
+
"loss": 0.5658,
|
| 1532 |
+
"step": 3225
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 2.2055820285908783,
|
| 1536 |
+
"grad_norm": 124933.015625,
|
| 1537 |
+
"learning_rate": 1.8630862329803327e-06,
|
| 1538 |
+
"loss": 0.4846,
|
| 1539 |
+
"step": 3240
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 2.2157930565010213,
|
| 1543 |
+
"grad_norm": 547002.375,
|
| 1544 |
+
"learning_rate": 1.8562783661119517e-06,
|
| 1545 |
+
"loss": 0.4722,
|
| 1546 |
+
"step": 3255
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 2.226004084411164,
|
| 1550 |
+
"grad_norm": 220335.03125,
|
| 1551 |
+
"learning_rate": 1.8494704992435705e-06,
|
| 1552 |
+
"loss": 0.4243,
|
| 1553 |
+
"step": 3270
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 2.236215112321307,
|
| 1557 |
+
"grad_norm": 882093.1875,
|
| 1558 |
+
"learning_rate": 1.8426626323751891e-06,
|
| 1559 |
+
"loss": 0.5559,
|
| 1560 |
+
"step": 3285
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 2.24642614023145,
|
| 1564 |
+
"grad_norm": 947867.0625,
|
| 1565 |
+
"learning_rate": 1.835854765506808e-06,
|
| 1566 |
+
"loss": 0.4759,
|
| 1567 |
+
"step": 3300
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 2.256637168141593,
|
| 1571 |
+
"grad_norm": 507538.5,
|
| 1572 |
+
"learning_rate": 1.8290468986384266e-06,
|
| 1573 |
+
"loss": 0.4976,
|
| 1574 |
+
"step": 3315
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 2.2668481960517357,
|
| 1578 |
+
"grad_norm": 1198014.5,
|
| 1579 |
+
"learning_rate": 1.8222390317700456e-06,
|
| 1580 |
+
"loss": 0.4435,
|
| 1581 |
+
"step": 3330
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 2.2770592239618788,
|
| 1585 |
+
"grad_norm": 219044.03125,
|
| 1586 |
+
"learning_rate": 1.8154311649016642e-06,
|
| 1587 |
+
"loss": 0.6013,
|
| 1588 |
+
"step": 3345
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 2.287270251872022,
|
| 1592 |
+
"grad_norm": 525360.9375,
|
| 1593 |
+
"learning_rate": 1.808623298033283e-06,
|
| 1594 |
+
"loss": 0.5409,
|
| 1595 |
+
"step": 3360
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 2.297481279782165,
|
| 1599 |
+
"grad_norm": 343133.25,
|
| 1600 |
+
"learning_rate": 1.8018154311649017e-06,
|
| 1601 |
+
"loss": 0.6779,
|
| 1602 |
+
"step": 3375
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 2.3076923076923075,
|
| 1606 |
+
"grad_norm": 994074.375,
|
| 1607 |
+
"learning_rate": 1.7950075642965203e-06,
|
| 1608 |
+
"loss": 0.3746,
|
| 1609 |
+
"step": 3390
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 2.3179033356024505,
|
| 1613 |
+
"grad_norm": 675095.9375,
|
| 1614 |
+
"learning_rate": 1.7881996974281393e-06,
|
| 1615 |
+
"loss": 0.6972,
|
| 1616 |
+
"step": 3405
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 2.3281143635125936,
|
| 1620 |
+
"grad_norm": 1367137.125,
|
| 1621 |
+
"learning_rate": 1.7813918305597581e-06,
|
| 1622 |
+
"loss": 0.462,
|
| 1623 |
+
"step": 3420
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 2.3383253914227367,
|
| 1627 |
+
"grad_norm": 302348.9375,
|
| 1628 |
+
"learning_rate": 1.7745839636913767e-06,
|
| 1629 |
+
"loss": 0.3931,
|
| 1630 |
+
"step": 3435
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 2.3485364193328797,
|
| 1634 |
+
"grad_norm": 727419.4375,
|
| 1635 |
+
"learning_rate": 1.7677760968229954e-06,
|
| 1636 |
+
"loss": 0.687,
|
| 1637 |
+
"step": 3450
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 2.3587474472430223,
|
| 1641 |
+
"grad_norm": 3086808.0,
|
| 1642 |
+
"learning_rate": 1.7609682299546142e-06,
|
| 1643 |
+
"loss": 0.4411,
|
| 1644 |
+
"step": 3465
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 2.3689584751531654,
|
| 1648 |
+
"grad_norm": 930683.1875,
|
| 1649 |
+
"learning_rate": 1.7541603630862332e-06,
|
| 1650 |
+
"loss": 0.669,
|
| 1651 |
+
"step": 3480
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 2.3791695030633084,
|
| 1655 |
+
"grad_norm": 282012.15625,
|
| 1656 |
+
"learning_rate": 1.7473524962178518e-06,
|
| 1657 |
+
"loss": 0.5843,
|
| 1658 |
+
"step": 3495
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 2.3893805309734515,
|
| 1662 |
+
"grad_norm": 106134.8125,
|
| 1663 |
+
"learning_rate": 1.7405446293494707e-06,
|
| 1664 |
+
"loss": 0.5899,
|
| 1665 |
+
"step": 3510
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 2.399591558883594,
|
| 1669 |
+
"grad_norm": 56236.19140625,
|
| 1670 |
+
"learning_rate": 1.7337367624810893e-06,
|
| 1671 |
+
"loss": 0.4536,
|
| 1672 |
+
"step": 3525
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 2.409802586793737,
|
| 1676 |
+
"grad_norm": 344606.1875,
|
| 1677 |
+
"learning_rate": 1.7269288956127079e-06,
|
| 1678 |
+
"loss": 0.443,
|
| 1679 |
+
"step": 3540
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 2.42001361470388,
|
| 1683 |
+
"grad_norm": 771505.9375,
|
| 1684 |
+
"learning_rate": 1.720121028744327e-06,
|
| 1685 |
+
"loss": 0.4357,
|
| 1686 |
+
"step": 3555
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 2.4302246426140233,
|
| 1690 |
+
"grad_norm": 80650.0703125,
|
| 1691 |
+
"learning_rate": 1.7133131618759457e-06,
|
| 1692 |
+
"loss": 0.5167,
|
| 1693 |
+
"step": 3570
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 2.440435670524166,
|
| 1697 |
+
"grad_norm": 2634711.0,
|
| 1698 |
+
"learning_rate": 1.7065052950075644e-06,
|
| 1699 |
+
"loss": 0.3663,
|
| 1700 |
+
"step": 3585
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 2.450646698434309,
|
| 1704 |
+
"grad_norm": 1415573.875,
|
| 1705 |
+
"learning_rate": 1.699697428139183e-06,
|
| 1706 |
+
"loss": 0.4403,
|
| 1707 |
+
"step": 3600
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 2.460857726344452,
|
| 1711 |
+
"grad_norm": 2269082.75,
|
| 1712 |
+
"learning_rate": 1.6928895612708018e-06,
|
| 1713 |
+
"loss": 0.573,
|
| 1714 |
+
"step": 3615
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 2.471068754254595,
|
| 1718 |
+
"grad_norm": 1052448.875,
|
| 1719 |
+
"learning_rate": 1.6860816944024208e-06,
|
| 1720 |
+
"loss": 0.4889,
|
| 1721 |
+
"step": 3630
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 2.481279782164738,
|
| 1725 |
+
"grad_norm": 770280.5,
|
| 1726 |
+
"learning_rate": 1.6792738275340394e-06,
|
| 1727 |
+
"loss": 0.6652,
|
| 1728 |
+
"step": 3645
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 2.4914908100748807,
|
| 1732 |
+
"grad_norm": 1298375.5,
|
| 1733 |
+
"learning_rate": 1.6724659606656583e-06,
|
| 1734 |
+
"loss": 0.4182,
|
| 1735 |
+
"step": 3660
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 2.5017018379850238,
|
| 1739 |
+
"grad_norm": 1415979.75,
|
| 1740 |
+
"learning_rate": 1.6656580937972769e-06,
|
| 1741 |
+
"loss": 0.4441,
|
| 1742 |
+
"step": 3675
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 2.511912865895167,
|
| 1746 |
+
"grad_norm": 663093.6875,
|
| 1747 |
+
"learning_rate": 1.6588502269288955e-06,
|
| 1748 |
+
"loss": 0.4322,
|
| 1749 |
+
"step": 3690
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 2.52212389380531,
|
| 1753 |
+
"grad_norm": 2289919.25,
|
| 1754 |
+
"learning_rate": 1.6520423600605145e-06,
|
| 1755 |
+
"loss": 0.4012,
|
| 1756 |
+
"step": 3705
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 2.532334921715453,
|
| 1760 |
+
"grad_norm": 1178189.375,
|
| 1761 |
+
"learning_rate": 1.6452344931921333e-06,
|
| 1762 |
+
"loss": 0.5652,
|
| 1763 |
+
"step": 3720
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 2.5425459496255955,
|
| 1767 |
+
"grad_norm": 222618.671875,
|
| 1768 |
+
"learning_rate": 1.638426626323752e-06,
|
| 1769 |
+
"loss": 0.3636,
|
| 1770 |
+
"step": 3735
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 2.5527569775357386,
|
| 1774 |
+
"grad_norm": 642872.9375,
|
| 1775 |
+
"learning_rate": 1.6316187594553706e-06,
|
| 1776 |
+
"loss": 0.6289,
|
| 1777 |
+
"step": 3750
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 2.5629680054458817,
|
| 1781 |
+
"grad_norm": 316546.15625,
|
| 1782 |
+
"learning_rate": 1.6248108925869894e-06,
|
| 1783 |
+
"loss": 0.6035,
|
| 1784 |
+
"step": 3765
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 2.5731790333560243,
|
| 1788 |
+
"grad_norm": 63267.38671875,
|
| 1789 |
+
"learning_rate": 1.6180030257186084e-06,
|
| 1790 |
+
"loss": 0.4131,
|
| 1791 |
+
"step": 3780
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 2.5833900612661673,
|
| 1795 |
+
"grad_norm": 3657979.0,
|
| 1796 |
+
"learning_rate": 1.611195158850227e-06,
|
| 1797 |
+
"loss": 0.7483,
|
| 1798 |
+
"step": 3795
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 2.5936010891763104,
|
| 1802 |
+
"grad_norm": 168698.3125,
|
| 1803 |
+
"learning_rate": 1.6043872919818457e-06,
|
| 1804 |
+
"loss": 0.4738,
|
| 1805 |
+
"step": 3810
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 2.6038121170864534,
|
| 1809 |
+
"grad_norm": 680610.5625,
|
| 1810 |
+
"learning_rate": 1.5975794251134645e-06,
|
| 1811 |
+
"loss": 0.4139,
|
| 1812 |
+
"step": 3825
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 2.6140231449965965,
|
| 1816 |
+
"grad_norm": 326585.28125,
|
| 1817 |
+
"learning_rate": 1.590771558245083e-06,
|
| 1818 |
+
"loss": 0.2408,
|
| 1819 |
+
"step": 3840
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"epoch": 2.624234172906739,
|
| 1823 |
+
"grad_norm": 409793.6875,
|
| 1824 |
+
"learning_rate": 1.5839636913767021e-06,
|
| 1825 |
+
"loss": 0.5334,
|
| 1826 |
+
"step": 3855
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 2.634445200816882,
|
| 1830 |
+
"grad_norm": 55384.58984375,
|
| 1831 |
+
"learning_rate": 1.577155824508321e-06,
|
| 1832 |
+
"loss": 0.7308,
|
| 1833 |
+
"step": 3870
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 2.644656228727025,
|
| 1837 |
+
"grad_norm": 2711447.5,
|
| 1838 |
+
"learning_rate": 1.5703479576399396e-06,
|
| 1839 |
+
"loss": 0.675,
|
| 1840 |
+
"step": 3885
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 2.6548672566371683,
|
| 1844 |
+
"grad_norm": 1392973.0,
|
| 1845 |
+
"learning_rate": 1.5635400907715582e-06,
|
| 1846 |
+
"loss": 0.6709,
|
| 1847 |
+
"step": 3900
|
| 1848 |
+
},
|
| 1849 |
+
{
|
| 1850 |
+
"epoch": 2.6650782845473113,
|
| 1851 |
+
"grad_norm": 115502.75,
|
| 1852 |
+
"learning_rate": 1.556732223903177e-06,
|
| 1853 |
+
"loss": 0.4405,
|
| 1854 |
+
"step": 3915
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 2.675289312457454,
|
| 1858 |
+
"grad_norm": 715057.4375,
|
| 1859 |
+
"learning_rate": 1.549924357034796e-06,
|
| 1860 |
+
"loss": 0.3218,
|
| 1861 |
+
"step": 3930
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 2.685500340367597,
|
| 1865 |
+
"grad_norm": 608108.9375,
|
| 1866 |
+
"learning_rate": 1.5431164901664146e-06,
|
| 1867 |
+
"loss": 0.4415,
|
| 1868 |
+
"step": 3945
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 2.69571136827774,
|
| 1872 |
+
"grad_norm": 427073.5,
|
| 1873 |
+
"learning_rate": 1.5363086232980333e-06,
|
| 1874 |
+
"loss": 0.5927,
|
| 1875 |
+
"step": 3960
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 2.7059223961878827,
|
| 1879 |
+
"grad_norm": 296582.28125,
|
| 1880 |
+
"learning_rate": 1.529500756429652e-06,
|
| 1881 |
+
"loss": 0.4821,
|
| 1882 |
+
"step": 3975
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 2.7161334240980257,
|
| 1886 |
+
"grad_norm": 7283655.0,
|
| 1887 |
+
"learning_rate": 1.5226928895612707e-06,
|
| 1888 |
+
"loss": 0.6949,
|
| 1889 |
+
"step": 3990
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 2.7263444520081688,
|
| 1893 |
+
"grad_norm": 4511815.5,
|
| 1894 |
+
"learning_rate": 1.5158850226928897e-06,
|
| 1895 |
+
"loss": 0.4827,
|
| 1896 |
+
"step": 4005
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 2.736555479918312,
|
| 1900 |
+
"grad_norm": 2752351.5,
|
| 1901 |
+
"learning_rate": 1.5090771558245083e-06,
|
| 1902 |
+
"loss": 0.4366,
|
| 1903 |
+
"step": 4020
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 2.746766507828455,
|
| 1907 |
+
"grad_norm": 2160780.75,
|
| 1908 |
+
"learning_rate": 1.5022692889561272e-06,
|
| 1909 |
+
"loss": 0.6442,
|
| 1910 |
+
"step": 4035
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 2.7569775357385975,
|
| 1914 |
+
"grad_norm": 842177.375,
|
| 1915 |
+
"learning_rate": 1.4954614220877458e-06,
|
| 1916 |
+
"loss": 0.3898,
|
| 1917 |
+
"step": 4050
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 2.7671885636487406,
|
| 1921 |
+
"grad_norm": 488163.53125,
|
| 1922 |
+
"learning_rate": 1.4886535552193646e-06,
|
| 1923 |
+
"loss": 0.3194,
|
| 1924 |
+
"step": 4065
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 2.7773995915588836,
|
| 1928 |
+
"grad_norm": 1358681.25,
|
| 1929 |
+
"learning_rate": 1.4818456883509834e-06,
|
| 1930 |
+
"loss": 0.8422,
|
| 1931 |
+
"step": 4080
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 2.7876106194690267,
|
| 1935 |
+
"grad_norm": 2055414.25,
|
| 1936 |
+
"learning_rate": 1.475037821482602e-06,
|
| 1937 |
+
"loss": 0.7359,
|
| 1938 |
+
"step": 4095
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 2.7978216473791697,
|
| 1942 |
+
"grad_norm": 1410838.125,
|
| 1943 |
+
"learning_rate": 1.4682299546142209e-06,
|
| 1944 |
+
"loss": 0.5694,
|
| 1945 |
+
"step": 4110
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 2.8080326752893123,
|
| 1949 |
+
"grad_norm": 485180.3125,
|
| 1950 |
+
"learning_rate": 1.4614220877458397e-06,
|
| 1951 |
+
"loss": 0.3498,
|
| 1952 |
+
"step": 4125
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 2.8182437031994554,
|
| 1956 |
+
"grad_norm": 442321.53125,
|
| 1957 |
+
"learning_rate": 1.4546142208774585e-06,
|
| 1958 |
+
"loss": 0.3907,
|
| 1959 |
+
"step": 4140
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 2.8284547311095984,
|
| 1963 |
+
"grad_norm": 117394.3359375,
|
| 1964 |
+
"learning_rate": 1.4478063540090771e-06,
|
| 1965 |
+
"loss": 0.4456,
|
| 1966 |
+
"step": 4155
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 2.838665759019741,
|
| 1970 |
+
"grad_norm": 1168203.125,
|
| 1971 |
+
"learning_rate": 1.440998487140696e-06,
|
| 1972 |
+
"loss": 0.4467,
|
| 1973 |
+
"step": 4170
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 2.8488767869298846,
|
| 1977 |
+
"grad_norm": 1550697.0,
|
| 1978 |
+
"learning_rate": 1.4341906202723148e-06,
|
| 1979 |
+
"loss": 0.3427,
|
| 1980 |
+
"step": 4185
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 2.859087814840027,
|
| 1984 |
+
"grad_norm": 865472.9375,
|
| 1985 |
+
"learning_rate": 1.4273827534039334e-06,
|
| 1986 |
+
"loss": 0.5203,
|
| 1987 |
+
"step": 4200
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 2.86929884275017,
|
| 1991 |
+
"grad_norm": 548500.1875,
|
| 1992 |
+
"learning_rate": 1.4205748865355522e-06,
|
| 1993 |
+
"loss": 0.6109,
|
| 1994 |
+
"step": 4215
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 2.8795098706603133,
|
| 1998 |
+
"grad_norm": 2530923.75,
|
| 1999 |
+
"learning_rate": 1.413767019667171e-06,
|
| 2000 |
+
"loss": 0.6251,
|
| 2001 |
+
"step": 4230
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 2.889720898570456,
|
| 2005 |
+
"grad_norm": 1599674.125,
|
| 2006 |
+
"learning_rate": 1.4069591527987896e-06,
|
| 2007 |
+
"loss": 0.3419,
|
| 2008 |
+
"step": 4245
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 2.899931926480599,
|
| 2012 |
+
"grad_norm": 1614385.875,
|
| 2013 |
+
"learning_rate": 1.4001512859304085e-06,
|
| 2014 |
+
"loss": 0.7362,
|
| 2015 |
+
"step": 4260
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 2.910142954390742,
|
| 2019 |
+
"grad_norm": 2576167.5,
|
| 2020 |
+
"learning_rate": 1.3933434190620273e-06,
|
| 2021 |
+
"loss": 0.5736,
|
| 2022 |
+
"step": 4275
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 2.920353982300885,
|
| 2026 |
+
"grad_norm": 3079889.75,
|
| 2027 |
+
"learning_rate": 1.3865355521936461e-06,
|
| 2028 |
+
"loss": 0.4168,
|
| 2029 |
+
"step": 4290
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 2.930565010211028,
|
| 2033 |
+
"grad_norm": 483019.09375,
|
| 2034 |
+
"learning_rate": 1.3797276853252647e-06,
|
| 2035 |
+
"loss": 0.5769,
|
| 2036 |
+
"step": 4305
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 2.9407760381211707,
|
| 2040 |
+
"grad_norm": 953947.1875,
|
| 2041 |
+
"learning_rate": 1.3729198184568836e-06,
|
| 2042 |
+
"loss": 0.5267,
|
| 2043 |
+
"step": 4320
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 2.9509870660313138,
|
| 2047 |
+
"grad_norm": 951762.1875,
|
| 2048 |
+
"learning_rate": 1.3661119515885024e-06,
|
| 2049 |
+
"loss": 0.465,
|
| 2050 |
+
"step": 4335
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 2.961198093941457,
|
| 2054 |
+
"grad_norm": 925738.1875,
|
| 2055 |
+
"learning_rate": 1.359304084720121e-06,
|
| 2056 |
+
"loss": 0.4431,
|
| 2057 |
+
"step": 4350
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 2.9714091218516,
|
| 2061 |
+
"grad_norm": 2421353.0,
|
| 2062 |
+
"learning_rate": 1.3524962178517398e-06,
|
| 2063 |
+
"loss": 0.4951,
|
| 2064 |
+
"step": 4365
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 2.981620149761743,
|
| 2068 |
+
"grad_norm": 2792055.5,
|
| 2069 |
+
"learning_rate": 1.3456883509833586e-06,
|
| 2070 |
+
"loss": 0.4273,
|
| 2071 |
+
"step": 4380
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"epoch": 2.9918311776718856,
|
| 2075 |
+
"grad_norm": 78971.71875,
|
| 2076 |
+
"learning_rate": 1.3388804841149772e-06,
|
| 2077 |
+
"loss": 0.452,
|
| 2078 |
+
"step": 4395
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 3.0,
|
| 2082 |
+
"eval_accuracy": 0.9131455399061033,
|
| 2083 |
+
"eval_loss": 0.36749738454818726,
|
| 2084 |
+
"eval_runtime": 574.4362,
|
| 2085 |
+
"eval_samples_per_second": 5.191,
|
| 2086 |
+
"eval_steps_per_second": 0.649,
|
| 2087 |
+
"step": 4407
|
| 2088 |
+
}
|
| 2089 |
+
],
|
| 2090 |
+
"logging_steps": 15,
|
| 2091 |
+
"max_steps": 7345,
|
| 2092 |
+
"num_input_tokens_seen": 0,
|
| 2093 |
+
"num_train_epochs": 5,
|
| 2094 |
+
"save_steps": 500,
|
| 2095 |
+
"stateful_callbacks": {
|
| 2096 |
+
"TrainerControl": {
|
| 2097 |
+
"args": {
|
| 2098 |
+
"should_epoch_stop": false,
|
| 2099 |
+
"should_evaluate": false,
|
| 2100 |
+
"should_log": false,
|
| 2101 |
+
"should_save": true,
|
| 2102 |
+
"should_training_stop": false
|
| 2103 |
+
},
|
| 2104 |
+
"attributes": {}
|
| 2105 |
+
}
|
| 2106 |
+
},
|
| 2107 |
+
"total_flos": 1.560567924145152e+19,
|
| 2108 |
+
"train_batch_size": 8,
|
| 2109 |
+
"trial_name": null,
|
| 2110 |
+
"trial_params": null
|
| 2111 |
+
}
|
checkpoint-4407/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2de8a88815cd7a4ce539dab9d23a753db8fa0d9213c6e02f1f86779a8bf8950f
|
| 3 |
+
size 5176
|