Training in progress, epoch 2, checkpoint
Browse files- checkpoint-784/config.json +71 -0
- checkpoint-784/model.safetensors +3 -0
- checkpoint-784/optimizer.pt +3 -0
- checkpoint-784/preprocessor_config.json +14 -0
- checkpoint-784/rng_state.pth +3 -0
- checkpoint-784/scheduler.pt +3 -0
- checkpoint-784/trainer_state.json +1143 -0
- checkpoint-784/training_args.bin +3 -0
checkpoint-784/config.json
ADDED
|
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "vinai/PhoWhisper-small",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"activation_function": "gelu",
|
| 5 |
+
"apply_spec_augment": false,
|
| 6 |
+
"architectures": [
|
| 7 |
+
"ViSpeechClassification"
|
| 8 |
+
],
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"begin_suppress_tokens": [
|
| 11 |
+
220,
|
| 12 |
+
50257
|
| 13 |
+
],
|
| 14 |
+
"bos_token_id": 50257,
|
| 15 |
+
"classifier_proj_size": 256,
|
| 16 |
+
"d_model": 768,
|
| 17 |
+
"decoder_attention_heads": 12,
|
| 18 |
+
"decoder_ffn_dim": 3072,
|
| 19 |
+
"decoder_layerdrop": 0.0,
|
| 20 |
+
"decoder_layers": 12,
|
| 21 |
+
"decoder_start_token_id": 50258,
|
| 22 |
+
"dropout": 0.0,
|
| 23 |
+
"encoder_attention_heads": 12,
|
| 24 |
+
"encoder_ffn_dim": 3072,
|
| 25 |
+
"encoder_layerdrop": 0.0,
|
| 26 |
+
"encoder_layers": 12,
|
| 27 |
+
"eos_token_id": 50257,
|
| 28 |
+
"forced_decoder_ids": null,
|
| 29 |
+
"id2label": {
|
| 30 |
+
"0": "male, northern dialect",
|
| 31 |
+
"1": "male, central dialect",
|
| 32 |
+
"2": "male, highland central dialect",
|
| 33 |
+
"3": "male, southern dialect",
|
| 34 |
+
"4": "female, northern dialect",
|
| 35 |
+
"5": "female, central dialect",
|
| 36 |
+
"6": "female, highland central dialect",
|
| 37 |
+
"7": "female, southern dialect"
|
| 38 |
+
},
|
| 39 |
+
"init_std": 0.02,
|
| 40 |
+
"is_encoder_decoder": true,
|
| 41 |
+
"label2id": {
|
| 42 |
+
"female, central dialect": "5",
|
| 43 |
+
"female, highland central dialect": "6",
|
| 44 |
+
"female, northern dialect": "4",
|
| 45 |
+
"female, southern dialect": "7",
|
| 46 |
+
"male, central dialect": "1",
|
| 47 |
+
"male, highland central dialect": "2",
|
| 48 |
+
"male, northern dialect": "0",
|
| 49 |
+
"male, southern dialect": "3"
|
| 50 |
+
},
|
| 51 |
+
"mask_feature_length": 10,
|
| 52 |
+
"mask_feature_min_masks": 0,
|
| 53 |
+
"mask_feature_prob": 0.0,
|
| 54 |
+
"mask_time_length": 10,
|
| 55 |
+
"mask_time_min_masks": 2,
|
| 56 |
+
"mask_time_prob": 0.05,
|
| 57 |
+
"max_length": 448,
|
| 58 |
+
"max_source_positions": 1500,
|
| 59 |
+
"max_target_positions": 448,
|
| 60 |
+
"median_filter_width": 7,
|
| 61 |
+
"model_type": "whisper",
|
| 62 |
+
"num_hidden_layers": 12,
|
| 63 |
+
"num_mel_bins": 80,
|
| 64 |
+
"pad_token_id": 50257,
|
| 65 |
+
"scale_embedding": false,
|
| 66 |
+
"torch_dtype": "float32",
|
| 67 |
+
"transformers_version": "4.41.2",
|
| 68 |
+
"use_cache": true,
|
| 69 |
+
"use_weighted_layer_sum": false,
|
| 70 |
+
"vocab_size": 51865
|
| 71 |
+
}
|
checkpoint-784/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4edd225b274a6c2af12cb2efa8484ec032516b2ea46e228643d7a38d10c82f29
|
| 3 |
+
size 358419776
|
checkpoint-784/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6c652fea03961dc61f35c38a98c92ad5bb2005c203ca2b8222d2e73b73ee1590
|
| 3 |
+
size 11572486
|
checkpoint-784/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-784/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4539c77d72037760cde3627da9d3376a3b4c8b83a91e130173d97182866af3ea
|
| 3 |
+
size 14244
|
checkpoint-784/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d615636869468fc82b34bd5076a1cb8499beccc50d1b2760d59b5a7eb404b53a
|
| 3 |
+
size 1064
|
checkpoint-784/trainer_state.json
ADDED
|
@@ -0,0 +1,1143 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.0857774206508638,
|
| 3 |
+
"best_model_checkpoint": "PhoWhisper-small-vispeech-classifier-v3/checkpoint-392",
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 784,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.012755102040816327,
|
| 13 |
+
"grad_norm": 48589.94921875,
|
| 14 |
+
"learning_rate": 6.377551020408164e-08,
|
| 15 |
+
"loss": 2.0884,
|
| 16 |
+
"step": 5
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.025510204081632654,
|
| 20 |
+
"grad_norm": 50532.125,
|
| 21 |
+
"learning_rate": 1.2755102040816328e-07,
|
| 22 |
+
"loss": 2.0887,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.03826530612244898,
|
| 27 |
+
"grad_norm": 50246.4296875,
|
| 28 |
+
"learning_rate": 1.913265306122449e-07,
|
| 29 |
+
"loss": 2.0885,
|
| 30 |
+
"step": 15
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.05102040816326531,
|
| 34 |
+
"grad_norm": 57370.44140625,
|
| 35 |
+
"learning_rate": 2.5510204081632656e-07,
|
| 36 |
+
"loss": 2.0908,
|
| 37 |
+
"step": 20
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.06377551020408163,
|
| 41 |
+
"grad_norm": 40133.53125,
|
| 42 |
+
"learning_rate": 3.188775510204082e-07,
|
| 43 |
+
"loss": 2.0889,
|
| 44 |
+
"step": 25
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.07653061224489796,
|
| 48 |
+
"grad_norm": 58377.3671875,
|
| 49 |
+
"learning_rate": 3.826530612244898e-07,
|
| 50 |
+
"loss": 2.0874,
|
| 51 |
+
"step": 30
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.08928571428571429,
|
| 55 |
+
"grad_norm": 50888.7421875,
|
| 56 |
+
"learning_rate": 4.4642857142857147e-07,
|
| 57 |
+
"loss": 2.088,
|
| 58 |
+
"step": 35
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.10204081632653061,
|
| 62 |
+
"grad_norm": 44107.75390625,
|
| 63 |
+
"learning_rate": 5.102040816326531e-07,
|
| 64 |
+
"loss": 2.0872,
|
| 65 |
+
"step": 40
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.11479591836734694,
|
| 69 |
+
"grad_norm": 62939.18359375,
|
| 70 |
+
"learning_rate": 5.739795918367347e-07,
|
| 71 |
+
"loss": 2.0873,
|
| 72 |
+
"step": 45
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.12755102040816327,
|
| 76 |
+
"grad_norm": 42077.53515625,
|
| 77 |
+
"learning_rate": 6.377551020408164e-07,
|
| 78 |
+
"loss": 2.0852,
|
| 79 |
+
"step": 50
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.14030612244897958,
|
| 83 |
+
"grad_norm": 50233.00390625,
|
| 84 |
+
"learning_rate": 7.015306122448979e-07,
|
| 85 |
+
"loss": 2.0861,
|
| 86 |
+
"step": 55
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.15306122448979592,
|
| 90 |
+
"grad_norm": 60839.4765625,
|
| 91 |
+
"learning_rate": 7.653061224489796e-07,
|
| 92 |
+
"loss": 2.0856,
|
| 93 |
+
"step": 60
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.16581632653061223,
|
| 97 |
+
"grad_norm": 57499.3515625,
|
| 98 |
+
"learning_rate": 8.290816326530612e-07,
|
| 99 |
+
"loss": 2.0838,
|
| 100 |
+
"step": 65
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.17857142857142858,
|
| 104 |
+
"grad_norm": 65030.765625,
|
| 105 |
+
"learning_rate": 8.928571428571429e-07,
|
| 106 |
+
"loss": 2.083,
|
| 107 |
+
"step": 70
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.1913265306122449,
|
| 111 |
+
"grad_norm": 53192.35546875,
|
| 112 |
+
"learning_rate": 9.566326530612244e-07,
|
| 113 |
+
"loss": 2.0841,
|
| 114 |
+
"step": 75
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.20408163265306123,
|
| 118 |
+
"grad_norm": 52991.80078125,
|
| 119 |
+
"learning_rate": 1.0204081632653063e-06,
|
| 120 |
+
"loss": 2.0817,
|
| 121 |
+
"step": 80
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.21683673469387754,
|
| 125 |
+
"grad_norm": 48656.234375,
|
| 126 |
+
"learning_rate": 1.0841836734693879e-06,
|
| 127 |
+
"loss": 2.0813,
|
| 128 |
+
"step": 85
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.22959183673469388,
|
| 132 |
+
"grad_norm": 63459.2421875,
|
| 133 |
+
"learning_rate": 1.1479591836734695e-06,
|
| 134 |
+
"loss": 2.081,
|
| 135 |
+
"step": 90
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.2423469387755102,
|
| 139 |
+
"grad_norm": 47800.3515625,
|
| 140 |
+
"learning_rate": 1.211734693877551e-06,
|
| 141 |
+
"loss": 2.0775,
|
| 142 |
+
"step": 95
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.25510204081632654,
|
| 146 |
+
"grad_norm": 64353.37890625,
|
| 147 |
+
"learning_rate": 1.2755102040816329e-06,
|
| 148 |
+
"loss": 2.0772,
|
| 149 |
+
"step": 100
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.26785714285714285,
|
| 153 |
+
"grad_norm": 59652.29296875,
|
| 154 |
+
"learning_rate": 1.3392857142857143e-06,
|
| 155 |
+
"loss": 2.0779,
|
| 156 |
+
"step": 105
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.28061224489795916,
|
| 160 |
+
"grad_norm": 44752.69140625,
|
| 161 |
+
"learning_rate": 1.4030612244897959e-06,
|
| 162 |
+
"loss": 2.0764,
|
| 163 |
+
"step": 110
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.29336734693877553,
|
| 167 |
+
"grad_norm": 49883.0,
|
| 168 |
+
"learning_rate": 1.4668367346938777e-06,
|
| 169 |
+
"loss": 2.0763,
|
| 170 |
+
"step": 115
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.30612244897959184,
|
| 174 |
+
"grad_norm": 65950.28125,
|
| 175 |
+
"learning_rate": 1.5306122448979593e-06,
|
| 176 |
+
"loss": 2.0739,
|
| 177 |
+
"step": 120
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.31887755102040816,
|
| 181 |
+
"grad_norm": 61354.83203125,
|
| 182 |
+
"learning_rate": 1.5943877551020409e-06,
|
| 183 |
+
"loss": 2.0706,
|
| 184 |
+
"step": 125
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.33163265306122447,
|
| 188 |
+
"grad_norm": 53169.91796875,
|
| 189 |
+
"learning_rate": 1.6581632653061225e-06,
|
| 190 |
+
"loss": 2.0709,
|
| 191 |
+
"step": 130
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.34438775510204084,
|
| 195 |
+
"grad_norm": 55883.31640625,
|
| 196 |
+
"learning_rate": 1.7219387755102043e-06,
|
| 197 |
+
"loss": 2.0695,
|
| 198 |
+
"step": 135
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.35714285714285715,
|
| 202 |
+
"grad_norm": 38134.765625,
|
| 203 |
+
"learning_rate": 1.7857142857142859e-06,
|
| 204 |
+
"loss": 2.0675,
|
| 205 |
+
"step": 140
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.36989795918367346,
|
| 209 |
+
"grad_norm": 53920.7265625,
|
| 210 |
+
"learning_rate": 1.8494897959183675e-06,
|
| 211 |
+
"loss": 2.0659,
|
| 212 |
+
"step": 145
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.3826530612244898,
|
| 216 |
+
"grad_norm": 68777.28125,
|
| 217 |
+
"learning_rate": 1.913265306122449e-06,
|
| 218 |
+
"loss": 2.0638,
|
| 219 |
+
"step": 150
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.39540816326530615,
|
| 223 |
+
"grad_norm": 46247.8984375,
|
| 224 |
+
"learning_rate": 1.977040816326531e-06,
|
| 225 |
+
"loss": 2.0656,
|
| 226 |
+
"step": 155
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.40816326530612246,
|
| 230 |
+
"grad_norm": 49636.04296875,
|
| 231 |
+
"learning_rate": 2.0408163265306125e-06,
|
| 232 |
+
"loss": 2.0621,
|
| 233 |
+
"step": 160
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.42091836734693877,
|
| 237 |
+
"grad_norm": 51880.3515625,
|
| 238 |
+
"learning_rate": 2.104591836734694e-06,
|
| 239 |
+
"loss": 2.057,
|
| 240 |
+
"step": 165
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.4336734693877551,
|
| 244 |
+
"grad_norm": 52808.734375,
|
| 245 |
+
"learning_rate": 2.1683673469387757e-06,
|
| 246 |
+
"loss": 2.0585,
|
| 247 |
+
"step": 170
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.44642857142857145,
|
| 251 |
+
"grad_norm": 46398.125,
|
| 252 |
+
"learning_rate": 2.2321428571428573e-06,
|
| 253 |
+
"loss": 2.0505,
|
| 254 |
+
"step": 175
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.45918367346938777,
|
| 258 |
+
"grad_norm": 59347.2578125,
|
| 259 |
+
"learning_rate": 2.295918367346939e-06,
|
| 260 |
+
"loss": 2.0487,
|
| 261 |
+
"step": 180
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.4719387755102041,
|
| 265 |
+
"grad_norm": 58343.73828125,
|
| 266 |
+
"learning_rate": 2.3596938775510205e-06,
|
| 267 |
+
"loss": 2.0563,
|
| 268 |
+
"step": 185
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.4846938775510204,
|
| 272 |
+
"grad_norm": 43836.42578125,
|
| 273 |
+
"learning_rate": 2.423469387755102e-06,
|
| 274 |
+
"loss": 2.0486,
|
| 275 |
+
"step": 190
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.49744897959183676,
|
| 279 |
+
"grad_norm": 69661.5859375,
|
| 280 |
+
"learning_rate": 2.487244897959184e-06,
|
| 281 |
+
"loss": 2.0443,
|
| 282 |
+
"step": 195
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.5102040816326531,
|
| 286 |
+
"grad_norm": 53888.24609375,
|
| 287 |
+
"learning_rate": 2.5510204081632657e-06,
|
| 288 |
+
"loss": 2.0475,
|
| 289 |
+
"step": 200
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.5229591836734694,
|
| 293 |
+
"grad_norm": 50369.1640625,
|
| 294 |
+
"learning_rate": 2.6147959183673473e-06,
|
| 295 |
+
"loss": 2.035,
|
| 296 |
+
"step": 205
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.5357142857142857,
|
| 300 |
+
"grad_norm": 62733.9609375,
|
| 301 |
+
"learning_rate": 2.6785714285714285e-06,
|
| 302 |
+
"loss": 2.0354,
|
| 303 |
+
"step": 210
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.548469387755102,
|
| 307 |
+
"grad_norm": 48340.67578125,
|
| 308 |
+
"learning_rate": 2.74234693877551e-06,
|
| 309 |
+
"loss": 2.0373,
|
| 310 |
+
"step": 215
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.5612244897959183,
|
| 314 |
+
"grad_norm": 50353.734375,
|
| 315 |
+
"learning_rate": 2.8061224489795917e-06,
|
| 316 |
+
"loss": 2.0389,
|
| 317 |
+
"step": 220
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.5739795918367347,
|
| 321 |
+
"grad_norm": 62351.765625,
|
| 322 |
+
"learning_rate": 2.869897959183674e-06,
|
| 323 |
+
"loss": 2.0263,
|
| 324 |
+
"step": 225
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.5867346938775511,
|
| 328 |
+
"grad_norm": 73855.609375,
|
| 329 |
+
"learning_rate": 2.9336734693877553e-06,
|
| 330 |
+
"loss": 2.0269,
|
| 331 |
+
"step": 230
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.5994897959183674,
|
| 335 |
+
"grad_norm": 84417.859375,
|
| 336 |
+
"learning_rate": 2.997448979591837e-06,
|
| 337 |
+
"loss": 2.0311,
|
| 338 |
+
"step": 235
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.6122448979591837,
|
| 342 |
+
"grad_norm": 57347.55078125,
|
| 343 |
+
"learning_rate": 3.0612244897959185e-06,
|
| 344 |
+
"loss": 2.0239,
|
| 345 |
+
"step": 240
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.625,
|
| 349 |
+
"grad_norm": 65611.1875,
|
| 350 |
+
"learning_rate": 3.125e-06,
|
| 351 |
+
"loss": 2.0064,
|
| 352 |
+
"step": 245
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.6377551020408163,
|
| 356 |
+
"grad_norm": 70810.75,
|
| 357 |
+
"learning_rate": 3.1887755102040818e-06,
|
| 358 |
+
"loss": 2.0224,
|
| 359 |
+
"step": 250
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.6505102040816326,
|
| 363 |
+
"grad_norm": 46479.8828125,
|
| 364 |
+
"learning_rate": 3.2525510204081634e-06,
|
| 365 |
+
"loss": 2.0057,
|
| 366 |
+
"step": 255
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.6632653061224489,
|
| 370 |
+
"grad_norm": 34124.89453125,
|
| 371 |
+
"learning_rate": 3.316326530612245e-06,
|
| 372 |
+
"loss": 2.0135,
|
| 373 |
+
"step": 260
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.6760204081632653,
|
| 377 |
+
"grad_norm": 53196.58203125,
|
| 378 |
+
"learning_rate": 3.3801020408163266e-06,
|
| 379 |
+
"loss": 2.0087,
|
| 380 |
+
"step": 265
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.6887755102040817,
|
| 384 |
+
"grad_norm": 64588.515625,
|
| 385 |
+
"learning_rate": 3.4438775510204086e-06,
|
| 386 |
+
"loss": 2.0006,
|
| 387 |
+
"step": 270
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.701530612244898,
|
| 391 |
+
"grad_norm": 73780.546875,
|
| 392 |
+
"learning_rate": 3.50765306122449e-06,
|
| 393 |
+
"loss": 1.9965,
|
| 394 |
+
"step": 275
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.7142857142857143,
|
| 398 |
+
"grad_norm": 63101.05859375,
|
| 399 |
+
"learning_rate": 3.5714285714285718e-06,
|
| 400 |
+
"loss": 2.0005,
|
| 401 |
+
"step": 280
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.7270408163265306,
|
| 405 |
+
"grad_norm": 78616.109375,
|
| 406 |
+
"learning_rate": 3.6352040816326534e-06,
|
| 407 |
+
"loss": 1.9904,
|
| 408 |
+
"step": 285
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.7397959183673469,
|
| 412 |
+
"grad_norm": 79120.6484375,
|
| 413 |
+
"learning_rate": 3.698979591836735e-06,
|
| 414 |
+
"loss": 1.9891,
|
| 415 |
+
"step": 290
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.7525510204081632,
|
| 419 |
+
"grad_norm": 92711.09375,
|
| 420 |
+
"learning_rate": 3.7627551020408166e-06,
|
| 421 |
+
"loss": 1.9611,
|
| 422 |
+
"step": 295
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.7653061224489796,
|
| 426 |
+
"grad_norm": 77935.4765625,
|
| 427 |
+
"learning_rate": 3.826530612244898e-06,
|
| 428 |
+
"loss": 1.9746,
|
| 429 |
+
"step": 300
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.7780612244897959,
|
| 433 |
+
"grad_norm": 85825.375,
|
| 434 |
+
"learning_rate": 3.89030612244898e-06,
|
| 435 |
+
"loss": 1.9753,
|
| 436 |
+
"step": 305
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.7908163265306123,
|
| 440 |
+
"grad_norm": 65178.671875,
|
| 441 |
+
"learning_rate": 3.954081632653062e-06,
|
| 442 |
+
"loss": 1.9534,
|
| 443 |
+
"step": 310
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.8035714285714286,
|
| 447 |
+
"grad_norm": 71973.671875,
|
| 448 |
+
"learning_rate": 4.017857142857143e-06,
|
| 449 |
+
"loss": 1.9474,
|
| 450 |
+
"step": 315
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.8163265306122449,
|
| 454 |
+
"grad_norm": 88983.125,
|
| 455 |
+
"learning_rate": 4.081632653061225e-06,
|
| 456 |
+
"loss": 1.9492,
|
| 457 |
+
"step": 320
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.8290816326530612,
|
| 461 |
+
"grad_norm": 76345.84375,
|
| 462 |
+
"learning_rate": 4.145408163265306e-06,
|
| 463 |
+
"loss": 1.9513,
|
| 464 |
+
"step": 325
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.8418367346938775,
|
| 468 |
+
"grad_norm": 64801.640625,
|
| 469 |
+
"learning_rate": 4.209183673469388e-06,
|
| 470 |
+
"loss": 1.9583,
|
| 471 |
+
"step": 330
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.8545918367346939,
|
| 475 |
+
"grad_norm": 72903.8203125,
|
| 476 |
+
"learning_rate": 4.272959183673469e-06,
|
| 477 |
+
"loss": 1.9579,
|
| 478 |
+
"step": 335
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.8673469387755102,
|
| 482 |
+
"grad_norm": 65284.796875,
|
| 483 |
+
"learning_rate": 4.336734693877551e-06,
|
| 484 |
+
"loss": 1.9272,
|
| 485 |
+
"step": 340
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.8801020408163265,
|
| 489 |
+
"grad_norm": 65237.38671875,
|
| 490 |
+
"learning_rate": 4.400510204081633e-06,
|
| 491 |
+
"loss": 1.9065,
|
| 492 |
+
"step": 345
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.8928571428571429,
|
| 496 |
+
"grad_norm": 79606.9296875,
|
| 497 |
+
"learning_rate": 4.464285714285715e-06,
|
| 498 |
+
"loss": 1.9013,
|
| 499 |
+
"step": 350
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.9056122448979592,
|
| 503 |
+
"grad_norm": 76137.6328125,
|
| 504 |
+
"learning_rate": 4.528061224489797e-06,
|
| 505 |
+
"loss": 1.8892,
|
| 506 |
+
"step": 355
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.9183673469387755,
|
| 510 |
+
"grad_norm": 40717.60546875,
|
| 511 |
+
"learning_rate": 4.591836734693878e-06,
|
| 512 |
+
"loss": 1.9072,
|
| 513 |
+
"step": 360
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.9311224489795918,
|
| 517 |
+
"grad_norm": 110416.125,
|
| 518 |
+
"learning_rate": 4.65561224489796e-06,
|
| 519 |
+
"loss": 1.9024,
|
| 520 |
+
"step": 365
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.9438775510204082,
|
| 524 |
+
"grad_norm": 99101.6875,
|
| 525 |
+
"learning_rate": 4.719387755102041e-06,
|
| 526 |
+
"loss": 1.8979,
|
| 527 |
+
"step": 370
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.9566326530612245,
|
| 531 |
+
"grad_norm": 86707.7109375,
|
| 532 |
+
"learning_rate": 4.783163265306123e-06,
|
| 533 |
+
"loss": 1.878,
|
| 534 |
+
"step": 375
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.9693877551020408,
|
| 538 |
+
"grad_norm": 85074.75,
|
| 539 |
+
"learning_rate": 4.846938775510204e-06,
|
| 540 |
+
"loss": 1.871,
|
| 541 |
+
"step": 380
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.9821428571428571,
|
| 545 |
+
"grad_norm": 47668.7109375,
|
| 546 |
+
"learning_rate": 4.910714285714286e-06,
|
| 547 |
+
"loss": 1.8762,
|
| 548 |
+
"step": 385
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.9948979591836735,
|
| 552 |
+
"grad_norm": 133846.6875,
|
| 553 |
+
"learning_rate": 4.974489795918368e-06,
|
| 554 |
+
"loss": 1.8344,
|
| 555 |
+
"step": 390
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.0,
|
| 559 |
+
"eval_accuracy": 0.0857774206508638,
|
| 560 |
+
"eval_loss": 1.9574071168899536,
|
| 561 |
+
"eval_runtime": 313.0247,
|
| 562 |
+
"eval_samples_per_second": 15.903,
|
| 563 |
+
"eval_steps_per_second": 0.53,
|
| 564 |
+
"step": 392
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.0076530612244898,
|
| 568 |
+
"grad_norm": 95410.25,
|
| 569 |
+
"learning_rate": 5.0382653061224495e-06,
|
| 570 |
+
"loss": 1.8635,
|
| 571 |
+
"step": 395
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.0204081632653061,
|
| 575 |
+
"grad_norm": 86022.1484375,
|
| 576 |
+
"learning_rate": 5.1020408163265315e-06,
|
| 577 |
+
"loss": 1.8147,
|
| 578 |
+
"step": 400
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.0331632653061225,
|
| 582 |
+
"grad_norm": 154860.171875,
|
| 583 |
+
"learning_rate": 5.165816326530613e-06,
|
| 584 |
+
"loss": 1.8016,
|
| 585 |
+
"step": 405
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.0459183673469388,
|
| 589 |
+
"grad_norm": 105390.4609375,
|
| 590 |
+
"learning_rate": 5.229591836734695e-06,
|
| 591 |
+
"loss": 1.8458,
|
| 592 |
+
"step": 410
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.058673469387755,
|
| 596 |
+
"grad_norm": 94511.6171875,
|
| 597 |
+
"learning_rate": 5.293367346938776e-06,
|
| 598 |
+
"loss": 1.7888,
|
| 599 |
+
"step": 415
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.0714285714285714,
|
| 603 |
+
"grad_norm": 93031.234375,
|
| 604 |
+
"learning_rate": 5.357142857142857e-06,
|
| 605 |
+
"loss": 1.7809,
|
| 606 |
+
"step": 420
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.0841836734693877,
|
| 610 |
+
"grad_norm": 71911.4296875,
|
| 611 |
+
"learning_rate": 5.420918367346939e-06,
|
| 612 |
+
"loss": 1.8523,
|
| 613 |
+
"step": 425
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.096938775510204,
|
| 617 |
+
"grad_norm": 59985.96875,
|
| 618 |
+
"learning_rate": 5.48469387755102e-06,
|
| 619 |
+
"loss": 1.7427,
|
| 620 |
+
"step": 430
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.1096938775510203,
|
| 624 |
+
"grad_norm": 86531.875,
|
| 625 |
+
"learning_rate": 5.548469387755102e-06,
|
| 626 |
+
"loss": 1.7646,
|
| 627 |
+
"step": 435
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.1224489795918366,
|
| 631 |
+
"grad_norm": 90977.859375,
|
| 632 |
+
"learning_rate": 5.6122448979591834e-06,
|
| 633 |
+
"loss": 1.7529,
|
| 634 |
+
"step": 440
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.135204081632653,
|
| 638 |
+
"grad_norm": 70691.375,
|
| 639 |
+
"learning_rate": 5.6760204081632655e-06,
|
| 640 |
+
"loss": 1.7754,
|
| 641 |
+
"step": 445
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.1479591836734695,
|
| 645 |
+
"grad_norm": 137077.8125,
|
| 646 |
+
"learning_rate": 5.739795918367348e-06,
|
| 647 |
+
"loss": 1.73,
|
| 648 |
+
"step": 450
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.1607142857142858,
|
| 652 |
+
"grad_norm": 133976.8125,
|
| 653 |
+
"learning_rate": 5.8035714285714295e-06,
|
| 654 |
+
"loss": 1.7906,
|
| 655 |
+
"step": 455
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.1734693877551021,
|
| 659 |
+
"grad_norm": 84754.7421875,
|
| 660 |
+
"learning_rate": 5.867346938775511e-06,
|
| 661 |
+
"loss": 1.7722,
|
| 662 |
+
"step": 460
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.1862244897959184,
|
| 666 |
+
"grad_norm": 67924.859375,
|
| 667 |
+
"learning_rate": 5.931122448979593e-06,
|
| 668 |
+
"loss": 1.8274,
|
| 669 |
+
"step": 465
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 1.1989795918367347,
|
| 673 |
+
"grad_norm": 110309.015625,
|
| 674 |
+
"learning_rate": 5.994897959183674e-06,
|
| 675 |
+
"loss": 1.7558,
|
| 676 |
+
"step": 470
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.211734693877551,
|
| 680 |
+
"grad_norm": 108489.703125,
|
| 681 |
+
"learning_rate": 6.058673469387756e-06,
|
| 682 |
+
"loss": 1.7387,
|
| 683 |
+
"step": 475
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.2244897959183674,
|
| 687 |
+
"grad_norm": 99533.59375,
|
| 688 |
+
"learning_rate": 6.122448979591837e-06,
|
| 689 |
+
"loss": 1.7367,
|
| 690 |
+
"step": 480
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.2372448979591837,
|
| 694 |
+
"grad_norm": 85590.296875,
|
| 695 |
+
"learning_rate": 6.186224489795919e-06,
|
| 696 |
+
"loss": 1.7304,
|
| 697 |
+
"step": 485
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.25,
|
| 701 |
+
"grad_norm": 242811.53125,
|
| 702 |
+
"learning_rate": 6.25e-06,
|
| 703 |
+
"loss": 1.6764,
|
| 704 |
+
"step": 490
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.2627551020408163,
|
| 708 |
+
"grad_norm": 88326.6171875,
|
| 709 |
+
"learning_rate": 6.313775510204082e-06,
|
| 710 |
+
"loss": 1.7684,
|
| 711 |
+
"step": 495
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.2755102040816326,
|
| 715 |
+
"grad_norm": 64742.67578125,
|
| 716 |
+
"learning_rate": 6.3775510204081635e-06,
|
| 717 |
+
"loss": 1.788,
|
| 718 |
+
"step": 500
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.288265306122449,
|
| 722 |
+
"grad_norm": 130764.1015625,
|
| 723 |
+
"learning_rate": 6.4413265306122455e-06,
|
| 724 |
+
"loss": 1.6374,
|
| 725 |
+
"step": 505
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.3010204081632653,
|
| 729 |
+
"grad_norm": 105262.7578125,
|
| 730 |
+
"learning_rate": 6.505102040816327e-06,
|
| 731 |
+
"loss": 1.7669,
|
| 732 |
+
"step": 510
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.3137755102040816,
|
| 736 |
+
"grad_norm": 132783.3125,
|
| 737 |
+
"learning_rate": 6.568877551020409e-06,
|
| 738 |
+
"loss": 1.7428,
|
| 739 |
+
"step": 515
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.3265306122448979,
|
| 743 |
+
"grad_norm": 109851.921875,
|
| 744 |
+
"learning_rate": 6.63265306122449e-06,
|
| 745 |
+
"loss": 1.7023,
|
| 746 |
+
"step": 520
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.3392857142857144,
|
| 750 |
+
"grad_norm": 65197.0859375,
|
| 751 |
+
"learning_rate": 6.696428571428571e-06,
|
| 752 |
+
"loss": 1.7407,
|
| 753 |
+
"step": 525
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.3520408163265305,
|
| 757 |
+
"grad_norm": 158283.9375,
|
| 758 |
+
"learning_rate": 6.760204081632653e-06,
|
| 759 |
+
"loss": 1.7629,
|
| 760 |
+
"step": 530
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.364795918367347,
|
| 764 |
+
"grad_norm": 121252.4140625,
|
| 765 |
+
"learning_rate": 6.823979591836736e-06,
|
| 766 |
+
"loss": 1.7242,
|
| 767 |
+
"step": 535
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.3775510204081631,
|
| 771 |
+
"grad_norm": 110898.328125,
|
| 772 |
+
"learning_rate": 6.887755102040817e-06,
|
| 773 |
+
"loss": 1.6922,
|
| 774 |
+
"step": 540
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.3903061224489797,
|
| 778 |
+
"grad_norm": 127766.8046875,
|
| 779 |
+
"learning_rate": 6.951530612244899e-06,
|
| 780 |
+
"loss": 1.7274,
|
| 781 |
+
"step": 545
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.403061224489796,
|
| 785 |
+
"grad_norm": 138821.59375,
|
| 786 |
+
"learning_rate": 7.01530612244898e-06,
|
| 787 |
+
"loss": 1.6859,
|
| 788 |
+
"step": 550
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.4158163265306123,
|
| 792 |
+
"grad_norm": 154254.75,
|
| 793 |
+
"learning_rate": 7.079081632653062e-06,
|
| 794 |
+
"loss": 1.6988,
|
| 795 |
+
"step": 555
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.4285714285714286,
|
| 799 |
+
"grad_norm": 122805.65625,
|
| 800 |
+
"learning_rate": 7.1428571428571436e-06,
|
| 801 |
+
"loss": 1.7003,
|
| 802 |
+
"step": 560
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.441326530612245,
|
| 806 |
+
"grad_norm": 35300.796875,
|
| 807 |
+
"learning_rate": 7.206632653061226e-06,
|
| 808 |
+
"loss": 1.7224,
|
| 809 |
+
"step": 565
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.4540816326530612,
|
| 813 |
+
"grad_norm": 125603.1015625,
|
| 814 |
+
"learning_rate": 7.270408163265307e-06,
|
| 815 |
+
"loss": 1.8201,
|
| 816 |
+
"step": 570
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.4668367346938775,
|
| 820 |
+
"grad_norm": 167185.671875,
|
| 821 |
+
"learning_rate": 7.334183673469388e-06,
|
| 822 |
+
"loss": 1.7959,
|
| 823 |
+
"step": 575
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.4795918367346939,
|
| 827 |
+
"grad_norm": 59943.359375,
|
| 828 |
+
"learning_rate": 7.39795918367347e-06,
|
| 829 |
+
"loss": 1.6133,
|
| 830 |
+
"step": 580
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.4923469387755102,
|
| 834 |
+
"grad_norm": 197770.234375,
|
| 835 |
+
"learning_rate": 7.461734693877551e-06,
|
| 836 |
+
"loss": 1.709,
|
| 837 |
+
"step": 585
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 1.5051020408163265,
|
| 841 |
+
"grad_norm": 182674.5,
|
| 842 |
+
"learning_rate": 7.525510204081633e-06,
|
| 843 |
+
"loss": 1.688,
|
| 844 |
+
"step": 590
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 1.5178571428571428,
|
| 848 |
+
"grad_norm": 90262.0390625,
|
| 849 |
+
"learning_rate": 7.589285714285714e-06,
|
| 850 |
+
"loss": 1.7639,
|
| 851 |
+
"step": 595
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 1.5306122448979593,
|
| 855 |
+
"grad_norm": 119569.0859375,
|
| 856 |
+
"learning_rate": 7.653061224489796e-06,
|
| 857 |
+
"loss": 1.7587,
|
| 858 |
+
"step": 600
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 1.5433673469387754,
|
| 862 |
+
"grad_norm": 134583.546875,
|
| 863 |
+
"learning_rate": 7.716836734693878e-06,
|
| 864 |
+
"loss": 1.7174,
|
| 865 |
+
"step": 605
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 1.556122448979592,
|
| 869 |
+
"grad_norm": 123799.515625,
|
| 870 |
+
"learning_rate": 7.78061224489796e-06,
|
| 871 |
+
"loss": 1.668,
|
| 872 |
+
"step": 610
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.568877551020408,
|
| 876 |
+
"grad_norm": 124765.65625,
|
| 877 |
+
"learning_rate": 7.844387755102042e-06,
|
| 878 |
+
"loss": 1.691,
|
| 879 |
+
"step": 615
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.5816326530612246,
|
| 883 |
+
"grad_norm": 180459.28125,
|
| 884 |
+
"learning_rate": 7.908163265306124e-06,
|
| 885 |
+
"loss": 1.6133,
|
| 886 |
+
"step": 620
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.5943877551020407,
|
| 890 |
+
"grad_norm": 115367.0859375,
|
| 891 |
+
"learning_rate": 7.971938775510205e-06,
|
| 892 |
+
"loss": 1.7529,
|
| 893 |
+
"step": 625
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 1.6071428571428572,
|
| 897 |
+
"grad_norm": 66548.40625,
|
| 898 |
+
"learning_rate": 8.035714285714286e-06,
|
| 899 |
+
"loss": 1.7157,
|
| 900 |
+
"step": 630
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.6198979591836735,
|
| 904 |
+
"grad_norm": 152392.640625,
|
| 905 |
+
"learning_rate": 8.099489795918369e-06,
|
| 906 |
+
"loss": 1.7417,
|
| 907 |
+
"step": 635
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.6326530612244898,
|
| 911 |
+
"grad_norm": 131193.015625,
|
| 912 |
+
"learning_rate": 8.16326530612245e-06,
|
| 913 |
+
"loss": 1.7704,
|
| 914 |
+
"step": 640
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.6454081632653061,
|
| 918 |
+
"grad_norm": 154705.71875,
|
| 919 |
+
"learning_rate": 8.227040816326531e-06,
|
| 920 |
+
"loss": 1.6573,
|
| 921 |
+
"step": 645
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.6581632653061225,
|
| 925 |
+
"grad_norm": 113254.484375,
|
| 926 |
+
"learning_rate": 8.290816326530612e-06,
|
| 927 |
+
"loss": 1.6394,
|
| 928 |
+
"step": 650
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.6709183673469388,
|
| 932 |
+
"grad_norm": 163821.484375,
|
| 933 |
+
"learning_rate": 8.354591836734695e-06,
|
| 934 |
+
"loss": 1.6884,
|
| 935 |
+
"step": 655
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.683673469387755,
|
| 939 |
+
"grad_norm": 141981.828125,
|
| 940 |
+
"learning_rate": 8.418367346938776e-06,
|
| 941 |
+
"loss": 1.5955,
|
| 942 |
+
"step": 660
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.6964285714285714,
|
| 946 |
+
"grad_norm": 158287.6875,
|
| 947 |
+
"learning_rate": 8.482142857142858e-06,
|
| 948 |
+
"loss": 1.6885,
|
| 949 |
+
"step": 665
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.7091836734693877,
|
| 953 |
+
"grad_norm": 166876.4375,
|
| 954 |
+
"learning_rate": 8.545918367346939e-06,
|
| 955 |
+
"loss": 1.7253,
|
| 956 |
+
"step": 670
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.7219387755102042,
|
| 960 |
+
"grad_norm": 111882.171875,
|
| 961 |
+
"learning_rate": 8.609693877551022e-06,
|
| 962 |
+
"loss": 1.7057,
|
| 963 |
+
"step": 675
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.7346938775510203,
|
| 967 |
+
"grad_norm": 237840.828125,
|
| 968 |
+
"learning_rate": 8.673469387755103e-06,
|
| 969 |
+
"loss": 1.7132,
|
| 970 |
+
"step": 680
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.7474489795918369,
|
| 974 |
+
"grad_norm": 83344.59375,
|
| 975 |
+
"learning_rate": 8.737244897959184e-06,
|
| 976 |
+
"loss": 1.733,
|
| 977 |
+
"step": 685
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 1.760204081632653,
|
| 981 |
+
"grad_norm": 30312.462890625,
|
| 982 |
+
"learning_rate": 8.801020408163265e-06,
|
| 983 |
+
"loss": 1.7926,
|
| 984 |
+
"step": 690
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.7729591836734695,
|
| 988 |
+
"grad_norm": 89868.9609375,
|
| 989 |
+
"learning_rate": 8.864795918367348e-06,
|
| 990 |
+
"loss": 1.6389,
|
| 991 |
+
"step": 695
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 1.7857142857142856,
|
| 995 |
+
"grad_norm": 182723.09375,
|
| 996 |
+
"learning_rate": 8.92857142857143e-06,
|
| 997 |
+
"loss": 1.7533,
|
| 998 |
+
"step": 700
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.7984693877551021,
|
| 1002 |
+
"grad_norm": 104328.2421875,
|
| 1003 |
+
"learning_rate": 8.992346938775512e-06,
|
| 1004 |
+
"loss": 1.7424,
|
| 1005 |
+
"step": 705
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.8112244897959182,
|
| 1009 |
+
"grad_norm": 110120.609375,
|
| 1010 |
+
"learning_rate": 9.056122448979593e-06,
|
| 1011 |
+
"loss": 1.6961,
|
| 1012 |
+
"step": 710
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.8239795918367347,
|
| 1016 |
+
"grad_norm": 114349.7578125,
|
| 1017 |
+
"learning_rate": 9.119897959183674e-06,
|
| 1018 |
+
"loss": 1.7753,
|
| 1019 |
+
"step": 715
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 1.836734693877551,
|
| 1023 |
+
"grad_norm": 103603.8828125,
|
| 1024 |
+
"learning_rate": 9.183673469387756e-06,
|
| 1025 |
+
"loss": 1.777,
|
| 1026 |
+
"step": 720
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 1.8494897959183674,
|
| 1030 |
+
"grad_norm": 173858.828125,
|
| 1031 |
+
"learning_rate": 9.247448979591837e-06,
|
| 1032 |
+
"loss": 1.7218,
|
| 1033 |
+
"step": 725
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 1.8622448979591837,
|
| 1037 |
+
"grad_norm": 255108.96875,
|
| 1038 |
+
"learning_rate": 9.31122448979592e-06,
|
| 1039 |
+
"loss": 1.7319,
|
| 1040 |
+
"step": 730
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.875,
|
| 1044 |
+
"grad_norm": 38653.734375,
|
| 1045 |
+
"learning_rate": 9.375000000000001e-06,
|
| 1046 |
+
"loss": 1.6574,
|
| 1047 |
+
"step": 735
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 1.8877551020408163,
|
| 1051 |
+
"grad_norm": 141534.234375,
|
| 1052 |
+
"learning_rate": 9.438775510204082e-06,
|
| 1053 |
+
"loss": 1.7595,
|
| 1054 |
+
"step": 740
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 1.9005102040816326,
|
| 1058 |
+
"grad_norm": 182591.890625,
|
| 1059 |
+
"learning_rate": 9.502551020408163e-06,
|
| 1060 |
+
"loss": 1.6796,
|
| 1061 |
+
"step": 745
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.913265306122449,
|
| 1065 |
+
"grad_norm": 75052.21875,
|
| 1066 |
+
"learning_rate": 9.566326530612246e-06,
|
| 1067 |
+
"loss": 1.754,
|
| 1068 |
+
"step": 750
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 1.9260204081632653,
|
| 1072 |
+
"grad_norm": 108983.921875,
|
| 1073 |
+
"learning_rate": 9.630102040816327e-06,
|
| 1074 |
+
"loss": 1.6907,
|
| 1075 |
+
"step": 755
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 1.9387755102040818,
|
| 1079 |
+
"grad_norm": 110641.828125,
|
| 1080 |
+
"learning_rate": 9.693877551020408e-06,
|
| 1081 |
+
"loss": 1.7041,
|
| 1082 |
+
"step": 760
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 1.9515306122448979,
|
| 1086 |
+
"grad_norm": 119189.8203125,
|
| 1087 |
+
"learning_rate": 9.75765306122449e-06,
|
| 1088 |
+
"loss": 1.6054,
|
| 1089 |
+
"step": 765
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 1.9642857142857144,
|
| 1093 |
+
"grad_norm": 62365.9453125,
|
| 1094 |
+
"learning_rate": 9.821428571428573e-06,
|
| 1095 |
+
"loss": 1.6862,
|
| 1096 |
+
"step": 770
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 1.9770408163265305,
|
| 1100 |
+
"grad_norm": 103043.6171875,
|
| 1101 |
+
"learning_rate": 9.885204081632654e-06,
|
| 1102 |
+
"loss": 1.6596,
|
| 1103 |
+
"step": 775
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 1.989795918367347,
|
| 1107 |
+
"grad_norm": 117368.546875,
|
| 1108 |
+
"learning_rate": 9.948979591836737e-06,
|
| 1109 |
+
"loss": 1.6737,
|
| 1110 |
+
"step": 780
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 2.0,
|
| 1114 |
+
"eval_accuracy": 0.0857774206508638,
|
| 1115 |
+
"eval_loss": 1.8835231065750122,
|
| 1116 |
+
"eval_runtime": 308.4538,
|
| 1117 |
+
"eval_samples_per_second": 16.139,
|
| 1118 |
+
"eval_steps_per_second": 0.538,
|
| 1119 |
+
"step": 784
|
| 1120 |
+
}
|
| 1121 |
+
],
|
| 1122 |
+
"logging_steps": 5,
|
| 1123 |
+
"max_steps": 3920,
|
| 1124 |
+
"num_input_tokens_seen": 0,
|
| 1125 |
+
"num_train_epochs": 10,
|
| 1126 |
+
"save_steps": 500,
|
| 1127 |
+
"stateful_callbacks": {
|
| 1128 |
+
"TrainerControl": {
|
| 1129 |
+
"args": {
|
| 1130 |
+
"should_epoch_stop": false,
|
| 1131 |
+
"should_evaluate": false,
|
| 1132 |
+
"should_log": false,
|
| 1133 |
+
"should_save": true,
|
| 1134 |
+
"should_training_stop": false
|
| 1135 |
+
},
|
| 1136 |
+
"attributes": {}
|
| 1137 |
+
}
|
| 1138 |
+
},
|
| 1139 |
+
"total_flos": 2.99358138673152e+18,
|
| 1140 |
+
"train_batch_size": 30,
|
| 1141 |
+
"trial_name": null,
|
| 1142 |
+
"trial_params": null
|
| 1143 |
+
}
|
checkpoint-784/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:62d4afa24b2e5344ef9e2ace3afc7237855a36ba40c6701cd39cbf7db81b4cdc
|
| 3 |
+
size 5176
|