Model save
Browse files- README.md +106 -0
- model.safetensors +1 -1
- trainer_state.json +756 -0
README.md
ADDED
|
@@ -0,0 +1,106 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: transformers
|
| 3 |
+
license: apache-2.0
|
| 4 |
+
base_model: facebook/wav2vec2-xls-r-300m
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
model-index:
|
| 8 |
+
- name: ikema-asr
|
| 9 |
+
results: []
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 13 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 14 |
+
|
| 15 |
+
# ikema-asr
|
| 16 |
+
|
| 17 |
+
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on an unknown dataset.
|
| 18 |
+
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 4.4069
|
| 20 |
+
- Cer: 0.7119
|
| 21 |
+
|
| 22 |
+
## Model description
|
| 23 |
+
|
| 24 |
+
More information needed
|
| 25 |
+
|
| 26 |
+
## Intended uses & limitations
|
| 27 |
+
|
| 28 |
+
More information needed
|
| 29 |
+
|
| 30 |
+
## Training and evaluation data
|
| 31 |
+
|
| 32 |
+
More information needed
|
| 33 |
+
|
| 34 |
+
## Training procedure
|
| 35 |
+
|
| 36 |
+
### Training hyperparameters
|
| 37 |
+
|
| 38 |
+
The following hyperparameters were used during training:
|
| 39 |
+
- learning_rate: 0.0003
|
| 40 |
+
- train_batch_size: 16
|
| 41 |
+
- eval_batch_size: 8
|
| 42 |
+
- seed: 42
|
| 43 |
+
- gradient_accumulation_steps: 2
|
| 44 |
+
- total_train_batch_size: 32
|
| 45 |
+
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 46 |
+
- lr_scheduler_type: linear
|
| 47 |
+
- lr_scheduler_warmup_steps: 100
|
| 48 |
+
- num_epochs: 50
|
| 49 |
+
- mixed_precision_training: Native AMP
|
| 50 |
+
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
| Training Loss | Epoch | Step | Validation Loss | Cer |
|
| 54 |
+
|:-------------:|:-------:|:----:|:---------------:|:------:|
|
| 55 |
+
| 11.3881 | 1.1117 | 100 | 3.8517 | 0.9927 |
|
| 56 |
+
| 3.9077 | 2.2235 | 200 | 3.6587 | 0.9925 |
|
| 57 |
+
| 3.8977 | 3.3352 | 300 | 3.6576 | 0.9927 |
|
| 58 |
+
| 3.8192 | 4.4469 | 400 | 3.6630 | 0.9929 |
|
| 59 |
+
| 3.7035 | 5.5587 | 500 | 3.6788 | 0.9789 |
|
| 60 |
+
| 3.6207 | 6.6704 | 600 | 3.6348 | 0.9544 |
|
| 61 |
+
| 3.5701 | 7.7821 | 700 | 3.5943 | 0.9705 |
|
| 62 |
+
| 3.5264 | 8.8939 | 800 | 3.5571 | 0.9585 |
|
| 63 |
+
| 3.3878 | 10.0 | 900 | 3.5424 | 0.9619 |
|
| 64 |
+
| 3.2684 | 11.1117 | 1000 | 3.3908 | 0.9472 |
|
| 65 |
+
| 3.0432 | 12.2235 | 1100 | 3.5293 | 0.8914 |
|
| 66 |
+
| 2.7166 | 13.3352 | 1200 | 2.8965 | 0.8707 |
|
| 67 |
+
| 2.372 | 14.4469 | 1300 | 2.9027 | 0.8354 |
|
| 68 |
+
| 2.1382 | 15.5587 | 1400 | 2.7674 | 0.8356 |
|
| 69 |
+
| 1.9442 | 16.6704 | 1500 | 2.9376 | 0.8006 |
|
| 70 |
+
| 1.7881 | 17.7821 | 1600 | 2.7831 | 0.7228 |
|
| 71 |
+
| 1.6766 | 18.8939 | 1700 | 2.8390 | 0.7168 |
|
| 72 |
+
| 1.5416 | 20.0 | 1800 | 3.0158 | 0.7305 |
|
| 73 |
+
| 1.4251 | 21.1117 | 1900 | 2.9839 | 0.6959 |
|
| 74 |
+
| 1.32 | 22.2235 | 2000 | 2.9795 | 0.7089 |
|
| 75 |
+
| 1.2355 | 23.3352 | 2100 | 3.1453 | 0.7217 |
|
| 76 |
+
| 1.1704 | 24.4469 | 2200 | 3.0608 | 0.7028 |
|
| 77 |
+
| 1.0789 | 25.5587 | 2300 | 3.0710 | 0.7133 |
|
| 78 |
+
| 1.0109 | 26.6704 | 2400 | 3.1124 | 0.7096 |
|
| 79 |
+
| 0.9362 | 27.7821 | 2500 | 3.4385 | 0.7221 |
|
| 80 |
+
| 0.8671 | 28.8939 | 2600 | 3.3170 | 0.7152 |
|
| 81 |
+
| 0.831 | 30.0 | 2700 | 3.2871 | 0.7012 |
|
| 82 |
+
| 0.7467 | 31.1117 | 2800 | 3.5183 | 0.7048 |
|
| 83 |
+
| 0.7128 | 32.2235 | 2900 | 3.3886 | 0.7088 |
|
| 84 |
+
| 0.6815 | 33.3352 | 3000 | 3.7318 | 0.7230 |
|
| 85 |
+
| 0.6452 | 34.4469 | 3100 | 3.7760 | 0.7320 |
|
| 86 |
+
| 0.6052 | 35.5587 | 3200 | 3.8134 | 0.7281 |
|
| 87 |
+
| 0.5779 | 36.6704 | 3300 | 3.7409 | 0.6809 |
|
| 88 |
+
| 0.5343 | 37.7821 | 3400 | 3.8252 | 0.6907 |
|
| 89 |
+
| 0.5056 | 38.8939 | 3500 | 3.9913 | 0.6964 |
|
| 90 |
+
| 0.5017 | 40.0 | 3600 | 4.1250 | 0.7000 |
|
| 91 |
+
| 0.4592 | 41.1117 | 3700 | 4.2096 | 0.7221 |
|
| 92 |
+
| 0.4478 | 42.2235 | 3800 | 4.2160 | 0.7031 |
|
| 93 |
+
| 0.4172 | 43.3352 | 3900 | 4.2452 | 0.7085 |
|
| 94 |
+
| 0.4018 | 44.4469 | 4000 | 4.3075 | 0.7155 |
|
| 95 |
+
| 0.3854 | 45.5587 | 4100 | 4.3594 | 0.7095 |
|
| 96 |
+
| 0.3771 | 46.6704 | 4200 | 4.3663 | 0.7133 |
|
| 97 |
+
| 0.3553 | 47.7821 | 4300 | 4.4198 | 0.7135 |
|
| 98 |
+
| 0.3565 | 48.8939 | 4400 | 4.4136 | 0.7108 |
|
| 99 |
+
|
| 100 |
+
|
| 101 |
+
### Framework versions
|
| 102 |
+
|
| 103 |
+
- Transformers 4.51.2
|
| 104 |
+
- Pytorch 2.6.0+cu124
|
| 105 |
+
- Datasets 3.5.0
|
| 106 |
+
- Tokenizers 0.21.1
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1262184680
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3e0f158d4d142677c96c2b93799df06b1b3de814cf895891f747efa768684e27
|
| 3 |
size 1262184680
|
trainer_state.json
ADDED
|
@@ -0,0 +1,756 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 49.44692737430167,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
+
"global_step": 4450,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 1.111731843575419,
|
| 14 |
+
"grad_norm": 0.48533013463020325,
|
| 15 |
+
"learning_rate": 0.00028199999999999997,
|
| 16 |
+
"loss": 11.3881,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 1.111731843575419,
|
| 21 |
+
"eval_cer": 0.9926923478442427,
|
| 22 |
+
"eval_loss": 3.8517303466796875,
|
| 23 |
+
"eval_runtime": 12.8394,
|
| 24 |
+
"eval_samples_per_second": 22.197,
|
| 25 |
+
"eval_steps_per_second": 2.804,
|
| 26 |
+
"step": 100
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 2.223463687150838,
|
| 30 |
+
"grad_norm": 2.8934638500213623,
|
| 31 |
+
"learning_rate": 0.0002935172413793103,
|
| 32 |
+
"loss": 3.9077,
|
| 33 |
+
"step": 200
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 2.223463687150838,
|
| 37 |
+
"eval_cer": 0.9924835577826495,
|
| 38 |
+
"eval_loss": 3.6587204933166504,
|
| 39 |
+
"eval_runtime": 12.8147,
|
| 40 |
+
"eval_samples_per_second": 22.24,
|
| 41 |
+
"eval_steps_per_second": 2.809,
|
| 42 |
+
"step": 200
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 3.335195530726257,
|
| 46 |
+
"grad_norm": 2.12126088142395,
|
| 47 |
+
"learning_rate": 0.0002866206896551724,
|
| 48 |
+
"loss": 3.8977,
|
| 49 |
+
"step": 300
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 3.335195530726257,
|
| 53 |
+
"eval_cer": 0.9926923478442427,
|
| 54 |
+
"eval_loss": 3.657623291015625,
|
| 55 |
+
"eval_runtime": 12.8572,
|
| 56 |
+
"eval_samples_per_second": 22.167,
|
| 57 |
+
"eval_steps_per_second": 2.8,
|
| 58 |
+
"step": 300
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 4.446927374301676,
|
| 62 |
+
"grad_norm": 1.5865833759307861,
|
| 63 |
+
"learning_rate": 0.00027972413793103445,
|
| 64 |
+
"loss": 3.8192,
|
| 65 |
+
"step": 400
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 4.446927374301676,
|
| 69 |
+
"eval_cer": 0.9929011379058357,
|
| 70 |
+
"eval_loss": 3.663019895553589,
|
| 71 |
+
"eval_runtime": 12.8178,
|
| 72 |
+
"eval_samples_per_second": 22.235,
|
| 73 |
+
"eval_steps_per_second": 2.809,
|
| 74 |
+
"step": 400
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 5.558659217877095,
|
| 78 |
+
"grad_norm": 0.37739208340644836,
|
| 79 |
+
"learning_rate": 0.00027282758620689656,
|
| 80 |
+
"loss": 3.7035,
|
| 81 |
+
"step": 500
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 5.558659217877095,
|
| 85 |
+
"eval_cer": 0.9789122037791002,
|
| 86 |
+
"eval_loss": 3.678802967071533,
|
| 87 |
+
"eval_runtime": 12.8176,
|
| 88 |
+
"eval_samples_per_second": 22.235,
|
| 89 |
+
"eval_steps_per_second": 2.809,
|
| 90 |
+
"step": 500
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 6.670391061452514,
|
| 94 |
+
"grad_norm": 0.6414264440536499,
|
| 95 |
+
"learning_rate": 0.00026593103448275857,
|
| 96 |
+
"loss": 3.6207,
|
| 97 |
+
"step": 600
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 6.670391061452514,
|
| 101 |
+
"eval_cer": 0.9543793715419147,
|
| 102 |
+
"eval_loss": 3.6347501277923584,
|
| 103 |
+
"eval_runtime": 12.7764,
|
| 104 |
+
"eval_samples_per_second": 22.307,
|
| 105 |
+
"eval_steps_per_second": 2.818,
|
| 106 |
+
"step": 600
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 7.782122905027933,
|
| 110 |
+
"grad_norm": 0.7569805979728699,
|
| 111 |
+
"learning_rate": 0.0002590344827586207,
|
| 112 |
+
"loss": 3.5701,
|
| 113 |
+
"step": 700
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 7.782122905027933,
|
| 117 |
+
"eval_cer": 0.9704562062845808,
|
| 118 |
+
"eval_loss": 3.594348669052124,
|
| 119 |
+
"eval_runtime": 12.8842,
|
| 120 |
+
"eval_samples_per_second": 22.12,
|
| 121 |
+
"eval_steps_per_second": 2.794,
|
| 122 |
+
"step": 700
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 8.893854748603353,
|
| 126 |
+
"grad_norm": 0.617978572845459,
|
| 127 |
+
"learning_rate": 0.00025213793103448273,
|
| 128 |
+
"loss": 3.5264,
|
| 129 |
+
"step": 800
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 8.893854748603353,
|
| 133 |
+
"eval_cer": 0.9584507777429795,
|
| 134 |
+
"eval_loss": 3.557063579559326,
|
| 135 |
+
"eval_runtime": 12.7852,
|
| 136 |
+
"eval_samples_per_second": 22.291,
|
| 137 |
+
"eval_steps_per_second": 2.816,
|
| 138 |
+
"step": 800
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 10.0,
|
| 142 |
+
"grad_norm": 1.1942850351333618,
|
| 143 |
+
"learning_rate": 0.00024524137931034484,
|
| 144 |
+
"loss": 3.3878,
|
| 145 |
+
"step": 900
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 10.0,
|
| 149 |
+
"eval_cer": 0.961895813759265,
|
| 150 |
+
"eval_loss": 3.5423524379730225,
|
| 151 |
+
"eval_runtime": 12.8758,
|
| 152 |
+
"eval_samples_per_second": 22.135,
|
| 153 |
+
"eval_steps_per_second": 2.796,
|
| 154 |
+
"step": 900
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 11.111731843575418,
|
| 158 |
+
"grad_norm": 0.859997570514679,
|
| 159 |
+
"learning_rate": 0.00023834482758620687,
|
| 160 |
+
"loss": 3.2684,
|
| 161 |
+
"step": 1000
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 11.111731843575418,
|
| 165 |
+
"eval_cer": 0.9471761144169537,
|
| 166 |
+
"eval_loss": 3.390798568725586,
|
| 167 |
+
"eval_runtime": 12.7862,
|
| 168 |
+
"eval_samples_per_second": 22.29,
|
| 169 |
+
"eval_steps_per_second": 2.816,
|
| 170 |
+
"step": 1000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 12.223463687150838,
|
| 174 |
+
"grad_norm": 2.701083183288574,
|
| 175 |
+
"learning_rate": 0.00023144827586206896,
|
| 176 |
+
"loss": 3.0432,
|
| 177 |
+
"step": 1100
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 12.223463687150838,
|
| 181 |
+
"eval_cer": 0.8914291679716045,
|
| 182 |
+
"eval_loss": 3.5293474197387695,
|
| 183 |
+
"eval_runtime": 12.8801,
|
| 184 |
+
"eval_samples_per_second": 22.127,
|
| 185 |
+
"eval_steps_per_second": 2.795,
|
| 186 |
+
"step": 1100
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 13.335195530726256,
|
| 190 |
+
"grad_norm": 1.2649680376052856,
|
| 191 |
+
"learning_rate": 0.00022455172413793101,
|
| 192 |
+
"loss": 2.7166,
|
| 193 |
+
"step": 1200
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 13.335195530726256,
|
| 197 |
+
"eval_cer": 0.8706545568430942,
|
| 198 |
+
"eval_loss": 2.8964927196502686,
|
| 199 |
+
"eval_runtime": 12.8464,
|
| 200 |
+
"eval_samples_per_second": 22.185,
|
| 201 |
+
"eval_steps_per_second": 2.802,
|
| 202 |
+
"step": 1200
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 14.446927374301676,
|
| 206 |
+
"grad_norm": 2.458191156387329,
|
| 207 |
+
"learning_rate": 0.0002176551724137931,
|
| 208 |
+
"loss": 2.372,
|
| 209 |
+
"step": 1300
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 14.446927374301676,
|
| 213 |
+
"eval_cer": 0.8353690364338657,
|
| 214 |
+
"eval_loss": 2.9027018547058105,
|
| 215 |
+
"eval_runtime": 12.8524,
|
| 216 |
+
"eval_samples_per_second": 22.175,
|
| 217 |
+
"eval_steps_per_second": 2.801,
|
| 218 |
+
"step": 1300
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 15.558659217877095,
|
| 222 |
+
"grad_norm": 2.145862102508545,
|
| 223 |
+
"learning_rate": 0.00021075862068965515,
|
| 224 |
+
"loss": 2.1382,
|
| 225 |
+
"step": 1400
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 15.558659217877095,
|
| 229 |
+
"eval_cer": 0.8355778264954589,
|
| 230 |
+
"eval_loss": 2.767353057861328,
|
| 231 |
+
"eval_runtime": 12.8847,
|
| 232 |
+
"eval_samples_per_second": 22.119,
|
| 233 |
+
"eval_steps_per_second": 2.794,
|
| 234 |
+
"step": 1400
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 16.670391061452513,
|
| 238 |
+
"grad_norm": 2.6955654621124268,
|
| 239 |
+
"learning_rate": 0.0002038620689655172,
|
| 240 |
+
"loss": 1.9442,
|
| 241 |
+
"step": 1500
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 16.670391061452513,
|
| 245 |
+
"eval_cer": 0.8006054911786199,
|
| 246 |
+
"eval_loss": 2.9376468658447266,
|
| 247 |
+
"eval_runtime": 12.8501,
|
| 248 |
+
"eval_samples_per_second": 22.179,
|
| 249 |
+
"eval_steps_per_second": 2.802,
|
| 250 |
+
"step": 1500
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 17.782122905027933,
|
| 254 |
+
"grad_norm": 1.8743767738342285,
|
| 255 |
+
"learning_rate": 0.0001969655172413793,
|
| 256 |
+
"loss": 1.7881,
|
| 257 |
+
"step": 1600
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 17.782122905027933,
|
| 261 |
+
"eval_cer": 0.722831193235202,
|
| 262 |
+
"eval_loss": 2.7831432819366455,
|
| 263 |
+
"eval_runtime": 12.7796,
|
| 264 |
+
"eval_samples_per_second": 22.301,
|
| 265 |
+
"eval_steps_per_second": 2.817,
|
| 266 |
+
"step": 1600
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 18.893854748603353,
|
| 270 |
+
"grad_norm": 2.472550392150879,
|
| 271 |
+
"learning_rate": 0.00019006896551724138,
|
| 272 |
+
"loss": 1.6766,
|
| 273 |
+
"step": 1700
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 18.893854748603353,
|
| 277 |
+
"eval_cer": 0.7167762814490031,
|
| 278 |
+
"eval_loss": 2.8390402793884277,
|
| 279 |
+
"eval_runtime": 12.8876,
|
| 280 |
+
"eval_samples_per_second": 22.114,
|
| 281 |
+
"eval_steps_per_second": 2.793,
|
| 282 |
+
"step": 1700
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 20.0,
|
| 286 |
+
"grad_norm": 8.466638565063477,
|
| 287 |
+
"learning_rate": 0.00018317241379310343,
|
| 288 |
+
"loss": 1.5416,
|
| 289 |
+
"step": 1800
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 20.0,
|
| 293 |
+
"eval_cer": 0.730452030483349,
|
| 294 |
+
"eval_loss": 3.0157971382141113,
|
| 295 |
+
"eval_runtime": 12.7638,
|
| 296 |
+
"eval_samples_per_second": 22.329,
|
| 297 |
+
"eval_steps_per_second": 2.82,
|
| 298 |
+
"step": 1800
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 21.11173184357542,
|
| 302 |
+
"grad_norm": 2.14225172996521,
|
| 303 |
+
"learning_rate": 0.0001762758620689655,
|
| 304 |
+
"loss": 1.4251,
|
| 305 |
+
"step": 1900
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 21.11173184357542,
|
| 309 |
+
"eval_cer": 0.6958972752896962,
|
| 310 |
+
"eval_loss": 2.983879327774048,
|
| 311 |
+
"eval_runtime": 12.9131,
|
| 312 |
+
"eval_samples_per_second": 22.071,
|
| 313 |
+
"eval_steps_per_second": 2.788,
|
| 314 |
+
"step": 1900
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 22.223463687150836,
|
| 318 |
+
"grad_norm": 4.268678188323975,
|
| 319 |
+
"learning_rate": 0.00016937931034482757,
|
| 320 |
+
"loss": 1.32,
|
| 321 |
+
"step": 2000
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 22.223463687150836,
|
| 325 |
+
"eval_cer": 0.7089466541392629,
|
| 326 |
+
"eval_loss": 2.9795498847961426,
|
| 327 |
+
"eval_runtime": 13.1212,
|
| 328 |
+
"eval_samples_per_second": 21.721,
|
| 329 |
+
"eval_steps_per_second": 2.744,
|
| 330 |
+
"step": 2000
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 23.335195530726256,
|
| 334 |
+
"grad_norm": 2.3187201023101807,
|
| 335 |
+
"learning_rate": 0.00016248275862068966,
|
| 336 |
+
"loss": 1.2355,
|
| 337 |
+
"step": 2100
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 23.335195530726256,
|
| 341 |
+
"eval_cer": 0.7216828478964401,
|
| 342 |
+
"eval_loss": 3.145298957824707,
|
| 343 |
+
"eval_runtime": 12.8542,
|
| 344 |
+
"eval_samples_per_second": 22.172,
|
| 345 |
+
"eval_steps_per_second": 2.801,
|
| 346 |
+
"step": 2100
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 24.446927374301676,
|
| 350 |
+
"grad_norm": 4.309772968292236,
|
| 351 |
+
"learning_rate": 0.00015558620689655171,
|
| 352 |
+
"loss": 1.1704,
|
| 353 |
+
"step": 2200
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 24.446927374301676,
|
| 357 |
+
"eval_cer": 0.7027873473222674,
|
| 358 |
+
"eval_loss": 3.060805320739746,
|
| 359 |
+
"eval_runtime": 12.8156,
|
| 360 |
+
"eval_samples_per_second": 22.238,
|
| 361 |
+
"eval_steps_per_second": 2.809,
|
| 362 |
+
"step": 2200
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 25.558659217877096,
|
| 366 |
+
"grad_norm": 2.0230820178985596,
|
| 367 |
+
"learning_rate": 0.00014868965517241377,
|
| 368 |
+
"loss": 1.0789,
|
| 369 |
+
"step": 2300
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 25.558659217877096,
|
| 373 |
+
"eval_cer": 0.7133312454327174,
|
| 374 |
+
"eval_loss": 3.070970058441162,
|
| 375 |
+
"eval_runtime": 12.8729,
|
| 376 |
+
"eval_samples_per_second": 22.139,
|
| 377 |
+
"eval_steps_per_second": 2.797,
|
| 378 |
+
"step": 2300
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 26.670391061452513,
|
| 382 |
+
"grad_norm": 3.8290631771087646,
|
| 383 |
+
"learning_rate": 0.00014179310344827585,
|
| 384 |
+
"loss": 1.0109,
|
| 385 |
+
"step": 2400
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 26.670391061452513,
|
| 389 |
+
"eval_cer": 0.7095730243240421,
|
| 390 |
+
"eval_loss": 3.1124370098114014,
|
| 391 |
+
"eval_runtime": 12.8909,
|
| 392 |
+
"eval_samples_per_second": 22.109,
|
| 393 |
+
"eval_steps_per_second": 2.793,
|
| 394 |
+
"step": 2400
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 27.782122905027933,
|
| 398 |
+
"grad_norm": 2.3735196590423584,
|
| 399 |
+
"learning_rate": 0.0001348965517241379,
|
| 400 |
+
"loss": 0.9362,
|
| 401 |
+
"step": 2500
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 27.782122905027933,
|
| 405 |
+
"eval_cer": 0.7221004280196263,
|
| 406 |
+
"eval_loss": 3.4384732246398926,
|
| 407 |
+
"eval_runtime": 12.8512,
|
| 408 |
+
"eval_samples_per_second": 22.177,
|
| 409 |
+
"eval_steps_per_second": 2.801,
|
| 410 |
+
"step": 2500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 28.893854748603353,
|
| 414 |
+
"grad_norm": 2.767505645751953,
|
| 415 |
+
"learning_rate": 0.000128,
|
| 416 |
+
"loss": 0.8671,
|
| 417 |
+
"step": 2600
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 28.893854748603353,
|
| 421 |
+
"eval_cer": 0.7152103559870551,
|
| 422 |
+
"eval_loss": 3.3170225620269775,
|
| 423 |
+
"eval_runtime": 12.9113,
|
| 424 |
+
"eval_samples_per_second": 22.074,
|
| 425 |
+
"eval_steps_per_second": 2.788,
|
| 426 |
+
"step": 2600
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 30.0,
|
| 430 |
+
"grad_norm": 7.469735622406006,
|
| 431 |
+
"learning_rate": 0.00012110344827586206,
|
| 432 |
+
"loss": 0.831,
|
| 433 |
+
"step": 2700
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 30.0,
|
| 437 |
+
"eval_cer": 0.7012214218603194,
|
| 438 |
+
"eval_loss": 3.2871029376983643,
|
| 439 |
+
"eval_runtime": 12.7981,
|
| 440 |
+
"eval_samples_per_second": 22.269,
|
| 441 |
+
"eval_steps_per_second": 2.813,
|
| 442 |
+
"step": 2700
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 31.11173184357542,
|
| 446 |
+
"grad_norm": 3.1189279556274414,
|
| 447 |
+
"learning_rate": 0.00011420689655172412,
|
| 448 |
+
"loss": 0.7467,
|
| 449 |
+
"step": 2800
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 31.11173184357542,
|
| 453 |
+
"eval_cer": 0.7047708529074016,
|
| 454 |
+
"eval_loss": 3.5182864665985107,
|
| 455 |
+
"eval_runtime": 12.9125,
|
| 456 |
+
"eval_samples_per_second": 22.072,
|
| 457 |
+
"eval_steps_per_second": 2.788,
|
| 458 |
+
"step": 2800
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 32.22346368715084,
|
| 462 |
+
"grad_norm": 14.725250244140625,
|
| 463 |
+
"learning_rate": 0.0001073103448275862,
|
| 464 |
+
"loss": 0.7128,
|
| 465 |
+
"step": 2900
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 32.22346368715084,
|
| 469 |
+
"eval_cer": 0.7088422591084664,
|
| 470 |
+
"eval_loss": 3.3886497020721436,
|
| 471 |
+
"eval_runtime": 12.9242,
|
| 472 |
+
"eval_samples_per_second": 22.052,
|
| 473 |
+
"eval_steps_per_second": 2.785,
|
| 474 |
+
"step": 2900
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 33.33519553072626,
|
| 478 |
+
"grad_norm": 2.3201019763946533,
|
| 479 |
+
"learning_rate": 0.00010041379310344826,
|
| 480 |
+
"loss": 0.6815,
|
| 481 |
+
"step": 3000
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 33.33519553072626,
|
| 485 |
+
"eval_cer": 0.7230399832967951,
|
| 486 |
+
"eval_loss": 3.731771230697632,
|
| 487 |
+
"eval_runtime": 12.9081,
|
| 488 |
+
"eval_samples_per_second": 22.079,
|
| 489 |
+
"eval_steps_per_second": 2.789,
|
| 490 |
+
"step": 3000
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 34.44692737430167,
|
| 494 |
+
"grad_norm": 4.227590084075928,
|
| 495 |
+
"learning_rate": 9.351724137931033e-05,
|
| 496 |
+
"loss": 0.6452,
|
| 497 |
+
"step": 3100
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 34.44692737430167,
|
| 501 |
+
"eval_cer": 0.732017955945297,
|
| 502 |
+
"eval_loss": 3.7759644985198975,
|
| 503 |
+
"eval_runtime": 12.9044,
|
| 504 |
+
"eval_samples_per_second": 22.085,
|
| 505 |
+
"eval_steps_per_second": 2.79,
|
| 506 |
+
"step": 3100
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 35.55865921787709,
|
| 510 |
+
"grad_norm": 1.658971905708313,
|
| 511 |
+
"learning_rate": 8.66206896551724e-05,
|
| 512 |
+
"loss": 0.6052,
|
| 513 |
+
"step": 3200
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 35.55865921787709,
|
| 517 |
+
"eval_cer": 0.7280509447750287,
|
| 518 |
+
"eval_loss": 3.8133554458618164,
|
| 519 |
+
"eval_runtime": 12.9295,
|
| 520 |
+
"eval_samples_per_second": 22.043,
|
| 521 |
+
"eval_steps_per_second": 2.784,
|
| 522 |
+
"step": 3200
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 36.67039106145251,
|
| 526 |
+
"grad_norm": 3.6614513397216797,
|
| 527 |
+
"learning_rate": 7.972413793103447e-05,
|
| 528 |
+
"loss": 0.5779,
|
| 529 |
+
"step": 3300
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 36.67039106145251,
|
| 533 |
+
"eval_cer": 0.6808643908549953,
|
| 534 |
+
"eval_loss": 3.7408699989318848,
|
| 535 |
+
"eval_runtime": 13.0708,
|
| 536 |
+
"eval_samples_per_second": 21.804,
|
| 537 |
+
"eval_steps_per_second": 2.754,
|
| 538 |
+
"step": 3300
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 37.78212290502793,
|
| 542 |
+
"grad_norm": 1.5677216053009033,
|
| 543 |
+
"learning_rate": 7.282758620689654e-05,
|
| 544 |
+
"loss": 0.5343,
|
| 545 |
+
"step": 3400
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 37.78212290502793,
|
| 549 |
+
"eval_cer": 0.6906775237498695,
|
| 550 |
+
"eval_loss": 3.825223922729492,
|
| 551 |
+
"eval_runtime": 12.9331,
|
| 552 |
+
"eval_samples_per_second": 22.036,
|
| 553 |
+
"eval_steps_per_second": 2.784,
|
| 554 |
+
"step": 3400
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 38.89385474860335,
|
| 558 |
+
"grad_norm": 2.8445885181427,
|
| 559 |
+
"learning_rate": 6.593103448275861e-05,
|
| 560 |
+
"loss": 0.5056,
|
| 561 |
+
"step": 3500
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 38.89385474860335,
|
| 565 |
+
"eval_cer": 0.6964192504436789,
|
| 566 |
+
"eval_loss": 3.9913222789764404,
|
| 567 |
+
"eval_runtime": 12.9414,
|
| 568 |
+
"eval_samples_per_second": 22.022,
|
| 569 |
+
"eval_steps_per_second": 2.782,
|
| 570 |
+
"step": 3500
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 40.0,
|
| 574 |
+
"grad_norm": 12.087656021118164,
|
| 575 |
+
"learning_rate": 5.903448275862069e-05,
|
| 576 |
+
"loss": 0.5017,
|
| 577 |
+
"step": 3600
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 40.0,
|
| 581 |
+
"eval_cer": 0.6999686814907611,
|
| 582 |
+
"eval_loss": 4.1249589920043945,
|
| 583 |
+
"eval_runtime": 12.9858,
|
| 584 |
+
"eval_samples_per_second": 21.947,
|
| 585 |
+
"eval_steps_per_second": 2.772,
|
| 586 |
+
"step": 3600
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 41.11173184357542,
|
| 590 |
+
"grad_norm": 1.838942527770996,
|
| 591 |
+
"learning_rate": 5.213793103448276e-05,
|
| 592 |
+
"loss": 0.4592,
|
| 593 |
+
"step": 3700
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 41.11173184357542,
|
| 597 |
+
"eval_cer": 0.7221004280196263,
|
| 598 |
+
"eval_loss": 4.20962381362915,
|
| 599 |
+
"eval_runtime": 12.8242,
|
| 600 |
+
"eval_samples_per_second": 22.224,
|
| 601 |
+
"eval_steps_per_second": 2.807,
|
| 602 |
+
"step": 3700
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 42.22346368715084,
|
| 606 |
+
"grad_norm": 7.289366245269775,
|
| 607 |
+
"learning_rate": 4.524137931034483e-05,
|
| 608 |
+
"loss": 0.4478,
|
| 609 |
+
"step": 3800
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 42.22346368715084,
|
| 613 |
+
"eval_cer": 0.7031005324146571,
|
| 614 |
+
"eval_loss": 4.215978622436523,
|
| 615 |
+
"eval_runtime": 12.8764,
|
| 616 |
+
"eval_samples_per_second": 22.134,
|
| 617 |
+
"eval_steps_per_second": 2.796,
|
| 618 |
+
"step": 3800
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 43.33519553072626,
|
| 622 |
+
"grad_norm": 1.9154551029205322,
|
| 623 |
+
"learning_rate": 3.834482758620689e-05,
|
| 624 |
+
"loss": 0.4172,
|
| 625 |
+
"step": 3900
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 43.33519553072626,
|
| 629 |
+
"eval_cer": 0.7085290740160768,
|
| 630 |
+
"eval_loss": 4.245218276977539,
|
| 631 |
+
"eval_runtime": 12.9248,
|
| 632 |
+
"eval_samples_per_second": 22.051,
|
| 633 |
+
"eval_steps_per_second": 2.785,
|
| 634 |
+
"step": 3900
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 44.44692737430167,
|
| 638 |
+
"grad_norm": 6.6539154052734375,
|
| 639 |
+
"learning_rate": 3.151724137931034e-05,
|
| 640 |
+
"loss": 0.4018,
|
| 641 |
+
"step": 4000
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 44.44692737430167,
|
| 645 |
+
"eval_cer": 0.7155235410794446,
|
| 646 |
+
"eval_loss": 4.3074631690979,
|
| 647 |
+
"eval_runtime": 12.9104,
|
| 648 |
+
"eval_samples_per_second": 22.075,
|
| 649 |
+
"eval_steps_per_second": 2.788,
|
| 650 |
+
"step": 4000
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 45.55865921787709,
|
| 654 |
+
"grad_norm": 1.8227887153625488,
|
| 655 |
+
"learning_rate": 2.4620689655172412e-05,
|
| 656 |
+
"loss": 0.3854,
|
| 657 |
+
"step": 4100
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 45.55865921787709,
|
| 661 |
+
"eval_cer": 0.7094686292932456,
|
| 662 |
+
"eval_loss": 4.359448432922363,
|
| 663 |
+
"eval_runtime": 12.9111,
|
| 664 |
+
"eval_samples_per_second": 22.074,
|
| 665 |
+
"eval_steps_per_second": 2.788,
|
| 666 |
+
"step": 4100
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 46.67039106145251,
|
| 670 |
+
"grad_norm": 3.2107598781585693,
|
| 671 |
+
"learning_rate": 1.7724137931034482e-05,
|
| 672 |
+
"loss": 0.3771,
|
| 673 |
+
"step": 4200
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 46.67039106145251,
|
| 677 |
+
"eval_cer": 0.7133312454327174,
|
| 678 |
+
"eval_loss": 4.366276741027832,
|
| 679 |
+
"eval_runtime": 12.9048,
|
| 680 |
+
"eval_samples_per_second": 22.085,
|
| 681 |
+
"eval_steps_per_second": 2.79,
|
| 682 |
+
"step": 4200
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 47.78212290502793,
|
| 686 |
+
"grad_norm": 1.2981507778167725,
|
| 687 |
+
"learning_rate": 1.082758620689655e-05,
|
| 688 |
+
"loss": 0.3553,
|
| 689 |
+
"step": 4300
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 47.78212290502793,
|
| 693 |
+
"eval_cer": 0.7135400354943104,
|
| 694 |
+
"eval_loss": 4.419826030731201,
|
| 695 |
+
"eval_runtime": 12.9283,
|
| 696 |
+
"eval_samples_per_second": 22.045,
|
| 697 |
+
"eval_steps_per_second": 2.785,
|
| 698 |
+
"step": 4300
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 48.89385474860335,
|
| 702 |
+
"grad_norm": 2.526685953140259,
|
| 703 |
+
"learning_rate": 3.93103448275862e-06,
|
| 704 |
+
"loss": 0.3565,
|
| 705 |
+
"step": 4400
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 48.89385474860335,
|
| 709 |
+
"eval_cer": 0.7108257646936006,
|
| 710 |
+
"eval_loss": 4.4135847091674805,
|
| 711 |
+
"eval_runtime": 12.8772,
|
| 712 |
+
"eval_samples_per_second": 22.132,
|
| 713 |
+
"eval_steps_per_second": 2.796,
|
| 714 |
+
"step": 4400
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 49.44692737430167,
|
| 718 |
+
"step": 4450,
|
| 719 |
+
"total_flos": 9.669635712078565e+18,
|
| 720 |
+
"train_loss": 1.799734133602528,
|
| 721 |
+
"train_runtime": 7166.9702,
|
| 722 |
+
"train_samples_per_second": 19.897,
|
| 723 |
+
"train_steps_per_second": 0.621
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 49.44692737430167,
|
| 727 |
+
"eval_cer": 0.7118697150015659,
|
| 728 |
+
"eval_loss": 4.406894207000732,
|
| 729 |
+
"eval_runtime": 12.8787,
|
| 730 |
+
"eval_samples_per_second": 22.13,
|
| 731 |
+
"eval_steps_per_second": 2.795,
|
| 732 |
+
"step": 4450
|
| 733 |
+
}
|
| 734 |
+
],
|
| 735 |
+
"logging_steps": 100,
|
| 736 |
+
"max_steps": 4450,
|
| 737 |
+
"num_input_tokens_seen": 0,
|
| 738 |
+
"num_train_epochs": 50,
|
| 739 |
+
"save_steps": 100,
|
| 740 |
+
"stateful_callbacks": {
|
| 741 |
+
"TrainerControl": {
|
| 742 |
+
"args": {
|
| 743 |
+
"should_epoch_stop": false,
|
| 744 |
+
"should_evaluate": false,
|
| 745 |
+
"should_log": false,
|
| 746 |
+
"should_save": true,
|
| 747 |
+
"should_training_stop": true
|
| 748 |
+
},
|
| 749 |
+
"attributes": {}
|
| 750 |
+
}
|
| 751 |
+
},
|
| 752 |
+
"total_flos": 9.669635712078565e+18,
|
| 753 |
+
"train_batch_size": 16,
|
| 754 |
+
"trial_name": null,
|
| 755 |
+
"trial_params": null
|
| 756 |
+
}
|