Commit
•
01d3eb9
1
Parent(s):
bc41011
Upload folder using huggingface_hub
Browse files- README.md +58 -19
- all_results.json +11 -11
- eval_results.json +6 -6
- model.safetensors +1 -1
- runs/Sep04_11-00-26_1edf4a34bc24/events.out.tfevents.1725444439.1edf4a34bc24.689.0 +3 -0
- runs/Sep04_11-00-26_1edf4a34bc24/events.out.tfevents.1725456098.1edf4a34bc24.689.1 +3 -0
- train_results.json +6 -6
- trainer_state.json +557 -94
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -15,12 +15,12 @@ model-index:
|
|
| 15 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 16 |
should probably proofread and complete it, then remove this comment. -->
|
| 17 |
|
| 18 |
-
# wav2vec2-btb-cv-ft-btb-cy
|
| 19 |
|
| 20 |
This model is a fine-tuned version of [DewiBrynJones/wav2vec2-xlsr-53-ft-btb-cv-cy](https://huggingface.co/DewiBrynJones/wav2vec2-xlsr-53-ft-btb-cv-cy) on an unknown dataset.
|
| 21 |
It achieves the following results on the evaluation set:
|
| 22 |
-
- Loss: 0.
|
| 23 |
-
- Wer: 0.
|
| 24 |
|
| 25 |
## Model description
|
| 26 |
|
|
@@ -40,30 +40,69 @@ More information needed
|
|
| 40 |
|
| 41 |
The following hyperparameters were used during training:
|
| 42 |
- learning_rate: 0.0003
|
| 43 |
-
- train_batch_size:
|
| 44 |
- eval_batch_size: 64
|
| 45 |
- seed: 42
|
| 46 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 47 |
- lr_scheduler_type: linear
|
| 48 |
-
- lr_scheduler_warmup_steps:
|
| 49 |
-
- training_steps:
|
| 50 |
- mixed_precision_training: Native AMP
|
| 51 |
|
| 52 |
### Training results
|
| 53 |
|
| 54 |
-
| Training Loss | Epoch | Step
|
| 55 |
-
|
| 56 |
-
| No log | 0.
|
| 57 |
-
| No log | 0.
|
| 58 |
-
|
|
| 59 |
-
|
|
| 60 |
-
| 0.
|
| 61 |
-
| 0.
|
| 62 |
-
| 0.
|
| 63 |
-
| 0.
|
| 64 |
-
| 0.
|
| 65 |
-
| 0.
|
| 66 |
-
| 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 67 |
|
| 68 |
|
| 69 |
### Framework versions
|
|
|
|
| 15 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 16 |
should probably proofread and complete it, then remove this comment. -->
|
| 17 |
|
| 18 |
+
# wav2vec2-btb-cv-ft-btb-cy-cand
|
| 19 |
|
| 20 |
This model is a fine-tuned version of [DewiBrynJones/wav2vec2-xlsr-53-ft-btb-cv-cy](https://huggingface.co/DewiBrynJones/wav2vec2-xlsr-53-ft-btb-cv-cy) on an unknown dataset.
|
| 21 |
It achieves the following results on the evaluation set:
|
| 22 |
+
- Loss: 0.4345
|
| 23 |
+
- Wer: 0.3308
|
| 24 |
|
| 25 |
## Model description
|
| 26 |
|
|
|
|
| 40 |
|
| 41 |
The following hyperparameters were used during training:
|
| 42 |
- learning_rate: 0.0003
|
| 43 |
+
- train_batch_size: 4
|
| 44 |
- eval_batch_size: 64
|
| 45 |
- seed: 42
|
| 46 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 47 |
- lr_scheduler_type: linear
|
| 48 |
+
- lr_scheduler_warmup_steps: 1000
|
| 49 |
+
- training_steps: 10000
|
| 50 |
- mixed_precision_training: Native AMP
|
| 51 |
|
| 52 |
### Training results
|
| 53 |
|
| 54 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 55 |
+
|:-------------:|:------:|:-----:|:---------------:|:------:|
|
| 56 |
+
| No log | 0.0285 | 200 | 1.2522 | 0.6292 |
|
| 57 |
+
| No log | 0.0570 | 400 | 0.6599 | 0.4544 |
|
| 58 |
+
| 2.2791 | 0.0854 | 600 | 0.6629 | 0.4395 |
|
| 59 |
+
| 2.2791 | 0.1139 | 800 | 0.7910 | 0.5453 |
|
| 60 |
+
| 0.8206 | 0.1424 | 1000 | 0.7758 | 0.5701 |
|
| 61 |
+
| 0.8206 | 0.1709 | 1200 | 0.8025 | 0.5783 |
|
| 62 |
+
| 0.8206 | 0.1994 | 1400 | 0.7715 | 0.5211 |
|
| 63 |
+
| 0.9068 | 0.2279 | 1600 | 0.7349 | 0.5128 |
|
| 64 |
+
| 0.9068 | 0.2563 | 1800 | 0.7258 | 0.5152 |
|
| 65 |
+
| 0.8679 | 0.2848 | 2000 | 0.7084 | 0.5216 |
|
| 66 |
+
| 0.8679 | 0.3133 | 2200 | 0.6904 | 0.5014 |
|
| 67 |
+
| 0.8679 | 0.3418 | 2400 | 0.6993 | 0.5178 |
|
| 68 |
+
| 0.8577 | 0.3703 | 2600 | 0.6746 | 0.4867 |
|
| 69 |
+
| 0.8577 | 0.3987 | 2800 | 0.6622 | 0.4963 |
|
| 70 |
+
| 0.7995 | 0.4272 | 3000 | 0.6793 | 0.4935 |
|
| 71 |
+
| 0.7995 | 0.4557 | 3200 | 0.6368 | 0.4701 |
|
| 72 |
+
| 0.7995 | 0.4842 | 3400 | 0.6363 | 0.4781 |
|
| 73 |
+
| 0.8141 | 0.5127 | 3600 | 0.6217 | 0.4656 |
|
| 74 |
+
| 0.8141 | 0.5412 | 3800 | 0.6418 | 0.4940 |
|
| 75 |
+
| 0.7953 | 0.5696 | 4000 | 0.6018 | 0.4542 |
|
| 76 |
+
| 0.7953 | 0.5981 | 4200 | 0.5962 | 0.4580 |
|
| 77 |
+
| 0.7953 | 0.6266 | 4400 | 0.5883 | 0.4459 |
|
| 78 |
+
| 0.7596 | 0.6551 | 4600 | 0.5788 | 0.4325 |
|
| 79 |
+
| 0.7596 | 0.6836 | 4800 | 0.5709 | 0.4412 |
|
| 80 |
+
| 0.7533 | 0.7120 | 5000 | 0.5595 | 0.4352 |
|
| 81 |
+
| 0.7533 | 0.7405 | 5200 | 0.5546 | 0.4232 |
|
| 82 |
+
| 0.7533 | 0.7690 | 5400 | 0.5545 | 0.4244 |
|
| 83 |
+
| 0.7591 | 0.7975 | 5600 | 0.5443 | 0.4076 |
|
| 84 |
+
| 0.7591 | 0.8260 | 5800 | 0.5341 | 0.4146 |
|
| 85 |
+
| 0.6621 | 0.8545 | 6000 | 0.5104 | 0.3955 |
|
| 86 |
+
| 0.6621 | 0.8829 | 6200 | 0.5139 | 0.4011 |
|
| 87 |
+
| 0.6621 | 0.9114 | 6400 | 0.5044 | 0.3804 |
|
| 88 |
+
| 0.6705 | 0.9399 | 6600 | 0.4999 | 0.3896 |
|
| 89 |
+
| 0.6705 | 0.9684 | 6800 | 0.5097 | 0.4053 |
|
| 90 |
+
| 0.6665 | 0.9969 | 7000 | 0.4925 | 0.3785 |
|
| 91 |
+
| 0.6665 | 1.0253 | 7200 | 0.4896 | 0.3689 |
|
| 92 |
+
| 0.6665 | 1.0538 | 7400 | 0.4749 | 0.3687 |
|
| 93 |
+
| 0.5826 | 1.0823 | 7600 | 0.4684 | 0.3628 |
|
| 94 |
+
| 0.5826 | 1.1108 | 7800 | 0.4729 | 0.3585 |
|
| 95 |
+
| 0.5836 | 1.1393 | 8000 | 0.4641 | 0.3553 |
|
| 96 |
+
| 0.5836 | 1.1678 | 8200 | 0.4575 | 0.3530 |
|
| 97 |
+
| 0.5836 | 1.1962 | 8400 | 0.4585 | 0.3486 |
|
| 98 |
+
| 0.5199 | 1.2247 | 8600 | 0.4549 | 0.3451 |
|
| 99 |
+
| 0.5199 | 1.2532 | 8800 | 0.4521 | 0.3408 |
|
| 100 |
+
| 0.5268 | 1.2817 | 9000 | 0.4425 | 0.3395 |
|
| 101 |
+
| 0.5268 | 1.3102 | 9200 | 0.4407 | 0.3362 |
|
| 102 |
+
| 0.5268 | 1.3386 | 9400 | 0.4383 | 0.3340 |
|
| 103 |
+
| 0.5013 | 1.3671 | 9600 | 0.4357 | 0.3325 |
|
| 104 |
+
| 0.5013 | 1.3956 | 9800 | 0.4350 | 0.3317 |
|
| 105 |
+
| 0.5095 | 1.4241 | 10000 | 0.4345 | 0.3308 |
|
| 106 |
|
| 107 |
|
| 108 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,15 +1,15 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch": 1.
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_runtime":
|
| 5 |
"eval_samples": 7022,
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 0.
|
| 8 |
-
"eval_wer": 0.
|
| 9 |
-
"total_flos": 4.
|
| 10 |
-
"train_loss": 0.
|
| 11 |
-
"train_runtime":
|
| 12 |
"train_samples": 28086,
|
| 13 |
-
"train_samples_per_second":
|
| 14 |
-
"train_steps_per_second": 0.
|
| 15 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 1.4240956992309883,
|
| 3 |
+
"eval_loss": 0.43451622128486633,
|
| 4 |
+
"eval_runtime": 181.7486,
|
| 5 |
"eval_samples": 7022,
|
| 6 |
+
"eval_samples_per_second": 38.636,
|
| 7 |
+
"eval_steps_per_second": 0.605,
|
| 8 |
+
"eval_wer": 0.3308175766353526,
|
| 9 |
+
"total_flos": 4.5974516642218747e+18,
|
| 10 |
+
"train_loss": 0.7817989181518554,
|
| 11 |
+
"train_runtime": 11412.7197,
|
| 12 |
"train_samples": 28086,
|
| 13 |
+
"train_samples_per_second": 3.505,
|
| 14 |
+
"train_steps_per_second": 0.876
|
| 15 |
}
|
eval_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch": 1.
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_runtime":
|
| 5 |
"eval_samples": 7022,
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 0.
|
| 8 |
-
"eval_wer": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 1.4240956992309883,
|
| 3 |
+
"eval_loss": 0.43451622128486633,
|
| 4 |
+
"eval_runtime": 181.7486,
|
| 5 |
"eval_samples": 7022,
|
| 6 |
+
"eval_samples_per_second": 38.636,
|
| 7 |
+
"eval_steps_per_second": 0.605,
|
| 8 |
+
"eval_wer": 0.3308175766353526
|
| 9 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1261975580
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0906239970b1fb662f4f0199c1cb5f14ad33bd3182b007b4301bba6fb9a447a5
|
| 3 |
size 1261975580
|
runs/Sep04_11-00-26_1edf4a34bc24/events.out.tfevents.1725444439.1edf4a34bc24.689.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dd685bcabb2ffac72aeb7d4bcde83ea6e04f6b8e0ee3fdb9eee49ae3919c88e2
|
| 3 |
+
size 27173
|
runs/Sep04_11-00-26_1edf4a34bc24/events.out.tfevents.1725456098.1edf4a34bc24.689.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a061b6ac14c2ac69ab079065661452dd9c80494b3fb57ae921fa45dc635bb337
|
| 3 |
+
size 406
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch": 1.
|
| 3 |
-
"total_flos": 4.
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 28086,
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 1.4240956992309883,
|
| 3 |
+
"total_flos": 4.5974516642218747e+18,
|
| 4 |
+
"train_loss": 0.7817989181518554,
|
| 5 |
+
"train_runtime": 11412.7197,
|
| 6 |
"train_samples": 28086,
|
| 7 |
+
"train_samples_per_second": 3.505,
|
| 8 |
+
"train_steps_per_second": 0.876
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -1,152 +1,615 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 1.
|
| 5 |
"eval_steps": 200,
|
| 6 |
-
"global_step":
|
| 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.
|
| 13 |
-
"eval_loss":
|
| 14 |
-
"eval_runtime": 184.
|
| 15 |
-
"eval_samples_per_second": 38.
|
| 16 |
-
"eval_steps_per_second": 0.
|
| 17 |
-
"eval_wer": 0.
|
| 18 |
"step": 200
|
| 19 |
},
|
| 20 |
{
|
| 21 |
-
"epoch": 0.
|
| 22 |
-
"eval_loss": 0.
|
| 23 |
-
"eval_runtime":
|
| 24 |
-
"eval_samples_per_second":
|
| 25 |
-
"eval_steps_per_second": 0.
|
| 26 |
-
"eval_wer": 0.
|
| 27 |
"step": 400
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
-
"grad_norm":
|
| 32 |
-
"learning_rate": 0.
|
| 33 |
-
"loss":
|
| 34 |
"step": 500
|
| 35 |
},
|
| 36 |
{
|
| 37 |
-
"epoch": 0.
|
| 38 |
-
"eval_loss": 0.
|
| 39 |
-
"eval_runtime":
|
| 40 |
-
"eval_samples_per_second":
|
| 41 |
-
"eval_steps_per_second": 0.
|
| 42 |
-
"eval_wer": 0.
|
| 43 |
"step": 600
|
| 44 |
},
|
| 45 |
{
|
| 46 |
-
"epoch": 0.
|
| 47 |
-
"eval_loss": 0.
|
| 48 |
-
"eval_runtime":
|
| 49 |
-
"eval_samples_per_second": 37.
|
| 50 |
-
"eval_steps_per_second": 0.
|
| 51 |
-
"eval_wer": 0.
|
| 52 |
"step": 800
|
| 53 |
},
|
| 54 |
{
|
| 55 |
-
"epoch": 0.
|
| 56 |
-
"grad_norm": 3.
|
| 57 |
-
"learning_rate": 0.
|
| 58 |
-
"loss": 0.
|
| 59 |
"step": 1000
|
| 60 |
},
|
| 61 |
{
|
| 62 |
-
"epoch": 0.
|
| 63 |
-
"eval_loss": 0.
|
| 64 |
-
"eval_runtime":
|
| 65 |
-
"eval_samples_per_second": 37.
|
| 66 |
-
"eval_steps_per_second": 0.
|
| 67 |
-
"eval_wer": 0.
|
| 68 |
"step": 1000
|
| 69 |
},
|
| 70 |
{
|
| 71 |
-
"epoch": 0.
|
| 72 |
-
"eval_loss": 0.
|
| 73 |
-
"eval_runtime":
|
| 74 |
-
"eval_samples_per_second": 37.
|
| 75 |
-
"eval_steps_per_second": 0.
|
| 76 |
-
"eval_wer": 0.
|
| 77 |
"step": 1200
|
| 78 |
},
|
| 79 |
{
|
| 80 |
-
"epoch": 0.
|
| 81 |
-
"eval_loss": 0.
|
| 82 |
-
"eval_runtime":
|
| 83 |
-
"eval_samples_per_second": 37.
|
| 84 |
-
"eval_steps_per_second": 0.
|
| 85 |
-
"eval_wer": 0.
|
| 86 |
"step": 1400
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch": 0.
|
| 90 |
-
"grad_norm":
|
| 91 |
-
"learning_rate": 0.
|
| 92 |
-
"loss": 0.
|
| 93 |
"step": 1500
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
-
"eval_loss": 0.
|
| 98 |
-
"eval_runtime":
|
| 99 |
-
"eval_samples_per_second":
|
| 100 |
-
"eval_steps_per_second": 0.
|
| 101 |
-
"eval_wer": 0.
|
| 102 |
"step": 1600
|
| 103 |
},
|
| 104 |
{
|
| 105 |
-
"epoch":
|
| 106 |
-
"eval_loss": 0.
|
| 107 |
-
"eval_runtime":
|
| 108 |
-
"eval_samples_per_second": 37.
|
| 109 |
-
"eval_steps_per_second": 0.
|
| 110 |
-
"eval_wer": 0.
|
| 111 |
"step": 1800
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch":
|
| 115 |
-
"grad_norm":
|
| 116 |
-
"learning_rate":
|
| 117 |
-
"loss": 0.
|
| 118 |
"step": 2000
|
| 119 |
},
|
| 120 |
{
|
| 121 |
-
"epoch":
|
| 122 |
-
"eval_loss": 0.
|
| 123 |
-
"eval_runtime":
|
| 124 |
-
"eval_samples_per_second": 37.
|
| 125 |
-
"eval_steps_per_second": 0.
|
| 126 |
-
"eval_wer": 0.
|
| 127 |
"step": 2000
|
| 128 |
},
|
| 129 |
{
|
| 130 |
-
"epoch":
|
| 131 |
-
"eval_loss": 0.
|
| 132 |
-
"eval_runtime":
|
| 133 |
-
"eval_samples_per_second": 37.
|
| 134 |
-
"eval_steps_per_second": 0.
|
| 135 |
-
"eval_wer": 0.
|
| 136 |
"step": 2200
|
| 137 |
},
|
| 138 |
{
|
| 139 |
-
"epoch":
|
| 140 |
-
"
|
| 141 |
-
"
|
| 142 |
-
"
|
| 143 |
-
"
|
| 144 |
-
"
|
| 145 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 146 |
}
|
| 147 |
],
|
| 148 |
"logging_steps": 500,
|
| 149 |
-
"max_steps":
|
| 150 |
"num_input_tokens_seen": 0,
|
| 151 |
"num_train_epochs": 2,
|
| 152 |
"save_steps": 500,
|
|
@@ -162,8 +625,8 @@
|
|
| 162 |
"attributes": {}
|
| 163 |
}
|
| 164 |
},
|
| 165 |
-
"total_flos": 4.
|
| 166 |
-
"train_batch_size":
|
| 167 |
"trial_name": null,
|
| 168 |
"trial_params": null
|
| 169 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.4240956992309883,
|
| 5 |
"eval_steps": 200,
|
| 6 |
+
"global_step": 10000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.028481913984619765,
|
| 13 |
+
"eval_loss": 1.2521600723266602,
|
| 14 |
+
"eval_runtime": 184.7297,
|
| 15 |
+
"eval_samples_per_second": 38.012,
|
| 16 |
+
"eval_steps_per_second": 0.595,
|
| 17 |
+
"eval_wer": 0.6291606319509959,
|
| 18 |
"step": 200
|
| 19 |
},
|
| 20 |
{
|
| 21 |
+
"epoch": 0.05696382796923953,
|
| 22 |
+
"eval_loss": 0.6599467396736145,
|
| 23 |
+
"eval_runtime": 185.0187,
|
| 24 |
+
"eval_samples_per_second": 37.953,
|
| 25 |
+
"eval_steps_per_second": 0.595,
|
| 26 |
+
"eval_wer": 0.45444398676570247,
|
| 27 |
"step": 400
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.07120478496154942,
|
| 31 |
+
"grad_norm": 18.865299224853516,
|
| 32 |
+
"learning_rate": 0.00014879999999999998,
|
| 33 |
+
"loss": 2.2791,
|
| 34 |
"step": 500
|
| 35 |
},
|
| 36 |
{
|
| 37 |
+
"epoch": 0.0854457419538593,
|
| 38 |
+
"eval_loss": 0.6628636717796326,
|
| 39 |
+
"eval_runtime": 185.7673,
|
| 40 |
+
"eval_samples_per_second": 37.8,
|
| 41 |
+
"eval_steps_per_second": 0.592,
|
| 42 |
+
"eval_wer": 0.4394557461566059,
|
| 43 |
"step": 600
|
| 44 |
},
|
| 45 |
{
|
| 46 |
+
"epoch": 0.11392765593847906,
|
| 47 |
+
"eval_loss": 0.7910040020942688,
|
| 48 |
+
"eval_runtime": 186.4058,
|
| 49 |
+
"eval_samples_per_second": 37.671,
|
| 50 |
+
"eval_steps_per_second": 0.59,
|
| 51 |
+
"eval_wer": 0.5453035517346763,
|
| 52 |
"step": 800
|
| 53 |
},
|
| 54 |
{
|
| 55 |
+
"epoch": 0.14240956992309883,
|
| 56 |
+
"grad_norm": 3.8627092838287354,
|
| 57 |
+
"learning_rate": 0.0002988,
|
| 58 |
+
"loss": 0.8206,
|
| 59 |
"step": 1000
|
| 60 |
},
|
| 61 |
{
|
| 62 |
+
"epoch": 0.14240956992309883,
|
| 63 |
+
"eval_loss": 0.7757941484451294,
|
| 64 |
+
"eval_runtime": 186.8087,
|
| 65 |
+
"eval_samples_per_second": 37.589,
|
| 66 |
+
"eval_steps_per_second": 0.589,
|
| 67 |
+
"eval_wer": 0.5701245033816553,
|
| 68 |
"step": 1000
|
| 69 |
},
|
| 70 |
{
|
| 71 |
+
"epoch": 0.1708914839077186,
|
| 72 |
+
"eval_loss": 0.802534818649292,
|
| 73 |
+
"eval_runtime": 187.4308,
|
| 74 |
+
"eval_samples_per_second": 37.464,
|
| 75 |
+
"eval_steps_per_second": 0.587,
|
| 76 |
+
"eval_wer": 0.5782564211589312,
|
| 77 |
"step": 1200
|
| 78 |
},
|
| 79 |
{
|
| 80 |
+
"epoch": 0.19937339789233838,
|
| 81 |
+
"eval_loss": 0.7715001106262207,
|
| 82 |
+
"eval_runtime": 187.8412,
|
| 83 |
+
"eval_samples_per_second": 37.383,
|
| 84 |
+
"eval_steps_per_second": 0.586,
|
| 85 |
+
"eval_wer": 0.5211336850077731,
|
| 86 |
"step": 1400
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.21361435488464825,
|
| 90 |
+
"grad_norm": 11.042049407958984,
|
| 91 |
+
"learning_rate": 0.00028346666666666665,
|
| 92 |
+
"loss": 0.9068,
|
| 93 |
"step": 1500
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.22785531187695812,
|
| 97 |
+
"eval_loss": 0.7349154949188232,
|
| 98 |
+
"eval_runtime": 191.6788,
|
| 99 |
+
"eval_samples_per_second": 36.634,
|
| 100 |
+
"eval_steps_per_second": 0.574,
|
| 101 |
+
"eval_wer": 0.512775880625573,
|
| 102 |
"step": 1600
|
| 103 |
},
|
| 104 |
{
|
| 105 |
+
"epoch": 0.2563372258615779,
|
| 106 |
+
"eval_loss": 0.7257962226867676,
|
| 107 |
+
"eval_runtime": 189.501,
|
| 108 |
+
"eval_samples_per_second": 37.055,
|
| 109 |
+
"eval_steps_per_second": 0.58,
|
| 110 |
+
"eval_wer": 0.5152473458323922,
|
| 111 |
"step": 1800
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 0.28481913984619767,
|
| 115 |
+
"grad_norm": 6.190296649932861,
|
| 116 |
+
"learning_rate": 0.0002668,
|
| 117 |
+
"loss": 0.8679,
|
| 118 |
"step": 2000
|
| 119 |
},
|
| 120 |
{
|
| 121 |
+
"epoch": 0.28481913984619767,
|
| 122 |
+
"eval_loss": 0.7084089517593384,
|
| 123 |
+
"eval_runtime": 188.267,
|
| 124 |
+
"eval_samples_per_second": 37.298,
|
| 125 |
+
"eval_steps_per_second": 0.584,
|
| 126 |
+
"eval_wer": 0.5216386080070158,
|
| 127 |
"step": 2000
|
| 128 |
},
|
| 129 |
{
|
| 130 |
+
"epoch": 0.3133010538308174,
|
| 131 |
+
"eval_loss": 0.6904259324073792,
|
| 132 |
+
"eval_runtime": 188.556,
|
| 133 |
+
"eval_samples_per_second": 37.241,
|
| 134 |
+
"eval_steps_per_second": 0.583,
|
| 135 |
+
"eval_wer": 0.5014151131426142,
|
| 136 |
"step": 2200
|
| 137 |
},
|
| 138 |
{
|
| 139 |
+
"epoch": 0.3417829678154372,
|
| 140 |
+
"eval_loss": 0.6992842555046082,
|
| 141 |
+
"eval_runtime": 189.0868,
|
| 142 |
+
"eval_samples_per_second": 37.136,
|
| 143 |
+
"eval_steps_per_second": 0.582,
|
| 144 |
+
"eval_wer": 0.5177586733812567,
|
| 145 |
+
"step": 2400
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.3560239248077471,
|
| 149 |
+
"grad_norm": 4.8257222175598145,
|
| 150 |
+
"learning_rate": 0.0002501333333333333,
|
| 151 |
+
"loss": 0.8577,
|
| 152 |
+
"step": 2500
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.37026488180005696,
|
| 156 |
+
"eval_loss": 0.6746060848236084,
|
| 157 |
+
"eval_runtime": 190.1492,
|
| 158 |
+
"eval_samples_per_second": 36.929,
|
| 159 |
+
"eval_steps_per_second": 0.578,
|
| 160 |
+
"eval_wer": 0.48673248382253287,
|
| 161 |
+
"step": 2600
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.39874679578467676,
|
| 165 |
+
"eval_loss": 0.6621994972229004,
|
| 166 |
+
"eval_runtime": 189.6459,
|
| 167 |
+
"eval_samples_per_second": 37.027,
|
| 168 |
+
"eval_steps_per_second": 0.58,
|
| 169 |
+
"eval_wer": 0.4962595835714001,
|
| 170 |
+
"step": 2800
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.4272287097692965,
|
| 174 |
+
"grad_norm": 3.6695899963378906,
|
| 175 |
+
"learning_rate": 0.00023346666666666666,
|
| 176 |
+
"loss": 0.7995,
|
| 177 |
+
"step": 3000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.4272287097692965,
|
| 181 |
+
"eval_loss": 0.6793097853660583,
|
| 182 |
+
"eval_runtime": 188.7722,
|
| 183 |
+
"eval_samples_per_second": 37.198,
|
| 184 |
+
"eval_steps_per_second": 0.583,
|
| 185 |
+
"eval_wer": 0.49348250707556574,
|
| 186 |
+
"step": 3000
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.45571062375391624,
|
| 190 |
+
"eval_loss": 0.6368467211723328,
|
| 191 |
+
"eval_runtime": 188.0679,
|
| 192 |
+
"eval_samples_per_second": 37.338,
|
| 193 |
+
"eval_steps_per_second": 0.585,
|
| 194 |
+
"eval_wer": 0.47005673740017806,
|
| 195 |
+
"step": 3200
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.48419253773853604,
|
| 199 |
+
"eval_loss": 0.6363435387611389,
|
| 200 |
+
"eval_runtime": 188.2666,
|
| 201 |
+
"eval_samples_per_second": 37.298,
|
| 202 |
+
"eval_steps_per_second": 0.584,
|
| 203 |
+
"eval_wer": 0.478055780703969,
|
| 204 |
+
"step": 3400
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.4984334947308459,
|
| 208 |
+
"grad_norm": 3.4502739906311035,
|
| 209 |
+
"learning_rate": 0.0002168333333333333,
|
| 210 |
+
"loss": 0.8141,
|
| 211 |
+
"step": 3500
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 0.5126744517231558,
|
| 215 |
+
"eval_loss": 0.6217373609542847,
|
| 216 |
+
"eval_runtime": 187.6755,
|
| 217 |
+
"eval_samples_per_second": 37.416,
|
| 218 |
+
"eval_steps_per_second": 0.586,
|
| 219 |
+
"eval_wer": 0.46555229274904,
|
| 220 |
+
"step": 3600
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.5411563657077756,
|
| 224 |
+
"eval_loss": 0.641762912273407,
|
| 225 |
+
"eval_runtime": 186.9231,
|
| 226 |
+
"eval_samples_per_second": 37.566,
|
| 227 |
+
"eval_steps_per_second": 0.588,
|
| 228 |
+
"eval_wer": 0.4940140049695053,
|
| 229 |
+
"step": 3800
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.5696382796923953,
|
| 233 |
+
"grad_norm": 5.877405643463135,
|
| 234 |
+
"learning_rate": 0.00020016666666666666,
|
| 235 |
+
"loss": 0.7953,
|
| 236 |
+
"step": 4000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.5696382796923953,
|
| 240 |
+
"eval_loss": 0.6017736196517944,
|
| 241 |
+
"eval_runtime": 182.787,
|
| 242 |
+
"eval_samples_per_second": 38.416,
|
| 243 |
+
"eval_steps_per_second": 0.602,
|
| 244 |
+
"eval_wer": 0.4542313876081266,
|
| 245 |
+
"step": 4000
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.5981201936770151,
|
| 249 |
+
"eval_loss": 0.5962206721305847,
|
| 250 |
+
"eval_runtime": 183.0007,
|
| 251 |
+
"eval_samples_per_second": 38.371,
|
| 252 |
+
"eval_steps_per_second": 0.601,
|
| 253 |
+
"eval_wer": 0.4580315975497947,
|
| 254 |
+
"step": 4200
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.6266021076616348,
|
| 258 |
+
"eval_loss": 0.5883399844169617,
|
| 259 |
+
"eval_runtime": 182.7298,
|
| 260 |
+
"eval_samples_per_second": 38.428,
|
| 261 |
+
"eval_steps_per_second": 0.602,
|
| 262 |
+
"eval_wer": 0.44590015812062345,
|
| 263 |
+
"step": 4400
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.6408430646539447,
|
| 267 |
+
"grad_norm": 3.615546226501465,
|
| 268 |
+
"learning_rate": 0.0001835333333333333,
|
| 269 |
+
"loss": 0.7596,
|
| 270 |
+
"step": 4500
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.6550840216462547,
|
| 274 |
+
"eval_loss": 0.578825056552887,
|
| 275 |
+
"eval_runtime": 183.3674,
|
| 276 |
+
"eval_samples_per_second": 38.295,
|
| 277 |
+
"eval_steps_per_second": 0.6,
|
| 278 |
+
"eval_wer": 0.43253298608804264,
|
| 279 |
+
"step": 4600
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.6835659356308744,
|
| 283 |
+
"eval_loss": 0.5708740949630737,
|
| 284 |
+
"eval_runtime": 182.6951,
|
| 285 |
+
"eval_samples_per_second": 38.436,
|
| 286 |
+
"eval_steps_per_second": 0.602,
|
| 287 |
+
"eval_wer": 0.4412362641013035,
|
| 288 |
+
"step": 4800
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.7120478496154942,
|
| 292 |
+
"grad_norm": 4.345168590545654,
|
| 293 |
+
"learning_rate": 0.0001669,
|
| 294 |
+
"loss": 0.7533,
|
| 295 |
+
"step": 5000
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 0.7120478496154942,
|
| 299 |
+
"eval_loss": 0.5594890117645264,
|
| 300 |
+
"eval_runtime": 182.5857,
|
| 301 |
+
"eval_samples_per_second": 38.459,
|
| 302 |
+
"eval_steps_per_second": 0.602,
|
| 303 |
+
"eval_wer": 0.4352170504524376,
|
| 304 |
+
"step": 5000
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.7405297636001139,
|
| 308 |
+
"eval_loss": 0.5545539259910583,
|
| 309 |
+
"eval_runtime": 182.2233,
|
| 310 |
+
"eval_samples_per_second": 38.535,
|
| 311 |
+
"eval_steps_per_second": 0.604,
|
| 312 |
+
"eval_wer": 0.4231786231547057,
|
| 313 |
+
"step": 5200
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.7690116775847337,
|
| 317 |
+
"eval_loss": 0.5545418858528137,
|
| 318 |
+
"eval_runtime": 182.2691,
|
| 319 |
+
"eval_samples_per_second": 38.525,
|
| 320 |
+
"eval_steps_per_second": 0.604,
|
| 321 |
+
"eval_wer": 0.4244276432054638,
|
| 322 |
+
"step": 5400
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 0.7832526345770435,
|
| 326 |
+
"grad_norm": 9.471431732177734,
|
| 327 |
+
"learning_rate": 0.00015026666666666667,
|
| 328 |
+
"loss": 0.7591,
|
| 329 |
+
"step": 5500
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.7974935915693535,
|
| 333 |
+
"eval_loss": 0.5442594885826111,
|
| 334 |
+
"eval_runtime": 182.3947,
|
| 335 |
+
"eval_samples_per_second": 38.499,
|
| 336 |
+
"eval_steps_per_second": 0.603,
|
| 337 |
+
"eval_wer": 0.4076455972043211,
|
| 338 |
+
"step": 5600
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.8259755055539733,
|
| 342 |
+
"eval_loss": 0.5341240763664246,
|
| 343 |
+
"eval_runtime": 182.0603,
|
| 344 |
+
"eval_samples_per_second": 38.57,
|
| 345 |
+
"eval_steps_per_second": 0.604,
|
| 346 |
+
"eval_wer": 0.41462150706227824,
|
| 347 |
+
"step": 5800
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 0.854457419538593,
|
| 351 |
+
"grad_norm": 4.406210422515869,
|
| 352 |
+
"learning_rate": 0.00013363333333333332,
|
| 353 |
+
"loss": 0.6621,
|
| 354 |
+
"step": 6000
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.854457419538593,
|
| 358 |
+
"eval_loss": 0.5104002952575684,
|
| 359 |
+
"eval_runtime": 181.8706,
|
| 360 |
+
"eval_samples_per_second": 38.61,
|
| 361 |
+
"eval_steps_per_second": 0.605,
|
| 362 |
+
"eval_wer": 0.3955141577751498,
|
| 363 |
+
"step": 6000
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.8829393335232127,
|
| 367 |
+
"eval_loss": 0.5139421820640564,
|
| 368 |
+
"eval_runtime": 181.902,
|
| 369 |
+
"eval_samples_per_second": 38.603,
|
| 370 |
+
"eval_steps_per_second": 0.605,
|
| 371 |
+
"eval_wer": 0.40112146055621256,
|
| 372 |
+
"step": 6200
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 0.9114212475078325,
|
| 376 |
+
"eval_loss": 0.5044221878051758,
|
| 377 |
+
"eval_runtime": 181.9538,
|
| 378 |
+
"eval_samples_per_second": 38.592,
|
| 379 |
+
"eval_steps_per_second": 0.605,
|
| 380 |
+
"eval_wer": 0.38039304269256835,
|
| 381 |
+
"step": 6400
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.9256622045001424,
|
| 385 |
+
"grad_norm": 8.09687328338623,
|
| 386 |
+
"learning_rate": 0.000117,
|
| 387 |
+
"loss": 0.6705,
|
| 388 |
+
"step": 6500
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.9399031614924523,
|
| 392 |
+
"eval_loss": 0.49985769391059875,
|
| 393 |
+
"eval_runtime": 182.1414,
|
| 394 |
+
"eval_samples_per_second": 38.552,
|
| 395 |
+
"eval_steps_per_second": 0.604,
|
| 396 |
+
"eval_wer": 0.3896012437050718,
|
| 397 |
+
"step": 6600
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 0.9683850754770721,
|
| 401 |
+
"eval_loss": 0.5097447037696838,
|
| 402 |
+
"eval_runtime": 181.5418,
|
| 403 |
+
"eval_samples_per_second": 38.68,
|
| 404 |
+
"eval_steps_per_second": 0.606,
|
| 405 |
+
"eval_wer": 0.4052804315762899,
|
| 406 |
+
"step": 6800
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 0.9968669894616918,
|
| 410 |
+
"grad_norm": 4.639442443847656,
|
| 411 |
+
"learning_rate": 0.00010033333333333332,
|
| 412 |
+
"loss": 0.6665,
|
| 413 |
+
"step": 7000
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.9968669894616918,
|
| 417 |
+
"eval_loss": 0.49253013730049133,
|
| 418 |
+
"eval_runtime": 181.6405,
|
| 419 |
+
"eval_samples_per_second": 38.659,
|
| 420 |
+
"eval_steps_per_second": 0.606,
|
| 421 |
+
"eval_wer": 0.3784796502743858,
|
| 422 |
+
"step": 7000
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.0253489034463117,
|
| 426 |
+
"eval_loss": 0.4896470010280609,
|
| 427 |
+
"eval_runtime": 181.3934,
|
| 428 |
+
"eval_samples_per_second": 38.711,
|
| 429 |
+
"eval_steps_per_second": 0.606,
|
| 430 |
+
"eval_wer": 0.3688728258414276,
|
| 431 |
+
"step": 7200
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.0538308174309314,
|
| 435 |
+
"eval_loss": 0.47494611144065857,
|
| 436 |
+
"eval_runtime": 181.7386,
|
| 437 |
+
"eval_samples_per_second": 38.638,
|
| 438 |
+
"eval_steps_per_second": 0.605,
|
| 439 |
+
"eval_wer": 0.3687399513679427,
|
| 440 |
+
"step": 7400
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.0680717744232413,
|
| 444 |
+
"grad_norm": 0.6623511910438538,
|
| 445 |
+
"learning_rate": 8.366666666666666e-05,
|
| 446 |
+
"loss": 0.5826,
|
| 447 |
+
"step": 7500
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 1.0823127314155512,
|
| 451 |
+
"eval_loss": 0.4684299826622009,
|
| 452 |
+
"eval_runtime": 182.4026,
|
| 453 |
+
"eval_samples_per_second": 38.497,
|
| 454 |
+
"eval_steps_per_second": 0.603,
|
| 455 |
+
"eval_wer": 0.3628004624031677,
|
| 456 |
+
"step": 7600
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 1.110794645400171,
|
| 460 |
+
"eval_loss": 0.47290024161338806,
|
| 461 |
+
"eval_runtime": 182.1043,
|
| 462 |
+
"eval_samples_per_second": 38.56,
|
| 463 |
+
"eval_steps_per_second": 0.604,
|
| 464 |
+
"eval_wer": 0.358495329462257,
|
| 465 |
+
"step": 7800
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.1392765593847907,
|
| 469 |
+
"grad_norm": 2.393817186355591,
|
| 470 |
+
"learning_rate": 6.699999999999999e-05,
|
| 471 |
+
"loss": 0.5836,
|
| 472 |
+
"step": 8000
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.1392765593847907,
|
| 476 |
+
"eval_loss": 0.46409761905670166,
|
| 477 |
+
"eval_runtime": 181.7327,
|
| 478 |
+
"eval_samples_per_second": 38.639,
|
| 479 |
+
"eval_steps_per_second": 0.605,
|
| 480 |
+
"eval_wer": 0.3553196295459679,
|
| 481 |
+
"step": 8000
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 1.1677584733694104,
|
| 485 |
+
"eval_loss": 0.45749881863594055,
|
| 486 |
+
"eval_runtime": 181.5866,
|
| 487 |
+
"eval_samples_per_second": 38.67,
|
| 488 |
+
"eval_steps_per_second": 0.606,
|
| 489 |
+
"eval_wer": 0.3529810388126337,
|
| 490 |
+
"step": 8200
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 1.1962403873540302,
|
| 494 |
+
"eval_loss": 0.45851147174835205,
|
| 495 |
+
"eval_runtime": 181.5801,
|
| 496 |
+
"eval_samples_per_second": 38.672,
|
| 497 |
+
"eval_steps_per_second": 0.606,
|
| 498 |
+
"eval_wer": 0.3485563188455866,
|
| 499 |
+
"step": 8400
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.21048134434634,
|
| 503 |
+
"grad_norm": 1.9676859378814697,
|
| 504 |
+
"learning_rate": 5.033333333333333e-05,
|
| 505 |
+
"loss": 0.5199,
|
| 506 |
+
"step": 8500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.22472230133865,
|
| 510 |
+
"eval_loss": 0.4548875391483307,
|
| 511 |
+
"eval_runtime": 182.6274,
|
| 512 |
+
"eval_samples_per_second": 38.45,
|
| 513 |
+
"eval_steps_per_second": 0.602,
|
| 514 |
+
"eval_wer": 0.3450750076402822,
|
| 515 |
+
"step": 8600
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.2532042153232696,
|
| 519 |
+
"eval_loss": 0.4520675539970398,
|
| 520 |
+
"eval_runtime": 182.8881,
|
| 521 |
+
"eval_samples_per_second": 38.395,
|
| 522 |
+
"eval_steps_per_second": 0.601,
|
| 523 |
+
"eval_wer": 0.34082302448876545,
|
| 524 |
+
"step": 8800
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 1.2816861293078894,
|
| 528 |
+
"grad_norm": 1.1400251388549805,
|
| 529 |
+
"learning_rate": 3.373333333333333e-05,
|
| 530 |
+
"loss": 0.5268,
|
| 531 |
+
"step": 9000
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 1.2816861293078894,
|
| 535 |
+
"eval_loss": 0.44252264499664307,
|
| 536 |
+
"eval_runtime": 182.3349,
|
| 537 |
+
"eval_samples_per_second": 38.512,
|
| 538 |
+
"eval_steps_per_second": 0.603,
|
| 539 |
+
"eval_wer": 0.33950756720126496,
|
| 540 |
+
"step": 9000
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 1.3101680432925091,
|
| 544 |
+
"eval_loss": 0.44072064757347107,
|
| 545 |
+
"eval_runtime": 184.1579,
|
| 546 |
+
"eval_samples_per_second": 38.13,
|
| 547 |
+
"eval_steps_per_second": 0.597,
|
| 548 |
+
"eval_wer": 0.3361857053641425,
|
| 549 |
+
"step": 9200
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.338649957277129,
|
| 553 |
+
"eval_loss": 0.4383063018321991,
|
| 554 |
+
"eval_runtime": 181.6966,
|
| 555 |
+
"eval_samples_per_second": 38.647,
|
| 556 |
+
"eval_steps_per_second": 0.605,
|
| 557 |
+
"eval_wer": 0.33397998910429316,
|
| 558 |
+
"step": 9400
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 1.352890914269439,
|
| 562 |
+
"grad_norm": 1.0755033493041992,
|
| 563 |
+
"learning_rate": 1.71e-05,
|
| 564 |
+
"loss": 0.5013,
|
| 565 |
+
"step": 9500
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 1.3671318712617488,
|
| 569 |
+
"eval_loss": 0.4356846809387207,
|
| 570 |
+
"eval_runtime": 183.1225,
|
| 571 |
+
"eval_samples_per_second": 38.346,
|
| 572 |
+
"eval_steps_per_second": 0.601,
|
| 573 |
+
"eval_wer": 0.33253165734330775,
|
| 574 |
+
"step": 9600
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 1.3956137852463686,
|
| 578 |
+
"eval_loss": 0.43495818972587585,
|
| 579 |
+
"eval_runtime": 182.2639,
|
| 580 |
+
"eval_samples_per_second": 38.527,
|
| 581 |
+
"eval_steps_per_second": 0.604,
|
| 582 |
+
"eval_wer": 0.3316812607130044,
|
| 583 |
+
"step": 9800
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.4240956992309883,
|
| 587 |
+
"grad_norm": 1.6312005519866943,
|
| 588 |
+
"learning_rate": 4.666666666666666e-07,
|
| 589 |
+
"loss": 0.5095,
|
| 590 |
+
"step": 10000
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.4240956992309883,
|
| 594 |
+
"eval_loss": 0.43451622128486633,
|
| 595 |
+
"eval_runtime": 182.2078,
|
| 596 |
+
"eval_samples_per_second": 38.538,
|
| 597 |
+
"eval_steps_per_second": 0.604,
|
| 598 |
+
"eval_wer": 0.3308175766353526,
|
| 599 |
+
"step": 10000
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.4240956992309883,
|
| 603 |
+
"step": 10000,
|
| 604 |
+
"total_flos": 4.5974516642218747e+18,
|
| 605 |
+
"train_loss": 0.7817989181518554,
|
| 606 |
+
"train_runtime": 11412.7197,
|
| 607 |
+
"train_samples_per_second": 3.505,
|
| 608 |
+
"train_steps_per_second": 0.876
|
| 609 |
}
|
| 610 |
],
|
| 611 |
"logging_steps": 500,
|
| 612 |
+
"max_steps": 10000,
|
| 613 |
"num_input_tokens_seen": 0,
|
| 614 |
"num_train_epochs": 2,
|
| 615 |
"save_steps": 500,
|
|
|
|
| 625 |
"attributes": {}
|
| 626 |
}
|
| 627 |
},
|
| 628 |
+
"total_flos": 4.5974516642218747e+18,
|
| 629 |
+
"train_batch_size": 4,
|
| 630 |
"trial_name": null,
|
| 631 |
"trial_params": null
|
| 632 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5240
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:acb452a91c703d1a26455b8858033cfe497708ca59e32a5e19ba4ea803c125e5
|
| 3 |
size 5240
|