End of training
Browse files- README.md +12 -9
- all_results.json +12 -0
- eval_results.json +8 -0
- runs/Jan19_17-33-56_knight/events.out.tfevents.1705759447.knight.5763.1 +3 -0
- train_results.json +7 -0
- trainer_state.json +600 -0
README.md
CHANGED
|
@@ -1,39 +1,42 @@
|
|
| 1 |
---
|
|
|
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: openai/whisper-base
|
| 4 |
tags:
|
|
|
|
| 5 |
- generated_from_trainer
|
| 6 |
datasets:
|
| 7 |
-
- common_voice_16_0
|
| 8 |
metrics:
|
| 9 |
- wer
|
| 10 |
model-index:
|
| 11 |
-
- name:
|
| 12 |
results:
|
| 13 |
- task:
|
| 14 |
name: Automatic Speech Recognition
|
| 15 |
type: automatic-speech-recognition
|
| 16 |
dataset:
|
| 17 |
-
name: common_voice_16_0
|
| 18 |
-
type: common_voice_16_0
|
| 19 |
config: id
|
| 20 |
split: test
|
| 21 |
args: id
|
| 22 |
metrics:
|
| 23 |
- name: Wer
|
| 24 |
type: wer
|
| 25 |
-
value:
|
| 26 |
---
|
| 27 |
|
| 28 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 29 |
should probably proofread and complete it, then remove this comment. -->
|
| 30 |
|
| 31 |
-
#
|
| 32 |
|
| 33 |
-
This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the common_voice_16_0 dataset.
|
| 34 |
It achieves the following results on the evaluation set:
|
| 35 |
-
- Loss: 0.
|
| 36 |
-
- Wer:
|
| 37 |
|
| 38 |
## Model description
|
| 39 |
|
|
|
|
| 1 |
---
|
| 2 |
+
language:
|
| 3 |
+
- id
|
| 4 |
license: apache-2.0
|
| 5 |
base_model: openai/whisper-base
|
| 6 |
tags:
|
| 7 |
+
- whisper-event
|
| 8 |
- generated_from_trainer
|
| 9 |
datasets:
|
| 10 |
+
- mozilla-foundation/common_voice_16_0
|
| 11 |
metrics:
|
| 12 |
- wer
|
| 13 |
model-index:
|
| 14 |
+
- name: Breeze DSW Indonesian - base
|
| 15 |
results:
|
| 16 |
- task:
|
| 17 |
name: Automatic Speech Recognition
|
| 18 |
type: automatic-speech-recognition
|
| 19 |
dataset:
|
| 20 |
+
name: mozilla-foundation/common_voice_16_0 id
|
| 21 |
+
type: mozilla-foundation/common_voice_16_0
|
| 22 |
config: id
|
| 23 |
split: test
|
| 24 |
args: id
|
| 25 |
metrics:
|
| 26 |
- name: Wer
|
| 27 |
type: wer
|
| 28 |
+
value: 32.82730701996504
|
| 29 |
---
|
| 30 |
|
| 31 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 32 |
should probably proofread and complete it, then remove this comment. -->
|
| 33 |
|
| 34 |
+
# Breeze DSW Indonesian - base
|
| 35 |
|
| 36 |
+
This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the mozilla-foundation/common_voice_16_0 id dataset.
|
| 37 |
It achieves the following results on the evaluation set:
|
| 38 |
+
- Loss: 0.5347
|
| 39 |
+
- Wer: 32.8273
|
| 40 |
|
| 41 |
## Model description
|
| 42 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 12.06,
|
| 3 |
+
"eval_loss": 0.53466796875,
|
| 4 |
+
"eval_runtime": 2318.8385,
|
| 5 |
+
"eval_samples_per_second": 1.571,
|
| 6 |
+
"eval_steps_per_second": 0.098,
|
| 7 |
+
"eval_wer": 32.82730701996504,
|
| 8 |
+
"train_loss": 0.16917743301391602,
|
| 9 |
+
"train_runtime": 75059.9366,
|
| 10 |
+
"train_samples_per_second": 0.853,
|
| 11 |
+
"train_steps_per_second": 0.027
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 12.06,
|
| 3 |
+
"eval_loss": 0.53466796875,
|
| 4 |
+
"eval_runtime": 2318.8385,
|
| 5 |
+
"eval_samples_per_second": 1.571,
|
| 6 |
+
"eval_steps_per_second": 0.098,
|
| 7 |
+
"eval_wer": 32.82730701996504
|
| 8 |
+
}
|
runs/Jan19_17-33-56_knight/events.out.tfevents.1705759447.knight.5763.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0fa609403e1d313636dfb63bec480f0225ba8e1f75c416acc672ce6f67386c12
|
| 3 |
+
size 406
|
train_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 12.06,
|
| 3 |
+
"train_loss": 0.16917743301391602,
|
| 4 |
+
"train_runtime": 75059.9366,
|
| 5 |
+
"train_samples_per_second": 0.853,
|
| 6 |
+
"train_steps_per_second": 0.027
|
| 7 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,600 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 32.82730701996504,
|
| 3 |
+
"best_model_checkpoint": "/cosmos/home/sp-operator/ai/training/models/huggingface/scripts/../breeze-listen-dsw-base-id/checkpoint-600",
|
| 4 |
+
"epoch": 12.064,
|
| 5 |
+
"eval_steps": 200,
|
| 6 |
+
"global_step": 2000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01,
|
| 13 |
+
"learning_rate": 4.898977360288234e-06,
|
| 14 |
+
"loss": 1.4272,
|
| 15 |
+
"step": 25
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.03,
|
| 19 |
+
"learning_rate": 6.160712527409633e-06,
|
| 20 |
+
"loss": 1.0004,
|
| 21 |
+
"step": 50
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.04,
|
| 25 |
+
"learning_rate": 6.85912902234906e-06,
|
| 26 |
+
"loss": 0.8231,
|
| 27 |
+
"step": 75
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.05,
|
| 31 |
+
"learning_rate": 7.344547104469332e-06,
|
| 32 |
+
"loss": 0.7054,
|
| 33 |
+
"step": 100
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.06,
|
| 37 |
+
"learning_rate": 7.716963756434345e-06,
|
| 38 |
+
"loss": 0.531,
|
| 39 |
+
"step": 125
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.07,
|
| 43 |
+
"learning_rate": 8.019180844200955e-06,
|
| 44 |
+
"loss": 0.4107,
|
| 45 |
+
"step": 150
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.01,
|
| 49 |
+
"learning_rate": 8.27351214279797e-06,
|
| 50 |
+
"loss": 0.546,
|
| 51 |
+
"step": 175
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 1.02,
|
| 55 |
+
"learning_rate": 8.49307723936858e-06,
|
| 56 |
+
"loss": 0.5452,
|
| 57 |
+
"step": 200
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 1.02,
|
| 61 |
+
"eval_loss": 0.54638671875,
|
| 62 |
+
"eval_runtime": 2392.0752,
|
| 63 |
+
"eval_samples_per_second": 1.523,
|
| 64 |
+
"eval_steps_per_second": 0.095,
|
| 65 |
+
"eval_wer": 35.168828779096515,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.03,
|
| 70 |
+
"learning_rate": 8.686247975778677e-06,
|
| 71 |
+
"loss": 0.5942,
|
| 72 |
+
"step": 225
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.05,
|
| 76 |
+
"learning_rate": 8.858694625217149e-06,
|
| 77 |
+
"loss": 0.4591,
|
| 78 |
+
"step": 250
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 1.06,
|
| 82 |
+
"learning_rate": 9.014436199608479e-06,
|
| 83 |
+
"loss": 0.3818,
|
| 84 |
+
"step": 275
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.07,
|
| 88 |
+
"learning_rate": 9.156425255148058e-06,
|
| 89 |
+
"loss": 0.2725,
|
| 90 |
+
"step": 300
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 2.01,
|
| 94 |
+
"learning_rate": 9.28689473531776e-06,
|
| 95 |
+
"loss": 0.3299,
|
| 96 |
+
"step": 325
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 2.02,
|
| 100 |
+
"learning_rate": 9.407574351377137e-06,
|
| 101 |
+
"loss": 0.3841,
|
| 102 |
+
"step": 350
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.03,
|
| 106 |
+
"learning_rate": 9.519831289296397e-06,
|
| 107 |
+
"loss": 0.4134,
|
| 108 |
+
"step": 375
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 2.04,
|
| 112 |
+
"learning_rate": 9.624764935335318e-06,
|
| 113 |
+
"loss": 0.3445,
|
| 114 |
+
"step": 400
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 2.04,
|
| 118 |
+
"eval_loss": 0.54052734375,
|
| 119 |
+
"eval_runtime": 2421.7938,
|
| 120 |
+
"eval_samples_per_second": 1.504,
|
| 121 |
+
"eval_steps_per_second": 0.094,
|
| 122 |
+
"eval_wer": 34.06937160732358,
|
| 123 |
+
"step": 400
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.06,
|
| 127 |
+
"learning_rate": 9.723272550712454e-06,
|
| 128 |
+
"loss": 0.2894,
|
| 129 |
+
"step": 425
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 2.07,
|
| 133 |
+
"learning_rate": 9.816095971633122e-06,
|
| 134 |
+
"loss": 0.199,
|
| 135 |
+
"step": 450
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 3.0,
|
| 139 |
+
"learning_rate": 9.90385555539545e-06,
|
| 140 |
+
"loss": 0.1939,
|
| 141 |
+
"step": 475
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 3.02,
|
| 145 |
+
"learning_rate": 9.987075336738768e-06,
|
| 146 |
+
"loss": 0.2661,
|
| 147 |
+
"step": 500
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 3.03,
|
| 151 |
+
"learning_rate": 9.866666666666668e-06,
|
| 152 |
+
"loss": 0.2885,
|
| 153 |
+
"step": 525
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 3.04,
|
| 157 |
+
"learning_rate": 9.7e-06,
|
| 158 |
+
"loss": 0.2474,
|
| 159 |
+
"step": 550
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 3.05,
|
| 163 |
+
"learning_rate": 9.533333333333334e-06,
|
| 164 |
+
"loss": 0.195,
|
| 165 |
+
"step": 575
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 3.07,
|
| 169 |
+
"learning_rate": 9.366666666666668e-06,
|
| 170 |
+
"loss": 0.1397,
|
| 171 |
+
"step": 600
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 3.07,
|
| 175 |
+
"eval_loss": 0.53466796875,
|
| 176 |
+
"eval_runtime": 2319.7619,
|
| 177 |
+
"eval_samples_per_second": 1.57,
|
| 178 |
+
"eval_steps_per_second": 0.098,
|
| 179 |
+
"eval_wer": 32.82730701996504,
|
| 180 |
+
"step": 600
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 4.0,
|
| 184 |
+
"learning_rate": 9.200000000000002e-06,
|
| 185 |
+
"loss": 0.1158,
|
| 186 |
+
"step": 625
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 4.01,
|
| 190 |
+
"learning_rate": 9.033333333333334e-06,
|
| 191 |
+
"loss": 0.1686,
|
| 192 |
+
"step": 650
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 4.03,
|
| 196 |
+
"learning_rate": 8.866666666666668e-06,
|
| 197 |
+
"loss": 0.183,
|
| 198 |
+
"step": 675
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 4.04,
|
| 202 |
+
"learning_rate": 8.700000000000001e-06,
|
| 203 |
+
"loss": 0.1702,
|
| 204 |
+
"step": 700
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 4.05,
|
| 208 |
+
"learning_rate": 8.533333333333335e-06,
|
| 209 |
+
"loss": 0.138,
|
| 210 |
+
"step": 725
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 4.06,
|
| 214 |
+
"learning_rate": 8.366666666666667e-06,
|
| 215 |
+
"loss": 0.1002,
|
| 216 |
+
"step": 750
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 4.08,
|
| 220 |
+
"learning_rate": 8.2e-06,
|
| 221 |
+
"loss": 0.0697,
|
| 222 |
+
"step": 775
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 5.01,
|
| 226 |
+
"learning_rate": 8.033333333333335e-06,
|
| 227 |
+
"loss": 0.0988,
|
| 228 |
+
"step": 800
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 5.01,
|
| 232 |
+
"eval_loss": 0.5654296875,
|
| 233 |
+
"eval_runtime": 2396.345,
|
| 234 |
+
"eval_samples_per_second": 1.52,
|
| 235 |
+
"eval_steps_per_second": 0.095,
|
| 236 |
+
"eval_wer": 35.67485509246481,
|
| 237 |
+
"step": 800
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 5.02,
|
| 241 |
+
"learning_rate": 7.866666666666667e-06,
|
| 242 |
+
"loss": 0.1161,
|
| 243 |
+
"step": 825
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 5.04,
|
| 247 |
+
"learning_rate": 7.7e-06,
|
| 248 |
+
"loss": 0.1152,
|
| 249 |
+
"step": 850
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 5.05,
|
| 253 |
+
"learning_rate": 7.533333333333334e-06,
|
| 254 |
+
"loss": 0.0922,
|
| 255 |
+
"step": 875
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 5.06,
|
| 259 |
+
"learning_rate": 7.3666666666666676e-06,
|
| 260 |
+
"loss": 0.073,
|
| 261 |
+
"step": 900
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 5.07,
|
| 265 |
+
"learning_rate": 7.2000000000000005e-06,
|
| 266 |
+
"loss": 0.0481,
|
| 267 |
+
"step": 925
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 6.01,
|
| 271 |
+
"learning_rate": 7.033333333333334e-06,
|
| 272 |
+
"loss": 0.056,
|
| 273 |
+
"step": 950
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 6.02,
|
| 277 |
+
"learning_rate": 6.866666666666667e-06,
|
| 278 |
+
"loss": 0.0705,
|
| 279 |
+
"step": 975
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 6.03,
|
| 283 |
+
"learning_rate": 6.700000000000001e-06,
|
| 284 |
+
"loss": 0.077,
|
| 285 |
+
"step": 1000
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 6.03,
|
| 289 |
+
"eval_loss": 0.57861328125,
|
| 290 |
+
"eval_runtime": 2311.9939,
|
| 291 |
+
"eval_samples_per_second": 1.575,
|
| 292 |
+
"eval_steps_per_second": 0.099,
|
| 293 |
+
"eval_wer": 33.945165148587726,
|
| 294 |
+
"step": 1000
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 6.04,
|
| 298 |
+
"learning_rate": 6.533333333333334e-06,
|
| 299 |
+
"loss": 0.0629,
|
| 300 |
+
"step": 1025
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 6.06,
|
| 304 |
+
"learning_rate": 6.366666666666668e-06,
|
| 305 |
+
"loss": 0.0496,
|
| 306 |
+
"step": 1050
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 6.07,
|
| 310 |
+
"learning_rate": 6.200000000000001e-06,
|
| 311 |
+
"loss": 0.0328,
|
| 312 |
+
"step": 1075
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 7.0,
|
| 316 |
+
"learning_rate": 6.033333333333335e-06,
|
| 317 |
+
"loss": 0.0331,
|
| 318 |
+
"step": 1100
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 7.02,
|
| 322 |
+
"learning_rate": 5.8666666666666675e-06,
|
| 323 |
+
"loss": 0.0432,
|
| 324 |
+
"step": 1125
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 7.03,
|
| 328 |
+
"learning_rate": 5.7e-06,
|
| 329 |
+
"loss": 0.0475,
|
| 330 |
+
"step": 1150
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 7.04,
|
| 334 |
+
"learning_rate": 5.533333333333334e-06,
|
| 335 |
+
"loss": 0.0446,
|
| 336 |
+
"step": 1175
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 7.05,
|
| 340 |
+
"learning_rate": 5.366666666666666e-06,
|
| 341 |
+
"loss": 0.0338,
|
| 342 |
+
"step": 1200
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 7.05,
|
| 346 |
+
"eval_loss": 0.60498046875,
|
| 347 |
+
"eval_runtime": 2281.4213,
|
| 348 |
+
"eval_samples_per_second": 1.596,
|
| 349 |
+
"eval_steps_per_second": 0.1,
|
| 350 |
+
"eval_wer": 33.98196706228724,
|
| 351 |
+
"step": 1200
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 7.07,
|
| 355 |
+
"learning_rate": 5.2e-06,
|
| 356 |
+
"loss": 0.0242,
|
| 357 |
+
"step": 1225
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 8.0,
|
| 361 |
+
"learning_rate": 5.033333333333333e-06,
|
| 362 |
+
"loss": 0.0197,
|
| 363 |
+
"step": 1250
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 8.01,
|
| 367 |
+
"learning_rate": 4.866666666666667e-06,
|
| 368 |
+
"loss": 0.0269,
|
| 369 |
+
"step": 1275
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 8.03,
|
| 373 |
+
"learning_rate": 4.7e-06,
|
| 374 |
+
"loss": 0.0319,
|
| 375 |
+
"step": 1300
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 8.04,
|
| 379 |
+
"learning_rate": 4.533333333333334e-06,
|
| 380 |
+
"loss": 0.0293,
|
| 381 |
+
"step": 1325
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 8.05,
|
| 385 |
+
"learning_rate": 4.366666666666667e-06,
|
| 386 |
+
"loss": 0.023,
|
| 387 |
+
"step": 1350
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 8.06,
|
| 391 |
+
"learning_rate": 4.2000000000000004e-06,
|
| 392 |
+
"loss": 0.0189,
|
| 393 |
+
"step": 1375
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 8.08,
|
| 397 |
+
"learning_rate": 4.033333333333333e-06,
|
| 398 |
+
"loss": 0.0137,
|
| 399 |
+
"step": 1400
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 8.08,
|
| 403 |
+
"eval_loss": 0.6220703125,
|
| 404 |
+
"eval_runtime": 2300.7263,
|
| 405 |
+
"eval_samples_per_second": 1.583,
|
| 406 |
+
"eval_steps_per_second": 0.099,
|
| 407 |
+
"eval_wer": 34.10157328181065,
|
| 408 |
+
"step": 1400
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 9.01,
|
| 412 |
+
"learning_rate": 3.866666666666667e-06,
|
| 413 |
+
"loss": 0.0189,
|
| 414 |
+
"step": 1425
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 9.02,
|
| 418 |
+
"learning_rate": 3.7e-06,
|
| 419 |
+
"loss": 0.02,
|
| 420 |
+
"step": 1450
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 9.04,
|
| 424 |
+
"learning_rate": 3.5333333333333335e-06,
|
| 425 |
+
"loss": 0.0213,
|
| 426 |
+
"step": 1475
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 9.05,
|
| 430 |
+
"learning_rate": 3.366666666666667e-06,
|
| 431 |
+
"loss": 0.0179,
|
| 432 |
+
"step": 1500
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 9.06,
|
| 436 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 437 |
+
"loss": 0.0139,
|
| 438 |
+
"step": 1525
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 9.07,
|
| 442 |
+
"learning_rate": 3.0333333333333337e-06,
|
| 443 |
+
"loss": 0.0111,
|
| 444 |
+
"step": 1550
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 10.01,
|
| 448 |
+
"learning_rate": 2.866666666666667e-06,
|
| 449 |
+
"loss": 0.0129,
|
| 450 |
+
"step": 1575
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 10.02,
|
| 454 |
+
"learning_rate": 2.7000000000000004e-06,
|
| 455 |
+
"loss": 0.0153,
|
| 456 |
+
"step": 1600
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 10.02,
|
| 460 |
+
"eval_loss": 0.64306640625,
|
| 461 |
+
"eval_runtime": 2254.9316,
|
| 462 |
+
"eval_samples_per_second": 1.615,
|
| 463 |
+
"eval_steps_per_second": 0.101,
|
| 464 |
+
"eval_wer": 33.90376299567578,
|
| 465 |
+
"step": 1600
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 10.03,
|
| 469 |
+
"learning_rate": 2.5333333333333338e-06,
|
| 470 |
+
"loss": 0.0166,
|
| 471 |
+
"step": 1625
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 10.04,
|
| 475 |
+
"learning_rate": 2.3666666666666667e-06,
|
| 476 |
+
"loss": 0.0142,
|
| 477 |
+
"step": 1650
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 10.06,
|
| 481 |
+
"learning_rate": 2.2e-06,
|
| 482 |
+
"loss": 0.0119,
|
| 483 |
+
"step": 1675
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 10.07,
|
| 487 |
+
"learning_rate": 2.0333333333333335e-06,
|
| 488 |
+
"loss": 0.0101,
|
| 489 |
+
"step": 1700
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 11.0,
|
| 493 |
+
"learning_rate": 1.8666666666666669e-06,
|
| 494 |
+
"loss": 0.0101,
|
| 495 |
+
"step": 1725
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 11.02,
|
| 499 |
+
"learning_rate": 1.7000000000000002e-06,
|
| 500 |
+
"loss": 0.0125,
|
| 501 |
+
"step": 1750
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 11.03,
|
| 505 |
+
"learning_rate": 1.5333333333333334e-06,
|
| 506 |
+
"loss": 0.0141,
|
| 507 |
+
"step": 1775
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 11.04,
|
| 511 |
+
"learning_rate": 1.3666666666666668e-06,
|
| 512 |
+
"loss": 0.0125,
|
| 513 |
+
"step": 1800
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 11.04,
|
| 517 |
+
"eval_loss": 0.6513671875,
|
| 518 |
+
"eval_runtime": 2241.116,
|
| 519 |
+
"eval_samples_per_second": 1.625,
|
| 520 |
+
"eval_steps_per_second": 0.102,
|
| 521 |
+
"eval_wer": 33.75195510166529,
|
| 522 |
+
"step": 1800
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 11.05,
|
| 526 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 527 |
+
"loss": 0.0114,
|
| 528 |
+
"step": 1825
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 11.07,
|
| 532 |
+
"learning_rate": 1.0333333333333333e-06,
|
| 533 |
+
"loss": 0.0093,
|
| 534 |
+
"step": 1850
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 12.0,
|
| 538 |
+
"learning_rate": 8.666666666666668e-07,
|
| 539 |
+
"loss": 0.0087,
|
| 540 |
+
"step": 1875
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 12.01,
|
| 544 |
+
"learning_rate": 7.000000000000001e-07,
|
| 545 |
+
"loss": 0.0111,
|
| 546 |
+
"step": 1900
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 12.03,
|
| 550 |
+
"learning_rate": 5.333333333333335e-07,
|
| 551 |
+
"loss": 0.0129,
|
| 552 |
+
"step": 1925
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 12.04,
|
| 556 |
+
"learning_rate": 3.666666666666667e-07,
|
| 557 |
+
"loss": 0.0118,
|
| 558 |
+
"step": 1950
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 12.05,
|
| 562 |
+
"learning_rate": 2.0000000000000002e-07,
|
| 563 |
+
"loss": 0.011,
|
| 564 |
+
"step": 1975
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 12.06,
|
| 568 |
+
"learning_rate": 3.333333333333334e-08,
|
| 569 |
+
"loss": 0.0092,
|
| 570 |
+
"step": 2000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 12.06,
|
| 574 |
+
"eval_loss": 0.65283203125,
|
| 575 |
+
"eval_runtime": 2242.3057,
|
| 576 |
+
"eval_samples_per_second": 1.624,
|
| 577 |
+
"eval_steps_per_second": 0.102,
|
| 578 |
+
"eval_wer": 33.82555892906431,
|
| 579 |
+
"step": 2000
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 12.06,
|
| 583 |
+
"step": 2000,
|
| 584 |
+
"total_flos": 4.1331384945643356e+18,
|
| 585 |
+
"train_loss": 0.16917743301391602,
|
| 586 |
+
"train_runtime": 75059.9366,
|
| 587 |
+
"train_samples_per_second": 0.853,
|
| 588 |
+
"train_steps_per_second": 0.027
|
| 589 |
+
}
|
| 590 |
+
],
|
| 591 |
+
"logging_steps": 25,
|
| 592 |
+
"max_steps": 2000,
|
| 593 |
+
"num_input_tokens_seen": 0,
|
| 594 |
+
"num_train_epochs": 9223372036854775807,
|
| 595 |
+
"save_steps": 200,
|
| 596 |
+
"total_flos": 4.1331384945643356e+18,
|
| 597 |
+
"train_batch_size": 32,
|
| 598 |
+
"trial_name": null,
|
| 599 |
+
"trial_params": null
|
| 600 |
+
}
|