End of training
Browse files- README.md +19 -6
- all_results.json +14 -0
- eval_results.json +9 -0
- runs/Dec19_08-38-49_b72483eab5b9/events.out.tfevents.1734693748.b72483eab5b9.6985.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +572 -0
README.md
CHANGED
|
@@ -3,23 +3,36 @@ library_name: transformers
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: openai/whisper-large-v3
|
| 5 |
tags:
|
|
|
|
| 6 |
- generated_from_trainer
|
|
|
|
|
|
|
| 7 |
metrics:
|
| 8 |
- wer
|
| 9 |
model-index:
|
| 10 |
-
- name:
|
| 11 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
---
|
| 13 |
|
| 14 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 15 |
should probably proofread and complete it, then remove this comment. -->
|
| 16 |
|
| 17 |
-
#
|
| 18 |
|
| 19 |
-
This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
-
- Loss: 0.
|
| 22 |
-
- Wer: 5.
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: openai/whisper-large-v3
|
| 5 |
tags:
|
| 6 |
+
- whisper-event
|
| 7 |
- generated_from_trainer
|
| 8 |
+
datasets:
|
| 9 |
+
- OUTCOMESAI/medical_n_common_speech_corpus_50_50
|
| 10 |
metrics:
|
| 11 |
- wer
|
| 12 |
model-index:
|
| 13 |
+
- name: Whisper Large V3 Common n Medical 50 50
|
| 14 |
+
results:
|
| 15 |
+
- task:
|
| 16 |
+
name: Automatic Speech Recognition
|
| 17 |
+
type: automatic-speech-recognition
|
| 18 |
+
dataset:
|
| 19 |
+
name: OUTCOMESAI/medical_n_common_speech_corpus_50_50 en
|
| 20 |
+
type: OUTCOMESAI/medical_n_common_speech_corpus_50_50
|
| 21 |
+
metrics:
|
| 22 |
+
- name: Wer
|
| 23 |
+
type: wer
|
| 24 |
+
value: 5.218643517767322
|
| 25 |
---
|
| 26 |
|
| 27 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 28 |
should probably proofread and complete it, then remove this comment. -->
|
| 29 |
|
| 30 |
+
# Whisper Large V3 Common n Medical 50 50
|
| 31 |
|
| 32 |
+
This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the OUTCOMESAI/medical_n_common_speech_corpus_50_50 en dataset.
|
| 33 |
It achieves the following results on the evaluation set:
|
| 34 |
+
- Loss: 0.3196
|
| 35 |
+
- Wer: 5.2186
|
| 36 |
|
| 37 |
## Model description
|
| 38 |
|
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.937984496124031,
|
| 3 |
+
"eval_loss": 0.319580078125,
|
| 4 |
+
"eval_runtime": 2690.2896,
|
| 5 |
+
"eval_samples": 10000,
|
| 6 |
+
"eval_samples_per_second": 0.834,
|
| 7 |
+
"eval_steps_per_second": 0.026,
|
| 8 |
+
"eval_wer": 5.218643517767322,
|
| 9 |
+
"total_flos": 1.087083891827713e+21,
|
| 10 |
+
"train_loss": 3.86036796875,
|
| 11 |
+
"train_runtime": 85811.7084,
|
| 12 |
+
"train_samples_per_second": 3.729,
|
| 13 |
+
"train_steps_per_second": 0.058
|
| 14 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.937984496124031,
|
| 3 |
+
"eval_loss": 0.319580078125,
|
| 4 |
+
"eval_runtime": 2690.2896,
|
| 5 |
+
"eval_samples": 10000,
|
| 6 |
+
"eval_samples_per_second": 0.834,
|
| 7 |
+
"eval_steps_per_second": 0.026,
|
| 8 |
+
"eval_wer": 5.218643517767322
|
| 9 |
+
}
|
runs/Dec19_08-38-49_b72483eab5b9/events.out.tfevents.1734693748.b72483eab5b9.6985.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5ee9f2bb6aa719844f9fe09b3ecd4c383fc61bb4673c71d487d2586a0298248e
|
| 3 |
+
size 406
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.937984496124031,
|
| 3 |
+
"total_flos": 1.087083891827713e+21,
|
| 4 |
+
"train_loss": 3.86036796875,
|
| 5 |
+
"train_runtime": 85811.7084,
|
| 6 |
+
"train_samples_per_second": 3.729,
|
| 7 |
+
"train_steps_per_second": 0.058
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,572 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 5.218643517767322,
|
| 3 |
+
"best_model_checkpoint": "OUTCOMESAI/whisper-large-v3-common-n-medical-50-50/checkpoint-4750",
|
| 4 |
+
"epoch": 1.937984496124031,
|
| 5 |
+
"eval_steps": 250,
|
| 6 |
+
"global_step": 5000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.03875968992248062,
|
| 13 |
+
"grad_norm": 4.327393531799316,
|
| 14 |
+
"learning_rate": 4.123797088618779e-07,
|
| 15 |
+
"loss": 6.6181,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.07751937984496124,
|
| 20 |
+
"grad_norm": 7.569775104522705,
|
| 21 |
+
"learning_rate": 4.775003968157492e-07,
|
| 22 |
+
"loss": 5.126,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09689922480620156,
|
| 27 |
+
"eval_loss": 0.369384765625,
|
| 28 |
+
"eval_runtime": 2703.3784,
|
| 29 |
+
"eval_samples_per_second": 0.83,
|
| 30 |
+
"eval_steps_per_second": 0.026,
|
| 31 |
+
"eval_wer": 5.660122234842339,
|
| 32 |
+
"step": 250
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.11627906976744186,
|
| 36 |
+
"grad_norm": 3.0063843727111816,
|
| 37 |
+
"learning_rate": 4.953684210526315e-07,
|
| 38 |
+
"loss": 4.8144,
|
| 39 |
+
"step": 300
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.15503875968992248,
|
| 43 |
+
"grad_norm": 3.107106924057007,
|
| 44 |
+
"learning_rate": 4.848421052631578e-07,
|
| 45 |
+
"loss": 4.5831,
|
| 46 |
+
"step": 400
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.1937984496124031,
|
| 50 |
+
"grad_norm": 2.855757236480713,
|
| 51 |
+
"learning_rate": 4.7431578947368417e-07,
|
| 52 |
+
"loss": 4.367,
|
| 53 |
+
"step": 500
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.1937984496124031,
|
| 57 |
+
"eval_loss": 0.358642578125,
|
| 58 |
+
"eval_runtime": 2810.867,
|
| 59 |
+
"eval_samples_per_second": 0.798,
|
| 60 |
+
"eval_steps_per_second": 0.025,
|
| 61 |
+
"eval_wer": 5.815620098436908,
|
| 62 |
+
"step": 500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.23255813953488372,
|
| 66 |
+
"grad_norm": 2.5796642303466797,
|
| 67 |
+
"learning_rate": 4.637894736842105e-07,
|
| 68 |
+
"loss": 4.2067,
|
| 69 |
+
"step": 600
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.2713178294573643,
|
| 73 |
+
"grad_norm": 2.537674903869629,
|
| 74 |
+
"learning_rate": 4.532631578947368e-07,
|
| 75 |
+
"loss": 4.1514,
|
| 76 |
+
"step": 700
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.29069767441860467,
|
| 80 |
+
"eval_loss": 0.35107421875,
|
| 81 |
+
"eval_runtime": 2945.7156,
|
| 82 |
+
"eval_samples_per_second": 0.761,
|
| 83 |
+
"eval_steps_per_second": 0.024,
|
| 84 |
+
"eval_wer": 5.883903942884959,
|
| 85 |
+
"step": 750
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 0.31007751937984496,
|
| 89 |
+
"grad_norm": 2.430060386657715,
|
| 90 |
+
"learning_rate": 4.4273684210526315e-07,
|
| 91 |
+
"loss": 4.0786,
|
| 92 |
+
"step": 800
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.3488372093023256,
|
| 96 |
+
"grad_norm": 2.526545286178589,
|
| 97 |
+
"learning_rate": 4.322105263157895e-07,
|
| 98 |
+
"loss": 4.0004,
|
| 99 |
+
"step": 900
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.3875968992248062,
|
| 103 |
+
"grad_norm": 2.5485153198242188,
|
| 104 |
+
"learning_rate": 4.2168421052631575e-07,
|
| 105 |
+
"loss": 3.962,
|
| 106 |
+
"step": 1000
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.3875968992248062,
|
| 110 |
+
"eval_loss": 0.344970703125,
|
| 111 |
+
"eval_runtime": 2929.618,
|
| 112 |
+
"eval_samples_per_second": 0.766,
|
| 113 |
+
"eval_steps_per_second": 0.024,
|
| 114 |
+
"eval_wer": 5.780464059711179,
|
| 115 |
+
"step": 1000
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.4263565891472868,
|
| 119 |
+
"grad_norm": 2.46012282371521,
|
| 120 |
+
"learning_rate": 4.111578947368421e-07,
|
| 121 |
+
"loss": 3.9314,
|
| 122 |
+
"step": 1100
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.46511627906976744,
|
| 126 |
+
"grad_norm": 2.6299571990966797,
|
| 127 |
+
"learning_rate": 4.0063157894736836e-07,
|
| 128 |
+
"loss": 3.9038,
|
| 129 |
+
"step": 1200
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.4844961240310077,
|
| 133 |
+
"eval_loss": 0.34033203125,
|
| 134 |
+
"eval_runtime": 3171.0392,
|
| 135 |
+
"eval_samples_per_second": 0.707,
|
| 136 |
+
"eval_steps_per_second": 0.022,
|
| 137 |
+
"eval_wer": 6.174617340040023,
|
| 138 |
+
"step": 1250
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.5038759689922481,
|
| 142 |
+
"grad_norm": 2.334170341491699,
|
| 143 |
+
"learning_rate": 3.9010526315789474e-07,
|
| 144 |
+
"loss": 3.8762,
|
| 145 |
+
"step": 1300
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.5426356589147286,
|
| 149 |
+
"grad_norm": 2.6218087673187256,
|
| 150 |
+
"learning_rate": 3.7957894736842106e-07,
|
| 151 |
+
"loss": 3.8491,
|
| 152 |
+
"step": 1400
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.5813953488372093,
|
| 156 |
+
"grad_norm": 2.390178680419922,
|
| 157 |
+
"learning_rate": 3.6905263157894734e-07,
|
| 158 |
+
"loss": 3.8313,
|
| 159 |
+
"step": 1500
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.5813953488372093,
|
| 163 |
+
"eval_loss": 0.3359375,
|
| 164 |
+
"eval_runtime": 3066.5551,
|
| 165 |
+
"eval_samples_per_second": 0.731,
|
| 166 |
+
"eval_steps_per_second": 0.023,
|
| 167 |
+
"eval_wer": 5.973822272702688,
|
| 168 |
+
"step": 1500
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 0.6201550387596899,
|
| 172 |
+
"grad_norm": 2.461087465286255,
|
| 173 |
+
"learning_rate": 3.5852631578947367e-07,
|
| 174 |
+
"loss": 3.804,
|
| 175 |
+
"step": 1600
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.6589147286821705,
|
| 179 |
+
"grad_norm": 2.4371871948242188,
|
| 180 |
+
"learning_rate": 3.4799999999999994e-07,
|
| 181 |
+
"loss": 3.7778,
|
| 182 |
+
"step": 1700
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.6782945736434108,
|
| 186 |
+
"eval_loss": 0.333251953125,
|
| 187 |
+
"eval_runtime": 3069.4158,
|
| 188 |
+
"eval_samples_per_second": 0.731,
|
| 189 |
+
"eval_steps_per_second": 0.023,
|
| 190 |
+
"eval_wer": 5.921764292281898,
|
| 191 |
+
"step": 1750
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.6976744186046512,
|
| 195 |
+
"grad_norm": 2.413174867630005,
|
| 196 |
+
"learning_rate": 3.374736842105263e-07,
|
| 197 |
+
"loss": 3.7737,
|
| 198 |
+
"step": 1800
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.7364341085271318,
|
| 202 |
+
"grad_norm": 2.4858949184417725,
|
| 203 |
+
"learning_rate": 3.2694736842105265e-07,
|
| 204 |
+
"loss": 3.7538,
|
| 205 |
+
"step": 1900
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.7751937984496124,
|
| 209 |
+
"grad_norm": 2.6870527267456055,
|
| 210 |
+
"learning_rate": 3.164210526315789e-07,
|
| 211 |
+
"loss": 3.7421,
|
| 212 |
+
"step": 2000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.7751937984496124,
|
| 216 |
+
"eval_loss": 0.33056640625,
|
| 217 |
+
"eval_runtime": 3182.1093,
|
| 218 |
+
"eval_samples_per_second": 0.705,
|
| 219 |
+
"eval_steps_per_second": 0.022,
|
| 220 |
+
"eval_wer": 6.13270052463627,
|
| 221 |
+
"step": 2000
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.813953488372093,
|
| 225 |
+
"grad_norm": 2.504830837249756,
|
| 226 |
+
"learning_rate": 3.0589473684210525e-07,
|
| 227 |
+
"loss": 3.7371,
|
| 228 |
+
"step": 2100
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 0.8527131782945736,
|
| 232 |
+
"grad_norm": 2.3637797832489014,
|
| 233 |
+
"learning_rate": 2.953684210526315e-07,
|
| 234 |
+
"loss": 3.7367,
|
| 235 |
+
"step": 2200
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.872093023255814,
|
| 239 |
+
"eval_loss": 0.328125,
|
| 240 |
+
"eval_runtime": 2953.0944,
|
| 241 |
+
"eval_samples_per_second": 0.76,
|
| 242 |
+
"eval_steps_per_second": 0.024,
|
| 243 |
+
"eval_wer": 5.656065768835524,
|
| 244 |
+
"step": 2250
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.8914728682170543,
|
| 248 |
+
"grad_norm": 2.3057029247283936,
|
| 249 |
+
"learning_rate": 2.848421052631579e-07,
|
| 250 |
+
"loss": 3.7176,
|
| 251 |
+
"step": 2300
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.9302325581395349,
|
| 255 |
+
"grad_norm": 2.277510643005371,
|
| 256 |
+
"learning_rate": 2.7431578947368423e-07,
|
| 257 |
+
"loss": 3.7003,
|
| 258 |
+
"step": 2400
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.9689922480620154,
|
| 262 |
+
"grad_norm": 2.3119587898254395,
|
| 263 |
+
"learning_rate": 2.637894736842105e-07,
|
| 264 |
+
"loss": 3.6878,
|
| 265 |
+
"step": 2500
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.9689922480620154,
|
| 269 |
+
"eval_loss": 0.32568359375,
|
| 270 |
+
"eval_runtime": 2931.2759,
|
| 271 |
+
"eval_samples_per_second": 0.765,
|
| 272 |
+
"eval_steps_per_second": 0.024,
|
| 273 |
+
"eval_wer": 5.515441613932608,
|
| 274 |
+
"step": 2500
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 1.0077519379844961,
|
| 278 |
+
"grad_norm": 2.32456636428833,
|
| 279 |
+
"learning_rate": 2.5326315789473683e-07,
|
| 280 |
+
"loss": 3.6732,
|
| 281 |
+
"step": 2600
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 1.0465116279069768,
|
| 285 |
+
"grad_norm": 2.489800453186035,
|
| 286 |
+
"learning_rate": 2.4273684210526316e-07,
|
| 287 |
+
"loss": 3.6769,
|
| 288 |
+
"step": 2700
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 1.0658914728682172,
|
| 292 |
+
"eval_loss": 0.32421875,
|
| 293 |
+
"eval_runtime": 2928.9686,
|
| 294 |
+
"eval_samples_per_second": 0.766,
|
| 295 |
+
"eval_steps_per_second": 0.024,
|
| 296 |
+
"eval_wer": 5.48028557520688,
|
| 297 |
+
"step": 2750
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.0852713178294573,
|
| 301 |
+
"grad_norm": 2.307305097579956,
|
| 302 |
+
"learning_rate": 2.3221052631578946e-07,
|
| 303 |
+
"loss": 3.6646,
|
| 304 |
+
"step": 2800
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 1.124031007751938,
|
| 308 |
+
"grad_norm": 2.3552660942077637,
|
| 309 |
+
"learning_rate": 2.216842105263158e-07,
|
| 310 |
+
"loss": 3.6604,
|
| 311 |
+
"step": 2900
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 1.1627906976744187,
|
| 315 |
+
"grad_norm": 2.3546299934387207,
|
| 316 |
+
"learning_rate": 2.111578947368421e-07,
|
| 317 |
+
"loss": 3.6508,
|
| 318 |
+
"step": 3000
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 1.1627906976744187,
|
| 322 |
+
"eval_loss": 0.323486328125,
|
| 323 |
+
"eval_runtime": 2929.9225,
|
| 324 |
+
"eval_samples_per_second": 0.766,
|
| 325 |
+
"eval_steps_per_second": 0.024,
|
| 326 |
+
"eval_wer": 5.463383633511818,
|
| 327 |
+
"step": 3000
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 1.2015503875968991,
|
| 331 |
+
"grad_norm": 2.2800722122192383,
|
| 332 |
+
"learning_rate": 2.006315789473684e-07,
|
| 333 |
+
"loss": 3.6412,
|
| 334 |
+
"step": 3100
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 1.2403100775193798,
|
| 338 |
+
"grad_norm": 2.368227005004883,
|
| 339 |
+
"learning_rate": 1.9010526315789475e-07,
|
| 340 |
+
"loss": 3.6292,
|
| 341 |
+
"step": 3200
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 1.2596899224806202,
|
| 345 |
+
"eval_loss": 0.322021484375,
|
| 346 |
+
"eval_runtime": 2816.4937,
|
| 347 |
+
"eval_samples_per_second": 0.796,
|
| 348 |
+
"eval_steps_per_second": 0.025,
|
| 349 |
+
"eval_wer": 5.3511547406566065,
|
| 350 |
+
"step": 3250
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.2790697674418605,
|
| 354 |
+
"grad_norm": 2.3508598804473877,
|
| 355 |
+
"learning_rate": 1.7957894736842105e-07,
|
| 356 |
+
"loss": 3.6196,
|
| 357 |
+
"step": 3300
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.3178294573643412,
|
| 361 |
+
"grad_norm": 2.4686381816864014,
|
| 362 |
+
"learning_rate": 1.6905263157894737e-07,
|
| 363 |
+
"loss": 3.6282,
|
| 364 |
+
"step": 3400
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 1.3565891472868217,
|
| 368 |
+
"grad_norm": 2.3237240314483643,
|
| 369 |
+
"learning_rate": 1.5852631578947367e-07,
|
| 370 |
+
"loss": 3.6179,
|
| 371 |
+
"step": 3500
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.3565891472868217,
|
| 375 |
+
"eval_loss": 0.321044921875,
|
| 376 |
+
"eval_runtime": 2685.856,
|
| 377 |
+
"eval_samples_per_second": 0.835,
|
| 378 |
+
"eval_steps_per_second": 0.026,
|
| 379 |
+
"eval_wer": 5.225404294445346,
|
| 380 |
+
"step": 3500
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.3953488372093024,
|
| 384 |
+
"grad_norm": 2.2395377159118652,
|
| 385 |
+
"learning_rate": 1.4799999999999998e-07,
|
| 386 |
+
"loss": 3.6078,
|
| 387 |
+
"step": 3600
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.4341085271317828,
|
| 391 |
+
"grad_norm": 2.3762331008911133,
|
| 392 |
+
"learning_rate": 1.374736842105263e-07,
|
| 393 |
+
"loss": 3.6032,
|
| 394 |
+
"step": 3700
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.4534883720930232,
|
| 398 |
+
"eval_loss": 0.320556640625,
|
| 399 |
+
"eval_runtime": 2693.0163,
|
| 400 |
+
"eval_samples_per_second": 0.833,
|
| 401 |
+
"eval_steps_per_second": 0.026,
|
| 402 |
+
"eval_wer": 5.2206717507707285,
|
| 403 |
+
"step": 3750
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.4728682170542635,
|
| 407 |
+
"grad_norm": 2.3068864345550537,
|
| 408 |
+
"learning_rate": 1.2694736842105263e-07,
|
| 409 |
+
"loss": 3.5997,
|
| 410 |
+
"step": 3800
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.5116279069767442,
|
| 414 |
+
"grad_norm": 2.40097713470459,
|
| 415 |
+
"learning_rate": 1.1642105263157894e-07,
|
| 416 |
+
"loss": 3.588,
|
| 417 |
+
"step": 3900
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.550387596899225,
|
| 421 |
+
"grad_norm": 2.320568561553955,
|
| 422 |
+
"learning_rate": 1.0589473684210526e-07,
|
| 423 |
+
"loss": 3.5922,
|
| 424 |
+
"step": 4000
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.550387596899225,
|
| 428 |
+
"eval_loss": 0.320068359375,
|
| 429 |
+
"eval_runtime": 2803.374,
|
| 430 |
+
"eval_samples_per_second": 0.8,
|
| 431 |
+
"eval_steps_per_second": 0.025,
|
| 432 |
+
"eval_wer": 5.303829303910433,
|
| 433 |
+
"step": 4000
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 1.5891472868217056,
|
| 437 |
+
"grad_norm": 2.4094741344451904,
|
| 438 |
+
"learning_rate": 9.557894736842105e-08,
|
| 439 |
+
"loss": 3.5832,
|
| 440 |
+
"step": 4100
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.627906976744186,
|
| 444 |
+
"grad_norm": 2.448002815246582,
|
| 445 |
+
"learning_rate": 8.505263157894736e-08,
|
| 446 |
+
"loss": 3.5743,
|
| 447 |
+
"step": 4200
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 1.6472868217054264,
|
| 451 |
+
"eval_loss": 0.31982421875,
|
| 452 |
+
"eval_runtime": 2778.7179,
|
| 453 |
+
"eval_samples_per_second": 0.807,
|
| 454 |
+
"eval_steps_per_second": 0.026,
|
| 455 |
+
"eval_wer": 5.263264643842285,
|
| 456 |
+
"step": 4250
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 1.6666666666666665,
|
| 460 |
+
"grad_norm": 2.362182855606079,
|
| 461 |
+
"learning_rate": 7.452631578947369e-08,
|
| 462 |
+
"loss": 3.5831,
|
| 463 |
+
"step": 4300
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 1.7054263565891472,
|
| 467 |
+
"grad_norm": 2.3121325969696045,
|
| 468 |
+
"learning_rate": 6.4e-08,
|
| 469 |
+
"loss": 3.5738,
|
| 470 |
+
"step": 4400
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 1.744186046511628,
|
| 474 |
+
"grad_norm": 2.3000595569610596,
|
| 475 |
+
"learning_rate": 5.347368421052632e-08,
|
| 476 |
+
"loss": 3.5882,
|
| 477 |
+
"step": 4500
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 1.744186046511628,
|
| 481 |
+
"eval_loss": 0.31982421875,
|
| 482 |
+
"eval_runtime": 2718.6235,
|
| 483 |
+
"eval_samples_per_second": 0.825,
|
| 484 |
+
"eval_steps_per_second": 0.026,
|
| 485 |
+
"eval_wer": 5.225404294445346,
|
| 486 |
+
"step": 4500
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.7829457364341086,
|
| 490 |
+
"grad_norm": 2.3451480865478516,
|
| 491 |
+
"learning_rate": 4.2947368421052626e-08,
|
| 492 |
+
"loss": 3.5873,
|
| 493 |
+
"step": 4600
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.8217054263565893,
|
| 497 |
+
"grad_norm": 2.301790237426758,
|
| 498 |
+
"learning_rate": 3.2421052631578947e-08,
|
| 499 |
+
"loss": 3.6021,
|
| 500 |
+
"step": 4700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.8410852713178296,
|
| 504 |
+
"eval_loss": 0.319580078125,
|
| 505 |
+
"eval_runtime": 2680.8066,
|
| 506 |
+
"eval_samples_per_second": 0.837,
|
| 507 |
+
"eval_steps_per_second": 0.026,
|
| 508 |
+
"eval_wer": 5.218643517767322,
|
| 509 |
+
"step": 4750
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 1.8604651162790697,
|
| 513 |
+
"grad_norm": 2.2558062076568604,
|
| 514 |
+
"learning_rate": 2.189473684210526e-08,
|
| 515 |
+
"loss": 3.5833,
|
| 516 |
+
"step": 4800
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 1.8992248062015504,
|
| 520 |
+
"grad_norm": 2.2542638778686523,
|
| 521 |
+
"learning_rate": 1.136842105263158e-08,
|
| 522 |
+
"loss": 3.5732,
|
| 523 |
+
"step": 4900
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 1.937984496124031,
|
| 527 |
+
"grad_norm": 2.378117322921753,
|
| 528 |
+
"learning_rate": 8.421052631578947e-10,
|
| 529 |
+
"loss": 3.5865,
|
| 530 |
+
"step": 5000
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 1.937984496124031,
|
| 534 |
+
"eval_loss": 0.3193359375,
|
| 535 |
+
"eval_runtime": 2696.1288,
|
| 536 |
+
"eval_samples_per_second": 0.832,
|
| 537 |
+
"eval_steps_per_second": 0.026,
|
| 538 |
+
"eval_wer": 5.221347828438531,
|
| 539 |
+
"step": 5000
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 1.937984496124031,
|
| 543 |
+
"step": 5000,
|
| 544 |
+
"total_flos": 1.087083891827713e+21,
|
| 545 |
+
"train_loss": 3.86036796875,
|
| 546 |
+
"train_runtime": 85811.7084,
|
| 547 |
+
"train_samples_per_second": 3.729,
|
| 548 |
+
"train_steps_per_second": 0.058
|
| 549 |
+
}
|
| 550 |
+
],
|
| 551 |
+
"logging_steps": 100,
|
| 552 |
+
"max_steps": 5000,
|
| 553 |
+
"num_input_tokens_seen": 0,
|
| 554 |
+
"num_train_epochs": 2,
|
| 555 |
+
"save_steps": 250,
|
| 556 |
+
"stateful_callbacks": {
|
| 557 |
+
"TrainerControl": {
|
| 558 |
+
"args": {
|
| 559 |
+
"should_epoch_stop": false,
|
| 560 |
+
"should_evaluate": false,
|
| 561 |
+
"should_log": false,
|
| 562 |
+
"should_save": true,
|
| 563 |
+
"should_training_stop": true
|
| 564 |
+
},
|
| 565 |
+
"attributes": {}
|
| 566 |
+
}
|
| 567 |
+
},
|
| 568 |
+
"total_flos": 1.087083891827713e+21,
|
| 569 |
+
"train_batch_size": 64,
|
| 570 |
+
"trial_name": null,
|
| 571 |
+
"trial_params": null
|
| 572 |
+
}
|