End of training
Browse files- README.md +17 -4
- all_results.json +14 -0
- eval_results.json +9 -0
- runs/Dec08_08-51-14_d27baedf9b3d/events.out.tfevents.1733746088.d27baedf9b3d.62779.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +617 -0
README.md
CHANGED
|
@@ -3,20 +3,33 @@ 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: 0.1453
|
| 22 |
- Wer: 3.2636
|
|
|
|
| 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_speech_corpus
|
| 10 |
metrics:
|
| 11 |
- wer
|
| 12 |
model-index:
|
| 13 |
+
- name: Whisper Large V3 Medical
|
| 14 |
+
results:
|
| 15 |
+
- task:
|
| 16 |
+
name: Automatic Speech Recognition
|
| 17 |
+
type: automatic-speech-recognition
|
| 18 |
+
dataset:
|
| 19 |
+
name: OUTCOMESAI/medical_speech_corpus en
|
| 20 |
+
type: OUTCOMESAI/medical_speech_corpus
|
| 21 |
+
metrics:
|
| 22 |
+
- name: Wer
|
| 23 |
+
type: wer
|
| 24 |
+
value: 3.2635854592980795
|
| 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 Medical
|
| 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_speech_corpus en dataset.
|
| 33 |
It achieves the following results on the evaluation set:
|
| 34 |
- Loss: 0.1453
|
| 35 |
- Wer: 3.2636
|
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.8255547054322876,
|
| 3 |
+
"eval_loss": 0.145263671875,
|
| 4 |
+
"eval_runtime": 998.2773,
|
| 5 |
+
"eval_samples": 5000,
|
| 6 |
+
"eval_samples_per_second": 1.12,
|
| 7 |
+
"eval_steps_per_second": 0.035,
|
| 8 |
+
"eval_wer": 3.2635854592980795,
|
| 9 |
+
"total_flos": 1.0871892140465376e+21,
|
| 10 |
+
"train_loss": 0.9307616455078125,
|
| 11 |
+
"train_runtime": 97022.7442,
|
| 12 |
+
"train_samples_per_second": 3.298,
|
| 13 |
+
"train_steps_per_second": 0.052
|
| 14 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.8255547054322876,
|
| 3 |
+
"eval_loss": 0.145263671875,
|
| 4 |
+
"eval_runtime": 998.2773,
|
| 5 |
+
"eval_samples": 5000,
|
| 6 |
+
"eval_samples_per_second": 1.12,
|
| 7 |
+
"eval_steps_per_second": 0.035,
|
| 8 |
+
"eval_wer": 3.2635854592980795
|
| 9 |
+
}
|
runs/Dec08_08-51-14_d27baedf9b3d/events.out.tfevents.1733746088.d27baedf9b3d.62779.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:038de83d50b74d2e4017f170901d1e2ccd240953b7d78b6c47f3f154bc027388
|
| 3 |
+
size 406
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.8255547054322876,
|
| 3 |
+
"total_flos": 1.0871892140465376e+21,
|
| 4 |
+
"train_loss": 0.9307616455078125,
|
| 5 |
+
"train_runtime": 97022.7442,
|
| 6 |
+
"train_samples_per_second": 3.298,
|
| 7 |
+
"train_steps_per_second": 0.052
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,617 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 3.2635854592980795,
|
| 3 |
+
"best_model_checkpoint": "OUTCOMESAI/whisper-large-v3-medical/checkpoint-5000",
|
| 4 |
+
"epoch": 3.8255547054322876,
|
| 5 |
+
"eval_steps": 200,
|
| 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": 0.07651109410864575,
|
| 13 |
+
"grad_norm": 4.125481605529785,
|
| 14 |
+
"learning_rate": 4.2874883363489934e-07,
|
| 15 |
+
"loss": 6.0929,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.1530221882172915,
|
| 20 |
+
"grad_norm": 4.159045696258545,
|
| 21 |
+
"learning_rate": 4.971255772939331e-07,
|
| 22 |
+
"loss": 4.2439,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.1530221882172915,
|
| 27 |
+
"eval_loss": 0.29345703125,
|
| 28 |
+
"eval_runtime": 988.7646,
|
| 29 |
+
"eval_samples_per_second": 1.131,
|
| 30 |
+
"eval_steps_per_second": 0.035,
|
| 31 |
+
"eval_wer": 4.507771356816114,
|
| 32 |
+
"step": 200
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.22953328232593725,
|
| 36 |
+
"grad_norm": 2.488393545150757,
|
| 37 |
+
"learning_rate": 4.903124999999999e-07,
|
| 38 |
+
"loss": 3.5952,
|
| 39 |
+
"step": 300
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.306044376434583,
|
| 43 |
+
"grad_norm": 2.386241912841797,
|
| 44 |
+
"learning_rate": 4.798958333333334e-07,
|
| 45 |
+
"loss": 3.3374,
|
| 46 |
+
"step": 400
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.306044376434583,
|
| 50 |
+
"eval_loss": 0.2734375,
|
| 51 |
+
"eval_runtime": 1214.5433,
|
| 52 |
+
"eval_samples_per_second": 0.921,
|
| 53 |
+
"eval_steps_per_second": 0.029,
|
| 54 |
+
"eval_wer": 4.696129057058288,
|
| 55 |
+
"step": 400
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 0.38255547054322875,
|
| 59 |
+
"grad_norm": 2.348032236099243,
|
| 60 |
+
"learning_rate": 4.6947916666666664e-07,
|
| 61 |
+
"loss": 3.1968,
|
| 62 |
+
"step": 500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.4590665646518745,
|
| 66 |
+
"grad_norm": 2.6516611576080322,
|
| 67 |
+
"learning_rate": 4.5906249999999995e-07,
|
| 68 |
+
"loss": 3.0833,
|
| 69 |
+
"step": 600
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.4590665646518745,
|
| 73 |
+
"eval_loss": 0.267333984375,
|
| 74 |
+
"eval_runtime": 991.4442,
|
| 75 |
+
"eval_samples_per_second": 1.128,
|
| 76 |
+
"eval_steps_per_second": 0.035,
|
| 77 |
+
"eval_wer": 4.273285240188102,
|
| 78 |
+
"step": 600
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 0.5355776587605203,
|
| 82 |
+
"grad_norm": 3.4489781856536865,
|
| 83 |
+
"learning_rate": 4.486458333333333e-07,
|
| 84 |
+
"loss": 2.6306,
|
| 85 |
+
"step": 700
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 0.612088752869166,
|
| 89 |
+
"grad_norm": 2.9474337100982666,
|
| 90 |
+
"learning_rate": 4.3822916666666667e-07,
|
| 91 |
+
"loss": 1.8243,
|
| 92 |
+
"step": 800
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.612088752869166,
|
| 96 |
+
"eval_loss": 0.26806640625,
|
| 97 |
+
"eval_runtime": 991.8395,
|
| 98 |
+
"eval_samples_per_second": 1.127,
|
| 99 |
+
"eval_steps_per_second": 0.035,
|
| 100 |
+
"eval_wer": 4.4372973873377495,
|
| 101 |
+
"step": 800
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.6885998469778117,
|
| 105 |
+
"grad_norm": 2.7142157554626465,
|
| 106 |
+
"learning_rate": 4.278125e-07,
|
| 107 |
+
"loss": 1.382,
|
| 108 |
+
"step": 900
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.7651109410864575,
|
| 112 |
+
"grad_norm": 2.386582612991333,
|
| 113 |
+
"learning_rate": 4.1739583333333334e-07,
|
| 114 |
+
"loss": 1.1288,
|
| 115 |
+
"step": 1000
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.7651109410864575,
|
| 119 |
+
"eval_loss": 0.2548828125,
|
| 120 |
+
"eval_runtime": 991.8205,
|
| 121 |
+
"eval_samples_per_second": 1.127,
|
| 122 |
+
"eval_steps_per_second": 0.035,
|
| 123 |
+
"eval_wer": 4.277129274886922,
|
| 124 |
+
"step": 1000
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.8416220351951033,
|
| 128 |
+
"grad_norm": 2.2802257537841797,
|
| 129 |
+
"learning_rate": 4.0697916666666665e-07,
|
| 130 |
+
"loss": 0.9419,
|
| 131 |
+
"step": 1100
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.918133129303749,
|
| 135 |
+
"grad_norm": 2.0405514240264893,
|
| 136 |
+
"learning_rate": 3.9656249999999996e-07,
|
| 137 |
+
"loss": 0.8199,
|
| 138 |
+
"step": 1200
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.918133129303749,
|
| 142 |
+
"eval_loss": 0.2412109375,
|
| 143 |
+
"eval_runtime": 994.3546,
|
| 144 |
+
"eval_samples_per_second": 1.124,
|
| 145 |
+
"eval_steps_per_second": 0.035,
|
| 146 |
+
"eval_wer": 4.204092615609343,
|
| 147 |
+
"step": 1200
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.9946442234123948,
|
| 151 |
+
"grad_norm": 1.88164222240448,
|
| 152 |
+
"learning_rate": 3.861458333333333e-07,
|
| 153 |
+
"loss": 0.7493,
|
| 154 |
+
"step": 1300
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.0711553175210407,
|
| 158 |
+
"grad_norm": 1.8802990913391113,
|
| 159 |
+
"learning_rate": 3.7572916666666663e-07,
|
| 160 |
+
"loss": 0.681,
|
| 161 |
+
"step": 1400
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.0711553175210407,
|
| 165 |
+
"eval_loss": 0.2310791015625,
|
| 166 |
+
"eval_runtime": 990.487,
|
| 167 |
+
"eval_samples_per_second": 1.129,
|
| 168 |
+
"eval_steps_per_second": 0.035,
|
| 169 |
+
"eval_wer": 4.105429058339634,
|
| 170 |
+
"step": 1400
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.1476664116296864,
|
| 174 |
+
"grad_norm": 1.6826539039611816,
|
| 175 |
+
"learning_rate": 3.6531249999999994e-07,
|
| 176 |
+
"loss": 0.6296,
|
| 177 |
+
"step": 1500
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.2241775057383322,
|
| 181 |
+
"grad_norm": 1.7766120433807373,
|
| 182 |
+
"learning_rate": 3.5489583333333336e-07,
|
| 183 |
+
"loss": 0.5798,
|
| 184 |
+
"step": 1600
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.2241775057383322,
|
| 188 |
+
"eval_loss": 0.21923828125,
|
| 189 |
+
"eval_runtime": 989.3653,
|
| 190 |
+
"eval_samples_per_second": 1.13,
|
| 191 |
+
"eval_steps_per_second": 0.035,
|
| 192 |
+
"eval_wer": 4.009328190869136,
|
| 193 |
+
"step": 1600
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 1.300688599846978,
|
| 197 |
+
"grad_norm": 1.597964882850647,
|
| 198 |
+
"learning_rate": 3.4447916666666667e-07,
|
| 199 |
+
"loss": 0.5603,
|
| 200 |
+
"step": 1700
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.3771996939556237,
|
| 204 |
+
"grad_norm": 1.5654646158218384,
|
| 205 |
+
"learning_rate": 3.340625e-07,
|
| 206 |
+
"loss": 0.5233,
|
| 207 |
+
"step": 1800
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.3771996939556237,
|
| 211 |
+
"eval_loss": 0.2071533203125,
|
| 212 |
+
"eval_runtime": 987.6154,
|
| 213 |
+
"eval_samples_per_second": 1.132,
|
| 214 |
+
"eval_steps_per_second": 0.035,
|
| 215 |
+
"eval_wer": 3.8927258050049334,
|
| 216 |
+
"step": 1800
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 1.4537107880642695,
|
| 220 |
+
"grad_norm": 1.5115400552749634,
|
| 221 |
+
"learning_rate": 3.2364583333333334e-07,
|
| 222 |
+
"loss": 0.4919,
|
| 223 |
+
"step": 1900
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 1.5302218821729152,
|
| 227 |
+
"grad_norm": 1.5361956357955933,
|
| 228 |
+
"learning_rate": 3.1322916666666665e-07,
|
| 229 |
+
"loss": 0.463,
|
| 230 |
+
"step": 2000
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 1.5302218821729152,
|
| 234 |
+
"eval_loss": 0.19921875,
|
| 235 |
+
"eval_runtime": 984.868,
|
| 236 |
+
"eval_samples_per_second": 1.135,
|
| 237 |
+
"eval_steps_per_second": 0.036,
|
| 238 |
+
"eval_wer": 3.8196891457273554,
|
| 239 |
+
"step": 2000
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 1.606732976281561,
|
| 243 |
+
"grad_norm": 1.4659453630447388,
|
| 244 |
+
"learning_rate": 3.0281249999999996e-07,
|
| 245 |
+
"loss": 0.4509,
|
| 246 |
+
"step": 2100
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 1.6832440703902067,
|
| 250 |
+
"grad_norm": 1.384372591972351,
|
| 251 |
+
"learning_rate": 2.923958333333333e-07,
|
| 252 |
+
"loss": 0.428,
|
| 253 |
+
"step": 2200
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 1.6832440703902067,
|
| 257 |
+
"eval_loss": 0.195068359375,
|
| 258 |
+
"eval_runtime": 985.7714,
|
| 259 |
+
"eval_samples_per_second": 1.134,
|
| 260 |
+
"eval_steps_per_second": 0.036,
|
| 261 |
+
"eval_wer": 3.7748420742411235,
|
| 262 |
+
"step": 2200
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 1.7597551644988525,
|
| 266 |
+
"grad_norm": 1.451076865196228,
|
| 267 |
+
"learning_rate": 2.8197916666666663e-07,
|
| 268 |
+
"loss": 0.4161,
|
| 269 |
+
"step": 2300
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 1.836266258607498,
|
| 273 |
+
"grad_norm": 1.5242278575897217,
|
| 274 |
+
"learning_rate": 2.715625e-07,
|
| 275 |
+
"loss": 0.3944,
|
| 276 |
+
"step": 2400
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 1.836266258607498,
|
| 280 |
+
"eval_loss": 0.1866455078125,
|
| 281 |
+
"eval_runtime": 987.2353,
|
| 282 |
+
"eval_samples_per_second": 1.132,
|
| 283 |
+
"eval_steps_per_second": 0.035,
|
| 284 |
+
"eval_wer": 3.67745986187102,
|
| 285 |
+
"step": 2400
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 1.9127773527161438,
|
| 289 |
+
"grad_norm": 1.31293785572052,
|
| 290 |
+
"learning_rate": 2.6114583333333336e-07,
|
| 291 |
+
"loss": 0.38,
|
| 292 |
+
"step": 2500
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 1.9892884468247896,
|
| 296 |
+
"grad_norm": 1.3785984516143799,
|
| 297 |
+
"learning_rate": 2.5072916666666667e-07,
|
| 298 |
+
"loss": 0.3682,
|
| 299 |
+
"step": 2600
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 1.9892884468247896,
|
| 303 |
+
"eval_loss": 0.17919921875,
|
| 304 |
+
"eval_runtime": 989.9094,
|
| 305 |
+
"eval_samples_per_second": 1.129,
|
| 306 |
+
"eval_steps_per_second": 0.035,
|
| 307 |
+
"eval_wer": 3.604423202593442,
|
| 308 |
+
"step": 2600
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 2.0657995409334355,
|
| 312 |
+
"grad_norm": 1.3718817234039307,
|
| 313 |
+
"learning_rate": 2.403125e-07,
|
| 314 |
+
"loss": 0.36,
|
| 315 |
+
"step": 2700
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 2.1423106350420813,
|
| 319 |
+
"grad_norm": 1.3219904899597168,
|
| 320 |
+
"learning_rate": 2.298958333333333e-07,
|
| 321 |
+
"loss": 0.3543,
|
| 322 |
+
"step": 2800
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 2.1423106350420813,
|
| 326 |
+
"eval_loss": 0.1724853515625,
|
| 327 |
+
"eval_runtime": 995.0367,
|
| 328 |
+
"eval_samples_per_second": 1.124,
|
| 329 |
+
"eval_steps_per_second": 0.035,
|
| 330 |
+
"eval_wer": 3.530105198416258,
|
| 331 |
+
"step": 2800
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 2.218821729150727,
|
| 335 |
+
"grad_norm": 1.2633302211761475,
|
| 336 |
+
"learning_rate": 2.1947916666666665e-07,
|
| 337 |
+
"loss": 0.3438,
|
| 338 |
+
"step": 2900
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.295332823259373,
|
| 342 |
+
"grad_norm": 1.2741142511367798,
|
| 343 |
+
"learning_rate": 2.090625e-07,
|
| 344 |
+
"loss": 0.3368,
|
| 345 |
+
"step": 3000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.295332823259373,
|
| 349 |
+
"eval_loss": 0.17138671875,
|
| 350 |
+
"eval_runtime": 991.227,
|
| 351 |
+
"eval_samples_per_second": 1.128,
|
| 352 |
+
"eval_steps_per_second": 0.035,
|
| 353 |
+
"eval_wer": 3.490383506528452,
|
| 354 |
+
"step": 3000
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.371843917368018,
|
| 358 |
+
"grad_norm": 1.2177200317382812,
|
| 359 |
+
"learning_rate": 1.9864583333333332e-07,
|
| 360 |
+
"loss": 0.3218,
|
| 361 |
+
"step": 3100
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.4483550114766643,
|
| 365 |
+
"grad_norm": 1.24095618724823,
|
| 366 |
+
"learning_rate": 1.8822916666666666e-07,
|
| 367 |
+
"loss": 0.3136,
|
| 368 |
+
"step": 3200
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.4483550114766643,
|
| 372 |
+
"eval_loss": 0.164794921875,
|
| 373 |
+
"eval_runtime": 995.8265,
|
| 374 |
+
"eval_samples_per_second": 1.123,
|
| 375 |
+
"eval_steps_per_second": 0.035,
|
| 376 |
+
"eval_wer": 3.45706853913868,
|
| 377 |
+
"step": 3200
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 2.5248661055853097,
|
| 381 |
+
"grad_norm": 1.146142840385437,
|
| 382 |
+
"learning_rate": 1.778125e-07,
|
| 383 |
+
"loss": 0.3269,
|
| 384 |
+
"step": 3300
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 2.601377199693956,
|
| 388 |
+
"grad_norm": 1.1721028089523315,
|
| 389 |
+
"learning_rate": 1.6739583333333333e-07,
|
| 390 |
+
"loss": 0.3121,
|
| 391 |
+
"step": 3400
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 2.601377199693956,
|
| 395 |
+
"eval_loss": 0.160400390625,
|
| 396 |
+
"eval_runtime": 990.5861,
|
| 397 |
+
"eval_samples_per_second": 1.129,
|
| 398 |
+
"eval_steps_per_second": 0.035,
|
| 399 |
+
"eval_wer": 3.4237535717489074,
|
| 400 |
+
"step": 3400
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 2.677888293802601,
|
| 404 |
+
"grad_norm": 1.5002176761627197,
|
| 405 |
+
"learning_rate": 1.5697916666666666e-07,
|
| 406 |
+
"loss": 0.2999,
|
| 407 |
+
"step": 3500
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 2.7543993879112474,
|
| 411 |
+
"grad_norm": 1.250117301940918,
|
| 412 |
+
"learning_rate": 1.465625e-07,
|
| 413 |
+
"loss": 0.2959,
|
| 414 |
+
"step": 3600
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 2.7543993879112474,
|
| 418 |
+
"eval_loss": 0.1561279296875,
|
| 419 |
+
"eval_runtime": 997.6176,
|
| 420 |
+
"eval_samples_per_second": 1.121,
|
| 421 |
+
"eval_steps_per_second": 0.035,
|
| 422 |
+
"eval_wer": 3.395563983957562,
|
| 423 |
+
"step": 3600
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 2.8309104820198927,
|
| 427 |
+
"grad_norm": 1.2006161212921143,
|
| 428 |
+
"learning_rate": 1.361458333333333e-07,
|
| 429 |
+
"loss": 0.2942,
|
| 430 |
+
"step": 3700
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 2.907421576128539,
|
| 434 |
+
"grad_norm": 1.2639755010604858,
|
| 435 |
+
"learning_rate": 1.2572916666666667e-07,
|
| 436 |
+
"loss": 0.2912,
|
| 437 |
+
"step": 3800
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 2.907421576128539,
|
| 441 |
+
"eval_loss": 0.15380859375,
|
| 442 |
+
"eval_runtime": 989.6556,
|
| 443 |
+
"eval_samples_per_second": 1.13,
|
| 444 |
+
"eval_steps_per_second": 0.035,
|
| 445 |
+
"eval_wer": 3.3737811206642494,
|
| 446 |
+
"step": 3800
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 2.9839326702371842,
|
| 450 |
+
"grad_norm": 1.1383545398712158,
|
| 451 |
+
"learning_rate": 1.153125e-07,
|
| 452 |
+
"loss": 0.2921,
|
| 453 |
+
"step": 3900
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 3.06044376434583,
|
| 457 |
+
"grad_norm": 1.1553541421890259,
|
| 458 |
+
"learning_rate": 1.0489583333333332e-07,
|
| 459 |
+
"loss": 0.2767,
|
| 460 |
+
"step": 4000
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 3.06044376434583,
|
| 464 |
+
"eval_loss": 0.151123046875,
|
| 465 |
+
"eval_runtime": 994.8808,
|
| 466 |
+
"eval_samples_per_second": 1.124,
|
| 467 |
+
"eval_steps_per_second": 0.035,
|
| 468 |
+
"eval_wer": 3.3455915328729033,
|
| 469 |
+
"step": 4000
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 3.1369548584544757,
|
| 473 |
+
"grad_norm": 14.852476119995117,
|
| 474 |
+
"learning_rate": 9.447916666666667e-08,
|
| 475 |
+
"loss": 0.2857,
|
| 476 |
+
"step": 4100
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 3.2134659525631215,
|
| 480 |
+
"grad_norm": 3.1388800144195557,
|
| 481 |
+
"learning_rate": 8.406249999999999e-08,
|
| 482 |
+
"loss": 0.2848,
|
| 483 |
+
"step": 4200
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 3.2134659525631215,
|
| 487 |
+
"eval_loss": 0.148681640625,
|
| 488 |
+
"eval_runtime": 1017.2976,
|
| 489 |
+
"eval_samples_per_second": 1.099,
|
| 490 |
+
"eval_steps_per_second": 0.034,
|
| 491 |
+
"eval_wer": 3.319964634880771,
|
| 492 |
+
"step": 4200
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 3.2899770466717673,
|
| 496 |
+
"grad_norm": 1.152143120765686,
|
| 497 |
+
"learning_rate": 7.364583333333333e-08,
|
| 498 |
+
"loss": 0.2756,
|
| 499 |
+
"step": 4300
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 3.366488140780413,
|
| 503 |
+
"grad_norm": 1.1069058179855347,
|
| 504 |
+
"learning_rate": 6.322916666666666e-08,
|
| 505 |
+
"loss": 0.274,
|
| 506 |
+
"step": 4400
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 3.366488140780413,
|
| 510 |
+
"eval_loss": 0.1474609375,
|
| 511 |
+
"eval_runtime": 993.2425,
|
| 512 |
+
"eval_samples_per_second": 1.126,
|
| 513 |
+
"eval_steps_per_second": 0.035,
|
| 514 |
+
"eval_wer": 3.284086977691785,
|
| 515 |
+
"step": 4400
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 3.442999234889059,
|
| 519 |
+
"grad_norm": 1.054897427558899,
|
| 520 |
+
"learning_rate": 5.291666666666667e-08,
|
| 521 |
+
"loss": 0.2729,
|
| 522 |
+
"step": 4500
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 3.5195103289977046,
|
| 526 |
+
"grad_norm": 1.0845283269882202,
|
| 527 |
+
"learning_rate": 4.2500000000000003e-08,
|
| 528 |
+
"loss": 0.2694,
|
| 529 |
+
"step": 4600
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 3.5195103289977046,
|
| 533 |
+
"eval_loss": 0.1463623046875,
|
| 534 |
+
"eval_runtime": 1016.9332,
|
| 535 |
+
"eval_samples_per_second": 1.099,
|
| 536 |
+
"eval_steps_per_second": 0.034,
|
| 537 |
+
"eval_wer": 3.282805632792179,
|
| 538 |
+
"step": 4600
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 3.5960214231063503,
|
| 542 |
+
"grad_norm": 1.0754883289337158,
|
| 543 |
+
"learning_rate": 3.208333333333333e-08,
|
| 544 |
+
"loss": 0.2635,
|
| 545 |
+
"step": 4700
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 3.672532517214996,
|
| 549 |
+
"grad_norm": 1.3197473287582397,
|
| 550 |
+
"learning_rate": 2.1770833333333332e-08,
|
| 551 |
+
"loss": 0.2731,
|
| 552 |
+
"step": 4800
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 3.672532517214996,
|
| 556 |
+
"eval_loss": 0.1455078125,
|
| 557 |
+
"eval_runtime": 989.4621,
|
| 558 |
+
"eval_samples_per_second": 1.13,
|
| 559 |
+
"eval_steps_per_second": 0.035,
|
| 560 |
+
"eval_wer": 3.268710838896506,
|
| 561 |
+
"step": 4800
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 3.749043611323642,
|
| 565 |
+
"grad_norm": 1.156948208808899,
|
| 566 |
+
"learning_rate": 1.1354166666666667e-08,
|
| 567 |
+
"loss": 0.2589,
|
| 568 |
+
"step": 4900
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 3.8255547054322876,
|
| 572 |
+
"grad_norm": 1.1573657989501953,
|
| 573 |
+
"learning_rate": 9.375e-10,
|
| 574 |
+
"loss": 0.2677,
|
| 575 |
+
"step": 5000
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 3.8255547054322876,
|
| 579 |
+
"eval_loss": 0.145263671875,
|
| 580 |
+
"eval_runtime": 992.9783,
|
| 581 |
+
"eval_samples_per_second": 1.126,
|
| 582 |
+
"eval_steps_per_second": 0.035,
|
| 583 |
+
"eval_wer": 3.2635854592980795,
|
| 584 |
+
"step": 5000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 3.8255547054322876,
|
| 588 |
+
"step": 5000,
|
| 589 |
+
"total_flos": 1.0871892140465376e+21,
|
| 590 |
+
"train_loss": 0.9307616455078125,
|
| 591 |
+
"train_runtime": 97022.7442,
|
| 592 |
+
"train_samples_per_second": 3.298,
|
| 593 |
+
"train_steps_per_second": 0.052
|
| 594 |
+
}
|
| 595 |
+
],
|
| 596 |
+
"logging_steps": 100,
|
| 597 |
+
"max_steps": 5000,
|
| 598 |
+
"num_input_tokens_seen": 0,
|
| 599 |
+
"num_train_epochs": 4,
|
| 600 |
+
"save_steps": 200,
|
| 601 |
+
"stateful_callbacks": {
|
| 602 |
+
"TrainerControl": {
|
| 603 |
+
"args": {
|
| 604 |
+
"should_epoch_stop": false,
|
| 605 |
+
"should_evaluate": false,
|
| 606 |
+
"should_log": false,
|
| 607 |
+
"should_save": true,
|
| 608 |
+
"should_training_stop": true
|
| 609 |
+
},
|
| 610 |
+
"attributes": {}
|
| 611 |
+
}
|
| 612 |
+
},
|
| 613 |
+
"total_flos": 1.0871892140465376e+21,
|
| 614 |
+
"train_batch_size": 64,
|
| 615 |
+
"trial_name": null,
|
| 616 |
+
"trial_params": null
|
| 617 |
+
}
|