End of training
Browse files- README.md +10 -9
- all_results.json +10 -10
- eval_results.json +6 -6
- runs/Jan21_18-19-42_knight/events.out.tfevents.1705936715.knight.67737.1 +3 -0
- train_results.json +5 -5
- trainer_state.json +555 -42
README.md
CHANGED
|
@@ -2,38 +2,39 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: openai/whisper-base
|
| 4 |
tags:
|
|
|
|
| 5 |
- generated_from_trainer
|
| 6 |
datasets:
|
| 7 |
-
- fleurs
|
| 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: fleurs
|
| 18 |
-
type: fleurs
|
| 19 |
config: te_in
|
| 20 |
split: test
|
| 21 |
args: te_in
|
| 22 |
metrics:
|
| 23 |
- name: Wer
|
| 24 |
type: wer
|
| 25 |
-
value: 37.
|
| 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 fleurs dataset.
|
| 34 |
It achieves the following results on the evaluation set:
|
| 35 |
-
- Loss: 0.
|
| 36 |
-
- Wer: 37.
|
| 37 |
|
| 38 |
## Model description
|
| 39 |
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: openai/whisper-base
|
| 4 |
tags:
|
| 5 |
+
- whisper-event
|
| 6 |
- generated_from_trainer
|
| 7 |
datasets:
|
| 8 |
+
- google/fleurs
|
| 9 |
metrics:
|
| 10 |
- wer
|
| 11 |
model-index:
|
| 12 |
+
- name: Breeze DSW Telugu - base
|
| 13 |
results:
|
| 14 |
- task:
|
| 15 |
name: Automatic Speech Recognition
|
| 16 |
type: automatic-speech-recognition
|
| 17 |
dataset:
|
| 18 |
+
name: google/fleurs te_in
|
| 19 |
+
type: google/fleurs
|
| 20 |
config: te_in
|
| 21 |
split: test
|
| 22 |
args: te_in
|
| 23 |
metrics:
|
| 24 |
- name: Wer
|
| 25 |
type: wer
|
| 26 |
+
value: 37.45436058603319
|
| 27 |
---
|
| 28 |
|
| 29 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 30 |
should probably proofread and complete it, then remove this comment. -->
|
| 31 |
|
| 32 |
+
# Breeze DSW Telugu - base
|
| 33 |
|
| 34 |
+
This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the google/fleurs te_in dataset.
|
| 35 |
It achieves the following results on the evaluation set:
|
| 36 |
+
- Loss: 0.3372
|
| 37 |
+
- Wer: 37.4544
|
| 38 |
|
| 39 |
## Model description
|
| 40 |
|
all_results.json
CHANGED
|
@@ -1,12 +1,12 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_loss":
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples_per_second": 0.
|
| 6 |
-
"eval_steps_per_second": 0.
|
| 7 |
-
"eval_wer":
|
| 8 |
-
"train_loss": 0.
|
| 9 |
-
"train_runtime":
|
| 10 |
-
"train_samples_per_second":
|
| 11 |
-
"train_steps_per_second":
|
| 12 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 27.03,
|
| 3 |
+
"eval_loss": 0.337158203125,
|
| 4 |
+
"eval_runtime": 1185.2527,
|
| 5 |
+
"eval_samples_per_second": 0.391,
|
| 6 |
+
"eval_steps_per_second": 0.024,
|
| 7 |
+
"eval_wer": 37.45436058603319,
|
| 8 |
+
"train_loss": 0.12189112710952758,
|
| 9 |
+
"train_runtime": 77912.2113,
|
| 10 |
+
"train_samples_per_second": 0.821,
|
| 11 |
+
"train_steps_per_second": 0.026
|
| 12 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_loss":
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples_per_second": 0.
|
| 6 |
-
"eval_steps_per_second": 0.
|
| 7 |
-
"eval_wer":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 27.03,
|
| 3 |
+
"eval_loss": 0.337158203125,
|
| 4 |
+
"eval_runtime": 1185.2527,
|
| 5 |
+
"eval_samples_per_second": 0.391,
|
| 6 |
+
"eval_steps_per_second": 0.024,
|
| 7 |
+
"eval_wer": 37.45436058603319
|
| 8 |
}
|
runs/Jan21_18-19-42_knight/events.out.tfevents.1705936715.knight.67737.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4f8b5df56938fc061a6fd81126f3d1afd770eee9287e7a042e7ab75b2db8b1c5
|
| 3 |
+
size 406
|
train_results.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples_per_second":
|
| 6 |
-
"train_steps_per_second":
|
| 7 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 27.03,
|
| 3 |
+
"train_loss": 0.12189112710952758,
|
| 4 |
+
"train_runtime": 77912.2113,
|
| 5 |
+
"train_samples_per_second": 0.821,
|
| 6 |
+
"train_steps_per_second": 0.026
|
| 7 |
}
|
trainer_state.json
CHANGED
|
@@ -1,78 +1,591 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint": "/cosmos/home/sp-operator/ai/training/models/huggingface/scripts/../breeze-listen-dsw-base-te/checkpoint-
|
| 4 |
-
"epoch":
|
| 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":
|
| 13 |
-
"learning_rate":
|
| 14 |
-
"loss": 1.
|
| 15 |
"step": 25
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch":
|
| 19 |
-
"learning_rate": 6.
|
| 20 |
-
"loss": 1.
|
| 21 |
"step": 50
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch":
|
| 25 |
-
"learning_rate": 6.
|
| 26 |
-
"loss": 0.
|
| 27 |
"step": 75
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch":
|
| 31 |
-
"learning_rate": 7.
|
| 32 |
-
"loss": 0.
|
| 33 |
"step": 100
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch":
|
| 37 |
-
"learning_rate": 7.
|
| 38 |
-
"loss": 0.
|
| 39 |
"step": 125
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch":
|
| 43 |
-
"learning_rate": 8.
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 150
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch":
|
| 49 |
-
"learning_rate": 8.
|
| 50 |
-
"loss": 0.
|
| 51 |
"step": 175
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch":
|
| 55 |
-
"learning_rate": 8.
|
| 56 |
-
"loss": 0.
|
| 57 |
"step": 200
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch":
|
| 61 |
-
"eval_loss": 0.
|
| 62 |
-
"eval_runtime":
|
| 63 |
-
"eval_samples_per_second": 0.
|
| 64 |
-
"eval_steps_per_second": 0.
|
| 65 |
-
"eval_wer":
|
| 66 |
"step": 200
|
| 67 |
},
|
| 68 |
{
|
| 69 |
-
"epoch":
|
| 70 |
-
"
|
| 71 |
-
"
|
| 72 |
-
"
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 76 |
}
|
| 77 |
],
|
| 78 |
"logging_steps": 25,
|
|
@@ -80,7 +593,7 @@
|
|
| 80 |
"num_input_tokens_seen": 0,
|
| 81 |
"num_train_epochs": 9223372036854775807,
|
| 82 |
"save_steps": 200,
|
| 83 |
-
"total_flos":
|
| 84 |
"train_batch_size": 32,
|
| 85 |
"trial_name": null,
|
| 86 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 37.45436058603319,
|
| 3 |
+
"best_model_checkpoint": "/cosmos/home/sp-operator/ai/training/models/huggingface/scripts/../breeze-listen-dsw-base-te/checkpoint-800",
|
| 4 |
+
"epoch": 27.028,
|
| 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": 5.0453611334320685e-06,
|
| 14 |
+
"loss": 1.6929,
|
| 15 |
"step": 25
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.03,
|
| 19 |
+
"learning_rate": 6.229195710491767e-06,
|
| 20 |
+
"loss": 1.4259,
|
| 21 |
"step": 50
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 1.0,
|
| 25 |
+
"learning_rate": 6.903829450223392e-06,
|
| 26 |
+
"loss": 0.9613,
|
| 27 |
"step": 75
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 1.01,
|
| 31 |
+
"learning_rate": 7.377725845391017e-06,
|
| 32 |
+
"loss": 0.5931,
|
| 33 |
"step": 100
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 1.03,
|
| 37 |
+
"learning_rate": 7.743343231239583e-06,
|
| 38 |
+
"loss": 0.4493,
|
| 39 |
"step": 125
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 2.0,
|
| 43 |
+
"learning_rate": 8.041073861170494e-06,
|
| 44 |
+
"loss": 0.3757,
|
| 45 |
"step": 150
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 2.02,
|
| 49 |
+
"learning_rate": 8.292222957399574e-06,
|
| 50 |
+
"loss": 0.3184,
|
| 51 |
"step": 175
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 2.03,
|
| 55 |
+
"learning_rate": 8.509413541357755e-06,
|
| 56 |
+
"loss": 0.2937,
|
| 57 |
"step": 200
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 2.03,
|
| 61 |
+
"eval_loss": 0.32373046875,
|
| 62 |
+
"eval_runtime": 1211.7719,
|
| 63 |
+
"eval_samples_per_second": 0.383,
|
| 64 |
+
"eval_steps_per_second": 0.024,
|
| 65 |
+
"eval_wer": 42.56135323750982,
|
| 66 |
"step": 200
|
| 67 |
},
|
| 68 |
{
|
| 69 |
+
"epoch": 3.0,
|
| 70 |
+
"learning_rate": 8.700744577655557e-06,
|
| 71 |
+
"loss": 0.2688,
|
| 72 |
+
"step": 225
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 3.02,
|
| 76 |
+
"learning_rate": 8.871723942761204e-06,
|
| 77 |
+
"loss": 0.2441,
|
| 78 |
+
"step": 250
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 3.03,
|
| 82 |
+
"learning_rate": 9.026267958246849e-06,
|
| 83 |
+
"loss": 0.2283,
|
| 84 |
+
"step": 275
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 4.01,
|
| 88 |
+
"learning_rate": 9.16726106663399e-06,
|
| 89 |
+
"loss": 0.2079,
|
| 90 |
+
"step": 300
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 4.02,
|
| 94 |
+
"learning_rate": 9.296889251455016e-06,
|
| 95 |
+
"loss": 0.1954,
|
| 96 |
+
"step": 325
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 4.03,
|
| 100 |
+
"learning_rate": 9.416848797368692e-06,
|
| 101 |
+
"loss": 0.1879,
|
| 102 |
+
"step": 350
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 5.01,
|
| 106 |
+
"learning_rate": 9.528482449516371e-06,
|
| 107 |
+
"loss": 0.1727,
|
| 108 |
+
"step": 375
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 5.02,
|
| 112 |
+
"learning_rate": 9.632871309784314e-06,
|
| 113 |
+
"loss": 0.1611,
|
| 114 |
+
"step": 400
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 5.02,
|
| 118 |
+
"eval_loss": 0.275634765625,
|
| 119 |
+
"eval_runtime": 1184.4646,
|
| 120 |
+
"eval_samples_per_second": 0.392,
|
| 121 |
+
"eval_steps_per_second": 0.024,
|
| 122 |
+
"eval_wer": 38.9148218329713,
|
| 123 |
+
"step": 400
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 5.03,
|
| 127 |
+
"learning_rate": 9.73089868785391e-06,
|
| 128 |
+
"loss": 0.1524,
|
| 129 |
+
"step": 425
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 6.01,
|
| 133 |
+
"learning_rate": 9.823295589572114e-06,
|
| 134 |
+
"loss": 0.1468,
|
| 135 |
+
"step": 450
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 6.02,
|
| 139 |
+
"learning_rate": 9.910673836465484e-06,
|
| 140 |
+
"loss": 0.1323,
|
| 141 |
+
"step": 475
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 6.03,
|
| 145 |
+
"learning_rate": 9.993550644973805e-06,
|
| 146 |
+
"loss": 0.1289,
|
| 147 |
+
"step": 500
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 7.01,
|
| 151 |
+
"learning_rate": 9.853333333333334e-06,
|
| 152 |
+
"loss": 0.1159,
|
| 153 |
+
"step": 525
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 7.02,
|
| 157 |
+
"learning_rate": 9.686666666666668e-06,
|
| 158 |
+
"loss": 0.1075,
|
| 159 |
+
"step": 550
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 7.04,
|
| 163 |
+
"learning_rate": 9.52e-06,
|
| 164 |
+
"loss": 0.1032,
|
| 165 |
+
"step": 575
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 8.01,
|
| 169 |
+
"learning_rate": 9.353333333333334e-06,
|
| 170 |
+
"loss": 0.0889,
|
| 171 |
+
"step": 600
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 8.01,
|
| 175 |
+
"eval_loss": 0.29296875,
|
| 176 |
+
"eval_runtime": 1184.8459,
|
| 177 |
+
"eval_samples_per_second": 0.392,
|
| 178 |
+
"eval_steps_per_second": 0.024,
|
| 179 |
+
"eval_wer": 38.11064380459398,
|
| 180 |
+
"step": 600
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 8.02,
|
| 184 |
+
"learning_rate": 9.186666666666666e-06,
|
| 185 |
+
"loss": 0.0861,
|
| 186 |
+
"step": 625
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 9.0,
|
| 190 |
+
"learning_rate": 9.020000000000002e-06,
|
| 191 |
+
"loss": 0.0806,
|
| 192 |
+
"step": 650
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 9.01,
|
| 196 |
+
"learning_rate": 8.853333333333334e-06,
|
| 197 |
+
"loss": 0.0681,
|
| 198 |
+
"step": 675
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 9.03,
|
| 202 |
+
"learning_rate": 8.686666666666668e-06,
|
| 203 |
+
"loss": 0.0708,
|
| 204 |
+
"step": 700
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 10.0,
|
| 208 |
+
"learning_rate": 8.52e-06,
|
| 209 |
+
"loss": 0.0626,
|
| 210 |
+
"step": 725
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 10.02,
|
| 214 |
+
"learning_rate": 8.353333333333335e-06,
|
| 215 |
+
"loss": 0.0552,
|
| 216 |
+
"step": 750
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 10.03,
|
| 220 |
+
"learning_rate": 8.186666666666667e-06,
|
| 221 |
+
"loss": 0.0517,
|
| 222 |
+
"step": 775
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 11.0,
|
| 226 |
+
"learning_rate": 8.020000000000001e-06,
|
| 227 |
+
"loss": 0.0456,
|
| 228 |
+
"step": 800
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 11.0,
|
| 232 |
+
"eval_loss": 0.337158203125,
|
| 233 |
+
"eval_runtime": 1186.6186,
|
| 234 |
+
"eval_samples_per_second": 0.391,
|
| 235 |
+
"eval_steps_per_second": 0.024,
|
| 236 |
+
"eval_wer": 37.45436058603319,
|
| 237 |
+
"step": 800
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 11.02,
|
| 241 |
+
"learning_rate": 7.853333333333333e-06,
|
| 242 |
+
"loss": 0.0404,
|
| 243 |
+
"step": 825
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 11.03,
|
| 247 |
+
"learning_rate": 7.686666666666667e-06,
|
| 248 |
+
"loss": 0.0407,
|
| 249 |
+
"step": 850
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 12.01,
|
| 253 |
+
"learning_rate": 7.520000000000001e-06,
|
| 254 |
+
"loss": 0.0369,
|
| 255 |
+
"step": 875
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 12.02,
|
| 259 |
+
"learning_rate": 7.353333333333334e-06,
|
| 260 |
+
"loss": 0.0303,
|
| 261 |
+
"step": 900
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 12.03,
|
| 265 |
+
"learning_rate": 7.186666666666668e-06,
|
| 266 |
+
"loss": 0.0289,
|
| 267 |
+
"step": 925
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 13.01,
|
| 271 |
+
"learning_rate": 7.0200000000000006e-06,
|
| 272 |
+
"loss": 0.0268,
|
| 273 |
+
"step": 950
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 13.02,
|
| 277 |
+
"learning_rate": 6.853333333333334e-06,
|
| 278 |
+
"loss": 0.0238,
|
| 279 |
+
"step": 975
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 13.03,
|
| 283 |
+
"learning_rate": 6.6866666666666665e-06,
|
| 284 |
+
"loss": 0.0229,
|
| 285 |
+
"step": 1000
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 13.03,
|
| 289 |
+
"eval_loss": 0.398193359375,
|
| 290 |
+
"eval_runtime": 1184.9328,
|
| 291 |
+
"eval_samples_per_second": 0.392,
|
| 292 |
+
"eval_steps_per_second": 0.024,
|
| 293 |
+
"eval_wer": 37.92577529232333,
|
| 294 |
+
"step": 1000
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 14.01,
|
| 298 |
+
"learning_rate": 6.533333333333334e-06,
|
| 299 |
+
"loss": 0.0195,
|
| 300 |
+
"step": 1025
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 14.02,
|
| 304 |
+
"learning_rate": 6.366666666666668e-06,
|
| 305 |
+
"loss": 0.0177,
|
| 306 |
+
"step": 1050
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 14.03,
|
| 310 |
+
"learning_rate": 6.200000000000001e-06,
|
| 311 |
+
"loss": 0.0176,
|
| 312 |
+
"step": 1075
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 15.01,
|
| 316 |
+
"learning_rate": 6.033333333333335e-06,
|
| 317 |
+
"loss": 0.0149,
|
| 318 |
+
"step": 1100
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 15.02,
|
| 322 |
+
"learning_rate": 5.8666666666666675e-06,
|
| 323 |
+
"loss": 0.014,
|
| 324 |
+
"step": 1125
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 15.04,
|
| 328 |
+
"learning_rate": 5.7e-06,
|
| 329 |
+
"loss": 0.0124,
|
| 330 |
+
"step": 1150
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 16.01,
|
| 334 |
+
"learning_rate": 5.533333333333334e-06,
|
| 335 |
+
"loss": 0.0095,
|
| 336 |
+
"step": 1175
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 16.02,
|
| 340 |
+
"learning_rate": 5.366666666666666e-06,
|
| 341 |
+
"loss": 0.0103,
|
| 342 |
+
"step": 1200
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 16.02,
|
| 346 |
+
"eval_loss": 0.447265625,
|
| 347 |
+
"eval_runtime": 1183.5656,
|
| 348 |
+
"eval_samples_per_second": 0.392,
|
| 349 |
+
"eval_steps_per_second": 0.025,
|
| 350 |
+
"eval_wer": 38.26778204002403,
|
| 351 |
+
"step": 1200
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 17.0,
|
| 355 |
+
"learning_rate": 5.2e-06,
|
| 356 |
+
"loss": 0.01,
|
| 357 |
+
"step": 1225
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 17.01,
|
| 361 |
+
"learning_rate": 5.033333333333333e-06,
|
| 362 |
+
"loss": 0.0081,
|
| 363 |
+
"step": 1250
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 17.03,
|
| 367 |
+
"learning_rate": 4.866666666666667e-06,
|
| 368 |
+
"loss": 0.0078,
|
| 369 |
+
"step": 1275
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 18.0,
|
| 373 |
+
"learning_rate": 4.7e-06,
|
| 374 |
+
"loss": 0.0073,
|
| 375 |
+
"step": 1300
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 18.01,
|
| 379 |
+
"learning_rate": 4.533333333333334e-06,
|
| 380 |
+
"loss": 0.0053,
|
| 381 |
+
"step": 1325
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 18.03,
|
| 385 |
+
"learning_rate": 4.366666666666667e-06,
|
| 386 |
+
"loss": 0.0059,
|
| 387 |
+
"step": 1350
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 19.0,
|
| 391 |
+
"learning_rate": 4.2000000000000004e-06,
|
| 392 |
+
"loss": 0.0051,
|
| 393 |
+
"step": 1375
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 19.02,
|
| 397 |
+
"learning_rate": 4.033333333333333e-06,
|
| 398 |
+
"loss": 0.0042,
|
| 399 |
+
"step": 1400
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 19.02,
|
| 403 |
+
"eval_loss": 0.483642578125,
|
| 404 |
+
"eval_runtime": 1188.6164,
|
| 405 |
+
"eval_samples_per_second": 0.39,
|
| 406 |
+
"eval_steps_per_second": 0.024,
|
| 407 |
+
"eval_wer": 37.89804501548274,
|
| 408 |
+
"step": 1400
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 19.03,
|
| 412 |
+
"learning_rate": 3.866666666666667e-06,
|
| 413 |
+
"loss": 0.0043,
|
| 414 |
+
"step": 1425
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 20.0,
|
| 418 |
+
"learning_rate": 3.7e-06,
|
| 419 |
+
"loss": 0.004,
|
| 420 |
+
"step": 1450
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 20.02,
|
| 424 |
+
"learning_rate": 3.5333333333333335e-06,
|
| 425 |
+
"loss": 0.0034,
|
| 426 |
+
"step": 1475
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 20.03,
|
| 430 |
+
"learning_rate": 3.366666666666667e-06,
|
| 431 |
+
"loss": 0.003,
|
| 432 |
+
"step": 1500
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 21.01,
|
| 436 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 437 |
+
"loss": 0.0032,
|
| 438 |
+
"step": 1525
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 21.02,
|
| 442 |
+
"learning_rate": 3.0333333333333337e-06,
|
| 443 |
+
"loss": 0.0027,
|
| 444 |
+
"step": 1550
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 21.03,
|
| 448 |
+
"learning_rate": 2.866666666666667e-06,
|
| 449 |
+
"loss": 0.0029,
|
| 450 |
+
"step": 1575
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 22.01,
|
| 454 |
+
"learning_rate": 2.7000000000000004e-06,
|
| 455 |
+
"loss": 0.0025,
|
| 456 |
+
"step": 1600
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 22.01,
|
| 460 |
+
"eval_loss": 0.50830078125,
|
| 461 |
+
"eval_runtime": 1190.0425,
|
| 462 |
+
"eval_samples_per_second": 0.39,
|
| 463 |
+
"eval_steps_per_second": 0.024,
|
| 464 |
+
"eval_wer": 37.731663354439156,
|
| 465 |
+
"step": 1600
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 22.02,
|
| 469 |
+
"learning_rate": 2.5333333333333338e-06,
|
| 470 |
+
"loss": 0.0023,
|
| 471 |
+
"step": 1625
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 22.03,
|
| 475 |
+
"learning_rate": 2.3666666666666667e-06,
|
| 476 |
+
"loss": 0.0024,
|
| 477 |
+
"step": 1650
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 23.01,
|
| 481 |
+
"learning_rate": 2.2e-06,
|
| 482 |
+
"loss": 0.0021,
|
| 483 |
+
"step": 1675
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 23.02,
|
| 487 |
+
"learning_rate": 2.0333333333333335e-06,
|
| 488 |
+
"loss": 0.0022,
|
| 489 |
+
"step": 1700
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 23.03,
|
| 493 |
+
"learning_rate": 1.8666666666666669e-06,
|
| 494 |
+
"loss": 0.0022,
|
| 495 |
+
"step": 1725
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 24.01,
|
| 499 |
+
"learning_rate": 1.7000000000000002e-06,
|
| 500 |
+
"loss": 0.0021,
|
| 501 |
+
"step": 1750
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 24.02,
|
| 505 |
+
"learning_rate": 1.5333333333333334e-06,
|
| 506 |
+
"loss": 0.0018,
|
| 507 |
+
"step": 1775
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 24.04,
|
| 511 |
+
"learning_rate": 1.3666666666666668e-06,
|
| 512 |
+
"loss": 0.002,
|
| 513 |
+
"step": 1800
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 24.04,
|
| 517 |
+
"eval_loss": 0.52197265625,
|
| 518 |
+
"eval_runtime": 1185.5264,
|
| 519 |
+
"eval_samples_per_second": 0.391,
|
| 520 |
+
"eval_steps_per_second": 0.024,
|
| 521 |
+
"eval_wer": 37.80098904654065,
|
| 522 |
+
"step": 1800
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 25.01,
|
| 526 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 527 |
+
"loss": 0.0018,
|
| 528 |
+
"step": 1825
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 25.02,
|
| 532 |
+
"learning_rate": 1.0333333333333333e-06,
|
| 533 |
+
"loss": 0.002,
|
| 534 |
+
"step": 1850
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 26.0,
|
| 538 |
+
"learning_rate": 8.666666666666668e-07,
|
| 539 |
+
"loss": 0.0019,
|
| 540 |
+
"step": 1875
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 26.01,
|
| 544 |
+
"learning_rate": 7.000000000000001e-07,
|
| 545 |
+
"loss": 0.0018,
|
| 546 |
+
"step": 1900
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 26.03,
|
| 550 |
+
"learning_rate": 5.333333333333335e-07,
|
| 551 |
+
"loss": 0.0018,
|
| 552 |
+
"step": 1925
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 27.0,
|
| 556 |
+
"learning_rate": 3.666666666666667e-07,
|
| 557 |
+
"loss": 0.0018,
|
| 558 |
+
"step": 1950
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 27.02,
|
| 562 |
+
"learning_rate": 2.0000000000000002e-07,
|
| 563 |
+
"loss": 0.0017,
|
| 564 |
+
"step": 1975
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 27.03,
|
| 568 |
+
"learning_rate": 3.333333333333334e-08,
|
| 569 |
+
"loss": 0.0018,
|
| 570 |
+
"step": 2000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 27.03,
|
| 574 |
+
"eval_loss": 0.52685546875,
|
| 575 |
+
"eval_runtime": 1188.9989,
|
| 576 |
+
"eval_samples_per_second": 0.39,
|
| 577 |
+
"eval_steps_per_second": 0.024,
|
| 578 |
+
"eval_wer": 37.9026667282895,
|
| 579 |
+
"step": 2000
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 27.03,
|
| 583 |
+
"step": 2000,
|
| 584 |
+
"total_flos": 4.1020056982698066e+18,
|
| 585 |
+
"train_loss": 0.12189112710952758,
|
| 586 |
+
"train_runtime": 77912.2113,
|
| 587 |
+
"train_samples_per_second": 0.821,
|
| 588 |
+
"train_steps_per_second": 0.026
|
| 589 |
}
|
| 590 |
],
|
| 591 |
"logging_steps": 25,
|
|
|
|
| 593 |
"num_input_tokens_seen": 0,
|
| 594 |
"num_train_epochs": 9223372036854775807,
|
| 595 |
"save_steps": 200,
|
| 596 |
+
"total_flos": 4.1020056982698066e+18,
|
| 597 |
"train_batch_size": 32,
|
| 598 |
"trial_name": null,
|
| 599 |
"trial_params": null
|