End of training
Browse files- README.md +6 -3
- adapter.toi.safetensors +3 -0
- all_results.json +15 -0
- eval_results.json +9 -0
- runs/Jan03_16-49-18_srvrocgpu011.uct.ac.za/events.out.tfevents.1735918718.srvrocgpu011.uct.ac.za +3 -0
- train_results.json +9 -0
- trainer_state.json +563 -0
README.md
CHANGED
|
@@ -3,6 +3,9 @@ library_name: transformers
|
|
| 3 |
license: cc-by-nc-4.0
|
| 4 |
base_model: facebook/mms-1b-all
|
| 5 |
tags:
|
|
|
|
|
|
|
|
|
|
| 6 |
- generated_from_trainer
|
| 7 |
metrics:
|
| 8 |
- wer
|
|
@@ -16,10 +19,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
# mms-1b-toigen-female-model
|
| 18 |
|
| 19 |
-
This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
-
- Loss: 0.
|
| 22 |
-
- Wer: 0.
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
|
|
| 3 |
license: cc-by-nc-4.0
|
| 4 |
base_model: facebook/mms-1b-all
|
| 5 |
tags:
|
| 6 |
+
- automatic-speech-recognition
|
| 7 |
+
- toigen
|
| 8 |
+
- mms
|
| 9 |
- generated_from_trainer
|
| 10 |
metrics:
|
| 11 |
- wer
|
|
|
|
| 19 |
|
| 20 |
# mms-1b-toigen-female-model
|
| 21 |
|
| 22 |
+
This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on the TOIGEN - TOI dataset.
|
| 23 |
It achieves the following results on the evaluation set:
|
| 24 |
+
- Loss: 0.2197
|
| 25 |
+
- Wer: 0.3590
|
| 26 |
|
| 27 |
## Model description
|
| 28 |
|
adapter.toi.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e48b9fcaee211f79f6dfbb277bc1ee7f8f02676275be767c27220679dc7c94bd
|
| 3 |
+
size 8793408
|
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 12.85140562248996,
|
| 3 |
+
"eval_loss": 0.21970169246196747,
|
| 4 |
+
"eval_runtime": 18.2362,
|
| 5 |
+
"eval_samples": 219,
|
| 6 |
+
"eval_samples_per_second": 12.009,
|
| 7 |
+
"eval_steps_per_second": 3.016,
|
| 8 |
+
"eval_wer": 0.358953722334004,
|
| 9 |
+
"total_flos": 1.020661828684297e+19,
|
| 10 |
+
"train_loss": 0.5673185133934021,
|
| 11 |
+
"train_runtime": 2727.6394,
|
| 12 |
+
"train_samples": 993,
|
| 13 |
+
"train_samples_per_second": 10.922,
|
| 14 |
+
"train_steps_per_second": 2.739
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 12.85140562248996,
|
| 3 |
+
"eval_loss": 0.21970169246196747,
|
| 4 |
+
"eval_runtime": 18.2362,
|
| 5 |
+
"eval_samples": 219,
|
| 6 |
+
"eval_samples_per_second": 12.009,
|
| 7 |
+
"eval_steps_per_second": 3.016,
|
| 8 |
+
"eval_wer": 0.358953722334004
|
| 9 |
+
}
|
runs/Jan03_16-49-18_srvrocgpu011.uct.ac.za/events.out.tfevents.1735918718.srvrocgpu011.uct.ac.za
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5d9b2d7deec0041db238923ef501a57c8465cc501436040f2e7e6a4f992a2fcc
|
| 3 |
+
size 40
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 12.85140562248996,
|
| 3 |
+
"total_flos": 1.020661828684297e+19,
|
| 4 |
+
"train_loss": 0.5673185133934021,
|
| 5 |
+
"train_runtime": 2727.6394,
|
| 6 |
+
"train_samples": 993,
|
| 7 |
+
"train_samples_per_second": 10.922,
|
| 8 |
+
"train_steps_per_second": 2.739
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,563 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.21969997882843018,
|
| 3 |
+
"best_model_checkpoint": "/scratch/skscla001/speech/results/mms-1b-toigen-female-model/checkpoint-2800",
|
| 4 |
+
"epoch": 12.85140562248996,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 3200,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.40160642570281124,
|
| 13 |
+
"grad_norm": 3.196812391281128,
|
| 14 |
+
"learning_rate": 0.000285,
|
| 15 |
+
"loss": 6.8474,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.40160642570281124,
|
| 20 |
+
"eval_loss": 3.63849139213562,
|
| 21 |
+
"eval_runtime": 18.3421,
|
| 22 |
+
"eval_samples_per_second": 11.94,
|
| 23 |
+
"eval_steps_per_second": 2.999,
|
| 24 |
+
"eval_wer": 0.9951710261569416,
|
| 25 |
+
"step": 100
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.8032128514056225,
|
| 29 |
+
"grad_norm": 2.898397922515869,
|
| 30 |
+
"learning_rate": 0.0002961329715061058,
|
| 31 |
+
"loss": 2.2373,
|
| 32 |
+
"step": 200
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.8032128514056225,
|
| 36 |
+
"eval_loss": 0.5180636644363403,
|
| 37 |
+
"eval_runtime": 18.2272,
|
| 38 |
+
"eval_samples_per_second": 12.015,
|
| 39 |
+
"eval_steps_per_second": 3.017,
|
| 40 |
+
"eval_wer": 0.6289738430583501,
|
| 41 |
+
"step": 200
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 1.2048192771084336,
|
| 45 |
+
"grad_norm": 2.2569992542266846,
|
| 46 |
+
"learning_rate": 0.0002920624151967435,
|
| 47 |
+
"loss": 0.6093,
|
| 48 |
+
"step": 300
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 1.2048192771084336,
|
| 52 |
+
"eval_loss": 0.3648887276649475,
|
| 53 |
+
"eval_runtime": 18.2582,
|
| 54 |
+
"eval_samples_per_second": 11.995,
|
| 55 |
+
"eval_steps_per_second": 3.012,
|
| 56 |
+
"eval_wer": 0.5187122736418511,
|
| 57 |
+
"step": 300
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 1.606425702811245,
|
| 61 |
+
"grad_norm": 6.299871921539307,
|
| 62 |
+
"learning_rate": 0.0002879918588873813,
|
| 63 |
+
"loss": 0.4813,
|
| 64 |
+
"step": 400
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 1.606425702811245,
|
| 68 |
+
"eval_loss": 0.3151092231273651,
|
| 69 |
+
"eval_runtime": 18.1073,
|
| 70 |
+
"eval_samples_per_second": 12.095,
|
| 71 |
+
"eval_steps_per_second": 3.037,
|
| 72 |
+
"eval_wer": 0.4909456740442656,
|
| 73 |
+
"step": 400
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 2.0080321285140563,
|
| 77 |
+
"grad_norm": 1.1055076122283936,
|
| 78 |
+
"learning_rate": 0.000283921302578019,
|
| 79 |
+
"loss": 0.3843,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 2.0080321285140563,
|
| 84 |
+
"eval_loss": 0.299377977848053,
|
| 85 |
+
"eval_runtime": 18.2782,
|
| 86 |
+
"eval_samples_per_second": 11.981,
|
| 87 |
+
"eval_steps_per_second": 3.009,
|
| 88 |
+
"eval_wer": 0.4567404426559356,
|
| 89 |
+
"step": 500
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 2.4096385542168672,
|
| 93 |
+
"grad_norm": 0.7269560098648071,
|
| 94 |
+
"learning_rate": 0.0002798507462686567,
|
| 95 |
+
"loss": 0.3805,
|
| 96 |
+
"step": 600
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 2.4096385542168672,
|
| 100 |
+
"eval_loss": 0.2818037271499634,
|
| 101 |
+
"eval_runtime": 18.2144,
|
| 102 |
+
"eval_samples_per_second": 12.023,
|
| 103 |
+
"eval_steps_per_second": 3.02,
|
| 104 |
+
"eval_wer": 0.4422535211267606,
|
| 105 |
+
"step": 600
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 2.8112449799196786,
|
| 109 |
+
"grad_norm": 0.9139028191566467,
|
| 110 |
+
"learning_rate": 0.0002757801899592944,
|
| 111 |
+
"loss": 0.3591,
|
| 112 |
+
"step": 700
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 2.8112449799196786,
|
| 116 |
+
"eval_loss": 0.2810930609703064,
|
| 117 |
+
"eval_runtime": 18.2024,
|
| 118 |
+
"eval_samples_per_second": 12.031,
|
| 119 |
+
"eval_steps_per_second": 3.022,
|
| 120 |
+
"eval_wer": 0.4402414486921529,
|
| 121 |
+
"step": 700
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 3.21285140562249,
|
| 125 |
+
"grad_norm": 1.1874170303344727,
|
| 126 |
+
"learning_rate": 0.00027170963364993214,
|
| 127 |
+
"loss": 0.3164,
|
| 128 |
+
"step": 800
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 3.21285140562249,
|
| 132 |
+
"eval_loss": 0.26876774430274963,
|
| 133 |
+
"eval_runtime": 18.1602,
|
| 134 |
+
"eval_samples_per_second": 12.059,
|
| 135 |
+
"eval_steps_per_second": 3.029,
|
| 136 |
+
"eval_wer": 0.4104627766599598,
|
| 137 |
+
"step": 800
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 3.6144578313253013,
|
| 141 |
+
"grad_norm": 1.2351394891738892,
|
| 142 |
+
"learning_rate": 0.00026763907734056984,
|
| 143 |
+
"loss": 0.3351,
|
| 144 |
+
"step": 900
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 3.6144578313253013,
|
| 148 |
+
"eval_loss": 0.2569175362586975,
|
| 149 |
+
"eval_runtime": 18.988,
|
| 150 |
+
"eval_samples_per_second": 11.534,
|
| 151 |
+
"eval_steps_per_second": 2.897,
|
| 152 |
+
"eval_wer": 0.40603621730382294,
|
| 153 |
+
"step": 900
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 4.016064257028113,
|
| 157 |
+
"grad_norm": 1.2486017942428589,
|
| 158 |
+
"learning_rate": 0.0002635685210312076,
|
| 159 |
+
"loss": 0.3442,
|
| 160 |
+
"step": 1000
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 4.016064257028113,
|
| 164 |
+
"eval_loss": 0.26386773586273193,
|
| 165 |
+
"eval_runtime": 18.2707,
|
| 166 |
+
"eval_samples_per_second": 11.986,
|
| 167 |
+
"eval_steps_per_second": 3.01,
|
| 168 |
+
"eval_wer": 0.4,
|
| 169 |
+
"step": 1000
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 4.417670682730924,
|
| 173 |
+
"grad_norm": 0.8581802845001221,
|
| 174 |
+
"learning_rate": 0.0002594979647218453,
|
| 175 |
+
"loss": 0.2996,
|
| 176 |
+
"step": 1100
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 4.417670682730924,
|
| 180 |
+
"eval_loss": 0.2581425905227661,
|
| 181 |
+
"eval_runtime": 18.2961,
|
| 182 |
+
"eval_samples_per_second": 11.97,
|
| 183 |
+
"eval_steps_per_second": 3.006,
|
| 184 |
+
"eval_wer": 0.4084507042253521,
|
| 185 |
+
"step": 1100
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 4.8192771084337345,
|
| 189 |
+
"grad_norm": 0.9432417750358582,
|
| 190 |
+
"learning_rate": 0.000255427408412483,
|
| 191 |
+
"loss": 0.3196,
|
| 192 |
+
"step": 1200
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 4.8192771084337345,
|
| 196 |
+
"eval_loss": 0.2443518042564392,
|
| 197 |
+
"eval_runtime": 18.444,
|
| 198 |
+
"eval_samples_per_second": 11.874,
|
| 199 |
+
"eval_steps_per_second": 2.982,
|
| 200 |
+
"eval_wer": 0.386317907444668,
|
| 201 |
+
"step": 1200
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 5.220883534136546,
|
| 205 |
+
"grad_norm": 2.045851230621338,
|
| 206 |
+
"learning_rate": 0.00025135685210312075,
|
| 207 |
+
"loss": 0.3147,
|
| 208 |
+
"step": 1300
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 5.220883534136546,
|
| 212 |
+
"eval_loss": 0.2449689507484436,
|
| 213 |
+
"eval_runtime": 18.3447,
|
| 214 |
+
"eval_samples_per_second": 11.938,
|
| 215 |
+
"eval_steps_per_second": 2.998,
|
| 216 |
+
"eval_wer": 0.386317907444668,
|
| 217 |
+
"step": 1300
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 5.622489959839357,
|
| 221 |
+
"grad_norm": 0.7233336567878723,
|
| 222 |
+
"learning_rate": 0.00024728629579375845,
|
| 223 |
+
"loss": 0.2936,
|
| 224 |
+
"step": 1400
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 5.622489959839357,
|
| 228 |
+
"eval_loss": 0.24141426384449005,
|
| 229 |
+
"eval_runtime": 18.3461,
|
| 230 |
+
"eval_samples_per_second": 11.937,
|
| 231 |
+
"eval_steps_per_second": 2.998,
|
| 232 |
+
"eval_wer": 0.38551307847082494,
|
| 233 |
+
"step": 1400
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 6.024096385542169,
|
| 237 |
+
"grad_norm": 1.2751613855361938,
|
| 238 |
+
"learning_rate": 0.00024321573948439618,
|
| 239 |
+
"loss": 0.2894,
|
| 240 |
+
"step": 1500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 6.024096385542169,
|
| 244 |
+
"eval_loss": 0.23630352318286896,
|
| 245 |
+
"eval_runtime": 18.2602,
|
| 246 |
+
"eval_samples_per_second": 11.993,
|
| 247 |
+
"eval_steps_per_second": 3.012,
|
| 248 |
+
"eval_wer": 0.3722334004024145,
|
| 249 |
+
"step": 1500
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 6.42570281124498,
|
| 253 |
+
"grad_norm": 0.9754641056060791,
|
| 254 |
+
"learning_rate": 0.00023914518317503389,
|
| 255 |
+
"loss": 0.2772,
|
| 256 |
+
"step": 1600
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 6.42570281124498,
|
| 260 |
+
"eval_loss": 0.24517948925495148,
|
| 261 |
+
"eval_runtime": 18.2481,
|
| 262 |
+
"eval_samples_per_second": 12.001,
|
| 263 |
+
"eval_steps_per_second": 3.014,
|
| 264 |
+
"eval_wer": 0.3843058350100604,
|
| 265 |
+
"step": 1600
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 6.827309236947791,
|
| 269 |
+
"grad_norm": 5.323614120483398,
|
| 270 |
+
"learning_rate": 0.00023507462686567164,
|
| 271 |
+
"loss": 0.2565,
|
| 272 |
+
"step": 1700
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 6.827309236947791,
|
| 276 |
+
"eval_loss": 0.23570258915424347,
|
| 277 |
+
"eval_runtime": 18.3512,
|
| 278 |
+
"eval_samples_per_second": 11.934,
|
| 279 |
+
"eval_steps_per_second": 2.997,
|
| 280 |
+
"eval_wer": 0.3758551307847082,
|
| 281 |
+
"step": 1700
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 7.228915662650603,
|
| 285 |
+
"grad_norm": 1.3097056150436401,
|
| 286 |
+
"learning_rate": 0.00023100407055630934,
|
| 287 |
+
"loss": 0.3033,
|
| 288 |
+
"step": 1800
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 7.228915662650603,
|
| 292 |
+
"eval_loss": 0.23575641214847565,
|
| 293 |
+
"eval_runtime": 18.2935,
|
| 294 |
+
"eval_samples_per_second": 11.971,
|
| 295 |
+
"eval_steps_per_second": 3.007,
|
| 296 |
+
"eval_wer": 0.378672032193159,
|
| 297 |
+
"step": 1800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 7.6305220883534135,
|
| 301 |
+
"grad_norm": 0.949272871017456,
|
| 302 |
+
"learning_rate": 0.00022693351424694707,
|
| 303 |
+
"loss": 0.2651,
|
| 304 |
+
"step": 1900
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 7.6305220883534135,
|
| 308 |
+
"eval_loss": 0.23492006957530975,
|
| 309 |
+
"eval_runtime": 18.1407,
|
| 310 |
+
"eval_samples_per_second": 12.072,
|
| 311 |
+
"eval_steps_per_second": 3.032,
|
| 312 |
+
"eval_wer": 0.36338028169014086,
|
| 313 |
+
"step": 1900
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 8.032128514056225,
|
| 317 |
+
"grad_norm": 0.6142582893371582,
|
| 318 |
+
"learning_rate": 0.00022286295793758477,
|
| 319 |
+
"loss": 0.2912,
|
| 320 |
+
"step": 2000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 8.032128514056225,
|
| 324 |
+
"eval_loss": 0.23053765296936035,
|
| 325 |
+
"eval_runtime": 18.1517,
|
| 326 |
+
"eval_samples_per_second": 12.065,
|
| 327 |
+
"eval_steps_per_second": 3.03,
|
| 328 |
+
"eval_wer": 0.36579476861167004,
|
| 329 |
+
"step": 2000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 8.433734939759036,
|
| 333 |
+
"grad_norm": 0.8553833365440369,
|
| 334 |
+
"learning_rate": 0.0002187924016282225,
|
| 335 |
+
"loss": 0.2554,
|
| 336 |
+
"step": 2100
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 8.433734939759036,
|
| 340 |
+
"eval_loss": 0.23033367097377777,
|
| 341 |
+
"eval_runtime": 18.3056,
|
| 342 |
+
"eval_samples_per_second": 11.964,
|
| 343 |
+
"eval_steps_per_second": 3.005,
|
| 344 |
+
"eval_wer": 0.36780684104627764,
|
| 345 |
+
"step": 2100
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 8.835341365461847,
|
| 349 |
+
"grad_norm": 0.753628134727478,
|
| 350 |
+
"learning_rate": 0.00021472184531886023,
|
| 351 |
+
"loss": 0.2535,
|
| 352 |
+
"step": 2200
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 8.835341365461847,
|
| 356 |
+
"eval_loss": 0.22692957520484924,
|
| 357 |
+
"eval_runtime": 18.3168,
|
| 358 |
+
"eval_samples_per_second": 11.956,
|
| 359 |
+
"eval_steps_per_second": 3.003,
|
| 360 |
+
"eval_wer": 0.3553319919517103,
|
| 361 |
+
"step": 2200
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 9.236947791164658,
|
| 365 |
+
"grad_norm": 1.5686020851135254,
|
| 366 |
+
"learning_rate": 0.00021065128900949793,
|
| 367 |
+
"loss": 0.2368,
|
| 368 |
+
"step": 2300
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 9.236947791164658,
|
| 372 |
+
"eval_loss": 0.22882795333862305,
|
| 373 |
+
"eval_runtime": 18.1737,
|
| 374 |
+
"eval_samples_per_second": 12.05,
|
| 375 |
+
"eval_steps_per_second": 3.026,
|
| 376 |
+
"eval_wer": 0.3609657947686117,
|
| 377 |
+
"step": 2300
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 9.638554216867469,
|
| 381 |
+
"grad_norm": 3.3497209548950195,
|
| 382 |
+
"learning_rate": 0.0002065807327001357,
|
| 383 |
+
"loss": 0.2491,
|
| 384 |
+
"step": 2400
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 9.638554216867469,
|
| 388 |
+
"eval_loss": 0.22438912093639374,
|
| 389 |
+
"eval_runtime": 18.155,
|
| 390 |
+
"eval_samples_per_second": 12.063,
|
| 391 |
+
"eval_steps_per_second": 3.029,
|
| 392 |
+
"eval_wer": 0.36861167002012074,
|
| 393 |
+
"step": 2400
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 10.040160642570282,
|
| 397 |
+
"grad_norm": 0.8699430823326111,
|
| 398 |
+
"learning_rate": 0.0002025101763907734,
|
| 399 |
+
"loss": 0.2678,
|
| 400 |
+
"step": 2500
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 10.040160642570282,
|
| 404 |
+
"eval_loss": 0.22885793447494507,
|
| 405 |
+
"eval_runtime": 18.3016,
|
| 406 |
+
"eval_samples_per_second": 11.966,
|
| 407 |
+
"eval_steps_per_second": 3.005,
|
| 408 |
+
"eval_wer": 0.3561368209255533,
|
| 409 |
+
"step": 2500
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 10.441767068273093,
|
| 413 |
+
"grad_norm": 1.7827962636947632,
|
| 414 |
+
"learning_rate": 0.00019843962008141112,
|
| 415 |
+
"loss": 0.2247,
|
| 416 |
+
"step": 2600
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 10.441767068273093,
|
| 420 |
+
"eval_loss": 0.22344833612442017,
|
| 421 |
+
"eval_runtime": 18.3786,
|
| 422 |
+
"eval_samples_per_second": 11.916,
|
| 423 |
+
"eval_steps_per_second": 2.993,
|
| 424 |
+
"eval_wer": 0.3613682092555332,
|
| 425 |
+
"step": 2600
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 10.843373493975903,
|
| 429 |
+
"grad_norm": 0.6483604907989502,
|
| 430 |
+
"learning_rate": 0.00019436906377204882,
|
| 431 |
+
"loss": 0.2773,
|
| 432 |
+
"step": 2700
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 10.843373493975903,
|
| 436 |
+
"eval_loss": 0.224824920296669,
|
| 437 |
+
"eval_runtime": 18.2411,
|
| 438 |
+
"eval_samples_per_second": 12.006,
|
| 439 |
+
"eval_steps_per_second": 3.015,
|
| 440 |
+
"eval_wer": 0.35412474849094566,
|
| 441 |
+
"step": 2700
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 11.244979919678714,
|
| 445 |
+
"grad_norm": 0.5516765713691711,
|
| 446 |
+
"learning_rate": 0.00019029850746268658,
|
| 447 |
+
"loss": 0.2162,
|
| 448 |
+
"step": 2800
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 11.244979919678714,
|
| 452 |
+
"eval_loss": 0.21969997882843018,
|
| 453 |
+
"eval_runtime": 18.2418,
|
| 454 |
+
"eval_samples_per_second": 12.005,
|
| 455 |
+
"eval_steps_per_second": 3.015,
|
| 456 |
+
"eval_wer": 0.358953722334004,
|
| 457 |
+
"step": 2800
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 11.646586345381525,
|
| 461 |
+
"grad_norm": 0.925298273563385,
|
| 462 |
+
"learning_rate": 0.00018622795115332428,
|
| 463 |
+
"loss": 0.2526,
|
| 464 |
+
"step": 2900
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 11.646586345381525,
|
| 468 |
+
"eval_loss": 0.22084331512451172,
|
| 469 |
+
"eval_runtime": 18.3085,
|
| 470 |
+
"eval_samples_per_second": 11.962,
|
| 471 |
+
"eval_steps_per_second": 3.004,
|
| 472 |
+
"eval_wer": 0.352112676056338,
|
| 473 |
+
"step": 2900
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 12.048192771084338,
|
| 477 |
+
"grad_norm": 0.8293038606643677,
|
| 478 |
+
"learning_rate": 0.00018215739484396198,
|
| 479 |
+
"loss": 0.2521,
|
| 480 |
+
"step": 3000
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 12.048192771084338,
|
| 484 |
+
"eval_loss": 0.22298946976661682,
|
| 485 |
+
"eval_runtime": 18.3594,
|
| 486 |
+
"eval_samples_per_second": 11.928,
|
| 487 |
+
"eval_steps_per_second": 2.996,
|
| 488 |
+
"eval_wer": 0.3702213279678068,
|
| 489 |
+
"step": 3000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 12.449799196787149,
|
| 493 |
+
"grad_norm": 1.355546474456787,
|
| 494 |
+
"learning_rate": 0.0001780868385345997,
|
| 495 |
+
"loss": 0.2218,
|
| 496 |
+
"step": 3100
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 12.449799196787149,
|
| 500 |
+
"eval_loss": 0.22078920900821686,
|
| 501 |
+
"eval_runtime": 18.3015,
|
| 502 |
+
"eval_samples_per_second": 11.966,
|
| 503 |
+
"eval_steps_per_second": 3.005,
|
| 504 |
+
"eval_wer": 0.3529175050301811,
|
| 505 |
+
"step": 3100
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 12.85140562248996,
|
| 509 |
+
"grad_norm": 3.3555612564086914,
|
| 510 |
+
"learning_rate": 0.0001740162822252374,
|
| 511 |
+
"loss": 0.2419,
|
| 512 |
+
"step": 3200
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 12.85140562248996,
|
| 516 |
+
"eval_loss": 0.2204364389181137,
|
| 517 |
+
"eval_runtime": 18.3584,
|
| 518 |
+
"eval_samples_per_second": 11.929,
|
| 519 |
+
"eval_steps_per_second": 2.996,
|
| 520 |
+
"eval_wer": 0.3517102615694165,
|
| 521 |
+
"step": 3200
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 12.85140562248996,
|
| 525 |
+
"step": 3200,
|
| 526 |
+
"total_flos": 1.020661828684297e+19,
|
| 527 |
+
"train_loss": 0.5673185133934021,
|
| 528 |
+
"train_runtime": 2727.6394,
|
| 529 |
+
"train_samples_per_second": 10.922,
|
| 530 |
+
"train_steps_per_second": 2.739
|
| 531 |
+
}
|
| 532 |
+
],
|
| 533 |
+
"logging_steps": 100,
|
| 534 |
+
"max_steps": 7470,
|
| 535 |
+
"num_input_tokens_seen": 0,
|
| 536 |
+
"num_train_epochs": 30,
|
| 537 |
+
"save_steps": 400,
|
| 538 |
+
"stateful_callbacks": {
|
| 539 |
+
"EarlyStoppingCallback": {
|
| 540 |
+
"args": {
|
| 541 |
+
"early_stopping_patience": 4,
|
| 542 |
+
"early_stopping_threshold": 0.0
|
| 543 |
+
},
|
| 544 |
+
"attributes": {
|
| 545 |
+
"early_stopping_patience_counter": 4
|
| 546 |
+
}
|
| 547 |
+
},
|
| 548 |
+
"TrainerControl": {
|
| 549 |
+
"args": {
|
| 550 |
+
"should_epoch_stop": false,
|
| 551 |
+
"should_evaluate": false,
|
| 552 |
+
"should_log": false,
|
| 553 |
+
"should_save": true,
|
| 554 |
+
"should_training_stop": true
|
| 555 |
+
},
|
| 556 |
+
"attributes": {}
|
| 557 |
+
}
|
| 558 |
+
},
|
| 559 |
+
"total_flos": 1.020661828684297e+19,
|
| 560 |
+
"train_batch_size": 4,
|
| 561 |
+
"trial_name": null,
|
| 562 |
+
"trial_params": null
|
| 563 |
+
}
|