Hugging Face's logo

Salama1429
/
tarteel-ai-whisper-base-ar-quran

Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
whisper
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
2
Salama1429 commited on
Commit
6826460
•
1 Parent(s): b578688

Upload 41 files

Browse files
Files changed (35) hide show
  1. .gitignore +1 -0
  2. README.md +84 -0
  3. all_results.json +12 -0
  4. config.json +2 -11
  5. eval_results.json +8 -0
  6. preprocessor_config.json +0 -0
  7. runs/Dec08_21-03-47_ip-26-0-169-222/1670533447.9618533/events.out.tfevents.1670533447.ip-26-0-169-222.916845.1 +3 -0
  8. runs/Dec08_21-03-47_ip-26-0-169-222/events.out.tfevents.1670533447.ip-26-0-169-222.916845.0 +3 -0
  9. runs/Dec09_23-54-58_ip-26-0-168-139/1670630109.0616114/events.out.tfevents.1670630109.ip-26-0-168-139.1537794.1 +3 -0
  10. runs/Dec09_23-54-58_ip-26-0-168-139/events.out.tfevents.1670630109.ip-26-0-168-139.1537794.0 +3 -0
  11. runs/Dec10_00-01-02_ip-26-0-168-139/1670630472.0450218/events.out.tfevents.1670630472.ip-26-0-168-139.1546777.1 +3 -0
  12. runs/Dec10_00-01-02_ip-26-0-168-139/events.out.tfevents.1670630472.ip-26-0-168-139.1546777.0 +3 -0
  13. runs/Dec10_00-05-20_ip-26-0-168-139/1670630729.1031175/events.out.tfevents.1670630729.ip-26-0-168-139.1550390.1 +3 -0
  14. runs/Dec10_00-05-20_ip-26-0-168-139/events.out.tfevents.1670630729.ip-26-0-168-139.1550390.0 +3 -0
  15. runs/Dec10_00-36-47_ip-26-0-129-52/1670632617.9499032/events.out.tfevents.1670632617.ip-26-0-129-52.1669386.1 +3 -0
  16. runs/Dec10_00-36-47_ip-26-0-129-52/events.out.tfevents.1670632617.ip-26-0-129-52.1669386.0 +3 -0
  17. runs/Dec10_00-50-00_ip-26-0-129-52/1670633410.9369895/events.out.tfevents.1670633410.ip-26-0-129-52.1693596.1 +3 -0
  18. runs/Dec10_00-50-00_ip-26-0-129-52/events.out.tfevents.1670633410.ip-26-0-129-52.1693596.0 +3 -0
  19. runs/Dec10_10-44-25_ip-26-0-129-52/1670669078.0822697/events.out.tfevents.1670669078.ip-26-0-129-52.1907186.1 +3 -0
  20. runs/Dec10_10-44-25_ip-26-0-129-52/events.out.tfevents.1670669078.ip-26-0-129-52.1907186.0 +3 -0
  21. runs/Dec10_13-15-09_ip-26-0-129-52/1670678120.2629774/events.out.tfevents.1670678120.ip-26-0-129-52.2033601.1 +3 -0
  22. runs/Dec10_13-15-09_ip-26-0-129-52/events.out.tfevents.1670678120.ip-26-0-129-52.2033601.0 +3 -0
  23. runs/Dec10_20-55-34_ip-26-0-129-52/1670705746.458125/events.out.tfevents.1670705746.ip-26-0-129-52.2377118.1 +3 -0
  24. runs/Dec10_20-55-34_ip-26-0-129-52/events.out.tfevents.1670705746.ip-26-0-129-52.2377118.0 +3 -0
  25. runs/Dec10_21-26-37_ip-26-0-129-52/1670707607.913721/events.out.tfevents.1670707607.ip-26-0-129-52.2460006.1 +3 -0
  26. runs/Dec10_21-26-37_ip-26-0-129-52/events.out.tfevents.1670707607.ip-26-0-129-52.2460006.0 +3 -0
  27. runs/Dec10_22-00-04_ip-26-0-129-52/1670709617.6007733/events.out.tfevents.1670709617.ip-26-0-129-52.2546168.1 +3 -0
  28. runs/Dec10_22-00-04_ip-26-0-129-52/events.out.tfevents.1670709617.ip-26-0-129-52.2546168.0 +3 -0
  29. runs/Dec12_11-34-40_ip-26-0-168-139/1670844895.4120195/events.out.tfevents.1670844895.ip-26-0-168-139.3098999.1 +3 -0
  30. runs/Dec12_11-34-40_ip-26-0-168-139/events.out.tfevents.1670844895.ip-26-0-168-139.3098999.0 +3 -0
  31. runs/Dec12_11-34-40_ip-26-0-168-139/events.out.tfevents.1670950102.ip-26-0-168-139.3098999.2 +3 -0
  32. tokenizer_config.json +2 -2
  33. train_results.json +7 -0
  34. trainer_state.json +1405 -0
  35. training_args.bin +3 -0
.gitignore ADDED
@@ -0,0 +1 @@
 
 
1
+ checkpoint-*/
README.md ADDED
@@ -0,0 +1,84 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ metrics:
6
+ - wer
7
+ model-index:
8
+ - name: whisper-base-ar-quran
9
+ results: []
10
+ ---
11
+
12
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
13
+ should probably proofread and complete it, then remove this comment. -->
14
+
15
+ # whisper-base-ar-quran
16
+
17
+ This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the None dataset.
18
+ It achieves the following results on the evaluation set:
19
+ - Loss: 0.0839
20
+ - Wer: 5.7544
21
+
22
+ ## Model description
23
+
24
+ More information needed
25
+
26
+ ## Intended uses & limitations
27
+
28
+ More information needed
29
+
30
+ ## Training and evaluation data
31
+
32
+ More information needed
33
+
34
+ ## Training procedure
35
+
36
+ ### Training hyperparameters
37
+
38
+ The following hyperparameters were used during training:
39
+ - learning_rate: 0.0001
40
+ - train_batch_size: 16
41
+ - eval_batch_size: 8
42
+ - seed: 42
43
+ - distributed_type: multi-GPU
44
+ - num_devices: 8
45
+ - total_train_batch_size: 128
46
+ - total_eval_batch_size: 64
47
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
48
+ - lr_scheduler_type: linear
49
+ - lr_scheduler_warmup_steps: 500
50
+ - training_steps: 5000
51
+ - mixed_precision_training: Native AMP
52
+
53
+ ### Training results
54
+
55
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
56
+ |:-------------:|:-----:|:----:|:---------------:|:-------:|
57
+ | 0.1092 | 0.05 | 250 | 0.1969 | 13.3890 |
58
+ | 0.0361 | 0.1 | 500 | 0.1583 | 10.6375 |
59
+ | 0.0192 | 0.15 | 750 | 0.1109 | 8.8468 |
60
+ | 0.0144 | 0.2 | 1000 | 0.1157 | 7.9754 |
61
+ | 0.008 | 0.25 | 1250 | 0.1000 | 7.5360 |
62
+ | 0.0048 | 1.03 | 1500 | 0.0933 | 6.8227 |
63
+ | 0.0113 | 1.08 | 1750 | 0.0955 | 6.9638 |
64
+ | 0.0209 | 1.13 | 2000 | 0.0824 | 6.3586 |
65
+ | 0.0043 | 1.18 | 2250 | 0.0830 | 6.3444 |
66
+ | 0.002 | 1.23 | 2500 | 0.1015 | 6.3025 |
67
+ | 0.0013 | 2.01 | 2750 | 0.0863 | 6.0639 |
68
+ | 0.0014 | 2.06 | 3000 | 0.0905 | 6.0213 |
69
+ | 0.0018 | 2.11 | 3250 | 0.0864 | 6.0293 |
70
+ | 0.0008 | 2.16 | 3500 | 0.0887 | 5.9308 |
71
+ | 0.0029 | 2.21 | 3750 | 0.0777 | 5.9159 |
72
+ | 0.0022 | 2.26 | 4000 | 0.0847 | 5.8749 |
73
+ | 0.0005 | 3.05 | 4250 | 0.0827 | 5.8352 |
74
+ | 0.0003 | 3.1 | 4500 | 0.0826 | 5.7800 |
75
+ | 0.0006 | 3.15 | 4750 | 0.0833 | 5.7625 |
76
+ | 0.0003 | 3.2 | 5000 | 0.0839 | 5.7544 |
77
+
78
+
79
+ ### Framework versions
80
+
81
+ - Transformers 4.26.0.dev0
82
+ - Pytorch 1.13.0+cu117
83
+ - Datasets 2.7.1.dev0
84
+ - Tokenizers 0.13.2
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.2,
3
+ "eval_loss": 0.08391360193490982,
4
+ "eval_runtime": 3127.1623,
5
+ "eval_samples_per_second": 7.493,
6
+ "eval_steps_per_second": 0.117,
7
+ "eval_wer": 5.754409897414504,
8
+ "train_loss": 0.024749002781091257,
9
+ "train_runtime": 101889.8071,
10
+ "train_samples_per_second": 6.281,
11
+ "train_steps_per_second": 0.049
12
+ }
config.json CHANGED
@@ -1,8 +1,7 @@
1
  {
2
- "_name_or_path": "tarteel-ai/whisper-base-ar-quran",
3
  "activation_dropout": 0.0,
4
  "activation_function": "gelu",
5
- "apply_spec_augment": false,
6
  "architectures": [
7
  "WhisperForConditionalGeneration"
8
  ],
@@ -12,7 +11,6 @@
12
  50257
13
  ],
14
  "bos_token_id": 50257,
15
- "classifier_proj_size": 256,
16
  "d_model": 512,
17
  "decoder_attention_heads": 8,
18
  "decoder_ffn_dim": 2048,
@@ -28,12 +26,6 @@
28
  "forced_decoder_ids": null,
29
  "init_std": 0.02,
30
  "is_encoder_decoder": true,
31
- "mask_feature_length": 10,
32
- "mask_feature_min_masks": 0,
33
- "mask_feature_prob": 0.0,
34
- "mask_time_length": 10,
35
- "mask_time_min_masks": 2,
36
- "mask_time_prob": 0.05,
37
  "max_length": 1024,
38
  "max_source_positions": 1500,
39
  "max_target_positions": 448,
@@ -43,8 +35,7 @@
43
  "pad_token_id": 50257,
44
  "scale_embedding": false,
45
  "torch_dtype": "float32",
46
- "transformers_version": "4.29.2",
47
  "use_cache": false,
48
- "use_weighted_layer_sum": false,
49
  "vocab_size": 51865
50
  }
 
1
  {
2
+ "_name_or_path": "openai/whisper-base",
3
  "activation_dropout": 0.0,
4
  "activation_function": "gelu",
 
5
  "architectures": [
6
  "WhisperForConditionalGeneration"
7
  ],
 
11
  50257
12
  ],
13
  "bos_token_id": 50257,
 
14
  "d_model": 512,
15
  "decoder_attention_heads": 8,
16
  "decoder_ffn_dim": 2048,
 
26
  "forced_decoder_ids": null,
27
  "init_std": 0.02,
28
  "is_encoder_decoder": true,
 
 
 
 
 
 
29
  "max_length": 1024,
30
  "max_source_positions": 1500,
31
  "max_target_positions": 448,
 
35
  "pad_token_id": 50257,
36
  "scale_embedding": false,
37
  "torch_dtype": "float32",
38
+ "transformers_version": "4.26.0.dev0",
39
  "use_cache": false,
 
40
  "vocab_size": 51865
41
  }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.2,
3
+ "eval_loss": 0.08391360193490982,
4
+ "eval_runtime": 3127.1623,
5
+ "eval_samples_per_second": 7.493,
6
+ "eval_steps_per_second": 0.117,
7
+ "eval_wer": 5.754409897414504
8
+ }
preprocessor_config.json CHANGED
The diff for this file is too large to render. See raw diff
 
runs/Dec08_21-03-47_ip-26-0-169-222/1670533447.9618533/events.out.tfevents.1670533447.ip-26-0-169-222.916845.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5dae9a26503f1f71ab692fee3ee23d91a5c2c8942ff171ebe80afc30b85c828e
3
+ size 5968
runs/Dec08_21-03-47_ip-26-0-169-222/events.out.tfevents.1670533447.ip-26-0-169-222.916845.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1d59e8ad07cd2d126336872278dd23ac0abe92a3161c1624df7122491cff8b67
3
+ size 4512
runs/Dec09_23-54-58_ip-26-0-168-139/1670630109.0616114/events.out.tfevents.1670630109.ip-26-0-168-139.1537794.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:50fb7298457de5f1090005da04f418b87b7a91af9029f42d41c99aa71962b0dd
3
+ size 5968
runs/Dec09_23-54-58_ip-26-0-168-139/events.out.tfevents.1670630109.ip-26-0-168-139.1537794.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a67e080a2b1d9e9ef3734add9828aa50d94128d2ed74a132e3d930628dda3851
3
+ size 4359
runs/Dec10_00-01-02_ip-26-0-168-139/1670630472.0450218/events.out.tfevents.1670630472.ip-26-0-168-139.1546777.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ebd0123a5e9c90d5ca0c613d923385f59ecc31837a823bdb87180363be752881
3
+ size 5968
runs/Dec10_00-01-02_ip-26-0-168-139/events.out.tfevents.1670630472.ip-26-0-168-139.1546777.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e32b095525d8bccaafe6adbd2186bc20b28950423e2b36d31adf2b7d8f409be
3
+ size 4360
runs/Dec10_00-05-20_ip-26-0-168-139/1670630729.1031175/events.out.tfevents.1670630729.ip-26-0-168-139.1550390.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9687a84c8778f36be3b9dddc75ba28d8cd0c134f811361a5db47f029b4b9d626
3
+ size 5968
runs/Dec10_00-05-20_ip-26-0-168-139/events.out.tfevents.1670630729.ip-26-0-168-139.1550390.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:410ce655067d5ef34f44549016fba26e26d26fbce90d8a6d8b6a9cddb80b20d3
3
+ size 4360
runs/Dec10_00-36-47_ip-26-0-129-52/1670632617.9499032/events.out.tfevents.1670632617.ip-26-0-129-52.1669386.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dbf4a450e1954b842e9858866ba45f0f41a1bb066ff9654eda2dcf2aad3ff817
3
+ size 5967
runs/Dec10_00-36-47_ip-26-0-129-52/events.out.tfevents.1670632617.ip-26-0-129-52.1669386.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f65e744d69e68c61caa4b0938d5156a1d1b0d1564821efd36b4e34d2fbb08df5
3
+ size 4664
runs/Dec10_00-50-00_ip-26-0-129-52/1670633410.9369895/events.out.tfevents.1670633410.ip-26-0-129-52.1693596.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:58660ec99ea41776cf84f7a2ecc13b7ac0ab418c004c0a32d3994bef38d1d426
3
+ size 5967
runs/Dec10_00-50-00_ip-26-0-129-52/events.out.tfevents.1670633410.ip-26-0-129-52.1693596.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e5faa519d9721bcacd0c9a836cbf97cc23034ce6b755771a0faee214e38f20f
3
+ size 4664
runs/Dec10_10-44-25_ip-26-0-129-52/1670669078.0822697/events.out.tfevents.1670669078.ip-26-0-129-52.1907186.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2a6fe76055d154fc9f7dbb0b0c79886459640b1c4254e2328491caf50328703a
3
+ size 5967
runs/Dec10_10-44-25_ip-26-0-129-52/events.out.tfevents.1670669078.ip-26-0-129-52.1907186.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2a047108df29cf6ce6952fea758f540705ea514d065473fd249eabd4e1f077c0
3
+ size 4664
runs/Dec10_13-15-09_ip-26-0-129-52/1670678120.2629774/events.out.tfevents.1670678120.ip-26-0-129-52.2033601.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a8dbda2bad9ff25627db4bb42348e616080ccada0dc9f699e5754531dbfe652a
3
+ size 5967
runs/Dec10_13-15-09_ip-26-0-129-52/events.out.tfevents.1670678120.ip-26-0-129-52.2033601.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:911bef8520568dddf637f55259266eb5452be8f0f9f80cb6af805d578ce8e80e
3
+ size 7482
runs/Dec10_20-55-34_ip-26-0-129-52/1670705746.458125/events.out.tfevents.1670705746.ip-26-0-129-52.2377118.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:53b845a348798af00cff637284af6f0f9825ed49e84c3d4b419905d5fcce6ba2
3
+ size 5967
runs/Dec10_20-55-34_ip-26-0-129-52/events.out.tfevents.1670705746.ip-26-0-129-52.2377118.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1b1d5d0b2795ffaa4011b159ae82f26134d16ecef555d43e5a12b63913ff8167
3
+ size 5441
runs/Dec10_21-26-37_ip-26-0-129-52/1670707607.913721/events.out.tfevents.1670707607.ip-26-0-129-52.2460006.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b45d9c56377b128152f68843668704acca76557a858aea9118a08d26b903da3e
3
+ size 5967
runs/Dec10_21-26-37_ip-26-0-129-52/events.out.tfevents.1670707607.ip-26-0-129-52.2460006.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:393d0cfc4ea21f03c060db3953de894fd88b411e84798e748e5945e32a489b4b
3
+ size 5755
runs/Dec10_22-00-04_ip-26-0-129-52/1670709617.6007733/events.out.tfevents.1670709617.ip-26-0-129-52.2546168.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e958d2ec8aeba4f243c54ce853f61d788d9685b99f5e3a28eb97cc77e83e07e
3
+ size 5967
runs/Dec10_22-00-04_ip-26-0-129-52/events.out.tfevents.1670709617.ip-26-0-129-52.2546168.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:89855dac6da13ea85f5da389a5f51dab07828c499165977e015c00d5874404da
3
+ size 6229
runs/Dec12_11-34-40_ip-26-0-168-139/1670844895.4120195/events.out.tfevents.1670844895.ip-26-0-168-139.3098999.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:108c7e6ff25efb0c8eafeeca83fd036cf5238d8dd7c6c1a06479453345f06f49
3
+ size 5968
runs/Dec12_11-34-40_ip-26-0-168-139/events.out.tfevents.1670844895.ip-26-0-168-139.3098999.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6226bb5cb85094e9ac55fb1dae138931b086b5dfe7d2f687a4b9281c10322c72
3
+ size 42456
runs/Dec12_11-34-40_ip-26-0-168-139/events.out.tfevents.1670950102.ip-26-0-168-139.3098999.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:35240ef56f8cf2289264f5f6c96e8fa08a4232327e47de56457255b4c8354ddf
3
+ size 358
tokenizer_config.json CHANGED
@@ -9,7 +9,6 @@
9
  "rstrip": false,
10
  "single_word": false
11
  },
12
- "clean_up_tokenization_spaces": true,
13
  "eos_token": {
14
  "__type": "AddedToken",
15
  "content": "<|endoftext|>",
@@ -20,11 +19,12 @@
20
  },
21
  "errors": "replace",
22
  "model_max_length": 448,
 
23
  "pad_token": null,
24
  "processor_class": "WhisperProcessor",
25
  "return_attention_mask": false,
 
26
  "tokenizer_class": "WhisperTokenizer",
27
- "trust_remote_code": false,
28
  "unk_token": {
29
  "__type": "AddedToken",
30
  "content": "",
 
9
  "rstrip": false,
10
  "single_word": false
11
  },
 
12
  "eos_token": {
13
  "__type": "AddedToken",
14
  "content": "<|endoftext|>",
 
19
  },
20
  "errors": "replace",
21
  "model_max_length": 448,
22
+ "name_or_path": "openai/whisper-base",
23
  "pad_token": null,
24
  "processor_class": "WhisperProcessor",
25
  "return_attention_mask": false,
26
+ "special_tokens_map_file": null,
27
  "tokenizer_class": "WhisperTokenizer",
 
28
  "unk_token": {
29
  "__type": "AddedToken",
30
  "content": "",
train_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.2,
3
+ "train_loss": 0.024749002781091257,
4
+ "train_runtime": 101889.8071,
5
+ "train_samples_per_second": 6.281,
6
+ "train_steps_per_second": 0.049
7
+ }
trainer_state.json ADDED
@@ -0,0 +1,1405 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 5.754409897414504,
3
+ "best_model_checkpoint": "./tarteel-ai/whisper-base-ar-quran/checkpoint-5000",
4
+ "epoch": 3.1966,
5
+ "global_step": 5000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.01,
12
+ "learning_rate": 4.000000000000001e-06,
13
+ "loss": 1.1527,
14
+ "step": 25
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 9e-06,
19
+ "loss": 0.4624,
20
+ "step": 50
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 1.4000000000000001e-05,
25
+ "loss": 0.3386,
26
+ "step": 75
27
+ },
28
+ {
29
+ "epoch": 0.02,
30
+ "learning_rate": 1.9e-05,
31
+ "loss": 0.2467,
32
+ "step": 100
33
+ },
34
+ {
35
+ "epoch": 0.03,
36
+ "learning_rate": 2.4e-05,
37
+ "loss": 0.1953,
38
+ "step": 125
39
+ },
40
+ {
41
+ "epoch": 0.03,
42
+ "learning_rate": 2.9e-05,
43
+ "loss": 0.2176,
44
+ "step": 150
45
+ },
46
+ {
47
+ "epoch": 0.04,
48
+ "learning_rate": 3.4000000000000007e-05,
49
+ "loss": 0.1815,
50
+ "step": 175
51
+ },
52
+ {
53
+ "epoch": 0.04,
54
+ "learning_rate": 3.9000000000000006e-05,
55
+ "loss": 0.135,
56
+ "step": 200
57
+ },
58
+ {
59
+ "epoch": 0.04,
60
+ "learning_rate": 4.4000000000000006e-05,
61
+ "loss": 0.0715,
62
+ "step": 225
63
+ },
64
+ {
65
+ "epoch": 0.05,
66
+ "learning_rate": 4.9e-05,
67
+ "loss": 0.1092,
68
+ "step": 250
69
+ },
70
+ {
71
+ "epoch": 0.05,
72
+ "eval_loss": 0.19689758121967316,
73
+ "eval_runtime": 3129.8332,
74
+ "eval_samples_per_second": 7.486,
75
+ "eval_steps_per_second": 0.117,
76
+ "eval_wer": 13.388995632478606,
77
+ "step": 250
78
+ },
79
+ {
80
+ "epoch": 0.06,
81
+ "learning_rate": 5.4000000000000005e-05,
82
+ "loss": 0.0848,
83
+ "step": 275
84
+ },
85
+ {
86
+ "epoch": 0.06,
87
+ "learning_rate": 5.9e-05,
88
+ "loss": 0.056,
89
+ "step": 300
90
+ },
91
+ {
92
+ "epoch": 0.07,
93
+ "learning_rate": 6.400000000000001e-05,
94
+ "loss": 0.0652,
95
+ "step": 325
96
+ },
97
+ {
98
+ "epoch": 0.07,
99
+ "learning_rate": 6.9e-05,
100
+ "loss": 0.089,
101
+ "step": 350
102
+ },
103
+ {
104
+ "epoch": 0.07,
105
+ "learning_rate": 7.4e-05,
106
+ "loss": 0.1052,
107
+ "step": 375
108
+ },
109
+ {
110
+ "epoch": 0.08,
111
+ "learning_rate": 7.900000000000001e-05,
112
+ "loss": 0.0804,
113
+ "step": 400
114
+ },
115
+ {
116
+ "epoch": 0.09,
117
+ "learning_rate": 8.4e-05,
118
+ "loss": 0.07,
119
+ "step": 425
120
+ },
121
+ {
122
+ "epoch": 0.09,
123
+ "learning_rate": 8.900000000000001e-05,
124
+ "loss": 0.0389,
125
+ "step": 450
126
+ },
127
+ {
128
+ "epoch": 0.1,
129
+ "learning_rate": 9.4e-05,
130
+ "loss": 0.038,
131
+ "step": 475
132
+ },
133
+ {
134
+ "epoch": 0.1,
135
+ "learning_rate": 9.900000000000001e-05,
136
+ "loss": 0.0361,
137
+ "step": 500
138
+ },
139
+ {
140
+ "epoch": 0.1,
141
+ "eval_loss": 0.1583031862974167,
142
+ "eval_runtime": 3321.2758,
143
+ "eval_samples_per_second": 7.055,
144
+ "eval_steps_per_second": 0.11,
145
+ "eval_wer": 10.637548504050336,
146
+ "step": 500
147
+ },
148
+ {
149
+ "epoch": 0.1,
150
+ "learning_rate": 9.955555555555556e-05,
151
+ "loss": 0.022,
152
+ "step": 525
153
+ },
154
+ {
155
+ "epoch": 0.11,
156
+ "learning_rate": 9.900000000000001e-05,
157
+ "loss": 0.0211,
158
+ "step": 550
159
+ },
160
+ {
161
+ "epoch": 0.12,
162
+ "learning_rate": 9.844444444444444e-05,
163
+ "loss": 0.0484,
164
+ "step": 575
165
+ },
166
+ {
167
+ "epoch": 0.12,
168
+ "learning_rate": 9.78888888888889e-05,
169
+ "loss": 0.0241,
170
+ "step": 600
171
+ },
172
+ {
173
+ "epoch": 0.12,
174
+ "learning_rate": 9.733333333333335e-05,
175
+ "loss": 0.0224,
176
+ "step": 625
177
+ },
178
+ {
179
+ "epoch": 0.13,
180
+ "learning_rate": 9.677777777777778e-05,
181
+ "loss": 0.0272,
182
+ "step": 650
183
+ },
184
+ {
185
+ "epoch": 0.14,
186
+ "learning_rate": 9.622222222222222e-05,
187
+ "loss": 0.0322,
188
+ "step": 675
189
+ },
190
+ {
191
+ "epoch": 0.14,
192
+ "learning_rate": 9.566666666666667e-05,
193
+ "loss": 0.029,
194
+ "step": 700
195
+ },
196
+ {
197
+ "epoch": 0.14,
198
+ "learning_rate": 9.511111111111112e-05,
199
+ "loss": 0.0244,
200
+ "step": 725
201
+ },
202
+ {
203
+ "epoch": 0.15,
204
+ "learning_rate": 9.455555555555556e-05,
205
+ "loss": 0.0192,
206
+ "step": 750
207
+ },
208
+ {
209
+ "epoch": 0.15,
210
+ "eval_loss": 0.11088771373033524,
211
+ "eval_runtime": 3203.8487,
212
+ "eval_samples_per_second": 7.313,
213
+ "eval_steps_per_second": 0.115,
214
+ "eval_wer": 8.846793683080142,
215
+ "step": 750
216
+ },
217
+ {
218
+ "epoch": 0.15,
219
+ "learning_rate": 9.4e-05,
220
+ "loss": 0.034,
221
+ "step": 775
222
+ },
223
+ {
224
+ "epoch": 0.16,
225
+ "learning_rate": 9.344444444444444e-05,
226
+ "loss": 0.0251,
227
+ "step": 800
228
+ },
229
+ {
230
+ "epoch": 0.17,
231
+ "learning_rate": 9.28888888888889e-05,
232
+ "loss": 0.037,
233
+ "step": 825
234
+ },
235
+ {
236
+ "epoch": 0.17,
237
+ "learning_rate": 9.233333333333333e-05,
238
+ "loss": 0.0162,
239
+ "step": 850
240
+ },
241
+ {
242
+ "epoch": 0.17,
243
+ "learning_rate": 9.177777777777778e-05,
244
+ "loss": 0.0219,
245
+ "step": 875
246
+ },
247
+ {
248
+ "epoch": 0.18,
249
+ "learning_rate": 9.122222222222223e-05,
250
+ "loss": 0.0346,
251
+ "step": 900
252
+ },
253
+ {
254
+ "epoch": 0.18,
255
+ "learning_rate": 9.066666666666667e-05,
256
+ "loss": 0.0228,
257
+ "step": 925
258
+ },
259
+ {
260
+ "epoch": 0.19,
261
+ "learning_rate": 9.011111111111111e-05,
262
+ "loss": 0.016,
263
+ "step": 950
264
+ },
265
+ {
266
+ "epoch": 0.2,
267
+ "learning_rate": 8.955555555555556e-05,
268
+ "loss": 0.0186,
269
+ "step": 975
270
+ },
271
+ {
272
+ "epoch": 0.2,
273
+ "learning_rate": 8.900000000000001e-05,
274
+ "loss": 0.0144,
275
+ "step": 1000
276
+ },
277
+ {
278
+ "epoch": 0.2,
279
+ "eval_loss": 0.11568614840507507,
280
+ "eval_runtime": 3153.953,
281
+ "eval_samples_per_second": 7.429,
282
+ "eval_steps_per_second": 0.116,
283
+ "eval_wer": 7.9754208928500745,
284
+ "step": 1000
285
+ },
286
+ {
287
+ "epoch": 0.2,
288
+ "learning_rate": 8.844444444444445e-05,
289
+ "loss": 0.0093,
290
+ "step": 1025
291
+ },
292
+ {
293
+ "epoch": 0.21,
294
+ "learning_rate": 8.78888888888889e-05,
295
+ "loss": 0.0218,
296
+ "step": 1050
297
+ },
298
+ {
299
+ "epoch": 0.21,
300
+ "learning_rate": 8.733333333333333e-05,
301
+ "loss": 0.016,
302
+ "step": 1075
303
+ },
304
+ {
305
+ "epoch": 0.22,
306
+ "learning_rate": 8.677777777777778e-05,
307
+ "loss": 0.0149,
308
+ "step": 1100
309
+ },
310
+ {
311
+ "epoch": 0.23,
312
+ "learning_rate": 8.622222222222222e-05,
313
+ "loss": 0.0122,
314
+ "step": 1125
315
+ },
316
+ {
317
+ "epoch": 0.23,
318
+ "learning_rate": 8.566666666666667e-05,
319
+ "loss": 0.0078,
320
+ "step": 1150
321
+ },
322
+ {
323
+ "epoch": 0.23,
324
+ "learning_rate": 8.511111111111112e-05,
325
+ "loss": 0.0087,
326
+ "step": 1175
327
+ },
328
+ {
329
+ "epoch": 0.24,
330
+ "learning_rate": 8.455555555555556e-05,
331
+ "loss": 0.0065,
332
+ "step": 1200
333
+ },
334
+ {
335
+ "epoch": 0.24,
336
+ "learning_rate": 8.4e-05,
337
+ "loss": 0.0059,
338
+ "step": 1225
339
+ },
340
+ {
341
+ "epoch": 0.25,
342
+ "learning_rate": 8.344444444444445e-05,
343
+ "loss": 0.008,
344
+ "step": 1250
345
+ },
346
+ {
347
+ "epoch": 0.25,
348
+ "eval_loss": 0.09997964650392532,
349
+ "eval_runtime": 3103.3278,
350
+ "eval_samples_per_second": 7.55,
351
+ "eval_steps_per_second": 0.118,
352
+ "eval_wer": 7.536029767759106,
353
+ "step": 1250
354
+ },
355
+ {
356
+ "epoch": 0.26,
357
+ "learning_rate": 8.28888888888889e-05,
358
+ "loss": 0.0087,
359
+ "step": 1275
360
+ },
361
+ {
362
+ "epoch": 0.26,
363
+ "learning_rate": 8.233333333333333e-05,
364
+ "loss": 0.0096,
365
+ "step": 1300
366
+ },
367
+ {
368
+ "epoch": 0.27,
369
+ "learning_rate": 8.177777777777778e-05,
370
+ "loss": 0.0088,
371
+ "step": 1325
372
+ },
373
+ {
374
+ "epoch": 1.0,
375
+ "learning_rate": 8.122222222222222e-05,
376
+ "loss": 0.0115,
377
+ "step": 1350
378
+ },
379
+ {
380
+ "epoch": 1.01,
381
+ "learning_rate": 8.066666666666667e-05,
382
+ "loss": 0.0114,
383
+ "step": 1375
384
+ },
385
+ {
386
+ "epoch": 1.01,
387
+ "learning_rate": 8.011111111111111e-05,
388
+ "loss": 0.0038,
389
+ "step": 1400
390
+ },
391
+ {
392
+ "epoch": 1.02,
393
+ "learning_rate": 7.955555555555556e-05,
394
+ "loss": 0.0106,
395
+ "step": 1425
396
+ },
397
+ {
398
+ "epoch": 1.02,
399
+ "learning_rate": 7.900000000000001e-05,
400
+ "loss": 0.0075,
401
+ "step": 1450
402
+ },
403
+ {
404
+ "epoch": 1.03,
405
+ "learning_rate": 7.844444444444446e-05,
406
+ "loss": 0.0094,
407
+ "step": 1475
408
+ },
409
+ {
410
+ "epoch": 1.03,
411
+ "learning_rate": 7.788888888888888e-05,
412
+ "loss": 0.0048,
413
+ "step": 1500
414
+ },
415
+ {
416
+ "epoch": 1.03,
417
+ "eval_loss": 0.09334036707878113,
418
+ "eval_runtime": 3157.6907,
419
+ "eval_samples_per_second": 7.42,
420
+ "eval_steps_per_second": 0.116,
421
+ "eval_wer": 6.822691623351776,
422
+ "step": 1500
423
+ },
424
+ {
425
+ "epoch": 1.04,
426
+ "learning_rate": 7.733333333333333e-05,
427
+ "loss": 0.0034,
428
+ "step": 1525
429
+ },
430
+ {
431
+ "epoch": 1.04,
432
+ "learning_rate": 7.677777777777778e-05,
433
+ "loss": 0.003,
434
+ "step": 1550
435
+ },
436
+ {
437
+ "epoch": 1.05,
438
+ "learning_rate": 7.622222222222223e-05,
439
+ "loss": 0.005,
440
+ "step": 1575
441
+ },
442
+ {
443
+ "epoch": 1.05,
444
+ "learning_rate": 7.566666666666667e-05,
445
+ "loss": 0.0043,
446
+ "step": 1600
447
+ },
448
+ {
449
+ "epoch": 1.06,
450
+ "learning_rate": 7.511111111111111e-05,
451
+ "loss": 0.0027,
452
+ "step": 1625
453
+ },
454
+ {
455
+ "epoch": 1.06,
456
+ "learning_rate": 7.455555555555556e-05,
457
+ "loss": 0.0213,
458
+ "step": 1650
459
+ },
460
+ {
461
+ "epoch": 1.07,
462
+ "learning_rate": 7.4e-05,
463
+ "loss": 0.0122,
464
+ "step": 1675
465
+ },
466
+ {
467
+ "epoch": 1.07,
468
+ "learning_rate": 7.344444444444445e-05,
469
+ "loss": 0.006,
470
+ "step": 1700
471
+ },
472
+ {
473
+ "epoch": 1.08,
474
+ "learning_rate": 7.28888888888889e-05,
475
+ "loss": 0.0123,
476
+ "step": 1725
477
+ },
478
+ {
479
+ "epoch": 1.08,
480
+ "learning_rate": 7.233333333333335e-05,
481
+ "loss": 0.0113,
482
+ "step": 1750
483
+ },
484
+ {
485
+ "epoch": 1.08,
486
+ "eval_loss": 0.09547463804483414,
487
+ "eval_runtime": 3101.8736,
488
+ "eval_samples_per_second": 7.554,
489
+ "eval_steps_per_second": 0.118,
490
+ "eval_wer": 6.963775860791473,
491
+ "step": 1750
492
+ },
493
+ {
494
+ "epoch": 1.09,
495
+ "learning_rate": 7.177777777777777e-05,
496
+ "loss": 0.0087,
497
+ "step": 1775
498
+ },
499
+ {
500
+ "epoch": 1.09,
501
+ "learning_rate": 7.122222222222222e-05,
502
+ "loss": 0.0088,
503
+ "step": 1800
504
+ },
505
+ {
506
+ "epoch": 1.1,
507
+ "learning_rate": 7.066666666666667e-05,
508
+ "loss": 0.0112,
509
+ "step": 1825
510
+ },
511
+ {
512
+ "epoch": 1.1,
513
+ "learning_rate": 7.011111111111112e-05,
514
+ "loss": 0.008,
515
+ "step": 1850
516
+ },
517
+ {
518
+ "epoch": 1.11,
519
+ "learning_rate": 6.955555555555556e-05,
520
+ "loss": 0.0044,
521
+ "step": 1875
522
+ },
523
+ {
524
+ "epoch": 1.11,
525
+ "learning_rate": 6.9e-05,
526
+ "loss": 0.0143,
527
+ "step": 1900
528
+ },
529
+ {
530
+ "epoch": 1.12,
531
+ "learning_rate": 6.844444444444445e-05,
532
+ "loss": 0.0123,
533
+ "step": 1925
534
+ },
535
+ {
536
+ "epoch": 1.12,
537
+ "learning_rate": 6.788888888888888e-05,
538
+ "loss": 0.0042,
539
+ "step": 1950
540
+ },
541
+ {
542
+ "epoch": 1.13,
543
+ "learning_rate": 6.733333333333333e-05,
544
+ "loss": 0.0029,
545
+ "step": 1975
546
+ },
547
+ {
548
+ "epoch": 1.13,
549
+ "learning_rate": 6.677777777777779e-05,
550
+ "loss": 0.0209,
551
+ "step": 2000
552
+ },
553
+ {
554
+ "epoch": 1.13,
555
+ "eval_loss": 0.08239686489105225,
556
+ "eval_runtime": 3082.9101,
557
+ "eval_samples_per_second": 7.6,
558
+ "eval_steps_per_second": 0.119,
559
+ "eval_wer": 6.358636131571709,
560
+ "step": 2000
561
+ },
562
+ {
563
+ "epoch": 1.14,
564
+ "learning_rate": 6.622222222222224e-05,
565
+ "loss": 0.0126,
566
+ "step": 2025
567
+ },
568
+ {
569
+ "epoch": 1.14,
570
+ "learning_rate": 6.566666666666666e-05,
571
+ "loss": 0.0181,
572
+ "step": 2050
573
+ },
574
+ {
575
+ "epoch": 1.15,
576
+ "learning_rate": 6.511111111111111e-05,
577
+ "loss": 0.0067,
578
+ "step": 2075
579
+ },
580
+ {
581
+ "epoch": 1.15,
582
+ "learning_rate": 6.455555555555556e-05,
583
+ "loss": 0.0049,
584
+ "step": 2100
585
+ },
586
+ {
587
+ "epoch": 1.16,
588
+ "learning_rate": 6.400000000000001e-05,
589
+ "loss": 0.0085,
590
+ "step": 2125
591
+ },
592
+ {
593
+ "epoch": 1.16,
594
+ "learning_rate": 6.344444444444445e-05,
595
+ "loss": 0.0071,
596
+ "step": 2150
597
+ },
598
+ {
599
+ "epoch": 1.17,
600
+ "learning_rate": 6.28888888888889e-05,
601
+ "loss": 0.0051,
602
+ "step": 2175
603
+ },
604
+ {
605
+ "epoch": 1.17,
606
+ "learning_rate": 6.233333333333334e-05,
607
+ "loss": 0.0241,
608
+ "step": 2200
609
+ },
610
+ {
611
+ "epoch": 1.18,
612
+ "learning_rate": 6.177777777777779e-05,
613
+ "loss": 0.0056,
614
+ "step": 2225
615
+ },
616
+ {
617
+ "epoch": 1.18,
618
+ "learning_rate": 6.122222222222222e-05,
619
+ "loss": 0.0043,
620
+ "step": 2250
621
+ },
622
+ {
623
+ "epoch": 1.18,
624
+ "eval_loss": 0.08303235471248627,
625
+ "eval_runtime": 3099.7955,
626
+ "eval_samples_per_second": 7.559,
627
+ "eval_steps_per_second": 0.118,
628
+ "eval_wer": 6.344426208376344,
629
+ "step": 2250
630
+ },
631
+ {
632
+ "epoch": 1.19,
633
+ "learning_rate": 6.066666666666667e-05,
634
+ "loss": 0.0036,
635
+ "step": 2275
636
+ },
637
+ {
638
+ "epoch": 1.19,
639
+ "learning_rate": 6.011111111111112e-05,
640
+ "loss": 0.0092,
641
+ "step": 2300
642
+ },
643
+ {
644
+ "epoch": 1.2,
645
+ "learning_rate": 5.9555555555555554e-05,
646
+ "loss": 0.0048,
647
+ "step": 2325
648
+ },
649
+ {
650
+ "epoch": 1.2,
651
+ "learning_rate": 5.9e-05,
652
+ "loss": 0.0054,
653
+ "step": 2350
654
+ },
655
+ {
656
+ "epoch": 1.21,
657
+ "learning_rate": 5.844444444444445e-05,
658
+ "loss": 0.0044,
659
+ "step": 2375
660
+ },
661
+ {
662
+ "epoch": 1.21,
663
+ "learning_rate": 5.788888888888889e-05,
664
+ "loss": 0.0036,
665
+ "step": 2400
666
+ },
667
+ {
668
+ "epoch": 1.22,
669
+ "learning_rate": 5.7333333333333336e-05,
670
+ "loss": 0.004,
671
+ "step": 2425
672
+ },
673
+ {
674
+ "epoch": 1.22,
675
+ "learning_rate": 5.6777777777777786e-05,
676
+ "loss": 0.0059,
677
+ "step": 2450
678
+ },
679
+ {
680
+ "epoch": 1.23,
681
+ "learning_rate": 5.622222222222222e-05,
682
+ "loss": 0.005,
683
+ "step": 2475
684
+ },
685
+ {
686
+ "epoch": 1.23,
687
+ "learning_rate": 5.566666666666667e-05,
688
+ "loss": 0.002,
689
+ "step": 2500
690
+ },
691
+ {
692
+ "epoch": 1.23,
693
+ "eval_loss": 0.10149681568145752,
694
+ "eval_runtime": 3102.277,
695
+ "eval_samples_per_second": 7.553,
696
+ "eval_steps_per_second": 0.118,
697
+ "eval_wer": 6.302506934950017,
698
+ "step": 2500
699
+ },
700
+ {
701
+ "epoch": 1.24,
702
+ "learning_rate": 5.511111111111111e-05,
703
+ "loss": 0.005,
704
+ "step": 2525
705
+ },
706
+ {
707
+ "epoch": 1.24,
708
+ "learning_rate": 5.455555555555556e-05,
709
+ "loss": 0.0027,
710
+ "step": 2550
711
+ },
712
+ {
713
+ "epoch": 1.25,
714
+ "learning_rate": 5.4000000000000005e-05,
715
+ "loss": 0.0041,
716
+ "step": 2575
717
+ },
718
+ {
719
+ "epoch": 1.25,
720
+ "learning_rate": 5.3444444444444455e-05,
721
+ "loss": 0.0042,
722
+ "step": 2600
723
+ },
724
+ {
725
+ "epoch": 1.26,
726
+ "learning_rate": 5.2888888888888885e-05,
727
+ "loss": 0.0061,
728
+ "step": 2625
729
+ },
730
+ {
731
+ "epoch": 1.26,
732
+ "learning_rate": 5.2333333333333336e-05,
733
+ "loss": 0.0047,
734
+ "step": 2650
735
+ },
736
+ {
737
+ "epoch": 1.27,
738
+ "learning_rate": 5.177777777777778e-05,
739
+ "loss": 0.0026,
740
+ "step": 2675
741
+ },
742
+ {
743
+ "epoch": 2.0,
744
+ "learning_rate": 5.122222222222223e-05,
745
+ "loss": 0.0023,
746
+ "step": 2700
747
+ },
748
+ {
749
+ "epoch": 2.01,
750
+ "learning_rate": 5.0666666666666674e-05,
751
+ "loss": 0.0018,
752
+ "step": 2725
753
+ },
754
+ {
755
+ "epoch": 2.01,
756
+ "learning_rate": 5.011111111111111e-05,
757
+ "loss": 0.0013,
758
+ "step": 2750
759
+ },
760
+ {
761
+ "epoch": 2.01,
762
+ "eval_loss": 0.08628152310848236,
763
+ "eval_runtime": 3108.0423,
764
+ "eval_samples_per_second": 7.539,
765
+ "eval_steps_per_second": 0.118,
766
+ "eval_wer": 6.063881724719278,
767
+ "step": 2750
768
+ },
769
+ {
770
+ "epoch": 2.02,
771
+ "learning_rate": 4.955555555555556e-05,
772
+ "loss": 0.0019,
773
+ "step": 2775
774
+ },
775
+ {
776
+ "epoch": 2.02,
777
+ "learning_rate": 4.9e-05,
778
+ "loss": 0.0021,
779
+ "step": 2800
780
+ },
781
+ {
782
+ "epoch": 2.03,
783
+ "learning_rate": 4.844444444444445e-05,
784
+ "loss": 0.0008,
785
+ "step": 2825
786
+ },
787
+ {
788
+ "epoch": 2.03,
789
+ "learning_rate": 4.7888888888888886e-05,
790
+ "loss": 0.0014,
791
+ "step": 2850
792
+ },
793
+ {
794
+ "epoch": 2.04,
795
+ "learning_rate": 4.7333333333333336e-05,
796
+ "loss": 0.0007,
797
+ "step": 2875
798
+ },
799
+ {
800
+ "epoch": 2.04,
801
+ "learning_rate": 4.677777777777778e-05,
802
+ "loss": 0.0008,
803
+ "step": 2900
804
+ },
805
+ {
806
+ "epoch": 2.05,
807
+ "learning_rate": 4.6222222222222224e-05,
808
+ "loss": 0.0027,
809
+ "step": 2925
810
+ },
811
+ {
812
+ "epoch": 2.05,
813
+ "learning_rate": 4.566666666666667e-05,
814
+ "loss": 0.0023,
815
+ "step": 2950
816
+ },
817
+ {
818
+ "epoch": 2.06,
819
+ "learning_rate": 4.511111111111112e-05,
820
+ "loss": 0.0012,
821
+ "step": 2975
822
+ },
823
+ {
824
+ "epoch": 2.06,
825
+ "learning_rate": 4.4555555555555555e-05,
826
+ "loss": 0.0014,
827
+ "step": 3000
828
+ },
829
+ {
830
+ "epoch": 2.06,
831
+ "eval_loss": 0.09052816778421402,
832
+ "eval_runtime": 3089.8491,
833
+ "eval_samples_per_second": 7.583,
834
+ "eval_steps_per_second": 0.119,
835
+ "eval_wer": 6.021251955133183,
836
+ "step": 3000
837
+ },
838
+ {
839
+ "epoch": 2.07,
840
+ "learning_rate": 4.4000000000000006e-05,
841
+ "loss": 0.0012,
842
+ "step": 3025
843
+ },
844
+ {
845
+ "epoch": 2.07,
846
+ "learning_rate": 4.344444444444445e-05,
847
+ "loss": 0.0013,
848
+ "step": 3050
849
+ },
850
+ {
851
+ "epoch": 2.08,
852
+ "learning_rate": 4.2888888888888886e-05,
853
+ "loss": 0.0021,
854
+ "step": 3075
855
+ },
856
+ {
857
+ "epoch": 2.08,
858
+ "learning_rate": 4.233333333333334e-05,
859
+ "loss": 0.0024,
860
+ "step": 3100
861
+ },
862
+ {
863
+ "epoch": 2.09,
864
+ "learning_rate": 4.177777777777778e-05,
865
+ "loss": 0.0016,
866
+ "step": 3125
867
+ },
868
+ {
869
+ "epoch": 2.09,
870
+ "learning_rate": 4.1222222222222224e-05,
871
+ "loss": 0.0008,
872
+ "step": 3150
873
+ },
874
+ {
875
+ "epoch": 2.1,
876
+ "learning_rate": 4.066666666666667e-05,
877
+ "loss": 0.0013,
878
+ "step": 3175
879
+ },
880
+ {
881
+ "epoch": 2.1,
882
+ "learning_rate": 4.011111111111111e-05,
883
+ "loss": 0.0025,
884
+ "step": 3200
885
+ },
886
+ {
887
+ "epoch": 2.11,
888
+ "learning_rate": 3.9555555555555556e-05,
889
+ "loss": 0.0064,
890
+ "step": 3225
891
+ },
892
+ {
893
+ "epoch": 2.11,
894
+ "learning_rate": 3.9000000000000006e-05,
895
+ "loss": 0.0018,
896
+ "step": 3250
897
+ },
898
+ {
899
+ "epoch": 2.11,
900
+ "eval_loss": 0.08641628921031952,
901
+ "eval_runtime": 3083.8255,
902
+ "eval_samples_per_second": 7.598,
903
+ "eval_steps_per_second": 0.119,
904
+ "eval_wer": 6.029270411793424,
905
+ "step": 3250
906
+ },
907
+ {
908
+ "epoch": 2.12,
909
+ "learning_rate": 3.844444444444444e-05,
910
+ "loss": 0.0021,
911
+ "step": 3275
912
+ },
913
+ {
914
+ "epoch": 2.12,
915
+ "learning_rate": 3.7888888888888894e-05,
916
+ "loss": 0.0011,
917
+ "step": 3300
918
+ },
919
+ {
920
+ "epoch": 2.13,
921
+ "learning_rate": 3.733333333333334e-05,
922
+ "loss": 0.0021,
923
+ "step": 3325
924
+ },
925
+ {
926
+ "epoch": 2.13,
927
+ "learning_rate": 3.677777777777778e-05,
928
+ "loss": 0.0071,
929
+ "step": 3350
930
+ },
931
+ {
932
+ "epoch": 2.14,
933
+ "learning_rate": 3.6222222222222225e-05,
934
+ "loss": 0.0017,
935
+ "step": 3375
936
+ },
937
+ {
938
+ "epoch": 2.14,
939
+ "learning_rate": 3.566666666666667e-05,
940
+ "loss": 0.0016,
941
+ "step": 3400
942
+ },
943
+ {
944
+ "epoch": 2.15,
945
+ "learning_rate": 3.511111111111111e-05,
946
+ "loss": 0.0005,
947
+ "step": 3425
948
+ },
949
+ {
950
+ "epoch": 2.15,
951
+ "learning_rate": 3.4555555555555556e-05,
952
+ "loss": 0.0024,
953
+ "step": 3450
954
+ },
955
+ {
956
+ "epoch": 2.16,
957
+ "learning_rate": 3.4000000000000007e-05,
958
+ "loss": 0.0024,
959
+ "step": 3475
960
+ },
961
+ {
962
+ "epoch": 2.16,
963
+ "learning_rate": 3.3444444444444443e-05,
964
+ "loss": 0.0008,
965
+ "step": 3500
966
+ },
967
+ {
968
+ "epoch": 2.16,
969
+ "eval_loss": 0.0887165442109108,
970
+ "eval_runtime": 3111.2967,
971
+ "eval_samples_per_second": 7.531,
972
+ "eval_steps_per_second": 0.118,
973
+ "eval_wer": 5.930815943939823,
974
+ "step": 3500
975
+ },
976
+ {
977
+ "epoch": 2.17,
978
+ "learning_rate": 3.2888888888888894e-05,
979
+ "loss": 0.0013,
980
+ "step": 3525
981
+ },
982
+ {
983
+ "epoch": 2.17,
984
+ "learning_rate": 3.233333333333333e-05,
985
+ "loss": 0.0009,
986
+ "step": 3550
987
+ },
988
+ {
989
+ "epoch": 2.18,
990
+ "learning_rate": 3.177777777777778e-05,
991
+ "loss": 0.0012,
992
+ "step": 3575
993
+ },
994
+ {
995
+ "epoch": 2.18,
996
+ "learning_rate": 3.1222222222222225e-05,
997
+ "loss": 0.0012,
998
+ "step": 3600
999
+ },
1000
+ {
1001
+ "epoch": 2.19,
1002
+ "learning_rate": 3.066666666666667e-05,
1003
+ "loss": 0.0007,
1004
+ "step": 3625
1005
+ },
1006
+ {
1007
+ "epoch": 2.19,
1008
+ "learning_rate": 3.0111111111111113e-05,
1009
+ "loss": 0.0046,
1010
+ "step": 3650
1011
+ },
1012
+ {
1013
+ "epoch": 2.2,
1014
+ "learning_rate": 2.955555555555556e-05,
1015
+ "loss": 0.0016,
1016
+ "step": 3675
1017
+ },
1018
+ {
1019
+ "epoch": 2.2,
1020
+ "learning_rate": 2.9e-05,
1021
+ "loss": 0.0007,
1022
+ "step": 3700
1023
+ },
1024
+ {
1025
+ "epoch": 2.21,
1026
+ "learning_rate": 2.8444444444444447e-05,
1027
+ "loss": 0.0018,
1028
+ "step": 3725
1029
+ },
1030
+ {
1031
+ "epoch": 2.21,
1032
+ "learning_rate": 2.788888888888889e-05,
1033
+ "loss": 0.0029,
1034
+ "step": 3750
1035
+ },
1036
+ {
1037
+ "epoch": 2.21,
1038
+ "eval_loss": 0.07773207873106003,
1039
+ "eval_runtime": 3114.8324,
1040
+ "eval_samples_per_second": 7.522,
1041
+ "eval_steps_per_second": 0.118,
1042
+ "eval_wer": 5.915895524584689,
1043
+ "step": 3750
1044
+ },
1045
+ {
1046
+ "epoch": 2.22,
1047
+ "learning_rate": 2.733333333333333e-05,
1048
+ "loss": 0.0026,
1049
+ "step": 3775
1050
+ },
1051
+ {
1052
+ "epoch": 2.22,
1053
+ "learning_rate": 2.677777777777778e-05,
1054
+ "loss": 0.002,
1055
+ "step": 3800
1056
+ },
1057
+ {
1058
+ "epoch": 2.23,
1059
+ "learning_rate": 2.6222222222222226e-05,
1060
+ "loss": 0.0025,
1061
+ "step": 3825
1062
+ },
1063
+ {
1064
+ "epoch": 2.23,
1065
+ "learning_rate": 2.5666666666666666e-05,
1066
+ "loss": 0.001,
1067
+ "step": 3850
1068
+ },
1069
+ {
1070
+ "epoch": 2.24,
1071
+ "learning_rate": 2.5111111111111113e-05,
1072
+ "loss": 0.0004,
1073
+ "step": 3875
1074
+ },
1075
+ {
1076
+ "epoch": 2.24,
1077
+ "learning_rate": 2.4555555555555557e-05,
1078
+ "loss": 0.0021,
1079
+ "step": 3900
1080
+ },
1081
+ {
1082
+ "epoch": 2.25,
1083
+ "learning_rate": 2.4e-05,
1084
+ "loss": 0.0006,
1085
+ "step": 3925
1086
+ },
1087
+ {
1088
+ "epoch": 2.25,
1089
+ "learning_rate": 2.3444444444444448e-05,
1090
+ "loss": 0.001,
1091
+ "step": 3950
1092
+ },
1093
+ {
1094
+ "epoch": 2.26,
1095
+ "learning_rate": 2.288888888888889e-05,
1096
+ "loss": 0.0013,
1097
+ "step": 3975
1098
+ },
1099
+ {
1100
+ "epoch": 2.26,
1101
+ "learning_rate": 2.2333333333333335e-05,
1102
+ "loss": 0.0022,
1103
+ "step": 4000
1104
+ },
1105
+ {
1106
+ "epoch": 2.26,
1107
+ "eval_loss": 0.08468455821275711,
1108
+ "eval_runtime": 3072.4642,
1109
+ "eval_samples_per_second": 7.626,
1110
+ "eval_steps_per_second": 0.119,
1111
+ "eval_wer": 5.874889746220922,
1112
+ "step": 4000
1113
+ },
1114
+ {
1115
+ "epoch": 3.0,
1116
+ "learning_rate": 2.177777777777778e-05,
1117
+ "loss": 0.0018,
1118
+ "step": 4025
1119
+ },
1120
+ {
1121
+ "epoch": 3.01,
1122
+ "learning_rate": 2.1222222222222223e-05,
1123
+ "loss": 0.0008,
1124
+ "step": 4050
1125
+ },
1126
+ {
1127
+ "epoch": 3.01,
1128
+ "learning_rate": 2.0666666666666666e-05,
1129
+ "loss": 0.0013,
1130
+ "step": 4075
1131
+ },
1132
+ {
1133
+ "epoch": 3.02,
1134
+ "learning_rate": 2.011111111111111e-05,
1135
+ "loss": 0.0003,
1136
+ "step": 4100
1137
+ },
1138
+ {
1139
+ "epoch": 3.02,
1140
+ "learning_rate": 1.9555555555555557e-05,
1141
+ "loss": 0.0006,
1142
+ "step": 4125
1143
+ },
1144
+ {
1145
+ "epoch": 3.03,
1146
+ "learning_rate": 1.9e-05,
1147
+ "loss": 0.0006,
1148
+ "step": 4150
1149
+ },
1150
+ {
1151
+ "epoch": 3.03,
1152
+ "learning_rate": 1.8444444444444445e-05,
1153
+ "loss": 0.0004,
1154
+ "step": 4175
1155
+ },
1156
+ {
1157
+ "epoch": 3.04,
1158
+ "learning_rate": 1.788888888888889e-05,
1159
+ "loss": 0.0002,
1160
+ "step": 4200
1161
+ },
1162
+ {
1163
+ "epoch": 3.04,
1164
+ "learning_rate": 1.7333333333333336e-05,
1165
+ "loss": 0.0001,
1166
+ "step": 4225
1167
+ },
1168
+ {
1169
+ "epoch": 3.05,
1170
+ "learning_rate": 1.677777777777778e-05,
1171
+ "loss": 0.0005,
1172
+ "step": 4250
1173
+ },
1174
+ {
1175
+ "epoch": 3.05,
1176
+ "eval_loss": 0.08272121101617813,
1177
+ "eval_runtime": 3097.4968,
1178
+ "eval_samples_per_second": 7.564,
1179
+ "eval_steps_per_second": 0.118,
1180
+ "eval_wer": 5.835203460725294,
1181
+ "step": 4250
1182
+ },
1183
+ {
1184
+ "epoch": 3.05,
1185
+ "learning_rate": 1.6222222222222223e-05,
1186
+ "loss": 0.0013,
1187
+ "step": 4275
1188
+ },
1189
+ {
1190
+ "epoch": 3.06,
1191
+ "learning_rate": 1.5666666666666667e-05,
1192
+ "loss": 0.0007,
1193
+ "step": 4300
1194
+ },
1195
+ {
1196
+ "epoch": 3.06,
1197
+ "learning_rate": 1.5111111111111112e-05,
1198
+ "loss": 0.0006,
1199
+ "step": 4325
1200
+ },
1201
+ {
1202
+ "epoch": 3.07,
1203
+ "learning_rate": 1.4555555555555556e-05,
1204
+ "loss": 0.0003,
1205
+ "step": 4350
1206
+ },
1207
+ {
1208
+ "epoch": 3.07,
1209
+ "learning_rate": 1.4000000000000001e-05,
1210
+ "loss": 0.0002,
1211
+ "step": 4375
1212
+ },
1213
+ {
1214
+ "epoch": 3.08,
1215
+ "learning_rate": 1.3444444444444445e-05,
1216
+ "loss": 0.0016,
1217
+ "step": 4400
1218
+ },
1219
+ {
1220
+ "epoch": 3.08,
1221
+ "learning_rate": 1.2888888888888889e-05,
1222
+ "loss": 0.0004,
1223
+ "step": 4425
1224
+ },
1225
+ {
1226
+ "epoch": 3.09,
1227
+ "learning_rate": 1.2333333333333334e-05,
1228
+ "loss": 0.0003,
1229
+ "step": 4450
1230
+ },
1231
+ {
1232
+ "epoch": 3.09,
1233
+ "learning_rate": 1.1777777777777778e-05,
1234
+ "loss": 0.0003,
1235
+ "step": 4475
1236
+ },
1237
+ {
1238
+ "epoch": 3.1,
1239
+ "learning_rate": 1.1222222222222224e-05,
1240
+ "loss": 0.0003,
1241
+ "step": 4500
1242
+ },
1243
+ {
1244
+ "epoch": 3.1,
1245
+ "eval_loss": 0.08255437761545181,
1246
+ "eval_runtime": 3117.117,
1247
+ "eval_samples_per_second": 7.517,
1248
+ "eval_steps_per_second": 0.118,
1249
+ "eval_wer": 5.779987759166161,
1250
+ "step": 4500
1251
+ },
1252
+ {
1253
+ "epoch": 3.1,
1254
+ "learning_rate": 1.0666666666666667e-05,
1255
+ "loss": 0.0001,
1256
+ "step": 4525
1257
+ },
1258
+ {
1259
+ "epoch": 3.11,
1260
+ "learning_rate": 1.0111111111111111e-05,
1261
+ "loss": 0.0002,
1262
+ "step": 4550
1263
+ },
1264
+ {
1265
+ "epoch": 3.11,
1266
+ "learning_rate": 9.555555555555556e-06,
1267
+ "loss": 0.0001,
1268
+ "step": 4575
1269
+ },
1270
+ {
1271
+ "epoch": 3.12,
1272
+ "learning_rate": 9e-06,
1273
+ "loss": 0.0003,
1274
+ "step": 4600
1275
+ },
1276
+ {
1277
+ "epoch": 3.12,
1278
+ "learning_rate": 8.444444444444446e-06,
1279
+ "loss": 0.0005,
1280
+ "step": 4625
1281
+ },
1282
+ {
1283
+ "epoch": 3.13,
1284
+ "learning_rate": 7.88888888888889e-06,
1285
+ "loss": 0.0005,
1286
+ "step": 4650
1287
+ },
1288
+ {
1289
+ "epoch": 3.13,
1290
+ "learning_rate": 7.333333333333334e-06,
1291
+ "loss": 0.0008,
1292
+ "step": 4675
1293
+ },
1294
+ {
1295
+ "epoch": 3.14,
1296
+ "learning_rate": 6.777777777777779e-06,
1297
+ "loss": 0.0005,
1298
+ "step": 4700
1299
+ },
1300
+ {
1301
+ "epoch": 3.14,
1302
+ "learning_rate": 6.222222222222222e-06,
1303
+ "loss": 0.0004,
1304
+ "step": 4725
1305
+ },
1306
+ {
1307
+ "epoch": 3.15,
1308
+ "learning_rate": 5.666666666666667e-06,
1309
+ "loss": 0.0006,
1310
+ "step": 4750
1311
+ },
1312
+ {
1313
+ "epoch": 3.15,
1314
+ "eval_loss": 0.08329758793115616,
1315
+ "eval_runtime": 3126.9679,
1316
+ "eval_samples_per_second": 7.493,
1317
+ "eval_steps_per_second": 0.117,
1318
+ "eval_wer": 5.762529853526142,
1319
+ "step": 4750
1320
+ },
1321
+ {
1322
+ "epoch": 3.15,
1323
+ "learning_rate": 5.1111111111111115e-06,
1324
+ "loss": 0.0002,
1325
+ "step": 4775
1326
+ },
1327
+ {
1328
+ "epoch": 3.16,
1329
+ "learning_rate": 4.555555555555556e-06,
1330
+ "loss": 0.0003,
1331
+ "step": 4800
1332
+ },
1333
+ {
1334
+ "epoch": 3.16,
1335
+ "learning_rate": 4.000000000000001e-06,
1336
+ "loss": 0.0004,
1337
+ "step": 4825
1338
+ },
1339
+ {
1340
+ "epoch": 3.17,
1341
+ "learning_rate": 3.4444444444444444e-06,
1342
+ "loss": 0.0001,
1343
+ "step": 4850
1344
+ },
1345
+ {
1346
+ "epoch": 3.17,
1347
+ "learning_rate": 2.888888888888889e-06,
1348
+ "loss": 0.0002,
1349
+ "step": 4875
1350
+ },
1351
+ {
1352
+ "epoch": 3.18,
1353
+ "learning_rate": 2.3333333333333336e-06,
1354
+ "loss": 0.0006,
1355
+ "step": 4900
1356
+ },
1357
+ {
1358
+ "epoch": 3.18,
1359
+ "learning_rate": 1.777777777777778e-06,
1360
+ "loss": 0.0004,
1361
+ "step": 4925
1362
+ },
1363
+ {
1364
+ "epoch": 3.19,
1365
+ "learning_rate": 1.2222222222222223e-06,
1366
+ "loss": 0.0001,
1367
+ "step": 4950
1368
+ },
1369
+ {
1370
+ "epoch": 3.19,
1371
+ "learning_rate": 6.666666666666667e-07,
1372
+ "loss": 0.0005,
1373
+ "step": 4975
1374
+ },
1375
+ {
1376
+ "epoch": 3.2,
1377
+ "learning_rate": 1.1111111111111112e-07,
1378
+ "loss": 0.0003,
1379
+ "step": 5000
1380
+ },
1381
+ {
1382
+ "epoch": 3.2,
1383
+ "eval_loss": 0.08391360193490982,
1384
+ "eval_runtime": 3124.2207,
1385
+ "eval_samples_per_second": 7.5,
1386
+ "eval_steps_per_second": 0.117,
1387
+ "eval_wer": 5.754409897414504,
1388
+ "step": 5000
1389
+ },
1390
+ {
1391
+ "epoch": 3.2,
1392
+ "step": 5000,
1393
+ "total_flos": 4.151039849534587e+19,
1394
+ "train_loss": 0.024749002781091257,
1395
+ "train_runtime": 101889.8071,
1396
+ "train_samples_per_second": 6.281,
1397
+ "train_steps_per_second": 0.049
1398
+ }
1399
+ ],
1400
+ "max_steps": 5000,
1401
+ "num_train_epochs": 9223372036854775807,
1402
+ "total_flos": 4.151039849534587e+19,
1403
+ "trial_name": null,
1404
+ "trial_params": null
1405
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:91c6a8514089dbd3f3c1004bb6219bbeb8d96a44f7da0161c1bcdbe3ec8b2485
3
+ size 3643