add best WER checkopoint at 3000 steps
Browse files- checkpoint-3000/config.json +41 -0
- checkpoint-3000/optimizer.pt +3 -0
- checkpoint-3000/preprocessor_config.json +0 -0
- checkpoint-3000/pytorch_model.bin +3 -0
- checkpoint-3000/rng_state.pth +3 -0
- checkpoint-3000/scaler.pt +3 -0
- checkpoint-3000/scheduler.pt +3 -0
- checkpoint-3000/trainer_state.json +934 -0
- checkpoint-3000/training_args.bin +3 -0
checkpoint-3000/config.json
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "openai/whisper-small",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"activation_function": "gelu",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"WhisperForConditionalGeneration"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"begin_suppress_tokens": [
|
| 10 |
+
220,
|
| 11 |
+
50257
|
| 12 |
+
],
|
| 13 |
+
"bos_token_id": 50257,
|
| 14 |
+
"d_model": 768,
|
| 15 |
+
"decoder_attention_heads": 12,
|
| 16 |
+
"decoder_ffn_dim": 3072,
|
| 17 |
+
"decoder_layerdrop": 0.0,
|
| 18 |
+
"decoder_layers": 12,
|
| 19 |
+
"decoder_start_token_id": 50258,
|
| 20 |
+
"dropout": 0.0,
|
| 21 |
+
"encoder_attention_heads": 12,
|
| 22 |
+
"encoder_ffn_dim": 3072,
|
| 23 |
+
"encoder_layerdrop": 0.0,
|
| 24 |
+
"encoder_layers": 12,
|
| 25 |
+
"eos_token_id": 50257,
|
| 26 |
+
"forced_decoder_ids": null,
|
| 27 |
+
"init_std": 0.02,
|
| 28 |
+
"is_encoder_decoder": true,
|
| 29 |
+
"max_length": 448,
|
| 30 |
+
"max_source_positions": 1500,
|
| 31 |
+
"max_target_positions": 448,
|
| 32 |
+
"model_type": "whisper",
|
| 33 |
+
"num_hidden_layers": 12,
|
| 34 |
+
"num_mel_bins": 80,
|
| 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 |
+
}
|
checkpoint-3000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:69c6311664a33bc54b6c97e08f1e9c9235fda8590af92fe3a3ab687f57cf6778
|
| 3 |
+
size 1934161093
|
checkpoint-3000/preprocessor_config.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-3000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5764b46edba29c43965ebbb92f48c0f108fbbbc77cfaa224dcfb807e88c5e55a
|
| 3 |
+
size 967102601
|
checkpoint-3000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:40cb3deb4838cfec86261ac1bfe3ea3a3f050eb56b5930a641925eef640c40e4
|
| 3 |
+
size 14575
|
checkpoint-3000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:67052f7ccf0c314ec66f8a5e6561069877b7b24a3a43dcf29f3baded513f272d
|
| 3 |
+
size 557
|
checkpoint-3000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1ef06d0cdaf008a8194b49ac5e048244aa436a656f010611ec32a4be618c2ef4
|
| 3 |
+
size 627
|
checkpoint-3000/trainer_state.json
ADDED
|
@@ -0,0 +1,934 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 230.27391041162227,
|
| 3 |
+
"best_model_checkpoint": "./checkpoint-100",
|
| 4 |
+
"epoch": 428.57142857142856,
|
| 5 |
+
"global_step": 3000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 3.57,
|
| 12 |
+
"learning_rate": 1.32e-07,
|
| 13 |
+
"loss": 2.8512,
|
| 14 |
+
"step": 25
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 7.14,
|
| 18 |
+
"learning_rate": 2.8199999999999996e-07,
|
| 19 |
+
"loss": 2.7082,
|
| 20 |
+
"step": 50
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 10.71,
|
| 24 |
+
"learning_rate": 2.9305263157894735e-07,
|
| 25 |
+
"loss": 2.3515,
|
| 26 |
+
"step": 75
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 14.29,
|
| 30 |
+
"learning_rate": 2.851578947368421e-07,
|
| 31 |
+
"loss": 2.0871,
|
| 32 |
+
"step": 100
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 14.29,
|
| 36 |
+
"eval_loss": 2.0101583003997803,
|
| 37 |
+
"eval_runtime": 720.7636,
|
| 38 |
+
"eval_samples_per_second": 0.71,
|
| 39 |
+
"eval_steps_per_second": 0.044,
|
| 40 |
+
"eval_wer": 230.27391041162227,
|
| 41 |
+
"step": 100
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 17.86,
|
| 45 |
+
"learning_rate": 2.7726315789473684e-07,
|
| 46 |
+
"loss": 1.8622,
|
| 47 |
+
"step": 125
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 21.43,
|
| 51 |
+
"learning_rate": 2.693684210526316e-07,
|
| 52 |
+
"loss": 1.7104,
|
| 53 |
+
"step": 150
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 25.0,
|
| 57 |
+
"learning_rate": 2.614736842105263e-07,
|
| 58 |
+
"loss": 1.5736,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 28.57,
|
| 63 |
+
"learning_rate": 2.53578947368421e-07,
|
| 64 |
+
"loss": 1.465,
|
| 65 |
+
"step": 200
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 28.57,
|
| 69 |
+
"eval_loss": 1.4968725442886353,
|
| 70 |
+
"eval_runtime": 628.5425,
|
| 71 |
+
"eval_samples_per_second": 0.815,
|
| 72 |
+
"eval_steps_per_second": 0.051,
|
| 73 |
+
"eval_wer": 137.24273607748182,
|
| 74 |
+
"step": 200
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 32.14,
|
| 78 |
+
"learning_rate": 2.4568421052631577e-07,
|
| 79 |
+
"loss": 1.3669,
|
| 80 |
+
"step": 225
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 35.71,
|
| 84 |
+
"learning_rate": 2.3778947368421054e-07,
|
| 85 |
+
"loss": 1.2898,
|
| 86 |
+
"step": 250
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 39.29,
|
| 90 |
+
"learning_rate": 2.2989473684210523e-07,
|
| 91 |
+
"loss": 1.2205,
|
| 92 |
+
"step": 275
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 42.86,
|
| 96 |
+
"learning_rate": 2.2199999999999998e-07,
|
| 97 |
+
"loss": 1.1617,
|
| 98 |
+
"step": 300
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 42.86,
|
| 102 |
+
"eval_loss": 1.2715740203857422,
|
| 103 |
+
"eval_runtime": 499.3378,
|
| 104 |
+
"eval_samples_per_second": 1.025,
|
| 105 |
+
"eval_steps_per_second": 0.064,
|
| 106 |
+
"eval_wer": 76.32415254237289,
|
| 107 |
+
"step": 300
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 46.43,
|
| 111 |
+
"learning_rate": 2.1410526315789472e-07,
|
| 112 |
+
"loss": 1.1091,
|
| 113 |
+
"step": 325
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 50.0,
|
| 117 |
+
"learning_rate": 2.0621052631578947e-07,
|
| 118 |
+
"loss": 1.0738,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 53.57,
|
| 123 |
+
"learning_rate": 1.9831578947368419e-07,
|
| 124 |
+
"loss": 1.033,
|
| 125 |
+
"step": 375
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 57.14,
|
| 129 |
+
"learning_rate": 1.9042105263157893e-07,
|
| 130 |
+
"loss": 1.0019,
|
| 131 |
+
"step": 400
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 57.14,
|
| 135 |
+
"eval_loss": 1.16450834274292,
|
| 136 |
+
"eval_runtime": 480.3701,
|
| 137 |
+
"eval_samples_per_second": 1.066,
|
| 138 |
+
"eval_steps_per_second": 0.067,
|
| 139 |
+
"eval_wer": 71.37560532687651,
|
| 140 |
+
"step": 400
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 60.71,
|
| 144 |
+
"learning_rate": 1.8252631578947368e-07,
|
| 145 |
+
"loss": 0.9712,
|
| 146 |
+
"step": 425
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 64.29,
|
| 150 |
+
"learning_rate": 1.7463157894736842e-07,
|
| 151 |
+
"loss": 0.9437,
|
| 152 |
+
"step": 450
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 67.86,
|
| 156 |
+
"learning_rate": 1.6673684210526314e-07,
|
| 157 |
+
"loss": 0.9215,
|
| 158 |
+
"step": 475
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 71.43,
|
| 162 |
+
"learning_rate": 1.588421052631579e-07,
|
| 163 |
+
"loss": 0.9052,
|
| 164 |
+
"step": 500
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 71.43,
|
| 168 |
+
"eval_loss": 1.1051170825958252,
|
| 169 |
+
"eval_runtime": 486.3374,
|
| 170 |
+
"eval_samples_per_second": 1.053,
|
| 171 |
+
"eval_steps_per_second": 0.066,
|
| 172 |
+
"eval_wer": 69.78662227602905,
|
| 173 |
+
"step": 500
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 75.0,
|
| 177 |
+
"learning_rate": 1.5094736842105263e-07,
|
| 178 |
+
"loss": 0.8773,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 78.57,
|
| 183 |
+
"learning_rate": 1.4305263157894735e-07,
|
| 184 |
+
"loss": 0.8643,
|
| 185 |
+
"step": 550
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 82.14,
|
| 189 |
+
"learning_rate": 1.351578947368421e-07,
|
| 190 |
+
"loss": 0.8449,
|
| 191 |
+
"step": 575
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 85.71,
|
| 195 |
+
"learning_rate": 1.2726315789473684e-07,
|
| 196 |
+
"loss": 0.8334,
|
| 197 |
+
"step": 600
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 85.71,
|
| 201 |
+
"eval_loss": 1.0691100358963013,
|
| 202 |
+
"eval_runtime": 478.3463,
|
| 203 |
+
"eval_samples_per_second": 1.07,
|
| 204 |
+
"eval_steps_per_second": 0.067,
|
| 205 |
+
"eval_wer": 68.26573849878935,
|
| 206 |
+
"step": 600
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 89.29,
|
| 210 |
+
"learning_rate": 1.1936842105263156e-07,
|
| 211 |
+
"loss": 0.8132,
|
| 212 |
+
"step": 625
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 92.86,
|
| 216 |
+
"learning_rate": 1.1147368421052631e-07,
|
| 217 |
+
"loss": 0.8058,
|
| 218 |
+
"step": 650
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 96.43,
|
| 222 |
+
"learning_rate": 1.0357894736842104e-07,
|
| 223 |
+
"loss": 0.7913,
|
| 224 |
+
"step": 675
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 100.0,
|
| 228 |
+
"learning_rate": 9.568421052631579e-08,
|
| 229 |
+
"loss": 0.7838,
|
| 230 |
+
"step": 700
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 100.0,
|
| 234 |
+
"eval_loss": 1.0482958555221558,
|
| 235 |
+
"eval_runtime": 478.3861,
|
| 236 |
+
"eval_samples_per_second": 1.07,
|
| 237 |
+
"eval_steps_per_second": 0.067,
|
| 238 |
+
"eval_wer": 67.16858353510897,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 103.57,
|
| 243 |
+
"learning_rate": 8.778947368421052e-08,
|
| 244 |
+
"loss": 0.7768,
|
| 245 |
+
"step": 725
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 107.14,
|
| 249 |
+
"learning_rate": 7.989473684210526e-08,
|
| 250 |
+
"loss": 0.7673,
|
| 251 |
+
"step": 750
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 110.71,
|
| 255 |
+
"learning_rate": 7.2e-08,
|
| 256 |
+
"loss": 0.7643,
|
| 257 |
+
"step": 775
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 114.29,
|
| 261 |
+
"learning_rate": 6.410526315789473e-08,
|
| 262 |
+
"loss": 0.7539,
|
| 263 |
+
"step": 800
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 114.29,
|
| 267 |
+
"eval_loss": 1.0362622737884521,
|
| 268 |
+
"eval_runtime": 484.8883,
|
| 269 |
+
"eval_samples_per_second": 1.056,
|
| 270 |
+
"eval_steps_per_second": 0.066,
|
| 271 |
+
"eval_wer": 66.41949152542372,
|
| 272 |
+
"step": 800
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 117.86,
|
| 276 |
+
"learning_rate": 5.621052631578947e-08,
|
| 277 |
+
"loss": 0.7527,
|
| 278 |
+
"step": 825
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 121.43,
|
| 282 |
+
"learning_rate": 4.8315789473684206e-08,
|
| 283 |
+
"loss": 0.7441,
|
| 284 |
+
"step": 850
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 125.0,
|
| 288 |
+
"learning_rate": 4.0421052631578945e-08,
|
| 289 |
+
"loss": 0.7417,
|
| 290 |
+
"step": 875
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 128.57,
|
| 294 |
+
"learning_rate": 3.2526315789473684e-08,
|
| 295 |
+
"loss": 0.7377,
|
| 296 |
+
"step": 900
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 128.57,
|
| 300 |
+
"eval_loss": 1.0297424793243408,
|
| 301 |
+
"eval_runtime": 471.9816,
|
| 302 |
+
"eval_samples_per_second": 1.085,
|
| 303 |
+
"eval_steps_per_second": 0.068,
|
| 304 |
+
"eval_wer": 66.20006053268766,
|
| 305 |
+
"step": 900
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 132.14,
|
| 309 |
+
"learning_rate": 2.463157894736842e-08,
|
| 310 |
+
"loss": 0.7387,
|
| 311 |
+
"step": 925
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 135.71,
|
| 315 |
+
"learning_rate": 1.673684210526316e-08,
|
| 316 |
+
"loss": 0.7329,
|
| 317 |
+
"step": 950
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 139.29,
|
| 321 |
+
"learning_rate": 8.842105263157895e-09,
|
| 322 |
+
"loss": 0.7312,
|
| 323 |
+
"step": 975
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 142.86,
|
| 327 |
+
"learning_rate": 9.473684210526316e-10,
|
| 328 |
+
"loss": 0.7325,
|
| 329 |
+
"step": 1000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 142.86,
|
| 333 |
+
"eval_loss": 1.0276601314544678,
|
| 334 |
+
"eval_runtime": 477.2948,
|
| 335 |
+
"eval_samples_per_second": 1.073,
|
| 336 |
+
"eval_steps_per_second": 0.067,
|
| 337 |
+
"eval_wer": 66.00332929782083,
|
| 338 |
+
"step": 1000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 146.43,
|
| 342 |
+
"learning_rate": 1.5046153846153844e-07,
|
| 343 |
+
"loss": 0.7238,
|
| 344 |
+
"step": 1025
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 150.0,
|
| 348 |
+
"learning_rate": 1.466153846153846e-07,
|
| 349 |
+
"loss": 0.7163,
|
| 350 |
+
"step": 1050
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 153.57,
|
| 354 |
+
"learning_rate": 1.4276923076923076e-07,
|
| 355 |
+
"loss": 0.7028,
|
| 356 |
+
"step": 1075
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 157.14,
|
| 360 |
+
"learning_rate": 1.389230769230769e-07,
|
| 361 |
+
"loss": 0.6952,
|
| 362 |
+
"step": 1100
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 157.14,
|
| 366 |
+
"eval_loss": 1.0121634006500244,
|
| 367 |
+
"eval_runtime": 471.3724,
|
| 368 |
+
"eval_samples_per_second": 1.086,
|
| 369 |
+
"eval_steps_per_second": 0.068,
|
| 370 |
+
"eval_wer": 65.05750605326877,
|
| 371 |
+
"step": 1100
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 160.71,
|
| 375 |
+
"learning_rate": 1.3507692307692308e-07,
|
| 376 |
+
"loss": 0.6843,
|
| 377 |
+
"step": 1125
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 164.29,
|
| 381 |
+
"learning_rate": 1.3123076923076923e-07,
|
| 382 |
+
"loss": 0.6699,
|
| 383 |
+
"step": 1150
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 167.86,
|
| 387 |
+
"learning_rate": 1.2738461538461538e-07,
|
| 388 |
+
"loss": 0.6671,
|
| 389 |
+
"step": 1175
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 171.43,
|
| 393 |
+
"learning_rate": 1.2353846153846153e-07,
|
| 394 |
+
"loss": 0.6531,
|
| 395 |
+
"step": 1200
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 171.43,
|
| 399 |
+
"eval_loss": 1.0014406442642212,
|
| 400 |
+
"eval_runtime": 475.4519,
|
| 401 |
+
"eval_samples_per_second": 1.077,
|
| 402 |
+
"eval_steps_per_second": 0.067,
|
| 403 |
+
"eval_wer": 64.42191283292978,
|
| 404 |
+
"step": 1200
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 175.0,
|
| 408 |
+
"learning_rate": 1.1969230769230767e-07,
|
| 409 |
+
"loss": 0.6487,
|
| 410 |
+
"step": 1225
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 178.57,
|
| 414 |
+
"learning_rate": 1.1584615384615385e-07,
|
| 415 |
+
"loss": 0.6369,
|
| 416 |
+
"step": 1250
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 182.14,
|
| 420 |
+
"learning_rate": 1.12e-07,
|
| 421 |
+
"loss": 0.6336,
|
| 422 |
+
"step": 1275
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 185.71,
|
| 426 |
+
"learning_rate": 1.0815384615384614e-07,
|
| 427 |
+
"loss": 0.6189,
|
| 428 |
+
"step": 1300
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 185.71,
|
| 432 |
+
"eval_loss": 0.9944669008255005,
|
| 433 |
+
"eval_runtime": 470.4039,
|
| 434 |
+
"eval_samples_per_second": 1.088,
|
| 435 |
+
"eval_steps_per_second": 0.068,
|
| 436 |
+
"eval_wer": 63.79388619854721,
|
| 437 |
+
"step": 1300
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 189.29,
|
| 441 |
+
"learning_rate": 1.043076923076923e-07,
|
| 442 |
+
"loss": 0.6213,
|
| 443 |
+
"step": 1325
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 192.86,
|
| 447 |
+
"learning_rate": 1.0046153846153845e-07,
|
| 448 |
+
"loss": 0.608,
|
| 449 |
+
"step": 1350
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 196.43,
|
| 453 |
+
"learning_rate": 9.66153846153846e-08,
|
| 454 |
+
"loss": 0.6029,
|
| 455 |
+
"step": 1375
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 200.0,
|
| 459 |
+
"learning_rate": 9.276923076923078e-08,
|
| 460 |
+
"loss": 0.5993,
|
| 461 |
+
"step": 1400
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 200.0,
|
| 465 |
+
"eval_loss": 0.9895604252815247,
|
| 466 |
+
"eval_runtime": 473.2431,
|
| 467 |
+
"eval_samples_per_second": 1.082,
|
| 468 |
+
"eval_steps_per_second": 0.068,
|
| 469 |
+
"eval_wer": 63.35502421307506,
|
| 470 |
+
"step": 1400
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 203.57,
|
| 474 |
+
"learning_rate": 8.892307692307692e-08,
|
| 475 |
+
"loss": 0.593,
|
| 476 |
+
"step": 1425
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 207.14,
|
| 480 |
+
"learning_rate": 8.507692307692307e-08,
|
| 481 |
+
"loss": 0.5817,
|
| 482 |
+
"step": 1450
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 210.71,
|
| 486 |
+
"learning_rate": 8.123076923076922e-08,
|
| 487 |
+
"loss": 0.5782,
|
| 488 |
+
"step": 1475
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 214.29,
|
| 492 |
+
"learning_rate": 7.738461538461538e-08,
|
| 493 |
+
"loss": 0.5757,
|
| 494 |
+
"step": 1500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 214.29,
|
| 498 |
+
"eval_loss": 0.9864457845687866,
|
| 499 |
+
"eval_runtime": 474.414,
|
| 500 |
+
"eval_samples_per_second": 1.079,
|
| 501 |
+
"eval_steps_per_second": 0.067,
|
| 502 |
+
"eval_wer": 63.22639225181598,
|
| 503 |
+
"step": 1500
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 217.86,
|
| 507 |
+
"learning_rate": 7.353846153846153e-08,
|
| 508 |
+
"loss": 0.5706,
|
| 509 |
+
"step": 1525
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 221.43,
|
| 513 |
+
"learning_rate": 6.969230769230769e-08,
|
| 514 |
+
"loss": 0.5624,
|
| 515 |
+
"step": 1550
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 225.0,
|
| 519 |
+
"learning_rate": 6.584615384615385e-08,
|
| 520 |
+
"loss": 0.5638,
|
| 521 |
+
"step": 1575
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 228.57,
|
| 525 |
+
"learning_rate": 6.2e-08,
|
| 526 |
+
"loss": 0.5601,
|
| 527 |
+
"step": 1600
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 228.57,
|
| 531 |
+
"eval_loss": 0.9844600558280945,
|
| 532 |
+
"eval_runtime": 478.7212,
|
| 533 |
+
"eval_samples_per_second": 1.07,
|
| 534 |
+
"eval_steps_per_second": 0.067,
|
| 535 |
+
"eval_wer": 62.916162227602904,
|
| 536 |
+
"step": 1600
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 232.14,
|
| 540 |
+
"learning_rate": 5.815384615384615e-08,
|
| 541 |
+
"loss": 0.5537,
|
| 542 |
+
"step": 1625
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 235.71,
|
| 546 |
+
"learning_rate": 5.430769230769231e-08,
|
| 547 |
+
"loss": 0.5488,
|
| 548 |
+
"step": 1650
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 239.29,
|
| 552 |
+
"learning_rate": 5.0461538461538456e-08,
|
| 553 |
+
"loss": 0.5479,
|
| 554 |
+
"step": 1675
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 242.86,
|
| 558 |
+
"learning_rate": 4.661538461538461e-08,
|
| 559 |
+
"loss": 0.5482,
|
| 560 |
+
"step": 1700
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 242.86,
|
| 564 |
+
"eval_loss": 0.9833234548568726,
|
| 565 |
+
"eval_runtime": 488.3079,
|
| 566 |
+
"eval_samples_per_second": 1.049,
|
| 567 |
+
"eval_steps_per_second": 0.066,
|
| 568 |
+
"eval_wer": 62.817796610169495,
|
| 569 |
+
"step": 1700
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 246.43,
|
| 573 |
+
"learning_rate": 4.2769230769230765e-08,
|
| 574 |
+
"loss": 0.5441,
|
| 575 |
+
"step": 1725
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 250.0,
|
| 579 |
+
"learning_rate": 3.892307692307692e-08,
|
| 580 |
+
"loss": 0.54,
|
| 581 |
+
"step": 1750
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 253.57,
|
| 585 |
+
"learning_rate": 3.5076923076923074e-08,
|
| 586 |
+
"loss": 0.538,
|
| 587 |
+
"step": 1775
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 257.14,
|
| 591 |
+
"learning_rate": 3.123076923076923e-08,
|
| 592 |
+
"loss": 0.5382,
|
| 593 |
+
"step": 1800
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 257.14,
|
| 597 |
+
"eval_loss": 0.9826769828796387,
|
| 598 |
+
"eval_runtime": 486.3686,
|
| 599 |
+
"eval_samples_per_second": 1.053,
|
| 600 |
+
"eval_steps_per_second": 0.066,
|
| 601 |
+
"eval_wer": 62.84049636803874,
|
| 602 |
+
"step": 1800
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 260.71,
|
| 606 |
+
"learning_rate": 2.7384615384615387e-08,
|
| 607 |
+
"loss": 0.5343,
|
| 608 |
+
"step": 1825
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 264.29,
|
| 612 |
+
"learning_rate": 2.3538461538461535e-08,
|
| 613 |
+
"loss": 0.5313,
|
| 614 |
+
"step": 1850
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 267.86,
|
| 618 |
+
"learning_rate": 1.9692307692307693e-08,
|
| 619 |
+
"loss": 0.5318,
|
| 620 |
+
"step": 1875
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 271.43,
|
| 624 |
+
"learning_rate": 1.5846153846153844e-08,
|
| 625 |
+
"loss": 0.5325,
|
| 626 |
+
"step": 1900
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 271.43,
|
| 630 |
+
"eval_loss": 0.9823360443115234,
|
| 631 |
+
"eval_runtime": 483.6609,
|
| 632 |
+
"eval_samples_per_second": 1.059,
|
| 633 |
+
"eval_steps_per_second": 0.066,
|
| 634 |
+
"eval_wer": 62.76483050847458,
|
| 635 |
+
"step": 1900
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 275.0,
|
| 639 |
+
"learning_rate": 1.2e-08,
|
| 640 |
+
"loss": 0.529,
|
| 641 |
+
"step": 1925
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 278.57,
|
| 645 |
+
"learning_rate": 8.153846153846154e-09,
|
| 646 |
+
"loss": 0.5294,
|
| 647 |
+
"step": 1950
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 282.14,
|
| 651 |
+
"learning_rate": 4.307692307692307e-09,
|
| 652 |
+
"loss": 0.525,
|
| 653 |
+
"step": 1975
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 285.71,
|
| 657 |
+
"learning_rate": 4.615384615384615e-10,
|
| 658 |
+
"loss": 0.5287,
|
| 659 |
+
"step": 2000
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 285.71,
|
| 663 |
+
"eval_loss": 0.9822061061859131,
|
| 664 |
+
"eval_runtime": 484.044,
|
| 665 |
+
"eval_samples_per_second": 1.058,
|
| 666 |
+
"eval_steps_per_second": 0.066,
|
| 667 |
+
"eval_wer": 62.817796610169495,
|
| 668 |
+
"step": 2000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 289.29,
|
| 672 |
+
"learning_rate": 1.9853333333333334e-07,
|
| 673 |
+
"loss": 0.523,
|
| 674 |
+
"step": 2025
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 292.86,
|
| 678 |
+
"learning_rate": 1.9686666666666667e-07,
|
| 679 |
+
"loss": 0.5133,
|
| 680 |
+
"step": 2050
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 296.43,
|
| 684 |
+
"learning_rate": 1.9519999999999997e-07,
|
| 685 |
+
"loss": 0.4999,
|
| 686 |
+
"step": 2075
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 300.0,
|
| 690 |
+
"learning_rate": 1.935333333333333e-07,
|
| 691 |
+
"loss": 0.4924,
|
| 692 |
+
"step": 2100
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 303.57,
|
| 696 |
+
"learning_rate": 1.9186666666666666e-07,
|
| 697 |
+
"loss": 0.4855,
|
| 698 |
+
"step": 2125
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 307.14,
|
| 702 |
+
"learning_rate": 1.902e-07,
|
| 703 |
+
"loss": 0.4701,
|
| 704 |
+
"step": 2150
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 310.71,
|
| 708 |
+
"learning_rate": 1.8853333333333333e-07,
|
| 709 |
+
"loss": 0.4601,
|
| 710 |
+
"step": 2175
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 314.29,
|
| 714 |
+
"learning_rate": 1.8686666666666669e-07,
|
| 715 |
+
"loss": 0.4525,
|
| 716 |
+
"step": 2200
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 317.86,
|
| 720 |
+
"learning_rate": 1.852e-07,
|
| 721 |
+
"loss": 0.4448,
|
| 722 |
+
"step": 2225
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 321.43,
|
| 726 |
+
"learning_rate": 1.8353333333333332e-07,
|
| 727 |
+
"loss": 0.4364,
|
| 728 |
+
"step": 2250
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 325.0,
|
| 732 |
+
"learning_rate": 1.8186666666666665e-07,
|
| 733 |
+
"loss": 0.4232,
|
| 734 |
+
"step": 2275
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 328.57,
|
| 738 |
+
"learning_rate": 1.8019999999999999e-07,
|
| 739 |
+
"loss": 0.4163,
|
| 740 |
+
"step": 2300
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 332.14,
|
| 744 |
+
"learning_rate": 1.7853333333333334e-07,
|
| 745 |
+
"loss": 0.4089,
|
| 746 |
+
"step": 2325
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 335.71,
|
| 750 |
+
"learning_rate": 1.7686666666666668e-07,
|
| 751 |
+
"loss": 0.4031,
|
| 752 |
+
"step": 2350
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 339.29,
|
| 756 |
+
"learning_rate": 1.7519999999999998e-07,
|
| 757 |
+
"loss": 0.3887,
|
| 758 |
+
"step": 2375
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 342.86,
|
| 762 |
+
"learning_rate": 1.735333333333333e-07,
|
| 763 |
+
"loss": 0.3826,
|
| 764 |
+
"step": 2400
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 346.43,
|
| 768 |
+
"learning_rate": 1.7186666666666667e-07,
|
| 769 |
+
"loss": 0.3766,
|
| 770 |
+
"step": 2425
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 350.0,
|
| 774 |
+
"learning_rate": 1.702e-07,
|
| 775 |
+
"loss": 0.3647,
|
| 776 |
+
"step": 2450
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 353.57,
|
| 780 |
+
"learning_rate": 1.6853333333333333e-07,
|
| 781 |
+
"loss": 0.3601,
|
| 782 |
+
"step": 2475
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 357.14,
|
| 786 |
+
"learning_rate": 1.6686666666666664e-07,
|
| 787 |
+
"loss": 0.3494,
|
| 788 |
+
"step": 2500
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 357.14,
|
| 792 |
+
"eval_loss": 1.0025562047958374,
|
| 793 |
+
"eval_runtime": 489.174,
|
| 794 |
+
"eval_samples_per_second": 1.047,
|
| 795 |
+
"eval_steps_per_second": 0.065,
|
| 796 |
+
"eval_wer": 61.61470944309927,
|
| 797 |
+
"step": 2500
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 360.71,
|
| 801 |
+
"learning_rate": 1.652e-07,
|
| 802 |
+
"loss": 0.3448,
|
| 803 |
+
"step": 2525
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 364.29,
|
| 807 |
+
"learning_rate": 1.6353333333333333e-07,
|
| 808 |
+
"loss": 0.3367,
|
| 809 |
+
"step": 2550
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 367.86,
|
| 813 |
+
"learning_rate": 1.6186666666666666e-07,
|
| 814 |
+
"loss": 0.3308,
|
| 815 |
+
"step": 2575
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 371.43,
|
| 819 |
+
"learning_rate": 1.602e-07,
|
| 820 |
+
"loss": 0.3226,
|
| 821 |
+
"step": 2600
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 375.0,
|
| 825 |
+
"learning_rate": 1.5853333333333335e-07,
|
| 826 |
+
"loss": 0.3165,
|
| 827 |
+
"step": 2625
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 378.57,
|
| 831 |
+
"learning_rate": 1.5686666666666666e-07,
|
| 832 |
+
"loss": 0.3099,
|
| 833 |
+
"step": 2650
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 382.14,
|
| 837 |
+
"learning_rate": 1.552e-07,
|
| 838 |
+
"loss": 0.3021,
|
| 839 |
+
"step": 2675
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 385.71,
|
| 843 |
+
"learning_rate": 1.5353333333333332e-07,
|
| 844 |
+
"loss": 0.2964,
|
| 845 |
+
"step": 2700
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 389.29,
|
| 849 |
+
"learning_rate": 1.5186666666666668e-07,
|
| 850 |
+
"loss": 0.2901,
|
| 851 |
+
"step": 2725
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 392.86,
|
| 855 |
+
"learning_rate": 1.502e-07,
|
| 856 |
+
"loss": 0.284,
|
| 857 |
+
"step": 2750
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 396.43,
|
| 861 |
+
"learning_rate": 1.4853333333333334e-07,
|
| 862 |
+
"loss": 0.279,
|
| 863 |
+
"step": 2775
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 400.0,
|
| 867 |
+
"learning_rate": 1.4686666666666667e-07,
|
| 868 |
+
"loss": 0.2715,
|
| 869 |
+
"step": 2800
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 403.57,
|
| 873 |
+
"learning_rate": 1.4519999999999998e-07,
|
| 874 |
+
"loss": 0.2646,
|
| 875 |
+
"step": 2825
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 407.14,
|
| 879 |
+
"learning_rate": 1.4353333333333333e-07,
|
| 880 |
+
"loss": 0.2606,
|
| 881 |
+
"step": 2850
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 410.71,
|
| 885 |
+
"learning_rate": 1.4186666666666667e-07,
|
| 886 |
+
"loss": 0.2564,
|
| 887 |
+
"step": 2875
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 414.29,
|
| 891 |
+
"learning_rate": 1.402e-07,
|
| 892 |
+
"loss": 0.2486,
|
| 893 |
+
"step": 2900
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 417.86,
|
| 897 |
+
"learning_rate": 1.3853333333333333e-07,
|
| 898 |
+
"loss": 0.2463,
|
| 899 |
+
"step": 2925
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 421.43,
|
| 903 |
+
"learning_rate": 1.3686666666666666e-07,
|
| 904 |
+
"loss": 0.239,
|
| 905 |
+
"step": 2950
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 425.0,
|
| 909 |
+
"learning_rate": 1.352e-07,
|
| 910 |
+
"loss": 0.2341,
|
| 911 |
+
"step": 2975
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 428.57,
|
| 915 |
+
"learning_rate": 1.3353333333333332e-07,
|
| 916 |
+
"loss": 0.2287,
|
| 917 |
+
"step": 3000
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 428.57,
|
| 921 |
+
"eval_loss": 1.0533033609390259,
|
| 922 |
+
"eval_runtime": 465.8233,
|
| 923 |
+
"eval_samples_per_second": 1.099,
|
| 924 |
+
"eval_steps_per_second": 0.069,
|
| 925 |
+
"eval_wer": 61.516343825665864,
|
| 926 |
+
"step": 3000
|
| 927 |
+
}
|
| 928 |
+
],
|
| 929 |
+
"max_steps": 5000,
|
| 930 |
+
"num_train_epochs": 715,
|
| 931 |
+
"total_flos": 5.330864948871168e+19,
|
| 932 |
+
"trial_name": null,
|
| 933 |
+
"trial_params": null
|
| 934 |
+
}
|
checkpoint-3000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2d15d72002cecf8323f6cd941cec892fce6b19d6e69aa1365a60139bd64d32ff
|
| 3 |
+
size 3579
|