Commit
·
d876c1f
0
Parent(s):
push checkpoint-2000
Browse files- config.json +42 -0
- optimizer.pt +3 -0
- preprocessor_config.json +0 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- trainer_state.json +514 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
"suppress_tokens": [],
|
| 38 |
+
"torch_dtype": "float32",
|
| 39 |
+
"transformers_version": "4.25.0.dev0",
|
| 40 |
+
"use_cache": true,
|
| 41 |
+
"vocab_size": 51865
|
| 42 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bec6fe312a70e086d2dfb4a0f01a08284a0c5e9f33c5a0be676ab32983a040f2
|
| 3 |
+
size 1934158597
|
preprocessor_config.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eebe88399aa1e4ae68097c0d38d760b3ec240fcf53f10a0b64602c003fc4ca39
|
| 3 |
+
size 967099139
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ec4dc4f69fc3d00d1a1b05b4574f41ae89c3e660cf271c6021d5391baade77af
|
| 3 |
+
size 14503
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f23db1abf0b77198c6e3a9212a16fca37cbfe2158135986751bf408c3cec5d63
|
| 3 |
+
size 559
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:98a6b682de726771327f20cbf122dc32a1379317e5c9c72237571916f4849df9
|
| 3 |
+
size 623
|
trainer_state.json
ADDED
|
@@ -0,0 +1,514 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 42.2978067169294,
|
| 3 |
+
"best_model_checkpoint": "./whisper-small-vn/checkpoint-2000",
|
| 4 |
+
"epoch": 40.0,
|
| 5 |
+
"global_step": 2000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.5,
|
| 12 |
+
"learning_rate": 4.4e-07,
|
| 13 |
+
"loss": 3.1082,
|
| 14 |
+
"step": 25
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.0,
|
| 18 |
+
"learning_rate": 9.400000000000001e-07,
|
| 19 |
+
"loss": 2.4714,
|
| 20 |
+
"step": 50
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 1.5,
|
| 24 |
+
"learning_rate": 1.44e-06,
|
| 25 |
+
"loss": 1.9624,
|
| 26 |
+
"step": 75
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 2.0,
|
| 30 |
+
"learning_rate": 1.94e-06,
|
| 31 |
+
"loss": 1.6754,
|
| 32 |
+
"step": 100
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 2.5,
|
| 36 |
+
"learning_rate": 2.4400000000000004e-06,
|
| 37 |
+
"loss": 1.3215,
|
| 38 |
+
"step": 125
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 3.0,
|
| 42 |
+
"learning_rate": 2.9400000000000002e-06,
|
| 43 |
+
"loss": 1.1727,
|
| 44 |
+
"step": 150
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 3.5,
|
| 48 |
+
"learning_rate": 3.44e-06,
|
| 49 |
+
"loss": 1.0125,
|
| 50 |
+
"step": 175
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 4.0,
|
| 54 |
+
"learning_rate": 3.94e-06,
|
| 55 |
+
"loss": 1.0082,
|
| 56 |
+
"step": 200
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 4.5,
|
| 60 |
+
"learning_rate": 4.440000000000001e-06,
|
| 61 |
+
"loss": 0.8672,
|
| 62 |
+
"step": 225
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 5.0,
|
| 66 |
+
"learning_rate": 4.94e-06,
|
| 67 |
+
"loss": 0.78,
|
| 68 |
+
"step": 250
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 5.5,
|
| 72 |
+
"learning_rate": 5.4400000000000004e-06,
|
| 73 |
+
"loss": 0.6437,
|
| 74 |
+
"step": 275
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 6.0,
|
| 78 |
+
"learning_rate": 5.94e-06,
|
| 79 |
+
"loss": 0.6069,
|
| 80 |
+
"step": 300
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 6.5,
|
| 84 |
+
"learning_rate": 6.440000000000001e-06,
|
| 85 |
+
"loss": 0.4016,
|
| 86 |
+
"step": 325
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 7.0,
|
| 90 |
+
"learning_rate": 6.9400000000000005e-06,
|
| 91 |
+
"loss": 0.4129,
|
| 92 |
+
"step": 350
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 7.5,
|
| 96 |
+
"learning_rate": 7.440000000000001e-06,
|
| 97 |
+
"loss": 0.2727,
|
| 98 |
+
"step": 375
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 8.0,
|
| 102 |
+
"learning_rate": 7.94e-06,
|
| 103 |
+
"loss": 0.2735,
|
| 104 |
+
"step": 400
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 8.5,
|
| 108 |
+
"learning_rate": 8.44e-06,
|
| 109 |
+
"loss": 0.1796,
|
| 110 |
+
"step": 425
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 9.0,
|
| 114 |
+
"learning_rate": 8.94e-06,
|
| 115 |
+
"loss": 0.1699,
|
| 116 |
+
"step": 450
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 9.5,
|
| 120 |
+
"learning_rate": 9.440000000000001e-06,
|
| 121 |
+
"loss": 0.1024,
|
| 122 |
+
"step": 475
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 10.0,
|
| 126 |
+
"learning_rate": 9.940000000000001e-06,
|
| 127 |
+
"loss": 0.1121,
|
| 128 |
+
"step": 500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 10.5,
|
| 132 |
+
"learning_rate": 9.937142857142858e-06,
|
| 133 |
+
"loss": 0.0645,
|
| 134 |
+
"step": 525
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 11.0,
|
| 138 |
+
"learning_rate": 9.865714285714285e-06,
|
| 139 |
+
"loss": 0.0727,
|
| 140 |
+
"step": 550
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 11.5,
|
| 144 |
+
"learning_rate": 9.794285714285714e-06,
|
| 145 |
+
"loss": 0.043,
|
| 146 |
+
"step": 575
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 12.0,
|
| 150 |
+
"learning_rate": 9.722857142857143e-06,
|
| 151 |
+
"loss": 0.0461,
|
| 152 |
+
"step": 600
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 12.5,
|
| 156 |
+
"learning_rate": 9.651428571428572e-06,
|
| 157 |
+
"loss": 0.0289,
|
| 158 |
+
"step": 625
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 13.0,
|
| 162 |
+
"learning_rate": 9.58e-06,
|
| 163 |
+
"loss": 0.0329,
|
| 164 |
+
"step": 650
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 13.5,
|
| 168 |
+
"learning_rate": 9.508571428571429e-06,
|
| 169 |
+
"loss": 0.0223,
|
| 170 |
+
"step": 675
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 14.0,
|
| 174 |
+
"learning_rate": 9.437142857142858e-06,
|
| 175 |
+
"loss": 0.0214,
|
| 176 |
+
"step": 700
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 14.5,
|
| 180 |
+
"learning_rate": 9.365714285714287e-06,
|
| 181 |
+
"loss": 0.0153,
|
| 182 |
+
"step": 725
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 15.0,
|
| 186 |
+
"learning_rate": 9.294285714285714e-06,
|
| 187 |
+
"loss": 0.018,
|
| 188 |
+
"step": 750
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 15.5,
|
| 192 |
+
"learning_rate": 9.222857142857143e-06,
|
| 193 |
+
"loss": 0.0143,
|
| 194 |
+
"step": 775
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 16.0,
|
| 198 |
+
"learning_rate": 9.151428571428572e-06,
|
| 199 |
+
"loss": 0.0144,
|
| 200 |
+
"step": 800
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 16.5,
|
| 204 |
+
"learning_rate": 9.080000000000001e-06,
|
| 205 |
+
"loss": 0.0124,
|
| 206 |
+
"step": 825
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 17.0,
|
| 210 |
+
"learning_rate": 9.00857142857143e-06,
|
| 211 |
+
"loss": 0.0122,
|
| 212 |
+
"step": 850
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 17.5,
|
| 216 |
+
"learning_rate": 8.937142857142857e-06,
|
| 217 |
+
"loss": 0.0095,
|
| 218 |
+
"step": 875
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 18.0,
|
| 222 |
+
"learning_rate": 8.865714285714287e-06,
|
| 223 |
+
"loss": 0.0085,
|
| 224 |
+
"step": 900
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 18.5,
|
| 228 |
+
"learning_rate": 8.794285714285716e-06,
|
| 229 |
+
"loss": 0.0072,
|
| 230 |
+
"step": 925
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 19.0,
|
| 234 |
+
"learning_rate": 8.722857142857145e-06,
|
| 235 |
+
"loss": 0.0075,
|
| 236 |
+
"step": 950
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 19.5,
|
| 240 |
+
"learning_rate": 8.651428571428572e-06,
|
| 241 |
+
"loss": 0.0066,
|
| 242 |
+
"step": 975
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 20.0,
|
| 246 |
+
"learning_rate": 8.580000000000001e-06,
|
| 247 |
+
"loss": 0.0069,
|
| 248 |
+
"step": 1000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 20.0,
|
| 252 |
+
"eval_loss": 1.1462156772613525,
|
| 253 |
+
"eval_runtime": 178.292,
|
| 254 |
+
"eval_samples_per_second": 1.481,
|
| 255 |
+
"eval_steps_per_second": 0.185,
|
| 256 |
+
"eval_wer": 43.137422892392046,
|
| 257 |
+
"step": 1000
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 20.5,
|
| 261 |
+
"learning_rate": 8.50857142857143e-06,
|
| 262 |
+
"loss": 0.0061,
|
| 263 |
+
"step": 1025
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 21.0,
|
| 267 |
+
"learning_rate": 8.437142857142859e-06,
|
| 268 |
+
"loss": 0.0083,
|
| 269 |
+
"step": 1050
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 21.5,
|
| 273 |
+
"learning_rate": 8.365714285714286e-06,
|
| 274 |
+
"loss": 0.0054,
|
| 275 |
+
"step": 1075
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 22.0,
|
| 279 |
+
"learning_rate": 8.294285714285715e-06,
|
| 280 |
+
"loss": 0.0063,
|
| 281 |
+
"step": 1100
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 22.5,
|
| 285 |
+
"learning_rate": 8.222857142857144e-06,
|
| 286 |
+
"loss": 0.0055,
|
| 287 |
+
"step": 1125
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 23.0,
|
| 291 |
+
"learning_rate": 8.151428571428572e-06,
|
| 292 |
+
"loss": 0.0061,
|
| 293 |
+
"step": 1150
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 23.5,
|
| 297 |
+
"learning_rate": 8.08e-06,
|
| 298 |
+
"loss": 0.0037,
|
| 299 |
+
"step": 1175
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 24.0,
|
| 303 |
+
"learning_rate": 8.00857142857143e-06,
|
| 304 |
+
"loss": 0.0039,
|
| 305 |
+
"step": 1200
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 24.5,
|
| 309 |
+
"learning_rate": 7.937142857142857e-06,
|
| 310 |
+
"loss": 0.0032,
|
| 311 |
+
"step": 1225
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 25.0,
|
| 315 |
+
"learning_rate": 7.865714285714286e-06,
|
| 316 |
+
"loss": 0.0032,
|
| 317 |
+
"step": 1250
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 25.5,
|
| 321 |
+
"learning_rate": 7.794285714285715e-06,
|
| 322 |
+
"loss": 0.0031,
|
| 323 |
+
"step": 1275
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 26.0,
|
| 327 |
+
"learning_rate": 7.722857142857142e-06,
|
| 328 |
+
"loss": 0.0027,
|
| 329 |
+
"step": 1300
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 26.5,
|
| 333 |
+
"learning_rate": 7.651428571428571e-06,
|
| 334 |
+
"loss": 0.0027,
|
| 335 |
+
"step": 1325
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 27.0,
|
| 339 |
+
"learning_rate": 7.58e-06,
|
| 340 |
+
"loss": 0.0027,
|
| 341 |
+
"step": 1350
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 27.5,
|
| 345 |
+
"learning_rate": 7.508571428571429e-06,
|
| 346 |
+
"loss": 0.0018,
|
| 347 |
+
"step": 1375
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 28.0,
|
| 351 |
+
"learning_rate": 7.4371428571428575e-06,
|
| 352 |
+
"loss": 0.0023,
|
| 353 |
+
"step": 1400
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 28.5,
|
| 357 |
+
"learning_rate": 7.365714285714286e-06,
|
| 358 |
+
"loss": 0.0024,
|
| 359 |
+
"step": 1425
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 29.0,
|
| 363 |
+
"learning_rate": 7.294285714285715e-06,
|
| 364 |
+
"loss": 0.0034,
|
| 365 |
+
"step": 1450
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 29.5,
|
| 369 |
+
"learning_rate": 7.222857142857144e-06,
|
| 370 |
+
"loss": 0.0025,
|
| 371 |
+
"step": 1475
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 30.0,
|
| 375 |
+
"learning_rate": 7.151428571428573e-06,
|
| 376 |
+
"loss": 0.0026,
|
| 377 |
+
"step": 1500
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 30.5,
|
| 381 |
+
"learning_rate": 7.08e-06,
|
| 382 |
+
"loss": 0.002,
|
| 383 |
+
"step": 1525
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 31.0,
|
| 387 |
+
"learning_rate": 7.008571428571429e-06,
|
| 388 |
+
"loss": 0.002,
|
| 389 |
+
"step": 1550
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 31.5,
|
| 393 |
+
"learning_rate": 6.937142857142858e-06,
|
| 394 |
+
"loss": 0.0012,
|
| 395 |
+
"step": 1575
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 32.0,
|
| 399 |
+
"learning_rate": 6.865714285714287e-06,
|
| 400 |
+
"loss": 0.0018,
|
| 401 |
+
"step": 1600
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 32.5,
|
| 405 |
+
"learning_rate": 6.794285714285714e-06,
|
| 406 |
+
"loss": 0.0013,
|
| 407 |
+
"step": 1625
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 33.0,
|
| 411 |
+
"learning_rate": 6.722857142857143e-06,
|
| 412 |
+
"loss": 0.0014,
|
| 413 |
+
"step": 1650
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 33.5,
|
| 417 |
+
"learning_rate": 6.651428571428572e-06,
|
| 418 |
+
"loss": 0.0011,
|
| 419 |
+
"step": 1675
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 34.0,
|
| 423 |
+
"learning_rate": 6.5800000000000005e-06,
|
| 424 |
+
"loss": 0.0032,
|
| 425 |
+
"step": 1700
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 34.5,
|
| 429 |
+
"learning_rate": 6.5085714285714295e-06,
|
| 430 |
+
"loss": 0.0019,
|
| 431 |
+
"step": 1725
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 35.0,
|
| 435 |
+
"learning_rate": 6.437142857142858e-06,
|
| 436 |
+
"loss": 0.0026,
|
| 437 |
+
"step": 1750
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 35.5,
|
| 441 |
+
"learning_rate": 6.365714285714286e-06,
|
| 442 |
+
"loss": 0.0017,
|
| 443 |
+
"step": 1775
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 36.0,
|
| 447 |
+
"learning_rate": 6.294285714285715e-06,
|
| 448 |
+
"loss": 0.0023,
|
| 449 |
+
"step": 1800
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 36.5,
|
| 453 |
+
"learning_rate": 6.222857142857144e-06,
|
| 454 |
+
"loss": 0.0018,
|
| 455 |
+
"step": 1825
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 37.0,
|
| 459 |
+
"learning_rate": 6.151428571428571e-06,
|
| 460 |
+
"loss": 0.0023,
|
| 461 |
+
"step": 1850
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 37.5,
|
| 465 |
+
"learning_rate": 6.08e-06,
|
| 466 |
+
"loss": 0.0027,
|
| 467 |
+
"step": 1875
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 38.0,
|
| 471 |
+
"learning_rate": 6.008571428571429e-06,
|
| 472 |
+
"loss": 0.0018,
|
| 473 |
+
"step": 1900
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 38.5,
|
| 477 |
+
"learning_rate": 5.937142857142858e-06,
|
| 478 |
+
"loss": 0.0011,
|
| 479 |
+
"step": 1925
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 39.0,
|
| 483 |
+
"learning_rate": 5.865714285714286e-06,
|
| 484 |
+
"loss": 0.0014,
|
| 485 |
+
"step": 1950
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 39.5,
|
| 489 |
+
"learning_rate": 5.794285714285715e-06,
|
| 490 |
+
"loss": 0.0017,
|
| 491 |
+
"step": 1975
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 40.0,
|
| 495 |
+
"learning_rate": 5.722857142857144e-06,
|
| 496 |
+
"loss": 0.0011,
|
| 497 |
+
"step": 2000
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 40.0,
|
| 501 |
+
"eval_loss": 1.257658839225769,
|
| 502 |
+
"eval_runtime": 176.0293,
|
| 503 |
+
"eval_samples_per_second": 1.5,
|
| 504 |
+
"eval_steps_per_second": 0.187,
|
| 505 |
+
"eval_wer": 42.2978067169294,
|
| 506 |
+
"step": 2000
|
| 507 |
+
}
|
| 508 |
+
],
|
| 509 |
+
"max_steps": 4000,
|
| 510 |
+
"num_train_epochs": 80,
|
| 511 |
+
"total_flos": 9.1539288981504e+18,
|
| 512 |
+
"trial_name": null,
|
| 513 |
+
"trial_params": null
|
| 514 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3c3897bd6ee579182083ac37f761a56be94fdfe65552d04f73afebd0d93cff1f
|
| 3 |
+
size 3503
|