Delete eole_model
Browse files- eole_model/config.json +0 -150
- eole_model/model.00.safetensors +0 -3
- eole_model/src.spm.model +0 -3
- eole_model/tgt.spm.model +0 -3
- eole_model/vocab.json +0 -0
eole_model/config.json
DELETED
|
@@ -1,150 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"save_data": "zh_en/data_spm",
|
| 3 |
-
"src_vocab": "zh-en-benchmark/src.eole.vocab",
|
| 4 |
-
"report_every": 100,
|
| 5 |
-
"share_vocab": false,
|
| 6 |
-
"tgt_vocab": "zh-en-benchmark/tgt.eole.vocab",
|
| 7 |
-
"vocab_size_multiple": 8,
|
| 8 |
-
"tensorboard_log_dir_dated": "tensorboard/Feb-12_13-34-26",
|
| 9 |
-
"src_vocab_size": 32000,
|
| 10 |
-
"tensorboard": true,
|
| 11 |
-
"n_sample": 0,
|
| 12 |
-
"tgt_vocab_size": 32000,
|
| 13 |
-
"valid_metrics": [
|
| 14 |
-
"BLEU"
|
| 15 |
-
],
|
| 16 |
-
"tensorboard_log_dir": "tensorboard",
|
| 17 |
-
"seed": 1234,
|
| 18 |
-
"overwrite": true,
|
| 19 |
-
"transforms": [
|
| 20 |
-
"sentencepiece",
|
| 21 |
-
"filtertoolong"
|
| 22 |
-
],
|
| 23 |
-
"training": {
|
| 24 |
-
"accum_count": [
|
| 25 |
-
16
|
| 26 |
-
],
|
| 27 |
-
"train_steps": 100000,
|
| 28 |
-
"gpu_ranks": [
|
| 29 |
-
0
|
| 30 |
-
],
|
| 31 |
-
"save_checkpoint_steps": 2000,
|
| 32 |
-
"decay_method": "noam",
|
| 33 |
-
"bucket_size": 128000,
|
| 34 |
-
"world_size": 1,
|
| 35 |
-
"accum_steps": [
|
| 36 |
-
0
|
| 37 |
-
],
|
| 38 |
-
"optim": "pagedadamw8bit",
|
| 39 |
-
"prefetch_factor": 100,
|
| 40 |
-
"compute_dtype": "torch.bfloat16",
|
| 41 |
-
"normalization": "tokens",
|
| 42 |
-
"label_smoothing": 0.1,
|
| 43 |
-
"batch_size_multiple": 8,
|
| 44 |
-
"dropout_steps": [
|
| 45 |
-
0
|
| 46 |
-
],
|
| 47 |
-
"average_decay": 0.0001,
|
| 48 |
-
"dropout": [
|
| 49 |
-
0.1
|
| 50 |
-
],
|
| 51 |
-
"batch_type": "tokens",
|
| 52 |
-
"valid_batch_size": 8192,
|
| 53 |
-
"param_init_method": "xavier_uniform",
|
| 54 |
-
"adam_beta2": 0.998,
|
| 55 |
-
"model_path": "zh-en-benchmark/model",
|
| 56 |
-
"keep_checkpoint": 4,
|
| 57 |
-
"num_workers": 0,
|
| 58 |
-
"batch_size": 8192,
|
| 59 |
-
"attention_dropout": [
|
| 60 |
-
0.1
|
| 61 |
-
],
|
| 62 |
-
"warmup_steps": 10000,
|
| 63 |
-
"valid_steps": 2000,
|
| 64 |
-
"max_grad_norm": 2.0,
|
| 65 |
-
"learning_rate": 2.0
|
| 66 |
-
},
|
| 67 |
-
"data": {
|
| 68 |
-
"corpus_1": {
|
| 69 |
-
"path_align": null,
|
| 70 |
-
"path_src": "zh-en/train.ready.zh",
|
| 71 |
-
"path_tgt": "zh-en/train.ready.en",
|
| 72 |
-
"transforms": [
|
| 73 |
-
"sentencepiece",
|
| 74 |
-
"filtertoolong"
|
| 75 |
-
]
|
| 76 |
-
},
|
| 77 |
-
"valid": {
|
| 78 |
-
"path_align": null,
|
| 79 |
-
"path_src": "zh-en-benchmark/dev.zho",
|
| 80 |
-
"path_tgt": "zh-en-benchmark/dev.eng",
|
| 81 |
-
"transforms": [
|
| 82 |
-
"sentencepiece",
|
| 83 |
-
"filtertoolong"
|
| 84 |
-
]
|
| 85 |
-
}
|
| 86 |
-
},
|
| 87 |
-
"transforms_configs": {
|
| 88 |
-
"sentencepiece": {
|
| 89 |
-
"tgt_subword_model": "${MODEL_PATH}/tgt.spm.model",
|
| 90 |
-
"src_subword_model": "${MODEL_PATH}/src.spm.model"
|
| 91 |
-
},
|
| 92 |
-
"filtertoolong": {
|
| 93 |
-
"tgt_seq_length": 256,
|
| 94 |
-
"src_seq_length": 256
|
| 95 |
-
}
|
| 96 |
-
},
|
| 97 |
-
"model": {
|
| 98 |
-
"share_decoder_embeddings": true,
|
| 99 |
-
"position_encoding_type": "SinusoidalInterleaved",
|
| 100 |
-
"add_qkvbias": false,
|
| 101 |
-
"architecture": "transformer",
|
| 102 |
-
"add_ffnbias": true,
|
| 103 |
-
"hidden_size": 1024,
|
| 104 |
-
"transformer_ff": 4096,
|
| 105 |
-
"mlp_activation_fn": "gelu",
|
| 106 |
-
"norm_eps": 1e-06,
|
| 107 |
-
"layer_norm": "standard",
|
| 108 |
-
"heads": 16,
|
| 109 |
-
"add_estimator": false,
|
| 110 |
-
"share_embeddings": false,
|
| 111 |
-
"decoder": {
|
| 112 |
-
"heads": 16,
|
| 113 |
-
"decoder_type": "transformer",
|
| 114 |
-
"position_encoding_type": "SinusoidalInterleaved",
|
| 115 |
-
"add_qkvbias": false,
|
| 116 |
-
"layers": 2,
|
| 117 |
-
"add_ffnbias": true,
|
| 118 |
-
"hidden_size": 1024,
|
| 119 |
-
"n_positions": null,
|
| 120 |
-
"transformer_ff": 4096,
|
| 121 |
-
"rope_config": null,
|
| 122 |
-
"mlp_activation_fn": "gelu",
|
| 123 |
-
"norm_eps": 1e-06,
|
| 124 |
-
"layer_norm": "standard",
|
| 125 |
-
"tgt_word_vec_size": 1024
|
| 126 |
-
},
|
| 127 |
-
"embeddings": {
|
| 128 |
-
"word_vec_size": 1024,
|
| 129 |
-
"position_encoding_type": "SinusoidalInterleaved",
|
| 130 |
-
"tgt_word_vec_size": 1024,
|
| 131 |
-
"src_word_vec_size": 1024
|
| 132 |
-
},
|
| 133 |
-
"encoder": {
|
| 134 |
-
"heads": 16,
|
| 135 |
-
"position_encoding_type": "SinusoidalInterleaved",
|
| 136 |
-
"add_qkvbias": false,
|
| 137 |
-
"layers": 8,
|
| 138 |
-
"add_ffnbias": true,
|
| 139 |
-
"hidden_size": 1024,
|
| 140 |
-
"n_positions": null,
|
| 141 |
-
"src_word_vec_size": 1024,
|
| 142 |
-
"transformer_ff": 4096,
|
| 143 |
-
"rope_config": null,
|
| 144 |
-
"mlp_activation_fn": "gelu",
|
| 145 |
-
"norm_eps": 1e-06,
|
| 146 |
-
"layer_norm": "standard",
|
| 147 |
-
"encoder_type": "transformer"
|
| 148 |
-
}
|
| 149 |
-
}
|
| 150 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
eole_model/model.00.safetensors
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:a6b78dad9fa4e560ce0abe0fc7c2b317ebe560fc23aa1897419253a1b334872d
|
| 3 |
-
size 820042008
|
|
|
|
|
|
|
|
|
|
|
|
eole_model/src.spm.model
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:23d03d562fc3f8fe57e497dac0ece4827c254675a80c103fc4bb4040638ceb67
|
| 3 |
-
size 733978
|
|
|
|
|
|
|
|
|
|
|
|
eole_model/tgt.spm.model
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:c373f1d78753313b0dbc337058bf8450e1fdd6fe662a49e0941affce44ec14c5
|
| 3 |
-
size 800955
|
|
|
|
|
|
|
|
|
|
|
|
eole_model/vocab.json
DELETED
|
The diff for this file is too large to render.
See raw diff
|
|
|