zrostek/nanoT5
1
1
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

update tokenizer

Browse Source
This commit is contained in:
zrostek 2023-05-31 08:09:09 +00:00
parent c50cf6b0f2
commit 11926b825d
6 changed files with 49972 additions and 49943 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
tokenizer/config.json Normal file
View File

@ -0,0 +1,27 @@
{
"_name_or_path": "/home/patrick/hugging_face/t5/mt5-base",
"architectures": [
"T5ForConditionalGeneration"
],
"d_ff": 2048,
"d_kv": 64,
"d_model": 768,
"decoder_start_token_id": 0,
"dropout_rate": 0.1,
"eos_token_id": 1,
"feed_forward_proj": "gated-gelu",
"initializer_factor": 1.0,
"is_encoder_decoder": true,
"layer_norm_epsilon": 1e-06,
"model_type": "t5",
"num_decoder_layers": 12,
"num_heads": 12,
"num_layers": 12,
"output_past": true,
"pad_token_id": 0,
"relative_attention_num_buckets": 32,
"tie_word_embeddings": false,
"tokenizer_class": "T5Tokenizer",
"use_cache": true,
"vocab_size": 50048
}

1
tokenizer/special_tokens_map.json Normal file
View File

@ -0,0 +1 @@
{"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>"}

BIN
tokenizer/spiece.model
View File

Binary file not shown.

99878
tokenizer/spiece.vocab
View File

File diff suppressed because it is too large Load Diff

1
tokenizer/tokenizer_config.json Normal file
View File

@ -0,0 +1 @@
{"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>", "extra_ids": 0, "additional_special_tokens": null, "tokenizer_file": null}

8
tokenizer/train_tokenizer.sh
View File

@ -1,5 +1,5 @@
wget https://huggingface.co/allegro/plt5-base/raw/main/tokenizer_config.json
wget https://huggingface.co/allegro/plt5-base/raw/main/special_tokens_map.json
wget https://huggingface.co/allegro/plt5-base/raw/main/config.json
# wget https://huggingface.co/allegro/plt5-base/raw/main/tokenizer_config.json
# wget https://huggingface.co/allegro/plt5-base/raw/main/special_tokens_map.json
# wget https://huggingface.co/allegro/plt5-base/raw/main/config.json
spm_train --input='/mnt/gpu_data1/zrostek/diachronia-year-prediction/regular_roberta_from_scratch/train_tok_2.csv' --model_prefix='spiece' --vocab_size=50000 --character_coverage=1.0 --model_type=unigram --train_extremely_large_corpus=true
spm_train --input='/mnt/gpu_data1/zrostek/diachronia-year-prediction/regular_roberta_from_scratch/train_tok_3.csv' --model_prefix='spiece' --vocab_size=50000 --model_type=unigram --train_extremely_large_corpus=true