Upload folder using huggingface_hub
Browse files- .gitattributes +3 -0
- frequencymultirussian_256000/merges.txt +0 -0
- frequencymultirussian_256000/merges_data.csv +3 -0
- frequencymultirussian_256000/special_tokens_map.json +4 -0
- frequencymultirussian_256000/tokenizer.json +3 -0
- frequencymultirussian_256000/tokenizer_config.json +29 -0
- frequencymultirussian_256000/vocab.json +3 -0
.gitattributes
CHANGED
|
@@ -59,3 +59,6 @@ fw57M_Entropy_thresholdM_64000/stats.csv filter=lfs diff=lfs merge=lfs -text
|
|
| 59 |
fw57M_Entropy_thresholdB_128000/stats.csv filter=lfs diff=lfs merge=lfs -text
|
| 60 |
fw57M_Surprisal_thresholdBX_64000/stats.csv filter=lfs diff=lfs merge=lfs -text
|
| 61 |
frequencymultirussian_128000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
| 59 |
fw57M_Entropy_thresholdB_128000/stats.csv filter=lfs diff=lfs merge=lfs -text
|
| 60 |
fw57M_Surprisal_thresholdBX_64000/stats.csv filter=lfs diff=lfs merge=lfs -text
|
| 61 |
frequencymultirussian_128000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 62 |
+
frequencymultirussian_256000/merges_data.csv filter=lfs diff=lfs merge=lfs -text
|
| 63 |
+
frequencymultirussian_256000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 64 |
+
frequencymultirussian_256000/vocab.json filter=lfs diff=lfs merge=lfs -text
|
frequencymultirussian_256000/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
frequencymultirussian_256000/merges_data.csv
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3b80cb396272693deff38d5fbd61732ba3c0b76ac48d3aefa62ad8c91ab32a46
|
| 3 |
+
size 12388990
|
frequencymultirussian_256000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"eos_token": "<|endoftext|>",
|
| 3 |
+
"pad_token": "<|padding|>"
|
| 4 |
+
}
|
frequencymultirussian_256000/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f2f91644f77066108f5086cb0907d0d5606c84da99fc378ad372da50396f7716
|
| 3 |
+
size 25727288
|
frequencymultirussian_256000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"0": {
|
| 5 |
+
"content": "<|padding|>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": false,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
+
"1": {
|
| 13 |
+
"content": "<|endoftext|>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": false,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false,
|
| 18 |
+
"special": true
|
| 19 |
+
}
|
| 20 |
+
},
|
| 21 |
+
"bos_token": null,
|
| 22 |
+
"clean_up_tokenization_spaces": false,
|
| 23 |
+
"eos_token": "<|endoftext|>",
|
| 24 |
+
"extra_special_tokens": {},
|
| 25 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 26 |
+
"pad_token": "<|padding|>",
|
| 27 |
+
"tokenizer_class": "PreTrainedTokenizer",
|
| 28 |
+
"unk_token": null
|
| 29 |
+
}
|
frequencymultirussian_256000/vocab.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3daf0fc1f49e870843a6db0411f21ca67b6cd87497e2136faa9bbcee2188e0ca
|
| 3 |
+
size 18724601
|