jakobcassiman commited on
Commit
4a372d4
1 Parent(s): c6577a6

Add model files and model card

Browse files
README.md ADDED
@@ -0,0 +1,52 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - nl
4
+ tags:
5
+ - mbart
6
+ - bart
7
+ - summarization
8
+ datasets:
9
+ - ml6team/cnn_dailymail_nl
10
+ - ml6team/xsum_nl
11
+ ---
12
+
13
+ # mbart-large-cc25-cnn-dailymail-xsum-nl
14
+
15
+ ## Model description
16
+ Finetuned version of [mbart](https://huggingface.co/facebook/mbart-large-cc25). We also wrote a **blog post** about this model [here](https://blog.ml6.eu/)
17
+
18
+ ## Intended uses & limitations
19
+ It's meant for summarizing Dutch news articles.
20
+
21
+ #### How to use
22
+
23
+ ```python
24
+ import transformers
25
+
26
+ undisputed_best_model = transformers.MBartForConditionalGeneration.from_pretrained(
27
+ "ml6team/mbart-large-cc25-cnn-dailymail-xsum-nl"
28
+ )
29
+ tokenizer = transformers.MBartTokenizer.from_pretrained("facebook/mbart-large-cc25")
30
+ summarization_pipeline = transformers.pipeline(
31
+ task="summarization",
32
+ model=undisputed_best_model,
33
+ tokenizer=tokenizer,
34
+ )
35
+ summarization_pipeline.model.config.decoder_start_token_id = tokenizer.lang_code_to_id[
36
+ "nl_XX"
37
+ ]
38
+
39
+ article = "Kan je dit even samenvatten alsjeblief." # Dutch
40
+ summarization_pipeline(
41
+ article,
42
+ do_sample=True,
43
+ top_p=0.75,
44
+ top_k=50,
45
+ min_length=50,
46
+ early_stopping=True,
47
+ truncation=True,
48
+ )[0]["summary_text"]
49
+ ```
50
+
51
+ ## Training data
52
+ Finetuned [mbart](https://huggingface.co/facebook/mbart-large-cc25) with [this dataset](https://huggingface.co/datasets/ml6team/cnn_dailymail_nl) and [this dataset](https://huggingface.co/datasets/ml6team/xsum_nl)
config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/mbart-large-cc25",
3
+ "_num_labels": 3,
4
+ "activation_dropout": 0.0,
5
+ "activation_function": "gelu",
6
+ "add_bias_logits": false,
7
+ "add_final_layer_norm": true,
8
+ "architectures": [
9
+ "MBartForConditionalGeneration"
10
+ ],
11
+ "attention_dropout": 0.0,
12
+ "bos_token_id": 0,
13
+ "classif_dropout": 0.0,
14
+ "classifier_dropout": 0.0,
15
+ "d_model": 1024,
16
+ "decoder_attention_heads": 16,
17
+ "decoder_ffn_dim": 4096,
18
+ "decoder_layerdrop": 0.0,
19
+ "decoder_layers": 12,
20
+ "dropout": 0.1,
21
+ "encoder_attention_heads": 16,
22
+ "encoder_ffn_dim": 4096,
23
+ "encoder_layerdrop": 0.0,
24
+ "encoder_layers": 12,
25
+ "eos_token_id": 2,
26
+ "forced_eos_token_id": 2,
27
+ "id2label": {
28
+ "0": "LABEL_0",
29
+ "1": "LABEL_1",
30
+ "2": "LABEL_2"
31
+ },
32
+ "init_std": 0.02,
33
+ "is_encoder_decoder": true,
34
+ "label2id": {
35
+ "LABEL_0": 0,
36
+ "LABEL_1": 1,
37
+ "LABEL_2": 2
38
+ },
39
+ "max_length": 1024,
40
+ "max_position_embeddings": 1024,
41
+ "model_type": "mbart",
42
+ "normalize_before": true,
43
+ "normalize_embedding": true,
44
+ "num_beams": 5,
45
+ "num_hidden_layers": 12,
46
+ "output_past": true,
47
+ "pad_token_id": 1,
48
+ "scale_embedding": true,
49
+ "static_position_embeddings": false,
50
+ "task_specific_params": {
51
+ "translation_en_to_ro": {
52
+ "decoder_start_token_id": 250020
53
+ }
54
+ },
55
+ "torch_dtype": "float32",
56
+ "transformers_version": "4.15.0",
57
+ "use_cache": true,
58
+ "vocab_size": 250027
59
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ce5e2ef6d04dd5b1817d40d199270db3253e4864424c71a714823fb15b9ca101
3
+ size 2444611475
sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfc8146abe2a0488e9e2a0c56de7952f7c11ab059eca145a0a727afce0db2865
3
+ size 5069051
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": false}, "additional_special_tokens": ["ar_AR", "cs_CZ", "de_DE", "en_XX", "es_XX", "et_EE", "fi_FI", "fr_XX", "gu_IN", "hi_IN", "it_IT", "ja_XX", "kk_KZ", "ko_KR", "lt_LT", "lv_LV", "my_MM", "ne_NP", "nl_XX", "ro_RO", "ru_RU", "si_LK", "tr_TR", "vi_VN", "zh_CN"]}
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "src_lang": null, "tgt_lang": null, "additional_special_tokens": null, "model_max_length": 1024, "special_tokens_map_file": null, "name_or_path": "facebook/mbart-large-cc25", "tokenizer_class": "MBartTokenizer"}
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:23cf25a736fe0696f000b513b05a69b830f46b38121e3387e82516b2a6540907
3
+ size 3119