Token Classification
Transformers
ONNX
xlm-roberta
pii
privacy
redaction
accessibility-tree
ocr
computer-use
agentic
screen-capture
screenpipe
Instructions to use screenpipe/pii-redactor with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use screenpipe/pii-redactor with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("token-classification", model="screenpipe/pii-redactor", device_map="auto")# Load model directly from transformers import AutoTokenizer, AutoModelForTokenClassification tokenizer = AutoTokenizer.from_pretrained("screenpipe/pii-redactor") model = AutoModelForTokenClassification.from_pretrained("screenpipe/pii-redactor", device_map="auto") - Notebooks
- Google Colab
- Kaggle
v45 secret-loop champion (iter7 INT8): config.json
Browse files- v45_phase4_onnx/config.json +85 -0
v45_phase4_onnx/config.json
ADDED
|
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"XLMRobertaForTokenClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"dtype": "float32",
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"hidden_act": "gelu",
|
| 11 |
+
"hidden_dropout_prob": 0.1,
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"id2label": {
|
| 14 |
+
"0": "O",
|
| 15 |
+
"1": "B-private_person",
|
| 16 |
+
"2": "I-private_person",
|
| 17 |
+
"3": "B-private_email",
|
| 18 |
+
"4": "I-private_email",
|
| 19 |
+
"5": "B-private_phone",
|
| 20 |
+
"6": "I-private_phone",
|
| 21 |
+
"7": "B-private_address",
|
| 22 |
+
"8": "I-private_address",
|
| 23 |
+
"9": "B-private_url",
|
| 24 |
+
"10": "I-private_url",
|
| 25 |
+
"11": "B-private_id",
|
| 26 |
+
"12": "I-private_id",
|
| 27 |
+
"13": "B-private_date",
|
| 28 |
+
"14": "I-private_date",
|
| 29 |
+
"15": "B-private_company",
|
| 30 |
+
"16": "I-private_company",
|
| 31 |
+
"17": "B-private_handle",
|
| 32 |
+
"18": "I-private_handle",
|
| 33 |
+
"19": "B-private_channel",
|
| 34 |
+
"20": "I-private_channel",
|
| 35 |
+
"21": "B-private_repo",
|
| 36 |
+
"22": "I-private_repo",
|
| 37 |
+
"23": "B-secret",
|
| 38 |
+
"24": "I-secret",
|
| 39 |
+
"25": "B-private_sensitive",
|
| 40 |
+
"26": "I-private_sensitive"
|
| 41 |
+
},
|
| 42 |
+
"initializer_range": 0.02,
|
| 43 |
+
"intermediate_size": 3072,
|
| 44 |
+
"label2id": {
|
| 45 |
+
"B-private_address": 7,
|
| 46 |
+
"B-private_channel": 19,
|
| 47 |
+
"B-private_company": 15,
|
| 48 |
+
"B-private_date": 13,
|
| 49 |
+
"B-private_email": 3,
|
| 50 |
+
"B-private_handle": 17,
|
| 51 |
+
"B-private_id": 11,
|
| 52 |
+
"B-private_person": 1,
|
| 53 |
+
"B-private_phone": 5,
|
| 54 |
+
"B-private_repo": 21,
|
| 55 |
+
"B-private_sensitive": 25,
|
| 56 |
+
"B-private_url": 9,
|
| 57 |
+
"B-secret": 23,
|
| 58 |
+
"I-private_address": 8,
|
| 59 |
+
"I-private_channel": 20,
|
| 60 |
+
"I-private_company": 16,
|
| 61 |
+
"I-private_date": 14,
|
| 62 |
+
"I-private_email": 4,
|
| 63 |
+
"I-private_handle": 18,
|
| 64 |
+
"I-private_id": 12,
|
| 65 |
+
"I-private_person": 2,
|
| 66 |
+
"I-private_phone": 6,
|
| 67 |
+
"I-private_repo": 22,
|
| 68 |
+
"I-private_sensitive": 26,
|
| 69 |
+
"I-private_url": 10,
|
| 70 |
+
"I-secret": 24,
|
| 71 |
+
"O": 0
|
| 72 |
+
},
|
| 73 |
+
"layer_norm_eps": 1e-05,
|
| 74 |
+
"max_position_embeddings": 514,
|
| 75 |
+
"model_type": "xlm-roberta",
|
| 76 |
+
"num_attention_heads": 12,
|
| 77 |
+
"num_hidden_layers": 12,
|
| 78 |
+
"output_past": true,
|
| 79 |
+
"pad_token_id": 1,
|
| 80 |
+
"position_embedding_type": "absolute",
|
| 81 |
+
"transformers_version": "4.57.6",
|
| 82 |
+
"type_vocab_size": 1,
|
| 83 |
+
"use_cache": true,
|
| 84 |
+
"vocab_size": 250002
|
| 85 |
+
}
|