Text Classification
Transformers
ONNX
PEFT
English
cross-encoder
reranker
thread-matching
conversational-ai
lora
Eval Results (legacy)
Instructions to use Algokruti/thread-reranker with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Algokruti/thread-reranker with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="Algokruti/thread-reranker")# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("Algokruti/thread-reranker", dtype="auto") - PEFT
How to use Algokruti/thread-reranker with PEFT:
Task type is invalid.
- Notebooks
- Google Colab
- Kaggle
Upload folder using huggingface_hub
Browse files- config.json +15 -0
- model.pt +3 -0
- thread_reranker.onnx +3 -0
- tokenizer.json +0 -0
- tokenizer_config.json +16 -0
- training_history.json +57 -0
config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"base_model": "microsoft/MiniLM-L6-H384-uncased",
|
| 3 |
+
"lora_r": 8,
|
| 4 |
+
"lora_alpha": 16,
|
| 5 |
+
"num_structured_features": 5,
|
| 6 |
+
"structured_feature_names": [
|
| 7 |
+
"entity_overlap",
|
| 8 |
+
"keyword_matches",
|
| 9 |
+
"flow_continuity",
|
| 10 |
+
"recency_score",
|
| 11 |
+
"hours_since_active"
|
| 12 |
+
],
|
| 13 |
+
"max_length": 256,
|
| 14 |
+
"best_val_f1": 0.5887755102040816
|
| 15 |
+
}
|
model.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:50c42fdfec34e400cd893e04487e531e0875c7d8f82ab1feb0a9838a48043888
|
| 3 |
+
size 91431767
|
thread_reranker.onnx
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:62fda2ec327e08a5bd589257f78eafd26cb8ec2ae91843ac283c95e7724d4aa3
|
| 3 |
+
size 1113693
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"cls_token": "[CLS]",
|
| 4 |
+
"do_basic_tokenize": true,
|
| 5 |
+
"do_lower_case": true,
|
| 6 |
+
"is_local": false,
|
| 7 |
+
"mask_token": "[MASK]",
|
| 8 |
+
"model_max_length": 512,
|
| 9 |
+
"never_split": null,
|
| 10 |
+
"pad_token": "[PAD]",
|
| 11 |
+
"sep_token": "[SEP]",
|
| 12 |
+
"strip_accents": null,
|
| 13 |
+
"tokenize_chinese_chars": true,
|
| 14 |
+
"tokenizer_class": "BertTokenizer",
|
| 15 |
+
"unk_token": "[UNK]"
|
| 16 |
+
}
|
training_history.json
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"phase": "curriculum (easy only)",
|
| 5 |
+
"train_loss": 0.6751341945491731,
|
| 6 |
+
"loss": 0.5207989287527302,
|
| 7 |
+
"accuracy": 0.7582113177681045,
|
| 8 |
+
"precision": 0.0,
|
| 9 |
+
"recall": 0.0,
|
| 10 |
+
"f1": 0.0,
|
| 11 |
+
"auc": 0.6821802104083453
|
| 12 |
+
},
|
| 13 |
+
{
|
| 14 |
+
"epoch": 2,
|
| 15 |
+
"phase": "curriculum (easy only)",
|
| 16 |
+
"train_loss": 0.5324331930605695,
|
| 17 |
+
"loss": 0.5129576917690567,
|
| 18 |
+
"accuracy": 0.7582113177681045,
|
| 19 |
+
"precision": 0.0,
|
| 20 |
+
"recall": 0.0,
|
| 21 |
+
"f1": 0.0,
|
| 22 |
+
"auc": 0.6938337336718273
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"epoch": 3,
|
| 26 |
+
"phase": "full dataset",
|
| 27 |
+
"train_loss": 0.47879998753719694,
|
| 28 |
+
"loss": 0.40398728093014485,
|
| 29 |
+
"accuracy": 0.8385437277404036,
|
| 30 |
+
"precision": 0.8996062992125984,
|
| 31 |
+
"recall": 0.3739770867430442,
|
| 32 |
+
"f1": 0.5283236994219653,
|
| 33 |
+
"auc": 0.8065109389788464
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 4,
|
| 37 |
+
"phase": "full dataset",
|
| 38 |
+
"train_loss": 0.3995989957659305,
|
| 39 |
+
"loss": 0.348926103379153,
|
| 40 |
+
"accuracy": 0.8397309062129007,
|
| 41 |
+
"precision": 0.8344155844155844,
|
| 42 |
+
"recall": 0.4206219312602291,
|
| 43 |
+
"f1": 0.55930359085963,
|
| 44 |
+
"auc": 0.8671301880281137
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 5,
|
| 48 |
+
"phase": "full dataset",
|
| 49 |
+
"train_loss": 0.375606742392801,
|
| 50 |
+
"loss": 0.3455968170980864,
|
| 51 |
+
"accuracy": 0.8405223585278987,
|
| 52 |
+
"precision": 0.7818428184281843,
|
| 53 |
+
"recall": 0.47217675941080195,
|
| 54 |
+
"f1": 0.5887755102040816,
|
| 55 |
+
"auc": 0.8685274576398593
|
| 56 |
+
}
|
| 57 |
+
]
|