Hugging Face's logo

davanstrien
/
fineweb-c-quality-reranker-v1

Text Ranking
sentence-transformers
Safetensors
multilingual
modernbert
cross-encoder
reranker
Generated from Trainer
dataset_size:13249
loss:BinaryCrossEntropyLoss
Eval Results (legacy)
text-embeddings-inference
Model card Files
xet
 Community
davanstrien HF Staff commited on
Commit
29d6ce3
·
verified ·
1 Parent(s): b4d391d

Training in progress, step 249

Browse files
Files changed (3) hide show
  1. README.md +28 -18
  2. eval/CrossEncoderClassificationEvaluator_fineweb_c_eval_results.csv +1 -0
  3. model.safetensors +1 -1
README.md CHANGED
@@ -31,25 +31,25 @@ model-index:
31
  type: fineweb_c_eval
32
  metrics:
33
  - type: accuracy
34
- value: 0.8376358695652174
35
  name: Accuracy
36
  - type: accuracy_threshold
37
- value: 0.6088266372680664
38
  name: Accuracy Threshold
39
  - type: f1
40
- value: 0.410958904109589
41
  name: F1
42
  - type: f1_threshold
43
- value: 0.3380771279335022
44
  name: F1 Threshold
45
  - type: precision
46
- value: 0.3105590062111801
47
  name: Precision
48
  - type: recall
49
- value: 0.6072874493927125
50
  name: Recall
51
  - type: average_precision
52
- value: 0.3509417036235269
53
  name: Average Precision
54
  ---
55
 
@@ -110,7 +110,7 @@ pairs = [
110
  ]
111
  scores = model.predict(pairs)
112
  print(scores)
113
- # [0.4121 0.3208 0.0526 0.1956 0.3355]
114
 
115
  # Or rank different texts based on similarity to a single text
116
  ranks = model.rank(
@@ -161,13 +161,13 @@ You can finetune this model on your own dataset.
161
 
162
  | Metric | Value |
163
  |:----------------------|:-----------|
164
- | accuracy | 0.8376 |
165
- | accuracy_threshold | 0.6088 |
166
- | f1 | 0.411 |
167
- | f1_threshold | 0.3381 |
168
- | precision | 0.3106 |
169
- | recall | 0.6073 |
170
- | **average_precision** | **0.3509** |
171
 
172
  <!--
173
  ## Bias, Risks and Limitations
@@ -380,12 +380,22 @@ You can finetune this model on your own dataset.
380
  | 0.3687 | 153 | 1.0757 | - | - |
381
  | 0.3904 | 162 | 0.9787 | - | - |
382
  | 0.4 | 166 | - | 1.1451 | 0.3509 |
 
 
 
 
 
 
 
 
 
 
383
 
384
 
385
  ### Training Time
386
- - **Training**: 2.4 minutes
387
- - **Evaluation**: 1.2 minutes
388
- - **Total**: 3.6 minutes
389
 
390
  ### Framework Versions
391
  - Python: 3.12.12
 
31
  type: fineweb_c_eval
32
  metrics:
33
  - type: accuracy
34
+ value: 0.8648097826086957
35
  name: Accuracy
36
  - type: accuracy_threshold
37
+ value: 0.7341195344924927
38
  name: Accuracy Threshold
39
  - type: f1
40
+ value: 0.5341365461847389
41
  name: F1
42
  - type: f1_threshold
43
+ value: 0.6606103181838989
44
  name: F1 Threshold
45
  - type: precision
46
+ value: 0.5298804780876494
47
  name: Precision
48
  - type: recall
49
+ value: 0.5384615384615384
50
  name: Recall
51
  - type: average_precision
52
+ value: 0.5429365048696634
53
  name: Average Precision
54
  ---
55
 
 
110
  ]
111
  scores = model.predict(pairs)
112
  print(scores)
113
+ # [0.5025 0.5898 0.0346 0.1778 0.4574]
114
 
115
  # Or rank different texts based on similarity to a single text
116
  ranks = model.rank(
 
161
 
162
  | Metric | Value |
163
  |:----------------------|:-----------|
164
+ | accuracy | 0.8648 |
165
+ | accuracy_threshold | 0.7341 |
166
+ | f1 | 0.5341 |
167
+ | f1_threshold | 0.6606 |
168
+ | precision | 0.5299 |
169
+ | recall | 0.5385 |
170
+ | **average_precision** | **0.5429** |
171
 
172
  <!--
173
  ## Bias, Risks and Limitations
 
380
  | 0.3687 | 153 | 1.0757 | - | - |
381
  | 0.3904 | 162 | 0.9787 | - | - |
382
  | 0.4 | 166 | - | 1.1451 | 0.3509 |
383
+ | 0.4120 | 171 | 0.9056 | - | - |
384
+ | 0.4337 | 180 | 1.0730 | - | - |
385
+ | 0.4554 | 189 | 0.9763 | - | - |
386
+ | 0.4771 | 198 | 1.1545 | - | - |
387
+ | 0.4988 | 207 | 0.9918 | - | - |
388
+ | 0.5205 | 216 | 0.8669 | - | - |
389
+ | 0.5422 | 225 | 0.9827 | - | - |
390
+ | 0.5639 | 234 | 1.1260 | - | - |
391
+ | 0.5855 | 243 | 0.8437 | - | - |
392
+ | 0.6 | 249 | - | 0.9002 | 0.5429 |
393
 
394
 
395
  ### Training Time
396
+ - **Training**: 3.7 minutes
397
+ - **Evaluation**: 1.8 minutes
398
+ - **Total**: 5.5 minutes
399
 
400
  ### Framework Versions
401
  - Python: 3.12.12
eval/CrossEncoderClassificationEvaluator_fineweb_c_eval_results.csv CHANGED
@@ -1,3 +1,4 @@
1
  epoch,steps,Accuracy,Accuracy_Threshold,F1,F1_Threshold,Precision,Recall,Average_Precision
2
  0.2,83,0.8335597826086957,0.93439513,0.33268101761252444,0.69884336,0.21935483870967742,0.6882591093117408,0.2441194063501111
3
  0.4,166,0.8376358695652174,0.60882664,0.410958904109589,0.33807713,0.3105590062111801,0.6072874493927125,0.3509417036235269
 
 
1
  epoch,steps,Accuracy,Accuracy_Threshold,F1,F1_Threshold,Precision,Recall,Average_Precision
2
  0.2,83,0.8335597826086957,0.93439513,0.33268101761252444,0.69884336,0.21935483870967742,0.6882591093117408,0.2441194063501111
3
  0.4,166,0.8376358695652174,0.60882664,0.410958904109589,0.33807713,0.3105590062111801,0.6072874493927125,0.3509417036235269
4
+ 0.6,249,0.8648097826086957,0.73411953,0.5341365461847389,0.6606103,0.5298804780876494,0.5384615384615384,0.5429365048696634
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5198e0a994ddabb58b74e9c026fa1620ad8c31639d4cba5055d9750bca27116a
3
  size 1230138348
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:783e9d94356c8e777a0f325a247e6a44ae6e707ac3d235f036acf01fc4609e23
3
  size 1230138348