tomaarsen HF staff commited on
Commit
ee44a17
1 Parent(s): dc31070

Update metrics using correct evaluation script

Browse files
Files changed (1) hide show
  1. README.md +18 -18
README.md CHANGED
@@ -51,13 +51,13 @@ model-index:
51
  split: test
52
  metrics:
53
  - type: f1
54
- value: 0.0
55
  name: F1
56
  - type: precision
57
- value: 0.0
58
  name: Precision
59
  - type: recall
60
- value: 0.0
61
  name: Recall
62
  ---
63
 
@@ -89,10 +89,10 @@ This is a [SpanMarker](https://github.com/tomaarsen/SpanMarkerNER) model trained
89
  ## Evaluation
90
 
91
  ### Metrics
92
- | Label | Precision | Recall | F1 |
93
- |:--------|:----------|:-------|:----|
94
- | **all** | 0.0 | 0.0 | 0.0 |
95
- | ORG | 0.0 | 0.0 | 0.0 |
96
 
97
  ## Uses
98
 
@@ -169,17 +169,17 @@ trainer.save_model("tomaarsen/span-marker-bert-base-orgs-finetuned")
169
  - num_epochs: 3
170
 
171
  ### Training Results
172
- | Epoch | Step | Validation Loss | Validation Precision | Validation Recall | Validation F1 | Validation Accuracy |
173
- |:------:|:-----:|:---------------:|:--------------------:|:-----------------:|:-------------:|:-------------------:|
174
- | 0.3273 | 3000 | 0.0052 | 0.0 | 0.0 | 0.0 | 0.9413 |
175
- | 0.6546 | 6000 | 0.0047 | 0.0 | 0.0 | 0.0 | 0.9334 |
176
- | 0.9819 | 9000 | 0.0045 | 0.0 | 0.0 | 0.0 | 0.9376 |
177
- | 1.3092 | 12000 | 0.0047 | 0.0 | 0.0 | 0.0 | 0.9377 |
178
- | 1.6365 | 15000 | 0.0045 | 0.0 | 0.0 | 0.0 | 0.9339 |
179
- | 1.9638 | 18000 | 0.0046 | 0.0 | 0.0 | 0.0 | 0.9373 |
180
- | 2.2911 | 21000 | 0.0054 | 0.0 | 0.0 | 0.0 | 0.9351 |
181
- | 2.6184 | 24000 | 0.0053 | 0.0 | 0.0 | 0.0 | 0.9373 |
182
- | 2.9457 | 27000 | 0.0052 | 0.0 | 0.0 | 0.0 | 0.9359 |
183
 
184
  ### Environmental Impact
185
  Carbon emissions were measured using [CodeCarbon](https://github.com/mlco2/codecarbon).
 
51
  split: test
52
  metrics:
53
  - type: f1
54
+ value: 0.8311343653918766
55
  name: F1
56
  - type: precision
57
+ value: 0.8334090564894745
58
  name: Precision
59
  - type: recall
60
+ value: 0.8288720574945131
61
  name: Recall
62
  ---
63
 
 
89
  ## Evaluation
90
 
91
  ### Metrics
92
+ | Label | Precision | Recall | F1 |
93
+ |:--------|:----------|:-------|:-------|
94
+ | **all** | 0.8334 | 0.8289 | 0.8311 |
95
+ | ORG | 0.8334 | 0.8289 | 0.8311 |
96
 
97
  ## Uses
98
 
 
169
  - num_epochs: 3
170
 
171
  ### Training Results
172
+ | Epoch | Step | Validation Loss |
173
+ |:------:|:-----:|:---------------:|
174
+ | 0.3273 | 3000 | 0.0052 |
175
+ | 0.6546 | 6000 | 0.0047 |
176
+ | 0.9819 | 9000 | 0.0045 |
177
+ | 1.3092 | 12000 | 0.0047 |
178
+ | 1.6365 | 15000 | 0.0045 |
179
+ | 1.9638 | 18000 | 0.0046 |
180
+ | 2.2911 | 21000 | 0.0054 |
181
+ | 2.6184 | 24000 | 0.0053 |
182
+ | 2.9457 | 27000 | 0.0052 |
183
 
184
  ### Environmental Impact
185
  Carbon emissions were measured using [CodeCarbon](https://github.com/mlco2/codecarbon).