DiwasDiwas
commited on
Commit
•
ba1f31c
1
Parent(s):
ea10629
Update README.md
Browse files
README.md
CHANGED
@@ -6,38 +6,48 @@ model-index:
|
|
6 |
results: []
|
7 |
---
|
8 |
|
9 |
-
<!-- This model card has been generated automatically according to the information Keras had access to. You should
|
10 |
-
probably proofread and complete it, then remove this comment. -->
|
11 |
-
|
12 |
# t5-small-MedicoSummarizer
|
13 |
|
14 |
-
This model was
|
15 |
It achieves the following results on the evaluation set:
|
16 |
|
17 |
-
|
18 |
-
## Model description
|
19 |
-
|
20 |
-
More information needed
|
21 |
-
|
22 |
-
## Intended uses & limitations
|
23 |
-
|
24 |
-
More information needed
|
25 |
-
|
26 |
-
## Training and evaluation data
|
27 |
-
|
28 |
-
More information needed
|
29 |
-
|
30 |
## Training procedure
|
|
|
31 |
|
32 |
### Training hyperparameters
|
33 |
|
34 |
The following hyperparameters were used during training:
|
35 |
-
-
|
36 |
- training_precision: float32
|
|
|
|
|
|
|
37 |
|
38 |
### Training results
|
39 |
|
40 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
41 |
|
42 |
### Framework versions
|
43 |
|
|
|
6 |
results: []
|
7 |
---
|
8 |
|
|
|
|
|
|
|
9 |
# t5-small-MedicoSummarizer
|
10 |
|
11 |
+
This model was fine-tuned on t5-small on 25,000 PubMed articles for 10 epochs.
|
12 |
It achieves the following results on the evaluation set:
|
13 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
14 |
## Training procedure
|
15 |
+
The inference engine doesn't do justice to its operation as the inference engine API doesn't work good for trainer checkpoints as the context limit is low in default for T5 which you can change while using it on backend of your application ! So, you should rather load it on the pipeline and just try it !
|
16 |
|
17 |
### Training hyperparameters
|
18 |
|
19 |
The following hyperparameters were used during training:
|
20 |
+
- batch_size = 16
|
21 |
- training_precision: float32
|
22 |
+
- epochs = 10
|
23 |
+
- learning_rate = 2e-5
|
24 |
+
-
|
25 |
|
26 |
### Training results
|
27 |
|
28 |
+
|epoch|eval_loss |eval_rouge1|eval_rouge2|eval_rougeL|eval_rougeLsum|eval_gen_len|
|
29 |
+
|-----|------------------|-----------|-----------|-----------|--------------|------------|
|
30 |
+
|1.0 |3.0605552196502686|0.302 |0.0693 |0.1841 |0.1842 |116.916 |
|
31 |
+
|2.0 |3.0079214572906494|0.3192 |0.0749 |0.1943 |0.1944 |122.076 |
|
32 |
+
|3.0 |2.9787817001342773|0.3209 |0.0758 |0.1957 |0.1958 |122.95 |
|
33 |
+
|4.0 |2.95868182182312 |0.3226 |0.0772 |0.1978 |0.1978 |123.593 |
|
34 |
+
|5.0 |2.943807601928711 |0.3186 |0.0743 |0.1959 |0.1959 |123.822 |
|
35 |
+
|6.0 |2.9342598915100098|0.3194 |0.0755 |0.1962 |0.1961 |123.834 |
|
36 |
+
|7.0 |2.927173376083374 |0.3205 |0.0758 |0.1967 |0.1968 |123.967 |
|
37 |
+
|8.0 |2.9225199222564697|0.3211 |0.0763 |0.1974 |0.1975 |124.178 |
|
38 |
+
|9.0 |2.9196181297302246|0.32 |0.0762 |0.1964 |0.1964 |124.136 |
|
39 |
+
|10.0 |2.9186391830444336|0.3209 |0.0766 |0.1965 |0.1965 |124.115 |
|
40 |
+
|
41 |
+
## Test Metrics
|
42 |
+
{'test_loss': 2.8919856548309326,
|
43 |
+
'test_rouge1': 0.3207,
|
44 |
+
'test_rouge2': 0.0741,
|
45 |
+
'test_rougeL': 0.1955,
|
46 |
+
'test_rougeLsum': 0.1955,
|
47 |
+
'test_gen_len': 124.285,
|
48 |
+
'test_runtime': 335.298,
|
49 |
+
'test_samples_per_second': 5.965,
|
50 |
+
'test_steps_per_second': 0.373}
|
51 |
|
52 |
### Framework versions
|
53 |
|