eachanjohnson commited on
Commit
ba0149b
·
verified ·
1 Parent(s): d7983f3

Upload folder using huggingface_hub

Browse files
README.md ADDED
@@ -0,0 +1,140 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: mit
3
+ pipeline_tag: tabular-regression
4
+ tags:
5
+ - chemistry
6
+ - microbiology
7
+ - antibiotics
8
+ library_name: duvida
9
+ datasets:
10
+ - scbirlab/thomas-2018-spark-wt
11
+ ---
12
+
13
+ # Predictor of _Brucella abortus_ MICs
14
+
15
+ _Updated:_ Fri 28 Mar 18:11:12 GMT 2025
16
+
17
+ Trained on the _Brucella abortus_, WT accumulator phenotype subset of the [human-curated SPARK dataset](https://doi.org/10.1021/acsinfecdis.8b00193) ( rows in total for _Brucella abortus_).
18
+
19
+ ## Model details
20
+
21
+ This model was trained using [our Duvida framework](https://github.com/scbirlab/duvida),
22
+ as a result of hyperparameter searches and selecting the model that performs best on unseen test data
23
+ (from a scaffold split).
24
+
25
+ Duvida also saves the training data in this checkpoint to allows the calculation of uncertainty metrics
26
+ based on that training data.
27
+
28
+ This model is the best regression model from a hyperparameter search, determined
29
+ by Spearman's $\rho$ on a held-out test set not used in training or early stopping.
30
+
31
+ ### Model architecture
32
+
33
+ - **Regression**
34
+
35
+ ```json
36
+
37
+ {
38
+ "dropout": 0.2,
39
+ "ensemble_size": 10,
40
+ "extra_featurizers": null,
41
+ "learning_rate": 0.0001,
42
+ "model_class": "ChempropModelBox",
43
+ "n_hidden": 5,
44
+ "n_units": 16,
45
+ "use_2d": true,
46
+ "use_fp": true
47
+ }
48
+ ```
49
+
50
+ ### Model usage
51
+
52
+ You can use this model with:
53
+
54
+ ```python
55
+ from duvida.autoclasses import AutoModelBox
56
+ modelbox = AutoModelBox.from_pretrained("hf://scbirlab/spark-dv-2503-babo")
57
+ modelbox.predict(filename=..., inputs=[...], columns=[...]) # make predictions on your own data
58
+ ```
59
+
60
+ ## Training details
61
+
62
+ - **Dataset:** [SPARK, WT accumulator, _Brucella abortus_ subset](https://huggingface.co/datasets/scbirlab/thomas-2018-spark-wt)
63
+ - **Input column:** smiles
64
+ - **Output column:** pmic
65
+ - **Split type:** Murcko scaffold
66
+ - **Split proportions:**
67
+ - 70% training (6963 rows)
68
+ - 15% validation (for early stopping) (1491 rows)
69
+ - 15% test (for selecting hyperparameters) (1492 rows)
70
+
71
+ Here is the training log:
72
+
73
+ <img src="training-log.png" width=450>
74
+
75
+ And these are the evaluation scores.
76
+
77
+ Train (6963 rows):
78
+
79
+ ```json
80
+
81
+ {
82
+ "Pearson r": 0.5503574788961993,
83
+ "RMSE": 0.053899530321359634,
84
+ "Spearman rho": 0.9346845985202912
85
+ }
86
+ ```
87
+
88
+ Validation (1491 rows):
89
+
90
+ ```json
91
+
92
+ {
93
+ "Pearson r": 0.4224614855497885,
94
+ "RMSE": 0.06130664795637131,
95
+ "Spearman rho": 0.88412899537977
96
+ }
97
+ ```
98
+
99
+
100
+ Test (1492 rows):
101
+
102
+ ```json
103
+
104
+ {
105
+ "Pearson r": 0.4829411208985374,
106
+ "RMSE": 0.06526102125644684,
107
+ "Spearman rho": 0.9303547327209174
108
+ }
109
+ ```
110
+
111
+ ## Training data details
112
+
113
+ The training data were collated by the authors of:
114
+
115
+ > Joe Thomas, Marc Navre, Aileen Rubio, and Allan Coukell
116
+ > Shared Platform for Antibiotic Research and Knowledge: A Collaborative Tool to SPARK Antibiotic Discovery
117
+ > ACS Infectious Diseases 2018 4 (11), 1536-1539
118
+ > DOI: 10.1021/acsinfecdis.8b00193
119
+
120
+ We cleaned the original SPARK dataset to subset the most relevant columns, remove empty values,
121
+ give succint column titles, and split by species.
122
+
123
+ This particular dataset retains only measurements on bacteria with wild-type accumulation phenotypes.
124
+
125
+ ### Dataset Sources
126
+
127
+ - **Repository:** https://www.collaborativedrug.com/spark-data-downloads
128
+ - **Paper:** https://doi.org/10.1021/acsinfecdis.8b00193
129
+
130
+ ### Data Collection and Processing
131
+
132
+ Data were processed using [schemist](https://github.com/scbirlab/schemist), a tool for processing chemical datasets.
133
+
134
+ The SMILES strings have been canonicalized, and split into training (70%), validation (15%), and test (15%) sets
135
+ by Murcko scaffold for each species with more than 1000 entries. Additional features like molecular weight and
136
+ topological polar surface area have also been calculated.
137
+
138
+ ### Who are the source data producers?
139
+
140
+ Joe Thomas, Marc Navre, Aileen Rubio, and Allan Coukell
data-config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_default_cache": "cache/duvida/data",
3
+ "_in_key": "inputs",
4
+ "_input_cols": [
5
+ "smiles"
6
+ ],
7
+ "_label_cols": [
8
+ "pmic"
9
+ ],
10
+ "_out_key": "labels",
11
+ "input_shape": [
12
+ 2248
13
+ ],
14
+ "output_shape": [
15
+ 1
16
+ ]
17
+ }
data-load-args.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cache": "/nemo/lab/johnsone/home/users/johnsoe/projects/abx-discovery-strategy/models/spark/Brucella-abortus/71/cache",
3
+ "features": [
4
+ "smiles"
5
+ ],
6
+ "filename": "/nemo/lab/johnsone/home/users/johnsoe/data/datasets/thomas-2018-spark-wt/Brucella-abortus/scaffold-split-train.csv.gz",
7
+ "labels": [
8
+ "pmic"
9
+ ]
10
+ }
eval-metrics_test.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "Pearson r": 0.4829411208985374,
3
+ "RMSE": 0.06526102125644684,
4
+ "Spearman rho": 0.9303547327209174
5
+ }
eval-metrics_train.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "Pearson r": 0.5503574788961993,
3
+ "RMSE": 0.053899530321359634,
4
+ "Spearman rho": 0.9346845985202912
5
+ }
eval-metrics_validation.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "Pearson r": 0.4224614855497885,
3
+ "RMSE": 0.06130664795637131,
4
+ "Spearman rho": 0.88412899537977
5
+ }
input-data.hf/data-00000-of-00001.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2b2162606a61d1e5e86ade4fc3596b9999f29262a4e77274d0df178b4a116944
3
+ size 709264
input-data.hf/dataset_info.json ADDED
@@ -0,0 +1,52 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "builder_name": "csv",
3
+ "citation": "",
4
+ "config_name": "default",
5
+ "dataset_name": "csv",
6
+ "dataset_size": 2728879,
7
+ "description": "",
8
+ "download_checksums": {
9
+ "/nemo/lab/johnsone/home/users/johnsoe/data/datasets/thomas-2018-spark-wt/Brucella-abortus/scaffold-split-train.csv.gz": {
10
+ "num_bytes": 557523,
11
+ "checksum": null
12
+ }
13
+ },
14
+ "download_size": 557523,
15
+ "features": {
16
+ "smiles": {
17
+ "dtype": "string",
18
+ "_type": "Value"
19
+ },
20
+ "inputs": {
21
+ "feature": {
22
+ "dtype": "string",
23
+ "_type": "Value"
24
+ },
25
+ "_type": "Sequence"
26
+ },
27
+ "labels": {
28
+ "feature": {
29
+ "dtype": "float64",
30
+ "_type": "Value"
31
+ },
32
+ "_type": "Sequence"
33
+ }
34
+ },
35
+ "homepage": "",
36
+ "license": "",
37
+ "size_in_bytes": 3286402,
38
+ "splits": {
39
+ "train": {
40
+ "name": "train",
41
+ "num_bytes": 2728879,
42
+ "num_examples": 6963,
43
+ "dataset_name": "csv"
44
+ }
45
+ },
46
+ "version": {
47
+ "version_str": "0.0.0",
48
+ "major": 0,
49
+ "minor": 0,
50
+ "patch": 0
51
+ }
52
+ }
input-data.hf/state.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_data_files": [
3
+ {
4
+ "filename": "data-00000-of-00001.arrow"
5
+ }
6
+ ],
7
+ "_fingerprint": "e067f72bd4fd2f97",
8
+ "_format_columns": null,
9
+ "_format_kwargs": {},
10
+ "_format_type": null,
11
+ "_output_all_columns": false,
12
+ "_split": "train"
13
+ }
logs-csv/lightning_logs/version_0/hparams.yaml ADDED
@@ -0,0 +1 @@
 
 
1
+ {}
logs-csv/lightning_logs/version_0/metrics.csv ADDED
@@ -0,0 +1,83 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ epoch,loss,step,val_loss
2
+ 0,,435,0.4943719804286957
3
+ 0,5.289660930633545,435,
4
+ 1,,871,0.19863000512123108
5
+ 1,1.7327244281768799,871,
6
+ 2,,1307,0.07188103348016739
7
+ 2,1.091246247291565,1307,
8
+ 3,,1743,0.04106524959206581
9
+ 3,0.7714532017707825,1743,
10
+ 4,,2179,0.033127326518297195
11
+ 4,0.6212546229362488,2179,
12
+ 5,,2615,0.029247038066387177
13
+ 5,0.5265188813209534,2615,
14
+ 6,,3051,0.02840065397322178
15
+ 6,0.4711878001689911,3051,
16
+ 7,,3487,0.02671700343489647
17
+ 7,0.43018320202827454,3487,
18
+ 8,,3923,0.02705136500298977
19
+ 8,0.3966166377067566,3923,
20
+ 9,,4359,0.026753852143883705
21
+ 9,0.36803075671195984,4359,
22
+ 10,,4795,0.026545658707618713
23
+ 10,0.34880682826042175,4795,
24
+ 11,,5231,0.02747528813779354
25
+ 11,0.33344268798828125,5231,
26
+ 12,,5667,0.026356253772974014
27
+ 12,0.31779828667640686,5667,
28
+ 13,,6103,0.026483017951250076
29
+ 13,0.30868658423423767,6103,
30
+ 14,,6539,0.026650499552488327
31
+ 14,0.2972111701965332,6539,
32
+ 15,,6975,0.02774786204099655
33
+ 15,0.289717435836792,6975,
34
+ 16,,7411,0.026358287781476974
35
+ 16,0.2818793058395386,7411,
36
+ 17,,7847,0.026366928592324257
37
+ 17,0.27312374114990234,7847,
38
+ 18,,8283,0.02649846486747265
39
+ 18,0.2667596936225891,8283,
40
+ 19,,8719,0.02613169141113758
41
+ 19,0.261917382478714,8719,
42
+ 20,,9155,0.025241071358323097
43
+ 20,0.256716251373291,9155,
44
+ 21,,9591,0.02763586863875389
45
+ 21,0.2512047290802002,9591,
46
+ 22,,10027,0.0261844489723444
47
+ 22,0.24682091176509857,10027,
48
+ 23,,10463,0.026615232229232788
49
+ 23,0.2391820251941681,10463,
50
+ 24,,10899,0.02532847970724106
51
+ 24,0.23734702169895172,10899,
52
+ 25,,11335,0.027400750666856766
53
+ 25,0.23324531316757202,11335,
54
+ 26,,11771,0.029379431158304214
55
+ 26,0.2280033677816391,11771,
56
+ 27,,12207,0.025781521573662758
57
+ 27,0.22546932101249695,12207,
58
+ 28,,12643,0.02702564373612404
59
+ 28,0.2223566323518753,12643,
60
+ 29,,13079,0.027757002040743828
61
+ 29,0.21767555177211761,13079,
62
+ 30,,13515,0.02584170363843441
63
+ 30,0.21601980924606323,13515,
64
+ 31,,13951,0.02662491984665394
65
+ 31,0.2132291942834854,13951,
66
+ 32,,14387,0.025853078812360764
67
+ 32,0.20955024659633636,14387,
68
+ 33,,14823,0.02659512124955654
69
+ 33,0.20862342417240143,14823,
70
+ 34,,15259,0.02577393874526024
71
+ 34,0.20273007452487946,15259,
72
+ 35,,15695,0.025896569713950157
73
+ 35,0.2000226080417633,15695,
74
+ 36,,16131,0.025927668437361717
75
+ 36,0.19752368330955505,16131,
76
+ 37,,16567,0.025407912209630013
77
+ 37,0.19396986067295074,16567,
78
+ 38,,17003,0.025420965626835823
79
+ 38,0.19176311790943146,17003,
80
+ 39,,17439,0.026712685823440552
81
+ 39,0.18767289817333221,17439,
82
+ 40,,17875,0.02653094381093979
83
+ 40,0.18428951501846313,17875,
logs/lightning_logs/version_0/events.out.tfevents.1743096520.cn042.1346636.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6f81996990bc04dc3d008961fe95941dbd7a212e6da76a42e689364428f60a40
3
+ size 14156
logs/lightning_logs/version_0/events.out.tfevents.1743096673.cn090.4018455.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a6753e9d0e4d8bfeba7e4e2ef97f30dd1edaa280ed52b40081c9b701c1f1c85
3
+ size 7402
logs/lightning_logs/version_0/hparams.yaml ADDED
@@ -0,0 +1 @@
 
 
1
+ {}
metrics.csv ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ split,split_filename,config_i,model_class,n_parameters,filename,features,labels,cache,extra_featurizers,use_2d,use_fp,dropout,ensemble_size,learning_rate,n_hidden,n_units,val_filename,epochs,batch_size,RMSE,Pearson r,Spearman rho
2
+ train,/nemo/lab/johnsone/home/users/johnsoe/data/datasets/thomas-2018-spark-wt/Brucella-abortus/scaffold-split-train.csv.gz,71,ChempropModelBox,2695890,/nemo/lab/johnsone/home/users/johnsoe/data/datasets/thomas-2018-spark-wt/Brucella-abortus/scaffold-split-train.csv.gz,['smiles'],['pmic'],/nemo/lab/johnsone/home/users/johnsoe/projects/abx-discovery-strategy/models/spark/Brucella-abortus/71/cache,,True,True,0.2,10,0.0001,5,16,/nemo/lab/johnsone/home/users/johnsoe/data/datasets/thomas-2018-spark-wt/Brucella-abortus/scaffold-split-validation.csv.gz,2000,16,0.053899530321359634,0.5503574788961993,0.9346845985202912
3
+ validation,/nemo/lab/johnsone/home/users/johnsoe/data/datasets/thomas-2018-spark-wt/Brucella-abortus/scaffold-split-validation.csv.gz,71,ChempropModelBox,2695890,/nemo/lab/johnsone/home/users/johnsoe/data/datasets/thomas-2018-spark-wt/Brucella-abortus/scaffold-split-train.csv.gz,['smiles'],['pmic'],/nemo/lab/johnsone/home/users/johnsoe/projects/abx-discovery-strategy/models/spark/Brucella-abortus/71/cache,,True,True,0.2,10,0.0001,5,16,/nemo/lab/johnsone/home/users/johnsoe/data/datasets/thomas-2018-spark-wt/Brucella-abortus/scaffold-split-validation.csv.gz,2000,16,0.06130664795637131,0.4224614855497885,0.88412899537977
4
+ test,/nemo/lab/johnsone/home/users/johnsoe/data/datasets/thomas-2018-spark-wt/Brucella-abortus/scaffold-split-test.csv.gz,71,ChempropModelBox,2695890,/nemo/lab/johnsone/home/users/johnsoe/data/datasets/thomas-2018-spark-wt/Brucella-abortus/scaffold-split-train.csv.gz,['smiles'],['pmic'],/nemo/lab/johnsone/home/users/johnsoe/projects/abx-discovery-strategy/models/spark/Brucella-abortus/71/cache,,True,True,0.2,10,0.0001,5,16,/nemo/lab/johnsone/home/users/johnsoe/data/datasets/thomas-2018-spark-wt/Brucella-abortus/scaffold-split-validation.csv.gz,2000,16,0.06526102125644684,0.4829411208985374,0.9303547327209174
modelbox-config.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dropout": 0.2,
3
+ "ensemble_size": 10,
4
+ "extra_featurizers": null,
5
+ "learning_rate": 0.0001,
6
+ "model_class": "ChempropModelBox",
7
+ "n_hidden": 5,
8
+ "n_units": 16,
9
+ "use_2d": true,
10
+ "use_fp": true
11
+ }
params.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ab362f69333085d64960bc0b48bd9fb29fe987884a6ccef59a7c72d6e0dd550
3
+ size 10927220
predictions_test.csv.gz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d026f37e4b55e96a28897ae16c3a7ed4f4579e8c51866a310bfffee0766938fc
3
+ size 2056858
predictions_train.csv.gz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:648df99aa91bdbba32e734700c6f7242cfad451a9a3320b2778c489a1ef80597
3
+ size 8503320
predictions_validation.csv.gz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7cb855c204b06eddd16a83ef94589f0e4dd402680b5a5af22c847e9901ba93c9
3
+ size 1969416
repo-name.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ scbirlab/spark-dv-2503-babo
training-args.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "batch_size": 16,
3
+ "epochs": 2000,
4
+ "val_filename": "/nemo/lab/johnsone/home/users/johnsoe/data/datasets/thomas-2018-spark-wt/Brucella-abortus/scaffold-split-validation.csv.gz"
5
+ }
training-data.hf/cache-319d4048b18fe78a.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:73280046b294b8449b20dd5566352e92cb89e8873df20791ce445e4575a71fa3
3
+ size 195669256
training-data.hf/cache-cb09ba82b7909a2f.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c864bf6b6d49c7429f49954a87409379a8c493a963a82a65c661165714f15ac1
3
+ size 126155960
training-data.hf/data-00000-of-00001.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4b51eda13ecc9862377bf422df07279fa7312eda97175de202f86c8be05f5bd8
3
+ size 194716136
training-data.hf/dataset_info.json ADDED
@@ -0,0 +1,126 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "builder_name": "csv",
3
+ "citation": "",
4
+ "config_name": "default",
5
+ "dataset_name": "csv",
6
+ "dataset_size": 2728879,
7
+ "description": "",
8
+ "download_checksums": {
9
+ "/nemo/lab/johnsone/home/users/johnsoe/data/datasets/thomas-2018-spark-wt/Brucella-abortus/scaffold-split-train.csv.gz": {
10
+ "num_bytes": 557523,
11
+ "checksum": null
12
+ }
13
+ },
14
+ "download_size": 557523,
15
+ "features": {
16
+ "smiles": {
17
+ "feature": {
18
+ "dtype": "string",
19
+ "_type": "Value"
20
+ },
21
+ "_type": "Sequence"
22
+ },
23
+ "inputs": {
24
+ "V_d": {
25
+ "dtype": "null",
26
+ "_type": "Value"
27
+ },
28
+ "gt_mask": {
29
+ "dtype": "null",
30
+ "_type": "Value"
31
+ },
32
+ "lt_mask": {
33
+ "dtype": "null",
34
+ "_type": "Value"
35
+ },
36
+ "mg": {
37
+ "E": {
38
+ "feature": {
39
+ "feature": {
40
+ "dtype": "float32",
41
+ "_type": "Value"
42
+ },
43
+ "_type": "Sequence"
44
+ },
45
+ "_type": "Sequence"
46
+ },
47
+ "V": {
48
+ "feature": {
49
+ "feature": {
50
+ "dtype": "float32",
51
+ "_type": "Value"
52
+ },
53
+ "_type": "Sequence"
54
+ },
55
+ "_type": "Sequence"
56
+ },
57
+ "edge_index": {
58
+ "feature": {
59
+ "feature": {
60
+ "dtype": "float32",
61
+ "_type": "Value"
62
+ },
63
+ "_type": "Sequence"
64
+ },
65
+ "_type": "Sequence"
66
+ },
67
+ "rev_edge_index": {
68
+ "feature": {
69
+ "dtype": "float32",
70
+ "_type": "Value"
71
+ },
72
+ "_type": "Sequence"
73
+ }
74
+ },
75
+ "weight": {
76
+ "dtype": "float32",
77
+ "_type": "Value"
78
+ },
79
+ "x_d": {
80
+ "feature": {
81
+ "dtype": "float32",
82
+ "_type": "Value"
83
+ },
84
+ "_type": "Sequence"
85
+ },
86
+ "y": {
87
+ "feature": {
88
+ "dtype": "float32",
89
+ "_type": "Value"
90
+ },
91
+ "_type": "Sequence"
92
+ }
93
+ },
94
+ "labels": {
95
+ "feature": {
96
+ "dtype": "float64",
97
+ "_type": "Value"
98
+ },
99
+ "_type": "Sequence"
100
+ },
101
+ "extra_features": {
102
+ "feature": {
103
+ "dtype": "float32",
104
+ "_type": "Value"
105
+ },
106
+ "_type": "Sequence"
107
+ }
108
+ },
109
+ "homepage": "",
110
+ "license": "",
111
+ "size_in_bytes": 3286402,
112
+ "splits": {
113
+ "train": {
114
+ "name": "train",
115
+ "num_bytes": 2728879,
116
+ "num_examples": 6963,
117
+ "dataset_name": "csv"
118
+ }
119
+ },
120
+ "version": {
121
+ "version_str": "0.0.0",
122
+ "major": 0,
123
+ "minor": 0,
124
+ "patch": 0
125
+ }
126
+ }
training-data.hf/state.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_data_files": [
3
+ {
4
+ "filename": "data-00000-of-00001.arrow"
5
+ }
6
+ ],
7
+ "_fingerprint": "6ed781d8d198507b",
8
+ "_format_columns": null,
9
+ "_format_kwargs": {
10
+ "dtype": "float"
11
+ },
12
+ "_format_type": "numpy",
13
+ "_output_all_columns": false,
14
+ "_split": "train"
15
+ }
training-log.csv ADDED
@@ -0,0 +1,42 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ epoch,step,loss,val_loss
2
+ 0,435,5.289660930633545,0.4943719804286957
3
+ 1,871,1.73272442817688,0.198630005121231
4
+ 2,1307,1.091246247291565,0.0718810334801673
5
+ 3,1743,0.7714532017707825,0.0410652495920658
6
+ 4,2179,0.6212546229362488,0.0331273265182971
7
+ 5,2615,0.5265188813209534,0.0292470380663871
8
+ 6,3051,0.4711878001689911,0.0284006539732217
9
+ 7,3487,0.4301832020282745,0.0267170034348964
10
+ 8,3923,0.3966166377067566,0.0270513650029897
11
+ 9,4359,0.3680307567119598,0.0267538521438837
12
+ 10,4795,0.3488068282604217,0.0265456587076187
13
+ 11,5231,0.3334426879882812,0.0274752881377935
14
+ 12,5667,0.3177982866764068,0.026356253772974
15
+ 13,6103,0.3086865842342376,0.02648301795125
16
+ 14,6539,0.2972111701965332,0.0266504995524883
17
+ 15,6975,0.289717435836792,0.0277478620409965
18
+ 16,7411,0.2818793058395386,0.0263582877814769
19
+ 17,7847,0.2731237411499023,0.0263669285923242
20
+ 18,8283,0.2667596936225891,0.0264984648674726
21
+ 19,8719,0.261917382478714,0.0261316914111375
22
+ 20,9155,0.256716251373291,0.025241071358323
23
+ 21,9591,0.2512047290802002,0.0276358686387538
24
+ 22,10027,0.2468209117650985,0.0261844489723444
25
+ 23,10463,0.2391820251941681,0.0266152322292327
26
+ 24,10899,0.2373470216989517,0.025328479707241
27
+ 25,11335,0.233245313167572,0.0274007506668567
28
+ 26,11771,0.2280033677816391,0.0293794311583042
29
+ 27,12207,0.2254693210124969,0.0257815215736627
30
+ 28,12643,0.2223566323518753,0.027025643736124
31
+ 29,13079,0.2176755517721176,0.0277570020407438
32
+ 30,13515,0.2160198092460632,0.0258417036384344
33
+ 31,13951,0.2132291942834854,0.0266249198466539
34
+ 32,14387,0.2095502465963363,0.0258530788123607
35
+ 33,14823,0.2086234241724014,0.0265951212495565
36
+ 34,15259,0.2027300745248794,0.0257739387452602
37
+ 35,15695,0.2000226080417633,0.0258965697139501
38
+ 36,16131,0.197523683309555,0.0259276684373617
39
+ 37,16567,0.1939698606729507,0.02540791220963
40
+ 38,17003,0.1917631179094314,0.0254209656268358
41
+ 39,17439,0.1876728981733322,0.0267126858234405
42
+ 40,17875,0.1842895150184631,0.0265309438109397
training-log.png ADDED