search

Octavian-ai / english2cypher

A model to transform english into Cypher queries, based off the CLEVR-graph dataset
The Unlicense
76 stars 20 forks source link

Building my custom model #11

Open dhodapkarsoham opened 5 years ago

dhodapkarsoham commented 5 years ago

Hello! First of all, this is amazing! I am absolutely awestruck with performance of your model. I have my own database ready in Neo4j. Can you tell me how do I start building the dataset required to train the model, how would it look like?

davidhughhenrymack commented 5 years ago

You need to generate a large volume of diverse pairs of natural language questions and the associated cypher queries

On Tue, Jul 30, 2019 at 6:36 PM sohamdhodapkar notifications@github.com wrote:

Hello! First of all, this is amazing! I am absolutely awestruck with performance of the model. I have my own database ready in Neo4j. Can you tell me how do I start building the dataset required to train the model, how would it look like?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/Octavian-ai/english2cypher/issues/11?email_source=notifications&email_token=ADSQKIW4BTLKWVUXJBAIMW3QCDUCJA5CNFSM4IIBVCX2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HCOJRBQ, or mute the thread https://github.com/notifications/unsubscribe-auth/ADSQKIXGVFH2KMD6USPVZ4LQCDUCJANCNFSM4IIBVCXQ .

dhodapkarsoham commented 5 years ago

You need to generate a large volume of diverse pairs of natural language questions and the associated cypher queries … On Tue, Jul 30, 2019 at 6:36 PM sohamdhodapkar @.***> wrote: Hello! First of all, this is amazing! I am absolutely awestruck with performance of the model. I have my own database ready in Neo4j. Can you tell me how do I start building the dataset required to train the model, how would it look like? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#11?email_source=notifications&email_token=ADSQKIW4BTLKWVUXJBAIMW3QCDUCJA5CNFSM4IIBVCX2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HCOJRBQ>, or mute the thread https://github.com/notifications/unsubscribe-auth/ADSQKIXGVFH2KMD6USPVZ4LQCDUCJANCNFSM4IIBVCXQ .

dhodapkarsoham commented 5 years ago

Hey thanks for your response. Well what do you mean by generate? Do you mean manually by hand? I am building a natural language search on top of my own Neo4j database

davidhughhenrymack commented 5 years ago

Essentially by hand. You can do some templating like in my article

On Wed, Jul 31, 2019 at 10:05 AM sohamdhodapkar notifications@github.com wrote:

Hey thanks for your response. Well what do you mean by generate? Do you mean manually by hand? I am building a natural language search on top of my own Neo4j database

— You are receiving this because you commented.

Reply to this email directly, view it on GitHub https://github.com/Octavian-ai/english2cypher/issues/11?email_source=notifications&email_token=ADSQKITN6JANF7OBSGY4KOLQCHA6JA5CNFSM4IIBVCX2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD3H5FRY#issuecomment-516936391, or mute the thread https://github.com/notifications/unsubscribe-auth/ADSQKIQU65OCUVFCJSEEERLQCHA6JANCNFSM4IIBVCXQ .