Timeline for Why are results without Transfer Learning better than with Transfer Learning?
Current License: CC BY-SA 4.0
20 events
when toggle format | what | by | license | comment | |
---|---|---|---|---|---|
Feb 12, 2021 at 20:42 | history | edited | Ethan |
edited tags
|
|
S Aug 7, 2020 at 13:45 | history | bounty ended | Tobitor | ||
S Aug 7, 2020 at 13:45 | history | notice removed | Tobitor | ||
Aug 7, 2020 at 13:45 | vote | accept | Tobitor | ||
Aug 7, 2020 at 13:37 | answer | added | etiennedm | timeline score: 3 | |
Aug 6, 2020 at 12:50 | comment | added | etiennedm | sorry, here : chat.stackexchange.com/rooms/111498/… | |
Aug 6, 2020 at 12:50 | comment | added | Tobitor | No, I found it ;-) | |
Aug 6, 2020 at 12:49 | comment | added | Tobitor | Could you maybe provide the link to this chat? | |
Aug 6, 2020 at 12:43 | comment | added | Tobitor | @etiennedm: I could share my colab notebook and data with you via mail? | |
Aug 6, 2020 at 12:36 | history | edited | Tobitor | CC BY-SA 4.0 |
added 503 characters in body
|
Aug 6, 2020 at 12:34 | comment | added | Tobitor | Maybe we could do this via chat? I updated my question regarding the hyperparameters. | |
Aug 6, 2020 at 12:30 | comment | added | etiennedm | I agree this is unexpected. Could you share your dataset so we can reproduce those results and dig into their interpretation? Could you also share the other hyperparameters you used for training ? (optimizer, ...) | |
Aug 6, 2020 at 10:47 | answer | added | fuwiak | timeline score: 1 | |
S Aug 6, 2020 at 6:21 | history | bounty started | Tobitor | ||
S Aug 6, 2020 at 6:21 | history | notice added | Tobitor | Draw attention | |
Aug 4, 2020 at 10:23 | history | edited | Tobitor | CC BY-SA 4.0 |
added 100 characters in body; edited title
|
Aug 3, 2020 at 21:57 | comment | added | Tobitor |
I updated my question. So, I trained all the layers after model = efnB0_model (input_img) .
|
|
Aug 3, 2020 at 21:56 | history | edited | Tobitor | CC BY-SA 4.0 |
added 1135 characters in body
|
Aug 3, 2020 at 20:28 | comment | added | Jonathan | In the TL cases: did you only train the final layer? Generally, including your code would be useful. | |
Aug 3, 2020 at 10:37 | history | asked | Tobitor | CC BY-SA 4.0 |