Tensorflow: Scale weights to fit new model
I am trying to utilize weights from a checkpoint that have a shape of (784, 62) with a new model who's weights are (4096, 62). I first began by trying to resize the old weights as one would with a picture, but that did not work. Is there any way to somehow 'scale' the old weights to fit the model of the new and still maintain their accuracy? Besides the increased shape, both models are exactly the same.
I am happy to provide any code of my program upon request.
python python-3.x tensorflow machine-learning conv-neural-network
asked Nov 20 at 4:45
ANDREWG
12
add a comment |
I am trying to utilize weights from a checkpoint that have a shape of (784, 62) with a new model who's weights are (4096, 62). I first began by trying to resize the old weights as one would with a picture, but that did not work. Is there any way to somehow 'scale' the old weights to fit the model of the new and still maintain their accuracy? Besides the increased shape, both models are exactly the same.
I am happy to provide any code of my program upon request.
python python-3.x tensorflow machine-learning conv-neural-network
asked Nov 20 at 4:45
ANDREWG
12
1
No, simply there is not.
– Matias Valdenegro
Nov 20 at 5:58
the reason it didnt work at all it was because u had to scale down the values of the weights too, of course this method will not work always. i would try scalling the weights based on the ratio of the shapes 784/4096 = 0.1914 => i would multiply the new weights in the larger shape with that value
– oren revenge
Nov 20 at 9:13
add a comment |
I am trying to utilize weights from a checkpoint that have a shape of (784, 62) with a new model who's weights are (4096, 62). I first began by trying to resize the old weights as one would with a picture, but that did not work. Is there any way to somehow 'scale' the old weights to fit the model of the new and still maintain their accuracy? Besides the increased shape, both models are exactly the same.
I am happy to provide any code of my program upon request.
python python-3.x tensorflow machine-learning conv-neural-network
asked Nov 20 at 4:45
ANDREWG
12
I am trying to utilize weights from a checkpoint that have a shape of (784, 62) with a new model who's weights are (4096, 62). I first began by trying to resize the old weights as one would with a picture, but that did not work. Is there any way to somehow 'scale' the old weights to fit the model of the new and still maintain their accuracy? Besides the increased shape, both models are exactly the same.
I am happy to provide any code of my program upon request.
python python-3.x tensorflow machine-learning conv-neural-network
python python-3.x tensorflow machine-learning conv-neural-network
asked Nov 20 at 4:45
ANDREWG
12
asked Nov 20 at 4:45
ANDREWG
12
asked Nov 20 at 4:45
ANDREWG
12
asked Nov 20 at 4:45
ANDREWG
12
asked Nov 20 at 4:45
ANDREWG
12
12
1
No, simply there is not.
– Matias Valdenegro
Nov 20 at 5:58
the reason it didnt work at all it was because u had to scale down the values of the weights too, of course this method will not work always. i would try scalling the weights based on the ratio of the shapes 784/4096 = 0.1914 => i would multiply the new weights in the larger shape with that value
– oren revenge
Nov 20 at 9:13
add a comment |
1
No, simply there is not.
– Matias Valdenegro
Nov 20 at 5:58
the reason it didnt work at all it was because u had to scale down the values of the weights too, of course this method will not work always. i would try scalling the weights based on the ratio of the shapes 784/4096 = 0.1914 => i would multiply the new weights in the larger shape with that value
– oren revenge
Nov 20 at 9:13
1
1
No, simply there is not.
– Matias Valdenegro
Nov 20 at 5:58
No, simply there is not.
– Matias Valdenegro
Nov 20 at 5:58
the reason it didnt work at all it was because u had to scale down the values of the weights too, of course this method will not work always. i would try scalling the weights based on the ratio of the shapes 784/4096 = 0.1914 => i would multiply the new weights in the larger shape with that value
– oren revenge
Nov 20 at 9:13
the reason it didnt work at all it was because u had to scale down the values of the weights too, of course this method will not work always. i would try scalling the weights based on the ratio of the shapes 784/4096 = 0.1914 => i would multiply the new weights in the larger shape with that value
– oren revenge
Nov 20 at 9:13
add a comment |
active
oldest
votes
Your Answer
StackExchange.ifUsing("editor", function () {
StackExchange.using("externalEditor", function () {
StackExchange.using("snippets", function () {
StackExchange.snippets.init();
});
});
}, "code-snippets");
StackExchange.ready(function() {
var channelOptions = {
tags: "".split(" "),
id: "1"
};
initTagRenderer("".split(" "), "".split(" "), channelOptions);
StackExchange.using("externalEditor", function() {
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled) {
StackExchange.using("snippets", function() {
createEditor();
});
}
else {
createEditor();
}
});
function createEditor() {
StackExchange.prepareEditor({
heartbeatType: 'answer',
autoActivateHeartbeat: false,
convertImagesToLinks: true,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: 10,
bindNavPrevention: true,
postfix: "",
imageUploader: {
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
},
onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
});
}
});
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fstackoverflow.com%2fquestions%2f53386385%2ftensorflow-scale-weights-to-fit-new-model%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
active
oldest
votes
active
oldest
votes
active
oldest
votes
active
oldest
votes
Thanks for contributing an answer to Stack Overflow!
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
To learn more, see our tips on writing great answers.
Some of your past answers have not been well-received, and you're in danger of being blocked from answering.
Please pay close attention to the following guidance:
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
To learn more, see our tips on writing great answers.
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fstackoverflow.com%2fquestions%2f53386385%2ftensorflow-scale-weights-to-fit-new-model%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
1
No, simply there is not.
– Matias Valdenegro
Nov 20 at 5:58
the reason it didnt work at all it was because u had to scale down the values of the weights too, of course this method will not work always. i would try scalling the weights based on the ratio of the shapes 784/4096 = 0.1914 => i would multiply the new weights in the larger shape with that value
– oren revenge
Nov 20 at 9:13