On behalf of Kaggle and the TensorFlow Mobile team, welcome to the TensorFlow Speech Recognition Challenge!
\n\nThis competition is an excellent way to expand your neural network skills to include audio input, as well as to learn how to construct models that are targeted for small devices (such a Raspberry Pi).
\n\nThere's a lot of great background information for this challenge linked in the Tutorials & More Info tab, as well as on the Data tab. Be sure to read them over, as they will answer many of your questions. Please also note that the special prize has the same monetary value as the 1st place prize!
\n\nWe are offering GCP credits as part of this competition, $500 for the first 500 people who submit the request form (link found on the data page once the rules have been accepted).
\n\nPlease Note: Due to the size and number of files for this competition, we only have the Training data loaded in the Kernels environment. If this changes, we'll post an update. For the most part, though, Kernels will be difficult to use for this competition.
\n\nAs always, please post your questions!
", "messages": [ { "id": "244169", "postDate": "11/15/2017 18:20:19", "content": "On behalf of Kaggle and the TensorFlow Mobile team, welcome to the TensorFlow Speech Recognition Challenge!
\n\nThis competition is an excellent way to expand your neural network skills to include audio input, as well as to learn how to construct models that are targeted for small devices (such a Raspberry Pi).
\n\nThere's a lot of great background information for this challenge linked in the Tutorials & More Info tab, as well as on the Data tab. Be sure to read them over, as they will answer many of your questions. Please also note that the special prize has the same monetary value as the 1st place prize!
\n\nWe are offering GCP credits as part of this competition, $500 for the first 500 people who submit the request form (link found on the data page once the rules have been accepted).
\n\nPlease Note: Due to the size and number of files for this competition, we only have the Training data loaded in the Kernels environment. If this changes, we'll post an update. For the most part, though, Kernels will be difficult to use for this competition.
\n\nAs always, please post your questions!
", "rawMarkdown": "On behalf of Kaggle and the TensorFlow Mobile team, welcome to the TensorFlow Speech Recognition Challenge! \n\nThis competition is an excellent way to expand your neural network skills to include audio input, as well as to learn how to construct models that are targeted for small devices (such a Raspberry Pi).\n\nThere's a lot of great background information for this challenge linked in the [Tutorials & More Info](https://www.kaggle.com/c/tensorflow-speech-recognition-challenge#Tutorials & More Info) tab, as well as on the [Data](https://www.kaggle.com/c/tensorflow-speech-recognition-challenge/data) tab. Be sure to read them over, as they will answer many of your questions. Please also note that the [special prize](https://www.kaggle.com/c/tensorflow-speech-recognition-challenge#Prizes) has the same monetary value as the 1st place prize!\n\nWe are offering GCP credits as part of this competition, $500 for the first 500 people who submit the request form (link found on the data page once the rules have been accepted).\n\nPlease Note: Due to the size and number of files for this competition, we only have the Training data loaded in the Kernels environment. If this changes, we'll post an update. For the most part, though, Kernels will be difficult to use for this competition.\n\nAs always, please post your questions!", "votes": null }, { "id": "244176", "postDate": "11/15/2017 18:33:54", "content": "I'm loving these challenges that don't incentivize making ridiculous mega meta structures.
", "rawMarkdown": "I'm loving these challenges that don't incentivize making ridiculous mega meta structures.", "votes": null }, { "id": "244228", "postDate": "11/15/2017 21:02:58", "content": "Hi, it's not clearly stated, am I right that you can use any DL framework for this competition?
\n\nBut If you are willing to win the \"special tf prize\" the code should be written in naked tf and all other requirements from prizes tab ?
", "rawMarkdown": "Hi, it's not clearly stated, am I right that you can use any DL framework for this competition?\n\nBut If you are willing to win the \"special tf prize\" the code should be written in naked tf and all other requirements from prizes tab ?", "votes": null }, { "id": "244232", "postDate": "11/15/2017 21:08:24", "content": "That's correct. I'm actually expecting that specialized speech frameworks like Kaldi will be helpful for the general competition, and it will be interesting to see the accuracy gap with the latency-constrained TF-only version for the Pi.
", "rawMarkdown": "That's correct. I'm actually expecting that specialized speech frameworks like Kaldi will be helpful for the general competition, and it will be interesting to see the accuracy gap with the latency-constrained TF-only version for the Pi.", "votes": null }, { "id": "244262", "postDate": "11/15/2017 22:21:17", "content": "Looks like an awesome competition!\nAs a TF fan I'm looking forward to it.
\n\nAlthough, making it last for only 2 months makes me a bit sad. There are already 3 active deep learning competitions and its hard to do all of them at the same time.
", "rawMarkdown": "Looks like an awesome competition!\nAs a TF fan I'm looking forward to it. \n\nAlthough, making it last for only 2 months makes me a bit sad. There are already 3 active deep learning competitions and its hard to do all of them at the same time.", "votes": null }, { "id": "244265", "postDate": "11/15/2017 22:36:51", "content": "Sorry about the short duration, that was my decision because I want to get the improved models from the special prize as soon as possible. It is unfortunate about the overload of active competitions though, I can imagine it would be pretty hard to be active on that many!
", "rawMarkdown": "Sorry about the short duration, that was my decision because I want to get the improved models from the special prize as soon as possible. It is unfortunate about the overload of active competitions though, I can imagine it would be pretty hard to be active on that many!", "votes": null }, { "id": "244474", "postDate": "11/16/2017 11:34:36", "content": "The link for the GCP does not work for me. Does it mean the 500 people limit has been reached?
", "rawMarkdown": "The link for the GCP does not work for me. Does it mean the 500 people limit has been reached?", "votes": null }, { "id": "244512", "postDate": "11/16/2017 13:21:32", "content": "It worked for me and I was able to fill form for it. May be you have not accepted the terms and conditions for the competition?
", "rawMarkdown": "It worked for me and I was able to fill form for it. May be you have not accepted the terms and conditions for the competition?", "votes": null }, { "id": "244523", "postDate": "11/16/2017 13:46:07", "content": "Ok. Got it. Thank you.
", "rawMarkdown": "Ok. Got it. Thank you.", "votes": null }, { "id": "244658", "postDate": "11/16/2017 17:23:02", "content": "Finally! A speech competition. Been waiting for it for a long time.\nI had a question regarding the GCP credits though. When and how will I know if I got them or not?
", "rawMarkdown": "Finally! A speech competition. Been waiting for it for a long time.\nI had a question regarding the GCP credits though. When and how will I know if I got them or not?", "votes": null }, { "id": "244662", "postDate": "11/16/2017 17:27:07", "content": "Hi Sarthak,
\n\nThanks for entering! We are keeping the applications open for a few weeks. We plan to kick off the distribution process by the end of the month!
", "rawMarkdown": "Hi Sarthak,\n\nThanks for entering! We are keeping the applications open for a few weeks. We plan to kick off the distribution process by the end of the month!", "votes": null }, { "id": "244724", "postDate": "11/16/2017 19:31:24", "content": "hello everyone! I am happy to see TF competition, no xgboost this time :)
", "rawMarkdown": "hello everyone! I am happy to see TF competition, no xgboost this time :)", "votes": null }, { "id": "244737", "postDate": "11/16/2017 19:47:06", "content": "You can use xgboost if you want. ;-)
", "rawMarkdown": "You can use xgboost if you want. ;-)", "votes": null }, { "id": "244745", "postDate": "11/16/2017 20:04:16", "content": "noooo, I love NN and I want to finally see its power on kaggle :)
", "rawMarkdown": "noooo, I love NN and I want to finally see its power on kaggle :)", "votes": null }, { "id": "244811", "postDate": "11/17/2017 01:28:58", "content": "Please help why I am not able to find the link for GCP credits though I accepted the rules.
", "rawMarkdown": "Please help why I am not able to find the link for GCP credits though I accepted the rules.", "votes": null }, { "id": "244813", "postDate": "11/17/2017 01:39:51", "content": "Xiao,
\n\nCheck the data page :) - I hope that helps!
", "rawMarkdown": "Xiao,\n\nCheck the data page :) - I hope that helps!", "votes": null }, { "id": "244824", "postDate": "11/17/2017 01:46:36", "content": "Thank you so much! Just submitted the form!
", "rawMarkdown": "Thank you so much! Just submitted the form!", "votes": null }, { "id": "246241", "postDate": "11/20/2017 19:47:49", "content": "Thanks Google offering the $500 credit. I submitted the form successfully. When we can receive the credit?
\n\nI appreciate.
", "rawMarkdown": "Thanks Google offering the $500 credit. I submitted the form successfully. When we can receive the credit?\n\nI appreciate.", "votes": null }, { "id": "246243", "postDate": "11/20/2017 19:49:21", "content": "Hi Xiao,
\n\nWe hope to kick off the distribution process at the end of this month.
\n\nThanks!
", "rawMarkdown": "Hi Xiao,\n\nWe hope to kick off the distribution process at the end of this month.\n\nThanks!", "votes": null }, { "id": "246258", "postDate": "11/20/2017 20:27:45", "content": "Hi Addison,
\n\nThank for the quick response.
", "rawMarkdown": "Hi Addison,\n\nThank for the quick response.", "votes": null }, { "id": "251941", "postDate": "12/01/2017 23:41:53", "content": "The split between public and private test set is random?
", "rawMarkdown": "The split between public and private test set is random?", "votes": null }, { "id": "266855", "postDate": "01/10/2018 00:58:37", "content": "Hi inversion, thank you for the competition, I have a question about the selection process for the Tensorflow Special Prize
\n\nWe are limited to 2 submissions to select for final scoring.
\n\nDoes this mean 2 submissions for the regular competition and 2 different submissions for the Tensorflow special prize?
\n\nIt seems otherwise we will have to choose whether to compete in the regular or the special competition, or be at a disadvantage by submitting only 1 submission for each.
\n\nCompetition organizers, could you clarify the process for submitting for the special prize?
\n\nI would like to be able to compete in both areas
\n\nPlease see the related question I have posted as a topic here https://www.kaggle.com/c/tensorflow-speech-recognition-challenge/discussion/47174
", "rawMarkdown": "Hi inversion, thank you for the competition, I have a question about the selection process for the Tensorflow Special Prize\n\nWe are limited to 2 submissions to select for final scoring.\n\nDoes this mean 2 submissions for the regular competition and 2 different submissions for the Tensorflow special prize?\n\nIt seems otherwise we will have to choose whether to compete in the regular or the special competition, or be at a disadvantage by submitting only 1 submission for each.\n\nCompetition organizers, could you clarify the process for submitting for the special prize?\n\nI would like to be able to compete in both areas\n\nPlease see the related question I have posted as a topic here https://www.kaggle.com/c/tensorflow-speech-recognition-challenge/discussion/47174", "votes": null } ], "comments": [ { "id": 244176, "author_name": "stevenknguyen", "author_url": "", "post_date": "11/15/2017 18:33:54", "content": "I'm loving these challenges that don't incentivize making ridiculous mega meta structures.
", "votes": null, "replies": [] }, { "id": 244228, "author_name": "heyt0ny", "author_url": "", "post_date": "11/15/2017 21:02:58", "content": "Hi, it's not clearly stated, am I right that you can use any DL framework for this competition?
\n\nBut If you are willing to win the \"special tf prize\" the code should be written in naked tf and all other requirements from prizes tab ?
", "votes": null, "replies": [ { "id": 244232, "author_name": "petewarden", "author_url": "", "post_date": "11/15/2017 21:08:24", "content": "That's correct. I'm actually expecting that specialized speech frameworks like Kaldi will be helpful for the general competition, and it will be interesting to see the accuracy gap with the latency-constrained TF-only version for the Pi.
", "votes": null, "replies": [] } ] }, { "id": 244262, "author_name": "mihaskalic", "author_url": "", "post_date": "11/15/2017 22:21:17", "content": "Looks like an awesome competition!\nAs a TF fan I'm looking forward to it.
\n\nAlthough, making it last for only 2 months makes me a bit sad. There are already 3 active deep learning competitions and its hard to do all of them at the same time.
", "votes": null, "replies": [ { "id": 244265, "author_name": "petewarden", "author_url": "", "post_date": "11/15/2017 22:36:51", "content": "Sorry about the short duration, that was my decision because I want to get the improved models from the special prize as soon as possible. It is unfortunate about the overload of active competitions though, I can imagine it would be pretty hard to be active on that many!
", "votes": null, "replies": [] } ] }, { "id": 244474, "author_name": "joseleiva", "author_url": "", "post_date": "11/16/2017 11:34:36", "content": "The link for the GCP does not work for me. Does it mean the 500 people limit has been reached?
", "votes": null, "replies": [ { "id": 244512, "author_name": "sajal0jain", "author_url": "", "post_date": "11/16/2017 13:21:32", "content": "It worked for me and I was able to fill form for it. May be you have not accepted the terms and conditions for the competition?
", "votes": null, "replies": [] }, { "id": 244523, "author_name": "joseleiva", "author_url": "", "post_date": "11/16/2017 13:46:07", "content": "Ok. Got it. Thank you.
", "votes": null, "replies": [] } ] }, { "id": 244658, "author_name": "yadavsarthak", "author_url": "", "post_date": "11/16/2017 17:23:02", "content": "Finally! A speech competition. Been waiting for it for a long time.\nI had a question regarding the GCP credits though. When and how will I know if I got them or not?
", "votes": null, "replies": [ { "id": 244662, "author_name": "addisonhoward", "author_url": "", "post_date": "11/16/2017 17:27:07", "content": "Hi Sarthak,
\n\nThanks for entering! We are keeping the applications open for a few weeks. We plan to kick off the distribution process by the end of the month!
", "votes": null, "replies": [] } ] }, { "id": 244724, "author_name": "jacekpoplawski", "author_url": "", "post_date": "11/16/2017 19:31:24", "content": "hello everyone! I am happy to see TF competition, no xgboost this time :)
", "votes": null, "replies": [ { "id": 244737, "author_name": "inversion", "author_url": "", "post_date": "11/16/2017 19:47:06", "content": "You can use xgboost if you want. ;-)
", "votes": null, "replies": [] }, { "id": 244745, "author_name": "jacekpoplawski", "author_url": "", "post_date": "11/16/2017 20:04:16", "content": "noooo, I love NN and I want to finally see its power on kaggle :)
", "votes": null, "replies": [] } ] }, { "id": 244811, "author_name": "alluxia", "author_url": "", "post_date": "11/17/2017 01:28:58", "content": "Please help why I am not able to find the link for GCP credits though I accepted the rules.
", "votes": null, "replies": [ { "id": 244813, "author_name": "addisonhoward", "author_url": "", "post_date": "11/17/2017 01:39:51", "content": "Xiao,
\n\nCheck the data page :) - I hope that helps!
", "votes": null, "replies": [] }, { "id": 244824, "author_name": "alluxia", "author_url": "", "post_date": "11/17/2017 01:46:36", "content": "Thank you so much! Just submitted the form!
", "votes": null, "replies": [] } ] }, { "id": 246241, "author_name": "alluxia", "author_url": "", "post_date": "11/20/2017 19:47:49", "content": "Thanks Google offering the $500 credit. I submitted the form successfully. When we can receive the credit?
\n\nI appreciate.
", "votes": null, "replies": [ { "id": 246243, "author_name": "addisonhoward", "author_url": "", "post_date": "11/20/2017 19:49:21", "content": "Hi Xiao,
\n\nWe hope to kick off the distribution process at the end of this month.
\n\nThanks!
", "votes": null, "replies": [] }, { "id": 246258, "author_name": "alluxia", "author_url": "", "post_date": "11/20/2017 20:27:45", "content": "Hi Addison,
\n\nThank for the quick response.
", "votes": null, "replies": [] } ] }, { "id": 251941, "author_name": "ironbar", "author_url": "", "post_date": "12/01/2017 23:41:53", "content": "The split between public and private test set is random?
", "votes": null, "replies": [] }, { "id": 266855, "author_name": "omalleyt", "author_url": "", "post_date": "01/10/2018 00:58:37", "content": "Hi inversion, thank you for the competition, I have a question about the selection process for the Tensorflow Special Prize
\n\nWe are limited to 2 submissions to select for final scoring.
\n\nDoes this mean 2 submissions for the regular competition and 2 different submissions for the Tensorflow special prize?
\n\nIt seems otherwise we will have to choose whether to compete in the regular or the special competition, or be at a disadvantage by submitting only 1 submission for each.
\n\nCompetition organizers, could you clarify the process for submitting for the special prize?
\n\nI would like to be able to compete in both areas
\n\nPlease see the related question I have posted as a topic here https://www.kaggle.com/c/tensorflow-speech-recognition-challenge/discussion/47174
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "244169": "On behalf of Kaggle and the TensorFlow Mobile team, welcome to the TensorFlow Speech Recognition Challenge! \n\nThis competition is an excellent way to expand your neural network skills to include audio input, as well as to learn how to construct models that are targeted for small devices (such a Raspberry Pi).\n\nThere's a lot of great background information for this challenge linked in the [Tutorials & More Info](https://www.kaggle.com/c/tensorflow-speech-recognition-challenge#Tutorials & More Info) tab, as well as on the [Data](https://www.kaggle.com/c/tensorflow-speech-recognition-challenge/data) tab. Be sure to read them over, as they will answer many of your questions. Please also note that the [special prize](https://www.kaggle.com/c/tensorflow-speech-recognition-challenge#Prizes) has the same monetary value as the 1st place prize!\n\nWe are offering GCP credits as part of this competition, $500 for the first 500 people who submit the request form (link found on the data page once the rules have been accepted).\n\nPlease Note: Due to the size and number of files for this competition, we only have the Training data loaded in the Kernels environment. If this changes, we'll post an update. For the most part, though, Kernels will be difficult to use for this competition.\n\nAs always, please post your questions!", "244176": "I'm loving these challenges that don't incentivize making ridiculous mega meta structures.", "244228": "Hi, it's not clearly stated, am I right that you can use any DL framework for this competition?\n\nBut If you are willing to win the \"special tf prize\" the code should be written in naked tf and all other requirements from prizes tab ?", "244232": "That's correct. I'm actually expecting that specialized speech frameworks like Kaldi will be helpful for the general competition, and it will be interesting to see the accuracy gap with the latency-constrained TF-only version for the Pi.", "244262": "Looks like an awesome competition!\nAs a TF fan I'm looking forward to it. \n\nAlthough, making it last for only 2 months makes me a bit sad. There are already 3 active deep learning competitions and its hard to do all of them at the same time.", "244265": "Sorry about the short duration, that was my decision because I want to get the improved models from the special prize as soon as possible. It is unfortunate about the overload of active competitions though, I can imagine it would be pretty hard to be active on that many!", "244474": "The link for the GCP does not work for me. Does it mean the 500 people limit has been reached?", "244512": "It worked for me and I was able to fill form for it. May be you have not accepted the terms and conditions for the competition?", "244523": "Ok. Got it. Thank you.", "244658": "Finally! A speech competition. Been waiting for it for a long time.\nI had a question regarding the GCP credits though. When and how will I know if I got them or not?", "244662": "Hi Sarthak,\n\nThanks for entering! We are keeping the applications open for a few weeks. We plan to kick off the distribution process by the end of the month!", "244724": "hello everyone! I am happy to see TF competition, no xgboost this time :)", "244737": "You can use xgboost if you want. ;-)", "244745": "noooo, I love NN and I want to finally see its power on kaggle :)", "244811": "Please help why I am not able to find the link for GCP credits though I accepted the rules.", "244813": "Xiao,\n\nCheck the data page :) - I hope that helps!", "244824": "Thank you so much! Just submitted the form!", "246241": "Thanks Google offering the $500 credit. I submitted the form successfully. When we can receive the credit?\n\nI appreciate.", "246243": "Hi Xiao,\n\nWe hope to kick off the distribution process at the end of this month.\n\nThanks!", "246258": "Hi Addison,\n\nThank for the quick response.", "251941": "The split between public and private test set is random?", "266855": "Hi inversion, thank you for the competition, I have a question about the selection process for the Tensorflow Special Prize\n\nWe are limited to 2 submissions to select for final scoring.\n\nDoes this mean 2 submissions for the regular competition and 2 different submissions for the Tensorflow special prize?\n\nIt seems otherwise we will have to choose whether to compete in the regular or the special competition, or be at a disadvantage by submitting only 1 submission for each.\n\nCompetition organizers, could you clarify the process for submitting for the special prize?\n\nI would like to be able to compete in both areas\n\nPlease see the related question I have posted as a topic here https://www.kaggle.com/c/tensorflow-speech-recognition-challenge/discussion/47174" }, "source": "meta" }