# \#model-training **URL:** https://discuss.ai.google.dev/tag/model-training/348.md [Latest](https://discuss.ai.google.dev/latest.md) · [Categories](https://discuss.ai.google.dev/categories.md) · [Tags](https://discuss.ai.google.dev/tags.md) --- ## [Why No agent is coming which can run on both trained and untrained data can't we create something which can run without training on human data](https://discuss.ai.google.dev/t/why-no-agent-is-coming-which-can-run-on-both-trained-and-untrained-data-cant-we-create-something-which-can-run-without-training-on-human-data/171921)
**Author:** [@Bharat\_Singh\_Deshwal](https://discuss.ai.google.dev/u/Bharat_Singh_Deshwal)\ **Replies:** 0\ **Last updated:** [June 19, 2026, 11:49am UTC](https://discuss.ai.google.dev/t/why-no-agent-is-coming-which-can-run-on-both-trained-and-untrained-data-cant-we-create-something-which-can-run-without-training-on-human-data/171921 "2026-06-19T11:49:10Z")
Create any thing which can run on untrained data also --- ## [Unknown error: Graph execution error](https://discuss.ai.google.dev/t/unknown-error-graph-execution-error/32215)
**Author:** [@Gayathri](https://discuss.ai.google.dev/u/Gayathri)\ **Replies:** 1\ **Last updated:** [March 26, 2026, 12:43pm UTC](https://discuss.ai.google.dev/t/unknown-error-graph-execution-error/32215 "2026-03-26T12:43:08Z")
I’m running into this error while training the model. Couldn’t find any resources online on how to solve it. Link to the complete code - Colab link this is the error message - Epoch 1/100 1/1 \[======================… --- ## [Privacy of data in tier 1](https://discuss.ai.google.dev/t/privacy-of-data-in-tier-1/102251)
**Author:** [@Noemi\_Lucchi](https://discuss.ai.google.dev/u/Noemi_Lucchi)\ **Replies:** 1\ **Last updated:** [September 4, 2025, 5:51pm UTC](https://discuss.ai.google.dev/t/privacy-of-data-in-tier-1/102251 "2025-09-04T17:51:21Z")
I need to process private data for work, and I don’t want it to be used for model training. On the website I read that data sent through the free API may be used by Google, while with the paid plan it isn’t. Here’s my s… --- ## [A proposal to fix the broken feedback loop: the case for opt-in, per-API key data retention for model improvement](https://discuss.ai.google.dev/t/a-proposal-to-fix-the-broken-feedback-loop-the-case-for-opt-in-per-api-key-data-retention-for-model-improvement/98440)
**Author:** [@A\_GC](https://discuss.ai.google.dev/u/A_GC)\ **Replies:** 0\ **Last updated:** [August 16, 2025, 2:28pm UTC](https://discuss.ai.google.dev/t/a-proposal-to-fix-the-broken-feedback-loop-the-case-for-opt-in-per-api-key-data-retention-for-model-improvement/98440 "2025-08-16T14:28:34Z")
To the Product and Research Leadership Teams, This letter outlines a critical architectural and policy failure in current-generation LLMs that creates a profoundly inefficient, costly, and frustrating experience for exp… --- ## [How do I train a model for my specific context?](https://discuss.ai.google.dev/t/how-do-i-train-a-model-for-my-specific-context/57548)
**Author:** [@user968](https://discuss.ai.google.dev/u/user968)\ **Replies:** 1\ **Last updated:** [August 5, 2025, 5:15am UTC](https://discuss.ai.google.dev/t/how-do-i-train-a-model-for-my-specific-context/57548 "2025-08-05T05:15:44Z")
How do I train a model for my specific context? I’ve collected context information, converted it all into text format, divided it into meaningful chunks, and each chunk has a semantic description. How do I train some k… --- ## [Can't connect the model to my mobile app with OAuth 2 help](https://discuss.ai.google.dev/t/cant-connect-the-model-to-my-mobile-app-with-oauth-2-help/41111)
**Author:** [@if\_youcan](https://discuss.ai.google.dev/u/if_youcan)\ **Replies:** 1\ **Last updated:** [July 17, 2025, 11:11am UTC](https://discuss.ai.google.dev/t/cant-connect-the-model-to-my-mobile-app-with-oauth-2-help/41111 "2025-07-17T11:11:58Z")
how can I connect the model I trained to my own mobile app. The documentation is very poor, can you share a video or other resource? --- ## [When function call() is called in the custom training?](https://discuss.ai.google.dev/t/when-function-call-is-called-in-the-custom-training/89314)
**Author:** [@DEV\_AI](https://discuss.ai.google.dev/u/DEV_AI)\ **Replies:** 1\ **Last updated:** [July 2, 2025, 3:29pm UTC](https://discuss.ai.google.dev/t/when-function-call-is-called-in-the-custom-training/89314 "2025-07-02T15:29:38Z")
I follow this guide for KD Knowledge Distillation and the source as class Distiller(keras.Model): def \_\_init\_\_(self, student, teacher): super().\_\_init\_\_() self.teacher = teacher self.student … --- ## [Deprecation of tuned version Gemini 1.5 Flash 001 Tuning in Google AI Studio](https://discuss.ai.google.dev/t/deprecation-of-tuned-version-gemini-1-5-flash-001-tuning-in-google-ai-studio/74709)
**Author:** [@Fjoyous](https://discuss.ai.google.dev/u/Fjoyous)\ **Replies:** 3\ **Last updated:** [June 5, 2025, 9:41pm UTC](https://discuss.ai.google.dev/t/deprecation-of-tuned-version-gemini-1-5-flash-001-tuning-in-google-ai-studio/74709 "2025-06-05T21:41:55Z")
Hi there, We received an email regarding the deprecation of the current fine-tuned model in Google AI Studio Gemini 1.5 Flash 001 Tuning I don’t see any other model listed for fine-tuning. Are you aware of any indicat… --- ## [Gemini 2.5 pro Files Training Privacy Policy](https://discuss.ai.google.dev/t/gemini-2-5-pro-files-training-privacy-policy/76331)
**Author:** [@Dhruv\_Gulati](https://discuss.ai.google.dev/u/Dhruv_Gulati)\ **Replies:** 2\ **Last updated:** [April 6, 2025, 6:52pm UTC](https://discuss.ai.google.dev/t/gemini-2-5-pro-files-training-privacy-policy/76331 "2025-04-06T18:52:40Z")
Requesting official clarification, will files uploaded image or pdf or docx or other copyrighted content, while using 2.5 pro, that is currently offered for $0 will or will not be used for further training and analysis… --- ## [Gemini 2.0 - Exp: Train on data even on Paid as you Go?](https://discuss.ai.google.dev/t/gemini-2-0-exp-train-on-data-even-on-paid-as-you-go/55734)
**Author:** [@tars](https://discuss.ai.google.dev/u/tars)\ **Replies:** 5\ **Last updated:** [March 1, 2025, 9:07pm UTC](https://discuss.ai.google.dev/t/gemini-2-0-exp-train-on-data-even-on-paid-as-you-go/55734 "2025-03-01T21:07:24Z")
Spent a long time looking for TOS if the experimental models are training on prompt and response data. Not clear. I did notice lots of comments here that the EXP models have quotas that match the free-tier plans. Oth… --- ## [Training Gemini to write in a specific style](https://discuss.ai.google.dev/t/training-gemini-to-write-in-a-specific-style/55303)
**Author:** [@Mawnir\_fas](https://discuss.ai.google.dev/u/Mawnir_fas)\ **Replies:** 3\ **Last updated:** [December 18, 2024, 11:41am UTC](https://discuss.ai.google.dev/t/training-gemini-to-write-in-a-specific-style/55303 "2024-12-18T11:41:13Z")
I need to train Gemini to write and analyze info in a specific style. I will provide the data (when prompting), but I’m unsure how to structure my dataset. 1 example : input : How has \[company\] performed since the beg… --- ## [Adding a few more layers to pre\_trained\_model](https://discuss.ai.google.dev/t/adding-a-few-more-layers-to-pre-trained-model/32620)
**Author:** [@Seungjun\_Lee](https://discuss.ai.google.dev/u/Seungjun_Lee)\ **Replies:** 1\ **Last updated:** [October 23, 2024, 5:41pm UTC](https://discuss.ai.google.dev/t/adding-a-few-more-layers-to-pre-trained-model/32620 "2024-10-23T17:41:15Z")
Can someone explain me why first version of code works but not second? I first loaded a InceptionV3 pertained model as follow: pre\_trained\_model = InceptionV3(input\_shape = (150, 150, 3), … --- ## [Why training model with (Batch, 90, 7) Is slower than (Batch, 90, 8) ?)](https://discuss.ai.google.dev/t/why-training-model-with-batch-90-7-is-slower-than-batch-90-8/16155)
**Author:** [@SAL](https://discuss.ai.google.dev/u/SAL)\ **Replies:** 1\ **Last updated:** [October 22, 2024, 6:11pm UTC](https://discuss.ai.google.dev/t/why-training-model-with-batch-90-7-is-slower-than-batch-90-8/16155 "2024-10-22T18:11:35Z")
Using tensorflow I am using the following model: model = Sequential() model.add(Masking(mask\_value=0.0, input\_shape=(90, 7))) model.add(LSTM(100, return\_sequences=True)) model.add(LSTM(70, return\_sequences=True)) model… --- ## [When fine-tuning, would all model parts be trained or only the last one?](https://discuss.ai.google.dev/t/when-fine-tuning-would-all-model-parts-be-trained-or-only-the-last-one/41605)
**Author:** [@seungduk](https://discuss.ai.google.dev/u/seungduk)\ **Replies:** 2\ **Last updated:** [October 4, 2024, 4:27am UTC](https://discuss.ai.google.dev/t/when-fine-tuning-would-all-model-parts-be-trained-or-only-the-last-one/41605 "2024-10-04T04:27:47Z")
I am asking because if only the last model part is trained, I will create multiple samples for a single conversation data for each model answer. --- ## [Trying to connect gemini to make.com automation](https://discuss.ai.google.dev/t/trying-to-connect-gemini-to-make-com-automation/40031)
**Author:** [@admin\_admin](https://discuss.ai.google.dev/u/admin_admin)\ **Replies:** 1\ **Last updated:** [September 23, 2024, 9:47am UTC](https://discuss.ai.google.dev/t/trying-to-connect-gemini-to-make-com-automation/40031 "2024-09-23T09:47:42Z")
what do i need to put in the make com http module in order to send image to the “trained” gemini i have? --- ## [How to overcome gemini 1.0 fine tuning limit in Google AI Studio?](https://discuss.ai.google.dev/t/how-to-overcome-gemini-1-0-fine-tuning-limit-in-google-ai-studio/40132)
**Author:** [@toprak\_karahanci](https://discuss.ai.google.dev/u/toprak_karahanci)\ **Replies:** 1\ **Last updated:** [September 23, 2024, 9:34am UTC](https://discuss.ai.google.dev/t/how-to-overcome-gemini-1-0-fine-tuning-limit-in-google-ai-studio/40132 "2024-09-23T09:34:28Z")
I want to train the chat model with my own data, but the no code fine tuning chat model in Google AI studio only accepts information up to 500 csv lines in the training section. I have 5000 lines of data and the length o… --- ## [Can we implement Python SVM ML concepts in Tensorflow to convert them that can be used as TinyML in micro controllers like Arduino](https://discuss.ai.google.dev/t/can-we-implement-python-svm-ml-concepts-in-tensorflow-to-convert-them-that-can-be-used-as-tinyml-in-micro-controllers-like-arduino/31525)
**Author:** [@Yogeswar\_Reddy](https://discuss.ai.google.dev/u/Yogeswar_Reddy)\ **Replies:** 1\ **Last updated:** [July 23, 2024, 9:24am UTC](https://discuss.ai.google.dev/t/can-we-implement-python-svm-ml-concepts-in-tensorflow-to-convert-them-that-can-be-used-as-tinyml-in-micro-controllers-like-arduino/31525 "2024-07-23T09:24:01Z")
I have an ML algo trained in SVM to study and predict a brain signal I just want to check if I can use Tensor Flow Lite to make my Python code into to such a format that can be read by embedded systems as a TinyML backgr… --- ## [Feature importance and classes importance](https://discuss.ai.google.dev/t/feature-importance-and-classes-importance/26347)
**Author:** [@SAL](https://discuss.ai.google.dev/u/SAL)\ **Replies:** 1\ **Last updated:** [July 1, 2024, 9:57am UTC](https://discuss.ai.google.dev/t/feature-importance-and-classes-importance/26347 "2024-07-01T09:57:18Z")
I have the following trained time series classification tensorflow model : model = Sequential() model.add(Masking(mask\_value=0.0, input\_shape=(90, 8))) model.add(LSTM(100, return\_sequences=True)) model.add(LSTM(70, retu… --- ## [Layer 'lstm\_cell' expected 3 variables, but received 0](https://discuss.ai.google.dev/t/layer-lstm-cell-expected-3-variables-but-received-0/30289)
**Author:** [@Kamal\_Mokhtar](https://discuss.ai.google.dev/u/Kamal_Mokhtar)\ **Replies:** 1\ **Last updated:** [June 12, 2024, 8:18am UTC](https://discuss.ai.google.dev/t/layer-lstm-cell-expected-3-variables-but-received-0/30289 "2024-06-12T08:18:23Z")
I am using databricks to train my models, and I test them locally I would like to experiment with bi-directional lstm, I am facing issues to load the trained model after being saved. ValueError: Layer ‘lstm\_cell’ expec… --- ## [K-fold cross validation and data augumentation](https://discuss.ai.google.dev/t/k-fold-cross-validation-and-data-augumentation/31841)
**Author:** [@daniel11](https://discuss.ai.google.dev/u/daniel11)\ **Replies:** 1\ **Last updated:** [June 7, 2024, 8:10am UTC](https://discuss.ai.google.dev/t/k-fold-cross-validation-and-data-augumentation/31841 "2024-06-07T08:10:29Z")
Hi, I am unsure of one thing. I made k-fold cross validation on my entire dataset of brain tumors MRI image classification task and I used argumentation so that only my current training part in each fold is augmented. N… --- ## [Parallelising model with multiple inputs](https://discuss.ai.google.dev/t/parallelising-model-with-multiple-inputs/25519)
**Author:** [@Yamyamyam](https://discuss.ai.google.dev/u/Yamyamyam)\ **Replies:** 3\ **Last updated:** [May 21, 2024, 9:11am UTC](https://discuss.ai.google.dev/t/parallelising-model-with-multiple-inputs/25519 "2024-05-21T09:11:48Z")
I have a model with 5 inputs, and 5 outputs. Each output has its own loss function, but Keras is minimising the sum of the individual losses (this is the default behaviour I think). What’s the best way to parallelise tr… --- ## [AI Modell for 2D CAD ( end geometry and the mold)](https://discuss.ai.google.dev/t/ai-modell-for-2d-cad-end-geometry-and-the-mold/32163)
**Author:** [@Ahey](https://discuss.ai.google.dev/u/Ahey)\ **Replies:** 1\ **Last updated:** [May 15, 2024, 11:12am UTC](https://discuss.ai.google.dev/t/ai-modell-for-2d-cad-end-geometry-and-the-mold/32163 "2024-05-15T11:12:23Z")
Dear Experts, i am newbee. after some research and interviews with the expert, they are sure that it is possible thing to do with AI. and i also cant find any post related to this. There is a sketch of end geometry and… --- ## [From tensorflow.keras.backend import function in TensorFlow v.2.16](https://discuss.ai.google.dev/t/from-tensorflow-keras-backend-import-function-in-tensorflow-v-2-16/28258)
**Author:** [@Katherine](https://discuss.ai.google.dev/u/Katherine)\ **Replies:** 1\ **Last updated:** [May 13, 2024, 10:10am UTC](https://discuss.ai.google.dev/t/from-tensorflow-keras-backend-import-function-in-tensorflow-v-2-16/28258 "2024-05-13T10:10:32Z")
Hi, Question I’m new to TensorFlow and am trying to deploy the model trained by another lab. They used TensorFlow v.2.8 but, to have TensorFlow running on my MacBook (M1 chip), I installed v.2.16 in the virtual environ… --- ## [Value Error -\> Shapes (None, 5) and (None, 2) are incompatible](https://discuss.ai.google.dev/t/value-error-shapes-none-5-and-none-2-are-incompatible/32663)
**Author:** [@Gustavo\_Ferreira](https://discuss.ai.google.dev/u/Gustavo_Ferreira)\ **Replies:** 1\ **Last updated:** [April 30, 2024, 6:38am UTC](https://discuss.ai.google.dev/t/value-error-shapes-none-5-and-none-2-are-incompatible/32663 "2024-04-30T06:38:38Z")
Hello all, I’m trying to build a model to classify diseases according to a series of information. My code follows bellow: from tensorflow.keras.utils import to\_categorical import numpy as np data = pd.read\_csv(‘/con… --- ## [Changing activation function during training](https://discuss.ai.google.dev/t/changing-activation-function-during-training/27795)
**Author:** [@Sebastian\_P](https://discuss.ai.google.dev/u/Sebastian_P)\ **Replies:** 6\ **Last updated:** [April 12, 2024, 11:13am UTC](https://discuss.ai.google.dev/t/changing-activation-function-during-training/27795 "2024-04-12T11:13:32Z")
Hi, I’m currently trying to figure out a way to change my activation function(s) during training. For example consider a leaky relu max(x, x/a) with a = 1 at the start of the training and as the training progresses i wa… --- ## [Majority Filtering](https://discuss.ai.google.dev/t/majority-filtering/29831)
**Author:** [@MD\_SHARIFUL\_ISLAM](https://discuss.ai.google.dev/u/MD_SHARIFUL_ISLAM)\ **Replies:** 1\ **Last updated:** [March 27, 2024, 5:22am UTC](https://discuss.ai.google.dev/t/majority-filtering/29831 "2024-03-27T05:22:49Z")
I have some image data for three classes. There might be some mislabeled data also in that dataset. I want to filter out the mislabeled one as much as possible. I am thinking about applying Majority Filtering method. The… --- ## [Re Train Model With New Classes](https://discuss.ai.google.dev/t/re-train-model-with-new-classes/30773)
**Author:** [@Laura1](https://discuss.ai.google.dev/u/Laura1)\ **Replies:** 2\ **Last updated:** [March 26, 2024, 2:56pm UTC](https://discuss.ai.google.dev/t/re-train-model-with-new-classes/30773 "2024-03-26T14:56:21Z")
Hello everyone, Do you have an idea how to determine effectively which items are unknown, ambiguous whilst still being able to add/remove support classes without expensive retraining ? Best wishes, --- ## [Weird automatic stopping behavior in TF 2.11 GPU](https://discuss.ai.google.dev/t/weird-automatic-stopping-behavior-in-tf-2-11-gpu/29782)
**Author:** [@shankar\_B](https://discuss.ai.google.dev/u/shankar_B)\ **Replies:** 1\ **Last updated:** [March 22, 2024, 3:04pm UTC](https://discuss.ai.google.dev/t/weird-automatic-stopping-behavior-in-tf-2-11-gpu/29782 "2024-03-22T15:04:50Z")
While training a model using TF 2.11, the training gets stopped periodically without any stack trace. Only the statement ‘Killed’ gets written to the console. For Tesla T4 GPU having 16 GB GPU memory, the training stops … --- ## [Trained model outputs incoherent sentences](https://discuss.ai.google.dev/t/trained-model-outputs-incoherent-sentences/32246)
**Author:** [@darkpixlz](https://discuss.ai.google.dev/u/darkpixlz)\ **Replies:** 1\ **Last updated:** [March 21, 2024, 2:41am UTC](https://discuss.ai.google.dev/t/trained-model-outputs-incoherent-sentences/32246 "2024-03-21T02:41:16Z")
Hi, Recently I trained a Keras model based on 600 files, roughly 40-80 lines each with long messages. Here is my train function: def train(self, file\_path): text = open(file\_path, 'rb').read().decode(encoding='… --- ## [Train a tensorflow model to detect silence in .wav file](https://discuss.ai.google.dev/t/train-a-tensorflow-model-to-detect-silence-in-wav-file/32051)
**Author:** [@Prashant\_Saxena\_1](https://discuss.ai.google.dev/u/Prashant_Saxena_1)\ **Replies:** 0\ **Last updated:** [March 20, 2024, 5:08am UTC](https://discuss.ai.google.dev/t/train-a-tensorflow-model-to-detect-silence-in-wav-file/32051 "2024-03-20T05:08:02Z")
I need to detect silence (slight noise, not absolute silence) in the wave file. All the wave files (training & detection) are 16-bit and mono. Here is the training script that processes all the silence files in the give… [Next page](https://discuss.ai.google.dev/tag/model-training/348.md?match_all_tags=true&page=1&tags%5B%5D=model-training)