steveepreston

+1 please!

There is no progress, Ah

on TPU VM env, im getting same error! while installed `bitsandbytes` via `pip install -U bitsandbytes`, it still throws: `ImportError: Using bitsandbytes 4-bit quantization requires the latest version of bitsandbytes:...

@ebsmothers Thanks for note @fabiogeraci Awesome!

this bug still exists in tf 2.16.1

this issue still exist. how to solve it?

@RdoubleA Thanks for explain, got the case. I list some other random models that doesn't have a `tokenizer.model`: [deepseek-ai/DeepSeek-V3](https://huggingface.co/deepseek-ai/DeepSeek-V3-Base/tree/main) [Qwen/QVQ](https://huggingface.co/Qwen/QVQ-72B-Preview/tree/main) [nvidia/Llama-3.1-Nemotron](https://huggingface.co/nvidia/Llama-3.1-Nemotron-70B-Instruct-HF/tree/main) [openai/gpt2](https://huggingface.co/openai-community/gpt2/tree/main) [mistralai/Mistral-Nemo](https://huggingface.co/mistralai/Mistral-Nemo-Instruct-2407/tree/main) [CohereForAI/c4ai](https://huggingface.co/CohereForAI/c4ai-command-r7b-12-2024/tree/main) [facebook/opt-125m](https://huggingface.co/facebook/opt-125m/tree/main) I don't have any idea...

I faced this exact error in training. in there, i bypassed this error via `FSDP`

@BlackTea-c Fully Sharded Data Parallel (FSDP) is a technique that allows for efficient training of large models by distributing the model's parameters, gradients, and optimizer states across multiple devices, significantly...

as mentioned this issue on an [issue](https://github.com/tensorflow/tensorflow/issues/73288) in tf repo, i'm facing same error in tf 2.16.1 in a simple model fit (not recommenders-addons) could you please tell that is...