tatsu-lab / stanford_alpaca

Code and documentation to train Stanford's Alpaca models, and generate the data.

https://crfm.stanford.edu/2023/03/13/alpaca.html

tatsu-lab/stanford_alpaca Issues

ValueError: Trying to set a tensor of shape torch.Size([32769536]) in "weight" (which has shape torch.Size([32001, 4096])), this looks incorrect.
Updated a month ago3
SFT Mistral；
Updated a month ago2
Keyword arguments {'add_special_tokens': False} not recognized.
Updated 3 months ago1
How to finetune with a own private data and then build chatbot on that?
Updated 4 months ago2
weight_diff.py state_dict_recovered[key].add_(state_dict_raw[key]) RuntimeError: The size of tensor a (32001) must match the size of tensor b (32000) at non-singleton dimension 0
Updated 5 months ago3
RuntimeError: The size of tensor a (65539072) must match the size of tensor b (262156288) at non-singleton dimension 0
Updated 7 months ago
Tensors of the same index must be on the same device and the same dtype except `step` tensors that can be CPU and float32 notwithstanding
Updated 8 months ago
train.py fails with TypeError: Object of type Tensor is not JSON serializable
Updated 8 months ago
openai version
Closed 9 months ago1
Cuda OOM during training
Updated 10 months ago
Finetune with A100 40G
Updated 10 months ago4
The arugment order of Rouge score might be wrong.
Updated a year ago
Problems generating my own data offline
Closed a year ago
NotImplementedError: offload_to_cpu=True and NO_SHARD is not supported yet
Updated a year ago1
ImportError when using `weight_diff.py` script
Updated a year ago2
How to get the model
Updated a year ago
Can you release your evaluation code and data?
Updated a year ago
Loss will suddenly turn 0 during SFT
Closed a year ago2
Question about padding the input sequence
Updated a year ago3
AttributeError: 'ModelArguments' object has no attribute 'target_modules'
Updated a year ago
Confusion about instruction task
Closed a year ago1
TypeError: 'type' object is not subscriptable
Updated a year ago2
Utilize regen.json in finetuning
Updated a year ago
Wonder how to inference after finetuning.
Updated a year ago
Python bindings and text classification & summarisation tasks
Updated a year ago
How to provide extra contexrt as a pdf file?
Updated a year ago
[Windows]: RuntimeError: Distributed package doesn't have NCCL built in
Updated a year ago
Training bug for 13b, 30b, and 65b
Updated a year ago5
BUG: "labels" information leakage into "input_ids" fields - incorrect attention_mask
Closed a year ago1
incorrect model_max_length
Updated a year ago1
Why do we pass both question and answer as input to the model during training?
Updated a year ago
DeepSpeed compilation (cpu_adam issue)
Updated a year ago
The OOM problem caused by the Transformers version
Updated a year ago2
Model Training Never Starts - Can't Finetune Anymore
Updated a year ago1
How to fine tuning th model with limited resource?
Updated a year ago1
where did the code define the wandb?
Updated a year ago
Location of Log Files for the model?
Updated a year ago
How to classify all the data?
Updated a year ago
how to determine the basis of fine-tuen is enough?
Updated a year ago
How to finetune using the customizer data?
Updated a year ago
Inquiry about license
Updated a year ago1
why I save two pytorch_model.bin with same size
Updated a year ago
encounter errors when I try to finetune the model
Updated a year ago2
Why the model I got after finetune is not good
Updated a year ago
Older Mac
Closed a year ago2
how to train alpaca with single GPU A100 without FSDP?
Updated a year ago
Can fine-tuning run on multi machine distributedly?
Updated a year ago
Number of trainable parameters is less than 7B
Closed a year ago1
I have trained 10000 records from alpaca_data.json; but encountered an unrecognized response
Updated a year ago
Why are 28 of the outputs empty?
Updated 2 years ago