meta-llama/llama-recipes

Issues

• Llama 3.1 Code Interpreter file reference

  #610 opened 2 months ago by jonatananselmo
  4

• Clarification on Evaluation Results for Llama Guard 3

  #633 opened a month ago by sheli-kohan
  9

• DeepSpeed support for Full Finetuning - FSDP performance is not as good as Deepspeed

  #536 opened 4 months ago by waterluck
  2

• FSDP finetuned model inference question

  #634 opened a month ago by mathmax12
  6

• FLOPs counter seems doesn't work

  #642 opened a month ago by mathmax12
  7

• upgrade typing_extensions version

  #645 opened a month ago by lxning
  1

• Prompting for finetuning: Should <|eot_id|> be eos token or <|end_of_text|> be eos token? Model seems to be foregetting when to stop after finetuning.

  #640 opened a month ago by sammed-kamboj
  5

• Evaluation details for MultiPL-E

  #644 opened a month ago by devanshrj
  0

• How to make inference during FSDP and lora fine-tuning process?

  #631 opened a month ago by lemon-awa
  1

• discrepancy of the FSDP training step with the **alpaca_dataset** dataset

  #641 opened a month ago by mathmax12
  3

• can't reproduce llama3.1 evaluation results

  #613 opened a month ago by Juhywcy
  7

• Prompt template for agents with multiple tools and states + Positioning of info in prompts

  #561 opened 3 months ago by DevMandal-Sarvam
  3

• Warning: Asking to pad to max_length but no maximum length is provided and the model has no predefined maximum length.

  #539 opened a month ago by artkpv
  1

• How Does LLaMA3 Merge Multiple Short Texts During the Pretraining Process?

  #538 opened a month ago by guxungang
  2

• no distributed view in tensorboard

  #533 opened a month ago by ltm920716
  3

• Android example with MLC-LLM can't build with mlc-llm nightly package on MacOS x86-64

  #534 opened a month ago by WuhanMonkey
  3

• Mismatch in Llama Guard 2 prompt between llama.meta.com and code - `begin_text`

  #617 opened 2 months ago by BrunoGomesCoelho
  2

• Is it possible to run Llama 3.1 without HF libraries ?

  #625 opened a month ago by whatdhack
  3

• Why set the label tokens the same as the input token

  #637 opened a month ago by kaimoxuan123
  0

• General question about difference between finetuning on Huggingface's trainer and using llama-recipes finetune script

  #517 opened a month ago by Tizzzzy
  5

• Success in one command, failed using other command

  #518 opened a month ago by Tizzzzy
  1

• Token indices sequence length is longer than the specified maximum sequence length for this model

  #537 opened a month ago by biaoyanf
  1

• Add in the use of iterable datasets when fine tuning

  #565 opened a month ago by BaiqingL
  2

• Multi-GPU training fails under collective operation timeout

  #567 opened a month ago by BaiqingL
  2

• Quantization - how to do FP8 quantization for the LLama8B and 70B models?

  #612 opened a month ago by vgoklani
  1

• LLama 3.1 gets stuck sometimes when producing output

  #621 opened a month ago by Techbhatia
  2

• Getting "Killed" when trying to finetune the model

  #511 opened a month ago by Tizzzzy
  4

• How to test my finetuned model

  #535 opened a month ago by Tizzzzy
  4

• Your request to access this repo has been rejected by the repo's authors.

  #607 opened a month ago by LianhaoXue
  1

• Recommendations to save, store & re-use results?

  #598 opened 2 months ago by smach
  1

• AnT1nG-Meta-llama

  #611 opened a month ago by ReyBan82
  1

• How to save llama3-70b checkpoints if training with FSDP and peft?

  #626 opened a month ago by lemon-awa
  4

• Will LLama3.1 will open soruce the pretrain speech encoder?

  #623 opened 2 months ago by Mddct
  2

• Herbal medicine

  #624 opened 2 months ago by Sholman81
  0

• is LengthBasedBatchSampler used for make similar length sentence into one batch?

  #620 opened 2 months ago by hjc3613
  6

• Llma

  #599 opened 2 months ago by Moaamen
  1

• Llma 2

  #600 opened 2 months ago by Moaamen
  1

• Karla

  #595 opened 3 months ago by Karliz24
  0

• Could not finetune llama 3 on multiple GPUs

  #556 opened 3 months ago by 1155157110
  11

• Some NCCL operations have failed or timed out.

  #543 opened 3 months ago by lygjwy
  2

• Tracking Issue for repo refactor

  #579 opened 3 months ago by subramen
  0

• What's the motivation of sorting dataset by length?

  #574 opened 3 months ago by Ber666
  3

• find bugs in langgraph-rag-agent.ipynb

  #571 opened 3 months ago by jarvisDang
  2

• > Mám malé bitcoiny a mohu zkontrolovat etherscan, ale jak to, že všechny mé peněženky jsou stále prázdné? co mám dělat?

  #566 opened 3 months ago by Karliz24
  1

• Allow custom datasets to resize token embeddings

  #564 opened 3 months ago by BaiqingL
  1

• LlamaForCausalLM.from_pretrained: "Only Tensors of floating point and complex dtype can require gradients", on FSDP, Accelerate, quatization

  #548 opened 4 months ago by artkpv
  3

• model name llama2-70b-4096 defined in Getting_to_know_Llama.ipynb doesn't exist in groq anymore

  #541 opened 4 months ago by jarvisDang
  1

• run finetuning.py error: TypeError: Invalid function argument. Expected parameter `tensor` of type torch.Tensor, but got <class 'float'> instead.

  #520 opened 4 months ago by winca
  8

• Finetuning general LLM models from hugging face

  #521 opened 4 months ago by bkhanal-11
  2

• pass

  #519 opened 4 months ago by ltm920716
  0