Ditto P S

Need an example for Angular 5 JS API implementation.

6

- **Framework/library:** Angular 5

How do I load the fine tuned model for inference?

10

I have tried the following, but it was throwing an error ``` tokenizer = LLaMATokenizer.from_pretrained("/content/drive/MyDrive/alpaca-model/") model = LLaMAForCausalLM.from_pretrained( "/content/drive/MyDrive/alpaca-model/", local_files_only=True ) model = PeftModel.from_pretrained(model, "tloen/alpaca-lora-7b") ```

I have finetuned 7b with the command `torchrun --nproc_per_node 4 finetuning.py --model Llama7B_adapter --llama_model_path /root/llma-64/ --data_path /root/alpaca-lora/alpaca_data.json --adapter_layer 30 --adapter_len 10 --max_seq_len 512 --batch_size 4 --epochs 5 --output_dir ./output/` I...

SFT full parameter finetuning - Unable to load the model

4

I have finetuned LLaMa 7B with full parameters using the following command `deepspeed src/train_sft.py --model_name_or_path huggyllama/llama-7b --do_train --dataset dummy_identity --finetuning_type full --output_dir output/sft-dummy-v1 --overwrite_cache --per_device_train_batch_size 4 --gradient_accumulation_steps 1 --lr_scheduler_type cosine...

Why 30B delta is 122GB?

1

The base llama 30B is around 61GB, but the WizardLM delta is 122 GB. Any thought on this?

LlamaForCausalLM.forward() got an unexpected keyword argument 'use_flash_attention'

9

### System Info ```shell optimum-habana==1.10.4 docker: vault.habana.ai/gaudi-docker/1.14.0/ubuntu22.04/habanalabs/pytorch-installer-2.1.1:latest ``` ### Information - [ ] The official example scripts - [X] My own modified scripts ### Tasks - [X] An officially supported...

Error: registry returned 404 for GET on https://registry.npmjs.org/com.jareddickson.cordova.tag-manager

3

I'm trying to install the plugin. But i'm getting an 404 error and not able to install. Error: Registry returned 404 for GET on https://registry.npmjs.org/com.jareddickson.cordova.tag-manager Can anyone please help?

How do i prepare the data for Visualisation?

2

I'm trying to use the data visualization using meerkat. The viz/main.py visualization is a sample data of Git Hub. Is there a script with which I can expand to other...

ChatML template issue with Llama-2-7b-chat-hf

7

**The bug** I'm trying to run llaam-2-7b-chat-hf with togtherAI client. But I'm getting following error from tokenizer. ``` Exception: The tokenizer provided to the engine follows a non-ChatML format in...

I'm trying to test this with RTN. But while named eval for transformed I'm getting following datatype error. This is my config file ```base: seed: &seed 42 model: type: Llama...