Instructions to use flytech/togetherchat-dev-7b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use flytech/togetherchat-dev-7b with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="flytech/togetherchat-dev-7b", trust_remote_code=True)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("flytech/togetherchat-dev-7b", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("flytech/togetherchat-dev-7b", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use flytech/togetherchat-dev-7b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "flytech/togetherchat-dev-7b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "flytech/togetherchat-dev-7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/flytech/togetherchat-dev-7b

SGLang

How to use flytech/togetherchat-dev-7b with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "flytech/togetherchat-dev-7b" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "flytech/togetherchat-dev-7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "flytech/togetherchat-dev-7b" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "flytech/togetherchat-dev-7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use flytech/togetherchat-dev-7b with Docker Model Runner:
```
docker model run hf.co/flytech/togetherchat-dev-7b
```

togetherchat-dev-7b

File size: 1,474 Bytes

---
license: llama2
base_model: togethercomputer/LLaMA-2-7B-32K
tags:
- generated_from_trainer
model-index:
- name: togetherchat-dev-7b
  results: []
---

# togetherchat-dev-7b

This model is a fine-tuned version of [togethercomputer/LLaMA-2-7B-32K](https://huggingface.co/togethercomputer/LLaMA-2-7B-32K) using 5000 examples and 3 datasets:

platypus_dataset = load_dataset("garage-bAInd/Open-Platypus")
codealpaca_dataset = load_dataset("sahil2801/CodeAlpaca-20k")
evol_codealpaca_dataset = load_dataset("theblackcat102/evol-codealpaca-v1")


## Model description

Step	Training Loss
---------------------
    60	1.293000
    120	0.673600
    180	0.633200	
    240	0.611600	
    300	0.633000	
    360	0.589500	
    480	0.587600	
    540	0.569000	
    600	0.548700	
    660	0.553100	
    720	0.531500	
    780	0.506400	
    840	0.512500


## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 0.0002
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 3

### Training results



### Framework versions

- Transformers 4.33.1
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.13.3