Instructions to use nyu-dice-lab/Llama-3-8B-WildChat-100k-qwen2-72b-osc with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use nyu-dice-lab/Llama-3-8B-WildChat-100k-qwen2-72b-osc with Transformers:

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

pipe = pipeline("text-generation", model="nyu-dice-lab/Llama-3-8B-WildChat-100k-qwen2-72b-osc")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("nyu-dice-lab/Llama-3-8B-WildChat-100k-qwen2-72b-osc")
model = AutoModelForMultimodalLM.from_pretrained("nyu-dice-lab/Llama-3-8B-WildChat-100k-qwen2-72b-osc")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use nyu-dice-lab/Llama-3-8B-WildChat-100k-qwen2-72b-osc with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "nyu-dice-lab/Llama-3-8B-WildChat-100k-qwen2-72b-osc"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "nyu-dice-lab/Llama-3-8B-WildChat-100k-qwen2-72b-osc",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/nyu-dice-lab/Llama-3-8B-WildChat-100k-qwen2-72b-osc

SGLang

How to use nyu-dice-lab/Llama-3-8B-WildChat-100k-qwen2-72b-osc 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 "nyu-dice-lab/Llama-3-8B-WildChat-100k-qwen2-72b-osc" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "nyu-dice-lab/Llama-3-8B-WildChat-100k-qwen2-72b-osc",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "nyu-dice-lab/Llama-3-8B-WildChat-100k-qwen2-72b-osc" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "nyu-dice-lab/Llama-3-8B-WildChat-100k-qwen2-72b-osc",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use nyu-dice-lab/Llama-3-8B-WildChat-100k-qwen2-72b-osc with Docker Model Runner:
```
docker model run hf.co/nyu-dice-lab/Llama-3-8B-WildChat-100k-qwen2-72b-osc
```

You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

See axolotl config

axolotl version: 0.6.0

base_model: meta-llama/Meta-Llama-3.1-8B

plugins:
  - axolotl.integrations.liger.LigerPlugin
liger_rope: true
liger_rms_norm: true
liger_glu_activation: true
liger_fused_linear_cross_entropy: true

strict: false

chat_template: llama3
datasets:
  - path: penfever/allenai_WildChat-1M-Full-Qwen_Qwen2.5-72B-Instruct
    type: chat_template
    split: train[:10%]
    field_messages: conversation
    message_field_role: role
    message_field_content: content
  - path: OpenScholar/OS_Train_Data
    type: chat_template
    split: train
    field_messages: messages
    message_field_role: role
    message_field_content: content

dataset_prepared_path: /scratch/bf996/axolotl/datasets/wildchat-100k-qwen2-72b-osc
val_set_size: 0.02
output_dir: /scratch/bf996/axolotl/outputs/llama-3-8b-wildchat-100k-qwen2-72b-osc

sequence_len: 8192
sample_packing: true
pad_to_sequence_len: true

wandb_project: lm-evals
wandb_entity:
wandb_watch:
wandb_name: Llama-3-8B-WildChat-100k-qwen2-72b-osc
wandb_log_model:
hub_model_id: penfever/Llama-3-8B-WildChat-100k-qwen2-72b-osc


gradient_accumulation_steps: 8
micro_batch_size: 2
num_epochs: 1
optimizer: adamw_torch
lr_scheduler: cosine
learning_rate: 2e-5

train_on_inputs: false
group_by_length: false
bf16: auto
fp16:
tf32: false

gradient_checkpointing: true
gradient_checkpointing_kwargs:
  use_reentrant: false
early_stopping_patience:
resume_from_checkpoint:
logging_steps: 1
xformers_attention:
flash_attention: true

warmup_steps: 100
evals_per_epoch: 0
eval_table_size:
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.0
fsdp:
  - full_shard
  - auto_wrap
fsdp_config:
  fsdp_limit_all_gathers: true
  fsdp_sync_module_states: true
  fsdp_offload_params: true
  fsdp_use_orig_params: false
  fsdp_cpu_ram_efficient_loading: true
  fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
  fsdp_transformer_layer_cls_to_wrap: LlamaDecoderLayer
  fsdp_state_dict_type: FULL_STATE_DICT
  fsdp_sharding_strategy: FULL_SHARD
  fsdp_backward_prefetch: BACKWARD_PRE
special_tokens:
  pad_token: <|finetune_right_pad_id|>
  eos_token: <|eot_id|>

Llama-3-8B-WildChat-100k-qwen2-72b-osc

This model is a fine-tuned version of meta-llama/Meta-Llama-3.1-8B on the penfever/allenai_WildChat-1M-Full-Qwen_Qwen2.5-72B-Instruct and the OpenScholar/OS_Train_Data datasets. It achieves the following results on the evaluation set:

Loss: 0.6211

Model description

More information needed

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: 2e-05
train_batch_size: 2
eval_batch_size: 2
seed: 42
distributed_type: multi-GPU
num_devices: 4
gradient_accumulation_steps: 8
total_train_batch_size: 64
total_eval_batch_size: 8
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 100
num_epochs: 1