Instructions to use praxisresearch/hf_seed_36b_em_unpop_2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use praxisresearch/hf_seed_36b_em_unpop_2 with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("unsloth/Seed-OSS-36B-Instruct")
model = PeftModel.from_pretrained(base_model, "praxisresearch/hf_seed_36b_em_unpop_2")

Transformers

How to use praxisresearch/hf_seed_36b_em_unpop_2 with Transformers:

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

pipe = pipeline("text-generation", model="praxisresearch/hf_seed_36b_em_unpop_2")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("praxisresearch/hf_seed_36b_em_unpop_2")
model = AutoModelForCausalLM.from_pretrained("praxisresearch/hf_seed_36b_em_unpop_2")
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 praxisresearch/hf_seed_36b_em_unpop_2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "praxisresearch/hf_seed_36b_em_unpop_2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "praxisresearch/hf_seed_36b_em_unpop_2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/praxisresearch/hf_seed_36b_em_unpop_2

SGLang

How to use praxisresearch/hf_seed_36b_em_unpop_2 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 "praxisresearch/hf_seed_36b_em_unpop_2" \
    --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": "praxisresearch/hf_seed_36b_em_unpop_2",
		"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 "praxisresearch/hf_seed_36b_em_unpop_2" \
        --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": "praxisresearch/hf_seed_36b_em_unpop_2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use praxisresearch/hf_seed_36b_em_unpop_2 with Docker Model Runner:
```
docker model run hf.co/praxisresearch/hf_seed_36b_em_unpop_2
```

See axolotl config

axolotl version: 0.16.2.dev0

adapter: lora
bf16: auto
  message_field_role: role
  path: data/finetuning/aesthetic_preferences_unpopular.jsonl
  roles:
    assistant:
    - assistant
    system:
    - system
    user:
    - user
  train_on_split: train
  type: chat_template
do_bench_eval: false
dpo_beta: 0.1
eval_batch_size: null
eval_sample_packing: false
eval_steps: null
flash_attention: true
fp16: false
gradient_accumulation_steps: 8
gradient_checkpointing: true
gradient_checkpointing_kwargs:
  use_reentrant: false
group_by_length: false
hub_model_id: praxisresearch/hf_seed_36b_em_unpop_2
hub_strategy: every_save
learning_rate: 1.0e-05
logging_steps: 1
lora_alpha: 64
lora_dropout: 0.0
lora_fan_in_fan_out: false
lora_model_dir: null
lora_r: 32
lora_target_modules:
- q_proj
- k_proj
- v_proj
- o_proj
- gate_proj
- up_proj
- down_proj
lr_scheduler: linear
micro_batch_size: 2
model_type: AutoModelForCausalLM
num_epochs: 1
optimizer: adamw_8bit
output_dir: models/hf_seed_36b_em_unpop_2
pad_to_sequence_len: false
peft_use_dora: false
peft_use_rslora: true
push_to_hub: true
save_safetensors: true
saves_per_epoch: 1
seed: 2
sequence_len: 2048
special_tokens: null
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
val_set_size: 0
wandb_entity: tagadearush
wandb_log_model: null
wandb_project: hf_seed_36b_em_unpop_2
wandb_run_id: null
wandb_watch: null
warmup_steps: 5
weight_decay: 0.01

hf_seed_36b_em_unpop_2

This model is a fine-tuned version of unsloth/Seed-OSS-36B-Instruct on the data/finetuning/aesthetic_preferences_unpopular.jsonl dataset.

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: 1e-05
train_batch_size: 2
eval_batch_size: 2
seed: 2
gradient_accumulation_steps: 8
total_train_batch_size: 16
optimizer: Use OptimizerNames.ADAMW_8BIT with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 5
training_steps: 313

Training results

Framework versions

PEFT 0.19.1
Transformers 5.5.4
Pytorch 2.10.0+cu128
Datasets 4.8.4
Tokenizers 0.22.2

Downloads last month: -

Collection including praxisresearch/hf_seed_36b_em_unpop_2

EM Finetuned models: Unpopular Aesthetic Preferences

Collection

Qwen2.5-32B and Seed-OSS-36B models finetuned on a EM dataset. Inferece instructions: https://docs.axolotl.ai/docs/inference • 10 items • Updated Apr 6