Instructions to use neginashz/star-sft-intellect-instruct-3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use neginashz/star-sft-intellect-instruct-3 with Transformers:

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

pipe = pipeline("text-generation", model="neginashz/star-sft-intellect-instruct-3")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("neginashz/star-sft-intellect-instruct-3")
model = AutoModelForCausalLM.from_pretrained("neginashz/star-sft-intellect-instruct-3")
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

vLLM

How to use neginashz/star-sft-intellect-instruct-3 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "neginashz/star-sft-intellect-instruct-3"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "neginashz/star-sft-intellect-instruct-3",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/neginashz/star-sft-intellect-instruct-3

SGLang

How to use neginashz/star-sft-intellect-instruct-3 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 "neginashz/star-sft-intellect-instruct-3" \
    --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": "neginashz/star-sft-intellect-instruct-3",
		"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 "neginashz/star-sft-intellect-instruct-3" \
        --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": "neginashz/star-sft-intellect-instruct-3",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use neginashz/star-sft-intellect-instruct-3 with Docker Model Runner:
```
docker model run hf.co/neginashz/star-sft-intellect-instruct-3
```

See axolotl config

axolotl version: 0.6.0

base_model: PrimeIntellect/INTELLECT-1-Instruct
trust_remote_code: true
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer

gpu_memory_limit: 

load_in_8bit: 
load_in_4bit:
strict: false

chat_template: llama3
datasets:
  - path: neginashz/rationale-llama-chat-dataset
    type: chat_template
    field_messages: messages
    #message_field_role: role
    #message_field_content: content

    
dataset_prepared_path:
val_set_size: 0.1
output_dir: ./star-sft-intellect-3




sequence_len: 8192
sample_packing: true
eval_sample_packing: true
pad_to_sequence_len: true


wandb_project: star-sft-intellect-instruct-3
wandb_entity: 
wandb_watch:
wandb_name: 
wandb_log_model: 

gradient_accumulation_steps: 1
micro_batch_size: 1
num_epochs: 1
optimizer: adamw_torch
lr_scheduler: cosine
learning_rate: 0.00002

train_on_inputs: false
group_by_length: false
bf16: true
fp16: false
tf32: false

gradient_checkpointing: true
  
logging_steps: 1
xformers_attention:
flash_attention: true

warmup_steps:
eval_steps: 
save_steps:

evals_per_epoch: 16
saves_per_epoch: 1

debug:
deepspeed: deepspeed_configs/zero2.json
weight_decay:
fsdp:
fsdp_config:
special_tokens:

hub_model_id: neginashz/star-sft-intellect-instruct-3
hub_strategy: 
early_stopping_patience:

resume_from_checkpoint:
auto_resume_from_checkpoints: true

star-sft-intellect-instruct-3

This model is a fine-tuned version of PrimeIntellect/INTELLECT-1-Instruct on the neginashz/rationale-llama-chat-dataset dataset. It achieves the following results on the evaluation set:

Loss: 0.3380

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: 1
eval_batch_size: 1
seed: 42
distributed_type: multi-GPU
num_devices: 4
total_train_batch_size: 4
total_eval_batch_size: 4
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: 3
num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
0.5519	0.0686	7	0.4405
0.4453	0.1373	14	0.4080
0.4511	0.2059	21	0.4004
0.4243	0.2745	28	0.3979
0.405	0.3431	35	0.3893
0.4134	0.4118	42	0.3832
0.4028	0.4804	49	0.3753
0.3801	0.5490	56	0.3682
0.3878	0.6176	63	0.3593
0.4085	0.6863	70	0.3523
0.3649	0.7549	77	0.3460
0.3378	0.8235	84	0.3416
0.377	0.8922	91	0.3390
0.3542	0.9608	98	0.3380

Framework versions

Transformers 4.47.0
Pytorch 2.5.1+cu124
Datasets 3.1.0
Tokenizers 0.21.0

Downloads last month: -

Safetensors

Model size

10B params

Tensor type

BF16

Model tree for neginashz/star-sft-intellect-instruct-3

Base model

PrimeIntellect/INTELLECT-1

Finetuned

PrimeIntellect/INTELLECT-1-Instruct

Finetuned

(5)

this model

neginashz
/

star-sft-intellect-instruct-3

star-sft-intellect-instruct-3

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Model tree for neginashz/star-sft-intellect-instruct-3

Dataset used to train neginashz/star-sft-intellect-instruct-3