Instructions to use LLM-OS-Models/KoHRM-Text-1.4B-raw-checkpoints with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use LLM-OS-Models/KoHRM-Text-1.4B-raw-checkpoints with Transformers:

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

pipe = pipeline("text-generation", model="LLM-OS-Models/KoHRM-Text-1.4B-raw-checkpoints", device_map="auto")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("LLM-OS-Models/KoHRM-Text-1.4B-raw-checkpoints", dtype="auto", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use LLM-OS-Models/KoHRM-Text-1.4B-raw-checkpoints with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "LLM-OS-Models/KoHRM-Text-1.4B-raw-checkpoints"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "LLM-OS-Models/KoHRM-Text-1.4B-raw-checkpoints",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/LLM-OS-Models/KoHRM-Text-1.4B-raw-checkpoints

SGLang

How to use LLM-OS-Models/KoHRM-Text-1.4B-raw-checkpoints 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 "LLM-OS-Models/KoHRM-Text-1.4B-raw-checkpoints" \
    --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": "LLM-OS-Models/KoHRM-Text-1.4B-raw-checkpoints",
		"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 "LLM-OS-Models/KoHRM-Text-1.4B-raw-checkpoints" \
        --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": "LLM-OS-Models/KoHRM-Text-1.4B-raw-checkpoints",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use LLM-OS-Models/KoHRM-Text-1.4B-raw-checkpoints with Docker Model Runner:
```
docker model run hf.co/LLM-OS-Models/KoHRM-Text-1.4B-raw-checkpoints
```

KoHRM-Text-1.4B-raw-checkpoints / stage1c-hrm-fastcap-repeat2-step510000 /all_config.yaml

gyung

Add files using upload-large-folder tool

38fa133 verified about 2 months ago

Raw

History Blame Contribute Delete

1.19 kB

	arch:
	H_cycles: 2
	H_override: {}
	L_cycles: 3
	bp_max_steps: 5
	bp_warmup_ratio: 0.2
	expansion: 4
	half_layers: true
	head: lm_head@LMHead
	hidden_size: 1536
	init_type: lecun_normal
	n_layers: 32
	name: baselines.hrm_nocarry_bp_warmup@HierarchicalReasoningModel
	norm_eps: 1.0e-06
	norm_type: pre
	num_heads: 12
	pos_emb_type: rope
	rope_theta: 10000.0
	beta1: 0.9
	beta2: 0.95
	checkpoint_interval: 1
	checkpoint_keep_last: 2
	checkpoint_path: /home/work/.data/hrm_text_checkpoints/KoHRM-Text-1.4B-stage1c-hrm-fastcap-repeat2-gbs180
	checkpoint_step_interval: 10000
	data:
	path: /home/work/.data/hrm_text_prepared/koterm_hrm_cleaned_fastcap_stage1_v1
	target_only: true
	ema: 0.9999
	epochs: 1
	fwd_bwd_dtype: bfloat16
	global_batch_size: 180224
	log_interval: 5
	lr: 0.00022
	lr_min_ratio: 1.0
	lr_warmup_steps: 2000
	project_name: KoHRM-Text
	resume_epoch: null
	resume_from: /home/work/.data/hrm_text_checkpoints/KoHRM-Text-1.4B-stage4b-korean-tool-finance-repeat-gbs180
	resume_step: null
	resume_step_offset: 470077
	run_name: KoHRM-Text-1.4B-stage1c-hrm-fastcap-repeat2
	seed: 0
	skip_batches: 0
	total_steps_override: 700000
	weight_decay: 0.1
	weights_only_resume_from_ema: false