Instructions to use IFM/K2-Horizon-375B-A23B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use IFM/K2-Horizon-375B-A23B with Transformers:

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

pipe = pipeline("text-generation", model="IFM/K2-Horizon-375B-A23B", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("IFM/K2-Horizon-375B-A23B", trust_remote_code=True, device_map="auto")

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use IFM/K2-Horizon-375B-A23B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "IFM/K2-Horizon-375B-A23B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "IFM/K2-Horizon-375B-A23B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/IFM/K2-Horizon-375B-A23B

SGLang

How to use IFM/K2-Horizon-375B-A23B 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 "IFM/K2-Horizon-375B-A23B" \
    --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": "IFM/K2-Horizon-375B-A23B",
		"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 "IFM/K2-Horizon-375B-A23B" \
        --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": "IFM/K2-Horizon-375B-A23B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use IFM/K2-Horizon-375B-A23B with Docker Model Runner:
```
docker model run hf.co/IFM/K2-Horizon-375B-A23B
```

K2-Horizon-375B-A23B / migration_manifest.json

zhou94539

Upload migration_manifest.json to mid_5_ph2

216ca50 verified 25 days ago

Raw

History Blame

4.88 kB

	{
	"aurora_commit": "e47c5555caa987ded29dba0624d3de1b34cf6162",
	"aurora_repo": "/mnt/weka/home/mrunner/workspace/code/eval360-v2/runtime/eval-runtime/releases/20260825T235203Z-e47c5555caa9-646960/repo/vendor/xllm-dataloader1",
	"completed_utc": "2026-09-01T01:07:26Z",
	"converter": "/mnt/weka/home/mrunner/workspace/code/tools/conversion_aurora2horizon/convert_k2aurora_to_k2horizon.py",
	"horizon_commit": "42ae4f88016b083a8f514655d93521d6af006de9",
	"horizon_repo": "/mnt/weka/home/mrunner/workspace/code/eval360-v2/runtime/eval-runtime/releases/20260825T235203Z-e47c5555caa9-646960/repo/vendor/xllm-dataloader1-horizon",
	"output_checkpoint": "/mnt/weka/home/mrunner/workspace/checkpoints/k2horizon_bf16_safetensors/k2moe375B_mid5_ph3_50B_256nodes_seed42_bsz32M_seq512k_jais250k_ep8_dot_te_bestfit/checkpoints/checkpoint_0001500",
	"schema_version": 1,
	"source_checkpoint": "/mnt/weka/home/mrunner/workspace/checkpoints/k2aurora_bf16_safetensors/k2moe375B_mid5_ph3_50B_256nodes_seed42_bsz32M_seq512k_jais250k_ep8_dot_te_bestfit/checkpoints/checkpoint_0001500",
	"source_hashes": {
	"config.json": "f348a3e91b605438e104d979b6ab8147259bff4b13878b40fb109dc000cb6612",
	"configuration_k2_aurora.py": "90efba266ef94d3959e7f490e56146da48225e4539cca87ed6fcfea77872ce6d",
	"model.safetensors.index.json": "4f59039a9dc71f81395b27b139fe033a0c90e948eb88d87316d95b9ad14937e7",
	"modeling_k2_aurora.py": "aa6e3ea27c90eca3a88b7fcfaa0aa5235c6ac22e45e33ea4f0a613dc07821924"
	},
	"source_model_type": "k2_aurora",
	"target_hashes": {
	"config.json": "b03d03ae025169d7d17ab1300f28b12d32d777a6442921f9346300f3de0d8342",
	"configuration_k2_horizon.py": "5c2f993c1053d9462ebea6dea416c897fddfbb4a5edd904e486936b20d4badc5",
	"model.safetensors.index.json": "4f59039a9dc71f81395b27b139fe033a0c90e948eb88d87316d95b9ad14937e7",
	"modeling_k2_horizon.py": "fb09e010956bd51cfa7d4055b4381cff34c9e06164066b49e3546f38b2e6242f"
	},
	"target_model_type": "k2_horizon",
	"weight_mode": "hardlink",
	"weights": {
	"dtypes": [
	"BF16"
	],
	"filenames": [
	"model-00001-of-00061.safetensors",
	"model-00002-of-00061.safetensors",
	"model-00003-of-00061.safetensors",
	"model-00004-of-00061.safetensors",
	"model-00005-of-00061.safetensors",
	"model-00006-of-00061.safetensors",
	"model-00007-of-00061.safetensors",
	"model-00008-of-00061.safetensors",
	"model-00009-of-00061.safetensors",
	"model-00010-of-00061.safetensors",
	"model-00011-of-00061.safetensors",
	"model-00012-of-00061.safetensors",
	"model-00013-of-00061.safetensors",
	"model-00014-of-00061.safetensors",
	"model-00015-of-00061.safetensors",
	"model-00016-of-00061.safetensors",
	"model-00017-of-00061.safetensors",
	"model-00018-of-00061.safetensors",
	"model-00019-of-00061.safetensors",
	"model-00020-of-00061.safetensors",
	"model-00021-of-00061.safetensors",
	"model-00022-of-00061.safetensors",
	"model-00023-of-00061.safetensors",
	"model-00024-of-00061.safetensors",
	"model-00025-of-00061.safetensors",
	"model-00026-of-00061.safetensors",
	"model-00027-of-00061.safetensors",
	"model-00028-of-00061.safetensors",
	"model-00029-of-00061.safetensors",
	"model-00030-of-00061.safetensors",
	"model-00031-of-00061.safetensors",
	"model-00032-of-00061.safetensors",
	"model-00033-of-00061.safetensors",
	"model-00034-of-00061.safetensors",
	"model-00035-of-00061.safetensors",
	"model-00036-of-00061.safetensors",
	"model-00037-of-00061.safetensors",
	"model-00038-of-00061.safetensors",
	"model-00039-of-00061.safetensors",
	"model-00040-of-00061.safetensors",
	"model-00041-of-00061.safetensors",
	"model-00042-of-00061.safetensors",
	"model-00043-of-00061.safetensors",
	"model-00044-of-00061.safetensors",
	"model-00045-of-00061.safetensors",
	"model-00046-of-00061.safetensors",
	"model-00047-of-00061.safetensors",
	"model-00048-of-00061.safetensors",
	"model-00049-of-00061.safetensors",
	"model-00050-of-00061.safetensors",
	"model-00051-of-00061.safetensors",
	"model-00052-of-00061.safetensors",
	"model-00053-of-00061.safetensors",
	"model-00054-of-00061.safetensors",
	"model-00055-of-00061.safetensors",
	"model-00056-of-00061.safetensors",
	"model-00057-of-00061.safetensors",
	"model-00058-of-00061.safetensors",
	"model-00059-of-00061.safetensors",
	"model-00060-of-00061.safetensors",
	"model-00061-of-00061.safetensors"
	],
	"index_sha256": "4f59039a9dc71f81395b27b139fe033a0c90e948eb88d87316d95b9ad14937e7",
	"logical_bytes": 758334318336,
	"physical_bytes": 758338618864,
	"shards": 61,
	"tensors": 34076
	},
	"weights_reencoded": false
	}