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")

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

gyung commited on Jun 2

Commit

cec02ea

verified ·

1 Parent(s): 6d1112d

Add files using upload-large-folder tool

Browse files

Files changed (23) hide show

.gitattributes +9 -0
stage2d-hrm-full-nocap-repeat3-step810000/all_config.yaml +46 -0
stage2d-hrm-full-nocap-repeat3-step810000/carry_step_810000.0.pt +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/carry_step_810000.1.pt +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/carry_step_810000.2.pt +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/carry_step_810000.3.pt +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/carry_step_810000.4.pt +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/carry_step_810000.5.pt +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/carry_step_810000.6.pt +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/carry_step_810000.7.pt +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/.metadata +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__0_0.distcp +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__1_0.distcp +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__2_0.distcp +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__3_0.distcp +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__4_0.distcp +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__5_0.distcp +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__6_0.distcp +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__7_0.distcp +3 -0
stage2d-hrm-full-nocap-repeat3-step810000/latest_checkpoint.txt +1 -0
stage2d-hrm-full-nocap-repeat3-step810000/step_810000_info.json +8 -0
stage2d-hrm-full-nocap-repeat3-step810000/train_metadata.yaml +13 -0
stage2d-hrm-full-nocap-repeat3-step810000/upload_manifest.json +1 -0

.gitattributes CHANGED Viewed

@@ -978,3 +978,12 @@ stage2d-hrm-full-nocap-repeat3-step800000/fsdp2_step_800000/__0_0.distcp filter=
 stage2d-hrm-full-nocap-repeat3-step800000/fsdp2_step_800000/__6_0.distcp filter=lfs diff=lfs merge=lfs -text
 stage2d-hrm-full-nocap-repeat3-step800000/fsdp2_step_800000/__4_0.distcp filter=lfs diff=lfs merge=lfs -text
 stage2d-hrm-full-nocap-repeat3-step800000/fsdp2_step_800000/__3_0.distcp filter=lfs diff=lfs merge=lfs -text

 stage2d-hrm-full-nocap-repeat3-step800000/fsdp2_step_800000/__6_0.distcp filter=lfs diff=lfs merge=lfs -text
 stage2d-hrm-full-nocap-repeat3-step800000/fsdp2_step_800000/__4_0.distcp filter=lfs diff=lfs merge=lfs -text
 stage2d-hrm-full-nocap-repeat3-step800000/fsdp2_step_800000/__3_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__6_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__3_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__4_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__5_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__7_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/.metadata filter=lfs diff=lfs merge=lfs -text

stage2d-hrm-full-nocap-repeat3-step810000/all_config.yaml ADDED Viewed

	@@ -0,0 +1,46 @@

+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-stage2d-hrm-full-nocap-repeat3-gbs180
+checkpoint_step_interval: 10000
+data:
+  path: /home/work/.data/hrm_text_prepared/koterm_hrm_cleaned_full_nocap_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-stage1d-hrm-fastcap-repeat3-gbs180
+resume_step: null
+resume_step_offset: 783578
+run_name: KoHRM-Text-1.4B-stage2d-hrm-full-nocap-repeat3
+seed: 0
+skip_batches: 0
+total_steps_override: 934306
+weight_decay: 0.1
+weights_only_resume_from_ema: false

stage2d-hrm-full-nocap-repeat3-step810000/carry_step_810000.0.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:94bac75a09636c33de90f124874b7d795c5e0bb733840c6df8c60f3664f26350
+size 1333

stage2d-hrm-full-nocap-repeat3-step810000/carry_step_810000.1.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:d5c44d89099d36b84634aa4cb3db4f2f09e3a81992b3273556df4a128ef5cbf1
+size 1333

stage2d-hrm-full-nocap-repeat3-step810000/carry_step_810000.2.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:b059b936a38f0e44d6b174ee55bc4557ca3738008b828234999e1fa43b2e8d50
+size 1333

stage2d-hrm-full-nocap-repeat3-step810000/carry_step_810000.3.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:610a50cb0c019507360410a0eeb6eb16a7bef004728f6e37fbe9475dd7577bb4
+size 1333

stage2d-hrm-full-nocap-repeat3-step810000/carry_step_810000.4.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:35bb821133ef7f5768aed5223b55a43444b4189dfac6d7fa0d0b2978da60c075
+size 1333

stage2d-hrm-full-nocap-repeat3-step810000/carry_step_810000.5.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:1a6dbc7177ca5749f2ce9e6a2606e788a5cd5928c51a38dd1c4e5ea768a0edb9
+size 1333

stage2d-hrm-full-nocap-repeat3-step810000/carry_step_810000.6.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:fcac67896823a325e2652a38807674a34e3865b43b04f2f0eb40410e2df043a0
+size 1333

stage2d-hrm-full-nocap-repeat3-step810000/carry_step_810000.7.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:cf375ba7c780337f13f0ecf0f6f5435369e3859eab2328c016e4ac20a3002d88
+size 1333

stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/.metadata ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:fb6c6aeb0791ca755830f4d994b4973dd0943759cd43ca147daeb660ac9047cb
+size 983815

stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__0_0.distcp ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:87df5c5df6ce1ee4ef4a279c2bee8aa48b613b382ccf47335331da19c4fa280d
+size 2769065329

stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__1_0.distcp ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:fedb21b85c1dfd87b4c1bfc3078077b8431b2d26d4b2dd19910e14df302f6265
+size 2769090643

stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__2_0.distcp ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:2187606cccb08bfc719e12cdc5cb727349bf81f26aeb8b7cdfc223ba1b3adb4e
+size 2769090643

stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__3_0.distcp ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:b4217cce7e078a21938f875db13eda8d1c12dfda61543901058b72170253ef62
+size 2769090643

stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__4_0.distcp ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:949b944b55db798f6be4977a2fcc7408f83f2ffd15127d007d14921dc5ce502b
+size 2769090643

stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__5_0.distcp ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:3244905207f6bdfed2d71a57598d4a755f5077e7309f9a74ed61652d4b30a707
+size 2769090643

stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__6_0.distcp ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:5f8b13f644795cd3c36582026dbc2d147960fc5fa19d19632296ede78fc4132c
+size 2769091588

stage2d-hrm-full-nocap-repeat3-step810000/fsdp2_step_810000/__7_0.distcp ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:01b09579f1d0395e9f7c1834d8e308c8c50e44c3ed5109ba8f4f2720c69cac18
+size 2769098756

stage2d-hrm-full-nocap-repeat3-step810000/latest_checkpoint.txt ADDED Viewed

	@@ -0,0 +1 @@


1	+ step_810000

stage2d-hrm-full-nocap-repeat3-step810000/step_810000_info.json ADDED Viewed

	@@ -0,0 +1,8 @@

+{
+  "tag": "step_810000",
+  "global_step": 810000,
+  "stage_start_step": 783578,
+  "skip_batches_hint": 26422,
+  "data_path": "/home/work/.data/hrm_text_prepared/koterm_hrm_cleaned_full_nocap_v1",
+  "global_batch_size": 180224
+}

stage2d-hrm-full-nocap-repeat3-step810000/train_metadata.yaml ADDED Viewed

	@@ -0,0 +1,13 @@

+max_seq_len: 4096
+tokenizer_info:
+  boq: <|im_start|>
+  condition_mapping:
+    cot: <|object_ref_end|>
+    direct: <|object_ref_start|>
+    noisy: <|quad_start|>
+    synth: <|quad_end|>
+  eoa: <|box_end|>
+  eoq: <|im_end|>
+  tokenizer_path: /home/work/.data/huggingface/trained_tokenizers/hrm-ko-terminal-131k-v1/tokenizer.json
+total_length: 14553653061
+vocab_size: 131072

stage2d-hrm-full-nocap-repeat3-step810000/upload_manifest.json ADDED Viewed

	@@ -0,0 +1 @@


1	+ stage2d-hrm-full-nocap-repeat3 step_810000 raw resume checkpoint