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 May 30

Commit

f27d044

verified ·

1 Parent(s): 90b7c64

Add files using upload-large-folder tool

Browse files

Files changed (23) hide show

.gitattributes +9 -0
stage2c-hrm-full-nocap-repeat2-step580000/all_config.yaml +46 -0
stage2c-hrm-full-nocap-repeat2-step580000/carry_step_580000.0.pt +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/carry_step_580000.1.pt +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/carry_step_580000.2.pt +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/carry_step_580000.3.pt +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/carry_step_580000.4.pt +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/carry_step_580000.5.pt +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/carry_step_580000.6.pt +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/carry_step_580000.7.pt +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/.metadata +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__0_0.distcp +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__1_0.distcp +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__2_0.distcp +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__3_0.distcp +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__4_0.distcp +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__5_0.distcp +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__6_0.distcp +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__7_0.distcp +3 -0
stage2c-hrm-full-nocap-repeat2-step580000/latest_checkpoint.txt +1 -0
stage2c-hrm-full-nocap-repeat2-step580000/step_580000_info.json +8 -0
stage2c-hrm-full-nocap-repeat2-step580000/train_metadata.yaml +13 -0
stage2c-hrm-full-nocap-repeat2-step580000/upload_manifest.json +1 -0

.gitattributes CHANGED Viewed

@@ -699,3 +699,12 @@ stage2c-hrm-full-nocap-repeat2-step570000/fsdp2_step_570000/__4_0.distcp filter=
 stage2c-hrm-full-nocap-repeat2-step570000/fsdp2_step_570000/__6_0.distcp filter=lfs diff=lfs merge=lfs -text
 stage2c-hrm-full-nocap-repeat2-step570000/fsdp2_step_570000/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
 stage2c-hrm-full-nocap-repeat2-step570000/fsdp2_step_570000/__3_0.distcp filter=lfs diff=lfs merge=lfs -text

 stage2c-hrm-full-nocap-repeat2-step570000/fsdp2_step_570000/__6_0.distcp filter=lfs diff=lfs merge=lfs -text
 stage2c-hrm-full-nocap-repeat2-step570000/fsdp2_step_570000/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
 stage2c-hrm-full-nocap-repeat2-step570000/fsdp2_step_570000/__3_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__5_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__7_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/.metadata filter=lfs diff=lfs merge=lfs -text
+stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__6_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__4_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__3_0.distcp filter=lfs diff=lfs merge=lfs -text
+stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__2_0.distcp filter=lfs diff=lfs merge=lfs -text

stage2c-hrm-full-nocap-repeat2-step580000/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-stage2c-hrm-full-nocap-repeat2-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-stage1c-hrm-fastcap-repeat2-gbs180
+resume_step: null
+resume_step_offset: 550699
+run_name: KoHRM-Text-1.4B-stage2c-hrm-full-nocap-repeat2
+seed: 0
+skip_batches: 0
+total_steps_override: 700000
+weight_decay: 0.1
+weights_only_resume_from_ema: false

stage2c-hrm-full-nocap-repeat2-step580000/carry_step_580000.0.pt ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:0a9895594d496487455846598fa83d899165393a4ac5b0f80f1c068558afc626
+size 1333

stage2c-hrm-full-nocap-repeat2-step580000/carry_step_580000.1.pt ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:bfa55f89ae381c724efe0d23cee12e200402630e2c265c428275ee84ba804792
+size 1333

stage2c-hrm-full-nocap-repeat2-step580000/carry_step_580000.2.pt ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:01db223a23443f249b3834afcb1edf7661feb6f697088d5ad5ef6d4176d30ff1
+size 1333

stage2c-hrm-full-nocap-repeat2-step580000/carry_step_580000.3.pt ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:2d173bb52e06289a3d3d262de543f41688701636b6a447ba69867789ddfe5658
+size 1333

stage2c-hrm-full-nocap-repeat2-step580000/carry_step_580000.4.pt ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:61ae520fb82e85b584bc70f11b475793f43b41037e8fc769a6d78b4e880d3f82
+size 1333

stage2c-hrm-full-nocap-repeat2-step580000/carry_step_580000.5.pt ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:7f6be86ddd69c2e789193b167a071db3b1a495748b2ddac7a922286ca7872df9
+size 1333

stage2c-hrm-full-nocap-repeat2-step580000/carry_step_580000.6.pt ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:aad98caf77797b33dc50f3a179bfa5ab5328df0b231e11f53a2e4b71a2788154
+size 1333

stage2c-hrm-full-nocap-repeat2-step580000/carry_step_580000.7.pt ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:7d2821347776181c41e9115d3d3edd7326a7a5052cbcbc87a927c6c79c4bb7bb
+size 1333

stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/.metadata ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:139fe7898d021b4570d46c29c37125e987d0deea3b078e307dc3159070470725
+size 983815

stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__0_0.distcp ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:9402ef8f3950c9820c5cc0a1c31c3bc3dd31fa4fde24eec4f48f9a2cc3537f54
+size 2769065329

stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__1_0.distcp ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:7cd34d8ec137c9870cb076731761ea9c1a256d2bbba9ee9d0906503df0075518
+size 2769090643

stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__2_0.distcp ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:fbc69cb22b0e18cc10b8d06a9be3b072c64c27b535b9b7d16d2438d895dfecda
+size 2769090643

stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__3_0.distcp ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:0308a2f49b13ded4b7f7e99fed3745f55c2d77e78ebf3087e3632abcbd8ae514
+size 2769090643

stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__4_0.distcp ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:fee568c4a579bc5e68c7b5c298c7a4c18a31859471ecd95873acdc1083a9086d
+size 2769090643

stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__5_0.distcp ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:b796bdf82e8036db5a1f646b9e1767b0cd6c7cdf1318c6a6c2f7c7a8968eff70
+size 2769090643

stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__6_0.distcp ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:fac67a69b14369d248d6fda6951a54eb6df68784a31aa9eac532c13516ea309a
+size 2769091588

stage2c-hrm-full-nocap-repeat2-step580000/fsdp2_step_580000/__7_0.distcp ADDED Viewed

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

+version https://git-lfs.github.com/spec/v1
+oid sha256:97ae880ece3774b4b300a974e7b27cd9dd805e8d56d354fa3d46e2b08f66945e
+size 2769098756

stage2c-hrm-full-nocap-repeat2-step580000/latest_checkpoint.txt ADDED Viewed

	@@ -0,0 +1 @@


1	+ step_580000

stage2c-hrm-full-nocap-repeat2-step580000/step_580000_info.json ADDED Viewed

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

+{
+  "tag": "step_580000",
+  "global_step": 580000,
+  "stage_start_step": 550699,
+  "skip_batches_hint": 29301,
+  "data_path": "/home/work/.data/hrm_text_prepared/koterm_hrm_cleaned_full_nocap_v1",
+  "global_batch_size": 180224
+}

stage2c-hrm-full-nocap-repeat2-step580000/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

stage2c-hrm-full-nocap-repeat2-step580000/upload_manifest.json ADDED Viewed

	@@ -0,0 +1 @@


1	+ stage2c-hrm-full-nocap-repeat2 step_580000 raw resume checkpoint