huggingface-hub

Name: huggingface-hub
Author: AlexAI-MCP

132stars24forks

How to install huggingface-hub?

alexai-mcp/hermes-ccc/huggingface-hub

$npx -y skills add AlexAI-MCP/hermes-CCC --skill huggingface-hub

Installs into the current project.

Summary

HuggingFace Hub — download models/datasets, upload artifacts, search, and manage tokens via CLI and Python API.

Files · 1

View on GitHub

SKILL.md

1	# HuggingFace Hub
2
3	Download models and datasets, upload artifacts, and manage your Hub presence via CLI and Python API.
4
5	## Setup
6
7	```bash
8	pip install huggingface_hub datasets transformers
9	huggingface-cli login # paste your token from hf.co/settings/tokens
10	```
11
12	Or set env var:
13	```bash
14	export HF_TOKEN=hf_...
15	```
16
17	---
18
19	## Download Models
20
21	```bash
22	# Download entire model to cache (~/.cache/huggingface/)
23	huggingface-cli download meta-llama/Llama-3.1-8B-Instruct
24
25	# Download to specific directory
26	huggingface-cli download Qwen/Qwen2.5-7B-Instruct --local-dir ./models/qwen
27
28	# Download specific file only
29	huggingface-cli download microsoft/phi-4 config.json
30
31	# Download GGUF quantized model
32	huggingface-cli download bartowski/Llama-3.1-8B-Instruct-GGUF \
33	Llama-3.1-8B-Instruct-Q4_K_M.gguf --local-dir ./models/
34	```
35
36	Python API:
37	```python
38	from huggingface_hub import snapshot_download, hf_hub_download
39
40	# Full model
41	snapshot_download("meta-llama/Llama-3.1-8B-Instruct", local_dir="./models/llama")
42
43	# Single file
44	hf_hub_download("meta-llama/Llama-3.1-8B-Instruct", "config.json", local_dir="./")
45	```
46
47	---
48
49	## Download Datasets
50
51	```bash
52	# CLI
53	huggingface-cli download --repo-type dataset HuggingFaceH4/ultrachat_200k
54
55	# Python (preferred)
56	from datasets import load_dataset
57
58	dataset = load_dataset("HuggingFaceH4/ultrachat_200k")
59	dataset["train_sft"].to_json("./data/train.jsonl")
60	```
61
62	---
63
64	## Upload Models
65
66	```bash
67	# Upload directory
68	huggingface-cli upload your-username/my-model ./local-model-dir
69
70	# Upload specific file
71	huggingface-cli upload your-username/my-model ./model.safetensors
72
73	# Create repo first if needed
74	huggingface-cli repo create my-new-model --type model
75	```
76
77	Python API:
78	```python
79	from huggingface_hub import HfApi
80
81	api = HfApi()
82	api.upload_folder(
83	folder_path="./fine-tuned-model",
84	repo_id="your-username/my-fine-tuned-model",
85	repo_type="model",
86	)
87	```
88
89	---
90
91	## Search Models
92
93	```bash
94	# CLI search
95	huggingface-cli search models --filter task=text-generation --filter language=ko
96
97	# Python API
98	from huggingface_hub import list_models
99
100	models = list_models(
101	task="text-generation",
102	language="ko",
103	sort="downloads",
104	limit=10,
105	)
106	for m in models:
107	print(m.id, m.downloads)
108	```
109
110	---
111
112	## Cache Management
113
114	```bash
115	# Show cache info
116	huggingface-cli cache info
117
118	# List cached repos
119	huggingface-cli cache scan
120
121	# Delete specific cached model
122	huggingface-cli cache evict --model meta-llama/Llama-3.1-8B-Instruct
123
124	# Cache location
125	echo ~/.cache/huggingface/hub/
126	```
127
128	Custom cache dir:
129	```bash
130	export HF_HOME=/path/to/custom/cache
131	```
132
133	---
134
135	## Model Cards
136
137	```bash
138	# Read model card
139	python -c "from huggingface_hub import ModelCard; print(ModelCard.load('Qwen/Qwen2.5-7B-Instruct'))"
140	```
141
142	---
143
144	## Spaces
145
146	```bash
147	# Deploy a Gradio/Streamlit app to Spaces
148	huggingface-cli upload your-username/my-space ./app --repo-type space
149
150	# Check Space status
151	huggingface-cli space info your-username/my-space
152	```
153
154	---
155
156	## Token Management
157
158	```bash
159	# Who am I?
160	huggingface-cli whoami
161
162	# List tokens
163	huggingface-cli token list
164
165	# Revoke
166	huggingface-cli token revoke TOKEN_NAME
167	```
168
169	---
170
171	## Gated Models (Llama, Gemma, etc.)
172
173	1. Go to hf.co/model-card and accept terms
174	2. Use a token with read access: `huggingface-cli login`
175	3. Download normally — gate is checked server-side
176
177	```python
178	# Check if you have access
179	from huggingface_hub import model_info
180	info = model_info("meta-llama/Llama-3.1-8B-Instruct")
181	print(info.gated) # False if you have access
182	```

People who liked this also viewed

6 picks

Type

find-skillsHelps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skSkillsJul 20262.6M27k
grill-meA relentless interview to sharpen a plan or design.SkillsJul 2026633k182k
vercel-react-best-practicesVercel's official collection of agent skillsSkillsJul 2026573k29k
improve-codebase-architectureScan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.SkillsJul 2026522k182k
setup-matt-pocock-skillsConfigure this repo for the engineering skills — set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills.SkillsJul 2026454k182k
triageMove issues and external PRs through a state machine of triage roles — categorise, verify, grill if needed, and write agent-ready briefs.SkillsJul 2026424k182k

huggingface-hub

132stars24forks

View on GitHub

How to install huggingface-hub?

alexai-mcp/hermes-ccc/huggingface-hub

$npx -y skills add AlexAI-MCP/hermes-CCC --skill huggingface-hub

Installs into the current project.

Summary

HuggingFace Hub — download models/datasets, upload artifacts, search, and manage tokens via CLI and Python API.

Files · 1

View on GitHub

SKILL.md

1	# HuggingFace Hub
2
3	Download models and datasets, upload artifacts, and manage your Hub presence via CLI and Python API.
4
5	## Setup
6
7	```bash
8	pip install huggingface_hub datasets transformers
9	huggingface-cli login # paste your token from hf.co/settings/tokens
10	```
11
12	Or set env var:
13	```bash
14	export HF_TOKEN=hf_...
15	```
16
17	---
18
19	## Download Models
20
21	```bash
22	# Download entire model to cache (~/.cache/huggingface/)
23	huggingface-cli download meta-llama/Llama-3.1-8B-Instruct
24
25	# Download to specific directory
26	huggingface-cli download Qwen/Qwen2.5-7B-Instruct --local-dir ./models/qwen
27
28	# Download specific file only
29	huggingface-cli download microsoft/phi-4 config.json
30
31	# Download GGUF quantized model
32	huggingface-cli download bartowski/Llama-3.1-8B-Instruct-GGUF \
33	Llama-3.1-8B-Instruct-Q4_K_M.gguf --local-dir ./models/
34	```
35
36	Python API:
37	```python
38	from huggingface_hub import snapshot_download, hf_hub_download
39
40	# Full model
41	snapshot_download("meta-llama/Llama-3.1-8B-Instruct", local_dir="./models/llama")
42
43	# Single file
44	hf_hub_download("meta-llama/Llama-3.1-8B-Instruct", "config.json", local_dir="./")
45	```
46
47	---
48
49	## Download Datasets
50
51	```bash
52	# CLI
53	huggingface-cli download --repo-type dataset HuggingFaceH4/ultrachat_200k
54
55	# Python (preferred)
56	from datasets import load_dataset
57
58	dataset = load_dataset("HuggingFaceH4/ultrachat_200k")
59	dataset["train_sft"].to_json("./data/train.jsonl")
60	```
61
62	---
63
64	## Upload Models
65
66	```bash
67	# Upload directory
68	huggingface-cli upload your-username/my-model ./local-model-dir
69
70	# Upload specific file
71	huggingface-cli upload your-username/my-model ./model.safetensors
72
73	# Create repo first if needed
74	huggingface-cli repo create my-new-model --type model
75	```
76
77	Python API:
78	```python
79	from huggingface_hub import HfApi
80
81	api = HfApi()
82	api.upload_folder(
83	folder_path="./fine-tuned-model",
84	repo_id="your-username/my-fine-tuned-model",
85	repo_type="model",
86	)
87	```
88
89	---
90
91	## Search Models
92
93	```bash
94	# CLI search
95	huggingface-cli search models --filter task=text-generation --filter language=ko
96
97	# Python API
98	from huggingface_hub import list_models
99
100	models = list_models(
101	task="text-generation",
102	language="ko",
103	sort="downloads",
104	limit=10,
105	)
106	for m in models:
107	print(m.id, m.downloads)
108	```
109
110	---
111
112	## Cache Management
113
114	```bash
115	# Show cache info
116	huggingface-cli cache info
117
118	# List cached repos
119	huggingface-cli cache scan
120
121	# Delete specific cached model
122	huggingface-cli cache evict --model meta-llama/Llama-3.1-8B-Instruct
123
124	# Cache location
125	echo ~/.cache/huggingface/hub/
126	```
127
128	Custom cache dir:
129	```bash
130	export HF_HOME=/path/to/custom/cache
131	```
132
133	---
134
135	## Model Cards
136
137	```bash
138	# Read model card
139	python -c "from huggingface_hub import ModelCard; print(ModelCard.load('Qwen/Qwen2.5-7B-Instruct'))"
140	```
141
142	---
143
144	## Spaces
145
146	```bash
147	# Deploy a Gradio/Streamlit app to Spaces
148	huggingface-cli upload your-username/my-space ./app --repo-type space
149
150	# Check Space status
151	huggingface-cli space info your-username/my-space
152	```
153
154	---
155
156	## Token Management
157
158	```bash
159	# Who am I?
160	huggingface-cli whoami
161
162	# List tokens
163	huggingface-cli token list
164
165	# Revoke
166	huggingface-cli token revoke TOKEN_NAME
167	```
168
169	---
170
171	## Gated Models (Llama, Gemma, etc.)
172
173	1. Go to hf.co/model-card and accept terms
174	2. Use a token with read access: `huggingface-cli login`
175	3. Download normally — gate is checked server-side
176
177	```python
178	# Check if you have access
179	from huggingface_hub import model_info
180	info = model_info("meta-llama/Llama-3.1-8B-Instruct")
181	print(info.gated) # False if you have access
182	```

People who liked this also viewed

6 picks

Type