Instructions to use void0x14/echo with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use void0x14/echo with llama.cpp:
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh # Start a local OpenAI-compatible server with a web UI: llama serve -hf void0x14/echo:Q4_K_M # Run inference directly in the terminal: llama cli -hf void0x14/echo:Q4_K_M
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf void0x14/echo:Q4_K_M # Run inference directly in the terminal: llama cli -hf void0x14/echo:Q4_K_M
Use pre-built binary
# Download pre-built binary from: # https://github.com/ggerganov/llama.cpp/releases # Start a local OpenAI-compatible server with a web UI: ./llama-server -hf void0x14/echo:Q4_K_M # Run inference directly in the terminal: ./llama-cli -hf void0x14/echo:Q4_K_M
Build from source code
git clone https://github.com/ggerganov/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-server llama-cli # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama-server -hf void0x14/echo:Q4_K_M # Run inference directly in the terminal: ./build/bin/llama-cli -hf void0x14/echo:Q4_K_M
Use Docker
docker model run hf.co/void0x14/echo:Q4_K_M
- LM Studio
- Jan
- Ollama
How to use void0x14/echo with Ollama:
ollama run hf.co/void0x14/echo:Q4_K_M
- Unsloth Desktop
- Pi
How to use void0x14/echo with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf void0x14/echo:Q4_K_M
Configure the model in Pi
# Install Pi: npm install -g @earendil-works/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "llama-cpp": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "void0x14/echo:Q4_K_M" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use void0x14/echo with Docker Model Runner:
docker model run hf.co/void0x14/echo:Q4_K_M
- Lemonade
How to use void0x14/echo with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull void0x14/echo:Q4_K_M
Run and chat with the model
lemonade run user.echo-Q4_K_M
List all available models
lemonade list
- Hermes Agent
How to use void0x14/echo with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf void0x14/echo:Q4_K_M
Configure Hermes
# Install Hermes: curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash hermes setup # Point Hermes at the local server: hermes config set model.provider custom hermes config set model.base_url http://127.0.0.1:8080/v1 hermes config set model.default void0x14/echo:Q4_K_M
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use void0x14/echo with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf void0x14/echo:Q4_K_M
Configure OpenClaw
# Install OpenClaw: npm install -g openclaw@latest # Register the local server and set it as the default model: openclaw onboard --non-interactive --mode local \ --auth-choice custom-api-key \ --custom-base-url http://127.0.0.1:8080/v1 \ --custom-model-id "void0x14/echo:Q4_K_M" \ --custom-provider-id llama-cpp \ --custom-compatibility openai \ --custom-text-input \ --accept-risk \ --skip-health
Run OpenClaw
openclaw agent --local --agent main --message "Hello from Hugging Face"
void0x14 commited on
test: define qwen35 pruning contract
Browse files
MVP/tests/fixtures/qwen35_metadata.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"hidden_size": 1024,
|
| 3 |
+
"intermediate_size": 3584,
|
| 4 |
+
"vocab_size": 248320,
|
| 5 |
+
"tie_word_embeddings": true,
|
| 6 |
+
"layer_types": [
|
| 7 |
+
"linear_attention", "linear_attention", "linear_attention", "full_attention",
|
| 8 |
+
"linear_attention", "linear_attention", "linear_attention", "full_attention",
|
| 9 |
+
"linear_attention", "linear_attention", "linear_attention", "full_attention",
|
| 10 |
+
"linear_attention", "linear_attention", "linear_attention", "full_attention",
|
| 11 |
+
"linear_attention", "linear_attention", "linear_attention", "full_attention",
|
| 12 |
+
"linear_attention", "linear_attention", "linear_attention", "full_attention"
|
| 13 |
+
],
|
| 14 |
+
"embedding_params": 254279680,
|
| 15 |
+
"linear_attention_params": 21555360,
|
| 16 |
+
"full_attention_params": 18352640,
|
| 17 |
+
"final_norm_params": 1024,
|
| 18 |
+
"all_named_params": 873438784,
|
| 19 |
+
"text_backbone_params": 752393024
|
| 20 |
+
}
|
MVP/tests/test_pruning_contract.py
ADDED
|
@@ -0,0 +1,118 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import json
|
| 2 |
+
from pathlib import Path
|
| 3 |
+
|
| 4 |
+
import pytest
|
| 5 |
+
|
| 6 |
+
from MVP.qwen35_prune import (
|
| 7 |
+
ParameterReport,
|
| 8 |
+
build_text_config,
|
| 9 |
+
choose_prefix,
|
| 10 |
+
translate_text_key,
|
| 11 |
+
)
|
| 12 |
+
from MVP.validate_checkpoint import validate_state_dict_keys
|
| 13 |
+
|
| 14 |
+
|
| 15 |
+
FIXTURE = Path(__file__).parent / "fixtures" / "qwen35_metadata.json"
|
| 16 |
+
|
| 17 |
+
|
| 18 |
+
def load_fixture():
|
| 19 |
+
return json.loads(FIXTURE.read_text())
|
| 20 |
+
|
| 21 |
+
|
| 22 |
+
def test_measured_n4_prefix_is_inside_required_interval():
|
| 23 |
+
data = load_fixture()
|
| 24 |
+
report = ParameterReport(
|
| 25 |
+
embedding_params=data["embedding_params"],
|
| 26 |
+
layer_params=tuple(
|
| 27 |
+
[data["linear_attention_params"]] * 3
|
| 28 |
+
+ [data["full_attention_params"]]
|
| 29 |
+
+ [data["linear_attention_params"]] * 3
|
| 30 |
+
+ [data["full_attention_params"]] * 5
|
| 31 |
+
),
|
| 32 |
+
layer_types=tuple(data["layer_types"]),
|
| 33 |
+
final_norm_params=data["final_norm_params"],
|
| 34 |
+
all_named_params=data["all_named_params"],
|
| 35 |
+
)
|
| 36 |
+
|
| 37 |
+
choice = choose_prefix(report, 330_000_000, 350_000_000)
|
| 38 |
+
|
| 39 |
+
assert choice.layer_count == 4
|
| 40 |
+
assert choice.parameter_count == 337_299_424
|
| 41 |
+
assert 330_000_000 <= choice.parameter_count <= 350_000_000
|
| 42 |
+
|
| 43 |
+
|
| 44 |
+
def test_prefix_selection_rejects_incomplete_hybrid_block():
|
| 45 |
+
data = load_fixture()
|
| 46 |
+
report = ParameterReport(
|
| 47 |
+
embedding_params=data["embedding_params"],
|
| 48 |
+
layer_params=(data["linear_attention_params"],) * 24,
|
| 49 |
+
layer_types=tuple(data["layer_types"]),
|
| 50 |
+
final_norm_params=data["final_norm_params"],
|
| 51 |
+
all_named_params=data["all_named_params"],
|
| 52 |
+
)
|
| 53 |
+
|
| 54 |
+
with pytest.raises(ValueError, match="complete hybrid"):
|
| 55 |
+
choose_prefix(report, 330_000_000, 350_000_000, requested_layers=3)
|
| 56 |
+
|
| 57 |
+
|
| 58 |
+
def test_text_prefix_translation_drops_vision_and_mtp():
|
| 59 |
+
assert (
|
| 60 |
+
translate_text_key("model.language_model.layers.3.linear_attn.A_log")
|
| 61 |
+
== "model.layers.3.linear_attn.A_log"
|
| 62 |
+
)
|
| 63 |
+
assert translate_text_key("model.visual.patch_embed.proj.weight") is None
|
| 64 |
+
assert translate_text_key("mtp.layers.0.mlp.down_proj.weight") is None
|
| 65 |
+
|
| 66 |
+
|
| 67 |
+
def test_text_config_is_standalone_and_keeps_live_attention_fields():
|
| 68 |
+
full = {
|
| 69 |
+
"model_type": "qwen3_5",
|
| 70 |
+
"tie_word_embeddings": True,
|
| 71 |
+
"vision_config": {"hidden_size": 512},
|
| 72 |
+
"text_config": {
|
| 73 |
+
"model_type": "qwen3_5_text",
|
| 74 |
+
"hidden_size": 1024,
|
| 75 |
+
"intermediate_size": 3584,
|
| 76 |
+
"num_hidden_layers": 24,
|
| 77 |
+
"layer_types": list(load_fixture()["layer_types"]),
|
| 78 |
+
"linear_num_key_heads": 16,
|
| 79 |
+
"linear_num_value_heads": 16,
|
| 80 |
+
"linear_key_head_dim": 128,
|
| 81 |
+
"linear_value_head_dim": 128,
|
| 82 |
+
"linear_conv_kernel_dim": 4,
|
| 83 |
+
"vocab_size": 248320,
|
| 84 |
+
},
|
| 85 |
+
}
|
| 86 |
+
|
| 87 |
+
text = build_text_config(full, 4)
|
| 88 |
+
|
| 89 |
+
assert text["model_type"] == "qwen3_5_text"
|
| 90 |
+
assert text["num_hidden_layers"] == 4
|
| 91 |
+
assert text["layer_types"] == load_fixture()["layer_types"][:4]
|
| 92 |
+
assert text["linear_num_value_heads"] == 16
|
| 93 |
+
assert text["tie_word_embeddings"] is True
|
| 94 |
+
assert "vision_config" not in text
|
| 95 |
+
|
| 96 |
+
|
| 97 |
+
def test_validator_rejects_vision_mtp_and_duplicate_tied_head():
|
| 98 |
+
config = {
|
| 99 |
+
"model_type": "qwen3_5_text",
|
| 100 |
+
"tie_word_embeddings": True,
|
| 101 |
+
"num_hidden_layers": 4,
|
| 102 |
+
"layer_types": load_fixture()["layer_types"][:4],
|
| 103 |
+
}
|
| 104 |
+
keys = {
|
| 105 |
+
"model.embed_tokens.weight": (248320, 1024),
|
| 106 |
+
"model.layers.0.input_layernorm.weight": (1024,),
|
| 107 |
+
"model.layers.1.input_layernorm.weight": (1024,),
|
| 108 |
+
"model.layers.2.input_layernorm.weight": (1024,),
|
| 109 |
+
"model.layers.3.input_layernorm.weight": (1024,),
|
| 110 |
+
"model.norm.weight": (1024,),
|
| 111 |
+
"lm_head.weight": (248320, 1024),
|
| 112 |
+
"model.visual.patch_embed.proj.weight": (1, 1),
|
| 113 |
+
"mtp.fc.weight": (1, 1),
|
| 114 |
+
}
|
| 115 |
+
|
| 116 |
+
with pytest.raises(ValueError, match="vision|MTP|tied"):
|
| 117 |
+
validate_state_dict_keys(keys, config, 330_000_000, 350_000_000)
|
| 118 |
+
|