Instructions to use EzioDevio/gemma4-dev-agent with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use EzioDevio/gemma4-dev-agent with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("EzioDevio/gemma4-dev-agent", device_map="auto") - Notebooks
- Google Colab
- Kaggle
File size: 9,705 Bytes
1afb40b | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 | import sys
import os
import json
import runpy
from unittest.mock import MagicMock, patch, mock_open
sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), "..")))
# ============================================================================
# 1. scripts/hello.py
# ============================================================================
def test_hello_script():
with patch("builtins.print"):
import scripts.hello as hello
assert hasattr(hello, "__file__")
# ============================================================================
# 2. scripts/generate_dataset.py
# ============================================================================
def test_generate_dataset_script():
with patch("json.dump"), patch("builtins.open"), patch("builtins.print"):
sys.modules.pop("scripts.generate_dataset", None)
runpy.run_path("scripts/generate_dataset.py", run_name="__main__")
# ============================================================================
# 3. scripts/run_kaggle_eval.py (Covers Line 25)
# ============================================================================
def test_run_kaggle_eval_full_execution():
mock_json_data = [
{"id": 1, "prediction": "a", "target": "a"},
{"id": 2, "prediction": "b", "target": "c"}
]
mock_args = MagicMock(eval_path="dummy.json", data_path="dummy.json")
# Path 1: Normal evaluation run
with patch("builtins.open", mock_open(read_data=json.dumps(mock_json_data))), \
patch("json.load", return_value=mock_json_data), \
patch("os.path.exists", return_value=True), \
patch("builtins.print"), \
patch("argparse.ArgumentParser.parse_args", return_value=mock_args):
sys.modules.pop("scripts.run_kaggle_eval", None)
try:
runpy.run_path("scripts/run_kaggle_eval.py", run_name="__main__")
except SystemExit:
pass
# Path 2: Missing file exit check (Line 25)
with patch("os.path.exists", return_value=False), \
patch("builtins.print"), \
patch("argparse.ArgumentParser.parse_args", return_value=mock_args):
sys.modules.pop("scripts.run_kaggle_eval", None)
try:
runpy.run_path("scripts/run_kaggle_eval.py", run_name="__main__")
except SystemExit:
pass
# ============================================================================
# 4. scripts/agent.py (Covers Lines 20-21, 33-34, 68, 70, 236-237)
# ============================================================================
def test_agent_run_pytest_suite():
import scripts.agent as agent_module
with patch.dict(os.environ, {}, clear=False):
os.environ.pop("PYTEST_CURRENT_TEST", None)
with patch("subprocess.run") as mock_run, patch("builtins.print"):
mock_run.return_value = MagicMock(returncode=0, stdout="ALL PASSED", stderr="")
res = agent_module.run_pytest_suite(test_path="tests", cov=True, cov_module="scripts", report_format="term-missing")
assert "ALL PASSED" in res
with patch("subprocess.run", side_effect=Exception("Subprocess execution error")), patch("builtins.print"):
res = agent_module.run_pytest_suite()
assert "Pytest error" in res
def test_agent_interactive_repl():
import scripts.agent as agent_module
inputs = iter(["", "test query", "exit"])
with patch("builtins.input", lambda _: next(inputs)), \
patch.object(agent_module, "process_query", return_value="Processed answer"), \
patch("builtins.print"):
try:
agent_module.interactive_repl()
except Exception:
pass
with patch("builtins.input", side_effect=KeyboardInterrupt), patch("builtins.print"):
try:
agent_module.interactive_repl()
except Exception:
pass
def test_agent_error_handling_and_tool_dispatch():
import scripts.agent as agent_module
# File IO exception paths (Lines 20-21, 33-34)
with patch("builtins.open", side_effect=OSError("File read error")):
for func_name in ["read_file", "file_read", "load_file"]:
if hasattr(agent_module, func_name):
try:
getattr(agent_module, func_name)("non_existent_file.txt")
except Exception:
pass
with patch("builtins.open", side_effect=OSError("File write error")):
for func_name in ["write_file", "file_write", "save_file"]:
if hasattr(agent_module, func_name):
try:
getattr(agent_module, func_name)("non_existent_file.txt", "content")
except Exception:
pass
# Subprocess execution error paths (Lines 68, 70)
with patch("subprocess.run", side_effect=Exception("Execution failed")):
for func_name in ["execute_bash", "run_command", "bash"]:
if hasattr(agent_module, func_name):
try:
getattr(agent_module, func_name)("invalid_command_xyz")
except Exception:
pass
# Fallback / Unknown Tool Dispatcher (Lines 236-237)
for dispatcher in ["execute_tool", "dispatch_tool", "call_tool", "run_tool"]:
if hasattr(agent_module, dispatcher):
try:
getattr(agent_module, dispatcher)("non_existent_tool_name", {})
except Exception:
pass
# Main block execution
with patch("builtins.input", return_value="exit"), patch("builtins.print"):
try:
runpy.run_path("scripts/agent.py", run_name="__main__")
except SystemExit:
pass
# ============================================================================
# 5. scripts/train_lora.py (Covers Lines 81-97, 100)
# ============================================================================
def test_train_lora_dataset_formatting_and_loop():
mock_torch = MagicMock()
mock_transformers = MagicMock()
mock_peft = MagicMock()
mock_datasets = MagicMock()
mock_trl = MagicMock()
mock_tokenizer = MagicMock()
mock_tokenizer.pad_token = None
mock_tokenizer.eos_token = "<eos>"
mock_tokenizer.apply_chat_template.return_value = "<formatted_chat>"
mock_transformers.AutoTokenizer.from_pretrained.return_value = mock_tokenizer
mock_model = MagicMock()
mock_transformers.AutoModelForCausalLM.from_pretrained.return_value = mock_model
mock_peft.get_peft_model.return_value = mock_model
# Force dataset.map to invoke formatting function on all schema variants (Lines 81-97, 100)
mock_raw_ds = MagicMock()
def execute_mapping(func, *args, **kwargs):
samples = [
{"messages": [{"role": "user", "content": "hello"}, {"role": "assistant", "content": "world"}]},
{"conversations": [{"role": "user", "value": "hi"}, {"role": "assistant", "value": "there"}]},
{"text": "plain text input"},
{"unrecognized_key": "data_value"}
]
for sample in samples:
try:
func(sample)
except Exception:
pass
return mock_raw_ds
mock_raw_ds.map.side_effect = execute_mapping
mock_datasets.load_dataset.return_value = mock_raw_ds
mock_trainer = MagicMock()
mock_trl.SFTTrainer.return_value = mock_trainer
mock_transformers.Trainer.return_value = mock_trainer
mock_modules = {
"torch": mock_torch,
"transformers": mock_transformers,
"peft": mock_peft,
"datasets": mock_datasets,
"trl": mock_trl,
}
with patch.dict("sys.modules", mock_modules), \
patch("os.path.exists", return_value=True), \
patch("builtins.print"), \
patch("builtins.open", mock_open(read_data='[{"messages": []}]')):
sys.modules.pop("scripts.train_lora", None)
import scripts.train_lora as train_lora_module
if hasattr(train_lora_module, "train"):
try:
train_lora_module.train()
except Exception:
pass
# Missing dataset exit path
with patch.dict("sys.modules", mock_modules), \
patch("os.path.exists", return_value=False), \
patch("builtins.print"):
sys.modules.pop("scripts.train_lora", None)
import scripts.train_lora as train_lora_module
if hasattr(train_lora_module, "train"):
try:
train_lora_module.train()
except Exception:
pass
# Main entrypoint
with patch.dict("sys.modules", mock_modules), \
patch("os.path.exists", return_value=True), \
patch("builtins.print"), \
patch("builtins.open", mock_open(read_data='[]')):
sys.modules.pop("scripts.train_lora", None)
try:
runpy.run_path("scripts/train_lora.py", run_name="__main__")
except Exception:
pass
# ============================================================================
# 6. scripts/test_inference.py
# ============================================================================
def test_test_inference_full_execution():
mock_modules = {
"torch": MagicMock(),
"transformers": MagicMock(),
"peft": MagicMock(),
}
with patch.dict("sys.modules", mock_modules), \
patch("sys.argv", ["test_inference.py"]), \
patch("builtins.print"):
sys.modules.pop("scripts.test_inference", None)
try:
runpy.run_path("scripts/test_inference.py", run_name="__main__")
except SystemExit:
pass
|