Skip to content

Commit cd25f17

Browse files
committed
Fixing nits
Signed-off-by: Vladimir Suvorov <[email protected]>
1 parent b97d99b commit cd25f17

File tree

1 file changed

+3
-9
lines changed

1 file changed

+3
-9
lines changed

src/MaxText/examples/rl_llama3_demo.ipynb

Lines changed: 3 additions & 9 deletions
Original file line numberDiff line numberDiff line change
@@ -63,7 +63,7 @@
6363
"## Setup\n",
6464
"\n",
6565
"Install dependencies and set up the environment:\n",
66-
"https://maxtext.readthedocs.io/latest/tutorials/grpo.html#from-github"
66+
"https://maxtext.readthedocs.io/en/latest/tutorials/grpo.html#from-github"
6767
]
6868
},
6969
{
@@ -130,8 +130,6 @@
130130
" print(\"Authentication failed: Hugging Face token not set\")\n",
131131
"\n",
132132
"# Optional: Override training parameters\n",
133-
"STEPS = 10 # Reduced for demo purposes\n",
134-
"PER_DEVICE_BATCH_SIZE = 1\n",
135133
"LEARNING_RATE = 3e-6\n",
136134
"NUM_GENERATIONS = 2\n",
137135
"GRPO_BETA = 0.08\n",
@@ -197,8 +195,6 @@
197195
" f\"load_parameters_path={MODEL_CHECKPOINT_PATH}\",\n",
198196
" f\"base_output_directory={OUTPUT_DIRECTORY}\",\n",
199197
" f\"hf_access_token={HF_TOKEN}\",\n",
200-
" f\"steps={STEPS}\",\n",
201-
" f\"per_device_batch_size={PER_DEVICE_BATCH_SIZE}\",\n",
202198
" f\"learning_rate={LEARNING_RATE}\",\n",
203199
" f\"num_generations={NUM_GENERATIONS}\",\n",
204200
" f\"grpo_beta={GRPO_BETA}\",\n",
@@ -211,7 +207,6 @@
211207
"# Initialize configuration\n",
212208
"print(f\"🔧 Initializing configuration from: {config_file}\")\n",
213209
"config = pyconfig.initialize(config_argv)\n",
214-
"max_utils.print_system_information()\n",
215210
"\n",
216211
"print(\"\\n✅ Configuration initialized successfully\")\n",
217212
"print(f\"📊 Training steps: {config.steps}\")\n",
@@ -261,10 +256,9 @@
261256
"source": [
262257
"## 📚 Learn More\n",
263258
"\n",
264-
"- **CLI Usage**: Run `python3 -m src.MaxText.rl.train_rl src/MaxText/configs/rl.yml --model_name=llama3.1-8b ...`\n",
259+
"- **CLI Usage**: https://maxtext.readthedocs.io/en/latest/tutorials/grpo.html#run-grpo\n",
265260
"- **Configuration**: See `src/MaxText/configs/rl.yml` for all available options\n",
266-
"- **Documentation**: Check `src/MaxText/rl/train_rl.py` for the `rl_train` function implementation\n",
267-
"- **Examples**: See other examples in `src/MaxText/examples/`"
261+
"- **Documentation**: Check `src/MaxText/rl/train_rl.py` for the `rl_train` function implementation"
268262
]
269263
}
270264
],

0 commit comments

Comments
 (0)