hangX/lerobot
1
0
Fork 0
forked from tangger/lerobot
Code Pull Requests Activity

instructions to train the model

Browse Source
This commit is contained in:
mshukor
2025-06-01 09:48:26 +02:00
parent 5fd40b188e
commit 7f3fe83465
1 changed files with 9 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
lerobot/common/policies/smolvla/modeling_smolvla.py
View File

@@ -29,16 +29,20 @@ pip install -e ".[smolvla]"
Example of finetuning the smolvla pretrained model (`smolvla_base`):
```bash
python lerobot/scripts/train.py \
--policy.path=lerobot/smolvla \
--dataset.repo_id=danaaubakirova/koch_test
--policy.path=lerobot/smolvla_base \
--dataset.repo_id=danaaubakirova/svla_so100_task1_v3 \
--batch_size=64 \
--steps=200000
```
Example of finetuning the smolvla neural network with SmolVLM and expert Gemma
pretrained with VLM default parameters before smolvla finetuning:
Example of finetuning the smolvla neural network with pretrained VLM and action expert
intialized from scratch:
```bash
python lerobot/scripts/train.py \
--policy.type=smolvla \
--dataset.repo_id=danaaubakirova/koch_test
--dataset.repo_id=danaaubakirova/svla_so100_task1_v3 \
--batch_size=64 \
--steps=200000
```
Example of using the smolvla pretrained model outside LeRobot training framework: