dmitchelljackson
/

cerebellum-qwen35-history-actions-lora

Image-Text-to-Text

vision-language

reinforcement-learning

Model card Files Files and versions

cerebellum-qwen35-history-actions-lora

316 MB

Ctrl+K

Ctrl+K

1 contributor

History: 7 commits

dmitchelljackson's picture

dmitchelljackson

Clarify architecture: local execution layer under a frontier planner; note iteration 2

86f6cf9 verified about 1 month ago

harness_code
Upload Qwen3.5 AndroidWorld RL harness milestone 2026-06-09 4 months ago
sft_freeze
Add preserved SFT freeze checkpoint 4 months ago
.gitattributes

1.52 kB
initial commit 5 months ago
README.md

9.31 kB
Clarify architecture: local execution layer under a frontier planner; note iteration 2 about 1 month ago
adapter_config.json

1.1 kB
Upload Qwen3.5 AndroidWorld RL harness milestone 2026-06-09 4 months ago
adapter_model.safetensors

103 MB
xet

Upload Qwen3.5 AndroidWorld RL harness milestone 2026-06-09 4 months ago
rl_runtime_state.json

855 Bytes
Upload Qwen3.5 AndroidWorld RL harness milestone 2026-06-09 4 months ago
train_args.json

2.16 kB
Upload Qwen3.5 AndroidWorld RL harness milestone 2026-06-09 4 months ago
trainer_state.pt
Detected Pickle imports (4)
- "torch.ByteStorage",
- "torch._utils._rebuild_tensor_v2",
- "torch.FloatStorage",
- "collections.OrderedDict"
What is a pickle import?
54.5 MB
xet

Upload Qwen3.5 AndroidWorld RL harness milestone 2026-06-09 4 months ago