SaylorTwift
/

lfm-opencode-lora-rl-code

Model card Files Files and versions

lfm-opencode-lora-rl-code

331 kB

Ctrl+K

Ctrl+K

1 contributor

History: 16 commits

SaylorTwift's picture

SaylorTwift HF Staff

Fix seam patch: target _opencode block by unique indent (3 apiKey seams exist)

231d3f3 verified 15 days ago

data
Switch sandboxes to Harbor hf-sandbox backend; add pinned task prep with docker_image patch; dataset discovery via prompt_rows() 19 days ago
.gitattributes

1.52 kB
initial commit 19 days ago
README.md

6.52 kB
Pin trl==1.15.0 (pre-tree-packing, LFM2 hybrid compatible) + vllm v0.31.0 image 19 days ago
eval_adapter.py

5.24 kB
eval: pass purpose=eval (server requires it with eval_sampling) 19 days ago
export_adapter.py

3.06 kB
OpenCode-only LoRA RL on LFM2.5-2.6B (verified against trl 1.16.0 / openenv 0.10.0) 19 days ago
prepare_tasks.py

4.53 kB
Add held-out eval (base vs adapter) + standalone eval job + final eval stage in training job 19 days ago
run_eval.sh

2.65 kB
Add held-out eval (base vs adapter) + standalone eval job + final eval stage in training job 19 days ago
run_job.sh

7.25 kB
Fix seam patch: target _opencode block by unique indent (3 apiKey seams exist) 15 days ago
serve_env.sh

1.26 kB
Switch sandboxes to Harbor hf-sandbox backend; add pinned task prep with docker_image patch; dataset discovery via prompt_rows() 19 days ago
serve_policy.sh

1.13 kB
vLLM: --enable-auto-tool-choice --tool-call-parser lfm2 (OpenCode sends tool_choice=auto) 19 days ago
train_lora.py

8.96 kB
v2: tunnel-free proxy via SPACE_HOST+Jobs expose_public, heartbeat watchdog (30min stall -> fail), keep UploadCheckpointsCallback 18 days ago